U.S. Department
of Commerce
Volume 90J
Number 1
January 1992
ine Biological Laboratory
LIBRARY
AY 5 mz
f"W(Wds Hole, Mass.
U.S. Department
of Commerce
Barbara Hackman Franklin
Secretary
National Oceanic
and Atmospheric
Administration
John A. Knauss
Under Secretary for
Oceans and Atmosphere
National Marine
Fisheries Service
William _W. Fox Jr.
Assistant Administrator
for Fisheries
^^ATci 0< *'
The Fishery Bulletin (ISSN 0090-0656)
is published quarterly by the Scientific
Publications Office, National Marine
Fisheries Service, NOAA, 7600 Sand
Point Way NE, BIN C15700, Seattle, WA
98115-0070. Second class postage is paid
in Seattle, Wash., and additional offices.
POSTMASTER send address changes for
subscriptions to Fishery Bulletin, Super-
intendent of Documents, Attn: Chief,
Mail List Branch, Mail Stop SSOM,
Washington, DC 20402.
Although the contents have not been
copyrighted and may be reprinted entire-
ly, reference to source is appreciated.
The Secretary of Commerce has deter-
mined that the publication of this period-
ical is necessary in the transaction of the
public business required by law of this
Department. Use of funds for printing of
this periodical has been approved by the
Director of the Office of Management and
Budget.
For sale by the Superintendent of
Documents, U.S. Government Printing
Office, Washington, DC 20402. Subscrip-
tion price per year: $16.00 domestic and
$20.00 foreign. Cost per single issue:
$9.00 domestic and $11.25 foreign. See
back page for order form.
D
D
3QaDD(SGDDD
Scientific Editor
Dr. Linda L. Jones
National Marine Mammal Laboratory
National Marine Fisheries Service, NOAA
7600 Sand Point Way NE
Seattle, Washington 981 15-0070
Editorial Committee
Dr. Andrew E. Dizon National Marine Fisheries Service
Dr. Charles W. Fowler National Marine Fisheries Service
Dr. Richard D. Methot National Marine Fisheries Service
Dr. Theodore W. Pietsch University of Washington
Dr. Joseph E. Powers National Marine Fisheries Service
Dr. Tim D. Smith National Marine Fisheries Service
Dr. Mia J. Tegner Scripps Institution of Oceanography
IVIanaging Editor
Nancy Peacocl<
National Marine Fisheries Service
Scientific Publications Office
7600 Sand Point Way NE, BIN CI 5700
Seattle, Washington 981 15-0070
The Fishery Bulletin carries original research reports and technical notes on investiga-
tions in fishery science, engineering, and economics. The Bulletin of the United States
Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries
in 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates
were issued as documents through volume 46; the last document was No. 1 103. Begin-
ning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate
appeared as a numbered bulletin. A new system began in 1963 with volume 63 in
which papers are bound together in a single issue of the bulletin. Beginning with
volume 70, number I, January 1972, the Fishery Bulletin became a periodical, issued
quarterly. In this form, it is available by subscription from the Superintendent of
Documents, U.S. Government Printing Office, Washington, DC 20402. It is also
available free in limited numbers to libraries, research institutions, State and Federal
agencies, and in exchange for other scientific publications.
U.S. Department
of Commerce
Seattle, Washington
Volume 90
Number 1
January 1992
Fishery
Bulletin
Contents
1 Buckland, Stephen T., Karen L.
Alejandro A. Anganuzzi
Estimating trends in abundance of dolphins associated with tuna in
the eastern tropical Pacific Ocean, using sightings data collected on
commercial tuna vessels
13 Collette, Bruce B., and Gary B. Gillis
Morphology, systematics, and biology of the double-lined mackerels
[Grammatorcynus, Scombridae)
54 Douglas, Michael E., Gary D. Schnell,
Daniel J. Hough, and William F. Perrin
Geographic variation in cranial morphology of spinner dolphins
Stenella longirostns in the eastern tropical Pacific Ocean
77 Gall, Graham A.E., Devin Bartley, Boyd Bentley,
Jon Brodziak, Richard Gomulkiewicz,
and Marc Mangel
Geographic variation in population genetic structure of Chinook
salmon from California and Oregon
101 Hunter, J. Roe, Beverly J. Macewicz,
N. Chyan-huei Lo, and Carol A. KImbrell
Fecundity, spawning, and maturity of female dover sole Microstomus
paaficus, with an evaluation of assumptions and precision
1 29 Kendall, Arthur W. Jr., and Toshikuni NakatanI
Comparisons of early-life-history characteristics of walleye pollock
Theragra chalcogramma in Shelikof Strait, Gulf of Alaska, and
Funka Bay, Hokkaido, Japan
139 McShane, Paul E.
Exploitation models and catch statistics on the Victorian fishery for
abalone Haliotis rubra
Fishery Bulletin 90(1). 1992
147 Sissenwine, Michael P., and Pamela M. Mace
ITQs in New Zealand: The era of fixed quota in perpetuity
161 Stoner, Allan W., Veronique J. Sandt, and Isabelle F. Boidron-Metairon
Seasonality in reproductive activity and larval abundance of queen conch Strombus gigas
171 Wainwright, Thomas C, David A. Armstrong, Paul A. Dinnel,
Jose M. Orensanz, and Katherine A. McGraw
Predicting effects of dredging on a crab population: An equivalent adult loss approach
Notes
183 Chen, Weihzong, John J. Govoni, and Stanley M. Warlen
Comparison of feeding and growth of larval round herring Etrumeus teres and gulf menhaden
Brevoortia patronus
190 D'Amours, Denis, and Francois Gregoire
Analytical correction for oversampled Atlantic mackerel Scomber scombrus eggs collected with oblique
plankton tows
1 97 Rajaguru, Arjuna, and Gopaisamy Shantha
Association between the sessile barnacle Xenobalanus globiapitis (Coronulidae) and the bottlenose
dolphin Tursiops truncatus (Delphinidae) from the Bay of Bengal, India, with a summary of previous
records from cetaceans
203 Safford, Susan E., and Henry Booke
Lack of biochemical genetic and morphometric evidence for discrete stocks of Northwest Atlantic herring
Clupea harengus harengus
211 Stergiou, Konstantinos I.
Variability of monthly catches of anchovy Engraulis encrasicolus in the Aegean Sea
Abstract.- We summarize the
methods for estimating relative abun-
dance of seven dolphin stocks in the
eastern tropical Pacific Ocean using
sightings data collected on commer-
cial tuna vessels by trained observ-
ers, developed by Buckland and
Anganuzzi (1988a) and Anganuzzi
and Buckland (1989). Their estimates
of relative abundance, which may
show large year-to-year fluctuations,
are smoothed to provide estimates of
the underlying trend in dolphin abun-
dance between 1976 and 1988. The
bootstrap method provides estima-
tion of precision in a way that allows
trend estimates to be used for man-
agement purposes, without the need
to assume that trends in abundance
are linear. Concerns about the valid-
ity of the estimates are addressed.
Estimating trends in abundance
of dolphins associated witli tuna in
the eastern tropical Pacific Ocean,
using sightings data collected
on commercial tuna vessels
Stephen T. Buckland
Karen L. Cattanach
SASS Environmental Modelling Unit, MLURI
Craigiebuckler, Aberdeen AB9 2QJ. United Kingdom
Alejandro A. Anganuzzi
Inter-American Tropical Tuna Commission
8604 La Jolla Shores Drive, La Jolla, California 92093
Manuscript accepted 27 November 1991.
Fishery Bulletin, U.S. 90:1-12 (1992).
Incidental mortality of dolphins in
the tuna fishery in the eastern trop-
ical Pacific since 1959 has been suf-
ficient to affect abundance of stocks
of at least two species of dolphin: the
spotted dolphin Stenella attenuata
and the spinner dolphin S. longiros-
tris (Smith 1983). Although there is
less information available on stocks
of the common dolphin Delphinus
delphis, mortality estimates (e.g.,
Hall and Boyer 1988) suggest that
abundance of stocks of this species
may also have been reduced. To mon-
itor possible effects of incidental mor-
tality on the size of dolphin stocks,
several attempts to estimate abun-
dance have been made, usually apply-
ing line-transect methodology to data
collected on either commercial tuna
vessels ("tuna vessel data") or re-
search vessels ("research vessel
data") or both. Holt and Powers
(1982) and Holt (1985, 1987) consid-
ered analyses of research vessel data
alone, and of tuna vessel data com-
bined with research vessel data.
More recently. Holt and Sexton
(1989, 1990a, b) analyzed data from
research vessels alone. Tuna vessel
data alone were analyzed by Ham-
mond and Laake (1983), by Polacheck
(1987), by Buckland and Anganuzzi
(1988a), and by Anganuzzi and Buck-
land (1989).
The tuna vessel data are collected
by scientific technicians placed by
two organizations onboard commer-
cial tuna purse seiners. The Inter-
American Tropical Tuna Commission
(lATTC) places technicians on vessels
of the international fleet (including
U.S. -registered vessels), and the
National Marine Fisheries Service
(NMFS) of the United States places
technicians on U.S. -registered ves-
sels only. Data were first collected by
NMFS in 1974, and by lATTC in
1979.
Tuna vessel data provide a large
database, with regular coverage of a
substantial portion of the area oc-
cupied by the dolphin stocks. How-
ever, due to the nature of the fishery
operations, the assumptions neces-
sary for line-transect sampling to
yield unbiased estimates of absolute
abundance are often violated. There-
fore, analytic procedures should as
far as possible be insensitive to those
violations. We summarize here the
procedures of Buckland and Anga-
nuzzi (1988a), as modified by Anga-
nuzzi and Buckland (1989). Since
these procedures are unlikely to re-
move all biases, the estimates should
Fishery Bulletin 90(1). 1992
be treated as indices of relative abundance, rather than
estimates of absolute abundance of the stocks. The
definition of a stock, and its boundaries, is problematic,
but we follow the recommendations of Au et al. (1979),
for reasons stated by Anganuzzi and Buckland (1989),
except in two cases. A more southerly southern bound-
ary was found to be necessary for the southern offshore
stock of spotted dolphins (Anganuzzi et al. 1991), and
we adopt the recommendation of Perrin et al. (1991)
to combine the northern and southern whitebelly stocks
of spinner dolphins. We also derive estimates for pooled
offshore stocks of spotted dolphins and pooled stocks
of common dolphins, since they are not differentiable
in the field.
Buckland and Anganuzzi (1988a) provided three
types of test for assessing whether abundance of a
stock had changed over time. For several stocks, the
tests failed to provide a clear indication of recent
changes, since the occasional large fluctuation in an-
nual estimates indicated that there were significant
changes in abundance that were biologically implaus-
ible. We present here a method of smoothing the
sequence of estimates of relative abundance. Used in
conjunction with the bootstrap, it yields a simple
method of assessing change over time which does not
require that trends are assumed to be linear, and which
does not yield biologically implausible rates of change.
Edwards and Kleiber (1989) have questioned the
validity of estimating trends in abundance from sight-
ings data collected on commercial tuna vessels. We
carry out a simple simulation study to assess their
assertions, and compare the relative abundance esti-
mates calculated from tuna vessel data with those
calculated from research vessel data for the years
1986-89, for which data from both sources are
available.
Methods
The number of dolphins A^ in an area for a given stock
and year is estimated by
N = A ■ S ■ D
where A is the size of the area,
s is the estimated average school size for the
stock in area A, and
D is the estimated density of schools in area A.
The line-transect method provides the estimate D
(Burnham et al. 1980). Suppose schools farther than
a distance w from the trackline are discarded from the
analyses. Then
D =
2L
(1)
where n is the number of schools detected in the area
that are within the truncation distance w,
/(O) is the estimated probability density function
of the n perpendicular distances, evaluated
at perpendicular distance zero, and
L is the total length of transect in nautical
miles within the area.
If we define the encounter rate E to be the expected
number of sightings detected within m' of the trackline
per nautical mile of search, then its estimate is given by
E = nIL.
Hence,
and
D
N
E-m
Ef{0)-s-A
(2)
(3)
UD and N were estimates of absolute abundance, then
the following assumptions would be required:
(i) Within each area or stratum, either the search effort
of the tuna vessels is random or the dolphin schools
are randomly distributed;
(ii) any movement of schools is slow relative to the
speed of the vessel, at least before detection;
(iii) all schools on or close to the trackline are detected
and identified;
(iv) sighting distances and angles are measured with-
out error;
(v) sightings of schools are independent events;
(vi) school size is recorded without error, and for mixed
schools percent of each species is recorded without
error;
(vii) probability of detection of a school is independent
of its size, at least out to perpendicular distance w.
If the estimates are used solely as indices of relative
abundance, as here, then any or all of the above
assumptions may fail without invalidating the esti-
mates, provided that bias arising from the failure of
an assumption is consistent across time. Even this pro-
viso may be relaxed when trends in abundance over
a long sequence of years are estimated; in this case it
is merely necessary to assume that bias shows no trend
with time. Catch-per-unit-effort methods for estimating
relative abundance are known to show trends in bias
over time in some instances, due to increased efficiency
of vessels (Cooke 1985). We attempt to avoid such prob-
lems by incorporating a parameter that measures the
Buckland et al.: Estimating abundance of tuna-associated dolphin stocks in the eastern tropical Pacific
efficiency of search of the tuna vessels. This parameter,
the effective search width, is estimated using line-
transect theory. It may be interpreted as twice the
distance at which the number of undetected dolphin
schools closer to the vessel is equal to the number of
detected schools further from the vessel, and is there-
fore the effective width of the strip of ocean searched
by the vessel. As efficiency of the fleet to detect dolphin
schools increases (e.g., through the use of helicopters,
high-resolution radar, etc.), the effective search width
increases, and bias in abundance estimates should re-
main unaffected.
We adopt a strategy of reducing bias as much as
possible, so that the effect of any trend in bias over
time on estimated trends in abundance is minimized.
To estimate the different components of the estimator
of Equation (3), separate stratification schemes are ap-
plied for encounter rate, effective search width, and
school size. In stratifying for a given component, our
aim is to define strata such that each stratum is
relatively homogeneous with respect to that compo-
nent, so that non-random search effort and non-random
distribution of schools generate only small bias in any
given stratum. Crude encounter rates, average school
sizes, and average detection distances are estimated
by 1° square. Where data are insufficient, the crude
estimates are smoothed, and the same smoothing pro-
cedure interpolates for squares in which there was no
tuna vessel effort. These estimates are used to allocate
1 ° squares to strata, yielding the separate stratifica-
tions for encounter rate, school size, and effective
search width, respectively. Full details are given by
Anganuzzi and Buckland (1989).
Thus the problem of abundance estimation has been
reduced to three simpler problems: For a random point
in the stock area, the expectations of encounter rate,
school size, and effective search width are estimated,
and the three estimates are multiplied together to ob-
tain the final abundance estimate. Lack of indepen-
dence between the three estimates does not bias the
overall estimate, and independence is not assumed
when estimating variance. A nonparametric bootstrap
technique is used to obtain variances. The resampling
unit in the bootstrap is the individual cruise, and for
each bootstrap replicate the full estimation procedure
is applied, thus generating bootstrap estimates of abun-
dance. The sample variance of these estimates yields
the required variance estimates, and confidence inter-
vals are obtained by the percentile method. (See Buck-
land and Anganuzzi 1988a, for details.)
Bias arising from rounding errors in the recorded
sighting distances r and angles 9 is reduced by smear-
ing the data, using the method favored by Buckland
and Anganuzzi (1988b). The recorded location of
each school relative to the tuna vessel at the time of
detection is defined by r and 9, and that location is
"smeared" over the sector defined by r • (1 ± d ) and
9 ± ^12. to allow for inaccuracy in the recorded values.
The smearing parameters d and I are estimated from
the data. When a small sighting angle is rounded to
zero, the calculated perpendicular distance is zero,
giving a spurious spike in the perpendicular distance
distribution at zero distance. Smearing yields more
robust estimation by removing or reducing this spike.
Here we take the estimates of Anganuzzi and Buck-
land (1989) and of Anganuzzi et al. (1991) and attempt
to estimate the underlying trends in dolphin abundance
by smoothing them. Various smoothing methods such
as moving averages, running medians, and polynomial
regression were investigated (Smith 1988). The chosen
method was a compound running median known as
"4253H, twice" (Velleman and Hoaglin 1981), which
is constructed as follows.
Suppose that {X{t )}, ^ = 1, . . . , A'^, is a time-series of
length A^, and let {5,(0} be a smoothed version of it,
found by calculating an i -period running median. We
can construct compound smoothing methods such as
{Sijit)}, which is simply {Sj{Si{t))}. Thus, a 4253 run-
ning median method smooths a time-series using a
4-period running median, which is in turn smoothed by
a 2-period running median, smoothed again by a
5-period running median, and then by a 3-period
running median (i.e., {54253(0} = {5'3(S5(S2(S4(0)))})-
Near the endpoints, where there are not enough values
surrounding a point to be smoothed using the spe-
cified running median, a shorter-period running median
may be used. The endpoints of the resultant time-series
are calculated by estimating X(0) and X(N + l), the
"observed" values at t = 0 and t=N + l, and then
calculating
54253(1) = median {1(0), X{1), 54,53(2)} and
54253(iV) = median {S4253(A^-1), XiN), X{N + 1)}.
X(0) is found by extrapolating from the straight line
which passes through the smoothed values att=2 and
i = 3, i.e., 1(0) = 3 -54953(2)- 2 -54953(3); similarly,
X{N + 1) = 3- 54953(iV - 1) - 2 - 54253(^ - 2).
The H in "4253H, twice" denotes a linear smoothing
method commonly used with running medians, which
is known as Banning. It is a 3-period weighted mov-
ing average iort=2,...,N-l, with weights {0.25, 0.5,
0.25}. The endpoints remain unchanged.
The pattern of the time-series may be recovered by
calculating the residuals of the series (i.e., the differ-
ences between the smoothed and unsmoothed esti-
mates), smoothing the residual series using the same
method as for the time-series, and then adding the
smoothed values of the residuals to the smoothed
Fishery Bulletin 90(1), 1992
values of the series. This is known as smoothing
"twice." For example, if we define the residuals of the
time-series smoothed by 4253H to be {E(t )} = {X(t ) -
•54253 (i )}. then the values of the times-series smoothed
by "4253H, twice" can be defined by
{•S4253H, twice (0} = {54253h(0 + 'S4253h(-E'(0)}-
Thus the "4253H, twice" running median method
uses a 4253 running median to smooth the time-series,
estimates the endpoints of the smoothed series, and
then smooths the resultant series by Manning. The
residuals of the series are calculated and are also
smoothed, using the same method as above. The
smoothed values of the residuals are then added to the
smoothed values of the time-series to produce a time-
series smoothed by "4253H, twice." The advantage of
using running medians is that the magnitude of an
extreme estimate does not affect the resultant
smoothed time-series. The above method is sufficient-
ly complex that its behavior cannot be readily under-
stood. However, simpler methods were found to suf-
fer from one or more of the following shortcomings:
Estimated trends were not always smooth; implausible
rates of change were sometimes indicated; trends near
the start or end of the sequence of estimates were often
poorly estimated.
Nonparametric bootstrap replicates are generated as
described by Anganuzzi and Buckland (1989). We select
here the bootstrap estimates that correspond to an 85%
confidence interval for relative abundance in each year.
The rationale for the choice of confidence level is that
if two 85% confidence intervals do not overlap, the
difference between the corresponding relative abun-
dance estimates is significant at roughly the 5% level
(P<0.05); whereas if they do, the difference is not
significant (P>0.05). If the abundance estimates are
assumed to be lognormally distributed, each with the
same coefficient of variation, then the exact confidence
level that gives this property is 83.4%. If one estimate
has twice the coefficient of variation of the other, the
confidence level increases slightly to 85.6%. Thus a
choice of 85% makes some allowance for variability in
the coefficient of variation.
. For each abundance estimate, 79 bootstrap replicates
are run, so that the 6th smallest and 6th largest boot-
strap estimates provide an approximate 85% confi-
dence interval (Buckland 1984). If this procedure is
carried out independently for each year, confidence
intervals are wide. Provided the assumed stock area
spans the whole range of the stock, numbers of dolphins
within it are unlikely to vary greatly in successive
years, and a procedure that calculates confidence in-
tervals for a given year incorporating information from
years immediately preceding and following that year
is more informative. For a given stock, we achieve this
by carrying out one bootstrap replication for each year
that a relative abundance estimate is available. These
estimates are smoothed using the routine described
above, and the process is repeated 79 times. For each
year, the 6th smallest and 6th largest smoothed
estimates provide approximate 85% confidence limits.
We use the sequence of medians of the smoothed boot-
strap estimates (i.e., the 40th estimate of each ordered
set of 79) as the "best" indicator of trend, so that it
is calculated in a comparable manner to the confidence
limits. Larger numbers of bootstrap replicates are
preferable, but available computer power was limited.
Repeat runs for the northern offshore stock of spotted
dolphins were carried out, to assess the Monte Carlo
variability.
By using overlapping confidence intervals to test for
a difference between years, independence between
smoothed estimates for different years is assumed.
Given the strong positive correlation in the smoothed
estimates between successive years, the test is unlike-
ly to detect a large change between one year and the
next, but should be reliable for detecting trends over
a period of perhaps five or more years, for which cor-
relations between smoothed estimates are small.
Results
Figures 1-10 show the estimates of underlying trend
for each of the main stocks associated with tuna in the
eastern tropical Pacific Ocean. Since stock boundaries
and stock identity are both uncertain, we also show
trend estimates after pooling data from stocks that are
not differentiable in the field. The broken horizontal
lines in these plots correspond to the upper and lower
85% confidence limits for the 1988 relative abundance
estimate. Years for which the entire confidence inter-
val lies outside the region between the broken horizon-
tal lines show a relative abundance significantly
different from that for 1988. Because the smoothed
estimate for the first or final year of a sequence can
be poor, we show the unsmoothed estimate and cor-
responding 85% confidence limits for the first and last
year on each plot.
Figures 1 and 2 show estimated trends for northern
offshore spotted dolphins, with and without the abnor-
mally low 1983 estimate, which corresponded with a
very strong El Nino event. It is clear that the 1983
estimate affects the smoothed estimate of trend, but
its effect is no greater than if it had been just smaller
than the 1984 estimate. Thus abnormal estimates may
be more safely retained when using this procedure, and
subjective decisions of whether to treat an estimate as
an outlier are avoided.
Buckland et al : Estimating abundance of tuna-associated dolphin stocks in the eastern tropical Pacific
7000
-6000
1
i
Jsooo
Itooo
1
"§3000
a 2000
1000
Q^^^
D
1975 1976 1977 1973 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
Figure 1
Smoothed abundance trends of northern offshore stock of
spotted dolphin Stenella attenuata in the eastern tropical
Pacific. Broken lines indicate approximate 85% confidence
limits. Horizontal lines correspond to 85% confidence limits
for the 1988 estimate. If lower limit lies above upper limit
for an earlier year, abundance has increased significantly
between that year and 1988 (P< 0.05); if upper limit lies below
lower limit for an earlier year, abundance has decreased
significantly.
The estimated trend from Figure 1 is downwards
until around 1983. Estimated abundance in 1976 and
1977 was significantly higher than in 1988 (P<0.05),
but there is some evidence of a recovery between 1983
and 1988 (P<0.05). Thus northern offshore spotted
dolphins appeared to decrease through the 1970s and
early 1980s, with numbers remaining stable or increas-
ing since.
Figure 3 suggests there may have been a marked
decline in numbers of southern offshore spotted dol-
phins since the late 1970s. The smoothed 1988 estimate
is significantly lower than the smoothed estimates for
1977 and 1978, but there is evidence of an increase
since 1986 (P<0.05), after a relatively high unsmoothed
estimate for 1989. As shown by Anganuzzi et al. (1991),
southern offshore spotted dolphins appear to occupy
appreciably different regions from one year to another,
and the extent of mixing with northern offshore
spotted dolphins remains unclear. We therefore believe
that trend estimates for this stock are unreliable. The
estimated trends obtained by pooling data from the off-
shore stocks are shown in Figure 4. The estimates are
dominated by the data from the larger northern off-
shore stock, and the plot is similar to Figure 1. The
1988 smoothed relative-abundance estimate is signifi-
cantly higher than the 1983 and 1984 estimates, and
significantly lower than all estimates preceding 1979.
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
Figure 2
Smoothed abundance trends of northern offshore stock of
spotted dolphin Stenella attenuata in the eastern tropical
Pacific, excluding 1983 estimate. Broken lines indicate approx-
imate 85% confidence limits. See Figure 1 for more details.
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
Figure 3
Smoothed abundance trends of southern offshore stock of
spotted dolphin Stenella attenuata in the eastern tropical
Pacific. Broken lines indicate approximate 85% confidence
limits. See Figure 1 for more details.
Figure 5 suggests that the eastern spinner dolphin
might have had a pattern of change similar to the
northern offshore spotted dolphin, although estimated
abundance in the late 1980s is roughly equal to that
in the mid-1970s, so depletion between 1975 and 1983
may have been less than for northern offshore spotted
dolphins. The 1988 smoothed estimate is just signifi-
cantly higher than the smoothed estimates for 1981 and
1982 (P<0.05).
Fishery Bulletin 90(1). 1992
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
Figure 4
Smoothed abundance trends of pooled northern and southern
offshore stocks of spotted dolphin Stenella attenuata in the
eastern tropical Pacific. Broken lines indicate approximate
85% confidence limits. See Figure 1 for more details.
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
Figure 6
Smoothed abundance trends of whitebelly stock of spinner
dolphin Stenella longirostris in the eastern tropical Pacific.
Broken lines indicate approximate 85% confidence limits. See
Figure 1 for more details.
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1985 1987 1998 1989
Figure 5
Smoothed abundance trends of eastern stock of spinner
dolphin Stenella longirostris in the eastern tropical Pacific.
Broken lines indicate approximate 85% confidence limits. See
Figure 1 for more details.
soo
•;r
I
^ .-' □
£ 500
£
X
?
' /
'"'
D-^
\
i 400
/
/
=^^-_^
X
^
/ ,
^^ ^
"""-^^
1
/
^/"■"^
°
■~-
*
/
^
^ i
200
\
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
Figure 7
Smoothed abundance trends of northern stock of common
dolphin Delphinus delphis in the eastern tropical Pacific.
Broken lines indicate approximate 85% confidence limits. See
Figure 1 for more details.
The estimated trend for whitebelly spinner dolphins
(Fig. 6) is similar to that for eastern spinner dolphins
and northern offshore spotted dolphins. There is some
evidence that abundance in 1988 was higher than in
1982 (P=0.05), but no other comparisons with 1988 are
significant. The 1982 smoothed estimate is significantly
lower than those for 1976-78.
End effects in Figure 7 give rise to an implausible
trend in numbers of northern common dolphins dur-
ing 1975-78. Since 1980, there may have been a decline
in this stock, but no smoothed estimates differ signif-
icantly. The central stock of common dolphins (Fig. 8)
shows evidence of a steep decline from 1977 to 1983,
with stability since. The smoothed estimate for 1988
is significantly lower than for all years preceding 1980
(P<0.05), but does not differ significantly from any
later estimates. Data on the southern stock of common
dolphins are sparse. There may have been a decreas-
ing trend (Fig. 9), but unsmoothed estimates fluctuate
widely and no smoothed estimates differ significantly.
Buckland et al Estimating abundance of tuna-associated dolphin stocks in the eastern tropical Pacific
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1983 1989
Figure 8
Smoothed abundance trends of central stock of common
dolphin Delphinus delphis in the eastern tropical Pacific.
Broken lines indicate approximate 85% confidence limits. See
Figure 1 for more details.
1975 1976 1977 1978 1979 1980 1981 1982 1985 1984 1985 1936 1987 1988 1989
Figure 10
Smoothed abundance trends of pooled northern, central, and
southern stocks of common dolphin Delphinus delphis in the
eastern tropical Pacific. Broken lines indicate approximate
85% confidence limits. See Figure 1 for more details.
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
Figure 9
Smoothed abundance trends of southern stock of common
dolphin Delphinus delphis in the eastern tropical Pacific.
Broken lines indicate approximate 85% confidence limits. See
Figure 1 for more details.
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
Figure 1 1
Smoothed abundance trends of northern offshore stock of
spotted dolphin Stenella attenuata in the eastern tropical
Pacific. Broken lines indicate approximate 85% confidence
limits. Estimates and limits were determined from four in-
dependent sets of 79 bootstrap replicates, so that the plot
indicates uncertainty in the estimates arising from Monte
Carlo variation.
If data are pooled across stocks of common dolphins
(Fig. 10), the 1988 smoothed estimate is significantly
lower than all those preceding 1981.
Four independent sets of 79 bootstrap replicates
were generated for the northern offshore stock of
spotted dolphins. The resulting plots, one of which
corresponds exactly to Figure 1, are superimposed in
Figure 11. If an infinite number of replicates could be
carried out for each set, the four plots would be iden-
tical. Thus Figure 11 indicates the imcertainty that can
be expected in the median and interval estimates due
to Monte Carlo variation.
Discussion
Unsmoothed estimates of relative abundance some-
times show larger year-to-year variation than is
Fishery Bulletin 90(1). 1992
plausible, even if full allowance is made for the preci-
sion of the estimates. An example is the 1983 estimate
for the northern offshore stock of spotted dolphins,
which is significantly lower than either the 1982 or the
1984 estimate. This has been attributed to the strong
El Nino event of that year (Buckland and Anganuzzi
1988a). The change in environmental conditions ap-
peared to cause spotted dolphins to split into smaller
schools and to disperse more widely than is normal, so
that tuna vessels were unable to locate areas of con-
centration. If, in normal years when concentrations
occur in known areas, there is positive bias in the abun-
dance index, then a relatively low estimate might be
expected for 1983. This effect would be enhanced if
many animals wandered beyond the normal range of
the stock, so that the abundance index for 1983 cor-
responded to only that portion of the stock remaining
within its normal bounds. Such effects may be regarded
either as bias that fluctuates over time or as an addi-
tional source of variability that is unaccounted for in
the variances of the abundance indices. Provided the
effects are essentially random, and do not exhibit a con-
sistent linear trend over time, the smoothing algorithm
described above smooths out the large fluctuations and,
in conjunction with the bootstrap, provides variance
and interval estimates for the smoothed abundance
indices that take full account of variability not allowed
for in the variance estimates of the unsmoothed indices.
The validity of estimating trends in dolphin abun-
dance from tuna-vessel sightings data has been ques-
tioned by Edwards and Kleiber (1989). They used a
simple simulation model of non-random search vessel
effort coupled with clustered distributions of dolphin
schools to investigate bias. By allowing the clustering
of schools to be slight in one year and extreme in the
next, they showed that bias in the relative abundance
estimates can be inconsistent between years. They
define a change estimate as the ratio of relative abun-
dance estimates for the two years. They state, "This
two-sample change estimate is only a rough approx-
imation to a trend estimate derived from a series of
measurements . . . However, conclusions about the ef-
fects of inconsistent biases on this change estimate will
be valid for trend estimates also, except for the unlikely
case in which effects of various inconsistent biases
cancel each other out, so that the trend estimate
reflects the actual trend, but only fortuitously." (The
emphasis on "change" and "trend" is theirs.) They also
note that "It is obvious. . .that even relatively small
changes of bias can lead to considerably inaccurate
estimates of change and, by implication, estimates of
trend." If this is so, there would be little value in
estimating trends in abundance from tuna-vessel sight-
ings data. We question whether the simulation model
of Edwards and Kleiber (1989), which is a considerable
Table
1
Actual abundance (millions), and expected and simulated |
relative-abundance estimate by j
ear for a hj-pothetical stock.
declining
at an annual
rate of 5%. Expected
abundance is
calculated
assuming estimates are biased down
by 20% in El
Nino years (*) and up by 100%
in other years
Actual
Expected
Simulated
Year
abundance
estimate
estimate
1975
4.00
8.00
8.04
1976*
3.80
3.04
3.37
1977
3.61
7.22
6.86
1978
3.43
6.86
5.86
1979
3.26
6.52
6.87
1980
3.10
6.19
8.66
1981
2.94
5.88
6.26
1982*
2.79
2.23
1.97
1983*
2.65
2.12
3.22
1984
2.52
5.04
4.98
1985
2.39
4.79
5.72
1986
2.28
4.55
4.02
1987*
2.16
1.73
1.65
1988
2.05
4.11
4.01
1989
1.95
3.90
4.75
simplification of reality, allows such strong conclusions.
However, we use their results to assess the validity of
their argTiments. We take their worst-case scenario of
a static environment, using the stratified and smoothed
option, and average across their four replicates for the
high-density case. The calculations indicate a down-
ward bias of about 20% for the "simple, gentle" en-
vironmental topography of year 1 and an upward bias
of about 100% for the "complex, steep" topography
of year 2. Thus, if the population comprised 2500
schools (as in their simulations), the expected estimate
would be around 2000 schools in the first year and 5000
in the second, a 2.5-fold estimated increase for a pop-
ulation that has constant size. Is this conclusion "valid
for trend estimates also"? Suppose a population com-
prised 4 million animals in 1975, and decreased at a
rate of 5% per annum until 1989. Suppose we again
take an extreme scenario in which the "simple, gentle"
environmental topography applied in El Nino years,
and the "complex, steep" topography applied in all
other years. The expectations of the estimates are
shown in Table 1. Also shown are simulated estimates,
for which errors were generated from a lognormal
distribution which yields a coefficient of variation of
15%, close to that observed for estimates based on tuna
vessel data. The errors were then added to the ex-
pected estimates. The estimated rate of decrease for
the expected estimates is 5.0% per annum (SE2.5%),
and that for the simulated estimates is 4.7% per annum
(SE 2.6%). Thus a scenario of extreme and inconsistent
Buckland et al : Estimating abundance of tuna-associated dolphin stocks in the eastern tropical Pacific
bias does not invalidate the procedures when applied
to a long sequence of estimates. In practice, a rate of
change in abundance is unlikely to be roughly constant
over such a long time-period, yet tests for trend over
a short time-period have low power. Figures 1-10 pro-
vide a simple method to test for change over longer
time-periods without the necessity of assuming the rate
of change is constant.
The smoothing procedure used for generating trend
estimates can perform poorly at the start (e.g.. Fig. 7)
or at the end of a sequence of estimates, so that sharp
increases or declines during the first or last year or two
should be treated with suspicion. The first and last
smoothed estimate in a sequence are especially un-
reliable, and are omitted from Figures 1-10. Thus,
changes in abundance are assessed relative to 1988
rather than 1989.
To assess the current status of dolphin stocks, and
the effects of recent levels of mortality, it is necessary
to determine whether trends in dolphin abundance are
best estimated from tuna vessel data or research vessel
data, or whether some combination of estimates from
both sources is preferable. Given sufficient data and
adequate coverage of the entire range of each stock,
research-vessel estimates of trend would be preferred,
since they are likely to be less biased. However, Holt
and Sexton (1989, 1990ab), to exploit fully the small
number of research vessel sightings, made assumptions
that might be seriously violated. Firstly, data are pool-
ed across all sightings of dolphin schools of at least 15
animals, irrespective of species, to improve precision
of effective search-width estimates. This may introduce
bias which is not consistent over time, especially if non-
target species (those which are seldom associated with
tuna, and are therefore seldom encircled by purse
seines) have a different effective search width and a
different rate of change in abundance than target
species. Secondly, although abundance estimates are
given by stock, encounter-rate estimates by stock area
are ignored for stocks that are not separated in the
field. Thus for offshore spotted dolphins, a single abun-
dance estimate per year is generated and then prorated
by stock area, to yield separate estimates for the north-
ern and southern offshore stocks. If the southern off-
shore stock became extinct, and the northern offshore
stock increased at a rate that ensured overall abun-
dance remained constant, the expected trend in re-
search vessel estimates would be zero for both stocks.
The same applies to common dolphin stocks. The esti-
mates of Holt and Sexton indicate that there are large
numbers of common dolphins in the western sector of
the eastern tropical Pacific, yet the species is seldom
recorded there. Using the estimation methods of Holt
and Sexton, valid trend estimates from research vessel
data are not available separately for northern and
southern offshore stocks of spotted dolphin or for the
main stocks of common dolphin.
In Figures 12-15 we show the valid estimates of
trend (i.e., those obtained after pooling data from
stocks that are not differentiable in the field) from the
research-vessel relative abundance estimates for
1986-89, taken from Sexton et al. (1991) and Gerro-
dette and Wade (1991). Also shown are the corre-
sponding unsmoothed trend estimates from tuna vessel
data. Vertical bars show ± 2 standard errors. Plots are
based on the relative abundance estimates and stan-
dard errors of Tables 2 and 3. The research vessel
estimates indicate changes in abundance that are
biologically implausible, even with full allowance for the
estimated precision of the estimates. Thus either the
precision of the surveys is appreciably worse than
estimated or there is strong and inconsistent bias in
the estimates from one year to the next. By contrast,
despite the concerns over the validity of tuna vessel
estimates, they yield biologically plausible rates of
change during 1986-89 when the precision of the
estimates is accounted for.
5
c
o
= 4
5
T
T
T
T
Abundance
■
^
•
Relative
•
0
86 87 88 89
Year
Figure 12
Unsmoothed abundance trends of northern and southern off-
shore stocks of spotted dolphin Stenella atteniiata in the
eastern tropical Pacific, estimated from research (solid line)
and tuna vessel data. Vertical bars are ± 2 standard errors.
Fishery Bulletin 90|l). 1992
Year
Figure 13
Unsraoothed abundance trends of eastern stock of spinner
dolphin Stenella longirostris in the eastern tropical Pacific,
estimated from research (solid line) and tuna vessel data. Ver-
tical bars are ±2 standard errors.
IX 8
o
c
(0
■□
c
<
>
^ 2
0)
a:
Figure 15
Unsmoothed abundance trends of northern, central, and
southern stocks of common dolphin Delphinus delphis in the
eastern tropical Pacific, estimated from research (solid line)
and tuna vessel data. Vertical bars are ± 2 standard errors.
1.5
(Millions)
fo
T
r
r
Abundance
o p
T
\
01
>
'i 0.3
lU
■
■
a:
0
86 87 88 89
Year
Figure 14
Unsmoothed abundance trends of whitebelly stock of spinner
dolphin Stenella longirostris in the eastern tropical Pacific,
estimated from research (solid line) and tuna vessel data. Ver-
tical bars are ±2 standard errors.
Acknowledgments
We are grateful to Dr. J. Joseph, Dr. M. Hall, and
Dr. M. Scott for comments on the methods outlined
here, and to two reviewers and Dr. L. Jones for their
constructive comments and criticisms. We also ac-
knowledge the recent and continuing efforts of the
Southwest Fisheries Science Center to evaluate
methods for analyzing tuna vessel and research vessel
sightings data; their program of work forced us to
address more carefully the issue of how to estimate and
test for trends in abundance.
Citations
Anganuzzi, A. A., and S.T. Buckland
1989 Reducing bias in estimated trends from dolphin abun-
dance indices derived from tuna vessel data. Rep. Int. Whal-
ing Comm. 39:323-334.
Anganuzzi, A. A., S.T. Buckland. and K.L. Cattanach
1991 Relative abundance of dolphins associated with tuna in
the eastern tropical Pacific, estimated from tuna vessel sight-
ings data for 1988 and 1989. Rep. Int. Whaling Comm. 41:
497-506.
Au, D., W.L. Perryman, and W. Perrin
1979 Dolphin distribution and the relationship to environmental
features in the eastern tropical Pacific. Admin. Rep. LJ-79-43,
Southwest Fish. Sci. Cent., NMFS, NOAA, La Jolla, CA 92038,
59 p.
Buckland et al.: Estimating abundance of tuna-associated dolphin stocks in the eastern tropical Pacific
1 I
Table 2
Unsmoothed relative-abundance estimates (standard errors in parentheses) of some
eastern tropical Pacific, calculated from research vessel data collected 1986-89.
stocks of dolphin in the
Offshore Eastern
Year spotted dolphin spinner dolphin
Whitebelly
spinner dolphin
Common
dolphin
1986 1527 (261)** 716
1987 2388 (377) 707
1988 2549 (476) 902
1989 3560 (634)** 1200
(152)
(138)
(191)
(254)
657 (140)
750 (159)
821 (174)
759 (248)
1810 (437)*
1026 (298)tTT
5263 (1368)*TT
2586 (587)t
•Estimates differ significantly (P<0.05)
t Estimates differ significantly (P<0.05)
••Estimates differ significantly (P<0.01)
TT Estimates differ significantly (P<0.01)
Table 3
Unsmoothed relative-abundance estimates (standard errors in
parentheses) of some
stocks of dolphin in the |
eastern
tropical Pacific
calculated from tuna vessel data collected 1986-89.
Offshore
Eastern
Whitebelly
Common
Year
spotted dolphin
spinner dolphin
spinner
dolphin
dolphin
1986
3484
(342)
590
(118)
595
(119)
532
(159)
1987
3627
(420)
363
(100)
937
(170)
271
(132)
1988
3048
(439)
665'
(119)
575
(109)
487
(167)
1989
3640
(337)
ficantly (P<0.05)
381*
(74)
748
(105)
408
(111)
* Estimates differ signi
Buckland, S.T.
1984 Monte Carlo confidence intervals. Biometrics 40:
811-817.
Buckland, S.T., and A. A. Anganuzzi
1988a Trends in abundance of dolphins associated with tuna
in the eastern tropical Pacific. Rep. Int. Whaling Comm. 38:
411-437.
1988b Comparison of smearing methods in the analysis of
minke sightings data from IWC/IDCR Antarctic cruises. Rep.
Int. Whaling Comm. 38:257-263.
Burnham, K.P., D.R. Anderson, and J.L. Laake
1980 Estimation of density from line transect sampling of
biological populations. Wildl. Monogr. 72, 202 p.
Cooke, J.G.
1985 On the relationship between catch per unit effort and
whale abundance. Rep. Int. WhaUng Comm. 35:511-519.
Edwards, E.F., and P.M. Kleiber
1989 Effects of nonrandomness on line transect estimates of
dolphin school abundance. Fish. Bull., U.S. 87:859-876.
Gerrodette, T., and P.R. Wade
1991 Monitoring trends in dolphin abundance in the eastern
tropical Pacific using research vessels over a long sampling
period: Analysis of 1989 data. Rep. Int. Whaling Comm. 41:
511-515.
Hall, M.A., and S.D. Beyer
1988 Incidental mortality of dolphins in the eastern tropical
Pacific. Rep. Int. Whaling Comm. 38:439-441.
Hammond. P.S.. and J.L. Laake
1983 Trends in estimates of abundance of dolphins (SteneUa
spp. and Delphinus delphis) involved in the purse-seine fishery
for tunas in the eastern Pacific Ocean, 1977-81. Rep. Int.
Whaling Comm. 33:565-588.
Holt, R.S.
1985 Estimates of abundance of dolphin stocks taken inciden-
tally in the eastern tropical Pacific yellowfin tuna fishery.
Admin. Rep. LJ-85-20, Southwest Fish. Sci. Cent., NMFS,
NOAA, La JoUa, CA 92038, 32 p.
1987 Estimating density of dolphin schools in the eastern
tropical Pacific Ocean by line transect methods. Fish. Bull.,
U.S. 85:419-434.
Holt, R.S., and J.E. Powers
1982 Abundance estimation of dolphin stocks involved in the
eastern tropical Pacific yellowfin tuna fishery determined from
aerial and ship surveys to 1979. Tech. Memo. 23, Southwest
Fish. Sci. Cent., NMFS, NOAA, La Jolla, CA 92038, 95 p.
Holt, R.S., and S.N. Sexton
1989 Monitoring trends in dolphin abundance in the eastern
tropical Pacific using research vessels over a long sampling
period: Analyses of 1987 data. Rep. Int. Whaling Comm.
39:347-351.
1990a Monitoring trends in dolphin abundance in the eastern
tropical Pacific using research vessels over a long sampling
period: Analyses of 1986 data, the first year. Fish. Bull., U.S.
88:105-111.
,2 Fishery Bulletin 90(1). 1992
1990b Monitoring trends in dolphin abundance in the eastern
tropical Pacific using research vessels over a long sampling
period: Analyses of 1988 data. Rep. Int. Whaling Comm.
40:471-476.
Perrin, W.F., P.A. Akin, and J.V. Kashiwada
1991 Geographic variation in external morphology of the spin-
ner dolphin Stenella longirostris in the eastern Pacific and im-
plications for conservation. Fish. Bull., U.S. 89:411-428.
Polacheck, T.
1987 Relative abundance, distribution and inter-specific rela-
tionship of cetacean schools in the eastern tropical Pacific.
Mar. Mammal Sci. 3:54-77.
Sexton, S.N., R.S. Holt, and D. DeMaster
1991 Investigating parameters affecting relative estimates in
dolphin abundance in the eastern tropical Pacific from research
vessel surveys in 1986, 1987, and 1988. Rep. Int. Whaling
Comm. 41:517-524.
Smith, K.L.
1988 Calibration and smoothing of relative dolphin abundance
estimates. MSc. diss., University of Strathclyde.
Smith, T.D.
1983 Changes in sizes of three dolphin {Sfenella spp.) popula-
tions in the eastern tropical Pacific. Fish. Bull., U.S. 81:1-14.
Velleman, P.F.. and D.C. Hoaglin
1981 Applications, basics and computing of exploratory data
analysis. Duxbury Press, Boston.
Abstract.- Osteological differ-
ences confirm the validity of two spe-
cies of Grammatorcynus, G. bicari-
natus (Quoy and Gaimard 1825) and
the long-recognized G. bilineattis
(Riippell 1836). In addition to having
fewer gill rakers (12-15 vs. 18-24),
a smaller eye (3.1-4.6% vs. 4.0-6.0%
FL), small black spots on the lower
sides of the body, and reaching a
larger size (110cm FL vs. 60cm), G.
bicarinatus differs from G. biline-
atus in having a shorter neurocra-
nium, shorter parasphenoid flanges,
lower posterior edge of maxillary
shank, shorter quadrate process,
narrower first postcleithrum, wider
ethmoid, wider vomer, wider lach-
rymal, longer teeth, wider palatine
tooth patch, wider opercle, and a thin
posttemporal shelf between the
anterior processes. All but one of the
16 osteological differences previous-
ly found between Grammatorcynus
bilineatus and Scomberomorus and
Acanthocybium are confirmed with
the inclusion of G. bicarinatus in the
genus. Grammatorcynus bilineatus
is widespread in tropical and sub-
tropical waters of the Indo-West
Pacific from the Red Sea to Tokelau
Islands in Oceania. The range of G.
bicarinatus is restricted to the west-
ern and eastern coasts of Australia
and southern Papua New Guinea.
Morphology, systematics, and biology
of the double-lined mackerels
[Grammatorcynus, Scombrldae)
Bruce B. Collette
Systematics Laboratory. National Marine Fisheries Service, NOAA
National Museum of Natural History, Washington, DC 20560
Gary B. Gillis
Observer Program, Alaska Fisheries Science Center, National Marine Fisheries Service
NOAA, 7600 Sand Point Way NE, Seattle, Washington 981 15-0070
Current address: Department of Ecology and Evolutionary Biology
University of California, Irvine, California 92715
Manuscript accepted 18 December 1991.
Fishery Bulletin, U.S. 90:13-53 (1992).
Until recently, most authors consid-
ered the genus Grammatorcynus to
be monotypic (Fraser-Brunner 1950,
Silas 1963, Zharov 1967, Collette
1979). Electrophoretic work (Lewis
1981, Shaklee 1983) indicated there
were two species of double-lined mack-
erels in Australia. This was confirmed
by Collette (1983) who showed there
are two species: the double-lined
mackerel or scad G. bilineatus, (Riip-
pell 1836), widespread in the Indo-
West Pacific, with more gill rakers
(18-24), a larger eye (4.0-6.0% FL),
and a smaller maximum size (60 cm
FL); and the shark mackerel G. bica-
rinatus (Quoy and Gaimard 1825),
restricted to the waters of northern
Australia and southern New Guinea,
with fewer gill rakers (12-15), a
smaller eye (3.1-4.6% FL), and a
larger maximum size (110 cm FL). All
morphological information concern-
ing Grafnimatorcynus in Collette
(1979) and Collette and Russo (1985b)
was based solely on G. bilineatus.
The purposes of this paper are to
describe osteological differences be-
tween the two species of Gramma-
torcynus, redefine the genus and
both species, and summarize the
literature on both species. The paper
is divided into two parts. Part 1,
Comparative Morphology, contains
descriptions and illustrations of mor-
phometry, meristic characters, soft
anatomy, and osteology of the two
species of Grammatorcynus; com-
parisons are made with Scombero-
morus and Acanthocybium- where
appropriate. Part 2, Systematics and
Biology, contains a generic descrip-
tion and accounts of both species, in-
cluding synonymy, types of nominal
species, diagnoses (based on char-
acters from the first section), size,
biology, interest to fisheries, geo-
graphic distribution, and material
examined.
Methods and materials
Methods are those used by Collette
and Russo (1985b) in a revision of
Scomberomorus, and by Collette and
Chao (1975) in a revision of the
bonitos (Sardini).
Material of Grammatorcynus is
listed at the end of each species ac-
count; 80 specimens of G. bilineatus
and 11 G. bicarinatus. Abbreviations
of institutions housing the material
follow Leviton et al. (1985). Com-
parative material oi Scomberomorus
and Acanthocybium was listed in the
species accounts in Collette and
Russo (1985b).
13
14
Fishery Bulletin 90(1). 1992
'^"•iirairi'^i-itlliiillftlM'i"'^-
B
Figure I
Species of Gr animator cynus. (A) G. bilineatus (from Evermann and Seale 1907. fig. 3, holotype oi Nesogrammus piersoni, 372mm
FL, Philippine Is.); (B) G. bicarinatus (from McCulloch, 1915, p. 1, fig. 1, 925 mm FL, New South Wales, Australia).
Part 1: Comparative morphology
Morphological characters useful for distinguishing be-
tween species of Grammatorcynus and for evaluating
phylogenetic relationships of the genus are divided into
six categories: lateral line, color pattern, morphometry,
meristic characters, soft anatomy, and osteology.
Lateral line
The genus Grammatorcynus differs from all other
genera of Scombridae in having two lateral lines, hence
their common name, double-lined mackerels. The
dorsal-most lateral line is slightly convex, originates
near the dorsal portion of the opercle, and continues
posteriorly until it converges with the second lateral
line, just anterior to the median caudal keel. The sec-
ond lateral line originates from the first at a point
below the first four spines of the dorsal fin. It starts
ventrally, running under, or just posterior to, the pec-
toral fin, and abruptly turns into a concave line that
continues posteriorly until meeting the dorsal lateral
line (Fig. 1). The function of this additional lateral line
is unknown. The characteristic two lateral lines are
discernible in specimens as small as 56.9mm SL
(Nishikawa 1979:133). Anomalies in the pattern of the
lateral lines are occasionally found, but none appear
to be species specific (Fig. 2; Silas 1963: fig. 3).
Color pattern
Dark spots are usually found on the ventral portion of
G. bicarinatus (Fig. IB). The spots are smaller than
the pupil, originate near the ventral border of the oper-
Collette and Gillis Osteological differences between two species of Grammatorcynus
15
Figure 2
Variations in lateral line pattern in Grammatorcynus. (a)
Usual pattern in G. bilineatus; (b-d) variations in pattern in
G. bilineatus; (b) Australia, 410mm FL; (c) Queensland,
^16mm FL; (d) Queensland, 400mm FL; (e) usual pattern in
G. bicarinatus; (f) variation in pattern in G. hicarinatus.
Western Australia, 765 mm FL.
28
22
ORBF, mm
0^ 'A
10
/ / "
)
1 1 1 1
15 46 77 ' 108 139 170
HDL, mm
Figure 3
Orbit length (ORBF) compared with head length (HDL) in
Grammatorcynus. Open circles = G. bilin£atus. squares =
G. bicarinatus.
shows the range and mean of all the characters as
thousandths of fork length, and eight of the characters
as thousandths of head length (Table 1). Scatter
diagrams, with regression lines, show two of the best
morphometric characters: G. hicarinatus has a smaller
orbit (Fig. 3), and a longer first dorsal fin base (Fig. 4).
MerJstJc characters
Numbers of fin rays (first dorsal spines, second dorsal
rays, dorsal fmlets, anal rays, anal finlets, and pectoral
rays), gill rakers, and teeth on the upper and lower jaws
are systematically valuable in Grammatorcynus. They
are discussed in the relevant osteological sections of
the paper.
culum, and continue posteriorly to the anal fin. They
are found below the ventral lateral line on both sides
of the fish. No spots were present in the two smallest
specimens examined (AMS IB.5207-8, 306-315mm
FL). Spots are never present in G. bilineatus (Fig. lA).
Morphometric characters
In addition to fork length, 26 measurements were
routinely made on all specimens. Several morphometric
characters separate the two species. A summary table
Soft anatomy
Viscera Emphasis was placed on the appearance of
the viscera in ventral view, after removal of an oval
segment of the belly wall. Previous descriptions of the
viscera of Grammatorcynus include Kishinouye (1923),
Silas (1963), and Collette and Russo (1985b).
The anterior end of the liver abuts the transverse
septum anteriorly in the body cavity. The liver has
three lobes. The right and left lobe are longer than the
middle lobe, with the right lobe being longest (Fig.
5c-d). The liver is similar in shape in Scomberomorus,
16
Fishery Bulletin 90(1). 1 992
rable 1
Morphometric comparison
of Grammatorcynus
bilineatus and G. bicarinatus.
Character
G
. bicarinatus
G
bilineatus
N
Min
Max
Mean
SD
N
Min
Max
Mean
SD
Fork len^h (thousandths)
10
306
825
551
186
64
226
575
408
77
Snout-A
7
596
626
613
10
61
581
641
606
13
Snout-2D
7
536
558
549
8
61
528
619
547
14
Snout- ID
9
267
301
280
11
64
276
322
295
9
Snout-P2
9
234
272
253
13
63
236
306
258
12
Snout-Pl
9
197
230
216
10
63
199
245
226
9
P1-P2
10
91
255
115
49
62
90
135
101
7
Head length
10
191
223
207
9
64
197
236
218
7
Max. body depth
8
177
210
192
13
57
164
234
196
14
Max. body width
8
105
129
115
8
56
91
136
114
9
PI length
10
118
137
127
5
63
106
142
126
8
P2 length
10
65
81
74
5
63
70
87
77
3
P2 insertion-vent
7
313
345
332
12
62
262
354
328
14
P2 tip-vent
9
238
281
260
15
61
228
275
251
10
Base ID
9
253
272
264
6
63
207
261
235
11
Height 2D
6
97
111
103
5
54
82
116
98
7
Base 2D
10
76
102
90
8
62
68
118
102
9
Height A
10
94
116
104
8
49
67
114
94
9
Base A
9
66
91
80
8
63
73
105
87
7
Snout (fleshy)
10
77
88
81
4
64
58
90
80
5
Snout (bony)
10
64
76
70
4
64
60
80
72
5
Maxilla length
10
91
110
102
6
63
89
108
98
5
Postorbital
10
87
98
92
3
62
78
98
91
3
Orbit (fleshy)
10
31
46
37
5
64
40
60
49
4
Orbit (bony)
10
48
69
59
8
64
53
88
68
6
Interorbital
9
59
71
64
4
62
56
74
62
3
2D-caudal
9
412
475
454
27
60
427
496
470
13
Head length (thousandths)
11
64
165
112
33
64
50
126
89
17
Snout (fleshy)
11
379
410
393
8
64
248
397
366
21
Snout (bony)
11
313
356
340
16
64
281
357
329
16
Maxilla length
11
475
510
495
12
63
420
480
448
15
Postorbital
11
412
471
446
17
62
350
450
419
15
Orbit (fleshy)
11
164
211
179
16
64
191
257
226
14
Orbit (bony)
11
238
319
282
25
64
252
381
313
24
Interorbit
10
274
322
308
13
62
253
327
283
13
but in Acanthoeybium the right and left lobes are about
the same size. Two efferent vessels lead directly from
the anterior surface of the liver into the sinus venosus.
The stomach is sometimes visible in ventral view,
partially covered by the liver and caecal mass, but often
completely hidden. Stomach contents included crusta-
ceans and small fishes.
The pyloric portion of the intestine arises from the
anterior end of the stomach, where the main branches
of the pyloric caeca join the intestine. The caeca branch
and form a dense dendritic conglomeration, the caecal
mass. The intestine continues posteriorly as a simple
straight tube to the anus. A straight intestine is also
found in Acanthoeybium (Fig. 5b) and S. niphonius, but
all other species of Scomberomorus have folds (2 or 4)
in the intestine (Fig. 5a).
Osteology
The osteological description is divided into five sections:
skull, axial skeleton, dorsal and anal fins, pectoral
girdle, and pelvic girdle. Osteological terminology and
organization generally follow that of Collette and Russo
(1985b).
Skull Description of the skull is presented in two sec-
tions: neurocranium (Figs. 6-9) and branchiocranium.
Neurocranium Following a general description of
the neurocranium, the four major regions are dis-
cussed: ethmoid, orbital, otic, and basicranial.
General characteristics In dorsal view (Fig. 6),
the neurocranium of Grammatorcynus is more or less
triangular in shape, narrow at its anterior margin.
Collette and Gillis: Osteological differences between two species of Grammatorcynus
17
205
200 290 380 470 560 o\ : :
FL , mm
Figure 4
Length of first dorsal fin base (BID) compared with
forl< length (FL) in Grammatorcynus. Open circles =
G. bilineatus, squares = G. bicarinatus.
widening posteriorly. It is intermediate in shape be-
tween the elongate neurocranium of Acanthocyhium,
Scomber, and Rastrelliger , and the shorter, wider
neurocranium of Thunnus. The posterodorsal surface
is marked by a median ridge (supraoccipital crest), with
two parallel ridges on either side. These five thin ridges
of bone form six grooves, three on each side: dilator
(very shallow), temporal (quite deep), and supratem-
poral (most easily seen in lateral view) (Allis 1903:49).
The median ridge originates just posterior to the thin,
oval pineal foramen located between the posterior,
median edges of the frontal bones. This ridge becomes
larger posteriorly, and forms the supraoccipital crest.
Internal or temporal ridges originate at the posterior
portion of the frontals (midlevel of the orbit), continu-
ing posteriorly to the epiotic. External or pterotic
ridges also originate near the posterior margin of the
frontals, continuing posteriorly to the pterotic.
Neurocrania of the two species of Grammatorcynus
differ in size, relative to fork length. Length of the
neurocranium, measured from the anterior tip of the
vomer to the posterior margin of the basioccipital, is
slightly longer in G. bilineatus (14-16% FL) than in
G. bicarinatus (13% FL).
Ethmoid region This region is composed of the
ethmoid, lateral ethmoid, and vomer. The nasal bone
lies lateral to the ethmoid and lateral ethmoid, and,
therefore, is included here.
Ethmoid The ethmoid (dermethmoid) has a
smooth flat dorsal surface that is partially overlapped
by the frontals. It connects ventrally to the vomer,
posteriorly to the lateral ethmoids, and anterolateral-
ly to the nasals. Its anterior border is nearly straight,
with an anteromedian projection, unlike the relatively
smooth, concave border in Scomberomorus and Acan-
thocybium. The ethmoid is clearly visible in dorsal view
(Fig. 6), and is wider, relative to the length of the
neurocranium, in G. bicarinatus (width 25-28% of
length) than in G. bilineatus (19-21%).
Lateral ethmoid The lateral ethmoids (pareth-
moids) are massive, paired bones that extend down-
ward from the middle region of the frontals and form
the anterior margin of the orbit and the posterior and
mesial walls of the nasal cavity. The ventral surface
of the lateral ethmoid bears an articulating surface for
the palatine, and the posterolateral process serves as
an articulation surface for the lachrymal. The lateral
expansion of the bone is greater in G. bicarinatus
(45-50% of neurocranium length) than in G. bilineatus
(39-42%) (Fig. 8).
Vomer The anterior process of the vomer bears
a circular or oval patch of fine teeth on its ventral sur-
face. Its pointed posterior end is firmly ankylosed dor-
sally with the parasphenoid. The anterior process is
wider in G. bicarinatus (16-18% of neurocranium
length) than in G. bilineatus (13-15%) (Fig. 8).
IMasal The nasal bones are flat, elongate bones
that articulate with the lateral edge of the frontals.
They project out beyond the ethmoid and, from a dor-
sal view, reach about as far anteriorly as the vomer.
There is no such projection of the nasal bones in
Scomberomorus or Acanthocybium. Length divided by
width is 2.8-3.4 in Grammatorcynus, which is inter-
mediate between the ranges oi Scomberomorus (2.0-
3.1) and Acayithocybiuyn (3.1-4.2). The anterior end
of the bone forms a short, slightly angled arm. No
differences were found between the nasals of the two
species of Grammatorcynus.
Orbital region The orbit (Fig. 7) is surrounded
by the posterior wall of the lateral ethmoid, the ven-
tral side of the frontal, the pterosphenoid, sphenotic,
prootic, suborbital, and lachrymal bones. The left and
right orbits are partially separated by the basisphenoid.
The sclerotic bones enclose the eyeballs.
The orbit of G. bilineatus is larger than that of
G. bicarinatus (Fig. 7), reflecting the difference in orbit
length (Fig. 3). The maximum height of the orbit
measured from the parasphenoid to the pterosphenoid
is 24-25% of neurocranium length in G. bilineatus vs.
16-17% in G. bicarinatus. Orbit length in G. bilineatus
is 51-54% of neurocranium length vs. 47-49% in
G. bicarinatus.
Fishery Bulletin 90(1), 1992
9
C
LIVER
^:::::>::-M CAECAL MASS
• • • • ^
■ • • • •
• • • • i
I NTESTI NE
GONAD
P?aS?X^
t^i:
STOMACH
GALL BLADDER
^ U
RINARY BLADDER
GAS BLADDER
Figure 5
Ventral view of viscera, (a) Scomberomorus
tnaculatus, Georgia, 290mm FL; (h) Acan-
thocybium solandri, Campeche Banks, Mex-
ico, 1280mm FL; (c) Grammatorcynus
bilineatus, Marshall Is.. 424 mm FL; (d) G.
bicarinatus, Australia.
Frontal The paired frontals form the largest
portion of the dorsal surface of the neurocranium. A
small, elongate oval pineal opening is present between
the posterior ends of the frontals. A larger and more
irregular foramen is present in Acanthocybium, but
Scomberomorus lacks this opening (Collette and Russo
1985b:figs. 11-12).
In Scomberomorus and Acanthocybium, the frontals
form a median ridge that increases in height posteriorly
and joins the supraocoipital crest. Grammatorcynus
lacks this ridge and the supraoccipital crest begins
posterior to the pineal opening, giving the top of the
skull a much flatter appearance than in the other two
genera.
In ventral view (Fig. 8), the left and right frontals
articulate with the pterosphenoids at the anterior end
of a median opening into the brain cavity. The ridge
around the anterior end of this space forms a point and
Collette and Gillis Osteological differences between two species of Grammatorcynus
19
SPHENOTIC
PTEROTIC
FRONTAL
NASAL
INTERCALAR
EPIOTIC
EXOCCIPITAL
VOMER
ETHMOID
_ LATERAL ETHMOID
a
FIRST VERTEBRA
SUPRAOCCIPITAL
PARIETAL
Figure 6
Dorsal view of skulls in Grammatorcynus. (a) G. bilirwatiis, Scott Reef, Timor Sea, 453mm FL; (b) G. bicarinatus.
Western Australia, Exmouth Gulf, 765 mm FL.
extends almost to the ethmoid in G. bilineatus. The
ridge curves around the anterior end of the space and
ends distinctly more posteriorly in G. bicarinatus. This
difference cannot be seen in the ventral view of the
skulls (Fig. 8) because the median part of the opening
is obscured by the parasphenoid, so a separate outline
figure has been made (Fig. 9).
Pterosphenoid The pterosphenoids (alisphe-
noids) form the posterodorsal margin of the orbit. They
serve as the base for the median basisphenoid, and abut
the prootics posteriorly and the frontals and sphenotics
laterally.
20
Fishery Bulletin 90(1). 1992
SUPRAOCCIPITAL CREST
PTEROSPHENOID
FRONTAL
ETHMOID
PTEROTIC
LATERAL ETHMOID
a
FIRST VERTEBRA
EXOCCIPITAL
PROOTIC
BASISPHENOID
Figure 7
Lateral view of skulls in Grammatorcynus. (a) G. hUineatus, Scott Reef, Timor Sea, 453 mm FL; (b) G. hkariiiatus.
Western Australia, Exmouth Gulf, 765mm FL.
Sclerotic The sclerotic bones consist of two
thickened, semicircular segments connected by carti-
lage on the inner surface and by corneal membranes
on the outside. The sclerotic bones of Grammatorcynus
are relatively larger and thinner compared with Scom-
beromorus and Acanthocybium.
Basisphenoid The basisphenoid is a small,
median, Y-shaped bone that connects the prootics and
pterosphenoids dorsally with the parasphenoid ventral-
ly (Fig. 7). The dorsal compressed vertical base bears
a slight anterior process, but no posterior process. This
is similar to the condition in Scomber omorus, but the
anterior process is much shorter in Grammatorcynus.
The basisphenoid is longer in G. bilineatus since the
height of the orbit is greater in this species compared
with G. bicarinatus.
Infraorbitals The bones of the infraorbital
series (Fig. 10) enclose the infraorbital branch of the
lateral sensory canal system. The canal enters the
infraorbital series at what is usually considered the last
element (dermosphenotic), and continues around the
orbit, terminating on the first infraorbital (lachrymal).
The lachrymal, the first and largest element, is
elongate with a mesially-directed articular process just
anterior to the middle of the bone. It covers part of
the maxilla, and articulates with the lateral ethmoid
Collette and Gillis. Osteological differences between two species of Grammatorcynus
SPHENOTIC
FRONTAL
LATERAL ETHMOID
PROOTIC
V — INTERCALAR
BASIOCCIPITAL
VOMER
a PARASPHENOID
FIRST VERTEBRA
EXOCCIPITAL
PTEROSPHENOID
PTEROTIG
Figure 8
Ventral view of skulls in Grammatorcynus. (a) G. hilineatus, Scott Reef, Timor Sea, 453mm FL; (b)
G. bicarinatus. Western Australia, Exmouth Gulf, 765 mm FL.
dorsally by the articular process. The process is larger
in G. bicarinatus, making the lachrymal wider (30-35%
of total bone length) than in G. hilineatus (27-30%).
The anterior portion has a small notch in it, much more
indistinct than the forked anterior region in Scorn-
beromorus (Fig. 10a). The posterior region is distinct-
ly forked, with the ventral arm being wider and longer
than the dorsal arm.
The second infraorbital connects to the forked pos-
terior region of the lachrymal. It is a small, elongate
bone. The third infraorbital is an elongate, tubular bone
that connects to the posterior portion of the second
22
Fishery Bulletin 90(1). 1992
a
Figure 9
Outline of pterosphenoid opening on ventral side of
skull in Grammatorcynus. (a) G. bilineatus, Scott
Reef, Timor Sea, 453 mm FL; (b) G. bicarinatus,
Western Australia, Exmouth Gulf, 765mm FL.
infraorbital. It has a large, mesial, shelflike extension
(subocular shelf of Smith and Bailey 1962). The fourth
through penultimate elements total 13 in a specimen
of G. bilineatus (Fig. 10c), are small, and are easily lost
with cheek scales during dissection. No special effort
was made to compare these bones in the two species.
Otic region This region encloses the otic cham-
ber inside the skull, and is formed by the parietal,
epiotic, supraoccipital, prootic, pterotic, sphenotic, and
intercalar (opisthotic) bones.
Parietals The parietals articulate with the
frontals anteriorly, the supraoccipital mesially, the
pterotics laterally, sphenotics ventrally, and epiotics
posteriorly. There is a short inner lateral crest on the
parietals and epiotics, but this crest does not originate
on the frontals as it does in Scomberomorus and
Acanthocybium.
Epiotics The epiotics are irregular bones
bounded by the parietals anteriorly, the supraoccipital
mesially, the exoccipitals posteriorly, and the pterotics
laterally. The medial process of the posttemporal bone
attaches to a distinct roughened process on the
posterior corner of the epiotic. Scomberomorus has a
roughened area at the posterior end of the fronto-
epiotic crest rather than a distinct process.
Supraoccipital The supraoccipital forms the
dorsomedian portion of the posterior end of the neuro-
cranium. It bears a well-developed crest that continues
forward onto the parietals but stops at the pineal
opening instead of extending all the way forward onto
the frontals as in Scomberomorus. The supraoccipital
consists of a thin crest on a roughly hexagonal base.
The crest extends down over the exoccipitals along the
median line where the dorsal walls of the exoccipitals
suture with each other. It extends posteriorly over the
first vertebral centrum (Fig. 7).
Prootics In ventral view (Fig. 8), the prootics
connect with all the bones in the posterior part of the
neurocranium. Each prootic is bordered ventrally by
the parasphenoid; posteriorly by the basioccipital, ex-
occipital, and intercalar; laterally by the pterotic and
sphenotic; and anteriorly by the parasphenoid and
basisphenoid. The prootics are irregular in shape and
meet each other along the ventromedian line of the
brain case to form the posterior portion of the
myodome.
Pterotics The pterotics form the lateral pos-
terior corners of the neurocranium. Each pterotic is
produced posteriorly to form a spine. A pterotic ridge
continues anteriorly onto the parietal, but does not
extend onto the posterior part of the frontal as it does
in Scomber om^orus. In ventral view (Fig. 8), the
pterotics articulate with the sphenotics anteriorly and
the prootics and intercalars medially.
Sphenotics The sphenotics form the most pos-
terior dorsolateral part of the roof of the orbit. They
continue the outer lateral shelf from the frontals, and
articulate with the pterosphenoid medially and the
prootic and pterotic posteriorly. A fossa at the junc-
ture of the sphenotic and pterotic receives the anterior
condyle of the hyomandibula. In dorsal or ventral view,
the distance between the tips of the two sphenotics is
the widest portion of the cranium, 60-67% the length
of the neurocranium in Grammatorcynus.
Intercalars The intercalars (opisthotics) are flat
bones that form part of the posterior border of the
neurocranium interposed between the pterotics and ex-
occipitals. The anterior portion on the dorsal surface
is concealed by the overlapping pterotic, thus expos-
ing the bone on the dorsal surface less than on the
ventral surface (compare in Figures 6 and 8). Each
intercalar has a roughened area on its dorsal surface
to receive the lateral arm of the posttemporal. There
is no posterior projection from the intercalars in Gram-
matorcynus or Acanthocybium as there is in eight
species oi Scomberomonis, such as S. commerson and
S. concolor (Collette and Russo 1985b: figs. 11a and
12b).
Basicranlal region This region consists of the
parasphenoid, basioccipital, and exoccipital bones, and
forms the posteroventral base of the skull.
Parasphenoid The parasphenoid is a long,
cross-shaped bone. It articulates with the vomer ante-
riorly and forms the ventral axis of the skull. It also
Collette and Gillis Osteological differences between two species of Crammatorcynus
23
dennosphenotic
anterior process
cheek scales
Figure 10
Left infraorbital bones in lateral view, (a) Scomberomorus maculatus, Cape Hatteras. NC, 534 mm FL; (b) Acanthocybium solandri,
Revillagigedos Is., 1068mm FL; (c) Grarmnatocrynus bilineatus, Timor Sea, 453mm FL.
forms the ventral border of the orbits and connects
with the lateral ethmoids, basisphenoid, prootics, and
basioccipital bones dorsally. The lateral wings of the
parasphenoid extend dorsolaterally along the ventral
ridge of the prootic bones on either side, and have
pointed ends which form part of the anteroventral wall
of the posterior myodome. Posteriorly, the parasphe-
noid bifurcates into two lateral flanges that attach
dorsally to the corresponding posteroventral flanges
of the basioccipital bone, and surround the posterior
24
Fishery Bulletin 90(1). 1992
opening of the posterior myodome. These flanges are
longer in G. bilineatus (18-21% of neurocranium
length) than in G. bicarinatus (14%), making the
posterior opening of the posterior myodome larger in
G. bilineatus (Fig. 8). A ventrally projecting median
keel is present in the area anterior to the origin of the
lateral flanges. In ventral view, the parasphenoid nar-
rows posteriorly until near the region of the median
keel, where it widens slightly before the lateral wings.
The anterior portion and the region just anterior to the
lateral wings are about equal in width. In Gramma-
torcynus, the shaft of the parasphenoid is narrower
than that of Scomberomorus and Acanthocybium. In
G. bilineatMS, the contour of the parasphenoid is con-
cave, making the orbit larger than in G. bicarinatus,
in which the parasphenoid is flat (Fig. 7).
BasioccJpital The basioccipital has lateral
flanges on either side of the skull and forms the roof
and lateral walls of the posterior myodome. The lateral
flanges expand ventrally to meet the flat posterior
flanges of the parasphenoid. Anteriorly, the basi-
occipital is attached to the prootics and dorsally with
the exoccipitals. The first vertebral centrum attaches
to the posterior surface of the basioccipital.
Exoccipital The exoccipitals connect the skull
with the first vertebra dorsally. The exoccipital artic-
ulates with the epiotic and supraoccipital bones antero-
dorsally, the intercalars laterally, and with the other
exoccipital posterodorsally. In ventral view, the ex-
occipital articulates with the prootic anteriorly, basi-
occipital ventromedially, and intercalar laterally. In
posterior view, the foramen magnum is framed by the
exoccipitals.
Branchiocranium The branchiocranium is divided
into five sections: mandibular arch, palatine arch,
hyoid arch, opercular apparatus, and branchial
apparatus.
Mandibular arch The man-
dibular arch is composed of the
upper jaw (premaxilla, maxilla,
and supramaxilla) and the lower
jaw (dentary, angular, and retro-
articular). Teeth are borne on the
premaxilla and dentary, and the
number of teeth on these bones
differs between species.
Dentition Long, thin,
slightly laterally compressed
teeth are present in a single row
in the upper and lower jaws of
Grammatorcynus. Scomberomo-
rus has large, triangular, lateral-
ly compressed teeth similar to
those of Acanthocybium, which
are blunter and more tightly compressed. The length
of the jaw teeth differs between the species: G. bica-
rinatus has longer teeth than G. bilineatus (maximum
length 6% vs. 4% dentary length). The number of jaw
teeth in Grammatorcynus also varies. Teeth are often
broken or lost, so the range in mean tooth count may
not reflect accurately the actual number of teeth.
However, the maximum number of teeth is useful.
Grammatorcynus bicarinatus has a lower maximum
tooth count on its upper jaw than G. bilineatus (25 vs.
37), and the same is true of the lower jaw, (23 vs. 32;
Table 2). The maximum number of jaw teeth present
in Scomberomorus is slightly higher than G. bilineatus
(39, range 5-39 in the upper jaw; and 37, range 4-37
in the lower jaw). Collette and Russo (1985b) noted that
in Scomberomorus, the species with the fewest teeth
has the fewest gill rakers and the species with the most
teeth has the most gill rakers. There is a similar cor-
relation in Grammatorcynus: G. bilineatus also has
more gill rakers (18-24 vs. 12-15 in G. bicarinatus).
Premaxilla The premaxilla (Fig. 11) is a long,
curved bone with an arrowhead-shaped anterior end
that extends dorsally and posteriorly as an ascending
process. The posterior shank of the premaxilla is
elongate and bears a row of 14-37 long, thin teeth on
its ventral margin. There are two articular facets for
the overlying maxilla at the junction of the posterior
margin of the ascending process with the shank por-
tion. Ascending processes of both premaxillae are
closely approximated to each other mesially and fit into
the median groove of the ethmoid bone. The ascend-
ing process forms an angle of 55-67° with the shank:
G. bilineatus has a slightly larger angle (60-67°, Fig.
lie) than G. bicarinatus {bb-bi° , Fig. lid). Gramma-
torcynus has a larger angle than any species of Scom-
beromorus excepts, guttatus (60-61°). The ascending
process is 33-40% of the total length of the premax-
Table 2
Number of teeth in
upper
and lower jaws
of Grammatorcynus.
Species
Side
Min
Max
X
Overall x
N
Upper jaw
G. biUnealiiS
L
R
14
14
37
36
23.5
24.5
24.0
39
G. bicarinatus
L
R
14
17
25
24
20.5
20.9
20.7
8
Lower jaw
G. bilineatus
L
R
12
14
32
30
18.6
19.1
18.8
36
G. tiicarinatus
L
R
16
15
20
23
17.5
17.6
17.5
7
Collette and Gillis. Osteological differences between two species of Grammatorcynus
25
'0mv^
Figure 1 1
Left premaxillae in lateral view, (a.) Scomberomorus lineolatus, Cochin, India, 786mm FL, 2x; (b) Acanthocybium solandri,
Miami, FL, 1403mm FL, 1 x ; (c) Grammatorcynus bilineatus, Marshall Is., 424mm FL, 2 x ; (d) G. bicarinatus, Great Barrier
Reef, 563mm FL.
a
Figure 12
Left maxillae in lateral view, (a) Scomber<ynwrus munroi, New Guinea, 512 mm FL, 2 x ; (b) Acanthocybium solandri, Miami, FL, 1403 mm
FL, Ix; (c) Grammatorcynus bilineatus. Timor Sea, 453mm FL, 2x; (d) G. bicarinatus, Great Barrier Reef, 563mm FL.
ilia. This is a small percentage relative to Acantho-
cybium (50%, Fig. lib). ScomberoTnorus is intermediate
(31-48%, Fig. 11a).
Maxilla The maxilla (Fig. 12) is a long, curved
bone surmounting the premaxilla dorsolaterally by
means of an anterior head and ventral sulcus. The head
consists of a thick, massive inner condyle and a small
lateral process. The inner condyle possesses a promi-
nent knob at its dorsolateral aspect that fits into the
articular surface of the vomer, and an anterior, deep
26
Fishery Bulletin 90(1), 1992
Figure 13
Supramaxillae in lateral view in Gramma-
torcyniis. (a) G. bilineatus. Western Aus-
tralia, 460 mm FL; (b) G. bicarinatus,
Australia, 625 mm FL.
concavity facing the inner wall of the premaxilla. Im-
mediately posterior to the head is a shallow depression
that receives the anterior articulating process of the
palatine. The shank of the maxilla is narrow and
somewhat flattened. It remains at a relatively constant
height along its entire length, unlike the shank of
Scomberomorus (Fig. 12a) in which the posterior end
of the shank expands into a flat plate. The posterior
end is distinctly thinner than the middle of the shank
in Acanthocyhium (Fig. 12b).
The head of the maxilla is longer (25-29% of total
maxilla length) in Grammatorcynus than in any species
oi Scomberomorus (18-25%) but shorter than in Acan-
thocyhium (33%, Fig. 12b).
The height of the posterior end of the shank, relative
to the total length of the maxilla, is less in G. bicari-
natus (6-8%, Fig. 12d) than in G. bilineatus (8-11%,
Fig. 12c). Grammatorcynus bilineatus is similar to
those species of Scomberomorus that have the least
well-developed (lowest) posterior expansions: S. multi-
radiatus (8-9%) and S. sinensis (9-11%). Other species
of Scomberomorus, such asS. munroi (Fig. 12a), have
larger posterior expansions (11-15%).
Supramaxilla The supramaxilla (Fig. 13) covers
the posterior end of the maxilla. It is a small, flat bone
that is expanded posterodorsally. The expansion is
much more pronounced in G. bicarinatus (59-76% of
bone length. Fig. 13b) than in G. bilineatus (35-42%,
Fig. 13a).
Figure 14
Left dentaries in lateral view, (a) Scomberoinorus semi-
fasciatiis, Port Moresby, New Guinea, 510mm FL, 2x; (b)
Acanthocybium solandri, Miami, FL, 1403mni FL, Ix; (c)
Grammatorcynus bilineatus, Marshall Is., 424 mm FL; (d)
G. bicarinatus, Western Australia, 765mm FL, Ix.
Dentary The dentary (Fig. 14) is laterally flat-
tened and bears a single row of 12-32 long, thin teeth
on the dorsal margin. Posteriorly, the dentary has two
arms of the same relative width (the ventral arm may
be slightly narrower), un\\ke Scomberomorus (Fig. 14a)
and Acanthocybium (Fig. 14b) where the ventral arm
Collette and Gillis: Osteological differences between two species of Grammatorcynus
27
Figure 15
Left anguiars and retroarticulars in lateral view,
(a) Scomberomorus semifasciatus. Port Moresby,
New Guinea, 510mm FL, 3.5 x ; (b) Acanthocybium
solandri, Miami, FL, 1403mm FL, Ix; (c) Gram-
matorcyniis bilineattis, Papua New Guinea, 382 mm
FL, 4.5 x; (d) G. bicarinatus, Western Australia.
765mm FL, 2x.
is much narrower than the dorsal arm.
The length of the dentary from its
anterior margin to the tip of the ven-
tral arm is 97-109% of the length of
the dorsal arm. The ventral arm is
longer in G. bilineatus (104-109% of
dorsal arm length, Fig. 14c) than in
G. bicarinatus (97-98%, Fig. 14d).
The ventral arm is longer in Gramma-
torcynus than it is in Scomberomorus
(86-97%) and Acanthocybium (91-
96%). Species of Grammatorcynus and
Scomberomorus have a notch on the
anteroventral margin of the dentary
that is absent in Acanthocybium. Acan-
thocybium has a prominent notch on
the anterior margin of the dentary
that is indistinct or absent in Gram-
matorcynus and Scomberomorus.
Angular (Fig. 15) The triang-
ular anterior end of the angular (fre-
quently called articular) fits into the
dentary anteriorly. The posterior end
of the angular bears three large pro-
cesses: the dorsal process, directed
forward and upward; the ventral pro-
cess, directed forward; and the pos-
terior process, directed backward and
upward. The posterior process is
hooked and carries a transverse artic-
ular facet for the quadrate. The length
from the tip of the posterior process
to the tip of the dorsal process is
40-47% of the total length of the bone.
The length from the tip of the poster-
ior process to the tip of the ventral pro-
cess is slightly longer, 44-52% of bone
length. The depth of the angular, mea-
sured from the tip of the dorsal pro-
cess to the tip of the ventral process,
is 36-48% of the total length, with the depth of G.
bicarinatus being greater (44-48%, Fig. 15d) than that
of G. bilineatus (36-41%, Fig. 15c). The ventral pro-
cess is approximately as long or longer than the dor-
sal process in Grammatorcynus. In G. bilineatus, the
ANGULAR
ventral process is 84-105% of the length of the dorsal
process, and in G. bicarinatus the ventral process is
longer than the dorsal process (153-200%). Only Acan-
thocybium (99-148%, Fig. 15b), S. commerson (99-
162%), and S. queenslandicus (115-136%) also have a
28
Fishery Bulletin 90(1), 1992
a
Figure 16
Left palatines in lateral view, slightly rotated to better show tooth patches, (a) Scomberomorus semifaseiatus. New Guinea.
740mm FL, 2x; (b) Scomberomorus commerson, New South Wales, 1155mm FL, Ix; (c) Acanthocybium solandri. Miami,
PL, 1403mm FL, Ix; (d) Grammatorcynus bilineatus. Timor Sea, 453mm FL, 2x.
ventral process as long or longer than the dorsal pro-
cess. All other species of Scomberomorus (Fig. 15a)
have ventral processes that are relatively shorter.
Retroarticular (Fig. 15) The retroarticular bone
(frequently called the angular) is rhomboid and at-
tached firmly, but not fused, to the posteroventral
margin of the angular. No differences were found
between the retroarticulars of the species of Gramma-
torcynus.
Palatine arch The palatine arch consists of four
pairs of bones in the roof of the mouth: palatine,
ectopterygoid, entopterygoid, and metapterygoid.
Palatine The palatine is forked both posteriorly
and anterolaterally (Fig. 16). The dorsal branch of the
anterolateral fork is hooked, and its anterior end
articulates v/ith a facet on the maxilla, immediately
ventral to the nasal. The ventral branch appears almost
indistinct in comparison with the longer ventral branch
oi Acanthocybium (Fig. 16c) and the even longer ven-
tral branch oi Scomberomorus (Fig. 16a-b). In Scom-
beromorus, the ventral branch is longer than the dorsal
branch, which is not true of Grammatorcynus or
Acanthocybium. The distance from the anterior end of
the ventral branch to the end of the external branch
divided by the distance from the tip of the dorsal hook
to the end of the external branch is 118-125% in Gram-
matorcynus, 112-121% in Acanthocybium, and only
87-107% in Scomberomorus. The distance from the tip
of the dorsal hook to the tip of the inner branch divided
by the distance to the tip of the outer branch is 71-75%
in Grammatorcynus, 54-84% in Scomberomorus, and
97-99% in Acanthocybium. Hence, Acanthocybium dif-
fers from both Grammatorcynus and Scomberomorus
in that its posteriorly directed inner branch is almost
as long as the outer branch. The tooth patch is short
and wide in Grammatorcynus (Fig. 16d), more so in
G. bicarinatus (wadth 38-42% of length. Fig. 17b) than
in G. bilineatus (width 26-32% of length. Fig. 17a),
long and narrow in Acanthocybium., and between the
two extremes in Scomberomorus. The teeth are fine in
all three genera, but are a little larger in Gramma-
torcynus and Acanthocybium than in most species of
Scomberomorus.
Ectopterygoid The ectopterygoid is a T-shaped
bone with the top of the T forming its posterior end.
It joins with the entopterygoid dorsolaterally, the
palatine laterally and anteriorly, and the quadrate and
metapterygoid posteriorly (Fig. 18). Dividing the dor-
sal distance (from the anterior end of the bone to the
tip of the dorsal arm) by the ventral distance (from the
anterior end to the tip of the ventral process) results
in a number that is greater than 100% in Gramma-
torcynus (107-116%, Fig. 18c) and Acanthocybium
(103-109%, Fig. 18b), but only 85-100% in Scombero-
Collette and Gillis: Osteological differences between two species of Crammatorcynus
29
Figure 17
Palatine tooth patches in Grammatorcynus.
(a) G. bilineatus, Queensland, 521 mm FL; (b)
G. bicarinatus. Western Australia, 765 mm
FL.
Figure 18 (right)
Left suspensoria in mesial view, (a) Scomberoirwrus
semifasciatus, New Guinea, 510 mm FL, 2.5 x ; (b) Acan-
thocybium solandri, Revillagigedos Is., 1068mm FL,
1.5 x; (c) Grammatorcynus bilineatus, Marshall Is.,
424mm FL, 2x.
morus (Fig. 18a). The shank is longer in
Acanthocybium than in the other two gen-
era. The posterior edge of the ectoptery-
goid (from the tip of the dorsal process to
the tip of the ventral process) relative to
the ventral distance is long, 63-72% in
Grammatorcynus, and relatively shorter in
Acanthocybium (41-47%) and Scombero-
morus (43-63%).
Entopterygoid The entopterygoid
is elongate and oval in shape (width 35-
41% of length, Collette and Russo 1985b:
fig. 28). The outer margin of the entop-
terygoid is the thickest part of the bone
and attaches to the inner margin of the ec-
topterygoid. The entopterygoid also con-
nects with the palatine anteriorly and the
metapterygoid posterolaterally. The
mesial and posterior borders are free from
contact with other bony elements. The
dorsal surface is roughly convex. The dorsal surface
is similarly convex in Acanthocybium, but the dorsal
surface in Scomberom^rcnis is concave. The ventral sur-
HYOMANDIBULA
METAPTERYGOID
face is convex in all three genera, and it forms the
major part of the buccal roof. Scomberomonts contains
species that have both narrower (S. commerson, width
30
Fishery Bulletin 90(1). 1992
23-28% of length) and wider (S. maculatus width
41-42% of length) entopterygoids. The entopterygoid
of Acanthocybium (30-35%) is slightly narrower than
that of Grammatorcynus (35-41%).
Metapterygoid The metapterygoid is a flat,
quadrangular or somewhat triangular bone (Fig. 18).
The posterodorsal margin of this bone is deeply
grooved to receive the hyomandibula. The dorsal por-
tion is strongly ankylosed to the lamellar region of the
hyomandibula. The ventroposterior margin abuts the
lowermost portion of the symplectic process of the
hyomandibula, but does not touch the hyomandibula.
There is a relatively long slit beween the two bones
through which the hyoidean artery passes (Allis 1903).
The ventral border is divided into two portions: the
horizontal portion in contact with the quadrate and the
anterior oblique portion ankylosed to the ectopterygoid.
On the mesial surface, the metapterygoid has a
triangular-shaped area that forms an interdigitating
articulation with the upper arm of the ectopterygoid.
The posteroventral margin of the metapterygoid articu-
lates with the dorsal end of the symplectic in Gram-
matorcynus (Fig. 18c) and Acanthocybium (Fig. 18b),
but does not do so in most species of Scomberomorus
(Fig. 18a). The posterior, horizontal part of the ven-
tral border is shorter than the anterior oblique part in
Grammatorcynus (anterior part 132-181% of posterior
part) and Acanthocybium (188-218%); however, in
Scomberomorus the posterior part is longer than the
anterior part (anterior part 39-86% of posterior part).
Hyoid arch The hyoid arch is composed of the
hyomandibula, symplectic, quadrate, hyoid complex
(hypohyals, ceratohyal, epihyal, interhyal, and the
seven branchiostegal rays), and two median unpaired
bones, the glossohyal and urohyal.
Hyomandibula The hyomandibula is an in-
verted L-shaped bone (Fig. 18) connecting the man-
dibular suspensorium and opercular bones to the
neurocranium. Dorsally, there are three prominent con-
dyles. The long dorsal condyle forms the base of the
L and fits into the fossa at the junction of the pterotic
and sphenotic bones. The anterior condyle articulates
with the ventral fossa of the pterotic, and the lateral
process is attached to the inside of the opercle.
Anterolaterally, the hyomandibula is drawn out into a
lamellar region that joins the metapterygoid. Postero-
laterally, it has a long articulation with the preopercle.
Ventrally, the hyomandibula has a long symplectic pro-
cess; at the posterodorsal corner there is a small,
sometimes almost indistinct spine. A strong vertical
ridge extends from the ventral margin to just below
the dorsal border, where it then curves anteriorly to
confluence with the anterior condyle. The areas lying
anterior and posterior to this ridge are grooved for
articulation with the metapterygoid and preopercle.
respectively; in situ, only the ridge and a portion of the
upper broader surface are visible exteriorly. The upper
surface of the symplectic is connected to the ventral
border of the hyomandibula by way of a cartilage, best
developed in Acanthocybium. There is one deep fossa
on the inner surface of the hyomandibula in Gramma-
torcynus (Fig. 18c) and Scomberomorus (Fig. 18a);
there are two such fossae m Acanthocybium (Fig. 18b).
The posterodorsal spine, which is quite small in
Grammatorcynus and in most species of Scomberomo-
rus, is best developed in Acanthocybium, S. commer-
son (Devaraj 1977), and S. queenslandicus. The max-
imum width (tip of anterior condyle to outer margin
of posterior condyle) of the hyomandibula is least
relative to the total length (ventral tip to dorsal margin
of dorsal condyle) in Grammatorcynus (width 34-39%
of length) and S. multiradiatus (36-39%). The hyoman-
dibula is widest, relative to length in S. sinensis
(45-52%). Acanthocybium and the other species of
Scomberomorus fall between these two extremes
(39-47%).
Symplectic The symplectic is a small bone that
fits into a groove on the inner surface of the quadrate
(Fig. 18). In Grammatorcynus the symplectic is slightly
wider than it is in Scomberomorus; however, the groove
into which the symplectic fits is narrower in Gramma-
torcynus than in Scomberomorus, so that the symplec-
tic nearly fills the groove in Grammatorcynus and does
not fill the groove in Scomberomoms (Fig. 18a). The
symplectic is greatly expanded at its dorsal end in
Acanthocybium (Fig. 18b). The symplectics in Gram-
matorcynus and Acanthocybium extend well beyond
the dorsal margin and even beyond the dorsal end of
the posterior process of the quadrate to make contact
with the metapterygoids, making them much longer
than the symplectics in most species oi Scomberomo-r-us.
The symplectic of G. bilineatus (Fig. 19a) is longer than
that of G. bicarinatus (Fig. 19b).
Quadrate The lower jaw is suspended from the
cranium by means of the articulating facet of the ven-
tral surface of the triangular quadrate (Fig. 18). The
broad dorsal margin of the quadrate abuts the ventral
border of the metapterygoid. There is a strong process
on the posterior margin of the quadrate that is attached
along the lower anterior arm of the preopercle. This
process is quite long in G. bilineatus (its length mea-
sured from the ventral facet to the tip of the process
is 134-145% of the distance from the ventral facet to
the dorsal margin; Fig. 19a) and Acanthocybium, but
shorter in G. bicarinatus (122-125%, Fig. 19b) and
most species of Scomberomorus.
Hyoid complex This complex includes the two
hypohyals (basihyal of Mago Leccia 1958), ceratohyal,
epihyal, and interhyal bones, and the seven branchio-
stegal rays (Collette and Russo 1985b: fig. 29). The
Collette and Gillis Osteological differences between two species of Orammatorcynus
Figure 19
Quadrate and symplectic in Grammatorcynus. (a) G. biliine-
atus, Western Australia, 460 mm FL; (b) G. bicarinatus,
Queensland, 563 mm FL.
hypohyals, ceratohyal, and epihyal are closely associ-
ated and form a functional unit.
Hypohyals The hypohyals comprise separate
dorsal and ventral elements joined longitudinally. In
lateral view, the ventral hypohyal is clearly larger than
the dorsal hypohyal in Grammatorcynus, but in Scom-
beromorus not quite as large relative to Grammatorcy-
nus. In Acanthocybium the ventral hypohyal is three
times larger than the dorsal hypohyal. Laterally, the
suture that runs between the dorsal and ventral hypo-
hyals curves ventrally at various angles in Gramma-
torcynus and Scomberomorus, but runs almost hori-
zontally in Acanthocybium. Mesially, a pointed lateral
process at the anterodorsal end of the dorsal hypohyal
forms a symphysis with the glossohyal, urohyal, basi-
branchial, and the process of the hypohyal from the
opposite side in Grammatorcynus and Scomberomorus.
Acanthocybium also has a pointed lateral process, but
it appears to be further posterior due to also having
an anterior pointed end to the hypohyals at the junc-
tion of the dorsal and ventral hypohyals. In addition,
Acanthocybium has a prominent anterolateral process
on the ventral hypohyal. The groove for the hyoidean
artery runs along the outer surface of the epihyal,
ceratohyal, and ventral portion of the dorsal hypo-
hyal. In Grammatorcynus the groove in the dorsal
hypohyal is relatively short, extending anteriorly
11-39% of the length of the dorsal hypohyal before
becoming a covered tunnel leading to the inner side of
the dorsal hypohyal. In Scomberomorus the groove
extends 32-53% before becoming a tunnel to the inner
side, and in Acanthocybium the groove extends 29-47%
before becoming a foramen leading to the inner side.
The opening on the inner side appears as a small to
moderate pit, usually located in the ventral portion of
the dorsal hypohyal in Grammatorcynus and Scom-
beromorus.
Ceratohyal The ceratohyal is a long flat bone,
broadest at the posterior end, and with an anteroven-
tral projection that articulates with the posteroventral
notch of the ventral hypohyal. It is the largest bone
of the hyoid complex. Posteriorly, the middle part of
the ceratohyal interlocks with the epihyal by means of
odontoid processes issuing from both elements
(ceratohyal-epihyal suture of McAllister 1968), while
the upper and lower portions are joined by cartilage.
Four acinaciform branchiostegal rays are attached to
the respective articular surfaces along the concave
middle portion of the ventral margin in Gramma-
torcynus and Acanthocybium.. In Scomberomorus the
fifth branchiostegal ray is also usually attached to the
ceratohyal (most posterior part) or on the space be-
tween the ceratohyal and epihyal, not on the anterior
part of the epihyal. In Grammatorcynus and Acantho-
cybium the fifth branchiostegal ray is on the anterior
part of the epihyal. The hyoidean groove runs the
length of the ceratohyal on its lateral surface. The
groove is so deep in 10 species of Scomberomorus
(brasiliensis, commerson, concolor, multiradiatus,
munroi, niphonius, queenslandicus, semifasciatus,
sierra, and tritor) that it forms a thin slit through the
bone, the ceratohyal window or "beryciform" foramen.
This slit is rare in Grammatorcynus and Acantho-
cybium, and is either rare or occasional in the other
eight species oi Scomberomorus . The dorsal margin of
the ceratohyal is usually concave, but sometimes flat
in Grammatorcynus. It is deeply concave in Acan-
thocybium, and varies from concave to convex in
Scomberomorus.
Epihyal The epihyal is a triangular bone that
interlocks anteriorly with the ceratohyal and has a
posterior process that articulates with the interhyal.
Three branchiostegal rays articulate wdth the epihyal
in Gramm.atorcyyius and Acanthocybium. Only two
branchiostegal rays are found on the ventral portion
of the epihyal in Scomberomorus. In Grammatorcynus
the depth of the epihyal is 66-80% of the length from
the smooth anterior margin of the bone to the tip of
the posterior process. Epihyal depth is narrowest in
Acanthocybium (58-62%), and in Scomberomorus it
varies from 68% in S. commersoyi and S. cavalla to 98%
in S. koreanus, vnth intermediate values for the other
species.
Interhyal The interhyal is a small flattened
bone that is attached to the epihyal dorsal to the
posterior process. It is directed obliquely upward and
links the hyoid complex to the hyomandibula and
symplectic. No differences in interhyals were noted.
Glossohyal The glossohyal (basihyal) is a
median bone that supports the tongue and overlies the
32
Fishery Bulletin 90(1), 1992
first basibranchial bone at the
anterior end of the branchial
arch. In Grammatorcynus there
is a quadrangular to oval tooth
plate fused to and covering the
dorsal surface of the glossohyal
(Collette and Russo 1985b: fig.
30). No tooth plate is present in
Acanthocybium or Scomberomo-
rus. The glossohyal of Gramma-
torcynus is slightly wider (width
47-55% of length) than the
roughly rod-shaped or conical-
shaped glossohyal of Scombero-
morus (35-54%) and the flat-
tened spatulate glossohyal of
Acanthocybium (42%).
Urohyal The urohyal is a
compressed, median, unpaired
bone (Fig. 20). The anterior end
of this element lies between, and
is connected with, the hypohyals
of the left and right sides. The
most prominent difference in the
urohyal of Grammatorcynus is
that in dorsal view, the posterior
end of the dorsal margin is tri-
partite (Fig. 20c-d) instead of
forked, as it is in Scomberomorus
and Acanthocybium (Fig. 20a-b).
The dorsal and anterior portions
of the ventral margins are thick-
ened in Grammatorcynus. The
anterior end has an articulation
head; the posterior end is deepest
in Scomberomorus, and much
less deep in Grammatorcynus
due to the convex shape of the
ventral margin. The maximum
depth of the urohyal in Gramma-
torcynus is 15-20% of the length
of the dorsal margin. The max-
imum depth in Acanthocybium is
not as great as this (13-15%), and in Scomberomorus
it is greater (16-24%). In Grammatorcynus the ven-
tral margin of the urohyal is relatively short, only
68-71% of the length of the dorsal margin, compared
with 80-91% in Acanthocybium and Scomber orfiorus.
Opercular apparatus Four wide, flat bones
(opercle, preopercle, subopercle, and interopercle) fit
together to form the gill cover.
Opercle The opercle (Fig. 21) is overlapped
laterally on its anterior margin by the posterior half
of the preopercle. The narrow, elongate, articular facet
Figure 20
Urohyals in left lateral view, (a) Scomberomorus queenslandicus, Queensland, 641 mm
FL; (h) Acanthocybium solandri, Indian 0., 1088mm FL; (c) Grammatorcymis bilineatiis,
New Guinea, 382 mm FL; (d) G. bicarinatus, Australia, 663 mm FL. Fig. (a) drawn
twice as large as (b), (c) three times as large. Inset is the posterior end of the dorsal
margin, in dorsal view.
for the opercular process of the hyomandibula is located
on the medial surface of the anterodorsal corner of the
opercle. Grammatorcynus and most species of Scom-
beromorus have a weak process at the posterodorsal
corner. This process is absent in Acanthocybium. In
Grammatorcynus the width of the opercle is 63-79%
of the total length of the bone; G. bicarinatus has a
wider opercle (width 72-79% of length. Fig. 21d) than
G. bilineatus (63-72%, Fig. 21c). Both species of Gram-
matorcynus have narrow opercles compared with the
extremely wide opercles found in Acanthocybium
(Fig. 21b).
Collette and Gillis: Osteological differences between two species of Orammatorcynus
33
a
Figure 21
Left opercles in lateral view, (a) Scomberomorns semifasciatus, New Guinea, 510mm FL; (b) Acan-
thocybium solandri, Revillagigecios Is., 1080mm FL; (c) Grammatorcynus bilirwatiis, Marshall Is., 424mm
FL; (d) G. bicarinatiis, Great Barrier Reef, 563 mm FL.
Preopercle The preopercle is a large, crescent-
shaped flat bone, broadest at the lower posterior angle
(Collette and Russo 1985b: fig. 33). The anterior por-
tion of the bone is thickened into a bony ridge. A series
of 5-7 pores along the lower margin of the ridge
represents the preoperculomandibular canal of the
lateral line system which continues onto the dentary.
On the mesial side, the bony ridge possesses a groove
for attachment to the hyomandibula and the quadrate.
There is a shelf mesial to the anteroventral end of the
preopercle in Acanthocybium that is absent in Gram-
matorcynus and Scomheromorus. The canals leading
to the preopercular pores are visible through the bone
in G. hilineatus and all species oi Scomheromorus, but
these canals could not be seen in the specimens of
G. bicarinatus and Acanthocybium due to the thickness
of the bone. The posterior margin of the preopercle is
distinctly concave in Grammatorcynus and most
species of Scomberomorus. However, it is only slight-
ly concave or flat in Acanthocybium and S. commer-
son. In Grammatorcynus the distance from the anterior
margin of the bony ridge to the posterior end of the
lower lobe is 64-75% of the height of the preopercle
measured from the ventral margin to the dorsal tip of
the bone. In Scomberomorus the lower lobe is 69-80%
of the height of the preopercle. The anterodorsal mar-
gin terminates in a pore similar to the preoperculo-
mandibular lateral-line canal pore at the anteroventral
margin of the bone.
Subopercle The subopercle is a flat, roughly
triangular bone with a prominent anterior projec-
tion (Collette and Russo 1985b: fig. 34). Two ridges
34
Fishery Bulletin 90|1). 1992
converge posteriorly from the anterior projection on
the lateral side of the bone. The upper ridge articulates
with the lower posterior projection of the opercle, and
the lower ridge connects to the posterodorsal margin
of the interopercle. The dorsal ridge is much more
prominent than the ventral ridge and extends over the
main part of the subopercle as a discrete shelf. The
weaker ventral ridge is more difficult to detect in most
specimens of Grammatorcynus. The angle between the
anterior projection and the anterior margin of the
subopercle is acute in Grammatorcynus and S. multi-
radiatus; however, in Acanthocybium and the other
species o{ Scomberomonis the angle is close to 90°. The
length of the anterior projection in Grammatorcynus
varies from 23 to 33% of the length of the anterior
margin dorsal to the projection. The projection is slight-
ly longer (28-33%) in G. bicarinatus than in G. biline-
atus (23-28%). The projection is longest in Acan-
thocybium (36-45%) and shortest in S. commerson
(20-25%), with the rest of the species of Scombero-
morus having projections between 21 and 43%.
Interopercle The interopercle is roughly oval
in shape, narrow at the anterior margin and widening
posteriorly, with a crest on the superior margin
(Collette and Russo 1985b: fig. 35). There is a well-
developed facet on the mesial side to receive the artic-
ular process of the interhyal. The maximum depth of
the interopercle relative to the length of the bone is
35-43% in Grammatorcymis. The maximum depth of
the interopercle is a little greater in Acanthocybium
(40-49%), and much greater in the species of Scom-
beromorus (45-61%). Often there is a well-formed
notch anterior to the crest on the sloping anterior
margin in Grammatorcynus and Scomberomorus,
which is not as well developed in Acanthocybium. The
posterior margin is rounded in Grammatorcynus and
Scomberomorus but divided into two by a notch in
Acanthocybium.
Branchial apparatus The branchial apparatus is
composed of five pairs of gill arches, gill filaments, gill
rakers, pharyngeal tooth patches, and supporting
bones. The general arrangement in Grammatorcynus
is similar to that found in other scombrids such as the
Sardini (Collette and Chao 1975), Thunnus (Iwai and
Nakamura 1964:22, fig. 1; de Sylva 1955:21, fig. 40),
Scomberomorus (Mago Leccia 1958:327, pi. 12; Collette
and Russo 1985b: fig. 36), and Rastrelliger (Gnana-
muttu 1971:14, fig. 6). Most branchial bones bear
patches of tiny teeth.
Baslbranchials The three basibranchials form
an anteroposterior chain. The first and second are
about the same size, and considerably shorter than the
third. The first is covered dorsally by the glossohyal.
In lateral view, the first basibranchial is narrowest
in the middle. In Grammatorcynus and Acantho-
cybium it is elongate. In Scomberomorus it is short with
a wide base where it joins with the second basibran-
chial. The second basibranchial has a prominent notch
in the ventral margin and a distinct groove laterally
that extends from the anteroventral margin to the mid-
dorsal region of the bone. This groove accepts the
anterior end of the first hypobranchial. The third
basibranchial has an expanded anterior end at its junc-
tion with the second basibranchial, and then tapers
posteriorly. A prominent groove is present anteriorly
that accepts the medial anterior end of the second
hypobranchial. A section of cartilage extends poster-
iorly to articulate with the fourth and fifth cerato-
branchials.
Hypobranchial Three hypobranchials are pres-
ent. The first is interposed beween the second basi-
branchial and the first ceratobranchial. The second is
about the same size as the first, fits into a groove on
the third basibranchial, and extends to the second
ceratobranchial. The third hypobranchial is smaller
than the first or second, fits snugly against the pos-
terolateral margin of the third basibranchial, and its
posterior end articulates with the third ceratobranchial.
Ceratobranchlals The five ceratobranchials
are the longest bones in the branchial arches. They
have a deep groove ventrally for the branchial arteries
and veins. The ceratobranchials support most of the
gill filaments and gill rakers. The first three are mor-
phologically similar and articulate with the posterior
ends of their respective hypobranchials. The fourth is
more irregular and attaches to a cartilage posterior to
the third basibranchial. The fifth ceratobranchial is also
attached to the cartilage, has a dermal tooth plate fused
to its dorsal surface, and the complex is termed the
lower pharyngeal bone. It is covered with small con-
ical teeth that are directed slightly posteriad.
Epibranchials The posterolateral end of each
of the four epibranchials is attached to the ends of the
first four ceratobranchials. Each epibranchial bears a
groove posterodorsally for the branchial arteries and
veins. The first epibranchial is the longest and bears
two processes mesially. The anterior process articulates
with the first pharyngobranchial, and the posterior
process attaches with the interarcual cartilage. The
second epibranchial is similar to the first, but slightly
shorter. The anterior end has two processes, an anter-
ior process that attaches to the second pharyngo-
branchial and a posterior process that is coupled with
the third pharyngobranchial by way of an elongate
cartilage. This process is relatively elongate in Gram-
matorcynus, but shorter in Acanthocybium and Scom-
beromorus. The third epibranchial is the shortest in the
series. Laterally, it is attached with the third cera-
tobranchial; mesially, it is attached with the third
pharyngobranchial. An elongate posterodorsal process
Collette and Gillis: Osteological differences between two species of Orammatorcynus
35
is present. This process joins with the fourth epibran-
chial, which is larger than the third and is interposed
between the fourth ceratobranchial and pharyngobran-
chial. It is a curved bone with the angle formed by the
lateral and medial arms being much more acute in
Grammatorcynus than in Acanthocybimn and Scom-
beromorus. A dorsal process arises from the middle of
the bone and attaches to the third epibranchial.
Pharyngobranchials There are four pharyngo-
branchials attached basally to the epibranchial of their
respective gill arch. The first is long and slender,
articulates dorsally with the prootic, and is frequently
called the suspensory pharyngeal (Iwai and Nakamura
1964). The elongate second pharyngobranchial bears
a patch of teeth. The third is the largest element in the
series; it has a broad patch of teeth on its ventral sur-
face, a broad posterior end, and tapers to a narrow
anterior end. In Grammatorcynus and Acanthocybium.
the third pharyngobranchial is shorter than in Scom-
beromorus. The fourth pharyngobranchial also bears
a ventral tooth plate, has a rounded posterior end, and
has an elongate strut (pharyngobranchial stay) mesially
which overlaps the third pharyngobranchial. This stay
is much shorter in Grammatorcynus and Acantho-
cybium than in Scomberomorus.
Gill rakers The hypobranchial, ceratobranchial,
and epibranchial of the first gill arch support a series
of slender, rigid gill rakers. The longest gill raker is
at or near the junction of the upper and lower arches,
between the ceratobranchial and epibranchial. Magnu-
son and Heitz (1971) have clearly shown that there is
a correlation between numbers of gill rakers, gap
between gill rakers, and size of food items in a number
of species of Scombridae.
The number of gill rakers is easily countable and is
an especially useful taxonomic character in differen-
tiating between the two species of Grammatorcynus:
G. bilineatus has more gill rakers (18-24) than
G. bicarinatus (12-15) (Table 3). Acanthocybium dif-
fers from all other genera of Scombridae in complete-
ly lacking gill rakers. Three species of Scomberomorus
have greatly reduced numbers of gill rakers: S. multi-
radiatus (1-4 gill rakers), S. commerson (1-8), and
S. queenslandicus (3-9). Scomberomorus concolor
stands out in having the most gill rakers (21-27). Other
species of Scomberomorus fall between these extremes.
There is a correlation between number of gill rakers
and number of jaw teeth in Grammatorcynus and
Scomberomorus. Species with the fewest gill rakers,
G. bicarinatus and S. multiradiatus, also have the
fewest jaw teeth, and species with the most gill rakers,
G. bilineatus and S. concolor, have the most teeth.
Axial skeleton This section is divided into four parts:
vertebral number, vertebral column, ribs and inter-
muscular bones, and caudal complex.
Vertebral number Vertebrae may be divided into
precaudal (abdominal) and caudal. The first caudal
vertebra is defined as the first vertebra that bears a
notably elongate haemal spine and lacks pleural ribs.
Vertebral counts include the urostyle which bears the
hypural plate. Grammatorcynus has 31 vertebrae,
which is less than Scomberomorus (41-56 vertebrae),
which in turn is less than Acanthocybium (62-64). The
same situation also applies to precaudal and caudal
vertebrae. Both species of Grammatorcynus have 12
precaudal and 19 caudal (except for one specimen of
G. bicarinatus with 1 1 plus 20 caudal). Scomberomorus
has 16-23 precaudal and 20-36 caudal, and Acantho-
cybium has 31-33 precaudal and 31-33 caudal. The
presence of only 31 vertebrae in Grammatorcynus is
a primitive condition agreeing with Scomber and
Rastrelliger, the most primitive members of the Scom-
brinae. The increased number of vertebrae in Acantho-
cybium is clearly a specialization.
Vertebral column The neural arches and spines
are stout and compressed on the first to the fourth
vertebra (especially the first 3) in Grammatorcynus.
They extend farther back, to the fifth or sixth verte-
brae, in most species of Scomberomorus, and extend
farthest, to the seventh vertebra, in Acanthocybium
and S. commerson. Posteriorly, toward the caudal
peduncular vertebrae and caudal complex, the neural
spines bend abruptly backward and cover most of the
neural groove; caudally they merge into the caudal
complex as in Thunnus (Kishinouye 1923, Gibbs and
Collette 1967) and the bonitos (Collette and Chao 1975).
Table 3
Number of gill rakers on first arch in Grammatorcynus.
12
13
14 15 16 17 18 19 20 21
22
23
24
N
X
G. bilineatus
G. bicarinatus
1
-
1 10 15 30
7 3
10
5
1
72
11
20.8
14.1
36
Fishery Bulletin 90(1). 1992
The neural prezygapophyses on the first vertebra are
modified to articulate with the exoccipital where the
vertebral axis is firmly articulated with the skull.
Neural postzygapophyses arise posterodorsally from
the centrum and overlap prezygapophyses posterior-
ly. The postzygapophyses progressively merge into the
neural spine in the peduncular region to disappear by
the last 5-6 vertebrae. The basic structure and ele-
ments of the neural arches and neurapophyses are
similar to those of other scombrids (Kishinouye 1923,
Conrad 1938, Mago Leccia 1958, Nakamura 1965,
Gibbs and Collette 1967, Collette and Chao 1975,
Potthoff 1975, Collette and Russo 1985b).
Variable characters are found on the haemal arches
and haemapophyses. Laterally directed parapophyses,
arising from the middle of the centrum, appear on the
4th-6th vertebrae where the intermuscular bones and
pleural ribs are encountered (see section on Ribs and
Intermuscular Bones). The parapophyses become
broader and longer posteriorly and gradually shift to
the anteroventral portion of the centra. In lateral view,
the first ventrally visible parapophyses are found on
the 6th-7th vertebra in Grammatorcynus, the 7th-9th
in Scomberomorus (usually the 8th), and on the
14th- 15th in Acanthocybium.
Posteriorly, the distal ends of the paired para-
pophyses meet, forming the first closed haemal arch.
The first closed haemal arch is on the 8th vertebra in
Grammatorcymis, 10th-16th vertebra in Scomberomo-
rus, and 25th-28th vertebra in Acanthocybium. This
location is correlated with the total number of verte-
brae. The haemal spines become elongate and point
posteriorly until they abruptly become more elongate
on the first caudal vertebra. The paired pleural ribs (see
section on Ribs and Intermuscular Bones) attach to the
distal ends of the parapophyses and arches and extend
posteriorly to the last precaudal vertebra. The haemal
arches and spines bend posteriorly at the caudal pedun-
cle and then merge into the caudal complex sym-
metrically with the neural arches and spines on the
caudal vertebrae.
Haemapophyses include pre- and postzygapophyses,
but their relative positions are different from those of
the neurapophyses, and they do not overlap. The first
haemal postzygapophyses arise posteroventrally from
the 6th-7th centrum in Grammatorcynus, the 6th-8th
in Scomberomorus, and the 9th-10th in Acantho-
cybium. They reach their maximum length at about the
junction of the precaudal and caudal vertebrae. The
haemal postzygapophyses fuse with the haemal spine
or disappear in the caudal peduncle region. The haemal
prezygapophyses arise from the anterior base of the
haemal arches on the 8th- 11th vertebra in Gramma-
torcynus, the 10th-22nd in Scomberomorus, and the
23rd-25th in Acanthocybium.
Struts between the haemal arch and the centrum
form the inferior foramina. Foramina are present from
the 17th-19th to the 25th-28th vertebra in Gram-
matorcynus, the 21st-33rd to the 35th-52nd in Scom-
beromorus, and the 49th-51st to the 56th-57th in
Acanthocybium.
Ribs and intermuscular bones Pleural ribs are
present from the second or third vertebra posterior to
the 12th-31st vertebra in the three genera. Inter-
muscular bones start on the back of the skull or the
first vertebra and extend to the 10th-30th vertebra.
Correlated with its low number of vertebrae. Gram-
matorcynus has the fewest pleural ribs (10 pairs).
Acanthocybium has the most pleural ribs (30 pairs) in
agreement with its many vertebrae. Species of Scom-
beromorus are intermediate in number of vertebrae and
also in number of pleural ribs (15-21). The first pleural
rib articulates with the centrum of the third vertebra
in Grammatorcynus and most species of Scombero-
morus, and articulates with the centrum of the second
vertebra in Acanthocybium., as noted by Devaraj
(1977:44), and in at least one specimen each of S. com-
m£rson, S. maculatus, and S. sinensis. Pleural ribs
usually extend posteriorly to about the last precaudal
vertebra: 12 in Grammatorcynus, 17-23 in Scombero-
morus, and 31 in Acanthocybium.
Intermuscular bones start on the first vertebra in
Grammatorcynus, Acanthocybium, and some species
of Scomberomorus. In some specimens of at least 13
species of Scomberoynonis, the first intermuscular bone
is attached to the exoccipital on the skull, and in S. con-
color, S. koreanus (also noted by Devaraj 1977), and
S. sierra, it appears to be the usual condition. Gram-
matorcynus has 19-21 pairs of intermuscular bones,
many more than Acanthocybium (only 10 pairs, which
seems odd given its high number of vertebrae and
pleural ribs), but fewer than most species oi Scombero-
morus (20-30 pairs).
Caudal complex Three preural centra support the
caudal fin in Grammatorcynus. In Scomberomorus four
or five preural centra support the caudal fin, and in
Acanthocybium there are five. The urostyle represents
a fusion of preural centrum 1 and the ural centrum
(Potthoff 1975). The urostyle is fused with the trian-
gular hypural plate posteriorly and articulates with the
uroneural dorsally. In Grammatorcynus there is very
little compression of the preural centra. Preural cen-
trum 4 is not shortened at all, preural centrum 3 is
shortened slightly, and preural centrum 2 is shortened
slightly more (Collette and Russo 1985b: fig. 39). In
Acanthocybium and Scomberomorus, preural centra
2-4 are compressed more than any of the preural
centra in Graynmatorcynus, but still not as much as the
centra in the bonitos and tunas (Collette and Chao
Collette and Gillis: Osteological differences between two species of Grammatorcynus
37
1975, Gibbs and Collette 1967).
In Grammatorcynus the poster-
ior-most neural and haemal spine
bend away from the vertebral
axis and parallel the dorsal and
ventral edges of the hypural plate.
In Acanthocybium and Scombero-
monts, three posterior neural and
haemal spines bend away from
the vertebral axis more abrupt-
ly than in Grammatorcynus.
The triangular hypural plate is composed of 5 fused
hypural bones (Potthoff 1975). In some specimens of
Grammatorcynus (G. bilineatus 453 and 521mm FL,
and G. bicarinatus 563mm FL) the dorsalmost (hypural
5) is partially fused with the dorsal part of the hypural
plate (hypurals 3-4). However, in smaller specimens
(382-424 mm FL) such fusion was absent, as is the case
in Scomberomorus and Acanthocybium. There is a
primitive hypural notch present on the middle of the
posterior margin of the hypural plate. This notch is a
remnant of the fusion of the dorsal part of the hypural
plate with the ventral part (hypurals 1-2). The notch
is absent in the more advanced bonitos and tunas
(Collette and Chao 1975).
The parhypural is separate from the ventral hypural
plate in Grammatorcynus and Scomberomorus but is
fused with it in Acanthocybium. This fusion was also
noted by Conrad (1938), Fierstine and Walters (1968),
and Devaraj (1977). The two haemal arches preceding
the parhypural are autogenous in the three genera,
although Devaraj (1977) stated that they were fused
with their centra in Acanthocybium..
The parhypural has a strongly-hooked process, the
parhypurapophysis (or hypurapophysis), at its proximal
end. The parhypurapophysis slopes slightly upwards
similarly in Grammatorcynus and Scomberomorus. In
Acanthocybium it has a right angle and then a level
projection.
There are two epurals as in other scombrids (Potthoff
1975). In shape and size, the anterior epural (1) resem-
bles the neural spine of adjacent preural centrum 3. The
smaller, posterior epural (2) is a free splint located
between the anterior epural and the uroneural and fifth
hypural, which are joined together.
Dorsal and anal fins Grammatorcynus usually has
12 dorsal spines, rarely 11 or 13 (Table 4), fewer than
either Scomberomorus (12-22) or Acanthocybium (23-
27). Dorsal spine counts are roughly correlated with
vertebral number: Grammatorcynus has the fewest
precaudal, caudal, and total vertebrae, and the fewest
dorsal spines, while Acanthocybium. has the most
precaudal and total vertebrae, and the most dorsal
spines.
Number of dorsal
Table 4
spines, second dorsal fin rays, and dorsal finlets
in Grammatorcynus.
Spines
Rays
Finlets
11 12 13
10
11
12
6 7 8
G. bilineatus
G. bicarinatus
4 65 1
10
10
10
55
4
61 9
1 9 1
Table 5
Number of anal fin rays and finlets in
Grammatorcynus .
Rays
Finlets
11 12 13
5 6 7
G. bilineatus 12 42 17
G. bicarinatus 5 3 1
1 61 8
3 7
The range in number of second dorsal fin rays is
10-25 in the three genera. Grammatorcynus has 10-12
rays, 10 in G. bicarinatus and usually 11 in G. biline-
atus (Table 4). There are usually more second dorsal
rays in Acanthocybium (11-16) and Scomberomorus
(15-25).
Dorsal finlets number 6-11 in the three genera.
Grammatorcynus has 6-8, usually 7 in G. bicarinatus,
and usually 6 in G. bilineatus (Table 4). Acanthocybium
has 7-10, and Scomberomomis has 6-11. The total
number of second dorsal elements is the same in both
species of Grammatorcynus, 11-1-6 = 17 in G. biline-
atus, 10-1-7 = 17 in G. bicarinatus.
Anal fin rays show a similar trend to that of dorsal
fin rays. The range in the three genera is 11-29. Gram-
matorcynus has 11-13 (Table 5), similar to Acantho-
cybium (11-14), but much fewer than Scomberomorus
(15-29).
Anal finlets range in number from 5 to 12 in the three
genera. Grammatorcynus has 5-7, usually 6 in
G. bilin£atus, and usually 7 in G. bicarinatus (Table 5),
generally fewer than Acanthocybium (7-10) or Scom-
beromorus (5-12). Again, the total number of anal
elements is the same in both species, 12-1-6 = 18 in
G. bilineatus, ll-i-7 = 18 in G. bicarinatus.
Pectoral girdle The pectoral girdle consists of the
girdle itself (cleithrum, coracoid, and scapula), the
radials to which the pectoral fin rays attach, and a chain
of bones that connect the girdle to the rear of the skull
(posttemporal, supracleithrum, supratemporal, and two
postcleithra).
38
Fishery Bulletin 90(1). 1992
Posttemporal The posttemporal (Fig. 22) is a flat
elliptical bone with two sturdy anterior processes that
attach the pectoral girdle to the neurocranium. The
median (dorsal) process articulates with the dorsal sur-
face of the epiotic. The lateral (ventral) process is
shorter, round in cross section, and its hollow anterior
end articulates with the dorsal protuberance of the
intercalar. There is a thin shelf visible between these
two processes in G. bicarinatus (Fig. 22d) and Scom-
beromorus (Fig. 22a), but this shelf is hidden behind
the flat, posterior portion of the bone in G. bilineatus
(Fig. 22c) and Acanthocybium (Fig. 22b). A variably-
sized notch is present at the middle of the posterior
edge of the flat body of the bone. Grammatorcynus
usually has a distinct, variably-sized anteriorly directed
spine on the ventral margin of the median process
about one-third of the distance from the body of the
bone to the anterior tip of the process. In Acantho-
cybium, there is a separate process extending anterior-
ly from the ventral wall of the median process. This
auxiliary process (Kishinouye 1923) is as long or almost
as long as the median process itself. It ends in a series
of several pointed processes. (Both Conrad 1938 and
Devaraj 1977 referred to the auxiliary process as the
median process.) The lengths of the median and lateral
processes vary among the species under discussion. The
lengths were measured from the midpoint of the shelf
that connects the two processes, to the end of the pro-
cesses. Both the median and lateral processes are
longer, relative to the length of the entire bone, in
G. bilineatus where the shelf is hidden posteriorly
(median process is 53-60% length of entire bone, and
lateral process is 35-40%) than in G. bicarinatus where
the shelf is not hidden, and is found more near the mid-
point of the bone (median process is 49% and lateral
process is 30%). In Acanthocybium (shelf hidden) the
median process is 56-65% the length of the entire bone,
and the lateral process is 27-37%. In Scomberomorus
(shelf evident) the median process is 36-51% and the
lateral process is 15-36%.
Another useful taxonomic character is the presence
(if present, shape is important) or absence of a spine
or process at the base of the lateral process on the inner
surface of the posttemporal. It is present as a wide flap
in Grammatorcynus (Fig. 22c, d), a blunt process in
Acanthocybium (Fig. 20b), and as a shelf with a point
in S. commerson, S. munroi, S. niphonius, S. pluri-
lineatus (Fig. 22a), and sometimes in S. sinensis. It is
absent or small and inconspicuous in the other 13
species of Scomberomorus.
Supracleithrum The supracleithrum is an ovate
bone, overlapped dorsolaterally by the posttemporal
and overlapping the anterior part of the dorsal wing-
like extension of the cleithrum. The anterior border of
the bone on the mesial side is thickened into a ridge.
median process
lateral process
auxiliary process
Figure 22
Lateral view of left posttemporals. (a) Scomberomorus
plurilineatus, South Africa, 910mm FL, Ix; (b) Acan-
thocybium solandri, Revillagigedos Is., 1068mm FL, 1 x ;
(c) Grammatorcynus bilineatus, Queensland, 521 mm
FL, 1.5 X ; (d) G. bicarinatus. Western Australia, 565 mm
FL.
Dorsally there is a small handle-shaped process that
curves into the posterior margin to end in a notch at
the posterodorsal aspect. Both the anterior and pos-
terior borders are extended so that they form humps
in Grammatorcynus (Collette and Russo 1985b: fig. 41).
A branch of the lateralis system extends from the
posterior notch of the posttemporal onto the supra-
cleithrum. This short canal lies ventral to the dorsal
process of the supracleithrum and extends to the
posterior edge of the bone.
Collette and Gillis: Osteological differences between two species of Orammatorcynus
39
The maximum width of the supracleithrum varies
from 43 to 75% of the total length of the bone in the
three genera. It is widest in Gram.matorcynus, width
72-82% (89% in one 475 mm FL specimen oibilineatus)
of length (due to the extensions of anterior and pos-
terior borders). Scomberomorus varies in width from
43% in S. multiradiatus to 62% in S. nipkonius. There
is no evidence that size is a factor in the size of the
supracleithrum in Grammatorcynus as was noted by
Collette and Russo (1985b) for Scomberomorus.
The dorsal process is prominent in Grammatorcynus,
S. cavalla, S. commerson, S. lineolatus, and especially
in Acanthocybium. In other species oi Scomberomorus,
it is either small or less sharply set off from the main
body of the supracleithrum.
Supratemporal The supratemporal is a thin flat
bone having three distinct arms and lying just under-
neath the skin where its lateral arm articulates with
a dorsal articular surface on the pterotic. The anterior-
most arm is the longest, while the ventrally directed
arm is the shortest. The arm directed posteriorly is
intermediate in length. The anterior margin is deeply
concave, and the greatly convex posterior margin
slightly overlaps the dorsal arm of the posttemporal.
The supratemporal bears a prominent lateral line
canal that extends out almost to the tips of all three
arms (Collette and Russo 1985b: fig. 42). In these three
genera, the canal along the anterior margin of the bone
is the longest, and the canal along the lateral side is
next longest. In Grammatorcynus, the first canal is not
branched like it is in most species of Scomberomorus,
and the second canal is relatively longer.
Cleithrum The main body of the cleithrum is cres-
cent-shaped with an anterodorsal spine and a posterior-
ly projecting plate at the upper end (Collette and Russo
1985b: fig. 43). The angle between the spine and the
plate is much smaller in Grammatorcynus and Scom-
beromorus than in Acanthocybium. In Grammatorcy-
nus, the spine does not extend as far dorsally as the
plate. In Acanthocybium and most species of Scombero-
morus, the spine extends about equally as far dorsally
as the plate, and in S. sinensis the spine extends well
beyond the dorsal margin of the plate. In Gramma-
torcynus and most species ot Scomberomorus, the plate
narrows posteriorly. The posterior plate is longer and
of uniform width in Acanthocybium.
The lower part of the cleithrum is large and folded
back upon itself as two walls: one lateral and the other
mesial, which meet at their anterior margins and run
parallel to each other. The mesial wall of the cleithrum
forms a large triangular slit with the coracoid. In
Grammatorcynus and Scomberomorus, the lateral wall
of the cleithrum is narrow enough to allow part of the
slit to be visible in a lateral view. This slit is hidden
in lateral view in the species oi Scomberomorus because
of the great width of the lateral wall of the cleithrum
(Devaraj 1977:46, Collette and Russo 1985b: figs.
43a-b).
Coracoid The coracoid is elongate and more or less
triangular in shape. It connects wath the scapula along
its dorsal edge and with the mesial shelf of the cleith-
rum anterodorsally and an tero ventrally. The coracoid
is relatively wider in Grammatorcynus and Scombero-
morus than in Acanthocybium.
Scapula The anterior margin of the scapula con-
nects to the mesial shelf of the cleithrum. This attach-
ment extends to the posterior projecting plate antero-
dorsally. The scapula is attached to the coracoid
posteriorly and with the first two radials posterodor-
sally. The posterodorsal margin of the scapula is drawn
out into a facet that accepts the most anterior ray of
the pectoral fin. The scapula is pierced by a large, usual-
ly round, foramen near the lateral margin with the in-
ner shelf of the cleithrum. A prominent suture leads
to the dorsal and ventral margin of the scapula from
the foramen. The suture is intermediate in size in
Gram.matorcynus relative to the large sutures present
in Acanthocybium, S. brasiliensis, andS. regalis, and
the small suture in S. koreanus.
Radials The four radials differ in size and shape and
attach directly to the thickened posterior edges of the
scapula and coracoid. The size of the radials increases
posteroventrally. Small foramina are located beween
the 2nd and 3rd and the 3rd and 4th radials counting
posteriorly. In Gramynatorcynus the first two radials,
and sometimes a small portion of the third, attach to
the scapula; the second two, sometimes only one and
a large portion of the second, attach to the coracoid.
In Acanthocybium. and Scomberomorus the upper one-
third of the third radial, along with the first two radials,
always attaches to the scapula, and the ventral two-
thirds of the third radial plus the fourth radial attach
to the coracoid. A much larger foramen is present
between the largest (fourth) radial and the coracoid.
Posteriorly, this foramen is framed by a posterior
process of the upper part of the fourth radial meeting
an anterior process from the posterior margin of the
coracoid. This process is only slightly developed in
Grammatorcynus. The foramen is about equal in size
to, or larger than the scapular foramen in Gramyna-
torcynus and Scomberomorus, whereas in Acantho-
cybium the scapular foramen is much larger.
Pectoral fin rays The first (uppermost and largest)
pectoral fin ray articulates directly with a posterior
process of the scapula. The other rays attach to the
radials. The number of pectoral rays varies from 19 to
26 in the three genera. Grammatorcynus has 21-26
40
Fishery Bulletin 90(1). 1992
pectoral fin rays, similar to Acanthocybium (22-26).
Scombermorus shows greater variation (19-26) in this
character and in most species averages less than either
Grammatorcynus or Acanthocybium. There is a slight
difference in number of pectoral fin rays between the
species of Grammatorcynus: G. bilineatu^ has a range
of 22-26, mode 25, x 24.4; G. bicarinatus 21-24,
mode 24, x 23.2 (Table 6).
First postcleithrum The posterior projecting plate
of the cleithrum has its posterior end attached to the
first postcleithrum which connects ventrally to the
second postcleithrum. The lamellar first postcleithrum
has a narrower upper end and a wider, rounded lower
margin (Fig. 23). The upper end is concave in Gram-
matorcynus (Fig. 23c-d) and pointed in both Scombero-
morus (Fig. 23a) and Acanthocybium (Fig. 23b). The
width of the postcleithrum varies from 46 to 62% of
the length of the bone in Grammatorcynus. It is nar-
rower in G. bicarinatus (width 46-52% of length, Fig.
23d) than in G. bilineatus (55-62%, Fig. 23c). In Acan-
thocybium (47-48%, Fig. 23b) the width is similar to
that of G. bicarinatus. Species of Scomberomorus (Fig.
23a) have narrower postcleithra (24-41%) than the
other two genera.
Second postcleithrum The second postcleithrum
is broad and lamellar at the upper part with a short
pointed ascending process and a long styliform
descending process. Grammatorcynus (Fig. 24d) dif-
fers strikingly from Acanthocybium (Fig. 24c) and
Scomberomorus (Fig. 24a-b) in having a distinct pro-
cess extending anteriorly from the broad lamellar por-
tion of the bone. The long descending process is so thin
in most specimens that an accurate measurement of
its length is nearly impossible because some portion of
it usually breaks off. No differences were detected in
this bone between the two species of Grammatorcynus.
Pelvic girdle The pelvic fin rays (1, 5) attach directly
to the paired basipterygia that make up the pelvic gir-
dle. The bones are united along the midline and are im-
bedded in the ventral abdominal wall, free from con-
tact with other bones. Each basipterygium is compos-
ed of three main parts: a wide anterodorsal plate, a
thin, flat anterior process, and a strong posterior
process.
To compare the pelvic girdles, the lengths of the
three parts were measured from their bases to their
tips. Grammatorcynus has the longest anterior process
(46-51% of the length of the anterodorsal plate. Fig.
25d), Acanthocybium has the next longest (35-47%,
Fig. 25c), and Scomberomorus the shortest (15-52%,
Fig. 25a-b). Grammatorcynus (29-33%, Fig. 25d) and
Acanthocybium (30-39%, Fig. 25c) have shorter pos-
terior processes than the species of Scomberomorus
Table 6
Number of pectoral fin rays in
Grammatorcynus
21
22 23
24
25
26
N
X
G.
G.
bilineatus
bicarinatus
1
1 11
2 2
19
6
27
4
62
11
24.4
23.2
a
Figure 23
Left first postcleithra in lateral view, (a) Scomberomarus
sinensis, Hong Kong, 677 mm FL, 1 x ; (b) Acanthocybium
solandri, Revillagigedos Is., 1068mm FL, 1 x ; (c) Gram-
matocynus bilineatus, Queensland, 521mm FL, 2x; (d)
G. bicarinatus, Queensland, 563mm FL.
(20-90%, Fig. 25a-b).
Grammatorcynus, some individuals of Acantho-
cybium, and several species of Scomberomorus have
longer anterior than posterior processes. The lengths
of the anterior process as a percentage of the posterior
process are: Grammatorcynus (154-158%), Acantho-
cybium (91-156%), and Scomberomorus species
(42-121%).
Collette and Gillis: Osteological differences between two species of Orammatorcynus
41
Figure 24
Left second postcleithra in lateral view, (a) Scomheromorus
queenslandieus, Great Barrier Reef, 641 mm FL, 1 x ; (b) S.
koreanus, Indonesia, 480mm FL, L5x; {c) Acanthocybium
solandri, Revillagigedos Is.. 1068mm FL, Ix; (d) Gram-
matorcynus bilineatus, Queensland, 382 mm FL, 2 x .
Grammatorcynus differs from most other
scombrids in having a single fleshy interpelvic
process. Auxis and Gymnosarda also have a
single interpelvic process; very large in the
former, moderate-sized in the latter.
Part 2: Systematics and biology
Grammatorcynus Gill 1862
Grammatorcynus Gill 1862: 125 (original
description; type-species Thynnus biline-
atus Riippell 1836 by original designation).
Nesogrammu^ Evermann and Seale 1907: 61
(original description; type-species Neso-
gramm.us piersoni Evermann and Seale
1907 by original designation, = Gram-
matorcynus bilineatus).
Figure 25 (below)
Right basipterygia of the pelvic girdle in mesial view, (a)
Scomheromorus regalis, Miami, FL, 469 mm FL, 1.5 x; (b)
S. lineolatus, Palk Strait, India, 428mm FL, 2x; {c)Acan-
ttiocybium solandri, Miami, FL, 1403 mm FL, 1 x ; (d) Gram-
maiorcynus bilineatus, Queensland, 521mm FL, 1.5 x.
anterodorsal plate
posterior process
anterior process
42
Fishery Bulletin 90(1). 1992
Diagnosis Grammatorcynus differs from all other
scombrid genera in having a second ventral lateral line,
and it differs from all other scombrids and billfishes
in lacking a triangular bony stay on the fourth pharyn-
geal toothplate (Johnson 1986). Like the Scombrini, it
has a low number of vertebrae (31) and the caudal fin
rays are supported by only the last three vertebrae.
Like the Scomberomorini, it has a well-developed
median keel on the caudal peduncle, but it lacks the
bony support for the keel that is present in bonitos and
higher tunas. Grammatorcynus differs from Scom-
beromorus in having a pineal window, a single inter-
pelvic process, and large scales.
Collette and Russo (1985b: 612) reported that Gram-
matorcynus bilineatus differed from Scomberomorus
and Acanthocybium in 16 osteological characters.
Grammatorcynus bicarinatus also differs in 15 of those
16 characters: (1) supracleithrum wide, 72-89% of
length (narrow, 42-62% in Scomberomorus and Acan-
thocybium); (2) pores absent along dorsal branch of
supratemporal (present); (3) nasal bones protrude far
beyond ethmoid region (do not protrude far beyond);
(4) posterior end of urohyal tripartite (forked); (5)
glossohyal with large tooth patch fused to dorsal sur-
face of bone (no fused tooth patch); (6) hyomandibula
narrow, 34-39% of length (wide, 39-52%); (7) angle
of lateral and medial arms of fourth epibranchial more
acute (less acute); (8) anterior process of second epi-
branchial elongate (shorter); (9) three vertebrae sup-
port caudal fin rays (four or five vertebrae); (10)
distinct anterior process on second postcleithrum (no
such process); (11) anterior end of first postcleithrum
notched (pointed); (12) first two pectoral radials attach
to scapula (upper one-third of third radial also attaches
to scapula); (13) jaw teeth conical (compressed and
triangular); (14) shaft of parasphenoid narrow and con-
cave or flat (wider and convex); and (16) posterior edge
of ectopterygoid long, 63-72% of ventral distance
(short, 41-63%). Unlike G. bilineatus, G. bicarinatus
resembles Scomberomorus and Acanthocybium in hav-
ing the upper margin of the dentary longer than the
lower margin (15).
Relationsiiips Larval characters of Grammatorcynus
bilineatus (as described by Wade 1951 from eight
specimens 8.5-17.5 mm FL) were used by Okiyama and
Ueyanagi (1977, 1978) and Ueyanagi and Okiyama
(1979) to construct an "index of primitiveness" that
divided the Scombrinae into four groups: mackerels,
Grammatorcynus, tunas, and Spanish mackerels and
bonitos. Nishikawa (1979) expanded the description of
larvae based on 62 specimens, 4.75-56.9 mm SL, from
Papua New Guinea. Nishikawa (1979) and Jenkins
(1989) noted that Grammatorcynus larvae have pre-
opercular spines characteristic of higher scombrids but
absent in Scomber and Rastrelliger.
Lewis (1981) examined Australian scombrids elec-
trophoretically and found that the two Gramma-
torcynus species showed fixed differences at 6 (23%)
of 26 loci (GPD, ADA, ADH, GDA, FKg, and PGMi).
Fixed differences were also observed at several other
loci not used in his study, namely AD2 and XO. He
analyzed the electrophoretic data phenetically and
cladistically. The two most parsimonious Wagner net-
works involved 308 steps. The species of Gramma-
torcynus were always paired and well-separated from
Figure 26
One of two equally parsimonious Wagner
networks for 23 Australian species of
Scombridae expressed in dendrogram form
(from Lewis 1981: fig. 6.4). (1) Scomber
autitralasinis, (2) Rastrelliger kanagurta,
(3) Grammatorcynus bicarinatus, (A) G.
bilineatus, (5) Scomberomonis commerson,
(6) S. queenslandicus, (7) S. midtiradiatus,
(8) S. semifasciatus, (9) S. munroi, (10)
Acanthocybium solandri, (11) Sarda aus-
trali.'i, (12) S. orientalis, (13) Cybiosarda
elegans, (14) Gymnosarda unicolor, (15)
Auxi^ sp., (16) Eulhynnus affinis. (17) Kat-
suivonus pelamis, (18) Thunnus albacares,
(19) T. tonggol, (20) T. obesus, (21) T. ala-
hmga, (22) T. maccoyii, (23) T. thynnus
orientalis.
Collette and Gillis: Osteological differences between two species of Crammatorcynus
43
other Scomberomorini (Fig. 26). They were most often
hnked with Acanthocybium and then with species of
Scomheromorus .
Based on both adult and larval morphological char-
acters and Lewis' electrophoretic data, Gramma-
torcynus is clearly more advanced than the mackerels
(Scombrini) and less advanced than the higher scom-
brids. Collette at al. (1984: fig. 312) placed it between
Gasterochisma and the Scombrini on the one hand, and
the more advanced Scomberomorini, Sardini, and
Thunnini on the other hand. Collette and Russo
(1985a, b) used Grammatorcynus as the primary out-
group in assessing relationships of the species of
Scomberomorus. In his reappraisal of scombroid rela-
tionships, Johnson (1986: fig. 1 and p. 38-39) placed
Grammatorcynus in its own tribe, Grammatorcynini,
above the Scombrini, as the sister group of higher
scombroids, which included the Sardini (including
the Thunnini), Scomberomorini, Acanthocybiini, and
billfishes.
Grammatorcynus bilineatus (Ruppell, 1836)
Double-lined or scad mackerel
Thynnus bilineatus Riippell 1836:39-40 (original
description. Red Sea), pi. 12, fig. 2. Giinther 1860:
366-367 (description). Klunzinger 1871:443 (Red
Sea). Meyer 1885:270 (Celebes).
Grammatorcynus bilineatiis. Gill 1862:125 (T. biline-
atus type species of new genus). Kishinouye 1923:
413-415 (description, anatomy; Ryukyu and Marshall
Is.), fig. 10 (skeleton); pi. 16, fig. 8 (transverse section
of vertebrae; pi. 34, fig. 62 (drawing). Hardenberg
1935:137-138 (description; W Java Sea). Okada
1938:170 (E. Indies, Red Sea; nijiyo saba). Morice
1953:36-40 (anatomy; after Kishinouye 1923).
Schultz 1960:411-412 (description; Bikini, Marshall
Is.). Kuronuma 1961:16 (listed, Vietnam). Lewis
1968:51 (Eniwetok, Marshall Is., infested with para-
sitic copepod Caligus asymmetricus). Collette 1983:
715-716 (distinguished from G. bicarinatus), fig. lA.
Collette and Nauen 1983:39-40 (description, range,
fig.). Collette et al. 1984:608 (fig. 326, larva after
Nishikawa 1979), 618 (larvae). McPherson 1984
(color pattern in Queensland, fig.). Masuda et al.
1984: 224-225 (description); color pi. 220A. Collette
and Russo 1985a:141-144 (outgroup for Scombero-
morus). Collette and Russo 1985b (anatomy, osteol-
ogy, figures, comparisons with Scomberomorus).
Allen and Russell 1986:101 (Scott Reef, NW Austra-
lia). Grant 1987:362-363 (scad mackerel; Queens-
land; color photo 769). Allen and Swainston 1988:
144 (Dampier Archipelago northwards, NW Austra-
lia), 145 (color painting 966). Bauchot et al. 1989:
657 (large brain, encephalization index of 226).
Zug et al. 1989:14 (Rotuma I.). Randall et al.
1990:443 (description, range), color plate VIII-14
(painting).
Nesogrammus piersoni Evermann and Seale 1907
(original description; Bulan, Sorsogon Province,
Luzon, Philippine Is.); pi. 1, fig. 3.
Grammatorcynus bicarinatus not of Quoy and Gai-
mard 1825. Herre 1931:33 (Balabac and Jolo, Philip-
pine Is.). Fraser-Brunner 1950:156 (synonymy), fig.
25. Umali 1950:9 (Zamboanga and Jolo, Philippine
Is.). Warfel 1950:18 (Philippine Is.), fig. 13 (draw-
ing of fish, gill arch, and liver). Wade 1951:456-458
(8 larvae, 8.5-17.5 mm; Philippine Is.), fig. 2 (8.5mm
specimen), fig. 3 (17.5 mm specimen), de Beaufort
1951:215-216 (description, synonymy), fig. 36.
Herre 1953:248 (synonymy). Dung and Royce 1953:
168-169, table 97 (morphometric data on 17 speci-
mens 408-580 mm FL, western Marshall Is.). Matsu-
bara 1955:519 (2 lateral lines; range), fig. 222B.
Munro 1958b:262-263 (New Guinea region records;
CSIRO C492, New Hanover, examined). Jones et
al. 1960:136 (Ross I., Port Blair, Andaman Is.).
Collette and Gibbs 1963a: 25 (monotypic genus).
Collette and Gibbs 1963b:27 (description), pi. 7.
Jones and Silas 1963:1781 (synonymy, Indian Ocean
references, range). Silas 1963:811-833 (description,
synonymy, synopsis of biological data). Kamohara
1964:34 (Miyako-jima, Ryukyu Is.). Jones and
Kumaran 1964:364-365, figs. 70-71 (larvae, after
Wade 1951). Jones and Silas 1964a:16, 18 (descrip-
tion, synonymy, range), pi. 4, fig. (449mm female
from Port Blair, Andaman Is.). Jones and Silas
1964b:258 (in key; Andaman Is.). Gorbunova 1965:
55 (references to Wade 1951 and Silas 1963). Tong-
yai 1966:6 (in key), 17 (pi. 1, outline fig. of specimen
from Phuket I.). Kamohara 1967:43 (description).
Munro 1967:197-198 (description of G. bicarinatus;
New Guinea specimens are G. bilineatus). Ben-
Tuvia 1968:35 (Entedibir Is., Red Sea), fig. 3g.
Ben-Yami 1968:40 (schools probably occur in region
of Sahlak Archipelago, southern Red Sea). Jones
1968:998 (occur in catches in Andaman area). Jones
1969:26 (Laccadive Archipelago, India). Tongyai
1970:558 (Thai common names; Indian Ocean coast
of Thailand). Tongyai 1971:3-5 (description. Thai
common names. Thai distribution). Shiino 1972:70
(common names). Richards and Klawe 1972:72
(references to larvae). Gushiken 1973:49 (color
photograph of 60 cm specimen from Okinawa). Helf-
man and Randall 1973:151 (Palau; common names
mokorokor and biturturch). Magnuson 1973:350
(correlation of size, pectoral fin length, and presence
of swim bladder). Lewis et al. 1974:83,87 (Bismarck
Archipelago, Papua New Guinea). Springer et al.
1974:40 (Indonesia). Romimohtairo et al. 1974:35
44
Fishery Bulletin 90(l|. 1992
(Gamber Bay, Gag I., Indonesia). Gorbunova
1974:26 (fig. 2, after Wade 1951). Orsi 1974:174
(listed, Vietnam). Masuda et al. 1975:256 (color
photograph F), 79 (description; Okinawa southward).
Cressey 1975: 216 (parasitic copepod Shiinoa occlusa
from nasal cavity of a specimen from N. Celebes).
Kailola 1975:235 (5 collections from Papua New
Guinea). Uyeno and Fujii 1975:14 (table 1, com-
parison of caudal complex with other scombrids).
Fourmanoir and Laboute 1976:183 (description; New
Caledonia), color photograph. Shiino 1976:229 (com-
mon names). Anonymous 1977:15 (table 4, Baga-
man I., Louisiade Archipelago, Papua New Guinea).
Klawe 1977:2 (table 1, range). Collette 1979:29
(characters, range). Ceng and Yang 1979:472-473
(description; Sisha Is., South China Sea), fig. 335.
Yamakawa 1979:43 (Miyako-jima, Ryukyu Is., after
Kamohara 1964). Joseph et al. 1979:38 (range, fig-
ure). Nishikawa 1979:125-140 (early development;
62 postlarval and juvenile specimens, mostly from
Papua New Guinea). Shirai 1980:64 (description,
Ryukyu Is.), color photograph. Cressey and Cressey
1980:46 (parasitic copepod fauna: Shiinoa occlusa
and Caligus asymmetricus). Rau and Rau 1980:
512-513 (description, Philippine Is.). Jones and
Kumaran 1981:581-582 (description; Laccadive
Archipelago), fig. 494. Wang 1981:161 (listed; S.
China Sea). Johannes 1981:156-157 (biology,
Palau). Lewis 1981:13 (species B, scad; maximum
size 60cm FL, 3 kg), photograph. Kyushin et al.
1982:249 (description, common name nijo-saba), color
photograph (specimen from Milne Bay, New Guinea).
Cressey et al. 1983:238 (systematic position of
genus), 264 (parasitic copepod fauna; 4 species of
Caligus added). Lewis et al. 1983:7 (table 2, 203
specimens, 380-630 mm FL; Fiji). Wass 1984:31
(Fiji; common name "namuauli"). Masuda et al.
1984:224-225 (description, Japan), pi. 220A. Gillett
1987:20 (caught by Satawal tuna fishermen, central
Caroline Is.). Nishikawa and Rimmer 1987:5 (larval
description; fig. 5, larva, postlarva, and juvenile from
Nishikawa 1979). Dyer et al. 1989:65 (monogenean
Caballerocotyla sp. from Okinawa specimen). Riva-
ton et al. 1989:67 (listed. New Caledonia).
Grammatorcynnus (sic) bicarinatus not of Quoy and
Gaimard 1825. Roux-Esteve and Fourmanoir
1955:201 (Abulat I., Red Sea).
Grommatorcynus (sic) bicarinatus not of Quoy and
Gaimard 1825. Zhang 1981:302 (description of 3
larvae, Sisha Is., South China Sea; fig. 1, 6.4mm
larva).
Grammatorcynos (sic) bilineatus. Myers 1988:168
(listed, Mariana Is.). Myers 1989:254 (description;
range), underwater photo 134A, 280 (listed; Caroline,
Mariana, and Marshall Is.).
Diagnosis Grammatorcynu^ bilineatus has more gill
rakers (18-24 vs. 12-15), a larger eye (4.1-6.0% vs.
3.1-4.6% FL), lacks black spots on the lower sides of
its body, and does not reach as large a size (max.
600mm FL) as G. bicarinatus.
Description Dorsal spines 11-13, usually 12; rays
10-12, usually 11; finlets 6, rarely 7 (Table 4). Anal fin
with one spine, 11-13, usually 12 rays; 6 finlets, rare-
ly 5 or 7 (Table 5). Pectoral fin rays 22-26, usually 24
or 25 (Table 6). Gill rakers on first arch 18-24, usually
21, X 20.8 (Table 3). Upper jaw teeth 14-37, x 23.5
(left), 24.5 (right) (Table 2); lower jaw teeth 12-32,
X 18.6 (left), 19.1 (right) (Table 2). Morphometric data
summarized in Table 1.
Grammatorcynus bilineatus has a longer neuro-
cranium (14-16% FL vs. 13%), longer parasphenoid
flanges (18-21% of neurocranium length vs. 14%),
higher maximum number of teeth on the upper (37 vs.
25) and lower (32 vs. 23) jaws, higher posterior expan-
sion of the maxilla (8-11% of maxilla length vs. 6-8%),
longer quadrate process (134-145% of quadrate length
vs. 122-125%), wider first postcleithrum (55-62% of
length vs. 46-52%), narrower ethmoid (19-21% of
length vs. 25-28%), narrower vomer (13-15% of length
vs. 16-18%), narrower lachrymal (27-30% of length vs.
30-35%), shorter teeth (up to 4% of dentary length vs.
up to 6%), narrower palatine tooth patch (26-32% of
length vs. 38-42%), narrower opercle (63-72% of
length vs. 72-79%), and the shelf between the post-
temporal processes is hidden behind the flat posterior
portion of the bone.
Color In life, the back is bright pale green, the upper
sides and belly silvery, and there are no black spots on
the belly as there are in G. bicarinatus (Grant 1987:
363). Underwater, it is reported to display a distinc-
tive white patch on the caudal peduncle (McPherson
1984). There are color photographs of fresh specimens
from Japan (Masuda et al. 1975:256, Shirai 1980:64),
New Caledonia (Fourmanoir and Laboute 1976:183),
South China Sea (Kyushin et al. 1982:249), and
Australia (Grant 1987: fig. 769). An underwater photo-
graph has been published from Micronesia (Myers 1989:
photo 134A). There is a color painting in Randall et al.
1990 (pi. VIII-14).
Size Maximum size is about 63cm FL, 3.3kg weight
(Lewis et al. 1983). Maturity seems to be attained at
about 40-43 cm FL (Silas 1963, Johannes 1981, Lewis
et al. 1983).
Biology The best summary of biological information
on G. bilineatus is Silas (1963). It is an epipelagic
species found mostly in shallow reef waters where it
Collette and Gillis: Osteological differences between two species of Grammatorcynus
45
120
150
180
• G. bilineatus
• G. bicarinatus
t
it}k
-^
^'
h
J0_
60
90
120
150
180
60
Figure 27
Distribution of Grammatorcynus based on specimens examined and literature records.
forms large schools. The spawning season in Fiji ex-
tends from October through March (Lewis et al. 1983).
Larvae and juveniles have been illustrated from the
Philippines (Wade 1951), Papua New Guinea (Nishi-
kawa 1979), and the South China Sea (Zhang 1981).
Food includes adult and juveniles of crustaceans and
fishes, particularly clupeoids such as Sardinella and
Thrissocles, but also includes other fishes such as
Sphyraena and Balistes (Silas 1963:831).
Parasites Six species of parasitic copepods have been
reported from G. bilineatus (Cressey and Cressey 1980,
Cressey et al. 1983): Shiinoidae: Shiinoa ocdusa Kaba-
ta; Caligidae: Caligus asymmetricus Kabata, C. rega-
lis Leigh-Sharpe, C. bonito Wilson, C. pelamydis
Kr^yer, and C. productus Dana. The monogenean
Caballerocotyla sp. was found on an Okinawan speci-
men (Dyer et al. 1989).
Interest to fisfieries Double-lined mackerel are taken
incidentally with hand lines off Port Blair, Andaman
Islands (Silas 1963). It is common in the offshore zones
of Fiji but is only occasionally seen in Fiji markets
(Lewis et al. 1983). The flesh is reported to be mild and
pleasantly flavored but it is necessary to remove the
kidney tissue before cooking to avoid the ammonia
smell. This characteristic has given rise to one of the
Palauan names for the species, biturchturch, which
means urine (Johannes 1981:187). It is valued for
marlin bait in Queensland (McPherson 1984).
i?ange Widespread near coral reefs in the tropical
and subtropical Indo-West Pacific (Fig. 27). Based on
literature, specimens examined, and photographs,
known from the Red Sea, Andaman Sea, East Indies,
Philippines, South China Sea, Ryukyu Islands, New
Guinea (New Britain, New Ireland, New Hanover, and
the Louisiade Archipelago), Australia (northern West-
ern Australia, from Dampier Archipelago north and
Queensland), Solomon Islands, New Caledonia, Caro-
line Islands, Marshall Islands, Fiji, Tonga, and Tokelau
Islands (photograph from Fakaofo Atoll received from
Robert Gillett, Regional Fishery Support Programme,
Suva, Fiji, Aug. 1985).
46
Fishery Bulletin 90(1), 1992
Table 7
Morphometric comparison
of Grammatorcynus bilineatus from the Red Sea and the western Pacific Ocean.
Character
Red Sea
Western Pacific Ocean
N
Min
Max
Mean
SD
N
Min
Max
Mean
SD
Fork len^h (thousandths)
15
264
432
364
53
44
226
575
430
72
Snout-A
15
592
628
615
11
41
581
641
603
12
Snout-2D
15
534
557
547
7
41
528
619
548
16
Snout-ID
15
287
306
296
7
44
276
322
295
10
Snout-P2
15
249
271
261
7
43
236
306
257
12
Snout-Pl
15
220
244
230
7
43
199
245
224
9
P1-P2
15
94
135
105
9
43
90
111
100
5
Head length
15
213
234
220
6
44
197
236
218
8
Max. body depth
13
182
213
201
8
40
164
234
193
15
Max. body width
12
97
122
111
8
40
91
136
115
9
PI length
14
110
142
122
9
44
106
142
128
7
P2 length
14
74
81
77
2
44
70
87
76
4
P2 insertion-vent
14
312
352
333
12
43
262
354
326
15
P2 tip-vent
13
242
275
256
11
43
228
273
249
9
Base ID
15
207
246
230
10
43
211
261
236
10
Height 2D
12
88
109
98
6
37
88
116
99
7
Base 2D
14
88
114
102
8
43
79
118
102
8
Height A
11
84
102
93
7
34
82
114
96
7
Base A
14
74
89
84
4
44
73
101
88
7
Snout (fleshy)
15
77
86
81
3
44
58
90
79
5
Snout (bony)
15
71
78
74
3
44
60
80
71
5
Maxilla length
15
95
108
101
4
43
89
107
97
5
Postorbital
15
89
98
93
2
43
78
98
91
4
Orbit (fleshy)
15
44
57
49
4
44
40
60
49
4
Orbit (bony)
15
60
75
68
5
44
53
88
68
7
Interorbital
15
58
67
62
3
43
56
74
62
3
2D-caudal
15
458
475
468
5
41
427
496
471
16
Head len^h (thousandths)
15
62
94
80
11
44
50
126
94
16
Snout (fleshy)
15
340
384
369
11
44
248
397
365
22
Snout (bony)
15
313
351
336
11
44
281
357
326
18
Maxilla length
15
443
469
459
8
43
420
480
443
14
Postorbital
15
397
433
420
12
43
350
450
419
16
Orbit (fleshy)
15
206
253
222
13
44
191
257
226
15
Orbit (bony)
15
283
336
307
15
44
252
381
313
27
Interorbit
15
268
298
283
9
43
253
327
284
14
Geographic variation The wide range of G. biline-
atus plus the gaps in distribution due to its preference
for coral reef habitats lead to the possibility that some
populations differ morphologically from others. How-
ever, comparison of frequency distributions by geo-
graphic areas of meristic characters summarized in
Tables 2-6 showed general uniformity in the range and
modes of these characters. The Red Sea population is
the most isolated from the rest, but it showed no
meristic differentiation. Comparisons of ranges and
means of morphometric data showed few differences
between the Red Sea and Pacific populations (Table 7).
Dissections 11 (382-521 mm FL). USNM 270386
(410), Australia, diss. 1-28-69. USNM 270390 (453), Scott
Reef, J. McCosker 73-8, diss. 4-1-76. USNM 270387 (424),
Marshall Is.. J. E. Randall, diss. 4-29-76. USNM 270384
(382), Kavieng, New Guinea, diss. 10-12-76. USNM 270385
(389), Cairns, Qld., G. McPherson, diss. 3-30-81. USNM
270389 (521). Port Douglas, Qld.. diss. 3-31-81. USNM
270388 (416). Cairns, Qld., G. McPherson, diss. 1-5-83.
USNM 270383 (475). Scott Reef, J. McCosker 72-18. diss.
1-10-83. USNM 270382 (399). Cairns. Qld.. G. McPherson.
diss. 1-1 1-83. USNM 270391 (400). Port Douglas. Qld.. diss.
1-13-83. USNM 316130 (460), Scott Reef, J. McCosker 73-8,
diss. 7-18-89.
Material examined 80 specimens (23.5-575 mm FL)
from 58 collections.
Red Sea 16 (264-440) from 10 collections. SMF 2755
(1, 287); Massua; E. Riippell; holotype of Thynmis bilirwatus;
stuffed. NMW uncat. (5. 362-432); Jambo; 1895-96; I.R.M.
Exped. 62c. NMW uncat. (1, 424); Hassani; 1895-96; I.R.M.
Collette and Gillis: Osteological differences between two species of Grammatorcynus
47
Exped. 62b. NMW uncat. (1, 360); Djeddah; 1895-96; I.R.M.
Exped. 62. NMW uncat. (2, 264-320); Rothes Meer;
1879-80; Klunzinger. NMW 16825 (2, 304-382); Rothes
Meer; Klunzinger. MNHN 52-28 (1, 422); Mer Rouge;
"Calypso". USNM 266928 (1, 327); near Entedebir; March
1962; ISRSE 4144. HUJ E62/4399 (1, 368); S Red Sea;
March-April 1962; Israel; S Red Sea Exped. BPBM 28388
(1, 440); Saudi Arabia, Jeddah market; 11 May 1982; J.E.
Randall.
Andaman Sea 4 (235-294) from 2 collections. MFLB
uncat. (2, 282-294); Thailand, Phuket Province; 23 Feb. 1966.
MFLB uncat. (2, 235-237); Thailand, Phuket Province; 27
Jan. 1970.
East Indies 4 (108-413) from 4 collections. BMNH
1872. 4.6.25(1, 413); N. Celebes; Meyer. AMNH 17583 (1,
108) Celebes. USNM 213564 (1, 395); Indonesia, Ambon fish
market; V.G. Springer; 19 March 1974. USNM 213565 (1,
360); Indonesia, Buton I., Teluk Buton; V.G. Springer and
M.F. Gomon; VGS 74-26; 28 March 1974.
Philippine Islands 5 (275-390) from 5 collections.
USNM 55899 (1, 372); Luzon, Sorsogon Province, Bulan; C.J.
Pierson; holotype oi Nesogrammics piersoni. USNM 195044
(1, 343); Cebu market; 6 April 1908; Albatross. CAS SU
13575 (1, 275); Balabac I.; A.W.C.T. Herre; 1929. CAS SU
13687 (1, 342); Jolo; A.W.C.T. Herre; 1931. CAS SU 40469
(1, 390); Gulf of Leyte, Leyte; R.F. Annereaux; 12 Sept. 1945.
Okinawa ZUMT 16738(1, 378); Okinawa. ZUMT 52381
(1, ca. 500); Okinawa, Ishigaki I.; 4 June 1966.
New Guinea 10 (23.5-410) from 6 collections. CSIRO
C.492 (1, 226); New Hanover, Drei Inseln Harbor, Kulineva
R. USNM 270384, 316162 (2, 363-382); New Ireland,
Kavieng; 20 March 1976. BASF 4247 (1, 23.5); New Brit-
ain; Borgen Bay; 13 April 1972. BASF 4248 (1, 37.9); New
Britain; Tavanatangir; 11 Oct. 1972. BASF 4250 (2,
54.3-65.5); New Britain; Bikarua I.; Cape Lambert; 28 Nov.
1972. USNM 320095 (3, 385-410); Hermit Is., E side Jalun
I.; 2 Nov. 1978.
Australia 9 (389-521) from 4 collections. USNM 270383
(1, 475); Western Australia, Scott Reef; J.E. McCosker 72-18.
USNM 270390, 316130 (2, 453-460); Western Australia,
Scott Reef, 14°05'S, 121°50'E; J.E. McCosker 73-8. USNM
270382, 270385, 270388, 316161 (4, 389-416); Queensland,
Cairns; G. McPherson and P. Cooper. USNM 270389,
270391 (2, 400-521); Queensland, off Port Bouglas; Sept.-
Oct. 1976.
Solomon Islands 2 (275-482) from 2 collections. USNM
205078 (1, 482); New Georgia, Gizo I.; W. Chapman; 30 May
1944. AMS 1.19435-020 (1, 275); Solomon Is.; G. Smith.
Caroline Islands 15 (327-575) from 13 collections. CAS
GVF 651 (1, 462); Palau Is., Rattakadakoru; Palau 145; 5
Sept. 1955. CAS GVF 933 (1, 439); Palau Is., Velasco Reef;
Palau 147; 6 Oct. 1956. CAS GVF 934 (1, 432); Palau Is.,
Velasco Reef; Palau 148; 6 Oct. 1956. CAS GVF 946 (1, 454);
Palau Is., Velasco Reef; Palau 149; 6 Oct. 1956. CAS GVF
1422 (1, 543); Palau Is., Ilruthapel I.; Palau 57-42; 20 Oct.
1957. CAS GVF 1867 (1, 461); Palau Is.; Palau 59-39; 15
April 1959. CAS GVF 1891 (1, 444); Palau Is., Angaur I.;
Palau sta. 59-63; 16 June 1959. CAS GVF 1970 (2, 543-575);
Palau Is., Kossol Passage, 7°56'18"N, 134°31'55"E; sta.
59-709; 30 July 1959. BPBM 10501 (2, 434-448); Palau; 23
April 1964. USNM 264910 (1, 510); between Ponape and
Ant Atoll; R.A. Croft; 1983. CAS GVF (1, 408); Kapinga-
marangi, 1°6'N, 154°44'W; sta. 108; 4 Aug. 1958. CAS
GVF 405 (1, 327); Kapingamarangi; sta. 102; 2 Aug. 1954.
CAS GVF 33 (1, 462); Ifaluk; 2 Oct. 1953.
Marshall Islands 9 (254-549) from 7 collections. USNM
140986 (2, 419-468); Bikini Atoll lagoon, V. Brock and J.
Marr; 2 April 1946. USNM 142054 (2, 503-549); BOdni Atoll
lagoon off Bikini I.; V. Brock and J. Marr; 25 March 1946.
USNM 142055 (1, 410) Bikini Atoll, W of Boro I., V. Brock;
6 April 1946. USNM 181932 (1, 382); Majuro Atoll; A.F.
Bartsch; 1958. BPBM uncat. (1, 254); Majuro Atoll; P.
Shiota; 30 Aug. 1972. USNM 270387 (1, 424); Enewetak
Atoll; J.E. Randall; 2 April 1976. BPBM 12800 (1, 330);
Enewetak; 4 April 1972; J.E. Randall.
Fiji Islands 2 (335-426) from 2 collections. USNM
176657 (1, 426); S of Suva; J.K. Howard; 4-15 Bee. 1952.
USNM 243969 (1, 335); reef NNE Malamala I., V.G. Springer
82-24; 24 May 1982.
Rotuma Island USNM 285517 (396); Rotuma; G.B.
Johnson; 14 May 1986.
Samoan Islands USNM 305080 (1, 465); American
Samoa, East Bank, 12 mi. off E end of Tutuila; 1 July 1989.
Grammatorcynus bicarlnatus
(Quoy and Gaimard, 1825)
Shark mackerel
Thynnus bicarinatus Quoy and Gaimard 1825:357
(original description; Baie des Chiens-Marins
( = Shark Bay), W. Australia), pi. 61, fig. 1.
Grammatorcynus bicarinatus. McCulloch 1915:266-
269 (description; off Cook I., near Tweed River
Heads, New South Wales; 925 mm FL, 18.75 lbs.),
pi. 1, fig. 1. Ogilby 1918:101 (reference to McCul-
loch 1915; caught off Moreton Bay, Queensland), 105
(30-lb. specimen in Queensland state fish market).
McCulloch 1922:106 (New South Wales; rarely cap-
tured; to 3 ft.). McCulloch and Whitley 1925:142
(Moreton Bay, Queensland). McCulloch 1929:263-
264 (synonymy). Anonymous 1945:7 (listed among
marketable fish of Cairns, Queensland area). Whit-
ley 1947:129 (W. Australia). Whitley 1948:24
(listed, W. Australia). Coates 1950:22 (Great Bar-
rier Reef; 25 lbs. maximum; "shark mackerel"),
fig. Serventy 1950:20 (common in W. Australia
from Geraldton northwards but not extending in
waters of the Kimberly Division of W. Australia).
Munro 1958a: 112 (description; Queensland, N New
South Wales, and W. Australia), fig. 748 (after
McCulloch). Whitley 1964a:232 (length to 48 in.,
weight 25 lb.), 239 (fig. 4f, range in Australia only),
pi. 4 (fig. b, after McCulloch). Whitley 1964b:48
(listed). Marshall 1964:367 (description, Queensland),
color plate 53, fig. 354. Grant 1965:176 (description
after Munro 1958a; sought-after market fish in
48
Fishery Bulletin 90(1). 1992
Queensland), fig. Marshall 1966: 205 (description),
color plate 53, fig. 354. Munro 1967:197-198 (text
is based on Australian G. bicarinatus; New Guinea
specimens are G. bilineatus), fig. 333. Grant 1972:
107 (same as 1965), fig. Coleman 1974:42 (color,
habits), 43 (underwater color photograph. Heron I.,
Queensland). Grant 1975:165 (same as 1972), fig.
Rohde 1976:50 (Lizard I., Queensland). Anonymous
1978:18 (listed among species being investigated by
Queensland Fisheries Service). Grant 1978:195
(same as 1975), fig. Hutchins 1979:83 (may visit
Rottnest I., W. Australia). Coleman 1981:268 (Aus-
tralia, habits), color underwater photo (from Coleman
1974). Lewis 1981:12 (species A, shark mackerel;
maximum size 110cm FL, 13.5kg) photograph.
Grant 1982:632 (same as 1978 plus comments on am-
monia smell of flesh). Collette 1983:716-718 (distin-
guished from G. bilineatus), fig. IB. Collette and
Nauen 1983:39-40 (description, range, fig.). Hut-
chins and Thompson 1983:62, 85 (W. Australia), p.
63 (fig. 290). Russell 1983:146 (Heron L, Barrier
Reef based on Coleman 1974). McPherson 1984 (color
pattern in Queensland, fig.). Collette and Russo
1985b:547 (in key). Hutchins and Swainston 1986:
102 (description, range), 103 (color painting 587), 141
(weight to 11.7kg). Grant 1987: 362 (shark mack-
erel; Queensland; color photo 768). Allen and Swains-
ton 1988:144 (Geographe Bay north. Western Aus-
tralia), 145 (color painting 965). Hutchins 1990:275
(sight record. Shark Bay, Western Australia). Ran-
dall et al. 1990:433 (description, range), color plate
Vn-13 (painting).
Diagnosis Grammatorcynus bicarinatus has fewer
gill rakers (12-15 vs. 18-24), a smaller eye (3.1-4.6%
vs. 4.1-6.0% FL), small black spots on the lower sides
of its body, and reaches a larger maximum size (1100
mm FL) than G. bilineatus.
Description Dorsal spines 12; rays 10, finlets usual-
ly 7, rarely 6 or 8 (Table 4). Anal spines 1, rays 11-13,
finlets 6 or 7, usually 7 (Table 5). Pectoral fin rays
21-24, X 23.2 (Table 6). Gill rakers on first arch 12-
15, X 14.1 (Table 3). Upper jaw teeth 14-25, x 20.5
(left), 20.9 (right) (Table 2); lower jaw 15-23, x 17.5
(left), 17.6 (right) (Table 2). Morphometric data sum-
marized in Table 1.
Grammatorcynus bicarinatus has a shorter neuro-
cranium (13% vs. 14-16% FL), shorter parasphenoid
flanges (14% vs. 18-21% neurocranium length), a lower
maximum number of teeth on the upper (25 vs. 37) and
lower (23 vs. 32) jaws, lower posterior edge of shank
of maxilla (6-8% vs. 8-11% maxilla length), shorter
quadrate process (122-125% vs. 134-145% quadrate
length), narrower first postcleithrum (46-52% vs.
55-62% length), wider ethmoid (25-28% vs. 19-21%
length), wider vomer (16-18% vs. 13-15% neurocra-
nium length), wider lachrymal (30-35% vs. 27-30%
length), longer teeth (maximum 6% vs. 4% dentary
length), wider palatine tooth patch (38-42% vs. 26-32%
length), wider opercle (72-79% vs. 63-72% length),
and a thin posttemporal shelf between the anterior
processes.
Color General color in life is bright, glowing green
above, grading into the silver of the sides and belly,
which is marked with scattered small black spots that
are absent in G. bilineatus (Grant 1987:362). Under-
water, it is reported to display a dark band along the
lower lateral line (McPherson 1984). Color photographs
have been published by Marshall (1964 and 1965: pi.
53, fig. 354) and Grant (1987: photo 768), and there are
color paintings in Hutchins and Swainston (1986:103),
Allen and Swainston (1988:965), and Randall et al.
(1990: plate Vni-13). An underwater photograph was
published by Coleman (1974:43 and 1981:268).
Size Maximum size is 110-130cm FL and 11.6-13.5
kg weight (Lewis 1981, Hutchins and Swainston 1986,
Allen and Swainston 1988).
Biology Shark mackerel form dense concentrations
near individual bays and reefs in Barrier Reef waters.
With the rising tide, they move into shallow water over
the reef flats, feeding on schools of clupeoid fishes
(Grant 1982).
Interest to fisheries Shark mackerel are fished off
Western Australia, the Northern Territory, Queens-
land, and northern New South Wales (Grant 1987). It
is regarded as a light-tackle sportsfish with commer-
cial value in Queensland (McPherson 1984). The name
shark mackerel comes from the ammonia-like smell
noticed when the fish is being filleted. This odor can
be masked by brushing the fillets with lemon juice prior
to cooking (Grant 1982, 1987).
Range Found over coastal reefs of all Australian
warm waters (Grant 1987) with occasional stragglers
south to 30° on both east (Cook I., New South Wales)
and west (Shark Bay, Western Australia) coasts (Fig.
27) and in the Gulf of Papua (A.D. Lewis, South Pacific
Comm., Noumea, pers. commun.). The apparent gap
in distribution may be due to ecological reasons, the
scarcity of reef habitats along the north coast of
Australia, or to historical reasons, as outlined by
Springer and Williams (1990).
Dissections 4 specimens (563-765 mm FL). USNM
270392 (563), Cairns. Qld., diss. 1-3-83. USNM 316126 (765),
Collette and Gillis: Osteological differences between Iaa/o species of Crammatorcynus
49
Exmouth Gulf, WA, B. Hutchins, diss. 1-18-83. USNM
316127-8 (2, 625-663), Australia, diss. 7-5-89.
Material examined 11 specimens (300-765 mm FL)
from 8 collections. USNM 316129 (1, 563), Queensland, Fitz-
roy I. S of Cairns; Jan. 1984. USNM 176832 (1, 525), Great
Barrier Reef; J.K. Howard; 8 April-29 May 1952. AMS
IB.5207-8 (2, 306-315), Queensland, Gladstone District; P.
Gibson. USNM 316126, uncat. (2, 607-765), Western
Australia, Exmouth Gulf. WAM-P 27343 (1, 825), Western
Australia, N. Muiron I., 21°39'S, 114°22'E. WAM-P 25821
(1, 320), Western Australia, S. Muiron I., 2r39'S, 114°20'E.
WAM-P 22974 (head only, 105 mm), Western Australia, Ken-
drew I., 20°29'S, 116°22'E; 21 Feb. 1973; B. Hutchins.
USNM 316127-8 (2, 625-663), Australia.
Acknowledgments
The original impetus for this study came from A.D.
Lewis, now with the South Pacific Commission, Nou-
mea, New Caledonia. In the course of his doctoral
dissertation on population genetics of Australian scom-
broids (Lewis 1981), he discovered that there were two
species in the genus Grammatorcynus, informed us of
the problem, and provided us with frozen material
needed to do a thorough anatomical study. We are
deeply appreciative of his interest and efforts through-
out the course of this study. Sally Rothwell spent the
month of February 1983 working in the Systematics
Laboratory on a Careers in Biology Program from Col-
gate University. Her efforts made completion of the
1983 preliminary paper possible. Frozen material, vital
for this study, was provided through the efforts of
A.D. Lewis, J.E. McCosker (CAS), B. Hutchins
(WAM), and G. McPherson (Northern Fisheries Re-
search Center, Queensland). Several assistants, stu-
dents, and colleagues participated in dissections over
the years: Linda Pushee Mercer, Frances Matthews
Van Dolah, Sally Rothwell, and Joseph Russo. Pre-
served material was made available by A. Ben-Tuvia
(HUJ), B.E. Bookheim (Department of Marine and
Wildlife Resources, Pago Pago, American Samoa),
W.E. Eschmeyer and P. Sonoda (CAS), R.A. Croft
(Ponape State Government), B. Herzig (NMW),
W. Klausewitz (SMF), A.D. Lewis (then in Papua New
Guinea), G.J. Nelson (AMNH), J.R. Paxton (AMS),
J.E. Randall and A.Y. Suzumoto (BPBM), J.B. Hut-
chins (WAM), and P.J.P. Whitehead (then at BMNH).
Assistance with analyzing the morphometric data was
provided by Ruth Gibbons. Osteological drawings were
painstakingly rendered by Keiko Hiratsuka Moore. The
map was produced by Martha S. Nizinski. Participa-
tion in this study by the second author was made pos-
sible by a Smithsonian Summer Internship in Verte-
brate Zoology in the summer of 1989. Drafts of the
manuscript were reviewed by G. David Johnson,
Thomas A. Munroe, Yosuo Nishikawa, and William
Smith-Vaniz.
Citations
Allen, G.R., and B.C. Russell
1986 Part VII. Fishes. In Faunal surveys of the Rowley
Shoals, Scott Reef and Seringapatam Reef, North-western
Australia. Rec. W. Aust. Mus. Suppl. 25:75-103.
Allen, G.R., and R. Swainston
1988 The marine fishes of north-western Australia. West.
Aust. Mus., Perth, 201 p.
Allis, E.P. Jr.
1903 The skull, and the cranial and first spinal muscles and
nerves in Scomber scomber. J. Morphol. 18(l-2):45-328.
Anonymous
1945 Marketable fish of the Cairns Area. N. Queens). Nat.
Club 2, 8 p.
1977 Report on Japan-Papua New Guinea joint skipjack tag-
ging in the PNG area, October-December, 1975 on the R/V
Fuji Mara. Res. Devel. Div., Fish. Agency Jpn., and Fish.
Div., Dep. Primary Ind., Papua New Guinea, 30 p.
1978 Major Queensland bid to track Spanish mackerel. Aust.
Fish. 37(3):18-19.
Bauchot, R., J.E. Randall, J.-M. Ridet, and M.-L. Bauchot
1989 Encephalization in tropical teleost fishes and comparison
with their mode of life. J. Hirnforsch. 30(6):645-669.
Ben-Tuvia. A.
1968 Report on the fisheries investigations of the Israel South
Red Sea Expedition, 1962. Bull. Sea Fish. Res. Stn. Haifa
52:21-25.
Ben-Yami, M.
1968 Observations on the distribution and behaviour of pelagic
schooling fish in the southern Red Sea. Bull. Sea Fish. Res.
Stn. Haifa 51:31-46.
Ceng, B-g., and Y. Yang
1979 Suborders Scombroidei. Thunnioidei. In Fishes of the
islands in the South China Sea, p. 470-488. Science Press,
Beijing.
Coates, G.
1950 Fishing on the Barrier Reef and inshore. T. Willmett
& Sons, Townsville, 72 p.
Coleman, N.
1974 Australian marine fishes in colour. A.H. & A.W. Reed,
Sydney, 108 p.
1981 Australian sea fishes north of 30''S. Doubleday, Sydney
and Auckland, 297 p.
Collette, B.B.
1979 Adaptations and systematics of the mackerels and tunas.
In Sharp, G.D., and A.E. Dizon (eds.), The physiological ecology
of tunas, p. 7-39. Academic Press, NY.
1983 Recognition of two species of double-lined mackerels
{Grammatorcynus: Scombridae). Proc. Biol. Soc. Wash. 96(4):
715-718.
Collette, B.B., and L.N. Chao
1975 Systematics and morphology of the bonitos (Sarda) and
their relatives (Scombridae, Sardini). Fish. Bull., U.S. 73(3):
516-625.
Collette, B.B., and R.H. Gibbs Jr.
1963a A preliminary review of the fishes of the family Scom-
bridae. FAO Fish. Rep. 6(l):23-32.
50
Fishery Bulletin 90(1). 1992
1963b Preliminary field guide to the mackerel- and tuna-like
fishes of the Indian Ocean (Scombridae). Smithson. Inst.,
Wash. DC, 48 p.
CoUette, B.B., and C.E. Nauen
1983 Scombrids of the world. FAG Species Catalogue. FAG
Fish. Synop. 125, Vol. 2, 137 p.
Collette, B.B., and J.L. Russo
1985a Interrelationships of the Spanish mackerels (Pisces:
Scombridae: Scomberomorus) and their copepod parasites.
Cladistics 1(2): 141-158.
1985b Morphology, systematics, and biology of the Spanish
mackerels {Scomberomorus, Scombridae). Fish. Bull., U.S.
82(4):545-692.
Collette. B.B., T. Potthoff, W.J. Richards, S. Ueyanagi,
J.L. Russo, and Y. Nishikawa
1984 Scombroidei: Development and relationships. In Moser,
H.G., et al. (eds.). Ontogeny and systematics of fishes, p.
591-620. Spec. Publ. 1, Am. Soc. Ichthyol. Herpetol. Allen
Press, Lawrence, KS.
Conrad, G.M.
1938 The osteology and relationships of the wahoo (Acan-
thocybium solandri), a scombrid fish. Am. Mus. Novit. 1000,
32 p.
Cressey, R.F.
1975 A new family of parasitic copepods (Cyelopoida, Shiinoi-
dae). Crustaceana 28(2):211-219.
Cressey, R.F., and H.B. Cressey
1980 Parasitic copepods of mackerel- and tuna-like fishes
(Scombridae) of the world. Smithson. Contrib. Zool. 311,
186 p.
Cressey, R.F., B.B. Collette. and J.L. Russo
1983 Copepods and scombrid fishes: A study in host-parasite
relationships. Fish. Bull., U.S. 81(2):227-265.
de Beaufort, L.F.
1951 Percomorphi (concluded). In The fishes of the Indo-
Australian Archipelago 9, p. 1-242. E.J. Brill, Leiden.
de Sylva. D.P.
1955 The osteology and phylogenetic relationships of the
blackfin tuna, Thunnus atlanticus (Lesson). Bull. Mar. Sci.
Gulf Caribb. 5(1):1-41.
Devaraj, M.
1977 Osteology and relationships of the Spanish mackerels and
seerfishes of the tribe Scomberomorini. Indian J. Fish.
22(l-2):l-67.
Dung, D.LY., and W.F. Royce
1953 Morphometric measurements of Pacific scombrids. U.S.
Fish Wildl. Serv. Spec. Sci. Rep. Fish. 95, 170 p.
Dyer, W.G., E.H. Williams Jr., and L.B. Williams
1989 Monogeneans from marine fishes of Okinawa, Japan.
Proc. Helminthol. Soc. Wash. 56(l);64-68.
Evermann, B.W., and A. Scale
1907 Fishes of the Philippine Islands. Bull. U.S. Bur. Fish.
26:49-110.
Fierstine, H.L., and V. Walters
1968 Studies in locomotion and anatomy of scombroid fishes.
Mem. South. Calif. Acad. Sci. 6:1-31.
Fourmanoir, P., and P. Laboute
1976 Poissons de Nouvelle Cal^donie et des Nouvelles
Hebrides. Editions Pacifique, Papeete, Tahiti, 376 p.
Fraser-Brunner, A.
1950 The fishes of the family Scombridae. Ann. Mag. Nat.
Hist., Ser. 12, 3:131-163.
Gibbs, R.H. Jr., and B.B. Collette
1967 Comparative anatomy and systematics of the tunas,
genus Thunnus. Fish. Bull., U.S. 66(1):65-130.
Gill, T.N.
1862 On the limits and arrangement of the family of scom-
broids. Proc. Acad. Nat. Sci. Phila. 14:124-127.
Gillett. R.
1987 Traditional tuna fishing: A study at Satawal, Central
Caroline Islands. Bishop Mus. Bull. Anthropol. 1, 48 p.
Gnanamuttu, J.C.
1971 Osteology of the Indian mackerel, Rastrelliger kanagurta
(Cuvier). Indian J. Fish. 13(l):l-26.
Gorbunova, N.N.
1965 Seasons and conditions of spawning of the scombroid
fishes (Pisces, Scombroidei). Tr. Inst. Okeanol. Akad. Nauk
SSSR 80:36-61 [in Russ., Engl. summ.].
1974 A review of larvae of scombrid fishes (Scombridae,
Pisces). Tr. Inst. Okeanol. Akad. Nauk SSSR 96:23-76 [in
Russ., Engl. summ.].
Grant, E.M.
1965 Guide to fishes. Queensl. Dep. Harbours and Marine,
Brisbane, 280 p.
1972 Guide to fishes. Queensl. Dep. Primary Ind., Brisbane,
2d ed., 472 p.
1975 Guide to fishes. Coordinator-General's Dep., Brisbane,
3d ed., 640 p.
1978 Guide to fishes. Dep. Harbours and Marine, Brisbane,
4th ed., 768 p.
1982 Guide to fishes. Dep. Harbours and Marine, Brisbane,
5th ed., 896 p.
1987 Fishes of Australia. E.M. Grant Pty Ltd., Queensl.,
480 p.
Giinther, A.
1860 Catalogue of the acanthopterygian fishes in the collec-
tion of the British Museum, Vol. 2. London, 548 p.
Gushiken, S.
1973 Fishes of the Okinawa Islands. Ryukyu Fish. Assoc,
Naha, 247 p.
Hardenberg, J.D.F.
1935 Some new or rare fishes of the Indo-Australian .A.rchi-
pelago IV. Treubia 15(2):131-140.
Helfman, G.S., and J.E. Randall
1973 Palauan fish names. Pac. Sci. 27(2):136-153.
Herre, A.W.C.T.
1931 The fishes of the Herre 1931 Philippine Expedition with
descriptions of 17 new species. Newspaper Enterprise Ltd.,
Hong Kong, 106 p. [reprinted by N.K. Gregg, Kentfield, CA].
1953 Check list of Philippine fishes. U.S. Fish Wildl. Serv.
Res. Rep. 20, 977 p.
Hutchins, B.
1979 A guide to the marine fishes of Rottnest Island. Creative
Research, Perth, 103 p.
1990 Fish survey of South Passage, Shark Bay, Western
Australia. In Berry, P.F., S.D. Bradshaw, and B.R. Wilson
(eds.). Research in Shark Bay, p. 263-278. Rep. France-Aust.
Bicent. Exped. Comm.. West. Aust. Mus., Perth.
Hutchins, B., and R. Swainston
1986 Sea fishes of southern Australia. Swainston Publ.,
Perth, 180 p.
Hutchins, B., and M. Thompson
1983 The marine and estuarine fishes of southwestern Aus-
tralia. A field guide for anglers and divers. West. Aust. Mus.,
Perth, 103 p.
Iwai, T., and I. Nakamura
1964 Branchial skeleton of the bluefin tuna, with special refer-
ence to the gill rays. Bull. Misaki Mar. Biol. Inst., Kyoto Univ.
6:21-25.
Collette and Gillis. Osteological differences between two species of Grammatorcynus
51
Jenkins, G.P.
1989 Scombridae— Mackerels, Spanish mackerels, bonitos,
tunas. In Leis, J.M., and T. Trnski (eds.), The larvae of Indo-
Pacific shorefishes, p. 291-297. New South Wales Univ.
Press, Kensin^on.
Johannes, R.E.
1981 Words of the lagoon. Fishing and marine lore in the Palau
District of Micronesia. Univ. Calif. Press, Berkeley, 245 p.
Johnson, G.D.
1986 Scombroid phylogeny: An alternate hypothesis. Bull.
Mar. Sci. 39(1):1-41.
Jones, S.
1968 The scombroid fishery of India— present and future. In
Proc, Symp. Scombroid Fishes, Mar. Biol. Assoc. India, Symp.
Ser. 1(3):994-1000.
1969 Catalogue of fishes from the Laccadive Archipelago in
the reference collections of the Central Marine Fisheries
Research Institute. Bull. Cent. Mar. Fish. Res. Inst. 8, 32 p.
Jones, S., and M. Kumaran
1964 Eggs, larvae and juveniles of Indian scombroid fishes. In
Proc, Symp. Scombroid Fishes, Mar. Biol. Assoc. India, Symp.
Ser. l(l):343-378.
1981 Fishesof the Laccadive Archipelago. Nature Conserv.
and Aquat. Sci. Serv., Trivandrum, 760 p.
Jones, S., and E.G. Silas
1963 Tuna and tuna-like fishes from the Indian Seas. FAO
Fish Rep. 3(6):1775-1796,
1964a A systematic review of the scombroid fishes of India. In
Proc, Symp. Scombroid Fishes, Mar. Biol. Assoc. India, Symp.
Ser. 1(1):1-105.
1964b Mackerel from the Andaman Sea. In Proc, Symp.
Scombroid Fishes, Mar. Biol. Assoc. India, Symp. Ser. 1(1):
255-282.
Jones, S., E.G. Silas, and E. Dawson
1960 New records of scombroid fishes from the Andaman-
Nicobar waters. J. Mar. Biol. Assoc. India 2(1):136-137.
Joseph, J., W. Klawe, and P. Murphy
1979 Tuna and billfish-Fish without a country. Inter-Am.
Trop. Tuna Comm., La Jolla, 46 p.
Kailola, P.J.
1975 A catalogue of the fish reference collection at the Kanudi
Fisheries Research Laboratory, Port Moresby. Dep. Agric,
Stock, Fish., Port Moresby, Res. Bull. 16, 277 p.
Kamohara, T.
1964 Fishes obtained in the Okinawa and the Yaeyama Islands.
Res. Rep. Kochi Univ. 13(5):31-43 [in Jpn.].
1967 Fishes of Japan in color. Hoikusha Publ. Co., Osaka,
135 p.
Kishinouye, K.
1923 Contributions to the comparative study of the so-called
scombroid fishes. J. Coll. Agric. Imp. Univ., Tokyo 8(3):
293-475.
Klawe, W.L.
1977 What is a tuna? Mar. Fish. Rev. 39(ll):l-5.
Klunzinger, C.B.
1871 Synopsis der Fische des Rothen Meeres. II. Theil. Verb.
Zool. Bot. Ges. Wien 21:441-668.
Kuronuma, K.
1961 A check list of the fishesof Vietnam. Di v. Agric, Nat.
Res., U.S. Oper. Mission Vietnam, ICA Contract IV-153, 66 p.
Kyushin, K., K. Amaoka, K. Nakaya, H. Ida, Y. Tanino, and
T. Senta
1982 Fishes of the South China Sea. Jpn. Mar. Fish. Res.
Cent., Tokyo, 333 p.
Leviton, A.E., R.H. Gibbs Jr., E. Heal, and C.E. Dawson
1985 Standards in herpetology and ichthyology: Part I. Stan-
dard symbolic codes for institutional resource collections in
herpetology and ichthyology. Copeia 1985:802-832.
Lewis, A.D.
1981 Population genetics, ecology and systematics of Indo-
Australian scombrid fishes, with particular reference to skip-
jack tuna {Katsuwonus pelam.is). Ph.D. thesis, Aust. Natl.
Univ. 314 p.
Lewis, A.D., B.R. Smith, and R.E. Kearney
1974 Studies on tunas and baitfish in Papua New Guinea
waters— II. Dep. Agric, Stock, Fish, Port Moresby, Res. Bull.
11, 113 p.
Lewis, A.D., L.B. Chapman, and A. Sesewa
1983 Biological notes on coastal pelagic fishes in Fiji. Fish.
Div. (MAF) Fiji Tech. Rep. 4:1-68.
Lewis, A.G.
1968 Copepod crustaceans parasitic on fishes of Eniwetok
Atoll. Proc U.S. Natl. Mus. 125(3656):l-78.
Magnuson, J.J.
1973 Comparative study of adaptations for continuous swim-
ming and hydrostatic equilibrium of scombroid and xiphoid
fishes. Fish. Bull, U.S. 71(2):337-356.
Magnuson, J.J., and J.G. Heitz
1971 Gill raker apparatus and food selectivity among mack-
erels, tunas, and dolphins. Fish. Bull., U.S. 69(2):361-370.
Mago Leccia, F.
1958 The comparative osteology of the scombroid fishes of the
genus Scomberomorus from Florida. Bull. Mar. Sci. Gulf
Caribb. 8(4):299-341.
Marshall, T.C.
1964 Fishes of the Great Barrier Reef and coastal waters of
Queensland. Angus &. Robertson, Sydney, 566 p.
1965 Fishes of the Great Barrier Reef and coastal waters of
Queensland. Livingston Publ. Co., Narberth, PA.
1966 Tropical fishes of the Great Barrier Reef. Angus &
Robertson. Sydney, 239 p.
Masuda, H., C. Araga, and T. Yoshino
1975 Coastal fishes of southern Japan. Tokai Univ. Press,
Tokyo, 379 p.
Masuda, H., K. Amaoka, C. Araga, T. Uyeno, and T. Yoshino
(editors)
1984 The fishes of the Japanese Archipelago. Tokai Univ.
Press, Tokyo, 437 p.
Matsubara, K.
1955 Fish morphology and hierarchy. Part 1. Ishizald Shoten,
Tokyo, 789 p. [in Jpn.].
McAllister, D.E.
1968 The evolution of branchiostegals and associated opercular,
gular, and hyoid bones and the classification of teleostome
fishes, hving and fossil. Bull. Natl. Mus. Canada 221, 239 p.
McCuUoch, A.R.
1915 Notes on, and descriptions of Australian fishes. Proc
Linn. Soc NSW 40:259-277.
1922 Check list of the fish and fish-like animals of New South
Wales. Part 3. Aust. Zool. 2(3):86-130.
1929 A check-list of the fishes recorded from Australia. Aust.
Mus. Mem. 5(1):1-144.
McCulloch, A.R., and G.P. Whitley
1925 A list of the fishes recorded from Queensland waters.
Mem. Queensl. Mus. 8(2):125-182.
McPherson, G.
1984 Two species of double-lined mackerel (family Scombridae)
in Queensland waters. Queensl. Dep. Primary Ind. Leafl., 2 p.
52
Fishery Bulletin 90(1). 1992
Meyer, A.B.
1885 Catalogo de los peces recolectados en el Archipielago de
las Indias Orientales. Ann. Soc. Espafi. Hist. Nat. 14:5-49.
Morice. J.
1953 Essai syst^matique sur les families des Cybiidae, Thun-
nidae at Katsuwonidae, poissons scombroides. Rev. Trav.
Inst. Peches Marit. 18(l):35-63.
Munro, I.S.R.
1958a Families Polynemidae to Scomberomoridae. Handbook
of Australian fishes 27:109-112. Aust. Fish. Newsl. 17(10):
17-20.
1958b The fishes of the New Guinea region. Papua New
Guinea Agric. J. 10(4):97-369.
1967 The fishes of New Guinea. Dep. Agric, Stock, Fish.,
Port Moresby. New Guinea, 650 p.
Myers, R.F.
1988 An annotated checklist of the fishes of the Mariana
Islands. Micronesica 21(1-2):115-180.
1989 Micronesian reef fishes. Coral Graphics, Guam, 298 p.
Nakamura, I.
1965 Relationships of fishes referable to the subfamily Thun-
ninae on the basis of the axial skeleton. Bull. MisaJd Mar. Biol.
Inst., Kyoto Univ. 8:7-38.
Nishikawa, Y.
1979 Early development of the double-lined mackerel, Gram-
matoreynus bicarinatiis (Quoy and Gaimard), from the western
tropical Pacific. Far Seas Fish. Res. Lab., Bull. 17:125-140.
Nishikawa, Y., and D.W. Rimmer
1987 Identification of larval tunas, billfishes and other scom-
broid fishes (suborder Scombroidei): An illustrated guide.
CSIRO Mar. Lab. Rep. 186, 20 p.
Ogilby, J.D.
1918 Ichthyological notes (no. 4). Mem. Queensl. Mus. 6:
97-05.
Okada, Y.
1938 A catalogue of vertebrates of Japan. Maruzen Co.,
Tokyo, 412 p.
Okiyama, M., and S. Ueyanagi
1977 Larvae and juvenile of the Indo-Pacific dogtooth tuna,
Gymnosarda unicolor (Riippell). Bull. Far Seas Fish. Res.
Lab., Shimizu 15:35-49.
1978 Interrelationships of scombroid fishes: An aspect from
larval morphology. Bull. Far Seas Fish. Res. Lab., Shimizu
16:103-113.
Orsi, J.J.
1974 A check list of the marine and freshwater fishes of Viet-
nam. Publ. Seto Mar. Biol. Lab. 21(3/4):153-177.
Potthoff, T.
1975 Development and structure of the caudal complex, the
vertebral column, and the pterygiophores in the blackfin tuna
(Thunnus atlanticus, Pisces, Scombridae). Bull. Mar. Sci.
2.5(2):205-231.
Quoy, J.R.C., and J. P. Gaimard
1824-1825 Voyage autour du Monde, . . . Execute sur les cor-
vettes de S. M. rUranie et la Physicienne, pendant les ann^es
1817, 1818, 1819 et 1820. 3, Zoologie, 712 p.
Randall, J.E., G.R. Allen, and R.C. Steene
1990 Fishes of the Great Barrier Reef and Coral Sea. Univ.
Hawaii Press, Honolulu, 507 p.
Rau, N., and A. Rau
1980 Commercial marine fishes of the Central Philippines (bony
fishes). German Agcy. for Tech. Coop. (GTZ), 629 p.
Richards, W.J., and W.L. Klawe
1972 Indexed bibliography of the eggs and young of tunas and
other scombrids (Pisces, Scombridae) 1880-1970. NOAA
Tech. Rep. NMFS SSRF-652, 107 p.
Rivaton, J., P. Fourmanoir, P. Bourret, and M. Kulbicki
1989 Catalogue des poissons de Nouvelle-Caledonie.
ORSTOM Rapp. Prov., 170 p.
Rohde, K.
1976 Monogenean gill parasites of Scomberomoms commer-
soni Lac^pede and other mackerel on the Australian east
coast. Z. Parasitenkde. 51:49-69.
Romimohtairo, K., M. Hutomo, and Sumarna
1974 Results of the biological observations in Camber Bay, Gag
Island with a special note on the tuna and skipjack fishing
ground in Irian Jaya waters. Oseanol. Indones. 3:29-45.
Roux-Esteve, R., and P. Fourmanoir
1955 Poissons captures par le mission de la Calypso en Mer
Rouge. Ann. Inst. Oceanogr., Paris 30:195-203.
Ruppell, E.
1836 Neue Wirbelthiere zu der fauna von Abyssinien gehorig.
Fische des rothen Meeres. Frankfurt am Main, 148 p. Lfg.
6:29-52.
Russell, B.C.
1983 Annotated checklist of the coral reef fishes in the
Capricorn-Bunker Group, Great Barrier Reef Australia. Great
Barrier Reef Marine Park Authority, Spec. Publ. Ser. 1, 184 p.
Schultz, L.P.
1960 Suborder Scombrina. Family Scombridae: Tunas. In
Schultz. L.P., et al. Fishes of the Marshall and Marianas
Islands. Bull. U.S. Nat. Mus. 202(2):410-417.
Serventy, D.L.
1950 Tuna survey of north Australia. Aust. Fish. Newsl.
9:18-20.
Shaklee, J.B.
1983 The utilization of isozymes as gene markers in fisheries
management and conservation. In Isozymes: Current topics
in biological and medical research, vol. 11:213-247. Liss, NY.
Shiino, S.M.
1972 List of English names of Japanese fishes with proposi-
tion of new names. Sci. Rep. Shima Marineland 1, 210 p. [in
Jpn. and Engl.].
1976 List of common names of fishes of the world, those
prevailing among English-speaking nations. Sci. Rep. Shima
Marineland 4, 262 p.
Shirai, S.
1980 (rev. ed.) Ecological encyclopedia of the marine animals
of the Ryukyu Islands in colour. Okinawa Kyoiku Shuppan,
636 p.
Silas, E.G.
1963 Synopsis of biological data on double-lined mackerel
Grammatorcynns bicarinatus (Quoy and Gaimard) (Indo-
Pacific). FAO Fish. Rep. 6, vol. 2:811-833.
Smith, C.L., and R.M. Bailey
1962 The subocular shelf of fishes. J. Morphol. 110(1):1-17.
Springer, V.G., and J.T. Williams
1990 Widely distributed Pacific Plate endemics and lowered
sea-level. Bull. Mar. Sci. 46(3):631-640.
Springer, V.G., Burhanuddin, and M.F. Gomon
1974 List of fishes collected during Rumphius Expedition
I. Oseanol. Indones. 1:39-45.
Tongyai. M.L.P.
1966 Background notes for Spanish mackerel, (Scomberomorus
spp.) from Thai waters, studies, 1954-1965. Thai Mar. Fish.
Lab., Scomberomorus Rep. Paper 1, 25 p.
1970 Plah in-see, Scomieromonts spp., of Thailand, 1967. In
Marr, J.D. (ed.). The Kuroshio— A symposium on the Japan
Current, p. 557-564. East- West Center Press, Honolulu.
1971 A taxonomic study of the pla in-si (Pisces: Scombroidei)
of Thailand. Div. Mar. Fish. Tech., Dep. Fish., Ministry
Agric, Bangkok, 36 p.
Collette and Gillis: Osteological differences between two species of Grammatorcynus 53
Ueyanagi, S., and M. Okiyama
1979 Larval fish characters and phylogeny of scombroid
fishes. Kaiyo Kagiiku ll(2):93-99 [in Jpn., transl. in NMFS
Proc. Rep. 82-14, NMFS, NOAA, Seattle, 24 p.].
Umali, A.F.
1950 Key to the families of common commercial fishes in the
Philippines. U.S. Fish Wildl. Serv. Res. Rep. 21, 47 p.
Uyeno, T., and S. Fujii
1975 A fish fossil of the family Scombridae from a Miocene
bed in Toyama Prefecture, Japan. Bull. Nat. Sci. Mus. Tokyo,
Ser. C(Geol.) 1(1):11-16.
Wade, C.B.
1951 Larvae of tuna and tuna-like fishes from Philippine
waters. Fish. Bull, U.S. 51:445-485.
Wang, C.
1981 Studies on the fish fauna of the South China Sea Islands,
Guangdong Province. China. Oceanol. Limnoi. Sin. (Suppl.)
p. 137-165. [in Chinese].
Warfel, H.E.
1950 Outlook for development of a tuna industry in the Philip-
pines. U.S. Fish Wildl. Serv. Res. Rep. 28. 37 p.
Wass. R.C.
1984 An annotated checklist of the fishes of Samoa. NOAA
Tech. Rep. NMFS SSRF-781, 43 p.
Whitley, G.P.
1947 New sharks and fishes from Western Australia. Part
3. Aust. Zool. 11(2):129-150.
1948 A list of the fishes of Western Australia. West. Aust.
Fish. Dep., Fish. Bull. 2, 35 p.
1964a Scombroid fishes of Australia and New Zealand. In
Proc, Symp. Scombroid Fishes, Mar. Biol. Assoc. India Symp.
Ser. l(l):221-253.
1964b A survey of Australian ichthyology. Proc. Linnean Soc.
NSW 89(1):11-127.
Yamakawa, T.
1979 Studies of the fish fauna around the Nansei Islands,
Japan. 1. Check list of fishes collected by Toshiji Kamohara
and Takeshi Yamakawa from 1954 to 1971. Rep. Usa Mar.
Biol. Stn., Suppl. 1:1-47.
Zhang, R.
1981 On the larvae of tuna-like fishes around waters of Xisha
Qundao of the Nan Hai. J. Fish. China 5(4):301-315 [in
Chinese, Engl. summ.].
Zharov, V.L.
1967 Classification of the scombroid fishes (suborder Seom-
broidei, order Perciformes). Vopr. Ikhtiol. 7(2):209-224 [in
Russ.].
Zug, G.R., V.G. Springer, J.T. Williams, and G.D. Johnson
1989 The vertebrates of Rotuma and surrounding waters.
Atoll Res. Bull. 316, 25 p.
Abstract.- The spinner dolphin
Stmella langirostris is widely distrib-
uted in the eastern tropical Pacific
Ocean. Geographic patterns in 30 cra-
nial features were determined from
246 museum specimens grouped in-
to 25 5° latitude-longitude blocks.
Statistically significant sexual dimor-
phism was demonstrated for one-half
of the cranial characters, with males
generally being larger. ANOVAs, as
well as principal components, canon-
ical variates, and cluster (UPGMA
and function-point) analyses demon-
strated geographic variation in all
characters. Patterns of geographic
variation in morphology were evalu-
ated for all S. longirostris specimens
using Mantel tests and matrix corre-
lations; 20 of 30 characters showed
significant "regional patterning,"
while most (25 of 30) exhibited "local"
patterning. The latitude-longitude
block with specimens of S. I. centro-
americana was distinctive in a num-
ber of features. Also, eastern spin-
ner dolphins (S. /. orientalis) were
smaller than spinners found to the
south, southwest, or west. Many of
the cranial characters exhibited a
concentric pattern of geographic
variation similar to that found by
previous investigators for several ex-
ternal characters. Hawaiian speci-
mens are the largest incorporated
into this study and, typically, are
more like those from southern local-
ities than animals from geographical-
ly closer blocks. The association be-
tween morphological characters and
13 environmental measures was as-
sessed with Mantel tests and product-
moment correlations, revealing sta-
tistical concordance of morphological
patterns for a number of cranial char-
acters with those for water depth,
sea surface temperature in January
and July, surface salinity, thermo-
cline depth, and surface dissolved
oxygen. Several of these environ-
mental variables manifest the same
distributional pattern found in many
of the cranial features.
Geographic variation in
cranial morphology of spinner
dolphins Stenella longirostris in
the eastern tropical Pacific Ocean
Michael E. Douglas
Oklahoma Biological Sun/ey and Department of Zoology
University of Oklahoma, Norman. Oklahoma 73019
Present address: Department of Zoology and Museum
Arizona State University, Tempe, Arizona 85287
Gary D. Schnell
Daniel J. Hough
Oklahoma Biological Survey and Department of Biology
University of Oklahoma, Norman, Oklahoma 73019
William F. Perrin
Southwest Fisheries Science Center, National Marine Fisheries Service, NOAA
PO Box 271, La Jolla, California 92038
Manuscript accepted 9 December 1991.
Fishery Bulletin, U.S. 90:54-76 (1992).
Information on geographic variation
of dolphins in the eastern tropical
Pacific is of intrinsic scientific inter-
est, but also has practical implica-
tions because fishermen in the region
kill dolphins in the course of purse-
seining for yellowfin tuna (Allen
1985). Tuna in the region associate
with schools of dolphins, primarily
Stenella spp. and Delphinus delphis,
and the fishermen set their nets on
the schools to capture the tuna below
them. In the process, many dolphins
die, as many as 80,000-125,000 aimu-
ally in recent years (Hall and Boyer
1988, 1989, 1990). The U.S. Govern-
ment has used a series of manage-
ment units, or stocks, in regulating
this exploitation of the dolphins by
U.S. vessels. For the spinner dolphin,
these have been the eastern spinner,
Costa Rican spinner, northern white-
belly spinner, and southern white-
belly spinner stocks (Perrin et al.
1985). These divisions are based on
morphology, including body length
and shape, color pattern, shape of the
dorsal fin, and cranial characters.
The Costa Rican form occurs close to
the coast of Central America and is
relatively large, with relatively long
beak, erect to forward-canted dorsal
fin, and monotonic gray coloration.
The eastern form is smaller, with
shorter beak; it also has the erect or
canted fin and is gray overall, but
with light patches in the axillary and
genital areas. The whitebelly forms
have a tripartite color pattern of dark
gray, light gray, and (ventrally) white,
and the dorsal fin is highly variable,
ranging in adults from falcate to
erect. The northern and southern
stocks were divided based on modal
differences in cranial measurements;
the boundary is at the Equator. The
eastern spinner and northern white-
belly spinner stocks overlap broadly;
overlap between the eastern spinner
and southern whitebelly spinner is
very slight (Perrin et al. 1985). Dol-
phins killed in the fishery are iden-
tified to stock based on the modal ap-
pearance of adults in the school and,
in the case of the two whitebelly
stocks, location.
Most recently, Perrin (1990) de-
scribed three subspecies of Stenella
54
Douglas et al : Geographic variation in cranial morphology of Stenella longirostns
55
longirostris: the pantropical spin-
ner dolphin S. I. longirostris
occurring in the Central, South,
and Western Pacific, Indian, and
Atlantic oceans; the Central
American spinner S. I. centra-
americana endemic to the coast
of Central America and corre-
sponding to the Costa Rican spin-
ner management stock; and the
eastern spinner S. I. orientalis
endemic to the eastern tropical
Pacific off Mexico, Central Amer-
ica, and northern South America
and corresponding to the eastern
spinner management stock. He
concluded that the more offshore
whitebelly forms constitute a
broad zone of hybridization or
intergradation between the east-
ern and pantropical forms. This
view has support from results of
a genetic study; Dizon et al.
(1991) found no unique haplo-
types in a restriction-enzyme ex-
amination of mitochondrial DNA
of animals of the eastern and
whitebelly morphological types.
Perrin et al. (1991) reexamined
color pattern, body size and
shape, and dorsal fin shape with-
out a priori assignment of speci-
mens to subspecies or management stock. They com-
pared specimens from 5° geographic blocks. The re-
sults of their analyses support the taxonomic treatment
by Perrin (1990); the whitebelly forms constitute a com-
plex zone of highly variable animals intermediate be-
tween the eastern and pantropical types. Perrin et al.
(1991) concluded that the pattern of geographic varia-
tion does not justify separation of northern and south-
ern units on morphological grounds alone.
The purpose of the studies reported here was to carry
out a parallel analysis of geographical cranial variation
in the eastern Pacific, again making no a priori assign-
ment of specimens to subspecies or management stock.
We also examined relationships between cranial varia-
tion and environmental variables, in an effort to better
understand the ecologies of the several forms of spin-
ner dolphins.
Materials and methods
Data from 246 adult museum specimens (maturity
judged by evaluating fusion of premaxilla with the max-
09
IBl
■ -.,'
.■'''"■■
1
1
1
1
^J
.p
m
^
OB
1
1
6
i'^ -'■ '
^
s
%
■'■'/
m
f^
i
07
S0i
1
\
X
k.
i
vv
<>
06
m
1
1
8
14
*
13
5
It
13
9
f
3
A
' 1
^
05
IS
2
1
2
1
2
4
3
1 1
14
*
2
2
5
5
3
6
3
1
]
W
1
"'■■■■■
- /
1
2
1
2
3
/
03
Slenella
■"•:;:.-
1
1
7
5
2-^-
4
02
Males
Females
1
,J'
■'-}'■-■■
9
4
»
12
14
*
3
V
01
kji '
. "
-
■ '■
7
5
1 1 1
■-:..■;■
■" --;.-.
--.•, : -.
■.--■^c.
•■■..;■■:
-»
\
Figure 1
Known range of Stenella longirostris in eastern tropical Pacific Ocean (modified from
Perrin 1990), with numbers of males (above) and females (below) available for each 5°
latitude-longitude block (total of 246 specimens). Asterisks indicate 10 blocks included
in analysis of sexual dimorphism. The 25 blocks with two or more specimens used as
basis for analyses of geographic variation; for some aspects, the 10 blocks with single
specimens projected onto axes based on the 25 blocks. Each block identified by numerical
code (numbers on left and bottom margins are combined; e.g., block 0812 is just to east
of southern tip of Baja California). One block (i.e., 06-02) is located off map to west (left).
ilia at distal end of rostrum; Dailey and Perrin 1973)
of spinner dolphins were used in this investigation (Fig.
1). We purposely included all appropriate specimens
available, including those from the three named sub-
species recognized from the region (Perrin 1990); fur-
thermore, we did not differentiate between those with
different color patterns ("eastern" and "whitebelly";
Perrin et al. 1985, 1991), in order to focus simply on
cranial features. The animals used included 188 of 199
specimens used in the earlier study of sexual dimor-
phism (Douglas et al. 1986; the 11 remaining specimens
not used had been incorrectly aged or had inadequate
locality data) and 58 new specimens.
The first set of specimens was measured by M.E.
Douglas and the new specimens by W.F. Perrin. In
addition, Perrin remeasured 81 specimens of spinner
dolphins and spotted dolphins S. attenuata measured
by Douglas. This allowed a comparison to determine
whether measurements were repeatable. Initially, 36
morphometric and meristic characters were evaluated
(illustrations and character definitions given in Schnell
et al. 1985a). Comparisons of measurements taken on
the same specimens by the two investigators indicated
56
Fishery Bulletin 90(1). 1992
Table
1
Geographic variation and sexual dimorphism in Stenella longriros
ris evaluated for 30 characters.
F-value*"
Mean"^
Correction
Percentage
Character"
Block
Sex
Male
Female
factor''
difference'
1 Condylobasal L.
22. 19***
0.05
405.9
404.6
0.22
0.32
2 L. Rostrum (frm.Base)
15.93***
0.50
258.9
259.4
-0.61
-0.21
3 L. Rostrum (frm. Pterygoid)
20.71***
0.01
299.6
299.3
-0.08
-0.11
4 W. Rostrum (at Base)
15.19***
2.88
74.2
73.1
0.37
1.44
5 W. Rostrum (at 1/4 L.)
10.17***
11.82***
52.1
50.6
0.67
2.84
6 W. Rostrum (at 1/2 L.)
9.97***
10.20**
44.2
42.8
0.65
3.16
7 W. Premax. (at 1/2 L.)
5.65***
8.31**
21.4
20.8
0.31
3.02
8 W. Rostrum (at 3/4 L.)
2.92**
24.96***
32.5
30.5
1.01
6.34
9 Preorbital W.
38.05***
8.67**
139.9
137.5
1.00
1.76
10 Postorbital W.
49.34***
8.19**
155.7
153.3
0.93
1.57
11 Skull W. (at Zygomatic P.)
49. II***
14.89***
154.4
151.3
1.27
2.04
12 Skull W. (at Parietals)
6.27***
20.36***
130.1
127.2
0.10
0.52
13 Ht. Braincase
16.56***
15.52***
89.1
87.1
0.89
2.28
14 L. Braincase
18.71***
8.09**
101.7
100.3
0.67
1.47
15 Max. W. Premax.
6.55***
0.22
62.9
62.6
0.04
0.36
16 W. External Nares
3.88***
0.09
41.6
41.5
1.40
2.27
17 L. Temporal Fossa
4.32***
9.27**
50.4
48.7
0.82
3.52
18 W. Temporal Fossa
9.24***
17.82***
40.2
38.2
0.96
5.16
19 Orbital L.
6.56***
0.00
40.7
40.6
-0.01
-0.13
20 L. Antorbital P.
12.41***
11.46***
42.7
41.4
0.67
3.27
21 W. Internal Nares
22.50***
3.85
43.5
42.7
0.31
1.81
22 L. Up. Toothrow
16.23***
1.12
224.3
225.5
-0.82
-0.54
23 No. Teeth (Up.Lf.)
3.39***
3.88
53.2
52.5
0.42
1.30
24 No. Teeth (Up.Rt.)
5.19***
1.15
52.7
52.3
0.21
0.76
25 No. Teeth (Low.Lf.)
2.33*
0.80
51.3
51.1
0.17
0.39
26 No. Teeth (Low.Rt.)
2.61**
0.34
51.0
50.9
0.11
0.29
27 L. Low. Toothrow
13.99***
1.00
218.4
219.5
-0.76
-0.51
28 Ht. Ramus
21.64***
13.02***
55.4
54.1
0.60
2.51
29 Tooth W.
3.74***
13.84***
2.6
2.5
0.07
5.10
30 L. Ramus
18.06*** 0.04
= height; L. = length; Lf. = left;
346.8
Low. = lower;
345.6 0.20
Max. = maximum; No. = number;
0.35
P. = process;
'Abbreviations: frm. = from; Ht.
Premax. = premaxillary; Rt. = right; Up. = upper;
W. = width.
•"F-values from main effects two-way analysis of variance (5° block i
/s. sex) involving 10 blocks ( *j
P<0.05; **P<0.01;
•**P<0.001).
Total of 170 individuals. Degrees
of freedom 9 for
among-block variation and 1 for between sexes.
"Unweighted mean for 10 blocks.
''Added to all individual female measurements and subtracted from all individual male measurements to correct for sexual differences. |
■■ Difference between sexes (males minus females) multiplied by 100, with the resulting value divided by
average of male and female means.
that 6 of the original 36 measurements (i.e., W. Lf.
Premax. [at midline of Nares], W. Rt. Premax. [at
midline of Nares], Separation of Pterygoids, L. Lf.
Tympanic Cavity, L. Rt. Tympanic Cavity, and W. at
Pterygobasioocipital Sutures; abbreviations used in
these and other character names are listed in footnote
a of Table 1) should be deleted, because we were not
able consistently to repeat these measurements. For
some other measurements, there were differences
between investigators, but the differences were con-
sistent (e.g., one obtained measurements that were
smaller than those reported by the other). Therefore,
we calculated regression equations for each of the
remaining characters based on the 81 jointly-measured
specimens. These regression equations were used to
convert the measurements from the rest of the initial
specimens to appropriate values for inclusion with the
measurements taken by Perrin. Through these pro-
cedures, we developed a data set of 30 characters (listed
in Table 1) for 246 specimens.
Only specimens that were largely complete were in-
cluded in the analysis. Missing values (1.34% of total)
were estimated by linear regression ("Missing Data
Estimator" program developed by Dennis M. Power,
Santa Barbara Mus. Nat. Hist., pers. commun.) onto
the character that explained the greatest proportion
of the variance for the variable under consideration.
Specimens then were assigned to 5° latitude-longi-
Douglas et al : Geographic variation in cranial morphology of Stenella longirostns
57
tude blocks, with each geographic block given a
numerical code (see Fig. 1). We had specimens from
35 blocks, although 10 were represented by only a
single specimen; the other 25 blocks were used as the
basis for most analyses of geographic variation. While
several of the remaining 25 blocks are represented by
relatively small samples, tests for geographic pattern-
ing (described below) suggest that, in general, sample
values are representative of what would be expected
for these blocks based on their geographic positions.
The 5° block size was selected, in part, because it was
judged that available sample sizes would not permit
detailed analysis of smaller geographic units. Further-
more, migratory movements and related factors were
less likely to significantly influence results when these
relatively large sampling areas were used.
Douglas et al. (1986) showed that 5. longirostris in
the eastern tropical Pacific was sexually dimorphic for
13 of 36 characters. Because some specimens used in
that analysis were removed and new specimens added
(see above), we reanalyzed the data with a two-way
analysis of variance (ANO VA) for block and sex based
on specimens in 10 blocks that had at least four of each
sex (Fig. 1). We then produced a series of correction
terms to adjust measurements of the larger sex down-
ward and the smaller sex upward, thus producing sex-
adjusted or "zwitter" measurements (for details on this
adjustment, see Schnell et al. 1985a). These corrections
enabled us to combine specimens for both sexes in an
overall analysis of geographic variation.
Correlation, ordination and clustering
After conversion to zwitters, characters were then
standardized so that means for blocks were zeros and
standard deviations ones. Product-moment correlations
were computed among characters, and the general
associations among characters were summarized by
clustering characters using the unweighted pair-group
method with arithmetic averages (UPGMA).
This type of hierarchical cluster analysis also was
performed to summarize average distance coefficients
(Sneath and Sokal 1973) calculated for all pairs of
blocks based on standardized data. Cophenetic correla-
tion coefficients were computed to indicate the degree
to which distances in the resulting dendrogram accu-
rately represented original interblock morphologic
distances.
In addition, we analyzed standardized data using a
nonhierarchical /T-group method called function-point
cluster analysis (Katz and Rohlf 1973; described in
Rohlf et al. 1979). Blocks are assigned to a series
of subgroups at a specified level. The value for the
w-parameter used by the function-point clustering
method was varied. A hierarchical (but not necessar-
ily non-overlapping) system of clusters can be obtained
by conducting the analysis at more than one cluster-
ing level. Results are presented in the form of a modi-
fied skyline diagram (Wirth et al. 1966) where, for a
given w-value, blocks joined in a common line are in
the same cluster.
Based on standardized data, we constructed scatter
diagrams of blocks projected onto the first two prin-
cipal components (Sneath and Sokal 1973) extracted
from a matrix of correlations among the 30 characters.
Canonical variates analysis also was applied to deter-
mine the subset of variables that show the greatest
degree of geographic variation— in this case, those that
provide the greatest interblock separation relative to
the degree of intrablock variation (Program P7M of
BMDP; Dixon 1990). Plots of the first two canonical
variables show the maximum separation of blocks in
two-dimensional space. The original variables, which
in combination exhibited maximum interblock variabil-
ity, were then subjected to additional analyses.
Mantel test for geographic patterning
Using a test devised by Mantel (1967) and described
by Sokal (1979), we analyzed interlocality variation in
each character to determine whether values are geo-
graphically patterned, or vary spatially at random. This
procedure enabled us to determine whether differences
in character values between all pairs of samples are
statistically associated in a linear manner with corre-
sponding geographic distances. The observed asso-
ciation between sets of character differences and
geographic distances was tested relative to its permu-
tational variance, and the resulting statistic was com-
pared against a Student's i -distribution with infinite
degrees of freedom. Computations were performed
using GEOVAR, a library of computer programs for
geographic variation analysis written by David M.
Mallis and furnished by Robert R. Sokal (State Univer-
sity of New York at Stony Brook).
Character differences were compared first with ac-
tual geographic distances (in nautical miles) between
centers of blocks and then wath reciprocals of distances.
In evaluations of reciprocals, where distances are
scaled in a nonlinear manner, longer distances are con-
sidered effectively to be equal, and the portion of the
scale involving smaller distances is expanded. Thus, use
of reciprocals of distances increases the power of anal-
yses to reveal geographic patterns that are "local" in
nature (i.e., involving closely placed blocks), whereas
tests involving nautical-mile distances evaluate
"regional" trends. Positive associations of character
differences and nautical-mile distances are indicated
by positive i -values, while negative ^-values denote such
associations when reciprocals of distances are used.
58
Fishery Bulletin 90(1). 1992
Table 2
Environmental measurements compiled for each 5° latitude-longitude block.*
1 Sea Current (N.. Winter)— Average northern component (in knots) of the surface water current in winter (Innis et al. 1979; their
fig. 2.2).
2 Sea Current (W., Winter)— Average western component (in knots) of the surface water current in winter (Innis et al. 1979; fig. 2.3).
3 Water Depth— Average sea depth (in m) (Bartholomew 1975; fig. 122).
4 Solar Insolation (Jan.)— Average incoming solar radiation for January (in gm. ■ cal/cm'; Brunt 1934; table 2).
5 Solar Insolation (Annual)— Average annual incoming solar radiation in gm. ■ cal/cm-; Brunt 1934; table 2).
6 Sea Surface Temp. (Jan.)— Average January sea surface temperature (in °C; Robinson 1976: fig. 2 north of 5°S; Wyrtki 1974:
fig. 2 south of 5°S).
7 Sea Surface Temp. (July)— Average July sea surface temperature (in °C; Robinson 1976: fig. 74 north of 5°S; Wyrtki 1974: fig.
8 south of 5°S).
8 Sea Surface Temp. (Ann. Var.)— Average annual sea surface temperature variation (in °C; Robinson 1976: fig. 148 north of 5°S;
Wyrtki 1974: fig. 26 south of 5°S).
9 Oxygen Min. Layer (Depth)— Annual mean depth (in m) of the absolute oxygen minimum surface with respect to the vertical (Levitus
1982: fig. 52).
10 Surface Salinity— Average salinity ("Ain) of surface sea water (Levitus 1982: microfiche F-02, frames 2-5).
1 1 Thermocline Depth (Winter)— Mean depths (in m) to the top of the thermocline for January, February, and March (Robinson 1976:
figs. 12, 24, and 36 north of 5°S; Cromwell 1958: fig. la south of 5°S).
12 ThermocHne Depth (Summer)— Mean depths (in m) to the top of the thermocline for July, August, and September (Robinson 1976:
figs. 84, 96, and 108 north of 5°S; Cromwell 1958: fig. Ic south of 5°S).
13 Surface Dissolved Oxygen— Annual mean dissolved oxygen (mL/L) of surface sea water (Levitus 1982: microfiche F-03, frames 2-5).
'Abbreviations: Ann. Var. = Annual variation; Jan. = January; Min. = Minimum; N. = North; Temp. = Temperature; W. = West.
As an example of the Mantel procedure, consider the
25 blocks for which two or more specimens were
available (Fig. 1). The geographic distances (in nautical
miles) between each pair of the 25 blocks (300 pairs
total) are computed. We then obtain the mean value
for a given morphological character for each block; con-
sider a character with large mean values in northern
blocks, a gradual change as one proceeds south, and
the smallest means in the most southerly blocks. We
calculate the absolute character difference for each pair
of blocks (300 difference values); in general, for this
hypothetical case, close blocks geographically exhibit
small differences in character means, while blocks far
apart (e.g., a northern and a southern block) have the
largest morphological differences. We and the Mantel
test would identify this morphological character as hav-
ing a strong regional pattern. We also compare recip-
rocals of geographic distance for each block pair with
corresponding morphological differences; this approach
indicates whether, in general, geographically close
blocks also are similar morphologically (a case of local
geographic patterning). The examplar morphological
character, thus, would be identified as displaying a
strong local pattern (in addition to the strong regional
pattern). In general, a character showing a regional
pattern (as we have defined it) also will exhibit a local
pattern, but the reverse is not necessarily true. For in-
stance, if the morphological character was large in both
the north and south, was small for blocks in the middle,
and had gradual changes between adjacent blocks, it
would have a strong local pattern but no regional pat-
tern (because many distant blocks are nearly identical
morphologically). Detailed computational examples of
the Mantel test can be found in Douglas and Endler
(1982), Schnell et al. (1985b), and Manley (1985).
We also computed matrix correlations (Sneath and
Sokal 1973) between character differences and the
associated geographic distances or reciprocals of dis-
tances between localities. The significance of these
coefficients cannot, however, be tested in the conven-
tional way, because all pairs of localities were used and
these are not statistically independent. However, the
resulting values are useful as descriptive statistics in-
dicating the degree of association of difference values.
Morphological-environmental covariation
Relatively little is knowm about the relationship (if any)
of geographic variation in morphological characteristics
of S. longirostris to differences in the environment.
Therefore, as an initial exploratory analysis of covari-
ation, we have calculated product-moment correla-
tions of block means for morphological characters with
environmental variables. Data were available for 13
environmental variables for the eastern tropical Pacific
Ocean (Table 2). We also used UPGMA to summarize
associations among these environmental variables
for 51 blocks with specimens of S. longirostris or
Douglas et al : Geographic variation in cranial morphology of Stenella longirostns
59
S. attenuata or both; since these
two dolphin species have broad-
ly overlapping distributions in
the eastern tropical Pacific, the
blocks used are representative of
areas inhabited by S. longirostris.
We conducted a principal com-
ponents analysis of the 13 envi-
ronmental variables for the 51
blocks in order to obtain sum-
mary variables that reflect over-
all environmental trends. Indi-
vidual blocks were projected onto
the resulting environmental prin-
cipal components based on stan-
dardized data. These block vari-
ables were used as composite
environmental variables for
comparisons with morphological
characteristics.
In addition to using matrix cor-
relations and the Mantel proce-
dure to test for local and regional
patterning of variation in individ-
ual morphological characters, we
compared difference patterns of
selected morphological measures
with those of environmental vari-
ables. In these tests, differences
between each pair of blocks for a morphological vari-
able were compared with those for an environmental
variable.
Sources for environmental data are expanded over
those used by Schnell et al. (1986: table 2) so as to ac-
commodate the broader geographic representation
resulting from increased numbers of specimens. Values
for depth of the oxygen minimum layer were taken for
all blocks from Levitus (1982). Data for sea surface
temperatures and thermocline depths were not avail-
able in the previously used source for blocks west of
120°. Data for these and other blocks north of 5°S were
taken from Robinson (1976). Overlapping blocks from
the two sources for each environmental variable were
used to produce regression equations. Previous data
for blocks south of 5°S were converted using these
regression equations. Overall, agreement of data for
overlapping blocks from the two sources was relative-
ly good. Correlations for sea surface temperatures
were: January, 0.956; July, 0.951; annual variation,
0.929. Thermocline depth in winter had a correlation
of 0.840, while that for summer values was lower
(0.767). All correlations were statistically significant
(P<0.001), and the associations of values from the two
sources were basically linear.
1 00
I 14
1 Condyiobasal L.
2 L Roslfum(frm Base)
30 L Ramus
3 L Rosl(um(tfm Plerygoid)
22 L Up Toothrow
27 L- Low Toolhrow
23 No Teeth{Up Lf )
24 No Teelh(UpRl)
J— 25 No Teelh(LowL()
1—26 No Teelh(LowRI )
— 19 Oibital L
4 W Roslfum(al Base)
5 W Rosltum{al 1/4 L)
6 W Rostrum(al 1/2L)
9 PreofbitalW.
Poslotbital W-
Skull W (al Zygomatic P }
L Braincase
28 HI Ramus
20 L Antorbilal P.
1 8 W Temporal Fossa
15 Max W Premax.
12 Skull W.(alPafJelals)
13 HI Braincase
21 W Internal Nares
7 W Premax (al 1/2L)
8 W Roslrum{al 3/4 L.)
17 L Temporal Fossa
16 W Exlernal Nares
29 Tooth W.
Figure 2
Correlations among characters based on character means for 25 blocks. Clustering per-
formed using UPGMA on absolute correlations among characters (i.e., negative signs
removed). Cophenetic correlation coefficient is 0.74.
Results
Sexual dimorphism
In the two-way ANOVA for block and sex, only three
measurements showed a significant interaction for
block and sex (W. Rostrum [at Base], L. Temporal
Fossa and No. Teeth [Up.Lf.]). All characters exhibited
significant variation by block (i.e., geographic varia-
tion), and 15 of the 30 characters displayed significant
sexual dimorphism (Table 1). For most characters,
males are larger than females. Character differences
between sexes range up to 6.34% (see Table 1), with
the most dimorphic character being W. Rostrum (at
3/4 L.).
Correlation, ordination and clustering
Figure 2 summarizes associations among characters
based on means for the 25 blocks. Virtually all of the
intercharacter correlations were positive in sign; a few
indicated weakly negative associations. For the cluster
analysis, absolute character correlations were analyzed
(i.e., sign of correlation ignored), because we wanted
to assess simply the degree of covariation. The char-
acter showing the most distinctive pattern relative to
60
Fishery Bulletin 90(1). 1992
Table 3
Principal component loadings
for Stenella longirostris involving character means
for 25 blocks.
Character
Component *
Character
Component *
I
II
I
II
1 Condylobasal L.
0.914
-0.325
16
W. External Nares
0.573
-0.045
2 L. Rostrum (frm.Base)
0.872
-0.397
17
L. Temporal Fossa
0.575
0.040
3 L. Rostrum (frm. Pterygoid)
0.885
-0.396
18
W. Temporal Fossa
0.647
0.599
4 W. Rostrum (at Base)
0.856
0.083
19
Orbital L.
0.782
-0.064
5 W. Rostrum (at 1/4 L.)
0.838
0.242
20
L. Antorbital P.
0.849
0.108
6 W. Rostrum (at 1/2 L.)
0.880
0.248
21
W. Internal Nares
0.631
0.427
7 W. Premax. (at 1/2 L.)
0.578
0.044
22
L. Up. Toothrow
0.864
-0.426
8 W. Rostrum (at 3/4 L.)
0.508
0.504
23
No. Teeth (Up.Lf.)
0.608
-0.680
9 Preorbital W.
0.938
0.255
24
No. Teeth (Up.Rt.)
0.688
-0.593
10 Postorbital W.
0.917
0.344
25
No. Teeth (Low.Lf.)
0.600
-0.666
11 Skull W. (at Zygomatic P.)
0.916
0.359
26
No. Teeth (Low.Rt.)
0.652
-0.630
12 Skull W. (at Parietals)
0.331
0.659
27
L. Low. Toothrow
0.814
-0.467
13 Ht. Braincase
0.625
0.719
28
Ht. Ramus
0.872
0.219
14 L. Braincase
0.881
0.326
29
Tooth W.
0.216
0.752
15 Max. W. Premax.
0.811 0.042
in bold as follows: (co
30
)>|0
L. Ramus
.81; (II)>|0.6|.
0.877
-0.366
* Relatively high loadings highlighted
mponent 1
other morphological characters is
Tooth W. In addition, correla-
tions of L. Temporal Fossa and
W. External Nares with other
characters are relatively low.
The rest of the characters are
placed in two groups. The first
cluster (characters listed between
1 and 19 at top of Fig. 2) includes
lengths involving the anterior
portion of the skull, tooth num-
bers, and Orbital L. The second
group (characters 4 to 8 as listed
in Fig. 2) includes a variety of
skull widths, dimensions of the
braincase, and Ht. Ramus.
Character loadings of a prin-
cipal components analysis using
25 blocks are presented in Table
3. The first component explained
57.0% of the total character vari-
ance and the second 18.5%
(cumulative total of 75.4%). Pro-
jections of all blocks onto these
components are shown in Figure
3, while Figure 4 is a map sum-
marizing geographically the pro-
jections onto the first component.
This component, which reflects
general skull size, has relatively
II III
1.5
_«0T
1.0-
-
i 0.5-
(D
0506
D " "
-
O
Q.
o O.Q-
O
— 0*14 0513
m"'" a""
■ ""
-0.5-
oei!
_06II ■ C40t
■ ■■>■"' D
_09I2
0°'°'
_OT02
g05ie
-
-1.0 -
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0
Component 1
Figure 3
Projections of blocks onto first two principal components based on 30 characters. Solid
symbols indicate 25 blocks on which analysis conducted. Open symbols represent blocks
with only single specimens, which were projected onto axes generated from 25 blocks
with two or more specimens.
Douglas et al : Geographic variation in cranial morphology of Stenella longirostris
61
high correlations (Table 3) with
all characters except Tooth W.
and Skull W. (at Parietals). Local-
ities to the right in Figure 3 are
from the Hawaiian Island area
(0702 and 0802; see Fig. 4), where
animals are larger. Specimens
from southern blocks (e.g., 0116,
0117, 0313) also are larger than
animals from other parts of the
range. Blocks to the left in Fig-
ure 3, with negative loadings on
component I, have snialler indi-
viduals. In general, S. longiros-
tris from the northeastern blocks
were the smallest (e.g., 0515,
0613, 0612).
Component II has its highest
positive correlations with Skull
W. (at Parietals), Ht. Braincase,
and Tooth W. ; it has negative
associations with characters 23-
26, which involve numbers of
teeth. Block 0507 is the most ex-
treme, with a positive projection
on this component (see Fig. 3);
animals from this block have
relatively wide skulls and re-
duced numbers of teeth. In con-
trast, block 0516 is at the other extreme,
with relatively narrow skulls and greater
numbers of teeth.
Figure 5 is a dendrogram depicting
results from a UPGMA cluster analysis
of the 25 geographic blocks. Four main
clusters are evident, with block 0702
being the most divergent and in its own
cluster. Block 0507 also is in a cluster by
itself. The first group in the diagram (i.e.,
blocks 0116 through 0516 at top of Fig.
5) includes predominantly southern and
western localities. Those in the largest
cluster (listed from block 0411 to 0509 in
Fig. 5) are situated to the north and/or
east.
A modified skyline diagram (Fig. 6A)
resulting from function-point clustering
for 25 blocks based on 30 characters in-
dicates an initial separation of block 0702
(which includes part of Hawaii) from the
others. At a w-value of 3.50, there are
three clusters: (1) block 0702; (2) the
southern blocks in addition to blocks 0508 and 0802;
and (3) the remaining northern and eastern blocks, in-
cluding those just north of the Equator. Further sub-
Figure 4
Geographic variation in principal component I. Bar represents range (low to high block
values), and midpoint is marked. Darkened part of bar indicates value for particular block.
00
I
^
<s
^
- 0116
-0117
- 0216
-0215
-0313
- 0508
- 0315
- 0002
-0516
-0411
- 0615
-0312
-0505
-0612
-0614
-0613
- 0417
-0512
-0514
-0513
-0506
-0505
-0509
-0507
-0702
Figure 5
Distance phenogram summarizing UPGMA clustering of 25 blocks based on
30 characters. Cophenetic correlation is 0.80.
division results with smaller w'-values (see Fig. 6A).
A similar analysis (Fig. 6B) was conducted using
the five characters— Postorbital W., L. Rostrum (frm.
62
Fishery Bulletin 90(1). 1992
Pterygoid), W. Internal Nares, W. Premax. (at 1/2 L.),
and W. Rostrum (at Base)— that, in combination, were
best for discriminating among blocks (based on canon-
ical variates analysis reported below). With a w -value
of 1.79, block 0702 is separated from the remaining
blocks. Note that four groups were formed when using
a 1.47 w-value; there are two single-block groups (i.e.,
0702 and 0802). When the w -value was lowered to 1.37,
the same groups were formed, except that 0315 was
in its own group and 0802 joined with a group of
predominantly southern localities. With a 1.26 M^-value,
the clusters are the same except that block 0505 joins
the northeastern blocks instead of those from the south.
Three groups were formed with a 1.15 w-value: (1)
block 0702; (2) a group of eight blocks, including south-
ern blocks in addition to 0508, 0505, and 0802; and (3)
the northern and eastern blocks, including 0411 and
0417, as well as 0506, 0507, and 0509. At smaller
w -values, there is further subdivision.
A canonical variates analysis, using as initial data the
information on all 30 measurements for 25 blocks, in-
corporated the five characters listed in Table 4. A two-
dimensional plot of the 25 block centroids on canonical
variables 1 and 2 is included as Figure 7; although not
used to generate the axes, 10 blocks with only single
specimens also are projected onto these variables. The
geographic pattern of canonical variable 1 is depicted
in Figure 8B. The eigenvalue for canonical variable 1
is 2.93 and that for the second is 0.55, with the two
summarizing 82.0% of the variance for the five char-
acters. These five characters in combination show the
Figure 6
Modified skyline diagrams for 25 blocks, indicating
groups formed using function-point clustering pro-
cedures and based on: (A) all 30 characters; (B) five
characters that, in combination, best discriminate
among blocks (Postorbital W, L. Rostrum [frm.
Pterygoid], W Internal Nares, W. Premax. [at 1/2
L.], and W Rostrum [at Base]). For given w-va.\ue
(i.e., row), blocks connected in common line are in
same cluster.
Table 4
Canonical variates analysis of all specimens from 25 blocks.
Character
F- value to
enter
Order of
entry
Coefficients*
1
3
4
7
10
21
L. Rostrum (frm. Pterygoid)
W Rostrum (at Base)
W Premax. (at 1/2 L.)
Postorbital W
W. Internal Nares
Constant
4.53
2.38
2.30
23.21
2.88
0.0242 (0.2830)
-0.0603 (-0.1622)
-0.0291 (-0.0393)
0.2113 (0.8570)
0.0958 (0.1870)
38.9175
* Unstandardized coefficients, with standardized values in parentheses, for canonical variates.
-0.0662 (-0.7734)
-0.1912 (-0.5145)
0.2419 (0.3273)
0.0648 (0.2628)
0.3324 (0.6489)
4.4931
Douglas et al : Geographic variation in cranial morphology of Stenella longirostns
63
greatest among-group variability relative to that
within groups and are used for more detailed com-
parisons with environmental variables (presented
below).
As indicated in Table 4, the first canonical
variable is most influenced by Postorbital W. (Fig.
8A). In Figure 7, blocks that are large for this
character are to the right, while those that are
small are to the left. When considering only those
blocks with more than one specimen (i.e., those
shown with solid symbols in Fig. 7), the two blocks
from the vicinity of the the Hawaiian Islands
(0802 and 0702) are to the right, as are blocks
predominantly from the southern portion of the
range. The blocks with single specimens (which
tend to be more westerly) also are to the right.
Specimens from blocks to the north and east are
smaller; they are depicted to the left in Figure
7. Some west-central blocks group with the
southern blocks, while others are intermediate or
group with those to the northeast. The second
canonical variable contrasts blocks from the
Hawaiian Island area (0702, 0802) with the others
(see Fig. 7); in the characters reflected by this
variable, values of block 0812 (which is northern,
but to the east) show some similarities to those
for 0702 and 0802.
Mantel test for geographic patterning
Individual characters were evaluated with respect
to geographic patterning using Mantel tests, as
well as matrix correlations that compare inter-
block geographic distances (or reciprocals of these
distances) and character differences between localities.
Of the 30 characters, 66.7% (20) show statistically
significant regional patterning indicating that geo-
graphic distances (in nautical miles) and interblock
character differences are interrelated (Table 5). For
measures showing significant f -values the greatest
character differences tend to be between blocks that
are farthest away from each other, while nearer
localities are more similar.
Local patterning, as indicated by a significant nega-
tive association of distance reciprocals and character
differences, was found in 83.3% (25) of the characters
(Table 5). All characters that showed regional pattern-
ing also exhibited local patterning.
Principal component projections also were assessed
in terms of geographic patterning. As indicated at the
bottom of Table 5, component I (Fig. 4) has strong
regional and local patterning; component II has signifi-
cant local patterning. Canonical variables 1 (Fig. 8B)
and 2 both exhibit marked regional and local pattern-
ing (Table 5).
o _
3
2-
„QJ1I
■
CSJ 1 "
0506
0515
*^ l-lmis OUT
■S 0-
■ 05.JJ1 °"'b
.
•^
> -1-
■ ""'
-
(0
o
b""
qKO.
^0803
1 -2-
■
-
c
to
O
-3 ■
-4-
■ ""=
-
-5-
\ 1 1 1 r 1 T
-3-2-1012345
Canonical Variable 1
Figure 7
Projections of blocks onto first two canonical variables based on 30
characters. Solid symbols indicate 25 blocks on which analysis con-
ducted. Open symbols represent blocks with only single specimens,
which were projected onto axes generated from 25 blocks with two
or more specimens.
Morphological-environmental covariation
Figure 9 is a dendrogram indicating absolute correla-
tions among the 13 environmental variables, sub-
dividing them into five clusters. Sea Current (N.,
Winter) is in a group by itself and quite different from
the others. Sea Current (W, Winter) and Oxygen Min.
Layer (Depth) are in the second cluster, which joins
with a group of five variables involving surface mea-
sures of temperature, oxygen, and salinity. The fourth
cluster involves two measures of solar insolation, and
the fifth reflects aspects of water depth.
The loadings of environmental variables on the first
three environmental principal components are given
in Table 6. The first component statistically explains
33.0% of the total character variance, the second
23.2%, and the third 15.8% (cumulatively 72.0%). Maps
(Fig. 10) depict projections of the 25 blocks with two
or more 5. longirostris onto the first two environmen-
tal components. Environmental component I has rela-
tively high values for blocks between 5° and 15 °N, with
intermediate values to the north and low values south
of the Equator (Fig. lOA). Sea Surface Temp. (July)
64
Fishery Bulletin 90|l). 1992
09
1
a'
'
^'
- ^
1
r
1
—
\
""^
5
\]
08
1
^
0
j
4^
^M-.
^
07
1
\
H
J.
L
"\
Ci
05
0
0
A
/
^
05
e
B
e
1
t
y
Ij
n
■
DU
^^~L
OJ
L'
fl
C3
A
Posiofbitai w
1
1
)
02
II
\
01
■
1
■15 9
i
1
01 03 05 07 03 n 13 15
09
1
iO°
'
0- ^
1
("
"■^
I —
yr
08
1
^
if]
!
-7
^
V
a:
07
1"
\
V
k
J.
/
~\
Ci
06
0
0
^
s
A
•
^
05
B
■
u
e
■^
t
\'
t
B
\
y
/
04
^
03
B
Canonical Variable 1
n::
1
tf
I
)
02
1
1
K
\
0.
■
J
2.12
1
1
Figure 8
Geographic variation in (A) Postorbital W. and (B) canonical variable 1. Darkened part
of bar indicates value for particular block.
has a high positive loading on component I, while that
for Sea Surface Temp. (Ann. Var.) is negative. Five
other variables have relatively high correlations with
this component (Table 6). The second environmental
component has high values for
the two blocks adjacent to the
Hawaiian Islands (Fig. lOB), with
intermediate values in other
western blocks. Strong negative
projections on this component
are found for blocks along the
coast of South and Central
America just north of the Equa-
tor. The most substantial load-
ings on this component are for
the two thermocline variables
(Table 6), while Water Depth and
Surface Salinity also exhibit
relatively high positive projec-
tions for component II. Envi-
ronmental component III reflects
mainly Solar Insolation (Annual),
with Solar Insolation (Jan.) also
having a relatively high positive
loading (Table 6). The most
extreme negative projection for
component III is for the north-
ern block near the coast (i.e.,
0812), with the highest positive
projects for blocks in the west-
central portion of the study area
(i.e., 0505 through 0509). In
general, other blocks have rela-
tively high projection values,
except for 0802 (which is some-
what lower). Other components
beyond the first three tended to
represent only single environ-
mental variables.
Several of the environmental
measures showed few or no sta-
tistical associations with mor-
phological characters (and result-
ing principal components or
canonical variables), while others
exhibited significant covariation
(Table 7). The first environmen-
tal variable. Sea Current (N.,
Winter), is not significantly cor-
related with any of the 30 mor-
phological measures. The other
character summarizing sea-cur-
rent information. Sea Current
(W, Winter), has a geographic
pattern showing relatively weak
statistical concordance with eight of the morpho-
logical characters. Toothrow lengths and three of the
four tooth counts are among those with significant
associations.
Douglas et al : Geographic variation in cranial morphology of Stenella longirostns
65
Table 5
Association of interlocality character differences
with geographic distances (in nautical
miles) and the reciprocals
of these
distances. Results from Mantel tests (t) and matrix correlations (r) for Stenella longirostris.
Character
Distance
Reciprocal of distance
t r
(
r
1 Condylobasal L.
4.49*'*
0.445
-4.86***
-0.348
2 L. Rostrum (frm.Base)
3.89'**
0.395
-4.35***
-0.315
3 L. Rostrum (frm. Pterygoid)
4.26**'
0.437
-4.67***
-0.340
4 W. Rostrum (at Base)
3.49***
0.485
-3.86***
-0.316
5 W. Rostrum (at 1/4 L.)
3.16**
0.327
-3.48***
-0.255
6 W. Rostrum (at 1/2 L.)
3.12**
0.297
-3.68***
-0.258
7 W. Premax. (at 1/2 L.)
-0.48
-0.046
0.23
0.017
8 W. Rostrum (at 3/4 L.)
0.33
0.028
-1.98*
-0.132
9 Preorbital W
4.59*"
0.403
-6.36***
-0.429
10 Postorbital W
3.57***
0.262
-5.97***
-0.373
11 Skull W (at Zygomatic P.)
3.43***
0.255
-5.98***
-0.357
12 Skull W (at Parietals)
1.83
0.199
-2.19*
-0.165
13 Ht. Braincase
1.51
0.146
-3.89***
-0.275
14 L. Braincase
3.62***
0.323
-4.85***
-0.329
15 Max. W Premax.
2.05*
0.209
-3.58***
-0.260
16 W External Nares
-0.48
-0.047
-0.38
-0.027
17 L. Temporal Fossa
1.87
0.193
-2.74**
-0.201
18 W Temporal Fossa
3.84**'
0.362
-4.69***
-0.327
19 Orbital L.
3.01**
0.325
-3.68***
-0.276
20 L. Antorbital P.
2.93*'
0.229
-4.32***
-0.277
21 W Internal Nares
0.56
0.047
-3.43***
-0.226
22 L. Up. Toothrow
3.96***
0.392
-4.47***
-0.319
23 No. Teeth (Up.Lf.)
3.46***
0.411
-3.61***
-0.286
24 No. Teeth (Up.Rt.)
3.46***
0.434
-3.30***
-0.271
25 No. Teeth (Low.Lf.)
1.62
0.178
-1.38
-0.105
26 No. Teeth (Low.Rt.)
1.52
0.158
-1.28
-0.093
27 L. Low. Toothrow
3.00**
0.285
-3.46***
-0.242
28 Ht. Ramus
3.03**
0.237
-4.77***
-0.306
29 Tooth W
-0.55
-0.067
-0.13
-0.011
30 L. Ramus
3.91***
0.375
-4.47***
-0.314
Component I
3.77***
0.363
-4.64***
-0.327
Component II
1.13
0.134
-2.72**
-0.214
Canonical Variable 1
2.77**
0.220
-5.08***
-0.327
Canonical Variable 2
4.05***
0.520
-4.12***
-0.343
*P<0.05; •♦P<0.01: ***P<0.001.
Correlation
050
1-00
- 1 Sea Current (N . Winler)
- 2 Sea Current {W . Winter)
- 9 Oxygen Min Layer (Depth)
- 6 Sea Surface Temp (Jan )
- 7 Sea Surlace Temp (July)
- 8 Sea Surface Temp (Ann Var )
-10 Surface Salinity
-13 Surface Dissolved Oxygen
- 4 Solar Insolation (Jan )
- 5 Solar Insolation (Annual)
- 3 Water Depth
-11 Thermocline Depth (Winler)
-12 Thermocline Depth (Summer)
Figure 9
Clustering by UPGMA on absolute correlation values among environmental variables.
Cophenetic correlation of 0.75.
66
Fishery Bulletin 90(1). 1992
Table 6
Principal component loadings for environmental variables.
Environmental variable
Component *
I
11
Ill
1 Sea Current (N. .Winter)
-0.126
0.307
0.431
2 Sea Current (W., Winter)
-0.495
-0.037
-0.091
3 Water Depth
-0.279
0.783
0.279
4 Solar Insolation (Jan.)
-0.683
-0.227
0.627
5 Solar Insolation (Ann.)
-0.274
-0.291
0.872
6 Sea Surface Temp. (Jan.)
0.768
-0.255
0.380
7 Sea Surface Temp. (July)
0.942
0.014
-0.101
8 Sea Surface Temp. (Ann.Var.)
-0.848
-0.224
-0.304
9 Oxygen Minimum Layer (Depth)
0.675
0.442
0.157
10 Surface Salinity
-0.608
0.600
-0.046
11 Thermocline Depth (Winter)
0.172
0.888
-0.269
12 Thermocline Depth (Summer)
-0.044
0.836
0.363
13 Surface Dissolved Oxygen
bold
-0.596
as follows: (component I) > |0.8|
0.089
(II and III) > |0.6|.
-0.380
* Relatively high loadings highlighted in
Table 7
Product-moment correlations of block
means for
morphological variables and components versus
environmental variab
es and com-
ponents based on 25 blocks of Stenella longirostris.^
Environmental
Environmental variable''
component
Character 1
2
3
4
5 6 7 8
9 10
11
12
13
I
II III
1 Condylobasal L.
— —
+ +
-
2 L. Rostrum (frm.Base)
— - +
+
+
-
3 L. Rostrum (frm. Pterygoid)
+
+
-
+ +
-
-
4 W Rostrum (at Base)
+ +
- -
+
+ + +
+ +
+ + +
5 W Rostrum (at 1/4 L.)
+ +
-
+
+ +
+ +
+ + +
6 W Rostrum (at 1/2 L.)
+ +
-
+
+
+ +
+ +
7 W Premax. (at 1/2 L.)
+
+ +
8 W Rostrum (at 3/4 L.)
-
+ +
+ +
+
9 Preorbital W
+ +
+ + +
+
+ +
+
—
+ +
10 Postorbital W
+ +
—
+ + +
+
+ +
+
—
+ +
11 Skull W (at Zygomatic P.)
+ +
—
+ + +
+
+ +
+
+ +
12 Skull W (at Parietals)
+
13 Ht. Braincase
+ +
—
+ + +
+ +
-
+
14 L. Braincase
+ + +
+ + +
+
+ +
+
15 Max. W Premax.
_ —
+
+
16 W External Nares
_ _
+ +
+ + +
+
-
17 L. Temporal Fossa
-
+
-
18 W Temporal Fossa
+ + +
-
+ +
+ +
+ + +
19 Orbital L.
+
— _
+
-
20 L. Antorbital P.
+
—
+ +
+
—
21 W Internal Nares
+
+ +
—
22 L. Up. Toothrow
+
_ __ _ +
-
+ +
-
-
23 No. Teeth (Up.Lf.)
+
—
-
24 No. Teeth (Up.Rt.)
+
—
-
25 No. Teeth (Low.Lf.)
26 No. Teeth (Low.Rt.)
+
27 L. Low. Toothrow
+
— - +
-
+
-
+
28 Ht. Ramus
+ +
_ —
+
+
+
+
-
+
29 Tooth W
+
30 L. Ramus
— - +
-
+ +
-
Component I
+
—
+ +
+
+
+
—
+
Component II
—
+ +
+
+
Canonical Variable 1
+ +
+
- + +
+
+ +
+
Canonical Variable 2
-
+ +
+ +
+ +
Douglas et al.: Geographic variation in cranial morphology of Stenella longiroscrls
67
Table 7 (continued)
^Blanks indicate nonsignificant correlations. Individual symbols refer to significant positive or negative correlations (P<0.05; >0.396);
double symbols indicate highly significant correlations (P<0.01; >0.505); and triple symbols represent very highly significant correla-
tions (P<0.001; >0.620).
'■Environmental variables: (1) Sea Current (N., Winter); (2) Sea Current (W., Winter); (3) Water Depth; (4) Solar Insolation (Jan.);
(5) Solar Insolation (Annual); (6) Sea Surface Temp. (Jan.); (7) Sea Surface Temp. (July); (8) Sea Surface Temp. (Ann. Var.); (9) Oxy-
gen Min. Layer (Depth); (10) Surface Salinity; (11) Thermocline Depth (Winter); (12) Thermocline Depth (Summer); and (13) Surface
Dissolved Oxygen.
Environmental
Principal
Component I
I
17 19
09
>6
>'
T
°
^
r
10
y
"^
^
08
1
^
[
7-
^
^
07
1
^
M
:^
D
C:>
06
u
■
■
P
{_
10"-
05
1
1
1
B
B
u
0
L
iY
/
04
■
— ,
03
B
Envifonmenlal
Principal
ComponenI II
u
-(
?
02
U
t
\
A
10=-
01
M
1 .
-0 80
■
B
1
Figure 10
Geographic variation in environmental vari-
ables as summarized in (A) principal compo-
nent I and (B) principal component II.
Darkened part of bar indicates value for par-
ticular block.
Water Depth (variable 3; Fig.
IIB) is positively correlated with
13 morphological measures, two
of which (L. Braincase and W.
Temporal Fossa) are very highly
significant (P< 0.001). The block
values for W. Temporal Fossa
(which have a 0.755 correlation
with Water Depth values) are
shown in Figure 11 A. For the 13
variables, relatively large values
typically were recorded in block
0117 and those in the vicinity of
the Hawaiian Islands (0702 and
0802), all of which have relative-
ly deep waters, while more shal-
low localities like 0516 and 0812
had individuals that were smaller
for these characters.
The fourth environmental
measure, Solar Insolation (Jan.),
changes from high to low values
uniformly from north to south. It
is statistically associated with
only one character, W. Internal
Nares, which has small values for
0812; values tend to get higher
as one proceeds south, but there
are exceptions (such as 0802,
which is relatively high). Canon-
ical variable 2 has a pattern
statistically similar to this envi-
ronmental variable (bottom of
Table 7).
68
Fishery Bulletin 90(1). 1992
Annual solar insolation (variable 5) has high values
at the earth's Equator, with decreasing values as one
proceeds toward either pole. Only two morphological
variables have significant correlations with this envi-
W Tempoial Fossa
144 0 mm
394
349
1 '
09
11
0°
0'
i;
•• :i
\V
1C
rj
V
-' "Vi
I —
H
oa
1
'^
f
7
^
V
07
1
\
J
•^
I
N,
Cb
06
B
u
B
B~
\
J
A-
05
1
1
1
1
1
B
u
■
■
04
■
03
B
Water Depth
B
•V.
u
{
)
02
1
fl
\
K
01
■
2500
J
r
1
1
Figure 1 1
Geographic variation in (A) W. Temporal Fossa and (B) Water Depth. Darkened part
of bar indicates value for particular block.
ronmental factor (Table 7), each of which are negative
and relatively weak.
The sixth environmental variable. Sea Surface Temp.
(Jan.), has significant negative correlations with 22
morphological characters, as well
as principal component I (Fig. 4)
and canonical variable 1 (Fig.
8B). Variable 7, which is Sea Sur-
face Temp. (July) (Fig. 12), has a
relatively high number (21) of
significant negative associations
i7^:;>^ with morphological measures, as
— LX^ v,'\ well as with principal component
I (Fig. 4) and canonical variable
1 (Fig. 8B). Postorbital W. (Fig.
8A) has the strongest correlation
(-0.681) of any of the morpho-
logical characters with Sea Sur-
face Temp. (July).
The eighth environmental vari-
able. Sea Surface Temp. (Ann.
Var.), exhibited relatively weak
geographic concordance with six
morphological characters, one of
which was a negative association
(Table 7). Also, only weak nega-
tive correlations of five morpho-
logical variables were found with
depth of the oxygen minimum
layer (variable 9).
Fourteen of the 30 morpholog-
ical measures are significantly
correlated with environmental
variable 10, Surface Salinity (Fig.
13B). In addition to east-west
changes from lower to higher
values at a given latitude, salin-
ity also exhibits a north-to-south
trend of increasing values (below
15 °N). The highest correlation
(0.661) is with L. Braincase (Fig.
ISA).
Thermocline Depth (Winter),
variable 11, was positively asso-
ciated with 13 morphological
variables (Table 8), while Thermo-
cline Depth (Summer), variable
12, has statistically significant
positive correlations with 11
morphological traits. The final
variable, Surface Dissolved Oxy-
gen, covaries with 16 morpholog-
ical variables. As suggested in
the dendrogram in Figure 9, this
environmental variable has a
Douglas et aL: Geographic variation in cranial morphology of Stenella longirostns
69
09
08
07
C6
05
04
03
02
01
1
<J°
I
lO"
'■
1
^
K
.J
- "V
\
1
"^
i)
[
/
-7
4^
X
i
\
/
"\
Ci
1
1
1
1
A
j
^
1
1
1
1
1
1
1
1
II
^0.
B
1,
Sea Surface
Temp (July)
.S7S0C
y-24 10
™ 19 SS
1 1 1
■^
■V.
)
Q
^
{
\
10^-
y
01 03 05 07 C9 11 13 15 17 13
Figure 12
Geographic variation in Sea Surface Temp. (Jul.). Darl<ened part of bar indicates value
for particular block.
pattern with similarities to those
for sea surface temperatures.
Table 7 indicates that the pat-
tern of correlations of environ-
mental principal component I
with cranial measures is similar
to that of Sea Surface Temp.
(July) (variable 7), which is ex-
pected given the strong loadings
of this environmental variable on
this component (see Table 6). En-
vironmental principal component
11 has positive correlations with
most skull-width measures (Table
7). The third environmental com-
ponent has relatively few signifi-
cant correlations with morpho-
logical characters; its strongest
association is with one of the
summary morphological vari-
ables, canonical variable 2.
In Table 8, we have summar-
ized Mantel i -values and matrix
correlations between selected
Table 8
Results of Mantel tests (t ) and matrix correlations (r ) for Stenella kmffirostris. Comparison of interiocality differences for 13 environmental
variables and 3 environmental components against those for 5 morphological variables selected in canonical variates analysis.
Environmental variable
Postorbital W.
(
10
11
12
13
Sea Current (N., Winter)
Sea Current (W., Winter)
Water Depth
Solar Insolation (Jan.)
Solar Insolation (Ann.)
Sea Surface Temp. (Jan.)
Sea Surface Temp. (July)
Sea Surface Temp.
(Ann.Var.)
Oxygen Minimum Layer
(Depth)
Surface Salinity
Thermocline Depth
(Winter)
Thermocline Depth
(Summer)
Surface Dissolved Oxygen
Environmental Component I
Environmental Component II
Environmental Component III
-0.56
-1.38
3.23*
4.46*
2.14*
5.47*
4.82*
1.64
1.22
4.57*
2.92*
2.52*
2.52*
3.63*
2.75*
1.39
L. Rostrum
(frm.Pterygoid)
t r
W Internal
Nares
t r
W Premax.
(at 1/2 L.)
t r
W Rostrum
(at Base)
t r
-0.050
-0.092
0.235
0.339
0.176
0.371
0.353
0.114
0.077
0.323
0.241
0.172
0.204
0.268
0.212
0.120
-0.59
0.12
1.27
2.60"
1.95
3.71**
1.73
1.22
2.12*
0.37
4.42*'
1.63
1.43
1.68
3.85*'
1.19
-0.083
0.010
0.128
0.283
0.241
0.327
0.176
0.113
0.158
0.035
0.548
0.146
0.172
0.174
0.429
0.160
1.99*
-0.72
2.25*
4.49***
1.86
3.63***
4.65***
3.80***
3.54***
1.03
-0.96
-0.15
1.88
4.09***
-1.11
1.99*
0.215
-0.053
0.186
0.393
0.180
0.271
0.387
0.236
0.082
-0.56
0.59
-0.76
0.77
-0.54
0.53
2..55*
-0.073
0.048
-0.073
0.079
-0.062
0.045
0.244
-0.79
-0.44
2.22*
1.81
1.43
1.63
0.64
0.294 1.40
0.123 -0.40
1.62
0.06
0.116
0.005
-0.61
0.62
-0.093 -0.55 -0.063 3.87***
-0.001
0.179
0.345
-0.099
0.207
-0.35
0.93
2.02*
-0.75
-0.97
-0.030
0.105
0.197
-0.078
-0.121
3.41***
-0.18
0.11
3.70***
0.32
-0.141
-0.044
0.271
0.241
0.219
0.170
0.079
-0.044
-0.051
0.072
0.598
0.362
-0.027
0.014
0.508
0.054
*P<0.05; •*P<0.01; •**F<0.001.
70
Fishery Bulletin 90(1). 1992
09
08
07
06
05
04
03
02
01
:0'
'
0°
'
.' ^
^^
1
"r
—
V
■^■"vt
\
1
\
0
-7
^
X
1
-N
o.
1
1
■
A
]
-f
L
i
1
e
■
t
^
u
i
L
t
/
A
L. Braincase
1 1 106 7 mm
^-101 5
" 963
1 < '
1
e^
1
le
{
w
1
1
01 03 05 07 09 11 13 i= 17 19
09
03
07
06
05
04
03
02
01
16
^
o"
12
'• .
1
I 1
10
Y
-'"V
\
1
'\
*1
V
f
-7'
^
V-
tf
\
X
J
I
^
Ci
ij
D
■
u
\
A
1
A-
e
1
1
a
e
Q
■
■
■
^
i
1
1
c
B
Surface Salinity
3545-.
y-34 15
32 84
I 1 1
1
i
[
1
1
\
\
1
1
Gee
ind
01 03 OS 07 09 11 13 ;5 17 19
Figure 1 3
)graphic variation in (A) L. Braincase and (B) Surface Salinity. Darkened part of bar
cates value for particular block.
morphological and environmental variables. Assessing
difference matrices using these techniques represents
an alternate method with which to evaluate covaria-
tion of geographic patterns. For Postorbital W. (Fig.
8A), there are nine significant associations using the
Mantel test (Table 8), with Sea
Surface Temp. (Jan.) being the
highest. The seven environmen-
tal variables displaying correla-
tions in Table 7 with Postorbital
W. also are judged concordant
using the Mantel test. In addi-
tion, based on interblock differ-
ence values, there are statistical-
ly significant associations with
the two measures of solar insola-
tion (variables 4 and 5; see Table
8). The concordance with these
two environmental variables is
primarily on the strength of pat-
tern similarities in the eastern
portion of the range. Postorbital
W. also shows significant associa-
tions with the first two environ-
mental principal components.
Based on correlation tests for
block means (Table 7), L. Ros-
trum (frm. Pterygoid) exhibited a
geographic distribution of mean
block values that was statistical-
ly associated with those for six
environmental measures. Four
significant associations were
identified using the Mantel test
(Table 8), only two of which were
found by both tests (Sea Surface
Temp. [Jan.] and Oxygen Mini-
mum Layer [Depth]). It has a
significant association with en-
vironmental component II.
In Table 8, a total of seven
significant associations of differ-
ence values are recorded for W.
Internal Nares with environmen-
tal variables, including the three
listed in Table 7 as having sta-
tistically significant associations
based on means. Difference
values for environmental com-
ponents I and III significantly
covary with those of W. Internal
Nares.
Using difference values, W.
Premax. (at 1/2 L.) has only a
weak correlation with a single
environmental variable (Table 8); only a single signifi-
cant association was found using correlations of mean
values, and this was with another environmental
variable (see Table 7). W. Premax. (at 1/2 L.) is weak-
ly associated with environmental component I.
Douglas et al : Geographic variation in cranial morphology of Stenella longirostns
71
The W. Rostrum (at Base) exhibits covariation with
three of the environmental variables based on the
Mantel test (Table 8), two of which involve thermocline
depth. The matrix correlation (Table 8) for W. Rostrum
(at Base) with Thermocline Depth (Winter) is substan-
tial (0.598), as it is with environmental component II
(0.508). Mean values also showed an association of W.
Rostrum (at Base) with the thermocline variables and
Water Depth, as well as with three other environmen-
tal variables (Table 7).
Discussion
Sexual dimorphism
Our analysis of sexual dimorphism extends the studies
of Douglas et al. (1986) in terms of additional specimens
and minor adjustments in previously collected data. We
found 15 of 30 variables were statistically dimorphic.
Douglas et al. (1986) identified 13 of these as showing
sexual differences, the increase of two characters
(Postorbital W. and Tooth W) being due primarily to
increased numbers of specimens available. In five
characters, including two rostral and two toothrow
lengths, measurements for females were slightly larger
than for males, although the differences were not
statistically significant. For all other measures, males
are larger than females, including all 15 where statis-
tical significance was found.
Geographic variation
In 1889, when the existence of spinner dolphins in the
eastern tropical Pacific was not yet known. True in-
dicated that the absence of adequate samples made
very difficult the task of taxonomically evaluating
species in the genus Stenella. By the 1970s, Perrin
(1975b) had available considerably more material for
a monographic treatment of S. attenuata and S. longi-
rostris from the eastern tropical Pacific and was able
to make significant advances with respect to our
understanding of morphological variation of Stenella.
However, his work on S. longirostris also was hindered
by the paucity of skeletal material from parts of the
range. For our study, many additional specimens of
S. longirostris were available. On the whole, our results
are strongly supportive of those obtained by Perrin
(1975b) for cranial characteristics, but we also have
been able to substantially extend his analyses.
Perrin (1972) conducted an initial analysis of geo-
graphic variation in color patterns of S. longirostris
in the eastern Pacific Ocean. He found geographic
variation, particularly in the "dorsal field system" of
coloration, which overlies a basic general pattern.
These and other data suggested differentiation into
eastern, whitebelly, and Hawaiian forms. The differ-
ences were analyzed further by Perrin (1975a,b), who
indicated that the whitebelly form was in some ways
similar to the Hawaiian form, but had a proportionately
smaller beak. Perrin (1975b) described but did not name
four races— the three mentioned above plus a Costa
Rican form which occurs off the coast of Central
America. Perrin et al. (1979b) evaluated possible dif-
ferentiation in S. longirostris involving animals found
south of the Equator in the eastern Pacific Ocean. They
concluded that these S. longirostris are morphological-
ly distinct from those to the northeast; characteristics
showing such a trend include coloration, size, shape,
and skeletal measures.
Recently, Perrin (1990) named and described three
subspecies: S. I. longirostris (Gray's spinner dolphin),
S. I. orientalis (eastern spinner dolphin), and S. I.
centroamericana (Central American spinner dolphin).
In our geographic variation assessment, we purposely
included all adult specimens available, without an
attempt a priori to differentiate previously described
forms. However, the differences among the named
forms are reflected in our results for cranial char-
acteristics.
In our analyses, the Central American spinner
dolphin of Perrin (1990) is shown to be different from
other S. longirostris by the positioning of block 0516
on principal component II; it had the lowest value of
any of the blocks (see Fig. 3). Three of the four speci-
mens in block 0516 exhibit characteristics of Central
American spinners (as does one of the nine from 0615).
Character associations with the second principal com-
ponent—summarized in Table 3— suggest that, after
taking into account general size (summarized in and
mathematically removed by component I), animals of
this subspecies have relatively longer toothrows,
greater numbers of teeth, a narrower skull at the
parietals, and a shallower braincase than S. longirostris
from other areas. Perrin (1990: table 2) provided com-
parative measurements and counts for S. I. longiros-
tris and S. I. centroamericana, which show the Cen-
tral American form to have longer toothrows and
greater numbers of teeth. He did not include data on
Ht. Braincase, but did characterize the Central Ameri-
can form as having a relatively long and narrow skull.
While Skull W. (at Parietals) is slightly greater for S. I.
centroamericana than S. I. longirostris (Perrin 1990:
table 2), the former has a relatively narrow skull given
its considerably greater length. Additional S. I. centro-
americana specimens are needed, since the diagnosis
and understanding of cranial variation in this sub-
species still is based on very few animals.
Perrin (1990) suggested the existence of a zone of
hybridization/intergradation between S. I. longirostris
12
Fishery Bulletin 90(1). 1992
and S. I. orientalis that may be about 2000km wide.
Stenella I. orientalis is found primarily in the north-
eastern blocks we assessed. Our analyses confirm that,
in general, adult spinner dolphins from this region are
smaller than those from areas to the south, southwest,
and west (for general trends, refer to block projections
onto canonical variable 1 in Fig. 8B). The nominate
subspecies, S. I. longirostris, of Perrin (1990) subsumes
a series of broadly distributed populations. He indicated
that S. I. longirostris likely includes areal entities (out-
side the eastern Pacific) worthy of formal taxonomic
recognition, but to date these have not been evaluated
properly because of a paucity of specimens from major
portions of the range.
Extensive data on geographic variation in external
morphology of S. longirostris in the eastern tropical
Pacific were assessed by Perrin et al. (1991). They
evaluated color patterns, dorsal-fin shapes, and total
lengths for 5. longirostris from throughout the geo-
graphic range covered in our study. Some external
characters (e.g., ventral field coloration pattern) ex-
hibited a "radial" or concentric pattern of variation,
where spinners to the south, southwest and west were
similar, but markedly different from those to the north-
east. This pattern also was prevalent among cranial
variables (e.g., see values for Postorbital W. [Fig. 8A],
canonical variable 1 [Fig. 8B], and L. Braincase [Fig.
13A]).
We were able to incorporate specimens into our
analyses from the general vicinity of the Hawaiian
Islands. Perrin (1975b) evaluated Hawaiian specimens
for cranial features and concluded that, in general, they
were strikingly larger than other spinners. However,
at the time, few specimens were available from south-
ern localities. When these southern blocks are incor-
porated into the analysis, the Hawaiian specimens are
not quite as extreme, although for most characters the
Hawaiian specimens remain the largest (see Postorbital
W. [Fig. 8A], W Temporal Fossa [Fig. 11 A] and L.
Braincase [Fig. 13A], as well as principal component
I [Figs. 3 and 4] and canonical variable 1 [Figs. 7 and
8B]). Also, for many of the characters the Hawaiian
specimens are more similar to far-southern ones than
to those from geographically closer western blocks
located between 5° and 10°N. When evaluating other
western single-specimen blocks that are situated closer
to the Hawaiian Islands, some analyses (e.g., canonical
variates analysis; see Fig. 7) indicate that spinners
similar to Hawaiian specimens are present; however,
additional specimens will be needed in order to clarify
the trends in variation in this part of the Pacific.
In some descriptive analyses (e.g., see Figs. 6 and
7), block 0702 to the south of the Hawaiian Islands is
depicted as quite distinct from other blocks, including
the adjacent block to the north (0802). This incongruity
is likely a statistical aberration related to the small
sample size for 0702 (n 2), rather than to a biological
difference. Checking the specimens from the two blocks
indicated that they were taken in relatively close prox-
imity, but were separated because of where the border
between the blocks happened to be located. This ap-
parent anomaly does not detract from the general con-
clusion that the spinners in the vicinity of the Hawaiian
Islands are among the largest found in the overall study
region.
Schnell et al. (1986) conducted an extensive analysis
of geographic variation of offshore S. attenuata, a
similar species that broadly overlaps in range with
S. longirostris. The two species frequently are seen in
mixed schools (Au and Perryman 1985, Reilly 1990).
Reilly (1990) noted that 73% of the S. longirostris
sightings from research vessels also included S. attenu-
ata; 49% of the records of the more common S. attenu-
ata involved schools that also had S. longirostris. While
detailed comparisons evaluating interspecific geo-
graphic covariation in morphology are beyond the scope
of this paper, our preliminary findings indicate that
about one-half of the individual morphological char-
acters show similar geographic patterns for blocks
where both species are represented. However, two
characters involving the temporal fossa (variables 17
and 18) exhibit negative correlations (the length cor-
relation is statistically significant and the width near-
ly so)— localities where the fossa is larger in S. attenu-
ata, it is smaller in S. longirostris. The temporal fossa
reflects the size of muscles involved in the feeding
apparatus. Also, the upper tooth counts show pro-
nounced negative correlations interspecifically.
Opposite trends in the two similar species for these
characters may be an example of ecological character
displacement related to differences in feeding and the
types of food taken by the two species in given localities
(Perrin 1984).
Genetic subdivision, management units,
and implications of cranial variation
Considerable attention has been given to definition of
stock units with a meaningful biological basis that can
be employed to manage iS. longirostris in the eastern
Pacific (Perrin 1975a, b; Perrin et al. 1979b, 1985,
1991). One of the important questions with respect to
the effectiveness or relevance of geographic manage-
ment units is the degree to which the species is
genetically subdivided. Perrin et al. (1991) noted a com-
plex patchwork pattern of geographic variation in ex-
ternal and other characteristics in S. longirostris sug-
gesting "that there is not a large amount of movement
between the various regions." They pointed out that
the complex geographic pattern of variation in the
Douglas et aL: Geographic variation in cranial morphology of Stenella longirostns
Ti
zone of intergradation/hybridization of S. I. orientalis
with S. l. longirostris is consistent with limited data
on movements from tag returns (Perrin et al. 1979a),
which indicate "a home range of a diameter of hun-
dreds rather than thousands of kilometers."
Our findings strongly support these conclusions. All
30 characters studied showed geographic variation,
with two-thirds having demonstrable regional pattern-
ing, and 25 of the 30 showing local patterning. These
patterns emerge even though data are based on
specimens pooled over season and for a number of
years; consistent geographic patterns largely would be
obscured if animals typically moved long distances
within or between years. Clearly, as found by Schnell
et al. (1986) for S. attenuata, in S. longirostris "there
are notable patterns of geographic variation ... in-
dicating that geographic subdivision exists among
populations."
We found concordance of geographic patterns in
S. longirostris for a number of cranial characters as
noted by Perrin et al. (1991) for external characters.
Yet, some patterns are not concordant; in fact, there
is a mosaic of patterns involving different characters
and/or character suites. For example, many of the tooth
counts, toothrow measurements, and rostrum and
ramus lengths show very similar patterns of variation
(as indicated in Fig. 2), while other characters like
Tooth W. have a pattern among blocks that is not close-
ly related to that of any other character. Not surpris-
ingly, a number of skull widths covary. Overall, the
findings for S. longirostris parallel the situation typical-
ly found in other mammals where geographic variation
in morphological characters has been studied. Some
observed patterns may be the consequence of action
by selective forces, while others simply result from and
are maintained because of isolation by distance. The
findings are consistent with S. longirostris being
genetically subdivided, stemming from individual
animals or groups of animals having relatively limited
home ranges.
For management stocks, Perrin et al. (1991) pro-
posed an alternative management scheme where "an
'eastern spinner conservation zone' could be devised
that would offer appropriate and unequivocal protec-
tion to the unique and coherent gene pool of the eastern
subspecies." For instance, a zone bounded on the south
by 10°N and and on the west by 125° W would encom-
pass 84% of the schools that were identified in the field
as being composed of "eastern" spinners, and would
include very few "whitebelly" animals (Perrin et al.
1991). Based on the cranial measures we employed,
spinners from the blocks in this portion of the range
are very similar; the blocks typically were closely linked
in cluster analyses and ordinations. Blocks from most
other parts of the range did not show the same degree
of consistency and concordance. Our data also provide
additional biological justification for establishing a
geographically defined management zone for S. I.
orientalis that, operationally, would be easily under-
stood and more effective for management purposes.
Perrin et al. (1991) also concluded that data on ex-
ternal characters do not support the division of white-
belly spinners into northern and southern stocks for
management purposes. For cranial features, if one con-
siders only eastern blocks, it is possible to achieve a
considerable degree of separation between northern
and southern whitebelly spinners. However, the situa-
tion becomes notably more complex when more wester-
ly blocks are added. For virtually all cranial characters,
the western blocks group with the more southerly
blocks even though they are at the same latitude as
blocks to the east containing northern whitebelly spin-
ners; the only possible exception is W. Internal Nares,
which shows a strong north to south gradient involv-
ing all blocks except for one in the vicinity of the
Hawaiian Islands (i.e., 0802). The addition of cranial
specimens from western locations has provided a more
sophisticated picture of geographic variation of
S. longirostris in the region under study.
Morphological-environmental covariation
Considerable heterogeneity exists in environmental
parameters over the range of S. longirostris in the
eastern tropical Pacific (see examples of environmen-
tal variation in Figs. 11-13). With two circulatory gyres
adjacent to the region, one to the north and the other
to the south, the eastern tropical Pacific has an
easterly-flowing equatorial counter-current from 3° to
10°N latitude, and a number of fronts and conver-
gences (Wyrtki 1966, 1967). These coupled with
latitudinal and other gradients result in substantial
spatial differences in environmental characteristics.
Spotted dolphin/environmental comparisons
Schnell et al. (1986) evaluated covariation in a similar
suite of environmental and cranial morphological
features for offshore S. attenuata in the eastern
tropical Pacific. The S. attenuata investigation was
focused in eastern areas (only 1 of 19 blocks was west
of 115°W). Our analysis of S. longirostris covers con-
siderably more of the ocean, and includes areas around
the Hawaiian Islands, which potentially could have
substantially different marine environments. The
importance of particular environmental variables, of
course, could be quite different when different geo-
graphic levels and different-sized areas are considered.
Furthermore, environmental influence could well vary
between species. Yet it can be instructive to compare
results of environmental-morphologic patterns for
74
Fishery Bulletin 90(1), 1992
these two dolphin species with broadly overlapping
geographic ranges in the tropical Pacific Ocean.
Sea surface temperatures (variables 6 and 7; July
values depicted in Fig. 12) have negative correlations
with a large number of morphological features in both
studies (Table 7 and Schnell et al. 1986: table 6), in-
dicating a general trend of larger animals in warmer
waters. Surface Salinity (Fig. 13B) exhibits relatively
strong morphologic correlations in both studies, reflect-
ing a pattern that has both east-west and north-south
components. Also, Thermocline Depth (Summer),
which has relatively low values in northern localities
and higher numbers in blocks as one proceeds to the
west and south, is positively associated with a number
of morphological measures in S. longirostris (Table 7),
and covaries with S. attenuata cranial features as well.
Solar Insolation (Jan.) registers a north-south gra-
dient. Our S. longirostris study produced virtually no
significant correlations with this measure, while there
were numerous positive correlations in the S. attenu-
ata investigation. In the eastern portion of the S. longi-
rostris range, a number of cranial features have north-
south gradients, but the overall statistical association
is negated with the addition of the western blocks,
where animals often (irrespective of latitude) exhibit
characteristics similar to those found in southern areas.
The same findings were obtained for Solar Insolation
(Ann.).
Three environmental variables— Water Depth (Fig.
IIB), Thermocline Depth (Winter), and Surface Dis-
solved Oxygen— are positively correlated with cranial
measures in S. longirostris, but show few of these
associations in S. attenuata. Again, the differences in
findings simply may reflect the inclusion of a wider
geographic range of blocks in the S. longirostris study.
Schnell et al. (1986) indicated that for S. attenuata
it would be helpful to have additional samples, par-
ticularly from western locations. They suggested that
"Such a geographic broadening of representation may
enable investigators to separate, at least in part,
environmental-morphological correspondences that
reflect causal relationships from trends [in morphology]
maintained primarily as a result of isolation by dis-
tance." For S. longirostris, where additional western
blocks are now represented (albeit in some cases with
very limited samples), it is clear that the gradients in
a relatively large number of cranial characteristics are
not simply north-south trends, but rather what Perrin
et al. (1991) described as a radial pattern. From north-
eastern blocks, these characters in S. longirostris
change more-or-less gradually as one moves to the
south, the southwest, or the west. There are several
environmental variables exhibiting this type of pattern
(e.g.. Surface Salinity; Fig. 13B). At the same time, the
January and July sea surface temperatures (for July
values, see Fig. 12) have a predominantly north-south
orientation (with the Hawaiian Island blocks being
lower than expected, given their latitude) and are cor-
related with the largest number of cranial characters
(see Table 7). When additional specimens of S. at-
tenuata become available from westerly blocks not
represented in samples available to Schnell et al. (1986),
it will be of interest to determine whether patterns of
cranial variation (and covariation with environmental
measures) in this species will mirror those we have
found for S. longirostris.
Significance of covariation with environmental
measures While previous literature has little infor-
mation on the relation of environmental and cranial
variation in S. longirostris, other investigators (Au and
Perryman 1985, Reilly 1990) have evaluated physical
environmental parameters with respect to distributions
of S. longirostris and several other species in the
eastern tropical Pacific. They pointed out that the
highest school densities for S. longirostris are in the
area off the Mexican coast, which also is the most
tropical and least seasonally variable portion of the
range. We have demonstrated notable associations of
cranial variation with physical environmental char-
acteristics. Analyses involving environmental-cranial
correlations, by their very nature, are descriptive and
do not provide direct information on causal factors per
se. Nevertheless, they clearly indicate that between
areas where animals are different cranially, there often
are marked habitat differences involving the physical
environment.
The first two environmental principal components
(Fig. 10) describe independent, orthogonal environmen-
tal patterns: component I has a general configuration
of high values between 5° and 15°N, slightly lower
values further to the north, and low values to the south
(Fig. lOA), which is overlain by basically an east-to-west
trend summarized in environmental principal compo-
nent II. The important individual environmental covari-
ates with morphological characters include surface
temperatures, salinity, and measures of water depth.
The physical environmental differences reflected by the
principal components, as well as by individual environ-
mental measures, describe basic habitat differences and
likely reflect, indirectly, geographic differences in
available prey species and their abundances. Given the
marked environmental differences exhibited in the
range of S. longirostris, the most surprising result
would have been if this species had been relatively
uniform geographically in cranial features— clearly, this
is not the case. Our initial assessment of morphologic-
environmental covariation further underscores the
appropriateness of treating different parts of the range
of S. longirostris in the eastern tropical Pacific
Douglas et al.: Geographic variation in cranial morphology of Stenella longirosCns
75
separately for management purposes. In particular, a
growing base of information suggests giving special
attention to the spinners from the relatively uniform
area of the Pacific just to the west of the Mexican/
Central American coast, and viewdng the pattern of
morphologic variation as being broadly concentric in
nature.
Acknowledgments
S.B. Reilly and A.E. Dizon reviewed the manuscript
and provided useful suggestions. We thank J. Gil-
patrick and J.V. Kashiwada for preparing and curating
some of the specimens, as well as assisting with prep-
aration of data for analysis. Computer programs and
assistance were provided by F.J. Rohlf, R.R. Sokal, and
M.A. Schene. Support for aspects of this research was
received by the University of Oklahoma through Con-
tract 79-ABC-00167 from the U.S. Department of
Commerce, National Oceanographic and Atmospheric
Administration, and Purchase Orders 84-ABA-02177
and 40JGNF0532 from the National Marine Fisheries
Service, Southwest Fisheries Science Center, La Jolla,
CA. The following individuals provided access to
museum specimens: S. Anderson, G.G. Musser, and
D. Russell (American Museum of Natural History, New
York, NY); M.C. Sheldrick (British Museum of Natural
History, Tring, United Kingdom); S. Bailey, L.C. Bin-
ford, and J. Schoenwald (California Academy of
Science, San Francisco, CA); P.J.H. van Bree (Instituut
voor Taxonomische Zoologie, University of Amster-
dam, Amsterdam, The Netherlands); D.R. Patten and
J.E. Heyning (Los Angeles County Museum of Natural
History, Los Angeles, CA); C.P. Lyman and M. Rutz-
moser (Museum of Comparative Zoology, Harvard
University, Cambridge, MA); W.Z. Lidicker Jr.
(Museum of Vertebrate Zoology, University of Califor-
nia, Berkeley, CA); C.G. van Zyll de Jong (Museum of
Natural Sciences, National Museum of Canada, Otta-
wa, Canada); J.G. Mead and C.W. Potter (National
Museum of Natural History, Washington D.C.); D.W.
Rice and A.A. Wolman (NOAA Northwest and Alaska
Fisheries Science Center, Seattle, WA); L.J. Hansen
(NOAA Southwest Fisheries Science Center, La Jolla,
CA); G.F. Mees (Rijksmuseum van Natuurlijke His-
toire, Leiden, The Netherlands); A. Rea (San Diego
Museum of Natural History, San Diego, CA); M.L.
Johnson and E. Kritzman (University of Puget Sound,
Tacoma, WA); and G.L. Worthen (Utah State Univer-
sity, Logan, UT).
Citations
Allen, R.A.
1985 Dolphins and the purse-seine fishery for yellowfin tuna.
In Beddington, J.R., R.J.H. Beverton, and D.M. Lavigne (eds.).
Marine mammals and fisheries, p. 236-252. Allen and Un-
win, London.
Au, D.W.K., and W.L. Ferryman
1985 Dolphin habitats in the eastern tropical Pacific. Fish.
Bull., U.S. 83:623-643.
Bartholomew, J., and Son
1975 The Times atlas of the world. Comprehensive Edition.
Times Books, London, 223 p.
Brunt, D.
1934 Physical and dynamical meteorology. University Press,
Cambridge. England, 411 p.
Cromwell, T.
1958 Thermocline topography, horizontal currents and "ridg-
ing" in the eastern tropical Pacific. Inter- Am. Trop. Tuna
Comm. Bull. 3:135-152.
Dailey, M.D., and W.F. Perrin
1973 Helminth parasites of porpoises of the genus Stenella in
the eastern tropical Pacific, with descriptions of two new
species: Mastigonema stenellae gen. et sp. n. (Nematoda:
Spiruroidea) and Zalophotrema pacijimm sp. n. (Trematoda:
Digenea). Fish. Bull., U.S. 71:455-471.
Dixon, W.D. (chief editor)
1990 BMDP statistical software, Vol. 1. Univ. Calif. Press,
Berkeley.
Dizon, A.E., S.O. Southern, and W.F. Perrin
1991 Molecular analysis of mtDNA types in exploited popula-
tions of spinner dolphins (Stenella longirostris). Rep. Int.
Whaling Comm. Spec. Issue 13:183-202.
Douglas, M.E., and J. A. Endler
1982 Quantitative matrix comparisons in ecological and evolu-
tionary investigations. J. Theor. Biol. 99:777-795.
Douglas, M.E., G.D. Schnell, and D.J. Hough
1986 Variation in spinner dolphins {Stenella longirostris) from
the eastern tropical Pacific Ocean: Sexual dimorphism in
cranial morphology. J. Mammal. 67:537-544.
Hall, M.A., and S.D. Boyer
1988 Incidental mortality of dolphins in the eastern tropical
Pacific tuna fishery in 1986. Rep. Int. Whaling Comm. 38:
439-441.
1989 Estimates of incidental mortality of dolphins in the
eastern Pacific fishery for tropical tunas in 1987. Rep. Int.
Whaling Comm. 39:321-322.
1990 Incidental mortality of dolphins in the tuna purse-seine
fishery in the eastern Pacific Ocean during 1988. Rep. Int.
Whaling Comm. 40:461-462.
Innis, G., J. Haefner, G. Worthen, and C. Fowler
1979 ETP ecosystem model documentation. Report for
Southwest Fish. Cent., NMFS, NOAA, La Jolla, CA, 277 p.
Katz, J.O., and F.J. Rohlf
1973 Function-point cluster analysis. Syst. Zool. 22:295-301.
Levitus, S.
1982 Climatological atlas of the world ocean. NOAA Prof.
Pap. 13, Rockville, MD.
Manley, B.J.F.
1985 The statistics of natural selection on animal popula-
tions. Chapman and Hall (Meuthen), NY.
Mantel, N.
1967 The detection of disease clustering and a generalized
regression approach. Cancer Res. 27:209-220.
76
Fishery Bulletin 90(1). 1992
Perrin, W.F.
1972 Color patterns of spinner porpoises (Stenella cf. S.
longirostris) of the eastern Pacific and Hawaii, with comments
on delphinid pigmentation. Fish. Bull., U.S. 70:983-1003.
1975a Distribution and differentiation of populations of
dolphins of the genus Stenella in the eastern tropical Pacific. J.
Fish. Res. Board Can. 32:1059-1067.
1975b Variation of spotted and spinner porpoise (genus
Stenella) in the eastern tropical Pacific and Hawaii. Bull.
Scripps Inst. Oceanogr. 21:1-206.
1984 Patterns of geographical variation in small cetaceans.
Acta Zool. Fenn. 172:137-140.
1990 Subspecies of Stenella longirostris (Mammalia: Cetacea:
Delphinidae). Proc. Biol. Soc. Wash. 103:453-463.
Perrin, W.F., W.E. Evans, and D.B. Holt
1979a Movements of pelagic dolphins (Stenella spp.) in the
eastern tropical Pacific as indicated by results of tagging, with
summary of tagging operations, 1969-1976. NOAA Tech.
Rep. NMFS SSRF-737, 14 p.
Perrin, W.F., P.A. Sloan, and J.R. Henderson
1979b Taxonomic status of the 'southwestern stocks' of spin-
ner dolphin, Stenella longirostris, and spotted dolphin, S. at-
tenuata. Rep. Int. Whaling Comm. 29:175-184.
Perrin, W.F., M.D. Scott, G.J. Walker, and V.L. Case
1985 Review of geographical stocks of tropical dolphins
(Stenella spp. and Delphinus delphis) in the eastern Pacific.
NOAA Tech. Rep. NMFS 28, 28 p.
Perrin, W.F., P.A. Akin, J.V. Kashiwada
1991 Geographic variation in external morphology of the spin-
ner dolphin Stenella longirostris in the eastern Pacific and im-
plications for conservation. Fish. Bull., U.S. 89:411-428.
Reilly. S.B.
1990 Seasonal changes in distribution and habitat differences
among dolphins in the eastern tropical Pacific. Mar. Ecol.
Prog. Ser. 66:1-11.
Robinson, M.K.
1976 Atlas of north Pacific Ocean monthly mean temperatures
and mean salinities of the surface layer. Naval Oceanogr. Of-
fice Ref. Publ. 2, Wash. DC.
Rohlf, F.J., J. Kishpaugh, and D. Kirk
1979 NT-SYS. Numerical taxonomy system of multivariate
statistical programs. State Univ. New York, Stony Brook.
Schnell, G.D., M.E. Douglas, and D.J. Hough
1985a Sexual dimorphism in spotted dolphins (Stenella at-
tenuata) in the eastern tropical Pacific Ocean. Mar. Mammal
Sci. 1:1-14.
1986 Geographic patterns of variation in offshore spotted
dolphins (Stenella attenuata) of the eastern tropical Pacific
Ocean. Mar. Mammal Sci. 2:186-213.
Schnell, G.D., D.J. Watt, and M.E. Douglas
1985b Statistical comparison of proximity matrices: Applica-
tions in animal behaviour. Anim. Behav. 33:239-253.
Sneath, P.H.A., and R.R. Sokal
1973 Numerical taxonomy. W.H. Freeman, San Francisco.
Sokal, R.R.
1979 Testing statistical significance of geographic variation
patterns. Syst. Zool. 28:227-232.
True, F.W.
1889 Contributions to the natural history of the cetaceans, a
review of the family Delphinidae. Bull. U.S. Natl. Mus.
36:1-191.
Wirth, M., G.F. Estabrook, and D.F. Rogers
1966 A graph theory model for systematic biology with an ex-
ample for the Oncidiinae (Orchidaceae). Syst. Zool. 15:59-69.
Wyrtki. K.
1966 Oceanography of the eastern equatorial Pacific Ocean.
Oceanogr. Mar. Biol. Annu. Rev. 4:36-68.
1967 Circulation and water masses in the eastern equatorial
Pacific Ocean. Int. J. Oceanol. Limnol. 1:117-147.
1974 The dynamic topography of the Pacific Ocean and its fluc-
tuations. Rep. H-16-74-S, Hawaii Inst. Geophysics, 19 p.
Abstract.- We analyzed the pro-
tein products of 78 isozyme loci in 37
populations of chinook salmon Onco-
rhynchus tshawytscha from Califor-
nia and Oregon. Allele frequencies at
47 polymorphic loci revealed substan-
tial genetic variability within the study
area. The collections of chinook salm-
on studied could be differentiated
into five major groups located in the
following geographical areas: (1)
Smith River- Southern Oregon area,
(2) Middle Oregon Rivers. (3) Kla-
math-Trinity Basin, (4) Eel River-
California Coastal area, and (5)
Sacramento-San Joaquin Basin.
Average heterozygosity estimates
were lowest in collections from the
Klamath-Trinity area and highest in
the Oregon populations. Gene diver-
sity analysis indicated that differ-
ences among fish within samples
accounted for 89.4% of the total
diversity, whereas intersample dif-
ferences accounted for 10.6 %. Esti-
mates of the average level of histor-
ical gene flow between populations
ranged from 15.57 migrants per
generation in the Sacramento-San
Joaquin River system to 3.97 in the
Klamath-Trinity Basin; an overall
estimate of number of salmon ex-
changing genes between populations
per generation was 2.11. Although
these data appeared to reflect pri-
marily population structures existing
prior to the 20th century, evidence
of some effects of hatchery manage-
ment and transplantations was
detected.
Geographic variation in population
genetic structure of cFiinool< salmon
from California and Oregon
Graham A.E. Gall
Devin Bartley
Boyd Bentley
Department of Animal Science
University of California. Davis, California 95616
Jon BrodzJak
Graduate Group in Applied Matfiematics and Institute of Theoretical Dynamics
University of California, Davis. California 95616
Richard Gomulkiewicz
Graduate Group in Applied Matfiematics and Institute of Thieoretical Dynamics
University of California, Davis, California 95616
Present address: Department of Zoology, University of Texas, Austin, Texas 78712
Marc Mangel
Department of Zoology and Center for Population Biology
University of California, Davis, California 95616
Manuscript accepted 13 August 1991.
Fishery Bulletin, U.S. 90:77-100 (1992).
Chinook salmon Oncorhynchus tsha-
wytscha is the most abundant and
commercially important species of
Pacific salmon native to California
and Oregon (Moyle 1976), but stocks
have dechned (Netboy 1974), in some
cases to near extinction. Efforts to
manage and preserve the chinook
fishery have involved traditional
methods such as tag and recapture
estimations and restrictive fishing
regtilations. Recently, however, pop-
ulation genetic analysis of Pacific
salmon has emerged as a major tool
in fishery management to estimate
population subdivision, migration,
gene flow, and stock composition of
ocean fisheries (Ryman and Utter
1987).
Genetic studies on chinook salmon
have refined our understanding of
these populations. Examination of
large numbers of polymorphic loci
revealed geographic associations
among populations of chinook salmon
(Gharrett et al. 1987, Utter et al.
1989, Bartley and Gall 1990, Shaklee
et al. 1990b). Genetic differences
among chinook salmon stocks from
different geographic areas are being
used to identify the stock composition
of mixed ocean salmon fisheries
(Pella and Milner 1987, Utter et al.
1987, Shaklee et al. 1990b, Brodziak
et al. 1992). In addition, genetic
studies have indicated the effects of
climate and geological events on the
population structure of chinook
salmon (Gharrett et al. 1987, Bartley
and Gall 1990).
Utter et al. (1989) and Bartley and
Gall (1990) recently described Cali-
fornia populations of chinook salmon
using data sets with 53 isozyme loci
for 35 populations, and 25 polymor-
phic loci for eight populations, respec-
tively. The objectives of the study
reported here were to further refine
the description of chinook salmon
populations in California and south-
ern Oregon, expand the baseline
genetic data available for genetic
stock-identification studies (Shaklee
et al. 1990b, Brodziak et al. 1992),
77
78
Fishery Bulletin 90(1), 1992
and provide estimates for heterozygosity, allele fre-
quencies, and genetic identities as used for optimum
estimation of stock composition of mixed fisheries.
Materials and methods
Samples
A total 37 samples of juvenile chinook salmon were col-
lected from northern California and southern Oregon
during 1987-88 (Fig. 1, Table 1). Fifteen of these
samples were from fish hatcheries and pond rearing
projects. All the samples represented fall-run fish with
the exception of the upper Sacramento sample (#33)
which represented winter run salmon. To collect out-
migrant chinook salmon from the wild, two fyke nets
(1.5x2.1 X 15m) were placed in a stream approximately
1.6km apart and allowed to set overnight. Juvenile
salmon were removed from the nets the following mor-
ning and frozen on dry ice. Juvenile chinook from
hatcheries were collected with dip nets. A small number
of salmon was taken from each raceway that contained
salmon until a total of 200 fish was collected. At the
laboratory, liver, muscle, heart, and eye tissue were
removed from 100 fish from each collection, placed in
individual tubes, and stored at -80°C. The remaining
100 salmon were frozen at -80°C in an archival
collection.
Electrophoresis
Tissue preparation and horizontal starch-gel electro-
phoresis followed standard procedures (Aebersold et al.
1987). Gels were made with 12% hydrolyzed potato
starch (Connaught Labs.) and one of the following
buffer solutions: CAM, an amine citrate buffer from
Clayton and Tretiak (1972) adjusted to pH 6.8; TBCL,
the discontinuous buffer system of Ridgway et al.
(1970) at pH 8.0; TC-4, a Tris citrate buffer of 0.223
M Tris, 0.083 M citric acid pH 5.8 as electrode buffer,
and a 3.7% mixture of buffer in distilled water for the
gel (Schaal and Anderson 1974); and TG, a Tris glycine
buffer of 0.025 Tris and 0.192 glycine pH 8.5 for both
gel and electrode buffers (Holmes and Masters 1970).
The protein systems analyzed, locus designations,
tissue distribution of isozymes, and buffer systems used
are presented in Table 2. Because of recent changes
in genetic nomenclature (Shaklee et al. 1990a), other
locus name synonyms are presented in Table 2 to
facilitate comparisons with other studies. Allele desig-
nations followed Allendorf and Utter (1979).
Histochemical staining procedures followed Shaw
and Prasad (1970) and Harris and Hopkinson (1976).
The data set described herein constitutes baseline data
Figure 1
Collection sites of 37 samples of chinook salmon Oncorhyn-
chus tshawytscha. Identification numbers are defined in
Table 1.
reported in Gall et al. (1989) and used in maximum-
likelihood estimates for the California mixed ocean
salmon fishery (Brodziak et al. 1992). The duplicated
isoloci AAT-1,2, IDH-3,4, MDH-1,2, MDH-3,4, and
PGM-3,4 each were treated as two loci. Variant alleles
were preferentially assigned to one locus, whereas
common alleles were assigned to the other (Gharrett
et al. 1987). Variation at the IDH-3,4 isoloci was
ascribed to specific loci as described by Shaklee et al.
(1990b). Our method of scoring isoloci is not the method
of choice for studies of genetic mechanisms, as it may
not reflect the true genetic distribution of alleles
Gall et al Geographic variation in population genetics of Oncorhynchus tshawytscha
79
Table I
Thirty-seven collections of juveni
e Chinook salmon from five areas of California and Oregon. Locations of collections
are designated
on Figure 1 by identification number (ID#).
N = number of fish analyzed.
Average
No. of
heterozygosity
Area
ID#
Collection site
N
loci scored
(Nei 1973)
Middle Oregon
1
Fall Creek Hatchery
100
78
0.072
2
Morgan Creek Hatchery
10
78
0.076
3
Millacoma River
100
78
0.072
4
Coquille River. South Fork
100
78
0.O73
5
Elk River Hatchery
100
78
0.076
6
Rock Creek Hatchery
100
78
0.054
S. Oregon/N. California Coastal
7
Rogue River
100
78
0.052
8
Applegate River
100
78
0.054
9
Chetco River Hatchery
100
78
0.063
10
Rowdy Creek Hatchery
62
77
0.067
11
Smith River, Middle Fork
99
77
0.059
Klamath-Trinity Basin
12
Blue Creek
100
77
0.059
13
Omagar Creek Pond-Rearing Facility
100
78
0.064
14
Irongate Hatchery
99
78
0.031
15
Bogus Creek
128
77
0.030
16
Shasta River
100
77
0.028
17
Salmon River
98
76
0.038
18
Camp Creek Pond-Rearing Facility
100
77
0.044
19
Horse Linto Creek
100
77
0.045
20
Trinity River, South Fork
100
77
0.039
21
Trinity River Hatchery
120
77
0.030
Eel River-California Coastal
22
Redwood Creek at Orick
95
77
0.050
23
Redwood Creek Lagoon
100
77
0.054
24
Mad River Hatchery
99
77
0.045
25
Mad River, North Fork
61
77
0.054
26
Eel River, Middle Fork
95
76
0.043
27
Eel River, South Fork
99
78
0.048
28
Van Duzen River
100
77
0.050
29
Redwood Creek, South Fork Eel
93
77
0.046
30
Hollow Tree Creek
100
78
0.045
31
Salmon Creek, South Fork Eel
96
77
0.044
32
Mattole River
100
77
0.049
Sacramento-San Joaquin
33
Upper Sacramento River
94
77
0.059
34
Coleman Hatchery
100
77
0.063
35
Feather River Hatchery
100
78
0.061
36
Nimbus Hatchery
100
78
0.064
37
Merced River Hatchery
100
78
0.057
(Allendorf and Thorgaard 1984, Waples 1988). How-
ever, our method of scoring increases the power of
maximum-Hkelihood estimates of stock composition by
equalizing the importance of variant alleles at isoloci
and non-duplicated loci. Furthermore, our system was
maintained for consistency with other research (Gall
et al. 1989, Brodziak et al. 1992).
A missing heteromeric isozyme between GPI-1 and
GPI-3 was observed in some fish. We scored this pat-
tern, as described in Bartley and Gall (1990), by as-
signing variation to an artificial locus named GPI-H and
labeling the common and variant alleles Gpi-H(lOO)
and Gpi-H(*), respectively. However, Utter et al. (1989)
described breeding data that indicated the variation
should be assigned to either GPI-1 or GPI-3.
Due to the difficulty of identifying heterozygote
banding patterns from GPI-H, LDH-1, and MDHP-2,
allele frequencies at these loci were calculated from the
square root of the frequency of the alternate homo-
zygote. The frequency of the Tpi-3(106) allele also was
calculated from the square root of the frequency of the
homozygous Tpi-3(106) pattern.
80
Fishery Bulletin 90(1), 1992
Table 2
Enzyme systems, lUBNC enzyme number,
sozyme loci
buffer systems, and tissues used in electrophoretic
analyses of chinook salmon.
For loci, m = mitochondrial. M = muscle
H = heart,
L = liver, E
= eye. Buffers explained
in the text.
Locus designations (synon-
yms) are locus names used by (1) present
study, (2) Bartley and Gall (1990), (3) American Fisheries Society (Shaklee
' et al. 1990a),
and (4) Utter et al. (1989).
Locus designations
Enzyme
Enzyme name
no.
1
2
3
4
Tissue
Buffer
Aspartate aminotransferase
2.6.1.1
AAT-1
AAT-1
sAAT-1,2*
Aat-1,2
M,H
TC-4
AAT-2
AAT-2
M,H
TC-4
AAT-3
sAAT-3'
Aat-3
E
TC-4
AAT-4
AAT-3
sAAT-J,'
L
TC-4
mAAT-1
)tiAAT-l'
M,H
CAM
mAAT-2
mAAT-2'
M,H,L
CAM, TC-4
mAAT-3
mAAT-3'
M,H
CAM, TC-4
Acid phosphatase
3.1.3.2
ACP-1
ACP-1'
M,L
CAM
ACP-2
ACP-2'
M
CAM
Adenosine deaminase
3.5.3.3
ADA-1
ADA-1'
M
TG
ADA-2
ADA--2'
M
TG
Alcohol dehydrogenase
1.1.1.1
ADH
ADH
ADH'
L
TC-4, TBCL
Aconitate hydratase
4.2.1.1
AH-1
AH
sAH'
L,M,E
CAM, TC-4
mAH-1
mAH-1'
E,H
CAM
mAH-2
mAH-2*
E,H
CAM
mAH-3
mAH-3'
M,H
CAM
mAH-4
mAH-lf*
M,H
CAM
Alanine aminotransferase
2.6.1.2
ALAT
ALAT*
M
TG
Creatine kinase
2.7.3.2
CK-1
CK-1
CK-Al*
M
TBCL, CAM
CK-2
CK-2
CK-A2'
M
TBCL, CAM
CK-4
CK-3
CK-A2'
E
CAM
Esterase
3.1.1.1
EST-3
EST-D'
M,E
TG, TBCL
Fructose-biphosphate aldolase
4.1.2.13
FBALD-4
FBA
FBALD-J,'
E
CAM, TC-4
Fumarate hydratase
4.2.1.2
FH
FH
FH'
M
CAM
Glycerol-3-phosphate dehydrogenase
1.1.1.8
G3PDH-1
GPDH-1
G3PDH-1'
M
CAM. TC-4
G3PDH-2
GPDH-2
G3PDH-2'
M
CAM, TC-4
G3PDH-3
GPDH-3
GsPDH-3'
M
CAM, TC-4
G3PDH-4
GPDH-4
G3PDH-J,'
M
CAM, TC-4
Glyceraldehyde-3-phosphate dehydrogenase 1.2.1.12
GAPDH-5
GAPDH-3
GAPDH-5'
E
CAM, TC-4
GAPDH-6
GAPDH-4
GAPDH-6'
E
CAM, TC-4
Glucose-6-phosphate isomerase
5.3.1.9
GPI-1
GPI-1
GPI-Bl'
Gpi-1
M
TG, TBCL
GPI-2
GPI-2
GPI-B2'
Gpi-2
M
TG, TBCL
GPI-3
GPI-3
GPI-A'
Gpi-3
M.E
TG, TBCL
GPI-H
GPI-H
GPIr'
Gpi-1
M
TG, TBCL
Glutathione reductase
1.6.4.2
GR
GR
GR'
Gr
M,E,L
TG TBCL
|3-Glucuronidase
3.2.1.31
GUS
GUS'
M
CAM, TC-4
Hydroacylglutathionine hydrolase
3.1.2.6
HAGH
HAGH'
L,M,E
TG
L-Iditol dehydrogenase
1.1.1.14
IDDH-1
IDDH-1
IDDH-1'
L
TBCL
IDDH-2
IDDH-2
IDDH-2'
L
TBCL
Isocitrate dehydrogenase
1.1.1.42
IDH-1
IDH-1
mIDHP-1'
M
CAM
IDH-2
IDH-2
mIDHP-2'
M
CAM
IDH-3
IDH-3
sIDHP-l*
Idh-3,4
M,E,L
CAM, TC-4
IDH-4
IDH-4
sIDHP-2'
E.L
CAM, TC-4
L-Lactate dehydrogenase
1.1.1.27
LDH-1
LDH-1
LDH-Al'
M
TBCL, TC-4
LDH-2
LDH-2
LDH-A2'
M
TBCL, TC-4
LDH-3
LDH-3
LDH-Bl'
H,E
TBCL, TC-4
LDH-4
LDH-4
LDH-B2'
Ldh-4
L,E
TC-4
LDH-5
LDH-5
LDH-C
Ldh-5
E
TC-4
o-Mannosidase
3.2.1.24
MAN
MAN
aMAN*
L
TC-4
Gall et aL: Geographic variation in population genetics of Oncorhynchus tshawytscha
Table 2 (continued)
Enzyme name
Enzyme
no.
Locus designations
Tissue
Buffer
1
2
3
4
Malate dehydrogenase (NADP)
1.1.1.40
MDHP-1
MDHP-2
sMEP-1'
sMEP-S*
M
M.E.L
TC-4
TC-4
mMDHP-1
TtMEP*
M
TC-4
Malate dehydrogenase (NAD)
1.1.1.37
MDH-1
MDH-1
sMDH-Al.2'
Mdh-1,2
E,M
TC-4
MDH-2
MDH-2
E,M
TC-4
MDH-3
MDH-3
sMDH-Bl.S'
Mdh-3,4
M,E
CAM, TC-4
MDH-4
MDH-4
M.E
CAM, TC-4
mMDH-1
mMDH-r
M,E
CAM
mMDH-2
mMDH-2*
M,H
CAM
Mannose-6-phosphate isomerase
5.3.1.8
MPT
MPI
MPI'
Mpi
E,M,L
CAM
Phosphogluconate dehydrogenase
1.1.1.44
PGDH
PGDH
PGDH'
M,E,L
TC-4
Phosphoglucokinase
2.7.2.3
PGK-1
PGK-2
PGK-2
PGK-1*
PGK-2'
Pgk-2
L
M,E,L
CAM
CAM
Phosphoglucosmutase
5.4.2.2
PGM-1
PGM-2
PGM-1
PGM-2
PGM-1*
PGM-2'
Pgm-1,2
M.E
M.E.L
CAM
TG. TC-4
PGM-3
PGM-3 A'
E.L.M
TG, TC-4
PGM-4
E,L,M
TC-4
Pyruvate kinase
2.1.7.40
PK-1
PK-1
PK-1'
M
TC-4
PK-2
PK-2
PK-2*
M
CAM
Superoxide dismutase
1.15.1.1
SOD-1
SOD-1
SOD-1*
Sod
L,M
CAM
mSOD
mSOD*
H,M,E
TG
Triosphosphate isomerase
5.3. 1.1
TPI-3
TPI-4
TPI-2.1*
TPI-2.2*
E
M,E,L,H
TC-4
TG, TBCl
P-N-Acetyl-D-glucosaminidase
3.2.1.30
a-GA
PBGLUA*
L
TG, TBCL
Peptidases (substrates)
Glycyl leucine
3.4.*. ♦
DPEP-1
PEPA-1
PEP-A*
Dpep-1
M.E.H
CAM, TG
DPEP-2
PEPA-2
PEP-C
Dpep-2
E
TG, TBCL
Phenylalanyl proHne
Prolyl leucine
PDPEP-2
PEPLT
PDPEP-2
PEP-D2'
PEP-LT'
M,E
M
TC-4
TG
Leucylglycyl glycine
TAPEP
PEPB
PEP-Bl*
Tapep-1
M.E
TBCL, TG
Analyses
Genetic variability for each collection of salmon was
assessed by calculating the frequencies of alleles at each
locus and average heterozygosity assuming Hardy-
Weinberg proportions (Nei 1973). A locus was con-
sidered variable if we observed polymorphism in at
least one sample. Analyses were based on a maximum
of 78 loci. If a sample was not scored for a particular
locus, the locus was retained for analyses involving
multiple samples. Deviations from expected Hardy-
Weinberg genotypic proportions were tested by chi-
square goodness-of-fit tests (Sokal and Rohlf 1981).
Variant allele frequencies were pooled so the expected
number of genotypes in a given class was always five
or greater. Some loci could not be tested for goodness-
of-fit because pooling allele frequencies to achieve a
minimum class-size reduced the degrees of freedom to
zero. In addition, the loci, PGM-3 and PGM-4, were ex-
cluded from goodness-of-fit tests due to the arbitrary
nature of assigning variation to a specific locus. GPI-H,
LDH-1, and MDHP-2 were excluded because of the
method of calculating allele frequencies from the fre-
quency of the alternate homozygotes.
Genetic identities (I) were calculated for each pair of
samples (Nei 1972) and a dendrogram was constructed
from estimates of I using the unweighted pair-group
method (UPGMA) (Sneath and Sokal 1973). Total gene
diversity (Hx) was partitioned to estimate within-
sample (Hs) and between-sample (Dgx) components,
and to estimate relative gene diversity (Ggx = Dgx/Hx)
(Nei 1973, Chakraborty and Leimar 1987). Total gene
diversity was partitioned into three hierarchical levels:
panmixia (T), area or drainage (D), and sample (S) based
on a priori geographic considerations (Table 1).
An estimate of average gene flow was calculated
from Wright's (1943) fixation index
FsT = l/(4Nm + 1)
(1)
82
Fishery Bulletin 90(1), 1992
where Nm is the average number of migrants exchang-
ing genes per generation. Equation (1) was solved for
Nm by setting Fst equal to the relative gene diversity
appropriate for the hierarchical level of interest. This
formulation provided an estimate of the number of
migrant fish exchanging genes among samples per
generation under the assumptions of selective neutral-
ity of alleles and Wright's (1943) island model of migra-
tion. Slatkin and Barton (1989) discussed the sensitivity
of equation (1) relative to various methods of estimating
Fgx in the presence of selection and alternative popu-
lation structures, and found it to be fairly robust.
Results
A total of 96 isozyme loci were examined. Thirty-one
loci were monomorphic, 47 were categorized as poly-
morphic (Appendix A), whereas variability of an un-
known and undefined nature was detected at 18 loci.
Details of genetic polymorphisms not described else-
where are outlined in Appendix B. The enzyme systems
involving the 18 loci for which evidence of probable
polymorphisms was detected (not listed in Table 2) and
warrant further study included: two adenylate kinase
loci, creatine kinase, four fructose biphosphate aldolase
loci, four glyceraldehyde-3-phosphate dehydrogenase
loci, two beta-galactosidase loci, alpha-glucoside, super-
oxide dismutase, two peptidase loci, and a highly anodal
acromatic band. Because of difficulties defining a gene-
tic model of inheritance, poor band resolution, or in-
complete data, these 18 loci were not included in the
analyses.
Tests of conformance to Hardy Weinberg genotypic
proportions revealed 37 out of 462 cases (8%) of dis-
equilibria. For wild samples of chinook salmon, 13 of
252 tests (5%) revealed disequilibrium, whereas in
hatchery samples, 24 of 210 tests (11%) showed non-
conformance to Hardy- Weinberg expectations. How-
ever, in the Klamath Basin, a higher percentage of
disequilibrium was found (13 of 97 cases or 13%) in
hatchery and wild samples. The proportion of disequi-
librium observed in Klamath and non-Klamath samples
was found to be significantly different (P<0.05) when
tested for equality by the generalized likelihood-ratio
test for binomial data (Larsen and Marx 1981) . The
proportion of disequilibrium observed in hatchery
(including pond rearing programs) and wild chinook
salmon populations also was significantly different
(P<0.05). The nature of the observed disequilibrium
appeared to be random. That is, we did not observe con-
sistent excesses or deficiencies of heterozygotes, nor
did we observe specific loci that consistently deviated
from Hardy- Weinberg expectations.
Estimates of average heterozygosity ranged from a
low value of 0.028 in Shasta River (#16) to a high of
0.076 in the Morgan Creek (#2) and Elk River (#5)
hatcheries. The Middle Oregon samples (#1-6) tended
to have high estimates of average heterozygosity,
whereas values for the Klamath-Trinity samples
(#12-21) tended to be lower (Table 1).
Although genetic identity indices between all pairs
of samples were greater than 0.982 (data not shown),
the geographic distribution of alleles suggested popula-
tion subdivision within the study area. For example,
we found the Aat-2(85), Aat-3(90), Aat-Ml30), and
Iddh-l(O) alleles predominantly in Oregon and north-
coastal California (collections 1-11). The mAh-Jt(112),
Gpi-H(*), and Pgdh<90) alleles were present mainly in
the Sacramento/San Joaquin system (collections 33-
37), whereas Mdhp-1(92) and Gpi-2(60) were less abun-
dant in the Sacramento Basin compared with more
northern areas. Mdhp-2(78) was a characteristic of the
Klamath-Trinity system and a few coastal samples.
Cluster analysis of genetic identities revealed a
strong geographic component to the grouping of
chinook salmon samples. Five distinct clusters that
reflected geographic areas were evident (Fig. 2): (1)
Smith River-Southern Oregon rivers, (2) Klamath-
Trinity Rivers, (3) Eel River system-California coastal
rivers, (4) Middle Oregon rivers, and (5) Sacramento-
San Joaquin system. The Smith River (#11) and the
Rowdy Creek Hatchery (#10) samples were the most
northern samples collected from California. Therefore,
it is reasonable that they would be genetically similar
to the southern Oregon samples. The sample from the
Fall Creek Hatchery (#1) was the only sample from
northern Oregon and therefore, appears as an indepen-
dent cluster. Three samples. Rock Creek Hatchery (#6,
middle Oregon), Blue Creek (#12, Klamath-Trinity
Basin), and Omagar Creek (#13, Klamath-Trinity
Basin), did not cluster in accordance with their geo-
graphic location.
Total gene diversity was 0.0620 (Hx) and average
sample diversity was 0.0554 (Hg). Therefore, approx-
imately 89.4% of the total genetic diversity was due
to intrasample variability and 10.6% was due to inter-
sample variation (Table 3). Further examination of the
intersample diversity showed that genetic differences
among samples within the five geographic groups iden-
tified from the dendrogram (see Table 1) accounted for
about 3.2% of the total variation and 7.4% of the total
diversity was due to differences between the major
geographic areas. Gene diversity analysis for each
geographic area treated separately revealed that
although the Klamath-Trinity system possessed the
lowest total gene diversity for a given area (Hd), rela-
tive gene diversity (Ggo) for this drainage was high
Gall et al : Geographic variation in population genetics of Oncorhynchus tshawytscha
83
S Oregon/
N. Cahl
Omagar Creek
Eel River/
Calif Coast
KlafTialh/
Trinity
Mid Oregon
Sacramento/
San Joaauin
1 000 996 992 98B
Genetic Identity
Figure 2
Dendrogram based on UPGMA clustering of genetic identi-
ty indices (Nei 1972). Identification numbers are defined in
Table 1. Brackets on left side indicate geographic grouping,
with Blue Creek and Omagar Creek as outliers (collection #6,
indicated as 6*, was from mid-Oregon).
and comparable to the middle Oregon area which
shared the highest total gene diversity (Table 3).
Based on an overall estimate of 0.106 for Gst (Table
3), the number of immigrant individuals contributing
genes to an average population, Nm, was estimated to
be 2.11 individuals per generation. Estimates of gene
flow within each geographic cluster were highest in the
Sacramento-San Joaquin system (Nm 15.57) and low-
est in the Klamath-Trinity drainage (Nm 3.97).
Discussion
The genetic structure of chinook salmon populations
reported here appears similar to that reported pre-
viously. Distributions of variant alleles at Mdh-4, AH-1
Pgdh, Pgm-2, GPI-H, and Gpi-2 were similar to those
reported by Bartley and Gall (1990). However, average
heterozygosity estimates for the Klamath-Trinity
Table 3
Hierarchical gene diversity analyses of 37 samples of chinook
salmon from Oregon and California.* HgQ = average gene
diversity of samples within areas; Hp and Gjq = total gene
diversity and relative gene diversity for a given area, respec-
tively; Nm = average number of migrants exchanging genes
per generation; H,, Hf, and Gg^ = within-sample, total, and
relative gene diversity, respectively.
Area
HsD
Hd
GsD
Nm
Middle Oregon
0.0704
0.0741
0.0502
4.70
South Oregon/
N. California Coast
0.0586
0.0599
0.0223
10.96
Klamath-Trinity
0.0402
0.0428
0.0592
3.97
Eel River/California Coast
0.0473
0.0486
0.0271
8.98
Sacramento-San Joaquin
0.0607
0.0616
0.0158
15.57
0.106, Nm 2.11
drainage were somewhat higher than reported by Utter
et al. (1989) and Bartley and Gall (1990). Bartley and
Gall (1990) observed a range of 0.008-0.016 for this
drainage, compared with the range of 0.028 for the
Shasta River sample to 0.064 for the sample from
Omagar Creek found in the present study. One reason
for the higher estimates in the present study was the
inclusion of the Mdhp-2 locus, which is highly polymor-
phic in the Klamath-Trinity drainage (Appendix A);
Bartley and Gall (1990) and Utter et al. (1989) did not
report data for this locus. Generally, comparisons of
heterozygosity estimates between this study and earlier
studies are difficult to interpret due to the improved
laboratory procedures that have greatly increased the
number of isozyme loci available for analysis.
Two samples from the Klamath-Trinity drainage.
Blue and Omagar Creeks, were genetically differen-
tiated from other samples from within the basin. For
example, Mdhp-2(78) had an average frequency of 0.32
in eight other samples from the drainage, whereas the
allele occurred at a frequency of 0.14 in Blue Creek and
was not found in the Omagar Creek sample. Further-
more, Omagar and Blue Creeks had higher frequencies
of the Tapep-l(130) and mMdh-l(-900) alleles than did
other Klamath-Trinity samples. These frequencies in-
dicated that fish from Omagar and Blue Creeks are
genetically closer to southern Oregon populations than
to Klamath-Trinity populations. This result was unex-
pected given the pattern of geographic clustering foimd
by Utter et al. (1989) and Bartley and Gall (1990).
However, earlier studies did not sample populations
near or below the confluence of the Trinity and
Klamath Rivers, as was done in the present study.
84
Fishery Bulletin 90(1). 1992
We do not know if the genetic structure of the Blue
and Omagar Creek samples is characteristic of the
lower Klamath-Trinity drainage. The Omagar Creek
sample consisted of progeny of broodstock captured by
instream gill nets at the mouth of Blue Creek and in
the main section of the Klamath River; the Blue Creek
sample was collected in the main stem of Blue Creek
and was presumed to represent progeny of natural
spawning. If accurate, our data suggest greater gene
exchange between the lower Klamath and coastal
populations of northern California-southern Oregon
than between the lower and upper Klamath basin. Ap-
parently northern California coastal populations of
Chinook salmon are genetically similar to southern
Oregon populations because the two samples from the
Smith River (samples 10 and 11) also clustered with
the Oregon populations. This genetic similarity may
have resulted from historical gene exchange in the form
of transplants into the Klamath basin (Snyder 1931).
Chinook salmon in the lower Klamath River are
thought to be similar to Oregon populations in other
characters, such as timing of spawning migration,
fecundity, and size (Snyder 1931; Craig Tuss, U.S. Fish
Wildl. Serv., Sacramento, CA 95616, pers. commun.,
Sept. 1990).
The relatively high incidence of Hardy-Weinberg
disequilibria in hatchery and pond rearing programs
may be the result of the limited number of broodstock
used in production or non-random sampling of a hatch-
ery's production, i.e., only sampling juveniles from a
few raceways. For example, the Coleman National Fish
Hatchery spawns approximately 10,000 fall-run Chi-
nook salmon. It is likely that our sample of 100 juveniles
may not be an adequate representation of the hatchery
output. The two samples with the highest number of
deviations from Hardy-Weinberg expectations were
both from pond rearing projects, Omagar and Camp
Creeks. These pond rearing projects can serve a useful
function by augmenting or establishing runs of chinook
salmon in specific streams. However, care must be
taken to maximize the effective population size of the
broodstock and to prevent changes in the genetic
variation.
The large number of significant departures from
Hardy-Weinberg expectations for the Klamath samples
compared with other samples was due primarily to the
samples from Camp Creek and Omagar Creek. These
two samples accounted for nine of the 13 significant
tests within the Klamath system. Deleting data for
these two Creeks from the comparison resulted in 6%
(4 of 72) significant deviations for Klamath system
samples versus 7% (24 of 349) for non-Klamath
samples.
Our results indicate a geographic basis for genetic
differentiation and subpopulation structure in chinook
salmon populations from California and Oregon. Geo-
graphic affinities among chinook salmon populations
have now been demonstrated along most of the western
coastline of North America (Gharrett et al. 1987, Utter
et al. 1989, Bartley and Gall 1990). Bartley and Gall
(1990) identified three major clusters of chinook salmon
populations in California that corresponded to the three
major river drainages: the Sacramento-San Joaquin,
the Eel, and the Klamath-Trinity. Utter et al. (1989)
identified nine population imits of chinook salmon over
a large area from British Columbia to California. They
found coastal populations from Oregon and Washing-
ton to be genetically similar to each other. Our data
indicate that some coastal populations in California are
differentiated from those in Oregon, but that northern
California coastal populations of chinook salmon are
similar to southern Oregon populations.
The level of intrasample gene diversity found in the
present study, 89.4%, is similar to the values of 82.3
and 87.7% reported by Bartley and Gall (1990) and
Utter et al. (1989), respectively. Overall estimates of
gene flow of 1.16 (Bartley and Gall 1990) and 2.11 (this
study) migrants per generation also are similar. The
slightly lower level of population subdivision and there-
fore, higher level of gene flow found in the present
study probably reflect a bias caused by the samples
analyzed. Bartley and Gall (1990) analyzed a greater
number of inland California populations than the pres-
ent study. Most of their samples were from the three
major drainages within California: the Klamath-
Trinity, the Sacramento-San Joaquin, and the Eel.
They suggested that straying and gene flow were
higher among coastal streams than among separate
drainages. Therefore, by including the large number
of coastal samples in the present study, slightly higher
overall estimates of gene flow and less apparent
subdivision were expected. Separate gene diversity
analyses of the groups from Oregon and northern
California revealed that approximately 6% of the total
diversity of the two Oregon groups was due to inter-
population differences compared with 12% for the
three California groups. These results further support
the expectation of lower levels of population subdivi-
sion when analyses involve many coastal samples.
The estimates of gene flow and population subdivi-
sion from hierarchical gene-diversity analyses varied
among geographic areas. The Klamath-Trinity system
would be expected to display lower levels of gene ex-
change if the lower and upper sections of the Klamath
are separate subpopulations. However, deletion of the
Blue Creek and Omagar Creek samples from the anal-
ysis changed the gene diversity estimates by less than
2%. The high level of estimated gene flow within the
Sacramento-San Joaquin system most likely reflects
the fact that four of the five samples were from
Gall et al : Geographic variation in population genetics of Oncorhynchus tshawytscha
85
hatcheries. Although egg and fingerHng transfers be-
tween areas have been reduced recently, a considerable
amount of historical mixing of the hatchery stocks has
occurred (Alan Baracco, Calif. Dep. Fish Game,
Sacramento, CA 95616, pers. commun. Dec. 1986). Ad-
ditionally, many salmon from the San Joaquin River
stray into the Sacramento River on their spawning
migration due to easier access and better water qual-
ity in the Sacramento River (Alan Baracco and For-
rest Reynolds, Calif. Dep. Fish Game, Sacramento, CA
95616, pers. commun. Dec. 1986).
Independent estimates of straying based on coded-
wire tagged fish indicate that chinook salmon in the
Sacramento River do stray within the system. Rough
estimates are that 2-5% of the Sacramento fall-run fish
are from hatcheries in the San Joaquin River system.
Approximately 1% of the fall-run chinook salmon
returning to the Feather River Hatchery is composed
of stray fish from the Nimbus (American River), Moku-
lumne, and Coleman Hatcheries. Straying also occurs
in northern streams because chinook salmon marked
on the Rogue River are recovered in the Klamath-
Trinity drainage (Fred Meyer, Calif. Dep. Fish Game,
Rancho Cordova, CA 95670, pers. commun. Feb. 1991).
Therefore, it is not surprising that gene flow esti-
mates for the Sacramento-San Joaquin drainage were
high and that southern coastal populations from
Oregon should resemble northern California coastal
populations.
Stability of allele frequencies over time is often
assumed in the methodology of genetic stock identifica-
tion. Although the present study was not intended to
uncover temporal variation of allele frequencies, some
samples we examined also had been analyzed earlier.
Eighteen locations from the present study were sam-
pled in 1984-86 by Bartley and Gall (1990). For the
interstudy comparison, loci chosen had to have a fre-
quency of less than 0.95 for the common allele in at
least two populations reported by Bartley and Gall
(1990); isoloci were not used. Twelve loci fit the cri-
terion: AH-1, DPEP-1, PDPEP-2, TAPER, GPI-2,
IDDH-2, IDH-2, MPI, PGDH, PGK-2, PGM-2, and
SOD-1.
We found 18 instances of significant change in allele
frequencies for seven hatchery samples (21.4%), 16
significant results for seven wild populations (19.0%),
and five instances of significant change for a pond rear-
ing project (41.7%) based on the G-statistic (Sokal and
Rohlf 1981). Interstudy comparisons of the samples
from Bogus Creek ( = Bogas Creek in Bartley and Gall
1990), Shasta Creek, and the Feather River Fish
Hatchery revealed no significant differences in allele
frequencies.
Six hatcheries sampled in the present study also had
been sampled by Utter et al. (1989). Loci selected to
compare allele frequencies for these studies had to have
a common allele frequency of less than 0.95 in one of
the studies. Eight loci met the frequency criterion:
AH, DPEP-1, TAPER, GPI-2, GR, MPI, PGK-2, and
SOD-1. Five of the six hatchery samples displayed
significant changes in allele frequency between the two
studies. Waples and Teel (1990) also reported signifi-
cant changes in allele frequencies in hatcheries sam-
pled in different years.
Although we observed differences in allele frequen-
cies between this and earlier studies, we do not know
if this represents temporal variation. It is tempting to
make statements on the temporal stability or instability
of allele frequencies in samples of chinook salmon from
a given area, but without estimates of sampling vari-
ability for a given year, it is not possible to separate
intrasample variation, random sampling error, and
temporal variation. Nevertheless, given the presumed
constancy of allele frequency data (Allendorf and Utter
1979), the number of significant G statistics uncovered
in comparisons between samples in this study and those
of Utter et al. (1989) and Bartley and Gall (1990) re-
quires some explanation.
Waples and Teel (1990:149) stated, "tests of the
equality of allele frequencies in temporally spaced
samples must be interpreted with caution." In addition,
Waples and Teel (1990) list inaccurate or artifactual
genetic data, nonrandom sampling of fish for genetic
analysis, selection, and migration as possible causes of
significant change in allele frequencies. For example,
large differences in allele frequencies at IDH-3 and
IDH-4 between the present study and Bartley and Gall
(1990) may be due to banding artifacts associated with
tissue breakdown. One of us (Bentley) has observed the
increased appearance of variant "alleles" at these loci
in samples that were not properly frozen and stored.
Therefore, the data for these two loci presented in
Bartley and Gall (1990) may be artifactual. In addition,
the analyses of Utter et al. (1989), Bartley and Gall
(1990), and the present study were done by different
personnel in different laboratories. Although standar-
dization was attempted, scoring of gel banding patterns
may have been inconsistent.
The level of temporal instability of allele frequencies
is an important issue in the use of GSI to manage and
conserve chinook salmon populations (Waples 1990,
Waples and Teel 1990). However, sampling design
should specifically address this question before one
draws conclusions concerning wild or hatchery popula-
tions. Although we documented differences in allele fre-
quencies between this and earlier studies, the overall
association between genetic similarity and geographic
location remains constant for populations of chinook
salmon in California and Oregon.
86
Fishery Bulletin 90(1). 1992
Acknowledgments
This research was funded by the California Department
of Fish and Game (Interagency Agreement No. C-1335,
Genetic Analysis of Chinook and Coho Salmon Popula-
tions) and the Institute for Theoretical Dynamics at the
University of California, Davis. We gratefully acknowl-
edge the support and assistance of A. Baracco and L.B.
Boydstun throughout the study. We thank personnel
from the California Department of Fish and Game, the
U.S. Fish and Wildlife Service, S. Downey, and
W. Shoals for assistance with fish collections. The
technical assistance from E. Childs, S. Fox, A. Mar-
shall, C. Panattoni, and C. Qi is also appreciated. The
valuable comments of F. Utter and two anonymous
referees also are appreciated. We are especially
grateful to the Northwest Fisheries Science Center of
the National Marine Fisheries Service and the
Washington Department of Fisheries Genetic Unit for
their contribution to the development of a coastwide
program of Genetic Stock Identification.
Citations
Aebersold, P.B., G.A. Winans, D.J. Teel, G.B. Milner, and
F.M. Utter
1987 Manual for starch gel electrophoresis: A method for the
detection of genetic variation. NOAA Tech. Rep. NMFS 61,
19 p.
Allendorf, F., and G.H. Thorgaard
1984 Tetraploidy and the evolution of salmonid fishes. In
Turner, B. (ed.). Evolutionary genetics of fishes, p. 1-53.
Plenum, NY.
Allendorf, F.W., and F.M. Utter
1979 Population genetics. In Hoar, W.J., and D.J. Randall
(eds.), Fish physiology, vol. 8, p. 407-454. Academic Press,
NY.
Bartley. D.M., and G.A.E. Gall
1990 Genetic structure and gene flow in chinook salmon
populations of California. Trans. Am. Fish. Soc. 119:55-71.
Brodziak, J., B. Bentley, D. Bartley, G.A.E. Gall,
R. Gomulkiewicz, and M. Mangel
1992 Tests of genetic stock identification using coded-wire
tagged fish. Can. J. Fish. Aquat. Sci. (In press).
Chakraborty, R., and 0. Leimar
1987 Genetic variation within a subdivided population. In
Ryman, N., and F. Utter (eds.). Population genetics and fishery
management, p. 89-120. Univ. Wash. Press, Seattle.
Clayton, J.W., and D.N. Tretiak
1972 Amine-citrate buffers for pH control in starch gel elec-
trophoresis. J. Fish. Res. Board Can. 29:1169-1172.
Gall, G.A.E., B. Bentley, C. Panattoni, E. Childs, C. Qi, S. Fox,
M. Mangel, J. Brodziak, and R. Gomulkiewicz
1989 Genetic stock identification: Chinook mixed fishery pro-
ject 1986-1989. Rep. to Calif. Dep. Fish Game, Sacramento,
by Univ. Calif., Davis, 420 p.
Gharrett, A.J., S.M. Shirley, and G.R. Tromble
1987 Genetic relationships among populations of Alaskan
chinook salmon (Oncorhynckus tshawytscha). Can. J. Fish.
Aquat. Sci. 44:765-774.
Harris, H., and D.A. Hopkinson
1976 Handbook of enzyme electrophoresis in human genetics.
North Holland Publ. Co., Amsterdam, var. pag.
Holmes, R.S., and C.J. Masters
1970 Epigenetic interconversions of the multiple forms of
mouse hver catalase. FEBS (Fed. Eur. Biochem. Soc.) Lett.
11:45-48.
Larsen, R.J., and M.L. Marx
1981 An introduction to mathematical statistics and its applica-
tions. Prentice Hall, Engelwood Cliffs, var. pag.
Moyle, P.B.
1976 Inland fishes of California. Univ. CalLf. Press, Berkeley.
Nei, M.
1972 Genetic distance between populations. Am. Nat. 106:
283-292.
1973 Analysis of genediversity in subdivided populations.
Proc. Natl. Acad. Sci. USA 70:3321-3323.
Netboy, A.
1974 The salmon: Their fight for survival. Houghton Mifflin,
Boston, 613 p.
Pella, J.J., and G.B. Milner
1987 Use of genetic marks in stock composition analysis. In
Ryman, N., and F. Utter (eds.), Population genetics and fishery
management, p. 247-276. Univ. Wash. Press, Seattle.
Ridg^way, G.J., S.W. Sherburne, and R.D. Lewis
1970 Polymorphisms in the esterase of Atlantic herring.
Trans. Am. Fish. Soc. 99:147-151.
Ryman, N., and F. Utter (editors)
1987 Population genetics and fishery management. Univ.
Wash. Press, Seattle, 420 p.
Schaal, B.A., and W.W. Anderson
1974 An outline of techniques for starch gel electrophoresis
of enzymes from the America oyster Crassostrea virginica
Gmelin. Tech. Rep. 74-3, Ga. Mar. Sci. Cent., 18 p.
Shaklee, J.B.. R.W. AIIendorf.D.C. Morizot. and G.S. Whitt
1990a Gene nomenclature for protein coding loci in fish.
Trans. Am. Fish. Soc. 119:2-15.
Shaklee, J.B., C. Busack, A. Marshall, M. Miller, and S.R. Phelps
1990b The electrophoretic analysis of mixed-stock fisheries of
Pacific salmon. In Ogita, Z-I., and C.L. Markert (eds).
Isozymes: Structure, function, and use in biology and medicine,
p. 235-265. Wiley-Liss, Inc., NY.
Shaw, C.R., and R. Prasad
1970 Starch gel electrophoresis of enzymes— a compilation of
recipes. Biochem. Genet. 4:297-320.
Slatkin, M., and N.H. Baron
1989 A comparison of three indirect methods for estimating
average levels of gene flow. Evolution 43:1349-1368.
Sneath, P.H.A., and R.R. Sokal
1973 Numerical taxonomy. W.H. Freeman, San Francisco,
573 p.
Snyder, J.O.
1931 Salmon of the Klamath River, California. Calif. Dep.
Fish Game, Fish. Bull. 34:1-130.
Sokal, R.R., and F.J. Rohlf
1981 Biometry. W.H. Freeman, NY, 859 p.
Utter, P.M., D. TeeL G. Milner. and D. Mclsaac
1987 Genetic estimates of stock comparisons of 1983 chinook
salmon, Oncorhynchiis tshawytscha, harvests off the Wash-
ington coasts and Columbia River. Fish. Bull., U.S. 85:12-23.
Gall et al .: Geographic variation in population genetics of Oncorhynchus tshawytscha
87
Utter, F.M.. G. Milner, G. Stahl. and D. Teel
1989 Genetic population structure of chinook salmon, Onco-
rhynchus tshawytscha, in the Pacific Northwest. Fish. Bull,
U.S. 87:239-264.
Waples. R.S.
1988 Estimation of allele frequencies at isoloci. Genetics 118:
371-384.
1990 Temporal changes of allele frequency in Pacific salmon:
Implications for mixed-stock fishery analysis. Can. J. Fish.
Aquat. Sci. 47:968-976.
Waples. R.S., and D.J. Teel
1990 Conservation genetics of Pacific salmon. I. Temporal
changes in allele frequency. Conserv. Biol. 4:144-156.
Wright, S.
1943 Isolation by distance. Genetics 28:114-138.
Appendix A
Allele frequencies at 47 variable isozyme loci. Identification numbers (ID#) defined in Table 1 and Figure 1;
N = number of fish scored. Allele designations of Bartley and Gall (1990) are included in parentheses.
A AT •>
Alleles
o
Alleles
A AT-/I
All«
100
(100)
QC 1 HK A AT
;les
AAl
ID#
N
(90)
ID#
N
100
100
1.000
90
ID#
1
N
100
100
0.755
130
Middle Oregon
1
100
0.990
0.010
1
0.245
2
100
0.930
0.070
2
100
0.995
0.005
2
100
0.785
0.215
3
100
0.890
0.110
3
100
1.000
3
100
0.875
0.125
4
100
0.920
0.080
4
100
0.995
0.005
4
100
0.835
0.165
5
100
0.910
0.090
5
100
1.000
5
100
0.880
0.120
6
100
1.000
6
100
0.975
0.025
6
100
1.000
S. Oregon/
7
100
1.000
7
100
0.965
0.035
7
100
0.995
0.005
N. California Coastal
8
100
1.000
8
100
0.965
0.035
8
100
1.000
9
100
0.995
0.005
9
100
1.000
9
100
1.000
10
62
1.000
10
62
0.960
0.040
10
62
1.000
11
99
0.970
0.030
11
99
0.990
0.010
11
99
0.995
0.005
Klamath-Trinity Basin
12
100
1.000
12
100
0.990
0.010
12
100
0.975
0.025
13
100
1.000
13
100
1.000
13
100
0.990
0.010
14
98
1.000
14
99
1.000
14
98
0.995
0.005
15
127
1.000
15
128
0.992
0.008
15
121
0.975
0.025
16
100
1.000
16
100
1.000
16
100
0.970
0.030
17
98
1.000
17
98
1.000
17
85
0.976
0.024
18
106
1.000
18
106
1.000
18
106
0.877
0.123
19
100
1.000
19
100
1.000
19
100
1.000
20
100
1.000
20
100
0.985
0.015
20
100
0.970
0.030
21
120
1.000
21
120
1.000
21
120
0.996
0.004
Eel River-California Coastal
22
95
0.968
0.032
22
95
1.000
22
87
1.000
23
100
0.965
0.035
23
100
1.000
23
100
1.000
24
99
0.995
0.005
24
99
1.000
24
99
1.000
25
61
1.000
25
61
1.000
25
60
1.000
26
95
1.000
26
95
1.000
26
95
1.000
27
99
1.000
27
99
1.000
27
97
1.000
28
100
1.000
28
100
1.000
28
88
0.994
0.006
29
93
1.000
29
93
1.000
29
93
1.000
30
100
0.995
0.005 30
100
1.000
30
94
1.000
31
96
1.000
31
96
1.000
31
93
0.984
0.016
32
100
1.000
32
100
1.000
32
100
1.000
Sacramento-San Joaquin
33
34
94
100
1.000
1.000
33
34
94
100
1.000
1.000
33
34
94
100
1.000
0.995
0.005
35
100
1.000
35
100
1.000
35
100
1.000
36
100
1.000
36
100
1.000
36
100
1.000
37
100
1.000
37
100
1.000
37
100
1.000
88
Fishery Bulletin 90(1). 1992
Appendix A (continued)
Alleles
Alleles
Allele.*?
m A AT" 1
m A AT.")
m A A'T-'l
ID#
N
-100
-77
-104
ID#
N
-100
-125
-90
ID#
N
-100
-450
Middle Oregon
1
100
1.000
1
100
0.985
0.015
1
100
1.000
2
100
0.970
0.030
2
100
0.960
0.040
2
100
0.965
0.035
3
100
0.990
0.010
3
100
0.985
0.015
3
100
0.970
0.030
4
100
1.000
4
100
0.975
0.025
4
100
0.955
0.045
5
100
0.990
0.010
5
100
1.000
5
100
0.925
0.075
6
100
0.985
0.015
6
100
0.945
0.055
6
100
1.000
S. Oregon/
7
100
0.980
0.020
7
100
0.945
0.005
0.050
7
100
1.000
N. California Coastal
8
100
0.980
0.020
8
100
0.945
0.055
8
100
1.000
9
100
0.985
0.015
9
100
0.975
0.025
9
100
0.995
0.005
10
62
0.984
0.016
10
62
0.911
0.089
10
0
11
99
0.955
0.005
0.040
11
70
1.000
11
0
Klamath-Trinity Basin
12
100
1.000
12
100
0.955
0.045
12
0
13
100
1.000
13
100
0.965
0.035
13
100
1.000
14
99
1.000
14
59
0.983
0.017
14
59
1.000
15
128
1.000
15
49
0.980
0.020
15
0
16
100
1.000
16
69
0.993
0.007
16
0
17
98
1.000
17
98
0.969
0.031
17
0
18
106
1.000
18
106
1.000
18
0
19
100
1.000
19
100
1.000
19
0
20
100
1.000
20
100
0.970
0.030
20
0
21
120
1.000
21
80
0.994
0.006
21
0
Eel River-California Coastal
22
95
1.000
22
95
1.000
22
0
23
100
1.000
23
100
1.000
23
0
24
99
0.990
0.010
24
99
0.980
0.020
24
0
25
61
1.000
25
61
0.967
0.033
25
0
26
95
0.979
0.021
26
95
1.000
26
0
27
98
1.000
27
46
0.989
0.011
27
40
1.000
28
100
0.995
0.005
28
40
1.000
28
0
29
93
1.000
29
93
1.000
29
0
30
100
1.000
30
40
1.000
30
40
1.000
31
96
1.000
31
96
1.000
31
0
32
100
1.000
32
100
0.995
0.005
32
0
Sacramento-San Joaquin
33
94
0.995
0.005
33
94
1.000
33
0
34
100
0.960
0.040
34
100
0.995
0.005
34
0
35
100
0.975
0.025
35
100
0.995
0.005
35
100
1.000
36
100
1.000
36
100
1.000
36
100
1.000
37
100
1.000
37
100
1.000
37
100
1.000
AriA 1
Alleles
ADH
ID#
[
Alleles
AH-1
Alleles
100
86
116
ID#
N
100
83
108
N
-100
-52
ID#
N
(100)
(90)
(110)
Middle Oregon
1
100
0.980
0.020
1
100
1.000
1
100
0.855
0.050
0.095
2
100
0.990
0.010
2
100
0.975
0.025
2
100
0.890
0.095
0.015
3
100
1.000
3
100
0.995
0.005
3
100
0.875
0.090
0.035
4
100
0.990
0.010
4
100
1.000
4
100
0.855
0.135
0.010
5
100
0.995
0.005
5
100
1.000
5
100
0.845
0.145
0.010
6
100
1.000
6
100
0.990
0.010
6
100
0.890
0.100
0.010
S. Oregon/
7
100
1.000
7
100
1.000
7
100
0.935
0.065
N. California Coastal
8
100
1.000
8
100
1.000
8
100
0.960
0.040
9
100
1.000
9
100
1.000
9
100
0.925
0.075
10
62
1.000
10
62
1.000
10
62
0.839
0.161
11
99
1.000
11
99
1.000
11
99
0.919
0.076
0.005
Klamath-Trinity Basin
12
100
0.995
0.005
12
100
1.000
12
100
0.940
0.060
13
100
1.000
13
100
1.000
13
100
1.000
14
99
1.000
14
99
1.000
14
99
0.990
0.005
0.005
15
128
1.000
15
118
1.000
15
128
1.000
16
100
1.000
16
100
1.000
16
100
0.995
0.005
17
0
17
97
1.000
17
98
1.000
18
106
1.000
18
106
1.000
18
106
0.953
0.047
19
100
1.000
19
100
1.000
19
100
1.000
Gall et at.: Geographic variation in population genetics of Oncorhynchus tshawytscha
89
Appendix A (continued)
ADA-l
Alleles
AHW
Alleles
AH.1
Alleles
100
(100)
86
(90)
116
(110)
ID#
N
100
83 108
ID#
N
-100
-52
ID#
N
Klamath-Trinity Basin
20
100
1.000
20
100
1.000
20
100
1.000
(continued)
21
120
1,000
21
120
1.000
21
120
1.000
Eel River-California Coastal
22
76
1,000
22
95
1.000
22
95
0.968
0.021
0.011
23
100
1,000
23
100
1.000
23
100
0.945
0.040
0.015
24
99
1.000
24
99
0.970
0.030
24
99
1.000
25
61
1,000
25
61
1.000
25
61
1.000
26
0
26
95
1.000
26
95
0.979
0.021
27
99
1.000
27
79
1.000
27
99
0.995
0.005
28
100
1.000
28
83
1.000
28
100
1.000
29
93
1,000
29
93
1.000
29
93
1.000
30
100
1.000
30
100
1.000
30
100
1.000
31
23
1,000
31
94
1.000
31
96
1.000
32
100
1,000
32
100
1.000
32
100
1.000
Sacramento- San Joaquin
33
94
1,000
33
94
1,000
33
94
0,862
0.128
0.011
34
100
1,000
34
100
1,000
34
100
0.775
0.200
0.025
35
100
0.955
0.045
35
100
1.000
35
100
0.885
0.105
0.010
36
100
0,960
0.040
36
100
1.000
36
100
0,835
0.130
0.035
37
100
0.870
0.130
37
100
1.000
37
100
0.765
0.165
0.070
Allele.'!
Alleles
Alleles
mAH-^
m AP-'J
m AP-^
ID#
N
100
65
111 riff
ID#
1
N
100
100
1,000
50
ID#
1
N
100
100
1.000
71
Middle Oregon
1
100
1.000
2
100
1.000
2
100
1,000
2
100
1.000
3
100
1.000
3
100
1,000
3
100
1.000
4
100
1.000
4
100
0,985
0.015
4
100
1.000
5
100
1.000
5
100
0,995
0.005
5
100
1.000
6
100
1.000
6
100
1,000
6
100
1.000
S. Oregon/
7
100
1.000
7
100
1,000
7
100
0.995
0.005
N. California Coastal
8
100
0,980
0.020
8
100
1,000
8
100
0.995
0.005
9
100
1,000
9
100
1,000
9
100
1.000
10
61
0,992
0.008
10
61
1.000
10
62
1.000
11
99
1,000
11
99
1.000
11
99
0.990
0.010
Klamath-Trinity Basin
12
100
0,995
0.005
12
100
1.000
12
100
0.995
0.005
13
100
1,000
13
100
1.000
13
100
1.000
14
99
1,000
14
99
1.000
14
99
1.000
15
128
0.980
0,020
15
128
1.000
15
128
1.000
16
100
0.990
0,010
16
100
1.000
16
100
1.000
17
98
0.995
0.005
17
98
1.000
17
98
1,000
18
87
1.000
18
87
1.000
18
106
1,000
19
100
1.000
19
100
1.000
19
100
1,000
20
100
1,000
20
100
1.000
20
100
1.000
21
120
0,975
0.025
21
120
1.000
21
120
1.000
Eel River-California Coastal
22
95
0.947
0.053
22
95
1.000
22
95
1.000
23
100
0.955
0.045
23
100
1.000
23
100
1.000
24
99
0.894
0.106
24
99
1.000
24
99
1.000
25
61
0.893
0.107
25
61
1.000
25
61
1.000
26
95
0.974
0.026
26
95
0.989
0.011
26
95
1.000
27
99
0.965
0.035
27
99
1.000
27
99
1.000
28
100
0.935
0.065
28
100
1.000
28
100
1.000
29
93
0,984
0.016
29
93
1.000
29
93
1.000
30
100
0.920
0.080
30
100
1.000
30
100
1.000
31
96
0.990
0.010
31
96
0.984
0.016
31
96
1.000
32
99
0.909
0,091
32
100
1.000
32
100
1.000
Sacramento-San Joaquin
33
94
0.973
0,027
33
94
1.000
33
94
1.000
34
100
0.975
0,025
34
100
1.000
34
100
1.000
35
100
0.995
0,005
35
100
1.000
35
100
1.000
36
100
1.000
36
100
1,000
36
100
1.000
37
100
1.000
37
100
1.000
37
100
1.000
90
Fishery Bulletin 90(1), 1992
Appendix A (continued)
Allf^lfts
Alleles
mAH-4
ID#
1
CV-^
N
100
119
112
123
ID#
1
N
100
100
1.000
10
15 95 98
Middle Oregon
1
100
1.000
2
100
1.000
2
100
1.000
3
100
0.950
0.050
3
100
1.000
4
100
0.975
0.020
0.005
4
100
0.955
0.045
5
100
0.960
0.010
0.030
5
100
1.000
6
100
0.940
0.045
0.015
6
100
1.000
S. Oregon/N. California Coastal
7
100
0.980
0.015
0.005
7
100
1.000
8
100
0.950
0.025
0.025
8
100
1.000
9
100
0.915
0.080
0.025
9
100
1.000
10
62
0.952
0.008
0.040
10
62
1.000
11
99
0.894
0.106
11
99
1.000
Klamath-Trinity Basin
12
100
0.985
0.015
12
100
0.995
0.005
13
100
0.775
0.225
13
80
0.988
0.013
14
99
0.899
0.101
14
99
1.000
15
128
0.938
0.051
0.012
15
118
1.000
16
100
0.955
0.030
0.015
16
100
1.000
17
98
0.929
0.046
0.005
0.020
17
98
1.000
18
106
0.943
0.028
0.028
18
106
1.000
19
100
0.905
0.095
19
100
1.000
20
100
0.980
0.015
0.005
20
100
1.000
21
120
0.942
0.054
0.004
21
120
1.000
Eel River-California Coastal
22
95
0.874
0.121
0.005
22
95
1.000
23
100
0.900
0.100
23
100
1.000
24
99
0.924
0.076
24
99
0.985
0.015
25
61
0.828
0.172
25
61
1.000
26
95
0.868
0.132
26
95
1.000
27
99
0.874
0.126
27
99
1.000
28
100
0.835
0.165
28
100
1.000
29
93
0.871
0.129
29
93
1.000
30
99
0.778
0.222
30
100
1.000
31
96
0.786
0.214
31
96
1.000
32
100
0.900
0.100
32
100
1.000
Sacramento— San Joaquin
33
94
0.957
0.011
0.032
33
94
1.000
34
100
0.925
0.020
0.055
34
100
1.000
35
100
0.860
0.035
0.105
35
100
1.000
36
100
0.925
0.020
0.055
36
100
1.000
37
100
0.905
0.065
0.030
37
100
1.000
EST-5
Alleles
f;pi.9
Alleles
GPI-
ID#
H
N
Alleles
100
(100)
60
(50)
135
(150)
100
(common) (*)
ID#
N
100
97
107
ID#
N
Middle Oregon
1
100
1.000
1
100
0.315
0.685
1
100
1.000
2
100
1.000
2
100
0.585
0.415
2
100
1.000
3
100
0.995
0.005
3
100
0.565
0.420
0.015
3
100
1.000
4
100
0.985
0.015
4
100
0.335
0.665
4
100
1.000
5
100
0.975
0.025
5
100
0.465
0.535
5
100
1.000
6
100
1.000
6
100
0.805
0.195
6
100
1.000
S. Oregon/
7
100
1.000
7
100
0.720
0.280
7
100
1.000
N. California Coastal
8
100
0.980
0.020
8
100
0.805
0.195
8
100
1.000
9
100
0.990
0.010
9
100
0.715
0.265
0.020
9
100
1.000
10
62
1.000
10
62
0.750
0.185
0.065
10
62
1.000
11
99
0.990
0.010
11
99
0.758
0.227
0.015
11
99
1.000
Klamath-Trinity Basin
12
100
0.995
0.005
12
100
0.765
0.235
12
100
1.000
13
60
0.967
0.033
13
100
0.615
0.385
13
100
1.000
14
99
0.980
0.020
14
99
0.949
0.051
14
99
1.000
15
58
0.991
0.009
15
128
0.945
0.055
15
128
1.000
16
90
0.083
0.017
16
100
0.945
0.055
16
80
1.000
17
98
0.995
0.005
17
98
0.888
0.112
17
98
1.000
18
106
1.000
18
106
0.769
0.231
18
106
1.000
19
100
0.985
0.015
19
100
0.915
0.085
19
100
1.000
Gall et al : Geographic variation in population genetics of Oncorhynchus tshawytscha
91
Appendix A (continued)
Alleles
Alleles
EST-'
Alleles
GPI ''
100
(100)
60
135
GPI-H
inn
*
ID#
N
100
97
107 ID#
N
(50)
(150)
ID# N (common)
(•)
Klamath-Trinity Basin
20
100
1.000
20
100
0.885
0.115
20 100 1.000
(continued)
21
120
1.000
21
120
0.929
0.071
21 120 1.000
Eel River-California Coastal
22
95
1.000
22
95
0.542
0.458
22
95 1.000
23
100
1.000
23
100
0.570
0.430
23 100 1.000
24
99
0.995
0.005
24
99
0.556
0.444
24
99 1.000
25
61
1.000
25
61
0.484
0.516
25
61 1.000
26
95
1.000
26
95
0.432
0.568
26
95 1.000
27
99
1.000
27
99
0.535
0.465
27
99 1.000
28
100
1.000
28
100
0.570
0.430
28 100 1.000
29
93
1.000
29
93
0.586
0.414
29
93 1.000
30
100
1.000
30
100
0.545
0.455
30 100 1.000
31
96
1.000
31
96
0.693
0.307
31
96 1.000
32
100
0.995
0.005 32
100
0.570
0.430
32 100 1.000
Sacramento-San Joaquin
33
92
0.989
0.011
33
94
0.777
0.064
0.160
1
33
94 0.643
0.357
34
100
0.995
0.005
34
100
0.940
0.040
0.020
1
34 100 0.717
0.283
35
100
0.995
0.005
35
100
0.925
0.065
0.010
1
35 100 0.613
0.387
36
100
1.000
36
100
0.930
0.070
36 100 0.654
0.346
37
100
1.000
37
100
0.965
0.035
37 100 0.755
0.245
Alleles
Alleles
Alleles
GR
ID#
TT AfJlJ
IDpH.l
N
100
85
ID#
N
100
100
1.000
143
78
ID#
1
N
100
100
0.950
0
0.050
Middle Oregon
1
96
1.000
1
2
100
0.895
0.105
2
100
0.980 0.015 0.005
2
99
0.712
0.288
3
97
0.943
0.057
3
100
0.985 0.015
3
99
0.864
0.136
4
99
0.975
0.025
4
100
1.000
4
100
0.710
0.290
5
80
1.000
5
100
1.000
5
99
0.934
0.066
6
100
0.995
0.005
6
100
1.000
6
100
0.995
0.005
S Oregon/
7
100
0.995
0.005
7
100
1.000
7
100
1.000
N. California Coastal
8
100
1.000
8
100
1.000
8
100
0.995
0.005
9
100
1.000
9
100
1.000
9
99
0.919
0.081
10
62
0.895
0.105
10
62
1.000
10
62
0.992
0.008
11
99
0.975
0.025
11
99
1.000
11
99
0.990
0.010
Klamath-Trinity Basin
12
100
0.995
0.005
12
100
1.000
12
100
0.990
0.010
13
100
1.000
13
100
1.000
13
100
0.995
0.005
14
99
1.000
14
99
1.000
14
92
1.000
15
128
1.000
15
98
1.000
15
128
1.000
16
100
0.995
0.005
16
100
1.000
16
100
1.000
17
98
1.000
17
98
1.000
17
95
1.000
18
106
1.000
18
106
1.000
18
106
1.000
19
100
1.000
19
100
1.000
19
100
1.000
20
100
1.000
20
100
1.000
20
100
1.000
21
120
1.000
21
120
1.000
21
120
1.000
Eel River-California Coastal
22
95
1.000
22
95
1.000
22
95
0.979
0.021
23
100
1.000
23
100
1.000
23
100
0.990
0.010
24
99
0.995
0.005
24
99
1.000
24
99
1.000
25
61
1.000
25
45
1.000
25
58
1.000
26
95
1.000
26
95
1.000
26
95
1.000
27
99
1.000
27
99
1.000
27
97
1.000
28
60
1.000
28
54
1.000
28
85
1.000
29
93
1.000
29
93
1.000
29
93
1.000
30
100
1.000
30
63
1.000
30
73
1.000
31
96
1.000
31
96
1.000
31
92
1.000
32
100
1.000
32
46
1.000
32
99
1.000
Sacramento-San Joaquin
33
94
1.000
33
94
1.000
33
93
1.000
34
100
1.000
34
100
1.000
34
100
1,000
35
100
1.000
35
100
1.000
35
100
1.000
36
100
1.000
36
100
0.990
0.010
36
100
1.000
37
100
1.000
37
100
1.000
37
100
1.000
92
Fishery Bulletin 90(1). 1992
Appendix A (continued)
Alleles
Alleles
IDDH-2
ID# JV
100
(100)
61
(50
20
IDH-2
1D# N
100
(100)
Middle Oregon
S Oregon/N. California Coastal
Klamath-Trinity Basin
Eel River-California Coastal
Sacramento-San Joaquin
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
100
99
99
100
99
100
100
100
99
61
99
100
100
92
128
100
95
104
93
100
120
95
100
99
55
95
97
83
93
73
92
99
93
100
100
100
100
1.000
0.995
0.990
0.990
0.990
0.940
0.975
0.945
0.939
0.861
0.929
0.975
0.925
0.978
0.988
0.985
0.937
0.976
0.892
0.945
1.000
0.974
0.990
0.939
0.945
0.995
0.985
0.982
0.995
1.000
1.000
0.909
0.984
0.990
0.975
0.990
0.990
0.005
0.010
0.010
0.010
0.060
0.025
0.055
0.061
0.139
0.071
0.025
0.075
0.022
0.012
0.015
0.063
0.024
0.108
0.055
0.026
0.010
0.061
0.055
0.005
0.015
0.018
0.005
0.091
0.016
0.010
0.025
0.010
0.010
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
100
100
100
100
100
100
100
100
100
62
99
100
100
99
127
100
98
106
100
100
120
95
100
99
61
95
98
100
93
100
96
100
94
100
100
100
100
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
0.995
0.995
1.000
1.000
1.000
1.000
1.000
1.000
1.000
0.975
0.974
0.990
0.990
1.000
1.000
1.000
1.000
0.941
0.905
0.950
0.830
0.885
Alleles
IDH-3
ID# N
100
(100)
74
(80)
142
94
(80)
83
129
(120)
Middle Oregon
S. Oregon/N. California Coastal
Klamath-Trinity Basin
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
100
100
100
100
100
100
100
100
100
62
99
100
100
99
124
99
98
106
100
1.000
1.000
0.985
0.995
1.000
1.000
1.000
0.990
0.995
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
0.015
0.005
0.010
154
(120)
0.005
0.005
0.025
0.026
0.010
0.010
0.059
0.095
0.050
0.170
0.115
136
0.005
Gall et al.: Geographic variation in population genetics of Oncorhynchus tshawytscha
93
Appendix A (continued)
IDH-3
Alleles
100
74
142
94
33
129
136
ID# N
(100)
(80)
(80)
(120)
Klamath-Trinity
Basin
20 100
1.000
(continued)
n 120
0.992
0.008
Eel River-California Coastal
12 95
0.995
0.005
IZ 100
1.000
24 99
1.000
25 61
1.000
26 95
1.000
27 99
1.000
28 100
1.000
29 93
1.000
30 100
1.000
SI 96
1.000
B2 100
1.000
Sacramento-San
Joaquin
?3 94
0.949 0.005
0.048
54 100
0.995
0.005
35 100
1.000
36 100
0.990
0.010
37 100
1.000
THH-'i
Alleles
.1
Alleles
I TW-A
Alleles
100
(100)
127
(120)
50 LDH
100
(100)
1.000
112
(115)
134 71
(75)
ID#
N
"1
N
100
100
1.000
800
ID#
1
N
100
Middle Oregon
1
100
0.935
0.065
1
2
100
0.995
0.005
2
100
1.000
2
100
0.985
0.015
3
100
0.975
0.025
3
100
1.000
3
100
1.000
4
100
0.970
0.030
4
100
1.000
4
100
0.990
0.010
5
100
0.950
0.050
5
100
0.900
0.100
5
100
0.985
0.015
6
100
0.930
0.070
6
100
1.000
6
100
1.000
S. Oregon/
7
100
0.975
0.025
7
100
1.000
7
100
1.000
N. California Coastal
8
100
0.945
0.055
8
100
1.000
8
100
0.980
0.010
0.010
9
100
0.975
0.025
9
100
1.000
9
100
1.000
10
62
0.879
0.121
10
62
1.000
10
62
1.000
11
99
0.985
0.015
11
99
1.000
11
99
1.000
Klamath-Trinity Basin
12
100
0.980
0.020
12
100
0.859
0.141
12
100
1.000
13
100
0.900
0.100
13
100
1.000
13
100
1.000
14
99
1.000
14
99
1.000
14
99
1.000
15
128
0.996
0.004
15
127
1.000
15
128
1.000
16
99
1.000
16
100
1.000
16
100
1.000
17
98
0.980
0.020
17
98
1.000
17
98
1.000
18
102
1.000
18
106
1.000
18
106
1.000
19
100
0.990
0.010
19
100
1.000
19
100
1.000
20
100
0.980
0.020
20
100
1.000
20
100
1.000
21
120
1.000
21
120
1.000
21
120
1.000
Eel River-California Coastal
22
95
0.868
0.132
22
95
0.897
0.103
22
95
1.000
23
100
0.845
0.155
23
100
0.900
0.100
23
100
1.000
24
99
0.899
0.101
24
99
1.000
24
99
1.000
25
61
0.885
0.115
25
61
1.000
25
61
1.000
26
95
0.900
0.100
26
95
1.000
26
95
1.000
27
99
0.859
0.141
27
99
1.000
27
99
1.000
28
100
0.865
0.135
28
100
1.000
28
100
1.000
29
93
0.785
0.215
29
93
1.000
29
93
1.000
30
100
0.810
0.190
30
100
1.000
30
100
1.000
31
96
0.859
0.141
31
96
1.000
31
96
1.000
32
100
0.765
0.235
32
100
1.000
32
100
1.000
Sacramento-San Joaquin
33
94
0.915
0.085
33
94
1.000
33
94
1.000
34
100
0.905
0.090 0.005 34
100
1.000
34
100
1.000
35
100
0.895
0.105
35
100
1.000
35
100
1.000
36
100
0.875
0.125
36
100
1.000
36
100
1.000
37
100
0.995
0.005
37
100
1.000
37
100
1.000
94
Fishery Bulletin 90(1). 1992
Appendix A (continued)
Alleles
Aliplf^s
Alleles
T nw c
\fnuii 1
MnHt*-*?
LiUtl
ID#
N
100
90
95 ID#
M.M. -X
N
100
92
ID#
N
100
78
Middle Oregon
1
100
1.000
1
100
0.260
0.740
1
100
1.000
2
100
0.970
0.030
2
100
0.375
0.625
2
100
1.000
3
100
0.975
0.025
3
100
0.470
0.530
3
100
1.000
4
100
0.990
0.010
4
100
0.325
0.675
4
100
1.000
5
100
1.000
5
100
0.380
0.620
5
100
1.000
6
100
0.995
0.005
6
100
0.465
0.535
6
100
1.000
S Oregon/
7
100
0.975
0.015
0.010 7
100
0.450
0.550
7
100
0.900
0.100
N. California Coastal
8
100
0.990
0.010
8
100
0.415
0.585
8
100
0.900
0.100
9
100
1.000
9
100
0.325
0.675
9
100
0.900
0.100
10
62
1.000
10
62
0.282
0.718
10
62
0.746
0.254
11
99
1.000
11
98
0.362
0.638
11
98
1.000
Klamath-Trinity Basin
12
100
0.985
0.015
12
100
0.315
0.685
12
100
0.859
0.141
13
100
0.890
0.110
13
100
0.390
0.610
13
100
1.000
14
99
1.000
14
99
0.247
0.753
14
99
0.598
0.402
15
127
1.000
15
123
0.228
0.772
15
123
0.558
0.442
16
100
1.000
16
99
0.212
0.788
16
99
0.562
0.438
17
98
1.000
17
98
0.245
0.755
17
98
0.622
0.378
18
106
1.000
18
105
0.333
0.667
18
105
0.564
0.436
19
100
1.000
19
100
0.465
0.535
19
100
0.827
0.173
20
100
0.975
0.025
20
100
0.330
0.670
20
100
0.859
0.141
21
120
1.000
21
120
0.150
0.850
21
120
0.726
0.274
Eel River-California Coastal
22
95
1.000
22
95
0.374
0.626
22
95
1.000
23
100
1.000
23
100
0.460
0.540
23
100
1.000
24
99
1.000
24
99
0.470
0.530
24
99
1.000
25
61
1.000
25
60
0.450
0.550
25
60
1.000
26
95
1.000
26
95
0.532
0.468
26
95
1.000
27
99
1.000
27
79
0.557
0.443
27
79
0.841
0.159
28
100
1.000
28
100
0.480
0.520
28
100
0.900
0.100
29
93
1.000
29
93
0.505
0.495
29
93
1.000
30
100
1.000
30
100
0.425
0.575
30
100
1.000
31
96
1.000
31
96
0.500
0.500
31
96
1.000
32
100
1.000
32
100
0.400
0.600
32
100
1.000
Sacramento-San Joaquin
33
94
1.000
33
94
0.851
0.149
33
94
1.000
34
100
1.000
34
100
0.805
0.195
34
100
1.000
35
100
1.000
35
100
0.775
0.225
35
100
1.000
36
100
1.000
36
100
0.810
0.190
36
100
1.000
37
100
1.000
37
100
0.860
0.140
37
100
1.000
MDH-'
Alleles
Mnv-'i
Alleles
100
(100)
1.000
121 70 i26
(120) (70)
ID#
N
100
120 27
45
111 UKr
ID#
1
N
100
Middle Oregon
1
100
1.000
2
100
1.000
2
100
0.980
0.020
3
100
0.995
0.005
3
100
0.995
0.005
4
100
0.995
0.005
4
100
0.995
0.005
5
100
0.880
0.075
0.045
5
100
0.980
0.020
6
100
1.000
6
100
0.935
0.065
S Oregon/N. California Coastal
7
100
1.000
7
100
1.000
8
100
1.000
8
100
0.975
0.025
9
100
0.990
0.005 0.005
9
100
0.950
0.045
0.005
10
62
1.000
10
62
1.000
11
99
1.000
11
99
0.975
0.015 0.010
Klamath-Trinity Basin
12
100
1.000
12
100
0.985
0.015
13
100
1.000
13
100
1.000
14
99
1.000
14
99
1.000
15
128
1.000
15
128
1.000
16
100
0.995
0.005
16
100
1.000
17
98
1.000
17
98
1.000
18
106
1.000
18
106
1.000
19
100
1.000
19
100
1.000
Gall et aL: Geographic variation in population genetics of Oncorhynchus tshawytscha
95
Appendix A (continued)
Alleles
MriR ■>
Alleles
MDH-'i
100
121 70 l?-fi
ID#
N
100
120 27
45
ID#
20
N
100
(100)
0.995
(120) (7
0.005
0)
Klamath-Trinity Basin
20
100
1.000
(continued)
21
120
1.000
21
120
1.000
Eel River-California Coastal
22
95
1.000
22
95
0.995
0.005
23
100
0.995
0.005
23
100
0.985
0.015
24
99
1.000
24
99
1.000
25
61
1.000
25
61
1.000
26
95
1.000
26
95
1.000
27
99
1.000
27
99
1.000
28
100
1.000
28
100
1.000
29
93
1.000
29
93
1.000
30
100
1.000
30
100
1.000
31
96
1.000
31
96
1.000
32
100
1.000
32
100
1.000
Sacramento-San Joaquin
33
94
1.000
33
94
0.979
0.021
34
100
1.000
34
100
0.920
0.070
0.010
35
100
1.000
35
100
0.955
0.045
36
100
1.000
36
100
0.905
0.065
0.030
37
100
1.000
37
100
0.935
0.040 0.025
»«Mrm 1
Alleles
mMDM-O
Alleles
MPI
Alleles
100
109
mML
ID#
FXl-Jl
N
-100
-900
ini¥iL
ID#
N
100
100
1.000
200
ID#
1
N
99
(100)
0.581
(110)
Middle Oregon
1
100
1.000
1
0.419
2
100
0.980
0.020
2
100
0.995
0.005
2
100
0.695
0.305
3
100
0.990
0.010
3
100
1.000
3
100
0.575
0.425
4
100
0.995
0.005
4
100
1.000
4
100
0.505
0.495
5
100
0.960
0.040
5
100
1.000
5
100
0.690
0.310
6
100
0.915
0.085
6
100
0.995
0.005
6
100
0.900
0.100
S. Oregon/
7
100
0.940
0.060
7
100
1.000
7
100
0.890
0.110
N. California Coastal
8
100
0.940
0.060
8
100
0.995
0.005
8
99
0.828
0.172
9
100
0.865
0.135
9
80
1.000
9
100
0.660
0.340
10
62
0.960
0.040
10
62
1.000
10
62
0.815
0.185
11
99
0.899
0.101
11
99
1.000
11
99
0.818
0.182
Klamath-Trinity Basin
12
100
0.910
0.090
12
100
0.995
0.005
12
100
0.860
0.140
13
100
0.795
0.205
13
100
1.000
13
100
0.860
0.140
14
99
1.000
14
99
1.000
14
99
0.970
0.030
15
128
0.996
0.004
15
80
1.000
15
128
1.000
16
60
1.000
16
60
1.000
16
100
1.000
17
98
0.990
0.010
17
98
0.995
0.005
17
98
0.959
0.041
18
70
1.000
18
106
1.000
18
106
0.953
0.047
19
100
0.990
0.010
19
100
0.905
0.095
19
100
0.940
0.060
20
100
0.970
0.030
20
100
1.000
20
100
0.975
0.025
21
120
1.000
21
120
1.000
21
120
0.992
0.008
Eel River-California Coastal
22
95
0.995
0.005
22
95
0.995
0.005
22
95
0.805
0.195
23
100
0.990
0.010
23
100
1.000
23
100
0.765
0.235
24
99
0.995
0.005
24
99
1.000
24
99
0.904
0.096
25
61
1.000
25
61
1.000
25
61
0.787
0.213
26
95
0.989
0.011
26
95
1.000
26
95
0.853
0.147
27
99
1.000
27
99
1.000
27
99
0.818
0.182
28
100
1.000
28
73
1.000
28
99
0.808
0.192
29
93
1.000
29
93
1.000
29
93
0.785
0.215
30
100
1.000
30
100
1.000
30
100
0.800
0.200
31
96
1.000
31
96
1.000
31
96
0.901
0.099
32
100
1.000
32
100
1.000
32
100
0.610
0.390
Sacramento-San Joaquin
33
94
1.000
33
94
1.000
33
94
0.617
0.383
34
100
1.000
34
100
1.000
34
100
0.585
0.415
35
100
1.000
35
100
1.000
35
100
0.580
0.420
36
100
1.000
36
100
1.000
36
100
0.545
0.455
37
100
1.000
37
100
1.000
37
100
0.700
0.300
96
Fishery Bulletin 90(1), 1992
Appendix A (continued)
PGDW
Alleles
vaK.")
Alleles
PGM-'
Alleles
100
90
1 HF.
100
90
ID#
N
(100)
(90) (90) ID#
N
(100)
(90)
ID#
N
100 210
50
Middle Oregon
1
100
1.000
1
100
0.660
0.340
1
100
0.855 0.065
0.080
2
100
1.000
2
100
0.445
0.555
2
100
0.870 0.070
0.060
3
100
1.000
3
100
0.435
0.565
3
100
0.910 0.070
0.020
4
100
1.000
4
100
0.355
0.645
4
100
0.870 0.090
0.040
5
100
1.000
5
100
0.465
0.535
5
100
0.880 0.090
0.030
6
100
1.000
6
100
0.430
0.570
6
60
1.000
S. Oregon/
7
100
1.000
7
100
0.395
0.605
7
100
1.000
N. California Coastal
8
100
0.985
0.015 8
100
0.345
0.655
8
100
1.000
9
100
0.990
0.010 9
100
0.515
0.485
9
100
0.980 0.020
10
62
1.000
10
62
0.468
0.532
10
62
1.000
11
99
1.000
11
98
0.439
0.561
11
99
1.000
Klamath-Trinity Basin
12
100
1.000
12
100
0.400
0.600
12
80
1.000
13
100
0.910
0.090 13
100
0.380
0.620
13
100
1.000
14
99
1.000
14
99
0.146
0.854
14
99
1.000
15
128
0.996
0.004
15
127
0.185
0.815
15
128
1.000
16
100
1.000
16
100
0.155
0.845
16
100
1.000
17
98
1.000
17
98
0.189
0.811
17
98
1.000
18
106
1.000
18
105
0.186
0.814
18
106
1.000
19
100
1.000
19
100
0.380
0.620
19
100
0.950
0.050
20
100
1.000
20
100
0.320
0.680
20
100
1.000
21
120
1.000
21
120
0.292
0.708
21
120
1.000
Eel River-California Coastal
22
95
1.000
22
95
0.379
0.621
22
95
1.000
23
100
1.000
23
100
0.345
0.655
23
80
0.994 0.006
24
99
1.000
24
99
0.525
0.475
24
99
1.000
25
61
1.000
25
61
0.459
0.541
25
61
1.000
26
95
1.000
26
95
0.242
0.758
26
95
1.000
27
99
1.000
27
99
0.480
0.520
27
99
1.000
28
100
1.000
28
99
0.439
0.561
28
100
1.000
29
93
1.000
29
93
0.392
0.608
29
93
1.000
30
100
1.000
30
100
0.245
0.755
30
100
1.000
31
96
1.000
31
96
0.365
0.635
31
96
1.000
32
100
1.000
32
100
0.315
0.685
32
100
1.000
Sacramento-San Joaquin
33
94
0.979
0.021
33
94
0.590
0.410
33
94
1.000
34
100
0.975
0.025
34
100
0.495
0.505
34
100
1.000
35
100
0.960
0.040
35
100
0.490
0.510
35
100
1.000
36
100
0.920
0.080
36
100
0.605
0.395
36
100
1.000
37
100
0.900
0.100
37
100
0.670
0.330
37
100
1.000
PGM-''
Alleles
-
PGM-3
ID#
1
Alleles
100
(100)
166
(166)
144
120
ID#
N
-
N
100
94
0.290
Middle Oregon
1
100
1.000
100
0.710
2
100
1.000
2
100
0.945
0.055
3
100
0.970
0.030
1
3
100
0.885
0.115
4
100
0.975
0.025
4
100
0.925
0.075
5
100
1.000
5
100
0.900
0.100
6
100
1.000
6
100
0.945
0.055
S. Oregon/N. California Coastal
7
100
0.995
0.005
7
100
0.970
0.030
8
100
1.000
8
100
0.970
0.030
9
100
0.965
0.030
0.005
9
100
0.950
0.050
10
62
0.927
0.073
10
62
0.968
0.032
11
99
0.995
0.005
11
99
0.934
0.066
Klamath-Trinity Basin
12
100
0.915
0.085
12
100
0.945
0.055
13
100
0.975
0.025
13
100
0.930
0.070
14
99
0.929
0.071
14
99
0.980
0.020
15
128
0.902
0.098
15
114
0.987
0.013
16
100
0.965
0.035
16
98
0.964
0.036
17
98
0.964
0.036
17
98
0.923
0.077
18
106
1.000
18
106
0.981
0.019
19
100
0.860
0.135
0.005
19
100
0.970
0.030
Gall et al,: Geographic variation in population genetics of Oncorhynchus tshawytscha 97
Appendix A (continued)
Alleles
Alleles
PGM-2
100
166
144 120
PGM-3
ID#
N
(100)
(166)
ID#
20
N
100
100
0.95C
94
1 0.050
Klamath-Trinity
Basin
20
100
1.000
(continued)
21
120
1.000
21
120
0.90C
1 0.100
Eel River-California Coastal
22
95
1.000
22
95
0.984
1 0.016
23
100
1.000
23
100
0.96E
) 0.035
24
99
0.970
0.025
0.005
24
99
0.99E
i 0.005
25
61
0.967
0.033
25
61
l.OOC
26
95
1.000
26
95
l.OOC
27
99
1.000
27
99
l.OOC
28
100
1.000
28
100
l.OOC
29
93
1.000
29
93
l.OOC
30
100
1.000
30
100
1.000
31
96
1.000
31
96
1.000
32
100
0.995
0.005
32
100
1.000
Sacramento-San
Joaquin
33
94
0.995
0.005
33
94
0.995 0.005
34
100
0.990
0.010
34
100
0.970 0.030
35
100
0.995
0.005
35
100
0.970 0.030
36
100
1.000
36
100
0.980 0.020
37
100
1.000
37
100
0.975 0.025
pr:vr^
Alleles
SOD-i
Alleles
- 100
-260
580
1260
ID#
N
100
94
108
88
90 97
ID#
1
N
99
(-100) 1
0.788
(-260)
0.202
(580)
0.010
Middle Oregon
1
100
0.100
0.520
0.015
0.285
0.080
2
100
0.325
0.565
0.050
0.010
0.035 0.015
2
100
0.770
0.230
3
100
0.330
0.610
0.030
0.005
0.020 0.005
3
100
0.765
0.230
0.005
4
100
0.385
0.540
0.055
0.005
0.015
4
100
0.785
0.215
5
100
0.265
0.675
0.030
0.030
5
100
0.570
0.430
6
100
0.505
0.430
0.055
0.010
6
100
0.715
0.270
0.015
S. Oregon/
7
100
0.505
0.435
0.060
7
100
0.730
0.255
0.005
0.010
N. California Coastal
8
100
0.535
0.415
0.045
0.005
8
100
0.780
0.210
0.010
9
100
0.370
0.630
9
100
0.810
0.190
10
62
0.315
0.685
10
62
0.782
0.218
11
98
0.464
0.536
11
98
0.760
0.240
Klamath-Trinity Basin
12
100
0.490
0.495
0.015
12
100
0.755
0.230
0.015
13
100
0.565
0.435
13
100
0.815
0.185
14
99
0.586
0.414
14
99
0.990
0.010
15
114
0.667
0.333
15
128
1.000
16
98
0.592
0.408
16
100
1.000
17
98
0.495
0.505
17
94
0.968
0.027
0.005
18
106
0.528
0.472
18
105
0.852
0.148
19
100
0.665
0.290
0.045
19
99
0.904
0.010
0.086
20
100
0.505
0.495
20
100
0.845
0.090
0.060
0.005
21
120
0.363
0.638
21
120
0.917
0.046
0.021
0.017
Eel River-California Coastal 22
95
0.726
0.268
0.005
22
92
0.750
0.250
23
100
0.675
0.325
23
100
0.635
0.365
24
99
0.763
0.227
0.010
24
99
0.798
0.202
25
61
0.877
0.115
0.008
25
59
0.636
0.364
26
95
0.753
0.247
26
95
0.700
0.300
27
99
0.813
0.187
27
99
0.778
0.222
28
100
0.800
0.200
28
87
0.793
0.207
29
93
0.892
0.108
29
92
0.837
0.163
30
100
0.855
0.145
30
99
0.798
0.202
31
96
0.760
0.240
31
91
0.714
0.286
32
100
0.880
0.120
32
100
0.715
0.270
0.015
Sacramento- San Joaquin
33
94
0.500
0.495
0.005
33
93
0.661
0.339
34
100
0.555
0.415
0.005
0.025
34
100
0.790
0.210
35
100
0.575
0.335
0.090
35
100
0.755
0.240
0.005
36
100
0.550
0.435
0.015
36
100
0.690
0.300
0.010
37
100
0.605
0.375
0.005
0.015
37
100
0.715
0.270
0.015
98
Fishery Bulletin 90(1). 1992
Appendix A (continued)
Alleles
TPT o
Alleles
TPI.;!
Alleles
nPFP-i
100
(100)
90
(90)
ID#
N
100
106
104
1 ri"
ID#
N
100
104
102
101
U± Ml
ID#
N
Middle Oregon
1
99
0.783
0.217
1
100
1.000
1
100
0,715
0.285
2
100
0.970
0.030
2
100
1.000
2
100
0,595
0.405
3
100
0.960
0.040
3
100
1.000
3
100
0,660
0.340
4
100
0.905
0.095
4
100
1.000
4
100
0,630
0.370
5
100
0.890
0.110
5
100
1.000
5
100
0,715
0,285
6
100
0.950
0.050
6
100
0.995
0.005
6
100
0.920
0.080
S. Oregon/
7
100
0.920
0.080
7
100
0.995
0.005
7
100
0.925
0.075
N. California Coastal
8
100
0.890
0.110
8
100
1.000
8
100
0.905
0.095
9
100
0.840
0.160
9
100
0.975
0.025
9
100
0.810
0.190
10
62
0.903
0.097
10
62
1.000
10
62
0,871
0.129
11
99
0.753
0.101
0.146
11
99
0.970
0.030
11
99
0,848
0,152
Klamath-Trinity Basin
12
100
0.865
0.135
12
100
1.000
12
100
0,895
0,105
13
100
0.965
0.035
13
100
1.000
13
100
0.770
0,230
14
99
0.970
0.030
14
99
1.000
14
99
0.990
0.010
15
128
1.000
15
128
1.000
15
128
1.000
16
100
1.000
16
100
1.000
16
100
1.000
17
98
0.964
0.036
17
98
0.995
0.005
17
98
0.964
0.036
18
106
0.967
0,033
18
106
1.000
18
105
0.824
0.176
19
100
0.940
0.060
19
100
1.000
19
100
0.940
0.060
20
100
0.970
0.030
20
100
1.000
20
100
0.930
0.070
21
120
0.979
0.021
21
120
1.000
21
120
1.000
Eel River-California Coastal
22
95
0.984
0.016
22
95
0.989
0.005
0.005
22
95
0.942
0.058
23
100
0.960
0.040
23
100
0.995
0.005
23
100
0.950
0.050
24
99
1.000
24
99
0.995
0,005
24
99
0.965
0.035
25
61
0.714
0.286
25
61
0.959
0.041
25
61
0.967
0.033
26
95
1.000
26
95
0.968
0.032
26
95
0.963
0.037
27
99
0.899
0.101
27
99
0.975
0.025
27
99
0.965
0.035
28
100
0.859
0.141
28
100
0.975
0.025
28
100
0.955
0.045
29
93
0.805
0.147
0.048
29
93
0.892
0.086
0.022
29
93
0.957
0.043
30
100
1.000
30
100
0.875
0.125
30
95
0.937
0.063
31
96
0.995
0.005
31
96
0.974
0.026
31
96
0,880
0.120
32
100
0.895
0.100
0.005
32
100
0.955
0.045
32
100
1.000
Sacramento-San Joaquin
33
94
0.936
0.064
33
94
1.000
33
94
0.894
0.106
34
100
0.945
0.055
34
100
0.930
0.070
34
100
0,810
0.190
35
100
0.915
0.085
35
100
0.960
0.040
35
100
0,850
0.150
36
100
0.870
0.130
36
100
0.960
0.035
0.005
36
100
0,875
0.125
37
100
0.830
0.170
37
100
0.965
0.035
37
100
0.950
0.050
pnpii"D-o
Alleles
PPPI T
Alleles
TAPPP-I
Alleles
100
(100)
107
(107)
83
100
(100)
0.725
130
(140)
0,275
ID#
N
ID#
1
N
100
100
1.000
110
1 I\X
ID#
1
N
Middle Oregon
1
100
0.995
0.005
100
2
100
0.955
0.045
2
100
1.000
2
100
0,865
0.135
3
100
0.980
0.020
3
100
1.000
3
100
0,880
0.120
4
100
0.995
0.005
4
100
1.000
4
100
0,945
0.055
5
100
1.000
5
100
1.000
5
100
0,895
0.105
6
100
0.990
0.010
6
100
0.945
0.055
6
100
0.950
0.050
S. Oregon/
7
100
0.990
0.005
0.005
7
100
0.965
0,035
7
100
0.925
0.075
N. California Coastal
8
100
0.995
0.005
8
100
0.995
0,005
8
100
0.975
0.025
9
100
1.000
9
100
1.000
9
100
0.940
0.060
10
62
1.000
10
62
1.000
10
62
0.847
0.153
11
99
1.000
11
99
1.000
11
99
0.955
0.045
Klamath-Trinity Basin
12
100
1.000
12
100
0.985
0.015
12
100
0,925
0.075
13
100
1.000
13
100
1.000
13
100
0,860
0.140
14
99
1.000
14
99
1.000
14
99
1.000
15
128
1.000
15
128
1.000
15
125
0,996
0.004
16
100
1.000
16
100
1.000
16
100
1.000
17
98
1.000
17
98
1.000
17
98
1.000
18
106
1.000
18
106
1.000
18
106
1.000
Gall et al.: Geographic variation in population genetics of Oncorhynchus tshawytscha
99
Appendix A (continued)
Alleles
Alleles
pnpi^i> •>
100
(100)
107 8^ PTTTJT T
Alleles
TAPl^P-1
100
(100)
1.000
130
(140)
ID#
N
(107)
ID#
N
100
100 110
1.000
ID#
19
N
100
Klamath-Trinity Basin
19
100
1.000
19
(continued)
20
100
1.000
20
60
1.000
20
100
0.980
0.020
21
120
1.000
21
120
1.000
21
120
1.000
Eel River-California Coastal
22
95
1.000
22
95
1.000
22
95
0.974
0.026
23
100
1.000
23
100
1.000
23
100
0.960
0.040
24
99
1.000
24
99
1.000
24
99
0.985
0.015
25
61
1.000
25
61
1.000
25
61
0.992
0.008
26
95
1.000
26
95
1.000
26
95
0.979
0.021
27
98
1.000
27
60
1.000
27
99
0.965
0.035
28
100
0.995
0.005
28
100
1.000
28
100
0.995
0.005
29
93
1.000
29
93
1.000
29
93
1.000
30
100
1.000
30
100
1.000
30
100
0.990
0.010
31
96
1.000
31
96
1.000
31
96
1.000
32
100
1.000
32
100
1.000
32
100
1.000
Sacramento-San Joaquin
33
94
1.000
33
94
1.000
33
94
0.862
0.138
34
100
0.995
0.005
34
100
1.000
34
100
0.890
0.110
35
100
1.000
35
100
1.000
35
100
0.950
0.050
36
100
0.990
0.010
36
100
1.000
36
100
0.940
0.060
37
100
1.000
37
100
1.000
37
100
0.955
0.045
Appendix B
Recently discovered allozyme variability
Two monomeric mitochondrial loci of aconitate hydra-
tase, mAH-1 and mAH-4, are polymorphic in chinook
salmon. The rnAh-l(65) allele was observed primarily
in coastal California samples, although it is also pres-
ent in the Sacramento system. Three alleles at mAH-4
were important in differentiating coastal and inland
samples. Shaklee et al. (Wash. Dep. Fish., Olympia,
WA 98504, pers. commun., Feb 1991) have recently
performed breeding studies which confirmed the
Mendelian model of inheritance for these loci.
Iditol dehydrogenase is coded by two loci in liver
tissue. The enzyme is a tetramer for which both loci
are assumed to be polymorphic. Variants were assigned
to a particular locus based on relative staining inten-
sities. The Iddh-l(O) allele was observed in Oregon and
coastal northern California populations. The Iddh-2(61)
allele was observed throughout the study area except
in samples from the Sacramento system, whereas the
Iddh-2(20) allele was only observed in the Sacramento
samples.
Variation in NADP-dependent malate dehydrogenase
was expressed at two cytosolic loci using chinook
salmon muscle and heart tissue. MDHP-2 is also ex-
pressed in liver and eye tissue in juvenile fish. MDHP-1
variation has been described by Shaklee et al. (1990b).
Due to the low levels of variability found in the
Klamath-Trinity system, these MDHP loci wil be ex-
tremely important in the identification of fish from this
area. The Mdhp-2(78) allele has nearly the same mobil-
ity as the Mdhp-l(lOO) allele, thus making identifica-
tion of heterozygous samples difficult.
A duplicated and highly polymorphic monomeric
PGM locus was designated by two loci, PGM-3 and
PGM-4. These isoloci present particular difficulties
when estimating allele and genotypic frequencies
(Robin Waples and Paul Aebersold, NMFS Northwest
Fish Sci. Cent., Seattle, WA 98115, pers. commun.,
June 1990). Six alleles have been identified in this
system and several individuals with three and four dif-
ferent alleles were observed. Therefore, standards are
required for correct analysis of banding patterns.
Similar expressions of variants are seen in both liver
and eye tissues. Conformance to Hardy-Weinberg pro-
portions at these loci has been found using goodness-
of-fit tests of expected and observed genotypes (Waples
and Aebersold, pers. commun.) and a protocol for
estimating allele frequencies from isoloci was presented
by Waples (1988).
Triosphosphate isomerase is coded by four loci in
Chinook salmon. The products of TPI-1 and TPI-2
migrate cathodally, and those of TPI-3 and TPI-4
migrate anodally. Two variant alleles, Tpi-3(10A) and
Tpi-3(106), were observed from eye tissue, and TPI-4
variation has been described by Shaklee (pers. com-
mun.). Because Tpi-3(106) migrates close to Tpi-MlOO),
only fish homozygous for the Tpi-3(106) allele can be
100
Fishery Bulletin 90(1), 1992
reliably scored. The Tpi-3(106) allele was observed in
California coastal samples and samples from the Eel
River.
The newly discovered alleles, Ldh-1(800), Mpdh-2(78),
and Tpi-3(106), could be visualized only in their
homozygous form. If these alleles occur at low frequen-
cies in samples of chinook salmon, they may not be
detected because of the low probabOity of sampling the
rare homozygote. This may account for the discon-
tinuous distribution observed for some of these alleles
(Appendix A). Consequently, Ldh-1(800) may be pres-
ent at low frequency in more than just four samples.
Abstract.- The potential annual
fecundity of Dover sole becomes fixed
before the spawning season when the
average diameter of the advanced
stock of yolked oocytes exceeds 0.86
mm; hence potential annual fecun-
dity is determinate. More central
California females had atretic ad-
vanced oocytes than Oregon females,
but rates of atresia were not suffi-
ciently high to have an important ef-
fect on the potential annual fecundity
of the population. A 1-kg female ma-
tured about 83,000 advanced yolked
oocytes at the beginning pf the sea-
son. Vitellogenesis continued for the
advanced yolked oocytes during most
of the spawning season whUe batches
were repetitively matured and
spawned. About nine batches were
spawned over a six-month spawning
season (December-May), and spawn-
ing ceased when the standing stock
of advanced oocytes was exhausted.
A 1-kg female released about 10,000
eggs per spawning, except for the
first and last batches which were
smaller than the rest. Near the end
of the season, females may spawn
more frequently than earlier in the
year, increasing the daily production
of eggs by the population even
though fewer females are reproduc-
tively active. Annual reproductive ef-
fort of Dover sole was equivalent to
about 14% of body wet weight per
year. Fifty percent of the females
had become sexually mature when
they reached 332 mm total length.
Various methodological issues were
also treated in this paper, including
validation of key assumptions under-
lying estimates of annual fecundity;
fecundity sample-size requirements;
evaluation of criteria and bias in
estimating female sexual maturity;
and comparisons of classification by
histology and gross anatomy.
Fecundity, spawning, and
maturity of female Dover sole
Microstomus pacificus, with an
evaluation of assumptions
and precision
J. Roe Hunter
Beverly J. Macewjcz
N. Chyan-huei Lo
Carol A. Kimbrell
Southwest Fisheries Science Center, National Marine Fisheries Service, NOAA
P,0. Box 271, La Jolla, California 92038
Manuscript accepted 15 January 1992.
Fishery Bulletin, U.S. 90:101-128 (1992).
Fecundity and sexual maturity esti-
mates are staples of fishery science.
Inevitably, they will be estimated for
every species of economic conse-
quence because of their importance
in the dynamics of the population. A
second reason for studying fecundity
is that when fecundity estimates are
combined with estimates of the abun-
dance of eggs in the sea, they can be
used to estimate the biomass of a
stock. Our laboratory is currently
evaluating such ichthyoplankton me-
thods for estimating the biomass of
Dover sole Microstomus pacificus, a
large demersal resource occurring
along the upper continental slope of
the west coast of North America. The
fecundity of Dover sole from Oregon
has been estimated (Yoklavich and
Pikitch 1989), but no estimate exists
for the segment of the stock living in
central California waters, nor have
the assumptions underlying fecun-
dity and sexual maturity assessments
been studied with the thoroughness
necessary for accurate estimates of
adult biomass. Thorough analysis of
these assumptions is usually lacking
in the fecundity literature.
Our objectives were to describe the
reproduction of Dover sole off central
California and Oregon, and evaluate
the assumptions underlying fecun-
dity and sexual maturity estimates.
We describe changes in the reproduc-
tive state of female Dover sole dur-
ing the spawning season, estimate
annual fecundity, batch fecundity,
rates of atresia, annual rates of
spawning, and length at 50% mature
(ML,5o).
Evaluation of the assumptions under-
lying annual fecundity estimates re-
quires defining six fecundity terms,
and those underlying maturity esti-
mates require defining four terms for
reproductive state.
Fecundity
Annual fecundity Total number of
eggs spawned by a female per year.
Total fecundity Standing stock of
advanced yolked oocytes.
Potential annual fecundity Total
advanced yolked oocytes matured per
year, uncorrected for atretic losses.
In species with determinate fecun-
dity, potential annual fecundity is
considered to be equivalent to the
total fecundity prior to the onset of
spawning.
Determinate fecundity Annual
fecundity is determinate when the
potential annual fecundity becomes
fixed prior to the onset of spawning.
In fishes with determinate fecundity,
total fecundity decreases with each
10!
102
Fishery Bulletin 90(1), 1992
spawning because the standing stocl< of advanced
yolked oocytes is not replaced during the spawning
season.
Indeterminate annual fecundity Annual fecundity
is indeterminate when the potential annual fecundity
of a female is not fixed prior to the onset of spawning
and unyolked oocytes continue to be matured and
spawned during the spawning season.
Batch fecundity Number of hydrated oocytes re-
leased in one spawning; usually determined by count-
ing the number of hydrated oocytes in the ovary.
Relative fecundity Fecundity divided by female
weight.
Reproductive states
Active Females capable of spawning at the time of
capture or in the near future (by the end of the survey
or of a season, or other temporal end point). Ovaries
of active females contain sufficient number of yolked
oocytes for a spawning.
Inactive Females not capable of spawning at the time
of capture nor in the near future, although some may
have been mature in the past.
IViature Females that have spawned in the current
reproductive season or can be expected to do so.
Immature Females that have not spawned in the ciu--
rent reproductive season nor can be expected to do so.
The central methodological issue in fishes with deter-
minate fecundity (Hunter and Macewicz 1985a, Hor-
wood and Greer Walker 1990) is to establish that poten-
tial annual fecundity is an unbiased estimate of annual
fecundity. For this to be true in Dover sole requires
four key assumptions. The first and most important
assumption is that fecundity is determinate in Dover
sole. This means that potential annual fecundity
becomes fixed before spawning begins. Estimation of
the standing stock of advanced oocytes (total fecundity)
is meaningless if, during the spawning season, oocytes
are added to that stock.
The second assumption is that the potential annual
fecundity is equivalent to annual fecundity. Strictly
speaking, this probably never happens because in any
fish population some of the females resorb some of their
advanced yolked oocytes rather than spawn them, a
process known as atresia. If many females resorbed
many of their advanced oocytes, potential annual fecun-
dity would be a serious overestimate of annual fecun-
dity in the population. In addition, not all ovulated
oocytes are spawned; a few remain in the ovigerous
folds of the ovary after spawning and are later re-
sorbed. Retention of ovulated oocytes is probably
seldom a serious bias.
The third assumption is that the females used to
estimate potential annual fecundity have not spawned
during the current reproductive season. Dover sole
females that have spawned some of their stock of ad-
vanced oocjftes cannot always be distinguished from
those that have not begun spawning. Inclusion of par-
tially spawned females in an estimate of potential
annual fecundity of the population could be a signifi-
cant bias.
The fourth assumption is that one is able to identify
with certainty the oocytes that constitute the poten-
tial annual fecundity. An ovary may not be sufficient-
ly developed to identify all of the oocytes destined to
be spawned. On the other hand, if the ovary is highly
advanced, spawning may have begun and some ad-
vanced oocytes lost. Clearly, an optimal range of
ovarian development exists where these risks are
minimized.
In addition to evaluating the above four assumptions
(determinate fecundity, atresia, spawning, and im-
maturity) we consider several other methodological
issues related to assessment of fecundity and female
sexual maturity. These issues are (1) validation of our
gross anatomical and histological classification of
ovaries into active or inactive and mature or immature
states; (2) four precision issues related to total fecun-
dity estimates (number of tissue samples per ovary,
number of females, location of ovarian tissue samples,
and within-trawl and between-trawl variance); and
(3) an evaluation of bias in the assessment of female
sexual maturity.
Methods
Collections and shipboard measurements
Dover sole were collected along the central California
coast (Point Conception to San Francisco Bay) during
six research trawl cruises (Table 1). Dover sole were
taken off the Oregon coast between Cape Lookout and
Heceta Head during two cruises in 1988-89; miscel-
laneous collections provided by E. Pikitch off the
Oregon coast in 1985 and 1986 were also used. Re-
search trawls were one-half hour or one hour long,
depending on depth. In central California waters, we
used a 400-mesh Eastern trawl (mouth opening ~15m
wide and 1.5 m high; Wathne 1977). In Oregon waters,
either an Alaska Fisheries Science Center (AFSC)
modified 5-inch mesh, 90/120, high-rise "poly
Nor'Eastern" trawl (fishing dimensions ~4.6m high
and 13.5 m wide at wing tips), a 5-inch mesh, 92/83, poly
Nor'Eastern trawl, or a 5V2-inch mesh, 75/90, high-rise
Aberdeen trawl was used. Up to 100 Dover sole from
Hunter et al : Fecundity, spawning, and maturity of Microstomus paaficus
103
Table 1
Sources of reproductive data on female Dover sole Microstomus pacificus. Number of specimens in three levels of ovarian analysis
and number of level-3 females with batch fecundity estimates.
Sampling protocol
Levels of ovarian analysis**
Batch
fecundity
Date
(Begin/End)
State
No. positive
trawl collections
Selection
of females*
1
2
3 Total
(no. females)
3 Dec 85
12 Dec 85
CA
11
A
39
65
104
4 Nov 85
14 Dec 85
OR
4
Unknown
_
73
73
6 Feb 86
7 Feb 86
OR
3
Unknown
37
37
3 May 86
OR
2
Unknown
-
27
-
34
7
5 Mar 86
7 Mar 86
CA
8
A
1
135
3
139
2 May 86
4 May 86
CA
3
A
59
1
60
11 Jan 87
24 Jan 87
CA
27
B
45
387
103
535
5 Feb 87
15 Feb 87
CA
22
B
14
391
92
500
3
23 Feb 88
9 Apr 88
CA
51
C
1716
120
62
1941
43
28 Nov 88
14 Dec 88
OR
53
C
667
620
152
1439
21 Feb 89
31 Mar 89
OR
21
C
104
151
34
292
3
All Oregon
83
771
908
186
1875
10
All California
122
1776
1131
326
3279
46***
Oregon + California
205 2547 2039 512 5154 56"*
3th females and males until 25 females were collected,
length (A^ 5) in < 275 mm class, 10 in 275-424 mm class, and 10 in > 425 mm.
100 fish (either females or males).
Level 2 = histological with anatomical; Level 3 = total fecundity with anatomical and histological.
oocytes provided by W.W. Wakefield were included in estimate of batch fecundity.
* A = Random selection of b
B = Selection stratified by
C = Random selection of <
* 'Level 1 = gross anatomical;
***Five females with hydrated
each trawl haul were measured (total length) to the
nearest millimeter, sexed, and their gonads classified;
some females immediately after capture were also in-
dividually weighed to the nearest gram and their
ovaries preserved in 10% neutral buffered formalin.
Females selected for ovarian preservation were either
taken randomly from the trawl catch or selected by
length according to a quota for each of three length
classes (<275mm, 275-424 mm, and > 425 mm) (see
Table 1). The preserved ovaries were used to validate
our shipboard classification of ovaries, to estimate
fecundity, and to provide material for histological
descriptions.
Gross aneitomical classification of ovaries
Ovaries that were examined onboard the ship were
assigned to one of three classes: no yolked oocytes
present; yolked oocytes present; and translucent
hydrated oocytes present. Ovaries with hydrated
oocytes or other yolked oocytes were considered to be
in the active state, while those ovaries in which
observers saw no yolked oocytes were considered to
be in the inactive state. This simple system based on
gross anatomical examination of the ovary is more
germane for biomass estimation work than are the
more complicated systems which involve many more
reproductive stages: for example, the seven-stage scale
of Hjort (1910), or the five-stage scale of Hagerman
104
Fishery Bulletin 90(1). 1992
(1952). Eighty percent of the females that we classified
using gross anatomical criteria were also classified as
active or inactive using histological criteria. The results
were compared to determine the accuracy of identify-
ing active and inactive females by gross anatomical
classification.
ALL FEMALES
Before spawning; N - 949
During spawning: N - 1.658
HISTOLOGICAL
CRITERIA
ACTIVE FEMALES
Capable of spawning now or in near future.
ACTIVE - MATURE
Advanced yotked oocytes preseni.
Before 61 4%
During: 39.4%
NONSPAWNING
No hydrated oocytes.
No postovulatory foilicle&.
Before: 60.4%
During: 33.3%
SPAWNI^4G
Hydraled oocytes, or post-
ovulatory follicles are present
Before: 1.0%
During: 6.1%
T
NO ATRESIA
No a atresia of advarx»d
yolked oocytes.
Before: 38.4%
During: 16.4%
MINOR ATRESIA
a atresia of advarx:ed yolked
oocytes Is between 1% arxi 49%.
Before: 22.0%
During: 16.9%
NO ATRESIA
No a atresia of advanced
yolked oocytes.
Before: 0.4%
During: 4,0%
MINOR ATRESIA
a atresia of advanced yolked
oocytes Is between 1% arxJ 49%
Before: 0.6%
During: 2,1%
INACTIVE FEMALES
Not capable of spawning now or In near future.
Before 38.6%
During: 60.6%
IMMATURE
Unyolked oocytes
No atresia.
Before: 4.7%
During: 12.2%
UNCERTAIN MATURITY
No advarx»d yolked oocytes.
Before: 31.3%
During: 43,1%
I
INACTIVE - MATURE
Before: 2.5%
During: 5.3%
ATRETIC UNYOLKED
Unyoll^ed oocytes preseni.
With atresia.
Before: 21 .0%
During: 31 .0%
I
EARLY YOLKED
Early yolked oocytes present.
May ftave atresia.
Before: 10.2%
During: 12.1%
MAJOR ATRESIA
Advanced yolked oocytes
present, a atresia of
advanced yolked Is > 50%.
Before: 2.5%
During: 2.0%
POSTSPAWNING
Postovulalory
lollldes preseni.
Before: 0 0%
During: 3 3%
WITH ALPHA
Only a atresia of
unyolked preseni.
Before: 15.5%
During: 17.3%
WfTH BETA
^ atresia must be preseni;
a of unyolked may be preseni.
Before: 5.6%
During: 13.8%
WITH ALPHA OR NONE
Only a atresia of early yolked
present. Or no atresia present.
Before: 4.7%
During: 1.6%
WTTH BETA
^ atresia must be present;
a of early yolked may tM present.
Before: 5.4%
During: 10.1%
WriH ADVANCED
Advanced yolked oocytes
present a atresia of
advanced yolked is > 50%
Before: 0.0%
During: 0.8%
WITHOUT ADVANCED
No advanced yolked
oocytes present
Before: 0.0%
During: 2.5%
Figure 1
Dendrogram illustrating hierarchial classes of histological criteria of active and inactive ovaries of Dover sole Microstomiis paciflcus.
Percentages of females (California and Oregon combined) in each class and subclass taken before (November-December) and during
(January- May) the spawning season.
Hunter et al : Fecundity, spawning, and maturity of Microstomus pacificus
105
Histological methods
All of the preserved ovaries, regardless of develop-
ment, had a piece removed for histological analysis. The
pieces were dehydrated and then embedded in Para-
plast. Subsequently histological sections were cut at
5-6 /im and stained with Harris hematoxylin followed
by eosin counterstain (H&E). Each ovary was classified
histologically in the manner developed for northern
anchovy Engraulis mordax by Hunter and Goldberg
(1980) and Hunter and Macewicz (1980, 1985ab), with
a few modifications appropriate for Dover sole ovarian
structure. In the ovary we identified the presence or
absence of the following: oocjrtes that have not begun
vitellogenesis; oocytes in the first vitellogenic stages
(0.15-0. 55mm diameter); advanced yolked oocytes
(0.47-1. 4mm diameter) noting any stages of nucleus
migration (precursor to hydration); hydrated oocytes;
two stages of postovulatory follicles; and the different
stages of atresia. The rate at which postovulatory
follicles are resorbed in Dover sole is unknown. Hence
no ages were assigned to postovulatory follicles.
Histological classification
We used histological analysis of the ovaries to assess
the accuracy of our gross anatomical classification into
active and inactive states, to define the optimal criteria
for distinguishing mature from immature females, and
to calculate various indices of spawning activity and
postspawning states. The dendrogram (Fig. 1) indicates
the histological characteristics used to classify ovaries
into active and inactive states. The dendrogram also
gives the frequency of the classes in each state for the
prespawning period (November-December) and for the
spawning season (January-May) using combined data
from California and Oregon. The data are also given
by cruise and region in Table 2.
Females were classed as active when histological
analysis indicated that the ovary contained the suffi-
cient number of advanced yolked oocytes for one
spawning. Active females were then separated into
spawning and nonspawning classes using additional
histological criteria. Spawning females were those
which showed histological evidence of past spawning
Table 2
Numbers of female Dover sole Microstomus pacificus in various histological subclasses. Listed by location, before or
during the
spawn-
mg season,
and mean cruise date (year and
month).
Inactive
Active
Immature
Uncertain maturity
Mature
Mature
Atretic unyolked
Early yolked
Major
Postspa
wning
Nonspawning
Spawning
atresia
Atresia present
No
Atresia present
Advanced yolk
No
Minor
No
Minor
atresia
atresia
atresia
atresia
atresia
a
/3
o
o
P
a
All
Cruise
only
only
and p
only
only
and P
With Without
females
Oregon
Before
8512
1
4
0
9
0
0
2
9
5
0
0
20
23
0
0
73
8812
32
133
2
37
5
37
4
34
15
0
0
316
150
3
4
772
During
8602
0
1
0
9
0
0
2
3
4
0
0
6
12
0
0
37
8903
23
70
2
16
1
3
7
18
2
1
3
24
14
3
1
188
8605
1
0
3
4
0
0
5
3
0
4
2
0
0
9
3
34
Total
57
208
7
75
6
40
20
67
26
5
5
366
199
15
8
1104
California
Before
8512
12
10
1
4
0
3
0
3
4
0
0
28
36
1
2
104
During
8701
56
65
11
32
0
7
3
37
11
0
0
115
138
6
9
490
8702
43
99
8
70
2
11
7
36
10
0
1
91
86
12
10
486
8603
38
24
8
40
0
1
7
4
3
3
7
1
2
0
0
138
8803
30
18
4
11
0
4
14
17
3
4
15
31
28
35
11
225
8605
12
9
3
7
2
1
5
0
0
1
13
4
0
2
1
60
Total
191
225
35
164
4
27
36
97
31
8
36
270
290
56
33
1503
106
Fishery Bulletin 90(1). 1992
(postovulatory follicles present) or imminent spawning
(hydrated oocytes or migratory nucleus-stage oocytes
present), while the ovaries of nonspawning females
showed no evidence of recent or imminent spawming
but were capable of spawning in the near future. The
fraction of active females classed as spawning was used
as a spawming rate index. Spawning performance was
also assessed by calculating the mean number of spawn-
ing states (postovulatory follicles, hydrated oocytes,
migratory nucleus) per female in the spawning class.
Females with ovaries classified as active are con-
sidered mature. On the other hand, females with in-
active ovaries could be either immature or mature
because an ovary may have regressed to an inactive
state after the female had attained sexual maturity. We
designed our histological classification of inactive
ovaries to distinguish as best as possible between
mature and immature conditions. Inactive females
were grouped into three classes (Fig. 1): immature,
uncertain maturity, and inactive-mature. The inactive-
mature class included ovaries showing clear histological
evidence of past spawning (postspawning subclass) or
past maturation of advanced yolked oocytes (major-
atresia subclass). Postspawning ovaries contained
either postoviilatory follicles and no advanced yolked
oocytes or postovulatory follicles and mostly atretic ad-
vanced yolked oocytes. The fraction of inactive females
identified as postspawning was used as an index of the
rate at which females passed from the active to the in-
active state during the spawning season.
The five major histological classes of active and in-
active females were subdivided into atretic subclasses
using the first (a) and second {ft) stages of resorption
as defined by Bretschneider and Duyvene de Wit (1947)
and Lambert (1970). One can identify the developmen-
tal stage of the oocyte only during the a stage of atresia
because the oocyte is completely absorbed by the end
of this stage. Subsequent stages (fi, y and 6) involve the
resorption of the follicle. Thus, a atresia is of key im-
portance to fecundity studies since the oocyte class can
be identified. Subsequent stages may be useful for iden-
tifying past spawning activity.
For ovaries containing early yolked or only unyolked
oocytes, classification was based solely on presence or
absence of the following atresia: /?, a of unyolked
oocytes, and a of early yolked oocytes (classes im-
mature and uncertain maturity of the inactive females)
(Fig. 1).
Ovaries with advanced yolked oocytes were sub-
divided into two atretic subgroups using the extent of
the a atresia of the advanced yolked oocytes: minor
atresia, i.e., females with one oocyte to 49% of their
advanced yolked oocytes in a; and major atresia, i.e.,
50% or more of the advanced yolked oocytes in a. We
showed in anchovy that the probability of spawning
was very low when more than 50% of the advanced
oocytes were atretic (Hunter and Macewicz 1985b).
Therefore, ovaries with major atresia of advanced
yolked oocytes were considered inactive (inactive-
mature class) although the ovary contained some ad-
vanced yolked oocytes.
Estimation of total fecundity
We used the gravimetric method to estimate total
fecundity of Dover sole. Total fecundity (Yp) was the
standing stock of advanced yolked oocytes in the ovary:
Yp = Z ■ C, where Z is the ovary weight in grams, and
C is oocyte density (number of advanced yolked oocytes
per gram of ovarian tissue). We also measured diam-
eters of 30 of the advanced yolked oocytes in at least
one of the 2-5 tissue samples analyzed for each female
for which fecundity was estimated. Advanced yolked
oocytes were identified, counted, and measured using
a digitizer linked by a video camera system to a dis-
section microscope.
We used the apparent density of yolk in whole
oocytes after preservation, when viewed on the televi-
sion monitor, to discriminate between developmental
stages of yolked oocytes. We defined three stages of
yolked oocytes: (1) only an initial layer of yolk along
the periphery of the oocyte, appearing as a narrow
band but not extending over 20% of the distance be-
tween the nucleus and the zona pellucida; (2) lightly-
packed yolk possibly extending from the periphery to
the nucleus with the nuclear area still evident; and (3)
yolk dense enough to occlude the nucleus (Fig. 2) which
is histologically equivalent to advanced yolked oocytes.
Counts of stage-3 oocytes were used to estimate fecun-
dity and measurements to estimate mean diameter of
these advanced yolked oocytes.
Alpha atresia of stage-3 yolked oocytes were dis-
tinguished from other whole oocytes viewed on the
television screen. The yolk within these a-atretic
stage-3 oocytes appeared mottled and lighter due to
yolk liquefaction and subsequent resorption, whereas
in normal yolked oocytes it appeared dense, dark, and
in compact globules (Fig. 2). In addition, the zona
radiata (chorion, or membrane layers surrounding the
oocyte) of the atretic oocytes was indistinct and irreg-
ular in appearance. It was not possible to accurately
identify atretic oocytes in frozen, thawed, or poorly
preserved ovaries. Atretic oocytes were not included
in counts of advanced yolked oocytes used to estimate
fecundity. To estimate rates of atresia, we recorded the
number of a-atretic yolked oocytes in the random
sample of 30 stage-3 oocytes measured. The number
of a-atretic advanced yolked oocytes divided by 30 was
used as an index of the intensity of atresia in all females
used for fecundity estimation.
Hunter et al . Fecundity, spawning, and maturity of Microstomus paaficus
107
Figure 2
Three stages of preserved whole
yolked oocytes of Dover sole Micro-
stOTTMS pacificus (stages defined in
text). Lower panel also shows migra-
tory nucleus (MN) oocytes, a hydrated
(H) oocyte, and an o-atretic advanced
yolked oocyte (A). The small air bubble
on the hydrated oocyte is an artifact.
Batch fecundity
Batch fecundity was considered to be the number of
migratory nucleus-stage oocytes or number of hydrated
oocytes in the ovary. We used the gravimetric method
to estimate numbers of these oocytes. Migratory
nucleus-stage and hydrated oocytes stand out as
discrete and easily identified oocyte maturity-classes
(Fig. 2). Hydrated ovaries that contained new post-
ovulatory follicles were not used to estimate batch
fecundity.
We assigned each spawning batch to one of a pos-
sible five batch-order designations [1, 2, (2<B<U-1),
(U- 1), and U], where B is the batch-order number and
U is the total number of spawning batches. The five
batch-order designations were defined as follows: first
batch (where B = l), nonhydrated, advanced yolked
oocytes present and postovulatory follicles absent; sec-
ond batch (where B = 2), one class of postovulatory
follicles and nonhydrated, advanced yolked oocytes
present; intermediate batches (where B>2 but less
than U- 1), two classes of postovulatory follicles and
nonhydrated, advanced yolked oocytes present; the
penultimate batch (U - 1), only two batches were pres-
ent, one of hydrated and one of migratory nucleus
oocytes, with no other advanced yolked oocytes pres-
ent; and last batch (U), no advanced yolked oocytes
present other than a single hydrated batch.
In this classification scheme, we assumed that (1) the
presence of a single class of postovulatory follicles in-
dicated one spawning had occurred; (2) the presence
of two classes indicated at least two spawnings had
occurred; and (3) the absence of postovulatory follicles
indicated no spawning had occurred. The assumption
of no spawning would not hold if the interval between
spawnings was sufficiently long for postovulatory
108
Fishery Bulletin 90(1). 1992
follicles to be resorbed. We used this batch-order clas-
sification system to determine if batch fecundity varied
with the order of the spawnings, as it does in some
species (Alheit 1986, Hunter et al. 1989).
Estimation of length at 50% mature
We estimated the total length (mm) of female Dover
sole when 50% had become mature using histological
criteria. The fraction of females considered to be
mature was estimated for 10 mm or for 50 mm length-
classes, and the data were fit to a logistic curve (Dixon
et al. 1988). We estimated the maturity threshold for
females taken off central California and off Oregon,
before and during the spawning season. In our analysis,
we evaluated the extent to which changes in histo-
logical criteria affected the maturity estimate using six
sets of histological criteria: (1) advanced yolked
Table 3
Conversion equations for Dover sole Microstomus paeifieus
by state, sex, or
season.
Variable
Frozen to fresh for length and weight
Linear equation Y =
a -^ bX
Range of
independent variable
Dependent Independent
Y X
State Sex a
b r
2
F
N
Fresh length Frozen
Fresh weight Frozen
Fresh weight** Frozen
length
weight
weight**
Cal All 9.47
Cal All 0*
Cal F 0*
1.01 0.99
1.22 -
1.29
25,550
10,229
19,575
251
111
147
196-512mm
54-1551g
76-1263g
Variable
Length to weight
Exponential equation
W = aL"
Range of
independent variable
Dependent Indepen
W L
dent
State
Sex a
SE
b
SE
N
Fresh weight Fresh length Cal
(g) (cm) Cal
Cal
Cal
Ore
Ore
Ore
Fresh weight** Fresh length Cal
(g) (cm) Ore
F 0.00198
M 0.00173
Unknown
F,M 0.00198
F 0.00141
M 0.00156
All 0.00159
F 0.0038
F 0.0012
0.00011
0.00018
0.00009
0.00013
0.00015
0.00011
0.00048
0.00026
3.45
3.49
3.45
3.53
3.51
3.50
3.27
3.58
0.016
0.029
0.013
0.026
0.027
0.018
0.033
0.056
1245
264
4
1509
991
457
1448
1198
430
11.8-54.7cm
18.5-47.8cm
12.8-23.5cm
11.8-54.7cm
18.8-57.7cm
20.0-52.2cm
18.8-57.7cm
11.8-54. 7cm
26.8-56.4 cm
Female weight and oocyte v
Variable
olume to ovary weight
Linear equation Y = a -i- bXj
-I- CX2
Range of
independent variable
Dependent
Season Y
Independent
Independent
X2 State
a b
c
r' F
N
Prespawning Ovary weight
(g)
Ovary weight
(g)
Spawning Ovary weight
(g)
Fish weight** Ore 9.07 0.013
(g)
Fish weight** Spher. vol.*** Cal -4.67 0.027
(g) (mm^)
Ore -34.05 0.036
Cal+ -26.06 0.036
Ore
Fish weight** Cal -7.05 0.032
(g) Ore 21.79 0.010
Cal+ -5.88 0.031
Ore
rom 0, line forced through origin,
r-'; for mean oocyte diameters >OMmm.
37.7
95.3
76.3
0.09 38.9 388
0.,54 17.9 30
0.59 45.85 64
0.59 68.2 94
0.41 826.7 1198
0.07 4.2 42
0.39 788.0 1240
122-2017g
202-1124g
0.33-0. 70 mm-''
236-1816g
0.33-0.59 mm^
202-1816g
0.33-0. 70mm3
14-1736g
148-1597g
14-1736g
* Intercept not different f
•* Without ovary.
*** Spherical volume = 4/3 n
Hunter et al Fecundity, spawning, and maturity of Microstomus paaficus
109
oocytes or postovulatory follicles present; (2) early
yolked oocytes with /3 atresia; (3) early yolked oocytes
with only a atresia of the early yolked oocytes or no
a or ^ atresia; (4) unyolked oocytes with /5; (5) unyolked
oocytes with only a of the unyolked oocjrtes; and (6)
unyolked oocytes with no atresia. The sexual maturity
for females identified by criterion 1 is certain, but some
females may be excluded if only criterion 1 is used.
Criteria 2-5, if added to criterion 1, broaden the
maturity definition but increase the risk of misclassi-
fication. Criterion 6 is considered by definition to be
immature. We evaluated these criteria to determine the
optimal histological definition of maturity using a
regression analysis of the lengths of females identified
by each criterion.
Length, weight, and
gonad weight relationships
To enable the reader to convert from one measurement
to another, equations are provided to estimate fresh
wet weight from frozen wet weight and from length
for Dover sole taken in Oregon and central California
waters (Table 3). Analysis of covariance indicated that
the slope of the regression of the natural logarithms
of weight on length did not differ between sexes for
either state. The adjusted group mean for males
differed from that for females in Oregon (A^ 1421,
Fi, 1418 64.87, P< 0.005 for length range 225-522 mm)
but not in California. The slope of the regression of the
natural logarithms of weight on length did not differ
between central California and Oregon females but the
adjusted group means were different (N 2215, F-y 2212
79.18, P<0.005 for length range 188-547mm).' No
difference existed between states in the equations for
males. We do not attach too much biological importance
to these differences; they could be related to differ-
ences in the timing of annual reproductive cycle or our
sampling of it. Nonetheless, it seemed preferable to use
the relationship for a specific sex or region, so all are
listed.
An exponential model was fit to these data sets using
a statistical program of weighted nonlinear regression
(Dixon et al. 1988) where the weighting factor was
the inverse of the variance of fish weight because the
variance of fish weight increased with fish length. To
compute the variance, fish lengths were divided into
several segments, chosen so that within each segment
the variance of fish weight was homogeneous. We pre-
ferred to obtain the estimates of coefficients directly
from the nonlinear fitting so that fish weight could
be directly estimated from the exponential model
(Table 3).
Freezing of Dover sole caused a 9.47mm shrinkage
in total length, independent of fish length (Table 3). A
sample of 251 Dover sole was measured just after cap-
ture and again, after thawing, four months later. The
slope of the regression of fresh total length on frozen
total length (after thawing) was not statistically differ-
ent from 1, but the intercept, 9.47mm, was significant.
Freezing of females, with ovary removed, resulted in
about a 22% loss in wet weight (0.22 = 1 - ;
1 29
see Table 3). ^"^^
We also provided equations to estimate ovary wet
weight (g) from female wet weight (g, without ovary).
This conversion is important if one wishes to express
fecundity as a function of the total weight of the female,
because all fecundity relations in this study are ex-
pressed as a function of female weight without an
ovary. As ovary weight is a function of the developmen-
tal state of the ovary as well as the weight of the
female, separate equations are provided for the pre-
spawning period (November- December) when ovaries
are less developed and for the spawning season when
they are more fully developed. We also provided multi-
ple regression equations to estimate ovary weight from
female weight and the spherical volume of the average
advanced yolked oocyte (computed from the mean
diameter). These equations are used in the discussion
to estimate ovary weight when an ovary contains an
entire complement of fully matured advanced yolked
oocytes.
Reproductive condition
Accuracy of gross anatomical classification
We rarely misclassified inactive ovaries using gross
anatomical criteria. Of the 1272 females classified as
inactive, only 14 (1.1%) were identified as active using
histological criteria. This error rate is so low that dif-
ferences could be attributable to clerical errors alone.
A more common error in gross anatomical classifica-
tion was to misclassify females as having active ovaries.
One hundred and fifty-nine females (11.9%) were
visually classified as having advanced yolked oocytes
and were believed to be capable of spawning, while
histological analysis indicated that their ovaries were
inactive and future spawning was unlikely. The 159
females misclassified as active fell predominantly into
two classes: females with ovaries in the early stages
of vitellogenesis (40.8%), and females with advanced
yolked oocytes with high levels of atresia (30.1%)
(Table 4).
Misclassification of the early stages of vitellogenesis
as active is expected because the gross anatomical
criterion, "yolked oocytes visible," is not exact;
observers are bound to differ on whether to include or
exclude females that fall near the visible threshold
110
Fishery Bulletin 90(1). 1992
Table 4
Histological classification of female Dover sole Microstomus pacificus with inactive ovaries
that were misclassified using gross anatomical criteria as having active ovaries. Data
from central California and Oregon are combined.
Collection
period
Percentage
as active per
of females misclassified
inactive histological class
Total no.
misclassed
females
Immature
Unyolk
atretic
Early
yolked
Major atresia
of adv. yolked
Post-
spawning
Nov-Dec
Jan-May
Nov-May
0.0
4.5
3.1
2.0
23.6
17.0
48.9
37.2
40.8
48.9
21.8
30.1
0.0
12.7
8.8
49
110
159
Figure 3 (below)
Seasonal change in four indices of reproduction in Dover sole Microstomus pacificus. Indices
are plotted as a function of elapsed time since 1 November; data are combined from different
years; California (solid circles) and Oregon (open circles) data were combined to fit trend lines;
numbers are the sample size of females, (upper left) Percentage of Dover sole with active
ovaries; trend line is a weighted Weibull model (see text Eq. 1). (upper right) Percentage of
females with active ovaries which had one or more spawning states; trend line is logistic model
ga + bL
P = , where a = - 5.678 and b = 0.036. (lower left) Mean number of spawning states
l + e"*''''
in active ovaries; bars are two standard errors of the mean, (lower right) Percentage of females
with inactive ovaries identified as postspawning; trend line is logistic model where a = 8.495
and b = 0.042.
for detection of yolked oocytes in
the ovary. An exact criterion,
such as oocyte diameter, would
be more accurate but would be
impractical for production work
on the ship. Misclassification of
highly atretic ovaries as active is
also expected, since a-atretic ad-
vanced yolked oocytes are diffi-
cult if not impossible to see with
the unaided eye. As highly-atre-
tic advanced ovaries were rare in
this study, our failure to detect
them was a minor systematic
error. Under environmental con-
ditions unfavorable to reproduc-
tion, however, this could be an
error of consequence.
Changes in ovarian
condition during tine
spawning season
The fraction of females anatom-
ically classed with active ovaries
70
iJJ 60
<
ff 50
UJ
>
I-
o
■
1 1
1 I I ■
1
.
1439^ ^^
.
\"'
o"
.
\
. • 500
■
• CALIFORNIA
O OREGON
\
O 30
\^0?52
• 60
138 ^\,.„^
1 • 1
50 JOO 150
DAYS ELAPSED SINCE NOVEMBER 1
FEB
MONTH
100
r
I
I
1 1
lol2
60
SO
■
105 /
/
40
■
• /
• 7
20
•
G7
26S
199^
O
0
■
473
4
43
O
16
42
•
3
50 100 150
DAYS ELAPSED SINCE NOVEMBER 1
40
-
A'
20
■
115
• •
135 y'^
/ "
-
37
222
287
^
146
1
299 0»
30
19
JAN FEB
MONTH
Hunter et al : Fecundity, spawning, and maturity of Microstomus pacificus
1 1
declined over the spawning sea-
son as females expended their
stock of advanced yolked oocytes.
We fitted a weighted Weibull
function to the combined Califor-
nia and Oregon data, yielding the
equation
0.656 e
-(111.5)
(Eq. 1)
0.8
0.6
I
u
o
o
z
<
UJ
s
1.2
1.0
0.8
0.6
20
where t is days elapsed since 1
November; P, the fraction active,
is weighted by , and
P(l-P)
pseudo r^ is 0.96. According to
the equation, the percentage of
females with active ovaries de-
clined from 65% at the onset of
the spawning season (about 6
December) to 40% by the end of
January; by the end of February
only 18% of the females had ac-
tive ovaries (Fig. 3, upper left).
In California and Oregon, the
mean diameter of the stock of ad-
vanced yolked oocytes increased
steadily from December through
April (Fig. 4). Thus reproductive-
ly active females continued
vitellogenesis throughout most of
the spawning season. By March
or April the average advanced
yolked oocyte is closer to the
minimum size at which hydration
begins (diameter 1.35 mm). Thus
Dover sole may be able to spawn
at a higher rate late in the
spawning season, because less
yolk would have to be added to
the advanced oocytes for them to
attain the size at hydration.
Only 10 females taken in No-
vember-December were classed
as spawning on the basis of their
ovarian histology (Oregon and California data com-
bined, N 949). They comprised only 1.0% of all females
with preserved ovaries taken during this time and only
1.7% of those classed as active. Clearly, spawning is
just beginning in November-December in California
and Oregon waters. The spawning rate index increased
from 1.3% in November-December to 12% by early
February; it accelerated at the end of the season with
spawning females comprising about 70% of all active
females (percent calculated by the trend line in Fig. 3,
Hydration
Z
<
5
ui
o
o
o
1.2
1.0
40
Dec
60
I
80
Jan
100
Feb
120
I
160
Apr
180 200
I May
Hydration
I
40
Dec
60
I
80
Jan
100
Feb
120
I
140
Mar
160
Apr
180
May
200
ELAPSED TIME (days)
Figure 4
Increase in mean oocyte diameter (D) of the advanced yolked oocytes of Dover sole
Microstomus pacificus from Oregon (top) and California (bottom) as a function of elapsed
time (T) since 1 November. Data from different spawning seasons are combined; shaded
area indicates size range of oocytes at the onset of hydration; trend lines are Oregon
(D = 0.742 + 0.00282T, r' 0.658, A^ 195) and California (D = 0.761 + 0.0022T, r- 0.41,
N 365).
upper right). Thus at the end of the season, most of
the females with active ovaries had spawned recently.
We believe this sharp increase in the index near the
end of the season is evidence for a seasonal increase
in spawning frequency.
A late seasonal increase also existed in the occur-
rence of multiple spawning stages within the same
ovary. As many as five different past or potential
spawning stages could be distinguished histologically
in the same ovary: two stages of postovulatory folli-
Fishery Bulletin 90(1), 1992
cles, hydrated oocytes, migratory nucleus oocytes, and
other advanced yolked oocytes. These data were ex-
pressed as the number of spawning states per spawner
(spawns per spawner, Fig. 3, lower left). The average
number of spawns per spawner increased from about
one in mid-March to about three by early April. These
data also indicated that spawning frequency may in-
crease near the end of the spawning season.
The fraction of females with inactive ovaries that
were classed as postspawning also increased late in the
season (Fig. 3, lower right). This index can be con-
sidered a measure of the rate females in the popula-
tion pass from the active to the inactive state. Although
the duration of this stage was unknown, we were cer-
tain that it was ephemeral because there were always
many fewer females classed as postspawning than the
cumulative total of females that had passed from the
active to inactive state. This index increased sharply
in late-March through April, indicating that the rate
females passed into the inactive stage accelerated
during the last part of the season.
The sharp increases in the three indices described in-
dicated that the daily production of eggs by the popula-
tion may be higher in March than February even
though fewer fish were spawning. For example, by mid-
March (13 March), a half to a third as many females
had reproductively active ovaries than in mid-February
(10 Feb.). On the other hand, in mid-March as compared
Table 5
Effect of location of tissue samples within the ovary of Dover sole Microstomiis pacificus
on oocyte density (number of advanced yolked oocytes per unit sample weight) with mean
and standard deviation (SD) and, below, two-way analysis of variance on results.
Location in ovary of sample*
Oocyte
density
Plane of section
no.
Lobe
Long.
Cross.
N
Mean
SD
1
Rt
Post
Int & Ext
10
1803,7
468.1
2
Rt
Mid
Ext
10
1801.1
532.0
3
Rt
Mid
Int
10
1754.1
488.6
4
Rt
Ant
Ext
10
1886,1
440.2
5
Lt
Mid
Int
10
1839.3
537.6
Analysis of variance on five locations
Source
DF
SS
MS
F
Fish
Position
Error
9
4
36
10,379,034
96,638
633,493
1,153,226
24,160
17,597
1.373
Total
49 11,109,165
Rt = right lobe, Lt = left lobe. Post = posterior end, Mid =
or end, Int = internal, and Ext = external.
•Long. =
middle,
= longitudinal
Ant = anteri
with mid-February, about twice as many females with
active ovaries were classed as spawning, and the
ovaries of the spawners contained evidence of about
twice as many past or potential spawnings. Thus the
reproductive output of the reproductively active fe-
males in the population in mid-March might be four
times that of the active fish in mid-February. If this
is true, half the number of active females could pro-
duce twice as many eggs per day. This is, of course,
sheer speculation because the duration of these spawn-
ing stages is unknown. Nevertheless, the data pre-
sented in this section collectively suggest that the daily
production of eggs by the population may increase near
the end of the season even though fewer females are
spawning.
Total fecundity
Location of tissue samples
A key assumption underlying the gravimetric method
of fecundity estimation is that oocytes are randomly
distributed in the ovary. To determine if advanced
yolked oocytes are randomly distributed in the ovary,
we compared the densities of advanced yolked oocytes
in tissue samples taken from five different locations
in the ovary of ten females. The location of a tissue
sample within the ovary was defined in terms of three
characteristics: longitudinal
plane of the ovary (anterior end,
middle, and posterior end); cross-
sectional plane (interior near the
lumen, exterior near the ovarian
wall, or interior and exterior
combined); and right and left
lobes of the ovary. The char-
acteristics of the five ovarian
locations along with the mean
oocyte density of each location
are indicated in Table 5.
Initially we tested the overall
effect of location of the tissue
sample on oocyte density using
two-way ANOVA; the effect of
position was insignificant at the
5% level of significance (Table 5,
lower). We also tested for pos-
sible differences between pairs of
location characteristics: posterior
end vs. middle; posterior vs.
anterior ends of the ovary; right
vs. left lobes of the ovary; and in-
terior and exterior sections of the
ovary. No significant differences
were detectable between any of
Hunter et al,: Fecundity, spawning, and maturity of Microstomus pacificus
113
these four comparisons; F values ranged from 0.01 to
0.14, with the degrees of freedom being 1 and 45 for
each comparison. Thus the advanced yolked oocytes in
Dover sole are randomly distributed within the ovary,
and tissue samples can be taken from any location or
lobe without bias.
Optimal number of tissue samples
To develop a procedure for estimating the number of
tissue samples needed for estimating total fecundity,
we first considered the general fecundity model. The
true total fecundity (Yp) is the condition where all the
advanced yolked oocytes in the ovary are counted, and
the relation between female weight (W) and fecundity
is defined as
Yp = f(W) + A
(Eq. 2)
where f(W) = a + bW, and A is the error term. The
variance of A, o^a. measures the deviation of the data
set (Yp.W) to the model f(W). As it was impractical
to count all advanced yolked oocytes in the ovary, Yp-
is estimated from counts of oocytes in weighed tissue
samples, expressed as oocytes per gram of tissue or
oocyte density. The precision of a fecundity estimate
can be increased by increasing the number of tissue
samples taken per female. On the other hand, if the
amount of labor for fecundity work is fixed, then in-
creasing the number of tissue samples per fish would
reduce the number of fish that can be sampled. Thus
we needed to know the minimum number of tissue
samples necessary to guarantee a goodness-of-fit of the
model to the data set.
We determined the optimum number of tissue
samples by minimizing the variance of sample variance
of A (o2(s2a))- This procedure led to using the ratio of
the variance of oocyte counts between tissue samples
within fish (o^e) to the variance around the regression
line (o^a), i.e., Q = a'^Ja'^t^. The smaller the 0, the
fewer tissue samples are needed.
Let's denote for the ith fish, i = 1, . . ,n,
Wj = fish weight,
Ypi = total number of advanced yolked oocytes in the
ovary,
yij = advanced yolked oocyte count in the jth tissue
sample, j = l,. .,m,
Zjj = weight of the jth tissue sample,
Zj = formalin wet weight of ovary,
m = number of tissue samples from an ovary,
Mj = maximum number of tissue samples in an
ovary,
Ypij = estimate of total number of advanced yolked
oocytes in the ovary from the jth tissue sample
■^ Z„
Ypi = estimated total number of advanced yolked
oocytes in the ovary and is used for all analyses
in fish fecundity in later sections
, and
m
Ypi = estimate of total fecundity from the regression
model.
We write Ypij as
Ypij = Ypi + (Ypij - Ypi)
= f(Wi) + Ai + ei3 (Eq. 3)
where e^ = Ypjj - Ypi . The estimated total number of
advanced yolked oocytes in the ovary is
m
I Ypij
Ypi = — = f(W) + Ai + e,
m
and
o^ = o2a +
= f(W) + I
(M-m\ „
m
(Eq. 4)
(Eq. 5)
Thus the variance around the regression line o-^ based
upon the data set (Ypi,Wi) is composed of two vari-
ance components: one is o^a ^"d the other is o^g- The
sample counterparts for o-^ and o^g ^^re s^i and s-g:
s2, =
[YF-f(W)]2
n-q
(Eq. 6)
is the mean square error from a regression analysis on
(Ypi, Wi) where q is the number of regression coeffi-
cients and n is the number of fish, and
1 l(YFij-Yp,)2
n(m-l)
(Eq. 7)
is the within-sample variance (Hunter et al. 1985). The
estimate (s^a) of the variance around f(W) when Yp
is known (o^a) can be estimated by subtraction:
Fishery Bulletin 90(1). 1992
s\
M-m
M
ing the sample size (n) for a linear regression was
(Eq. 8)
m
n- 1 =
According to Hunter et al. (1985), the optimum number
of tissue samples can be determined for a given 6
(= s^e/s^A). the cost of processing a tissue sample, and
the cost of processing a fish. The ratio, K = s^i/s^a.
measured the excess variance which is contributed by
taking tissue samples rather than counting every ad-
vanced oocyte in the ovary.
We used Dover sole collected during January-Feb-
ruary 1987 to determine the optimal number of tissue
samples. Two tissue samples were taken from the
ovaries of 99 Dover sple. The within-sample variance
of oocyte density (s^g = 1053 x lO'*) was obtained from
an ANOVA, and the linear regression of Yp on W
(Yp = 20,255 -H 40.54 W) gave the MSE (s2^ = 18,469 x
10^) (Table 6; Fig. 5, lower middle). Thus 0 is 0.058
when calculated from equations (7 and 8) where m = 2.
Because M was large (range 200-700), was
M
assumed to equal 1, and s^a was computed as s^^ -
(s^e^^) = 17,942 X 10^. Hence when two tissue samples
are used, the variance within tissue samples is only
5.8% of the variance around the fecundity-fish weight
regression line. To quantify the
excess variance due to subsam-
pling, we computed K = s-^/s^^
= 1.03. This means the variance
around the regression line which
was based on two tissue samples
per fish (Eq. 4) is about 1.03
times that of an equation based
on counts of all advanced yolked
oocytes in the ovary (Eq. 2). Al-
though the vnthin-ovary variance
was small, we recommend count-
ing two tissue samples per female
because the cost of processing
the second sample was minimal.
Cy(b)2
(Eq. 9)
Table 6
Within-sample variance (s%xlO"^) from ANOVA and the
MSE (s-(XlO"'') from the regression analysis of fecundity
(Yx 10"-) and weight of Dover sole Microstomiis pacificus.
California females taken January-February 1987.
Analysis of variance on total fecundity
Source
DF
SS
MS
Fish
Error
Total
98
99
197
8,559,008
104,201
8,663,209
87,337
1,053
Analysis of variance on linear regression
Source
DF
SS
MS
Regression
Residual
1
97
2,488,019
1,791,468
2,488,019
18,469
134.72
Predictor
Coeff.
SD
Constant
Fish wt.
202.55
40.54
32.03
3.49
6.32
11.61
Optimal number of females
In addition to the sample alloca-
tion based on cost of processing
fish and cost of processing tissue
samples, the number of females
needed for a regression estimate
of total fecundity was determined
by modifying a procedure sug-
gested by Thigpen (1987) to the
1987 fecundity data for Dover
sole. The equation for determin-
OREGON
N = 67
0 1-^
100
200 ] —
o
o
o
_i
<
I-
o
OREGON
N = 36
OREGON & CALIFORNIA
CAL DECEMBER
CAL JANUARY
CAL MARCH
ORE DECEMBER
ORE MARCH
100
200, —
CALIFORNIA
N = 31
DECEMBER
100
50C
CALIFORNIA
JANUARY
N =
173
.P-'^-
^t0^.
CALIFORNIA
MARCH
N t= 103
;-'^^?Tt
^
■*—■■-'""'
FEMALE WEIGHT (g)
Figure 5
Total fecundity of Dover sole Microstomus pacificus as a function of female wet weight
in grams (without ovary) for various months taken in central California and Oregon.
Each point represents a single female; equations are given in Table 9.
Hunter et al . Fecundity, spawning, and maturity of Microstomus paclflcus
15
where r^ is the coefficient of determination, and
CV(h) is the coefficient of variation for the regression
coefficient (b). The coefficient of determination {r~)
for Dover sole total fecundity and fish weight was 0.58.
Thus 73 females are required for a CV(h) = 0.10. In con-
clusion, two tissue samples from each of 70-80 Dover
sole females were adequate for expressing the relation
between weight and total fecun-
dity, if a CV(h) of about 0.10 is
desired.
qualitative analysis that recruitment of stage-2 oocytes
into the advanced stock probably ends when the mean
diameter of stage 3 is between 0.8 and 0.9mm.
We conducted a stepwise multiple regression analysis
of total fecundity (Yp) on mean diameter of the ad-
vanced oocytes (D) and female weight (W) for females
taken off Oregon in November-December. The coeffi-
Relation to ovarian
development
The optimum time for estimating
potential annual fecundity is
early in the spawning season
when the probability of spawning
is low. Estimates taken in this
period may be biased because all
oocytes may not have been
recruited into the advanced stock
of yolked oocytes. In this section,
we use Oregon Dover sole taken
in November-December to ex-
amine the recruitment of oocytes
into the advanced stock of yolked
oocytes.
If substantial numbers of
oocytes are maturing from early-
yolked to more advanced stages,
one would expect an overlap in
the size distributions of oocytes
in different development stages.
When vitellogenesis of the early
yolked oocytes does not continue,
one would expect that a gap
would develop between the less-
advanced and the most-advanced
oocytes as the yolked oocytes
continued their maturation. In
Dover sole, the diameter distri-
butions of stage- 1, -2, and -3
oocytes broadly overlap when the
mean diameter of the advanced
yolked oocytes (stage 3) is less
than 0.7 mm (Fig. 6). The extent
of overlap declines as the stage-3
oocytes grow from 0.7mm to
0.8mm. Separation of the ad-
vanced stock (stage 3) from the
other vitellogenic stages (1 and
2) becomes complete as the ad-
vanced stock grows from 0.8 to
0.9 mm. It appears from this
T3
ra
D
O
,
0.87
J\
0.88
0.96
1.00
' '
' ■. \
/ \
•. - ..
r T •^ 1 1
^,.^
■I'l "T "T" T"r ' I "I'TT T 'T -T
0.2 0.3 0.4 0.5 0.8 0.7 0.8 0.9 1.0 1.1 ' 0.2 0.3 0.4 0 5 0.6 0.7 0.8 0.9 1.0 1.1
OOCYTE DIAMETER (mm)
Figure 6
Frequency distribution of oocyte diameter of three vitellogenic oocyte stages in ovaries
of Dover sole Microstomus ipacijicus. Each panel represents the ovary from one
female. Fish arranged in order of the mean diameter of the advanced yolked oocytes,
stage 3 (which is shaded); numbers and arrows indicate mean diameter of stage-3 oocytes.
Mean total length was 420 mm and mean ovary-free weight of the females was 868 g.
Fish were taken November-December 1988 along the Oregon coastline.
1 16
Fishery Bulletin 90(1). 1992
Table 7
Analysis of the relation between total fecundity (Yp) of Dover sole Microstormis
padfieiis and gonad-free body weight (W) and the average diameter of the ad-
vanced oocytes (D) using stepwise regression with analysis of variance. Specimens
from Oregon in November-December 1988.
Stepwise regression
Step
1
2
Constant 22,398
Weight (W) 45.4
t* 7.56
Diameter (D)
f
S 25,654
R^ 27.58
-88,768
47.8
8,80
129,893
6.01
23,096
41.69
Analysis of variance
Source
DF
SS MS F
P
Regression
Error
Total
2
149
151
5.68x10'° 2.84
7.95x10'° 5.33
1.36x10'°
X 10'° 53.27
xl0'°
< 0.000
Source
DF
Sequential SS
Weight
Egg diameter
1 3.76x10'°
1 1.92x10'"
1.96<«1.98., df>120.
• For P = 0.05,
cient for diameter, as well as the one
for weight, was positive and signifi-
cant (Table 7). Thus the potential an-
nual fecundity was not fully recruited
as stage-3 oocytes in some of the Ore-
gon Dover sole taken in November-
December, since total fecundity in-
creased with the mean diameter of
the oocytes used to estimate total
fecundity.
To determine the level of ovarian
development (oocyte diameter) at
which the full complement of oocytes
was recruited into the advanced yolked
oocyte class (stage 3), we conducted a
series of stepwise multiple regression
analyses by successively removing the
data by 0.01 mm decrements from the
lowermost oocyte diameter class start-
ing at 0.71 mm. This analysis indicated
that the threshold for a significant ef-
fect of oocyte diameter on total fecun-
dity was between mean diameters of
0.85 and 0.86mm (Table 8). The multi-
ple regression coefficient for oocyte
diameter was significant and positive
when females with oocyte diameter
Table 8
Results of stepwise
multiple regression
of the total
fecundity (Yp) of Dover sole Microstomus padfieus on gonad-free body weight 1
(W) and mean oocyte diameter (D) for a succession of oocyte diameter-classes using the model Yp = a -i-
b,W -f b.,D. Specimens taken
along Oregon coast November-December 1988. Line
separates oocyte diameter-classes where diameter
is a significant variable, from
those where it is not.
Multiple
regression
coefficients and their < -ratios for:
Oocyte diameter
class (mm)
Constants
Fish weight
Oocyte
diameter
r^
N
a
b,
t
b.
r
0.71-1.04
152
-88,768
47.8
8.80
129,893
6.01
0,417
0.72-1.04
148
-90,147
48.3
8.90
130,860
5.87
0,424
0.74-1.04
133
-94,375
49.6
8.34
134,295
5.14
0,415
0.76-1.04
119
-90,710
51.3
7.98
128,688
4.17
0,407
0.78-1.04
105
-83,087
53.4
7.53
118,385
3.16
0.396
0.80-1.04
91
-90,344
55.9
7,29
123,891
2.60
0.400
0.82-1.04
85
-87,274
59.2
7,34
117,459
2.20
0.405
0.83-1.04
81
-111,317
56.9
7.44
145,193
2.77
0.440
0.84-1.04
77
- 103,009
58.7
7.55
134,678
2.42
0.460
0.85-1.04
72
-90,464
60.1
7.33
120,166
1.95
0.457
0.86-1.04
67
-38,172
65.3
7.96
60,555
0.94
0.502
0.87-1.04
64
-33,982
63.5
7.30
57,891
0.87
0.482
0,88-1,04
60
-15,721
69.4
7.63
33,456
0.48
0.514
0,90-1,04
46
45,073
82.0
7.92
-40,601
-0.47
0.594
0.92-1,04
34
1.98 for df =
60,339 84.0
120, 2.00 for df = 60, and 2.06 for df
6.92
= 25.
-58,185
-0.49
0.616
• For P = 0.05, t is
Hunter et al,: Fecundity, spawning, and maturity of Microstomus paoficus
1 17
equal to or less than 0.86mm were included, but was
insignificant when only those having a diameter
greater than 0.86mm were considered. We concluded
that ovaries in which the advanced stock of yolked
oocytes has an average diameter of 0.85 mm or less are
not sufficiently developed to be certain that the annual
stock is fully recruited. Consequently, to estimate the
potential annual fecundity, we used only females in
which the average oocyte diameter of the advanced
stock exceeded 0.85 mm.
No relationship between oocyte diameter and total
fecundity was detected in the females taken off cen-
tral California during November-December. Oocyte
diameter may not have been a significant variable in
central California because fewer females were exam-
ined and their ovaries were more advanced. In 48% of
females from California, advanced yolked oocytes
averaged more than 0.85mm in diameter {N 65),
whereas only 34% of the fish taken off Oregon had
oocytes that did so {N 128).
Seasonal variation in total fecundity
Total fecundity of Dover sole decreased during the
spawning season off both central California and Oregon
(Fig. 5). Analysis of covariance indicated that equations
expressing the relation between female weight and
fecundity differed within the spawning season (Table
9). The total fecundity for a 1kg female declined from
about 80,000 advanced oocytes in December to about
50-60,000 during the spawning season (Table 9).
To further describe the decline in total fecundity over
the season, we also regressed fecundity on female
weight and elapsed time since 1 November. In both cen-
tral California and Oregon the negative coefficients for
elapsed time were significant, indicating that total
fecundity declined wath elapsed time (Table 9). Analysis
of covariance indicated that multiple regression equa-
tions for California and Oregon were not different
(analysis over a similar weight range of 174-1542 g;
Fi 388 1.59, P 0.208; adjusted mean fecundity for
Oregon 57,849, SE 2092; adjusted mean fecundity for
California 54,733, SE 1152). When we combined data
for the two regions, we found that total fecundity
declined on the average about 12% per month. This
computation underestimated the actual rate of decline,
since it did not take into account females that had
spawned all of their advanced yolked oocytes.
Potential annual fecundity
Potential annual fecundity was considered to be equi-
valent to the standing stock of advanced yolked oocytes
in fully developed, prespawning females. We consider
Table 9
Linear regression coefficients, confidence intervals, and estimates for the relationship between female weight (W, ovary-free, in g)
and total fecundity (Yp ) of Dover sole Murostomus pacificus from California and Oregon. Analysis of covariance for the effect of season
on the relation between total fecundity and weight. Multiple regression coefficients are also given for the effect of elapsed time (T;
days since 1 Nov.) and female weight on total fecundity.
State
Mean date of
cruise
Linear regression by month and state
Linear equation Yp = a -h bW
Regression
estimate for
Analysis of covariance for
effect of month with
weight ranges similar
95% CI b 95% CI r-
N 1 kg female Variables df
Oregon
Central California
7 Dec 88
3 Mar 89
8 Dec 85
31 Jan 87
23 Mar 88
17,640 ±15,460 65.5 ±16.4
14,492 ±8,530 42.9 ±10.6
29,497
20,154
12,072
±6,121
±6,344
±16,924
51.6
38.9
40.7
±16.1
±6.8
±18.6
0.49
0.66
0.58
0.42
0.15
63.9
67.9
42.9
127.0
18.9
67
36
31
173
103
83,140
57,392
81,097
59,022
52,772
Weight
Month
Error
Total
1 67.85
1 20.06
95
97
Weight 1 25.20
Month 2 18.91
Error 217
Total 220
< 0.005
< 0.005
< 0.005
< 0.005
Multiple recession of total fecundity on weight and days elapsed since 1 November
Multiple regression equation Yp = a + b, W -h bjT
State
a 95% CI bi 95% CI
95% CI
N
W W T T
min max ^ nun * max
(g) (g) (d) (d)
Oregon 35,162
Central California 41,552
±13,378 55.2
±8,142 40.3
±11.2
±6.4
-237
-224
±104
±63
0.59
0.37
74.9
92.3
103
307
147.7
120.0
1815.9
1690.3
33
34
151
160
Fishery Bulletin 90|l). 1992
Table 10
Relationship between total fecundity (Yp) and gonad-free body weight (g) or total length (mm) for California and Oregon Dover sole
Microstomus pacifieus females meeting specifications for potential annual fecundity estimation (females taken in November-December
with average oocyte diameter > 0.85 mm and no evidence of past or imminent spawning). Data are compared with Yoklavich and Pikitch
(1989) estimates for Oregon.
Fecundity and weight
Linear equation Yp = a + bW
Estimate for
1 kg female
Gonad-free weight (g)
Mean Range
State a
95% CI b 95% CI r^
F N
Oregon 17,640
California 29,871
Oregon + California 21,124
±15,453 65.5 ±16.4 0.488
±12,996 50.9 ±17.1 0.544
±9,248 62.0 ±10.4 0.504
63.86 67
37.06 30
98.42 97
83,140
80,771
83,124
870.5 236.0-1815.9
704.7 202.1-1124.2
819.2 202.1-1815.9
Fecundity and leng^th
Exponential equation Yp = aL*"
Estimate for
453mm female*
Total length (mm)
Mean Range
State a
95% CI b 95% CI Pseudo r^
F N
Oregon 5.667-10-'=
California 6.101-10-''
Yoklavich & Pikitch** 1.637-10-'=
±1.943-10-== 3.806 ±0.713 0.431
+ 5.042-10-^ 3.020 ±1.380 0.084
±6.928-10-"= 4.021 ±0.684 0.818
ation in Table 3.
ikitch (1989) original data.
47.03 64
2.57 30
135.13 32
72,856
64,094
78,382
436 298-551
423 296-526
448 358-550
* Weight is about 1000 g using equ
** Estimated from Yoklavich and P
Dover sole to be developed when the average diameter
of the advanced oocytes (stage 3) exceeds 0.85mm;
prespawning females are those taken in November-
December which show no histological evidence of re-
cent past or of imminent spawning (no postovulatory
follicles nor hydrated oocytes present).
Q
Z
3
O
UJ
<
O
Using only specimens that met these specifications,
we regressed total fecundity on female weight (without
ovary) for females taken in central California and
Oregon. The two regression equations were quite
similar; when the data are truncated so that the ranges
of female weights were equal, no statistical difference
existed between California and
Oregon. Combining all data, we
obtained the general equation
200
180
160
140
120
100
80
60
40
20
0
-
o
-
■
-
o o
/ /
-
o
-
ca o ■,
/
-
oo
O 0 o
° 8 y 9
On O'^n' .^ ^ J?0 „
.-l^*-^ % o o o o
J*-"^ 0 0
— 0 o
0
0
o
o
o
o
1 1 1
_J 1_. t 1 I 1 1
1111
280 320 360 400 440 480
TOTAL LENGTH (mm)
520
560
Yp = 21,124 + 62.0W (Eq. 10)
where Yp is estimated total fecun-
dity from the regression line, and
W is ovary-free female weight in
grams. Therefore, the potential
annual fecundity for a 1 kg Dover
sole is about 83,000 oocytes
(Table 10).
Fecundity of Dover sole was
estimated recently by Yoklavich
Figure 7
Total fecundity as a function of total length
of Dover sole along the Oregon coast for our
November-December 1988 data (open circles
and solid line) and for data given by Yokla-
vich and Pikitch (1989) (filled squares and
dashed line). Equations are given in Table 10.
Hunter et al.: Fecundity, spawning, and maturity of Microstomus paaficus
I 19
Figure 8
Relative frequency distribution of the fraction of the random
sample of advanced yolked oocytes measured that were atretic {a
stage) from the Dover sole Microstomus pacifieus females used in
estimates of total fecundity shown by state: Oregon females. A' 189;
California females, N 361.
and Pikitch (1989) for females caught along the Oregon
coast. Those authors used an exponential model and
expressed annual fecundity as a function of length. The
distribution of points in our data broadly overlapped
the data of Yoklavich and Pikitch (1989) (Fig. 7). To
compare our Oregon data with theirs, we truncated
ours so that the length ranges of the two sets coincided
and applied an analysis of covariance to log-trans-
formed data. Analysis of covariance indicated that no
significant difference existed between the two equa-
tions (Fi 81 2.03, P 0.158). In Table 10, the exponen-
tial equation for fecundity as a function of length is
given for our data (not truncated) and for that of
Yoklavich and Pikitch (1989). In summary, we found
no statistical difference between California and
Oregon, nor between our Oregon data and that of
Yoklavich and Pikitch (1989).
Atretic losses
In fishes with determinate fecundity, a
key question is whether atretic losses dur-
ing a season constitute an important frac-
tion of the potential annual fecundity. We
identified whole atretic oocytes under a
microscope while doing our fecundity
work. To measure atretic losses, we
counted the number of atretic oocjrtes (a
advanced yolked oocytes) occurring in a
random sample of 30 advanced yolked
oocytes for each of the females used to
estimate total fecundity (N 550).
In the fish used to estimate fecundity,
the average fraction of advanced yolked
oocytes that were atretic was low with
the mean 0.015 (SD 0.032, N 361) in Cali-
fornia, and even lower in Oregon females
(mean 0.0033, SD 0.014, A^ 189). Atretic
oocytes were observed in only 26% of
California females and in only 6% of
Oregon females (Fig. 8).
The total fecundity of California fe-
males was negatively correlated with the
fraction of oocytes in the ovary that were
atretic. A stepwise multiple regression of
female weight, elapsed time, and fraction
atretic on total fecundity (Table 11) indi-
cated that the coefficient for the fraction
us
<
tu
0.9
0.8
0.7
0.6
A
OREGON
li-
O
0.5
z
O
0.4
h-
o
<
DC
u.
0.3
0.2
0.1
•
0.00 0.05 0,10 0.15 0.20
CO
LU
_J
<
5
LU
0.9
0.8
0.7
0.6
■
B
CALIFORNIA
U-
o
0.5
■
z
O
0.4
■
o
<
CC
u_
0.3
0.2
0.1
•
1 . - _
0.00 0.05 0.10 0.15 0.20
FRACTION OOCYTES THAT WERE ATRETIC
Table 1 1
Analysis of the relation between total fecundity (Yp) of Dover sole Micro-
stomus pacificus ovary-free body weight, elapsed days since 1 November, and
fraction of atretic oocytes using stepwise regression. Specimens from California.
Step
Stepwise regression
1 2
3
Constant
23,335
40,296
40,976
Weight
37.7
2.11
40.8
14.03
41.4
14.23
Elapsed days
t
-205
-7.85
-206
-7.94
Fraction atretic
t
-63,278
-2.17
S
r'
19,317
28.99
17,868
39.41
17,776
40.20
Source
DF
Analysis of variance
SS MS
Regression
Error
Total
3
357
360
7.58x10'"
1.13x10"
1.89x10"
2.53 xlC"
3.16x10'
79.99 < 0.001
Source
DF Sequential SS
Weight
Elapsed days
Fraction atretic
6.28x10'"
1.15x10'"
1.49x10'
*ForP = 0.05, t 1.97.
120
Fishery Bulletin 90(1). 1992
State
of oocytes that were atretic was sig-
nificant and negative. According to the
equation, when 10% of the advanced
oocytes were atretic, total fecundity in
a 1 kg female was about 8% lower than
when no advanced yolked oocytes
were atretic. This analysis indicated
that atretic losses of potential annual
fecundity occurred, but on a popula-
tion basis such losses were negligible.
No relation between fecundity and
atresia existed for Oregon females,
probably because atresia was less
prevalent in Oregon, with only 6% of
females effected compared with 26%
in central California.
The ovaries of many more females
were examined histologically for
atresia (N 2607) than were examined
using the anatomical method, because
we restricted the anatomical work to
the fish used to measure total fecun-
dity. Only 2% of all females examined
histologically (Table 2) had ovaries in
which 50% or more of the advanced
yolked oocytes were in a atresia, but
minor atresia was more common.
Minor atresia occurred in 52% of the
nonspawning California females and in
35% of the nonspawning Oregon
females (Table 2).
The histological method was more
sensitive than the anatomical one.
Alpha atresia of advanced yolked
oocytes was detected at least twice as
frequently using histological tech-
niques. The histological method was
more sensitive because we could detect
more subtle changes in oocyte struc-
ture and because we scanned about 150 oocj^es per
ovary, compared with 30 oocytes in the anatomical
method. Despite the lack of sensitivity, the anatomical
method was valuable because the standing stock of
atretic oocytes could be easily estimated and directly
related to total fecundity.
The histological evidence indicated that females with
a-atretic advanced yolked oocytes were more common
in central California waters than off Oregon. However,
season and locality were confounded because most
females from Oregon were taken prior to the spawn-
ing season while most females from California were
taken during the season. To determine if either season
or locality affected the relative frequency of atretic
females, we combined the minor-atresia and major-
atresia classes for California and Oregon and fit the
Table 12
Histological determination of number of Dover sole Microstomus pacificus, with
a atresia of advanced yolked oocytes expressed as a percentage of all females with
advanced yolked oocytes taken in central California and in Oregon, beginning
(November-December) and during (January-May) the spawning season.
Spawning season
Beginning
During
Beginning + During
%
N
N
%
95% CI
N
Central California 59.2 71 51.9 617 52.5 47.9-57.2 688
Oregon 36.8 536 49.4 83 38.4 33.8-43.3 619
Table 13
Mean relative batch fecundity for five batch-order numbers (B), from the first batch
to the last batch.
Relative batch fecundity
Batch-order no.^
(oocytes/female wt(g))
No. of
(B)
females Mean SD
1
4 2.421 3.810
2
12 11.661 4.410
2<B<U-1
19 10.489 3.184
U-1
11 12.378 8.035
U
9 7.835 5.723
Analysis
of
variance on five batch orders
Source DF
SS MS F
P
Batch order 4
371.2 92.8 3.44
0.015
Contrast'' 1
352.6 352.6 13.08
0.001
Error 50
1347.6 27.0
Total 54
-1;
1718.8
last batch spawned, B = U; penultimate batch,
'First batch spawned, B =
B = U-1.
''Comparison of relative fecundity of the first and last batch to the other batches.
stepwise logistic model
l + eCo+PiXi+zJoXz
(Eq. 11)
to the data (Table 12), where P is the fraction of females
with atretic oocytes. The independent variables for
location (Xi) are - 1 for California and 1 for Oregon,
and for season (X2) are - 1 for prespawning and 1 for
during spawning. The estimates of coefficients for the
equation are /Jo = -0.183 (SE 0.056) and /3, = -0.288
(SE 0.056) {P2 is not given because effect of season
was not significant in an early regression analysis); the
estimate of the atresia rate P for California is 0.525
(95% CI 0.479-0.572; Carter et al. 1986) and for
Oregon is 0.384 (95% CI 0.338-0.433).
Hunter et al : Fecundity, spawning, and maturity of Microstomus paaficus
121
Table 14
Comparison of fecundity between the penultimate batch and |
the last
batch within
the same Dover sole Microstomus \
pacificiu
■ female.
Batch fecundity*
(no. of oocytes)
(g)
Penultimate
Last
270.00
4634
2121
270.00
3716
323
324.50
9374
1025
539.00
11445
2903
703.86
3906
411
713.24
12213
6455
752.11
2843
124
793.54
6134
5856
824.66
7022
3820
1017.70
3621
29
1247.82
11047
250
Mean
677.9
6905
2120
SD
309.8
df 10,
3525
P 0.0005.
2355
'Paired «-test: t 4.99,
•
0
•
•
• .
•
•
•
•
y*
.
•
• /
• •
-
•
•
0
-
•
• •
-
/>
•
-
•
•
o
• c
^y
•
-
y
•
. -
•
•
• o
o
•
-
•
,o
^ ,
o
i
0 200 400 600 aOO 1.000 1,200 1.400 1,600
FEMALE WEIGHT (gm)
Figure 9
Batch fecundity of Dover sole Microstomus paaficus as a func-
tion of female weight (without ovary). Line is Yg = 10. IW for
the second through the penultimate spawning batches (filled
circles); triangles = first spawning batch; open circles = last
spawning batch.
This computation indicated that the occurrence of
females with a atresia of advanced oocytes was
significantly affected by locality of the samples but not
by season. In short, more California females had one
or more a-atretic advanced oocytes in their ovary than
did Oregon females.
Batch fecundity
The first step in our analysis of batch fecundity (Yg)
was to determine if the batch size varied with the order
of spawning. Analysis of variance indicated that a
significant batch-order effect existed (Table 13). The
mean relative fecundity of the first (1) and last spawn-
ing batch (U) were significantly lower than the other
batches (Table 13).
In eleven females, the only advanced oocytes left in
the ovary were two "hydrated" spawning batches
(Table 14). Each was in a different stage of develop-
ment: one was fully hydrated (last batch), and the
other was in the migratory nucleus stage (penultimate
batch). In all of the eleven females, the last batch was
always lower than the penultimate batch. The t-test
for paired differences confirmed the effect of batch
order on fecundity indicated by the ANOVA. The t-test
also had less potential for bias because we used absolute
rather than relative batch fecundity. The (-test in-
dicated that the fecundity of the last batch differed
from the penultimate batch {t 4.99, df 10, P 0.0005).
We concluded that the batch size of a female Dover sole
did change over the spawning season, with the last and
the first batch being lower than the rest.
We determined the relation between batch fecundity
and weight using regression analysis. We did not use
the first and last batches since they were lower than
the rest. The intercept for the regression of batch
fecundity on female weight did not differ from zero (a
2142; t 1.87, df 40, P 0.07). Therefore we forced the
regression line through 0, yielding the relationship
Yb = 10. IW, where female weight ranged from 148 to
1464 g (Fig. 9). This analysis indicated that the relative
batch fecundity of Dover sole is about 10 oocytes per
gram ovary-free female weight, except for the first and
last batch. The relative fecundity for the first and last
batches combined was also about 10 oocytes per gram
(1 and U in Table 13). Thus the number of potential
spawnings (S) per year can be calculated using S =
(Yfr/10)-i- 1, where Yp^ is the relative potential an-
nual fecundity (Yp/W; Yp from Eq. 10). This means
that the average 1kg female spawns its 83,000 ad-
vanced yolked oocytes in about nine batches.
Sexual maturity
To determine the optimal criteria for sexual maturity
in female Dover sole, we established six sets of histo-
logical criteria for maturity (Table 15). The first set of
criteria selects females with either advanced yolked
122
Fishery Bulletin 90(1). 1992
Table 1 5
Six sets of histological criteria for female sexual maturity in Dover sole Microstomus pacifiais,
with the mean
length of the females |
in each set. (o) not present; ( + ) present; (-) not considered.
Advanced
Post-
Early yolked
No. of females
yolked oocytes
ovulatory
oocytes
Unyolked oocytes
Mean
Calif.
+ Oregon
foHicles
length (mm)
(N 2595) 1
only a
Criteria
with a no a
with /? atresia
with /? only a
no
No.
Cumulative
set no. Certainty
atresia atresia
atresia or none
atresia atresia
atresia
X
±2 SE
in class
percent
1 Certain
maturity
+ +
+
-
-
-
434
±3
1343
52
2 Uncertain
0 0
0
+ —
-
414
±7
218
60
3 Uncertain
0 0
0
0 +
-
-
397
±11
77
63
4 Uncertain
0 0
0
0 0
+ -
-
379
±6
279
74
5 Uncertain
0 0
0
0 0
0 +
-
350
±5
432
90
6 Certain
0 0
0
0 0
0 0
+
297
±10
246
100
immaturity*
laturity because
no histological evidence exists for maturity.
* Defined as certain imn
oocytes or postovulatory follicles. The sexual matur-
ity of these females is certain, but some mature females
may be excluded if only the first set of criteria are used.
Criteria sets 2 to 5, if added to the first set, broaden
the maturity definition to include females having
ovaries in the earliest stages of vitellogenesis and those
showing possible signs of past reproductive activity (/?-
or a-stage atresia). Each additional criteria set that one
might add to the first set increases the risk that im-
mature fish will be classed as mature. Females in set
6 are considered to be immature because they have
none of the characteristics mentioned in the other five
sets.
Use of p atresia as a possible sign of past reproduc-
tive activity seems justified. Females with early yolked
oocytes and p atresia (set 2) were larger on the average
than those with no /3 atresia (set 3; t 2.45, P 0.015, df
293); and females with unyolked oocytes and ft atresia
(set 4) were larger than those with only a atresia of
the unyolked (set 5; t 7.69, P<0.001, df 709). In addi-
tion, the ranking of criteria sets based on our intuitive
appraisal of the risk of classification error is largely
borne out by the length distributions of the females
identified by the criteria set, since mean length de-
creased with criteria set number (Fig. 10).
To estimate the length at which 50% of the Dover
sole are mature (ML50) using all six histological
criteria sets, we first used a maturity algorithm to
estimate the fraction of fish that were mature in a given
length-class. This algorithm is a regression method
similar to those used to construct age-length keys
(Bartoo and Parker 1983, Kimura and Chikuni 1987).
Figure 10
Length distribution of female Dover sole Microstomus pacif-
mis identified by six sets of histological maturity criteria, rang-
ing from criteria set 1 where maturity is certain, to set 6 where
all females are considered to be immature (see Table 15).
Females captured in November-December in California and
Oregon; filled triangles indicate mean length of females.
This analysis was based on two equations. The first
equation was
Qjji = Qmii qjjm + (1 " Qniji) ^jji
(Eq. 12)
where qj | i is the fraction of fish of length-class j in the
ith criteria set; qjim = Ijii and qjiim = qj|6 because the
criteria set 1 consists of all mature (m) fish and the
criteria set 6 consists of all immature (im) fish; and
q^ii the overall fraction of mature fish in the ith cri-
Hunter et al.: Fecundity, spawning, and maturity of Microstomus paaficus
123
teria set. The second equation assumed that qn,|i
changed hnearly with criteria set number i:
Cbiili
b, + boi.
(Eq. 13)
Combining equations (12) and (13) results in the final
equation
Yij = bixij + b2X2j (Eq. 14)
where yij =qj|i - qj|™,
xij =qj|m - qjiim, and
X2j =i(qj|m - qjlim)-
For each criteria set i, we obtained the estimate of the
fraction of mature fish (qm|j|i) in each length-class j as
ImlJli
qm.jii/qjii = [qmiiqjim]/qj
ii-
We then obtained the estimated number of mature fish
of length-class j in the criteria set i (iVmijii) as the pro-
duct of the total number of fish at length-class j (A^j | i )
and the estimated fraction of mature fish at length-
class j (q^ijii):
■^mljli = -^jli qniljli-
The summation of iVm|j|i over all criteria (i) is the total
number of mature fish at length-class j (7^m|j = Zi
iV^ijii). The total number offish in length-class j (ATj =
Zi iVjii) and the number of mature fish (iV^ij) were
used to estimate ML50 for all females taken before the
onset of spawning (California and Oregon data com-
bined) using BMDPLR (Dixon et al. 1988).
We compared the above estimate with the ML50 for
each of the five maturity definitions created by in-
cluding progressively more criteria sets (Table 16).
When the definition of sexual maturity is expanded by
progressively adding criteria sets 2 to 5 to the defini-
tion, the ML50 decreased for each additional set of
criteria added. Our estimate of ML50 from the model
was 332 mm and is most similar to maturity definition
IV in Table 16. Thus, definition IV is the preferred
histological definition of maturity because it is probably
the least biased.
Inspection of Table 16 also indicated that the ML50
is always greater when measurements are made dur-
ing the reproductive season than before it begins,
regardless of the number of criteria sets used to define
sexual maturity. This implied that during the spawn-
ing season the ovaries of some postspawning females
are reabsorbed to the extent that they become indis-
tinguishable from females defined as immature. Thus
maturity should be estimated prior to the onset of
spawning, and the definition of maturity should be
broader than definition I.
We believe the preferable estimate of ML50 is one
based on the maturity algorithm because it uses all the
histological data, while those based on definitions use
only a portion of it. The maturity algorithm should be
applied only to data taken before the spawning season,
since data collected later in the season will be biased.
This method demands detailed histological classifica-
tion which may be too costly for many purposes. Defini-
tion III could be used if tissue were examined micro-
scopically or with a powerful hand lens, and it gives
an ML50 value close to that provided by the model.
Table 16
Estimated length at which 50% of Dover sole females are sexually mature, using six histological definitions of ovarian maturity and
a maturity algorithm that uses all data. California and Oregon data are combined; length at 50% mature estimated using logistic model
(Dixon et al. 1988).
Definition no.
Histological
criteria sets
incl. in maturity
definition"
Before spawning
(N 854 females)
During spawning
(N 1321 females)
Length at 50%
mature (mm)
No. of mature females
No. of mature females
N
% of
females
Length at 50%
mature (mm) N
%of
females
I
II
III
IV
V
Maturity algorithm
1
1,2
1, 2, 3
1, 2, 3, 4
1, 2, 3, 4, 5
1, 2, 3, 4, 5, 6
turity algorithm.
373
361
348
832
258
332
541
582
626
669
810
691"
63
68
73
78
95
81''
419 568
396 692
391 720
348 917
255 1184
389 742"
43
52
54
69
90
56"
'From Table 15.
''Estimated from ma
124
Fishery Bulletin 90(1). 1992
Table 1 7
Final maturity thresholds and logistic model parameters* for female Dover sole Microstomus pacifieus
season in California, Oregon, and the two states combined.
taken before the spawning
Region Maturity definition
50% mature
95% CI
a
SE
b
SE N
California IV
Oregon IV
California + Oregon Maturity algorithm
298
336
332
215-391
322-351
305-367
-14.412
-9.268
- 14.960
4.374
0.806
1.239
0.0483
0.0276
0.0450
0.0149 104
0.0022 750
0.0036 854
„a + bL
* P
l+e-""-
We combined data from Oregon and California in this
analysis because sample sizes (before spawning) were
inadequate for application of the model separately.
However, in Table 17 we provide estimates and fitting
parameters based on definition IV for each region as
well as those based on the model using the combined
data. The ML50 estimated for definition IV was lower
in California than in Oregon. However, analysis of co-
variance of the log transformation of fraction mature,
Ln + 1 , on length and locality indicated that
the difference between states was not significant (P
0.625, F 0.26). Thus our ML50 estimate for Dover sole
along the California and Oregon coasts is 332 mm with
95% CI of 315-349 mm (Carter et al. 1986; Fig. 11).
The ML50 we estimated from data in Hagerman (1952)
for Dover sole from the Eureka California fishery is
high (363 mm) compared with our final estimate for
Oregon and California coasts (Fig. 11). However,
Hagerman collected his specimens during the spawn-
ing season, and his estimate is similar to the ML50 for
females taken during the spawning season (Definition
IV, Table 16).
Discussion
Validation of fecundity assumptions
In the Introduction, we specified four assumptions re-
quired for an unbiased estimate of annual fecundity in
Dover sole. These assumptions were that (1) fecundity
was determinate; (2) potential annual fecundity was
equivalent to actual fecundity; (3) females used to
estimate annual fecundity had not spawned; and (4)
recruitment of oocytes into the advanced stock of
yolked oocytes had ceased for the season. The follow-
ing is a review of the evidence for the four assumptions.
Five lines of evidence support the assumption of
determinate fecundity for Dover sole: (1) in mature
FRACTION OF FEMALES MATURE
0000 —
00000
CURRENT STUDY / /
7 / HAGERMAN
/ " /
y/ /-Xil 363
" "" ' 220 260 300 340 380 420 460 500 540
TOTAL LENGTH (mm)
Figure 1 1
Fraction of female Dover sole Microstomus pacificits that were
sexually mature as a function of total length. Points are for
10 mm length-class intervals; the fraction mature per length-
class was assigned using maturity algorithm; data were from
California and Oregon in November-December; logistic
parameters are in Table 17. Present study is compared with
Hagerman (1952), line only (logistic curve parameters, a
-17.854, b 0.0450).
ovaries (mean diameter of advanced oocytes >0.85
mm), a hiatus existed between the advanced stock
of mature oocytes and smaller, less mature oocytes;
(2) total fecundity declined over the spawning season;
(3) total fecundity was lower in females having
postovulatory follicles; (4) the mean diameter of the
advanced oocytes increased over the spawning sea-
son; and (5) our analysis of the order of spawning
batches was consistent with the determinate fecundity
assumption.
The second assumption, lack of significant atresia,
also was supported by our analysis. Overall, atretic
losses of advanced oocytes were negligible during the
years of our study. Multiple regression analysis in-
dicated that atresia had a small but significant effect
on total fecundity of the California females that had
Hunter et al : Fecundity, spawning, and maturity of Microstomus pacificus
125
atretic oocytes. A few females suffered substantial
losses in total fecundity, but such fish were rare and
they had little effect on population means. Histological
and anatomical evidence indicated that females with
a-atretic advanced yolked oocytes were more common
in central California than in Oregon waters. Atresia
might be more common in central California Dover sole
because bottom sediments are contaminated. Alter-
natively, females with atretic ovaries may be more com-
mon in central California waters because they are living
near the southern end of their range where food supply
and other habitat conditions may be less than optimal.
Both explanations seem equally plausible at present.
The third assumption, that females used to estimate
potential annual fecutidity have not spawned in the cur-
rent reproductive year, would be rejected for females
taken in January through May. The assumption prob-
ably held for the females used to estimate annual fecun-
dity in November- December because only 2.9% of the
females from California and only 1% of the females in
Oregon showed any histological signs of past or immi-
nent spawning. The few females that showed histo-
logical signs of spawning were not used, of course, to
estimate annual fecundity. Spawning may have gone
undetected in some of the females used to estimate
fecundity since postovulatory follicles are eventually
resorbed. This does not seem likely for the November-
December case because the spawning season had just
begun and resorption is probably slow at the low tem-
peratures of Dover sole spawning habitat.
Our fourth assumption, that all the oocytes that con-
stitute the potential annual fecundity were included in
our oocyte counts, is supported by two lines of evi-
dence. The first is that no positive correlation existed
between the mean diameter of the advanced oocytes
and total fecundity. Such positive correlations were
eliminated by excluding all ovaries in which the mean
diameter of the advanced oocytes was less than
0.86 mm. A positive correlation between diameter and
fecundity existed when all ovaries were considered
(range in mean diameter of the advanced oocytes,
0.71-1.04 mm). This is evidence that recruitment of
oocytes into the advanced class continued until the ad-
vanced stock was well separated from early vitellogenic
oocytes (stages 1 and 2, Fig. 6). The second source of
evidence is the form of the oocyte size-frequency
distribution. A prominent gap between stage-2 and
stage-3 oocytes existed when the mean diameter of
stage-3 oocjd;es was between 0.84 and 0.96mm (Fig.
6). The absence of significant numbers of oocytes in the
intervening diameter classes (0.55-0.65 mm) indicates
maturation of oocytes across this range either had
ceased or was proceeding at a very slow pace. We con-
clude that recruitment of significant numbers of
oocytes into the advanced stock probably ceases in
Dover sole when the mean diameter of the advanced
stock is between 0.86 and 0.96 mm.
Some authors working with other species (Hislop and
Hall 1974 on Melangius merlangus (L.), Horwood and
Greer Walker 1990 on Solea solea) consider all yolking
oocytes to comprise the potential annual fecundity. In
Dover sole this would mean that in addition to stage
3, the most advanced yolked oocytes, stages 1 and 2
would also be used to estimate annual fecundity. Such
broad criteria are acceptable if all oocytes that began
vitellogenesis ultimately become a part of the mature
stock of oocytes that are spawned. This was not the
case in Dover sole because oocytes in the early stages
of vitellogenesis (stages 1 and 2) occurred in nearly all
mature ovaries, including those in which some of the
batches had already been spawned. The fate and
dynamics of these small partially-yolked oocytes in ad-
vanced ovaries is uncertain; their numbers might either
decrease due to resorption, increase and become part
of next year's production, or remain in stable numbers
until later in the year. It would seem impractical to ad-
just estimates of potential annual fecundity based on
all vitellogenic oocytes for the fraction of those oocytes
which do not continue vitellogenesis. Therefore, we
believe use of the more mature yolked oocytes for
estimating the potential annual fecundity is preferable.
An important implication of our discussions of the
third and fourth assumptions is that timing the sam-
pling of females is a critical element in estimating
potential annual fecundity: Sample too early in the
reproductive cycle and the ovaries are not sufficiently
mature; sample too late and spawning is prevalent. The
optimal time to sample Dover sole ovaries is when the
average diameter of the advanced stock is between 0.86
and 1.1mm (Fig. 12). When the diameter is less than
0.86 mm, the numbers of advanced oocytes are still in-
creasing (indicated by the t value for the diameter coef-
ficient in the fecundity equation. Fig. 12). When the
diameter exceeds 1.1mm, 20% or more of the females
show histological signs of past or imminent spawning,
and the assumption of no spawning cannot be safely
made.
Spawning rates and reproductive energetics
The spawning season of Dover sole was protracted with
postovulatory follicles occurring as early as December
and hydrated oocytes as late as May, indicating a
season of six months. This is a long season for a fish
of determinate fecundity, since typically they are high-
latitude species with short, 1-2 month spawning
seasons. Batch fecundity was low, averaging about 10
oocytes per gram female weight, except for the first
and last batch which average about 5 oocytes per gram.
Dover sole spawn about nine times during their
126
Fishery Bulletin 90|l). 1992
protracted spawning season. Vitellogenesis
does not cease when spawning begins, but
rather it continues throughout most of the
season as the advanced stocks of yolked
oocytes are matured and spawned. Spawn-
ing frequency appears to increase near the
end of the season. This may cause a higher
daily production of eggs by the population
in late-March or April than in February,
even though fewer females have active
ovaries in April.
To estimate reproductive effort of Dover
sole, we calculated the hypothetical weight
of the ovary when the entire advanced stock
of oocytes had completed vitellogenesis and
hydration had begun. The weight is hypo-
thetical because a Dover sole ovary never
contains a full complement of completely
yolked oocytes, since vitellogenesis of the
smaller advanced yolked oocytes continues
after a female begins spawning. To compute
the hypothetical weight, we assumed that
all oocytes completed vitellogenesis when
their average diameter was 1.5mm. Hydra-
tion begins when the advanced yolked
oocytes have a mean diameter of 1.3-1.7
mm. We estimated the gonad weight of a
lOOOg female with oocytes having a mean
diameter of 1.5 mm, using an equation in
which gonad weight was expressed as a
function of fish weight and volume of the
average advanced oocyte (1.5 mm diameter
has a volume of 1.77 mm^; California plus
Oregon data; Table 3). The ovary was esti-
mated to weigh 144g, or about 14% of the
body weight. In other words, the annual
reproductive effort of Dover sole was about
14% per year, and this effort was distrib-
uted over about nine spawnings averaging
about 1.6% of their body weight per spawning. Gonad
weight was considered to be a measure of reproduc-
tive effort by Gunderson and Dygert (1988); but they
did not adjust the gonad weight for the full complement
of yolk, and consequently their estimates are not com-
parable to these.
Assessment of sexual maturity
Our estimates of length at 50% mature (ML50) were
higher when females were taken during the spaviming
season than when they were sampled before spawning
began, regardless of the histological criteria used. Thus,
during the spawning season ovaries of some post-
spawning females had regressed far enough that they
were histologically indistinguishable from immature
X
Lli
Q
<
a.
O
z
3 0 5 10 15 2 0
MEAN DIAMETER OF ADVANCED STOCK (STAGE 3) (mm)
Figure 12
Optimal range of mean diameter of advanced yolked oocytes (stage 3) for
determining potential annual fecundity of Dover sole (shaded area, 0.86-
1.1mm). Open circles, Student's t as a function of the minimum mean
diameter of stage-3 oocytes included in the fecundity data set: when t » 2
(P 0.05), the mean oocyte diameter had a positive correlation viith fecun-
dity, indicating the stock of yolked oocytes for the season was not fully
recruited into stage 3. Spawning rate index (filled circles) for females used
in fecundity estimates; index = 0 when no females show signs of past or
imminent spawning, and index = 1 when all females have postovulatory
follicles or hydrated oocytes. Upper panel indicates diameter range of
oocytes in three yolked oocyte stages and hydrated oocytes. Only stage-3
oocytes were used to estimate potential annual fecundity.
females. This finding has two important implications:
First, it indicates that even the broadest histological
criteria, based on analysis of H&E sections, will not
identify all postspawning females; second, it means that
estimates of length or age at first maturity should
always be conducted prior to the onset of spawning,
when postspawning females with highly regressed
ovaries are rare.
Another limit to our ability to assess sexual matur-
ity is that we do not know how many of the females
that begin vitellogenesis actually complete it during
the current reproductive season. Dover sole ovaries
with oocytes in the early vitellogenic stage occurred
throughout the spawning season as well as before it
began, indicating some females that begin vitellogen-
esis may not reach sexual maturity in the current
Hunter et al : Fecundity, spawning, and maturity of Microstomus pacificus
127
season. At this time, it is an arbitrary choice to con-
sider as mature all females with vitellogenic ovaries or
only those with advanced yolked oocytes. Our analysis
showed that this arbitrary decision had a pronounced
effect on ML50 estimates. Thus the criteria for matur-
ity estimates should be precisely specified. It is par-
ticularly important to specify the minimum level of
oocyte development necessary for a female to be con-
sidered as mature. Our preferred definition of matur-
ity included females in the early stages of vitellogen-
esis with yolked oocytes as small as 0.18mm diameter,
and also included some females without vitellogenic
oocytes (maturity IV, Table 16). Those females without
vitellogenic oocytes had ft atresia in the ovary. We
believe that the presence of some /3 atresia is an in-
evitable consequence of the resorption of an active
ovary or ovulation.
No discussion of sexual maturity would be complete
without mentioning the gross anatomical systems used
to classify ovaries, because they are the chief method
used by fishery biologists to measure sexual maturity
in marine fishes. Using gross anatomical criteria, we
accurately separated active ovaries (advanced yolked
oocytes present) from inactive ovaries (no advanced
oocytes) with classification errors of 1-12%. Deter-
mining sexual maturity is a far more difficult task,
however. Identification of mature females using gross
anatomical methods has the same problems with post-
spawning and early vitellogenesis criteria as histo-
logical methods, but the potential for bias is greater.
Anatomical criteria are less accurate and may be
detectable for shorter periods than histological ones.
For these reasons, differences between maturity
studies should be interpreted with caution, especially
when done by different observers, or with different
methods, or when sampling at different times of the
year. Many investigators have not been particularly
careful to restrict sampling to early in the spawning
season. The tendency will be to overestimate the ML50
using anatomical methods, especially when samples are
taken midseason.
In an earlier paper on Dover sole. Hunter et al. (1990)
concluded that size at 50% mature in Dover sole from
central California in the 1980s differed from that of
Dover sole in northern California in the late 1940s as
determined by Hagerman (1952). Although a statistical
difference existed between these two data sets, we are
inclined to dismiss this difference, since it could be due
to differences in criteria and sampling times. Similar-
ly, Yoklavitch and Pikitch (1989) speculated that size
at 50% maturity of Oregon Dover sole has changed
because their estimate of maturity differed from Harry
(1959). We beheve that this difference also could easily
be due to differences in criteria and timing of sampling.
Our analysis of histological criteria for maturity clearly
shows that differences in criteria or timing of sampl-
ing can produce differences in the ML50 as large as
any of those seen in the Dover sole literature.
Acknowledgments
We appreciate and thank M. Yoklavich and E. Pikitch
for providing original data and supplying some speci-
mens, and W.W. Wakefield for providing some speci-
mens. We thank all on shipboard who helped collect
Dover sole ovaries: E. Lynn, W. Flerx, R. Dotson,
R. Leong, E. Acuna, and D. Squires. We thank all
others who served on the scientific crews and the crews
of NOAA ships David Starr Jordan and Miller Free-
man; they contributed greatly to the success of the
cruises. Processing of laboratory specimens and estima-
tion of fecundities were assisted by W. Kicklighter,
M. Drawbridge, R. Leong, E. Lynn, D. Ramon, and
S. Swailes. Computer programs were written by
C. Vedovato, R. Young, and J. Butler. L. Jacobson pro-
vided suggestions on the modeling of sexual maturity.
Illustrations were produced by R. Allen and H. Orr.
K. Schaefer, J. Zeldis, and an anonymous reviewer
reviewed the manuscript.
Citations
Alheit, J.
1986 Reproductive biologj' of sprat, Sprattus sprattus : Fac-
tors determining annual egg production. ICES CM 1986/H:58,
16 p.
Bartoo, N.W., and K.R. Parker
1983 Stochastic age-frequency estimation using the von Ber-
talanffy growth equation. Fish. Bull., U.S. 81:91-96.
Bretschneider, L.H., and J.J. Duyvene de Wit
1947 Sexual endocrinology of non-mammalian vertebrates.
Monogr. Prog. Res., Vol. II. Elsevier, NY.
Carter, W.H. Jr., V.M. Chinchilli, J.D. Wilson, E.D. Campbell,
F.K. Kessler, and R.A. Carchman
1986 An asymptotic confidence region for the EDm^ from the
logistic response surface for a combination of agents. Am.
Stat. 40:124-128.
Dixon, W.J., M.B. Brown, L. Engelman, M.A. Hill, and
R.l. Jennrich
1988 BMDP statistical software manual, Vol 2. Univ. Calif.
Press, Los Angeles, 1234 p.
Gunderson, D.R., and P.H. Dygert
1988 Reproductive effort as a predictor of natural mortality
rate. J. Cons. Cons. Int. Explor. Mer 44:200-209.
Hagerman, F.B.
1952 The biology of the Dover sole, Microstomus pacificus
(Lockington). Calif. Fish Game, Fish Bull. 85:1-48.
Harry, G.Y.
1959 Time of spawning, length at maturity, and fecundity of
the English, petrale, and Dover soles (Parophrys vetulaus,
Eopsetta jordani, and Microstomus pacificus, respective-
ly). Fish Comm. Oreg. Res. Briefs 7:5-13.
128 Fishery Bulletin 90(1). 1992
Hislop, J.R.G., and W.B. Hall
1974 The fecundity of whiting, Merlangius merlangus (L.) in
the North Sea, the Minch and at Iceland. J. Cons. Cons. Int.
Explor. Mer 36:42-49.
Hjort, J.
1910 Report on herring investigations until January 1910.
Publ. Circ. Cons. Explor. Mer 53.
Horwood, J.W., and M. Greer Walker
1990 Determinacy of fecundity in sole (Solea solea) from the
Bristol Channel. J. Mar. Biol. Assoc. U.K. 70:803-813.
Hunter, J.R., and S.R. Goldberg
1980 Spawning incidence and batch fecundity in northern an-
chovy, Engraulis mordax. Fish. Bull., U.S. 77:641-652.
Hunter, J.R., and B.J. Macewicz
1980 Sexual maturity, batch fecundity, spawning frequency,
and temporal pattern of spawning for the northern anchovy,
Engraulis mordax, during the 1979 spawning season. Calif.
Coop. Oceanic Fish. Invest. Rep. 21:139-149.
1985a Measurement of spavining frequency in multiple spawn-
ing fishes. In Lasker, R. (ed.). An egg production method for
estimating spawning biomass of pelagic fish: Application to
the northern anchovy, Engraulis mordax, p. 79-94. NOAA
Tech. Rep. NMFS 36.
1985b Rates of atresia in the ovary of captive and wild north-
em anchovy, Engraulis mordax. Fish Bull., U.S. 83:119-136.
Hunter, J.R., N.C.H. Lo, and R.J. Leong
1985 Batch fecundity in multiple spawning fishes. In Lasker,
R. (ed.), An egg production method for estimating spawning
biomass of pelagic fish: Application to the northern anchovy,
Engraulis mordax, p. 67-78. NOAA Tech. Rep. NMFS 36.
Hunter, J.R., B.J. Macewicz, and C.A. Kimbrell
1989 Fecundity and other aspects of the reproduction of sable-
fish, Anopkrpoma fimbria, in Central California waters. Calif.
Coop. Oceanic Fish. Invest. Rep. 30:61-72.
Hunter, J.R., J.L. Butler, C. Kimbrell, and E.A. Lynn
1990 Bathymetric patterns in size, age, sexual maturity, water
content, and caloric density of Dover sole, Microstomus pacif-
ieus. Calif. Coop. Oceanic Fish. Invest. Rep. 31:132-144.
Kimura, D.K., and S. Chikuni
1987 Mixtures of empirical distributions: An iterative applica-
tion of the age-length key. Biometrics 43:23-25.
Lambert, J.G.D.
1970 The ovary of the guppy Poesilia reticulata. The atretic
follicle, a Corpus atreticum or a Corpus luteum praeovula-
tionis. Z. Zellforsh. 107:54-67.
Thigpen, C.C.
1987 A sample size problem in simple linear regression. Am.
Stat. 41:214-215.
Wathne, F.
1977 Performance of trawls used in resource assessment.
Mar. Fish. Rev. 39(6):16-23.
Yoklavich, M.M., and E.K. Pikitch
1989 Reproductive status of Dover sole, Mi^-rostomus paeificus,
off northern Oregon. Fish. Bull., U.S. 87:988-955.
Abstract.- Over the past sever-
al years researchers in Japan and the
United States have independently
been conducting extensive studies on
the early life history of two discrete
populations of walleye pollock Thera-
gra chalcogramma, trying to under-
stand recruitment variation. The
population of interest to Japanese
researchers spawns near Funka Bay,
Hokkaido, Japan, while the popu-
lation of interest to American re-
searchers spawns in Shelikof Strait,
Gulf of Alaska. This paper summar-
izes and compares characteristics of
spawning and ecology of eggs, lar-
vae, and early juveniles of the spe-
cies in these two areas. Although the
species has apparently adapted its
early-life-history pattern to environ-
mental differences in the two areas,
some underlying similarities exist.
The adults mainly spawn at a partic-
ular time of year following a spawn-
ing migration to a specific location
so that the eggs and larvae can reach
specific areas for subsequent devel-
opment. In both areas oceanographic
conditions are favorable for larval
food production (copepod nauplii)
when the walleye pollock larvae are
present. Drift of the eggs into the
bay, where copepod production is
enhanced, seems important in Funka
Bay, and drift of the larvae toward
juvenile nursery grounds on the con-
tinental shelf as opposed to being
swept offshore, seems important in
Shelikof Strait. Interannual differ-
ences in larval drift and food pro-
duction because of varying oceano-
graphic conditions may contribute
significantly to variations in year-
class size.
Comparisons of early-life-history
characteristics of walleye pollock
Theragra chalcogramma in
Shelikof Strait, Gulf of Alaska,
and Funka Bay, Hokkaido, Japan*
Arthur W. Kendall Jr.
Alaska Fisheries Science Center, National Marine Fisheries Service. NOAA
7600 Sand Point Way NE, Seattle, Washington 981 15-0070
Toshikuni Nakatani
Laboratory of Principles of Fishing Grounds, Faculty of Fisheries
Hokkaido University, Hakodate, Japan 041
Walleye pollock Theragra chalcogram-
ma is a dominant fish in the North
Pacific Ocean and in the Bering Sea,
both in terms of population size and
importance to commercial fisheries.
It is a major fishery resource in the
Funka Bay area on the Pacific Ocean
side of Hokkaido, Japan, and in Sheli-
kof Strait, Gulf of Alaska. In both
areas, most fishing is done just prior
to and during the spawning season.
In Funka Bay, walleye pollock are
caught in bottom gillnets, while in
Shelikof Strait midwater trawls are
used. In Funka Bay the catch varied
from about 4.3x10^ metric tons (t)
to about 10.7x10-4 t from 1976 to
1986. In Shelikof Strait, an intense
fishery on the spawmers existed from
1981 through 1988, although harvest
has been severely restricted since
1986 because of reduced abundance
of the population (Megrey 1989). The
harvest in Shelikof Strait peaked in
1984 at about SlxlC* t.
There is a growing interest in
understanding recruitment in this
species, and considerable work has
been conducted independently by
Japanese researchers in Funka Bay
and by U.S. researchers in Shelikof
Strait over the past several years.
Manuscript accepted 27 November 1991.
Fishery Bulletin, U.S. 90:129-138 (1992).
* Contribution FOCI- 138 to Fisheries Oceanog-
raphy Coordinated Investigations, NOAA.
This paper compares the results of
these studies (Table 1). While these
studies reveal that the early-life-
history strategy of walleye pollock
allows this species to adapt to differ-
ent environments, they also indicate
that underlying similarities exist be-
tween populations. Although under-
standing causes of recruitment varia-
tion in either area is a distant goal,
testable hypotheses have been devel-
oped in both areas. The comparisons
presented in this paper may help
researchers in both areas focus their
studies toward an understanding of
the recruitment process. They may
also guide future studies of the spe-
cies in other areas such as the Ber-
ing Sea.
Environmental
comparisons
Physical setting
Funka Bay is located in the southern
part of Hokkaido, Japan, at about
42°N (Fig. 1). Depths within the bay
are generally less than 80 m, although
there is a small area of water deeper
than 100 m in the center of the bay.
Immediately outside the bay the bot-
tom slopes evenly to 500 m within
45km. The area of the bay is 2270
km2.
129
130
Fishery Bulletin 90(1). 1992
Shelikof Strait is located in the
northern Gulf of Alaska between
the Alaska Peninsula and the
Kodiak Archipelago at about 57°N
(Fig. 1). Water depths within
Shelikof Strait exceed 300 m in
some areas. At the northeast and
southwest ends of the strait there
are sill depths of about 200 m.
Depths of greater than 500 m are
reached on the continental slope im-
mediately beyond the southwestern
sill. The southern part of the strait
and waters to the south comprising
about 12,450 km^, are the areas oc-
cupied by eggs and larvae of wall-
eye pollock originating in Shelikof
Strait.
Physical oceanography
The water of Funka Bay originates
from the seasonal influx of two
water masses: The Tsugaru Warm
Water and the Oyashio Water.
Tsugaru Warm Water enters the
Bay in late-summer when surface
waters exceed 15°C and there is a strong thermocline
in the upper 20 m (Nakatani 1988). Autumnal cooling
produces isothermal conditions and cooling to about
4°C (Winter Funka Bay Water: Ohtani and Kido 1980).
In late-winter or early-spring, the cold (<2°C), less
saline (<33.0"/oo) Oyashio Water usually intrudes into
the Bay above the Winter Funka Bay Water, produc-
ing a stratified condition with a temperature inversion.
In late-spring and early-summer, seasonal warming of
surface waters occurs and a thermocline develops.
Throughout the year, bottom temperatures remain at
3-6°C.
Shelikof Strait has an estuarine type circulation, with
less seasonal variation than Funka Bay. In its upper
layers, the Alaska Coastal Current (ACC) flows to the
southwest and is particularly pronounced on the Alaska
Peninsula side of the Strait. During runoff seasons
Gate-spring to early-fall), substantial amounts of fresh-
water enter the strait, primarily from Cook Inlet, and
flow along the Peninsula until thoroughly mixed with
the ACC. From approximately 150m to the bottom,
more saline water flows into the strait over the sill to
the southwest (Kim 1987). During April and May (when
walleye pollock eggs and larvae are present), near-
surface water temperatures in the ACC are generally
0-4°C, warming to 7°C by late May, while the deeper
waters are generally 4-5. 5°C. Salinity varies from
about 31 to 33.5''/oo.
Table 1
Comparisons of early-life-history characteristics of walleye pollock Theragra chalco-
gramma and their spawning environments in Funka Bay, Japan, and Shelikof Strait,
Gulf of Alaska.
Funka Bay
Shelikof Strait
Latitude
45°50'-42°35'N
56°00'-59°00'N
Area
2270 km=
12,450km2
Nominal annual catch
''70,000t
i" 100,000 1
Spawning season
'Dec-Mar.
'' Early April
Spawning depth
'■100-120m
" 200-300 m
Temperature at spawning depth
'■2-6°C
'5.5°C
Depth of maximum egg
'0-40m
'I50-200m
concentrations
Egg specific gravity
'1.020-1.026g/cm^
'1.024-1.031g/cm='
Depth of maximum larval
n0-20m
•■ 15-50 m
occurrence
Length of larvae when copepod
'^< 7-8 mm
''<llmm
nauplii are predominant
components of diet
Larval growth rate
1989, "^Maeda et al. 1976,
'■0.21 mm/day
'' Kendall and Picquelle
"Nakayama et ai. 1987, ''Megrey
1990, 'Nakatani 1988, 'Kendall and Kim 1989, ^Kamba 1977, "Kendall et al. 1987.
In Fimka Bay there is considerable interannual varia-
tion in the date when the Oyashio Water intrudes and
in the length of time that surface temperatures remain
cold (<3°C). In Shelikof Strait, interannual variation
in the frequency, intensity, and track of storms affects
water properties and transport.
Biological comparisons
Spawning
In Funka Bay, adult walleye pollock mature and spawn
from November to March, with peak spawning activ-
ity occurring in January and February (Maeda et al.
1976 and 1981, Yoon 1981). whereas in Shelikof Strait
most fish mature in February and March and spawn-
ing peaks in early April (Kim 1989, Kendall and Pic-
quelle 1990). Pelagic eggs are present in Funka Bay
from December until March, and in Shelikof Strait eggs
are present mainly in April. There is some interannual
variation in time of spawning in Funka Bay, and eggs
have even been collected in November and April
(Maeda et al. 1980). Thus the spawning season seems
to occur earlier in the year and lasts longer in Funka
Bay than in Shelikof Strait. Spawning occurs mainly
at depths of 100- 120 m near the entrance of Funka Bay
(Maeda et al. 1976, Nakatani 1988, Nakatani and
Maeda 1989). In Shelikof Strait, spawning is concen-
Kendall and Nakatani: Early life history of Theragra chalcogramma in Shelikof Strait and Funka Bay
131
OOE ISO ODE ISO OOE r70 OOE 160 OOU 170 OOW 160 OOW tSO OOW 140 OOU 130 OOW
I'l I I I I I I I ! I I t I I I I 1 I I I I I I I I I I Ill
158 OOW 1S7 OOW ISe OOW ISS OOW 1S4 OOU 153 OOW 152
40 OOH
56 30N
Figure 1
Location of Funka Bay,
Japan, and Shelikof Strait,
Gulf of Alaska. Insets are
enlargements of the areas
with pertinent bathymetry.
trated in a small area of deep water (> 250 m) near Cape
Kekurnoi (Fig. 1) (Kendall and Picquelle 1990).
Field surveys of adult walleye pollock in Funka Bay
and Shelikof Strait show that fish congregate and
migrate to a particular part of their range just prior
to the spawning season. Final migration to a restricted
spawning area takes place quickly. In Shelikof Strait,
hydroacoustic surveys show that the fish separate into
vertical strata, presumably by sex (females below
males) and readiness to spawn (Muigwa 1989).
Although the fish move to the spawning area as a
large group, spawning itself is by pairs. Behavior of
spawning walleye pollock has been investigated using
captive fish from Funka Bay (Sakurai 1982, 1989), as
well as from Puget Sound, Washington (Baird and 011a
1991). The shallow tanks used by Sakurai (1989) may
have prevented some of the vertical aspects of spawn-
ing behavior observed by Baird and 011a (1991).
Although no such studies have been conducted on fish
from Shelikof Strait, similarities between the behavior
of fish from near the eastern (Puget Sound) and
western (Funka Bay) extremes of the species distribu-
tion may indicate that spawning behavior varies little
geographically. In experimental tanks, the fish form
loose aggregations near the surface. Males frequently
follow other males and females. Sakurai (1989) related
male-male interaction to the agonistic behaviors
associated with dominance; Baird and 011a (1991) con-
sidered the male's following behavior as a searching
behavior for potential mates. Sakurai (1989) also
observed courtship displays by males toward prospec-
tive mates. At the onset of a spawning, a female would
swim down with a male following her. The male then
made contact with her by rubbing his ventral surface
first against her dorsum or side and then he swam
beneath her, with their two vents in contact. Other
males occasionally followed the pair closely and also
made contact with the female. During vent-to- vent con-
tact, the male rubbed his body rapidly against the
female's abdomen, and presumably gametes were
released at this time (they could not be seen in the
water, but were found in the tank overflow within an
hour). Most spawning took place in evening or morn-
ing twilight (Baird and 011a 1991).
Female walleye pollock characteristically spawn a
number of batches of eggs over a fairly short period
132
Fishery Bulletin 90(1), 1992
each year. The interval between batches is a few days.
The number of eggs per batch and size of eggs decrease
with successive batches. These patterns have been
observed both in Funka Bay (Sakurai 1982) and
Shelikof Strait (Hinckley 1990).
Fecundity
Miller et al. (1986) related fecundity of walleye pollock
from Shelikof Strait to gutted weight and fork length,
while Sakurai (1982) related fecundity of walleye
pollock from Funka Bay to whole weight and body
length. Conversions were applied here to the Funka
Bay length and weight data so fecundity could be com-
pared with Shelikof Strait values based on
Y = 0.7634X + 23.4472 (r^ 0.96628, N 40)
where X = body weight and Y = gutted weight
(Y. Sakurai, unpubl.); and
Y = 1.0659X + 4.050 {r^ 0.9959, N 53)
where X = body length and Y = fork length (T. Maeda,
unpubl.).
The relative fecundity of Funka Bay fish is repre-
sented by the relationship F = 8.73x lO-^^L^-ss and
F = 106.2 Wi-21, where L = body length in mm and W
= body weight in grams (A^ 94) (Sakurai 1982); there-
fore a 300 g (gutted weight) fish produces 129,000 eggs
and a 1000 g fish yields 589,000 eggs. In Shelikof Strait,
the relationship was found to be F = 1.2604L2-2169 a,nd
F = 387.4551 W'oieo (N 60), where L = fork length
in cm and W = gutted weight in grams; this yields
127,000 eggs for a 300g fish and 433,000 eggs for a
lOOOg fish (Miller et al. 1986). Thus small fish from
Funka Bay have about the same number of eggs, but
larger fish have more eggs than those from Shelikof
Strait (Fig. 2).
Eggs
Development Eggs from Funka Bay are more vari-
able in size and slightly larger than those from Sheli-
kof Strait. In Funka Bay, eggs are 1.15-1. 68mm (i
1.46mm) in diameter (Nakatani and Maeda 1984,
T. Nakatani, unpubl.). In Shelikof Strait, egg diameter
ranges from 1.30 to 1.41mm, and egg size has been
shown to vary interannually and decrease during the
spawning season (Hinckley 1990).
Eggs from Funka Bay develop at a rate dependent
on temperature according to the relationship
D = 31.70 exp(-0.12T),
where D is days to 50% hatch and T is temperature
^ 600
c
(0
-•- Funka Bay — H Shelikof Strait
3 500
1—
— 400
Ui
U)
" 300
"o
+
S 200
E
+
1-
^ 100
*
c
200 400 600 800 1000 1200 1400 1600
Gutted weight (gm)
Figure 2
Fecundity-weight relationship for walleye pollock in Funka
Bay, Japan (based on Sakurai 1982), and Shelikof Strait, Gulf
of Alaska (based on Miller et al. 1986).
(°C). Thus 50% hatch times are 22.1 days at 3°C, 17.4
days at 5°C, and 15.4 days at 6°C (Nakatani and Maeda
1984). No measurements of incubation time are avail-
able for eggs from Shelikof Strait; however, reared
eggs from Auke Bay in southeast Alaska (58°20'N) re-
quired 19.2 days at 3°C, 14.1 days at 5°C, and 12.2
days at 6°C for 50% hatch (Haynes and Ignell 1983).
Thus eggs from southeast Alaska developed to hatching
more quickly, by about 2-3 days, than those from
Funka Bay (Fig. 3).
Vertical distribution The vertical distribution and
buoyancy of eggs have been investigated in both Funka
Bay and Shelikof Strait. In Funka Bay, eggs rise in the
water column as they develop. Stage-1 (fertilization to
morula) eggs were found at a depth of roughly 30 m
(10-40 m), whereas Stage-5 (embryo more than three-
fourths yolk circumference) eggs were mainly at depths
of 10-20m (Nakatani 1988). The specific gravity of
Funka Bay eggs throughout development was within
a range of 1.020-1. 025 g/cm^ (x 1.0226g/cm3). This
resulted in an upward velocity of 4.9 m/h in ambient
water through the homogenized water column early in
the spawning season (o' 26.41-27.17), and is consis-
tent with field observations of shallower depths for
older eggs compared with those recently spawned
(Nakatani and Maeda 1984, Nakatani 1988).
In Shelikof Strait, the vertical distribution of eggs
changes during development in response to their
changing specific gravity. Newly spawned eggs are
positively buoyant, and thus rise from the deep loca-
tions where they are spawned. In middle stages of
development, the eggs become heavier and sink until
just before hatching when they again rise toward the
surface (Kendall and Kim 1989). The specific gravity
Kendall and Nakatani: Early life history of Theragra cha/cogramma in Shelikof Strait and Funka Bay
133
-•- Nahalan.
+
■
— t— Haynes S Igoell
-*- Paul
■
+
-:;
Incubation temperature C
Figure 3
Incubation period of walleye pollock eggs from Funka Bay,
Japan (based on Nakatani 1988); Resurrection Bay (based on
A.J. Paul, Univ. Alaska, Seward, pers. commun.) and Auke
Bay, Alaska (based on Haynes and Ignell 1983).
all stages was 350,000 eggs/m^ in the area of max-
imum concentration; about 15 times the maximum
abundance observed in Funka Bay. By late April, egg
abundance is reduced as eggs are spread more evenly
throughout the southern two-thirds of the strait and
the area immediately to the southwest of the strait. By
late May, egg abundance is further reduced, but the
area of occurrence is still similar to that seen in late
April. As opposed to Funka Bay, there is little evidence
of drift of eggs in Shelikof Strait. It appears that the
adults spawn some eggs in the southwestern part of
the strait as they move toward the main spawning area
off Cape Kekurnoi. Later spawning in late April and
May seems to be dispersed throughout the strait and
occurs at a much reduced level. Measurements of cur-
rents in Shelikof Strait also indicate that little drift
would be expected in the deep waters (>150m) where
most eggs occur (Kendall and Kim 1989).
of eggs from Shelikof Strait varied from 1.0243 to
1.031g/cm3, whereas the water density varied from
1.0256 to 1.0259g/cm3 (in 1985). Less than 20% of
eggs of all ages occurred above 162 m in Shelikof Strait.
Over 80% of early- (fertilization to morula) and late-
stage eggs (embryo more than one-half circumference
of yolk to hatching) occurred between 216 and 277 m
(near bottom), while over 60% of middle-stage eggs
(gastrula) occurred between 162 and 216 m (Kendall
and Kim 1989). Thus eggs in Shelikof Strait are heavier
and occur deeper than those in Funka Bay.
Horizontal distribution The horizontal distribution
pattern of eggs in Funka Bay was fairly consistent
among the 3 years (1977, 1978, and 1987) for which
data are presented (Nakatani 1988, Nakatani and
Maeda 1981 and 1989). Younger eggs are mainly found
just outside the entrance to the bay and older eggs are
found inside the bay, indicating that spawning occurs
outside the bay and the eggs drift into the bay as they
develop. During the period 24 January to 11 February
1978, egg abundance reached 13,424 Stage-4 eggs/m^
at a station just south of the entrance to the bay where
large numbers of Stage 2-5 eggs were also present,
producing a total of 23,817 eggs/m^.
The egg distribution pattern in Shelikof Strait was
most intensively examined in 1981; however, sampling
in other years (1978-86) indicates similar patterns. The
first appearance of low numbers of eggs occurs in
March and early April, mainly in the southern part of
the strait (Kendall and Picquelle 1990). The highest con-
centrations of eggs occur off Cape Kekurnoi in early
April, where abundances of Stage-2 and -3 eggs ex-
ceeded lOOO/m^ in 1981. The combined abundance for
Larvae*
Vertical distribution The ecology of walleye pollock
larvae has been investigated in both Funka Bay and
Shelikof Strait. In both areas most larvae occur above
50 m in the water column and exhibit limited diel ver-
tical migration (Kamba 1977, Kendall et al. 1987). Few
larvae are collected at the surface, but some larvae
move up to 1 0-20 m depth in the evening. At night they
are fairly evenly distributed throughout the upper 50 m,
and in the early morning they are again concentrated
above 20 m. During midday they are most abundant at
20 m and deeper to 50 m. The larvae sampled by Ken-
dall et al. (1987) in Shelikof Strait averaged 11.0mm
(SD 1.7mm), while those in Funka Bay sampled by
Kamba (1977) had a wide range of lengths from 4.6 to
26.4mm, although most were 4. 6-12. 8mm. Kamba
(1977) indicated that larger larvae (> 13.7 mm) were
more often collected in shallow tows at night and in
deep tows during the day, suggesting that either the
larger larvae migrated more than the smaller ones or
that the larger larvae were more successful at avoiding
the shallow nets during the day. No large larvae were
collected by Kendall et al. (1987). Kamba (1977) con-
cluded that the diel vertical movements of pollock
larvae in Funka Bay corresponded to those of their
zooplankton prey. Both Kamba (1977) and Kendall
et al. (1987) found a diel pattern in gut fullness, with
little food found in guts at night and most food found
in guts during the day.
* Lengths of larvae and juveniles are reported here as standard length
(SL, from the tip of the snout to the end of the notochord or base
of the hypural plate), although in the Japanese literature they were
given as total length (TL). Conversion from TL to SL is based on
our paired measurements of 1048 fish (4.2-103 mm SL) which
resulted in the relationship: SL(mm) = 0.108 + 0.907 TL(mm).
134
Fishery Bulletin 90(1). 1992
Horizontal distribution In the Funka Bay region,
walleye pollock larvae are generally concentrated
inside the bay from late January through early April
(Nakatani 1988, Nakatani and Maeda 1989). Their
abundance decreases during this time from >5000
larvae/m- in the area of maximum concentration in
late January to 200-400 larvae/m^ in early April. In
many cases, surveys have disclosed more than one area
of abundance within the bay. Their occurrence general-
ly overlaps that of the Oyashio Water. For example,
in 1980 the Tsugaru Warm Water remained in the bay
longer than usual, and the Oyashio Water did not enter
the bay until mid-March; before then, the larvae were
concentrated at the mouth of the bay. It is possible that
larvae entering the bay before the invasion of the
Oyashio Water would experience low survival because
of inadequate prey production.
In Shelikof Strait, most larvae are concentrated in
one large patch that can be followed as it drifts to the
southwest with the prevailing currents from April
through May (Kendall et al. 1987). The velocity of drift
may vary interannually and depend on weather pat-
terns in the area as well as the strength of the ACC.
In some years, it appears that most of the larvae drift
out of the strait within 2-4 weeks after hatching, but
in other years they remain for several more weeks
because of the influence of nearshore eddies (Incze
et al. 1989). There is considerable cross-strait shear in
the current, so the drift of larvae is influenced by where
they reach the surface layer from their deep incuba-
tion area (Kim and Kendall 1989). Larval abundances
as high as lO.OOO/m- were observed in the patch in
late April 1981, and by late May abundances of 2400/
m^ were present (Bates and Clark 1983).
Feeding Copepod nauplii, which were not identified
to species, are the major prey item of first-feeding
walleye pollock larvae (Kamba 1977, Kendall et al.
1987, Nakatani and Maeda 1983). Copepodids are the
most important prey item in the diet of 1 1 mm larvae
in Shelikof Strait and 8 mm larvae in Funka Bay.
Copepod eggs were more prevalent in guts of larvae
in Funka Bay than in Shelikof Strait (Nakatani and
Maeda 1983, Kendall et al. 1987). Their digestibility and
nutritional value for walleye pollock larvae are un-
known. Pseudocalanus spp. was the most abundant
copepod taxon in the water column in Shelikof Strait
and Funka Bay when larvae were present (Kendall
et al. 1987, Nakatani 1988). The nauplii in the guts of
small larvae were probably mostly Pseudocalanus spp.
and Oithona spp., and most of the copepodids in larger
larvae were Pseudocalanus spp. Copepodids oi Pseudo-
calanus minutus and Oithona similis were most
abundant in larger larvae up to 30 mm in Funka Bay
(Nakatani and Maeda 1983). The maximum prey size
increases with growth of the larvae, but the minimum
size remains fairly constant through fish up to about
73 mm (Kamba 1977).
Based on laboratory and field studies, naupliar abun-
dances of about 10 per liter seem to be required to sup-
port growth of small (<8mm) walleye pollock larvae
(Paul 1983, Dagg et al. 1984). Prey densities above this
threshold have been observed associated with the larval
patch in Shelikof Strait before and during a storm
(Incze et al. 1990). Naupliar abundances below this
threshold were seen in Funka Bay throughout most of
the larval period in 1987, but they were above 10 per
liter in several other years (Nakatani and Maeda 1989).
However, naupliar densities were probably underesti-
mated, since they were collected on 100/jm sieves.
AvaUabOity of smaller nauplii as larval food will require
further observations.
Age and growth Daily growrth increments on oto-
liths have been used to determine the age of larvae and
early juveniles from both Shelikof Strait and Funka
Bay. Based on a series of 109 larvae (6.0-14.6 mm SL)
collected in Shelikof Strait in May 1983, the linear
growth equation SL = 4.29mm -t- 0.21 d (r- 0.75),
where d = age in days, was fit (Kendall et al. 1987).
Growth based on 357 larvae and early juveniles 3.9-
30.0mm SL from the Shelikof spawning collected May
through July 1987 fit a Laird-Gompertz function: SL
at age t = 4.505 (e''-854(i-e-"'*'"))^ where t = days after
hatch (Yoklavich and Bailey 1989). The growth of
larvae and juveniles from Funka Bay fit the function:
TL = 121.5/(1 -He-o-026(t-i24.5ii))^ with TL in mm
(Nishimura and Yamada 1984). Thus larvae 50 days old
from Funka Bay were about 14.0mm SL (see footnote)
while those from Shelikof Strait would range from 14.8
mm SL (Kendall et al. 1987) to 18.7mm SL (Yoklavich
and Bailey 1989) (Fig. 4).
Larval population length-frequency distributions de-
pend on time of spawning, mortality of larvae, growth
of larvae, and sampling bias. Except for sampling bias,
these factors represent population processes occurring
to the annual cohort of larvae. In Funka Bay, even
though spawning takes place over a protracted period,
larval survival appears low except during periods when
adequate food is present. Mortality due to starvation
is high for larvae that hatch before the spring increase
of nauplii in Funka Bay (Nakatani and Maeda 1989,
Nakatani 1991). Thus variations in size of larvae may
depend more on differences in the birth dates of sur-
viving larvae than on differences in growth rates.
In Shelikof Strait, spawning peaked during the first
week of April in several years. By the end of April
1981, most larvae were about 4.8mm. By the third
week in May 1981, they were mostly 7-8mm (Dunn
et al. 1984), as they were in 1982 (Kendall et al. 1987).
Kendall and Nakatani: Early life history of Theragra chalcogramma in Shelikof Strait and Funka Bay
135
30
■
-•- Yoklavich & Bailey
H— Nisfiimura & Yamada
CO 25
E
E 20
c
£ 15
oi
-*- Kendall et al
+
„^*=-'
5
.-=-^=^'
C
10 20 30 40 50 60
Age in days after hatch
70 80
Figure 4
Growth of walleye pollock larvae and juveniles from Funka
Bay, Japan (Nishimura and Yamada 1984), and larvae from
Shelikof Strait, Gulf of Alaska (Kendall et al. 1987, Yoklavich
and Bailey 1989).
However, in 1983 larvae averaged 11.23mm in late
May (Kendall et al. 1987). No interannual differences
in larval growth rates were discerned for larvae col-
lected in Shelikof Strait in late May 1983, 1985, 1986,
or 1987. Because larvae were larger in late May 1983
than in 1985, 1986, or 1987, they may have been sur-
vivors of an earlier spawning than those observed in
the other years (Yoklavich and Bailey 1989).
Early juveniles
Young-of-the-year juvenile walleye pollock (18-73 mm)
have been sampled extensively in Funka Bay using mid-
water and bottom trawl nets (Nakatani and Maeda
1987). The juveniles are about 34 mm in late May,
36 mm in June, and 55-80 mm in late July. In June,
juveniles (22-66 mm) are found mainly at 25-30 m at
night and at 10-15 m during the day (Nakatani and
Maeda 1987). The juveniles move deeper in the water
column in May and June, and by late July most are on
the bottom (Nakatani and Maeda 1987). In July, larger
fish are caught in bottom trawls while smaller fish are
still in the water column. As the juveniles grow and
move toward deeper water and the bottom, they also
move from inside the bay toward the entrance (in June)
and to the shelf (100-300 m) just outside the bay (by
August) (Nakatani and Maeda 1987).
Some variation in size-at-date of juveniles among
years has been observed (Fukuchi 1976, Nakatani and
Maeda 1987), which may be due to interannual differ-
ences in growth rates, or differences in hatch dates of
surviving juveniles.
Food organisms changed during juvenile growth with
Neocalanus plumchrus being most important in fish
>27mm long in midwater. Juveniles collected on the
bottom fed on large-sized copepodids of Neocalanus
cristatus and Eucalanus bungii, Euphausia pacifica
(a euphausid), and Parathemisto japonica (an amphi-
pod) (Nakatani and Maeda 1987).
Young-of-the-year juveniles from the Shelikof Strait
spawning were sampled with a Methot midwater frame
trawl (Methot 1986) in June and July 1987 (Hinckley
et al. 1989), and by small-mesh midwater and bottom
trawl surveys in late-summer of several years (Bailey
and Spring, in review). Data from these studies have
not yet been completely analyzed. However, in June
and July the early juveniles (mainly 20-30 mm) were
found on the shelf along the Alaska Peninsula. As with
the eggs and larvae, they formed a large discrete patch
surrounded by a large area with lower abundances.
From their pattern of distribution, it appears that at
this size and time of year they still inhabit midwater
depths and are not schooling (Hinckley et al. 1989). Fish
were found to feed mainly on various life stages of
Pseudocalanus spp., smaller fish ate primarily nauplii
and copepodids, while adults became more important
in larger fish. Differences in diet between fish sampled
at different locations indicated that the food organisms
were patchily distributed (Grover 1990).
Sampling in late-summer has concentrated mainly on
the bays around Kodiak Island and along the Alaska
Peninsula. Considerable interannual variation in sam-
pling and pattern of distribution of juveniles character-
ized these surveys. In 1987, when the sampling area
in late-summer included the shelf west of the Shumagin
Islands, a concentrated patch of juveniles was found
that was likely the product of the Shelikof Strait spawn-
ing, i.e., the eggs and larvae that had been followed
through the spring during their drift to the southwest
from Shelikof Strait (Bailey and Spring, in review).
Year-class determinants
Studies of walleye pollock early life history in both
Funka Bay and Shelikof Strait have been designed to
determine causes of year-class fluctuations. The basic
premise is that these fluctuations result from events
during early life history and have little relation to the
abundance or other characteristics of the spawning
population. The influence of hydrography and its ef-
fect on larval food supply has been the most intensive-
ly studied factor in both areas, but predation has also
been considered in Shelikof Strait research.
In Funka Bay, walleye pollock early life history
seems to be closely tied to the timing and extent of
the influx of Oyashio Water (Nakatani 1984). This
cold, low-salinity water carries Psevdocalanics minutus
into the bay where they produce nauplii, the primary
diet of small larvae in nearsurface waters. Walleye
136
Fishery Bulletin 90(1). 1992
pollock spawning seems to be timed and positioned to
correspond to this influx. In years when this influx is
delayed or absent, survival of larvae may be reduced
(Nakatani and Maeda 1989). Years with an early inva-
sion of the Oyashio Water have resulted in large year-
classes of walleye pollock (Nakatani 1988). However,
a strong year-class was also observed in 1980 when
there was a late invasion (Nakatani and Maeda 1983,
Nakayama et al. 1987). To predict population size fluc-
tuations will require further studies on the causes of
larval mortality.
Besides factors influencing larval food production in
Shelikof Strait (Incze et al. 1990), the complex dynam-
ics of the ACC as it exits the strait seem important in
determining the rate of drift of the larval patch and
its resultant position when the larvae are ready to settle
(Reed et al. 1989). If the larvae are in the center of the
ACC as it exits the strait, they may be carried quickly
offshore through the sea valley between the Semidi
Islands and Chirikof Island, as apparently happened
in 1985 (Incze et al. 1989). Some of these larvae may
remain offshore where larval feeding conditions are
probably not ideal. The return of offshore larvae to the
shelf for demersal settlement is also problematical. If
the larvae are on the Alaska Peninsula side of the core
of the ACC as it exits the strait, their drift will be
slower, and they should remain in the coastal region
where food production is probably enhanced. Their tra-
jectory should carry them west along the Alaska Penin-
sula to shelf areas suitable for demersal settlement.
Storm winds blowing offshore from Wide Bay may
displace the ACC as it exits the strait, and eddies have
been observed in this area. The influence of such fac-
tors on the larval patch and larval food production may
be important in determining the numbers of larvae
reaching the juvenile stage.
Conclusions
It appears that within large areas of distribution, wall-
eye pollock populations have evolved to spawn in very
specific areas and during brief times of the year. Adults
migrate to these areas annually for spawning. This
spawning pattern produces concentrations of plank-
tonic eggs and larvae that far exceed those reported
for any other fish (> 20,000 eggs/m^; <5000 larvae/
m^). These spawnings are such that the eggs and
larvae find themselves in areas where suitable food is
abundant and where currents later carry larvae to
suitable nursery areas. It appears that interannual
variations in oceanographic conditions responsible for
food production and larval drift impact larval survival,
and hence year-class strength. Although there are
marked differences in the geography and oceanography
of Shelikof Strait and Funka Bay, walleye pollock have
adapted to reproduce successfully in both areas. Adap-
tations in the early life history of walleye pollock to
these differences in environment include timing and
duration of the spawning season, specific gravity of the
eggs, and differences in prey size in relation to larval
size.
Time of spawning in both areas corresponds to sea-
sonal transitions in hydrographic conditions (Nakatani
1988, Kim 1987). The spawning season is several
months long in the lower-latitude Funka Bay area
where there is considerable interannual variation in
timing of the intrusion of the cold Oyashio Water,
which increases copepod naupliar production. The
Shelikof Strait area spawning is very peaked, taking
place mainly over a few weeks and during the same
time each year, early April. This is the time when
currents are at an annual minimum due to reduced
precipitation and weak winds. We do not know if low
current strength is the seasonal signal that fish respond
to, but presumably the signal is less variable than the
intrusion of Oyashio Water.
Eggs are less dense in Funka Bay where water
depths are only about one-third those of Shelikof Strait.
In Funka Bay, the eggs rise in the water column after
spawning and drift into the inner part of the bay. In
Shelikof Strait, the eggs remain in the nearbottom
water where they are spawned and show no appreciable
drift. This difference in transport of eggs may relate
to the desired location of hatching. Copepod produc-
tion is enhanced when Oyashio Water enters Funka
Bay and the egg drift pattern enables the eggs to hatch
there. In Shelikof Strait, the upper layers of water dur-
ing the spawning season are moving to the southwest
at a rate that would flush eggs in surface waters out
of the strait and into the offshore Alaska Stream in
a few weeks. By remaining in the sluggish bottom
waters, hatching is more likely to occur in southwest
Shelikof Strait where larval prey may be more abun-
dant. Interannual variations in storms in this area may
effect copepod production and thereby larval condition.
In both areas, nauplii of species of small copepods,
Pseudocalanus and Oithona, are dominant in the diet
of first-feeding larvae. Eating small prey is energetical-
ly costly for larger larvae, so it may be critical for them
to encounter more advanced stages of copepods (Incze
et al. 1984). This may be more important in Shelikof
Strait than in Funka Bay because larvae in Funka Bay
start eating larger prey at a smaller size than do larvae
in Shelikof Strait.
Drift of larvae to nursery grounds is more important
in Shelikof Strait than it is in Funka Bay. It appears
that most juveniles that result from spawning in Sheli-
kof Strait inhabit shelf and nearshore areas 100-200 km
from the spawning location by the age of 4 months
Kendall and Nakatani: Early life history of Theragra chalcogramma in Shelikof Strait and Funka Bay
137
(Hinckley et al. 1989). Juveniles from the Funka Bay
spawning are mostly found in waters just outside the
bay during their first summer (Nakatani 1988). In the
following winter, some of them remain in the center
of the bay (T. Maeda and T. Nakatani, unpubl. data).
Acknowledgments
Many investigators in Japan and the United States are
studying the early life history of walleye pollock. Many
have generously shared ideas and data with us and we
thank them. In particular. Dr. Kevin Bailey, AFSC, and
Dr. Jim Schumacher, PMEL, reviewed early drafts of
this manuscript and gave numerous helpful sugges-
tions, as well as Dr. Tatsuaki Maeda who conducted
many of the Funka Bay studies that formed the basis
for this paper with Dr. Toshikuni Nakatani. Dr. Lew
Haldorson, University of Alaska, and Dr. Svein Sund-
by. Institute of Marine Research, Norway, provided
valuable reviews of an earlier draft.
Citations
Bailey, K., and S. Spring
In review Comparison of larval, age-0 juvenile, and age-2
recruit abundance indices of walleye pollock Theragra chalco-
gramma in the western Gulf of Alaska. ICES J. Mar. Sci.
Baird. T.A., and B.L. 011a
1991 Social and reproductive behavior of walleye pollock
Theragra chalcogramma of a captive group. Environ. Biol.
Fish. 30:295-301.
Bates. R.D., and J. Clark
1983 Ichthyoplankton off Kodiak Island and the Alaskan Penin-
sula during spring 1981. NWAFC Proc. Rep. 83-89, North-
west & Alaska Fish. Cent., NMFS, NOAA. Seattle, WA
98115-0070, 105 p.
Dagg, M.J., M.E. Clarke. T. Nishiyama, and S.L. Smith
1984 Production and standing stock of copepod nauplii, food
items for larvae of walleye pollock Theragra chalcogramma
in the southeastern Bering Sea. Mar. Ecol. Prog. Ser. 19:
7-16.
Dunn, J.R.. A.W. Kendall Jr., and R.D. Bates
1984 Distribution and abundance patterns of eggs and larvae
of walleye pollock (Theragra chalcogramma) in the western Gulf
of Alaska. NWAFC Proc. Rep. 84-10, Northwest & Alaska
Fish. Cent., NMFS, NOAA, Seattle, WA 98115-0070, 66 p.
Fukuchi, M.
1976 Some aspects of bioenergetics of walleye pollock (Thera-
gra chalcogramma Pallas) at early life stages. Ph.D. thesis,
Univ. Hokkaido, Hokkaido, Japan.
Grover, J.J.
1990 Feeding ecology of late-larval and early-juvenile walleye
pollock Theragra chalcogramma from the Gulf of Alaska in
1987. Fish. Bull., U.S. 88:463-470.
Haynes, E.B., and S.E. Ignell
1983 Effect of temperature on rate of embryonic development
of walleye pollock, Theragra chalcogramma. Fish. Bull., U.S.
81:890-894.
Hinckley, S.
1990 Variation in egg size of walleye pollock (Theragra chal-
cogramma) with a preliminary examination of the effect of egg
size on larval size. Fish. Bull., U.S. 88:471-483.
Hinckley, S., K. Bailey, J. Schumacher. S. Picquelle, and
P. Stabeno
1989 Preliminary results of a survey for late stage larval
walleye pollock and observations on larval drift in the western
Gulf of Alaska. 1987. In Proc, Int. symp. on biology and
management of walleye pollock, p. 297-306. Alaska Sea Grant
Rep. 89-1, Univ. Alaska, Fairbanks.
Incze, L.S., M.E. Clarke, J.J. Goering, T. Nishiyama. and
A.J. Paul
1984 Eggs and larvae of walleye pollock and relationships to
the planktonic environment. In Ito, D.H. (ed.), Proceedings
of the workshop on walleye pollock and its ecosystem in the
eastern Bering Sea, p. 109-159. NOAA Tech. Memo. NMFS-
F/NWC-62, Northwest & Alaska Fish. Cent.. NMFS, Seattle,
WA 98115-0070.
Incze, L.S.. A.W. Kendall Jr.. J.D. Schumacher, and R.K. Reed
1989 Interactions of a mesoscale patch of larval fish (Theragra
chalcogramma) with the Alaska Coastal Current. Continen-
tal Shelf Res. 9:269-284.
Incze. L.S.. P.B. Ortner, and J.D. Schumacher
1990 Microzooplankton, vertical mixing and advection in a
larval fish patch. J. Plankton Res. 12:365-379.
Kamba, M.
1977 Feeding habits and vertical distribution of walleye pollock,
Theragra chalcogramma (Pallas), in early life stage in Uchiura
Bay, Hokkaido. Res. Inst. N. Pac. Fish., Hokkaido Univ.,
Spec. Vol., p. 175-197.
Kendall, A.W. Jr., and S. Kim
1989 Buoyancy of walleye pollock (Theragra chalcogramma)
eggs in relation to water properties and movement in Shelikof
Strait, Gulf of Alaska. In Beamish, R. J., and G.A. McFarlane
(eds.), Effects of ocean variability on recruitment and evalua-
tion of parameters used in stock assessment models, p.
169-180. Can. Spec. Publ. Fish Aquat. Sci. 108.
Kendall, A.W. Jr., and S.J. Picquelle
1990 Egg and larval distributions of walleye pollock, Theragra
chalcogramma, in Shelikof Strait, Gulf of Alaska. Fish. Bull.,
U.S. 88:133-154.
Kendall, A.W. Jr., M.E. Clarke. M.M. Yoklavich. and
G.W. Boehlert
1987 Distribution, feeding, and growth of larval walleye pol-
lock, Theragra chalcogramma, from Shelikof Strait, Gulf of
Alaska. Fish. Bull., U.S. 85:499-521.
Kim, S.
1987 Spawning behavior and early life history of walleye
pollock, Theragra chalcogramma, in Shelikof Strait, Gulf of
Alaska, in relation to oceanographic features. Ph.D. diss.
Univ. Wash., Seattle, 221 p.
1989 Early life history of walleye pollock, Theragra chalco-
gramma, in the Gulf of Alaska. In Proc, Int. symp. on biology
and management of walleye pollock, p. 117-139. Alaska Sea
Grant Rep. 89-1, Univ. Alaska, Fairbanks.
Kim, S.. and A.W. Kendall Jr.
1989 Distribution and transport of larval walleye pollock
(Theragra chalcogramma) in Shelikof Strait, Gulf of Alaska,
in relation to water movement. Rapp. P.-V. Reun. Cons. Int.
Explor. Mer 191:127-136.
Maeda, T., T. Takahashi, M. Ijichi, H. Hirakawa, and M. Ueno
1976 Ecological studies on the Alaska pollock in the adjacent
waters of the Funka Bay, Hokkaido-II spawning season. Bull.
Jpn. Soc. Sci. Fish. 42:1213-1222 [in Jpn., Engl, abstr.].
138
Fishery Bulletin 90(1). 1992
Maeda, T., T. Nakatani, T. Takahashi. and M. Ueno
1980 On the population and developmental fluctuations of the
Alaska pollock at the southern area off the Hokkaido in the
Pacific Ocean. Bull. Jpn. Soc. Fish. Oceanogr. 37:29-33 [in
Jpn.].
Maeda, T., T. Takahashi, and M. Ueno
1981 Annual life period of the adult Alaska pollock in the ad-
jacent waters of the Funka Bay, Hokkaido. Bull. Jpn. Soc.
Sci. Fish. 47:741-746 [in Jpn., Engl, abstr.].
Megrey, B.A.
1989 Exploitation of walleye pollock resources in the Gulf of
Alaska, 1964-88: Portrait of a fishery in transition. /nProc,
Int. symp. on biology and management of walleye pollock, p.
33-58. Alaska Sea Grant Rep. 89-1, Univ. Alaska, Fairbanks.
Methot, R.D.
1986 Frame trawl for sampling pelagic juvenile fish. Calif.
Coop. Oceanic Fish. Invest. Rep. 27:267-278.
Miller, B.S., D.R. Gunderson, D. Glass, D.B. Powell, and
B.A. Megrey
1986 Fecundity of walleye pollock (Theragra chalcogramma)
from the Shelikof Strait, Gulf of Alaska. FRI-UW-8608, Coll.
Fish., Univ. Wash., Seattle. 40 p.
Muigwa, N.M.
1989 Vertical distribution patterns of prespawning and spawn-
ing pollock {Theragra chalcogramma) in Shelikof Strait. In
Proc, Int. symp. on biology and management of walleye
pollock, p. 403-432. Alaska Sea Grant Rep. 89-1, Univ.
Alaska, Fairbanks.
Nakatani, T.
1984 The early life history of walleye pollock Theragra chalco-
gramma (Pallas) in Funka Bay and the adjacent waters, Hok-
kaido. Aquabiology 34:341-347 [in Jpn., Engl, abstr.].
1988 Studies on the early life history of walleye pollock in
Funka Bay and vicinity, Hokkaido. Mem. Fac. Fish.. Hok-
kaido Univ. 35:1-46.
1991 Copepod nauplii as food of walleye pollock larvae (Pisces:
Gadidae) in Funka Bay, Japan. Proc, Fourth int. conf. cope-
poda. Bull. Plankton Soc. Jpn., Spec. Vol., p. 515-525.
Nakatani, T., and T. Maeda
1981 Transport process of the Alaska pollack eggs in Funka
Bay and the adjacent waters, Hokkaido. Bull. Jpn. Soc. Sci.
Fish. 47:1115-1118.
1983 Distribution of walleye pollock larvae and their food sup-
ply in Funka Bay and the adjacent waters, Hokkaido. Bull.
Jpn. Soc. Sci. Fish. 49:183-187 [in Jpn., Engl, abstr.].
1984 Thermal effect on the development of walleye pollock
eggs and their upward speed to the surface. Bull. Jpn. Soc.
Sci. Fish. 50:937-942 [in Jpn., Engl, abstr.].
1987 Distribution and movement of walleye pollock larvae
Theragra chalcogramma in Funka Bay and the adjacent waters,
Hokkaido. Bull. Jpn. Soc. Sci. Fish. 53:1585-1591 [in Jpn.,
Engl, abstr.; Engl, transl. avail. Alaska Fish. Sci. Cent., NMFS,
NOAA, Seattle, WA 98115-0070].
1989 Distribution of copepod nauplii during the early life stages
of walleye pollock in Funka Bay and vicinity, Hokkaido. In
Proc, Int. symp. on biology and management of walleye
pollock, p. 217-240. Alaska Sea Grant Rep. 89-1, Univ.
Alaska, Fairbanks.
Nakayama, N., Y. Watanabe, H. Ida, and K. Nagasawa
1987 Observation report of walleye pollock fishery in the south
region of Hokkaido, the Pacific Ocean, 1987, no. 1, p. 1-13 [in
Jpn.].
Nishimura, A., and J. Yamada
1984 Age and growth of larval and juvenile walleye pollock,
Theragra chaUMgramma (Pallas), as determined by otolith daily
growth increments. J. Exp. Mar. Biol. Ecol. 82:191-205.
Ohtani, K., and K. Kido
1980 Oceanographic structure in Funka Bay. Bull. Fac. Fish.,
Hokkaido Univ. 31:84-114 [in Jpn., Engl, abstr.].
Paul, A.J.
1983 Light, temperature, nauplii concentrations, and prey cap-
ture by first feeding pollock larvae, Theragra chalcogramma.
Mar. Ecol. Prog. Ser. 13:175-179.
Reed, R.K., L.S. Incze, and J.D. Schumacher
1989 Estimation of the effects of the flow on dispersion of
larval pollock, Theragra chalcogramma, in Shelikof Strait,
Alaska. In Beamish, R.J., and G.A. McFarlane (eds.), Effects
of ocean variability on recruitment and evaluation of param-
eters used in stock assessment models, p. 239-246. Can. Publ.
Fish. Aquat. Sci. 108.
Sakurai, T.
1982 Reproductive ecology of walleye pollock Theragra chalco-
gramma (Pallas). Ph.D. diss., Hokkaido Univ., 178 p. [Engl,
transl. avail. Alaska Fish. Sci. Cent., NMFS, NOAA, Seattle,
WA 98115-0070.
1989 Reproductive characteristics of walleye pollock with
special reference to the ovarian development, fecundity and
social behavior. In Proc, Int. symp. on biology and manage-
ment of walleye pollock, p. 97-115. Alaska Sea Grant Rep.
89-1, Univ. Alaska, Fairbanks.
Yoklavich, M.M., and K. Bailey
1989 Growth of larval and juvenile walleye pollock from Sheli-
kof Strait, Gulf of Alaska, as determined from daily increments
in otoliths. In Proc, Int. symp. on biology and management
of walleye pollock, p. 241-251. Alaska Sea Grant Rep. 89-1,
Univ. Alaska, Fairbanks.
Yoon, T.
1981 Reproductive cycle of female walleye pollock Theragra
chalcogramma (Pallas), in the adjacent waters of Funka Bay,
Hokkaido. Bull. Fac. Fish. Hokkaido Univ. 32:22-38 [in Jpn.,
Engl, abstr.].
Abstract.- Catch and effort data
for the abalone Haliotis rub7-a fish-
ery off Victoria, Austraha, revealed
that catches were alHed to incentive
(price); annual catch was proportion-
al to effort. The robustness of the
fishery can be attributed to low fish-
ing mortality (F around 0.1) and a
relatively high minimum length of
capture (120 mm). Exploitation
models showed that egg production
was at least 50% that of unfished
stocks. The analyses also showed
that egg production was sensitive to
variation in the growth parameters;
fast-growing populations were more
vulnerable to recruitment overfish-
ing than slow-growing populations.
For slow-growing populations, yields
could be considerably increased
without endangering recruitment. It
is suggested, from the available
evidence, that overfishing has been
overemphasized in the collapse of
abalone fisheries.
Exploitation modeis and catch
statistics of thie Victorian fisliery
for abalone Haliotis rubra
Paul E. McShane
Marine Science Laboratories, Fisheries Division, Ministry for Conservation and Environment
P.O. Box 1 14, Queenscliff, Victoria 3225, Australia
Present address: Fisheries Research Division, Ministry of Agriculture and Fisheries
P.O, Box 297, Wellington, New Zealand
Manuscript accepted 18 December 1991.
Fishery Bulletin, U.S. 90:139-146 (1992).
Recent and comprehensive descrip-
tions of the state of the world's aba-
lone {Haliotis spp.) fisheries include
reviews of the abalone fisheries in
California (Tegner 1989, Tegner et
al. 1989 and 1992), British Columbia
(Breen 1986, Sloan and Breen 1988),
Australia (Prince and Shepherd 1992),
Mexico (Guzman del Proo 1992) and
Japan (Mottet 1978). A unifying
theme of these reviews is that aba-
lone fisheries are characterised by
initial high productivity followed by
irreversible decline. Australia has
developed an abalone fishery only
recently by world standards. In Vic-
toria, the fishery for the abalone
Haliotis rubra is productive, valu-
able, and apparently stable (McShane
1990). The government limits the
number of operators in the fishery
(71), the annual catch (1460 metric
tons) since 1988, and minimum
length of capture (120 mm).
A fundamental objective of fish-
eries science is to predict the produc-
tion from a fishery under varying
management strategies. A common
approach is to consider the yield from
an individual or year-class of in-
dividuals under different fishing con-
ditions (Beverton and Holt 1957,
Ricker 1975, Gulland 1988, Megrey
and Wespestad 1988). Such exploita-
tion models treat populations as the
sum total of their individual mem-
bers; yield is expressed as yield-per-
recruit because the absolute level of
recruitment is rarely known. Yield-
per-recruit models have been applied
to several abalone fisheries including
those for H. discus discus (Ishibashi
and Kojima 1979), H. iris (Sainsbury
1982a), H. laevigata (Sluczanowski
1984), H. kamtschatkana (Breen
1986, Sloan and Breen 1988), H. rufe-
scens and H. corrugata (Tegner et al.
1989), and the Tasmanian fishery for
H. rubra (Nash 1992).
Although yield-per-recruit models
can provide information on appropri-
ate harvest strategies to maximize
yield, the results provide no indica-
tion of the sustainability of a par-
ticular harvest regime. Because of
the historical tendency of abalone
fisheries to collapse, increasing atten-
tion has been focused on manage-
ment strategies which maintain egg
production as well as yield (Sluc-
zanowski 1984 and 1986, Breen 1986,
Sloan and Breen 1988, Tegner et al.
1989, Nash 1992).
In the present paper, the produc-
tivity of the fishery for abalone Ha-
liotis rubra off Victoria, Australia, is
described. To investigate the effect
of growth rate, the relative yields of
weight and eggs for two hypothetical
populations of H. rubra, fast- and
slow-growing, are examined. Man-
agement implications of my results
are discussed for H. rubra as well as
for other abalone species generally.
Materials and methods
Fishery statistics
Data on annual catch, effort, and
price (whole weight) for the Victorian
abalone fishery were obtained from
139
140
Fishery Bulletin 90(1), 1992
fishermen's returns and unpublished information
supplied by the Victorian Fisheries Division. In-
formation on the history of the Victorian abalone
fishery was extracted from unpublished records
supplied by the Victorian Fisheries Division (Dep.
Conserv. Environ., 240 Victoria Pde, Melbourne
3002; see also McShane 1990).
Yield-per-recruit
Generalised fisheries exploitation models such as
yield-per-recruit rely heavily on several assump-
tions. For any "unit stock":
1 Growth rates do not vary with time or density of
the exploitable stock. Thus growth can be modeled with
one set of parameters, e.g., the von Bertalanffy growth
equation (Ricker 1975). Departures from these assump-
tions are known for abalone (e.g., Newman 1968, Sloan
and Breen 1988, Day and Fleming 1992). However, for
stocks of H. rubra the assumptions are reasonable
(McShane et al. 1988a).
2 The rate of natural mortality is known and does not
vary with age, time or density of the stock. Natural
mortality is an important parameter in yield-per-recruit
models, yet it is often the most difficult to estimate ac-
curately. Natural mortality of//, rubra is constant with
age after the first year (Shepherd et al. 1982, McShane
1991, Shepherd and Breen 1992). Estimates of natural
mortality are in Table 1.
3 Fishing (F) and natural (M) mortality are indepen-
dent of each other. For abalone fisheries, fishing mor-
tality cannot be considered applicable to the entire
fishery. Individual exploitation rates are applied to
substocks opportunistically according to weather and
incentive (Sluczanowski 1984, McShane and Smith
1989a). Incidental mortality can be caused by fishing,
for example, wounding of undersize individuals (Sloan
and Breen 1988, Tegner 1989, Shepherd and Breen
1992).
4 Recruitment is constant. Recruitment measured as
the density of post-settlement individuals is highly
variable for//, rubra (McShane et al. 1988b, McShane
and Smith 1991). However, variation in growth rates
of prerecruit individuals within a population acts to
smooth out year-to-year variation in those //. rubra
reaching harvestable size (McShane 1991).
5 Individuals of the same age have the same weight
and susceptibility to capture. Individual variation in the
relationships of weight to length and length to age has
been demonstrated for H. rubra, but reasonable
Table 1
Estimates of rates of natural mortality (M) for Haliotis rubra.
Reference
Location
M(yr-')
Beinssen and Powell (1979)
Nash (1992)
Shepherd et al. (1982)
Prince et al. (1988)
northeast Victoria 0.20
northern Tasmania 0.24-0.29
South Australia 0.21-0.36
southeast Tasmania 0.1-0.7
generalizations of these relationships can be made for
the stock (McShane et al. 1988a, McShane and Smith
1992).
To investigate the effects of various rates of fishing,
the yield-per-recruit equation of Ricker (1975:237) was
used. The increase in length with age of//, rubra was
computed using the von Bertalanffy growth equation
Lt = L^(l
-Kt-
to)
where Lt is the shell length in mm of H. rubra at age
t years, L^ is the hypothetical maximum length, K is
the Brody growth constant, and to is the hypothetical
age when length is zero.
In calculating the yield-per-recruit of H. rubra at
various ages, I assumed that individuals were recruited
in the year corresponding to the minimum length at
capture. The biomass of an individual of age t years,
Wt (g), was assumed to be 0.00016 Lt"^, where Lt is in
mm (McShane et al. 1988a).
Egg-per-recruit
A simple age-structured model was used in which the
relative abundance of females of age t years (A'^t ) was
computed as
where Z is total mortality (F -i- M). The egg production
of a female of age t years (Ej) has a linear relationship
with length (Lt) for H. rubra (McShane et al. 1988b)
such that
Et = 0.03 Lt - 2.4
where E is fecundity in millions of eggs, and L is shell
length in mm; Lt is derived from the von Bertalanffy
growth equation.
Total egg production (Etot) is given by
t = 25
Etot = 1 iVt • Et
t=o
where t = 25 years is assumed to be the maximum age
McShane: Exploitation models and catch statistics for Haliotis rubra off Victoria, Australia
141
Table 2
Parameters used in computations of yield and egg-per-recruit.
Values for slow- and fast-growing populations of Haliotis
rubra are derived from mark-recapture studies in Victoria,
Australia (McShane 1990). Estimates provided are the von
Bertalanffy growth parameters (see text for details).
(McShane et al. 1988a). Egg production under various
combinations of F, M, and minimum length-at-capture
was compared with egg production of an unfished
population (F = 0).
Fast- and slow-growing populations ofH. rubra were
modeled. The generalised growth parameters (Table
2) were based on empirical estimates (McShane et al.
1988a, McShane 1990, McShane and Smith 1992). Both
yield and egg-per-recruit were expressed graphically
as a function of minimum length-at-capture (i.e., length-
at-recruitment) and F, using two rates of natural mor-
tality estimated for H. rubra (M = 0.1 and 0.2, Table
1). Length-at-recruitment was varied (in 10 mm incre-
ments) from 100 mm to 140 mm for fast-growing popu-
lations and from 70 mm to 130 mm for slow-growing
populations. The value of F was varied from 0.1 to 1.5.
A smooth surface was interpolated through points in
3-dimensional plots of yield and egg-per-recruit, follow-
ing the method of McLain (1974) in which negative
exponential weights are computed from distances be-
tween points in a regular grid and the irregularly
spaced data points in the X-Y plane (Wilkinson 1990).
Results
Fishery statistics
Annual variations in catch, effort, and value of the Vic-
torian abalone fishery are described in Figure 1. Catch
is highly correlated with effort (r 0.98, n 25, P< 0.001).
Although the catch rose in 1965-66 (accompanying
development of export markets), the trend in both
catch and effort is one of a slight but unalarming
decrease followed by a slight increase during the 1980s.
The introduction of catch quotas in 1988 is reflected
in the decrease in catch in that year (Fig. 1). It is
noteworthy that prior to 1988, price of abalone is a
significant factor influencing the catch of the Victorian
abalone fishery. Allowing for inflationary increases, the
price of abalone doubled between 1967 and 1976-77
accompanying development of Japanese markets. Four
i
965 1970 1975 1980 1985 1990
3.
1965 1970 1975 1980 1985 1990
YEAR
Figure 1
Comparison of (upper) annual catch and effort (dashed line),
and (lower) actual price and CPI adjusted price (dashed line)
of abalone Haliotis rubra in the fishery off Victoria, Australia,
1965-88.
exceptions to the steady rise in price have occurred.
In 1967-68, a slight fall in price resulted from ship-
ments of poor-quality abalone. Processing techniques
were, at the time, in a developmental phase. Second,
in 1976 an increase in price occurred concomitant with
high demand by export markets and increased competi-
tion between processors for supply. The introduction
of a competitive product, the Chilean "loco" Concho-
lepas concholepas, on Asian markets coupled with buyer
resistance to elevated prices of Australian abalone
resulted in a decrease in price during 1977 (Stanistreet
1978). Note that there is a lag between price variation
and catch and effort; the relative decrease in catch and
effort in 1978 reflects the price drop in 1977 (Fig. 1).
Buyer resistance also affected the price of abalone in
1981-82 and led to a decrease in effort and catch dur-
ing this period.
More recently, the collapse of the large Mexican
abalone fishery and the imposition of catch quotas on
the Tasmanian and South Australian abalone fisheries
(Prince and Shepherd 1992) decreased the world supply
of abalone and increased the competitiveness of Vic-
torian suppliers (McShane 1990). This and a decrease
in the relative value of the Australian currency against
142
Fishery Bulletin 90(1). 1992
that of export markets resulted
in a rapid increase in price of
abalone during the 1980s.
Exploitation models
The yield and egg-production-
per-recruit for individuals of vari-
ous lengths under various levels
of exploitation are shown for
fast-growing (Fig. 2) and slow-
growing (Fig. 3) populations of
H. rubra. It can be seen that the
relative yield-per-recruit is great-
er for fast-growing compared
with slow-growing populations of
H. rubra. The minimum lengths
producing maximum yield-per-
recruit are 130 mm for fast-grow-
ing populations and 120 mm for
slow-growing populations. These
maxima occurred at high exploi-
tation rates (F~l) and were in-
dependent of the natural mortal-
ity rates applied to the model
(0.1, 0.2). Natural mortality had
an obvious effect on the decline
of yield-per-recruit with mini-
mum length. For M = 0.2, yield-
per-recruit was less sensitive to
variation in minimum length
compared with M = 0.1. Note that
at realistic levels of F (0.1,
McShane and Smith 1989a),
yield-per-recruit is comparative-
ly low (Figs. 2 and 3). For such
low rates of exploitation, the
model shows that for fast-grow-
ing (in contrast to slow-growing)
populations, yield-per-recruit is relatively insensitive
to variation in the minimum length-at-capture. Similar
results were obtained from yield-per-recruit analyses
of other species of abalone, provided that rates of
fishing mortality are relatively low (F<0.3) (Ishibashi
and Kojima 1979, Sainsbury 1982a, Sluczanowski 1984,
Breen 1986, Sloan and Breen 1988, Tegner et al. 1989,
Nash 1992).
For fast-growing populations, exploitation rates pro-
ducing maximum yield-per-recruit are associated with
minimum egg production. Indeed, values of F>0. 3 are
associated with egg production of less than 50% of an
unfished population. Low egg production may cause
recruitment failure in abalone stocks (Sloan and Breen
1988, Tegner et al. 1989). Egg production increases
with minimum length; results of other studies show
Figure 2
Variation in egg and yield-per-recruit with shell length and fishing mortality (F) for a
fast-growing population oiHaliotis rubra subject to natural mortality (M) of 0.1 and 0.2.
that fecundity ofH. rubra is directly related to length
(McShane et al. 1988b, Prince et al. 1988).
Egg production is less sensitive to variation in fishing
mortality in slow-growing than in fast-growing popula-
tions oiH. rubra. For minimum lengths over 120mm,
egg production rates are over 50% of an unfished
population. At F = 0.1, egg production is over 60% of
that of an unfished population for both M = 0. 1 and 0.2.
Discussion
Catch levels for the Victorian abalone fishery suggest
a robust fishery. But catch data are poor indicators of
the stock abundance of abalone because fishermen can
maintain catch rates by exploiting substocks (Breen
McShane: Exploitation models and catch statistics for Hahotis rubra off Victoria. Australia
143
-\«^o
ic>s"»
Figure 3
Variation in egg and yield-per-recruit with shell length and fishing mortality (F) for a
slow-growing population ofHaliotis 'rubra subject to natural mortality (M) of 0.1 and 0.2.
1980, Sloan and Breen 1988, McShane and Smith
1989a). Yet the available evidence is that the fishery
is underexploited (McShane and Smith 1989ab;
McShane 1990). Catches ofH. rubra can be adjusted
opportunistically by increasing effort when incentive
(price) is high. Although Victorian abalone fishermen
have the capacity to serially deplete substocks, rates
of exploitation of H. rubra are generally low (see
McShane and Smith 1989a). A surplus of harvestable
individuals is maintained in substocks by the conser-
vative fishing practices employed by Victorian abalone
fishermen (McShane and Smith 1989a). With low ex-
ploitation rates (F<0.3), the egg-per-recruit model
shows that there is adequate egg production by in-
dividuals above the present legal minimum length of
120 mm. However, as a consequence of reaching har-
vestable size in about 4 years, prerecruit individuals
from fast-growing populations have fewer years of egg
production than those H. rubra
from slow-growing populations
which reach harvestable size in
about 10 years. Fast-growing
populations of H. rubra are
therefore vulnerable to recruit-
ment overfishing should exploita-
tion rates increase (F>0.3). This
is unlikely in the Victorian aba-
lone fishery because both the
number of operators and the an-
nual catch are controlled.
Most abalone fisheries are gen-
erally subject to pulse fishing.
Substocks are fished, then left to
recover (Sluczanowski 1984,
McShane and Smith 1989a).
Fast-growing populations are
important in this regard because
they can be fished at a higher fre-
quency than slow-growing popu-
lations (Sluczanowski 1984).
Thus fast-growang populations
are subject to higher exploitation
rates than slow-growing popula-
tions. Slow-growing populations
are often characterised by large
accumulations of prerecruit aba-
lone which are food-limited
(Sloan and Breen 1988, McShane
1990). Egg-per-recruit analysis
shows that even at extraordinar-
ily high rates of fishing mortal-
ity (F> 1) egg production in slow-
growing populations of H. rubra
is above the assumed "safe" level
of 50% of an imfished population
(Sloan and Breen 1988, Tegner et al. 1989). The model
shows that slow-growing populations could be "fished
down" at a reduced size limit and high exploitation rate
without endangering egg production. Such a harvesting
strategy could reduce the abundance of the accum-
ulated stock to a level where food is no longer a limiting
factor (McShane and Smith 1989b). To date in Victoria,
such slow-growing stocks are rarely fished because of
a paucity of abalone of harvestable size (McShane and
Smith 1989b).
Why is the Victorian fishery for H. rubra apparent-
ly robust in contrast to other abalone fisheries? The
viability of the Victorian fishery can be attributed to
a relatively low number of operators (limited entry has
operated in Victoria since 1968) and an associated low
exploitation rate (see McShane and Smith 1989a). A
minimum length that maintains a safe level of egg pro-
duction provides further safeguards against recruit-
144
Fishery Bulletin 90|l|. 1992
ment overfishing, as does an annual catch quota.
However, size limits introduced to the California aba-
lone fishery were also conservative but failed to arrest
the decline of the fishery (Tegner 1989, Tegner et al.
1992). The combination of high commercial effort and
intense recreational and illegal harvest resulted in a
removal of surplus stocks in the California (Tegner et
al. 1992) and Mexican (Guzman del Proo 1992) abalone
fisheries. Unrestrained recreational and illegal harvest
remains a threat to the Victorian fishery, but with a
comparatively low human population and a relatively
inaccessible coastline the Victorian abalone fishery is
less vulnerable to noncommercial overfishing than
the California or Mexican abalone fisheries (McShane
1990).
While fishing can deplete stocks, there are a multi-
tude of other factors that affect the abundance of aba-
lone. For example, overfishing could not explain the
recruitment failure which occurred in the abalone
fishery of British Columbia (Breen 1986, Sloan and
Breen 1988). Recruitment failure in various species of
abalone has been attributed to sea temperature anom-
alies (Hayashi 1980, Forster et al. 1982, Shepherd et al.
1985) or natural variation (Sainsbury 1982b; see also
McShane and Smith 1991). The collapse of the Califor-
nia abalone fishery for H. rufescens coincided with
predator release (Lowry and Pearse 1973, Hines and
Pearse 1982, Tegner 1989, Tegner et al. 1989, 1992).
The importance of predation in controling abalone
abundance is further exemplified by the recovery of
stocks of H. cracherodii concomitant with a reduction
in the abundance of major predators (Davis et al. 1992).
The decrease in abundance of some California popula-
tions of abalone (H. rufescens, H. cracherodii) was
attributed to low food availability caused by El Nino
events (Tegner and Dayton 1987, Tegner et al. 1989),
competition with other herbivores, and kelp harvest
(Davis et al. 1992). Starvation in abalone causes a
decrease in reproductive effort (Cox 1962) and an in-
creased susceptibility to disease, both of which can
cause a severe decline in stocks (Haaker et al. 1992,
Tissot 1992). A major factor in the reduced abundance
of abalone stocks in Japan is nearshore pollution
(I. Hayashi, Igarashi-Jutaku 2-205, Niigata, Japan,
pers. commun. 1990), a factor also implicated in the
decline of California abalone stocks (see Tegner et al.
1992).
Variation in abiotic factors such as temperature have
demonstrable effects on the survival and growth of ex-
ploited species (Cushing 1988). Such factors, apart from
seasonal variation, vary stochastically and introduce
uncertainty in fisheries management (Megrey and
Wespestad 1988, Walters and Collie 1988). Faced with
this uncertainty, fishery managers must proceed
cautiously and gain a better understanding of the
ecology of exploitable species, particularly of abalone
which have a history of unexplained stock collapse.
Acknowledgments
I thank Dr. Paul Breen for helpful discussion and con-
structive comments on the manuscript. Drs. Linda
Jones and Scoresby Shepherd offered helpful sugges-
tions as did an anonymous reviewer.
Citations
Beinssen, K., and D. Powell
1979 Measurement of natural mortality in a population of
blacklip abalone, Notohaliotis rvbei-. Rapp. P.-V. Reun. Cons.
Int. Explor. Mer 175:23-26.
Beverton, R.J.H., and S.J. Holt
1957 On the dynamics of exploited fish populations. Fish.
Invest. Ser. II Mar. Fish. G.B. Minist. Agric. Fish. Food 19,
533 p.
Breen, P.A.
1980 Measuring fishing intensity and annual production in the
abalone fishery of British Columbia. Fish. Res. Board Can.
Tech. Rep. 947, 49 p.
1986 Management of the British Columbia fishery for northern
abalone (Haliotis kamschatkajia). In Jameison, G.S., and
N. Bourne (eds.), North Pacific workshop on stock assessment
and management of invertebrates, p. 300-312. Can. Spec.
Publ. Fish. Aquat. Sci. 92.
Cox, K.W.
1962 California abalones, family Haliotidae. Calif. Fish Game,
Fish Bull. 118, 138 p.
Cushing. D.H.
1988 The study of stock and recruitment. In Gulland, J. A.
(ed.), Fish population dynamics, 2d ed.. p. 105-128. Wiley,
NY.
Davis, G.E., D.V. Richards, P.L. Haaker, and D.O. Parker
1992 Mass mortality and fishery management of black aba-
lone, Haliotis cracherodii, in Southern California. In Shep-
herd, S.A., M.J. Tegner, and S.A. Guzman del Proo (eds.),
Abalone of the world: Biology, fisheries and culture. Black-
wells, Oxford (in press).
Day, R.W., and A. Fleming
1992 The determinants and measurement of abalone growth.
In Shepherd, S.A., M.J. Tegner, and S.A. Guzman del Proo
(eds.), Abalone of the world: Biologj', fisheries and culture.
Blackwells, Oxford (in press).
Forster, G.R.. G.W. Potts, and R. Swinfen
1982 Changes in the ormer populations of Guernsey and
Jersey. J. Mar. Biol. Assoc. U.K. 62:717-727.
Gulland, J.A.
1988 Fish population dynamics, 2d ed. Wiley-Interscience,
NY.
Guzman del Proo, S.A.
1992 Una revision sobre la biologia y pesqueria del abulon
(Haliotis spp.) en Mexico. In Shepherd, S.A., M.J. Tegner,
and S.A. Guzman del Proo (eds.), Abalone of the world: Biol-
ogy, fisheries and culture. Blackwells, Oxford (in press).
McShane: Exploitation models and catch statistics for Haliotis rubra off Victoria. Australia
145
Haaker, P.L., D.V. Richards, C. Friedman, G. Davis, D.O. Parker,
and H. Togstad
1992 Abalone withering syndrome and mass mortality of black
abalone, Haliotis cracherodii, in California. In Shepherd,
S.A., M.J. Tegner, and S.A. Guzman del Proo (eds.), Abalone
of the world: Biology, fisheries and culture. Blackwells, Ox-
ford (in press).
Hayashi, I.
1980 Structure and growth of a shore population of the ormer,
Haliotis tuberculata. J. Mar. Biol. Assoc. U.K. 60:431-437.
Hines, A.H., and J.S. Pearse
1982 Abalones, shells and sea otters: Dynamics of prey popula-
tions in central California. Ecology 63:1547-1560.
Ishibashi, K., and H. Kojima
1979 A study on the stock of Japanese black abalone, Haliotis
disc-US disms in Tokushima Prefecture— III. Analysis on
abalone fisheries management. Bull. Tokai Reg. Fish. Res.
Lab. 98:65-75.
Lowry, L.F., and J.S. Pearse
1973 Abalone and sea urchins in an area inhabited by sea
otters. Mar. Biol. (Beri.) 23:213-219.
McLain, D.H.
1974 Drawing contours from arbitrary data points. Computer
J. 17:318-324.
McShane, P.E.
1990 The fisheries ecology of Victorian abalone. Ph.D. thesis,
La Trobe Univ, Melbourne, 364 p.
1991 Density dependent mortality of the abalone Haliotis
rubra (Mollusca:Gastropoda). Mar. Biol. (Beri.) 110:385-389.
McShane, P.E., and M.G. Smith
1989a Direct measurement of fishing mortality in abalone
{Haliotis rubra) off southeastern Australia. Fish. Res. 8:
93-102.
1989b Victorian abalone fishery in good shape. Aust. Fish.
48:20-25.
1991 Recruitment variation in sympatric populations of the
abalone Haliotis rubra (Mollusca:Gastropoda). Mar. Ecol.
Prog. Ser. 73:203-210.
1992 Shell growth checks are unreliable indicators of age of
the abalone Haliotis rubra (Mollusca:Gastropoda). Aust. J.
Mar. Freshwater Res. (in press).
McShane, P.E., M.G. Smith, and K.H.H. Beinssen
1988a Growth and morphometry in abalone (Haliotis rubra)
from Victoria. Aust. J. Mar. Freshwater Res. 39:161-6.
McShane, P.E., K.P. Black, and M.G. Smith
1988b Recruitment processes in Haliotis rubra (Mollusca:
Gastropoda) and regional hydrodynamics in southeast Australia
imply localised dispersal of larvae. J. Exp. Mar. Biol. Ecol.
124:175-203.
Megrey, B.A., and V.G. Wespestad
1988 A review of biological assumptions underlying fishery
assessment models. In Wooster, W.S. (ed.), Fishery science
and management— Objectives and limitations, p. 31-74. Lect.
Notes Coastal Estuarine Stud. 28.
Mottett, M.G.
1978 A review of the fishery biology of abalones. Wash. Dep.
Fish. Tech. Rep. 37, 78 p.
Nash, W.J.
1992 Determination of appropriate size limits by egg-per-
recruit analysis, with an assessment of size limits as a manage-
ment tool for blacklip abalone {Haliotis rubra) in Tasmania. In
Shepherd, S.A., M.J. Tegner, and S.A. Guzman del Proo (eds.),
Abalone of the world: Biology, fisheries and culture. Black-
wells, Oxford (in press).
Newman, G.G.
1968 Growth of the South African abalone Haliotis midae. S.
Afr. Div. Sea Fish. Invest. Rep. 67:1-24.
Prince, J.D., and S.A. Shepherd
1992 Australian fisheries for abalone and their management.
In Shepherd, S.A., M.J. Tegner, and S.A. Guzman del Proo
(eds.), Abalone of the world: Biology, fisheries and culture.
Blackwells, Oxford (in press).
Prince, J.D., T.L. Sellers, W.B. Ford, and S.R. Talbot
1988 Recruitment, growth, mortality and population structure
in a southern Australian population oi Haliotis rubra (Mollusca:
Gastropoda). Mar. Biol. (Beri.) 100:75-82.
Ricker, W.E.
1975 Computation and interpretation of biological statistics
of fish populations. Bull. Fish. Res. Board Can. 191, 382 p.
Sainsbury, K.J.
1982a Population dynamics and fishery management of the
paua, Haliotis iris. 2. Dynamics and management as examined
using a size class population model. N.Z. J. Mar. Freshwater
Res. 16:163-173.
1982b Population dynamics and fishery management of the
paua, Haliotis iiis. 1. Population structure, growth, reproduc-
tion and mortality. N.Z. J. Mar. Freshwater Res. 16:147-161.
Shepherd, S.A., and P. A. Breen
1992 Mortality in abalone: Its estimation, variability and
causes. In Shepherd, S.A., M.J. Tegner, and S.A. Guzman
del Proo (eds.), Abalone of the world: Biology, fisheries and
culture. Blackwells, Oxford (in press).
Shepherd, S.A., G.P. Kirkwood, and R.L. Sandland
1982 Studies on southern Australian abalone {genns Haliotis).
III. Mortality of H. laevigata and H. ruber. Aust. J. Mar.
Freshwater Res. 33:265-272.
Shepherd, S,A., P.S. Clarkson, and J.A. Turner
1985 Studies on southern Australian abalone (genus Haliotis).
V. Spawning, settlement and early growth of H. scalaris.
Trans. R. Soc. S. Aust. 109:61-62.
Sloan, N.A., and P. A. Breen
1988 Northern abalone, Haliotis kamtschatkana, in British
Columbia: Fisheries and synopsis of life history information.
Can. Spec. Publ. Fish. Aquat. Sci. 103, 46 p.
Sluczanowski, P.R.
1984 A management oriented model of an abalone fishery
whose substocks are subject to pulse fishing. Can. J. Fish.
Aquat. Sci. 41:1008-1014.
1986 A disaggregate model for sedentary stocks: The case of
South Australian abalone. In Jamieson, G.S., and N. Bourne
(eds.), North Pacific workshop on stock assessment and man-
agement of invertebrates, p. 393-401. Can. Spec. Publ. Fish.
Aquat. Sci. 92.
Stanistreet, K.
1978 Abalone sales prospects may be on the mend. Aust. Fish.
37:29-30.
Tegner, M.J.
1989 The California abalone fishery: Production, ecological
interactions, and prospects for the future. In Caddy, J.F. (ed.),
Scientific approaches to management of invertebrate stocks,
p. 401-420. J. Wiley & Sons, NY.
Tegner, M.J., and P.K. Dayton
1987 El Nino effects on southern California kelp forest com-
munities. Adv. Ecol. Res. 17:243-279.
Tegner M.J., P.A. Breen, and C.E. Lenner
1989 Population biology of red abalones, Haliotis rufescens,
in southern California and management of the red and pink,
H. corrugata, abalone fisheries. Fish. Bull., U.S. 87:313-339.
146
Fishery Bulletin 90(1). 1992
Tegner, M.J., J.D. DeMartini, and K.A. Karpov
1992 The California red abalone fishery: A case study in com-
plexity. In Shepherd, S.A., M.J. Tegner, and S.A. Guzman
del Proo (eds.), Abalone of the world: Biology, fisheries and
culture. Blackwells, Oxford (in press).
Tissot, B.N.
1992 Mass mortality of the black abalone (Haliotis cracherodit)
in the California Channel Islands. In Shepherd, S.A., M.J.
Tegner, and S.A. Guzman del Proo (eds.), Abalone of the
world: Biology, fisheries and culture. Blackwells, Oxford (in
press).
Walters, C.J., and J.S. Collie
1988 Is research on environmental factors useful to fisheries
management? Can. J. Fish. Aquat. Sci. 45:1848-1854.
Wilkinson, L.
1990 SYGRAPH: The system for graphics. SYSTAT, Inc.,
Evanston, IL.
Abstract.- On l October 1986,
New Zealand introduced Individual
Transferrable Quota (ITQ) manage-
ment for most of its fisheries. ITQ
management was implemented to
address overfishing, overcapitaliza-
tion, and excess government regula-
tion. Quotas were based on catch
histories, with a quota "buyback"
(costing $42.4 million NZ) and pro-
rated cuts to achieve total allowable
catch (TAG) levels indicated by pre-
liminary stock assessments. Fixed
amounts of quota (defined by weight)
were issued in perpetuity. Annual
stock assessments are conducted.
Government stated that it would buy
or sell quota at market-determined
prices in order to adjust TACs. On
1 April 1990, ITQs were redefined as
proportions of annual TACs (known
as "proportional ITQs"). Govern-
ment extracts resource rent.
To date, there is little evidence of
improvement in the condition of the
fisheries resources. It is difficult to
determine the economic effects of
ITQ management; however, econom-
ic conditions have worsened due to
factors which are unrelated to ITQ
management. Revenues to govern-
ment from the ITQ system have ex-
ceeded total costs, but there would
have been a deficit if government
had purchased quota to reduce TACs
to the levels indicated by stock as-
sessments. Government regulation
has not been reduced.
Although there is general support
for ITQ management in New Zea-
land, many problems have been en-
countered: quota overruns resulting
from bycatch; inadequate stock as-
sessment capability; disagreement
over the level of resource rentals;
and failure of government to enter
the marketplace to reduce TACs
when necessary.
ITQs in Mew Zealand: The era
of fixed quota in perpetuity
Michael P. Sissenwine
Headquarters, National Marine Fisheries Service, NOAA, Silver Spring, Maryland 20910
Pamela M. Mace
P,0 Box 7357, Silver Spring, Maryland 20907
The idea of managing fisheries by In-
dividual Transferable Quotas (ITQs)
is not new. Christy (1973) suggested
the method, and Maloney and Pearce
(1979) provided the economic ration-
ale for it. Until recently, there were
only a few applications of ITQ man-
agement (e.g., southern bluefin tuna,
Geen and Nayar 1988; Lake Erie
freshwater fisheries, Muse and
Schelle 1989). One application that
has received considerable attention is
the ITQ management of fisheries in
New Zealand. Two reasons for this
attention are that (1) New Zealand is
applying ITQ management on a more
comprehensive national scale than
ever before, and (2) New Zealand
officials have done a good job of de-
scribing their ITQ system to the rest
of the world (e.g., Clark et al. 1988,
Crothers 1988). New Zealand's early
experience with ITQ management
is of interest to the United States
because ITQ management is being
planned or discussed for several fish-
eries (e.g., Pacific sablefish and hali-
but, South Atlantic wreckfish, and
East Bering Sea groundfish). It has
recently been implemented for Mid-
Atlantic surf clams and ocean qua-
hogs. This paper reviews the poten-
tial benefits and problems of New
Zealand's ITQ management system
based on firsthand observations of
the authors. 1 The main body of the
paper was completed in mid-1990. A
postscript has been added to reflect
more recent events through 1991.
Before describing the fisheries man-
agement situation in New Zealand,
the authors want to caution that by
pointing out problems, they are not
condemning the ITQ system. Despite
problems, there seems to be a gen-
eral acceptance that ITQs are the
way New Zealand fisheries will be
managed. There is no widespread
sentiment, either within government
or the industry, to repeal ITQs. A
regional poll conducted shortly after
implementation of the ITQ system
(Dewees 1989) found that the major-
ity of the fishing industry favored it.
It would be interesting to repeat the
poll nationwide now. The authors are
of the opinion that the industry would
not want to return to the fisheries
management situation (or lack there-
of) that preceded ITQs.
New Zealand
fisheries setting
Fisheries have always been impor-
tant to New Zealand. Legend has it
that a Maori (the native people of
New Zealand) pulled up the North
Island of New Zealand from the sea
on a hook-and-line while fishing.
Fishing was so important that the
Manuscript accepted 16 December 1991.
Fishery Bulletin, U.S. 90:147-160 (1992).
' The authors of this paper were fortunate to
have the opportunity to observe ITQ manage-
ment in New Zealand firsthand. The first
author made six trips to New Zealand dur-
ing the first three and a half years of ITQ
management, including approximately seven
months employed by the New Zealand Fish-
eries Research Centre. The second author
was employed by the New Zealand Fisheries
Research Centre from August 1986 until May
1989. Both authors maintain contact with the
fisheries management situation in New Zea-
land through their previous affiliations.
147
148
Fishery Bulletin 90(l|, 1992
Treaty of Waitangi between the Maoris and the British,
signed in 1840, deeds the Maoris' rights to their tradi-
tional fisheries.2
Although New Zealand is a small nation in terms of
population and land area, its Exclusive Economic Zone
(EEZ) of 1.3 million nm^ (more than 15 times the land
mass) is the fourth largest in the world. Most of the
EEZ is deep; 72% of the zone has waters deeper than
1000 m, so it is difficult to judge the total potential
yield.
Historically, New Zealand fisheries were restricted
to coastal waters (<200m in depth) and yielded less
than 50,000 tons annually (Fig. 1). Deepwater fisheries
(to 1500m) developed during the 1970s, and the yield
increased rapidly to a peak of about 500,000 tons in
1977. Most of the increase was due to foreign fishing.
In 1978, New Zealand extended its jurisdiction to 200
miles. The yield decreased sharply for a few years, but
it has since returned to about 500,000 tons. Since
extended jurisdiction, domestic fishing has replaced
almost all of the foreign fishing. However, it should be
noted that much of the catch recorded as domestic is
actually taken by foreign vessels and foreign crews
imder contract to New Zealand firms. In 1987, the first
sale value of the catch was about $350 million NZ.'^
The export value of New Zealand fisheries products in-
creased from $50 million NZ in 1977 to $676 million
NZ in 1987. The 1987 figure represented about 6% of
New Zealand's total exports (Bevin et al. 1989).
Fisheries management began with the Fisheries Act
of 1908 which established authority for input controls,
such as limited entry licensing, closed areas and sea-
sons, controls on minimum fish sizes, and requirements
for vessels to land at specific ports. The actual basis
for the number of licenses allowed in the fisheries is
unclear. Restrictive licensing was repealed in 1963.
New Zealand established authority for output con-
trols (i.e., total allowable catches, or TACs) in 1978
when it extended jurisdiction. At the same time, a
moratorium was placed on new fishing permits for rock
lobsters and scallops. In 1980 the moratorium was ex-
tended to finfish permits. In 1983, a Deepwater Enter-
prise Allocation system was established. Deepwater
Enterprise Allocations were a forerunner of ITQs.
Quota for each of the species fished in deep water
(below about 200 m) was allocated to nine companies
which had already invested in deepwater harvesting
and shoreside processing capability. The motivation for
the Deepwater Enterprise Allocations was not over-
1 'T 1 1 1 1 1 1 1
90 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990
Figure 1
Annual yields from New Zealand EEZ, 1890-1990.
fishing or overcapitalization. It was intended to pre-
vent these ills from occurring (Clark et. al. 1988)
Presumably, it also encouraged investment in the deep-
water fisheries and hastened the replacement of for-
eign fishing activity by domestic fishing. The quotas
were initially awarded for a period of ten years, but
were made permanent in 1985. Although the govern-
ment had no authority to make quotas transferable,
there was considerable de facto trading and leasing of
shares among the nine companies.
New Zealand implemented ITQs for most of its fish-
eries in October 1986. The Government gave several
reasons for introducing ITQs. According to Crothers
(1988), "Fishery managers were faced with an open
access inshore fishery under severe biological and
economic pressure . . . many of the prime species were
experiencing growth and probably recruitment over-
fishing. . .and the industry was overcapitalized, crip-
pled by excessive government management interven-
tion, and rapidly declining economic performance." A
government publication titled "Inshore Finfish Fish-
eries: Proposed Policy for Future Management"
(Anonymous 1984) stated that "... a broad description
of the problem of the inshore fishery is that the major
fish stocks are too low as a result of overfishing. . .
there has been a moratorium on new entries to the
inshore . . . part-time fisherman were removed admin-
istratively . . . this had a negligible effect on fishing
effort or catch . . . the harvesting sector remains over-
capitalized. "■* In summary, the government turned to
^ The fishing heritage of the Maori people and the Treaty of Waitangi
are more than a matter of passing interest. As will be discussed
later in the paper, the Treaty of Waitangi has complicated implemen-
tation of ITQ management.
^Economic values are expressed in New Zealand dollars which equal
about $0.58 U.S.
* While the removal of part-time fishermen may have had a negli-
gible effect on fishing effort or catch, it did have social ramifica-
tions. Many of the part-time fishermen were Maoris. It could be
argued that their removal was one of the factors that stimulated
them to attempt to regain access to the fisheries through the courts
under the Treaty of Waitangi.
Sissenwine and Mace: ITQ management In New Zealand
149
ITQs because of perceived overfishing, overcapitaliza-
tion, and crippling excess regulation.
Undoubtedly, the success of the Deepwater Enter-
prise Allocation system contributed to the decision to
use ITQs to solve the perceived problems in the inshore
fisheries. Clark et. al. (1988) labeled it as a model for
inshore fisheries management. There was also a belief
that problems could be solved only by applying some
form of output controls (Sandrey and O'Donnell 1985),
and that input controls had already been attempted and
had failed (Crothers 1988). In fact, it is unclear how
seriously input controls had been attempted, or how
severely the fisheries were overfished or overcapital-
ized.^ Of course, the failures of input controls or over-
fishing and overcapitalization are not prerequisites for
ITQ management. It is better to put in place a prop-
erty rights system, such as ITQs, before problems
occur.
Implementation of ITQ management
in New Zealand
The idea behind ITQ management of fisheries is quite
simple. ITQs are intended to conserve the fisheries
resource by setting a TAG. They increase economic
efficiency by assigning ownership of portions of the
TAG, thus eliminating competition between harvesters
to obtain the largest possible share of the TAG. By
making quota transferable, ownership should eventual-
ly rest with the most efficient harvesters, since they
should be able to afford to pay the highest price to pur-
chase quota. Excess capital is likely to be removed from
the fishery as more efficient operators buy up enough
quota to make optimal use of the capital that remains
in the fishery.
In New Zealand, implementation of the ITQ manage-
ment system began with stock assessments of all of the
^It is interesting that there were virtually no input controls on New
Zealand fisheries during 1963-78 for rock lobsters and scallops and
1963-80 for finfish. Even after moratoria on new licenses were im-
plemented in 1978 and 1980, there were no additional direct con-
trols on fishing effort (e.g., limits on the number of days that could
be fished), although there were some indirect controls (e.g., closed
areas).
With regard to overcapitalization, the government estimated that
the harvesting sector was overcapitalized by $28 million NZ in 1983,
although details of how overcapitalization was defined and how it
was estimated are lacking (Anonymous 1984). Investment (book
value) in the harvesting sector in 1983 was estimated as $142 million
NZ (Bevin et al. 1989). This indicates that the harvesting sector
was overcapitalized by about 20%, which is almost certainly less
than some North American fisheries (e.g., Mid-Atlantic surf clams,
New England groundfish, Pacific halibut).
Clearly some inshore resources were overfished (e.g., snapper),
but it is difficult to evaluate how serious the overfishing problem
was in general. Stock assessment information is quite limited, as
will be discussed later in this paper.
fisheries resources to be managed. Initially, this in-
volved assessments of 153 management units, com-
posed of 26 species-groups in up to 10 management
areas per species-group. By April 1990, there were 169
management units, composed of 29 species-groups (45
species) and 10 major management areas. Forty-seven
of these management units are of minor importance
(in terms of amount of quota) with TAGs established
for administrative purposes only. There are insufficient
data to conduct meaningful assessments for most
management units. Initially, most of the TAGs were
based on one of two methods of estimation: (1) They
were equated to landings in the most recent year(s) for
which information was available, or (2) they were
equated to the product of a trawl-survey biomass
estimate and a stock productivity value in the range
0.05-0.15. The first method probably produced overly-
optimistic estimates of sustainable yields since recent
landings were often the highest on record. On the other
hand, the second method may have resulted in overly-
conservative estimates, since biomass estimates were
conservative (due to conservative assumptions about
the vulnerability of fish to trawl gear) and a maximum
productivity level of 0.15 is low (although there are
notable exceptions such as orange roughy). Other
methods used to estimate a few of the initial TAGs may
have produced reasonable results. These included use
of tagging data, yield-per-recruit analysis, and stock
reduction analysis.
For the deepwater fisheries, TAGs generally
matched the sum of quota allocations under the Deep-
water Enterprise Allocation system. These Deepwater
Enterprise Allocations were converted directly to
ITQs. In the inshore, a provisional maximum allocation
was determined separately for each fishing permit
holder as the average catch of that individual's best two
out of the three fishing years of October- September
1981-82, 1982-83, and 1983-84. These catch histories
were the basis for the initial allocation of quota defined
in fixed amounts by weight. Since the allocations were
based on the average of the best two-out-of -three years,
it was likely that the "Sum of Gatch Histories" (SGH)
would exceed the maximum annual catch that had oc-
curred during the base period. In addition, fishermen
were given the right to appeal their allocations if they
felt it did not represent their true share of the fisheries.
Of the 1800 fishermen notified of their catch histories,
about 1400 appealed, and many of these have subse-
quently increased their allocations. The appeals pro-
cess is still ongoing even though the ITQ system has
been fully implemented for more than three years.
If the SGH was equal to or less than the TAG, per-
mit holders were allocated their catch histories as ITQ
in perpetuity. TAGs in excess of the SGHs were offered
for sale. When the SGH exceeded the TAG, there was
150
Fishery Bulletin 90(1). 1992
a Government buyback of quota. Crothers (1988) in-
dicates that the buyback was to facilitate an orderly
"rationalization" of the industry, and to help create a
climate of support for ITQ management. Clark et al.
(1988) indicates the buyback was to reduce the mis-
match of fleet capacity to available catch. If the Govern-
ment was not able to buy back as much quota as was
necessary, prorated cuts in quota were made. This
threat of proration probably encouraged permit holders
to be more reasonable in determining the selling price
of their provisional allocation of quota.
The buyback cost the Government $42.4 million NZ
to purchase 15,700 tons of quota (the annual amount
the owners would have been entitled to catch in
perpetuity). Prorated cuts were made to reduce quota
by an additional 9500 tons. Presumably, the Govern-
ment felt that the potential increase in value of the
fishery when overfished stocks recovered merited the
cost of the buyback and the short-term losses that
resulted from prorated cuts.
Relatively few stocks accounted for most of the cost
of the buyback. Table 1 indicates that more than 85%
of funds spent on the buyback were used to buy quota
for four species (mostly in one management area where
traditional inshore fisheries are prosecuted). Nearly
50% were used for the snapper fisheries. The total
reduction from SCHs to TACs for the 1986-87 fishing
year (which began 1 October 1986) was 6%. For the
21 species that were involved in the buyback and pro-
rated cuts, the reduction was about 24%. For the four
primary species involved, the reduction was 54%.
Table 2 gives detailed information for the four
primary species affected by the buyback and prorated
costs. It is noteworthy that, in all cases, the SCH great-
ly exceeded the actual catch in the year just prior to
ITQs (1985-86). This means that a portion of the quota
that was bought back probably would not have been
caught. In fact, in all cases the actual catch in the first
year of the ITQ system (1986-87) was lower than the
TAG. This suggests there may have been a declining
trend in the resource condition from the base period
when SCHs were established to the point in time when
ITQs were implemented. It also seems likely, in the
authors opinion, that SCHs were inflated by the in-
dustry (i.e., a moral hazard phenomenon) in anticipa-
tion of ITQs. As a result, the government may have
spent much of the $42.4 million NZ to buy back quota
which would not have been caught; therefore, the
buyback may have had relatively little effect on fishing
mortality rates.
Since ITQ management was implemented in 1986,
stock assessments have been conducted annually for
each management unit, to the extent that the available
data allow. These assessments are conducted in Fish-
eries Assessment Meetings (FAMs) during the middle
Table 1
Buybacks and
prorated cuts
for implementation of New |
Zealand ITQ management.
Tons reduced
Payments
Species
(1000 s)
($NZ millions)
% Total $
Snapper
5.7
19.4
45.7
Rig
3.0
7.7
18.1
School shark
3.7
4.3
10.0
Hapuku bass
1.7
5.1
12.0
17 other
11.0
5.9
14.2
Total
25.1
42.4
100.0
Table 2
Relevant information for the foxir main
species
included in
the buyback and
prorated cuts under
New Zealand ITQ 1
management. Values in thousands of tons or $miliions NZ.
Hapuku
School
bass
Rig
shark
Snapper
Tons reduced
1.7
3.0
3.7
5.7
Cost of buyback
5.1
7.7
4.3
19.4
SCH (sum of
3.3
4.4
6.0
12.2
catch histories)
TAG 1986-87
1.7
1.4
2.4
6.5
(total allowable
catch)
Catch 1985-86
1.7
2.9
3.7
8.6
Catch 1986-87
1.1*
i.r
1.9*
5.4*
* Provisional
of the fishing year (April or May) in order to recom-
mend TAG adjustments for the next fishing year
(beginning in October). New Zealand law requires that
the TAG be set to produce the maximum sustainable
yield (MSY), as qualified by relevant factors including
economic and environmental considerations and
regional or global standards. Methods for estimating
yields have been refined since 1985 when the initial
TACs were calculated. New Zealand scientists now in-
terpret MSY in two alternative ways: a static inter-
pretation in which MSY is the maximum constant yield
(MC Y) that can be taken year after year from a fishery,
and a dynamic interpretation in which MSY is the max-
imum average yield (MAY) that can be attained by
varying the current annual yield (CAY) in response to
fluctuations in stock size (Annala 1989 and 1990, Mace
and Sissenwine 1989). MCY estimates are based on
historic estimates of stock biomass from resource
surveys, stock production models, or landings statistics.
CAY estimates are generally based on recent estimates
Sissenwine and Mace: ITQ management in New Zealand
151
of stock biomass and a target level of fishing mortal-
ity which is expected to produce MAY. Although the
dynamic (CAY) strategy leads to higher average yields,
the static (MCY) option has received the most atten-
tion for two reasons. First, the ITQ system was initially
specified in terms of fixed weights of quota, valid in
perpetuity. In practice, most TACs were constant.
Second, the facilities for fisheries research are inade-
quate for providing frequent updates of stock status
for all but a few of the more important fish stocks.
It should be recognized that FAMs are only part of
the process of determining the level of TACs. The ac-
tual advice to the Minister of Fisheries on the setting
of TACs is given by senior government officials who
integrate stock assessment information with other
considerations, including an evaluation of the risk to
the resource of not adjusting a TAG. But the authors
consider FAMs the best source of information on the
condition of the fisheries resources, since they are open
scientific meetings which formally document their
deliberations and conclusions.
When ITQ management was implemented, the gov-
ernment stated that it would adjust the TAC by enter-
ing the market to buy or sell quota at market-deter-
mined prices. Government also reserved the option to
make prorated cuts in quota. During the first three
years of ITQ management, government either sold
quota in perpetuity or leased annual quota for barra-
cuda, hake, ling, orange roughy, hoki, and stargazer
(Table 3). Most transactions were in the first year. A
total of $84.2 million NZ was collected in quota sales
and lease fees. But since the initial buyback when ITQs
were implemented, government has not entered the
marketplace to reduce any TAGs,^ despite the fact
that the need for reductions has been indicated by
several stock assessments (Annala 1989 and 1990; see
next section).
Since ITQ management should increase resource
rent, government charges an annual royalty (known as
a resource rental) on quota holdings. In order to dis-
courage speculation on quota (i.e., owning it without
using it), resource rentals are charged on quota hold-
ings rather than landings. This practice is an implied
guarantee that fish are abundant enough for all quota
to be caught without dissipating rent, which may not
be the case due to assessment errors, failure to adjust
TACs when assessments indicate TACs are too high,
and because of varying economic conditions.
Gilbert (1988) estimated that the ITQ system could
result in resource rents (referred to as surpluses in his
Table 3
Revenues from sale/lease of quota under New Zealand ITQ |
management, 1986-
-89.
Tons
$NZ
Species
(1000s)
(millions)
% Total $
Barracuda
1.7
1.7
2.0
Hake
1.3
2.2
2.6
Ling
2.1
2.2
2.7
Orange roughy
7.8
23.4
27.8
Hoki
131.0
58.2
63.2
Stargazer
1.8
1.5
1.8
Total
145.7
84.2
100.1
*The TAC for orange roughy on the Chatham Rise was reduced by
exchanging quota in that area for quota on the Challenger Plateau.
This was a temporary reduction for 1988-89, although stock
assessments indicated that a permanent reduction was necessary.
paper) of 15-45% of the first sale value of the catch,
depending on the species. His estimates reflect only the
benefits of reducing effort relative to the open-access
equilibrium (although the validity of an open-access
equilibrium baseline is questionable for some of New
Zealand's fisheries). They do not include the benefits
of eliminating competition for shares of an overall TAG.
If the average rent is 25% of the first sale value of the
fishing, then there is the potential for government to
extract at least $90 million NZ annually (i.e., 25%
of the 1987 first sale) as resource rentals. Resource
rentals averaged about $20 million NZ annually dur-
ing the first three years of ITQ management.
On 1 April 1990, ITQs were redefined as portions of
annual TACs. This eliminated the need for government
to adjust TACs by entering the marketplace to buy and
sell quota, and makes it more practical to vary TACs
in response to the inherent variability in fisheries
resources, and other factors (e.g., new scientific infor-
mation). The change to proportional ITQs came at a
time when government was facing a large liability
(discussed further below) to buy quota to adjust TACs.
Therefore, government agreed to freeze the rate of
resource rentals for five years and redistribute the
resource rentals to industry as compensation for TAG
reductions.
What has happened
under ITQ management
It is probably too early to conduct a formal evaluation
of ITQ management in New Zealand. A transition
period of 3-5 years, or longer, is to be expected. Many
of the species in New Zealand are long-lived, and it is
likely that adjustments in the condition of the resource,
which ultimately affect the economic benefits, will be
protracted. However, since some authors have already
declared New Zealand's ITQ management a success
152
Fishery Bulletin 90(1). 1992
(Clark et al. 1988), it is worth considering what has
happened to date, to the extent this is possible given
limitations in available information. As discussed
earlier, government authors and government publica-
tions indicate that the ITQ system was put in place to
address three problems: (1) conservation, (2) economic
performance, and (3) government intervention. The
initial effects of the ITQ system with respect to these
problems are discussed below.
Conservation
There is little evidence of improvement in the condi-
tion of fisheries resources; but since stock assessment
information is limited, it is difficult to know. The in-
crease in TACs that lead to the revenues reported in
Table 3 resulted from a reassessment of the stocks, and
not an increase in abundance.'' There is evidence that
some stocks have declined, most notably orange
roughy, which has been found to be much less produc-
tive than previously believed (Mace et al. 1990). The
current TAG for the largest stock of orange roughy ex-
ceeds even the most optimistic estimates of long-term
sustainable yield by a factor of three. ITQs are not
responsible for the problem, but have done little to
resolve it.
There are several species in addition to orange
roughy in need of TAC reductions. There is accum-
ulating evidence that TACs are too high in the long
term for valuable species such as hoki, squid, paua, and
rock lobster (Annala 1990). At the 1989 FAM (Annala
1989), MCY was estimated for 110 management units.
Twenty-one of the estimates were within 10% of the
TACs, 82 were less than 90% of the TAC, and only 7
were greater than 110% of the TAC. CAY was esti-
mated for nine management units. One estimate was
within 10% of the TAC, seven were less than 90% of
the TAC, and one was greater than 110% of the TAC.
In 36 cases, yield estimates were less than 50% of the
TAC. Reductions in TACs, either immediate or grad-
ually toward MCY or CAY estimates, were recom-
mended for several species. In other cases, reductions
were not recommended because of uncertainty in MCY
or CAY estimates, because accumulated biomass was
still being fished down (in new or developing fisheries),
or because recent catches indicated it was unlikely the
"In the case of hoki, the increase in TAC from 100,000 tons in
1985-86 to 250,000 tons in 1986-87 was controversial. Some com-
ponents of industry were skeptical of the assessment which was
in part based on a single hydroacoustics survey. The hydroacoustics
survey results were later found to be gross overestimates. So far,
the hoki resource has sustained the increase in TAC, but stock
assessment results (Annala 1990) suggest that a catch of 250,000
tons may not be sustainable over the longer term. Government is
giving high priority to monitoring the stock.
TAC would be reached. It should also be noted that
"actual" TACs are now almost invariably higher than
"official" TACs, mostly as a result of successful appeals
to the Quota Appeals Authority. Some of the differ-
ences are trivial, but a comparison between actual and
official TACs from Annala (1989) indicates that of the
122 scientifically-based TACs (i.e., excluding the 47 ad-
ministrative TACs), 25% of the actual TACs exceeded
the official TACs by more than 10%, and 6% were
higher by more than 20%.
There are also many species for which the TAC
greatly exceeds the catch. For example, in the 1987-88
fishing year, the TAC was undercaught by more than
10% in 122 (out of 169) management units (including
47 "administrative" management units that have TACs
of only 10-30 tons), and by more than 20% in 104 man-
agement units (Annala et al. 1991). For the 1988-89
fishing year, the total catch for all management units
was 66% of the sum of the actual TACs. In situations
in which TACs are nonrestrictive, they have little con-
servation benefit. In these cases, the stocks are either
being overfished (because TACs are too high), or they
would not be overfished without the ITQ system. There
are other cases in which TACs have been overrun (17
of the 169 management units exceeded the TAC by
more than 10% in the 1987-88 fishing year; Annala
et al. 1991). There are a number of mechanisms by
which fishermen can legally exceed their quota. Most
of these mechanisms were established in order to deal
with bycatch in multispecies trawl fisheries.
The general conclusion is that TACs are not closely
tied to the best available assessments of the fisheries
resources, nor are catches strongly controled by the
TACs. Some valuable stocks have probably declined in
abundance. To date, the track record of ITQ manage-
ment with respect to conservation is not good.
Economic effects
There is even less information on the economic effects
of ITQ management. ITQ management could increase
economic benefits through several mechanisms: (1)
Conservation could lead to an increase in resource
abundance and a decrease in harvesting costs; (2) the
initial buyback of quota and prorated cuts might have
caused some excess capital and labor to move to seg-
ments of the economy where they could add produc-
tion; (3) transfer of quota might have led to consolida-
tion of ownership by the most efficient operators, and
resulted in some excess capital being removed from the
fishery; and (4) elimination of competition for TACs
might have resulted in a more efficient harvest and an
increase in the value of product.
As discussed earlier, it is unlikely that ITQ man-
agement has resulted in an increase in population
Sissenwine and Mace: ITQ management in New Zealand
153
abundance. On the other hand, the dedine in the abun-
dance of orange roughy probably has not increased
harvesting costs so far. Although orange roughy abun-
dance has decreased considerably, the catch has been
stable. Since orange roughy are fished in dense, spatial-
ly and temporally predictable aggregations, the catch
rate is probably relatively insensitive to overall popula-
tion size (see Paloheimo and Dickie 1964, for a general
discussion of the phenomenon).
It is difficult to determine whether the initial buyback
of quota and prorated cuts reduced excess capital, but
it seems unlikely. As noted earlier, it probably did not
reduce fishing mortality in most cases because the
quota that was bought back would probably not have
been caught. Fishing mortality is a function of capital
investment in the harvesting sector (e.g., number of
vessels), labor inputs (e.g., number of days the vessels
are operated), and technology. It seems unlikely that
capital would have been removed from the fishery
unless fishing mortality were reduced.
There is evidence that quota holdings have been con-
solidated, presumably to more efficient owners.* Dur-
ing the period October 1986-April 1988, there were
15,580 quota sales involving 453,000 tons, and 3417
leases of quota involving 253,000 tons, the sum of
which exceeds the total amount of quotas (494,000 tons
owned privately and 64,000 owned by government);
therefore, some quota was involved in multiple trans-
actions (Muse and Schelle 1988). According to Bevin
et al. (1989), the total number of quota holders de-
creased by 5.7% during the first two years of ITQ
management. The amount of quota held by the top ten
quota owners increased from 57% to 80% of the total.
The number of quota holders with more than 50 tons
decreased by 37%. This consolidation in ownership of
quota does not necessarily mean that vessel ownership
has also been consolidated. Apparently, a number of
vessel owners who have sold their ITQ allotments to
fishing companies have also entered contracts to fish
that quota for periods of several years.
Unfortunately, the authors have not been able to ob-
tain reliable data on the number of vessels in the fishery
prior to and since ITQ management. There are some
data available (e.g., Anonymous 1987, Bevin et al.
1989), but the information is inconsistent. There are
* There is a legal limit to how much consolidation can occur. It is illegal
for a company to own more than 35% of the quota for a species
in any management area, or more than 20% of the quota for a
species overall. It is interesting that some segments of the fishing
industry have viewed the potential of consolidation of ownership
of quota negatively, while government fisheries managers have
generally viewed it as part of the process of increasing economic
efficiency (i.e., efficient harvesters can afford to buy quota from
less efficient harvesters). New Zealand government officials also
note that consolidation should reduce the cost of managing the ITQ
system.
Table 4
Investment anc
employment (in
larvesting sector and total;
processing-sector values
can be obtained by difference) in New
Zealand fisheries, 1983
-87 (from Bevin et al. 1989). Values
are in $millions
NZ (book-value) and numbers of employees.
1983
1984
1985
1986 1987
Investment
Harvesting
142
170
182
223 213
Total
353
405
437
510 550
Employment
Harvesting
3700
4000
4450
3800 4240
Total
7500
8000
8650
9200 10240
data that indicate a slight decrease in investment in
the harvesting sector in 1987, after several years
of steady growth (Bevin et al. 1989). On the other hand,
the data indicate that employment and investment
in the fisheries increased steadily through 1987
(Table 4).^
It is also difficult to evaluate the effects of eliminat-
ing competition for TACs, but there are some positive
signs. In informal discussions with members of the
fishing industry, the authors have been told that har-
vesters have modified their fishing practices to reduce
costs and/or increase the market value of their catches.
At this stage, it is unclear what economic effects ITQ
management has had. But, all other things being equal,
it seems reasonable that ITQ management should have
increased economic benefits. Unfortunately, all other
things are not equal.
Two events unrelated to ITQ management have
adversely affected the economic condition of the New
Zealand fishing industry. They are a weakening of the
price of product in export markets (particularly orange
roughy in the USA) and unfavorable exchange rates.
As a result, the industry had only a 4.3% return on in-
vestment (before income taxes) during the one-year
period beginning 1 April 1987 (Bevin et al. 1989). i^
While the overall economic benefit of ITQ manage-
ment to New Zealand is unclear so far, it was profitable
for the government. As noted earlier, the government's
revenues from sale or lease of quota was $84.2 million
NZ. It also collected about $60 million dollars in
"Note that there was a high rate of inflation during this period (3.6.
9.4, 15.3, 18.2, and 9.6% in 1983-87, respectively, or 69% overall)
which approximately offsets the increase in nominal value of capital
investment.
"It should be recognized that the economic condition of the New
Zealand industry is a controversial matter because of resource
rentals and fuel excise taxes. Bevin et al. (1989) indicate that in
1987 the industry paid $55 million NZ in resource rentals and fuel
excise taxes which reduced the rate of return on investment from
16.2% to 4.3% (before income taxes).
154
Fishery Bulletin 90|1). 1992
resource rentals during the first three years of ITQ
management. This income exceeds the cost of the
buyback ($42.4 million NZ) and the entire cost of the
government's fisheries research, management, and en-
forcement programs (about $30 million NZ per year).
And there is the potential for resource rentals to in-
crease substantially (see previous discussion). On the
other hand, the authors are of the opinion that govern-
ment should increase fisheries research considerably
if it is to produce adequate stock assessments to sup-
port ITQ management (i.e., to conserve without being
too restrictive). Furthermore, if government had
entered the marketplace and purchased quota to im-
plement the reductions suggested by yield calculations
performed at the 1989 Fisheries Assessment Meetings
(Annala 1989), the cost would have far exceeded the
revenue from the ITQ system (e.g., the reductions for
orange roughy alone would have cost in the range of
$60-150 million NZ).
Government intervention
The third problem that ITQ management was intended
to solve was excess government intervention. To date,
it has not reduced government intervention except by
removing the moratorium on new licenses. The mora-
torium was replaced by the requirement to own quota.
In addition, there are new recordkeeping/reporting
requirements and complicated rules that are intended
to cope with bycatch (Annala et al. 1991).
One form of government intervention that probably
hampered the fishing industry was restrictions on the
port at which harvesters were allowed to land their
catch. However, this restriction was removed prior to
ITQ management. Other forms of input controls, such
as minimum fish size restrictions and closed areas or
seasons, have usually not been removed. Some of these
restrictions are necessary, in addition to a quota, in
order to conserve the fisheries resources and to pre-
vent potential yield from being wasted." In other
cases, regulations were put in place to aid one segment
of the fishing industry relative to another. For exam-
ple, large factory trawlers are restricted from fishing
within 25 miles of the coast, which reduces direct com-
petition with smaller vessels.
General reaction
It is not surprising that implementation of ITQs in New
Zealand has been accompanied by controversy. The
" Fisheries management needs to consider two control variables: the
fishing mortality rate which can be regulated by a quota, and the
age- or size-at-first-capture which can be regulated by gear restric-
tions, area/season closures, or minimum fish size (Sissenwine and
Shepherd 1987).
newspapers report numerous charges by the industry
against the government. The industry is upset about
the level of resource rentals. There are complaints
about the fairness of the Quota Appeals Authority.
There were complaints that government had over-
estimated the productivity of the hoki resource when
it sold quota, and there are complaints that it has
overestimated the severity of the problem with orange
roughy now that it is attempting to reduce the quota.
Although there is strong support from industry and
government for ITQ management, many specific as-
pects of implementation are unpopular. This is probably
unavoidable for a system that is relatively complex and
so radically different from previous management.
Potential problems
From a theoretical perspective, ITQ management is an
ideal method which generates maximum net economic
returns, under some simplifying assumptions; but as
Copes (1986) points out, there are many potential
problems. Instead of reviewing Copes' list of potential
problems that apply to ITQ management in general,
this paper reviews actual and potential problems that
apply specifically to New Zealand. They are (1) prob-
lems arising from redefinition of quota ownership, (2)
implications of the Treaty of Waitangi, (3) inadequacy
of the scientific basis of TACs, (4) bycatch, (5) high-
grading, (6) enforcement, and (7) an adequate basis for
setting resource rentals.
Redefinition of quota ownership
The need to redefine ITQs from fixed quantities in
weight to proportions of the TAC resulted from gov-
ernment's failure to enter the marketplace to reduce
TACs when necessary. Early versions of the proposed
ITQ system included a "revolving fund" that would be
administered by the New Zealand Treasury. Resource
rentals and revenues from the sale of quota would have
gone into the fund which could then be used to buy back
quota as necessary. In fact, Crothers (1988) actually
reported that the revolving fund existed. However, the
fund never materialized and revenues paid to govern-
ment by the fishing industry were used for other
government functions. When faced with the over-
whelming cost of buying back quota to reduce the TAC
for orange roughy, the government announced its in-
tention to change the ITQ system from fixed to pro-
portional ITQ. The authors were surprised at how
rapidly government was able to obtain the legal
authority from Parliament to make such a fundamen-
tal, and economically significant, change in the system.
It took approximately one year from the time that
Sissenwine and Mace. ITQ management in New Zealand
155
government announced its intentions to convert the
system to proportional ITQs until the change became
effective on 1 April 1990.
The actual details of how the conversion will be im-
plemented had not been determined at the time this
paper was written, but some difficulties are almost
certain to be encountered. In order to gain industry
acceptance of the change, government agreed to freeze
resource rental rates for five years, and redistribute
these funds to compensate industry for quota reduc-
tions. Industry may have misjudged the amount of com-
pensation it will receive, since several of the species
that are most likely to have large quota reductions are
also the species that generate most of the resource
rentals (e.g., orange roughy, hoki, squid). Therefore,
the greater the reductions, the smaller the pool of funds
available for compensation.
One implication of converting from ITQ in fixed
amounts to proportional ITQ is that there will be
pressure to change the method of yield estimation from
an MCY strategy to a CAY strategy, with consequent
increases in the amount and variety of assessment in-
formation required. With quota as a fixed amount,
there was little change in TACs from year to year. With
ITQs as a proportion of the TAG, there will be greater
pressure from the industry to change TACs (particular-
ly to increase them when stock size is perceived to be
high).
Treaty of Waitangi
The Maori people have sued for rights to the fisheries
under the terms of the Treaty of Waitangi. There are
several related cases which had not been settled at
the time this paper was written, but it appears that
the Maori people are entitled to a significant amount
of quota. Prior to the ITQ system, when there was
no ownership of the fisheries, there was less incentive
for the Maoris to exercise provisions of the Treaty
of Waitangi. But when property rights were estab-
lished, and many Maoris were excluded from the
system because they were part-time fish harvesters
who had already been removed from the fishery, it
was inevitable that a controversy would follow. Bevin
et al. (1989) reported that industry has delayed major
investments in the fisheries because of uncertainty
about Maori fishing rights. Industry is concerned that
the eventual settlement with the Maoris will be at
their expense (i.e., they will not be compensated for
quota that is transferred to Maori ownership). The
dispute over the Treaty of Waitangi has also caused
government to delay adding important species into the
ITQ system.
Stock assessments
The scientific basis for assessing fish stocks, setting
TACs, and evaluating the overall performance of the
ITQ system is generally inadequate. New Zealand had
relatively little need for stock assessment capability
prior to ITQs. For the most part, their fisheries man-
agement was laissez-faire. In the case of data for
assessing deepwater species, New Zealand relied heav-
ily on foreign research vessels. When ITQs were im-
plemented, they were ill-prepared, in the opinion of the
authors, to conduct stock assessments for all of the
management units included in the system. The situa-
tion has improved since the implementation of ITQ
management as New Zealand scientists have developed
and refined the scientific basis for stock assessments,
but they have had inadequate support (e.g., research
vessels, data collection systems, and personnel). Inade-
quate assessment databases mean that the ITQ system
is operating under high levels of uncertainty. The price
of uncertainty is either conservative quotas or a high
risk of stock collapses.
Bycatch
Some bycatch is inevitable in multispecies fisheries.
This means harvesters will catch some fish for which
they do not own quota. New Zealand planned to
manage bycatch with a taxation scheme (referred to
as surrendering catch to the government or "Crown"),
which was intended to produce a neutral incentive for
bycatch. The tax was supposed to be high enough so
that harvesters would have no incentive to catch
species for which they did not hold quota, but if they
caught them as bycatch, it would be worth their while
to land them for sale. The problem is knowing what
the proper tax level is in order to result in a neutral
incentive. In some cases, even taxing 100% of the ex-
vessel value does not discourage fishing for species for
which harvesters do not hold quota. This is because of
vertical integration in the fishing industry and a very
high value added during processing.
There are several other provisions for dealing with
bycatch. Quota holders may overcatch by up to 10%
in exchange for next year's quota. They may trade
retrospectively for quota to cover catch they have
already taken. They may trade quota of certain species
to cover bycatch of certain other species (for specified
combinations of species, often involving one-way trades
only).
Another aspect of the bycatch problem is that it is
difficult to distinguish between bycatch problems that
are a conservation threat to the bycatch species and
those that result from setting the wrong TAC, as a
result of imprecise assessments. Regardless of whether
156
Fishery Bulletin 90(1). 1992
it is a conservation problem or not, bycatch constitutes
a management problem. It also constitutes a problem
for members of the fishing industry when they try to
adjust their portfolios of quota holdings to match their
landings. In theory, this can be done by buying and sell-
ing quota, assuming that the overall TACs match the
relative catch rates experienced by the fishing industry
in aggregate; but this may not be so.
Annala et al. (1991) reviewed the bycatch situation
in detail. In the 1987-88 fishing year, the quota was
overcaught for 33 (out of 169) management units, by
up to 74%. Nine management units were overcaught
by more than 20%. The frequency and magnitude of
overcatching increased from 1986-87 to 1987-88.
Hjghgradjng
Highgrading is the discarding or dumping of a lower
valued size or species of fish, in favor of keeping more
valuable fish. Although highgrading is illegal under the
New Zealand ITQ system, it is known to occur (Annala
et al. 1991). For example, it probably occurs in the
snapper fishery where there is a premium paid for high
quality fish for the Japanese "iki jime" (killed by spik-
ing the brain) market, and in the oreo dory fishery
where three species (spiky, and black and smooth oreo
dory) with significantly different values are managed
by a combined TAG. The amount of highgrading in
New Zealand fisheries has not been quantified.
Clark and Duncan (1986) felt that highgrading would
be " . . .a short term, transitional problem and should
disappear once the fishery recovers and product value
differential within the same stock diminish. . . " There
is little evidence that the fishery has recovered. Nor
should recovery of the fishery eliminate the incentive
for highgrading, unless the ITQ system is administered
such that TACs do not limit catch. If so, then other ad-
vantages of ITQ management would be undetermined.
Nor are the authors aware of reasons why ITQ manage-
ment should reduce value differences between species
or levels of quality.
Enforcement
ITQ management is potentially difficult to enforce.
New Zealand has some advantages over the United
States when it comes to enforcement. First, the popula-
tion is small, and therefore there is less scope for the
development of a domestic black market, although
black markets may be significant for some inshore
species consumed domestically. Second, the country is
remote, so that it is difficult to smuggle fish elsewhere.
Third, most fish are exported, which involves record-
keeping that helps to check the accuracy of quota
reports. Finally, fisheries enforcement is carried out
entirely by a single, coordinated agency.
New Zealand placed a high priority on establishing
enforcement capability when it implemented ITQs. It
reoriented enforcement from at-sea operations to
shoreside investigations. The emphasis moved from
conservation officers to accountants and investigators
and "electronic surveillance" (computerized data re-
cording). The industry is required to maintain and
submit several different types of records that are
necessary for monitoring catch and product flow.
Penalties for quota violations are heavy. They may in-
volve forfeiture of catch, vessel, and quota holdings,
in addition to fines of up to $10,000 NZ. A second of-
fense within seven years may result in prohibition from
participation in any aspect of the fishing industry for
up to three years. In addition, the fisheries enforcement
agency passes information on to the tax department,
which may then be used in income tax prosecutions.
It is difficult to assess how well this enforcement
approach is working.
Resource rentals
The New Zealand fishing industry is concerned about
the basis of setting resource rentals, although it does
not seem to dispute them in principle. The government
planned to gradually increase resource rentals^- until
the fair market value of quota was reduced to approx-
imately zero. In theory, government is extracting all
of the resource rent from the fisheries at the point in
time that there is no longer incentive to enter the
fisheries. The industry argued that not all of the re-
source rent should be extracted, since investment in
fishing is inherently risky.
It is arguable whether the market value of quota
reflects resource rent in the fisheries. The price paid
for quota should reflect the buyer's estimate of its net
present value. However, the buyer's estimate may be
incorrect (i.e., a bad investment). Even if the price paid
for quota is correct, it may not reflect rent in a par-
ticular year. In practice, the price paid for quota has
been extremely variable (e.g., from $13 per ton to
$16,500 per ton for snapper; Bevin et al. 1989) for a
variety of reasons (e.g., imperfect knowledge, inclusion
of other assets in the price of quota, different discount
rates, noncompetitive price setting). This makes it dif-
ficult to use the sales price of quota as a criterion for
setting resource rentals.
'-The law limits increases in resource rental rates to 20% per year.
Sissenwine and Mace: ITQ management in New Zealand
157
Table 5
Problems and benefits of fisheries management by input |
controls, quotas (Q), and ITQs. The symbol "0"
is used as the
standard. The symbol '
+ " means a more difficult problem |
or greater benefit than
"0." The symbol
' + +'
' means even
greater problems or more benefit than '
+ ."
Type of management
Input
lACs
ITQs
Problems
Stock assessments
0
+
+
Catch statistics
0
+
+ +
Enforcement
0
+
+ +
Bycatch
0
+
+ +
Benefits
Conservation
0
0
0( + )
Economics
0
0
+
General issues
Many potential problems of ITQ management are prob-
lems associated with TAG management in general. In
some cases they are exacerbated by individual quotas.
Table 5 compares the problems and benefits associated
with input controls (e.g., effort limits, closed areas or
seasons), TACs, and ITQs. TAG management requires
more frequent and timely stock assessments than
management by most input controls (Sissenwine and
Kirkley 1982). The problem is particularly severe for
short-lived species (Gopes 1986). The problem of pro-
viding stock assessments for ITQ management is about
the same as that for TAG management. Gatch statistics
are one component of stock assessments. The need for
catch statistics is generally greater for TAG manage-
ment than for management by input controls. The need
is even greater for ITQs because statistics on individual
quota holders are the basis of management. Both TAG
and ITQ management encourages "data fouling" or
misreporting (Gopes 1986), although the incentive is
greater for ITQs. Similarly, enforcement is generally
more of a problem for TAG management (although this
is not universally true) because the catch has to be ac-
curately enumerated. For ITQs, it must be accurately
enumerated for individual quota owners, some of whom
may have developed successful methods for circum-
venting the system. The bycatch problem is more dif-
ficult for TAG management than for input controls. For
ITQs, the bycatch problem is even more difficult
because individual quota owners must adjust their port-
folios to match their multispecies catch rates.
In terms of the conservation benefits, input controls,
TAGs, and ITQs are all potentially effective (Sissen-
wine and Kirkley 1982). ITQs may have a potential ad-
vantage over TAG management because, with owner-
ship, there should be greater incentive for the industry
to cooperate. But limited-entry licensing (a form of in-
put control) also conveys privileges that may encourage
industry cooperation. In terms of economic benefits,
ITQs are superior in theory. Both input controls and
TAG management eventually allow dissipation of
resource rent. For both forms of management, there
is an incentive for fishermen to increase their cost of
fishing, in order to gain a larger share of the resource,
until the rent is dissipated. In practice, the actual
economic benefits of input controls, TAGs, and ITQs
are probably fishery-specific.
Learning from
New Zealand's experience
There is much to be learned from New Zealand's ex-
perience with ITQ management. New Zealand took a
systems approach. Gomprehensive new legislation was
introduced. Enforcement needs, penalty schedules,
reporting and recordkeeping requirements (including
wholesalers and retailers), a quota trading system, a
process for appealing initial allocations, a buyback
scheme for "rationalization" of some fisheries, mech-
anisms for controling bycatch, the principle of resource
rentals, and public and fishery industry education were
all considered. New Zealand made some mistakes, but
it would have probably made more if its approach had
been piecemeal.
The authors are of the opinion that one mistake made
by New Zealand fisheries managers was to establish
ITQs in fixed amounts, valid in perpetuity. This method
was used because it was thought that ITQs in fixed
amounts would create a more certain environment for
industry; they would provide a mechanism for govern-
ment revenue-raising, since government believed TAGs
were conservative and future quota sales were likely;
and the trading price for fixed amounts of quota would
be the most effective method to obtain information to
set resource rentals (Glark et. al. 1988).
Apparently, the government did not recognize how
uncertain TAGs might be (due, for example, to errors
in stock assessments) or how often TAGs might need
to be adjusted (due, for example, to the inherent
variability in the size of fish stocks) by entering the
market to buy and sell quota, since the revolving fund
(or some other method) was not established. It is also
possible government did not expect the price of quota
to be so high as to make it prohibitively expensive for
the government to buy it to reduce quotas. In fact, the
sales price of quota may not have been economically
rational, in which case government would not want
158
Fishery Bulletin 90(1), 1992
to overpay to adjust TACs downward. But it should
be noted that the Government did sell quota for similar-
ly high prices. In any case, it seems more practical to
define quota as a portion of the TAG, in an uncertain
and dynamic environment.
In the authors' opinion, New Zealand fisheries man-
agers underestimated the complexity of the bycatch
problem. In a multispecies setting, the apparent in-
dependent fluctuations of each species complicate the
bycatch problem. In general, insufficient information,
variabOity between harvesters, and the complex organ-
ization of fisheries mean that it will be difficult to solve
the bycatch problem by adjusting a tax on bycatch.
Many fisheries are essentially single-species (e.g., surf
clams, herring, scallops, lobsters). These are the best
candidates for ITQ management with respect to by-
catch. If ITQ management is to be applied to multi-
species fisheries (e.g., New England groundfish), it
might be better to exclude some of the minor species
from the scheme, or to recognize that they may need
to be "sacrificed" in order to optimize fishing on the
more valuable species.
New Zealand lacked adequate stock assessment data
for a quota-based management system such as ITQs.
And, unfortunately, it will take time to develop appro-
priate time-series of data. In addition, there is much
that needs to be learned about the basic biology of the
deepwater species, many of which have only recently
been discovered in commercially viable quantities. The
basis for stock assessments is better in some other
places (e.g., throughout North America and Europe),
but the expectations for a high degree of precision may
still make stock assessment capability problematic.
ITQ management requires adequate monitoring and
enforcement capabOity to track individual catches. New
Zealand's enforcement of ITQs is geared towards in-
vestigations by accountants and auditors, instead of
traditional fisheries officers. In order for these inves-
tigators to be effective, the New Zealand fishing in-
dustry is required to maintain detailed "paper trails"
for products. Penalties for violations are severe. It is
too early to say whether this scheme is working, but
it is obvious that it will be necessary to impose addi-
tional recordkeeping to enforce ITQs in most cases in
the United States.
It is unclear how serious the overcapitalization prob-
lem was in New Zealand, but there are U.S. fisheries
that are severely overcapitalized (e.g., New England
groundfish). The buyback scheme in New Zealand pro-
bably did little to reduce overcapitalization. If a buy-
back scheme is intended to reduce overcapitalization,
funds should be used to reduce capital, and not hypo-
thetical catches that might not have been taken
anyway.
A positive lesson that should be learned from New
Zealand is the need to be clear about objectives when
applying an ITQ system. Glearly, one of the intentions
of New Zealand's fisheries managers was to increase
resource rent in the fisheries and to extract the rent
(through annual royalty payments^^) for the general
benefit of the country. What will be the objective for
applying ITQ management elsewhere? If the objective
is conservation, then quota management (or other
forms of management) is sufficient in theory, although
pressure from an overcapitalized fishing industry may
prevent TAGs from being set conservatively enough.
If the objective is economic efficiency, then it is impor-
tant to address distributional issues (resource rents,
producer surplus, and consumer surplus).
There is a great potential for ITQ management, but
it is not a panacea. When ITQ management is applied,
it is important that it be approached with realism and
based on adequate experience and data.
Postscript
Approximately 20 months have passed since New
Zealand converted its ITQ program from one of fixed
quota valid in perpetuity to one based on quota speci-
fied as a proportion of an annual TAG (also referred
to as a percentage ITQ system in New Zealand or a
percentage quota share system in the United States).
As predicted in this paper, the transition has been con-
troversial, in part because compensation available to
the industry in the form of resource rentals has not
been as large as anticipated. As a result, the fishing
industry filed a $150 million NZ court action against
the government. The lawsuit has since been settled out
of court.
In spite of the change from fixed to variable quota,
most TAGs have remained unchanged from one year
to the next. This is partly a result of inadequate infor-
mation for stock assessments. However, there have
been three notable reductions in TACs. The total hoki
TAG has been reduced from 250,000 to 200,000 tons,
Ghallenger orange roughy from 12,000 to 1900 tons,
and Chatham Rise orange roughy from 32,800 to
23,800 tons. The reduction in hoki quota was a reflec-
tion of new stock assessment results suggesting that
then-current TACs were unlikely to be sustainable; the
reductions in orange roughy TACs resulted from
assessment results suggesting that stock collapse was
imminent.
The anticipated need for large reductions in the
Chatham Rise orange roughy TAG was one of the
"At present, a legal basis for resource rentals in an ITQ system is
lacking in the Unites States.
Sissenwine and Mace: ITQ management in New Zealand
159
major factors that precipitated the change from fixed
to variable ITQs, since it could have cost the govern-
ment more than $100 million NZ to buy back sufficient
quota to reduce the TAG to the estimated long-term
sustainable level. After the change, it was agreed that
the quota would be reduced at the rate of 5000 tons
per year to the sustainable level, the latter being re-
calculated periodically as new data became available.
Recent assessments (Francis and Robertson 1991) in-
dicate a sustainable level of 7000-9000 tons and show
that the risks of stock collapse under the proposed
reduction schedule have increased due to the accumula-
tion of new data which has resulted in a decrease in
the point estimates of stock size and a decrease in
uncertainty of the estimates. The results clearly indi-
cate the need for a faster rate of reduction. However,
the fishing industry continues to oppose quota reduc-
tions, and at this point in time the government has
postponed the 5000-ton reduction schedule. The dis-
covery of new orange roughy aggregations in the
southern portion of the management area may alleviate
the problem in the short term, but the low productiv-
ity of orange roughy stocks means that any accum-
ulated biomass can be quickly fished down. Long-term
sustainable yields from orange roughy stocks are esti-
mated to be only about 1 .5-2.5% of the recruited virgin
biomass.
The problem of not reducing quotas when reductions
are indicated by assessments is exacerbated by wide-
spread rumors of quota busting, in spite of New Zea-
land's efforts to tailor enforcement to ITQ manage-
ment. Some of these rumors have been confirmed by
government sources.
New Zealand is now considering further evolution in
its fisheries management system towards a form of co-
management. Topics being debated include the need
to incorporate recreational fisheries into the manage-
ment system, the need to include all remaining ex-
ploited species-stocks, and the pros and cons of elim-
inating the current limits on aggregation of quota
(Pearse 1991). One objective is to transfer the costs of
management and responsibility for the resource to the
users of the resource, under the assumption that with
ownership comes motivation for conservation. Stay
tuned.
Citations
Annala, J.H.
1989 Report from the Fishery Assessment Plenary, May 1989:
Stock Assessments and Yield Estimates. Fish. Res. Cent.,
N.Z. Minist. Agric. Fish., Wellington, 158 p.
1990 Report from the Fishery Assessment Plenary, April-May
1990: Stock Assessments and Yield Estimates. Fish. Res.
Cent., N.Z. Minist. Agric. Fish., Welhng^ton, 165 p.
Annala, J.H., K.J. Sullivan, and A. Hore
1991 Management of multispecies fisheries in New Zealand
by individual transferable quotas. In Daan, N., and M.P.
Sissenwine (eds.), Multispecies models for management of
living resources. ICES Mar. Sci. Symp. 193:321-330.
Anonymous
1984 Inshore finfish fisheries: Proposed policy for future man-
agement. N.Z. Minist. Agric. Fish., WeUington, 31 p.
1987 Economic review of New Zealand fishing industry, 1986-
1987. N.Z. Fish. Ind. Board. Wellington, 56 p.
Bevin, G., P. Maloney, and P. Roberts
1989 Economic review of the New Zealand fishing industry,
1987-1988. N.Z. Fish. Ind. Board, Wellington, 56 p.
Christy, F.T.
1973 Fishermen quotas: A tentative suggestion for domestic
management. Occas. Pap. 19, Law of the Sea Inst., Univ. R.I..
Narragansett.
Clark, I.N., and A.J. Duncan
1986 New Zealand's fisheries management policies— Past,
present and future: The implementation of an ITQ based
management system. In Fishery assess control programs
worldwide, p. 107-141. Alaska Sea Grant Rep. 86-4, Univ.
Alaska, Fairbanks.
Clark, I.N., P.J. Major, and N. MoUet
1988 Development and implementation of New Zealand's ITQ
management system. Mar. Resour. Econ. 5:325-349.
Copes, P.
1986 A critical review of the individual quotas as a device in
fisheries management. Land Econ. 62(3):278-291.
Crothers, S.
1988 Individual transferable quotas: The New Zealand ex-
perience. Fisheries (Bethesda) 13(1):10-12.
Dewees, CM.
1989 Assessment of the implementation of individual trans-
ferable quotas in New Zealand's inshore fishery. N. Am. J.
Fish. Manage. 9:131-139.
Francis, R.I.C.C, and D.A. Robertson
1991 Assessment of the Chatham Rise (ORH 3B) orange
roughy fishery for the 1991/92 season. N.Z. Fish. Assess. Res.
Doc. 91/3, N.Z. Minist. Agric. Fish., Wellington, 36 p.
Geen, G., and M. Nayar
1988 Individual transferable quotas in the southern bluefin
tuna fishery: An economic appraisal. Mar. Resour. Econ. 5:
365-388.
Gilbert, D.J.
1988 Use of a simple age structured bioeconomic model to
estimate optimal long run surpluses. Mar. Resour. Econ.
5:23-42.
Mace, P.M.. and M.P. Sissenwine
1989 Biological reference points for New Zealand fisheries
assessments. N.Z. Fish. Assess. Res. Doc. 89/11, N.Z. Minist.
Agric. Fish., Wellington, 10 p.
Mace, P.M.. J.M. Fenaughty, R.P. Coburn, and LJ. Doonan
1990 Growth and productivity of orange roughy (Hoplostethus
atlanticus) on the north Chatham Rise. N.Z. J. Mar.
Freshwater Res. 24:105-119.
Maloney. D.G.. and P.H. Pearce
1979 Quantitative rights as an instrument for regulating com-
mercial fisheries. J. Fish. Res. Board Can. 36:859-866.
Muse. B., and K. Schelle
1988 New Zealand's ITQ Program. Alaska Commer. Fish.
Entry Comm. (CFEC 88-3), Juneau, 47 p.
1989 Individual fisherman's quotas: A preliminary review of
some recent programs. Alaska Commer. Fish. Entry Comm.
(CFEC 89-1), Juneau.
160
Fishery Bulletin 90(1). 1992
Paloheimo, J.E., and L.M. Dickie
1964 Abundance and fishing success. Rapp. P.-V Reun. Cons.
Int. Explor. Mer 155:152-163.
Pearse, P.H.
1991 Building on progress: Fisheries policy development in
New Zealand. Unpubl. rep. prepared for Minist. Agric. Fish.,
Wellington, NZ.
Sandrey, R.A., and D.K. O'Donnell
1985 New Zealand's inshore fishery: A perspective on the cur-
rent debate. Agric. Econ. Res. Unit Res. Rep. 164, Lincoln
College, Canterbury, NZ, 46 p.
Sissenwine, M.P., and J.E. Kirkley
1982 Fishery management techniques: Practical aspects and
limitations. Mar. Policy 6:43-58.
Sissenwine, M.P., and J. Shepherd
1987 An alternative perspective on biological reference points
and recruitment overfishing. Can. J. Fish. Aquat. Sci. 44;
913-918.
Abstract.— Reproductive behav-
ior and larval abundance of queen
conch Stromhus gigas L. were inves-
tigated near Lee Stocking Island,
Bahamas, with the primary purpose
of determining relationships between
physical variables, spawning frequen-
cy, and larval abundance. Monthly
observations made by divers at the
offshore spawning site showed that
copulation increased as a linear func-
tion of bottom water temperature
from April until the end of July,
when maximum summer tempera-
ture was reached. Pairing, copula-
tion, and egg-laying were all posi-
tively correlated with photoperiod
throughout the study period. The last
pairing and copulating conch were
observed in the middle of the warm-
est period in August suggesting that
stimuli other than temperature, such
as declining photoperiod, induce the
end of reproductive activity. The last
egg mass was found in early October.
There was a significant correlation
between spawning activity at the off-
shore reproductive site and larval
abundance in the adjacent downcur-
rent inlet. The first conch veligers
were found in plankton tows made
in early June, five weeks after the
first egg masses were observed at
the end of April. High larval density
was confined to July and August. Ad-
vanced-stage larvae, close to meta-
morophosis, were found only in the
vicinity of a shallow, benthic nursery
habitat. Comparison of reproductive
season in queen conch populations of
the Caribbean region showed no lati-
tudinal trend. In all areas, reproduc-
tion was associated with long days
and warm temperatures. Production
of conch larvae at the time of high
water temperature and steady trade
wind conditions may promote rapid
larval development and facilitate
transport of the vehgers to inshore
nursery habitats.
Seasonality in reproductive activity
and larval abundance of queen
conch Strombus gigas
Allan W. Stoner
Veronique J. Sandt
Isabelle F. Boidron-Metairon
Caribbean Marine Research Center
805 46th Place East, Vero Beach, Florida 32963
Manuscript accepted 31 January 1992.
Fishery Bulletin, U.S. 90:161-170 (1992).
The queen conch Strombus gigas L.
is the second most important fisher-
ies species in the Caribbean region,
after spiny lobster Panulirus argus
(Brownell and Stevely 1981). Conse-
quently, its general life history is well
known (Randall 1964, Brownell and
Stevely 1981, Berg and Olsen 1989).
Sexes are separate and sexual matur-
ity occurs at about 3V2 years of age,
a few months after the flared lip is
formed (Egan 1985, Wilkins et al.
1987, Appeldoorn 1990). Fertilization
is internal and copulation may pre-
cede spawning by several weeks
(D'Asaro 1965). An individual female
may spawn six to eight times during
a single reproductive season (Davis
and Hesse 1983). An egg mass, usu-
ally laid on clean, coral sand, takes
24-36 hours to produce and consists
of a single continuous egg-filled tube
folded upon itself to form a kidney-
shaped aggregate of eggs and sand
about 15cm in length. Robert-
son (1959) estimated that between
385,000 and 430,000 eggs were laid
in a single egg mass. Eggs hatch
after 5-6 days; pelagic veligers re-
main in the water column for 18-40
days prior to metamorphosis (Randall
1964, D'Asaro 1965, Brownell 1977,
Davis et al. 1987, Boidron-Metairon
1988, Mianmanus 1988).
Reproductive seasonality in queen
conch has been reported for different
sites within the Caribbean region (see
Fig. 6), but the mechanisms which
regulate reproductive behavior are
poorly known. In this study, we pro-
vide the first report on abundance
and seasonality of queen conch veli-
gers in the field, and examine re-
lationships between adult habitat,
reproductive activity, temperature,
photoperiod, and larval abundance.
Methods and materials
Study site
This study was conducted near Lee
Stocking Island (southern Exuma
Cays), Bahamas, an area known for
high abundance of queen conch (Fig.
1). The islands and cays of the Exuma
chain are bordered on the west by the
shallow Great Bahama Bank (mean
depth ~3m) and on the east by the
deep Exuma Sound. Waters from the
Exuma Sound flow onto the Bank
through numerous passes on the
flood tide and are mixed with Bank
water by wind-driven circulation.
Surface drogue studies (N.P. Smith,
Harbor Branch Oceanogr. Inst., Fort
Pierce, FL 34946, unpubl. data) in-
dicate that at the north end of Lee
Stocking Island, water flows through
Adderley Cay cut toward Shark
Rock. At the south end of the Island,
water flows through Rat Cay cut to
the west between Barraterre Island
and Children's Bay Cay. Most juve-
nile queen conch are located in shal-
low seagrass habitats on the Exuma
Bank; largest populations are found
near Shark Rock and southwest of
Children's Bay Cay.
In Exuma Sound, approximately
1km to the east of Lee Stocking
161
162
Fishery Bulletin 90(1), 1992
23°42
ye-os
Figure 1
Map of the Lee Stocking Island, Bahamas, study site showing plankton sampling sta-
tions (*). RS = location of the Reproductive Site, where observations and plankton col-
lections were made; RC = Rat Cay cut; AC = Adderley Cay cut; CBC = Children's
Bay Cay nursery site.
(called "mounds"), each sur-
rounded completely by bare
sand, were examined. All of the
mounds (designated with the
letter "M" in Fig. 2) were located
at depths of 18 m at the base with
tops between depths of 12 and
14m. (2) Sand habitats were
divided into two major regions.
Si is the extensive sand flat be-
tween the 10 m reef front and the
mound zone. S2 is the sand area
within the mound zone. (3) Rub-
ble and boulder areas are found
at the base of the 10 m reef in a
narrow band, with an extensive
boulder field (Bl) at the south-
east end of the study site. The
mounds and rubble, particularly
in the Bl area, are covered with
a turf of green algae (primarily
Cladophoropsis spp.), plus abun-
dant erect forms such as Hali-
meda spp. An area of mixed
hardground, sand, and coral
heads (HI) extends to the north
and east of the study site.
Island, there is a coral ledge at which depths increase
rapidly from 10 to 18 m. Beyond the ledge is a 1km-
wide platform with a gradual slope from 18 to 24 m.
Seaward from the platform, depth increases rapidly to
the deep basin of Exuma Sound. This geomorphology
is typical of the western side of the Exuma Sound.
Highest number and density of adult S. gigas occur on
the 18m-deep platform, which is beyond the normal
free-diving range of conch fishermen. In this area, more
than 99% of the conch are sexually mature (Stoner and
Sandt 1992). In the colder months, the conch are found
on algae-covered hardbottom; they move to sand for
mating and egg-laying in the summer.
A study site of approximately 12 ha surface area on
the 18 m platform was chosen for the investigation of
reproductive behavior and habitat association in adult
conch (Fig. 2). The particular location, north of the 10 m
coral ledge, was selected because of an abundance of
adult conch and close proximity of feeding and spawn-
ing habitats (Stoner and Sandt 1992). A scale map of
the site was constructed from compass bearings and
distances measured by scuba divers along the sides of
primary habitat features or boundaries. Figure 2 shows
all prominent features between the coral ledge and the
23m isobath.
Observations on reproductive behavior were made
in three habitat types (1) Five hard-bottom domes
Reproductive activity
Reproductive behavior was surveyed for 14 months,
on a monthly basis during the period of highest activ-
ity (March-October 1988) and at 6-8 week intervals
during January-February 1988 and November 1988-
February 1989. Longer sampling intervals were used
in the winter because preliminary observations near
Lee Stocking Island in previous years indicated that
no reproductive behavior occurs between November
and March. During each survey, spanning 5-15 days,
a scuba diver search for adult conch was made on
mounds Ml, M3, M4, and MS, in the boulder area (Bl),
in the rubble area (at the base of the coral ledge), and
in both sand zones SI and S2. During each survey
period, all conch were counted on each of the mounds
and at least one-half of the Bl area was examined. Very
few conch were found on M2 and this mound was aban-
doned early in the study. After determining that most
reproductive activity occurred on sand and not on hard-
ground or rubble (Table 1), the sampling protocol was
modified to locate at least 100 individuals on sand for
each survey. During winter months, less than 100
conch were located on sand in several days searching;
however, 100-300 animals were observed per month
during most of the reproductive season.
Stoner et al : Reproduction and larval abundance in queen conch
163
32 ^„sr N
^'^WfW *"^^
^^^-'-9 HI
^M2 ^
^ y
"^^^^4^
10 rnX^
100 m
^^^W^
Figure 2
Map of the Reproductive Site (see Fig. 1) showing elevated
Mounds (M), sand habitat (SI and S2), boulder area (Bl), and
area of mixed hardground and sand (HI). Sl-1 and Sl-2 are
transects over which density of conch were determined.
Table 1
Numbers and (percentages) of queen conch engaged in repro-
ductive activity on three substratum types near Lee Stock-
ing Island, Bahamas, 1988. Values for pairing and copulating
represent number of male/female pairs.
Behavioral type
Substratum
Pairing Copulating Egg-laying
Sand
Rubble
Hardground
51 (94.4)
0 (0.0)
3 (5.6)
28 (84.4)
2 (6.1)
3 (9.1)
148 (99.3)
0 (0.0)
1 (0.7)
(transects Sl-1 and Sl-2; Fig. 2) were examined each
survey period. Tiie transect surveys were made by a
scuba diver who counted all adult conch within a known
range while being towed 5 m above the sediment. High
water transparency resulted in a mean transect width
of 29 m (SD 6; range 20-40 m), which was measured
with a tape on each survey date. The total survey area
for each transect was calculated on the basis of horizon-
tal visibOity and the fixed distance of each transect line.
For additional information on the abundance of queen
conch on sand during the reproductive season, all adult
conch were counted in circles of 20 m radii at locations
of highest conch density in August 1987 (n 7 circles),
and in June (n 2) and July 1988 (n 2).
Each individual conch was classified in one of the
following reproductive categories. (1) Pairing: Two
conch were aligned, with the anterior part of the shell
of one animal overlapping the posterior part of the shell
of the other; but copulation was not observed. (2)
Copulating: Animals were engaged in copulation, with
the verge of the male beneath the mantle of the female.
(3) Egg-laying: A female was actively laying an egg
mass. (4) Non-reproductive: Conch was not engaged
in reproductive behavior.
Seasonality in reproductive behavior was quantified
by recording the percentage of total animals on sand
in each behavioral category. Notes were made on the
locations and substratum types (sand, rubble, hard-
ground) where pairing, copulating, and egg-laying
conch were found. Conch were measured for total shell
length (spire to siphonal groove) and greatest shell lip
thickness (approximately two-thirds of the distance
posterior from the siphonal groove). Shell measure-
ments were made to the nearest mm.
To estimate seasonal abundance of conch on sand,
two quantitative transects across the SI sand area
Physical measurements
To provide information on sediment grain-size and
organic content in the spawning habitat, sediment
samples were taken from the surface adjacent to
females laying eggs in August. Only eight samples were
collected; however, the sediment in sand areas SI and
S2 appeared to be of uniform grain size. An effort was
made to collect sediment samples from throughout the
study site. Sediments were frozen until laboratory
analysis. Organic content was determined by drying
a subsample (~100g wet wt) at 80 °C to constant
weight and incinerating at 500°C for 4 hours. Organic
content was quantified as the percent difference be-
tween dry weight and ash-free dry weight. Another
subsample (~50g wet wt) was examined for granulo-
metric properties. The sample was washed to remove
salts and extract the silt-clay fraction (<62^m). Silt-
clay was analyzed with standard pipette procedures
(Galehouse 1971), and the sand fraction with standard
dry sieve procedures (Folk 1966).
Bottom-water temperature was recorded with a
Ryan Instruments Temp Mentor placed at 17 m depth,
near the base of the coral ledge. The thermograph
recorded temperature with a precision of 0.2° C every
164
Fishery Bulletin 90|l). 1992
Figure 3
(A) Bottom- water temperature at 17 m depth
with 7-day averages (soHd line) and photo-
period in number of hours between sunrise
and sunset (dashed line). (B) Number of
queen conch Strombus gigas females on
sand engaged in various reproductive activ-
ities. (C) Number of conch larvae in Rat Cay
and Adderley Cay passes, January 1988-
February 1989.
30 minutes; 7-day averages were
generated and plotted (Fig.
3A). Surface-water temperature
and weather conditions were re-
corded each time that plankton
was collected.
To examine potential correla-
tion between reproductive sea-
sonality and photoperiod, a year-
long photoperiod curve (Fig. 3A)
was constructed for the study
site. Numbers of hours and min-
utes between sunrise and sunset
were calculated for local latitude
at 9-day intervals using the Nau-
tical Almanac.
Plankton collections
o
<
LJ
CL
00
UJ o
=3 2
Q ^
O o
CL ^
q:
<
>
en
5'
0\,
in
z
UJ
a
30
29
28
27
26
25
24
23
15
10
.'-'""''•->.
A
/ /'
/ J
•^•, y
'<
" v-v-../-'
-- 13
-■ 12
FMA MJJASONDJF
11
10
"0
I
o
—I
o
T)
m
O
o
o
c
5--
-o PAIRING
-A COPULATING
-■ EGGS
B
«-l f« 1 !-■—
FMAMJJ A SO NDJF
-•ADDERLEY CUT
-O RAT ISLAND "?
Daytime plankton collections
were made for queen conch veli-
gers from mid-March to October
1988. For seasonal analysis of
larval abundance, collections
were made every 2 to 3 weeks in
the pass between Lee Stocking
Island and Adderley Cay (Adder-
ley Cay cut) and in the pass be-
tween Rat Cay and Children's
Bay Cay (Rat Cay cut) (Fig. 1).
Additionally, collections were
made over the area surveyed for
reproductive activity (Reproduc-
tive Site) with the primary pur-
pose of detecting low densities of conch larvae at the
onset and end of the reproductive season. Collections
were not made at the Reproductive Site during peak
reproduction, between July and mid-August. To ex-
amine densities and size-frequency of larvae on Exuma
Bank, four collections were made over a known nursery
for S. gigas, west of Children's Bay Cay (Fig. 1). This
site is approximately 3.4 m deep and vegetated with the
seagrass Thalassia testudinum.
•m» « i» I
c
—1 1 r-
FMA MJ JA SON DJF
1988 1989
MONTH
In the passes, plankton were sampled during the first
2 hours of the flood tide; on the bank, tows were
scheduled during the last 2 hours of flood tide. Plankton
collections were made by tovvang a 0.5m diameter con-
ical net, 5 m long, with 202 ^m mesh. Two tows were
made at each site. Because the location of larvae in the
water column was unknown, collections at the Repro-
ductive Site were made by towing the net at 9 m depth
(midwater column) for 10 minutes, then raised near
Stoner et al,: Reproduction and larval. abundance in queen conch
165
the surface at 1.5 m depth for another 10 minutes. At
the other three sites where there was considerable ver-
tical mixing and shallow depth, the net was towed for
20 minutes in the upper 1.5 m of the water column.
Water volume sampled was calculated using a cali-
brated General Oceanics flowmeter, and larval abun-
dance was expressed in numbers of veligers per lOm*^.
To identify larvae, samples were refrigerated, sorted
live (within 4 hours), and compared with laboratory-
cultured larvae of the two most abundant Strombus
spp. in the central Bahamas, S. gigas and S. costatus.
Two other strombids occur in the Lee Stocking Island
area {S. galliis and S. raninus); however, both are very
rare relative to S. gigas and neither has been observed
on the windward side of the island or in the inlets. Shell
length, shell width, and shape of the shell tip were the
principle criteria used to identify early-stage larvae.
Number and shape of shell whorls and other shell
characteristics were used to identify advanced larval
stages. Measurement of shell length, from apex to
siphonal edge, was made with an ocular micrometer
and reported in microns for all intact shells.
Results
Conch reproduction
The reproductive season for Strombus gigas at Lee
Stocking Island extended from mid-April to early
October. The beginning of the season was marked
by a massive migration of conch from hardground
(mounds, rubble, and boulder areas) to sand habitats
(F'ig. 4) where first copulation, pairing, and spawning
were observed on 14, 15, and 25 April 1988, respec-
tively. In subsequent months, virtually all reproductive
behavior occurred on sand (see later). The number of
females engaged in reproductive activity increased
gradually from April (9.7% of total sampled population)
to July (18%) (Fig. 3B). In August, 13.8% of the popula-
tion were reproductively active females; the percentage
declined to less than 1.0% in September and October.
Last copulation and pairing were seen in August, but
egg-laying was observed through September. The last
egg mass was discovered on 5 October 1988.
The number of reproductive conch increased with
conch density on sand (Fig. 4) from January and Feb-
ruary (0 conch/1000m2) to July (10 conch/1000 m^).
Density decreased after the beginning of August and
was 0.61 conch/1000 m^ in October. Conch were ag-
gregated on some dates and not distributed evenly
along the transect lines. Large error bars in Figure 4
show that the two transect lines frequently had dif-
ferent densities of conch during the primary reproduc-
tive season. In August 1987, measurements in areas
with high conch densities ranged from 11.1 to 20.7
1 °
Q
§ 16.
LO
Z 1"
o
I^ 12.
O E
Z o 10-
4
'
O o
/
\
"2 8-
/
\
\
u. \
/
\
2 2
■
i/
•
■
\
0
^— • ' • — • • II
JFMAMJ JASONDJF
MONTH
Figure 4
Density of queen conch Strombus gigas on sand at the Repro-
ductive Site, off Lee Stocking Island, Bahamas, January 1988-
February 1989. Values are mean ±SE for two transects.
conch/1000 m2. Values as high as 29.7 conch/1000 m2
(SE 2.0) were found in June 1988.
Low bottom-water temperatures were observed from
early March to early April 1988 (near 23.6°C) (Fig. 3A).
First pairing and copulating conch were seen at a
temperature of 24 °C in mid-April, and the first egg-
laying female was found at 24.5 °C. The number of
copulating females increased as a linear function of bot-
tom water temperature until the reproductive max-
imum (r 0.916, F 15.726, p 0.029; March through July
1988). There was no significant correlation between
egg-laying and temperature (p 0.061) and pairing and
temperature (p 0.285). Bottom temperature was rela-
tively constant (28.3-28.8°C) from the end of July
through September; the last pairing and copulating
conch were observed during this period. Temperature
decreased rapidly after September, and the last egg
mass was found on 5 October. Water temperature was
26.5°C by late November 1988, decreasing to 25.1°C
in late December.
All pairing and copulation were confined to the
season with photoperiod greater than 12 hours, while
egg-laying was observed until day length declined
to 11 hours (Fig. 3A,B). Highest correlation oc-
curred between length of day and copulation (r 0.870,
F 24.838, p 0.001), but significant correlations were
also found between photoperiod and both pairing
(r 0.709, F 8.064, p 0.022) and egg-laying (r 0.838,
F 18.896, p 0.002).
A few conch were buried partially in sand in mid-
October 1987 and again in January and early February
1988. Burrowing was not seen again until mid-Sep-
tember 1988. In November, a few conch were buried
166
Fishery Bulletin 90(1). 1992
Table 2
Density of queen
conch larvae at the Reproductive Site off- 1
shore from Lee Stocking Island, Bahamas, and at the nursery
site near Children's Bay Cay on Exuma Bank, 1988. Values
are numbers of conch larvae/lOm' ±
SD(n
2).
Date
Location
Reproductive Site
Children's Bay Cay
15 March
0±0
—
28 March
0±0
—
7 April
0±0
—
20 April
0±0
—
19 May
0±0
—
2 June
0.26±0.11
—
6 June
0.04 ±0.03
—
16 June
0.99 ±0.25
—
29 June
0.30 ±0.09
—
13 July
—
0.82±0.23
28 July
—
1.35±0.47
12 August
—
0.77±0.41
23 August
0.77±0.32
—
14 September
-
0.17±0.10
23 September
0±0
—
6 October
0±0
~
in sand-filled depressions on the mounds. Some were
almost entirely covered with sand and the shells were
devoid of algae. It is possible that conch in the sand
habitat were underestimated during winter months
because of burial behavior; however, tag return data
(Stoner and Sandt 1992) suggest that most adult conch
move to hardground or rubble for the winter months.
The mean shell length of pairing, copulating, and
egg-laying females (i 226mm, SD 23.6, n 180) was
2.3% larger than that for males in reproductive be-
havior (x 221mm, SD 17.4, n 180). However, pairwise
ANOVA, using female-male pairs as statistical blocks,
indicated no significant differences in shell length
among pairs (F 1.155, p 0.358), or between females and
males (F 0.847, p 0.366). Results were similar in the
case of copulating conch (among pairs, F 1.105, p 0.430;
between females and males, F 0.112, p 0.743).
Reproductive activity in Stromhus gigas was rare on
hardbottom substrata (i.e., mounds, rubble and boulder
areas). Ninety-four percent of the pairing conch were
observed on sand, none were observed on rubble, and
only 5.6% were found on hardbottom (Table 1). Eighty-
five percent of copulating conch were found on sand,
with small percentages found on rubble and hard-
bottom.
A total of 149 egg-laying females were observed be-
tween April and October 1988; except for one female
found laying eggs on hardbottom in area Bl, all were
found spawning on sand (Table 1). Nine observations
100
UJ 90
<
> se-
ct:
<
o
a:
LU
CD
70
60-1-
50
40
30
20
10
0
N = 397
h _
SHELL LENGTH (urn)
Figure 5
Length-frequency distribution for veligers collected near Lee
Stocking Island, Bahamas, May-September 1988 (n 397).
of simultaneous pairing and egg-laying were made; only
one simultaneous copulation and egg-laying was ob-
served. All 148 females on sand were oriented perpen-
dicular to sand waves, with the anterior end of the shell
elevated near the crest of the wave and the egg mass
near the trough. Mean grain size of sediments collected
immediately adjacent to egg-laying females was 0.389
B (774/im) (SD 0.248, n 8), which is in the coarse-sand
classification. Sediments were poorly sorted as indi-
cated by a mean sorting coefficient of 0.967 Q (SD
0.302, n 8). Organic content was 3.45% of dry weight
(SD 0.69, n 8).
Larval abundance
Conch larvae were first collected at the Reproductive
Site on 2 June 1988 at a density of 0.26 larvae/10 m^
(Table 2), 5 weeks after the first egg mass was dis-
covered (Fig. 3B). Surface- and bottom-water temper-
atures were 27.5°C and 25.8°C, respectively. Veliger
density at the Reproductive Site ranged from 0.04
larvae/10 m'^ on 6 June to 0.99 larvae/10 m^ on 16
June. No plankton collections were made at this site
between 29 June (0.30 larvae/10 m^) and 23 August
(0.77/10 m^); during this interval, emphasis was
shifted to the Children's Bay Cay site on Exuma Bank.
Larvae were not found until 6 June in Rat Cay cut
and 20 June in Adderley Cay cut (Fig. 3C). By the end
of June, surface-water temperature was near maximum
(29.5°C and 29°C) in the two inlets. Highest density
in the tidal passes was 4.46 larvae/10 m^ on 21 July at
the Rat Cay cut, concurrent with maximum egg-laying
frequency (13%) and surface and bottom temperatures
Stoner et aL; Reproduction and larval abundance in queen conch
167
LOCATIONS
Bermuda
Florida Keys
Bahamas
Turks and Caicos
Mexico — — -
Jamaica
Puerto Rico —
U.S. Virgin Islands
St. Kitts / Nevis
Venezuela
REFERENCES
Berg ec al. (in press)
D'Asaro(l965)
Glazer (pers. comm.)
This study
Davis etal. (1987)
Cruz (1986)
Corral and Ogawa ( 1 987)
Salley (1986)
Appeldoom etal. (1987)
Randall (1964)
Coulston etal. (1987)
Willdns etal. (1987)
Brownell (1977)
Weil and Uughlin ( 1 984)
MONTHS
F M A M J
H 1 1 1 1 —
A S O N D
I 1 1 1 1 1
* no data prior to July
Figure 6
Reproductive seasons reported for Stromkis gigas in the Caribbean region. Seasonality
refers to any observations of reproductive behavior (copulating or egg-laying), and does
not include histological results (see text). Locations are arranged in order of latitude
from north (top) to south (bottom). Data for Mexico are for Banco Chinchorro on the
Caribbean coast.
density was 0.82/10 m^ (Table 2)
and surface-water temperature
was 31 °C. Highest density at
this site (1.35 larvae/lOm^) oc-
curred on 28 July, when numbers
of copulating and spawning
females were highest at the Re-
productive Site. On 12 August,
larval density declined to 0.77
larvae/lOm^, concurrent with
declines in reproductive activity.
Larvae were last collected over
the nursery area on 14 Septem-
ber at a density of 0.17 larvae/
10 m^; at this time, surface-
water temperature on the Exu-
ma Bank was 30 °C and repro-
ductive activity was near zero.
During the reproductive sea-
son, all but three of the conch
veligers collected were between
340 and 600 pim in shell length (x
384Mm, SD 64, n 394) (Fig. 5).
The largest three larvae (1350
^m) were removed alive from
samples collected at the bank site
in mid-July. Metamorphosis oc-
curred within 24 hours in all
three larvae.
of 29.8°C and 28.1°C, respectively. Larvae continued
to be found at the pass sites until the end of September,
but were not present at the Reproductive Site after 23
August. No veligers were collected at any of the sites
in October, concurrent with observation of the last egg
mass. At this time, surface-water temperature had de-
clined to 27.2°C and bottom temperature was 27.5°C.
Density of larvae at the Adderley Cay cut site showed
a direct correlation with the percentage of females
copulating (r 0.952, F 68.312, jo<0.0001) and egg-
laying (r 0.860, F 19.889, p 0.003). Densities of larvae
at the Reproductive Site and the Rat Cay cut site were
not correlated with copulation or egg-laying (p>0.05).
Maxima in larval abundance occurred during months
with highest water temperature, but there was no
significant correlation between abundance of larvae
and surface-water temperature at Adderley Cay cut (F
5.232, p 0.056) or Rat Cay cut (F 0.514, p 0.494) dur-
ing the reproductive season (June-October). Log-
transformation of the data did not improve the correla-
tion coefficients.
Plankton collections over the nursery area, west of
Children's Bay Cay, were begun on 13 July 1988; larval
Discussion
At Lee Stocking Island, the reproductive season for
queen conch began in April and ended in early October.
Although differences by a few months were found in
the occurrence of reproductive behavior, there was no
apparent trend related to latitude in beginning, end,
or length of reproductive season in queen conch from
Bermuda to Venezuela (Fig. 6).* The longest reproduc-
tive season was reported for the Caribbean coast of
Mexico (Banco Chinchorro) (Cruz 1986, Corral and
Ogawa 1987), where egg masses were found year-
round. One of the shortest reproductive seasons was
reported by D'Asaro (1965) for the Florida Keys, but
more recent, intensive observations have shown that
queen conch may spawn over at least a 9-month period
in Florida (R. Glazer, Dep. Nat. Resour., Marathon, FL
33050, pers. commun., Sept. 1990).
*For geographic comparison, "reproductive seasonality" refers to
any reported observation of pairing, copulation, or egg-laying in
queen conch, except where noted in the text. Histological data are
not included.
168
Fishery Bulletin 90(1). 1992
Geographic comparisons of seasonality in reproduc-
tion must be interpreted cautiously due to different
methods, frequency and number of observations, an-
nual variation, and different habitat types. For exam-
ple, Brownell (1977) found that egg-laying in Los
Roques, Venezuela, extended later into the season in
deep water than in shallow water. Quantitative mea-
sures of reproductive activity provide a basis for ex-
amining mechanisms of seasonality, which is more
useful than records of reproductive occurrence. In all
studies that present seasonal curves for reproductive
behavior or numbers of egg masses (e.g., Davis et al.
1984, Weil and Laughlin 1984, Corral and Ogawa 1987,
and this study), maximum reproductive activity was
reported during the warmest months of the year.
Control of gametogenesis and the physiology of egg
production are still unknown for 5. gigas, but histo-
logical studies of queen conch from Belize showed that
mature eggs and sperm were in the gonads year-round
(Egan 1985). External factors, therefore, are likely to
mediate seasonality in the intensity of reproductive
behavior and egg-laying.
Emphasis in the past has been placed on the poten-
tial role of water temperature in reproductive activ-
ity. Similar to observations in Los Roques, Venezuela
(Brownell 1977, Weil and Laughlin 1984), reproductive
activity at Lee Stocking Island began with rise in
temperature. At both locations, reproductive activity
intensified with increasing temperature to reach its
maximum during the warmest period. Brownell (1977)
suggested that a sharp temperature decline of 1.1°C
from November to December was responsible for the
termination of queen conch egg-laying in Los Roques.
Similarly, egg-laying at Lee Stocking Island ended as
bottom-water temperature began to decline steadily
from 28.6°C in late September to 25.1°C in December.
On the other hand, pairing, copulation, and egg-laying
all decreased suddenly between August and Septem-
ber, during a period of high and relatively-stable water
temperature, near 28.5°C.
Unlike the partial (early summer) correlation be-
tween reproductive behavior and temperature, pairing,
copulation, and egg laying were all positively correlated
with length of day throughout the year. Photoperiod,
therefore, may be one of the important environmental
variables which mediates the timing and length of
reproductive season. Synergistic interaction between
photoperiod and water temperature is possible.
In addition to decreasing length of day in late sum-
mer, increasing frequency and intensity of winds from
the northeast produce a surge reaching the bottom at
the Lee Stocking Island study site in the fall (Caribb.
Mar. Res. Cent., Vero Beach, FL 32963, unpubl. data).
The significance of wave disturbance is suggested
by our own anecdotal observations of short-term de-
creases in reproductive activity concurrent with 1-2
day periods of reduced temperature and increased
surge which occurred during the survey periods in early
summer. Reductions in reproductive activity with in-
creasing water turbulence have been noted for queen
conch in the Caicos Islands (Davis et al. 1984) and for
milk conch Stromhus costatus in Puerto Rico (R.S.
Appeldoorn, Dep. Mar. Sci., Univ. Puerto Rico, Maya-
guez, PR 00709, pers. commun.. May 1990).
As with temperature, photoperiod may influence the
production of mature gametes or have a direct effect
on the behavior of conch. It is likely that the combina-
tion of increasing water temperature, coupled with in-
creasing length of day, triggers the mass migration of
adult conch from hardground to sand habitats and to
search for mates. Decreasing length of day and increas-
ing wave surge appear to provide the best explanation
for termination of the reproductive season, as pairing
and copulation ended while bottom-water temperature
was high. Experimental analysis will be required to
determine the mechanisms involved in seasonal re-
productive rates. Temperature, rates of temperature
change, photoperiod, physical turbulence, and other
seasonally variable environmental factors will need to
be considered.
Similar to the findings of several others (D'Asaro
1965, Robertson 1959, Brownell 1977), egg-laying oc-
curred primarily on clean coral sand with coarse grain
size. Davis et al. (1984) noted that this type substrate
may be critical for reproductive activity. Copulation
and spawning stopped when they placed conch on a
bottom type other than coral sand. At Lee Stocking
Island, mating on hardbottom was observed, but was
rare. Given that only one egg mass was found on sub-
strate other than coral sand, it is clear that this is the
preferred, if not critical, substrate for egg-laying.
This study provides the first report on abundance and
distribution of queen conch veligers in the field. Veli-
gers were present in the water column from 2 June to
the end of September, in concordance with relatively
constant rates of egg-laying from April through Aug-
ust. Despite a spawning season spanning 7 months,
high numbers of larvae were present in the two inlets
only during a 2-month period (July and August).
Although mechanisms involved in seasonality of
larval production and survival are unknown as yet, it
is clear that larvae were most abundant during the
period of warmest water conditions. Summer spawn-
ing in Exuma Sound has adaptive significance. First,
high temperatures are associated with higher develop-
mental rates in pelagic larvae (Thorson 1950, McEd-
wards 1985, Boidron-Metairon 1987), decreasing the
time larvae spend in the plankton and probably reduc-
ing larval mortality (Strathmann 1980). However, in-
crease in temperature needs to be coupled with a food
Stoner et al : Reproduction and larval abundance in queen conch
169
supply sufficient to provide for higher feeding and
metaboHc rates (Scheltema and WilHams 1982). Sec-
ond, midsummer months are characterized by prevail-
ing tradewind conditions (i.e., relatively constant winds
and moderate seas from the southeast) in the Exuma
Cays. General circulation over the reproductive site
during this period was to the northwest, parallel to the
Exuma island chain (N.P. Smith, Harbor Branch
Oceanogr. Inst., Fort Pierce, FL 34946, unpubl. data).
This would facilitate transport of pelagic larvae past
the numerous inlets which veligers must enter to reach
primary nursery habitats on Exuma Bank. As veligers
are carried alongshore on the island shelf, they would
readily be drawn through the inlets on flood tides.
Northwest drift over the reproductive site may, in fact,
explain the close correlation between larval abundance
in Adderley Cay cut and reproductive activity occur-
ring upcurrent. Winter weather patterns, with fre-
quent passage of cold fronts and shifting winds and cur-
rents, would be less favorable for transport of conch
larvae to the Exuma Bank nurseries.
On the basis of laboratory growth curves (Boidron-
Metairon, unpubl. data), all but the three largest lar-
vae collected in this study were less than approximately
2 weeks old in a larval life stage near 30 days. There
are several possible explanations for the scarcity of ad-
vanced stage larvae: Late stages occupy habitats dif-
ferent from those of early-stage larvae (on or near the
bottom), the abundance of older stages in the water col-
umn is reduced due to natural mortality, and/or the late
stages are advected to different locations. Virtually
nothing is known about transport or behavior of queen
conch larvae in the field. Given the great significance
of recruitment processes to management of this rapidly
depleted fishery species, future research should include
studies of larval transport and settlement.
In summary, highest reproductive activity occurred
near Lee Stocking Island at a time of stable circula-
tion patterns, high temperature (28-30°C), and long
photoperiod. Maximum larval abundance in July and
August placed high numbers of veligers in the water
column at a time favorable for both high rates of
development and transport to nursery habitats. Prox-
imal mechanisms affecting short-term and seasonal
variation in reproduction in queen conch may include
temperature, rates of temperature change, photo-
period, wave-induced turbulence, and other variables
associated primarily with season.
Acknowledgments
This research was supported by a grant from the
Undersea Research Program of the National Oceanic
and Atmospheric Administration (U.S. Department of
Commerce) to the Caribbean Marine Research Center.
We thank R.I. Wicklund, Director of the Caribbean
Marine Research Center, for providing bottom-water
temperature data for the reproductive site. Thanks to
P. Bergman, N. Christie, K. McCarthy, 0. Monterrosa
and E. Wishinski for assistance in the field. R. Appel-
doorn, P. Colin, L. Jones, J. Shenker, J.-P. Thonney,
and anonymous reviewers provided helpful comments
on the manuscript.
Citations
Appledoorn, R.S.
1990 Fishing pressure and reproductive potential in strombid
conchs: Is there a critical stock density for reproduction?
Mem. Soc. Cienc. Nat. La Salle (In press).
Appeldoorn, R.S., G.D. Dennis, and O. Monterrosa Lopez
1987 Review of shared demersal resources of Puerto Rico and
the Lesser Antilles region. In Mahon, R. (ed.), Report and
proceedings of the expert consultation on shared Fishery
resources of the Lesser Antilles region. FAO Fish. Rep. 383:
36-57.
Berg, C.J. Jr., and D.A. Olsen
1989 Conservation and management of queen conch {Strom-
bits gigas) fisheries in the Caribbean. In Caddy, J.F. (ed.),
Marine invertebrate fisheries: Their assessment and manage-
ment, p. 421-442. Wiley, NY.
Berg, C.J. Jr., J. Ward. B. Luckhurst, K. Nisbet, and F. Couper
In press Observations of breeding aggregations of the queen
conch, Strombus gigas, in Bermuda. Proc. Gulf Caribb. Fish.
Inst.
Boidron-Metairon, I.F.
1987 Effet de la temperature sur la duree des periodes lar-
vaires de quatre especes d'Echinides des Caraibes. Bull. Soc.
Sci. Nat. Quest Fr. H. 8:75-79.
1988 A new approach to comparative studies of Strombus gigas
L. (Gastropoda: Strombidae) larvae at the developmental and
nutritional levels. Proc. Gulf Caribb. Fish. Inst. (In press).
Brownell, W.N.
1977 Reproduction, laboratory culture and growth of Strom-
bus gigas. S. costatus, and S. pugilis in Los Roques, Venezuela.
Bull. Mar. Sci. 27:668-680.
Brownell, W.N., and J.M. Stevely
1981 The biology, fisheries, and management of the queen
conch, Strombus gigas. Mar. Fish. Rev. 43(7):1-12.
CorraL J.L., and J. Ogawa
1987 Cultivo masivo de larva de caracol Strombus gigas en
estanques de concreto. Proc. Gulf Caribb. Fish. Inst. 38:
345-352.
Coulston, M.L., R.W. Berey, A.C. Dempsey, and P. Odum
1987 Assessment of queen conch (Strombus gigas). Population
and predation studies of hatchery reared juveniles in Salt River
Canyon, St. Croix, U.S. Virgin Islands. Proc. Gulf Caribb.
Fish. Inst. 38:294-306.
Cruz, R.S.
1986 Avances en la experimentacion de produccion massiva
de caracol en Quintana Roo, Mexico. Proc. Gulf Caribb. Fish.
Inst. 37:12-20.
D'Asaro, C.N.
1965 Organogenesis, development and metamorphosis in the
queen conch, Strombus gigas, with notes on breeding habits.
Bull. Mar. Sci. 15:359-416.
170
Fishery Bulletin 90(1). 1992
Davis, M., and C. Hesse
1983 Third world level conch mariculture in the Turks and
Caicos Islands. Proc. Gulf Caribb. Fish. Inst. 35:73-82.
Davis, M., B.A. Mitchell, and J.L. Brown
1984 Breeding behavior of the queen conch Strombiis gigas
Linne, held in a natural enclosed habitat. J. Shellfish Res.
4:17-21.
Davis, M., C. Hesse, and G. Hodgkins
1987 Commercial hatchery produced queen conch, Strombus
gigas, seed for research and grow-out market. Proc. Gulf
Caribb. Fish, Inst. 38:326-335.
Egan, B.D.
1985 Aspects of the reproductive biology of Strombus gigas.
M.S. thesis, Univ. British Columbia, Vancouver, 147 p.
Folk, R.L.
1966 A review of grain-size parameters. Sedimentology 6:
73-93.
Galehouse, J.S.
1971 Sediment analysis, /w Carver, R.E. (ed.), Procedures in
sedimentology petrology, p. 69-94. Wiley, NY.
McEdwards, L.R.
1985 Effects of temperature on the body form, growth, elec-
tron transport system activity, and development rate of an
echinopluteus. J. Exp. Mar. Biol. Ecol. 93:169-181.
Mianmanus, R.T.
1988 Induction of settlement and metamorphosis in larvae of
Aplysia brasiliana and Strombus gigas (Mollusca: Gastro-
poda). Ph.D. diss., Univ. Miami, Coral Gables, 171 p.
Randall, J.E.
1964 Contributions to the biology of the queen conch, Strom-
bus gigas. Bull. Mar. Sci. 14:246-295.
Robertson, R.
1959 Observations on the spawn and veligers of conchs (Strom-
bus) in the Bahamas. Proc. Malacol. Soc. 33:164-171.
Salley, S.
1986 Development of the statocyst of the queen conch larva,
Strombus gigas L. (Gastropoda: Prosobranchia). M.S. thesis,
McGill Univ., Montreal, 116 p.
Scheltema, R.S., and LP. Williams
1982 Significance of temperature to larval survival and length
of development in Balanus eburneus (Crustacea: Cirripedia).
Mar. Ecol. Prog. Ser. 9:43-49.
Stoner, A.W., and V.J. Sandt
1992 Population structure and seasonal movements in deep-
water queen conch, Strombus gigas. Bull. Mar. Sci. 51 (In
press).
Strathmann, R.R.
1980 Why does a larva swim so long? Paleobiology 6:373-376.
Thorson, G.
1950 Reproductive and larval ecology of marine bottom in-
vertebrates. Biol. Rev, 25:1-45.
Weil, E., and R. Laughlin
1984 Biology, population dynamics, and reproduction of the
queen conch, Strombus gigas Linne, in the Archipielago de Los
Roques National Park. J. Shellfish Res. 4:45-62.
Wilkins, R.M., M.H. Goodwin, and D.M. Reid
1987 Research applied to conch resource management in St.
Kitts/Nevis. Proc. Gulf Caribb. Fish. Inst. 38:370-375.
Abstract.- The effect of benthic
dredging on coastal fisheries has
been of concern for several decades,
but little work quantifying direct
population impacts has been pub-
lished. Modeling approaches have
been used extensively to assess ef-
fects of power plant entrainment on
fishery stocks. Several important dif-
ferences between power plant and
dredge operations prevent direct ap-
plication of these models to dredge
problems: Entrainment by dredges
is short-term, has a moving intake,
and affects all age-classes of the
population. We present an equiva-
lent adult loss model of impacts to
the Washington coast 'Dungeness
crab Cancer magister Dana fishery
from dredging of a navigation chan-
nel in Grays Harbor, Washington.
The model is driven by empirical
population data to account for spatial
and temporal variation in abundance
and age-class structure. Results
show that impacts are quite sensitive
to the type of dredge used and the
season in which dredging occurs.
Contrary to initial expectations, the
0 -I- age-group loss was unimportant
relative to losses from older age-
classes. Despite many limitations,
the model has proven useful for
focusing impact assessment work, as
a basis for scheduling construction to
reduce impacts, and as a basis for
scaling mitigation projects.
Predicting effects of
dredging on a crab population:
An equivalent adult loss approach
Thomas C. Wainwright
David A. Armstrong
Paul A. Dinnel
Jose M. Orensanz
Katherine A. McGraw
School of Fisheries, WH-10
University of Washington, Seattle, Washington 98195
Manuscript accepted 17 January 1992.
Fishery Bulletin, U.S. 90:171-182 (1992).
The effect of dredging on marine or-
ganisms has been an issue of environ-
mental concern for several decades.
Most studies on the impact of dredg-
ing and disposal of dredged material
are concerned with changes in in-
faunal species assemblages and com-
munity characteristics, and generally
measure effects by pre- and post-
dredging comparisons. Very little
work has been done on the direct ef-
fects of entrainment on populations
of mobile epibenthic invertebrates or
demersal fish, in part because such
species are difficult to quantify. The
reviews by Morton (1977) and Poiner
and Kennedy (1984) indicate a strong
research emphasis on habitat modifi-
cation (by either dredging or disposal
of sediments) and water column ef-
fects (turbidity, release of chemical
pollutants) during dredging opera-
tions. Water column effects were also
the focus of a workshop on anadro-
mous fish and dredging (Simenstad
1990). Virtually no published works
report on direct population losses
due to entrainment or burial during
dredging, except Stevens (1981) and
Armstrong et al. (1982). There are
few predictive models of dredging
impacts other than that of Bella and
Williamson (1980), who developed a
model of dredging effects in Coos
Bay, Oregon. Their model focused
on water chemistry and sediments,
but also gave some consideration to
broad categories of animals.
In sharp contrast, power plant en-
trainment and impingement of fish
has generated a large quantitative
modeling literature (e.g. van Winkle
1977). Among the methods used in
power plant assessments, the "equiv-
alent adult loss" (Horst 1975, Good-
year 1977) and "production fore-
gone" (Rago 1984) approaches are
transferable to dredging operations,
if sufficient biological and operational
data are available. There are, how-
ever, several noteworthy differences
between power plant and dredging
operations which require different
considerations in their analyses.
Firstly, power plant water intakes
operate continuously at a fixed loca-
tion, while dredging operations are
generally short-term, with a moving
intake. This means that continuous,
equilibrium approaches (e.g., MacCall
et al. 1982) are not appropriate for
dredging. Secondly, mobile benthic
invertebrate populations are char-
acterized by spatial aggregations and
seasonal shifts in distribution which
must be taken into account in esti-
mating entrainment by a moving
dredge. Finally, power plant entrain-
ment is usually restricted to a single
age-class (larvae or early juveniles),
whereas dredging removes all age-
classes present in the dredged habi-
tat, but may kill age-classes at dif-
ferent rates.
17!
172
Fishery Bulletin 90(1). 1992
The work we describe here ap-
plies an equivalent adult loss
model (the "Dredge Impact
Model" or "DIM") to assessing
entrainment loss to the Dunge-
ness crab Cancer magister Dana
fishery in and around Grays Har-
bor, Washington. The Grays
Harbor navigation channel (Fig.
1) extends from the harbor
mouth to the city of Aberdeen, a
distance of about 25 km. The U.S.
Army Corps of Engineers cur-
rently removes an average of 1.2
million m^ of sediment annually
from the channel during main-
tenance dredging. To improve
accessibility for deep draft ves-
sels, the Corps planned to vnden
and deepen the channel by re-
moving about 8.7 million m^ of
material over a two-year period
(McGraw et al. 1988). Based on
results and predictions of DIM,
the Corps changed their original
dredging program by modifying
gear, volume dredged, and location/season combina-
tions to minimize impact on crab within operational
constraints (including weather and protection of other
resources). Project construction took place throughout
1990, ending in January 1991. This paper extends an
initial analysis (Armstrong et al. 1987), incorporating
two additional years of biological data and providing
a more thorough analysis of year-to-year variation. The
study was undertaken in response to concerns of crab
fishermen and resource managers that Grays Harbor
is important as a juvenile crab nursery.
Dungeness crab provide major fisheries along the
west coast of North America, from central California
to southern Alaska (Botsford et al. 1989). Since 1945,
annual Washington coast crab landings have fluctuated
between 1.2 and 9.5 thousand metric tons per year (Fig.
2). The general life-history pattern of Dungeness crab
along the Washington coast is as follows (Gunderson
et al. 1990, Jamieson and Armstrong 1991). Females
molt to maturity along the open coast, generally in the
spring. Mating occurs at this time, but eggs are not
extruded until the following winter. Eggs generally
hatch between December and March, and larvae re-
main in the water column for a few months. Late-stage
larvae are found onshore in late-spring and summer,
where they settle to the bottom and metamorphose.
Settlement occurs both in nearshore coastal waters and
in estuaries; within estuaries, crab settle in both sub-
tidal and intertidal habitats. Crab settling in intertidal
GRAYS HARBOR
WASHINGTON
SCALE IN RiLOWETERS
Figure 1
Map of Grays Harbor, Washington, showing existing navigation channel (heavj' solid
line) and sampling strata (separated by dashed lines).
areas may remain there during their first summer, but
move into the subtidal zone in fall. Few older crab are
resident in the intertidal, but move on and off the tidal
flats with the tides (Stevens et al. 1984). Crab settling
in nearshore waters may remain there for life, but
there is evidence of some migration into the estuary
between their first and second summers. Crab remain
in estuarine subtidal areas for up to two years, but late-
juvenile and early-adult crab leave the estuary before
reproduction, which occurs mainly along the open
coast. Both female and male crab reach sexual maturity
at about 2 years of age. but males may not breed until
age-3 or older (Butler 1960 and 1961, Hankin et al.
1989).
Methods
Model structure
The calculation of crab loss is driven by two variables:
crab abundance (uncontrolled) and volume dredged
(controlled). Both of these vary in both space and time.
The two types of data are related through an entrain-
ment function that describes the number of crab en-
trained by each type of dredging gear as a function of
local crab density and volume dredged. Not all crab en-
trained are killed, so a second relationship describes
the number killed as a function of crab age and dredge
type. To apply the model, crab abundance is measured
Wainwnght et al.: Effects of dredging on a crab population
173
Figure 2
Historical landings in the Washington coastal Dungeness crab
fishery. Sources: 1920-47, Cleaver (1949); 1948-50, Wash.
Dep. Fish. (1951); 1951-87, Pac. Mar. Fish. Comm. (1989);
1988-91 are preliminary estimates (S. Barry, Wash. Dep.
Fish., Olympia, pers. commun.).
as density stratified by age, season, and location.
Dredging is described as the volume dredged by a par-
ticular gear in a location during a given season. Unad-
justed loss figures are converted to equivalent adult
loss by multiplying by the expected survival of crab
from a certain age-class and season to adulthood. This
approach is shown schematically in Figure 3, and
described in detail below. Because we could not resolve
older age-classes within our survey data, a crab was
considered to reach adulthood in winter of its age 2 +
year (i.e., approaching the end of its third year post-
settlement).
Calculating losses in this manner requires an underly-
ing concept of population dynamics and several simpli-
fying assumptions. Creating a detailed model of local
dynamics for a mobile benthic animal is difficult; there
is continuous mortality and migration among habitats,
the rates of which may vary with season, age, and
locality. This may be summarized by the usual mass-
balance equation for change in the population in a local
area over a discrete time period:
N(ti) = N(to) + R-M-E-HI,
(1)
where N is population abundance, to and tj are two
times, R is recruitment to the population (settlement),
M is mortality, E is emigration, and I is immigration.
Mortality and migration rates are rarely known ac-
curately (certainly not in our problem), so we have
taken an empirical approach to defining population
Volume Dredged
(gear, season, location]
Unadjusted Loss
E
Natural MortaJity
(age. season)
Equrvaleni Adurt Loss
Figure 3
Flowchart of Dungeness crab adult loss model, showing main
variables and structural categories (in parentheses).
abundance. The approach is similar to, but simpler
than, that taken by Boreman et al. (1981) for power
plant entrainment in an estuary. The model is a discrete
time, discrete age-population model with discrete
habitat structure. To allow for seasonal changes in
abundance or population structure, the year is sub-
divided into four seasons. Thus the population can be
described as the numbers in various age-classes pres-
ent in various habitat areas during particular seasons.
In our model, abundance of any age-class in an area
during a single time-step is taken to be the average
abundance estimated from field surveys. We assume
that all changes in abundance (i.e., mortality or migra-
tion) occur between time-steps, so that populations are
constant throughout a step. This assumption introduces
little error if the change during a step is small Oess than
about 10%), which will be true if time steps are relative-
ly short and rates of change are relatively low. To meet
this assumption in our application, we defined variable-
length seasons of relatively constant population struc-
ture (see Data and Estimation section below).
The starting point for our calculations is estimated
total crab density (D) for locations (1) and seasons (s),
combined with age-class proportions (P). (Variables are
fully defined in Table 1.) The second set of informa-
tion needed for the calculation is the dredging schedule,
expressed as volume dredged (V) by a specific gear type
(g) in a specific location and season. For planning
174
Fishery Bulletin 90(1). 1992
Table 1
Model notation.
Symbol
Description
Subscripts
a
1
age-class
location
s
season
g
dredge gear
Population
D,3
density
age class proportions
natural survival to adulthood
Dredging
Vis,
volume dredged
Entrainment
Sg
entrainment rate
f".sg
dredge-induced mortality
proportion
Loss
E|,g
total entrainment
Lalsg
unadjusted loss
equivalent adult loss
purposes, volume was measured as thousands of cubic
yards (key) of dredged material (1 kcy = 765m3).
To obtain crab loss due to dredging from these two
sets of information, we require crab entrainment rates
(e), measured as numbers of crab entrained per unit
volume dredged. Total entrainment (E) is
El so- = D
I s ' 6g " V 1 sg
(2)
Postentrainment mortality (m), expressed as a pro-
portion of those entrained, varies with gear type, age,
and season. Age-specific loss (L) of crab in a single
season, location, and gear combination will be
■'-'al sg ~ •'^ 1 sg ' •''^al s ' tl^asg •
(3)
To compare the relative importance of losses from dif-
ferent age-classes, equivalent adult loss (E AL) for any
season-location-gear combination is calculated as
EAL
Isg
^ '-' al sg ' "as I
(4)
Data and estimation
Population abundance Crab population surveys
were conducted over a six-year period (1983-88) in
Grays Harbor and along the adjacent coast. Stratified
random sampling was done with a small beam trawl
at biweekly or monthly intervals during spring and
summer (May-September) with occasional sampling
during fall and winter. From these surveys, crab den-
sities were estimated for each stratum, and total
population estimates were computed separately for
Grays Harbor and the adjacent coast using the National
Marine Fisheries Service BIOMASS program (Alaska
Fish. Sci. Cent., 7600 Sand Point Way NE, Seattle,
WA 98115), which uses standard stratified random
survey statistical methods (Cochran 1962). Details of
the survey design and population estimates can be
found in Armstrong and Gunderson (1985) and Gunder-
son et al. (1990). In addition to the trawl surveys,
intertidal crab were sampled in 0.25 m^ quadrats at
several locations within the harbor, and total intertidal
population was estimated as described by Dumbauld
and Armstrong (1987).
Growth and age-classes In general, age-class iden-
tification is difficult in crustaceans (Hartnoll 1982). The
lack of retained hard parts prohibits direct aging
techniques (such as scale analysis in fish), so age must
be estimated from size. We relied on visual separations
of age-classes in size-frequency plots from the popula-
tion surveys, but molting and individual variability in
growth obscure age-class modes except for young,
rapidly growing crab. In all cases, young-of-the-year
(age 0 -I- ) crab were easily identifiable as a separate size-
group. Age 1 -I- size distributions sometimes overlapped
older ages; in these cases, visual estimates of the sep-
aration point were supplemented by projecting growth
from earlier observations. No reasonable separations
could be made for older ages. For this reason, our
analysis uses three age-classes: O-i-, l + , and >l + .
Within Grays Harbor, we believe that most crab leave
the estuary before their third year, so that almost all
crab within the estuary identified as > 1 -i- are actually
age 2 + , and this assumption is made in our analysis.
Proportions in each age-class were then calculated from
the total size-frequency distribution of each sampling
stratum.
where Sag is the total natural survival to adulthood
from age-class i in season k (assumed equal in all
habitats). Total loss for the project is then
EAL tot = Z EALug.
Ug
(5)
Definition of model seasons Seasons were defined
to reflect important biological processes and major
changes in crab abundance through the year. The
spring season (April and May) reflects the start of
settlement of the 0 + age-class and a period of migra-
tion into the estuary by age 1 + coastal crab; summer
(June-September) is a period of continued settiement.
Wainwright et al. Effects of dredging on a crab population
175
rapid growth, and steady mortality for 0 + crab and
relative stability for older age-classes. Fall (October-
December) and winter (January-March) are periods for
which we have little sampling data, but both are periods
of general population decline, migration from intertidal
to subtidal areas within the estuary by 0 + crab, and
emigration from the estuary by older age-classes.
Where data were lacking during fall and vdnter, values
were projected from late-summer populations accord-
ing to the trends in numbers observed in years for
which winter data were available.
Definition of geograplnic strata The population
survey design had four strata within Grays Harbor:
Outer Harbor, North Bay, South Bay, and Inner Har-
bor (Fig. 1). The navigation channel passes through two
of these (Inner and Outer Harbor), and crab densities
within various reaches of the channel were assumed
to be the average densities for the corresponding
sampling strata. In fact, crab densities estimated within
the channel during entrainment studies are quite com-
parable with those estimated from the corresponding
strata of the regular surveys (Dinnel et al. 1986, Dum-
bauld et al. 1988, Wainwright et al. 1990). Thus calcula-
tions for Bar, Entrance, and South Reaches used crab
densities for the Outer Harbor, while Inner Harbor
values were used from Crossover Reach to Aberdeen
Reach. Crab densities decline upriver, and South Aber-
deen Reach was assumed to have no crab.
Mortality Mortality estimates were calculated by
regressing logarithm of population abundance on age.
This method was applied separately for early juveniles
(age 0 + ) and for older juveniles and adults (age 1 -i- and
older). Because substantial migration of 0 -i- crab to or
from the estuary does not occur, mortality rates spe-
cific to Grays Harbor could be calculated for this age-
group. To estimate mortality, total estuarine 0 -i- and
1 -I- populations were calculated from the six years of
trawl survey data. Estimates for 0+ subtidal popula-
tions were supplemented with intertidal estimates to
provide a complete representation of the estuarine
population. Direct calculation of mortality requires
analysis of a population with no recruitment or migra-
tion. Settlement had essentially ended by July of each
year, so we chose July of the 0 + year as the starting
point for calculations. During the 1 + year, migration
begins near the end of the summer as crab leave the
estuary. Because of this, we chose June of the 1 -i- year
as the endpoint for estimating first-year survival. First-
year mortality estimates were calculated for each of
five cohorts (1983-87 year-classes).
Estimation of mortality for older ages is more dif-
ficult for two reasons: age-class separation is difficult
and inacctu*ate, and migration to and from the estuary
occurs. Because of these problems, a different approach
was used. To reduce problems of migration, population
estimates for the estuary and adjacent coast were
combined. Age-class separations were made using an
instar analysis technique (Armstrong et al. 1987, Oren-
sanz and Gallucci 1988) to identify instar composition
of the population. Instar abundances were then as-
signed to year-classes. To reduce errors from sampling
and age-class identification, monthly abundance esti-
mates were averaged over all year-classes, then aver-
aged over months within each survey season to give
a single estimate for each age-class (a):
Na = mean(Namv).
(6)
where Na^y is the abundance estimate for age a in
month m of sample year y. Then survival from age a
to a -I- 1 was calculated as
N,
a+l
-"a, a+1
N,
(7)
Because a single strong year-class biases estimates
calculated in this way, the very strong 1984 year-class
was excluded. The calculated age-specific natural mor-
tality rates were then combined to produce the survival
schedule (Sas in Eq. 4) used to calculate equivalent
adult loss from unadjusted loss.
Estimating entrainment rate Numerous studies
have been conducted to estimate the rate of entrain-
ment of crab by various kinds of dredges, and the
subsequent damage and mortality to entrained crab
(McGraw et al. 1988). Entrainment and subsequent
mortality are discussed separately below.
A regression relationship was used to predict the en-
trainment rate (crab entrained/key dredged; e in Eq.
2) from trawl-based density estimates (crab/ha). This
approach was used by Armstrong et al. (1987) and
McGraw et al. (1988) to estimate entrainment rates for
a hopper dredge. More data have been collected since
those studies, so a new relationship has been calculated.
Sampling during the entrainment surveys consisted of
two parts: sampling of the dredged material stream
aboard a hopper dredge, and concurrent trawl surveys
within the channel section being dredged. During each
survey, sampling occurred over a two- to three-day
period and covered several stations wathin the naviga-
tion channel. For each survey, mean entrainment (crab
per key dredged) and mean density (crab per ha) were
calculated over all samples within each station. This
provided a total of 14 points which were used to
calculate the regression. Details of survey methods are
given in McGraw et al. (1988).
176
Fishery Bulletin 90(1). 1992
800
Irowl colch (cfob/lvj)
Figure 4
Relationship between trawl catch and entrainment of Dunge-
ness crab by a hopper dredge. The line was fit by least-squares
and non-parametric regression. Arrow indicates two outliers
which were excluded from the least-squares regression.
To relate crab entrainment to crab density, several
regression models were tried. The selection of a final
model was based on both statistical measures of fit and
biological reasonableness (i.e., an expectation that en-
trainment should increase with increasing crab den-
sity). First, a test for linearity ("XLOF" in the Minitab
package; Minitab Inc., University Park, PA) was per-
formed, and no significant nonlinearity was detected
(p>0.10). Second, a linear least-squares regression was
calculated; neither the slope nor the intercept were
significantly different from zero for this model. How-
ever, this relationship was heavily influenced ("Cook's
Distance Measure"; Weisberg 1985) by two points.
When these two points were excluded, the best least-
squares model was (Fig. 4)
Y = 0.27X,
(8)
where Y is entrainment by the dredge (crab/key), and
X is trawl-estimated density (crab/ha). Finally, a non-
parametric median-slope regression (Conover 1980)
was calculated using all 1 4 data points. This method
returned the same slope as the 12-point least-squares
regression.
Entrainment for the other dredge types was calcu-
lated from this model based on relative entrainment
factors given by Stevens (1981); entrainment by a
pipeline dredge is assumed to be 100% of the hopper
dredge value (this value is controversial, but is conser-
vative), while a clamshell dredge entrains only about
5% of the hopper dredge value.
Table 2
Postentrainment mortality rates for Dungeness
crab by age.
season, and dredge type.
Dredge
Age-
Size range
Mortality
type
class
Season
(mm)
(%)
Hopper
0 +
Apr-May
7-10
5
Jun-Sep
11-30
10
Oct-Dec
31-40
20
Jan-Mar
41-50
40
1-H
Apr-Sep
51-75
60
Oct-Mar
>75
86
>l-^
All
>75
86
Clamshell
All
All
All
10
Pipeline
All
All
All
100
Postentrainment mortality After entrainment, crab
may be killed due to physical trauma during transport
through pipes and pumps, burial under excessive sedi-
ment weight, or confined disposal in landfill by a
pipeline dredge. Several estimates of postentrainment
mortality (m in Eq. 3) have been made. For a hopper
dredge, Stevens (1981) reported approximately 75%
mortality, all sizes of crab combined. Armstrong et al.
(1982) reported mortality rates by crab size for a hop-
per dredge, with 86% mortality for crab larger than
50 mm carapace width (CW) and 46% mortality for
those smaller than 50 mm CW. Other studies indicate
that hopper dredge mortality rates for small (< 10 mm)
0-1- age-class crab range from 1% to 5% (K. Larson,
Portland Dist., U.S. Army Corps of Eng., pers. com-
mun., 1987). Gross mortality observations were also
made during later entrainment studies (McGraw et al.
1988, Wainwright et al. 1990), but these recorded only
obvious mutilations and so underestimate total mortal-
ity. We adopted a set of size-dependent mortality rates
for a hopper dredge based on these studies (Table 2).
Little information is available concerning mortality
of crab entrained by a clamshell dredge. Stevens (1981)
reported an overall mortality rate of less than 10%,
which seems reasonable considering the operation of
the gear. We have used a 10% mortality rate for a clam-
shell dredge for all age-classes. Because its effluent
goes to confined upland disposal, 100% mortality was
assumed for all crab entrained by the pipeline dredge.
Simulations Scheduling of dredge operations was
based on engineering constraints, weather limitations,
avoidance of salmon migration periods, and avoidance
of seasons and areas with high predicted crab loss. To
help in this planning process, loss rates (expressed as
crab per volume dredged) were calculated for each area
and each season, based on average seasonal crab den-
sities and age-class composition.
Wainwright et al : Effects of dredging on a crab population
177
Table 3
Hypothetical project scenarios
for Grays Harbor
WA, show-
ing volume
to be
dredged by each dredge type in each area |
and season
Harbor section
Season
Dredge Volume (key)
Scenario 1:
Full confined disposal
Outer
Jan-Mar
Hopper
1698
Outer
Apr- May
Hopper
1132
Outer
Apr-May
Hopper
330
Outer
Jun-Sep
Hopper
2800
Inner
Jun-Sep
Hopper
1000
Inner
Jun-Sep
Pipeline
434
Inner
Oct-Dec
Hopper
2036
Inner
Oct-Dec
PipeHne
2224
Inner
Jan-Mar
Hopper
1714
Inner
Jan-Mar
Pipeline
Total
670
14,038
Scenario 2:
Limited confined
disposal
Outer
Apr-May
Hopper
1462
Outer
Jun-Sep
Hopper
2800
Outer
Jan-Mar
Hopper
1698
Inner
Apr-May
Clamshell
771
Inner
Jun-Sep
Hopper
1000
Inner
Jun-Sep
Clamshell
579
Inner
Oct-Dec
Hopper
2036
Inner
Oct-Dec
Clamshell
778
Inner
Oct-Dec
Pipeline
374
Inner
Jan-Mar
Hopper
1714
Inner
Jan-Mar
Clamshell
Total
826
14,038
2000
Ouler Hotbor
ssrw ssfw ssfw ssfw ssfw ssrw ssrw
1983 1981 1985 1986 1987 1988 Meon
A 1 5000
1500
1000
I
,Ofl=.^,D,g,...il,lL
.BI.
ssrw ssrw ssrw ssrw ssrw ssrw ssrw
1983 1984 1985 1986 1987 1988 Meon
Figure 5
Seasonal abundance (catch per hectare) of Dungeness crab
in the Outer and Inner Harbor strata of Grays Harbor, Wash-
ington, by age-class. Solid bars, age 0-r; white, age 1 + ;
hatching, age > 1 -r .
Once project scheduling was determined, predictions
of total crab loss were needed, which we calculated by
simulating entrainment for planned construction sce-
narios. The scenarios we have used for calculating crab
losses reflect the project as planned in 1987 (Table 3).
There was some conflict between project costs and crab
protection, particularly regarding the tradeoff between
using gear that is economically efficient (hopper and
pipeline) and that which minimizes loss (clamshell).
Throughout most of the estuary, the efficiency of the
hopper dredge makes alternatives uneconomic. In cer-
tain areas of the Inner Harbor, the pipeline dredge is
economically most efficient but results in high post-
entrainment mortality. The alternative dredge in those
areas is a clamshell, which is generally more costly. To
better evaluate this tradeoff, two scenarios are con-
trasted. Scenario 1 includes full use of a pipeline dredge
where it is most effective; in Scenario 2, a clamshell
dredge is substituted where feasible. Table 3 shows
volumes dredged under each scenario by gear type,
location, and season.
As initially planned, construction was to occur over
two calendar years, extending through seven seasons.
To simplify calculations, we compressed the project into
a single model year (from spring of a given calendar
year through winter of the next), and calculated en-
trainment and losses for each scenario separately based
on each of the six years of survey-based crab abundance
estimates. This produced a set of 12 (six years by two
construction scenarios) model runs.
Because the project was revised in several ways since
these calculations were made, results presented here
do not reflect actual expected losses resulting from the
project, and are presented only to illustrate the method.
Results
Population parameters
Age-class abundance Densities of crab in the Inner
and Outer Harbor strata varied considerably among
years and seasons (Fig. 5). Average seasonal total den-
sity ranged from 73 ha-^ to 13,000 ha ^ Age O-i- crab
were most abundant in 1984, and were usually more
abundant in the Inner Harbor. Older crab were more
178
Fishery Bulletin 90(1). 1992
Table 4
Estimates of instantaneous mortality (Z) |
and annual
survival (S) for
age 0 +
Dungeness
;;rab, Grays Harbor
WA.
Year-class
Z(yr-')
S(%)
83
3.4
3.4
84
2.2
11.5
85
3.5
3.0
86
1.6
19.8
87
1.9
15.2
Average
2.5
8.1
Table 5
Estimates of instantaneous mortality (Z)
and annual survival (S) for older age-
classes of Dungeness crab, Grays Harbor,
WA, and adjacent coast combined. Esti-
mates are for July-July, average for
several years.
Table 6
Survival schedule: percent of Dungeness crab surviving from
each season to midwinter (15 Feb.) of the 2 -i- year.
Season
Midpoint
Age-class
0-1-
1-1-
2-^
Apr-May
Jun-Sep
Oct-Dec
Jan-Mar
30 Apr
31 Jul
15 Nov
15 Feb
0.87
1.65
3.40
6.35
10.7
16.0
25.5
38.0
53.2
64.9
81.9
100.0
abundant in the Outer Harbor, where they reached
peak densities in the summer season.
Mortality Estimated instantaneous mortahty rates
for age 0+ crab within Grays Harbor ranged from 1.6
to 3.5 yr- ', with a mean of 2.5 yr" ', corresponding to
an annual survival of 8.1% (Table 4). For older crab,
estimated mortality rates (Eq. 7) decreased to age 3 -t- ,
then increased slightly between ages 3 -i- and 4 -i- (Table
5). These two results were combined to derive the
seasonal survival schedule (Table 6) used in the model.
Gear and season comparisons
The results of gear/season comparison simulations are
presented in Figures 6 and 7. These data show the
300
. 200
100
Unodjusted Loss
tquivoleni Adull Loss
n
p
C H P
Apr-Moy
CUP
Jun-S€p
C H P
Ocl-Dcc
CUP
Jon- Mar
Figure 6
Entrainment rates of Dungeness crab by season and dredge
type for the Outer Harbor, by age-class, (upper) Unadjusted
losses; (lower) age 2 + equivalent losses. Dredge types: C =
clamshell, H = hopper, P = pipeline. Age-classes as in Fig-
ure 5.
strong contrast between the pipeline and clamshell
dredges: the clamshell dredge has negligible impact.
Comparing the unadjusted losses (Eq. 3) with age 2 -i-
equivalent losses (Eq. 4) shows the relative unimpor-
tance of 0 -(- crab. Also notable are the high age 2 -i-
equivalent losses in the Outer Harbor during summer
and fall, when there are concentrations of age 1 + and
older crab in this area (Fig. 5).
Impact estimates
Calculations of total age 2-i- equivalent loss (Eq. 5)
for the two project scenarios are shown in Figure 8.
As expected, Scenario 1 (full use of the pipeline
dredge with confined disposal) shows higher losses
than Scenario 2. For both scenarios, a large part of
the total loss occurs during the June-September sea-
son, due to large volumes being dredged in the Outer
Harbor where older crab are concentrated at this time.
The results indicate strong year-to-year variation in
Wainwright et al.: Effects of dredging on a crab population
179
30O T
15 T
Unodjuslcd Loss
.DM^-
n n
M. DM.
C H P C H P C H P
Jun-Scp Ocl-Dcc Jon-Wot
fl
tquivalenl Adull Loss
CHP Clip CMP CHP
Apr-May Jun-Sep Ocl-Oec Jon-
Uat
Figure 7
Entrainment rates of Dungeness crab by season and dredge
type for the Inner Harbor, as in Figure 6.
SIX)
|600-
1
|400
f-200
0
^^M Scenario 1
miifmi
1983 1984
1985 1986 1987 1988 Mean
800 T
" 600
400
5-200
Scenorio 2
I
L3 'i
1983 1984 1985 1986
987 1988 Meon
Figure 8
Total estimated age 2+ equivalent losses for hypothetical
dredging in six years for two project scenarios, (upper) Full
use of confined disposal; (lower) limited confined disposal. Age-
classes as in Figure 5.
impacts, with 1983 construction resulting in impacts
nearly three times the average for the other years. This
is apparently because 1983 followed two years of
strong settlement, as evidenced by the high abundance
of both age 1 + and > 1 -i- crab in that year (Fig. 5; see
also Gunderson et al. 1990). This emphasizes the im-
portance of population monitoring during construction
to accurately assess impacts.
Discussion
Gear and season comparisons made with DIM provided
several results which were subsequently used to sched-
ule construction gear, season, and location combina-
tions so as to reduce crab losses. As expected, the clam-
shell dredge (which moves slowly and does little
mechanical damage to organisms) had insignificant im-
pact in all seasons and areas. Comparing pipeline and
hopper dredge effects, our initial impression was that,
with confined disposal (resulting in 100% loss of all age-
classes), the pipeline dredge would cause extremely
high losses relative to the hopper dredge. This is true
when one considers the imadjusted losses (Figs. 6A and
7A). However, when viewed on an equivalent adult loss
basis (Figs. 6B and 7B), the pipeline dredge loss rate
is only 10-50% higher than that of the hopper dredge.
The equivalent adult loss viewpoint was also important
in seasonal comparisons, especially in the Inner Har-
bor (Fig. 7) where unadjusted loss was highest in
spring, but equivalent adult loss peaked in fall.
During any modeling endeavor in applied ecology,
certain decisions must be made to limit the scope and
applicability of the model. Many decisions are made
simply on the basis of information or time available,
while others reflect the biases and experiences of the
authors. One of the major decisions in this project was
the choice between predicting short-term losses via the
equivalent adult loss approach, or accounting for poten-
tial longer-term losses due to reduction of the local
reproductive stock via "production foregone" (Rago
1984) techniques. For local, short-term entrainment to
have longer-term population effects requires a strong
influence of current stock size on future recruitment.
180
Fishery Bulletin 90(1). 1992
For Dungeness crab, there is little evidence of stock-
dependence. In fact, it is not clear whether a local stock,
such as that in Grays Harbor, is self-reproducing or
depends on larval drift from other areas. For this rea-
son, we chose to use only short-term loss predictions.
The choice of slope for the regression of crab entrain-
ment on trawl catch will strongly influence model
results. We gave long consideration to the choice of
regression models. Problems arise because there are
few data points and large measurement errors asso-
ciated with both variables. Costs of sampling (which
involved simultaneous operations of a specially modi-
fied hopper dredge and a chartered trawler) prohibited
any increase in data quantity or precision. Initially, we
chose to use least-squares regression (LSR) with its
underlying assumptions of normal errors with equal
variances. There are two forms of LSR in common use:
predictive LSR which assumes that all error is in mea-
surement of the Y (dependent) variate, and functional
LSR which incorporates errors in both X and Y vari-
ates. In the overall context of DIM, the entrainment
regression serves the role of a calibration curve predic-
ting entrainment from a set of observed trawl catches.
For this reason, we used a predictive regression con-
ditional on the observed trawl catches. (This implies
that the result is not generalizable to any other method
of crab density estimation, but such generalization is
not needed here.) Two outliers were dropped from the
LSR analysis; both points were from the same station
in different years, and both were influenced by one or
two extremely high trawl catches. Because we were
not entirely satisfied with the assumptions of the LSR
analysis, the data was reanalyzed using a nonpara-
metric regression technique which is robust to non-
normality, inequality of variances, and errors in mea-
surement of the X variate. Because this analysis agreed
with the final LSR model (Eq. 8), we accepted that
model as the most reasonable.
Another limitation was our inability to reliably dis-
tinguish age-classes beyond 1 -i- and obtain mortality
estimates for older age-groups. Because of this, we
stopped our calculations at age 2 -i- , but there is a
strong desire to relate the results to fishery stocks with
recruitment at 3-5 years of age. It is possible to per-
form some rough calculations of actual impact to
fisheries, if we are willing to make some assumptions.
Using Scenario 2 (limited confined disposal) as an ex-
ample, estimated age 2 -i- equivalent losses ranged from
166 to 587 thousand crab (Fig. 8). The fishery harvests
males only, so with a 50% sex ratio these numbers
become 83-298 thousand age 2+ male crab lost. To
relate these to the fishery, we need to know survival
from age 2 -i- to recruitment. We have rough estimates
of mortality from age 2 -i- to 3 h- and from age 3 -i- to
4 -H (Table 6) calculated from the trawl survey data set.
These estimates are confounded with the decline in
gear efficiency with crab size, and so are probably
underestimates of true survival. They also depend on
tenuous assumptions about size-at-age. Accepting these
estimates and assuming the bulk of the fishery recruits
at age 3 -i- , our estimates of age 2 -i- loss correspond to
losses to the fishery of 37-134 thousand age 3-i- male
crab. As exploitation rates are quite high (~70-90%;
Methot and Botsford 1982), these numbers can be
related directly to annual catch. The ten-year average
catch for the Washington coast has been about 3000
metric tons, which corresponds to 3.3 million crab
(average individual weight of 0.9kg). So, losses for this
hypothetical scenario would be on the order of 1-4%
of the average annual catch by the Washington coast
fishery.
The model was limited by several other factors, par-
ticularly problems of data quality and parameter esti-
mates. Primary among these was lack of data on beam
trawl efficiency and size selectivity (Gunderson and
Ellis 1986). We have implicitly assumed that the trawl
sampling was 100% efficient for all sizes of crab, which
is certainly not the case. The gear was designed for cap-
turing juvenile crab, and we believe it to be relatively
efficient for juvenile sizes, but crab approaching legal
size are able to avoid or escape the small net. For
estimating absolute numbers entrained, this is not a
problem because the entrainment function is essentially
a calibration of entrainment against trawl catch, re-
gardless of trawl efficiency. However, to the extent
that gear efficiency is below 100%, we underestimate
total populations within the estuary. Calculations of en-
trainment as a proportion of the local population are
thus biased upward. Trawl efficiency also affects
natural mortality rate estimates, to which equivalent
adult loss calculations are extremely sensitive.
Overall, DIM has proved useful even with its limita-
tions. In project planning, the model allowed schedul-
ing gear and work seasons to reduce impacts on the
crab population, and provided some quantitative predic-
tions of loss on which to base mitigation programs. DIM
is now being used in conjunction with crab survey data
gathered during construction to estimate actual crab
losses and to fully define levels and type of mitigation.
Beyond these intended uses, the model served to focus
concerns about crab impacts, which tended to be some-
what ill-defined, onto specific questions of data qual-
ity and reliability of predictions, providing all sides a
common basis for argument.
Ackno\A/ledgments
This work was done under a combination of support
from the Seattle District, U.S. Army Corps of Engi-
Wainwright et al,: Effects of dredging on a crab population
neers (#DACW67-85-C-0033), Battelle Pacific North-
west Laboratories, Sequim, Washington, and Wash-
ington Sea Grant (#NA86AA-D-SG044 Project R/F-68).
We appreciate the contributions of the Corps staff,
notably Fred Weinmann, Gail Arnold, James Waller,
and Ann Uhrich, and the valuable comments and dis-
cussion of the Grays Harbor Crab Study Panel con-
vened by Walter Pearson. Louis Botsford provided
invaluable suggestions for the initial model design.
Loveday Conquest provided analysis of the entrain-
ment regression and other statistical advice. The pro-
ject could not have been completed without the coop-
eration of the crews of the COE dredge YAQUINA and
the fishing vessel Karelia (Vern Heikkila at the helm).
We also thank those (too numerous to name) who con-
tributed greatly to field work, data analysis, and ad-
ministrative support.
Citations
Armstrong, D.A., and D.R. Gunderson
1985 The role of estuaries in Dungeness crab early life history:
A case study in Grays Harbor, Washington. In Melteff, B.R.
(ed.), Proceedings, Symposium on Dungeness crab biology and
management, p. 145-170. Alaska Sea Grant Rep. 85-3, Univ.
Alaska, Fairbanks.
Armstrong, D.A., B.G. Stevens, and J.C. Hoeman
1982 Distribution and abundance of Dungeness crab and
Crangon shrimp, and dredging-related mortality of inverte-
brates and fish in Grays Harbor, Washington. Tech. Rep. to
Wash. Dep. Fish, and U.S. Army Corps of Eng., Seattle Dist.
by School of Fisheries, Univ. Wash., Seattle, 349 p.
Armstrong, D.A., T.C. Wainwright, J. Orensanz, P. A. Dinnel,
and B.R. Dumbauld
1987 Model of dredging impact on Dungeness crab in Grays
Harbor, Washington. Fish. Res. Inst. Rep. FRI-UW-8702,
School of Fisheries, Univ. Wash., Seattle, 167 p.
Bella, D.A.. and K.J. Williamson
1980 Diagnosis of chronic impacts of estuarine dredging. J.
Environ. Syst. 9:289-311.
Boreman, J., C.P. Goodyear, and S.W. Christensen
1981 An empirical methodology for estimating entrainment
losses at power plants sited on estuaries. Trans. Am. Fish.
Soc, 110:253-260.
Botsford, L.W., D.A. Armstrong, and J.M. Shenker
1989 Oceanographic influences on the dynamics of commer-
cially fished populations. In Landry, M.R., and B.M. Hickey
(eds.), Coastal oceanography of Washington and Oregon, p.
511-565. Elsevier, Amsterdam.
Butler, T.H.
1960 Maturity and breeding of the Pacific edible crab, Cancer
magister Dana. J. Fish. Res. Board Can. 17:641-646.
1961 Growth and age determination of the Pacific edible crab
Canter magister Dana. J. Fish. Res. Board Can. 18:873-890.
Cleaver, F.C.
1949 Preliminary results of the coastal crab (Cancer magister)
investigation. Wash. Dep. Fish. Biol. Rep. 49A:47-82.
Cochran, W.G.
1962 Sampling techniques, 2d ed. Wiley, NY, 413 p.
Conover, W.J.
1980 Practical nonparametric statistics, 2d ed. Wiley, NY,
462 p.
Dinnel, P.A., D.A. Armstrong, B.R. Dumbauld, and
T.C. Wainwright
1986 Impact of dredging on Dungeness crab. Cancer magister,
in Grays Harbor, Washington during August 1986. Fish. Res.
Inst. Rep. FRI-UW-8611, School of Fisheries, Univ. Wash.,
Seattle, 34 p.
Dumbauld, B.R., and D.A. Armstrong
1987 Potential mitigation of juvenile Dungeness crab loss dur-
ing dredging through enhancement of intertidal shell habitat
in Grays Harbor, Washington. Fish. Res. Inst. Rep. FRI-
UW-8714, School of Fisheries, Univ. Wash., Seattle, 64 p.
Dumbauld, B., D. Armstrong, P. Dinnel. and T. Wainwright
1988 Impact of dredging on Dungeness crab. Cancer magister.
in Grays Harbor, Washington during August 1987. Fish. Res.
Inst. Rep. FRI-UW-8820, School of Fisheries, Univ. Wash.,
Seattle, 25 p.
Goodyear, C.P.
1977 Mathematical methods to evaluate entrainment of aquatic
organisms by power plants. Rep. FWS/OBS-76/20.3, U.S.
Fish Wildl. Serv., Wash., DC.
Gunderson, D.R., and I.E. Ellis
1986 Development of a plumb staff beam trawl for sampling
demersal fauna. Fish. Res. (Amst.) 4:35-41.
Gunderson. D.R., D.A. Armstrong, Y.-B. Shi. and
R.A. McConnaughey
1990 Patterns of estuarine use by juvenile English sole (Paro-
phrys vetulus) and Dungeness crab {Cancer magister). Estu-
aries 13:59-71.
Hankin, D.G., N. Diamond, M.S. Mohr, and J. lanelli
1989 Growth and reproductive dynamics of adult female
Dungeness crabs (Cancer magister) in northern California. J.
Cons. Cons. Int. Explor. Mer 46:94-108.
Hartnoll, R.G.
1982 Growth. In Abele, L.G. (ed.). Embryology, morphology,
and genetics (The Biology of Crustacea, vol. 2), p. 111-196.
Academic Press, NY.
Horst, T.J.
1975 The assessment of impact due to entrainment of ichthyo-
plankton. In Saila, S.B. (ed.), Fisheries and energy produc-
tion: A symposium, p. 107-118. Le.xington Books, Lexington,
MA.
Jamieson. G.S., and D.A. Armstrong
1991 Spatial and temporal recruitment patterns of Dungeness
crab in the northeast Pacific. Mem. Queensl. Mus. 31:365-381.
MacCall, A.D., K.R. Parker, R. Leithiser, and B. Jessee
1982 Power plant impact assessment: A simple fishery pro-
duction model approach. Fish. Bull.. U.S. 81:613-619.
McGraw, K.A., L.L. Conquest. J.O. Waller. P.A. Dinnel, and
D.A. Armstrong
1988 Entrainment of Dungeness crabs, Cancer magister Dana,
by hopper dredge in Grays Harbor, Washington. J. Shellfish
Res. 7(2):219-231.
Methot. R.D. Jr., and L.W. Botsford
1982 Estimated preseason abundance in the California Dunge-
ness crab (Cancer magister) fisheries. Can. J. Fish. Aquat.
Sci. 39:1077-1083.
Morton. J.W.
1977 Ecological effects of dredging and dredge spoil disposal:
A literature review. Tech. Pap. U.S. Fish Wildl. Serv. 94,
Wash., DC, 33 p.
Orensanz. J.M.. and V.F. Gallucci
1988 A comparative study of postlarval life history schedules
in four sympatric Cancer species. J. Crust. Biol. 8:187-220.
182
Fishery Bulletin 90(1). 1992
Pacific Marine Fisheries Commission
1989 Data series: Crab and shrimp section (Rev. 1989). Pac.
Mar. Fish. Comm., Portland.
Poiner, I.R., and R. Kennedy
1984 Complex patterns of change in the macrobenthos of a
large sandbank following dredging. Mar. Biol. 78:335-352.
Rago, P.J.
1984 Production forgone: An alternative method for assess-
ing the consequences of fish entrainment and impingement
losses at power plants and other water intakes. Ecol. Modell.
24:79-111.
Simenstad, C.A. (editor)
1990 Effects of dredging on anadromous Pacific Coast fishes:
Workshop proceedings, Seattle, September 8-9, 1988. Wash.
Sea Grant Prog., Univ. Wash., Seattle, 160 p.
Stevens, E.G.
1981 Dredging-related mortality of Dungeness crabs associated
with four dredges operating in Grays Harbor, Washington.
Seattle Dist., U.S. Army Corps of Eng. DA-79-45, Seattle,
148 p.
Stevens, B.G., D.A. Armstrong, and J.C. Hoeman
1984 Diel activity of an estuarine population of Dungeness
crabs, Cancer magister, in relation to feeding and environmen-
tal factors. J. Crust. Biol. 4:390-403.
van Winkle, W. (editor)
1977 Proceedings of the conference on assessing the effects
of power-plant-induced mortality in fish populations, Gatlin-
burg, Tennessee, May 3-6, 1977. Pergamon Press, NY, 380 p.
Wainwright, T.C., K.A. McGraw, D.A. Armstrong,
B.R. Dumbauld, and L.L. Conquest
1990 Impact of dredging on Dungeness crab. Cancer magister,
in Grays Harbor, Washington, during August 1989. Fish. Res.
Inst. Rep. UW-FRI-9005, Univ. Wash., Seattle, 18 p.
Washington Department of Fisheries
1951 Washington commercial fisheries 1951 statistics. Wash.
Dep. Fish., Seattle, 46 p.
Weisberg, S.
1985 Applied linear regression, 2d ed. Wiley, NfY, 324 p.
Comparison of feeding and growtfi
of iarval round Fierring Etrumeus teres
and guif meniiaden Brevoortia patronus
Weihzong Chen
East China Sea Fisheries Research Institute
300 Jun Gong Road, Shanghai, Peoples Republic of China
John J. Govoni
Stanley M. Warlen
Beaufort Laboratory, Southeast Fisheries Science Center
National Marine Fisheries Service, NOAA, Beaufort, North Carolina 285 1 6
The round herring Etrumeus teres
is one of several clupeid fishes,
abundant in continental shelf
waters of the Gulf of Mexico, that
presently is not commercially ex-
ploited by the United States, al-
though its sibling species E. white-
headi is a fishery resource for South
Africa (Roel and Melo 1990). The
potential annual yield of this latent
resource is estimated as 3.3 x lO"* to
4.2 xlO^ metric tons for the east-
em Gulf (Houde 1977) and 1.1 x 10^
to 1.1 X 10'' metric tons for the en-
tire GuJf (Reintjes 1980). Details
relevant to the distribution and
population dynamics of round her-
ring, including elements of its early
life history, are presently sketchy.
Houde (1977) reported that round
herring in the eastern Gulf of Mex-
ico spawn from mid-October to the
end of May between the 30 and 200
m isobaths. He surmised that there
is a major spawning area about 150
km west-southwest of Tampa Bay,
Florida, and a minor area just north
of the Dry Tortugas. Off Texas and
Louisiana, spawning occurs from 50
to 200 km offshore and may extend
to the edge of the continental shelf
(Fore 1971). Round herring and
another clupeid, the gulf menhaden
Brevoortia ipatronus, are sympatric;
the latter spawns in inshore waters
of the northern Gulf at least as far
offshore as 130 km with a focus of
spawning off Mississippi between
mid-October and late March (Christ-
mas and Waller 1975).
Differences between adult round
herring and gulf menhaden are so
obvious that systematists once re-
ferred these two species to separate
families, Dussumieriidae and Glu-
peidae (Whitehead 1963), but their
larvae are morphologically similar
with one major exception, their jaw
structure. (The misperception that,
unlike other clupeids, round herring
larvae do not possess a swimblad-
der (Fahay 1983) has been perpetu-
ated in the literature.) At hatching
larvae of both species are about 3.0
mm notochord length (NL) and are
slender and elongate with a straight
alimentary canal and a posterior
anus (Houde and Fore 1973). Trans-
formation to the juvenile form
begins at about 18 mm standard
length (SL) (Houde and Fore 1973).
Round herring larvae develop teeth
on their long, spatulate upper and
lower jaws at about 6 mm SL
(Houde and Fore 1973); but gulf
menhaden do not develop teeth on
their shorter, less compressed jaws
until they are about 10 mm SL (Hef-
tier 1984).
The diets of the larvae of these
species might reflect differences in
jaw structure and dentition. In ad-
dition, differences in diet quality
and quantity may register different
growth between these species.
While feeding and growth of gulf
menhaden larvae are docimiented
(Govoni et al. 1983, Stoecker and
Govoni 1984, Warlen 1988), similar
information on the early life history
of round herring is unavailable. In
this paper, we compare the feeding
and growth of larval round herring
and gulf menhaden.
Materials and methods
Round herring and gulf menhaden
larvae used in this study were re-
moved from ichthyoplankton collec-
tions obtained during two cruises
(December 1980 and February 1981)
using MOCNESS gear (multiple
opening/closing nets and environ-
mental sensing system). Three sta-
tions were occupied, one each at the
18, 91, and 183 m isobaths, along
three transects (off Cape San Bias,
FL; off the Mississippi Delta, LA;
and off Galveston Bay, TX (Sogard
et al. 1987)). The objective of the
sampling plan was to broadly can-
vass the continental shelf of the
northern Gulf for larval gulf men-
haden and two other species (So-
gard et al. 1987); larval round her-
ring were collected incidentally.
Sampling at three discrete depths
(surface, in the middle of the upper
mixed layer, and within or below
the thermocline) assured the collec-
tion of adequate numbers of speci-
mens. In addition, larvae from a
single collection taken at the Mis-
sissippi River plume front (Govoni
et al. 1989) in December 1982 were
examined to augment gut content
data. Larvae were preserved in 5%
formalin for food analysis and in
70% ethanol for growth studies
(Table 1). To provide an indication
of true dietary differences between
species encountering the same food
assortment, only those larvae from
a single vertically and horizontally
discrete collection (Govoni et al.
1986) that produced both species
were used for diet comparisons
Manuscript accepted 25 November 1991.
Fishery Bulletin, U.S. 90:183-189 (1992).
183
184
Fishery Bulletin 90(1). 1992
Table 1
Time, location, and depth of collection in
the northern Gulf of Mexico of round herring
Etrumeus teres larvae examined for diet composition and growth determination.
Water column
Sample
No. of
No. of
depth
depth
larvae
larvae
Date
Time
Transect
(m)
(m)
collected
with food
6 Dec 80'
0645
Mississippi Delta
183
1
2
0
8 Dec 80^
0013
Cape San Bias
183
1
8
8 Dec 80^
0013
Cape San Bias
183
51
2
8 Dec 80'
0040
Cape San Bias
183
50
6
0
8 Dec 80'
0056
Cape San Bias
183
1
10
0
8 Dec 80'
0621
Cape San Bias
183
50
3
0
8 Dec 80^
1800
Cape San Bias
183
1
20
8 Dec 80^
1800
Cape San Bias
183
102
1
8 Dec 80'
1821
Cape San Bias
183
49
3
0
8 Dec 80'
1834
Cape San Bias
183
1
20
1
9 Dec 80^
0600
Cape San Bias
91
1
7
9 Dec 80-
0600
Cape San Bias
91
35
1
9 Dec 80^
0600
Cape San Bias
91
74
3
9 Dec 80'
0617
Cape San Bias
91
35
13
0
9 Dec 80'
1237
Cape San Bias
91
37
2
0
9 Dec 80^
1800
Cape San Bias
91
1
19
0
9 Dec 80'
1806
Cape San Bias
91
75
3
0
9 Dec 80'
1817
Cape San Bias
91
35
3
0
9 Dec 80'
1827
Cape San Bias
91
1
20
4
10 Dec 80^
0005
Cape San Bias
91
1
19
10 Dec 80^
0005
Cape San Bias
91
35
1
0
10 Dec 80^
0010
Cape San Bias
91
12
4
0
10 Dec 80'
0025
Cape San Bias
91
74
6
0
10 Dec 80'
0025
Cape San Bias
91
1
1
0
10 Dec 80'
0035
Cape San Bias
91
46
16
0
10 Dec 80'
0045
Cape San Bias
91
1
11
0
10 Dec 80'
1812
Cape San Bias
18
6
2
1
10 Dec 80'
1821
Cape San Bias
18
1
17
2
11 Dec 80'
0019
Cape San Bias
18
12
11
1
11 Dec 80'
0031
Cape San Bias
18
1
3
0
12 Feb 81^
1900
Galveston Bay
18
6
2
13 Feb 81^
0600
Galveston Bay
91
35
5
14 Feb 81'
0033
Galveston Bay
91
61
13
1
14 Feb 81'
0047
Galveston Bay
91
35
20
1
14 Feb 81'
0100
Galveston Bay
91
1
12
0
14 Feb 81^
0600
Galveston Bay
91
37
13
14 Feb 81'
0620
Galveston Bay
91
77
8
0
14 Feb 81'
0635
Galveston Bay
91
27
20
0
14 Feb 81'
0644
Galveston Bay
91
1
20
0
14 Feb 81'
1823
Galveston Bay
91
74
20
1
14 Feb 81'
1846
Galveston Bay
91
1
20
7
18 Feb 81'
0014
Mississippi Delta
91
75
2
0
18 Feb 81'
0025
Mississippi Delta
91
1
20
8
18 Feb 81^
0600
Mississippi Delta
91
25
20
0
18 Feb 81'
0607
Mississippi Delta
91
25
20
2
18 Feb 81'
0625
Mississippi Delta
91
1
4
1
18 Feb 81-
1800
Mississippi Delta
91
40
8
10 Dec 82'
0830
Mississippi Delta
mination of diet.
18
1
88
26
'Collections for exa
^Collections for determination of growth
(Table 1). Adequate numbers of
larvae allowed growth compari-
sons of larvae collected on the
Cape San Bias transect in De-
cember 1980 and larvae collected
on the Mississippi Delta and Gal-
veston Bay transects in Febru-
ary 1981.
Larvae were measured to the
nearest 0.1mm (NL before and
SL after the formation of hypu-
ral plates). Guts were dissected
and all gut contents were ex-
cised, identified, and measured.
Percent similarity (Schoener
1970) was used to compare the
diets of larval round herring
and gulf menhaden from single
collections.
Sagittal otoliths were removed
from larvae, cleaned in distilled
water, and mounted on glass
microscope slides with clear acry-
lic resin; no griding or sectioning
was necessary to resolve daily
growth increments. Otoliths of
round herring were semi-opaque
and similar to those of gulf men-
haden. Presumed daily incre-
ments were clearly discernable
as bipartite structures consisting
of adjoining incremental and
discontinuous zones (Campana
and Neilson 1985).
In describing the growth of
larval round herring, we did not
experimentally verify that their
first otolith increment appeared
5 days after hatching or that sub-
sequent increments were added
daily as Warlen (1988) has done
for gulf menhaden. We assumed
that initial and subsequent incre-
ment deposition in round herring
was similar to gulf menhaden.
This assumption is justified, in
part, by similarities in the period
of some key developmental
events. Incubation takes 36
hours at 20.5°C for round her-
ring (O'Toole and King 1974),
and 40-42 hours at 19-20°C for
gulf menhaden (Hettler 1984).
Complete adsorption of the yolk
occurs in 4 days for round her-
NOTE Chen et al.: Feeding and growth of larval Etrumeus teres and Brevoortia patronus
185
ing reared in the laboratory at 24-26°C (Miller et al.
1979), as well as for gulf menhaden reared at 18-22°C
(Hettler 1984). Further, we used alternative empirical
methods to support our assumption that otolith growth
increment formation occurs daOy (Hales 1987). By com-
paring the width of marginal increments with the width
of the proximal completely-formed increment, we
determined the percentage of larvae with partially-
formed or completely-formed marginal increments over
a 24-hour period (8-10 December 1981; Table 1). The
frequency of increment formation was inferred from
these percentages and from the relationship of otolith
radius and larval length.
The Laird version of the Gompertz growth model was
used to describe growth from the logarithm, of length
and the estimated age of larvae (Zweifel and Lasker
1976). Growth curves of round herring and gulf men-
haden larvae were compared by using the predictive,
resampling method described by Kappenman (1981).
Data for gulf menhaden growth were taken from
Warlen (1988) for comparisons with the growth of
round herring.
Results
Distribution and co-occurrence
In all, 419 round herring larvae were identified in the
present collections, four fewer than gulf menhaden
(Sogard et al. 1987). Collections of the larvae of both
species indicate that they co-occur infrequently. Round
herring and gulf menhaden larvae occurred together
at 15 of 45 locations where collections produced either
species. Larval round herring were collected most fre-
quently throughout the water column at the offshore
stations in water 91m deep, although one of the largest
single collections was made at 18 m (Table 1). Larval
gulf menhaden were collected mostly inshore at the
18 m stations along each transect. The larvae of these
species co-occurred mainly at the 91m stations along
each transect.
Diet comparisons
Only 56 round herring larvae had food in their guts.
Larval round herring had eaten primarily copepod
nauplii, copepodites, and adults, with pteropods (mainly
Limacina trochiformis), tintinnids, invertebrate eggs,
and Eucalanus spp. nauplii contributing lesser percent-
ages (Table 2). Eucalanus nauplii were considered a
discrete food organism separate from other copepod
nauplii, because its form and size differed markedly;
Eucalanus spp. nauplii have long, paddle-like appen-
dages and are more than three times larger than the
other copepod nauplii observed in the guts of larvae.
Table 2
The diet composition of 56 round
herring Etrumeus teres
larvae in the northern Gulf of Mexico.
Percent
frequency
Percent
of occurrence
total no.
Centric diatoms
1.8
1.3
Tintinnids
3.6
5.2
Pteropods
8.9
6.5
Pelecypods
1.8
1.3
Unidentified copepod nauplii
25.0
36.4
Unidentified Eucalanus nauplius
3.6
3.9
Copepodid and adult copepods
21.4
16.9
Calanoid copepodites and adults
5.4
3.9
Harpacticoid copepodites and
1.8
1.3
adults
Cyclopoid copepodites and adults
16.1
11.7
Invertebrate eggs
10.7
11.7
Table 3
Comparison of the percent frequency
of occurrence of food
organisms in the diet of 26 larval round herring and gulf
menhaden larvae collected simultaneously in the northern Gulf
of Mexico.
Percent
frequency
of occurrence
Round
Gulf
Food organism
herring
menhaden
Tintinnid
3.8
4.2
Pteropods
19.2
2.1
Unidentified copepod nauplii
7.7
5.3
Unidentified Eucalanus nauplius
3.8
5.3
Unidentified copepodites and adult
15.4
31.6
copepods
Calanoid copepodites and adults
3.8
33.7
Harpacticoid copepodites and adults
3.8
0
Cyclopoid copepodites and adults
30.8
11.6
Invertebrate eggs
11.5
6.3
The width of food organisms ranged from 40 to 280 f^m,
a width range comparable to that found for gulf men-
haden (Govoni et al. 1983).
Of the 88 round herring larvae collected simultan-
eously with gulf menhaden, 26 had food in their guts.
There were differences in the gut contents of these 26
round herring and 26 randomly selected gulf menhaden
larvae collected simultaneously (Table 3). Larval round
herring had eaten cylopoid copepods {Oncaea spp. and
Corycaeus spp.) and pteropods more frequently, but
calanoid copepodites and adult copepods less frequent-
ly, than had larval gulf menhaden. Percent similarity
of the diets of these larvae was 52.2, a value that in-
dicates marginal overlap in diet (Schoener 1970).
186
Fishery Bulletin 90(1). 1992
Growth comparisons
Marginal growth increments seemed to form from
evening through early morning (Table 4). The allo-
metric relationship (logio radius = 0.126 logioSL +
1.413; r^ 0.91) between otolith radius and standard
length of 131 larvae also suggested a daily periodicity
in otolith increment formation (Hales 1987).
Estimates of the length at hatching (L(0)) for gulf
menhaden provided by the Laird-
Gompertz model 3.4mm SL (Fig.
1; Table 5), closely approximate
the length-at-hatching of larvae
incubated in the laboratory at a
temperature of 20°C, 2.6-3.0mm
SL (Hettler 1984). Estimates of
L(0) for round herring, about 1.2
mm SL, however, are consider-
ably lower than the lengths re-
ported for larvae hatched in the
laboratory: 3. 8-4. 0mm body
length from eggs collected in the
South Atlantic and incubated at
20.5°C (O'Toole and King 1974)
and 6.0 mm SL from eggs col-
lected in the Pacific and incu-
bated at 24-26°C (Miller et al.
1979). If the interval from hatch-
ing to deposition of the first
growth increment is shorter in
reality than the assumed 5 days,
the Laird Gompertz growth
curve would shift to the left,
yielding a greater value for L,o),
but the form of the growth curve,
i.e., the growth rate, would re-
main the same.
Round herring grew faster
than gulf menhaden through the
first 20-40 days; gulf menhaden
exhibited faster growth than
round herring thereafter (Fig. 2).
The fastest growth rate (=0.85
mm/day) for round herring lar-
vae occurred at about 15 days.
Average growth rates through
27 days for December 1980 were
0.71 and 0.46 mm/day for round
herring and gulf menhaden;
average rates through 50 days in
•February 1981 were 0.45 and
0.34 mm/day. Annual differences
in larval gulf menhaden growth
are discussed in Warlen (1988).
Table 4
Percentage of round herring larvae with partially formed (nar-
row) or completed (wide) marginal otolith growth increments
collected at three different times of day.
Time of capture
(h)
No. of
fish
Percentage
Partially formed Completed
1800
2400
0600
9
10
6
22
40
100
78
60
0
Etrumeus teres
12 14 16 18 20 22 24 26
27 5
B
.
25 0
22 5
-/^^
20 0
. --y^
17.5
*''i/ 3
15,0
/
(■:.
12 5
""/*
'
10 0
' Y
7 5
/
E. teres
5 0
2 5
27.5
•
25.0
22 5
20.0
17.5
I
.^ y^g^
• •!
• •3 331:
36p^ 3 3 .
]
!•• 5 S*'
•«49^M3
3* '
23
#«*a33. . .
Uijl I..
•45 3* •
J
■J^-
4*3>
33-
B, paironua
5.0
^7
•
2-5
L .
20 25 30 35
20 25 30 35 40
ESTIMATED AGE (DAYS)
Figure 1
Growth of larval round herring Etrumeus teres and gulf menhaden Brevoortia patronus
collected in December 1980 (A) and February 1981 (B) in the northern Gulf of Mexico.
The log form of the Laird-Gompertz model was used to describe the growth of both species
(numbers indicate location of coincident data points).
NOTE Chen et al : Feeding and growth of larval Etrumeus teres and Brevoortia patronus
187
Table 5
Estimates of Laird-Gompertz growth model parameters* and mean age (d) and SL (mm) for larval round herring and gulf menhaden
collected in the northern Gulf of Mexico during December 1980 and February 1981.
Grovrth model parameters
Mean
estimated
age (d)
Mean
SL
(mm)
Date
observations
^0)
A<0)
a
December 1980
Round herring
81
1.184
0.259
0.081
16.025
12.056
(0.310)
(0.055)
(0.015)
(0.590)
(0.501)
Gulf menhaden
80
3.418
0.056
< 0.001
14.862
8.034
(0.995)
(0.037)
(0.407)
(0.337)
(0.191)
February 1981
Round herring
50
1.240
0.232
0.077
25.200
15.620
(0.710)
(0.089)
(0.018)
(1.318)
(28.871)
Gulf menhaden
561
3.401
0.087
0.045
23.401
11.493
(0.246) (0.009) (0.005)
ling, A(„| = specific growth rate at hatching, a = exponential decay of the specific
ic standard errors.
(0.333)
growth rate. Values
(0.121)
in paren-
• L,o, = length at hate
theses are asymptot
In two comparisons, the growth
curves of these species differed,
i.e., the sum of squares of the dif-
ferences between observed and
predicted lengths was greater
when data for the two species
were pooled than when the data
were considered separately. In
the comparison of larvae col-
lected in December 1980, the
sum of squares of deviations was
6.327 for pooled data and 2.736
for data considered separately
(total observations = 161). In the
comparison of larvae collected in
February 1981, from two tran-
sects, the sum of squares of devi-
ations was 13.255 for pooled data
and 10.477 for data considered
separately (total observations =
611).
Elmmam lerat
Brevoofll* palronu*
ESTIMATED AQE (DAYS)
Figure 2
Age-specific growth rates of larval round herring Etrumeus teres and gulf menhaden
Brevoortia patronus collected in December 1980 (A) and February 1981 (B) in the north-
em Gulf of Mexico. Specific grovrth rates-at-age were derived from the log form of the
Laird-Gompertz growth model parameters.
Discussion
The large, spatulate, and toothed jaws of larval round
herring might enable them to eat larger food organisms
than gulf menhaden, but while the diets of larval round
herring and gulf menhaden differed, the width of food
organisms coincided. Diets, then, do not directly reflect
differences in jaw structure and dentition. The ptero-
pods eaten by both species were Limacina trochifor-
mis, the cyclopoid copepods were primarily of the
genera Oncaea and Corycaeus, and the calanoid cope-
pods were primarily of the genera Paracalanus and
Acartia. Round herring larvae ate more pteropods and
cyclopoid copepods, but fewer calanoid copepods than
did gulf menhaden larvae.
The more offshore distribution of larval round her-
ring in the central and western northern Gulf of Mex-
ico (Shaw and Drullinger 1990; present data) may
explain differences in diet and growth. All of the food
organisms eaten by larval round herring and gulf
menhaden are broadly distributed in continental shelf
waters, but some of the copepods have different pat-
terns of distribution across the shelf in the northern
Fishery Bulletin 90(1). 1992
Gulf of Mexico (Ortner et al. 1989). Acartia, for exam-
ple, occurs in greater abundance inshore, in less saline
waters, whereas Oncaea and Corycaeus are more abun-
dant in water of traditional salinities offshore (Ortner
et al. 1989). Prior experience and learning can influence
the capture efficiency, food selection, and ingestion
rates of larval fishes (see review in Stoecker and Govoni
1984); and because larval round herring occupy more
offshore waters, they may be conditioned to feed pref-
erentially on the cyclopoids Oncaea and Corycaeus.
The difference in larval growth between these two
species may reflect differences in the physical environ-
ment where these larvae grow. Offshore water in the
northern Gulf of Mexico is typically warmer than in-
shore water during the winter. Inshore-offshore gra-
dients in average water column temperature among the
three stations along the three transects were 19.2 to
20.7 to 22.1°C for the Cape San Bias transect in
December 1980; 16.1 to 15.0 to 18.1, 16.9 to 19.5 to
19.8, and 12.9 to 18.5 to 19.0°C for the Mississippi
Delta, Cape San Bias, and Galveston Bay transects in
February 1981. Temperature differences of this mag-
nitude can account for intraspecific differences in
growth rates among larval fish (Jones 1986, Warlen
1988) as seen here in the slower growth of round her-
ring larvae in the cooler water of February 1981. The
faster, early growth of round herring larvae, overall,
probably results from the warmer waters of its offshore
occurrence.
Acknowledgments
The collection of specimens examined in this paper was
supported by a contract to the Beaufort Laboratory
of the Southeast Fisheries Science Center, National
Marine Fisheries Service, NOAA, from the Ocean
Assessment Division, National Ocean Services, NOAA.
This paper was developed while the senior author was
a visiting scientist at the National Marine Fisheries
Service, NOAA, Beaufort Laboratory.
Citations
Campana, S.E., and J.D. Neilson
1985 Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci.
42:1014-1032.
Christinas, J.Y., and R.S. Waller
1975 Location and time of menhaden spawning in the Gulf of
Mexico. Gulf Coast Res. Lab., Ocean Springs, MS, 20 p.
Fahay, M.P.
1983 Guide to the eariy stages of marine fishes occurring in
the western North Atlantic Ocean, Cape Hatteras to the
southern Scotian Shelf. J. Northwest Atl. Fish. Sci. 4:1-423.
Fore, P.L.
1971 The distribution of eggs and larvae of the round herring,
Etrumeus teres, in the northern Gulf of Mexico. Am. Soc. Biol.
Bull. 18, 34 p.
Govoni, J.J., D.E. Hoss, and A.J. Chester
1983 Comparative feeding of three species of larval fishes in
the northern Gulf of Mexico: Brevoortia patronus, Lewstomus
xanthwrus, and Micropogonias undulatus. Mar. Ecol. Prog.
Ser. 13:189-199.
Govoni, J.J., P.B. Ortner, F. AI-Yamani, and L.C. Hill
1986 Selective feeding of spot, Leiostomus xanthurus, and
Atlantic croaker, Micropogomas undulatus, larvae in the north-
ern Gulf of Mexico. Mar. Ecol. Prog. Ser. 28:175-183.
Govoni, J.J., D.E. Hoss, and D.R. Colby
1989 The spatial distribution of larval fishes about the Missis-
sippi River plume. Limnol. Oceanogr. 34:178-187.
Hales, L.S.
1987 Distribution, abundance, reproduction, food habits, age,
and growth of round scad, Decapterus punctatus, in the South
Atlantic Bight. Fish. Bull, U.S. 85:251-268.
Hettler, W.F.
1984 Description of eggs, larvae, and early juveniles of gulf
menhaden, Brevoortia patronus, and comparisons with Atlantic
menhaden, B. tyrannus, and yellowfin menhaden, B. smithi.
Fish. Bull., U.S. 82:85-95.
Houde, E.D.
1977 Abundance and potential yield of the round herring,
Etrumeus teres, and aspects of its early history in the eastern
Gulf of Mexico. Fish. Bull., U.S. 75:61-89.
Houde, E.D., and P.L. Fore
1973 Guide to identity of eggs and larvae of some Gulf of Mex-
ico clupeid fishes. Fia. Dep. Nat. Resour., Mar. Res. Lab.,
Leafl. ser. IV, pt. 1, no. 23, 14 p.
Jones, C.
1986 Determining age of larval fish with the otolith increment
technique. Fish. Bull., U.S. 84:91-103.
Kappenman, R.F.
1981 A method for growth curve comparisons. Fish. Bull,
U.S. 79:95-101.
Miller, J.M., W. Watson, and J.M. Leis
1979 An atlas of common nearshore marine fish larvae of the
Hawaiian Islands. Misc. Rep. PO-02, Univ. Hawaii Sea Grant
Prog., Honolulu, 179 p.
Ortner, P.B., L.C. Hill, and S.R. Cummings
1989 Zooplankton community structure and copepod species
composition in the northern Gulf of Mexico. Continental Shelf
Res. 9:387-402.
O'Toole, M.J., and D.P.F. King
1974 Early development of the round herring, Etumeus teres
(De Kay) from the south-east Atlantic. Vie Milieu 24:443-452.
Reintjes, J.W.
1980 Marine herring and sardine resources of the northern Gulf
of Mexico. In Flandorfer, M., and L. Skuplen (eds.), Workshop
for potential fishery resources of the northern Gulf of Mex-
ico, New Orleans, 1980. Publ. MASGP-80-012, Miss.-Alabama
Sea Grant Consort., 14 p.
Roel, B.A., and Y.C. Melo
1990 Reproductive biology of the round herring, Etrumeus
whitekeadi. S. Air. J. Mar. Sci. 9:177-187.
Schoener, T.W.
1970 Nonsynchronous spatial overlap of lizards in patchy
habitats. Ecology 51:408-418.
NOTE Chen et al,: Feeding and growth of larval Etrumeus teres and Brevoortia patronus
189
Shaw. R.F., and D.L. Drullinger
1990 Early-life history profiles, seasonal abundance, and dis-
tribution of four species of clupeid larvae from the northern
Gulf of Mexico, 1982 and 1983. NOAA Tech. Rep. NMFS 88,
60 p.
Sogard, S.M., D.E. Hoss, and J.J. Govoni
1987 Density and depth distribution of larval gulf menhaden,
Brevoortia patronus, Atlantic croaker, Micropogonias undu-
latus, and spot, Leiostomus xanthunis, in the northern Gulf
of Mexico. Fish. Bull., U.S. 85:601-609.
Stoecker, D.K., and J.J. Govoni
1984 Food selection by young gulf menhaden {Brevoortia
patronus). Mar. Biol. (Beri.) 80:299-306.
Warlen, S.M.
1988 Age and growth of larval gulf menhaden, Brevoortia
patronus, in the northern Gulf of Mexico. Fish. Bull., U.S.
86:77-90.
Whitehead, P.J.
1963 A revision of the recent round herrings (Pisces: Dus-
sumieriidae). Bull. Br. Mus. (Nat. Hist.) Zool. 10:305-380.
Zweifel, J.R., and R. Lasker
1976 Prehatch and posthatch growth of fishes— a general
model. Fish. Bull., U.S. 74:609-621.
Analytical correction for
oversampled Atlantic mackerel
Scomber scombrus eggs collected
with oblique plankton tows
Denis D'Amours
Francois Gregoire
Division de la Recherche sur les Peches
Minist^re des Peches et des Oceans, Institut Maurice-Lamontagne
C.P. 1000, Mont-Joli, Quebec G5H 3Z4, Canada
Atlantic mackerel Scomber scom-
brtLS is a pelagic species spawning
on both sides of the North Atlantic
ocean. In the east, mackerel spawn
off the British Isles and in the North
Sea (as reviewed by Lockwood 1988
and Daan et al. 1990). In the west,
mackerel spawn in the Middle At-
lantic Bight (Berrien 1978) and in
the Gulf of St. Lawrence (Ware
1977). Atlantic mackerel is a mod-
erately prolific species (Bigelow
and Schroeder 1953); its fecundity
has been estimated at 255,000 eggs
for a medium-size female (30 cm
FL) in the northeast Atlantic (Lock-
wood et al. 1981), and at 243,000
eggs for a similar-size female in
the Middle Atlantic Bight (Morse
1980).
Mackerel eggs are concentrated
near the surface when the water
column is thermally stratified dur-
ing spawning (Coombs et al. 1983,
Ware and Lambert 1985). Sette's
(1943) data from the Middle Atlan-
tic Bight indicated that 80% of
mackerel eggs were in the top 10 m.
In the North Sea, Coombs et al.
(1981) reported that 91% of mack-
erel eggs were above 26 m, and that
more than 85% were between 0 and
16 m. In the Gulf of St. Lawrence,
deLafontaine and Gascon (1989) in-
dicated that 89% of mackerel eggs
were within 15 m of the surface. The
Manuscript accepted 9 December 1991.
Fishery Bulletin, U.S. 90:190-196 (1992).
distribution of mackerel eggs is thus
characteristically non-homogeneous
in the vertical plane.
In the Gulf of St. Lawrence and
Middle Atlantic Bight, mackerel
eggs are routinely surveyed for
stock assessment purposes (e.g.,
Castonguay and Gregoire 1989, Ber-
rien 1990). Surveys are carried out
with oblique plankton tows, with
bongo nets as described by Posgay
and Marak (1980). However, accu-
racy of oblique plankton tows is
known to be sensitive to nonhomog-
eneous vertical distribution of the
sampled organisms (Smith and Rich-
ardson 1977). Ideally, there should
be no hesitation at the surface when
retrieving the net, as it would lead
to a severe oversampling of the sur-
face layer where the eggs are con-
centrated (Posgay and Marak 1980,
Smith et al. 1985). In practice, it is
difficult not to drag the plankton
net at the surface for at least a few
seconds; when the net is retrieved,
it reaches the surface several meters
behind the block, and is dragged at
the surface until directly under the
block, where it can be lifted out.
During such dragging at the sur-
face, the mouth of the net is typical-
ly nearly all submerged and samples
the surface layer. It is usually as-
sumed that such oversampling at
the surface leads to a negligible bias
in estimates of abundance.
In this paper, the bias caused by
an oversampling of surface water
on the calculated abundance of
mackerel eggs is analyzed. An ana-
lytical correction for this bias is de-
rived and applied to empirical data
from a mackerel egg survey held in
the Gulf of St. Lawrence in 1990 to
reevaluate the annual production of
eggs. Also, some potential effects
of oversampling surface water are
evaluated when computing total
abundance and mortality rates of
near surface organisms.
Bias in computed egg
abundance caused by
oversampled surface
water
Distribution
of macl<erel eggs
Concentrations of mackerel eggs
are highest near the surface and
decrease rapidly with depth. Sund-
by (1983) reported that under as-
sumptions applicable in the present
study, a negative exponential model
(as Eq. 1, below) was appropriate to
describe the vertical distribution of
mackerel eggs. Ware and Lambert
(1985) also concluded that the ver-
tical distribution of mackerel eggs
was best described by a negative ex-
ponential model. Data on the ver-
tical distribution of mackerel eggs
presented by Sette (1943) and by
deLafontaine and Gascon (1989)
were fitted to negative exponential
models; in both cases, over 90% of
the variance in the egg distribution
was explained. Therefore, a nega-
tive exponential model is appropri-
ate to describe the distribution of
mackerel eggs in the vertical plane.
Sampiing macl<erei eggs
Let the abundance of a population
of eggs in a body of water decrease
from the surface following an ex-
ponential model:
dN(z)
dz
= - kN(z)
(1)
190
NOTE D'Amours and Gr^goire: Analytical correction for oversampled Scomber scombrus eggs
where N(z) = concentration of eggs in number per
volume at depth z, and k = rate constant. Upon in-
tegration of Eq. 1, the concentration of eggs at depth
z is given by
N(z) = N^e-^,
(2)
where Nq = concentration of eggs at the surface.
When integrating Eq. 2, the total number of eggs (N^)
in number per surface area in this body of water is
given by
Na= f
■J (
No
N^e-k^dz = — .
k
(3)
If an oblique plankton tow (Fig. lA) is made through
this distribution of eggs with a net of radius a, and a
centered depth of a, the total number of eggs collected
(Nh) will be equal to
N.
(4)
;Li _ o + a +\/a--{z-ay
[ ( N(z) dydzdxi
Va--(z-a)
/Lj o + a + \/a? - (z-a)'
I I N(z) dydzdx2,
■\Jar-(z-aY
Figure 1
Path (broken line) of net during an oblique plankton tow: L's
represent length components of the tow along horizontal ref-
erences X] and X; ; D is the maximum depth of sampling, and
e's represent angles between path of net and horizon. In (A),
the net is recovered upon reaching the surface; in (B), the
net is dragged at the surface along Lp before recovery.
where Xj is the horizontal distance from the start of the tow and X2 is the horizontal distance from the end of
the tow (Fig. lA), and where z and y represent the vertical and horizontal openings of the net, respectively. In-
tegrating Eq. 4 over the limits on z and y (as in D'Amours 1988),
N.
1 +
(ka):
I Noe-k^dxi + I
[-' 0 -^ 0
N„ e-"^ dxzl.
(5)
The term ka originates from the slight difference between the position of the geometric center of the net and
the position of the center of abundance of the eggs within its opening (D'Amours 1988). Defining,
tan 0; = — , where D is the maximum depth of the tow, Eq. 5 can be rewritten as
Li
Nh =
1 +
(ka)=
I N«
e-ktanfliXi (Jxj +
/,
M e-ktane2X, dxp
(6)
192
Fishery Bulletin 90(1), 1992
Evaluating Eq. 6 for kD»0,
Nh =
No Ll
k Di
1 +
(ka)2
(7)
Nt =
No Ll
k D,
1 +
(ka)2
+ Noe-k-'Lc
1 +
(ka)2
(10)
where L = Lj + L2, and is the total horizontal length
of the tow (Fig. 1 A). It can now be seen that when the
number of eggs collected in an oblique plankton tow
(Eq. 7) is multiplied by the ratio D/L, and when the pro-
duct ka is small, an approximation of Eq. 3 is obtained,
which is a measure of the local abundance of eggs in
number per surface area. The procedure of multiply-
ing the total number of eggs collected in the oblique
tow (equation 7) by the ratio of its maximum depth to
its total horizontal length (D/L), is equivalent to the
standardization procedure described by Smith and
Richardson (1977). To obtain estimates of abundance
in number per surface area, the standardization pro-
cedure consists of multiplying the number of eggs
collected by the ratio of the maximun depth attained
during the tow to the volume of water filtered. This
standardization procedure is valid if all depth strata are
sampled equally.
Now assume the same population of eggs, but where
the net is dragged at the surface while being readied
for recovery at the end of the tow. The length of drag
at the surface is represented by Lq in Figure IB.
Along Lp, the mouth of the net is centered at depth
a, which is equal to its radius a; i.e., oversampling
occurs in the layer of water immediately below the
surface, as deeply as the diameter of the net. Over
Lo, the net will collect a number of eggs (Ng) equal to
;o + a +\fa'-{z-a)''
\ N(z)dydz, (8)
o-a -* - ya- - (z-a)-
which is approximated by (as in D'Amours 1988)
Ns = Noe-kaLD
1 -H
(ka)2
(9)
The total number of eggs (Nx) collected during an
oblique tow, where the net is dragged at the surface
at the end, will then be equal to the sum of Nh (Eq.
7) and Ng (Eq. 9):
The component Ng will add to the number of eggs col-
lected, and its inclusion in the standardization pro-
cedure will result in a systematic overestimation of the
abundance of eggs per surface area. When Eq. 10 is
standardized with Ld included in the total length of
the tow (Lx in Fig. IB), and the result divided by the
true theoretical abundance of eggs (Eq. 3), an expres-
sion is obtained which is the ratio (B) of the biased abun-
dance to the true abundance of eggs:
B = L -H D
-ko
1 -I-
(ka)2
(11)
Removal of bias from computed
abundance of mackerel eggs
Assumption of constant filtration efficiency
In Eq. 11, L and Lq can be replaced by the proportion
of the total duration of the tow they represent, under
the assumption of constant filtration efficiency. This
assumption is required to use tow time as a measure
directly proportionnal to amount of water filtered,
so as to separate L and L^ in Eq. 11. However, as
pointed out by Smith and Richardson (1977), the filtra-
tion efficiency of a plankton tow declines typically with
the duration of the tow, as the accumulated plankton
reduces the porosity of the net. They warned that the
diminishing efficiency of a net could result in an under-
sampling of surface water. To verify whether such
undersampling of surface water occurred, which would
offset oversampling at the end of the tow, the time-
course of the efficiency of the plankton net must be
assessed.
If filtration efficiency diminishes with time, the vol-
ume filtered per unit time will diminish with increas-
ing tow duration. The residuals about a straight line
fitted on the values of volume filtered against tow dura-
tion would then show a decreasing pattern of depar-
ture from linearity. The volumes filtered for the tows
in the Gulf of St. Lawrence in 1990 were regressed
against their respective total duration. The residuals
of this regression did not indicate a decreasing de-
NOTE D'Amours and Gr^goire: Analytical correction for oversampled Scomber scombrus eggs
193
-100 -■
200
400
600
800
TOW DURATION (s)
Figure 2
Residuals of a linear regression of volume filtered against tow
duration of oblique plankton tows, from a mackerel egg survey
held in the Gulf of St. Lawrence in 1990.
parture from linearity; somewhat unexpectedly, a
tendency towards an increasing departure from linear-
ity could be detected (Fig. 2). Therefore, it can be con-
cluded that no surface undersampling occurred as a
result of diminishing filtration efficiency. The apparent
increasing departure from linearity can be explained
by the fact that long tows (e.g., duration of 10 minutes)
are deeper, i.e., well below the stratum where mackerel
eggs are abundant. During short tows (e.g., duration
of 6 minutes), the net is towed mainly in the stratum
were eggs are present, and the filtration efficiency is
less, though stationary, than in water devoid of eggs.
During long, deep tows, more time is spent below the
stratum containing mackerel eggs, and proportionally
more free-flowing water is filtered there.
Correction of survey data
In Eq. 11, a rate constant k must be introduced to
describe the distribution of the sampled organisms in
the vertical plane. For the purpose of the demonstra-
tion, a rate constant k = 0.1 5/m was selected as repre-
sentative of all mackerel egg stages at all stations; as
discussed below, this rate constant is a representative
value extracted from the literature on mackerel egg
distribution. During the mackerel egg survey carried
out in the Gulf of St. Lawrence in late-June and early-
July 1990, the total duration of each oblique tow was
measured, as well as the duration of the period during
which the Bongo net was dragged at the surface before
recovey (F. Gregoire, unpubl. data). The period of drag
at the surface started when the net was visually spotted
at the surface and ended when the net was lifted out
of the water. From those measurements, values of L
and Ld were calculated in percent of total tow time.
With a rate constant k = 0.15/m, a net radius a = 0.305
m, a centered depth a = 0.305 m along Lq, and a mea-
sured maximum depth D, a value of the degree of bias
B was calculated for each tow as per Eq. 11. The cor-
rected abundance of eggs was obtained by multiplying
the computed biased abundance by [100%/B]. Using un-
corrected and corrected abundances of eggs at each sta-
tion, two total annual productions of mackerel eggs
were computed for the Gulf of St. Lawrence in 1990
following the procedures of Ouellet (1987). The totals
were 6.77 xlO^"* eggs with uncorrected abundance,
and 5.63 xlO^^ eggs with corrected abundance. The
difference of 1.14 x 10^^ eggs, with a mean fecundity
of 300,000 eggs and a sex ratio of 1:1, amounted to
7.6x10* mature mackerel.
The parameter D used in the above calculations was
measured accurately with a bathymeter mounted on
the plankton net. If triangulation had been used, where
D is estimated by the amount of wire paid out and the
angle subtended at the block, another source of bias
would have been introduced owing to the approx-
imative nature of the method. Assume a population of
mackerel eggs in a body of water where k = 0.15 and
No = 750; if sampled to a depth D of 50 m with a net
of radius a = 0.305 on a transect where L = 1000 m, a
total of 100,000 eggs will be collected (Eq. 7). Standar-
dization of this result by the ratio of D to L shows that
the abundance of eggs is 5000 eggs/m-. Had D been
underestimated by 10% at 45 m, the abundance of eggs
would have been underestimated also by 10% at 4500
eggs/m-. If the same tow is repeated, but with Ld =
75m and a = 0.305m, a total of 153,775 eggs will be col-
lected. Standardization of this result vdth D correctly
evaluated at 50 m indicates an abundance of 7152
eggs/m^; with D underestimated by 10% at 45 m, stan-
dardization indicates an abundance of 6437 eggs/m^.
These examples show how an underestimation of 10%
of D results in an abundance of eggs equal to 90% of
the real value, and how a 7% (75m/1075m) oversam-
pling at the surface results in an abundance of eggs
equal to 143% of real value. Also, they show that when
both an underestimation of D and an oversampling of
the surface layer occur during a tow, the effects of both
biases on the estimate of abundance are opposite, but
not symmetrical, with the effect of the oversampling
at the surface much more important than that from the
underestimation in D.
A degree of bias (B in Eq. 11) was computed for
various combinations of L^ (with L = 100%-Ld) and
rate constant k, with a = a = 0.305 m, and D = 50 m (Fig.
3A). The degree of bias caused by an oversampling of
surface water is a function of the time of sampling at
the surface, and of the degree of contagion of the eggs
near the surface, as described by the parameter k. For
194
Fishery Bulletin 90(1). 1992
E
I-
z
Z
o
o
UJ
cr
0 2
4 6
8 10 12 14
0.40
0.35
0.30
\f
\\\\
\\V\v
\ V \ \ N
0.25
\
\\V\'^
0.20
0.15
- \\
\ \ ''''^^\S\^^^^^^^
0.10
■\
'<^ \^
^^^^^^
0.05
^^^ —
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0 2 4 6 8 10 12 14
% TIME AT SURFACE
B.
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
0.40| I I M I in t, V n ^ A M ^ s ^ K — r— n0.40
0.5 1.0 1.5 2.02.53.0 3.54.0 4.5
% TIME AT SURFACE
example, with a sampling time at the surface represent-
ing 7% of the total duration of the tow, and with a rate
constant of 0.15/m, the calculated abundance will be
140% of the real value. With eggs highly concentrated
near the surface, that is with high values of k, even
briefer towing times at the surface will still result in
severe bias.
The degree of bias B was also computed for similar-
ily varying k and Ld, again with a=a = 0.305m, but
with maximum depth D increased to 200 m (Fig. 3B).
For a rate constant of 0.15/m as in the previous exam-
ple, but with towing time at the surface representing
Figure 3
Isopleths (in %) of the ratio of biased abundance to true abun-
dance per surface area of a theoretical population of fish eggs.
The isopleths were computed for variable rate constants k of
a negative exponential model describing the vertical distribu-
tion of the eggs, and for varying degree of oversampling sur-
face water during an oblique plankton tow, L^ , expressed as
percent of the total duration of the tow. In (A), the maximum
sampling depth was set at 50m; in (B), the maximum sam-
pling depth was set at 200 m. All other parameters equal in
(A) and (B).
only 2% of the total duration (e.g., 12 seconds at the
surface for a total tow time of 10 minutes), the calcu-
lated abundance of eggs will again be 140% of the real
value. This somewhat counterintuitive result stems
from the fact that by sampling deep strata, the frac-
tion of the tow occurring in the stratum where eggs
are present is proportionally smaller. As a result, brief
times of oversampling at the surface have proportional-
ly more effect on the calculated abundances of eggs
than when the tow extends only to shallow strata.
Effects of oversampling surface water
on variance of total abundance
and on mortality rate
Effect on variance
Since the length of dragging at the surface is likely to
vary (as a function of weather, crew handling of the
net, etc.), a variance will be introduced in the computa-
tion of the total abundance of eggs over the studied
body of water. It was assumed that the abundance of
eggs (5000 eggs/m-) was constant over the surface of
a theoretical body of water where numerical experi-
ments were carried out. Ten oblique tows were made
in this theoretical body of water, with L = 1000 m,
a = a = 0.564 m, and k = 0.2/m, which are convenient
values for illustrative purposes. A different length of
drag at the surface (Lp) was assigned randomly to
each tow; ten random numbers were multiplied by an
arbitrary length of 6 m and the resulting Lp's were 6
(x2), 12, 24 (x3), 30, 36, 48, and 54 m. Ten estimates
of abundance of eggs per unit surface area were
calculated, and the mean was 5994 eggslrrr, with 95%
confidence intervals of 5646-6342 eggs/m-. This illus-
trates that small and variable lengths of drag at the
surface bias the estimated abundance over the whole
body of water, and add a substantial margin of uncer-
tainty to the estimate of local abundance.
In another numerical experiment, the abundance of
eggs was again assumed constant throughout at 5000
eggs/m^, except this time the degree of contagion
NOTE D'Amours and Gr^goire: Analytical correction for oversampled Scomber scombrus eggs
195
near the surface was made variable, i.e., the rate con-
stant k varied randomly within bounds. Ten transects
were carried out through these distributions of eggs,
with L = 1000m, a = a = 0.564m, and the length of drag
at the surface (Lp) was held constant at 50 m. For
each tow, a random value was assigned to the rate con-
stant k: ten random numbers were multiplied by 0.1/m,
and added to 0.15. The resulting values were 0.15, 0.16,
0.17 (x2), 0.20 (x2), 0.21 (x2), 0.22, 0.23 (x2), and
the corresponding values of Nq were adjusted so that
No/k = 5000 eggs/m^. Ten estimates of abundance of
eggs per unit surface area were calculated, and the
mean was 6876/m-, with 95% confidence interval of
6713-7039 eggs/m^. This indicates that even when
maintaining a constant length of drag at the surface,
similar problems of bias and variance still arise when
the degree of contagion of the eggs varies.
Effect on mortality rates
Again for numerical experiments, the abundance of
eggs at a theoretical station was assumed to be 5000
eggs/m^, with No = 475/m3, and k = 0.095/m (i.e.,
475/0.095 = 5000). An oblique tow with L= 1000 m,
D = 50m, LD = 50m, and a = a' = 0.564m, was made
through this concentration of eggs at time to, and the
biased abundance was calculated to be 5873/m2. For
the purpose of the demonstration, it was assumed that
the eggs suffered no mortality. Some time later at time
tj , the abundance of eggs was still the same, but they
were closer to the surface, with No = 950/m^ and k =
0.19/m (i.e., 950/0.19 = 5000). The same oblique tow in
this slightly rearranged concentration of eggs yielded
a biased estimate of abundance of 6806/m2, a relative
increase of nearly 16% compared with the value at
to, and an absolute bias of over 36%. In real situa-
tions, then, an increase in the degree of vertical con-
tagion of the eggs over a sampling period could lead
to an underestimation of the mortality rate if the sur-
face water is oversampled, or to an overestimation if
the degree of contagion decreases.
Conclusion
The oblique tow is a convenient method to obtain an
estimate of abundance of eggs over a body of water.
However, in actual operating conditions, it is rarely
possible to carry out an oblique tow without dragging
the net at the surface for some period of time, which
may introduce a large bias in the estimate of abun-
dance. The first practical recommendation to avoid
such bias is to evaluate the assumption that a brief drag
time at the surface will cause only a small bias in the
estimation of abundance. If eggs are equally distrib-
uted over considerable depth, or concentrated in deeper
water, this assumption is valid. If not, action should
be taken to avoid dragging the net at the surface. If
this is impossible, a measure of the amount of over-
sampling at the surface, and of the rate constant k,
should be used to remove the bias from the data follow-
ing Eq. 11.
In this study, the percent time of the tow spent at
the surface was used as a measure of the amount of
oversampling at the surface; this information is read-
ily recorded in the field. However, the constant k
describing the distribution of eggs had to be approx-
imated from data available in the literature. Ware and
Lambert (1985) reported values of k ranging from 0.1
to 1.1; they further indicated that variations in k were
related to the steepness of the thermal gradient in the
water column, the development stage of the eggs, and
the degree of wind-induced mixing. Data on mackerel
egg distribution by deLafontaine and Gascon (1989) in-
dicated a mean value k = 0.1, with the lowest values for
the most recently spawned eggs. Data on mackerel egg
distribution by Sette (1943) indicated a higher mean
value k = 0. 17, but with the highest values for the most
recently spawned eggs. Differences in mean values of
k as well as development-stage specific values may
result from differences in local wind conditions as well
as in differences in local water density. As discussed
by Sundby (1983), the shape of a vertical distribution
of mackerel eggs will be determined by the difference
of density between the egg and the surrounding water,
and by the degree of wind-induced mixing. The rela-
tionship reported by Sundby (1983) between wind
velocity and vertical eddy diffusivity coefficient of
mackerel eggs indicates that the rate constant k should
diminish as the state of the sea increases. The definite
application of the analytical correction proposed herein
will require more site-specific studies on the factors af-
fecting the vertical distribution of mackerel eggs and
determining the value of k. Nonetheless, the value
k = 0.15 used in this study is representative of realistic
conditions in the field, and can be considered as a con-
servative estimate of the degree of vertical contagion
of mackerel eggs. With more reliable values of k, the
simple correction procedure suggested in this study
could help increase the accuracy of biological
parameters based on data from fish egg surveys where
the technique of the oblique plankton tow has been
used.
Acknowledgments
Dr. D. Booth and Mr. P. Gagnon reviewed an early ver-
sion of the manuscript. This work followed a study by
D'Amours (1988) during which help was provided by
196
Fishery Bulletin 90|l). 1992
Dr. Brian Petkau and Mr. Brian Leroux of the Statis-
tical Consulting and Research Laboratory, Department
of Statistics, University of British Columbia, Van-
couver. I thank an anonymous reviewer and Dr. L.L.
Jones for helpful reviews and comments.
Citations
Berrien, P.L.
1978 Eggs and larvae of Scomber scombrus and Scomber
japonicus in continental shelf waters between Massachusetts
and Florida. Fish. Bull., U.S. 76:95-115.
Berrien, P.
1990 Atlantic mackerel egg production and spawner biomass
estimates for the Gulf of St. Lawrence and Northeastern
United States waters in 1987. ICES CM 1990/H:18, 17 p.
Bigelow, H.B., and W.C. Schroeder
1953 Fishes of the Gulf of Maine. U.S. Fish. Wildl. Serv.,
Fish. Bull. 74, vol. 53, 577 p.
Castonguay, M., and F. Gregoire
1989 Le maquereau bleu {Scomber scombrus Linn6) du nord-
ouest de I'Atlantique, sous-r6gions 2 a 6 de I'OPANO en
1988. CAFSAC (Can. Atl. Fish. Sci. Advis. Comm.) Res. Doc.
89/39, 25 p.
Coombs, S.H., R.K. Pipe, and C.E. Mitchell
1981 The vertical distribution of eggs and larvae of blue whiting
(Micromesistius poutassou) and mackerel {Scomber scombrus)
in the eastern North Atlantic and North Sea. Rapp. P.-V.
R^un. Cons. Int. Explor. Mer 178:188-195.
Coombs, S.H., J. A. Lindley, and C.A. Fosh
1983 Vertical distribution of larvae of mackerel Scomber scom-
brus and microplankton, with some conclusions on feeding con-
ditions and survey methods. FAO Fish. Rep. 291(3):939-956.
Daan, N., P.J. Bromley, J.G.R. Hislop, and N.A. Nielsen
1990 Ecology of North Sea fish. Neth. J. Sea Res. 26:
343-386.
D'Amours, D.
1988 Vertical distribution and abundance of natant harpac-
ticoids on a vegetated tidal flat. Neth. J. Sea Res. 22:161-170.
deLafontaine, Y., and D. Gascon
1989 Ontogenetic variation in the vertical distribution of eggs
and larvae of Atlantic mackerel {Scomber scombrus). Rapp.
P.-V. R6un. Cons. Int. Explor. Mer 191:137-145.
Lockwood, S.J.
1988 The mackerel— Its biology, assessment and the manage-
ment of a fishery. Fishing News Books Ltd., Farnham, Sur-
rey, England, 181 p.
Lockwood, S.J., LG. Baxter, J.C. Gueguen, G. Joakimsson,
R. Grainger, A. Eltink, and S.H. Coombs
1981 The western mackerel spawning stock estimate for 1980.
Cons. Int. Explor. Mer, CM 1981/H:13, 20 p.
Morse, W.W.
1980 Spawning and fecundity of Atlantic mackerel, Scomber
scombrus, in the Middle Atlantic Bight. Fish. Bull., U.S. 78:
103-108.
Ouellet, P.
1987 Mackerel {Scomber scombrus) egg abundance in the
southern gulf of St. Lawrence from 1979 to 1986, and the use
of the estimate for stock assessment. CAFSAC (Can. Atl.
Fish. Sci. Advis. Comm.) Res. Doc. 87/62, 40 p.
Posgay, J. A., and R.R. Marak
1980 The MARMAP Bongo zooplankton samplers. J. North-
west Atl. Fish. Sci. 1:91-99.
Sette, O.E.
1943 Biology of the Atlantic mackerel {Scomber scombrus) of
North America. Part 1. Early life history. Fish. Bull., U.S.
50:149-237.
Smith, P.E., and S.L. Richardson
1977 Standard techniques for pelagic fish egg and larva sur-
veys. FAO Fish. Tech. Pap. 175, 100 p.
Smith, P.E., W. Flerx, and R.P. Hewitt
1985 The CalCOFI vertical egg tow (CalVET) net. /reLasker,
R. (ed.). An egg production method for estimating spawning
biomass of pelagic fish: Application to the northern anchovy,
Engraulis mordax. NOAA Tech. Rep. NMFS 36.
Sundby, S.
1983 A one-dimensional model for the vertical distribution of
pelagic fish eggs in the mixed layer. Deep-Sea Res. 30:
645-661.
Ware, D.
1977 Spawning time and egg size of Atlantic mackerel.
Scomber scombrus, in relation to the plankton. J. Fish. Res.
Board Can. 34:2308-2315.
Ware, D., and T.C. Lambert
1985 Early life history of Atlantic mackerel {Scomber scom-
brus) in the southern Gulf of St. Lawrence. Can. J. Fish.
Aquat. Sci. 42:577-592.
Association JDetween tiie sessile barnacle
Xenobalanus globicip'itis (Coronulidae)
and the bottlenose dolphin Tursiops
truncatus (Delphinidae) from the
Bay of Bengal, India, with a summary
of previous records from cetaceans
Arjuna Rajaguru
Gopalsamy Shantha
Systematics Laboratory, National Marine Fisheries Service, NOAA
National Museum of Natural History, Washington, DC 20560
Several instances of association be-
tween cetaceans and cirripeds have
been reported in the hterature.
Among the barnacles, Coronula
spp., Conchoderma spp., and Xeno-
balanus sp. have been reported
from various species of cetaceans of
both temperate and tropical waters
(Mackintosh and Wheeler 1929,
Mackintosh 1942). Devaraj and
Bennet (1974) reported a single spe-
cimen of Xenobalanus globicipitis
found attached to the fluke of a fin-
less black porpoise Neophocaena
phocaenoides caught off Karwar,
west coast of India. This type of
phoretic partnership (i.e., transpor-
tation by one promotes well-being
of the other) between 14 specimens
of a sessile barnacle Xenobalanus
globicipitis and a host, the bottle-
nose dolphin Tursiops truncatus, is
recorded here from the Bay of Ben-
gal on the east coast of India. This
is the first record of the bottlenose
dolphin as a host for Xenobalanus
globicipitis from the central and
northern Indian Ocean.
Five spinner dolphins Stenella
longirostris (Gray 1828) (113.0-
177.5cm TL) and six bottlenose dol-
phins Tursiops truncatus (Montagu
1821) (95.3-367.5cm TL), were col-
lected from the Bay of Bengal, off
Porto Novo (11°29'N; 79°46'E),
southeast coast of India, between 15
March 1982 and 1 September 1987.
These specimens were entangled
accidentally in bottom-set gillnets
(called Motha Valai, in Tamil ver-
nacular) set mainly for sharks. The
net is made of thick (no. 7-12) nylon
thread (monofilament). The stretched
mesh size is 10-12 cm, and there are
about 120 meshes from the head to
the foot rope; hence the net is about
12 m deep. Total length of the net
is about 800 m {'^ one-half mile).
Fishing operations, which were car-
ried out mostly at night, were con-
fined to the upper continental shelf,
up to 4 km from the coast, to depths
of 18-22 m. The dolphins became
entangled in the nets both day and
night.
All entangled dolphins were ex-
amined for external and internal
parasites (several dolphins had in-
ternal parasites). No barnacles were
found on the spinner dolphins. One
small bottlenose dolphin (148 cm
male) caught on 28 January 1985
had numerous Xenobalanus globici-
pitis attached (Fig. lA, B). None of
the four larger (>150cm) bottlenose
dolphins had any barnacles. All bar-
nacles were collected (four from the
left fluke, eight from the right fluke,
and one from each flipper) and
preserved in formalin. The barna-
cles were still alive after more than
12 hours out of water. Measure-
ments to the nearest millimeter
(Fig. 2) are given in Table 1.
In the sessile barnacles, extreme
reduction of plates is fotmd in Xeno-
balanus. The shell is thin, small,
white, irregularly star-shaped, and
vestigial, containing only the basal
parts of the animal. Connected to
this thin, star-shaped shell is a cylin-
drical, smooth, flexible, peduncle-
like body (Fig. 3). At the distal end
of this greatly elongated pseudo-
peduncle is a reflexed hood, which
bears two stumpy outpushings or
'horns,' but terga and scuta are ab-
sent. Cirri, mouth, a probosciform
penis, and associated organs project
from the reflexed hood. The wall
plates of this barnacle are em-
bedded in the skin of the dolphin,
with feeding appendages (cirri) and
associated organs suspended by the
long fleshy stalk. The body of X.
globicipitis was dark -brown in live
specimens, with a lighter colored
hood; the penis was whitish.
Although belonging to the sessile
group of the Cirripedia, this bar-
nacle closely resembles stalked
barnacles, especially Conchoderma
auritum which is also found on ceta-
ceans though never attached direct-
ly to the skin of its host. Xenobala-
nus globicipitis is always attached
directly to the skin of its host
(Pilsbry 1916, Barnard 1924). The
resemblance is superficial, and is
likely adaptive to being dragged
through the water by the host. The
closest affinities with Xenobalanus
are the genera Coronula, Platyle-
pas, and Tubicinella (Darwin 1854,
Pope 1958).
The barnacles are found only
around the rear margins of flippers
and flukes. It is hypothesized that
those that settle elsewhere are
more easily swept off. A single im-
mature barnacle (15 mm TL) was
found attached to each flipper of the
dolphin (Fig. lA). All 12 mature
barnacles (30-39 mm TL) (Table 1)
were aggregated at the rear margin
of the flukes (Fig. IB). Pilsbry
Manuscript accepted 19 December 1991.
Fishery Bulletin, U.S. 90:197-202 (1992).
197
198
Fishery Bulletin 90(1), 1992
B
Figure 1
Sessile barnacles Xenobalanus globieipitis on a bottlenose dolphin Tursiops truncattis,
caught in the Bay of Bengal, India. (A) Small barnacles on flippers. (B) Adult bar-
nacles on tail flukes. Barnacles from the left fluke were removed prior to photographing;
however, shell remnants are visible.
WBP
Figure 2
External measurements of the ses-
sile barnacle Xenobalanus globiei-
pitis. TL = total length, WB =
width of body, WBP = width of
basal plate, WH = width of hood.
(1916) reported that these barnacles grow in close
groups. This aggregation permits cross-fertilization,
which is common in hermaphroditic crustaceans
(Barnes 1986).
Xenobalanus globieipitis occurs on about 19 species
of cetaceans, from the small harbor porpoise Phocoena
phocoena to the large blue whale Balaenoptera mus-
culus (Table 2). The present record is the seventh
report from a bottlenose dolphin. Six previous reports
were from the central Atlantic
coast of the United States (True
1891, Mead and Potter 1990),
Gibraltar (Dollfus 1968, Pilleri
1970), and the east coast of
South Africa (Barnard 1924,
Ross 1984).
Based on a review of the lit-
erature, Rappe and Waerebeek
(1988) suggested that X. globiei-
pitis is an inhabitant of tropical
and warm-temperate waters.
They reported that occurrence of this species in the
northeast Atlantic and Mediterranean is erratic,
possibly related to sporadic incursions from adjacent
tropical warm-temperate waters. Their information
was based on only 23 reported localities. Our study
shows 87 localities (Table 2) reported for X. globieipi-
tis: 28 (32.2%) are located north of 40°N; 27 (31.0%)
between 35° and 40°N; and 32 (36.8%) between 30°N
and 30°S. From this it is clear thatX globieipitis is
NOTE Rajaguru and Shantha: Association between Xenobalanus globicipitis (Coronulldae) and Tursiops truncatus (Delphinidae) 1 99
Table 1
External measurements of the sessile barnacle Xenobalanus \
globicipitis collected from
a bottlenose dolphin
Tursiops
trun-
catus (0-; 148cm
TL), en
tangled in a
gillnet off Porto Novo, |
southeast coast of India
28 January
1985
Attachment
Specimen TL
WH
WB
WBP
area
no.
(mm) - - - -
Left fluke
1
37
11
7
5
2
37
10
6
5
3
37
11
6
5
4
32
9
6
5
Right fluke
5
35
11
7
5
6
30
11
7
5
7
39
12
7
5
8
35
12
7
5
9
30
10
7
5
10
34
11
8
8
11
37
10
6
4
12
30
9
6
5
Left flipper
13
15
5
3
3
Right flipper
14 15 4 3 3
): Basal plate to highest point of hood
tVB): Maximum width of elongated body
ate (WBP): Maximum width of basal plate
\VH): Maximum width of hood
Total length (TL
Width of body C
Width of basal p
Width of hood 0
a cosmopolitan species, occurring in temperate, warm-
temperate, and tropical waters. In relation to the
distribution of X. globicipitis, the bottlenose dolphin
is distributed widely in temperate and tropical oceans.
It is common from at least the north coast of Argen-
tina to northern Norway (Kenney 1990).
HO
HD
PP
10 mm
Figure 3
The sessHeha.ma.cle Xenobalaniis globicipitis. C = cirri, HD
= hood, HO = horn, P = penis, PP = pseudopeduncle, S =
shell.
Table 2
Distribution anc
reported hosts of the sessile barnacle Xenobalanus globicipitis. Identification of host species names are updated.
Host
Reported by
Year
Locality
Host
Reported by
Year
Locality
Order: Cetacea
Balaenoptera
Nilsson-Cantell 1930
Saldanha Bay
Suborder: Mysticeti
rmisculus
Mackintosh
1942
South Africa
Family: Balaenopteridae
(Blue whale)
Cornwall
1955a Pacific Canada
Balaenoptera
Broch
1924
Faroe Is., Greenland;
(continued)
Pike (in
1955b No locality
borealis
Ingoy, Norway
Cornwall)
(Sei whale)
Cornwall
1927
Vancouver
Boxshall (in
Shetland Is., Scotland
Nilsson-Cantell 1921
Faroe Is.
Rapp6 and
1930
Saldanha Bay, S. Afr.
Waerebeek)
Matthews
1938
Saldanha Bay
Balaenoptera
Caiman
1920
Shetland Is.
Mackintosh
1942
Saldanha Bay
physalus
Barnard
1924
North Atlantic
Heldt
1950
Tunis
(Fin whale)
Barnard
1924
Antarctic
Boxshall (in
Finnmark, Norway
Mackintosh
1929
South Africa
Rapp6 and
and Wheeler
Waerebeek)
Nilsson-Cantell 1930
S. Shetland Is.
Balaenoptera
Barnard
1924
Saldanha Bay
Nilsson-Cantell 1930
Saldanha Bay
musculus
Mackintosh
1929
South Africa
Mackintosh
1942
South Africa
(Blue whale)
and Wheeler
Raga and
Sanpera
1986
Galicia, Spain
200
Fishery Bulletin 90(1). 1992
Table 2 (continued)
Host
Reported by
Year
Locality
Host
Reported by
Year
Locality
Suborder: Odontoceti
Globicephala melas
Steenstrup
1852
Faroe Is.
Family: Delphinidae
(Long-finned
Hoek
1883
Faroe Is.
Delphiniis delphis
Hoek
1883
Atlantic
pilot whale)
Weltner
1897
Faroe Is.
(Common dolphin]
Gruvel
1905
South Africa
Gruvel
1912
Monaco
Gruvel
1920
South Africa
Pilsbry
1916
Chesapeake Bay
Richard
1936
Oran, Algeria
Gruvel
1920
Gibraltar
Stubbings
1965
Gor6e, Senegal
Nilsson-Cantell
1921
Faroe Is.
Pilleri
1970
W. Mediterranean
Richard
1936
Gibraltar
Rapp6
1988
Belgium
Zullo
Pilleri and
1963
1969
Woods Hole
Spanish coast,
Grampus griseus
Gruvel
1920
Azores
Knuckey
Medit. Sea
(Risso's dolphin)
Richard
1936
Azores
Kinze (in
Faroe Is.
Richard
1936
Azores
Rapp6 and
Pilleri and
1969
Barcelona, Spain
Waerebeek)
Gihr
Raga et. al.
1983
Faroe Is.
Ross
1984
SE coast, S. Afr.
Pilsbry
1916
New England
Sotalia sp.
Siciliano et al.
1988
Rio de Janeiro,
Ghbicephalus sp.
Barnard
1924
North Atlantic
(Tucuxi)
Brazil
Richard
1936
Mid-Atlantic
Stenella attenuata
Ross
1984
Mpelane, S. Afr.
Richard
1936
BaMares, Spain
(Spotted dolphin)
Orcinus orca
Gruvel
1920
Mediterranean
Stemlla
Raga et. a!.
1982
Spanish coast,
(Killer whale)
Richard
1936
Monaco
coeruleoalba
Medit. Sea
Richard
1936
Gibraltar
(Striped dolphin)
Raga et. al.
1983
Spanish coast
Pseudorca
Gruvel
1912
Monaco
Medit. Sea
crassidens
Gruvel
1920
Miguel, Azores
Ross
1984
SE coast, S. Afr.
(False killer
Richard
1936
Miguel, Azores
Raga and
1985
W. Mediterranean
whale)
Pilleri
1967
Spanish coast,
Carbonell
Medit. Sea
Boxshall (in
Mallorca
Rapp6 and
Family: Phocoenidae
Waerebeek)
Neophocaena
Devaraj and
1974
Karwar, India
Stenella euphrosyne
Pilleri
1970
Str. of Gilbraltar
phocaenoides
Bennet
(Euphrosyne
(Finless black
dolphin)
porpoise)
Tursiops truncatus
True
1891
N. Carolina
Phocoena phocoena
Stubbings
1965
Hann. Senegal
(Bottlenose
Dollfus
1968
Gibraltar
(Harbor porpoise)
dolphin)
Pilleri
1970
Gibraltar
Present study
1985
Porto Novo, India
Family: Ziphiidae
Mead and
1990
Central Atlantic
Potter
coast, USA
Mesoplodon mirus
Ross
1984
SE coast, S. Afr.
Ross
1984
Natal, S. Africa
(True's Beaked
Barnard
1924
Natal, S. Africa
whale)
Unidentified
Pope
1958
Heron I., Queensl.
Zipkius cavirostris
Bane and
1980
North Carolina
delphinid
Relini
1979
Ligurian Sea
(Goosebeaked
whale [or]
Zullo
Feresa attenuata
Stubbings
1965
Yenn, Senegal
Cuvier's Beaked
(Pygmy killer
whale)
whale)
Unidentified whale
Broch
1924
Greenland
Globicephala
Spivey
1977
Florida, Atlantic
Nilsson-Cantell
1930
W. Afr. (14°45'N;
macrorhynchus
coast
18°34'W)
(Short-finned
pilot whale)
Unknown host
Nilsson-Cantell
1978
Bay of Biscay
Acknowledgments
Bruce B. Collette, Director, NMFS Systematics Lab-
oratory, provided laboratory facilities and reviewed the
manuscript, and the Fish Division, USNM, provided
computer facilities. R. Natarajan and the authorities
of Annamalai University, Tamil Nadu, India, provided
facilities for this work. The senior author is grateful
NOTE Raiaguru and Shantha: Association between Xenobalanus globiapitis (Coronulidae) and Tursiops truncatus (Delphinidae) 201
to CSIR, New Delhi, for a Senior Research Fellowship
for the Marine Mammal Project. William A. Newman
verified the identification of X globicipitis. Drafts of
the manuscript were read by James G. Mead, William
F. Perrin, and Austin B. Williams.
Citations
Bane. G.W., and V.A. ZuUo
1980 (1981) Observations on a stranded goosebeaked whale
{Ziphim cavirostris, Cuvier, 1823) and its ectocommensal bar-
nacles (Xenobalanus globicipitis). J. Elisha Mitchell Sci. Soc.
96(l):l-3.
Barnard, K.H.
1924 Contributions .to the crustacean fauna of South Africa.
No. 7. Cirripedia. Ann. S. Afr. Mus. 20(1):1-103.
Barnes, R.D.
1986 Invertebrate zoology. W.B. Saunders, Philadelphia,
868 p.
Brocli, Hj.
1924 Cirripedia Thoracica von Norwegen und dem norwe-
gischen Nordmeere. Ein Systematik und biologischtier-
geographische Studie. Skr. Vidensk. Selsk. Christiana, mat.
naturv. KL, 17:1-121.
Caiman, W.T.
1920 A whale-barnacle of the genus Xenobalanus from Antarc-
tic seas. Ann. Mag. Nat. Hist., ser. 9, 6:165-166.
Cornwall, I.E.
1927 Some north Pacific whale barnacles. Contrib. Can. Biol.
Fish., Toronto, new ser., 3(23):501-517.
1955a Cirripedia. Canada Pacific Fauna, Ottawa, fasc. IOC,
49 p.
1955b The barnacles of British Columbia. B.C. Prov. Mus.,
Handbk. 7:1-69.
Darwin, C.H.
1854 A monograph on the sub-class Cirripedia. II. The Bala-
nidae, Verrucidae, etc. Ray Soc. Publ., 440 p.
Devaraj, M., and P.S. Bennet
1974 Occurrence of Xenobalanus globicipitis (Steenstrup) on
the finless black porpoise, Neomeris phocoenoides in Indian
seas. Indian J. Fish. 21(2):579-581.
DoUfus, R.Ph.
1968 Xenobalanus globicipitis Steenstrup (Cirripedia: Thora-
cica) recolte de la Cote Nord du Maroc. Bull. Inst. Peches
Marit. Maroc 16:55-60.
Gruvel, A.
1905 Monographic des Cirrhipedes ou Thecostraces. Masson
et Cie, Paris, 472 p.
1912 Sur le d^veloppement de la muraille chez le Xenobala-
nus globicipitis Steenstrup. C.R. Acad. Sci., Paris, 155:
1266-1268.
1920 Cirrhipedes provenant des campagnes scientifiques de
S.A.S. le Prince de Monaco (1885-1913). Result. Campagnes
Sci. accomp. sur son yacht par Albert I''', Prince Souverain
de Monaco, 53:1-88.
Heldt, J.H.
1950 Note au sujet de Xenabalamis globicipitis Steenstrup sur
Balaenoptera borealis Lessou en M^diterran^e. Bull. Soc.
Hist. Nat. Tunis 3(l):25-28.
Hoek, P.P.C.
1883 Report on the Cirripedia collected by H.M.S. Challenger
during years 1873-76. Rep. Sci. Res. Voyage H.M.S.
Challenger 8(25):1-169.
Kenney, R.D.
1990 Bottlenose dolphins off the northeastern United States.
In Leatherwood, S., and R.R. Reeves (eds.). The bottlenose
dolphin, p. 369-386. Academic Press. San Diego.
Mackintosh, N.A.
1942 The southern stocks of whalebone whales. Discovery
Rep. 22:199-298.
Mackintosh, N.A., and J.F.G. Wheeler
1929 Southern blue and fin whales. Discovery Rep. 1:
257-540.
Matthews, L.H.
1938 The sei whale, Balaenoptera borealis. Discovery Rep.
17:183-290.
Mead, J.G.. and C.W. Potter
1990 Natural history of bottlenose dolphins along the central
Atlantic coast of the United States. In Leatherwood, S., and
R.R. Reeves (eds.). The bottlenose dolphin, p. 165-195. Aca-
demic Press, San Diego.
Nilsson-Cantell, C.A.
1921 Cirripeden-Studien. Zur Kenntnis der Biologie, Anatomie
und Systematik dieser Gruppe. Zool. Bidr. Upps. 7:75-378.
1930 Thoracic cirripedes collected in 1925-27. Discovery Rep.
2:223-260.
1978 Cirripedia Thoracica and Acrothoracica. Marine Inverte-
brates of Scandinavia. Universistetes-forlaget, Oslo (5):1-135.
Pilleri, G.
1967 Behaviour of the Pseudorca crassidens (Owen) off the
Spanish Mediterranean coasts. Rev. Suisse Zool. 74:679-683.
1970 Xenobalanus globicipitis Steenstrup on Delphinus delphis.
Stenella styx, and Tursiops truncatus of the western Mediter-
ranean. In Pilleri, G. (ed.), Investigations on cetacea, vol. II,
p. 247-249. Himanatomisches Inst, der Univ., Berne,
Switzerland.
Pilleri, G., and M. Gihr
1969 On the anatomy and behaviour of Risso's dolphin (Gram-
pus griseus G. Cuvier). In Pilleri, G. (ed.), Investigations on
cetacea, vol. I, p. 74-93. Himanatomisches Inst, der Univ.,
Berne, Switzerland.
Pilleri, G.. and J. Knuckey
1969 Behaviour patterns of some Delphinidae observed in the
western Mediterranean. Z. Tierpsychol. 26:48-72.
Pilsbry, H.A.
1916 The sessile barnacles (Cirripedia) contained in the col-
lections of the U.S. National Museum, including a monograph
of the American species. Bull. U.S. Nat. Mus. 93, 366 p.
Pope, E.C.
1958 The barnacle, Xenobalanus globicipitis Steenstrup in
Australian Seas. Proc. R. Zool. Soc. N.S.W. 1956-57:159-161.
Raga, J. A., and E. Carbonell
1985 New dates about parasites on Stenella coeruleoalba
(Meyen, 1833) (Cetacea: Delphinidae) in the western Mediter-
ranean Sea. Invest. Cetacea 17:207-213.
Raga, J. A., and C. Sanpera
1986 Ectoparasites y epizoitos de Balaenoptera physalvs (L.,
1758) en aguas atlanticas ibericas. Invest. Pesq. (Barcelona)
50(4):489-498.
Raga, J. A., E. Carbonell, and M.A. Raduan
1982 Incidencias de parasites en los cetaceos varados en las
costas espanolas del Mediterraneo. Mem. Mus. Mar. Ser. Zool.
2(19):1-11.
Raga, J.A., M.A. Raduan, C. Blanco, and E. Carbonell
1983 Etude parasitologique du dauphin bleu et blanc Stenella
coeruleoalba dans la Mediterran^e occidentale. Rapp. P.-V.
Reun. Comm. Int. Explor. Sci. Mer Medit. 28(5):211-212.
202
Fishery Bulletin 90(1). 1992
Rappe, G.
1988 Een vondst van Xenobalanus ghbicipitis in de Noord-
zee? Strandvlo 8(2);100-101.
Rappe, G., and K. van Waerebeek
1988 Xenobalanus globicipitis (Crustacea: Cirripedia) on ceta-
ceans in the northeast Atlantic and the Mediterranean: A
review. In Evans, P.G.H. (ed.), European research on ceta-
ceans, p. 75-78. PrOc, Second annual conf. European Ceta-
cean Society, Troia, Portugal, 5-7 Feb. 1988. European
Cetacean Society, Lisbon.
Relini, G.
1979 Remarks on cirripeds of Ligurian Sea. Rapp. Comm.
Int. Mer. M6dit. 25/26(4):131-132.
Richard, J.
1936 Notes sur les c6tac6s et les pinnipedes. Result. Cam-
pagnes. Sci., Monaco 94:34-71.
Ross, G.J.B.
1984 The smaller cetaceans of the southeast coast of southern
Africa. Ann. Cape Prov. Mus. Nat. Hist., 15(2): 173-410.
Siciliano, S., L. Capistrano, and L. Lodi
1988 Xenobalanus globicipitis registrado em Sotalia sp. na bata
de Guanabara - Rio de Janeiro - Brasil. Reunion Trabajo
Especialistas MamLferos Acuaticos, America del Sur, 25-30
de Julio. (Unpubl. Abstract #III-CL-36). Projeto Baleia,
Jubarte, Rua Jos6 Higino 340, Ap. 102, Tijuca, Rio de Janiero,
RJ 20520, Brazil.
Spivey, H.R.
1977 Those tenacious travelers of Florida's Atlantic coast.
Fla. Nat. 50(6):6-10.
Steenstrup, J.
1852 Om Xenobalanus globicipitis, en ny Cirriped-Slaegt af
Coronulafamilien. Vidensk. Medd. Naturhist. Foren. (1852):
62-64.
Stubbings, H.G.
1965 West African Cirripedia in the collections of the Institute
FranQais d' Afrique Noire, Dakar, Senegal. Bull. Inst. Fr. Afr.
Noire 27, ser. A (3):876-905.
True, F.W.
1891 Observations on the life history of the bottlenose por-
poise. Proc. U.S. Nat. Mus. 13:197-203.
Weltner, W.
1897 Verzeichniss der bisher beschriebenen recenten Cir-
ripeden Arten. Arch. Naturgesch. 63:227-280.
Zullo, V.A.
1963 A preliminary report on systematics and distribution of
barnacles (Cirripedia) of the Cape Cod region. Publ. Mar. Biol.
Lab., Woods Hole, 33 p.
Lack of biochemical genetic and
morpiiometric evidence for discrete
stocl<s of IMorthwest Atlantic herring
Clupea harengus harengus
Susan E. Safford
U.S. Fish and Wildlife Service, P.O. Box 796 Turner's Falls, Massachusetts 01376
Present address. Graduate Center of Toxicology, 204 Funkhouser Building
University of Kentucky, Lexington, Kentucky 40506-0054
Henry Booke
us Fish and Wildlife Service, P.O Box 796, Turner's Falls, Massachusetts 01376
Historically, herring stock delinea-
tion has been based on spawning
site because herring are presumed
to return to their natal beds to
spawn (Sindermann 1979). For ex-
ample, Wheeler and Winters (1984)
have estimated homing fidelity of
spawning herring at 90%. Further-
more, some recognition of these
historic stocks has been achieved
through meristic studies (Anthony
1972, Parsons 1975, Cote et al.
1980), though these meristic dif-
ferences disappear after several
years, probably from environmen-
tal perturbations (Sindermann
1979). These means of defining a
stock imply genetic differentiation,
but do not measure it. A valid stock
definition such as that in Booke
(1981), "a species group, or popula-
tion, offish that maintains and sus-
tains itself over time in a definable
area," should include both genetic
and geographic isolation. However,
for managerial purposes it is often
useful to divide large groups of a
species into smaller groups, even if
genetic or permanent geographic
isolation cannot be demonstrated.
Managerial units have sometimes
been defined as stocks as in An-
thony (1972), "a group of fish that
remain sufficiently isolated so it can
be managed as a unit separate from
another one." A population can sub-
divide itself into discrete groups,
which can be individually managed
during the period of subdivision,
such as a spawning season, even if
these groups aren't genetically dif-
ferentiated. Therefore, the goal of
this study was to determine if the
two spawning groups investigated
constitute genetically differentiated
stocks, and whether these groups
could be identified either genotyp-
ically or phenotypically, regardless
of stock status, outside the spawn-
ing grounds.
The first objective was to deter-
mine if herring which spawn in two
geographically well-defined areas-
Trinity Ledge, Nova Scotia, and
Jeffries' Ledge, MA— constituted
separate stocks through the demon-
stration of genetic differentiation
by starch gel electrophoresis of en-
zymes. Electrophoretic studies on
herring, including specimens from
the two spawning grounds sampled
in the present study, have been
published (Komfield et al. 1981 and
1982, Grant 1981 and 1984, King
1984). However, lack of standar-
dization in technique, which has led
to differences in the number and
frequency of alleles at the same
locus in different studies, makes it
difficult to assess the true amount
of electrophoretic differentiation
among spawning groups. The sec-
ond objective was to determine if
these same groups of herring were
separable phenotypically, whether
or not genetic differences were
detected. Included in this objective
was the assessment of the temporal
stability of a set of phenotypic char-
acters measured over two years.
This was important, as most mor-
phometric and meristic studies
which have indicated that signifi-
cant phenotypic differences do exist
between spawning groups of her-
ring consist of only one year's data
(Parsons 1975, Cote et al. 1980,
Meng and Stocker 1984). The third
and most important objective was
to simultaneously measure the
amount of electrophoretic and mor-
phometric variation in the two
spawning groups. Simultaneous
performance of both kinds of anal-
yses, previously done only by Ry-
man et al. (1984) on Northeast
Atlantic herring, permits a better
understanding of the level of varia-
tion between herring spawning
groups.
Materials and methods
Sampling
Trinity Ledge (TL) fish were col-
lected on 31 August 1983 and 5 Sep-
tember 1984, and Jeffries' Ledge
(JL) fish on 1 November 1983 and
11 October 1984. All fish were
taken on spawning grounds (Fig. 1)
by commercial fishermen. The fish
were transported frozen or packed
in ice, and stored at -20°C for 1
week to 9 months imtil white mus-
cle tissue samples were excised. The
tissue samples were stored at
-80°C until analyzed electropho-
retically. A sample of 100 fish from
each collection (400 total) was ana-
lyzed electrophoretically. These
same fish were also analyzed mor-
phometrically, except for 50 TL fish
collected in 1983. Poor packing con-
ditions made these 50 fish difficult
Manuscript accepted 23 September 1991,
Fishery Bulletin, U.S. 90:203-210 (1992).
203
204
Fishery Bulletin 90(1), 1992
to measure accurately, so an ad-
ditional 50 herring were taken
from the remaining TL 1983
sample for the morphometric
analysis.
Morphometries
Measurements Initially, 25 mor-
phometric characters described
by Meng and Stocker (1984) in
their analysis of Pacific herring
were measured on 100 Atlantic
herring, 50 from each location,
from the 1983 sample. The mea-
surements followed Hubbs and
Lagler (1958). Standard length
(SL) was measured to the near-
est 0.5mm on a measuring board.
The other measurements were
taken with vernier calipers to the
nearest thousandth of an inch
and converted to millimeters.
Multivariate analyses were used
to determine if the groups were
different from one another and
which characters contributed to
these differences. To address
length bias, multiple analysis of
covariance, performed under the
MANOVA subroutine in the sta-
tistical package for the social
sciences, was used to remove the
effect of SL on the other vari-
ables (Sokal and Rohlf 1969,
Steel and Torrie 1980). The re-
sults of multiple analysis of vari-
ance of the adjusted measure-
ments versus spawning group,
performed under the same sub-
routine, showed that the two
groups were significantly differ-
ent from each other and iden-
tified eight characters whose
means were significantly dif-
ferent (Snedecor and Cochran
1967, Safford 1985) (Fig. 2). The
binomial distribution predicted
the probability of eight signifi-
cant characters out of 25 as 1.77
xlO"^, given a probability of
0.05 that a single character
would be significantly different
due to chance alone.
44'
42'
40'
44°N
66°E,
Figure 1
Atlantic herring spawning grounds in the Gulf of Maine region.
PVI
Figure 2
Eight morphometric measurements used to derive the discriminant function: MXL =
maxillary length; PCD = pectoral to dorsal fin; PVD = pelvic to dorsal fin; AD = anal
to dorsal fin; AH = anal fin height; INL = interorbital distance; PCI = distance be-
tween insertions of pelvic fins; PVI = distance between insertions of pectoral fins.
NOTE Safford and Booke Stock delineation of Clupea harengus harengus
205
Dfscrimlnant function These eight char-
acters were used to derive a discriminant
function (Snedecor and Cochran 1967, Sokal
and Rohlf 1969). Each measurement in sub-
sequent samples was adjusted to the SL of
the original sample to eliminate bias due to
differences in the SL. Details of the con-
struct of the discriminant function and the
formulae used to adjust the subsequent
measures can be found in Appendix A and
Safford (1985).
The discriminant function was tested for
spatial and temporal stability with addi-
tional samples from both 1983 and 1984.
The additional sample data were treated as
described in Appendix A to yield a z-score
so the fish could be classified according to
spawning group. The cut-off value for the
z-score was set at zero, where fish wnth a
z-score >0 were classified as Trinity Ledge
fish and those with a z-score <0 were classified as
Jeffries' Ledge fish (Norusis 1979, Safford 1985).
Statistics A stepwise function employing the F-value
of each character, (p<0.05), to accept or reject a char-
acter was derived to rank the variables. The distribu-
tion of phenotypic variation was measured by a nested
analysis of variance (ANOVA), with years nested
within groups, generated by nested procedures using
PC-SAS packaged programs (SAS 1985). One-way
ANOVA generated by the general linear models pro-
cedure in PC-SAS packaged programs (SAS 1985) was
used to analyze differences in morphometric measure-
ments, both between years within a spawming group
and between spawning groups within a year.
Electrophoresis
Enzyme visualization Traditional starch gel elec-
trophoresis of white muscle tissue samples as described
by Utter et al. (1974), with some modifications, was
used to resolve the enzymes. A detailed description of
the gel composition and running conditions can be
found in Safford (1985). Four polymorphic loci— phos-
phoglucomutase, PGM-2* (5.4.2.2), glucose-6-phos-
phate isomerase, GPI-2* (5.3.1.9), and two of lactate
dehydrogenase, LDH-1 * andLDH-2* (1.1.1.27)-were
analyzed. The enzyme abbreviations and numbers
follow the suggestions of Shaklee et al. (1989). Two
buffer systems, Ridgway et al. (1970) and Markert and
Faulhaber (1965), were used. The Ridgway gel buffer,
used for LDH and GPL was modified by doubling the
amount of Tris (Sigma Chemical Co., St. Louis) in the
recipe, which raised the pH to 8.5 and made the bands
more distinct. The Markert-Faulhaber buffer was used
Table 1
Means (mm), unadjusted for standard length, and 95% confidence inter-
vals of the eight morphometric characters used in the discriminant func-
tion derived from 100 Atlantic herring from the 1983
samples.
Jeffries' Ledge
Trinity Ledge
Character
(n50)
(w50)
Maxillary length (MXL)
27.90±0.57
31.0610.51
Pectoral to dorsal fin (PCD)
81.4611.98
93.4411.82
Pelvic to dorsal fin (PVD)
48.41 + 1.53
59.4311.63
Anal to dorsal fin (AD)
70.56 ±1.88
79.6211.59
Height of anal fin (AH)
14.3510.59
13.1310.47
Interorbital distance (INL)
10.3410.31
10.10l0.25
Distance between insertions
13.16l0.49
17.3710.71
of pectoral fins (PCI)
Distance between insertions
8.4110.31
10.6310.35
of pelvic fins (PVI)
Standard length (SL)
233.50118.03
251.50114.19
for PGM because it improved band resolution. Stain
recipes and techniques followed Shaw and Prasad
(1970) v^nth modifications which are detailed in Safford
(1985). Photographs were taken immediately upon
staining.
Statistics Allelic frequencies were compared between
samples by chi-square contingency table analysis.
Genotypic frequencies were tested for conformation to
the Castle-Hardy-Weinberg (C-H-W) equilibrium with
a chi-square goodness-of-fit test (Zar 1974). Gene diver-
sity analyses were conducted according to Nei (1973),
Chakraborty (1980), and Chakraborty et al. (1982).
Results
Morphometries
Group means of eight morphometric characters were
found to be significantly different (p<0.01) between
samples taken from the two spawning areas in 1983.
No overlap in range was found within the 95% con-
fidence interval for seven of these variables, and over-
lap at the eighth variable was very small (Table 1).
Therefore, it was concluded for this study that 50 fish
from each sample were sufficient. The stepwise dis-
criminant function accepted the first seven of these
eight variables. Multiple analysis of variance of the
eight characters versus locality for the 1984 samples
revealed that only three of these characters— distance
between insertions of the pelvic and dorsal fins (PVD),
anal fin height (AH), and distance between the inser-
tions of the pectoral fins (PCI)— were significantly dif-
ferent (p<0.01) between the two groups. Two of these
characters, PVD and PCL were among the three which
206
Fishery Bulletin 90(1). 1992
Table 2
Discriminant function analysis results of different Atlantic herring samples from known spawning grounds.
Sample
1983 discriminant function
construction sample
(N = N, + N2 = 100)
Jeffries' Ledge fNi = 50)
Trinity Ledge (N2=50)
1983 Sample (N = Ni + N2 = 100)
Jeffries' Ledge (Ni = 50)
Trinity Ledge (N2 = 50)
1984 Sample (N = Ni + N, = 198)
Jeffries' Ledge (Ni = 99)
Trinity Ledge (N, = 99)
Number from each spawning
ground classified as:
Percent from each spawning
ground classified as:
Jeffries' Ledge Trinity Ledge Jeffries' Ledge Trinity Ledge
Overall correct
classification
44
6
88%
12%
7
43
14%
86%
38
12
76%
24%
11
39
22%
78%
92
7
93%
7%
84
15
85%
15%
87%
77%
54%
Table 3
Phenotypic
variation of Atlantic herring in geographic
and temporal hierarchies for each of eight morphometric
:haracters. (*)p<
0.01; ( + )p< 0.001. See Figure 2 for definitions of morphometric characters.
Between
years within
Source of
variation:
Between spawning groups
a spawning group
Within
samples
Character
df square
component (%)
df
square
component (%)
df
square
component (%)
MXL
1 120.3*
0.0
2
597.3*
59.0
390
4.2
41.0
AH
1 38.2*
0.7
2
29.0*
3.3
390
6.6
96.0
PCD
1 2136.1*
0.7
2
7778.3*
58.5
390
55.7
41.5
PVD
1 2655.8*
0.7
2
4483.7*
57.1
390
34.0
42.9
AD
1 1379.7
0.0
2
4475.3
53.4
390
41.9
46.6
INL
1 0.0
0.0
2
131.0*
63.6
390
0.6
36.4
PCI
1 282.3*
0.0
2
383.9*
46.5
390
4.4
53.5
PVI
1 45.8*
0.0
2
136.2*
46.3
390
1.6
53.7
accounted for most of the between-group variation in
the 1983 sample. The percent correct classification by
spawning group of three sets of samples (two from
1983, one from 1984) separated by the derived discrimi-
nant function is found in Table 2. Overall misclassifica-
tion of fish collected in 1983 was 18%, while that of
fish collected in 1984 was 46%.
The phenotypic variation of the unadjusted measure-
ments was partitioned similarly within each morpho-
metric character, except AH (Table 3). The partition-
ing of the phenotypic variation averaged across all
characters is found in Table 4. None of the variation
was explained by differences between spawning
groups, while approximately one-half was partitioned
within a spawning group between years. The remain-
der of the variation was within a sample. One-way
ANOVA of between-year differences within a spawn-
ing group showed that within the TL group the means
of all the characters, except AH (;?< 0.02), were highly
significantly different (p< 0.0001) between 1983 and
1984. In contrast, within the JL group three char-
acters—distance between the insertions of the pelvic
fins (PVI), AH, and PCI— were not significantly dif-
ferent between years. The remaining characters were
significantly different (p<.05) between years.
Electrophoresis
Allelic frequencies within each sample for the four loci
chosen for analysis are found in Table 5. Other enzyme
systems were also investigated, but few specimens ex-
pressed enzyme activity at these loci (Safford 1985).
We chose these loci because they had previously been
shown to be polymorphic and to follow Mendelian in-
NOTE Safford and Booke: Stock delineation of Clupea harengus harengus
207
heritance in herring (Grant 1981, Kornfield et al. 1981
and 1982, King 1984). The designation of alleles is
taken from Kornfield et al. (1982) and is based on direc-
tion of migration of the enzymes on the gel and their
distance from the origin. LDH* is encoded by three
anodally migrating loci. The fastest locus, LDH-3 * , was
present only in eye tissue and activity was found in only
some of the fish, so this locus was not used in the pres-
ent analysis. Each of the other two loci were repre-
sented by two alleles. PGM* is encoded by two anod-
ally migrating loci. The slower locus, PGM-1*, was
fixed for the same allele in all samples. The polymor-
phic locus, PGM-2*, was represented by three alleles.
However, the slowest allele was found in only one
specimen. GPI* is encoded by one anodally migrat-
ing locus, GPI-2*, which was represented by five alleles
in our samples.
Chi-square contingency table analysis revealed that
GPI-2* was significantly different (p<0.05) between
JL 1984 and both TL 1984 and JL 1983. This was pro-
bably due to a greater frequency of allele 150 in the
JL 1984 sample. Its frequency
was double (0.18) that found in
the JL 1983 sample (0.09) and
the TL 1984 sample (0.10). The
chi-square goodness-of-fit test
showed that the JL 1984 sample
was not in C-H-W equilibrium at
the GPI-2* locus, which con-
tained an excess of 150/ -3 het-
erozygotes. The gene-diversity
analysis results for the individual
loci were similar within each
locus, and revealed that more
than 99% of the genetic diver-
sity was found within a single sample. A comparison
between the partitioning of the average gene diversity
and the average phenotypic variation is showm in Table
4. The large between-year phenotypic variation is not
reflected in the between-year genetic diversity index,
as <1% of the gene diversity can be explained by
between-year differences within a spawning group.
Discussion
In-depth discussions of the historical construct of her-
ring stocks and the implications of recent electropho-
retic findings can be found in Jorstad and Naevdal
(1981), Smith and Jamieson (1986), and Kornfield and
Bogdanovricz (1987). The traditional herring stock con-
struct has not been supported by genetic stock struc-
ture analyses as none of the electrophoretic studies,
including the present one, have found a large amount
of genetic differentiation (Andersson et al. 1981, Grant
1981 and 1984, Jorstad and Naevdal 1981 and 1983,
Table 4
Distribution of relative gene diversity and phenotypic variation of Atlantic herring in
geographic and temporal hierarchies. Gene diversity is based on four individual or pooled
samples and four polymorphic loci. Phenotypic variation is averaged over 393 individuals,
four individual or pooled samples, and eight morphometric characters.
Gene diversity
Absolute %
Phenotypic variation
Source of variation
df
Mean Variance
square component (%)
Between spawning groups
Between years within
spawning groups
Within samples
0.00024
0.00074
0.20138
0.1
0.4
99.5 390
823.28
2289.34
18.63
0.09
48.46
51.45
Allelic
Table 5
frequencies of samples of Atlantic herring.
Loci and alleles
LDH-1
LDH-2
PGM-2
GPI-2
Location/Year *100
0
100
72
112
100
95
150
100
40
-3
-75
Jeffries' Ledge (A^)
1983 (100) 0.980
1984 (100) 0.985
Trinity Ledge (N)
1983 (100) 0.965
1984 (100) 1.000
0.020 0.955 0.045
0.015 0.955 0.045
0.035 0.955 0.045
0 0.945 0.055
ate direction of migration ( +
0.025
0.055
0.065
0.025
anodal,
0.975
0.945
0.925
0.975
- cathodal),
0 0.090 0.630
0 0.180 0.540
0.005 0.110 0.560
0 0.100 0.630
and relative distance from
0.140 0.120 0.020
0.105 0.160 0.015
0.210 0.115 0.005
0.120 0.130 0.020
origin (the farther away, the
* Allelic designations indie
larger the number).
208
Fishery Bulletin 90(1). 1992
Kornfield et al. 1982, King 1984, Ryman et al. 1984).
The conclusion that herring spawning groups are not
discrete genetically -distinct stocks is further supported
by the results of a recent study by Kornfield and
Bogdanowicz (1987). They investigated the genetic
relationships of ripe female herring from three loca-
tions, including Jeffries' Ledge and some of the 1983
Trinity Ledge samples analyzed in this study, by
restriction endonuclease analysis of mitochondrial
DNA (mtDNA). In other species, this technique has
revealed genetic differentiation not uncovered by tradi-
tional enzyme electrophoresis (A vise et al. 1986). Korn-
field and Bogdanowicz (1987) found that these spawn-
ing groups were not completely distinguished by the
composite mtDNA digestion patterns generated, and
no consistent geographic patterns were found for the
unique composites. Therefore, they concluded that this
approach also provided no evidence for the existence
of genetically distinct stocks in the Gulf of Maine.
The significant departures from C-H-W equilibrium
found in this and previous studies (Grant 1981, Ryman
et al. 1984) may be considered contradictory to the
hypothesis of the existence of a genetically homogenous
herring population. However, these departures seem
to be a feature of pelagic fish stocks (Smith et al. 1989).
These disequilibria have been variously attributed to
chance due to the low frequency of occurrence (Grant
1981, Ryman et al. 1984) and assortative mating (Smith
et al. 1989). The significant departure in the present
data has derived from an excess number of hetero-
zygotes of one particular allelic combination in the JL
1984 data. One significant departure in 16 tests is
slightly higher than would be expected by chance alone
at the 5% probability level. An excess of heterozygotes
can result from negative assortative mating; however,
the data are not sufficient to support that hypothesis.
Importantly, the C-H-W equilibrium applies to all gen-
erations in a population, thus significant departures
may occur if sampling does not measure all generations
in the same proportion in which they occur in the
population. Based on SL, few immature and old fish
were included, so this sample bias may have con-
tributed to the significance level. Thus, the departure
from C-H-W equilibrium is probably due to chance and
perhaps some sampling bias. However, the distribution
of alleles across generations within a population may
warrant further investigation as disequilibrium, though
explicable, is a feature of herring populations and some
age-based selection may be occurring.
Although the genetic evidence argues for a single
population of herring, significant phenotypic differ-
ences between spawning groups have been demon-
strated (Parrish and Saville 1965, Burd 1969, Anthony
1972, Cote et al. 1980, Ryman et al. 1984). Morpho-
metric and meristic characters, which have a complex
underlying genetic structure, are believed to be great-
ly influenced by environmental parameters (Sinder-
mann 1979, Ryman et al. 1984). Thus phenotypic dif-
ferences may not reflect genetic differentiation, and
small but detectable genetic differences may not sig-
nificantly alter phenotypic characters. Differences in
biochemical genetic and phenotypic variation can best
be demonstrated when genetic and phenotypic analyses
are performed on the same specimens. In their study,
Ryman et al. (1984) screened 17 loci from herring
caught in 17 locations ranging from the Gulf of Bothnia
to the northeast Atlantic off Norway's western coast,
and found significant allelic heterogeneity at only 4 loci.
They concluded that the results resembled those of
samples drawn from a single breeding population, as
both the genetic diversity index and genetic distances
were very small. They chose numbers of vertebrae and
keeled scales as morphological characters. Morpho-
logical distances were used to construct a dendogram
which differentiated herring in central Baltic fall
spawning groups from a spring spawning Baltic group
and the other fall spawning groups. Thus these meristic
characters differed to some extent despite genetic
similarities. Morphologic variation was partitioned by
nested ANOVA with localities nested within larger geo-
graphic areas, and genetic variation was partitioned
by genetic diversity analysis. They found over 99% of
the gene diversity within a locality, compared with 50%
of the phenotypic variation. Most important, <1%
of the gene diversity was explained by between-geo-
graphic-group differences, while these differences
explained 40% of the phenotypic variation.
The partitioning of variance in our samples was
similar in many respects to that of Ryman et al. (1984).
Over 99% of the genetic variance in our samples also
occurred within a locality within a year, compared with
approximately 50% of the morphometric variance com-
ponent. However, the percent of the morphometric
variance component explained by differences between
spawning groups was similar for both the genetic and
morphometric components (0.1%), in contrast to the
large between-group morphometric variation found by
Ryman et al. (1984). Results from both these studies
demonstrate that most genetic diversity lies within a
single locality at one point in time, further supporting
the hypothesis that herring form a single panmictic
population. Thus the current situation seems to be that
despite the existence of discrete, defined spawning
groups and apparent high homing fidelity, enough gene
flow exists between spawning groups to prevent North-
west Atlantic herring from evolving into genetically
distinct stocks. Alternatively, herring may have begun
this process in recent geographic time, so that genetic
differences have not had time to evolve. This lack
of genetic differentiation also means that observed
NOTE Safford and Booke: Stock delineation of Clupea harengus harengus
209
phenotypic differences are most likely due primarily
to differences in environmental conditions during
development, and therefore will not be reliable in-
dicators of stock identity. Further, if all measurable
phenotypic characters are distributed similarly to those
in the present study and Ryman et al. (1984), then the
use of phenotypic characters to distinguish herring
groups may be proscribed, as the large within-group
variation would mask the subtler between-group
differences.
These ideas need to be incorporated into current her-
ring management policy. The results show that in-
dividuals from discrete spawning groups can not be
reliably identified off the spawning grounds. Therefore,
the contribution of each spawning group to various
fisheries cannot be estimated. These results also sug-
gest that the demise of a single spawning ground will
not adversely affect the underlying genetic structure
of the herring population, as few unique genes should
be found exclusively within a spawning location. How-
ever, small discrete spawning grounds are apparently
necessary to support a large population. Small spawn-
ing grounds may be necessary for appropriate spawn-
ing behavior or to ensure proper conditions for the
larvae. Therefore, until the relationship between dis-
crete spawning grounds and a healthy herring popula-
tion is understood, management policy should include
the maintenance of existing spawning grounds.
Acknowledgments
The authors thank Pamela Mace (Fisheries and Oceans,
Canada) and the captain and crew of the FV Bamegat
(Gloucester, MA) for field collection assistance, as well
as three anonymous reviewers whose comments
strengthened the paper. This work was supported by
a two-year stipend to S. Safford from the Massachu-
setts Division of Marine Fisheries and additional funds
from the Massachusetts Division of Fisheries and
Wildlife and The Masssachusetts Co-operative Re-
search Unit of Fisheries and Wildlife Biology.
Citations
Andersson, L., N. Ryman, R. Rosenberg, and G. Stahl
1981 Genetic variability in Atlantic herring {Clupea harengiis
harengus): Description of protein loci and population data.
Hereditas 95:69-73.
Anthony, V.
1972 Population dynamics of the Atlantic herring in the Gulf
of Maine. Ph.D. diss., Univ. Wash., Seattle, 266 p.
Avise, J.C, G.S. Helftnan, N.C. Saunders, and L.S. Hales
1986 Mitochondrial DNA differentiation in North Atlantic
eels: Population consequences of an unusual life history pat-
tern. Proc. Natl. Acad. Sci. 83:4350-4354.
Booke, H.E.
1981 The conundrum of the stock concept— Are nature and
nurture definable in fishery science? Can. J. Fish. Aquat. Sci.
38:1479-1480.
Burd, A.C.
1969 Trials with principal component analysis for herring racial
studies. Int. Counc. Explor. Sea CM 1969/30, Pelagic Fish
Comm., 11 p.
Chakraborty, R.
1980 Gene-diversity analysis in nested subdivided populations.
Genetics 96:721-726.
Chakraborty, R., M. Haag, N. Ryman, and G. Stahl
1982 Hierarchical gene diversity analysis and its application
to brown trout population data. Hereditas 97:17-21.
Cote, G., P. Lamoureux, J. Boulva, and G. Lacroix
1980 Separation des populations de hareng de I'Atlantique
(Clupea harengus harengus) de I'estuaire du Saint-Laurent et
de la peninsule gaspienne. Can. J. Fish. Aquat. Sci. 37:66-71.
Grant, W.S.
1981 Biochemical genetic variation, population structure, and
evolution of Atlantic and Pacific herring. Ph.D. diss., Univ.
Wash., Seattle, 135 p.
1984 Biochemical population genetics of Atlantic herring,
Clupea harengus. Copeia 1984(2):357-364.
Hubbs, C, and K. Lagler
1958 Fishes of the Great Lakes region. Bull. 26, Cranbrook
Inst. Sci., Bloomfield Hills, MI, 186 p.
Jorstad, K., and G. Naevdal
1981 Significance of population genetics on management of
herring stocks. Int. Counc. Explor. Sea CM 1981/4:64.
Jorstad, K.E., and G. Naevdal
1983 Genetically distinct populations of herring. Contrib. XV
Int. Cong. Genetics, New Delhi, India.
King, D.P.F.
1984 Genetic analysis of enzyme polymorphisms in herring.
Heredity 52:121-131.
Kornfield, I., and S.M. Bogdanowicz
1987 Differentiation of mitochondrial DNA in Atlantic herring,
Clupea harengus. Fish. Bull., U.S. 85(3):561-568.
Kornfield, L, P. Gagnon, and B. Sidell
1981 Inheritance of allozymes in Atlantic herring. Can. J.
Genet. Cytol. 23:715-720.
Kornfield, I., B. Sidell, and P. Gagnon
1982 Stock definition in Atlantic herring: Genetic evidence for
discrete fall and spring spawning populations. Can. J. Fish.
Aquat. Sci. 39:1610-1621.
Markert, C.L., and I. Faulhaber
1965 Lactate dehydrogenase patterns in fish. J. Exp. Zool.
159:319-332.
Meng, H.J., and M. Stocker
1984 An evaluation of morphometries and meristics for stock
separation of Pacific herring (Clupea harengus pallasi). Can.
J. Fish. Aquat. Sci. 41:414-422.
Nei, M.
1973 Analysis of gene diversity in subdivided populations.
Proc. Natl. Acad. Sci. 70(12)Pt. 1:3321-3323.
Norusis, M.J.
1979 SPSS. Statistical algorithms. Release 8.0. SPSS, Inc.,
Chicago.
Parrish, B.B., and A. Saville
1965 The biology of the Northeast Atlantic herring populations.
Oeeanogr. Mar. Biol. Annu. Rev. 3:323-373.
Parsons, L.S.
1975 Morphometric variation in Atlantic herring from New-
foundland and adjacent waters. Int. Comm. Northwest Atl.
Fish. Res. Doc. 11:73-92.
210
Fishery Bulletin 90(1). 1992
Ridgway, G.J., S.W. Sherburne, and R.P. Lewis
1970 Polymorphism in the esterases of Atlantic herring.
Trans. Am. Fish. Soc. 99(1):147-151.
Ryman, N., U. Lagercrantz, I. Andersson, R. Chakraborty, and
R. Rosenberg
1984 Lack of correspondence between genetic and morpho-
logical variability patterns in Atlantic herring (Clupea haren-
yiis). Heredity (53)3:687-704.
Safford, S.E.
1985 Stock identification through electrophoretic and mor-
phometric analyses of two spawning populations of Northwest
Atlantic herring {Clupea harengus harengus, Linnaeus).
Master's thesis, Univ. Mass., Amherst, 111 p.
SAS
1985 SAS/STAT guide for personal computers, Version 6
Ed. SAS Inst., Inc., Gary, NC.
Shaklee, J.B., F.W. Allendorf, D.C. Morizot, and G.S. Whitt
1989 Genetic nomenclature for protein-coding loci in fish; Pro-
posed guidelines. Trans. Am. Fish. Soc. 118:218-227.
Shaw, C.R., and R. Prasad
1970 Starch gel electrophoresis of enzymes: A compilation of
recipes. Biochem. Genet. 4:297-320.
Sindermann, C.J.
1979 Status of Northwest Atlantic herring stocks of concern
to the United States. Tech. Ser. Rep. 23, Sandy Hook Lab.,
NMFS, NOAA, Highlands, NJ, 449 p.
Smith, P.J., and A. Jamieson
1986 Stock discreteness in herrings: A conceptual revolution.
Fish. Res. 4:223-234.
Smith, P.J., A.J. Birley, and A. Jamieson
1989 Electrophoretic studies and the stock concept in marine
fishes. J. Fish. Biol. 35(Suppl. A):345-346.
Snedecor, G.W., and W.G. Cochran
1967 Statistical methods, 6th ed. Iowa State Univ. Press,
Ames, 593 p.
Sokal, R.R., and F.J. Rohlf
1969 Biometry. W.H. Freeman, San Francisco, 776 p.
Steel, R.G.D., and J.H. Torrie
1980 Principles and procedures of statistics, biometrical ap-
proach, 2d ed. McGraw-Hill, NY.
Utter, P.M., H.O. Hodgins, and F.W. Allendorf
1974 Biochemical genetic studies of fishes: Potentialities and
limitations. In Malins, D.C., and J.R. Sargent (eds.). Bio-
chemical and biophysical perspectives in marine biology. Vol.
1, p. 213-238. Academic Press, NY.
Wheeler, J. P., and G.H. Winters
1984 Homing of Atlantic herring in Newrfoundland waters as
indicated by tagging data. Gan. J. Fish. Aquat. Sci. 41:
108-117.
Zar, J.H.
1974 Biostatistical analysis. Prentice-Hall, Englewood Cliffs,
NJ, 620 p.
Appendix A
The values of the eight morphometric characters were
Hnearly combined to form a single value, a z-score,
which was used to classify an individual into a group
(Sokal and Rohlf 1969, Norusis 1979). The general for-
mula for a z-score is
Z = ClXl + C2X2 + ... + CnXn + e
where each C is the unstandardized canonical discrimi-
nant function coefficient for each character (Norusis
1979).
The general formula used to adjust each measure-
ment was
Rv = Vn - bo - b,*SL
where Ry = adjusted measure,
Vn = original measure,
bo = intercept,
bj = slope of the univariate covariance equa-
tion, with
SL = 242.5 (mean SL of original sample).
Variability of monthly catches of
anciiovy Engraulis encrasicolus
in tiie Aegean Sea
Konstantinos I. Stergiou
National Centre for Marine Research
Agios Kosmas, Hellinlkon, Athens 16604, Hellas
In a recent paper, Stergiou (1990a)
showed that the autoregressive
terms of an ARIMA model describ-
ing the monthly fishery of the an-
chovy Engraulis encrasicolus in
Hellenic waters indicated a 2- to
3-year periodicity in catches. A
similar cycle has also been shown
for anchovy in the Azov (Demen-
t'eva 1987) and Adriatic Seas (S.
Regner 1985). In comparison, long-
term periodicities have been shown
for Engraulis mordax off California
(Soutar and Isaacs 1974). In this
present study, I examined the vari-
ability of the Hellenic monthly
catches of anchovy during the
period 1964-87 using spectral
analysis.
The purse-seine fishery landed
51,282 t of fish which comprised
49% of the total Hellenic catch in
1987 (Stergiou 1990b). Anchovy
comprised 46.6% of the 1987 purse-
seine catch; the remainder included
sardine Sardina pilchardus, horse
mackerel Trachurus sp., bogue
Boops boops, chub mackerel Scom-
ber japonicus, and bonito Sarda
sarda (Stergiou 1990b). Ninety per-
cent of the mean annual anchovy
catch (1964-85) was caught in the
northern, northwestern, and west-
ern Aegean Sea (Stergiou unpubl.
data; no data are available for
monthly catches per major fishing
area). A recent study of genetic
distances, based on electrophoretic
o
o
o
3 -
B 2
Jan 1964 Jan 1969 Jan 1974 Jan 1979
Month
Jan 1984 Dec 1987
Figure 1
Monthly commercial catches (in 1000 tons) of anchovy Engraulis encra-
sicolus in Hellenic waters, 1964-87.
variation, and of morphometric and
meristic characters using multivari-
ate analysis does not indicate sep-
arate stocks of anchovy (or sardine)
in the Aegean Sea (Spanakis et al.
1989).
Monthly catches of anchovy (1964-
87, 288 data points) and annual fish-
ing effort (in horsepower, HP) of
pelagic seiners were gathered from
the Bulletins of the Hellenic Na-
tional Statistical Service (1968-89).
The monthly series was log-trans-
formed and detrended to become
stationary. The seasonal component
was removed by differencing with
lag=12 (Chatfield 1984). To avoid
a discontinuity at the end of the
data, the resulting series was tap-
ered by 20%. The Fast Fourier
Transform was used to compute
power spectral estimates, and
smoothed (5 moving averages)
squared amplitudes of the sinusoids
were plotted.
Anchovy catches show a marked
seasonal pattern (Fig. 1) and an
increasing trend for the years fol-
lowing 1980. The increased trend
in catch in recent years has raised
concern about whether these high
catches are sustainable. Due to
higher prices of anchovy since the
late 1970s, purse-seine fishing in
Hellenic waters is anchovy -oriented
rather than sardine-oriented (Ster-
giou 1986a, 1990a, b). Monthly fish-
ing effort by pelagic seiners is not
available. However, annual fishing
effort of pelagic seiners increased
considerably between 1964 and
1987 (from 363 boats, 20,316 HP,
and 6152 tonnage of boats, to 502
boats, 112,310 HP, and 18,922 ton-
nage of boats; Hellenic Natl. Stat.
Serv., 1968-89). Annual catches of
anchovy are highly positively cor-
related with annual horsepower of
the pelagic seiners [Ln(annual
catch) = 8.25 -H0.000013HP; n 24, r
-1-0.89, p<0.001)(Fig. 2), indicating
that the highly significant linear
Manuscript accepted 26 July 1991.
Fishery Bulletin, U.S. 90:211-215 (1992).
21 I
212
Fishery Bulletin 90(1). 1992
10.3
9.9
ns
o
15
c
§ 9.1
8 7
8.3
; ' ' ' — [ ' ' ' T ' ' ' i ' ' ' 1 ' ' ' I ' ' '
Ln(a«ual catch)^. 25^ 000013 *
-
. r.0.89. n.24 ,-• •
■
--■'"' •••V
•
''' /^^ -''
-
.'/,-•
. ■
.//.••
■
-' Z^'
_
..--■ X -•/>>-' • .--'
■
.--' •^■y--' •-••■
-•-'•-<•'' • ,.--'
..■■■'• // ,•'
/ ,-•
-
2 4 6 8 10
Fishing Effort in 10000 HP
12
Figure 2
Regression line (solid) and 95% and 99% confidence limits
(dotted) of the log annual commercial catch of anchovy and
fishing effort in horsepower (HP), in Hellenic waters, 1964-87.
3
_ '
i 6yr
sinusoids
r 1 9yr
-
Squared amplitude of
:
1/1
[1
-
0
I .... 1 ... 1 1 , ... 1 .
,vA^-^_
0 0.1 0.2 0.3
0.4 0.5
Cycles/montti
Figure 3
Spectrum of logged/detrended seasonally corrected monthly
anchovy catches.
trend in monthly catches [Ln(monthly catch) = 5.38 +
0.0053T, n 288, r 0.38, p<0.01, where T = 1-288] is
most likely attributed to increased fishing effort.
The spectrum of the resulting series (not shown
here), which may be postulated to be free of any an-
nual changes in effort, revealed a large major peak at
12 months (frequency 0.0833). This marked seasonal
pattern is most likely related to the seasonal offshore-
inshore migrations of anchovy and the nature of the
purse-seine fishery (Stergiou 1990a). Purse-seine fish-
ing in Hellenic waters does not occur in the open sea
but is mainly restricted to coastal areas where schools
of anchovy migrate seasonally. The anchovy starts its
inshore migration in early spring, but peak abundance
occurs in coastal waters in May-August. Offshore
migration probably occurs in late summer-fall.
The smoothed spectrum of the seasonally corrected
and detrended series (Fig. 3) reveals a prominent peak
at 4.6 years (frequency 0.018) and a probable second-
ary peak at 1.9 years (frequency 0.043) (95% confidence
intervals of the spectrum for 10 df: 0.4882-3.0798
squared amplitude of sinusoids). In contrast, non-
sinusoidal periodic variability generates harmonics with
■periods of less than 1 year (Fig. 3).
Cycles of 2-3 and 4-5 years have also been identified
in the air temperature in the northern (Thessaloniki)
and western Aegean (Athens) (Table 1) and in different
biotic (zooplankton, phytoplankton, fish eggs/larvae,
fish) and abiotic variables (air temperature/pressure,
sea temperature/salinity) in different areas of the
Mediterranean, Black, and Azov Seas (Table 1). These
cycles have also been suggested for annual anchovy
catches and eggs/larvae, temperature, salinity, and
zooplankton in the Adriatic Sea but the data set is
limited (annual, 1962-76) and the cycles may not be
statistically significant (D. Regner 1985). Correlations
have been found between biotic/abiotic variables (pri-
mary production, zooplankton biomass, winds, river
flow, air/sea temperature, salinity) and various abun-
dance indices of the Mediterranean anchovy (Azov-
Black Sea: Dement'eva 1987, Dekhnik and Rass 1988,
Porumb and Marinescu 1979; Hellenic waters: Ster-
giou 1986b; Adriatic Sea: S. Regner 1985; western
Mediterranean: Palomera and Lleonart 1989) and
other species oiEngraulis (see Bakun 1985).
Cycles with periods of 2-4 and 4-7 years have also
been identified in the physical environment and marine
populations in other areas of the world (e.g., Kort 1970,
Shuntov et al. 1981, Colebrook and Taylor 1984, Mysak
1986). Such cycles have frequently been related to
short-term ocean-atmosphere interactions (e.g., surface
heat-exchange phenomena: Zupanovich 1968, Cole-
brook and Taylor 1984; advection: Kort 1970, Mysak
1986).
A comprehensive discussion of the mechanisms
underlying such variability requires adequate biological
NOTE Stergiou: Monthly catch variability of Engraulis enaasicolus
213
Table 1
Cycles identified in the variability of
various climatic and
biological parameters in
the Mediterranean-Black Sea ecosystem. T = type 1
of data (A = annual,
M = monthly,
D =
daily); P = time perioc
of available data; Me =
method of analysis of data (sa = spectral
analysis, acf = autocorrelation function,
cgm = composite graphic method
, i.e.
comparison of graphs; see Dement'eva 1987).
Variable
Area
T
P
Me
Cycles
in years
Source
Sardine catch*
Adriatic
A
1853-1960
sa
2.3
3-3.5
Zupanovic 1968
Regner and Gacic 1974
Sardine catch
Hellas
M
1964-1982
sa
3.3
Stergiou 1988
Anchovy catch
Hellas
M
1964-1987
sa
1.9
3.3
4.6
This study
Anchovy catch*
Azov Sea
A
1955-1981
cgm
2-3
Dement'eva 1987
Carp catch*
Hellas^
A
1947-1983
sa
3-4
Economidis et al. 1988
Perch catch*
Hellas"
A
1947-1983
sa
3-4
Economidis et al. 1988
Copepods
Adriatic
M
1970-1974
acf
2-3
D. Regner 1985
Fish larvae
Adriatic
M
1971-1977
acf
3
S. Regner 1982
Fish eggs
Adriatic
M
1970-1974
acf
2-3
D. Regner 1985
Primary production
Adriatic
M
1970-1974
acf
2-3
D. Regner 1985
Diatoms*
Black Sea
A
1954-1987
sa
2.9
4.5
Petrova-Karadjova and Apostolov 1988
Air temp.*
Hellas"
A
1892-1981
sa
2.2-2.3
4
Flocas and Giles 1984
Air temp.*
Hellas'
A
1859-1981
sa
2.2-2.3
4
Flocas and Giles 1984
Air temp.*
Trieste
A
*
sa
2-2.9
4
Polli 1955
Air pressure*
Trieste
A
*
sa
2.3
4
Polli 1955
Air pressure*
Venice
A
♦
sa
2.1-2.8
4
Polli 1955
Sea surface temp.
Adriatic
M
1970-1974
acf
2-3
D. Regner 1985
Sea surface temp.
Monaco
D
1946-1961
sa
1.8
4.4
Bethoux and Ibanez 1979
Sea surface salinity
Adriatic M 1970-1974 acf 2-3
s of 8-12 years (frequently related to the 11 -year eye
e in
sunspot
D. Regner 1985
number, e.g., Gnevyshen and 01' 1977).
* Together with cyck
" Lake Koronia, ""Thessaloniki in summer, 'Athens in summer.
and physical oceanographic information, probably on
time scales of a few days and spatial scales of < 1 km
{sensu Leggett 1986). This information is not current-
ly available. The distribution and biology of larval,
juvenile, and adult anchovy and larval dispersal
patterns have not been studied in Hellenic waters.
However, some preliminary, conjectural discussion is
presented here.
The anchovy spawning season in the eastern Mediter-
ranean extends from April to September with a peak
in the summer months (Demir 1965, S. Regner 1985).
Anchovy larvae and postlarvae occur in the plankton
between May and September with a peak in July-
September (S. Regner 1985). This corresponds to the
predictable period of the etesians winds. These dry
northern, northeastern, and eastern winds blow each
year over the Aegean Sea from the end of May until
the end of October with a maximum frequency in July-
August (Fig. 4; Carapiperis 1962, Mariopoulos 1961).
Since anchovy spawning in the eastern Mediterranean
does not seem to be affected by abiotic factors such as
temperature or salinity (Demir 1965, S. Regner 1985),
the summer spawning habit of anchovy may represent
an important adaptation to the highly oligotrophic con-
ditions of the stratified coastal Aegean waters in sum-
mer. By spawning in summer, anchovy larvae (1) do
winds
Wl
1 -
J ^\.
days with
u
/ \
"o
/ N,
S 2
/\/ \
2
^
1 3 5 7 9 11 13 15 17
May Jun Jul Aug Sept Oct
10-<tay periods (1-1-10 May)
Figure 4
Mean number of days with etesians winds in Athens for May-
October, 1893-1960 (data from Carapiperis 1962).
not compete with sardine larvae which occur in the
plankton mainly in winter and spring (Yannopoulos
1977, Daoulas and Economou 1986, Regner et al. 1987),
and (2) are released in a relatively food-rich environ-
ment due to the effect of the etesians winds. The in-
creased frequency and intensity of the etesians winds
214
Fishery Bulletin 90(1). 1992
over the Aegean Sea in July- August when they fre-
quently reach gale force (Carapiperis 1962) would
probably deepen the mixed layer, and hence entrain
nutrient-rich water from below the thermocline. Mullin
et al. (1985) have shown that microzooplankton biomass
and chlorophyll a levels can be doubled after wind-
related events. In addition, an increase in the frequency
and intensity of etesians winds may also result in an
intensification of upwelling in the northern, northeast-
em, and eastern part of the Aegean Sea (Metaxas 1973,
Theocharis et al. 1988). Hence, periods dominated by
higher-than-average frequency of etesians in July-
August may be associated with favorable feeding con-
ditions for anchovy larvae which may be subject to
lesser mortalities through starvation and predation, the
main factors affecting larval mortality in Mediterra-
nean anchovy (Azov-Black Sea: see Dekhnik and Rass
1988 for a review; Adriatic Sea: see S. Regner 1985
for a review; western Mediterranean: Palomera and
Lleonart 1989).
Other factors may also affect variability in the an-
chovy abundance. For example, climatically-mediated
long-term changes in production and plankton species
composition in the eastern Mediterranean, changes in
larval dispersion due to changing patterns of currents,
as well as other factors, intrinsic or extrinsic, may
affect the egg/larval/postlarval/juvenile phases. It has
been maintained that in periods of increased air pres-
sure gradient over the eastern Mediterranean, the
water exchange between its basins intensifies (Pucher-
Petkovic et al. 1971, Vucetic 1981). As a result, the
salinity, nutrient content, temperature, and primary
productivity of the Adriatic Sea and of the eastern
Mediterranean basin rise, and the species composition
of the phytoplankton community changes. These
changes were accompanied by changes in the total
biomass of small pelagic fish (sardine, anchovy, horse
mackerel, etc). Such climate-plankton-small pelagic fish
interactions in the eastern Mediterranean involve time
lags of 2-3 years (Pucher-Petkovic et al. 1971). Last-
ly, cycles in anchovy catches may also be the result of
social-economic factors (Stergiou 1991) and/or a change
in the anchovy availability to purse seiners (changes
in the distribution and/or density of schools as a
response to changes in atmospheric and/or marine
climatic patterns) rather than to changes in the abun-
dance of anchovy itself.
Incorporation into management schemes of these
cycles in abundance (e.g., Taylor and Prochaska 1984)
is particularly important for anchovy and other small
pelagic fish which are prone to collapse under intense
fishing pressure and poor recruitment.
Acknowledgments
The author wishes to thank two anonymous reviewers
for their constructive criticisms.
Citations
Bakun, A.
1985 Comparative studies and the recruitment problem:
Searching for generalizations. Calif. Coop. Oceanic Fish. In-
vest. Rep. 26:30-40.
Bethoux, N., and F. Ibanez
1979 Analyse temporelle de la temperature de surface de la mer
dans la bale de Monaco; Relation avec la temperature de
I'air. Rapp. Comm. Int. Mer Medit. 25/26:47-49.
Carapiperis, L.N.
1962 The etesian winds II. On the frequency of the etesian
winds. Notes, Natl. Observatory of Athens, Greece, Ser. II,
Meteorologia 10:1-16 [Engl.].
Chatfield, C.
1984 The analysis of time series: An introduction. Chapman
and Hall, NY.
Colebrook, J.M., and A.H. Taylor
1984 Significant time scales of long-term variability in the
plankton and the environment. Rapp. P.-V. Reun. Cons. Int.
Explor. Mer 183:20-26.
Daoulas, Ch., and A.N. Economou
1986 Seasonal variation of egg size in the sardine, Sardina
pilchardus Walb., of the Saronikos Gulf: Causes and a prob-
able explanation. J. Fish Biol. 28:449-457.
Dekhnik, T.V., and T.S. Rass
1988 Main aspects of the study on icthyoplankton. J. Ichthyol.
28:51-57.
Dement'eva, T.F.
1987 A method for correlation of environmental factors and
year-class strength of fishes. J. Ichthyol. 27:55-59.
Demir, N.
1965 Synopsis of biological data on anchovy, Engraulis en-
crasicoltis, (Linnaeus) 1758 (Mediterranean and adjacent
seas). FAO Fish. Synop. 26, Rev. 1, 48 p.
Flocas, A.A., and B.D. Giles
1984 Air temperature variations in Greece. Pt. 2. Spectral
analysis. J. Climatol. 4:541-546.
Economidis, P.S., A. Sinis, and G.P. Stamou
1988 Spectral analysis of exploited fish populations in lake
Koronia (Macedonia, Greece) during the years 1947-1983.
Cybium 12:151-159.
Gnevyshen, M.N., and A.L 01' (editors)
1977 Effects of solar activity on the earth's atmosphere and
biosphere. Israel Prog. Sci. Transl.. Jerusalem, 290 p.
Hellenic National Statistical Service
1968-1989 Results of the sea fishery survey by motor vessels.
Sect. G, 20 issues (for years 1964-1987). Athens, Hellas.
Kort, V.G.
1970 Large-scale interaction between the ocean and the at-
mosphere using the North Pacific as an example. Oceanology
10:171-183.
Leggett, W.C.
1986 The dependence of fish larval survival on food and
predator densities. In Skreslet, S. (ed.), The role of freshwater
outflow in coastal marine ecosystems. NATO ASI Ser. G7:
117-137.
NOTE Stergiou: Monthly catch variability of Engraulis enaasicolus
215
Mariopoulos, E.G.
1961 An outline of the climate of Greece. Publ. Meteor. Inst.
Univ. Athens (Greece) 6:1-51 [Engl.].
Metaxas, D.A.
1973 Air-sea interaction in the Greek seas and resulted ete-
sian wind characteristics. Tech. Rep. 5, Univ. loannina,
Greece, School of Physics and Math., p. 1-32 [Engl.].
Mullin, M.M., E.R. Brooks, F.M.H. Reid, J. Napp, and
E.F. Stewart
1985 Vertical structure of nearshore plankton off southern
California: A storm and a larval fish food web. Fish. Bull.,
U.S. 83:151-167.
Mysak, L.A.
1986 El Nino, interannual variability and fisheries in the north-
west Pacific Ocean. Can. J. Fish. Aquat. Sci. 43:464-497.
Palomera, I., and J. Lleonart
1989 Field mortality estimates of anchovy larvae, Engraulis
encrasicoltis, in the western Mediterranean. J. Fish. Biol. 35
(Suppl.):133-138.
Petrova-Karadjova, V.J., and E.M. Apostolov
1988 Influence of solar activity upon the diatoms of Black Sea
plankton. Rapp. Comm. Int. Mer Medit. 31:224.
PoUi, S.
1955 I cicli climatici di 5.6 e 8 anni e la loro realta' fisca. Riv.
Meteor. Aueronaut. 2:1-12 [not seen, from Regner and Gacic
1974].
Porumb, 1. 1., and F. Marinescu
1979 Influence du facteur nourriture sur la formation des
agglomerations d'anchois dans la zone neritique roumaine de
la Mer Noire. Rapp. Comm. Int. Mer Medit. 25/26:47-48.
Pucher-Petkovie, T., M. Zore-Armanda, and I. Kacic
1971 Primary and secondary production of the Middle Adriatic
in relation to climatic factors. Thallasia Jugosl. 7:301-311.
Regner, D.
1985 Seasonal and multiannual dynamics of copepods in the
middle Adriatic Sea. Acta Adriat. 26:11-99.
Regner, S.
1982 Investigations of qualitative and quantitative composi-
tion of the larval fish stages in the plankton at the high sea
of the central Adriatic. Stud. Mar. Fauna 11/12:45-60.
1985 Ecology of planktonic stages of anchovy, Engraulis en-
cra^icolus (Linnaus, 1758), in the central Adriatic. Acta
Adriat. 26:5-113.
Regner, S., and J. Gacic
1974 The fluctuation of sardine catch along the eastern Adriatic
coast and solar activity. Acta Adriat. 15:1-15.
Regner, S., D. Regner, I. Marasovic, and F. Krsinic
1987 Spawning of sardine, Sardina yilchardus (Walbaum,
1792), in the Adriatic under upwelling conditions. Acta Adriat.
28:161-198.
Shuntov, B.P., N.A. Shurunov, and V.P. Vasil'kov
1981 Periodicity in the reproduction of fish from the New
Zealand plateau based on data from commercial statistics. J.
Ichthyol. 22:32-43.
Soutar, A., and J.D. Isaacs
1974 Abundance of pelagic fish during the 19th and 20th cen-
turies as recorded in anaerobic sediments off California. Fish.
Bull., U.S. 72:257-275.
Spanakis, E., N. Tsimenidis, and E. Zouros
1989 Genetic differences between populations of sardine, Sar-
dina pilchardu^. and anchovy, Engraulis encrasicolus, in the
Aegean and Ionian seas. J. Fish. Biol. 35:417-437.
Stergiou, K.I.
1986a On the anchovy and sardine fishery in Greek waters,
1964-1982. Rapp. Comm. Int. Mer Medit. 30:241.
1986b Pre-war changes in the abundance of Pagellics erythri-
nus, Merlucciiis merluccius, and Engraulis encrasicolus in
Greek waters and climatic change. FAO Fish. Rep. 361:
99-105.
1988 Spectral analysis of sardine fishery in Greek waters,
1964-1982. Rapp. Comm. Int. Explor. Mer Medit. 31:282.
1990a An autoregressive model of the anchovy Engraulis en-
crasicolus fishery in the eastern Mediterranean. Fish. Bull.,
U.S. 88:411-414.
1990b On the Greek fishery production. Greek Fish. News.
103:31-38 [in Greek].
1991 Describing and forecasting the sardine-anchovy complex
in the eastern Mediterranean using vector autoregressions.
Fish. Res. (Amst.) 11:127-141.
Taylor, G.T., and F.J. Prochaska
1984 Incorporating unobserved cyclical stock movements in
fishery catch equations: An application to the Florida blue
crab fishery. N. Am. J. Fish. Manage. 4:67-74.
Theocharis, A., D. Georgopoulos, Y. Krestenitis, and C. Koutitas
1988 Observations and modelling of upwelling in the Aegean
Sea. Rapp. Comm. Int. Mer Medit. 31:212.
Vucetic, T.
1981 Biological production of the Adriatic plankton. FAO
Fish. Rep. 253:65-71.
Yannopoulos, C.
1977 The reproduction and the egg size variations of Sardina
■pilchardus Walb. in the Thermaikos Gulf, Aegean Sea. Rapp.
Comm. Int. Mer Medit. 24:73-75.
Zupanovic, S.
1968 Causes of fluctuations in sardine catches along the eastern
coast of the Adriatic Sea. Anali Jandranskog Inst. 4:401-489.
Superintendent of Documents Subscriptions Order Form
Order Processing Code:
*5178
Charge your order.
It's Easy!
nsc
n YES, enter my subscripUon as follows: j„ f^ yo„, orders (202) 512-2233
subscriptions to FISHERY BULLETIN (FB) for $24.00 per year ($30.00 foreign).
The total cost of my order is $_
Price includes regular domestic postage and handling and is subject to change.
(Company or Personal Name)
(Please type or print)
(Additional address/attention line)
Please Choose Method of Payment:
I I Check Payable to the Superintendent of Documents
□ GPO Deposit Account | | | | | | | \-\~\
(Street address)
(City, State, ZIP Code)
(Daytime phone including area code)
VISA or MasterCard Account
Thank you for
(Credit card expiration date)
your order!
(Purchase Order No.) ^^ ^^
May we make your name/address available to other mailers? r~| r~|
(Authorizing Signature) '^
Mail To: New Orders, Superintendent of Documents
RO. Box 371954, Pittsburgh, PA 15250-7954
The National Marine Fisheries Service (NMFS) does not approve, recommend or en-
dorse any proprietary product or proprietary material mentioned in this publication.
No reference shall be made to NMFS, or to this publication furnished by NMFS, in
any advertising or sales promotion which would indicate or imply that NMFS ap-
proves, recommends or endorses any proprietary product or proprietary material
mentioned herein, or which has as its purpose an intent to cause directly or indirect-
ly the advertised product to be us«>d or purchased because of this NMFS publication.
Fishery Bulletin
Guidelines for Contributors
Form of manuscript
The document should be in the following
sequence: Title Page, Abstract (not re-
quired for Note), Tfext, Acknowledgments,
Citations, Text footnotes. Appendices,
Tables, Figure legends, and Figures.
Title page should include authors' full
names and mailing addresses and the
senior author's telephone and FAX
numbers.
Abstract Not to exceed one double-
spaced typed page. Should include a
sentence or two explaining to the gen-
eral reader why the research was under-
taken and why the results should be
viewed as important. Abstract should
convey the main point of the paper and
outline the results or conclusions. No
footnotes or references.
Text A brief introduction should por-
tray the broad significance of the paper.
The entire text should be intelligible to
readers from different disciplines. All
technical terms should be defined, as
well as all abbreviations, acronyms, and
sjmnbols in text, equations, or formulae.
Abbreviate units of measure only when
used with numerals or in tables and
figures to conserve space. Measure-
ments should be expressed in metric
units, with other equivalent units given
in parentheses. Follow the U.S. Govern-
ment Printing Office Style Manual,
1984 ed., and the CBE Style Manual,
5th ed. Fishery and invertebrate no-
menclature should follow the American
Fisheries Society Special Publication 12
(for fishes), 16 (for mollusks), and 77 (for
decapod crustaceans).
Text footnotes should be numbered
in Arabic numerals and typed on a sep-
arate sheet from the text. Footnotes are
not used for reference material or per-
sonal communications, but rather to
explain or define terms in the text and
for contribution numbers on the title
page.
Informal sources Personal communi-
cations, unpublished data, and untitled
manuscripts in preparation are noted
parenthetically in the text (full name.
affiliation, brief address including zip
code, and month and year when appro-
priate).
Acknowledgments Gather all ac-
knowledgments into a brief statement
at the end of the text. Give credit only
for exceptional contributions and not to
those whose contributions are part of
their normal duties.
Citations All titled sources should be
listed in the Citations section, including
unpublished and processed material. In
text, cite as Smith and Jones (1977) or
(Smith and Jones 1977); if more than
one citation, list chronologically (Smith
1936, Jones 1975, Doe In press). All
sources cited in the text should be listed
alphabetically by the senior authors'
surnames under the heading CITA-
TIONS. Abbreviations of periodicals
and serials should conform to Serial
Sources for the BIOSIS Data Base'"',
Indicate whether sources are in a langu-
age other than English. For informal
literature, include address of author or
publisher. Authors are responsible for
the accuracy of all citations.
Tables should supplement, not dupli-
cate, the text. Each table should be
numbered and cited consecutively, with
headings short but amply descriptive so
that the reader need not refer to the
text. For values less than one, zeros
should precede all decimal points. In-
dicate units of measure in column head-
ings; do not deviate from the unit of
measure within a column. Table foot-
notes should be noted consecutively in
Roman letters across the page from left
to right and then down. Since all tables
are typeset, they need not be submitted
camera- ready.
Figures Photographs and line draw-
ings should be of professional quality-
clear and concise— and reducible to 42
picas for full-page width or to 20 picas
for a single-column width, and to a max-
imum 55 picas high. All graphic ele-
ments in illustrations must be propor-
tioned to insure legibility when reduced
to fit the page format. Line weight and
lettering should be sharp and even. Let-
tering should be upper and lower case,
and vertical lettering should be avoided
whenever possible (except for vertical,
y, axis). Zeros should precede all deci-
mal points for values less than one. Re-
productions of line artwork are accepted
in the form of high-quality photographic
prints from negatives or photomechan-
ical transfer (PMT). Halftones should be
sharply focused with good contrast.
Micron rules should be inserted on elec-
tron micrographs, even when magnifi-
cation is included in the figure legend.
There should be clear distinction be-
tween identifying letters (press-on or
overlay) and background of photograph.
Label each figure in pencil on the back.
Send only xerox copies of figures to the
Scientific Editor; originals or photo-
graphic prints will be requested later
when the manuscript is accepted for
publication.
Copyright Government publications
are in the public domain, i.e., they are
not protected by copyright.
Submission of manuscript
Disks Authors are encouraged to re-
tain manuscripts on word-processing
storage media (diskettes, floppy disks)
and submit a double-spaced hardcopy
run from the storage media. Submit
disks as MS-DOS "print" or "non-
document" files (often called "ASCII
files"). If a disk cannot be converted to
an ASCII file, the author should indicate
on the disk the source computer and
software language along with the file
name. Either 5V4-inch or 3V2-inch disks
from IBM-compatible or Apple/Macin-
tosh systems (non-graphics only) can be
submitted, double-sided/double-density
or high-density, limiting each file to 300
kilobytes. All 8-inch word-processing
disks (e.g., Wang or NBI) must be con-
verted onto 5V4- or 3V2-inch MS-DOS
print disks.
Send original hardcopy and two dupli-
cated copies to:
Dr. Linda L. Jones, Scientific Editor
National Marine Mammal Laboratory
F/AKC3
National Marine Fisheries Service,
NOAA
7600 Sand Point Way NE
Seattle, WA 98115-0070
Copies of published articles and notes
The senior author and his/her organiza-
tion each receive 50 separates free-of-
charge. Additional copies may be pur-
chased in lots of 100.
'j:iqV.
U.S. Department
of Commerce
Volume 90
Number 2
April 1992
3riP'^ Biolog.ical Lateratory
LIBRARY
t5Ll lie Li iTr^^^' Hole, iVlass__ j
U.S. Department
of Commerce
Barbara Hackman Franklin
Secretary
National Oceanic
and Atmospfieric
Administration
John A. Knauss
Under Secretary for
Oceans and Atmosphere
National IVIarine
Fisheries Service
William W. Fox Jr.
Assistant Administrator
for Fisheries
©Gddd
Scientific Editor
Dr. Linda L. Jones
National Marine Mammal Laboratory
National Marine Fisheries Service, NOAA
7600 Sand Point Way NE
Seattle, Washington 981 15-0070
Editorial Committee
Dr. Andrew E. Dizon National Marine Fisheries Service
Dr. Charles W. Fowler National Marine Fisheries Service
Dr. Richard D. Methot National Marine Fisheries Service
Dr. Theodore W. Pietsch University of Washington
Dr. Joseph E. Powers National Marine Fisheries Service
Dr. Tim D. Smith National Marine Fisheries Service
Dr. IWia J. Tegner Scripps institution of Oceanography
The Fishery Bulktin aSSN 0090-0656)
is published quarterly by the Scientific
Publications Office, National Marine
Fisheries Service, NOAA, 7600 Sand
Point Way NE, BIN C15700, Seattle, WA
98115-0070. Second class postage is paid
in Seattle, Wash., and additional offices.
POSTMASTER send address changes for
subscriptions to Fishery Bulletin, Super-
intendent of Documents, Attn: Chief,
Mail List Branch, Mail Stop SSOM,
Washington, DC 20402-9373.
Although the contents have not been
copyrighted and may be reprinted entire-
ly, reference to source is appreciated.
The Secretary of Commerce has deter-
mined that the publication of this period-
ical is necessary in the transaction of the
public business required by law of this
Department. Use of funds for printing of
this periodical has been approved by the
Director of the Office of Management and
Budget.
For sale by the Superintendent of
Documents, U.S. Government Printing
Office, Washington, DC 20402. Subscrip-
tion price per year: $24.00 domestic and
$30.00 foreign. Cost per single issue;
$12.00 domestic and $15.00 foreign. See
back page for order form.
Managing Editor
Nancy Peacock
National Marine Fisheries Service
Scientific Publications Office
7600 Sand Point Way NE, BIN CI 5700
Seattle, Washington 981 15-0070
The Fishery Bvlletin carries original research reports and technical notes on investiga-
tions in fishery science, engineering, and economics. The Bulletin of the United States
Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries
in 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates
were issued as documents through volume 46; the last document was No. 1103. Begin-
ning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate
appeared as a numbered bulletin. A new system began in 1963 with volume 63 in
which papers are bound together in a single issue of the bulletin. Beginning with
volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued
quarterly. In this form, it is available by subscription from the Superintendent of
Documents, U.S. Government Printing Office, Washington, DC 20402. It is also
available free in limited numbers to libraries, research institutions, State and Federal
agencies, and in exchange for other scientific publications.
U.S. Department
of Commerce
Seattle, Washington
Volume 90
Number 2
April 1992
Fishery
Biological Lauoral
LIBRARY
VI J
UG - 3 1992
Contents
L.
VVUJUS MOle, iV.uSS.
iji Publications Awards, 1989-90
iv List of recent NOAA Technical Reports
217 Armstrong, Michael P., John A. Musicl<, and
James A. Colvocoresses
Age, growth, and reproduction of the goosefish Lophius amencanus
(Pisces.Lophiiformes)
231 Bowers, Michael J.
Annual reproductive cycle of oocytes and embryos of yellowtail
rockfish Sebastes flavidus (Family Scorpaenidae)
243 Bullock, Lewis H., Michael D. Murphy,
Mark F. Godcharles, and Michael E. Mitchell
Age, growth, and reproduction of jewfish Epinephelus itajara in the
eastern Gulf of Mexico
250 Campton, Donald E., Carl J. Berg Jr.,
Lynn M. Robison, and Robert A. Glazer
Genetic patchiness among populations of queen conch Strombus
gigas in the Florida Keys and Bimini
260 Dorn, Martin W.
Detecting environmental covariates of Pacific whiting Merluccius
productus growth using a growth-increment regression model
276 Hyndes, Glenn A., Nell R. Loneragan, and
Ian C. Potter
Influence of sectioning otoliths on marginal increment trends and
age and growth estimates for the flathead Platycephalus speculator
285 Markle, Douglas F., Phillip M. Harris, and
Christopher L. Toole
Metamorphosis and an overview of early-life-history stages in
Dover sole Microstomus pacificus
Fishery Bulletin 90(2). 1992
302 Parrack, Michael L.
Estimating stock abundance from size data
328 Rajaguru, Arjuna
Biology of two co-occurnng tonguefishes, Cynoglossus arel and C. Iida (Pleuronectiformes:Cynoglossidae),
from Indian waters
368 Somerton, David A., and Donald R. Kobayashi
Inverse method for mortality and growth estimation: A new method for larval fishes
376 Stone, Heath H., and Brian M. Jessop
Seasonal distribution of river herring Alosa pseudoharengus and A. aestivalis off the Atlantic coast of
Nova Scotia
IMotes
390 Bumguardner, Britt W., Robert L. Colura, and Gary C. Matlock
Long-term coded wire tag retention in juvenile Sciaenops ocellatus
395 Davis, Tim L.O., and Grant J. West
Growth and mortality of Lutjanus vittus (Quoy and Gaimard) from the North West Shelf of Australia
405 Hettler, William F.
Correlation of winter temperature and landings of pink shrimp Penaeus duorarum in North Carolina
407 Matlock, Gary C.
Growth of five fishes in Texas Bays in the 1 960s
412 Restrepo, Victor R.
A mortality model for a population in which harvested individuals do not necessarily die: The stone crab
417 Salvadd, Carlos A.M., Pierre Kleiber, and Andrew E. Dizon
Optimal course by dolphins for detection avoidance
421 Toole, Christopher L., and Roger L. Nielsen
Effects of microprobe precision on hypotheses related to otolith Sr:Ca ratios
U.S. Department
of Commerce
Seattle, Washington
Publications
Awards 1989-90
National Marine Fisheries Service, IMOAA
The Publications Advisory Committee of the National Marine Fisheries
Service is pleased to announce the awards for best publications authored
by NMFS scientists and published in the Fishery Bulletin volume 88 and
IVIarine Fisheries Review volume 5 1 . Eligible papers are nominated by
the Fisheries Science Centers and Regional Offices and are judged by the
NIVIFS Editorial Board. Only articles which significantly contribute to the
understanding and l<nowledge of NlVIFS-related studies are eligible. We
offer congratulations to the following authors for their outstanding efforts.
Fishery Bulletin 1 990
Joseph E. Hightower
Multispecies harvesting policies for Washington-Oregon-Caiifornia rockfish
trawl fisheries. Fishery Bulletin 88:6^5-656. Dr. Hightower is retired
from his position with the Southwest Fisheries Science Center, and is now
with the North Carolina Cooperative Fish and Wildlife Center, North
Carolina State University, Raleigh.
Marine Fistieries Review 1 989
Joseph M. Terry and Lewis E. Queirolo
U.S. fisheries management and foreign trade linkages: Policy implications
for groundfish fisheries in the North Pacific EEZ, Marine Fisheries Review
51(l):23-43. Drs. Terry and Oueirolo are with the Alaska Fisheries
Science Center, Seattle.
U.S. Department
of Commerce
Seattle, Washington
Recent publications in the
IMOAA Teciinical l^eports IMMFS Series
102 Svrjcek, Ralph S. (editor)
Marine ranching: Proceedings of the seventeenth U.S.-Japan meeting
on aquaculture, Ise, IVlie Prefecture, Japan, October 16, 17, and 18,
1988. IVIay 1991, 180 p.
103 Reid, Robert N., David J. Radosfi, Ann B. Frame,
Steven A. Fromm
Benthic macrofauna of the New York Bight, 1979-89. December
1991, 50 p.
104 Perez, IVIichael, and Tliomas R. Loughlin
Incidental catch of marine mammals by foreign and joint venture trawl
vessels in the U.S. EEZ of the North Pacific, 1 973-88. December 1 99 1 ,
57 p.
105 Wetherall, Jerry A. (editor)
Biology, oceanography, and fisheries of the North Pacific Transition
Zone and Subarctic Frontal Zone. December 1991, 111 p.
106 Svrjceic, Ralpli S. (editor)
Marine ranching: Proceedings of the eighteenth U.S.-Japan meeting
on aquaculture. Port Ludlow, Washington, 18-19 September 1989.
February 1992, 136 p.
107 Russell, Mike, Mark Grace, and Elmer J. Gutherz
Field guide to the searobins [Prionotus and Bellator] in the western
North Atlantic. March 1992, 26 p.
Some NCAA publications are avail-
able by purchase from the Superin-
tendent of Documents, U.S. Govern-
ment Printing Office, Washington,
DC 20402.
Abstract. - Age, growth, and
reproduction were studied in goose-
fish Lophius americarms collected
from National Marine Fisheries Ser-
vice groundfish surveys and com-
mercial fishing cruises between
Georges Bank and Cape Hatteras in
the western North Atlantic. Age and
growth of L. americanus were deter-
mined from vertebral annuli, which
became visible at the edge of the ver-
tebral centra in May. Maximum ages
of males and females were 9 and 11
years, respectively. Males appeared
to experience higher mortality than
females in the older age-classes. Von
Bertalanffy growth curves calcu-
lated for males and females had ex-
cellent agreement with back-calcu-
lated lengths. The growth rate of L.
americamcs was intermediate to its
eastern Atlantic congeners, L. pisca-
torius and L. budegassa. Male L.
americanus matured at 3-i- years
(~370mm TL) and females at 4 +
years ('^^485mm TL). Spawning took
place primarily in May and June.
Fecundity in 17 individuals of 610-
1048 mm TL ranged from 300,000 to
2,800,000 ova, and was linear with
total length in that size range. Histo-
logical examination of the ovaries
showed they are remarkably similar
to ovaries of other lophiiform spe-
cies. Females produced egg veils,
which may function in dispersion,
buoyancy, facilitating fertilization,
and protection of the eggs and
larvae.
Age, growth, and reproduction
of the goosefish Lophius americanus
(Pisces:Lophiif ormes) *
Michael P. Armstrong
School of Marine Science, Virginia Institute of Marine Science
College of William and Mary, Gloucester Point, Virginia 23062
Present address: Department of Zoology, University of New Hampshire
Durham, New Hampshire 03824
John A. Musick
James A. Colvocoresses
School of Marine Science, Virginia Institute of Marine Science
College of William and Mary, Gloucester Point, Virginia 23062
The goosefish Lophius americanus
(Valenciennes in Cuvier and Valen-
ciennes 1837) is a benthic fish which
occurs in the Northwest Atlantic
Ocean from the northern Gulf of
Saint Lawrence, southward to Cape
Hatteras, North Carolina (Bigelow
and Schroeder 1953, Scott and Scott
1988) and less commonly to Florida
(Caruso 1983). It has a eurybathic
depth distribution, having been col-
lected from the tideline (Bigelow and
Schroeder 1953) to approximately
840 m (Markle and Musick 1974), al-
though few large individuals occur
deeper than 400 m (Wenner 1978).
Goosefish have been taken in tem-
peratures of 0-24 °C (Grosslein and
Azarovitz 1982), but seem to be most
abundant in temperatures of about
9°C in the Mid-Atlantic Bight (Ed-
wards 1965), 3-9 °C in Canadian
waters (Jean 1965), and 7-11 °C on
the continental slope off Virginia
(Wenner 1978). The goosefish is sym-
patric with the black-finned goose-
fish L. gastrophysus in deep water
(>100-150m) from Cape Hatteras to
the Florida coast, although strays of
L. gastrophysus occur as far north as
Washington Canyon, off Virginia
(pers. observ., MPA).
Manuscript accepted 20 March 1992.
Fishery Bulletin, U.S. 90:217-230 (1992).
'Contribution 1735, Virginia Institute of Ma-
rine Science.
Lophitts americamis was confused
with L. piscatorius, a European spe-
cies, for many years. Thus all refer-
ences to L. piscatorius in the western
North Atlantic north of Cape Hat-
teras actually refer to L. americanus
(Caruso 1977). There are several
accounts of the species' life history
(Gill 1905, Connolly 1920, Dahlgren
1928, Hildebrand and Schroeder
1928, Proctor et al. 1928, McKen-
zie 1936, Bigelow and Schroeder
1953, Grosslein and Azarovitz 1982,
Scott and Scott 1988), but all are
general in nature. Much of the infor-
mation contained in these reports is
anecdotal.
Goosefish are a bycatch of ground-
fishing and scalloping operations and
are marketed under the name monk-
fish. They have traditionally been
considered "trash" fish in the United
States and discarded at sea or used
in the production of fish meal, with
a small amount being exported to
Europe where Lophius has been
highly esteemed as a food fish for
centuries. Goosefish have become
more popular with the American con-
sumer due to dwindling catches and
rising prices in recent years of the
more traditional fishery products.
Commercial landings have been in-
creasing yearly since 1970 (Northeast
Fisheries Science Center 1991). This
217
218
Fishery Bulletin 90(2). 1992
study describes age, growth, and reproduction of this
increasingly exploited fish.
Methods
Goosefish were collected during the spring and autumn
groundfish surveys (1982-85) conducted by the Na-
tional Marine Fisheries Service (NMFS) in the Mid-
Atlantic Bight and southern New England (for survey
methodology see Grosslein and Azarovitz 1982). Addi-
tional samples were obtained during the NMFS 1983
summer scallop survey off southern New England and
during cniises aboard commercial groundfish trawlers
and scallopers operating out of Hampton, Virginia.
Sampling effort was concentrated in the area from
southern New England to Virginia.
Goosefish greater than ~ 180 mm were examined at
sea. Smaller individuals were fixed in 10% formalin and
saved for examination in the laboratory. Examination
included measuring total and standard length and
weight, excising a section of the vertebral column,
removing both sagittal otoliths, recording stomach con-
tents, macroscopic staging and weighing of the gonads,
and preserving pieces of gonads for histological inspec-
tion and fecundity estimates.
Reproduction
Gonads were staged visually in the field and assigned
to one of the following classes: immature, resting,
developing, ripe, and spent. Both gonads were then
removed from the body cavity and weighed to the
nearest O.lg. A small representative piece was excised
from the midsection of selected gonads and preserved
in Davidson's fixative for histological study.
Late-developing and ripe ovaries were selected for
fecundity analyses. The extremely large size of goose-
fish ovaries precluded saving the entire organ. A sub-
sample of about lOOg was weighed to the nearest 0.1 g
and placed in modified GOson's solution (Simpson 1951).
After several months of storage, most of the ovarian
connective tissue had dissolved. Ova were removed
from the Gilson's solution, separated from any remain-
ing ovarian tissue, rinsed in water, blotted on absor-
bent paper, and weighed. Three subsamples, each con-
taining about 1000 ova, were removed and weighed to
the nearest 0.001 g. Ova in each sample were counted
using a dissecting microscope. Fecundity was calcu-
lated as:
Fecundity = (W)(P)(N)
where W = total weight of both ovaries,
weight of sample after Gilson's
N = mean number of ova/g from 3 subsamples.
Gonad portions preserved in Davidson's fixative for
histological preparations were dehydrated in a graded
series of ethanol baths and Technicon reagents (S-29
dehydrant VC-670 solvent). They were then embedded
in paraffin, sectioned at Yf^m and stained using Harris'
hematoxylin and counterstained with eosin Y. Gonad
sections were viewed at 40 x , 100 x , and 400 x to deter-
mine stages of oogenesis and spermatogenesis to verify
accuracy of macroscopic field staging and to examine
the histology of the goosefish ovary.
A gonasomatic index (GSI) was calculated for each
sex as:
GSI
gonad weight
total weight of fish
100.
P =
weight of sample before Gilson's
Age and growth
Weights were taken to the nearest gram in fish < 1200 g
and to the nearest 25 g increment in fish > 1200 g. Total
length (TL) in millimeters was measured from the tip
of the protruding lower jaw to the tip of the caudal fin
rays. Because of the large size and loose suspension
of the goosefish jaw apparatus, it was necessary to
hold the head in a standard position while length was
measured to reduce variation due to changes in head
and jaw configuration. This position was achieved
by applying light pressure to the top of the head,
thereby causing a maximal amount of dorsal-ventral
compression.
Vertebrae were chosen as the best method to age
L. americanus, based on a preliminary examination
which revealed that each vertebral centrum contained
concentric rings which appeared to be annuli. Sagittal
otoliths were also examined; however, otoliths from
larger fish were opaque and had extremely irregular
outer margins, which made it difficult or impossible to
discern annuli.
A section of the vertebral column containing verte-
brae numbers 3-11 was excised from each goosefish.
These were stored in 50% isopropanol for 1-12 months.
Vertebrae numbers 7-10 were similar in size and shape
and also had the largest diameters. Vertebra number
8 was used in aging, but number 9 was used if number
8 was damaged in preparation.
Vertebra number 8 was disarticulated from the rest
of the excised vertebral section. The neural and haemal
arches and all excess fat, muscle, connective tissue and
cartilage were removed by scalpel. The vertebra was
then sliced along the midsagittal line producing two
hourglass-shaped halves, similar to the method used by
Lyczkowski (1971) and Lawler (1976) for preparing
vertebrae from northern puffer Sphaeroides maculatus
Armstrong et al.: Age, growth, and reproduction of Lophius amencanus
219
and sandbar sharks Carcharinus plumbeus. These
halves were then heated in an oven at 200 °C for about
3 hours. Larger vertebra required one-half to 1 hour
further heating. This heating made the alternating
opaque and translucent bands of the vertebral centra
more distinct.
Annuli were counted on the posterior face of the
centrum. This was generally more concave than the
anterior face, thus allowing greater separation of the
rings. Each vertebra was read twice at an interval of
at least one month to insure independence of readings.
If they disagreed, a third reading was done. Agreement
between any two readings was considered as the true
count. If all three readings differed, the vertebra was
considered unreadable and not used in the analysis. A
random sample of fifty vertebrae was selected for
verification by an independent reader.
Measurements of the vertebral rings and radius were
made from the apex of the posterior and anterior faces
of the centrum along an oblique line that followed the
midline of the posterior centrum. All measurements
and counts were made with a binocular dissecting
microscope equipped with an ocular micrometer at 10 x
magnification using reflected light.
Regression analyses of vertebral radius on total
length and weight on total length were calculated by
the method of least squares. Length-at-age was back-
calculated by the Lee method (Lagler 1956):
L' = C + S' (L-C)/S
where L' = total length of the fish at time of annulus
formation,
L = total length of fish at time of capture,
S' = measurement to the annulus.
S = vertebral radius at time of capture,
C = correction factor; y-axis intercept of the
regression of total length on vertebral
radius.
Computation of the von Bertalanffy growth equations
followed Ricker (1975).
Results
Reproduction
External sexual dimorphism was not apparent in L.
americanus. Caruso (1975) noted sexual differences in
nostril morphology, but this was not a useable field
character. Sex was easily determined in mature in-
dividuals by examination of the gonads, which are
markedly different in appearance. Gonads from small
juveniles (<160-180mmTL) were indistinguishable
macroscopically. Both testes and ovaries from these
juveniles were small, translucent, and string-like.
In females larger than ~180mmTL the ovaries were
long, wide, and ribbon-like. They were greatly coiled
in the abdomen and supported by an extensive meso-
varium. The two ovaries were fused at their posterior
ends, forming a single, confluent organ. Dimensions
of the ovary varied greatly depending on the stage of
sexual development.
The testes were solid, sausage-like organs. A groove
was present along the medial aspect of each testis. This
groove contained blood vessels and served as the site
of attachment for mesentary connective tissue.
A physical description of the gonads in the five
developmental stages (immature, resting, developing,
ripe, and spent) is presented in Table 1.
Stage
Table 1
Description of gonads at five maturity stages assigned to Lophius americanus, based on macroscopic examination.
Description
Ovaries
Immature
Resting
Developing
Ripe
Spent
Testes
Immature
Resting
Developing
Ripe
Spent
Grayish-pink, relatively small, ribbon-like, appear almost empty, no vascularization.
Orangish-pink, contain material but no ova visible, larger than immature, little vascularization.
Pink, ova discernible by eye, abdominal cavity slightly bulging, highly vascular.
Straw-colored to almost clear as ovary approaches spawning, distinct ova present, abdominal cavity greatly bulging,
highly vascular.
Gray, extremely flaccid, appear almost empty, atretic ova appear as black or white dots, moderately vascular.
White to tan, similar in shape as mature testes but very small, medial groove less distinct.
White to tan, much larger than immature, medial groove distinct, small amount of milt sometimes present when dissected.
Blotchy cream to tan, moderate to large amount of milt produced when dissected, very firm in texture.
Blotchy cream to tan with areas of pink, extremely firm in texture, milt produced from genital pore when pressure
is applied on abdomen, copious amounts present when dissected.
Grayish-tan, edges appear translucent, extremely flaccid, small amount of milt sometimes present when dissected.
220
Fishery Bulletin 90(2), 1992
# ova = 4495.04(TL)-2.403.814.8 *
2.5-
r' = 0.67
D
5 2.0-
E
i ,.0-
0.5-
•
0 -
•
600 700 800 900 1000 1 1 00
Total Length (mm)
Figure I
Relationship of fecundity with total length for Lophius ameri-
camis. Dashed line indicates least-squares regression. Regres-
sion equation and coefficient of determination are given. Each
dot represents a single individual (re 17).
Fecundity in 17 individuals of 610-1048 mm TL
ranged from 301,150 to 2,780,632 ova. Fecundity in-
creased linearly with TL in that size range (Fig. 1), the
regression equation being
number of ova =
4495.04(TL)- 2,403,814.8
(r2 0.67).
Log transformations of one or both variables failed to
provide a better fit.
Goosefish reached sexual maturity (by macroscopic
staging) at 290-450 mm in males and 390-590 mm in
females (Fig. 2). Linear regressions of proportion
mature (arcsine-square root transformed) on TL for
these size intervals were:
Proportion of males mature =
0.0089(TL)- 2.498 (r2 0.96)
Proportion of females mature =
0.0079(TL)- 3.056 (r2 0.86).
Values for length at 50% maturity were 368.9mm in
males and 485.3mm in females.
Ovaries and testes followed similar patterns of devel-
opment, with the exception that testes changed from
a resting to developing state earlier in the year (Jan.-
Feb.) (Fig. 3). No resting gonads were found for either
sex in May or June. The percentage of spent gonads
was highest in July- August, indicating that spawning
had taken place in the previous time interval (May-
June). Although the percentage of ripe gonads was
100 -^ A A A A A
80 -
Males * *
60 -
A
A
40 -
A
J) 20 -
A
3
o 0 -
A
2 200 250 300 350 400 450 500 550
c
O
i] 100 -| A A A A
(X
80 -
Females * *
60 -
40 -
^
20 -
A *■
A *
300 350 400 450 500 550 600 650
Total Length (mm)
Figure 2
Percent of individuals sexually mature in relation
to total length (mm) for Lophius americamis, based
on macroscopic examination of the gonads. Lengths
at 50% maturity are 368.9 mm for males and
485.3mm for females.
highest in May-June, gonads in a near-spawning state
were also found in March-April and July-August.
Gonasomatic values were calculated for 117 mature
males and 98 mature females. The GSI peaked in May-
June for females and March-April and May-June for
males (Fig. 4). High index values in these months
corresponded with the greatest incidence of ripe in-
dividuals (Fig. 3). Again, similar to observations based
on gonad condition, males appeared to develop earlier
in the season and remain ripe longer. No mature
females were collected during the Jan. -Feb. interval.
GSI values for females were much greater than for
males (Table 2). Females showed a large increase in
GSI as the ovaries developed. The greatest value
recorded was 50.9, from a ripe female. This value in-
dicates that greater than half of the body weight was
composed of ovarian mass. However, only a relatively
small percentage of the ovarian weight from late-
developing and ripe females was composed of ova. The
actual percentage of the ovarian weight which was ova
ranged from 12.9% to 33.5% for the seventeen females
used for fecundity analysis. The remainder of the
weight was ovarian tissue, and more importantly, the
muco-gelatinous matrix surrounding the ova.
Slides were prepared from sections of 33 ovaries and
20 testes. Representatives from all the developmental
Armstrong et al.: Age. growth, and reproduction of Lophius amencanus
221
O 'o
Figure 3
Seasonal progression of gonad condition in mature male
and female Lophius americanus, based on macroscopic
examination.
classes (immature, resting, developing, ripe and spent)
were included.
Oogenesis proceeds through six distinguishable mor-
phological stages similar to other fishes, such as black
sea bass Centropristis striata (Mercer 1978):
Oogonia (4.5-1 l/.im) Densely packed, granular, deep-
ly basophilic cells.
Stage 1 Small (15-50 /jm) oocytes with a large nucle-
us, single nucleolus, and small amount of basophilic
cytoplasm.
Stage 2 (30-200 f.(m) Previtellogenic oocytes with
strongly basophilic cytoplasm and multiple nucleoli
around the nucleus margin.
Stage 3 (1 10-390f/m) Vitellogenesis begins with the
deposition of yolk vesicles in the less darkly-staining
cytoplasm. A thin zona radiata can be seen in late
stage-3.
Stage 4 (270-970Hm) Cytoplasm filled with yolk vesi-
cles and globules, lightly staining. Zona radiata well
developed and strongly acidophilic.
Stage 5 ( > 600 /.(m) Mature or nearly mature oocytes,
uniform in appearance due to the coalescence of yolk
globules. Often fractured or irregular in outline due
to fixation and sectioning.
o
o
Jon-Feb Mor-April Moy-June July-Auq Sept-Oct Nov-Dei
Month
Figure 4
Seasonal progression of mean gonasomatic index values for
male and female Lophius americanus. Numbers of mature in-
dividuals examined on each date are indicated.
Table 2
Gonasomatic index values at five gonad maturity stages for |
male and female Lophius americamts
based on macroscopic
staging.
Range
Mean(SE) n
Females
Immature
Trace-1.26
- 56
Resting
0.77-7.58
2.35(0.19) 53
Developing
3.82-22.12
12.26(1.18) 21
Ripe
18.23-50.90
33.96(2.73) 13
Spent
0.94-3.77
2.56(0.43) 12
Males
Immature
Trace-0.83
37
Resting
0.31-3.42
1.46(0.17) 36
Developing
0.46-6.18
2.44(0.27) 43
Ripe
0.84-5.72
3.20(0.22) 23
Spent
0.18-4.19
1.16(0.20) 21
Based on the occurrence of these oocyte stages, the
ovaries were placed in the following developmental
classes:
Immature Stage 1 and 2 oocytes present, atretic
bodies absent. The ovarian lamellae are pressed tight-
ly together and lumen is small.
Resting Stage 1, 2, and 3 oocytes are present with
stage 2 dominating.
222
Fishery Bulletin 90(2). 1992
Developing Oocyte stages 1, 2, 3,
present with 3 dominating.
Ripe Oocyte stages 1, 2, 3, 4, and
present with 4 dominating.
Spent Oocyte stages 1, 2, and
3 are present with 2 dominat-
ing. Atretic stage 4 and 5
oocytes and ruptured folHcles
are present.
Macroscopic and microscopic
maturity classifications showed
excellent agreement. Only two
(6%) needed to be reclassified
following histological examina-
tion. These included one reclas-
sified from ripe to developing,
and one from resting to imma-
ture.
Figures 5 and 6 show the his-
tology of the ovary. The lumen is
not centrally located but is at one
side (Fig. 5). The ovigerous tissue
extends into the lumen in the
form of lamellae from one wall
only. In late-developing and ripe
ovaries, the mucogelatinous
material that forms the egg veil
can be seen surrounding the
ovigerous lamellae and filling the
lumen (Fig. 6). This material is
produced by the epithelial cells
(Fulton 1898), which can be seen
lining the lumen and lamellae
(Fig. 6).
Spermatogenesis proceeds
through six distinct stages analo-
gous to those described for Tila-
pia spp. (Hyder 1969) and Cau-
lolatilus microps (Ross 1978).
These stages are primary and
secondary spermatogonia, pri-
mary and secondary spermato-
cytes, spermatids, and sperma-
tozoa. Spermatogenesis in goose-
fish is not notably different from
other teleosts, so the process is
not described here.
The 20 testes examined histo-
logically were placed in the fol-
lowing maturity classifications
based on a modification of the
system of Hyder (1969):
Immature Primary and/or
secondary spermatogonia are
and small 4 are
sometimes 5 are
present; primary and/or secondary spermatocytes
may also be present.
Resting Primary and/or secondary spermatogonia
and spermatocytes are present. Spermatids also
ow
Figure 5
Photomicrograph of Lopkius americanus ovary, classified as resting (40 x ): OL =
ovigerous lamella; L = lumen of ovary; OW = nonovigerous ovarian wall; 1-3 = stages
of oocyte development.
OW
Figure 6
Photomicrograph of Lophius americanus ovary , classified as late developing (40 x ): MG
= mucogelatinous matrix; EP = epithelial lining of lumen and lamellae; OW, =
nonovigerous ovarian wall; AR = artifact; 3-4 = stages of oocyte development.
Armstrong et al.: Age, growth, and reproduction of Lophius americanus
223
present. Small amount of spermatozoa
may be present in lumen.
Developing Few primary and/or sec-
ondary spermatogonia visible; primary
and/or secondary spermatocytes and
spermatids present; spermatozoa pres-
ent in lumen.
Ripe Few or no primary and/or second-
ary spermatogonia and spermatocytes
visible; lumen densely packed with
spermatozoa.
Spent No primary and/or secondary
spermatogonia or spermatocytes vis-
ible; no spermatids present; few sper-
matozoa remaining in lumen.
In all cases, maturity classifications based
on histological examination agreed with
visual classifications applied in the field.
Age and growth
Growth marks on the vertebrae of L.
americanus formed distinct steps on the
centrum surface. Under magnification in
reflected light, the surface texture of the
step appeared coarser than the rest of the
centrum. A narrow, dark, translucent
band was on the outer side of each step.
The step and the narrow band formed a
continuous ring around the centrum and
was considered to be the annulus. Broad-
er, lighter opaque bands with relatively
uniform surface texture were between
the annuli. A broad, opaque band com-
bined with a narrow, translucent band
and step was interpreted as one year's
growth. While these features were visible
on fresh vertebrae, they became much
more distinct when the vertebrae were
heated. The step became deeper and the narrow,
translucent band became opaque and dark relative to
the rest of the centrum (Fig. 7).
Annuli were counted on vertebrae from 635 goose-
fish. In 200 (31.5%) cases, the first and second reading
did not agree and a third reading was done. In most
cases, the second reading differed by only one. In 25
(3.9%) cases, the third reading was different from both
the first and second; these vertebrae were considered
unreadable and discarded from the analysis.
Differences between readings were due to the pres-
ence of false annuli or because the true annuli were not
distinct. False annuli appeared as dark bands but were
not associated with a step. Another extraneous mark
that sometimes occurred was a depression that formed
a continuous ring on the centrum but was not a defin-
2 3 4
Figure 7
Vertebra from a 4-year-old Lophius americanus, after heating. Annuli are
indicated.
itive step. This feature has also been found on black
bullhead (Lewis 1949) and northern puffer (Lyczkowski
1971) vertebrae.
Annuli counts determined by the independent reader
agreed with the original counts in 40 (80%) cases. In
no case did the counts differ by more than one.
Van Oosten (1929) established the following criteria
that must be met before checkmarks on scales or bones
can be considered annuli: (1) Scales or bones must re-
main constant in number and identity throughout the
life of the fish; (2) growth of the scale or bone must
be proportional to the overall growth of the fish; (8)
growth checkmarks must be formed at approximately
the same time each year; and (4) back-calculated lengths
should agree with empirical lengths. The first criteri-
on is fulfilled by using vertebrae as the aging tool.
224
Fishery Bulletin 90(2). 1992
6 ■
x:
12 46 h^
X)
f—i i. 5° /
^Tv r — i /
i\ '9 / \ 1
o-E 4 i
\? I T / \ /
.s^
> — i ''■■^ / \ /
CP
O D
\ TI363 yil
2 3
go ^ ■
\
Q>
:5
P
Jan Morch May July Sept Nov
Month
Figure 8
Monthly mean ( ± 1 SE of mean) marginal width for Lo-phius
americanns vertebrae. Number of vertebrae examined each
month is indicated. No vertebrae were collected during June.
100
80
□ 60 -
40 -
Jan March May
Month
Figure 9
Percentage of Lophius americanus vertebrae having a
marginal width less than one ocular unit by month. Number
of vertebrae examined is indicated. No vertebrae were col-
lected during June.
The regression of vertebral radius (VR) on TL re-
vealed a strong linear relationship between the two
variables. The regression equation based on 682
vertebrae from both sexes was as follows:
TL = 11.077(VR) + 40.018 (r2 0.97).
This indicates that growth of vertebrae is proportional
to growth of the fish, thereby satisfying the second
criterion.
Monthly mean marginal increments were plotted for
all age groups combined (Fig. 8). Sample size was not
large enough to plot the age-groups separately. How-
ever, inspection of the data indicated that the seasonal
progression of marginal increment was similar for all
age-groups. Percentage of vertebrae showing a very
small marginal increment (less than 1 ocular unit), in-
dicating that little or no growth had occurred since the
annulus was deposited, was also plotted (Fig. 9). The
annuli were found to be closest to the edge of the
vertebrae in May. Marginal increments were highest
in December-February, following a period of growth
during July-December. The percent of vertebrae with
thin margins showed less variation than marginal in-
crements. The percent was highest in May and de-
creased as the season progressed. These plots indicate
that May is the time of annulus formation, and only
one checkmark is formed per year. This appears to
fulfill the third criterion that states that growth checks
must be formed at approximately the same time each
year; however, because data were pooled from several
years, this cannot be stated with certainty. Although
there was a decrease in the marginal increment from
February to March, there was no corresponding rise
in the percentage of very small margins (i.e., the mean
800 -
Males
34
1
10 , /
49
--T^*
600 -
61
/
^
400 -
7
/
163/
200 -
/'
/ '
0 -
/
10 12
o
o
1000 -
800 -
Females
27
25
I.
17
13
++
2
600 -
400 -
-'
A
200 -
Y'
0 1
/
— 1 —
10 12
Age (yrs)
Figure 10
Mean observed lengths-at-age for Lophius americanus.
Vertical bars indicate ranges of total length observed for
each age. Sample sizes are indicated.
value of marginal width was not lowered by the pres-
ence of marginal widths < 1). Although the relatively
small sample sizes preclude making definitive conclu-
sions, these data suggest that some process is causing
Armstrong et al.: Age, growth, and reproduction of Lophius amencanus
225
Table 3
Observed, von Bertalanffy, and back-calculated lengths-at-age (TL, mm) for male and female Lophius americanus, based on counts
of vertebral annuli. The number examined for age 1 includes 142 unsexed individuals, which were used in the back-calculations for
both sexes
Number
Age examined
Mean von
observed Bertalanffy
length length
Mean back-calculated length
s at successive
annuli
I
II
III
IV
V
VI
VII
VIII
IX
X XI
Males
1
163
167
133
123
2
78
322
256
127
267
3
61
425
367
134
265
374
4
49
519
469
127
263
377
472
5
34
602
560
127
269
378
478
568
6
10
664
644
109
241
352
465
549
634
7
1
688
719
82
189
284
390
486
592
688
8
1
815
788
109
255
367
473
602
675
731
793
9
1
900
850
143
263
396
489
555
621
701
781
860
Mean
126
264
374
473
563
633
707
787
860
Annual growth
increment
126
138
110
100
90
70
74
80
73
Females
1
163
169
121
124
2
67
313
253
126
261
3
44
412
373
124
257
361
4
26
526
482
116
248
373
476
5
27
652
581
130
278
405
507
600
6
25
718
672
121
250
366
477
580
672
7
17
792
754
124
265
386
485
573
662
757
8
13
874
828
110
242
361
468
567
665
745
834
9
14
937
896
119
250
373
475
567
652
740
821
901
10
4
991
957
107
244
353
458
574
655
741
815
890
966
11
2
1024
1014
117
254
380
488
591
677
757
826
894
962 1013
Mean
123
258
374
483
581
664
748
826
898
965 1013
Annual growth
increment
123
135
116
109
98
»?.
S4
7S
72
t" 48
the vertebrae to decrease slightly in diameter, possibly
the resorbtion of the outer surfaces due to starvation
in late winter.
Mean lengths were back-calculated for 256 males and
260 females. One hundred forty-two individuals, whose
sex could not be determined because their gonads were
undifferentiated (94-239 mm TL) but who were deter-
mined to have one annulus, were included in the back-
calculations for each sex, bringing the total number
used in the analysis to 398 males and 402 females.
The observed lengths were consistently higher than
back-calculated or von Bertalanffy lengths for indi-
vidual age-groups (Table 3). However, the differences
are within the limits of seasonal growth, so the fourth
criterion appears to have been fulfilled.
Males and females had very similar lengths-at-age
until age 4. Above age 4, the mean lengths for females
were slightly greater than males, with the difference
becoming more pronounced with increasing age
(Fig. 10).
The data suggest a difference in maximum age for
the two sexes. The oldest male collected was 9 years
old. Males older than 6 were exceptionally rare. Only
one individual from each of the age groups 7, 8, and
9 was captured during the course of this study. The
oldest female sampled was 11 years old. Fifty females
greater than 6 years old were obtained. It appears that
the number of older males is much fewer than females,
indicating greater mortality of the males.
Mean back-calculated lengths-at-age were used to
develop the vonBertalanffy growth equations. The
resulting parameters and equation for females are:
K =0.095
L^ = 1576 mm
to =0.162
Lt =1576.0 (1-e -
0.095 (t-O.
162))_
The growth equation for males was calculated using
three slightly different data sets. It was first calculated
using all the mean back-calculated lengths available.
The equation was then formulated after eliminating the
two fish in age-groups 8 and 9 from the data set and
finally it was calculated without age-groups 7, 8, or 9.
226
Fishery Bulletin 90(2), 1992
Because there was only one individual in each of these
three oldest age-groups, these were possibly not good
estimates of length for these ages. The parameters and
equations are as follows.
All males:
K = 0.097
L^ = 1460.0
to = 0.015
Lt = 1460.0 (1-e -0097 (t-0.015))
Age-groups 8 and 9 eliminated:
K = 0.166
L^ = 1018.0
to = 0.211
Lt = 1018.0 (l-e-oi66(t-o.2ii))
Age-groups 7, 8, and 9 eliminated:
K = 0.157
L^ = 1059.0
to = 0.196
Lt = 1059.0 (1-e -0-157(1-0.196))
The length-weight relationships (Fig. 11) for 305
males and 311 females were:
Males
logio W = 2.833 Oogio TL) - 4.347 (r^ 0.95)
Females
logio W = 3.001 (logio TL) - 4.770 (r^ 0.98)
Discussion
Reproduction
All female members of the Lophiiformes are thought
to expel nonadhesive, mucoid egg rafts or veils with
the possible exception of one species of antenariid
angler fish (Pietsch and Grobecker 1980). These veils
are buoyant and have a complex structure consisting
of individual chambers, which each contain one to three
eggs and an opening providing water circulation
(Fulton 1898, Gill 1905, Rasquin 1958, Ray 1961). This
method of egg production appears to be unique among
the fishes.
The goosefishes, Lophitcs spp., have the most spec-
tacular egg veils because of their large size. The egg
veil of L. americanus can reach 6-12 m in length and
0.15-1. 5m in width (Martin and Drewry 1978). Several
authors have provided detailed description of the egg
veils of L. americanus (e.g., Agassiz and Whitman
1885, Connolly 1920, Dahlgren 1928) and L. pisca-
torius (Fulton 1898, Bowman 1919).
20000 -
Mole
- Female / /
^ 15000 -
//
//
^ 10000 -
u
5000 -
/
0 200 400 600 800 1000 1200
Total Length (mm)
Figure 1 1
Length-weight relationship for male (n 305) and
female (n 311) Lophius americanus.
Estimates of fecundity presented by other authors
are similar to those obtained in this study. Eaton et al.
(1954) estimated 543,000 ova in the ovary of a 660 mm
specimen. The regression of fecundity on TL presented
here predicts 563,000 ova for a female of this size.
Other estimates of fecundity range from 432,000 to
2,670,000 eggs, based on the examination of veils
released from females of unknown size (Baird 1871,
Nichols and Breder 1927, Berril 1929).
Female goosefish matured at a larger size and at a
greater age (487mm, age 4) than males (369mm, age
3). This is a common trend among teleosts (Moyle and
Cech 1982). In the case of goosefish, the female re-
quires a larger body size to accommodate the large egg
veil. Connolly (1920) was unable to determine size-at-
maturity because of small sample size, but he stated
that a goosefish 18 inches (457 mm) long (unstated sex)
was immature, and all individuals over 31 inches
(787mm) were mature. McBride and Brown (1980), in
a tabular summary of life-history parameters for
several demersal fish species, present the age-at-
maturity for L. americanus as 4 and 5 years for males
and females, respectively. The source of their data is
not stated. Martin and Drewry (1978) and several
others also suggest that the age of maturity is 4 and
5 years for males and females. They state the source
of this information as Connolly (1920). A review of Con-
nolly's paper shows that he was quoting a publication
by Fulton (1903), which deals with the growth of L.
■piscatorius, not L. americanus. At the time of Con-
nolly's paper, the two species were considered synon-
ymous. L. piscatorius is known to reach a larger max-
imum size and is larger at each age (based on data
presented in the following age and growth discussion).
The age-at-maturity cannot be considered the same
Armstrong et al . Age, growth, and reproduction of Lophius amencanus
221
for the two species; in fact, it would be expected that
the age- and length-at-maturity for L. piscatorius
would probably be greater, as suggested here.
Data on gonad condition and the gonasomatic index
indicate that spawning takes place in May-June in the
area from Cape Hatteras to Southern New England.
Because samples were collected and pooled from
throughout this entire region, a seasonal progression
of spawning from south to north as suggested in the
literature cannot be demonstrated. Testes appear to
develop earlier and remain ripe longer than ovaries.
Fulton (1898) found the same to be true for L. pisca-
torius. This suggests that males may be multiple
spawners. Multiple spawning in males would increase
the chances of a ripe female encountering a ripe male,
and thereby spawning successfully. It also serves to
equalize the energetic investment of the sexes in
reproduction. It appears that the investment of females
is relatively high. The GSI was as high as 50%. Tsi-
menidis (1980) found values as high as 37% for the
Mediterranean goosefish L. budegassa. A large part of
the ovarian weight is composed of the mucogelatinous
material that forms the veil. The caloric value of this
material is unknown, but probably is rather low because
of its low density and apparently high water content.
However, the large amount of this material, combined
with the great number of eggs produced, represents
a sizeable energetic contribution by the female to
reproduction.
Histological examination of the goosefish testes
showed that spermatogenesis and the internal struc-
ture are not remarkably different from other teleosts.
It also confirmed the validity of macroscopic staging
of testes in the field. Examination of ovaries showed
that oogenesis is similar to other teleosts but the struc-
ture of the ovary is somewhat different. The most
significant differences were the presence of stalk-like
lamellae containing the developing ova, and epithelium
lining the lumen which is responsible for secreting the
mucogelatinous matrix. Fulton (1898) was the first to
suggest this mechanism of veil formation in the
lophiids. His figures and descriptions of the histology
of the ovaries of L. piscatorius indicate they are iden-
tical to those from L. americanus seen here. Rasquin
(1958) provided detailed descriptions and photographs
of the ovaries of two species of antennariid anglers
(Antennariu^, Histrio) and one species of ogcocephalid
angler. These lophiiform species are known to produce
egg veils. Although they are all only a fraction of the
size of L. americanus and L. piscatorius, the histology
of their ovaries was virtually identical to their larger
relatives, including the presence of stalk-like ovigerous
lamellae and secretory epithelium. It is reasonable to
assume that all members of the order Lophiiformes
known to produce egg veils have similar ovaries. This
character may be useful in verifying veil production in
some of the deepwater lophiiform families for which
veil production has been assumed but not verified.
Pietsch and Grobecker (1980) suggest that the egg
veil is an excellent device for broadcasting a large
number of eggs over great geographical distances. In
addition, the buoyancy of the veil causes the eggs to
develop in relatively productive surface waters.
There seem to be additional selective advantages to
the egg veil as well. It may function in facilitating fer-
tilization of the eggs. When a veil is first extruded from
the female, it absorbs a large quantity of water. As
water is absorbed, sperm may be drawn into the egg
chambers through the small circulation pores in the
veil, thereby insuring fertilization. The veil likely func-
tions by several methods in the protection of the eggs
and embryos, since the embryos remain in the egg
chamber for 2-3 days after hatching (Dahlgren 1928).
Predators such as zooplankton are physically excluded
from the egg chambers by the small size of the circula-
tion pore. The veil may reduce or eliminate olfactory
cues, thereby eliminating predators locating food items
by this method. Wells (1977) suggests that the jelly coat
of yellow perch Perca flavescens spawn may act in a
similar manner. Finally, the mucogelatinous material
of goosefish egg veils may be toxic or repugnant to
potential predators. Newsome and Tompkins (1985)
found that the egg mass of yellow perch contain some
compound(s) that are not toxic but seem to deter pred-
ators. While such a protective device is rare among
teleosts (Fuhrman et al. 1969, Orians and Janzen 1974),
the presence of toxic or unpalatable compounds within
the jelly coat of amphibian egg masses is well known
(Licht 1969, Ward and Sexton 1981).
Age and growth
Females and males have about the same weight-at-
length before maturity. After maturity the females are
slightly heavier than males because of their large
ovaries. As the ovaries ripen, weight differences be-
tween males and females become greater. The regres-
sion slopes for males and females approximate 3, imply-
ing isometric growth in the length-weight relationship.
Tsimenidis and Ondrias (1980) calculated very similar
length-weight regressions for L. piscatorius in the
Mediterranean Sea.
Vertebrae appear to be valid aging tools for L. ameri-
canus. They satisfy all of Van Oosten's (1929) criteria.
Vertebrae can readily be located and removed from
goosefish and are relatively easy to prepare and read.
The annuli are readily discernible since only 3% of the
vertebrae were considered unreliable, and an inex-
perienced, independent reader agreed with the counts
in 80% of the readings he performed.
228
Fishery Bulletin 90(2), 1992
These data indicate that the annuli become discern-
ible in May. Because these rings are present on juve-
niles as well as adults, they appear to be related to
seasonal patterns of growth rather than reproduction.
The annuli are difficult to see when they are at the very
edge of the vertebral centra. For this reason, they are
probably not detected until some additional growth has
occurred after they are laid down. Yasuda (1940) has
shown that on vertebrae ofScombrops sp. annuli were
formed 1.5 months later than on the otoliths. So it is
likely that the annuli (composed of a step and a translu-
cent band) found on goosefish vertebrae represent the
end of fast growth (the step) in late-fall and a period
of slow winter growth (the translucent band).
While several authors have studied growth in L. pis-
catoritis and L. budegassa (Fulton 1903, Guillou and
Njock 1978, Tsimenidis and Ondrias 1980), only Con-
nolly (1920) has looked at growth in L. americaniis. He
based his growth estimates on vertebral annuli counts,
but his sample size was only six individuals. His results
were as follows: age 1, 114 mm; age 4, 457 mm; age
8, 737mm; age 9, 787mm; age 10, 940mm; age 12,
1016mm. These estimates are slightly lower than found
in this study, but a slower growth rate would be ex-
pected in the colder Canadian waters in which Connolly
conducted his study.
The growth rate of L. americanus is intermediate
to L. piscatorius and L. budegassa. Figure 12 compares
the mean back-calculated lengths for the two European
species (from Tsimenidis and Ondrias 1980) with data
presented here for L. americanus.
The differences in observed and back-calculated
mean lengths between males and females past age 4
are small, but appear to be real. This is the most com-
mon form of sexual dimorphism among fishes (Moyle
and Cech 1982). Tsimenidis and Ondrias (1980) found
similar small differences between the sexes for L. bu-
degassa and L. piscatorius.
More significant is the difference in mortality be-
tween the sexes implied by the data. The heavier mor-
tality of males may be caused by increased predation
due to their smaller size, but this does not seem likely.
Perhaps the males exhibit behavioral or distributional
differences which make them more susceptible to
predation or fishing effort. A final possibility is that
they simply reach senescence before females.
The von Bertalanffy growth equations fit the back-
calculated lengths extremely well. The values for L^
for both sexes seem somewhat inflated. The maximum
reported size for L. americanus is approximately
1220 mm (Bigelow and Schroeder 1953). The largest
female collected in this study was 1115 mm and the
calculated L^ was 1576 mm. The largest male collected
was 900 mm compared with a calculated L of 1018-
1460 mm. The inflation of L„ is caused by a lack of
L. piscatorius. ■
E 900 -
/^
E
// -^
J'"^ 'V^' arnericanus
c 600 -
V
—1
/ /
o
/ / -^ L. budegasso
1- 300 -
if^ Males
a' - Females
0 2 4 6 B 10 12
Age (yrs)
Figure 12
Back-calculated lengths-at-age for three spe-
cies of Lophius. Data for L. americanus from
present study; data for L. piscatorius and L.
budegassa from Tsimenidis and Ondrias
(1980).
representatives from the older age-classes. This is a
common problem in age and growth studies. The
asymptotic length is therefore not well defined for
either sex in this study. The sampling effort was be-
lieved to be intense enough to sample these larger
individuals if they were present in the population. It
is concluded that these individuals are simply not pres-
ent. This may be the result of commercial fishing
pressure (groundfishing and scalloping), which tends
to be selective towards larger individuals.
Acknowledgments
We are indebted to all the graduate students and staff
members at the Virginia Institute of Marine Science
who assisted at various points in this study. D. Sved,
M. Chittenden, and W. DuPaul provided helpful re-
views of this manuscript. Ship time was provided by
the Northeast Fisheries Science Center and the fishing
vessels Captain Wool, Vi7'ginia Queen, Virginia
Cavalier, and Cara Lyn. Funding was provided by Sea
Grant, National Oceanic and Atmospheric Administra-
tion, U.S. Department of Commerce, under Grant
NA86AA-D-SG042, through the Virginia Sea Grant
Program, Project RC/F-10, J.A. Musick, Principal In-
vestigator. This manuscript was based on a thesis
submitted by the senior author in partial fulfillment of
the M.A. degree, School of Marine Science, College of
William and Mary.
Armstrong et al.: Age, growth, and reproduction of Lophius amencanus
229
Citations
Agassiz, A., and CO. Whitman
1885 Studies from the Newport Marine Laboratory, XVI. The
development of the osseus fishes. Part I. The pelagic stages
of young fishes. Mem. Mus. Comp. Zool. 14(l):l-56.
Baird, S.F.
1871 Spawning of the goosefish L. americanus. Am. Nat.
5:785-786.
Berril, N.J.
1929 The validity of L. americanus as a species distinct from
L. piscatorius with notes on rate of developement. Contrib.
Can. Biol. Fish. 4(12):143-155.
Bigelow, H.B., and W.C. Schroeder
1953 Fishes of the Gulf of Maine. U.S. Fish. Wildl. Serv. Fish.
Bull. 53(74), 577 p.
Bowman, A.
1919 The eggs and larvae of the angler (Lophius piscatorius)
in Scottish waters. Rep. Fish. Board Scotl. Sci. Invest, for
1919, No. 1. 42 p.
Caruso, J.H.
1975 Sexual dimorphism of the olfactory organs of lophiids.
Copeia 1975:380-381.
1977 The systematics of the fish family Lophiidae. Ph.D. diss.,
Tulane Univ., New Orleans, 220 p.
1983 The systematics and distribution of the lophiid angler
fishes: II. Revision of the genera Lophiomus and Lophius.
Copeia 1983:11-30.
Connolly, C.J.
1920 History of the new food fishes. III. The angler. Bull.
Biol. Board Can. 3:1-17.
Cuvier, G., and A. Valenciennes
1837 Historie naturelle des poisons, vol. 12. F.G. Levrault,
Paris, 507 p.
Dahlgren, U.
1928 The habits and life history of Lophius, the angler fish.
Nat. Hist. 28(l):18-32.
Eaton, T.H. Jr., C.A. Edwards, M.A. Mcintosh, and J.P. Rowland
1954 The structure and relationships of the anglerfish, Lophius
americanus. J. Elisha Mitchell Sci. Soc. 70(2):205-218.
Edwards, R.L.
1965 Relation of temperature to fish abundance and distribu-
tion in the southern New England area. Int. Comm. North-
west Atl. Fish. Spec. Publ. 6:95-110.
Fuhrman, F.A., G.J. Fuhrman, D.L. Dull, and H.S. Moser
1969 Toxins from eggs of fishes and amphibia. J. Agric. Food
Chem. 17:417-424.
Fulton, T.W.
1898 The ovaries and ovarian eggs of the angler or frog fish
(Lophius piscatorius) and of the John Dory (Zeusfaber). Six-
teenth Annu. Rep. Fish. Board Scotl. 3:125-137.
1903 The distribution, growth, and food of the angler.
Twenty-first Annu. Rep. Fish. Board Scotl. 3:186-199.
Gill, T.
1905 The life history of the angler. Smithson. Misc. Collect.
47(4):500-516.
Grosslein, M.D., and T.R. Azarovitz
1982 Goosefish, Lophius americanus. In Fish distribution.
MESA N.Y. Bight Atlas Monogr. 15, New York Sea Grant
Inst., Stony Brook.
Guillou, A., and J.C. Njock
1978 Analyse des structures de la peche dans les parts de la
cote atlantique francaise de 1961 a 1975 et des incidences du
chalutage sur les stocks des principales especes concemees par
cette activite dans les mers adjacentes. Rev-ue Trav. Inst.
Peches Marit. 42(1,2):120-128 [in French].
Hildebrand, S.F., and W.C. Schroeder
1928 Fishes of Chesapeake Bay. Bull. U.S. Bur. Fish. 43, pt.
1, 388 p.
Hyder, M.
1969 Histological studies on the testes of Tilapia leucesticta
and other species of the genus Tilapia (Pisces:Teleostei).
Trans. Am. Microsc. Soc. 88(2):211-231.
Jean, Y.
1965 Seasonal distribution of monkfish along the Canadian
Atlantic mainland. J. Fish. Res. Board Can. 22(2):621-624.
Lagler, K.F.
1956 Freshwater fishery biology, 2d ed. W.C. Brown, Du-
buque, 248 p.
Lawler, E.F. Jr.
1976 The biology of the sandbar shark. Carcharinus plumbeus.
in the lower Chesapeake Bay and adjacent waters. M.A.
thesis, College of William and Mary, Williamsburg, 48 p.
Lewis, W.M.
1949 The use of vertebrae as indicators of the age of the north-
ern black bullhead, Ameriurus m. melas. Iowa State Coll. J.
Sci. 22(2):209-218.
Licht, L.E.
1969 Palatability of Rana and Hyla eggs. Am. Midi. Nat.
82:296-298.
Lyczkowski, J.M.
1971 Age and growth of the northern puffer, Sphoeroides
mactdatus. M.A. thesis. College of William and Mary,
Williamsburg, 26 p.
Markle, D.F., and J.A. Musick
1974 Benthic-slope fishes found at 900 m depth along a transect
in the western North Atlantic Ocean. Mar. Biol. (Berl.) 26:
225-233.
Martin, F.D., and G.E. Drewry
1978 Lophiidae. In Development of fishes of the mid-Atlantic
Bight— An atlas of egg, larval and juvenile stages, vol. 6,
Stromateidae through Ogocephalidae, p. 357-366. U.S. Fish
Wildl. Serv. OBS 78/12, 416 p.
McBride, M.M., and B.E. Brown
1980 The status of the marine fishery resources of the north-
eastern United States. NOAA Tech. Memo. NMFS-F/NEC-5,
Northeast Fish. Sci. Cent., Woods Hole, 13 p.
McKenzie, R.A.
1936 Some notes on the monkfish or angler (Lophius pisca-
torius). Can. Field Nat. 50:55-56.
Mercer, L.P.
1978 The reproductive biology and population dynamics of
black sea bass, Centropristis striata. Ph.D. diss.. College of
William and Mary, Williamsburg, 196 p.
Moyle, P.B., and J.J. Cech Jr.
1982 Fishes, an introduction to ichthyology. Prentice-Hall,
Englewood Cliffs, 593 p.
Newsome, G.E., and J. Tompkins
1985 Yellow perch egg masses deter predators. Can. J. Zool.
63:2882-2884.
Nichols, J.T., and CM. Breder
1927 The marine fishes of New York and southern New
England. Zoologica (NY) 9(1):1-192.
230
Fishery Bulletin 90(2). 1992
Northeast Fisheries Science Center
1991 Status of the fishery resources off the northeastern
United States. NCAA Tech. Memo. NMFS-F/NEC-86, North-
east Fish. Sci. Cent., Woods Hole, 132 p.
Orians, G.H., and D.H. Janzen
1974 Why are embryos so tasty? Am. Nat. 108:581-592.
Pietsch, T.W., and D.B. Grobecker
1980 Parental care as an alternative reproductive mode in an
antennariid angler fish. Copeia 1980:551-553.
Proctor, W., H.C. Tracy, E. Helwig, C.H. Blake, J.E. Morrison,
and S. Cohen
1928 Fishes— A contribution to the life history of the angler
(Lophiiis piscatorius). In Biological survey of the Mount
Desert Region, Part 2, p. 1-29, Philadelphia.
Rasquin, P.
1958 Ovarian morphology and early embryology of the pedicu-
late fishes Antennarius and Histrw. Bull. Am. Mus. Nat. Hist.
114(4):331-371.
Ray, C.
1961 Spawning behavior and egg raft morphology of the ocel-
lated fringed frogfish, Antennarivs numifer (Cuvier). Copeia
1961:230-231.
Ricker, W.E.
1975 Computation and interpretation of biological statistics
offish populations. Bull. Fish. Res. Board Can. 191, 382 p.
Ross, J.L.
1978 Life history aspects of the gray tilefish, Caulolatilus
microps (Goode and Bean, 1878). M.A. thesis, College of
William and Mary, Williamsburg, 120 p.
Scott, W.B., and M.G. Scott
1988 Atlantic fishes of Canada. Can. Bull. Fish. Aquat. Sci.
219, 731 p.
Simpson, A.C.
1951 The fecundity of the plaice. Fish. Invest. Ser. 2, Mar.
Fish. G.B. Minist. Agric. Fish. Food 17(5), 27 p.
Tsimenidis, N.Ch.
1980 Contribution to the study of the angler-fishes Lophius
budegassa Spinola, 1807 and L. piscatorius L., 1758 in Greek
seas. Inst. Oceanogr. Fish. Res. (Athens) Spec. Publ. 4:
180-190.
Tsimenidis, N.Ch., and J.Ch. Ondrias
1980 Growth studies on the angler-fishes L. piscatorius and
L. budegassa in Greek waters. Thalasographica 2(3):63-94.
Van Oosten, J.
1929 Life history of the lake herring {Leucichthys artedi,
LeSeur) of Lake Huron as revealed by its scales, with a cri-
tique of the scale method. Bull. U.S. Bur. Fish. 44:265-428.
Ward, D., and O.J. Sexton
1981 Anti-predator role of salamander egg membranes.
Copeia 1981:724-726.
Wells, L.
1977 Changes in yellow perch (Perca flavescens) populations
of Lake Michigan, 1954-75. J. Fish. Res. Board Can. 34:
1821-1829.
Wenner, C.A.
1978 Lffphius americanus. In Making a living on the continen-
tal slope and in the deep sea: Life history of some dominant
fishes of the Norfolk Canyon area, p. 253-263, 267-268. Ph.D.
diss.. College of William and Mary, Williamsburg.
Yasuda, H.
1940 On the rings formed on scale, vertebral centrum, and
otolith of the same individual fish. Jpn. Soc. Sci. Fish. Bull.
8:290-300.
Abstract. - Female yellowtall
rockfish Sebastes flavidus, a vivipa-
rous species employing intralumenal
gestation following fertilization of
ovulated eggs, were caught from
Cordell Bank (seamount 20 miles
west of Pt. Reyes, central California)
on a monthly basis from May 1985
through April 1986 to determine
their annual reproductive cycle.
Since histological methods provide
precise and detailed information, this
method was employed to (1) examine
oocytes and embryos to describe
developmental stages, and (2) pro-
vide temporal assessment of the an-
nual reproductive cycle. The descrip-
tion and staging scheme developed
provide a basis to compare reproduc-
tive developmental patterns between
cycles and populations.
Oogonia (Stage I) and early peri-
nucleolus (Stage II) oocytes were
present in samples from all months.
Progressive growth of oocytes from
early- to late perinucleolus (Stage III)
was evident in spent and recovering
ovaries, indicating the end of a repro-
ductive year and the beginning of a
new reproductive cycle. Initial yolk
accumulation (Stage IV) occurred in
July, and final yolk accumulation
(Stage V) was predominant from
September through January. In Feb-
ruary, the majority of samples dis-
played fertilized ova in early-celled
stages of embryonic development.
Gestation continued for about 30
days with parturition occurring be-
tween January and March. Mature
oocytes were also collected in March,
suggesting the Cordell Bank yellow-
tail population has a prolonged re-
productive season extending into
April.
Annual reproductive cycle of
oocytes and embryos of yellowtail
rockfish Sebastes flavidus
(Family Scorpaenidae)
Michael J. Bowers
Tiburon Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA
3 1 50 Paradise Drive, Tiburon, California 94920
Manuscript accepted 9 March 1992.
Fishery Bulletin, U.S. 90:231-242(1992).
Sixty species of rockfish (genus Se-
bastes) have been recorded in waters
off the California coast; twenty spe-
cies are utilized by commercial and
recreational fisheries (Lenarz 1986).
Rockfishes display a wide variety of
life history patterns with respect to
their habitat and seasonality of repro-
duction (Wyllie Echeverria 1987).
The majority of investigations on
rockfish reproduction have focused
on the development, occurrence, and
identification of larvae and juveniles.
Evaluating annual reproductive suc-
cess as a direct consequence of varia-
tions in oocyte viability has received
less attention.
The Sebastes complex contributed
approximately 37,806 mt to west
coast fisheries in 1985 (PFMC 1990)
and management of this resource is
heavily dependent upon predictions
of strong and weak year-classes.
Since no single trait accurately rep-
resents reproductive capacities of
fish populations (Eldridge et al.
1991), fisheries management is based
on a variety of information. Year-
class strength estimates may be en-
hanced by understanding factors
influencing annual fluctuations in
reproduction. Reproduction within
this genus is characterized by intra-
limienal gestation, followring fertiliza-
tion of ovulated, mature eggs. This
process is somewhat unique among
teleosts occurring only in scorpaenids
and zoarcids (Wourms et al. 1988).
The investigation described here
focuses on the development and tem-
poral occurrence of oocytes and em-
bryos within the ovary of yellowtail
rockfish Sebastes flavidus. Character-
ization of oocyte growth and embry-
onic development provides a basis for
assessing reproductive performance.
This study is part of a larger effort
to acquire information on the repro-
ductive biology of yellowtail rockfish
to ultimately determine factors that
influence reproductive success.
Although characteristics of oocyte
growth are generally similar among
teleosts (Wallace and Selman 1981),
numerous ovary maturity scales and
oocyte classification schemes exist
(Yamamoto 1956, Htun-Han 1978,
Robb 1982, Howell 1983). These clas-
sification schemes are useful for
determining reproductive strategies
(synchronous, group-synchronous, or
asynchronous) and evaluating aspects
of reproductive trends. Each oocyte
staging system is, however, less like-
ly to be adapted for teleosts outside
the genus of original study due to the
variety of reproductive strategies.
An oocyte classification scheme to
assess the reproductive status of the
marine, viviparous genus Sebastes in
the Eastern Pacific has not been
reported. Taking these factors under
consideration, the objectives of the
study reported here were two-fold:
(1) to describe oocyte and embryonic
development in Sebastes flavidus
through one complete reproductive
cycle, and (2) establish a staging
231
232
Fishery Bulletin 90(2). 1992
classification as a basis for the comparison of oocyte
and embryonic development between populations,
reproductive years, and other species of Sebastes. Such
data may be used to monitor reproductive development
during a particular year. In addition, descriptions of
oocyte and embryonic development provide a basis to
compare the impacts of environmental fluctuations and
physiological responses with the production of viable
offspring. An understanding of environmental and
physiological interactions influencing reproductive suc-
cess could provide valuable contributions to the under-
standing of recruitment dynamics and allow for more
efficient resource management.
Materials and methods
Specimens were collected from Cordell Bank (38°00'N,
123°25'W), a seamount 20 miles west of Pt. Reyes, at
monthly intervals. Adult female yellowtail rockfish
were captured by hook-and-line at depths of 50-150 m,
from May 1985 through April 1986. No samples were
obtained in June 1985 due to inclement weather. Mean
age and size of samples for each month are shown in
Table 1.
Fish were held on ice and transported to the labora-
tory where pieces of ovaries ("^^4 x 4 x 6 mm) were dis-
sected and fixed in 10% neutral buffered formalin.
Routine paraffin embedding followed the guidelines of
Humason (1967). Samples were sectioned at 6^ thick-
ness with a rotary microtome. Mounted sections were
stained in hematoxylin and counterstained in eosin
(H&E).
Cell measurements were made using a video coor-
dinate digitizer (Model 582 AVCD, H.E. Inc., Las
Vegas, NV) on cells sectioned through the nucleus.
Oocytes were measured and staged randomly. Mean
cell diameters were determined from a subsample of
10-20 cells for each stage. All cell diameters reported
are from fixed tissues.
In each monthly sample, the first 200-400 cells en-
countered were counted and staged. Percent frequen-
cy distributions of the various oocyte stages were
calculated by dividing the total number of a particular
stage by the total number of oocytes observed, ex-
pressed as a percentage. Because the probability of an
individual oocyte being sectioned is proportional to its
size as well as its abundance, larger oocytes tend to
be overestimated and smaller oocytes underestimated
(Howell 1983). Nonetheless, frequency distributions do
indicate seasonal changes within the ovary. Because
of the wide range of cell diameters, overlapping sizes
among oocyte stages, and shrinkage due to fixation,
criteria for staging oocytes was based on histological
appearances and cell structure.
Table 1
Monthly
means and standard errors for
age, standard length
(SL), and weight (Wt)
of Sebastes Jlaiyidus collected off cen-
tral California, May 1985-April 1986,
used for histological
analysis.
Month
Age (yr)
SL (cm)
Wt (g) n
May
14.7±2.0
37.411.0
14341115 6
June
—
—
0
July
17.9 + 2.3
39.511.5
16001146 9
Aug.
13.2±2.0
36.811.8
13871181 4
Sept.
30.0±2.6
44.410.6
23491111 6
Oct.
26.2 ±4.0
42.611.2
20171157 7
Nov.
21.1±2.2
40.910.6
17801 97 6
Dec.
18.7±2.1
39.311.1
18301122 10
Jan.
15.411.8
36.111.1
12721107 10
Feb.
14.611.8
36.111.1
13151 83 10
Mar.
19.311.8
37.511.2
14891102 10
Apr.
23.412.2
38.310.6
13871 52 10
Although ovaries in varying stages of postfertiliza-
tion development were observed, monthly collection of
ovaries was inadequate for a detailed study of rapid
embryonic development. Therefore, additional samples
of embryonic stages were collected by catheterization
from female yellowtail rockfish held in captivity. Ten
adult female yellowtail rockfish captured after copula-
tion were catheterized weekly for 6 weeks while being
maintained in a flow-through, sand-filtered (to 10f.<) sea-
water system. Photoperiod was ambient. The mean
temperature and salinity (10.4°C and 34.7"/oo) for the
2-month holding period (January and February) were
well within the range of parameters at the sampling
site.
Abbreviations used In figures
BC
blastodermal cap
MU
muscle
BR
brain
NC
notochord
C
capillary
NU
nucleoli
CH
chorion
N
nucleus
EB
embryonic body
OG
oil globule
EF
empty follicle
ON
oogonial nest
ER
erythrocyte
OP
optic vesicle
EY
eye
OV
oil vacuole
FOL
follicle
POF
postovulatory
G
granulosa
follicle
GR
germ ring
RE
retina
HG
hind gut
SO
somite
LC
lampbrush
T
theca
chromosome
VM
vitelline
LN
lens
membrane
LT
liver tissue
YG
yolk globule
MN
migratory nucleus
YM
yolk mass
Bowers: Reproductive cycle of oocytes and embryos of Sebastes flavidus
233
Table 2
Classification and temporal occurrence of oocytes from Sebastes flatiidus collected off central California, based on histological appearance.
See text for additional histological descriptions.
Stage
Major histological characteristics
Temporal occurrence
I Oogonia
II Early perinucleolus
III Late perinucleolus
IV Initial yolk accumulation
V Final yolk accumulation
VI Migratory nucleus
VII Ovulation & Fertilization
Small cells (5-25f/) found in clumps or "nests." Cytoplasm pale to All year
clear. Basophilic nucleus occupying most of cell volume.
Wide range of cell diameters (20-lOOfi). Intense basophilic cytoplasm. All year
One or two large nucleoli in nucleus.
Diameters 50-140(i. Small, clear vesicles present in cytoplasm. Feb. -Oct.
Cytoplasm pale-blue to light-gray. Several small nucleoli around inner
margins of nuclear membrane.
Cell diameters 120-210ti. Small spherical, eosinophilic yolk granules in July-Oct.
a distinct cortical zone in cytoplasm. Cytoplasm vesicular and light-
gray. Well-developed follicle surrounds a developing vitelline mem-
brane. Several small nucleoli around the inner margins of nuclear
membrane.
Large cells (200-600 jj). Cell volume one-half to entirely full of yolk Sept. -March
spheres. Lampbrush chromosomes visible in nucleoplasm. Lipid
vacuoles appear larger as they coalesce.
Cell diameters 600-750 ^i. A single, large lipid vacuole present. Dec. -March
Nuclear membrane indistinct or absent. Nuclear material irregularly
shaped and no longer centrally located. Follicle may be distorted and
irregularly shaped.
Mature oocyte free from follicle. The yolk mass is a single, large Dec-March
homogeneous mass, staining deep-purple. In fresh (unfixed) ovaries,
ova appear clear or translucent.
Results
Oocyte development
Major histological characteristics distinguishing stages
for S. flavidus are listed in Table 2. All cells could be
categorized into one of the seven stages. Terminology
and nomenclature follow Moser (1967a) and Howell
(1983).
Stage I: Oogonia These small cells were 5-25^ in
diameter and were found in clumps or "nests" along
the lamellar branches (Figs. lA, 4E). Larger oogonia
in the 20 fi range possessed a deeply-stained chromatin
network attached to a single, large basophilic nucleolus.
Stage II: Early perinucleolus These oocytes were
20-100^ in cell diameter. While still closely associated
with neighboring oogonia, there was noticeable move-
ment away from oogonial nests. The most obvious
feature of this cell was the intensely basophilic cyto-
plasm (Figs. IB, 4E).
Stage III: Late perinucleolus Diameters were 50-
140^. Clear vacuoles appeared in the cytoplasm of
oocytes as small as 50 ^i. Initially these vacuoles were
distributed as a poorly organized ring surrounding the
nucleus, but were seen randomly scattered throughout
the cytoplasm of larger oocj^es (Figs. IB, 4E). As
growth continued, they increased in size and number.
Stage IV: Initial yolk accumulation The earliest
signs of yolk accumulation were seen in oocytes of
120-210 p< in diameter. Small, spherical globules of yolk
were seen in a distinct cortical zone in the cytoplasm
(Fig. 1C,D). The follicle enclosing the oocyte is more
complex and composed of several identifiable struc-
tures (see below).
Stage V: Final yolk accumulation To simplify the
staging of yolked oocytes, cells with approximately one-
half their volume filled with yolk spheres, and cells
whose volumes were entirely filled with yolk, were
placed in Stage V. Yolk spheres increased in number
and size. By the end of this stage, the cell diameter in-
creased to about 650 fi. The cytoplasm was entirely
filled writh yolk spheres of various sizes (Fig. 2A). The
vacuoles which were distributed throughout the cyto-
plasm began to coalesce, forming larger vacuoles.
An eosinophilic nucleoplasm with lampbrush chromo-
somes was visible (Fig. 2B). In the late Stage-V cell,
the nucleus became irregularly shaped and the nuclear
membrane was often indistinct (Fig. 2A).
234
Fishery Bulletin 90(2), 1992
^^^^
cK^'-,
■■■■ ■^'' ^i /, ^ is-^,
■^■m
•fe
VM
fir
D
OV V
Bowers: Reproductive cycle of oocytes and embryos of Sebastes flavidus
235
Figure 1 (left page)
(A) Oogonial nest (Stage-I oocyte) from Sebastes jlavidiis, contain-
ing several primary oocytes collected Dec. 1985, 400 x . (B) Section
of oocytes in ovary of S. flamdus collected May 1985, 250 x . Baso-
philic properties of cytoplasm in Stage-II cells and their large nucleoli
are shown. Distribution of vacuoles are seen in the larger Stage-Ill
cells. (C) Cross-section of an ovary collected Aug. 1985 showing ar-
rangement of Stage-IV oocytes in grape-like clusters on outer
margins of a lamellar branch, 63 x . (D) Typical Stage-IV (initial yolk
accumulation) oocyte showing the first indications of yolk, 250 x .
The cellular composition of the mature follicle was
best observed in Stage-IV or Stage-V oocjdes (Fig. 2D).
A bilaminar vitelline membrane about 1^< in thickness
was next to the plasma membrane of the oocyte. Out-
side the vitelline membrane was a single inner epithelial
layer, the granulosa. Encapsulating the granulosa was
an intricate capillary network filled with erythrocytes.
The theca, a single epithelial layer consisting of squa-
mous cells with large nuclei, surrounded the profuse
capillary system.
Stage VI: Migratory nucleus Cell diameters ranged
from 600 to ~750^. Lipid material had coalesced to
form a single, large vacuole, usually centrally located.
Nuclear material was ameboid in appearance and no
longer occupied a centralized position in the cell (Fig.
2C). Nucleoli were small, indistinct, or entirely absent.
Stage VII: Ovulation/Fertilization Histological
evidence of ovulation was verified by observing the in-
tegrity of the surrounding follicle. Follicles appeared
either as irregularly shaped and shrunken away from
the oocyte or displayed a loss of continuity. Postovu-
latory follicles appeared throughout the sectioned ovary
(Fig." 2C).
Because fertilization of the mature oocyte occurs
rapidly after ovulation, distinction between ovulated
oocytes and recently-fertilized oocytes was unneces-
sary. Therefore, fertilization was considered an event
rather than a stage of histological distinction, and is
included in Stage VII to maintain logical continuity of
the developmental process. Following fertilization,
however, the yolk material became a single homog-
eneous mass staining bright-purple in histological
preparations, appearing clear or translucent in unfixed
samples (Fig. 3 A, B). This distinguishes fertilized from
recently ovulated ova.
Embryonic development
A complete series of sequential embryonic developmen-
tal stages was not obtained from field collections due
to the sampling interval and rapid development of
embryos. Embryos from field collections were, how-
ever, satisfactorily placed into one of three broad
categories: (1) early-celled, (2) embryonic body, or (3)
eyed-larvae (where retinal pigmentation was visible).
Early-celled The early celled stage of embryonic
development observed from field collections of yellow-
tail rockfish ovaries corresponded to stage 9 of Op-
penheimer's classification for Fundulus heteroclitus
(Oppenheimer 1937). The early-celled stage was seen
as an undifferentiated mass of cells (blastodermal cap)
on top of a large yolk mass (Fig. 3A). This stage was
first collected in January, most frequently seen in
February, and last occurred in March.
Embryos in a more advanced state (i.e., flattening
or expansion of the blastula) occasionally occurred
within an ovary primarily containing early-celled
embryos. This suggests rapid cellular divisions and
growth.
Embryonic body The appearance of an embryonic
body was first seen in an ovary collected in February
and last seen in March. This embryonic stage closely
corresponds to Oppenheimer's stages 14 or 15 (Op-
penheimer 1937). At the beginning of this stage, an un-
differentiated mass of cells (taking on the appearance
of tissue rather than individual cells) was located in a
high ridge lying over the yolk mass. The oil globule was
evident at the opposite pole (Fig. 3B). With further
development, embryos displayed optic vesicles orig-
inating from lateral buds, distinguishing the cephalic
region (Fig. 3C). By the end of this growth phase,
somites along the trunk were visible along with
lengthening of the tail. The head had fiuther developed
to include lens formation (Fig. 3D).
Figure 2 (overleaf, left page)
(A) Cross-section of ovary from Sebastes flavidus with clutch of
oocytes in late Stage V, 63 x . (B) Cross-section through nucleus of
a Stage-V oocyte showing distribution of nucleoli and lampbrush
chromosomes in the nucleoplasm, 400 x. (C) Section through two
nearly-mature oocytes in Stage VI (migratory nucleus). 63 x . (D)
Tangential section of Stage-V oocyte showing components of the
follicle outside of vitelline membrane, 400 x.
Figure 3 (overleaf, right page)
(A) Early-celled embryos from Sebastes flavidus ovary collected Feb.
1986, 63 X . (B) First appearance of embryonic body, showing cellular
differentiation. Whole embryo, formalin-fixed, 40 x. (C) Section
through developing embryo (embryonic body, late stage) showing
optic vesicle formation originating from lateral expansions in cephalic
region, 63 x . (D) Embryonic body stage further developed than in
Fig. 3C, with better definition of brain, retina, and lens formation,
63x.
236
Fishery Bulletin 90(2). 1992
Bowers: Reproductive cycle of oocytes and embryos of Sebastes flavidus
237
:<^*-
CH—H
■ymI
^^ft
OV ^
^H
^^r
^
^^^H^
N
ABBi^^
^ •'
Figure 3
238
Fishery Bulletin 90(2), 1992
YM HG
POF
J*.-
r^
«>■■ -X
D
«6'
.0^-
^<«i
^'fcC^
© oV
. t-s
»^ ^
ER
.^ -; '/
Figure 4 (above)
(A) Unhatched prolarvae of Sebastes jlavidus collected Feb. 1986.
Pigmentation of retina is apparent, as are well-formed somites,
63 X . (B) Tangential section of unhatched prolarvae with completed
pigmentation of the retina. Tail continues to lengthen and is seen
to pass the head slightly, 63 x . (C) Newly-hatched larva ofS-Jknidus
showing close association of liver with oil vacuole, developing jaw
and well-developed gut, 40 x . (D) Cross-section of a recently-spent
ovary oiS. jlavidus collected March 1986. Many empty and collapsed
follicles are being resorbed, 40 x . (E) Recovering and early develop-
ing ovary collected April 1986 showing reorganization of ovarian
stroma as resorption nears completion, 63 x .
Eyed-larvae Retinal pigmentation began as a black
deposit outlining the periphery of the retina. Concur-
rently, somites were well formed in the thoracic and
tail regions (Fig. 4A). Mature embryos (prehatching
larvae) exhibited complete pigmentation of the eyes and
a well-developed musculature system along the entire
length of the tail (Fig. 4B). Embryos in this broad
developmental category were in field samples collected
in January and February. Had ovaries containing
Bowers: Reproductive cycle of oocytes and embryos of Sebastes flavidus
239
hatched embryos (larvae) been collected in the field,
they would have been included in this stage. Larvae
were, however, taken from females held in the labora-
tory. These larvae were 4-6 mm in length and had open
mouths with functional jaws. The yolk mass appeared
to be reduced, and liver tissue was associated with a
persistent oil globule (Fig. 4C).
Seasonal oocyte cycle Oogonial nests were observed
in all samples, with about 25% frequency of occurrence
throughout the entire reproductive cycle (Fig. 5). These
Stage-I cells were most conspicuous early in the repro-
ductive season and in ovaries of spent females.
Stage II or early-perinucleolus oocytes were also
noted year-round and accounted for about one-third of
all oocyte types observed (Fig. 5). The large nucleoH
and dark cytoplasm were features that easily distin-
guish this stage. Mid- to late Stage-II cells were ob-
served either singly or in groups around oocytes of later
maturational stages (Stages III- VII) and were con-
sidered the 'resting' stage oocytes of other investi-
gators (Bowers and Holliday 1961, Howell 1983).
In early spring, a broader range of Stage-II cell
diameters was evident, indicating continued oocyte
growth. Stage-Ill cells rapidly increased during March
and April to a maximum frequency of 40% in April,
and decreased in frequency by August as this clutch
of oocytes developed (Fig. 5).
Copulation of yellowtail rockfish typically occurs over
three months beginning in August and ending in Octo-
ber (Eldridge et al. 1991). The incidence and frequency
of sperm in yellowtail rockfish ovaries were not evalu-
ated in this study. However, small clumps or 'packages'
of sperm were occasionally seen closely associated with
the stroma or in spaces between developing (yolked)
oocytes.
Initial yolk accumulation (Stage IV) was first docu-
mented in females collected in July, with all specimens
collected in August showing this stage. In August, 34%
of oocytes were Stage IV. Oocytes of Stage IV ap-
peared as grape-like clusters on the outer margins
of the lamellar branches, developing in a group-
synchronous manner (Fig. IC). The occurrence of
Stage-IV oocytes sharply declined from August to
November when no Stage-IV oocytes were observed
(Fig. 5). As yolk accumulation continued, yolk spheres
increased in size and number, filling the cytoplasm to
about one-half its volume. At this point, oocytes were
categorized as Stage V. Stage V was the most ad-
vanced oocyte observed from its first appearance in
September until December when the frequency de-
clined (Fig. 5). Oocytes in this developmental stage
were most prevalent in November when they accounted
for a mean of 48% of all oocytes. Stage-VI (Migratory
nucleus) oocytes were first observed in December.
so
25
0
50-
25
0-
50-
§ 25-
o
STAGE I
— 1 1 r
'-'-• *\.^*
T 1 1 1 1 1 1
STAGE II
*— »
+ +
— I 1 1 1 r
STAGE I
---i.
N^.->
- +
-f f
25-
0-
STAGE IV
1 ^-?— ?
50
25-
.i-i-
-4— f-
/-\ STAGE V
* + -
, , , s
50-
STAGE VI
25-
0-
— f
...,^'^^S
MJJASONDJFMA
MONTH
Figure 5
Mean monthly frequency distributions of oocyte
Stages I-VI in ovaries oi Sebastes flavidus during
1985-86. Error bars represent 1 SE. Plus signs =
present, minus = not present. Monthly sample size
same as in Table 1.
Ovaries with an advanced mode of Stage-VI oocytes
continued to be collected over the next 3 months
(January-March). This stage appeared to have a short
duration, as ovaries containing Stage VII were also
collected in some of the same months as Stage VI
(December-February).
Ovary maturation was determined by using the most
advanced oocyte or embryonic stage present in each
monthly sample, and their frequency of occurrence was
expressed as percent(s) (Fig. 6). Temporal ovarian
development is illustrated and reflects a prolonged
reproductive season.
While accurate frequency distributions on Stage- VII
oocytes were not possible, the peak month of ovulation
and fertilization appeared to be February. Sections of
samples with Stage-VII oocytes showed eggs free from
(e.g., outside) their follicular remnants. While continu-
ity of follicular components (theca and granulosa) was
disrupted, the integrity of the capillary network was
maintained and there was a close association with the
developing embryo.
Ovaries recently spawned (parturition) were seen as
early as January and most frequently collected in
March (Fig. 6). The ovary was greatly reduced in size,
240
Fishery Bulletin 90(2). 1992
reddish-blue in color, and very
soft in texture. Histologically,
the spent ovary displayed in-
creased vascularization, a thick-
ening of the tunica, and post-
ovulatory follicles undergoing
various stages of resorption (Fig.
4D). In addition, larvae remain-
ing after parturition and yolked
oocytes not reaching maturity
are frequently seen in various
stages of resorption.
Late-recovering and early-
developing ovaries possessed
reorganized lamellar branches
containing Stages I, II, and early
Stage-Ill oocytes as resorption
nears completion (Fig. 4E). All
samples collected in April were
in this condition (Fig. 6), which
marked the end of one reproduc-
tive season and the beginning of
the next reproductive cycle.
Discussion
MJJASON D
MONTH
Figure 6
Percent distribution of ovarian maturation stages (derived from the most advanced oocyte
or embryonic stage categorized histologically) from monthly collections of Sebastes
Jlavidus; n is same as presented in Table 1. Ill = late perinucleolus, IV = early yolk,
V = late yolk, VI = migratory nucleus, VII = ovulation/fertilization, EC = early-celled
embryo, EB = embryonic body, EL = eyed-larvae, SP = spent.
In the present study, I estab-
lished an oocyte/embryonic clas-
sification that allows rapid determination of a rockfish
population's status in the annual reproductive cycle.
The use of this staging system allows one to establish
oocyte frequency distributions and categorize ovaries
as to their seasonal development, both temporally and
spatially. This information, in turn, not only permits
interannual and interpopulational comparisons, but
may help reveal variations related to environmental
factors.
Developmental events that occur in the oocytes of
Sebastes Jlavidus are similar to those described for
other teleosts with group-synchronous development
(see review by Wallace and Selman 1981). Embryo-
genesis and the basic reproductive patterns follow
observations reported for other members of the genus
Sebastes (Moser 1967a, Wyllie Echeverria 1987). Tem-
poral occurrence of reproductive events and seasonal
variations of these events differ within the genus
(Wyllie Echeverria 1987).
In the present work, oocyte development in S. Jlavi-
dus has been categorized by separating oocyte growth
into seven distinct stages. Oogonia and early-peri-
nucleolus stages (Stages I and II, respectively) are
found in the ovaries throughout the year. These stages
appear to grow continuously, develop asynchronous-
ly, and, particularly in Stage-II cells, display a wide
range of cell diameters. Development of unyolked
oocytes in S. jlavidus is similar to that described for
S. paucispinus (Moser 1967a, b). However, seasonal
occurrence of Stage-Ill oocytes differs between the
two species. Stage-Ill oocytes in S. jlavidus decline
rapidly in number as yolk accumulation (Stage IV) is
initiated. They are not observed again in the ovaries
until after parturition and the beginning of the
reorganization of the lamellar branches. While Moser
(1967a) did not suggest a staging classification scheme,
his descriptions for S. paucispinus included oocytes
corresponding to Stage III (in the present study). In
contrast to S. jlavidus, these oocytes occurred in
ovaries of S. paucispinus throughout the year (Moser
1967a). The temporal difference in occurrence of Stage-
Ill oocytes between these two species is most likely a
reflection of the number of broods produced annually.
Viviparous species producing more than one brood an-
nually require a reserve of Stage-Ill oocytes. In rock-
fish where two or more broods of young are produced
in one reproductive season, a second clutch of yolked
oocytes develops concurrently with the initial brood of
gestating embryos. Moser (1967a) reported the second
clutch of yolked oocytes to occur in S. paucispinus
when the initial brood had reached eye-lens formation.
A distinct seasonal absence of Stage-Ill oocytes, or a
Bowers: Reproductive cycle of oocytes and embryos of Sebastes flavidus
241
clutch of yolked oocytes during embryonic gestation,
distinguishes single from multiple spawners.
There were approximately 30-40 days between the
appearance of fertilized ova and well-developed larvae
or recently-spawned females. Therefore, gestation ap-
pears to be 30-40 days in S. flavidus. Moser (1967b)
estimated 1-2 months gestation for the multiple-
spawner S. paucispinus, and Boehlert and Yoklavich
(1984) noted 37 days for S. melanops, a single-season
spawner. Similarly, Mizue (1959) compared a multiple-
spawner, S. marmoratus to a single-season spawner,
S. inermis. His data suggest approximately 30-45 days
for embryonic gestation in both species.
While the basic reproductive pattern among the
various Sebastes species is similar, variations exist in
reproductive strategy and life history (Boehlert and
Yoklavich 1984). Temporal variations in reproductive
seasonality of rockfishes are perhaps the most obvious
and, therefore, well documented. Releasing larvae over
an extended period of time increases the probability
that a portion of the reproductive population would en-
counter favorable environmental conditions for the sur-
vival of the progeny. Wyllie Echeverria (1987) listed
the peak parturition months for 34 species of Sebastes
and reported that larval extrusion occurs for up to 9
months in some species. In her study, from samples
collected over a 7-year period, the principal month of
parturition for yellowtail rockfish was February. In the
present study, and in more recent work (unpubl. data),
March was the peak month of parturition; however, the
samples were from a smaller geographical area. PhOlips
(1964), who sampled northern, central, and southern
California rockfish populations, determined S. flavidus
to be a "winter" spawner (November-March). Wyllie
Echeverria (1987) reported parturition for yellowtail
rockfish from north-central California to occur from
January to July. In the present study, a shorter par-
turition time was observed for the Cordell Bank
yellowtail population (January-March). This temporal
variance may reflect a clinal reproductive variation in
yellowtail rockfish populations. Care must be taken,
however, when interpreting and comparing results
where macroscopic characteristics are used. While field
assessments by microscopic staging of whole oocytes
or macroscopic examinations are less time-consuming,
validation by histological methods is required for
precise and detailed information (West 1990). Further-
more, studies on the impact of atresia and postovu-
latory follicles are relevant to understanding functional
relationships between yellowtail rockfish reproduction
and their environment. West (1990) suggests histology
as the appropriate method of use for these types of
studies.
A prolonged reproductive season is characteristic of
the genus Sebastes, but the factors regulating such a
mechanism are not clear. While temperature and
photoperiod appear to effect later spawning in higher-
latitude populations (Wooton 1984), inherent factors
may also play a key role in the prolonged seasonality
displayed by rockfishes. There is some evidence that
age, at least in yellowtail rockfish, may account for
some variation in parturition time within a season (M.J.
Bowers, unpubl. data; Eldridge et al. 1991). In addi-
tion, Boehlert and Yoklavich (1984) estimated 5 days
between hatching and birth in S. melanops, while par-
turition has been reported to occur immediately after
hatching in the ovary in the subgenus Sebasticus (Tsu-
kahara 1962). In this study, it could not be determined
if hatched larvae remained in the ovaries of yellowtail
rockfish. Further investigations are necessary to deter-
mine the occurrence, significance, and regulatory
mechansims of larval retention.
Rockfish are an important economic resource to the
Washington, Oregon, and California fisheries. Esti-
mates of total commercial rockfish landings in 1985
were 37,806 mt (PFMC 1990). In the same year, recrea-
tional anglers landed approximately 4000 mt of rock-
fish in California alone. Yellowtail, blue, and black
rockfishes represented 30% of the recreational landings
(Lenarz 1986). Fluctuations in year-class strength
cause the fishery to be somewhat unpredictable (Lenarz
1986), leaving it difficult for optimal management
strategies to protect stock depletion and establish
harvest guidelines. The earlier one can predict recruit-
ment success, the more precise management decisions
are likely to be. Leaman (1988) discussed the value of
directing management models toward biological prin-
ciples. Responses of yellowtail rockfish ovaries to
environmental fluctuations are early indicators of
reproductive performance. This study documents the
process of oocyte development in yellowtail rockfish
and provides a basis for interannual comparisons.
Acknowledgments
To Richard Powers, owner and operator of the New
Sea Angler, my sincerest appreciation for his skill and
assistance collecting samples. My greatest debt in con-
nection with this work is to Dr. Bruce MacFarlane
whose tireless enthusiasm, assistance, and patience are
beyond mortal explanation.
Citations
Boehlert, G.W., and M. Yoklavich
1984 Reproduction, embryonic energetics, and the maternal-
fetal relationships in the viviparous genus Sebastes (Pisces: Scor-
paenidae). Biol, Bull. (Woods Hole) 167:354-370.
242
Fishery Bulletin 90(2), 1992
Bowers, A.B., and F. Holliday
1961 Histological changes in the gonad associated with the
reproductive cycle of the herring (Clupea harengiis). Dep.
Agric. Fish. Scotl., Mar. Res. 5, p. 1-16.
Eldridge, M.B., J. Whipple, M. Bowers, B. Jarvis, and J. Gould
1991 Reproductive performances of yellowtail rockfish, Sebas-
tes flavidus. Environ. Biol. Fishes 30:91-102.
Howell, W.H.
1983 Seasonal changes in the ovaries of adult yellowtail
flounder, Limandaferraginea. Fish. Bull., U.S. 81:341-355.
Htun-Han, M.
1978 The reproductive biology of the dab Limanda limanda
(L.) in the North Sea: Seasonal changes in the ovary. J. Fish
Biol. 13:351-359.
Humason, G.L.
1967 Animal tissue techniques, 3d ed. W.H. Freeman, San
Francisco.
Leaman, B.M.
1988 Reproductive and population biology of Pacific ocean
perch (Sebastes alutus (Gilbert)). Ph.D. thesis, Univ. Brit. Col.,
Vancouver, 200 p.
Lenarz, W.H.
1986 A history of California rockfish fisheries. In Proc, Int.
Rockfish Symp., Oct. 1986, Anchorage, Alaska, p. 35-41.
Alaska Sea Grant Rep. 87-2, Univ. Alaska, Fairbanks.
Mizue, K.
1959 Studies on a scorpaenous fish Sebasticus marmoratiLs
Cuvier et Valenciennes. V. On the maturation and the seasonal
cycle of the ovaries of the marine ovoviviparous teleost. Bull.
Fac. Fish. Nagasaki Univ. 8:84-110 [in Jpn., Engl. summ.].
Moser, H.G.
1967a Seasonal histological changes in the gonads ofSebastodes
paucis-pinis (Ayres), an ovoviviparous teleost (family Scor-
paenidae). J. Morphol. 123:329-351.
1967b Reproduction and development of Sebastodes pauci-
spinis and comparison with other rockfishes off southern
California. Copeia 1967:773-797.
Oppenheimer, J.M.
1937 The noTTtisd stages of Fundulus heteroclitus. Anat. Rec.
68(1):1-15.
PFMC (Pacific Fishery Management Council)
1990 Status of the Pacific coast groundfish fishery through
1990 and recommended acceptable biological catches for
1991: Stock assessment and fishery evaluation. Document
prepared for the Council and its advisory entities. Pac. Fish.
Manage. Counc, Portland, OR 97201.
Phillips, J.B.
1964 Life history studies on ten species of rockfish (genus
Sebastodes). Calif. Dep. Fish Game, Fish Bull. 126, 70 p.
Robb, A.P.
1982 Histological observations on the reproductive biology of
the haddock, Melanogrammus aeglefirms (L.). J. Fish. Biol.
20:97-408.
Tsukahara, H.
1962 Studies on habits of coastal fishes in the Amakusa
Islands. 2. Early life history of the rockfish. Sebasticus mar-
moratus (Cuvier et Valenciennes). Rec. Oceanogr. Works
Jpn., Spec. 6:49-55.
Wallace, R.A., and K. Selman
1981 Cellular and dynamic aspects of oocyte growth in teleosts.
Am. Zool. 21:325-343.
West, G.
1990 Methods of assessing ovarian development in fishes: A
review. Aust. J. Mar. Freshwater Res. 41:199-222.
Wootton, R.J.
1984 Introduction: Strategies and tactics in fish reproduc-
tion. In Potts, G.W., and R.J. Wootton (eds.). Fish reproduc-
tion: Strategies and tactics, p. 1-12. Acad. Press, London.
Wourms, J. P., B.D. Grove, and J. Lombardi
1988 The maternal-embryonic relationship in viviparous fishes.
In Hoar, W.S., and D.J. Randall (eds.). Fish physiology. Vol.
XI, Part B. p. 1-34. Acad. Press, San Diego.
Wyllie Echeverria, T.
1987 Thirty-four species of California rockfishes: Maturity and
seasonality of reproduction. Fish. Bull, U.S. 85:229-250.
Yamamoto, K.
1956 Studies on the formation of fish eggs. I. Annual cycle
in the development of the ovarian eggs of the flounder, Liop-
setta obscura. J. Fac. Sci. Hokkaido Univ., Ser. VI Zool.
12:362-373.
Abstract. — We sampled jewfish
from recreational and commercial
catches in the eastern Gulf of Mex-
ico from November 1977 to January
1990 to obtain life history informa-
tion. A single annual minimum in
mean marginal increment ratios dur-
ing May- August supported the hy-
pothesis that jewfish up to at least
age 10 can be aged by counting the
opaque marks observed on otolith
sections. Annual opaque mark for-
mation was observed for a 3- and a
4-year-old jewfish using oxytetracy-
cline (OTC) reference marks on oto-
liths (sagittae). Male jewfish (A'' 41)
ranged 3-26 years old; females, 0-37
years (A'^ 85). Growth was similar for
males and females, averaging >100
mm/year through age 6, then slow-
ing to about 30 mm/year by age 15,
and finally declining to < 10 mm/year
after age 25. Observed total length
and age data were described well by
the following von Bertalanffy growth
model: total length (mm) = 2006
(1
3(-0.126
(age(yrs) + 0.49)))_ Jgwfish
spawned from June through Decem-
ber, with peak activity from July
through September. Male jewfish
matured at about 1100-1150mm
when 4-6 years old; females matured
at 1200-1350 mm when 6 or 7 years
old. The extensive overlap of length
and age distributions of males and
females, and the slight differences
between their sizes and ages at ma-
turity, prevent us from designating
jewfish as a protogynous herma-
phrodite. No transitional individuals
were found. Their relatively slow
growth, longevity, and behavioral
characteristics, such as the tenden-
cy to form spawning aggregations,
make jewfish populations highly sus-
ceptible to overfishing.
Age, growth, and reproduction of
jewfish Epinephelus itajara in
the eastern Gulf of Mexico
Lewis H. Bullock
Michael D. Murphy
Florida Marine Research Institute, Florida Department of Natural Resources
100 Eighth Avenue SE, St. Petersburg, Florida 33701-5095
Mark F. Godcharles
Southeast Region, National Marine Fisheries Service, NOAA
9450 Koger Boulevard, St. Petersburg, Florida 33702
Michael E. Mitchell
Florida Marine Research Institute, Florida Department of Natural Resources
1 481 -A Market Circle, Port Charlotte, Florida 33953
Manuscript accepted 11 March 1992.
Fishery Bulletin, U.S. 90:243-249 (1992).
The jevirfish Epinephelus itajara,
largest of the western North Atlan-
tic groupers (possibly reaching 455
kg; Robins et al. 1986), ranges from
the east coast of Florida throughout
the Gulf of Mexico, Caribbean Sea,
and south to Brazil (Smith 1971), and
also in the Pacific Ocean from Costa
Rica to Peru. Jewfish occur at depths
ranging from several meters (shallow
estuarine areas) to about 50 m. Juve-
niles can be found in holes and below
undercut ledges in swift tidal creeks
draining mangrove swamps. Large
adults occur both inshore around
structures such as piers and bridges,
and offshore around ledges and
wrecks (Bullock and Smith 1991).
Jewfish have recently been granted
protected status, eliminating harvest
in both the U.S. Exclusive Economic
Zone (NMFS 1990a, b) and Florida's
territorial waters (Florida Marine
Fisheries Commission 1990). Prior to
this designation, jewfish were cap-
tured by hook-and-line, speargun,
shark and grouper/snapper longlines,
and as a bycatch of shrimp trawling.
Historically, the majority of the U.S.
commercial catch has been landed
along the Florida Gulf coast, where
landings reached a high of approx-
imately 61,700 kg in 1988 (Fla. Dep.
Nat. Resour. Annual Landings Summ.,
Fla. Mar. Res. Inst., St. Petersburg,
unpubl. data).
A comprehensive study of jewfish
life history does not exist. Smith
(1971) discussed their systematics,
distribution, and ecology. Randall
(1967) described food habits from
nine individuals. Other researchers
have contributed incidental observa-
tions on diet (Beebe and Tee-Van
1928, Tabb and Manning 1961, Odum
1971), habitat (Smith 1976, Odum et
al. 1982), spawning (Schroeder 1924,
Colin 1990), and parasites/pseudo-
parasites (Breder and Nigrelli 1934,
Pearse 1934 and 1952, Manter 1947,
Olsen 1952). Bullock and Smith
(1991) provided basic life-history in-
formation on jewfish in the eastern
Gulf of Mexico, but did not dis-
cuss age and growth or size/age-at-
maturity. In this paper, we describe
age and growth, spawning seasonal-
ity, and approximate size- and age-
at-maturity for jewrfish in the east-
ern Gulf of Mexico. We also briefly
discuss the implications of these life-
history characteristics as they relate
to the jewfish's susceptibility to
overfishing.
243
244
Fishery Bulletin 90(2). 1992
Methods and materials
Jewfish were sampled aperiodically from recreational
and commercial catches from the eastern Gulf of Mex-
ico, November 1977 through January 1990. Fifty-six
percent (269/481) of the sampled jewfish were captured
using spearguns, 27% by hook-and-line, 8% by bottom
longline (either grouper/snapper or shark fisheries),
and the remaining 9% by shrimp trawl, trap, or un-
recorded methods. We attempted to determine sex,
whole (WW) and/or gutted (GW) weight (kg), and total
length (mmTL) for each specimen, although we could
not determine whole weight and sex when fish had been
eviscerated (A/" 271). Although eviscerated, unsexed fish
could not be included in our study of reproduction, they
were used in the age and growth analyses. If sagittae
could be located, they were removed from the otic cap-
sule (A'^ 384) and stored dry. A portion of the gonad,
if available (A'^ 173), was preserved in 10% formalin and
later transferred to 70% ethanol.
Otolith sections were examined for evidence of age
marks. Transverse sections, approximately 0.5mm
thick, were cut from each sagitta with a Buehler Isomet
low-speed saw. Sections were mounted on microscope
slides with Histomount mounting media and examined
for age marks under a dissecting microscope using
reflected light. Age marks were counted independent-
ly by two readers. Later, a joint reading was conducted
in an attempt to resolve differences between counts.
Monthly mean marginal increment ratios were cal-
culated for fish with 1-10 annuli to determine the
periodicity of mark formation. Marginal increment was
standardized for differences in growth among age-
classes by dividing the marginal increment for each fish
by the distance between its penultimate and outermost
annuli. We called this calculated value the 'marginal
increment ratio' (sensu percentage of marginal incre-
ment; Hood et al. Unpubl. manuscr.). Fish were as-
signed ages based on the number of annuli and a
biologically realistic hatching date of 1 September (time
of peak spawning; see Results). All fish were assigned
an age equal to their annulus count, except for fish
collected prior to 1 September and that had already
deposited an annulus during the most recent period of
mark deposition (April-August; see Results). The
assigned age for these fish was one less than their
number of annuli.
Observations to determine the validity of age marks
were made from two jewfish (290 mmTL, 509g; and
375 mmTL, 934 g) that were injected intramuscularly
with 50 mg oxytetracycline (OTC) per kg body weight
on 3 November 1990 and 21 October 1989, respective-
ly. These fish were maintained at ambient light and
temperature in flow-through 1038-gallon seawater
tanks located at the Keys Marine Laboratory in the
Florida Keys. The smaller specimen survived 11
months after OTC treatment; the larger fish was
sacrificed after 22 months.
Nonlinear regression of all available age and length
data (using FSAS; Saila et al. 1988) was used to esti-
mate parameters of the von Bertalanffy growth
equation,
It = L^(l-e(-K<t-t„))),
where If is total length (mm), t is age (years), L^ is
asymptotic length, K is the Brody growth coefficient,
and to is the age at zero length (von Bertalanffy 1957).
Likelihood ratio tests were used to compare male and
female von Bertalanffy parameter estimates (Kimura
1980, Cerrato 1990). Nonlinear regression was used to
fit the exponential equation, WW or GW = aTL'', to
whole- or gutted-weight and total-length data.
Histological preparations of gonads were made to
determine gonad developmental class, following the
criteria presented by Moe (1969) for red grouper Epi-
nephelus morio. Initially, gonad samples were em-
bedded in paraffin, but beginning with fish sampled in
1988, gonads were embedded in plastic (glycol meth-
acrylate) because of its superior tissue-infiltrating
abilities. Gonad samples were sectioned to a thickness
of 3.5f.im and stained with Weigert's hematoxylin and
eosin Y for microscopic examination. Spawning was in-
ferred from seasonal changes in the relative abundance
of fish having ovaries containing vitellogenic oocytes
or testes containing sperm in their efferent ducts. Sizes
or ages at maturity were determined from changes in
the proportion of mature fish over the entire age range
or across 50 mm size-classes.
Results
Age and growth
Opaque bands can be recognized and counted on thin-
sectioned jewfish sagittae. Initial counts of opaque
bands by two independent readers agreed on 62%
(237/384 fish) of sections analyzed, with 91% (348/384)
of all counts either in agreement or differing by one.
After a second, joint reading, agreement was reached
on opaque-band counts for all but two sections, leav-
ing 382 specimens for analysis of age and growrth.
The annual pattern of monthly mean marginal incre-
ment ratios and observations from two OTC-marked
jewfish support the hypothesis that annuli form once
each year. Mean marginal increment ratios were
greater than 70% during November- April and declined
to a minimum of 20% in June. The mean marginal in-
crement ratio remained less than 30% through August
(Fig. 1). For a large number of specimens captured
Bullock et al.: Age, growth, and reproduction of Epinephelus itajara
245
1 0^
09-
6
o
08-
\
2
5
]
3
^
3
1-
a:
0 7-
\
« -
r 4
\
/-^
^,
-1
\
/
EC
O
05-
\
^
<
04-
\
4
/
z
s
>
C3
03-
\
()
CC
<
18
■5.
0.2-
IB
0.1-
JAN
FEB
MAR
AP
R MAY JUN JUL
MONTH
AUG SEP
OCT NOV
DEC
Table 1
Number of aged jewfish Epinephel
us itajara from the eastern Gulf of Mexico, and 1
average observed and predicted total lengths for age groups
0-37 years
old. Pre-
dieted
lengths are
based on the von Bertalanffy growth equation mm
rL = 2006
(l-e<-
0.126(aee(>T)+0.49))^
Age
Average
observed total
N Female
length (mm)
N Unknown
Predicted total length (mm)
N
Male
Male
Female
Pooled*
0
0
1
338
4
170
1
0
—
3
517
0
—
434
382
344
2
0
—
5
717
1
711
605
563
541
3
2
863
4
708
4
751
759
725
714
4
1
1184
3
924
4
913
897
871
867
5
2
1080
1
1218
8
1067
1021
1002
1002
6
4
1078
0
—
5
1161
1132
1119
1121
7
5
1318
0
—
4
1423
1232
1225
1226
8
2
1476
2
1333
8
1437
1322
1319
1318
9
2
1400
2
1399
12
1368
1403
1404
1400
10
2
1398
6
1515
12
1516
1475
1481
1471
11
1
1660
6
1632
23
1544
1540
1549
1535
12
1
1690
7
1647
31
1612
1598
1611
1590
13
5
1620
10
1653
26
1644
1651
1666
1640
14
2
1849
7
1762
15
1723
1698
1715
1683
15
4
1828
4
1913
12
1737
1740
1760
1721
16
3
1909
4
1860
8
1735
1778
1800
1755
17
1
1770
2
1878
6
1879
1812
1836
1785
18
0
—
4
1820
8
1750
1843
1868
1811
19
0
—
0
—
6
1833
1870
1897
1834
20
0
—
2
1990
11
1842
1895
1923
1854
21
0
—
4
2023
8
1818
1917
1946
1872
22
0
—
2
2011
9
1820
1937
1967
1888
23
0
—
0
—
4
1938
1955
1986
1902
24
1
1905
1
1950
7
1936
1971
2003
1914
25
2
1955
0
—
4
1821
1985
2018
1925
26
1
1930
0
—
5
1891
1998
2032
1935
27
0
—
2
2065
3
1853
2010
2044
1943
28
0
—
1
1935
2
2006
2020
2055
1951
29
0
—
0
—
1
2090
2030
2065
1957
30
0
—
0
—
1
2040
2038
2073
1963
33
0
—
1
2015
2
1820
2058
2095
1977
34
0
—
0
—
1
2032
2064
2101
1980
36
0
—
0
—
1
1908
2073
2110
1986
37
0
—
1 1970
inknown sex
0
2077
2115
1988
* Including fish of i
Figure 1
Monthly mean marginal increment ratios for jewfish with 1-10
opaque marks on otolith sections. Vertical lines indicate range
of observations; sample size indicated by the number adjacent
to the mean (N 86).
during April-August, we observed an opaque band at
the outer edge of the otoHth section and interpreted
this as the deposition of a new annulus. After August,
the mean marginal increment ratio- increased until
reaching a maximum during November- April. The
observed annual minimum in monthly mean marginal
increment ratios suggests that
opaque bands form once each year
in the otoliths of jewfish < 10 years
of age. The validity of age marks
was confirmed for two OTC-marked
jewfish. The OTC reference mark
was clearly evident on otoliths of
each of the two specimens. The
otolith of the 3-year-old jewfish that
had survived for 1 1 months in cap-
tivity contained a single annulus
distal to the OTC mark. This fish,
injected with OTC in November
1990, had apparently deposited an
annulus prior to its death in October
1991. Its total length and weight
were 505mmTL and 2.7kg. The
4-year-old specimen, injected with
OTC in October 1989 and sacrificed
in August 1991 after 22 months in
captivity (total length and weight
of ~735mmTL and 9kg), had de-
posited two annuli distal to the OTC
reference mark.
A total of 481 jewfish were sam-
pled for life-history data. Age data
were determined for 382 individ-
uals. Age range was 3-26 years for
males {N 41), 0-37 for females (N
85), and 0-36 for fish of undeter-
mined sex (N 256). Total lengths of
jewfish sampled were 795-2057 mm
for males (N 75), 338-2155 mm for
females {N 131), and 75-2160mm
for fish of undetermined sex {N 275).
Jewfish grow slowly relative to
their potential maximum size. An-
nual growth was most rapid (aver-
aging > 100 mm/year) through age
6, then declined to about 30 mm/
year by age 15, and to less than
10 mm/year after age 25 (Table 1,
246
Fishery Bulletin 90(2|. 1992
AGE (yrs)
Figure 2
Observed ages (years) and total lengths (mm), and predicted
growth modeled for jewfish collected from the eastern Gulf
of Mexico. Parameters for the von Bertalanffy growth model
are L„=2006mmTL, K = 0.126/year, and t„ = -0.49 year
{N 382).
Fig. 2). Average observed and predicted (von Berta-
lanffy equation) sizes-at-age were similar between
sexes (Table 1). Results of likelihood ratio tests indi-
cated no significant differences between sex-specific
estimates of L^ (x^ 0.136, df 1, P>0.70), K (x^ 4.0 x
10-5, df 1, P>0.90), or to (jc^ 0.138, df 1, P>0.70).
Estimates of the growth equation parameters (asymp-
totic standard error) for pooled length and age data
were L^ =2006mmTL (23.3), K = 0.126/year (0.0057),
and to = -0.49 years (0.200).
The relationships of whole and gutted weight (kg) to
total length (mm) were
WW = 1.31 X 10-8TL3056 (jv 66, r^ 0.964)
GW = 2.94 X 10-8TL2911 (iv 402, r'^ 0.941).
Gutted and whole weights were linearly related (A^ 50,
r2 0.995) as follows:
WW =
-0.717 + 1.1039GW
GW =
1.001 -1- 0.9018WW.
Reproduction
Jewfish spawn during June-December in the eastern
Gulf of Mexico, with peak activity during July- Sep-
tember (Fig. 3). Ripe males and females first appeared
in our collections during June. Nearly all gonads col-
I MATURE RESTING
g RIPENING
□ RIPE
Q POST SPAWNING
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
MONTH
g MATURE RESTING
n RIPE
Q POST-SPAWNING
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
MONTH
Figure 3
Monthly distribution of gonad classes for mature (upper) male
(A^ 66) and (lower) female (A^ 86) jewfish collected from the
eastern Gulf of Mexico. Histological criteria for gonad classes
correspond to Moe's (1969) descriptions for red grouper.
lected from mature fish in July, August, and September
were classified as ripe. Most spawning appears to end
by October, although a ripe male was collected in
November and another in December. Spent fish were
collected in August (one male) and October (one
female). No transitional fish were found.
Male jewfish become sexually mature at a slightly
smaller size and younger age than females. Male jew-
fish were first mature when about 1100-1150mmTL
at 4-6 years of age. All males <1150mmTL (A'^ 6,
795-llOOmm) were immature, whereas all larger
males in our samples (A^ 55, 11 55-2057 mm) were
mature. Both 3-year-old males sampled were imma-
ture, whereas one large (1184 mmTL) 4-year-old male
was mature. Fifty percent of males 5 or 6 years old
(3 of 6 individuals) and all males age 7 or older (A'^ 31)
were mature. Female jewfish first mature when about
1200-1350 mmTL at age 6 or 7. All females <1225
mm (A^ 21, 338-1218mm) and <6 years old (A^ 17, 0-5
yr) were immature. All larger (A'^ 90, 1350-2155 mm)
and older (A'^ 68, 8-37 yr) females sampled were
mature.
Bullock et al : Age, growth, and reproduction of Epinephelus itajara
247
Discussion
Age and growth
The spring-summer period of annulus formation in
jewfish seems somewliat protracted. However, there
appears to be a considerable range for the duration of
annulus deposition in grouper populations: 2 or 3
months for Epinephelus morio, E. nigritis, E. drum-
mondhayi, and E. niveatus (Moe 1969, Matheson and
Huntsman 1984, Moore and Labisky 1984, Manooch
and Mason 1987) to 5-7 months for Mycteroperca
phenax and M. microlepis (Matheson et al. 1986, Hood
and Schlieder 1992). Moe (1969) discussed factors af-
fecting annulus formation and concluded that spawn-
ing and its associated physiological processes probably
caused annulus formation in red grouper. However, an-
nulus formation does not always occur in phase with
spawning in epinephelines. For example, Matheson et
al. (1986) found M. phenax to spawn during April-
August in the South Atlantic Bight, but annulus for-
mation occurred during December- April.
The annual deposition of opaque bands, seen in 3- and
4-year old OTC-marked jevirfish, needs to be validated
for fish older than 10 years. Due to the difficulty in
sampling large numbers of these older fish year-round,
it is probably not feasible to utilize indirect validation
techniques (i.e., marginal increment analysis). Valida-
tion will probably require direct observations of in-
dividuals that have been injected with OTC and recap-
tured after annulus deposition.
The growth rate of jewfish (i.e., K 0.13/year) falls
within or near the range observed for some of its con-
geners in the South Atlantic Bight and Gulf of Mex-
ico: speckled hind, 0.13/year (Matheson and Himtsman
1984) and red grouper, 0.11-0. 18/year (Moe 1969,
Muhlia-Melo 1975). However, jewfish growth is some-
what faster than that of the deepwater snowy grouper
E. niveatus {K 0.07-0.09/year; Matheson and Hunts-
man 1984, Moore and Labisky 1984) and considerably
greater than that of the second-largest grouper in the
western North Atlantic Ocean, the warsaw grouper
E. nigritis {K 0.05/year; Manooch and Mason 1987),
which may reach weights > 200 kg.
Reproduction
We foimd jewfish to be in peak spawning condition dur-
ing July- September in the eastern Gulf of Mexico. This
agrees with Schroeder's (1924) finding that jewfish
spawned during July- August, when heavily exploited
aggregations of jewfish appeared off the Florida
Keys. Furthermore, Colin (1990) observed what he
interpreted as courtship behavior in jewfish off south-
west Florida during the full moons of August and
September.
When compared with that of females, the slightly
smaller size and younger age of males at first matur-
ity is unexpected, given that jewfish are assumed to
be protogynous hermaphrodites (Smith 1971). Further-
more, whereas the youngest fish in our sample was
female, as would be expected for a protogynous fish,
so was the oldest. However, Sadovy and Shapiro (1987)
point out several factors that may obscure differences
in length, age, and maturity between males and females
of a protogynous fish: (1) Some females may never
change sex for lack of genetic or environmental cues
and therefore may attain sizes (ages) equal to or greater
than males, (2) a fraction of the population may initiate
female development but change to males prior to sex-
ual maturation, and (3) size at sex-reversal may differ
among subpopulations of the same species and thus
may obscure differences in length or age distribution
between the sexes. Conclusive evidence for proto-
gynous hermaphroditism in jewfish (i.e., the presence
of transitional individuals) was not found in this study.
Transitional individuals in confirmed protogynous
hermaphrodites, such as E. morio (Moe 1969) and
M. microlepis (Collins et al. 1987, Hood and Schlieder
1992), never represent a large percentage of the
population; therefore, more extensive collections than
ours may be needed to detect the presence of these
individuals.
Fisheries implications
The life-history characteristics that we describe imply
that jewrfish are highly vulnerable to overfishing. Their
slow growth, longevity, and presumed low natural mor-
tality specify a population composed of cohorts that
reach their maximum biomass at relatively old ages
(Alverson and Carney 1975). Thus the greatest yield
from a cohort of jewfish would be attained at either
low rates of fishing or when only large fish are har-
vested. If jewfish are indeed protogynous hermaphro-
dites, fishing may also disrupt their spawning and
recruitment by limiting the number of older males
available for spawning (Smith 1982, Bannerot et al.
1987, Huntsman and Waters 1987). In addition, be-
havioral traits exhibited by large jewfish, such as their
general unwariness of spearfishermen and apparent
site-specific spawning aggregations (Shroeder 1924,
Colin 1990), make them readily available for capture.
Fisheries managers of Florida territorial and U.S. Ex-
clusive Economic Zone waters have recognized the
jewfish's susceptibOity to overfishing and have recently
banned all harvest of jewrfish from waters under their
jurisdictions.
248
Fishery Bulletin 90(2), 1992
Acknowledgments
We would like to thank the following fishermen for
providing juvenile specimens: D. Bellamy, I. Bellamy,
J. Rhodes, and R. Woodring. This study would not
have been possible without the fine cooperation from
the following: T. Nachman, G. Brown, G. Migliano,
M. Nahon, W. Tappan, F. Devens, D. Harger,
E. McManus, R. Ruiz-Carus, J. Swanson, W. Gibbs,
W. Bell, and Capt. T. Reynolds. Special thanks go to
Capt. D. DeMaria, who not only permitted us to
thoroughly sample his catches but was also the major
impetus for bringing about protection for the jewfish
in Florida.
This manuscript benefited from review by R. Taylor,
R. Crabtree, and P. Hood of the Florida Marine
Research Institute, Department of Natural Resources.
This study was partially supported by funds provided
to the State of Florida under PL 99-659 of the Depart-
ment of Commerce.
Citations
Alverson. D.L., and M.J. Carney
1975 A graphic review of the growth and decay of population
cohorts. J. Cons. Cons. Int. Explor. Mer 36:133-143.
Bannerot, S., W.W. Fox Jr., and J.E. Powers
1987 Reproductive strategies and the management of snap-
pers and groupers in the Gulf of Mexico and Caribbean. In
Polovina, J.J., and S. Ralston (eds.), Tropical snappers and
groupers: Biology and fisheries management, p. 561-603.
Westview Press, Boulder.
Beebe, W., and J. Tee-Van
1928 The fishes of Port-au-Prince Bay, Haiti, with a summary
of the known species of marine fish of the Island of Haiti and
Santo Domingo. Zoologica (NY) 10:1-279.
Breder, CM. Jr., and R.F. Nigrelli
1934 The penetration of a grouper's digestive tract by a sharp-
tailed eel. Copeia 1934(4):162-164.
Bullock, L.H., and G.B. Smith
1991 Seabasses (Pisces: Serranidae). Mem. Hourglass Cruises
VIII (II), 243 p.
Cerrato, R.M.
1990 Interpretable statistical tests for growth comparisons
using parameters in the von Bertalanffy equation. Can. J.
Fish. Aquat. Sci. 47:1416-1426.
Colin, P.L.
1990 Preliminary investigations of reproductive activity of the
jewfish, Epinephelus itajara. In Abstr. Gulf Caribb. Fish.
Inst., 43rd Annu. Sess., Miami, 4-9 Nov. 1990, p. 9, 10.
Collins. M.R., C.W. Waltz, W.A. Roumillat, and D.L. Stubbs
1987 Contribution to the life history and reproductive biology
of gag, Mycteroperca microlepis (Serranidae), in the South
Atlantic Bight. Fish. Bull., U.S. 85:648-653.
Florida Marine Fisheries Commission
1990 Jewfish: Designation as protected species; prohibition
of harvest in state waters; possession, landing, and sale pro-
hibited. Florida Administrative Code 46-14.008, Tallahassee.
Hood, P.B., and R.C. Schlieder
1992 Age, growth and reproduction of gag, Mycteroperca
microlepis, in the eastern Gulf of Mexico. Bull. Mar. Sci. 51(3).
Hood, P.B., M.F. Godcharles, and R. Barco
Unpubl. Age, growth, reproduction, and the feeding ecology
of black sea bass, Ceritropristis striata (Pisces: Serranidae).
Fla. Mar. Res. Inst., Fla. Dep. Nat. Resour., St. Petersburg,
38 p.
Huntsman, G.R., and J.R. Waters
1987 Development of management plans for reef fishes— Gulf
of Mexico and U.S. South Atlantic. In Polovina, J. J., and
S. Ralston (eds.). Tropical snappers and groupers: Biology and
fisheries management, p. 533-560. Westview Press, Boulder.
Kimura, D.K.
1980 Likelihood methods for the von Bertalanffy growth curve.
Fish. Bull., U.S. 77:765-776.
Manooch, C.S. HI, and D.L. Mason
1987 Age and growth of the Warsaw grouper and black grouper
from the southeast region of the United States. Northeast
Gulf Sci. 9:65-75.
Manter, H.W.
1947 The digenetic trematodes of marine fishes of Tortugas,
Florida. Am. Midi. Nat. 38:257-416.
Matheson, R.H. HI, and G.R. Huntsman
1984 Growth, mortality, and yield-per-recruit models for
speckled hind and snowy grouper from the United States South
Atlantic Bight. Trans. Am. Fish. Soc. 113:607-616.
Matheson, R.H. Ill, G.R. Huntsman, and C.S. Manooch III
1986 Age, growth, mortality, food, and reproduction of the
scamp, Mycteroperca phenax, collected off North Carolina and
South Carolina. Bull. Mar. Sci. 38:300-312.
Moe, M.A. Jr.
1969 Biology of the red grouper Epinephelus morio (Valen-
ciennes) from the eastern Gulf of Mexico. Fla. Dep. Nat.
Resour. Mar. Res. Lab. Prof. Pap. Ser. 10, 95 p.
Moore, CM., and R.F. Labisky
1984 Population parameters of a relatively unexploited stock
of snowy grouper in the lower Florida Keys. Trans. Am. Fish.
Soc. 113:322-329.
Muhlia-Melo, A.
1975 Aspectos biol6gicos pesqueros de Epinephelus morio
(Val.). M.S. thesis, Univ. Nac. Aut6noma Mexico, Mexico City,
D.F., 68 p.
NMFS (National Marine Fisheries Service)
1990a Fishery management plan for the reef fish of the Gulf
of Mexico, Amendment 2. Gulf of Mexico Fish. Manage.
Counc, Tampa, 20 p.
1990b Fishery management plan for the snapper-grouper
fishery of the South Atlantic Region, Amendment 2. South
Atl. Fish. Manage. Counc, Charleston, 28 p.
Odum, W.E.
1971 Pathways of energy flow in a south Florida estuary.
Univ. Miami Sea Grant Prog. Sea Grant Tech. Bull. 7, 162 p.
Odum, W.E., CC Mclvor, and T.J. Smith HI
1982 The ecology of the mangroves of south Florida: A com-
munity profile. U.S. Fish Wildl. Serv. Biol. Serv. FWS-
OBS-81/24, 144 p.
Olsen, L.S.
1952 Some nematodes parasitic in marine fishes. Publ. Inst.
Mar. Sci. Univ. Tex. 2:173-215.
Pearse, A.S.
1934 Observations on the parasites and commensals found
associated with crustaceans and fishes at Dry Tortugas,
Florida. Carnegie Inst. Washington Publ. Pap. Tortugas Lab.
28(435):104-115.
Bullock et a\ Age, growth, and reproduction of Epinephelus ttajara
249
1952 Parasitic Crustacea from the Texas coast. Publ. Inst.
Mar. Sci. Univ. Tex. 2(2):5-42.
Randall, J.E.
1967 Food habits of reef fishes of the West Indies. Stud. Trop.
Oceanogr. Inst. Mar. Sci. Univ. Miami 5:665-847.
Robins, C.R., G.C. Ray, J. Douglass, and R. Freund
1986 A field guide to Atlantic coast fishes of North America.
Peterson Field Guide Ser. 32, Houghton Mifflin, Boston, 354 p.
Sadovy, Y., and D.Y. Shapiro
1987 Criteria for the diagnosis of hermaphroditism in fishes.
Copeia 1987:136-156.
Saila, S.B., C.W. Recksiek, and M.H. Prager
1988 Basic fishery science programs, a compendium of micro-
computer programs and manual of operation. Elsevier Sci.
Publ., NY, 230 p.
Schroeder, W.C.
1924 Fisheries of Key West and the clam industry of southern
Florida. Appendix XII to the Report of the U.S. Commissioner
of Fisheries for 1923. Bur. Fish. Doc. 962, 74 p.
Smith, C.L.
1971 A revision of the American groupers: Epinephelus and
allied genera. Bull. Am. Mus. Nat. Hist. 146:67-242,
1982 Patterns of reproduction in coral reef fishes, hi Hunts-
man, G.R., W.R. Nicholson, and W.W. Fox Jr. (eds.), The bio-
logical bases for reef fishery management, p. 49-66. NOAA
Tech. Memo. NOAA-TM-NMFS-SEFC-80, Beaufort Lab.,
Southeast Fish. Sci. Cent., Beaufort, NC.
Smith, G.B.
1976 Ecology and distribution of eastern Gulf of Mexico reef
fishes. Fla. Mar. Res. Publ. 19, 78 p.
Tabb, D.C., and R.B. Manning
1961 A checklist of the flora and fauna of northern Florida
Bay and adjacent brackish waters of the Florida mainland
collected during the period July 1957 through September
1960. Bull, Mar, Sci, Gulf Caribb, 11:552-649,
von Bertalanffy, L.
1957 Quantitative laws in metabolism and growth, Q, Rev,
Biol, 32:217-231,
Abstract.- Allozymes were used
tx) examine spatial and temporal com-
ponents of genetic variation among
populations of queen conch in the
Florida Keys and Bimini over a 4-
year period. Spatial and temporal
genetic variation were both signifi-
cant (F< 0.001) despite high levels of
genetic similarity among samples
(mean Nei's I, 0.994). However, no
consistent pattern of variation was
observed. The gene diversity among
localities (Glt 0.50%) did not differ
significantly (P>0.05) from the diver-
sity among years or samples within
locahties (Gsl 0.60%). In addition,
Florida Keys and Bimini populations
were very similar genetically to
those studied previously in the Carib-
bean Sea and Bermuda (mean Nei's
I, 0.988). In general, populations of
queen conch appear to be structured
as a mosaic of spatial and temporal
genetic patchiness within a continu-
um of high genetic similarity. This
genetic similarity is presumably main-
tained by larval drift and gene flow.
However, the observed patterns of
genetic variation suggest a dynamic
population structure. This structure
may reflect presettlement stochastic
events and processes in the marine
environment.
Genetic patchiness among
populations of queen concli
Strombus gigas in tlie
Florida Keys and Bimini*
Donald E. Campton
Department of Fisheries and Aquaculture, University of Florida
7922 NW 7 1 St Street, Gainesville, Florida 32606-0300
Carl J. Berg Jr.
Florida Marine Research Institute, Florida Department of Natural Resources
13365 Overseas Highway, Marathon, Florida 33050
Current address: P,0 Box 769, Kilauea, Hawaii 90754
Lynn M. Roblson
Department of Fisheries and Aquaculture, University of Florida
7922 NW 71st Street, Gainesville, Florida 32606-0300
Current address: Department of Fisheries and Allied Aquacultures
Auburn University, Auburn, Alabama 36849
Robert A. Glazer
Florida Marine Research Institute, Florida Department of Natural Resources
13365 Overseas Highway, Marathon, Florida 33050
The queen conch Strombus gigas is a
large marine gastropod of significant
economic importance to the Carib-
bean Sea area (reviewed by Berg and
Olsen 1989). The native range of the
species extends from south Florida
to Venezuela and eastward from
Central America to the Bahama and
West Indies Islands. An isolated
population also inhabits the coastal
waters of Bermuda. The species has
been heavily exploited in commercial,
recreational, and subsistance fisher-
ies throughout its geographic range.
Many populations are considered
depleted or overfished.
The life history of queen conch sug-
gests the potential for extensive gene
flow through larval dispersal (Schel-
tema 1971, 1986). Laboratory studies
indicate that larvae maintain the
planktonic stage for 12-35 days (x
21 days) before settling and meta-
Manuscript accepted 4 March 1992.
Fishery Bulletin, U.S. 90:250-259 (1992).
* Journal Series R-01534, Florida Agricultural
Experiment Station.
morphosis (Ballantine and Apple-
doorn 1983, Davis and Hesse 1983).
Larvae entrained in swift, Caribbean
currents (l-3km/h) could thus be
transported significant distances
(Kinder et al. 1985). However, disper-
sal and recruitment patterns of S.
gigas during the planktonic stage are
largely unknown. Effective manage-
ment and rehabilitation of the spe-
cies throughout its geographic range
necessitate an understanding of pop-
ulation structure, patterns of gene
flow, and genetic relationships.
In a recent allozyme study, Mitton
et al. (1989) found a high level of
genetic similarity among populations
of queen conch from eight localities
throughout the Caribbean Sea. How-
ever, significant spatial heterogen-
eity in allele frequencies indicated
that the sampled populations were
not totally panmictic. In addition,
allele frequencies for the geograph-
ically disjunct population of Bermuda
were distinctive at one locus.
250
Campton et al.: Genetic patchiness among Strombus gigas populations
251
In the study described here, allozymes were used to
examine the genetic structure of queen conch popula-
tions in the Florida Keys and Bimini. We collected
conch from the same localities in multiple years to com-
pare spatial and temporal components of genetic varia-
tion. Testing the relative significance of those two com-
ponents was a major objective of our study.
Materials and methods
Sampled populations
Queen conch were collected between 1987 and 1990
from four localities in the Florida Keys and from
Bimini, a linear distance of approximately 350 km
(Table 1, Fig. 1). Samples of conch were obtained in
multiple years from Ballast Key, Coffins Patch, and
Craig Key. Single samples were obtained from Key Bis-
cajme and Bimini. All animals were collected by scuba
diving or snorkeling.
The shell length, or major axis, of each conch was
measured with calipers to the nearest mm. Based on
size distributions, the Coffins Patch population (or ag-
gregation) appeared to be a single year-class or cohort
that we sampled in three consecutive years (Table 1).
All other populations represented mixtures of year-
classes with new recruits added each year.
Tissues
Conch collected in 1987 (three samples) were processed
according to the methods of Mitton et al. (1989): Only
the distal tip of the digestive gland, including gonad
and associated connective tissue, was retained for en-
zyme extraction. We were not able to resolve some of
the enzymes or presumptive loci reported by Mitton
et al. (1989) but were able to resolve some enzymes and
loci not examined previously. Consequently, from each
of 12 conch collected from Ballast Key in February,
1988 (Table 1), we dissected six tissues for further
screening of enzymes and loci: (1) foot muscle, (2)
proboscis with radula, (3) eyes and eyestalks, (4)
crystalline style, (5) mantle tissue, and (6) distal tip of
the digestive gland (Little 1965). Thirty-eight enzjmies
Table 1
Means ± standard errors (SE) and
ranges of shell
length for
samples of queer
conch Strombus
gigas collected from four 1
localities in the Florida
Keys and Bimini, 1987-90.
Locality
Year
N
Length (
mm)
Mean ± SE
Range
Ballast Key
1987
56
167 ± 6.2
96-241
1988"
12
125 ± 5.6
91-157
1988''
30
135 ± 4.2
69-167
Coffins Patch
1988
100
58 ± 0.4
46-71
1989
102
98 ± 0.6
87-112
1990
100
142 ± 0.9
117-170
Craig Key
1987
105
177 ± 2.5
127-267
1989
71
225 ± 3.5
137-257
1990
92
187 ± 3.5
125-252
Key Biscayne
1987
79
155 ± 3.5
99-216
Bimini
1989
tion.
96
194 ± 2.0
127-236
"February collec
''April collection.
Figure 1
Localities from which queen conch
Strombus gigas were collected. For
the study described here, conch were
collected from (1) Ballast Key, (2) Cof-
fins Patch, (3) Craig Key, (4) Key Bis-
cayne, and (5) Bimini. Localities 6-14
are from Mitton et al. (1989): (6) Ber-
muda, 1 site; (7) Turks and Caicos
Islands, 4 sites; (8) St. Kitts, 2 sites;
(9) Nevis, 1 site; (10) St. Lucia, 2 sites;
(11) Bequia, 1 site; (12) Barbados, 1
site; (13) Grenadines, 3 sites; and
(14) Belize, 2 samples, 1 each of the
normal and melanic forms.
252
Fishery Bulletin 90(2|. 1992
were assayed in each tissue using
a variety of electrophoresis buffers.
On the basis of the aforemen-
tioned analyses, three tissues
were retained from all conch col-
lected subsequently in 1988-90:
(1) foot muscle, (2) proboscis with
radula, and (3) digestive gland
with gonad. The three tissues
were dissected from each individ-
ual, placed in separate cryotubes
or plastic bags, and frozen in the
field with liquid nitrogen or dry
ice. All tissues were stored at
-80°C until prepared for en-
zyme extraction.
Electrophoresis
Allozymes were detected by hori-
zontal starch-gel electrophore-
sis following the procedures of
Aebersold et al. (1987). Enzymes
were extracted by homogenizing
each tissue separately in 0.5-1.0
volumes of 0.05M PIPES, 0.05%
Triton X-100, and 0.2 mM pyri-
doxal-5'-phosphate (adjusted to
pH 6.8 with l.OM NaOH). Gels
were prepared with a 12.5% mix-
ture (wt:vol) of Connaught starch
(Fisher Sci. Co.) and one of five
buffer solutions (Table 2). Histo-
chemical staining of gels followed
standard procedures (Morizot and Schmidt 1990). Gels
were stained by agar overlay for all enzymes except
AAT.
Presumptive loci and alleles were designated by the
nomenclature system outlined by Shaklee et al. (1990),
except peptidase loci were identified by their di- or tri-
peptide acronyms (DPEP, TPEP). Multiple loci of a
particular enzyme were designated numerically (1,2,
etc.) from fastest to slowest anodic mobility. Alleles of
a particular locus were designated by their relative,
anodic mobilities (most frequent allele = *100).
Statistics
Genotypic proportions at each locus were tested for
goodness-of-fit to Hardy-Weinberg expectations using
the likelihood-ratio test or G -statistic (Sokal and Rohlf
1981). Allele frequencies at each locus were tested for
homogeneity among samples by contingency table
(samples x alleles) G -tests (Sokal and Rohlf 1981). This
total G -statistic, or likelihood ratio, was then parti-
Table 2
Enzymes and loci resolved in queen
conch Strombus gigas.
Tissues
ire digestive
gland (D), foot muscle (F), and proboscis (P).
Enzyme
Optimum
Enzyme
number
Locus
Tissue
buffer-
Aspartate aminotransferase
2.6.1.1
AAT-1*
D,P
TC
AAT-2'
F,P
TC
Argenine kinase
2.7.3.3
ARGK'
F
TC
Dipeptidase""
3.4.13.11
"DPEP-l'
D,F,P
TLBC-2, TC
(substrates: Leu-Ala, Leu-Tyr)
DPEP-2'
D.F.P
TLBC-1
DPEP-S*
D.F.P
TLBC-1
Glucose-6-phosphate dehydrogenase
1.1.1.49
G6PDH'
D
TLBC-2
Glucose-6-phosphate isomerase
5.3.1.9
GPr
D,F,P
TC
Isocitrate dehydrogenase (NADP* )
1.1.1.42
IDHP-1*
D,F,P
AC
IDHP-2'
D.F.P
AC
Malate dehydrogenase''
1.1.1.37
MDH-r
F,P
AC
*'MDH-2'
D,P
AC
Octopine dehydrogenase
1.5.1.11
ODH'
F.P
TLBC-1
Phosphoglucomutase''
5.4.2.2
"PGM-l*
D,F,P
TC
^PGM-2'
D.F.P
TC
PGM-3'
F
TC
Phosphogluconate dehydrogenase''
1.1.1.44
^PGDH'
F.P
AC
Tripeptide aminopeptidase
3.4.11.4
TPEP-l"
D,F,P
TLBC-1
(substrate: Leu-Gly-Gly)
TPEP-2' D.F.P TLBC-1
7.5 with N-(3-aminopropyl) morpholine (Clayton
"AC: 0.04 M citric acid adjusted to p?
and Tretiak 1972); TBE: Tris-borate-EDTA, pH
8.6 (Boyeretal. 1963);
TC: TC buffer
of Siciliano and Shaw 1976; TLBC-1
: LiOH buffer of Ridgwayetal. (1970); TLBC-2:
LiOH buffer of Selander et al. (1971).
"•Enzymes and loci assayed also by Mitton et al.
(1989).
tioned into hierarchical components representing
temporal and spatial components of genetic variation
within and among localities, respectively (e.g., Smouse
and Ward 1978). An approximate F -ratio was then con-
structed as (G among localities/df)/(G among years
within localities/df) to test whether the genetic hetero-
geneity among localities was significantly greater than
the heterogeneity among years within localities. The
total gene diversity (Nei 1973) was similarly partitioned
into within- and among-locality components following
the algorithm of Chakraborty et al. (1982). In all tests
of statistical significance, significance probabilities
were adjusted for the number of tests (loci) evaluated
simultaneously (Rice 1989).
Nei's (1972) index of gene identity was calculated be-
tween all population samples. The genetic similarities
among all populations, including those sampled by Mit-
ton et al. (1989), were represented graphically in a
UPGMA dendrogram (Sneath and Sokal 1973). The
1987 sample from Ballast Key was excluded from these
latter analyses because of small sample size (n 12).
Campton et al,: Genetic patchiness among Strombus gigas populations
253
Table 3
Allele frequencies for
samples ol
queen conch Strombus gigas
from the Florida Keys (4 si
tes) and Bimini. "
MD" indicates
no data
for that locus.
Key
Locus
Alleles
Ballast Key
Coffins Patch
Craig Key
Biscayne
1987
Bimini
1987
1988'
1988"
1988
1989
1990
1987
1989
1990
1990
AAT-1'
100
0.713
1.00
0.62
0.723
0.755
0.696
0.663
0.671
0.678
0.648
0.591
120
0.287
—
0.38
0.277
0.245
0.304
0.338
0.329
0.322
0.352
0.403
130
—
—
—
—
—
—
—
—
-
-
0.005
AAT-2'
100
ND
1.00
1.00
0.975
0.980
0.955
ND
0.993
0.973
ND
0.973
150
—
—
0.025
0.020
0.045
0.007
0.022
0.016
31
—
_
_
—
—
—
_
0.011
180
—
—
—
—
-
—
0.005
-
DPEP-1*
100
0.472
0.42
0.57
0.536
0.536
0.460
0.567
0.521
0.522
0.513
0.565
108
0.528
0.58
0.43
0.464
0.464
0.540
0.433
0.479
0.478
0.487
0.435
GPI*
100
0.929
1.00
0.98
1.000
0.990
1.000
0.985
1.000
0.995
0.994
0.995
117
0.071
_
0.02
_
—
—
0.015
—
0.005
0.006
0.005
78
—
—
—
—
0.010
-
-
-
-
-
-
IDH-2'
100
ND
1.00
1.00
1.000
0.975
0.976
1.000
0.979
0.956
1.000
0.978
82
—
—
—
0.025
0.024
—
0.021
0.044
-
0.022
MDH-1'
100
ND
1.00
1.00
1.000
0.951
0.960
ND
0.944
0.944
ND
0.982
120
—
—
—
0.049
0.040
0.056
0.056
0.018
MDH-2'
100
0.794
1.00
1.00
1.000
1.000
1.000
0.955
1.000
1.000
1.000
1.000
138
0.206
-
-
-
-
-
0.005
-
-
-
-
ODH*
100
ND
1.00
1.00
0.990
0.990
1.000
ND
0.993
0.995
ND
0.995
68
—
—
0.005
0.010
—
0.007
0.005
0.005
m
—
-
0.005
-
-
-
-
-
PGM-l*
100
0.723
0.83
0.62
0.665
0.706
0.645
0.737
0.697
0.658
0.709
0.660
111
0.250
0.17
0.38
0.335
0.279
0.350
0.263
0.275
0.342
0.291
0.340
89
0.027
—
—
—
0.015
0.005
-
0.028
-
-
-
PGDH*
100
ND
0.59
0.63
0.686
0.721
0.695
ND
0.641
0.696
ND
0.681
150
0.41
0.37
0.314
0.279
0.305
0.359
0.293
0.319
200
(February collection)
- - - - - 0.011
were excluded from the statistical analyses because of small sample size.
'Data for thes
e conch
'■April collection.
Results
Nineteen presumptive loci encoding 11 enzymes were
resolved electrophoretically (Table 2). Ten loci were
polymorphic and were used exclusively in the popula-
tion analyses (Table 3).
Florida Keys and Bimini populations
Allele frequencies for samples of queen conch from the
Florida Keys and Bimini were very similar (Table 3).
The gene identity between samples, averaged over the
ten polymorphic loci, ranged from 0.978 to 0.999 and
averaged 0.994 for all pairwise comparisons. Most
alleles were present in all samples, but some rare {P<
0.01) alleles were detected as only one or two hetero-
zygotes (e.g., AAT-1*130). An exception to this latter
generalization was the presence of the MDH-2* 138
allele at a frequency of 0.206 (35 * 100/100, 11 * 100/138,
and 5 *138/138) among 51 scored individuals collected
from Ballast Key in 1987. Only one heterozygote for
this allele was observed elsewhere during the study.
Genotypes conformed (P>0.05) to Hardy- Weinberg
proportions at all loci except DPEP-1 * . At this latter
locus, significant (P<0.01) deficits of heterozygotes
were detected in 7 of 10 samples. Overall, 285, 279,
and 244 individuals had the *100/100, *100/108, and
*108/108 genotypes, respectively, at DPEP-1*. This
overall deficit of heterozygotes occurred despite similar
(P>0.05) allele frequencies among samples (Table 3).
Spatial and temporal variation in allele frequencies
accounted for minor but approximately equal amounts
of gene diversity. The total gene diversity (Hx) aver-
aged 0.202 for the ten polymorphic loci. Of this total,
0.60% and 0.50% were due to temporal and spatial
variation within and among localities, respectively
254
Fishery Bulletin 90(2), 1992
Table 4
Hierarchical likelihood-ratio tests (G -statistics) for homogeneity of allele frequencies among samples of queen conch Strombus gigas \
from the Florida Keys
and Bimini. Degrees
of freedom are
in parentheses.
Source of variation
AAT-1'
AAT-2'-
DPEP-1*
GPI*
lDH-2'
MDH-1'
Total
19.68(18)
20.03(18)
7.84(9)
43.22(18)**
21.09(8)*
24.51(6)**
Among localities
16.26(8)
15.21(9)
2.42(4)
32.66(8)*
9.48(4)
9.57(3)
Within localities
3.42(5)
4.82(4)
5.42(5)
10.57(5)
11.61(4)
14.95(3)*
Ballast Key
1.62(1)
—
1.38(1)
2.81(1)
—
—
Coffins Patch
1.71(2)
2.37(2)
3.04(2)
4.36(2)
4.41(2)
14.94(2)**
Craig Key
0.09(2)
2.45(2)
1.00(2)
3.40(2)
7.20(2)
0.01(1)
(Approx. F-ratio")
2.97(8,5)
1.40(9,4)
0.56(4,5)
1.93(8,5)
0.82(4,4)
0.64(3,3)
Source of variation
MDH-2'
ODH'
PGM-1*
PGDH*
Total
Total
116.83(9)***
7.03(12)
34.13(18)
10.97(12)
305.33(128)'**
Among localities
95.42(4)"*
2.02(6)
8.79(8)
6.65(6)
198.48(60)***
Within localities
21.41(3)***
5.00(5)
25.34(10)*
4.32(4)
106.86(48)***
Ballast Key
19.54(1)***
—
5.49(2)
—
30.84(6)***
Coffins Patch
—
4.97(4)
6.58(4)
0.61(2)
42.99(22)**
Craig Key
1.87(2)
0.03(1)
13.27(4)
3.71(2)
33.03(20)*
(Approx. F-ratio")
3.34(4,3) 0.37(6,5) 0.43(8,10)
= (G among localities/df)/(G within localities/df).
1.03(6,4)
1.49(60,48)
' Approximate F-ratio
•P<0.05; ••P<0.01; '
**P<0.001; after adjustment for number of tests (Rice
1989).
(GsL 0.0060, Glt 0.0050). The remaining 98.9% (Hg/
Hx) was due to within-sample heterozygosity.
Allele-frequency heterogeneity among samples was
significant (F<0.05) at several loci and was due to both
spatial and temporal components of variation (Table
4). Temporal variation at Ballast Key (P<0.001) and
Coffins Patch (P<0.01) was due primarily to variation
at MDH-2* and MDH-1*, respectively. On the other
hand, the heterogeneity among years at Craig Key
(P<0.05) was due primarily to the cumulative effects
of variation at IDH-2* and PGM-1*. Significant allele-
frequency variation also existed among localities, but
this latter variation did not exceed the temporal varia-
tion within localities as measured by F-ratio com-
parisons (P>0.05) at each locus.
Comparisons with Caribbean Sea
and Bermuda populations
Allele frequencies dX DPEP-1*, MDH-2*, PGM-1*, and
PGDH* for the Florida Keys and Bimini populations
of S. gigas can be compared directly with those for
populations sampled by Mitton et al. (1989). In that
previous study, queen conch were collected from 16
sampling sites representing eight major localities
throughout the Caribbean Sea area (Fig. 1). In addi-
tion, conch were collected from one site in Bermuda.
Patterns of genetic variation among populations in
the Caribbean Sea and Bermuda were similar to those
for populations in the Florida Keys and Bimini (Table
5). Total gene diversities (Hj) for the two groups of
populations were essentially equal (0.355 and 0.354,
respectively). However, the diversity within and among
localities was somewhat greater for Caribbean Sea and
Bermuda populations (Glt 1-69%, Gsl 1.14%) than for
populations from the Florida Keys and Bimini (Glt
0.39%, Gsl 0.68%). This latter result might be ex-
pected considering the relative geographic scales over
which populations were sampled in the two studies
(Fig. 1). In this context, summing Glt ^^d Gsl for
Florida Keys and Bimini populations yields a percent-
age of gene diversity (1.07%) that is approximately
equal to Gsl (sites within localities) for the Caribbean
Sea and Bermuda populations (1.14%).
A dendrogram based on Nei's index of gene identity
clearly reflected the high genetic similarity among
populations of S. gigas (Fig. 2). The average gene iden-
tity (based on the four aforementioned loci) among
populations sampled by Mitton et al. (1989) was 0.984,
among those sampled here was 0.993, and between
populations (samples) of the two studies was 0.988.
Twenty-three of these populations clustered together
at the 0.99 level or above. The Bermuda population and
the 1987 Ballast Key population (sample) formed a
separate subcluster, due primarily to divergent allele
frequencies at MDH-2*. The Vieux Fort (St. Lucia) and
Six Hill Cay (Turks and Caicos Islands) populations also
clustered separately, due primarily to slightly divergent
Campton et al . Genetic patchiness among Strombus gigas populations
255
Table 5
Percentages of total gene diversity (H^.) among localities
(Glt). among samples and sites within localities (Gsl), and
within samples and sites (Hj/Hj) for populations of queen
conch Strombus gigas from the Florida Keys and Bimini (this
study) and from the Caribbean Sea and Bermuda (Mitton et
al. 1989). Data represent the means for DPEP-l*, MDH-2*,
PGM-1*, PGDH*.
Gtene diversity (%)
Populations
Ht
Ho/Ht.
Florida Keys
and Bimini
Caribbean Sea
and Bermuda
All populations
0.354
0.355
0.354
0.39
1.69
1.24
0.68
1.14
1.01
98.94
97.17
97.75
allele frequencies at PGDH* and DPEP-l*, respective-
ly (Mitton et al. 1989).
In summary, populations of S. gigas are very similar
genetically and do not appear to be structured geo-
graphically. However, those populations cannot be con-
sidered totally panmictic.
Discussion
Population structure
Benthic marine invertebrates with planktonic larvae
often exhibit spatial and temporal genetic variation
similar to that described here for S. gigas. For exam-
ple. Watts et al. (1990) found significant allele-fre-
quency variation among three populations of sea urchin
Echinometra mathaei separated by only 4 km. More-
over, that heterogeneity over a 4 km distance was ap-
proximately equal to the genetic heterogeneity among
populations separated by over 1300km. Those inves-
tigators also detected significant allele-frequency varia-
tion among year-classes within each of the three micro-
spatial sample sites. Similar patterns of heterogene-
ity were reported for the limpet Siphonaria jeanae
(Johnson and Black 1982, 1984ab) and seastar Acan-
thaster planei (Nash et al. 1988, Nishida and Lucas
1988).
Significant microspatial genetic heterogeneity,
despite high macrospatial genetic similarity, has been
termed "genetic patchiness" (e.g., Johnson and Black
1984b). Such genetic patchiness could be due to either
postsettlement natural selection or genetic hetero-
geneity among groups of recruits that are spatially or
temporally separated (Watts et al. 1990). Under both
hypotheses, planktonic dispersal is believed to main-
^
— 23 population
E
— samples
Vieux Fort
St. LuCia
Ballast Key 1987
Bermuda
Six Hill Cay
Turks and Caicos
1 1 1 1 1 1
0.95 1.0
Gene Identity
Figure 2
UPGMA dendrogram of genetic similarities among popula-
tion samples of queen conch Strombus gigas based on allele
frequencies at DPEP-l*. MDH-2'. PGDH*. and PGM-l*.
and Net's (1972) index of genetic identity.
tain high genetic similarity among populations over
broad geographic areas. However, under postsettle-
ment natural selection, one would expect genetic varia-
tion among localities to be greater than the temporal
variation within localities because of local adaptation.
Conversely, under the model of presettlement genetic
heterogeneity, spatial and temporal components of
genetic variation are expected to be equal because the
population structure would result from presettlement
events that were independent of the specific localities
at which settlement occurred. Under this latter model,
spatial heterogeneity among localities would simply
reflect the temporal heterogeneity within localities.
Results obtained here for S. gigas are most consis-
tent with the presettlement hypothesis of genetic
patchiness. Populations of queen conch throughout
their geographic range are very similar genetically, yet
spatial and temporal components of genetic variation
appear significant and approximately equal. Mitton et
al. (1989) obtained similar results for macrospatial
(among-locality) and microspatial (within-locality) com-
ponents of genetic variation. These results suggest a
dynamic population structure in which allele-frequency
heterogeneity may exist among groups of recruits that
settle in different years at the same locality or at dif-
ferent localities in the same year. Johnson and Black
(1982, 1984ab) and Watts et al. (1990) reached similar
conclusions regarding genetic patchiness among pop-
ulations of limpet and sea urchin, respectively.
Several mechanisms can be invoked to explain genet-
ic patchiness due to presettlement events. Johnson and
Black (1984ab) and Watts et al. (1990) suggest that
256
Fishery Bulletin 90(2). 1992
selective mortality prior to settlement, possibly reflect-
ing stochastic variation in the marine environment
(e.g., water temperature, salinity), may be responsible
for the "chaotic genetic patchiness" that they observed.
Alternatively, temporal variation in the source of re-
cruits for each locality and/or genetic drift resulting
from a finite number of breeders could also generate
random genetic patchiness on both temporal and spatial
scales (e.g., Waples 1989). None of these aforemen-
tioned hypotheses can be excluded with the available
data.
The effective number of breeders (N^) contributing
to a cohort of larvae that settle together at a particular
location is unknown for S. gigas. Males and females
breed in aggregations at characteristic locations over
a 6-9 month period, and each female may produce
several egg masses of approximately 310,000-750,000
eggs each during the breeding season (Robertson 1959,
Randall 1964, Weil and Laughlin 1984, Berg and Olson
1989). Several females within an aggregation may lay
their egg masses simultaneously, and because the rate
of embryonic development is temperature-related,
hordes of larvae are released synchronously. These
larvae can thus be entrained together into the water
column and affected simultaneously by marine and
oceanic processes. Consequently, hordes of larvae from
a finite number of parents could potentially be pre-
sented simultaneously to a substrate that would induce
settlement and metamorphosis.
Recently, Bucklin et al. (1989) and Bucklin (1991)
obtained evidence that ocean currents and related
processes (e.g., upwellings, eddies, offshore jets) can
spatially and temporally maintain genetically discrete
cohorts of zooplankton in the marine environment. For
example, Bucklin et al. (1989) concluded that such pro-
cesses "prevented homogenization of the plankton
assemblages during transport" and that "plankton
populations in complex flow fields may show patchiness
in biological, biochemical, and/or genetic character at
small time/space scales." Their results suggest that
similar processes could affect significantly the distribu-
tion of pelagic larvae following their release into the
water column.
The source of S. gigas recruits for the Florida Keys
is unknown. The Florida Current, which sweeps east-
ward past the Florida Keys and subsequently forms the
Gulf Stream, is created by the massive flow of warm
water northward from the Caribbean Sea through the
Yucatan Channel. This current could entrain large
numbers of larvae from numerous locations prior to
flowing eastward past the Florida Keys (Mitton et al.
1989). Stochastic variations in water currents, surface
winds, and meteorological events (e.g., tropical storms)
could thus affect significantly the source of S. gigas
recruits for any particular locality. During the course
of our study, we attempted to gain permission to col-
lect conch from Cuba and Yucatan, Mexico— two pos-
sible sources of recruits for the Florida Keys— but were
unable to do so.
One potential shortcoming of our study was that the
temporal effects of recruitment were confounded with
other population processes; that is, temporal genetic
variation was measured among mixed aggregations of
conch sampled in different years at the same locality
and not among separate year-classes or cohorts. With
the exception of the Coffins Patch population or ag-
gregation (see below), all samples consisted of mixed
age- and size-classes with new recruits added each year.
In addition, some of the temporal genetic variation may
have been due to migration of juveniles and adults into
and out of the study areas (Hesse 1979, Weil and
Laughlin 1984, Stoner et al. 1988, Stoner 1989). Con-
sequently, we cannot separate the temporal effects of
recruitment from other population processes. How-
ever, our goal was not to estimate temporal genetic
variation among cohorts or year-classes per se, but
rather to provide a measure of within-population (i.e.,
within-locality) variation by which the significance of
genetic variation among localities could be evaluated.
Population processes causing temporal genetic varia-
tion within localities would similarly affect the genetic
variation among localities. Some measure of temporal
variation was thus needed before the microevolutionary
significance of genetic variation among localities could
be ascertained. Alternatively, some form of stratified
sampling of year- and/or size-classes would be required
to separate recruitment or year-class effects from other
potential sources of temporal genetic variation.
Possible evidence that recruitment, migration, or
similar population processes may significantly affect
the population structure of S. gigas was the presence
of the MDH-2*(138) allele at a frequency of 0.206 in the
1987 sample from Ballast Key but the near absence of
this allele in the 1988 sample and elsewhere during our
study. Mitton et al. (1989) similarly reported, for the
Bermuda population, a frequency of 0.30 for a "fast"
MDH-2* allele that was also rare elsewhere. However,
the Bermuda population is believed to be self-sustaining
with little planktonic recruitment from the Gulf Stream
or elsewhere (Mitton et al. 1989). Conversely, Ballast
Key is situated within the Florida Current and is the
most upstream locality from which we collected conch
for the present study. Two distinct aggregations of
S. gigas may have been sampled at Ballast Key in 1987
and 1988, respectively.
Anomalous results
Coffins Patch Size distributions suggest that the Cof-
fins Patch population was most likely a single year-class
Campton et al.: Genetic patchiness among Strombus gigas populations
257
or cohort that we sampled in three consecutive years
(1988-90). This population or cohort presumably re-
sulted from a large recruitment event during the sum-
mer and fall of 1987. We estimated that the 1987 ag-
gregation at Coffins Patch consisted of at least 250,000
animals covering an area of approximately 30 hectares
(Berg and Glazer, unpubl.).
Although the Coffins Patch aggregation appeared to
be a single cohort, we detected a significant allele-
frequency variation among years (1988-90) at MDH-1*.
This difference was due to the absence of the MDH-1 *
(120) allele in the 1988 sample {n 100) versus the pres-
ence of eight *100/120 heterozygotes in both the 1989
{n 102) and 1990 {n 100) samples. The 1989 sample also
had one * 120/120 homozygote. Sampling error does not
adequately explain those results because the probability
of obtaining all *100/100 homozygotes in the 1988
sample was only (0.9552)^"*^ = 0.0001 (assuming the
true frequency of the *120 allele was 0.045 [mean of
1989 and 1990 samples] and random mating). Similar-
ly, differential mortality among genotypes does not
adequately explain those results unless heterozygotes
were initially very rare and the subsequent mortality
of *100/100 homozygotes was extremely high.
Alternatively, recruitment to the Coffins Patch area
in 1987 may have been from more than one source
population. This could have resulted in an aggregation
that was not distributed randomly. Subsequent mixing
and/or possible immigration of juveniles (e.g., Stoner
1989) could thus explain changes in allele frequencies
between 1988 and 1989. None of these hypotheses can
be excluded with the available data.
Regardless of actual mechanism, the presence of only
one highly abundant year-class at Coffins Patch over
a 3-year period indicates that recruitment to specific
localities in the Florida Keys can be highly variable and
unpredictable. This observation thus supports the in-
terpretation that genetic patchiness may simply reflect
stochastic events prior to settlement.
DPEP-1 * We observed a consistent deficit of het-
erozygotes (with respect to Hardy- Weinberg expecta-
tions) at DPEP-1 * but not at other loci. Similar deficits
of heterozygotes have been reported often for marine
mollusks (reviewed by Gaffney et al. 1990). Such
deficits are frequently associated with positive correla-
tions between body size and individual heterozygosity.
We also observed a positive correlation between body
size and heterozygosity, but genotypic variation at
DPEP-1* did not contribute to that correlation. These
results will be described in detail elsewhere (Campton
et al. In press).
PGM-2* One possible point of inconsistency be-
tween the study described here and that of Mitton et
al. (1989) concerns data iov PGM-2*. Mitton et al. (1989)
presented only limited data for this latter locus (9 of
17 populations), but those investigators consistently
observed a high frequency (0.57-0.69) polymorphism
for a "slow" allele. In contrast, we found PGM-2* to
be fixed for a single allele. Only PGM-1* and PGM-2*
are expressed in digestive gland tissue, which was the
only tissue assayed by Mitton et al (1989). However,
we also scored PGM in foot muscle which clearly re-
vealed a third, more cathodal locus {PGM-3*). We also
observed three distinct loci in foot tissue of a second
conch species, S. costatus.
At least three possibilities could thus account for the
apparent difference between our results and those of
Mitton et al. (1989) at PGM-2*: (1) our inability to
resolve the variant electromorph at PGM-2*, (2) the
partial expression of the PGM-3* locus in digestive
gland tissue (e.g., Allendorf et al. 1983) of individuals
sampled by Mitton et al. (1989), thus giving false
readings of heterozygotes at PGM-2*, or (3) the re-
ported allele-frequency difference between the two
groups of populations are indeed real. Of the three
possibilities, we believe explanations (1) and (2) are the
most likely because of the high consistency of our allele
frequencies with those of Mitton et al. (1989) at all
other loci. Consequently, we believe that this apparent
discrepancy at PGM-2* most likely reflects laboratory-
specific adaptations of basic electrophoretic pro-
cedures. In this context, we were able to resolve several
loci not resolved by Mitton et al. (1989) and vice- versa.
Conclusions
The major finding of our study was the existence of
spatial and temporal genetic patchiness among popula-
tions of queen conch in the Florida Keys and Bimini.
We suggest that such genetic patchiness most likely
results from presettlement stochastic events and pro-
cesses in the marine environment. Nevertheless, these
populations are all very similar genetically, presumably
reflecting high levels of gene flow due to larval drift.
These interpretations are consistent with the results
of Mitton et al. (1989) and also explain similar patterns
of "chaotic genetic patchiness" in other taxa of marine
invertebrates.
Acknowledgments
We thank R. Estling, A. Kirkley, and W. Schumacher
for their assistance in the laboratory.
258
Fishery Bulletin 90|2). 1992
Citations
Aebersold, P.B., G.A. Winans. D.J. Teel, G.B. Milner, and
F.M. Utter
1987 Manual for starch gel electrophoresis: A method for the
detection of genetic variation. NOAA Tech. Rep. NMFS 61,
19 p.
Allendorf, F.W., R.F. Leary, and K.L. Knudsen
1983 Structural and regulatory variation of phosphogluco-
mutase in rainbow trout. Isozymes Curr. Top. Biol. Med. Res.
9:123-142.
Ballantine, D.L., and R.S. Appeldoorn
1983 Queen conch culture and future prospects in Puerto
Rico. Proc. Gulf Caribb. Fish. Inst. 35:57-63.
Berg, C.J. Jr., and D.A. Olsen
1989 Conservation and management of queen conch {Strom-
bus gigas) fisheries in the Caribbean. In Caddy, J.F. (ed.),
Marine invertebrate fisheries: Their assessment and manage-
ment, p. 421-442. John Wiley, NY.
Boyer, S.H., D.C. Fainer, and E.J. Watson-Williams
1963 Lactate dehydrogenase variant from human blood:
Evidence for molecular subunits. Science (Wash. DC) 141:
642-643.
Bucklin, A., M.M. Rienecker, and C.N.K. Mooers
1989 Genetic tracers of zooplankton transport in coastal
filaments off northern California. J. Geophys. Res. 94(C6:
8277-8288.
Bucklin, A.
1991 Population genetic responses of the planktonic copepod
Metridia paeifica to a coastal eddy in the California Cur-
rent. J. Geophys. Res. 96(C8):14,799-14,808.
Campton, D.E., C.J. Berg Jr., L.M. Robison. and R.A. Glazer
In press Deficits of heterozygotes and correlation between size
and heterozygosity in the marine mollusk, Strombus gigas
(Gastropoda: Strombidae). Evolution.
Chakraborty, R., M. Haag, N. Ryman, and G. Stahl
1982 Hierarchical gene diversity analysis and its application
to brown trout populations. Hereditas 97:17-21.
Clayton, J.W., and D.N. Tretiak
1972 Amine-citrate buffers for pH control in starch gel elec-
trophoresis. J. Fish. Res. Board Can. 29:1169-1172.
Davis, M., and C. Hesse
1983 Third worid level conch mariculture in the Turks and
Caicos Islands. Proc. Gulf Caribb. Fish Inst. 35:73-82.
Gaffney, P.M., T.M. Scott, R.K. Koehn, and W.J. Diehl
1990 Interrelationships of heterozygosity, growth rate and
heterozygote deficiencies in the coot clam, Mulinia lateralis.
Genetics 124:687-699.
Hesse, K.O.
1979 Movement and migration of the queen conch, Strombus
gigas, in the Turks and Caicos Islands. Bull. Mar. Sci. 29:
303-311.
Johnson, M.S., and R. Black
1982 Chaotic genetic patchiness in an intertidal limpet,
Siphonaria sp. Mar. Biol. (Beri.) 70:157-164.
1984a The Wahlund effect and the geographical scale of varia-
tion in the intertidal timpet Siphonaria sp. Mar. Biol. (Beri.)
79:295-302.
1984b Pattern beneath the chaos: The effect of recruitment
on genetic patchiness in an intertidal limpet. Evolution 38:
1371-1383.
Kinder, T.H., G.W. Hebrun, and A.W. Green
1985 Some aspects of the Caribbean circulation. Mar. Geol.
68:25-52.
Little, C.
1965 Notes on the anatomy of the queen conch, Strombus
gigas. Bull Mar. Sci. 15:338-358.
Mitton, J.B., C.J. Berg Jr., and K.S. Orr
1989 Population structure, larval dispersal, and gene flow in
the queen conch, Strombus gigas, of the Caribbean. Biol. Bull.
(Woods Hole) 177:356-362.
Morizot, D.C, and M.E. Schmidt
1990 Starch gel electrophoresis and histochemical visualiza-
tion of proteins. In Whitmore, D.H. (ed.), Electrophoretic and
isoelectric focusing techniques in fisheries management, p.
23-80. CRC Press, Boca Raton.
Nash, W.J., M. Goddard, and J.S. Lucas
1988 Population genetic studies of the crown-of-thorns star-
fish, Acanthasterplanci (L.), in the Great Barrier reef region.
Coral Reefs 7:11-18.
Nei, M.
1972 Genetic distance between populations. Am. Nat. 106;
283-292.
1973 Analysis of gene diversity in subdivided populations.
Proc. Natl, Acad. Sci. USA 70:3321-3323.
Nishida, M., and J.S. Lucas
1988 Genetic differences between geographic populations of
the crown-of-thorns starfish throughout the Pacific region.
Mar. Biol. (Beri.) 98:359-368.
Randall, J.E.
1964 Contributions to the biology of the queen conch, Strom-
bus gigas. Bull. Mar. Sci. 14:246-295.
Rice, W.R.
1989 Analyzing tables of statistical tests. Evolution 43:
223-225.
Ridgway, G.J., S.W. Sherburne, and R.D. Lewis
1970 Polymorphisms in the esterases of Atlantic herring.
Trans. Am. Fish. Soc. 99:147-151.
Robertson, R.
1959 Observations on the spawn and veligers of conchs {Strom-
bus) in the Bahamas, Proc. Malacol. Soc. Lond. 33:164-171.
Scheltema, R.S.
1971 Larval dispersal as a means of genetic exchange between
geographically separated populations of shallow-water benthic
marine gastropods, Biol Bull, (Woods Hole) 140:284-322.
1986 On dispersal and planktonic larvae of benthic inverte-
brates: An eclectic overview and summary of problems. Bull.
Mar. Sci. 39:241-256,
Selander, R.K.. M.H. Smith, S.H. Yang, W.E. Johnson, and
J.B. Gentry
1971 Biochemical polymorphism and systematics in the genus
Peromyscus. I. Variation in the old field mouse {Peromyscus
polionotus). Studies in Genetics VI. Univ. Texas Publ. 7103:
49-90.
Shaklee, J.B., F.W. Allendorf, D.C. Morizot, and G.S. Whitt
1990 Gene nomenclature for protein-coding loci in fish. Trans.
Am. Fish. Soc. 119:2-15.
Siciliano, M.J.. and C.R. Shaw
1976 Separation and visualization of enzymes on gels, hi
Smith, I. (ed.). Chromatographic and electrophoretic tech-
niques. Vol, II, Zone electrophoresis (4th ed.), p. 185-209.
Year Book Med, Publ,. Chicago.
Smouse, P.E., and R.H. Ward
1978 A comparison of the genetic infrastructure of the Ye'cu-
ana and the Yanomama: A likelihood analysis of genotypic
variation among populations. Genetics 88:611-631.
Sneath, P.H.A., and R.R. Sokal
1973 Numerical taxonomy, W.H, Freeman, San Francisco.
Campton et al,: Genetic patchiness among Strombus gigas populations
259
Sokal, R.R.. and F.J. Rohlf
1981 Biometry. W.H. Freeman, San Francisco.
Stoner, A.W.
1989 Winter mass migration of juvenile queen conch Strom-
bus gigas and their influence on the benthic environment. Mar.
Ecol. Prog. Ser. 56:99-104.
Stoner. A.W., R.N. Lipcius, L.S. Marshall Jr.. and A.T. Bardales
1988 Synchronous emergence and mass migration in juvenile
queen conch. Mar. Ecol. Prog. Ser. 49:51-55.
Waples. R.S.
1989 Temporal variation in allele frequencies: Testing the
right hypothesis. Evolution 43:1236-1251.
Watts. R.J., M.S. Johnson, and R. Black
1990 Effects of recruitment on genetic patchiness in the ur-
chin Echinometra mathaei in western Australia. Mar. Biol.
(Berl.) 195:145-151.
Weil. E.M.. and R.G. Laughlin
1984 Biology, population dynamics, and reproduction of the
queen conch Strombus gigas Linnaeus in the Archipielago de
Los Roques National Park. J. Shellfish Res. 4:45-62.
Abstract. -The growth patterns
of Pacific whiting Merluccius pro-
d/uctus, also known as Pacific hake,
were examined for the period 1978-
88 using fishery-based estimates of
length-at-age. Mean length-at-age
and a delta method variance esti-
mate of mean length-at-age were
computed for geographic and tem-
poral strata in the U.S. fishery.
These calculations took into account
the two-phase sampling design used
to sample the catch. A factorial anal-
ysis of variance of length found sig-
nificant differences due to age, year,
region, sex, and time-period during
the fishery. Length increases with
age and season. Pacific whiting
found in the north are larger, and
females are larger than males. The
mean length-at-age began declining
in 1978, and reached a minimum in
1984. From 1984 to 1986, there was
a slight rebound in length-at-age, but
after 1986 length-at-age again
declined. To investigate the influence
of population density and environ-
mental covariates on annual growth,
a generalized form of the von Ber-
talanffy growth model was devel-
oped. Deviations from a baseline
model for sex-specific asymptotic
growth were significantly correlated
with changes in sea-surface temper-
ature and adult biomass. Regression
results indicate that a 0.5°C increase
in mean summer sea-surface temper-
ature would reduce annual growth
by 24% at age 1 and 12% at age 4.
In contrast, the effect of adult bio-
mass on annual growth becomes
greater with age. An increase of
200,000 metric tons (approximately
10% of the mean population biomass)
would reduce annual growth by 5%
at age 4 and by 10% at age 7. It is
proposed that the effect of popula-
tion density is greater for the older
Pacific whiting because their diet
has shifted from euphausiids, whose
abundance is closely coupled with en-
vironmental processes, towards fish
species with multiyear life cycles that
can be affected by intense Pacific
whiting predation.
Detecting environmental
covariates of Pacific wliiting
Merluccius productus growth! using
a growtli-increment regression model
Martin W. Dorn
Alaska Fisheries Science Center, National Marine Fisheries Service. NOAA
7600 Sand Point Way NE. BIN CI 5700, Seattle, Washington 981 15-0070
Manuscript accepted 18 February 1992.
Fishery Bulletin, U.S. 90:260-275(1992).
This paper describes research on
the growth of Pacific whiting, also
known as Pacific hake, a gadoid spe-
cies that is an important component
of the California Current ecosystem
(Francis 1983). The coastal popula-
tion of Pacific whiting is currently
the target of a major fishery with an
average (1977-88) annual harvest of
147,000 metric tons (t) (Dorn and
Methot 1990). Adult Pacific whiting
migrate north in spring and summer,
feeding in the productive waters
along the continental shelf and slope
from northern California to Van-
couver Island, British Columbia. In
late autumn, Pacific whiting migrate
south to spawning areas from Point
Conception, California, to Baja Cali-
fornia (Bailey et al. 1982). The U.S.
fishery operates from April to No-
vember and in recent years has been
conducted primarily under joint-
venture arrangements, with U.S.
fishing boats delivering fish to pro-
cessing vessels from the Soviet
Union, Poland, Japan, and other
nations. The Canadian fishery for
Pacific whiting is conducted in sim-
ilar fashion, except that independent
fishing by the foreign fleet still ac-
counts for a significant portion of the
catch.
Hollowed et al. (1988) observed
that the mean length-at-age of Pacific
whiting had declined in recent years,
and hypothesized that the disruption
of normal circulation and tempera-
ture patterns associated with the
1983 El Nino may have been the
causative factor. The recniitment of
strong 1980 and 1984 year-classes
increased the population biomass
of Pacific whiting to a maximum in
1986. The decline in the length-at-
age could also have been a density-
dependent growth response to this
increase in population abundance.
Since the Pacific whiting resource is
managed by setting an annual quota
in tons based on a conversion using
weight-at-age from a projected yield
in numbers (Dorn and Methot 1990),
changes in growth must be taken
into account when making manage-
ment recommendations about the
resource.
The objective of this paper is to
examine the pattern of growth vari-
ability displayed by the coastal Pacif-
ic whiting population, and, in partic-
ular, to determine whether environ-
mental covariates or fluctuations in
population density could account for
the recent changes in length-at-age.
Analysis of variance, while useful as
an exploratory technique to identify
the sources of variability in length-
at-age, is inadequate to describe
changes in asymptotic growth. The
nonlinear regression model pre-
sented in this paper is a simple, bio-
logically realistic model for ex^ploring
the environmental determinants of
asymptotic growth. Its potential util-
ity is not limited to the application
described in this paper, i.e., growth
of Pacific whiting.
260
Dorn: Environmental covanates of Merlucous productus growth
261
Methods
The U.S. Foreign Fisheries Ob-
server Program at the Alaska
Fisheries Science Center (AFSC)
uses a two-phase sampling de-
sign to sample the catch of Pa-
cific whiting (French et al. 1981).
The first phase consists of obtain-
ing a large initial sample of fish
and recording the length and sex.
For the second phase of sam-
pling, a subsample of fixed size
is selected for each combination
of length category and sex. All
fish in these subsamples are aged
using otoliths.
Typically, each observer sam-
ples 2-3 hauls or joint-venture
deliveries per day for length and
sex, and 150 otoliths (5 per centi-
meter-length category per sex) are collected
over a two-month cruise. The numbers of aged
and measured fish from 1978-88 are given in
Table 1. This information resides in a data base
maintained by the Resource Ecology and Fish-
eries Management Division (REFM) at the
Alaska Fisheries Science Center.
Kimura and Chikuni (1987) point out that,
with a two-phase sampling design, estimates
of mean length-at-age are biased when ob-
tained simply by averaging the lengths of the
aged fish. To avoid this bias, stratified length-
at-age estimates were compiled from fishery
data for the years 1978-88 using separate age-
length keys for each stratum (see Appendix for
details). Three spatial strata were defined as:
(1) the area from lat. 39°00'N to lat. 43°00N,
including part of the International North Pa-
cific Fishery Commission (INPFC) Monterey
region and the Eureka INPFC region (EUR);
(2) the area from lat. 43°00N north to Cape
Falcon (lat. 46°45N) in the southern part of the
Columbia INPFC region (SCOL); and (3) the
area north of Cape Falcon to the U.S. -Canada
border including the northern part of the Co-
lumbia INPFC region and the U.S. portion of
the Vancouver INPFC region (VNC) (Fig. 1).
Figure 1
Spatial strata used to compile length-at-age
for midwater trawl fishery samples of Pacific
whiting Merluccius ■productus.
Table 1
Number of Pacific whiting Merlucciu^
productus sampled from the midwater trawl fishery 1
during the years 1978-88 off Washington, Oregon, and northern California.
Early period
Middle period
Late
period
Year
Apri
-June
July-August
Aged Measured
Sept
Aged
.-Nov.
Measured
Annual total
Aged
Measured
Aged
Measured
1978
2060
31,819
2801
66,153
978
26,799
5839
124,771
1979
1072
37,678
1552
83,584
500
52,094
3124
173,356
1980
844
15,674
2927
43,038
1565
43,536
5336
102,248
1981
1287
26.961
1928
55.174
1053
53,605
4268
135,740
1982
1913
77,529
1463
66,683
882
27,604
4258
171,816
1983
1480
82,186
1277
70,499
475
14,173
3232
166,858
1984
1344
70,888
1304
108,272
662
64,524
3310
243,684
1985
200
23,329
1690
142,592
550
101,089
2440
267,010
1986
1203
125,542
1393
238.779
474
109,786
3070
474,107
1987
1021
102,191
1414
188,361
740
140,902
3175
431,454
1988
1192
125,714
1349
194,246
502
100,184
3043
420,144
Total
13,616
719,511
19,098
1,257,381
8381
734,296
41,095
2,711,188
British Columbia
Nortlieasl Pacific Ocean
50°00'N
45°00'
40°00'
35°00'
30°00'
1 30°00'W
125°00'
1 20°00'
262
Fishery Bulletin 90(2). 1992
Each of these geographic regions encloses a center of
Pacific whiting abundance and a concentration of
fishing activity (Dorn and Methot 1990). Three time-
periods were also defined as strata: (1) early (April-
June), (2) middle (July- August), and (3) late (Septem-
ber-November). These time-periods divide the fishing
season into three roughly equal parts. Over the years
1978-88, 27.9% of the catch came from the early time-
period, 47.4% came from the middle time-period, and
24.7% came from the late time-period. In compiling the
length-at-age estimates for the spatial and temporal
strata, all data collected within that strata were ag-
gregated and assumed to originate from random sam-
pling of the catch within that strata.
Some of the detrimental effects of ageing error bias
and low sampling intensity of uncommon age groups-
common problems in analyzing fishery length-at-age
data— can be reduced if the precision of the length-
at-age estimates is known. A delta-method variance
estimator of length-at-age for a two-phase sampling
plan was derived and implemented for the U.S. fishery
samples. Details of this estimator and a procedure for
combining the length-at-age from different strata are
described in the Appendix.
Two general methods of analyzing the growth in
length of fish have been used widely in fishery research.
The first method interprets individual observations of
length-at-age or mean length-at-age in the population
by fitting asymptotic growth curves, most typically the
von Bertalanffy (or monomolecular) growth curve
(Boehlert and Kappenman 1980, Kimura 1980, Shep-
herd and Grimes 1983). Using this technique to study
environmental effects on growth on an annual scale is
difficult because growth curves summarize the growth
history of a year-class or a population over the lifespan
of the organism. One approach to generalizing growth
curves is to include seasonal environmental effects on
growth. An example of this is the work of Pauly and
Gaschiitz (1979); they incorporated a sine wave in the
von Bertalanffy growth curve to model the seasonal
growth cycle.
The second common approach to analyzing growth
data is analysis of variance (ANOVA). Factorial de-
signs have been used to investigate regional growth
variability (Francis 1983, Reish et al. 1985). Multiple
linear regression is often used to examine the effect
of the environment or population density on growth
(Kreuz et al. 1982, Ross and Almeida 1986, Peterman
and Bradford 1987). A factorial ANOVA of length
using age, year, region, sex, and time-period as fac-
tors is reported in the Results. It should be recog-
nized, however, that analysis of variance does not
account for changes in asymptotic growth, except by
fitting interaction terms that tend to obscure the
analysis. It is used in this paper only as an explora-
tory technique to identify the sources of variability in
length.
Because asymptotic growth is a universal feature of
fish growth, a model to examine the effect of the
environment on growth should account for this char-
acteristic. At the same time, such a model must be
general enough to allow for covariates to influence
annual growth. To meet this objective, a simple exten-
sion of the asymptotic von Bertalanffy growth model
was developed. The model has a framework similar to
analysis of covariance, in that it allows for the pos-
sibility of differences in growth between constituent
subgroups of the population and differences in growth
due to the influence of population density or environ-
mental covariates.
The von Bertalanffy growth model for the mean
length la of a year-class at age a is given by
la = U(l-e-K(a-a„)),
where 1^^^ is the asymptotic maximum length, k is a
growth coefficient, and ao is the hypothetical age at
length zero. Subtracting the length at age an- 1 from
the length at age a gives the first difference of this
equation, the annual growth increment from age a to
age a-i-l,
la.l - la = l„,(l-e-K)e-Ma-a„).
Defining go = ln[l^(l-e-'^)], and gi= -k, a simple ex-
pression for annual growth is obtained:
'a+l
la = exp(go + gi(a-ao)).
As might be expected, the parameter a^ becomes
redundant in this model for annual growth, since it is
confounded with the parameter go . One possibility is
simply to drop it from the equation. Another alterna-
tive, and the one used in this analysis, is to. use ao to
scale chronological age to some initial age for which
the growth model is intended to apply. In the Pacific
whiting data, there are growth increments from age
1 to age 2, so ao is set to 1. In this parameterization,
structural growth coefficients, go and g] , describe
simple elements of asymptotic growth: exp(go) is the
annual growth increment at age ao, and gi is the
exponential decline in the annual growth increment
(Fig. 2).
To assess the effect of an environmental covariate,
X, this model is augmented with an additional coeffi-
cient for that environmental variable,
la+l t+l s - lats = exp[go + ^ go, Xj
i
+ (gi + Z gij Xj)(a-ao)] + eats,
j
Dorn: Environmental covariates of Merluccius productus growth
263
where eats'^NCO, WatgO^). The additional sub-
scripts in this equation are: t for year, s for
sex, and i and j to index different environ-
mental variables (e.g., Xi and Xj). The case
weights, Wats, ^1"^ determined by the sum of
the estimation variance for the two length-at-
age estimates used to calculate the growth
increment.
In this regression model, environmental
variables can enter as either intercept or slope
terms. An intercept coefficient affects go and
indicates a constant percent change in the
growth increment regardless of age. Slope
coefficients affect gi and provide flexibility
for a varying percent change in the growth in-
crement with age. Together these two types
of coefficients, intercept and slope, cover a
wide range of different ways that environmen-
tal conditions can effect growth at different
ages. Note also that in this formulation, it is
possible to use indicator variables to param-
eterize growth differences between different
constituent groups of the population; for ex-
ample, sex differences or geographic differ-
ences in growth. This model resembles a linear
ANOVA model proposed by Weisberg (1986)
to analyze back-calculated fish lengths, though
he does not use von Bertalanffy growth to scale
the annual growth increments.
The general procedure for fitting a nonlinear
regression model in Ratkowsky (1983) was
followed using the PAR algorithm in the
BMDP statistical package for estimating a non-
linear regression model using weighted least-
squares (Dixon 1983). Mean-square error was
estimated by fitting a full model consisting of
the coefficients go and gi , and separate inter-
cept and slope coefficients for all environmen-
tal covariates (temperature, upwelling, bio-
mass, recruitment strength) assessed in the
analysis. Mean-square error was estimated by
dividing the residual sum of squares for this
model by the degrees of freedom. A full model
should account for all the explainable variabil-
ity, so that the residual error gives an estimate of
mean-square error. A P-value of < 0.05 was established
as the criteria for statistical significance. Because of
the presence of negatively-valued growth increments
due to measurement error, it was not possible to take
the logarithm of the growth increment and analyze the
model using linear regression.
The analysis with this model uses the change in mean
length of an age-group from the early period of the
fishery (April-June) of one year to the early period of
the following year. Geographic strata are not used in
Age
Figure 2
Families of asymptotic von Bertalanffy growth curves parameterized
by g„, the initial growth increment, and g, , the exponential dechne in
the annual grovrth increment with age. All curves were constrained to
pass through 20 cm at age 1.
the analysis because the migratory nature of the coastal
population of Pacific whiting would make any conclu-
sions regarding regional growth patterns impossible
to defend. It is assumed that the annual increment in
growth from one spring to the next is due to conditions
prevalent during the summer season of active growth.
Although growth increments could be studied for
shorter time-periods, this was considered inappropriate
for our study because of possible lags between envi-
ronmental conditions and the growth response of the
fish. In addition, the fishery estimates of length-at-
264
Fishery Bulletin 90(2). 1992
Table 2
Environmental and stock biomass covariates of Pacific whiting Merluccius prodiictiis growth, assessed using the growth-increment
regression model. Mean summer (April-August) sea-surface temperature (°C) is an average over lat. 40-50°N, and from the coast,
west to long. 125°W. The Bakun upwelling indices are mean summer coastal upwelling (April-August) from lat. 42-48°N. Stock biomass
is measured in millions of tons of age-2 and older fish in the coastal whiting population. Anomalies of temperature, upwelling index,
and biomass are calculated as the annual value minus the mean over 1978-87.
Temperature^
Temperature
Upwelling''
Upwelling
Biomass
Year
(°C)
anomaly
index
anomaly
Biomass"
anomaly
1978
13.1
0.6
49.5
0.3
1.503
-0.268
1979
12.5
0.0
49.2
0.1
1.709
-0.062
1980
12.2
-0.3
67.1
18.0
1.640
-0.131
1981
12.5
0.0
50.0
0.9
1.384
-0.387
1982
12.1
-0.3
55.9
6.7
2.000
0.230
1983
13.4
0.9
31.5
-17.6
1.805
0.035
1984
11.9
-0.5
45.7
-3.4
1.742
-0.029
1985
12.1
-0.4
45.3
-3.8
1.685
-0.086
1986
12.6
0.1
49.6
0.5
2.225
0.455
1987
12.3
-0.2
47.6
-1.5
2.012
0.242
Average 1978-87
12.47
49.15
1.771
'J.G. Norton, Pac. Fish. Environ. Group, P.O. Box 831, Monterey, CA 93942, pers. commun.,
''Mason and Bakun (1986).
"Dorn and Methot (1990).
Aug. 1989.
Table 3
Recruitment in billions of age-2 fish to the Pacific whiting
Merluccius produ^tvs population for year-classes 1965-84
(modified from Dom and Methot 1990).
The recruitment
anomaly is calculated as
the annual number of recruits minus |
the average over 1965-
-84.
Year
Recruits
Recruit anomaly
1965
0.692
-0.331
1966
0.786
-0.237
1967
1.110
0.087
1968
0.636
-0.387
1969
0.315
-0.708
1970
3.597
2.574
1971
0.169
-0.854
1972
0.306
-0.717
1973
1.432
0.409
1974
0.139
-0.884
1975
0.220
-0.803
1976
0.094
-0.929
1977
1.716
0.693
1978
0.035
-0.988
1979
0.145
-0.878
1980
4.604
3.581
1981
0.022
-1.001
1982
0.042
-0.981
1983
0.183
-0.840
1984
4.221
3.198
Average 1965-84
1.023
age are not point estimates in time, but averages over
2 or 3 months. As a result, relating the growth incre-
ment to the environment over a short time-period
would tend to blur environmentally -determined growth
differences.
The environmental variables and stock abundance
measures examined in the analysis were intentionally
limited to a few variables which would characterize the
environment of the population on the largest scale
possible. The environmental variables are summer
averages over the geographic range of the mature
stock. Mean summer (April- August) sea-surface tem-
perature (°C), provided by J.G. Norton (Pac. Fish. En-
viron. Group, Monterey, CA 93942, pers. commun.,
Aug. 1989), represents the mean value obtained from
ships of opportunity between lat. 40°N and 50°N, and
from the North American coast west to long. 125°W
(Table 2). The Bakun upwelling indices are a mean of
the monthly coastal upwelling indices during April-
August from lat. 42-48°N and are in units of metric
tons of water transported through the Ekman layer per
second per 100 m of coastline (Mason and Bakun 1986).
Stock biomass is measured in millions of tons of age-2
and older fish in the coastal whiting population, and
is estimated using the stock synthesis model (Dom and
Methot 1990). The estimates of year-class strength in
Table 3 come from the same source.
Results
Length analysis of variance
Annual length-at-age estimates by sex for 1978-88
were obtained using the procedures described in the
Dorn: Environmental covanates of Merluccius productus growth
265
55 n
47
37
32
Males
57
52
£ 47
Appendix for calculating strata esti-
mates of length-at-age, and for com-
bining the strata estimates to pro-
duce an annual estimate (Fig. 3).
The decline in length-at-age is evi-
dent in graphs of both male and
female length-at-age.
The factorial analysis of mean
length variance is given in Table 4.
Weighted analysis of variance was
used with the sampling variances as
weights. The F-tests for age, year,
region, sex, and time-period were
all highly significant (P< 0.0001).
Because of the large number of ob-
servations (2072), this result is not
surprising. The parameter esti-
mates in Table 4 are defined such
that the intercept term represents
the mean length of a 1-year-old
male in the early part of the season
in 1978 in the EUR region. Param-
eter estimates for the other factor
levels can be interpreted as the dif-
ference between mean length of
Pacific whiting identified by that
factor level and those identified by
the intercept characteristic with all
other factors being held constant.
The results of the ANOVA can be
summarized as follows. Length in-
creases with age to age 10, then
varies irregularly to age 15 (Table
4). Length increases 0.55 cm from
the early period (April- June) to the
middle period (July-August), and
increases an additional 0.46cm from
the middle period to the late period
(September-November). There is
an increase of 1.36 cm from the
EUR region in the south to the
VNC region in the north, indicating
that the larger Pacific whiting of an
age-group migrate farther north. On average, female
Pacific whiting are larger than males by 0.55cm. Since
the ANOVA model does not contain a sex-age interac-
tion, this difference in mean length would apply to all
ages. In general, these results are consistent with the
previously reported findings on the growth of Pacific
whiting (Dark 1975, Francis 1983). The ANOVA year
coefficients show the decline in length since 1978. Mean
length-at-age reached a minimum in 1984. There was
a slight rebound in length-at-age from 1984 to 1986,
but after 1986 length-at-age began to decline again.
Models containing interaction terms between factors
37
1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
Females
■ age 1 0
- age 9
- age 8
• age?
- age 6
- age 5
■ age 4
- age 3
1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
Year
Figure 3
Mean U.S. fishery length-at-age estimates by sex for Pacific v/h\tmg Merluccius pro-
ductus, 1978-89. The extremely low estimates of mean length for male and female
age-3 fish in 1979 were due to asymmetric error in the ageing of the strong 1977
year-class.
were also analyzed. All the two-way interaction terms
were statistically significant, but much less so than the
main effects. The addition of more than 200 parameters
to describe the two-way interactions would make inter-
pretation difficult. Yet a model with only main effects
is clearly inadequate, since it implies, for example, that
both the 2-year-old and 12-year-old whiting declined in
mean length by the same amount from 1983 to 1984.
Growth increment regression
Before considering the effects of environmental co-
266
Fishery Bulletin 90(2). 1992
variates, the growth increment
model was used to investigate
sex differences in growth. This
was done by fitting a model with
only go, gi, and indicator vari-
ables for sex. The P-value for the
intercept coefficient for sex was
0.059, but the P-value for the
slope term was 0.998. By the
criteria established earlier,
neither coefficient would be con-
sidered statistically significant,
although the P-value for the in-
tercept coefficient is close to the
critical value. The estimate of
the intercept coefficient for sex
(0.058) indicates that the females
grow approximately 6% (e^^ss)
more than males on an annual
basis, regardless of age. A differ-
ence of this magnitude is suffi-
cient to account for the greater
asymptotic size of the females.
Figure 4 shows the fitted sex-
specific curves for the annual
growth increment. Despite the
lack of statistical significance of
the intercept term for sex, it was
retained in the model while eval-
uating the significance of envi-
ronmental covariates on growth.
The larger size attained by the
female Pacific whiting is com-
pelling evidence that there are
sex-specific differences in Pacific
whiting growth. Including this
term in the baseline model is
important because it accounts
for this sex-specific variability in
growth.
Table 5 shows the analysis
of variance using the annual
growth-increment regression
model. The model was built in a
forward stepwise fashion, adding the environmental
term to the baseline model that resulted in the largest
reduction in the residual sum of squares. Temperature
and population biomass were significant covariates in
the model. Temperature had significant intercept (P<
0.001) and slope terms (P 0.026). For biomass, only the
slope term was significant (P 0.002).
The parameter estimates in Table 5 indicate that a
0.5°C increase in mean summer sea-surface tempera-
ture will bring about a 24% reduction in the annual
growth increment at age 1. At age 4, the same increase
Table 4
Factoria
analysis of variance
of Pacific whiting Merluccius productvs length-at
age usmg
midwater trawl fishery samp
es over 1978
-88. The model contains the factors
age, year,
geographic region,
sex, and
season. The
intercept term estimates the mean
length of
a 1 -year-
old male in the early part of the
season in 1978 in
the EUR region;
the other
terms estimate the difference in the mean
length of fish with that factor level and those |
with intercept characteristic
Source
df
SS
Mean square
F -value
P>F
Age
14
1,112,327.5
79,452.0
3611.5
<0.001
Year
10
33,126.7
3,312.7
150.6
< 0.001
Region
2
5,810.9
2,905.4
132.1
<0.001
Sex
1
2,239.2
2,239.2
101.8
<0.001
Season
2
3,531.0
1,765.5
80.3
<0.001
Error
2042
44,841.1
22.0
Parameter
Estimate (cm)
SE of estimate
Intercept
26.88
0.211
Age:
2
3
4
5
6
7
8
9
10
11
12
13
14
15
7.90
13.22
17.02
18.44
19.39
20.76
22.29
24.04
29.92
26.51
26.07
27.56
27.69
29.94
0.201
0.189
0.147
0.182
0.200
0.205
0.190
0.262
0.308
0.361
0.290
0.528
0.555
0.453
Year:
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
-0.82
-0.50
-1.54
-2.33
-3.31
-4.64
-3.00
-2.64
-3.03
-4.00
0.198
0.174
0.170
0.169
0.187
0.169
0.157
0.175
0.176
0.174
Region:
SCOL
VNC
0.57
1.36
0.071
0.089
Sex:
Females
0.55
0.052
Season:
Middle
Late
0.53
0.99
0.065
0.079
in temperature would be expected to produce a 12%
reduction in annual growth, and by age 7, the percent
reduction would be close to zero. The model predicts
an increase in growth due to increasing temperature
above age 7, but as annual growth is very slight by this
age, the consequences of this prediction are not impor-
tant. One concern about the reliability of these results
is that they may be overdependent on growth during
the 1983 El Nino, when sea-surface temperature was
the highest during the study. To investigate this pos-
sibility, a model was fit to the data excluding the
Dorn: Environmental covariates of Merlucaus productus growth
267
Age, yrs
Figure 4
Growth-increment regression curves with different initial growth-increment
coefficients (gf,) for male and female Pacific whiting Merluccius productus.
The distribution of actual growth increments for the years 1978-89 illustrates
the saddle-shaped variability of length-at-age statistics based on fishery sam-
pling (W 211). Points represent observed annual growth increments per age-
group per year.
Table 5
Analysis of variance of annual length increments using the nonlinear regres-
sion model. The coefficient g„ is an intercept term: exp(g„) estimates the
growth increment from age 1 to age 2. The coefficient g, determines the
slope of the exponential decline of the annual growth increment with age
under average environmental conditions. Terms relating to covariates are
identified as either intercept terms (gj) or as slope terms (g,).
Source
df
SS
Mean square F -value P>F
go (sex)
go (temp.)
gj (biomass)
gi (temp.)
Error
204
4067.8
12.6
93.5
36.3
17.4
711.4
4067.8
12.6
93.5
36.3
17.4
3.5
1166.5
3.6
26.8
10.4
5.0
Parameter
Estimate
SE of estimate
go
gi
go (sex)
go (temp.)
gi (biomass)
gi (temp.)
1.995
-0.383
0.058
-0.544
-0.086
0.099
0.038
0.015
0.029
0.098
0.027
0.043
growth increments from 1983. The parameter esti-
mates followed the same trend as the results in Table
5, with increases in sea-surface temperature associated
with large reductions in growth of younger fish, and
a decreasing percent reduction with age.
For population biomass, an increase of 200,000t
(~10% of the mean population biomass)
would cause a 5% reduction in annual
growth at age 4, and the percent reduction
would increase with age, reaching a 10%
reduction at age 7. Because the intercept
term for biomass was not significant,
growth of the age-1 fish would not be af-
fected by changes in population biomass.
During the 10 years studied, range in adult
biomass was from 26% above to 22% below
the mean of 1.771 million t (1978-87). This
lack of contrast in adult biomass makes
any interpretation very tentative, but the
results do suggest that the effect of pop-
ulation density on growth is relatively
small in comparison with the effect of
temperature.
None of the other environmental covari-
ates or measures of population density
tested in the model were significant. The
coefficient for upwelling was highly signifi-
cant in a model without temperature, but
when temperature was included in the re-
gression, upwelling was no longer signifi-
cant. Because water temperature and the
upwelling index are statistically correlated
off the coast of Oregon (Kruse and Huyer
1983), the parallel effects of temperature
and upwelling index are not unexpected.
The same parallel effect was found be-
tween year-class abundance and adult
biomass. However, adult biomass ac-
counted for more of the variability in
annual growth than did year-class abun-
dance. This suggests that the crowding
that occurs when a strong year-class
recruits to the population is experienced
by all the adults, and not just the individ-
uals which make up the strong year-class.
Figure 5 shows the standardized resid-
uals from the final model, plotted against
age and year. No trends are evident with
respect to age. This plot also shows that
the use of the estimation variances as
weighting terms in the least-squares fit
was successful in stabilizing the error vari-
ance with respect to age. However, there
still is a noticeable trend in the standard-
ized residuals by year, with positive resid-
uals associated with the earlier years in the time-series
(1979 and 1980), and negative residuals with the later
years (1985 and 1987). This indicates that the environ-
mental covariates examined thus far are not completely
successful in accounting for the decline in mean length-
at-age over the past decade.
< 0.001
0.059
<0.001
0.002
0.026
268
Fishery Bulletin 90(2). 1992
Residuals by year
-1 1 1 1 1 1 1 1 1 1
1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
Year
Residuals by age
10 11 12 13 14 15
Figure 5
Standardized residuals by age and by year for the final model, which
included gj, and gi (coefficients modeling asymptotic growth), and coef-
ficients for sex, sea-surface temperature, and population density. Lines
connect averages of the residuals at each cluster of points.
Discussion
Although density-dependent growth has been demon-
strated for many fishes (Shepherd and Grimes 1983,
Reish et al. 1985, Ware 1985, Ross and Almeida 1986,
Peterman and Bradford 1987, Overholtz 1989), few
researchers would argue that density-dependent
growth is an important characteristic of all fish popula-
tions. The age-structured yield models first developed
by Beverton and Holt (1957)— currently used to man-
age many temperate marine fish stocks— use a fixed
schedule of weight-at-age to calculate the yield, re-
gardless of the level of population abundance. For a
stock of Atlantic mackerel, however, the his-
torically observed variation in weight-at-age at-
tributable to density-dependent growth had a
significant effect on the projected yields from
the fishery (Overholtz et al. 1991). For Pacific
whiting, this potential for changes in weight-
at-age to influence yield is taken into account
by using the weight-at-age observed in recent
years to project the yield for the upcoming year
(Dorn and Methot 1990). To obtain a fishing
mortality rate that gives the long-term sus-
tainable yield, the average weight-at-age over
the history of the fishery is used. This strategy
tacitly assumes that the current decline in
weight-at-age is not a permanent change in the
population.
Parrish et al. (1981) state that the principal
resident species of the California Current
system do not exhibit density-dependent
growth. They contend that the population size
of these species is controlled by environmen-
tal variability during the larval stages. As a
result, the adults are seldom plentiful enough
to reach a food -limited carrjang capacity.
A contrasting viewpoint is found in Boehlert
et al. (1989) who present evidence that the
large biomass removals oiSebastes spp. in the
years 1966-70 off the west coast of the United
States resulted in increases in the annual
growth of two members of this genus: canary
rockfish S. pinniger and splitnose rockfish S.
diploproa. They maintain that the decline in
the total abundance of Sebastes spp. has had
an effect on food availability for individual
rockfish species. Although euphausiids are
shared by most Sebastes species as the principal
prey (Brodeur and Percy 1984), they are also
a major link in the food chain of the California
Current ecosystem, supporting numerous fish
and invertebrate populations. For this reason,
it is unlikely that changes in the abundance of
Sebastes spp. alone could have had a substan-
tial impact on the overall abundance of euphausiids
in the California Current ecosystem. However, since
rockfish are spatially restricted to habitats with limited
area, the density-dependent growth displayed by S.
pinniger and S. diploproa may be due to density-
dependent changes in the food availability within these
habitats. The possibility that cropping by Pacific
whiting and other species significantly affects the abun-
dance of euphausiids in the California Current eco-
system at large has not yet been adequately tested,
though Mullin and Conversi (1989) were unable to
detect any change in the abundance of euphausiids in
the California current system after the start of the
Dorn: Environmental covariates of Merlucaus productus growth
269
large-scale fishery for Pacific whiting Merluccius -pro-
ductus in 1966.
Environmental influence on growth has been ob-
served for many marine fish species (Kreuz et al. 1982,
Anthony and Fogarty 1985). Because it can be easily
measured and is associated with widespread changes
in the aquatic environment, water temperature is the
covariate most often studied. Water temperature may
have a direct physiological effect on the growth of fish,
or it may be indirectly linked to growth. For example,
decreases in water temperature occur with the onset
of upwelling in many coastal marine environments. On
the west coast of North America, the availability of
coastal upwelling indices on monthly, weekly, and daily
time scales (Mason and Bakun 1986) has made it pos-
sible to investigate the direct effect of upwelling on
growth, although convincing evidence of a link has not
yet been found (Kreuz et al. 1982, Francis 1983,
Boehlert et al. 1989).
Results from the Pacific whiting growth-increment
regression show that an environmental covariate, sea-
surface temperature, and population density could
explain the deviations from a simple baseline model for
asymptotic growth. The effect of temperature was
greatest on the youngest ages present in the fishery
samples and declined as age increased. In contrast, the
effect of population biomass on annual growth in-
creased with age. Temperature was the most impor-
tant covariate, both in terms of its statistical signif-
icance and its effect on growth. This association of
enhanced growth and reduced sea temperature is con-
sistent with what is known about the California Cur-
rent, a major eastern boundary current system. Kreuz
et al. (1982) found an identical inverse effect on the
growth of English sole Parophrys vetulus and Dover
sole Microstomus pacificus at two locations off the
Oregon coast.
Two mechanisms are believed to contribute to the
high productivity of the California Current system.
Coastal upwelling, generated by wind-driven offshore
transport in the surface Ekman layer, brings low
temperature, low salinity, and nutrient-rich water to
the surface (Bakun and Nelson 1977). A second
mechanism is the southward advection of water from
the Alaskan Subarctic Gyre. This water is characterized
by low temperatures, high nutrient content, and a large
standing stock of zooplankton (Roesler and Chelton
1987). Regardless of which mechanism is dominant dur-
ing a particular year, low mean sea-surface tempera-
ture during the summer can be expected to be asso-
ciated with high productivity.
The diet of Pacific whiting provides the link between
primary productivity and growth. The major prey of
Pacific whiting are euphausiids, primarily Thysanoessa
spinifera and Euphausia pacifica (Livingston 1983,
Rexstad and Pikitch 1986). In summer, the abundance
and pattern of distribution of these short-hved species
(1-2 yr) are closely tied to upwelling and primary pro-
ductivity (Simard and Mackas 1989). Rexstad and
Pikitch (1986) found that euphausiids comprised 90%
by weight of the diet of Pacific whiting 30-44 cm in
length collected during a trawl survey in 1983 off the
coasts of Oregon and Washington. Above 45 cm, a
length which corresponds approximately to ages 5-7,
this percentage drops to 20%. In the diet of the older
Pacific whiting, euphausiids are largely replaced by
small schooling fish and shrimp. These include northern
anchovy Engraulis mordax, Pacific herring Clupea
harengus pallasi, eulachon Thaleichthys pacificus, pink
shrimp Pandalus jordani, and rockfish Sebastes spp.
(Livingston 1983).
This shift in dietary preferences by Pacific whiting
may help explain the effect of population biomass on
growth. When the biomass of Pacific whiting is high,
predation on fish species with multiyear life cycles may
become intense enough to reduce their availability to
the whiting population. In contrast, euphausiid abun-
dance is closely coupled to annual variations in produc-
tivity, so whiting predation would likely have little
effect on their abundance.
Some supporting evidence for density-dependent
growth of Pacific whiting is found in Dark (1975), who
also documented a decline in length-at-age using fishery
samples from an earlier period in the fishery, 1964-69.
At the time of Dark's research, estimates of popula-
tion abundance were not available for Pacific whiting.
It is now believed that the 1961 year-class was excep-
tionally strong, nearly the same size as the record 1980
year-class (Dorn and Methot 1990). Consequently, this
earlier decline in length-at-age may also be partly at-
tributable to increases in population density as the 1961
year-class moved into the population.
Although weight-at-age is the measure of size typical-
ly used in stock assessment models, the analysis in this
paper focuses on length rather than weight. A prac-
tical reason for this strategy is that most at-sea sam-
pling platforms are not sufficiently stable to obtain
accurate individual weights offish. Indeed, the weights-
at-age used in stock assessment models for Pacific
whiting are obtained by first estimating length-at-age,
then converting to weight using a length-weight rela-
tionship (Dorn and Methot 1990). In addition, growth
in length has several characteristics that make it
amenable for analytical modeling. Except in very rare
instances, changes in fish length are always positive
or zero. The annual growth increment in length sum-
marizes the growth response of the organism to envi-
ronmental conditions that are prevalent throughout the
year, or are short-term. In contrast, weight-at-age has
a seasonal pattern of increase and decline, associated
270
Fishery Bulletin 90(2). 1992
with spawning, migration, and feeding, which
would have to be accounted for in a model
before analyzing environmental influences on
growth.
Underlying this seasonal pattern of variation
in weight is the length-weight relationship
characteristic to a species, determined by the
overall shape of the fish. Extreme departures
from the typical length-weight relationship are
unlikely to persist. Fish that are heavy in rela-
tion to their length in one year would tend to
grow faster in length than average, while
underweight fish would tend to experience
slower growth in length. Adjustments to an in-
dividual's annual reproductive effort can also
dampen departures from the typical length-
weight relationship (Tyler and Dunn 1976).
During the period covered by this analysis,
the length-weight relationship of Pacific whit-
ing has varied from year to year, most notice-
ably in 1983, when mean weight was extremely
low at a given length (Dorn and Methot 1990).
The link between anomalies in the length-
weight relationship and annual growi;h incre-
ments is best demonstrated by the results of
a trial model that used the anomaly in the
estimated weight at 45 cm from the annual length-
weight regression (Dorn and Methot 1990) as the only
predictor variable for the annual growth increment.
This variable was highly significant in the model
(P< 0.001), indicating that the annual growth incre-
ment is low during years where the length-weight rela-
tionship is below average. This result also supports the
hypothesis that variation in Pacific whiting length-at-
age is caused by environmental processes that affect
the availability of food.
The analysis presented in this paper is based ex-
clusively on fishery data. It should be acknowledged
that there are numerous problems associated with the
use of fishery statistics to infer growth patterns of fish
within a population. Incomplete, size-dependent re-
cruitment to a fishery can make fishery data on length-
at-age a biased estimate of population length-at-age.
Ageing error can distort the estimates of length-at-age
when the year-classes have large differences in abun-
dance. Shifts in the geographic and temporal pattern
of the fishery, or shifts in the geographic distribution
of the population itself, can cause spurious changes in
estimates of length-at-age. The lengths for less abun-
dant age-groups are not estimated as precisely as those
for abundant age-groups. This is particularly true of
extremely young and old fish, as these age-groups may
be represented in fishery samples by only one or two
individuals which determine the mean length for that
age-group.
100
1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
Year
Figure 6
Percent of the Pacific whiting Merliiccius productvs catch biomass taken
north of Cape Falcon flat. 46°45'N) near the mouth of the Columbia
River, during 1978-88. Catch of Pacific whiting from the Canadian zone
is included in the calculations.
The severity of some of these problems can be re-
duced by using the procedures described in the Appen-
dix for compiling strata estimates of length-at-age and
calculating variance estimates. Length-at-age for tem-
poral and geographic strata can be examined separately
before being combined to produce annual summary
statistics. Length-at-age estimates based on only a few
individuals can be discounted in the analysis by using
the estimated variances of length-at-age as weights.
Nevertheless, some factors affecting growth can only
be addressed by modeling fisheries as both a source of
information on the stock and a major influence on its
dynamics. The growth-increment regression model
used in this paper assumes that the fishery samples the
population without bias, so it is not the appropriate
framework for studying these processes. Models with
size-selective fishing mortality and stochastic growth
have been developed for exploited fish populations
(Deriso and Parma 1988, Parma and Deriso 1990). The
practical application of these models is limited by the
difficulty of distinguishing between different sources
of growth variability using only catch data.
Size-selective mortality may have played a role in
causing variation in length-at-age of Pacific whiting
over the years covered by this analysis. Since the length
ANOVA found a significant increase in length from
south to north, a northward shift in the fishery would
tend to increase length-at-age in the catch. At the same
time, however, the length-at-age of the survivors of the
Dorn: Environmental covanates of Merlucaus productus growth
271
2 -
0 -
?
u
CD
-4-
.^-^
^
'v
^\f^
1978 79 80 81 82 83 84 85 86 87 88
Year
Figure 7
Estimated coefficients for a region-year interaction in a factorial analysis
of variance, with the terms age, year, region, sex, season, and region-
year interaction, for U.S. fishery length-at-age estimates of Pacific
whiting Merliiccius productus over 1978-88.
fishery would decrease, and this would tend to decrease
the length-at-age in the catch in the following year from
what it would have been otherwise. From 1978 to 1982,
the fraction of the Pacific whiting catch taken north
of Cape Falcon (near the mouth of the Columbia River
at lat. 46°45'N) increased from 10% to 80%, and was
60% in 1988 (Fig. 6). The lack of fit of the growth-
increment regression model with respect to year in
Figure 5 may have been a result of this northward shift
in fishing mortality.
However, it is difficult to predict the long-term ef-
fects of a shift in the geographic pattern of exploita-
tion on length-at-age, because little is known about the
extent of mixing from one year to the next of fish
migrating from different regions. Without mixing be-
tween regions, an increase in fishing in the northern
part of the range would reduce the abundance of larger
individuals of an age-group, reducing the overall pop-
ulation length-at-age, while length-at-age of the south-
ern fish would be unaffected. A more likely hypothesis
is that some inter-regional mixing occurs from year to
year. In this case, the length-at-age in all regions would
decrease, though the magnitude of the decrease should
be greatest in the north where the higher fishing oc-
curred. Ultimately, this would tend to reduce latitudinal
variation in length-at-age.
To support this hypothesis, there is some
evidence of a change in the degree of latitu-
dinal segregation by size of Pacific whiting.
Figure 7 shows the coefficients for a region-
year interaction for a length ANOVA with
main effects being age, year, region, sex, and
season. The absence of interaction between
year and region would be identified by parallel
year effect lines for each region, and would in-
dicate that size-specific migratory pattern has
remained constant. From 1978 to 1985, the
region-year interaction does not appear promi-
nent. After 1985, however, the lengths-at-age
in the three regions become much closer
together; in particular, the lengths of the fish
in the VNC region, instead of being 1-2 cm
larger than the fish in the other regions, are
the same size or smaller.
Recently, Smith et al. (1990) examined the
length-at-age data from the fishery for Pacific
whiting in the Canadian waters over the same
years examined by this paper. They used
a generalized form of the von Bertalanffy
growth model that makes length-at-age a func-
tion of length-at-age in the previous year, plus
environmental covariates modeled in different
ways according to a hypothesized mechanism
by which the environmental covariate affects
growth or apparent growth. Significant covari-
ates in their model were population biomass, a suite
of oceanographic variables measuring the strength of
southward advection of water from the Alaskan Sub-
arctic Gyre, and several variables that model size-
selective mortality. Since fish younger than age 5 are
not common in the Canadian samples, their analysis
could not examine the sources of growth variability of
the younger fish. Consequently, the analysis presented
here on the environmental covariates of Pacific whiting
growth in U.S. waters is a necessary complement to
the paper by Smith et al. (1990). For example, an in-
verse relationship between temperature and growth,
which is most pronounced for the younger fish, was not
detected by Smith et al. (1990), and can partly account
for the fact that the fish currently recruiting to Cana-
dian waters at age 5 are much smaller than those
recruiting in the late 1970's.
A major contention of Smith et al. (1990) is that ex-
pansion of the Canadian fishery is largely responsible
for the decline in length-at-age observed since 1976.
They used the ratio of the Canadian catch (in biomass)
to the total population biomass during the current year
as a covariate in their nonlinear regression model, a
phenomenological approach that sidesteps the need to
model the population dynamics. Although the mono-
tonically increasing Canadian catch of Pacific whiting
272
Fishery Bulletin 90(2), 1992
since 1976 (except for 1985) and the declining length-
at-age over the same period guarantees a statistically
significant result, the real role of the Canadian fishery
in determining length-at-age can be established only
in a wider context that considers the magnitude and
the geographic pattern of the fishery for Pacific whit-
ing in U.S. waters. The region-year interaction coeffi-
cients in Figure 7 show that, up until 1984, the sever-
ity of the decline in length-at-age was similar in all
three geographic regions in U.S. waters, extending
from California north to the U.S. -Canada border. This
is difficult to reconcile with the contention that the
Canadian fishery is primarily responsible for the decline
in length-at-age.
Both Smith et al. (1990) and this study used growth
models that do not take into account the dynamics of
the population, and as a consequence both have short-
comings which limit the growth- related phenomena to
which they can be applied. An important direction for
further research is the development of models for
Pacific whiting that simultaneously model the growth
and the population dynamics of the stock, including
size-specific migratory behavior.
Acknowledgments
I thank Anne Hollowed, Daniel Kimura, Patricia
Livingston, and Richard Methot for their comments on
a preliminary version of this paper. In addition, sug-
gestions by two anonymous reviewers significantly
improved this paper.
Citations
Anthony, V.C, and M.J. Fogarty
1985 Environmental effects on recruitment, growth, and vul-
nerability of Atlantic herring (Clupea harengus) in the Gulf of
Maine region. Can. J. Fish. Aquat. Sci. 42(Suppl. 1);158-173.
Bailey, K.M., R.C. Francis, and P.R. Stevens
1982 The life history and fishery of Pacific whiting, Merluc-
ciiis productus. Calif. Coop. Oceanic Fish. Invest. Rep. 23:
81-98.
Bakun, A., and C.S. Nelson
1977 Climatology of upwelling related processes off Baja
California. Calif. Coop. Oceanic Fish. Invest. Rep. 19:107-127.
Beverton, R.J.H., and S.J. Holt
1957 On the dynamics of exploited fish populations. Fish.
Invest. Minist. Agric. Fish. Food (G.B.) Ser. 11(19), 533 p.
Boehlert, G.W., and R.F. Kappenman
1980 Variation of growth with latitude in two species of rock-
fish (Sebastes pinniger and S. diploproa) from the northeast
Pacific Ocean. Mar. Ecol. Prog. Ser. 3:1-10.
Boehlert, G.W., M.M. Yoklavich, and D.B. Chelton
1989 Time series of growth in the genus Sebastes from the
northeast Pacific Ocean. Fish. Bull., U.S. 87:791-806.
Brodeur, R.D., and W.G. Pearcy
1984 Food habits and dietary overlap of some shelf rockfishes
(genus Sebastes) from the northeastern Pacific Ocean. Fish.
Bull., U.S. 82:269-294.
Dark, T.A.
1975 Age and growth of Pacific hake, Merluccius productus.
Fish. Bull., U.S. 73:336-355.
Deriso, R.B., and A.M. Parma
1988 Dynamics of age and size for a stochastic population
model. Can. J. Fish. Aquat. Sci. 45:1054-1068.
Dixon, W.J. (editor)
1983 BMDP statistical software. Univ. Calif. Press, Berkeley,
734 p.
Dorn, M.W., and R.D. Methot
1990 Status of the Pacific whiting resource in 1989 and recom-
mendations for management in 1990. NOAA Tech. Memo.
NMFS F/NWC-182, Alaska Fish. Sci. Cent., Seattle, 84 p.
Francis, R.C.
1983 Population and trophic dynamics of Pacific hake (Merluc-
cius productus). Can. J. Fish. Aquat. Sci. 40:1925-1943.
French, R., R. Nelson Jr., and J. Wall
1981 Results of U.S. observer sampling on Japanese, U.S.S.R.,
Korean, and Polish fishing vessels, 1979. Int. North Pac. Fish.
Comm., Annu. Rep. 1980:96-122.
Hollowed, A.B., S.A. Adlerstein, R.C. Francis. M. Saunders,
N.J. Williamson, and T.A. Dark
1988 Status of the Pacific whiting resource in 1987 and recom-
mendations to management in 1988. NOAA Tech. Memo.
NMFS F/NWC-138, Alaska Fish. Sci. Cent., Seattle, 54 p.
Kimura, D.K.
1980 Likelihood methods for the von Bertalanffy growth
curve. Fish. Bull., U.S. 77:765-776.
1989 Variability in estimating catch in numbers-at-age and its
impact on cohort analysis. In Beamish, R.J., and G.A. McFar-
lane (eds.). Effects of ocean variability on recruitment and an
evaluation of parameters used in stock assessment models, p.
56-66. Can. Spec. Publ. Fish. Aquat. Sci. 108.
Kimura D.K., and S. Chikuni
1987 Mixtures of empirical distributions: An iterative applica-
tion of the age-length key. Biometrics 43:23-34.
Kreuz, K.F., A.V. Tyler, G.H. Kruse, and R.L. Demory
1982 Variation in growth of Dover soles and English soles as
related to upwelHng. Trans. Am. Fish. Soc. 111:180-192.
Kruse, G.H., and A. Huyer
1983 Relationships among shelf temperatures, coastal sea level,
and the coastal upwelling index off Newport, Oregon. Can.
J. Fish. Aquat. Sci. 40:238-242.
Livingston, P. A.
1983 Food habits of Pacific whiting, Merluccitcs productus, off
the west coast of North America, 1967 and 1980. Fish. Bull.,
U.S. 81:629-636.
Mason, J.E., and A. Bakun
1986 Upwelling index update, U.S. West Coast, 33N-48N
latitude. NOAA Tech. Memo. NMFS-SWFC-67, Southwest
Fish. Sci. Cent., La Jolla, 81 p.
Mullin, M.M.. and A. Conversi
1989 Biomasses of euphausiids and smaller zooplankton in the
California current— Geographic and interannual comparisons
relative to the Pacific whiting, Merluccius productus, fishery.
Fish. Bull., U.S. 87:633-644.
Overholtz, W.J.
1989 Density-dependent growth in the northwest Atlantic
stock of Atlantic mackerel {Scomber scombrus). J. Northwest
Atl. Fish. Sci. 9:115-121.
Dorn: Environmental covariates of Merlucaus productus growth
273
Overholtz, W.J., S.A. Murawski, and W.L. Michaels
1991 Impact of compensatory responses on assessment advice
for the Northwest Atlantic mackerel stock. Fish. Bull., U.S.
89:117-128.
Parma, A.M., and R.B. Deriso
1990 Dynamics of age and size composition in a population sub-
ject to size-selective mortality: Effects of phenotypic variability
in growth. Can. J. Fish. Aquat. Sci. 47:274-289.
Parrish, R.H., C.S. Nelson, and A. Bakun
1981 Transport mechanisms and reproductive success of fishes
in the California Current. Biol. Oceanogr. l(2):175-203.
Pauly, D., and G. Gaschiitz
1979 A simple method for fitting oscillating length growth data
wdth a program for pocket calculators. Int. Counc. Explor.
Sea CM 1979/0:24.
Peterman, R.M., and M.J. Bradford
1987 Density-dependent growth of age 1 English sole (Paro-
phrys vetidus) in Oregon and Washington coastal waters. Can.
J. Fish. Aquat. Sci. 44:48-53.
Ratkowsky, D.A.
1983 Nonlinear regression modeling. Marcel Dekker, NY,
276 p.
Reish, R.L., R.B. Deriso, D. Ruppert, and R.J. Carroll
1985 An investigation of the population dynamics of Atlantic
menhaden (Brevoortia tyrannvs). Can. J. Fish. Aquat. Sci.
42(Suppl. 1):147-157.
Rexstad, E.A., and E.K. Pikitch
1986 Stomach contents and food consumption estimates of
Pacific hake, Merluccius productus. Fish. Bull., U.S. 84:
947-956.
Roesler, C.S., and D.B. Chelton
1987 Zooplankton variability in the California Current, 1951-
1982. Calif. Coop. Oceanic Fish. Invest. Rep. 28:107-127.
Ross, M.R., and P.P. Almeida
1986 Density-dependent growth of silver hakes. Trans. Am.
Fish. Soc. 115:548-554.
Shepherd, G., and C.B. Grimes
1983 Geographic and historic variations in growth of weakfish,
Cynoscion regalis. in the Middle Atlantic Bight. Fish. Bull.,
U.S. 81:803-813.
Simard, Y., and D.L. Mackas
1989 Mesoscale aggregations of euphausiid sound scattering
layers on the continental shelf of Vancouver Island. Can. J.
Fish. Aquat. Sci. 46:1238-1249.
Singh, B.D., and D. Singh
1965 Some remarks on double sampling for stratification.
Biometrika 52:587-590.
Smith, B.D., G.A. McFarlane, and M.W. Saunders
1990 Variation in Pacific hake (Merluccius productus) summer
length-at-age near southern Vancouver Island and its relation-
ship to fishing and oceanography. Can. J. Fish. Aquat. Sci.
47:2195-2211.
Tyler, A.V., and R.S. Dunn
1976 Ration, growth, and measures of somatic and organ
condition in relation to meal frequency in winter flounder,
Pseudopleuronectes americanus, with hypotheses regarding
population homeostasis. J. Fish. Res. Board Can. 33:63-75.
Ware, D.M.
1985 Life history characteristics, reproductive value, and re-
silience of Pacific herring (Clupea harengus pallasi). Can J.
Fish. Aquat. Sci. 42(Suppl. 1):127-137.
Weisberg, S.
1986 A linear model approach to backcalculation of fish length.
J. Am. Stat. Assoc. 81:922-929.
Appendix: Variance estimates of
mean length-at-age using a two-phase
sampling procedure
The length and age samples collected by observers
in the Pacific whiting fishery are recorded by haul or
joint-venture delivery. The position and date of each
sampled haul and joint-venture delivery are also re-
corded. In compiling the length-at-age estimates for
spatial and temporal strata, all the data collected within
that strata are aggregated and assumed to originate
from random sampling from the catch within that
strata.
Sampling design
A large initial random sample is obtained from the
catch, and the length and sex of each fish is recorded.
For the second phase of sampling, a subsample of fixed
size is selected for each combination of length category
and sex. All fish in these subsamples are aged using
otoliths or other ageing structures.
Notation
i = 1,. .., I length categories
j = 1 , . . . , J age categories
n' = first-phase sample size
n'i = number of fish of n' in length category i
qi = probability that a fish is in length category i
nj = subsample sizes
njj = number of the subsample taken from length
category i of age j
q'ij = Pi"(j I i)' probability of age j given length i
qjj = pr(i|j), probability of length category i given
age j
Pj = pr(j), probability of age j
1; = midpoint of \th length category
Tj = mean length of age j fish
To simplify notation, subscripts for males and fe-
males are not defined. The variance estimator obtained
here is conditional on the first- and second-phase sam-
ple sizes. Separate estimates for males and females
can be obtained by separating the samples by sex and
conditioning on the number of each sex in the first- and
second-phase samples. The same variance estimator is
appropriate for separate sex estimates.
Sampling distributions
Assuming that the first-phase sample size n' is much
smaller than the size of the population being sampled,
the distribution of n'j can be modeled by the multi-
nomial distribution,
274
Fishery Bulletin 90(2). 1992
f({n'i}|n',(qi}) =
V I
n
n'i ! I .
n qi" .
where Z qi = 1. Estimators for {qi} and the elements
i
of the variance-covariance matrix of {qi} are
n'i
qi = — -
n
Var(qi) =
qi(i-qi)
n'
C6v(qj,qh) =
qiqh
To obtain a distribution for {njj}, it is convenient to
condition on the fixed subsamples nj. As in the case
of the first phase of samphng, it is assumed that n-, is
much less than the number in the population of that
length category, so that a product of multinomial
distributions is obtained for the second phase of
sampling,
f({nij} I (n,}, {q',j}) = fl
rin,!
n q'ij"'
where Z q'jj = 1 for all i. Estimators for (q'ij} and the
j
elements of the variance-covariance matrix of {q'jj}
are
Var(q,) = ^^X^^^M,
While they were able to obtain estimators which cor-
rectly modeled the sampling procedures, they also
found that their exercise in theoretical rigor did not
result in any appreciable difference in practice.
Estimation of mean length-at-age
An unbiased estimate of Ij is given by
^j = Z li qij.
where qjj is the probability of length i given age j. An
expression for qjj is obtained using Bayes theorem,
qu
qiqi
Z qi q'i
A variance approximation
for mean length-at-age
Because the above expression for mean length-at-age
is nonlinear in the observations {n'j} and (n|j}, a delta-
method approximation is derived. Delta-method esti-
mators can be algebraically complex but all have the
same simple structure. For mean length-at-age, a delta-
method approximation is given by
Var(Tj) = djT Vdj.
where dj is the vector of partial derivatives of Ij with
respect to {qj} and {q'ij}, and V is the variance-covari-
ance matrix of (qj} and {q'ij}. Defining
and
Aj = Z li qi q'ij.
Pj = Z qi q'ij.
C6v(q'ij, q'hk) =
q ij q hk
Hi
for i = h and zero otherwise.
A troublesome inconsistency with this approach is
that the n; are assumed to be predetermined quan-
tities. In fact, n; is necessarily less than or equal to n'i,
the number in the ith length category from the first-
phase sampling, and n'j is a random variable that can
take values between zero and the min(n',Ni) where
Ni is the number of fish of length category i in the
total catch. Singh and Singh (1965) address this issue
while developing variance estimators for what in this
fisheries application would correspond to mean age.
the elements of the vector of partial derivatives are
given by
and
3 1j ^ q'ij (liPj - Aj)
9 qi Pj'
dlj ^ qj (liPj - Aj)
a q'ij Pj^
Combining these expressions with the estimators for
the variance-covariance matrix of qi and q'ij given
earlier,
Dorn: Environmental covariates of Merluccius productus growth
275
Var(lj) =
1
2.(liPj-Aj)2 '- + ^ '- - 2.Zqiqhqijqhj(liPj-Aj) — '-^-^
i#h
Combining length-at-age from different strata
An estimate of combined mean length-at-age is
'-m'-
where Ijh is the length-at-age in the hth stratum, and Cjh is the catch-at-age for the same stratum.
Again using a delta-method approximation,
Var(lj) = IVar(Cjh)
h
(IjhCj - Xcjh Ijh)
h
+ IVar(ljh)
h
The complete expression should include a term involving Cov(Cjh, Ijh)- This term was always negligible com-
pared with the other two terms, and depended on the method used to calculate the catch-at-age. Consequently,
it is not included here. The variance estimator for combined length-at-age for several strata requires an estimate
of the variance of catch-at-age. A method for obtaining this is given in Kimura (1989).
Abstract. — Comparisons are
made between estimates of ages and
growth of the flathead Platycephalus
speculator Klunzinger, from a tem-
perate Western Austrahan estuary,
using data obtained from whole and
sectioned otoliths. The consistent an-
nual trends shown by the width of
the opaque zone on the periphery
(marginal increment) of sectioned
otoliths, irrespective of the number
of translucent zones, demonstrate
that the translucent zones in these
otoliths correspond to annuli. While
the marginal increments on whole
otoliths also showed a similar marked
and consistent annual trend when a
single translucent zone was present,
they were far less conspicuous when
two or more translucent zones were
observed. The large sample size and
strong trends shown by marginal in-
crements on otoliths exhibiting one
translucent zone accounts for the
fact that, when data for all whole
otoliths are pooled, the marginal in-
crement still shows a consistent an-
nual trend. Sectioning of otoliths
enhances the ability to differentiate
between the outer opaque and trans-
lucent zones, and also often reveals
one or more additional inner translu-
cent zones in older fish. The use of
whole otoliths frequently underesti-
mated age by one year in 2 -t- to 4 -i-
fish and two years in 5 -i- to 10 -i- fish,
and by as much as five or six years
in the oldest fish (1 1 -t- and 12 -i- ). The
respective 95% confidence limits for
the parameters La>, K, and tp in
the von Bertalanffy growth equa-
tions for males, calculated using data
from sectioned otoliths, overlapped
those calculated from data for whole
otoliths, and the same was true for
K with females. This similarity in
growth curves in particularly the
first four years of life can be attrib-
uted to the fact that approximately
74 and 65% of the growth of males
and females, respectively, occurred
in the first three years, when under-
estimates of age were limited.
Influence of sectioning otoliths
on marginal Increment trends
and age and growth estimates for
the flathead Platycephalus speculator
Glenn A. Hyndes
School of Biological and Environmental Sciences, Murdoch University
Murdoch, Western Australia 6150, Australia
Neil R. Loneragan
School of Biological and Environmental Sciences, Murdoch University
Murdoch, Western Australia 6150, Australia
Present address: CSIRO Division of Fisheries
P.O. Box 120, Cleveland, Queensland 4163, Australia
Ian C. Potter
School of Biological and Environmental Sciences, Murdoch University
Murdoch, Western Australia 6150, Australia
Manuscript accepted 18 February 1992.
Fishery Bulletin, U.S. 90:276-284 (1992).
Since assessments of fish stocks often
rely on information on age composi-
tion, it is crucial that any age esti-
mates used for such assessments are
validated (Beamish and McFarlane
1983, Casselman 1987). Validation
that growth zones on hard structures,
such as otoliths, scales, and spines,
are formed annually is often implied
by establishing that the pattern of
growth on the periphery of these
structures follows a consistent an-
nual trend (e.g., Johnson 1983, Ma-
ceina et al. 1987, Potter et al. 1988,
Beckman et al. 1989). In otoliths of
fish in temperate waters, an opaque
zone generally starts to form in the
spring immediately outside the trans-
lucent zone laid down during the
preceding winter. The width of this
opaque zone usually increases be-
tween spring and autumn. Although
a subsequent retardation of growth
during winter results in formation of
the translucent zone at the edge of
the otolith, this translucent zone fre-
quently cannot be readily detected
until the following spring, when it
becomes delineated by the formation
of a new opaque zone. The distance
outside the outer translucent zone
constitutes the marginal increment.
Therefore, if the outer opaque and
translucent zones are formed annu-
ally, the marginal increment should
decline only once during the year.
Verification that trends shown by the
marginal increments follow a pattern
consistent with annual growth is an
important method for establishing
that the alternating translucent and
opaque zones each correspond to an-
nuli and are thus appropriate for use
in ageing (Brothers 1983).
Many workers have presented data
which showed that an annual trend
was followed either by the marginal
increment, when data for all otoliths
in each sample were pooled, or by
the overall incidence of otoliths pos-
sessing either translucent or opaque
zones on their outer edge. When
these have shown a consistent annual
trend, it has often been concluded
that all translucent zones correspond
to annuli (e.g., Nel et al. 1985, Reis
1986, Rincon and Lobon-Cervia 1989,
Crozier 1990, Hayse 1990). However,
trends shown in such pooled data will
be strongly influenced by those of the
dominant groups, vis a vis the num-
ber of translucent zones, and may
276
277
Fishery Bulletin 90(2). 1992
not be representative of all groups. Furthermore,
trends shown by the marginal increment in some long-
lived fish become clear only after the otoliths have
either been sectioned or broken and burnt (Campana
1984, Collins et al. 1988). This accounts for estimates
of age sometimes being lower when whole otoliths have
been used than when either sectioned or broken and
burnt otoHths were employed (Beamish 1979a, Cam-
pana 1984, Collins et al. 1988).
Although the Platycephalidae occurs along the coasts
and within estuaries throughout the Indo-west Pacific
region, the majority of the 41 species of flathead found
in Australia are restricted to its southern waters (Sri-
ramachandra-Murty 1975, Paxton and Hanley 1989).
Despite wide distribution and, in some cases, the com-
mercial and recreational importance of the Platycepha-
lidae, estimates of the age and growth of represen-
tatives of this family are limited to those obtained for
Platycephalus bassensis, P. castelnaui, and P. specu-
lator by Brown (1977) and for P. richardsoni by Cole-
fax (1934), Fairbridge (1951), and Montgomery (1985),
the populations of which were all located in south-
eastern Australia. The most abundant species of flat-
head on the temperate southern coast of Western
Australia is P. speculator, a species which has been
shown to breed within Wilson Inlet, the largest estuary
of this region (Hyndes et al. In press).
Previous attempts to age platycephalids have used
whole sagittal otoliths (Colefax 1934, Fairbridge 1951,
Brown 1977, Montgomery 1985). However, a prelim-
inary investigation of the translucent zones in the sagit-
tal otoliths of P. speculator from southwestern Aus-
tralia showed that the outer opaque and translucent
zones on the otoliths of larger fish often became clear
only when the otoliths had been sectioned.
The present study was undertaken to determine the
age structure and growth of P. speculator in Wilson
Inlet, where this species is abundant and contributes
to the local commercial and recreational estuarine
fisheries (Lenanton and Potter 1987). Emphasis has
been placed on elucidating the degree to which section-
ing the otoliths influences marginal increment trends,
age estimates, and growth equations. In addition,
marginal increment data were pooled for both whole
and sectioned otoliths to examine whether the resul-
tant overall annual marginal increment trends were
strongly influenced by that of a group(s) of otoliths with
a particular number(s) of translucent zones.
Materials and methods
Sampling locality and regime
Wilson Inlet (117°25'E and 34°50'S) has a narrow en-
trance channel which opens into a wide basin (48 km^)
supplied by two main tributary rivers. Water depth in
the basin is generally less than 2m. Platycephalus
speculator was collected monthly from within the basin
of Wilson Inlet between September 1987 and April
1989 using beach seines (mesh size in pocket 9.5 mm)
during the day and gillnets (six stretched-mesh sizes,
38- 102 mm), otter trawls (mesh size in pocket 25 mm)
and plankton trawls at night (mesh size 1 mm).
Bottom water temperatures near the entrance chan-
nel of Wilson Inlet and 12 km further up the estuary
near the top end of the basin were recorded at the time
of sampling.
Age determination
Each fish was measured (total length) and weighed to
the nearest 1mm and O.lg, respectively. Sex was
recorded when the gonad could be identified as either
ovary or testis, which was usually possible in fish
> 100 mm in length. Both of the sagittal otoliths of 1305
juvenile and adult fish were cleaned, dried, and stored
in gelatin capsules.
Whole otoliths were placed in methyl salicylate solu-
tion and examined microscopically under reflected light
against a black background. The marginal increment,
i.e., the distance between the outer edge of the outer-
most translucent zone and the periphery, was mea-
sured on one of the otoliths of each fish and expressed
either as (1) a proportion of the distance between the
focus and the outer edge of the translucent zone when
only one translucent zone was present, or (2) as a
proportion of the distance between the outer edges
of the two outermost translucent zones when two or
more translucent zones were present. Measurements
were always made along the same axis, to the nearest
0.05 mm (Fig. 1). The number of translucent zones on
each otolith was recorded.
These otoliths were later mounted and embedded
in black epoxy resin (Bedford 1983, Augustine and
Kenchington 1987) and cut into 1.5-2 mm transverse
sections using the diamond saw described by Augustine
and Kenchington (1987). Sections were mounted on
glass slides and their surfaces ground on sequentially
finer grades (400-1200) of carborundum paper. Sec-
tions were then coated with clear nail polish and
examined microscopically under reflected light. Mea-
surements of the marginal increment and counts of the
number of translucent zones in these sectioned otoliths
were carried out in precisely the same manner as
described above for whole otoliths.
Mean marginal increment values were plotted sep-
arately for both whole and sectioned otoliths with
1-4 and > 5 translucent zones to ascertain if they follow
a consistent annual trend and thus permit the trans-
lucent zones to be considered as annuli. Width and
278
Fishery Bulletin 90(2). 1992
Transverse
Section
Figure 1
Schematic diagram showing the translucent and opaque zones
in a whole and sectioned sagittal otolith of Platycephalus
speculator. Broken line represents the axis along which
measurements were taken.
thickness of the whole otoliths of 123 fish, covering the
full range of sizes, were measured to the nearest 0.01
mm to examine the relationship between otolith width
and thickness. The number of translucent zones in 140
otoliths, of which up to 20 otoliths came from each age-
class (estimated from sectioned otoliths), were counted
in whole and sectioned otoliths by a second 'reader' who
had no previous experience in examining otoliths of this
species. The reproducibility of age estimates for each
method was determined by using the coefficient of
variation tSokal and Rohlf 1981, Chang 1982).
Von Bertalanffy growth curves were fitted to
individual lengths of males and females at the esti-
mated age-at-capture by a nonlinear technique (Gal-
lucci and Quinn 1979) using a nonlinear subroutine
in SPSS (SPSS 1988) and assuming a 'birth date' of
1 January. This date corresponds approximately to
the midpoint of the period when, on the basis of
gonadosomatic indices and trends shown by oocyte
development, P. speculator exhibited peak spawning
activity in Wilson Inlet (Hyndes et al. In press). The
von Bertalanffy equation is Lt = L^ [l-e^^('^'"'],
where Lt is the length at age t (yr), L„ is the mean
asymptotic length predicted by the equation, K is the
growth coefficient, and to is the hypothetical age at
which fish would have zero length if growth followed
that predicted by the equation. Comparisons have
been made between the age estimates and von Ber-
talanffy growth curves, calculated from data obtained
using whole and sectioned otoliths and assuming that,
in both cases, the translucent zones correspond to
annuli.
Results
Marginal increments
Annual trends in the mean marginal increments for
whole and sectioned otoliths with one translucent zone
were similar (Fig. 2). However, the sharp decline which
occurred in the marginal increment after the winter
(June- August) of 1988 was detected earlier in sectioned
otoliths (October) than in whole otoliths (December).
Although the data for 1987 were not as extensive, they
still exhibited a similar marked decrease at the same
time of year. In both years, the marginal increment on
both whole and sectioned otoliths subsequently rose
consistently through the summer, before leveling off
in the late autumn and winter (Fig. 2).
Annual trends in mean marginal increments of sec-
tioned otoliths with two, three, and four translucent
zones parallel those in sectioned otoliths with one
translucent zone, with marginal increments falling
sharply in the spring (October) of both 1987 and 1988
(Fig. 2). Although the marginal increment on whole
otoliths with two, three, and four translucent zones
also declined in spring, the decrease was far less pro-
nounced and the trends less consistent.
Since the number of otoliths with five or more
translucent zones was small, values for the marginal
increments on all such otoliths were pooled. Although
seasonal trends shown by the marginal increment in
sectioned otoliths with >5 translucent zones were
slightly less consistent than in those with 1-4 such
zones, they still followed a similar annual trend (Fig.
2). Furthermore, the translucent zones were still clearly
visible and had the same appearance as those in otoliths
with 1-4 translucent zones. No clear annual trend could
be seen in the marginal increments of whole otoliths
displaying >5 translucent zones (Fig. 2).
The above trends in marginal increments of sectioned
otoliths (with a sharp decline only occurring at one time
of the year, i.e., in the spring) show that the first four
translucent zones on otoliths of P. speculator are laid
down annually. Since the same trends were exhibited
in pooled data for the fifth and subsequent translucent
zones, these zones were presumably also, at least in
most of these cases, laid down annually. We thus con-
sider the translucent zone on sectioned otoliths as an
annulus which can be used for ageing P. speculator
from Wilson Inlet. The data also show that the outer
opaque zone starts to form when water temperatures
are rising from their winter (July) minima of about
11°C towards their summer (December-February)
maxima of ~22°C (c.f. Figs. 2,3).
The annual trend shown by the mean marginal in-
crement based on all sectioned otoliths, irrespective
of the number of translucent zones, was essentially
Hyndes et al.: Age and growth estimates of Ptatycephalus speculator
279
Figure 2
Mean marginal increments ± SE for whole
and sectioned sagittal otoliths of Platyce-
phalus speculator. Note that the marginal
increment is given as a relative value, i.e.,
as a percentage of the distance between the
focus and the outer translucent zone when
only one zone was present, or as a percent-
age of the distance between the outer edges
of the two outermost translucent zones when
two or more such zones were present. In this
Figure and in Figs. 3 and 6, the black bars
on the X-axis refer to winter (June-Aug.)
and summer (Dec. -Feb.) and the open bars
to spring (Sept. -Nov.) and autumn
(March-May).
the same as that of otoliths with
1-4 translucent zones (Fig. 2).
Mean marginal increments based
on all whole otoliths followed
similar annual trends, but they
were not as pronounced or con-
sistent, and the variation about
the means was greater (Fig. 2).
Length-frequency data
Few fish < 180 mm in length were
caught (Fig. 4), reflecting the
fact that smaller individuals of
this benthic species did not tend
to be collected by seine and gill-
nets.
The three fish caught in the
middle of spring (October) of
1987, which had otoliths with a
single translucent zone bounded
by a very narrow opaque zone,
measured 91-106 mm in length
(Fig. 4). This group is assumed to
represent the O-i- age-class, i.e.,
the result of spawning which
peaked in early January 1987
and can therefore be referred to
as the 1987 year-class. The larger
fish, which produced modal length-classes at 325-349
mm and 400-424 mm in October 1987 (Fig. 4), had
otoliths with a narrow opaque zone bounding two and
three translucent zones, respectively. The groups with
two and three translucent zones therefore represent
the 1 + and 2+ age-classes, or the 1986 and 1985 year-
classes, respectively. Fish viath otoliths exhibiting one,
two, and three translucent zones in December 1987 and
February 1988 are designated as representing the
Whole otoliths
Sectioned otoliths
1 translucent
zone
0.4
0
1.2 ■
0.8 ■
0.4
0"-
1.2
2 translucent
zones
3 translucent
zones
4 translucent
zones
5-12 translucent
zones
Pooled
data
A O D
1987
F A J A O D F A
1988 1989
AODFAJAODFA
1987 1988 1989
1987, 1986, and 1985 year-classes, which is consistent
with their length distributions (Fig. 4). The marked dif-
ference between the lengths of the 1987 year-class in
October and December 1987 suggests that this year-
class underwent remarkable growth between these
months. However, the three fish caught in October
were taken by beach seine in the shallows and are thus
presumed to represent the lower end of the length
range of this year-class, whereas those fish of the
280
Fishery Bulletin 90(2). 1992
25 r
y 20
g 15
10
A O D
1987
F A J
1988
Lower basin
Upper basin
O D F A
1989
Figure 3
Bottom-water temperatures recorded at sites at
lower and upper ends of the basin of Wilson Inlet,
Western Australia, August 1987-April 1989.
corresponding cohort caught in December were taken
in gillnets which, because of the mesh sizes of these
nets, would have taken only the larger members of that
year-class.
Larval P. speculator, ranging up to a length of 13
mm, were collected in plankton trawls in December
1987, February 1988, and February and April 1989
(Fig. 4). Otoliths of juveniles caught in April 1988
and 1989, and which from their lengths (23-89 mm)
clearly corresponded to the 0 + age-class, did not have
a translucent zone. The group of fish representing
the 1988 year-class had reached 198-231 mm by
December 1988, and 258-323 mm by April 1989.
These length ranges are similar to those attained
by the previous (1987) year-class in December 1987 and
April 1988 (Fig. 4). The modal length of the 1 -t- age-
class in October 1988 ( = 1987 year-class) was identical
to that of the 1+ age-class in October 1987 (= 1986
year-class). Six year-classes were usually found in
samples from each month, and as many as eleven year-
classes were present in February 1988 (older year-
classes are not shown in Fig. 4). Maximum lengths for
each sex were 696mm for a lO-i- female and 545 mm
for a 12-^ male.
Estimated ages and growth curves using
whole and sectioned otoliths
All of the 197 otoliths which showed one translucent
zone (= annulus) in sectioned otoliths also displayed
a single zone prior to sectioning (Fig. 5). However, 44%
of otoliths with two or more translucent zones after
sectioning produced underestimates using whole oto-
liths. Between 23% and 57% of the otoliths with 2-4
o
B
1987
Sepember
October
200 300 400 .SOO 600
Length (mm)
Figure 4
Length-frequency histograms for the 1985-89 year-
classes of Platycephalus speculator.
annuli each showed one less translucent zone prior to
sectioning and thus underestimated ages by 1 year.
Discrepancies between the number of translucent zones
in sectioned and whole otoliths were even more marked
in otoliths with 5 or more translucent zones. Indeed,
the numbers of annuli were as many as 5 or 6 less in
whole otoliths than the 11 or 12 annuli observed in sec-
tioned otoliths (Fig. 5). The proportion of underesti-
mates using whole otoliths was greatest in spring for
those otoliths which showed 2 or 3 translucent zones
after sectioning (Fig. 6).
Hyndes et al.: Age and growth estimates of Platycephalus speculator
281
Number of annuli using sectioned otoliths (O )
234 5678 9 10 11
12
c aw
u 2 ^
t/i .5 o
-3 -
^ °
-6
(197)
(309)
(370)
(275)
(70)
(20)
(13)
(5)
(4)
(6)
(1)
(2)
o
o
o
o
o
o
o
o
o
o
o
o
100
77.4
63.3
43.0
11.5
5.0
7.7
25.0
16.6
-
•
•
•
•
•
•
•
•
•
22.6
36.7
56.6
52.8
35.0
23.0
40.0
25.0
16.7
50.0
■
•
•
•
0.4
•
35.7
•
50.0
•
30.8
•
40.0
•
•
50.0
• -
■
•
•
•
10.0
•
30.8
•
20.0
50.0
•
16.7
■
■
•
•
7.7
•
•
•
■
■
•
50.0
100
• -
1
_l
_l
-.1
_l
•
_L_
60
r
^ 2 annuli
/ \ A A
40
-
20
" 23/-
^^^ \ju^2\/\
0
!>y» — ■ ■ ■'•■ ^2 \
100
r
3 annuli
83 20
s
80
■
o
60
-
/ \
-§
c
40
.
/ V^
3
J
■ 12 \
tiO
20
/
\ 52 13 14
0
/a 3
3oV-'N^o/Noi 6/"
100
r )
44
7*~~-^3 4 annuli j»3
o S
80
- /
x"^ / \
« —
y
o «
60
- i
\ /
•= •^-,
\ /
? o
40
-
63V "/
o
X. /
c
?0
-
^ \ /
<D
Ui
0
L
■ 1 1
Cu
100
\ 7
' ■ ■ 1^^ 5-12 annuli y- 6
80
■ /
22\ X
60
- /
\ >/
ii,
\ y**^
40
"
Vf
20
-
0
L
A
0 D F A J
Month
Figure 5
Number of otoliths with 1-12 annuli base(i
on sectioned sagittal otoliths oi Platycephxdus
speculator, and underestimates of the num-
ber of annuli observed on the same whole
sagittal otoliths. Numbers in parentheses in-
dicate the number of fish of different ages
based on sectioned otoliths, while numbers
above the closed circles indicate the percent-
age of underestimates using whole otoliths.
The coefficient of variation for
replicate age estimates between
readers was far less for sectioned
otoliths (1.2%) than for the same
otoliths prior to sectioning
(8.7%). While the estimated age
varied by only 1 year for each of
the six sectioned otoliths for
which there was disagreement,
the estimated ages varied by as
much as 3 years for the 53 whole
otoliths for which there were
discrepancies.
The relationship between width and thickness of
sagittal otoliths of P. speculator is curvilinear, demon-
strating that width does not increase proportionately
with thickness (Fig. 7). The relationship between otolith
width (W) and otolith thickness (T) is described by the
following polynomial equation:
W = -0.283 H- 4.635T - 1.175T2 (r'^ 0.82, n 123).
The von Bertalanffy growth parameters for both
male and female flathead were initially determined
using individual lengths at estimated age (Table 1).
Examination of the length-at-age plots showed that the
curve for both sexes fell below the majority of the
points for fish >5 years old, i.e., the asymptote was
too pronounced to accommodate lengths of the older
fish. Individual lengths of the fish were grouped into
intervals of 0.1 years and the curves determined again
by weighting the data by the inverse of the sample size
for each age interval (Beckman et al. 1990). This pro-
cedure resulted in a better fit of the curve (Fig. 8).
Although the values for to for males and females were
shifted slightly away from zero (namely from -0.134
to -0.332 and from -0.056 to -0.423, respectively),
differences in the lengths of males and females at
Figure 6
Seasonal incidence of underestimates by one or more an-
nuli when using whole vs. sectioned sagittal otoliths of
Platycephalus speculator.
282
Fishery Bulletin 90(2), 1992
5
.
B 4
£
■B 3
■ X- -'j^ — ^
0
1 .... 1
0 1 2
Otolith thickness (mm)
Relati
Figure 7
onship between width and thickness of the
sagitt
al otoliths of Platycephalus speculator.
ages 1, 2, 3, and 4 were never altered by
more than 22 mm and the change was
generally less than 12 mm. The coefficient
of determination (r^) for the von Berta-
lanffy curve was 0.85 for both sexes using
individual lengths-at-age, and 0.93 for males
and 0.89 for females using the weighted procedure.
The respective parameters L^, K, and to in the von
Bertalanffy growth equations for males determined
using weighted data obtained from sectioned otoliths
overlapped those when the same approach was em-
ployed for whole otoliths, and the same was true for
K with females (Table 1, Fig. 8). The von Bertalanffy
growth-equation parameters using weighted data from
sectioned otoliths show that female P. speculator grow
towards a larger asymptotic size (L^) than males
(Table 1, Fig. 8). Individual lengths of P. speculator in
December-February at the end of their first and sec-
ond years of life were 190-310mm and 210-370mm,
respectively, for males, and 210-300 mm and 250-400
mm, respectively, for females (Fig. 8).
Discussion
Marginal increments
Marginal increments of the otoliths of P. speculator
with two or more translucent zones exhibited con-
spicuous trends only after the otoliths were sectioned.
This is largely due to the fact that sectioning of otoliths
results in more accurate measurement of their periph-
eral and/or penultimate opaque zones, because one or
both of the two outermost translucent zones have
become more clearly delineated. This is similar to the
situation with starry flounder Platychthys stellatus, in
which annual trends in the marginal increments of
Table I
Parameter estimates (95% confidence limits) for the von Bertalanffy growth |
model fitted to 630 male and 711 female Platycephalus speculator
, deter-
mined from whole and sectioned otoliths using
individual lengths
-at-age
and weighted leng^ths of each age-group in each month
Individual lengths
Weighted lengths
Param- -
eter
Sex
Est.
Lower
Upper
Est.
Lower
Upper
Sectioned otoliths
Male
L„
429.2
419.9
438.5
477.4
468.6
486.2
K
0.573
0.525
0.621
0.408
0.380
0.437
to
-0.134
-0.201
-0.067
-0.332
-0.411
-0.253
Female
L„
481.8
469.4
494.2
601.0
588.2
619.8
K
0.593
0.547
0.639
0.309
0.283
0.335
t„
-0.056
-0.109
-0.003
-0.423
-0.515
-0.331
Whole otoliths
Male
L„
426.9
417.3
436.4
484.9
473.5
496.3
K
0.700
0.641
0.760
0.466
0.429
0.502
to
-0.068
-0.125
-0.012
-0.244
-0.292
-0.156
Female
L„
457.8
445.6
470.0
659.6
626.1
693.2
K
0.767
0.694
0.840
0.264
0.231
0.296
t„
-0.088
-0.148
-0.028
-0.698
-0.837
-0.560
600
500
Males
E
E
00
c
00
c
700
Females
600
500
-
■■•■^
^^^^"^
400
-
.■:i,jj|?pi: ■■
300
-
0^^ ■
— — Whole otoliths
200
//
f ■
100/
i
1 1 1 1 1
1 1 t 1 1 1 1
-1 0 1 2 3 4 5 6 7 8 9 10 11 12
Age (years)
Figure 8
Von Bertalanffy growth curves fitted to length-at-age data,
weighted by the inverse of the sample size of each 0.1 year
age-group, using whole and sectioned otoliths of male and
female Platycephahts stpeculator.
Hyndes et al.: Age and growth estimates of Platycephalus speculator
283
otoliths with four or more annuli were observed only
after otoliths had been broken and burnt (Campana
1984). Likewise, in the case of king mackerel Scom-
beromorus cavalla, the percentage of otoliths with an
opaque zone on their edge exhibited an annual trend
only after the otoliths were sectioned (Collins et al.
1988).
The consistency in annual trends of marginal incre-
ments among sectioned otoliths of P. speculator,
despite differing numbers of translucent zones, ac-
counts for the clear annual trend in marginal incre-
ments when data for all otoliths in each of the monthly
samples were pooled. The contrast between the con-
spicuous trend shown in pooled data for whole otoliths,
and the relatively poor trend exhibited in whole otoliths
with two or more translucent zones, shows how trends
can be unduly influenced by those of a relatively large
sample size of otoliths exhibiting a particularly strong
annual trend, such as was present with those otoliths
having one translucent zone. For validation of the use
of translucent zones as annuli, it is thus important to
establish that the trends shown by the marginal in-
crements on otoliths with differing numbers of translu-
cent zones each follow a consistent annual trend
(Johnson 1983, Maceina et al. 1987, Potter et al. 1988,
Beckman et al. 1989).
Age and growth estimates
Our results demonstrate that, while age estimates be-
tween sectioned and whole otoliths corresponded when
one translucent zone was present, ages of older fish
were underestimated by 2-4 years or as much as 5 or
6 years using whole otoliths. Increased resolution of
the growth zones after sectioning is reflected in the far
lower variability between age estimates made by two
independent readers using sectioned otoliths.
Although otoliths with two or three translucent zones
frequently yielded counts of one less zone prior to sec-
tioning, many underestimates occurred with the oto-
liths taken from fish between mid-spring and early
summer (October-December). In other words, they
were collected during the period when sectioning en-
abled the new opaque zone to be detected approximate-
ly 2 months earlier than was possible with whole
otoliths.
Our inability to detect all of the translucent zones in
whole otoliths can in part be attributed to the growth
pattern of the otolith. Whereas the first translucent
zone can be easily detected in whole otoliths, the dis-
proportionate increase in otolith thickness relative to
its width results in the translucent zones becoming
increasingly more closely apposed and therefore diffi-
cult to distinguish from one another. This parallels the
situation recorded by Beamish (1979a, b) for Pacific
hake Merluccius productus, and for several species of
rockfish (Sebastes), and also by Campana (1984) for
starry floimder Platychthys stellatus.
Despite the fact that a large proportion of ages were
underestimated using whole otoliths, the von Berta-
lanffy growth curves derived from data using whole
otoliths, particularly of males, did not differ markedly
from those obtained using sectioned otoliths. This can
be attributed to the fact that approximately 74 and 65%
of growth for males and females, respectively, occurred
in the first 3 years of life when underestimates of age
were limited.
Implications for management
The vast majority of male P. speculator reach sexual
maturity at the end of their first year of life (Hyndes
et al. In press).Since males have attained only 190-310
mm by this time (Fig. 8), they will only occasionally
have reached 300 mm, the minimum legal length for
capture of this species. However, the majority of
females do not first attain sexual maturity until they
are 2 years old (Hyndes et al. In press), by which time
they have reached 250-400 mm. Thus, the females of
P. speculator can be exploited before they have had the
opportunity to spawn.
In summary, this study has demonstrated that, in the
case of the flathead P. speculator, it is crucial to sec-
tion its otoliths in order to obtain an accurate estimate
of age. Sectioning reduces the problems of distinguish-
ing between peripheral translucent zones which, due
to the growth pattern of the otolith, become increas-
ingly more closely apposed with increasing size. While
the results presented in this paper refer only to P.
speculator, they parallel in some respects those ob-
tained for Platychthys stellatus and Scomberomorus
cavalla (Campana 1984, Collins et al. 1988). Such age
underestimates have obvious implications in estimating
mortalities for use in fisheries management. Our re-
sults also demonstrate the importance of plotting
marginal increments for otoliths with different num-
bers of translucent zones, to establish that such zones
are laid down annually on the otoliths of fish repre-
senting each presumed age-group. Since the females
of P. speculator are being caught before they have
spawned for the first time, there is a case for increas-
ing the minimum legal size for capture.
Acknowledgments
We thank F. Baronie, D. Gaughan, P. Geijsel, P. Hiun-
phries, L. Laurenson, and F. Neira for their assis-
tance with sampling. The sectioning saw was kindly
provided by S. Blaber of the CSIRO Marine Labora-
284
Fishery Bulletin 90(2). 1992
tones at Cleveland, Queensland. Helpful discussions
were provided by N. Hall, L. Laurenson and L. Pen.
L. Pen also estimated the number of translucent zones
in otoliths for the precision estimates. Gratitude is
expressed to J. Casselman, R. Fletcher, M. Moran,
N. Hall, two anonymous referees, and the scientific
editor for constructive criticism of our work and this
paper. Financial support was provided by the Western
Australian Department of Fisheries.
Citations
Augustine, O., and T.J. Kenchington
1987 A low cost saw for sectioning otoliths. J. Cons. Cons.
Int. Explor. Mer 43:296-298.
Beamish, R.J.
1979a Differences in the age of Pacific hake {Merluccius pro-
ducts) using whole and sections of otoliths. J. Fish. Res.
Board Can. 36:141-151.
1979b New information on the longevity of the Pacific hake
{Sebastes alutus). J. Fish. Res. Board Can. 36:1395-1400.
Beamish, R.J., and G.A. McFarlane
1983 The forgotten requirement for age validation in fisheries
biology. Trans. Am. Fish. Soc. 112:735-743.
Beckman, D.W., C.A. Wilson, and A.L. Stanley
1989 Age and growth of red drum, Sciaenops oceUatus. from
offshore waters of the northern Gulf of Mexico. Fish. Bull.,
U.S. 87:17-27.
Beckman, D.W., A.L. Stanley, J.H. Render, and C.A. Wilson
1990 Age and growth of black drum in Louisiana waters of
the Gulf of Mexico. Trans. Am. Fish. Soc. 119:537-544
Bedford, B.C.
1983 A method for preparing sections of large numbers of
otoliths embedded in black polyester resin. J. Cons. Cons. Int.
Explor. Mer 41:4-12.
Brothers, E.B.
1983 Summary of round table discussions on age valida-
tion. In Prince, E.D., and L.M. Pulos (eds.), Proceedings, In-
ternational workshop on age determination of oceanic pelagic
fishes: Tunas, billfishes, and sharks, p. 35-44. NOAA Tech
Rep. NMFS 8.
Brown, LW.
1977 Ecology of three sympatric flatheads (Platyeephalidae)
in Port Phillip Bay, Victoria. Ph.D. thesis, Monash Univ., Vic-
toria, 304 p.
Campana, S.E.
1984 Comparison of age-determination methods for the starry
flounder. Trans. Am. Fish. Soc. 113:365-369.
Casselman, J.M.
1987 Determination of age and growth, /re Weatherly, A.H.,
and H.S. Gill (eds.). The biology of fish growth, p. 209-242.
Academic Press, London.
Chang, W.Y.B.
1982 A statistical method for evaluating the reproducibilty of
age determination. Can. J. Fish. Aquat. Sci. 39:1208-1210
Colefax, A.N.
1934 A preliminary investigation of the natural history of the
tiger flathead (Neoplatycephalm macrodon) on the south-
eastern Australian coast. I. Distribution and supply; length
statistics. Proc. Linn. Soc. N.S.W. 49:71-91.
Collins, M.R., D.J. Schmidt, C.W. Waltz, and J.L. Pickney
1988 Age and growth of the king mackerel. Scomberomorus
cavalla, from the Atlantic coast of the United States. Fish.
Bull., U.S. 87:49-61.
Crozier, W.W.
1990 Age and growth of the angler-fish (Lophius piscatorius
L.) in the North Irish Sea. Fish. Res. (Amst.) 7:267-278.
Fairbridge, W.S.
1951 The New South Wales tiger flathead, Neoplatycephaltis
macrodon (Ogilby). Aust.J. Mar. Freshwater Res. 2:117-118.
Gallucci, V.F.. and T.J. Quinn
1979 Reparameterizing, fitting and testing a simple growth
model. Trans. Am. Fish. Soc. 10:14-25.
Hayse. J.W.
1990 Feeding habits, age, growth and reproduction of Atlan-
tic spadefish Ometodiptems faber (Pisces: Ephippidae) in South
Carolina. Fish. Bull., U.S. 88:76-83.
Hyndes, G.A., F.J. Nelra, and LC. Potter
In press Reproductive biology and early life history of the
marine teleost Platycephalus speculator Klunzinger (Platy-
eephalidae) in a temperate Australian estuary. J. Fish Biol.
Johnson, A.G.
1983 Age and growth of yellowtail snapper from South Florida.
Trans. Am. Fish. Soc. 112:173-177.
Lenanton, R.C.J., and LC. Potter
1987 Contribution of estuaries to commercial fisheries in
temperate Western Australia and the concept of estuarine
dependence. Estuaries 10:28-35.
Maceina, M.J., D.N. Hata, T.L. Linton, and A.M. Landry
1987 Age and growth analysis of spotted seatrout from Gal-
veston Bay, Texas. Trans. Am. Fish. Soc. 116:54-59.
Montgomery, S.S.
1985 Aspects of the biology of the tiger flathead P. richard-
soni and the associated fishery. Masters thesis, Univ. New
South Wales, Sydney, 228 p.
Nel, S.A., LC. Potter, and N.R. Loneragan
1985 The biology of the catfish Cnidoglanis macrocephalus
(Plotosidae) in an Australian Estuary. Estuarine Coastal Shelf
Sci. 21:895-909.
Paxton. J.R., and J.E. Hanley
1989 Platyeephalidae. In Walton, D.W., and R. Longmore
(eds.). Zoological catalogue of Australia, 7, Pisces, Petromyzon-
tidae to Carangidae, p. 465-472. Brown Prior Anderson, Bur-
wood, Victoria.
Potter, LC, A.J. Cheal, and N.R. Loneragan
1988 Protracted estuarine phase in the life cycle of the marine
pufferfish Torquigener pleurogramnia. Mar. Biol. (Berl.) 98:
317-329.
Reis, E.G.
1986 Age and growth of the estuarine catfish, Netuma barba
(Siluriformes, Ariidae), in the estuary of the Patos Lagoon
(Brasil). Fish. Bull., U.S. 84:679-686.
Rincon, P. A., and J. Lobon-Cervia
1989 Reproductive and growth strategies of the red roach,
Rutilus arcasii (Steindachner, 1866), in two contrasting tribu-
taries of the River Duero, Spain. J. Fish Biol. 34:687-705.
Sokal, R.R., and F.J. Rohlf
1981 Biometry, 2d ed. W.H. Freeman, NY. 859 p.
SPSS Inc.
1988 SPSS-X"' user's guide. SPSS Inc, Chicago, 1072 p.
Sriramachandra-Murty, V.
1975 On the fishes of the family Platyeephalidae of the seas
around India. J. Mar. Biol. Assoc. India 17:679-694.
Abstract.- Morphological changes
during metamorphosis of Dover sole
Microstomics padficus are described
from 2220 larvae and juveniles. Un-
like most flounders, initiation of eye
migration is uncoupled from meta-
morphosis and from the habitat
change from planktonic to benthic.
Dover sole larvae are optically asjrm-
metrical during most of their plank-
tonic life. Major features associated
with metamorphosis are reduction
in body depth with associated reduc-
tions in lengths of neural and hemal
spines, increase in relative eye diam-
eter, loss of canine-like teeth coin-
cidental with acquisition of incisor-
like teeth, resorption of posterior
process of coracoid, development of
body scales, change in body pigmen-
tation, and development of the gut
loop in the secondary body cavity.
From initiation to completion, meta-
morphosis appears to take about 9
months, during which time there is
little increase in body length.
Available evidence indicates that
most spawning off Oregon occurs in
spring, with April and May as peak
hatching months. Settlement from
the plankton occurs in winter, with
January to March as peak settlement
months. Duration of planktonic life
appears to be about 2 years, with
a minimum duration of about 18
months. Metamorphosing larvae
settle over a broad "landing" zone
(55-377 m), quantitatively distinct
from, but overlapping, the narrower
spring nursery zone (40-170m). As
yet, there is no evidence of delayed
metamorphosis. Metamorphosis is
protracted, seems to be seasonally-
triggered, and may involve a signif-
icant period during which larvae
switch between midwater and bot-
tom habitats.
Metamorphosis and an overview
of early-life-hiistory stages in Dover
sole Microstomus padficus*
Douglas F. Markle
Phillip IVI. Harris
Christopher L. Toole
Department of Fisheries and Wildlife, 104 Nash Hall
Oregon State University, Corvallis, Oregon 97331-3803
There is uncertainty about the length
of the pelagic life of Dover sole Micro-
stomus pacificus. Hagerman (1952)
noted that the "pelagic life is pro-
longed for several months and meta-
morphosis is delayed." Allen and
Mearns (1977) thought a 9-month
planktonic stage was "probably
not unusual." Pearcy et al. (1977a)
examined the early life history in
greater detail and concluded that
"Dover sole larvae are pelagic for at
least a year." Hayman and Tyler
(1980), although citing Pearcy et al.
(1977a), constructed a time-line in-
dicating a 9-month pelagic larval
stage.
There is also little agreement on
estimated body length at the end of
the first year of life. Pearcy et al.
(1977a) estimated growth to be 20-30
mm standard length (SL) during the
first year, but also concluded that
metamorphosis took place after
about 1 year at 30-50 mm SL (the
extra 10-30 mm of growth was not
explained). Hagerman (1952) and
Demory (1972), both limited by small
sample sizes, mention lengths of 66-
75 mm total length (TL) for nominal
1-year-old specimens.
Uncertainty about duration and
growth in the pelagic phase has im-
portant implications for age esti-
mates. Whether based on scales or
otolith sections, no researcher has
Manuscript accepted 18 March 1992.
Fishery Bulletin, U.S. 90:285-301 (1992).
'Technical paper 9837, Oregon State Univer-
sity Agricultural Experiment Station.
documented the age or size at which
the first nominal annulus forms
(Demory 1972, Chilton and Beamish
1982, Pikitch and Demory 1988,
Hunter et al. 1990).
From the large midwater trawl col-
lections made by W.G. Pearcy and
colleagues (OSU) from 1961 to 1982,
and juvenile bottom-trawl surveys
conducted off Oregon from 1988 to
1990, we describe metamorphosis
and other stages in the early life
history of Dover sole and address
questions relating to the duration and
timing of these stages.
Materials and methods
Midwater trawl collections
A total of 796 Dover sole larvae were
obtained from 425 midwater trawl
stations off Oregon. Details of sam-
pling methods are given in Pearcy
(1976, 1980) and Pearcy et al. (1977
a,b). Because the midwater trawls
were made for a variety of reasons,
there are constraints on interpreta-
tion of these data. The most impor-
tant constraints are seasonal, diel,
depth, and gear. Seasonal coverage
was best from June to September,
and poorest in May and October
(Table 1). There was a pronounced
diel bias. Relatively few samples
were collected between 0600 and
2000 hours (Table 2). Most samples
were collected at night between 2200
and 0500 hours. The range of collec-
285
286
Fishery Bulletin 90(2), 1992
Table I
Distribution of
midwater trawl stations by |
month, 1961-82
No. of
Percent of
Month
stations
total
January
94
3.8
February
127
5.1
March
157
6.4
April
174
7.0
May
85
3.4
June
306
12.4
July
397
16.1
August
237
9.6
September
479
19.4
October
74
3.0
November
165
6.7
December
173
7.0
Total
2468
99.9
Table 2
Distribution of midwater trawl stations by time of day, 1961
-82.
Hour
No. of
Hour
No. of
of set
stations
Percent
of set
stations
Percent
0100
192
7.8
1300
51
2.1
0200
158
6.4
1400
60
2.4
0300
177
7.2
1500
60
2.4
0400
184
7.4
1600
65
2.6
0500
154
6.2
1700
52
2.1
0600
82
3.3
1800
60
2.4
0700
68
2.8
1900
65
2.6
0800
61
2.5
2000
78
3.2
0900
61
2.5
2100
95
3.8
1000
69
2.8
2200
119
4.8
1100
45
1.8
2300
133
5.4
1200
55
2.2
2400
167
6.8
tion depths was 0-6000 m, but 81.8% of the samples
were from depths <500m. Eleven different gear types
were used: Tucker trawl, Cobb trawl, 0.9 m Isaacs-
Kidd midwater trawl (IKMT), 1.8m IKMT, 2.4m
IKMT, 3.0 m IKMT, 2.4 m rectangular midwater trawl
(RMT), 2.7m RMT, Im^ multiple plankton sampler,
65m2 midwater trawl, and lOOm^ midwater trawl.
Some gears were operated with and without opening-
closing devices (Pearcy 1980). Eighty-eight percent of
the collections were made with either a 1.8 m or 2.4 m
Isaacs-Kidd midwater trawl (IKMT). All specimens
were preserved in 10% formalin and transferred to
50% isopropanol.
Juvenile bottom-trawl collections
A bottom-trawl survey of juvenile Dover sole was con-
ducted bimonthly, January to November 1989, in three
areas off Oregon (Fig. 1). In March 1988 and 1990, a
more limited survey was conducted in the central
(Foulweather) area. Each area was 10 miles wide and
oriented to the coast such that the depth range of
50-400 m could be covered in the shortest distance.
Each area was subdivided into six strata bounded by
isobaths at 50, 80, 100, 120, 160, 220 and 400 m. Trawl
stations were randomly chosen such that a minimum
of three 5-minute trawls were attempted in each
stratum. When time permitted, additional stations
were added in strata with highest concentrations of
Dover sole (100-119 and 120-159m). All trawling was
conducted from the FV Olympic during daylight hours.
The gear was a commercial, 34.9mm mesh, two-seam
shrimp trawl with a 27.4m headrope, rigged with a
28.5m footrope and tickler chain. The posterior 3/4 of
the codend had a 6.4mm liner. The catch was sorted
on board, all fish species were counted and measured,
and all Dover sole <200mmSL were frozen or fixed
in 10% formalin and returned to the laboratory for mor-
phological analysis. All formalin-fixed specimens were
preserved in 50% isopropanol before measurement.
Morphological analysis
All measurements reported herein were made in the
laboratory on defrosted or formalin-fixed, isopropanol-
preserved specimens. We found no significant differ-
ences (P 0.93) between measurements of 89 defrosted
juvenile Dover sole (46. 9-71.0 mm SL) when remea-
sured over a year after fixation and preservation.
Measurements were taken on 2220 larvae, juve-
niles, and adults. Using the staging system developed
herein, the numbers examined in each stage were:
Stage 1, 811; Stage 2, 29; Stage 3, pelagic captures,
12; Stage 3, benthic captures, 409; Stage 4, pelagic
captures, 1; Stage 4, benthic captures, 461; and
Stage 5, 497. On all specimens returned to the labor-
atory, we measured TL, SL, body depth at anus
(BDIA), maximum body depth, snout to posterior
extent of intestine length (SINT), and body weight.
Length measurements were taken to the nearest
0.1mm using an ocular micrometer on specimens
<20mmSL and dial calipers on larger specimens. Body
weight was determined to the nearest O.OOlg for
Stages 1 and 2 and to the nearest O.lg for Stages 3,
4, and 5 (see staging description below). Weights were
taken from undamaged, pat-dried individuals. Weight
loss in isopropanol-preserved larvae was as great as
10% after 2 minutes of air exposure due to alcohol
evaporation. Although specimens were exposed for less
time before weighing, a 10% weighing error was
Markle et al.: Metamorphosis of Microstomus pacificus
287
125°
124° 123°
KWASH.
46°
—
JASTORIA.
46°
45°
\
N
y/:-
45°
\f\
1:
\ NEWPORT
44°
^
•XOOS BAY
44°
43°
—
1 "
43°
42°
—
\brookings
1 CALIF.
1 .t
42°
12
-5°
124° 12
3°
Three are
bottom-tra
sole Microt
Figure 1
as sampled during 1989
ivl survey for juvenile Dover
tomus pacificus. N = Ne-
tarts, F =
Foulweather, H = Heceta
sampling areas.
assumed for this study. Considering the change in
weight of three orders of magnitude between 10 and
SOmmSL, the weighing error was considered accept-
able for this study.
A smaller subset of 201 specimens was examined to
describe metamorphosis in greater detail, and all were
deposited in the Oregon State University Fish Collec-
tion (OS). These specimens were either cleared and
differentially stained with alizarin red S and alcian
blue (Potthoff 1984), radiographed, or both. This sub-
set included only postflexion Stage- 1, most Stage-2,
and representatives of Stages 3-5 larvae. In addition
to routine measurements listed above, we measured
right eye diameter, interorbital width, right upper jaw
length, length of gastrointestinal tract as measured
from anus to most posterior part of intestinal loop,
length of first caudal neural spine, length of dorsal fin
pterygiophore anterior to first caudal neural spine,
length of dorsal fin pterygiophore posterior to first
caudal neural spine, length of first hemal spine, length
of anal fin pterygiophore anterior to first caudal hemal
spine, and length of anal fin pterygiophore posterior
to caudal hemal spine. Counts of vertebrae and rays
of dorsal, anal, caudal, pectoral, and pelvic fins also
were made.
A staging system describing Dover sole ontogeny
was developed following the suggestions of Youson
(1988). Our terminology deviates from Balon (1979,
1984) and Youson (1988) in our use of five numbered
stages for early development, rather than numbered
stages for metamorphosis only. Dover sole have a pro-
tracted metamorphosis, and our stages can be related,
generally, to flatfish metamorphosis. We suggest term-
inology for each stage that incorporates traditional
concepts of larval and juvenile periods as well as the
metamorphic phase of the larval period. Metamorphosis
occurs in Stages 2-4.
We were especially concerned with describing the
beginning of metamorphosis, the initiation event, and
the completion of metamorphosis, the climax event
(Youson 1988). The initiation event was described
based on six characters that reach the adult state dur-
ing the plankton-to-benthos transition (see Results
below). Another character, body scale formation, could
be documented only in cleared and stained specimens
and was concordant with completion of the six ini-
tiation-event characters. Development of the intestinal
loop in the secondary body cavity, quantified by SINT,
is the last character to change in Dover sole meta-
morphosis. The climax event was described based on
the rate of change of the ratio of natural logarithms
of two measurements (SINT and SL). Both initiation
and climax events are further corroborated by body
shape changes.
We use the concept of competency, as developed in
the marine invertebrate developmental literature, as
part of our definition of stages. The term regrettably
has become a synonym for metamorphosis, as in the
phrase "competent to metamorphose" (Pechenik 1986).
Doyle (1975), using the term "delay" stage, noted that
the onset of competency included both developmental
criteria (strict metamorphosis as used herein) and a
behavioral criterion, the ability to settle. In some in-
vertebrates, attachment to a substrate is a prerequisite
to metamorphosis; thus settlement must occur prior to
metamorphosis. In fishes there is not necessarily a
connection between metamorphosis (Youson 1988) and
competence. However, Cowen (1991) applied the terms
to fish and kept the marine invertebrate connection
intact. Competency has been defined more narrowly
288
Fishery Bulletin 90(2). 1992
■ ■ ■ ■ '
240
-
B
200
1
1
160
-
1
120
-
J
80
-
i
b<
/
5 ~\
■
■
0
-
1 1 1 1-
i I. .. . ■ ■ 1
Depth of Capture (m)
Figure 2
Relationship between body depth at anus and standard length
in Dover sole Microstomus pacificus : (A) scatterplot of data
points, Stages 1-5; (B) polygons circumscribing areas bounded
by specimens in Stages 1-5.
as the ability to settle (Jackson and Strathmann 1981),
a conceptual improvement that removes predefined
connections to metamorphosis. We identify precompe-
tent, competent, and postcompetent stages during
metamorphosis of Dover sole.
Time-series analysis
Two approaches were used to construct a time-line of
early development: modal progression analysis (MPA,
Bhattacharya 1967) and an analysis of seasonality of
stages. MPA was facilitated using the computer pro-
Table 3
Characters used to quantify metamorphosis in Dover sole
Microstomus pacificus.
Character
Character state
Teeth
0
1
Canines
Canines, incisors developing
2
Incisors
Eye position
0
1
Left side of head or dorsal ridge
Right side of head, adult position
Position of
dorsal fin
0
1
First ray posterior to left eye
First ray equal with posterior margin
or anterior to left eye
Posterior process
of coracoid
0
1
2
Straight, angled posteriorly
Resorption beginning, tip curled into
hook
Resorption complete, process absent
Pectoral fin
shape
0
1
2
Round, paddlelike shape, < adult
shape complement of rays, no
radials formed
Intermediate shape, adult comple-
ment of rays, cartilaginous radials
Adult morphology
Pigmentation
0
1
Planktonic coloration
Benthic coloration
gram ELEFAN (Pauly 1987), but its utility was limited
by sample sizes. Analysis of seasonality of stages was
corroborated partly by monitoring growth of a single
metamorphosing individual held in the laboratory at
13°C. The Stage-3 specimen was 57.7 mm SL when cap-
tured on 20 March 1989. It was measured regularly and
progressed completely through Stage 4 to an early
Stage 5, when it was sacrificed on 15 June 1989.
Results
Morphology and development
Stage 1 jpremetamorphic larvae), 6. 1-58.5 mm SL
For convenience and because of our emphasis on meta-
morphosis, all premetamorphic planktonic specimens
are referred to as Stage 1 . However, the premetamor-
phic phase of the larval period could be usefully divided
into two intervals, the first approximating Stage I of
Pearcy et al. (1977a). A transition from the first inter-
val to the second occurs around 10-15mmSL, during
which eye migration begins, body depth increases (Fig.
2), the first dorsal and anal fin rays form, and caudal
fin flexion begins. During the second interval, speci-
mens acquire the adult numbers of vertebrae, and
dorsal, anal, caudal, and pelvic fin rays; the stomach
and intestine coil; 3-4 pyloric cecae develop; and a
pigmentation pattern of dashes develops into a solid
Markle et al.: Metamorphosis of Miaostomus pactficus
289
Figure 3
Jaw dentition during metamorphosis in Dover sole
Microstomus pacificus, left lateral views: (A) OS12578,
Stage 1 with canine teeth; (B) OS11377, Stage 2 with
canine and developing incisor teeth; and (C) OS11288,
Stage 5 with developed incisor teeth.
outline at the base of the dorsal and anal fins around 35-40
mmSL (Pearcy et al. 1977a).
Stage 2 (metamorphlc precompetent larvae), 42.3-60.4
mm SL Six morphological characters that define the ini-
tiation event of metamorphosis are, in their approximate
order of development: jaw dentition, completion of eye
migration, position of anterior margin of dorsal fm, posi-
tion or presence of posterior process of the coracoid,
pectoral fin morphology, and beginning of asymmetrical
coloration. Numerical scores given to the two or three
states of each character are shown in Table 3. A metamor-
phosing presettlement individual can have a metamorphic
score of 1 to 8. A score of 9 defines Stage 3, metamorphic
competent larvae.
Dover sole larvae have canine-like teeth on left and right
jaws (Fig. 3A). During Stage 2, incisor-like teeth develop
on the left premaxilla and dentary (Fig. 3B). The canine-
like teeth are lost from both jaws coincident with eruption
of incisors in the left jaw (Fig. 3C).
In Stage 1 larvae, anterior dorsal-fin pterygiophores and
fin rays are posterior to the orbit of the left eye which is
located on the dorsal ridge of the cranium (Fig. 4A). Dur-
ing Stage 2, these pterygiophores move anterior to the orbit
of the left eye (Fig. 4B).
The posterior process of the coracoid in larvae is a long,
slender element that projects posteriorly above the visceral
cavity, underneath the skin (Fig. 5A). During metamor-
phosis the process is resorbed. At the beginning of resorp-
tion, during Stage 2, the distal end of the process curls
anteriorly into a hook (Fig. 5B). In our samples there is
some indication that the process deteriorates (poor stain-
ing with alcian blue), but there is no gradual reduction in
length or thickness of the process. Specimens either have
the process or have lost it (Fig. 5C).
The pectoral fin in Stage-1 larvae is a paddle-shaped
membrane with a thin, fleshy base and without radials. Dur-
ing Stage 2, a fleshy rectangular base, cartilaginous radials,
and the adult complement of fin rays form (Fig. 5).
The Stage-1 larval color pattern consists of little or no
pigment on the midlateral areas. A transitional pattern,
in which melanophores aggregate along myosepta, is
followed by the first indication of melanophores aggre-
gating in two approximately circular groups anteriorly and
posteriorly along the lateral line (Fig. 6). We score larval
Premaxilla
Maxilla
Dentary -
1 mm
Articular
Angular
B
Premaxilla
Dentary
1 mm
Maxilla
Articular
Angular
Premaxilla
Dentary
1 mm
Maxilla
Articular
Angular
290
Fishery Bulletin 90(2). 1992
1 mm
1 mm
Figure 4
Position of dorsal fin rays and size of eyes in
developing Dover sole Microstomus pacificus :
(A) OS12558, Stage 1; and (B) OS12563, Stage 2.
Figure 5
Pectoral fin development in Dover sole Microstormis
pacificm: (A) OS12558, Stage 1 w\th straight ventral
process of coracoid; (B) OSl 1377, Stage 2 with hooked
tip on ventral process of coracoid; and (C) OS12563,
Stage 3 after resorption of ventral process of coracoid.
Postcleithrum
nor process coracoid
Posterior process coracoid
1 mm
B
Postcleithrum
Fin rays
f
Radials
Anterior process coracoid
Posterior process coracoid
1 mm
Postcleittirum
Fin rays
Radials
Anterior process coracoid
1 mn
Markle et al.: Metamorphosis of Microstomus paaftcus
291
Figure 6
Right-side midlateral pigmentation patterns during
development in Dover sole Microstomus pacificus :
(A) OS13115, Stage 2, developmental score 5,
larval pattern of no melanophores on myomeres;
(B) OS13118, Stage 2, developmental score 8, tran-
sitional pattern of melanophores on myosepta;
and (C) OS13117, Stage 3, aggregated pattern of
melanophores in circular area on caudal peduncle
and anterior trunk.
and transitional patterns equally and
consider the circular aggregations as
the first indication of asymmetrical
coloration.
Coincident with changes in these six
features are changes in features that
are not easily coded: gradual loss of
otic spines, reduction in body depth
(Figs. 3 and 8), reduction in interor-
bital width (Fig. 7), increase in right
eye diameter (Fig. 7), and increase in
right upper jaw length (Fig. 3). Devel-
opment of body scales also begins in
Stage-2 specimens with metamorphic
scores of 7 or 8. Body scales first form
above and below the lateral line,
anteriorly near the pectoral fin base,
and on the caudal peduncle.
Stage 3 (metamorphic competent
larvae), 40.7-74.9 mm SL Stage-3
specimens have a metamorphic score
of 9, indicating that all six initiation-
event features have either begun or
reached the adult state. These speci-
mens have a translucent appearance,
intermediate between the earlier
transparent stages and later opaque
stages. Stage-3 specimens have asym-
metrical coloration, retain the coiled,
larval gut configuration, and have
resorbed the posterior process of the
coracoid. Some morphometric features
initiated in Stage 2, such as increasing
right eye diameter and shrinkage in
body depth, continue in Stage 3 (Fig.
2). Ossification of pelvic-fin rays and
radials is initiated in Stage 3, appar-
ently after settlement (1 of 10 pelagic
specimens and 4 of 4 benthic Stage-3
specimens have ossified pelvic fin rays
and radials).
292
Fishery Bulletin 90(2). 1992
Figure 7
Dover sole MicrostoTOMA- poci/iCTis right lateral view: (A)OS13214, Stage 1, 20.4mmSL; (B)OS11377, Stage 2, 54.5mmSL; (C)OS13202,
Stage 3 benthic capture, 52.4 mm SL; (D) OS 13202, eariy Stage 4, 61.5 mm SL (intestinal loop is dark area above anal fin); (E) OS13203,
late Stage 4, 58.4 mm SL; (F) OS13204, early Stage 5, 78.4 mm SL, arrow points to posterior end of intestinal loop.
Stage 4 (metamorphic postcompetent larvae), 41 .7-
79.3 mm SL Adult Dover sole have a long intestinal
loop in the secondary body cavity above anal fin ptery-
giophores (Hagerman 1952). This intestinal loop (Fig.
7) forms after settlement, and its initiation is the de-
fining feature of Stage 4. Continuous metamorphic
changes in morphology, such as shrinkage in body
depth, are completed during Stage 4 (Fig. 2).
Markle et al : Metamorphosis of Microstomus pacificus
293
0.95 -
CO 090
c
tn
OJ 0.85
_C
*-•
3
o
c
CO
0.75
40 80 120 160 200
Standard Length (mm)
Stage 5 (juvenile), 48.9 mm SL to sexual maturity
We define the climax event, and Stage 5, as the point
at which length of the intestinal loop attains adult pro-
portions. The continuous nature of this process is il-
lustrated in the logarithm ratios of SINT/SL (Fig. 8).
We chose a cut-off ratio by calculating the ratio for
2mmSL increments and examining the rate at which
the ratio changes over length. The greatest rate of
change occurs between 67 and 69mmSL, during which
the mean ratio changes from 0.85 to 0.89. We chose
the midpoint of these ratios and therefore define Stage
5 as those individuals with a ratio of In (SINT)/ln (SL)
> 0.87. Coincident with this change is an overall
darkening of body color such that Stage 5 specimens
look like small adults.
General features of early development
and metamofphosis
Unlike most flounders, initiation of eye migration in
Dover sole is uncoupled from the change in habitat
from planktonic to benthic, as well as from the process
of metamorphosis (as defined herein). Eyes are sym-
metrical up to a maximum size of only 13.4 mm SL, and
the left eye can be on the midline in specimens as small
Figure 8
Relationship between the In SINT/ln SL ratio and stan-
dard length during development of Dover sole Microstomus
-pacificus. Symbols represent Stage 3(H), Stage 4 (O), and
Stage 5 (A).
as 9.5mmSL (Pearcy et al. 1977a). Eye migration in
Dover sole is arrested during planktonic growth,
with the left eye stopping at the dorsal margin of
the cranium at 15-20 mm SL (Fig. 4A). It remains in
this position until metamorphosis. Thus, during most
of their planktonic life, the eyes of Dover sole are
asymmetrical.
There is a complex relationship between body depth
and SL (Fig. 2), including (1) an interval of rapid in-
crease from about 10 mm to at least 60.4 mm SL in some
individuals, (2) a compensatory shrinkage phase over
the size range 40.7-74.9 mm SL, and (3) a more typical
linear growth phase that may begin in specimens as
small as 41.7mmSL. Body depth reduction is a regres-
sive process (Youson 1988) in which lengths of neural
and hemal spines and pterygiophores are reduced (Fig.
9). Two- and three-fold reductions occur in lengths of
first caudal neural and hemal spines and their imme-
diate anterior and posterior pterygiophores. Conse-
quently, metamorphosing Stage-3 specimens 40-50 mm
SL have neural and hemal elements comparable in
length to those of 20-30 mm SL Stage-1 larvae. Neural
and hemal elements and dorsal and anal pterygiophores
in Stage-1 larvae are cartilaginous or weakly ossified,
and vertebral centra lack zygopophyses. Complete
ossification of neural and hemal elements and forma-
tion of zygopophyses occurs in Stages 2 and 3.
During most of metamorphosis, especially in Stages
2 and 3, body length appears to be arrested (Fig. 10).
Although the sample size of Stage-2 larvae limits our
confidence in further analysis, the data show little in-
dication of growth between Stages 2 and 3 (Fig. 10).
Because metamorphosis occurs over a broad range of
sizes, similarity in size minima and maxima between
stages also suggests little or no growth in body length.
For example, the minimum sizes for Stages 2, 3, and
4 are almost identical (42.3, 40.7 and 41.7mmSL,
respectively). During Stage 4 there is finally some
indication of grovrth because the smallest Stage-5
juvenile is 48.9 mm SL, more than 7 mm larger than the
smallest Stage-4 larva. Yet, even this juvenile is 26 mm
smaller than the largest metamorphosing Stage-3
larva.
There is an apparent loss in body weight during
metamorphosis because of a decrease in mean weight
from 2.6 to 2.4g from Stage 2 to Stage 4 (Fig. 11).
However, the small sample size of Stage 2 and our
measuring error preclude attaching significance to the
294
Fishery Bulletin 90(2). 1992
E
E,
a>
c
Q.
CO
«
13
a>
c
T3
::
to
o
<19.9 20.0-29,9 30,0-39,9 40 0j»9.9 50 0-59,9 60.0.69,9 70,0-79 9 80 0-89 9
Standard Length (mm)
Figure 9
Change in length of first caudal neural spine during development
of Dover sole Microstcnmis pacificus. Symbols represent Stage 1 (A),
Stage 2 (•), Stage 3 (■), Stage 4 (O), and Stage 5 (A).
Standard Length (mnn)
.(k oi O) ^ CD
O O O O O
:
r-
J
Mean st
of obser
>40mni
Stage-3
pacificu
12 3 4
Stages
Figure 10
andard length (horizontal line), size interval for 50%
■vations (box), and size range for Stage-1 specimens
, all Stage-2 specimens, and all January and March
and Stage-4 specimens of Dover sole Microstomux
s.
apparent loss. During the second year of life in
the plankton (see next section), body weight in-
creases an order of magnitude from a mean of
about 0.30g for Stage 1 in February to 2.0-4.0g
for Stage 2. All individuals that reach a size of
40mmSL are at least 0.74g; Stage-2 specimens
are at least 1.39g; Stage 3, at least 0.80g; and
Stage 4, at least l.Og. Because our sample size
for Stage 3 is relatively large, our best estimate
of a weight threshold for metamorphosis is
~0.8g. However, if the suggestion of weight loss
during metamorphosis is not an artifact (Fig. 11),
the weight threshold may be closer to the mini-
mum weight of Stage 2. Further complicating an
estimate of that threshold is the observation that
our lightest Stage-2 specimen (42.6mm SL and
1.39g) was caught in January with a developmen-
tal score of 7, and presumably may already have
lost weight.
Because SINT increases during metamorphosis
and BDIA decreases, the SINT/BDIA ratio pro-
vides an additional means of visualizing the rela-
tionship between the metamorphic process and
developmental stages (Fig. 12).
Timing and duration of stages
Temporal change in size of Stage-1 larvae <40mmSL
was analyzed using MPA (Fig. 13). Small larvae, about
6-8 mm, were foimd from February to June. The small-
est identifiable mode was in April, and from November
to March modes were level around 22-25 mm. Two
notable features of these data are apparent accelerated
growth in June and reduced availability of larger
Figure 1 1
Mean weight (horizontal line), weight interval for 50% of
observations (box), and weight range for Stages 2-4 in Dover
sole MicTostomvs pacificus.
6
-
5
':
4
s
1
^ 3
1
D) ^
' —
1
1
(D
5 ,
L^
^ ^
0
1
1 1 1
2 3 4
stages
Markle et al : Metamorphosis of Microstomus pacificus
295
D
C
<
TO 2.0
Q.
(D
Q
>.
O
m
_c
V^
CO
03
o
C
05
^."^
Standard Length (mm)
Figure 12
Relationship between the SINT/BDIA ratio and standard
length during development of Dover sole Microstomus pacifi-
cus. Symbols represent Stage 1 ( • ), Stage 2 (A), Stage 3(B),
Stage 4 (O). and Stage 5 (A).
specimens after March (Fig. 13). Ail small larvae (< 10
mmSL) were collected on one day, 12 June 1971, be-
tween 50 and 67km offshore, whereas larger larvae
from June were collected considerably further offshore,
108-275km on various dates. There appears to be little
or no coherent size progression after the 24.5 mm mode
in March. Specimens >30mm are found in every
month, and weakly-defined modes can be visualized
around 50 mm in June, July, and September. Accel-
erated growth in April and May would seem to be re-
quired if the modal size were to double from about
25 mm in March to 50 mm in the second summer of life.
Paradoxically, April is a time when micronekton bio-
mass is normally low (Pearcy 1976).
Stage-2 specimens were caught from June to Febru-
ary (Table 4). The coherent progression of metamor-
phic scores for Stage-2 larvae indicates that metamor-
phosis begins as early as June; Stage-3 larvae are
present as early as December and as late as March.
About 6 months seems to be required to progress
4 5 6 7 8 9 10 11 12 1 2 3
Month of Capture
Figure 13
Relationship between standard length and month of capture
for Dover sole Microstomus pacificus, Stages 1-4 collected
in midwater trawls. Stars indicate modes determined by modal
progression analysis in larvae <40mmSL. Stage-2 and -3
specimens are circumscribed by lines.
Table 4
Seasonal catch-per-effort for planktonic Stage-2 and Stage-3
Dover sole Microstomtis pacificus larvae off Oregon. N =
number of trawls.
No. of Stage 2 based on
metamorphic score
Month of No. of
collection N 12345678 Stage 3
January 128 ------2- 2
February 167 -------2 4
March 147 ________ 2
April 201-------- 0
May 106-------- 0
June 278 1------- 0
July 291 -----1-- 0
August 303 1 1 ----- - 0
September 217-1-421-- 0
October 129 ----11-1 0
November 217 — -----2 1 0
December 126 ------26 2
through Stage 2.
The seasonal distribution of Stage-3 larvae in ben-
thic samples was consistent with their planktonic dis-
tribution (Table 5). During bimonthly sampling in 1989,
98.5% of Stage-3 specimens were caught in January
or March. Five Stage-3 specimens were caught in May,
and most new settlers appear to be in Stage 5 by July
296
Fishery Bulletin 90(2). 1992
Table 5
Seasonal distribution of Dover sole Microstomus padficus
stages in bottom-trawl samples off Oregon, 1989.
Month of
collection
N
Number (%)
Stage 3
Stage 4
Stage 5
January
371
177 (48)
12 (3)
182 (49)
March
655
155 (24)
347 (53)
153 (23)
May
222
5 (2)
113 (51)
104 (47)
July
60
0
11 (18)
49 (82)
September
154
0
5 (3)
149 (97)
November
267
0
3 (1)
264 (99)
(Table 5). Individuals appear to require about 45 days
to progress through Stage 3. Our laboratory -held
specimen progressed through Stage 4 in 43 days.
Overall, the progression through Stages 2-4 appears
to require about 1 year for the population as a whole
and about 9 months for an individual.
The length and weight of benthic Stage 3 larvae were
compared between January and March 1989. Mean
length in March (51.5mmSL) was significantly smaller
than the mean length in January (56.3 mm SL, P<
0.00001). March specimens were also significantly
lighter (2.0 g) than January specimens (2.7 g, P<
0.0001). A similar pattern (earliest individuals in a
stage being largest) was seen in Stage-2 larvae. The
mean length of Stage-2 larvae captured between
January and March Gate in the season. Table 4) was
smaller (53.0mm vs. 54.3mmSL) and lighter (2.4 vs.
2.8 g) than Stage-2 larvae captured between June and
December (early in the season. Table 4), but the dif-
ferences were not significant (length P< 0.4642; weight
P<0.1227).
Habitat of stages
On average, Stage-1 specintens were caught in nets
fished to a maximum depth of 338 m, Stage-2 specimens
in nets fished to 538 m, and planktonic Stage-3 speci-
mens in nets fished to 293 m. All planktonic Stage-3
larvae were caught at night, between 1835 and 0544
hours, and 93% of Stage-2 larvae were caught at night,
between 1802 and 0748 hours. Sampling effort was also
greatest at night (Table 1; opening-closing nets col-
lected 516 Dover sole but most (509) were Stage 1). The
minimum depth of capture in the discrete depth
samples was < 100m for 82%, and <300m for 95%, of
Stage-1 specimens. Only six Stage-2 specimens were
caught in discrete depth samples, and only two of these
in nets not fishing the surface. One Stage-2 larva was
caught at 100-150m and the other at 400-500m. The
240
-
A
200
-
■i B °.
"
160
-
° .° f*^i 8
120
•
m
■
80
-
■^Bp "
■
°°°lw
i 'b°.|
1 °
B ^ □
% I
' 1
B
40
•
0
-
■ 1 1 1 1
. 1 . .
. . ,
■ '
240
-
B
200
/
^
160
-
/
5
120
-
\
80
-
^
^
/
y~ -]
-
•
0
-
. r . .
. ■ 1 ■ . ■ .
Depth of Capture (m)
Figure 14
Relationship between standard length and depth of capture
of developing Dover sole Microstomus pacificus, caught in
bottom trawls off Oregon, January and March 1989: (A) scat-
terplot of data points. Stages 3-5; (B) polygons circumscribing
areas bounded by specimens in Stages 3-5.
single Stage-3 larva collected in a discrete depth sample
was collected at 0-330 m.
Benthic specimens were caught at depths shallower
than the maximum depth fished by non-closing mid-
water nets. Based on our stratified sampling, a com-
parison of depth of capture of stages shows that
Stage-3 specimens were caught at an average depth
of 146 m (SE 2.65, range 55-377 m), Stage-4 specimens
in January and March at an average depth of 118 m (SE
0.68, range 40-170m), and Stage-5 specimens in
January and March at an average depth of 110 m (SE
Markle et al.: Metamorphosis of Microstomus pacificus
297
0.48, range 75-188 m). Compared with Stages 4 and
5, the greater average depth and variance of benthic
Stage-3 larvae indicate a much broader depth distribu-
tion (Fig. 14).
Stage-3 larvae occupy a transitional "landing" zone
quantitatively distinct from, but overlapping, the late-
larval and juvenile nurserygrounds. Although Stage-3
larvae caught in bottom trawls quickly take to the bot-
tom when placed in aquaria (pers. observ.), their night-
time capture in midwater trawls and da3rtime capture
in bottom trawls suggest they may be engybenthic
(nearbottom) or benthopelagic, rather than exclusive-
ly benthic.
Behavior associated with metamorphosis presumably
includes some short-term (hours to days) switching be-
tween midwater and bottom habitats. In one individual
in our data set, the behavior continued into Stage 4.
A 53.0 mm Stage-4 specimen was caught 19 April 1963
off the mouth of the Columbia River at 0411 hours in
a midwater trawl fished to 73 m over a bottom depth
of about 125 m. Its gut loop was well developed and con-
tained sand grains. Additional evidence is provided by
midwater Cobb trawl samples collected by W. Lenarz
and colleagues (NMFS Southwest Fish. Sci. Cent.,
Tiburon, CA 94920) between Monterey and San Fran-
cisco, California, from 28 March to 2 April 1990. In
eight nighttime (2235-0447 hour) samples, fished at
0-1 10 m (most 0-30 m) over bottom depths of 33-1462
m, they collected 14 Stage-3 larvae (40. 4-51. 2 mm SL)
and 16 Stage-4 larvae (42.4-53.4 mm SL). Stage-3
larvae were collected over bottom depths of 73-1462 m,
and Stage-4 larvae were collected over bottom depths
of 33-91 m. Thus, settling Stage-3 larvae were found
in a "landing" zone at 55-377 m and in a wedge of the
water column above and seaward of that zone.
Discussion
Time-line
Dover sole spawn in deep water in winter, December
to February, according to Hagerman (1952), and
November to April according to the circumstantial
evidence of Harry (1959). Yoklavich and Pikitch (1989)
provide evidence that smaller Dover sole have an
earlier, shorter spawning season than larger fish, and
that Dover sole now mature at significantly smaller
sizes than reported by Hagerman (1952) or Harry
(1959). These observations suggest the possibility that
size-selective exploitation might have shifted the
spawning season to earlier dates.
However, other observations suggest that peak
hatching of Dover sole off Oregon is later, not earlier,
than indicated by Hagerman (1952) or Harry (1959).
Results of sampling the commercial Dover sole catch
off southern Oregon (43°N) from March 1990 to
September 1991 indicate running ripe females were
caught from February through July with a peak in
April (Mike Hosie, Oreg. Dep. Fish Wildl., Charleston,
OR 97420, pers. commun.). Spent females increased
from less than 10% of all females in April to 100% by
early August. However, these observations may be
biased towards later-spawning fish because the com-
mercial catch is culled of small fish (Yoklavich and
Pikitch 1989). Experiments performed in 1972 by
S. WOliams at Newport, Oregon, showed that hatching
took 18 days at 12.5°C, 27 days at 10.0°C, and 38 days
at 7.5°C (Mike Hosie, pers. commun.). In agreement
with these observations, small larvae (<10mmSL) in
this study were collected from February to July, with
most caught in April and May (Fig. 13, Pearcy et al.
1977a). In ten NMFS ichthyoplankton cruises con-
ducted at 40-48°N from 1980 to 1987, high densities
of Dover sole eggs were found in each of six cruises
conducted in March, April, or May; none or trace
amounts were found in four cruises conducted in
August, November, or January (Urena 1989; M. Doyle,
NMFS Alaska Fish. Sci. Cent., Seattle, WA 98115,
pers. commun.). Finally, "spawning" adults off Alaska
have been collected primarily in May and June
(Hirschberger and Smith 1983), and eggs are collected
in June (Kendall and Dunn 1985). Thus, the weight of
evidence seems to indicate that most Dover sole off
Oregon hatch from February (Fig. 13) to August
(Urena 1989), with a peak in April and May (see also
the time-line in Hayman and Tyler 1980).
Settlement is restricted to the period from January
to March or April (Table 5), whereas metamorphosis
requires a protracted period of up to one year, occurs
at sizes >40mmSL (Fig. 3), includes little growth in
body length, and may include loss of weight. Cessation
in growth of body length before and during metamor-
phosis has been documented in other flounders (Fuku-
hara 1986, 1988). If the modal size of Stage-1 larvae
is ~25 mm in March, then the average duration of the
planktonic period of Dover sole is about 21 months
(Fig. 15). However, the timing of settlement has a size
component; larger larvae tend to settle before smaller
larvae. It seems reasonable that larger larvae are those
that grow faster, but it is also possible that they are
slow growers or have otherwise delayed metamor-
phosis and, therefore, are more than 2 years old (see
Discussion below).
Distribution and relative abundance of metamorphic
planktonic stages provide additional insight. Larger
planktonic specimens were generally rare in midwater
trawl collections (Fig. 13). However, Stage-2 larvae,
with developmental scores of 7 and 8, and Stage-3
larvae were the most abundant of all metamorphic
stages found in midwater (Table 4), even though they
298
Fishery Bulletin 90(2). 1992
Yearl
Eggs
Stage 1
Year 2
stage 1
Stage 2
Stage 3
Years
stage 2
Stage 3
Stage 4 ■
Stage 5
were collected in months with few
samples (Table 1). Planktonic Stage-2
larvae were caught in nets fished
deeper than either Stage- 1 or plank-
tonic Stage-3 larvae. The rarity of
larger planktonic and early metamor-
phic stages may reflect movement
deeper into the mesopelagic zone and
lower relative sampling effort in
deeper water. Late in Stage 2 (devel-
opmental scores 7 and 8) and in Stage
3 this trend appears to be reversed, as
these stages were caught more fre-
quently. If metamorphosis is a time of
increased vulnerability, deeper water
may provide a predation refuge. Alter-
natively, the behavior may place meta-
morphosing specimens in a water mass
that facilitates late larval transport.
Settlement seems remarkably grad-
ual, coincides with the downwelling
season, ends with the spring transition
in the oceanographic regime (Huyer et
al. 1979), and occurs over a very broad
"landing" zone (Fig. 14). Stage-3 lar-
vae settling outside the nursery zone
may experience differential mortality,
or their broad depth distribution may
reflect a process of testing the habitat
in search of the preferred nurseryground. Capture of
Stage-3 and -4 specimens in both nighttime midwater
and daytime bottom trawls suggests a diel vertical
search pattern.
Egg and larval drift
A proposed recruitment mechanism for Dover sole
(Hayman and Tyler 1980, Parrish et al. 1981) focuses
on inshore-offshore transport. The long planktonic
period of Dover sole implies that alongshore transport
also may be important. Our data allow some first-order
generalizations about the distribution of early-life-
history stages and may give further insight into the
recruitment mechanism.
Urena (1989) found that greatest abundances of
Dover sole eggs were in neuston samples collected
beyond the 200m isobath. In April and May in the
upper 50 m, the current flows southward at about
10-15cm/second at the 200m isobath (Huyer 1977,
fig. 9; Huyer et al. 1979) and is even weaker further
offshore (Huyer and Smith 1978). At 10 cm/second,
eggs could be transported 260 km southward in 30
days, assuming that their transport was not inter-
rupted by offshore jets or gyres. Onshore-offshore
transport of eggs should be variable. During upwell-
Month
Figure 15
Hypothetical time-line of development for a cohort of Dover sole Microstomus
pacifiaxs, off Oregon. Solid lines represent presumed peak times for the different
stages, and dotted lines represent the ranges.
ing, the upper 20 m may experience an average offshore
velocity of 2-5 cm/second (Huyer 1983), and the upper
5 m may experience an average offshore velocity of
15 cm/second (Peterson et al. 1979). The short duration
of the egg stage, restricted area of high offshore
velocity, and the return inshore of water masses dur-
ing relaxation after upwelling (Peterson et al. 1979)
suggest that average offshore transport of eggs should
be slow, but nontrivial. Deepwater spawning may
help reduce both alongshore and onshore-offshore
transport.
Stage-1 larvae also are found beyond the 200 m
isobath (Pearcy et al. 1977a) and, like eggs, would be
vulnerable to the southward flow of the California Cur-
rent. In fact, the large surface area of the body of
Dover sole larvae might facilitate such transport.
Although there are important seasonal changes in
direction (Huyer et al. 1979), on an annual basis the
average surface current around 100 km offshore is
~0.5-1.0cm/second to the south (Hickey 1979, fig. 8b).
If Stage 1 lasts an average of 15 months, and assum-
ing a mean flow of 0.75 cm/second, these larvae would
travel an additional 295km southward.
We suggest that early Stage-2 larvae move into
deeper water. The California Undercurrent is a north-
ward-flowing countercurrent located below 200 m and
Markle et al,: Metamorphosis of Microstomus pacificus
299
influencing an area up to 500 km off the shelf (McLain
and Thomas 1983). Its velocity is < 10 cm/second over
the continental slope north of Cape Mendocino (Hickey
1979) and weaker seaward of the slope. Somewhat fur-
ther south, between Pt. Arena and Pt. Reyes, the cur-
rent is 3-10 cm/second from July to October and <1
cm/second from October to January (Huyer et al. 1989).
If eggs and Stage-1 larvae are displaced, on average,
555km (260-1-295) southward of their spawning site,
a northward-flowing undercurrent of 3.25cm/second
would be sufficient to return Stage-2 larvae to the
vicinity of their spawning site in 6 months. This does
not seem to be an unreasonable average velocity for
the undercurrent from July to January.
The depth range of the Stage 3 "landing" zone
(55-377 m) corresponds with the northward under-
current located at 200-300 m (Huyer and Smith 1985).
However, these larvae appear to need a mechanism to
bring them shoreward. The surface Ekman layer,
0-20 m, within which wind-driven transport occurs
(Huyer 1983), could be reached if larvae moved up in
the water column during storms. Dial offbottom migra-
tions could be part of this mechanism. Alternatively,
as the body surface area is reduced during this stage,
larvae may become less passive and move actively
inshore.
Delayed metamorphosis and settlement
The protracted process of metamorphosis in Dover sole
is contrary to expectations based on the ideas of salta-
tory ontogeny (Balon 1981). In general, ontogenetic
transformations are expected to occur rapidly because
intermediate forms are presumed to be maladapted.
For example, loss of teeth from the right side of the
jaw and development of incisors on the left side seem
to hold no advantage for a planktonic larva, yet this
is the situation in Dover sole for several months dur-
ing the precompetent Stage 2. Delayed metamorphosis
is also related to the concept of saltatory ontogeny;
because the transition is assumed to be quick, an
organism without the proper cues simply delays meta-
morphosis and settlement. In other words, it keeps the
morphology appropriate for the habitat. Typically, field
researchers identify a minimum threshold size or
developmental stage for metamorphosis and assume
that planktonic specimens greater than the threshold
size or in the threshold stage have delayed metamor-
phosis (Pechenik 1986). Others have used a minimum
age as a threshold (Cowen 1991).
Pearcy et al. (1977a) suggested that larger "hold-
over" Dover sole larvae (>50mmSL) delayed meta-
morphosis and few successfully recruited to the benthic
juvenile stage. Delayed metamorphosis is predicted
for coastal organisms subjected to offshore transport
(Jackson and Strathmann 1981) and there is some
evidence for delayed metamorphosis in fishes (Victor
1986, Cowen 1991). An advantage of delayed metamor-
phosis is extension of the settlement season beyond
what would be expected based on the spawning season
(Victor 1986). Contrary to this expectation, the dura-
tion of Dover sole settlement is seasonally restricted
and, off Oregon, no greater than the duration of the
spawning season. Because precompetent larvae are
probably a great distance from their settlement site,
cues for metamorphosis are likely to be seasonal rather
than site-related.
Experimental studies focusing on flounders have
showTi (1) fast-growing individuals metamorphose at
smaller sizes, (2) fast-growing individuals retain their
faster growth rate for at least several weeks after
metamorphosis, (3) age at metamorphosis (defined by
eye migration) is more variable than size at metamor-
phosis, and (4) a target size or threshold must be
reached prior to metamorphosis (Policansky 1982,
Chambers and Leggett 1987, Chambers et al. 1988).
Other fishes and organisms may have age-triggered,
size-triggered, or age- and size-triggered metamor-
phosis (Policansky 1983). Policansky (1983) points out
that a size threshold would be expected when there is
a size difference in available food between different
habitats or a minimum energy requirement to success-
fully function at a certain stage.
We suggest two contrasting interpretations of the
early life history of Dover sole. If size and age at
metamorphosis are positively correlated, as is the case
in winter flounder (Chambers et al. 1988), then larger,
earlier settlers are older and slower-growing than
smaller, later settlers. The difference in age could be
the difference between early and late spawners or be-
tween different years of spawning. Alternatively,
variation in size at metamorphosis may simply reflect
differential growth rates operating for a long time,
probably at least 2 years. As a consequence, larger,
earlier settlers would be the faster growers rather than
slower growers. One could distinguish between these
alternatives and demonstrate delayed metamorphosis
by documenting different year-classes among settlers.
In terms of life-history strategies, delayed metamor-
phosis and protracted metamorphosis may confer
similar advantages. Extension of settlement through
delayed metamorphosis allows for adaptive responses
to short-term oceanographic variability and avoidance
of settling during unfavorable conditions. If metamor-
phosis and settlement are cued to favorable seasons,
protracted metamorphosis and the ability of competent
metamorphosing individuals (Stage-3 larvae) to spend
several months moving between midwater and bottom
habitats should also compensate for any short-term im-
favorable oceanographic conditions.
300
Fishery Bulletin 90(2). 1992
Acknowledgments
This study was funded, in part, by Oregon Sea Grant
with funds from NOAA, Office of Sea Grant, Depart-
ment of Commerce, under grant NA85AA-D-SG095
(project R/OPF-29) and from appropriations made by
the Oregon State Legislature; by the Pacific Outer
Continental Shelf Region of the Minerals Management
Service, U.S. Department of the Interior, Washington,
D.C., under Contract 14-12-0001-30429; and by Na-
tional Marine Fisheries Service contract NA-87-ABH-
00014. Some ship time was generously donated by
T.N. Thompson. C. Ridgley, D. Nelson, S. Banks, and
R. Melendez helped collect data. Shipboard work was
made as pleasant as circumstances permitted by T.N.
Thompson and the crew of FV Olympic. A. Kendall
(NMFS, Seattle, WA), W. Lenarz (NMFS, Tiburon,
CA), and G. Moser and J. Butler (NMFS, La Jolla,
CA) shared results of ongoing research. W. Pearcy,
D. Stein, and collaborators saved, documented, and
made available 20 years of planktonic and midwater
trawl collections that were invaluable for this study.
Citations
Allen, M.J., and A.J. Mearns
1977 Life history of the Dover sole. Annu. Rep. South Calif.
Coastal Water Res. Proj. 77-1, El Segundo, p. 223-228.
Balon, E.K.
1979 The theory of saltation and its application to the ontogeny
of fishes: Steps and thresholds. Environ. Biol. Fishes 4:
97-101.
1981 Saltatory processes and altricial to precocial forms in the
ontogeny of fishes. Am. Zool. 21:573-596.
1984 Reflections on some decisive events in the early life of
fishes. Trans. Am. Fish. Soc. 113:178-185.
Bhattacharya, C.G.
1967 A simple method for resolution of a distribution into Gaus-
sian components. Biometrics 23:115-135.
Chambers, R.C., and W.C. Leggett
1987 Size and age at metamorphosis in marine fishes: An
analysis of laboratory-reared winter flounder (Pseudopleuro-
nectes americamis) with a review of variation in other species.
Can. J. Fish. Aquat. Sci. 44:1936-1947.
Chambers, R.C., W.C. Leggett, and J.A. Brown
1988 Variation in and among early life history traits of
laboratory-reared winter flounder Pseudopleuronectes
americanus. Mar. Ecol. Prog. Ser. 47:1-15.
Chilton, O.E., and R.J. Beamish
1982 Age determination methods for fishes studied by the
groundfish program at the Pacific Biological Station. Can.
Spec. Publ. Fish. Aquat. Sci. 60, 102 p.
Cowen, R.K.
1991 Variation in the planktonic larval duration of the tem-
perate wrasse Semicossy-phus pulcher. Mar. Ecol. Prog. Ser.
69:9-15.
Demory, R.L.
1972 Scales as a means of aging Dover sole {Microsto-mus
p(unficus). J. Fish. Res. Board Can. 29:1647-1650.
Doyle, R.W.
1975 Settlement of planktonic larvae: A theory of habitat
selection in varying environments. Am. Nat. 109:113-126.
Fukuhara, O.
198G Morphological and functional development of Japanese
flounder in early life stage. Bull. Jpn. Soc. Sci. Fish. 52:81-91.
1988 Morphological and functional development of larval and
juvenile Limanda yokohamae (Pisces: Pleuronectidae) reared
in the laboratory. Mar. Biol. (Beri.) 99:271-281.
Hagerman, F.B.
1952 The biology of the Dover sole. Calif. Dep. Fish Game,
Fish. Bull. 85:1-48.
Harry, G.Y. Jr.
1959 Time of spawning, length at maturity, and fecundity of
the English, Petrale, and Dover soles (Parophrys vetulus,
Eopsetta jordani, and Microstomus pacijlcus, respectively).
Greg. Fish Comm. Res. Briefs 7:5-13.
Hayman, R.A., and A.V. Tyler
1980 Environment and cohort strength of Dover sole and
English sole. Trans. Am. Fish. Soc. 109:54-70.
Hickey, B.M.
1979 The California Current System— Hypotheses and facts.
Prog. Oceanogr. 8:191-279.
Hirschberger, W.A., and G.B. Smith
1983 Spawning of twelve groundfish species in the Alaska and
Pacific Coast regions, 1975-81. NOAA Tech. Memo. NMFS
F/NWC-44, Northwest Fish. Sci. Cent., Seattle, WA 98112,
50 p.
Hunter, J.R., J.L. Butler, C. Kimbrell, and E.A. Lynn
1990 Bathymetric patterns in size, age, sexual maturity, water
content, and caloric density of Dover sole, Microstomus
pacificus. Calif. Coop. Oceanic Fish. Invest. Rep. 31:132-144.
Buyer, A.
1977 Seasonal variation in temperature, salinity and density
over the continental shelf off Oregon. Limnol. Oceanogr.
22:442-453.
1983 Coastal upwelling in the California Current system.
Prog. Oceanogr. 12:259-284.
Huyer, A., and R.L. Smith
1978 Physical characteristics of Pacific northwestern coastal
waters, hi Krauss, R. (ed.), The marine plant biomass of the
Pacific Northwest coast, p. 37-55. Oregon State Univ. Press,
Corvallis.
1985 The signature of El Nino off Oregon, 1982-1983. J.
Geophys. Res. 90:7133-7142.
Huyer, A., E.J.C. Sobey, and R.L. Smith
1979 The spring transition in currents over the Oregon con-
tinental shelf. J. Geophys. Res. 84:6995-7011.
Huyer, A., P.M. Kosro, S.J. Lentz. and R.C. Beardsley
1989 Poleward flow in the California Current system. In
Neshyba, S.J., Ch.N.K. Mooers, R.L. Smith, and R.T. Barber
(eds.), Poleward flows along eastern ocean boundaries, p.
144-159. Springer-Veriag, NY.
Jackson, G.A., and R.R. Strathmann
1981 Larval mortality from offshore mixing as a link between
precompetent and competent periods of development. Am.
Nat. 118:16-26.
Kendall, A.W. Jr., and J.R. Dunn
1985 Ichthyoplankton of the continental shelf near Kodiak
Island, Alaska. NOAA Tech. Rep. NMFS 20, 89 p.
McLain, D.R., and D.H. Thomas
1983 Year-to-year fluctuations of the California Countercur-
rent and effects on marine organisms. Calif, Coop. Oceanic
Fish. Invest. Rep. 24:165-181.
Markle et al : Metamorphosis of Microstomus pacificus
301
Parrish, R.H., C.S. Nelson, and A. Bakun
1981 Transport mechanisms and reproductive success of fishes
in the California Current. Biol. Oceanogr. l(2):175-203.
Pauly, D.
1987 A review of the ELEFAN system for analysis of length-
frequency data in fish and aquatic invertebrates. In Pauly,
D., and G. Moran (eds.), Length-based methods in fisheries
research, p. 7-34. Int. Cent. Living Aquat. Res. Manage.,
Manila.
Pearcy, W.G.
1976 Seasonal and inshore-offshore variations in the standing
stocks of micronekton and macrozooplankton off Oregon.
Fish. Bull., U.S. 74:70-80.
1980 A large, opening-closing midwater trawl for sampling
oceanic nekton, and comparison of catches with an Isaacs-Kidd
midwater trawl. Fish. Bull., U.S. 78:529-534.
Pearcy, W.G., M.J. Hosie, and S.L. Richardson
1977a Distribution and duration of pelagic life of larvae of
Dover sole, Microstomus pacificus; Rex sole, Glyptocephalus
zachirus; and Petrale sole, Eopsetta jordani, in waters off
Oregon. Fish. Bull., U.S. 75:173-183.
Pearcy, W.G., E.E. Krygier, R. Mesecar, and F. Ramsey
1977b Vertical distribution and migration of oceanic micro-
nekton off Oregon. Deep-Sea Res. 24:223-245.
Pechenik, J. A.
1986 Field evidence for delayed metamorphosis of larval
gastropods: Crepidula plana Say, C. fomicata (L.), and Bit-
tium altematum (Say). J. Exp. Mar. Biol. Ecol. 97:313-319.
Peterson, W.T., C.B. Miller, and A. Hutchinson
1979 Zonation and maintenance of copepod populations in the
Oregon upwelling zone. Deep-Sea Res. 26A:467-494.
Pikitch, E.K., and R.L. Demory
1988 An assessment of scales as a means of aging Dover sole.
Trans. Am. Fish. Soc. 117:345-349.
Policansky, D.
1982 Influence of age, size, and temperature on metamorphosis
in the starry flounder, Platichthys stellatiis. Can J. Fish.
Aquat. Sci. 39:514-517.
1983 Size, age and demography of metamorphosis and sexual
matuartion in fishes. Am. Zool. 23:57-63.
Potthoff, T.
1984 Clearing and staining techniques. In Moser, G., et al.
(eds.). Ontogeny and systematics of fishes, p. 35-37. Spec.
Publ. 1, Am. Soc. Ichthyol. Herpeteol. Allen Press, Law-
rence, KS.
Urena, H.M.
1989 Distribution of the eggs and larvae of some flatfishes
(Pleuronectiformes) off Washington, Oregon and Northern
California, 1980-1983. Unpubl. M.S. thesis, Oregon State
Univ., Corvallis, 192 p.
Victor, B.C.
1986 Delayed metamorphosis with reduced larval growth in
a coral reef fish (Thalassoma bifasciatum). Can. J. Fish.
Aquat. Sci. 43:1208-1213.
Yoklavich, M.M.. and E.K. Pikitch
1989 Reproductive status of Dover sole, Microstomvs pacificus,
off northern Oregon. Fish. Bull., U.S. 87:988-995.
Youson, J.H.
1988 First metamorphosis. In Hoar, W.S., and D.J. Randall
(eds.). Fish physiology, vol. XI. The physiology of developing
fish, Part B. Viviparity and posthatching juveniles, p.
135-196. Academic Press, San Diego.
Abstract . - The projection of re-
source production and the effect of
removals on fisheries populations are
based on abundance estimates, partic-
ularly estimates of the most current
abundance. Monte Carlo methods
were used to investigate a size-based
method of estimating abundance for
instances where the age of caught
fish cannot be established, but where
size samples and a growth schedule
exist. Neither process variabUity (re-
cruitment dates, growth rates, and
unobserved change rates) nor sam-
pling error (catch estimation, growth
rate estimation, and relative abun-
dance sampling) adversely affected
estimation, although low sampling
intensities often decreased precision.
Abundances of recently recruited
fish too small to occur in relative
abundance samples more than once
were estimated with large uncertain-
ty. Inappropriately wide size-class
widths caused uncertain abundance
estimates of larger size-classes. How-
ever, if size-classes were of suitable
width, the abundance of fish large
enough to occur in abundance sam-
ples more than once were accurate-
ly and precisely estimated even in
cases of high process variability and
small sample sizes. Sampling gear
efficiency (catchability) coefficients
were often estimated without large
bias but imprecisely. The exponent
of the unobserved change rate (in-
cluding natural mortality) was esti-
mated precisely, but estimates were
often biased. High correlations be-
tween estimates of the unobserved
change rate and sampling gear effi-
ciencies were not often observed.
Estimation characteristics were un-
like those based on virtual population
analysis calculations. Maximum-like-
lihood estimates of the most recent
abundances were accurate and pre-
cise, yet calculations of historical
abundances were biased and extreme-
ly imprecise.
Estimating stocic abundance
from size data
Michael L. Parrack
Miami Laboratory, Southeast Fisheries Science Center
National Marine Fisheries Service, NOAA
75 Virginia Beach Drive, Miami, Florida 33149-1099
Manuscript accepted 9 March 1992.
Fishery Bulletin, U.S. 90:302-325 (1992).
Most often, the objective of fisheries
regulations is to insure that stock
abundance does not decrease or, if
abundance is low, to increase it. The
welfare of the entire stock may be of
concern, or only a part of it such as
the adult portion (spawning stock).
These objectives are obtained by
limiting yields (weight caught) to
stock growth or, in instances were
abundance is low, to less than stock
growth. Abundance estimates are the
bases for this regulation strategy. An
opinion as to whether stock abun-
dance is currently depressed or not
is based on a comparison of an esti-
mate of current abundance with esti-
mates of previous abundances. Stock
production (growth) in the immediate
future is projected from the estimate
of current abundance. Since the pro-
duction projection is the basis for the
yield limit, the estimate of current
abundance determines the yield limit.
Because it is a critical element of
regulatory responsibility, abundance
estimation methodology is of major
interest.
Most estimation methods are based
on age data. These methods specify
that the population is entirely com-
posed of unique groups of fish of
equal age (cohorts) and that all mem-
bers of a cohort grow into the first
exploitable size (recruit) instanta-
neously before fishing begins once
each year. These two requirements
rarely, if ever, occur. Most popula-
tions spawTi during several months,
or sometimes throughout the entire
year, so that annual or even monthly
cohorts do not really exist. The
growth of the young fish to sizes
large enough to be caught is a con-
tinuous process so that recruitment
is typically an ongoing phenomena.
These biological realities are often ig-
nored, and age-based analysis meth-
ods are used anyway.
Since the primary data element of
age-based methods is the number of
caught fish of each age, the ages of
caught fish must be determined.
Sometimes this requirement is dif-
ficult to satisfy. Major circuli from
differing bone densities or the chem-
ical composition of skeletal structures
(scales, fin spines, or otoliths) have
been validated as age marks in only
3.4% of age determination studies
(Beamish and McFarlane 1983). Even
in cases where indirect evidence of
validation seems ample (Kreuz et al.
1982), direct measurement of growth
from mark and recapture data can
document a very different reality
(Pikitch and Demory 1988). Collect-
ing and processing samples can be so
difficult and time consuming that
large data voids occur. Frequent
molting and the absence of bony
tissue preclude the possibility of
using hardpart ageing methods for
many invertebrates, and the technol-
ogy to determine age from somatic
tissue does not currently exist.
These problems can be avoided by
methods that model populations in
terms of size and time rather than
age and years (or months). Size-based
methods need not require that the
population be composed of age-spe-
cific cohorts nor that recruitment be
an instantaneous, one-time event.
The first size-based methods, how-
ever, are not so constructed.
302
Parrack; Estimating stock abundance from size data
303
The original technique to assess fish stocks from
size instead of age data is a stepwise double-estimation
procedure (see Pauly et al. 1987 for an example). Size-
specific catches are first transformed to age-specific
catches by using an inverted growth equation (Ricker
1975:221) or statistical estimators based on growth
data (Clark 1981, Bartoo and Parker 1982, Shepherd
1985, Hoenig and Heisey 1987, Kimura and Chikuni
1987) so that the stock is assumed to be composed of
age-specific cohorts. Size-to-age transformation meth-
ods that require size-frequencies only (i.e., growth data
are not required) are available (Macdonald and Pitcher
1979, Pauly 1982, Fournier et al. 1990), but Monte
Carlo tests have shown pronounced weaknesses in
these methods (Hampton and Majkowski 1987, Rosen-
berg and Beddington 1987, Basson et al. 1988). Vir-
tual population analysis (Ricker 1948, Fry 1949, Jones
1961, Gulland 1965, Murphy 1965) is then applied to
the transformed catch, but the system of cohort-spe-
cific catch equations is underdetermined (Agger et al.
1971, Doubleday 1975, Ulltang 1977, Pope and Shep-
herd 1982). The inclusion of auxiliary data (total fish-
ing effort, catch effort, or other relative abundance
samples) using any of several statistical procedures
(Laurec and Bard 1980; Paloheimo 1980; Anon. 1981b,
1983, 1984, 1986; Parrack 1981, 1986; Collie and
Sissenwine 1983; Deriso 1985; Pope and Shepherd
1985; Mendelssohn 1988) eliminates that problem, so
abundances can be estimated. If based on actual age
data, virtual population analysis using auxiliary infor-
mation does estimate stock abundances and fishing
mortality rates reasonably well if the natural mortal-
ity rate is known (Deriso 1985, Pope and Shepherd
1985), but if the method is used without actual age data,
its statistical characteristics are unknown. If the
population is not composed of true age-specific cohorts
or if the ageing of caught fish is problematic, the
method is not appropriate. Spawning often is too pro-
tracted to establish cohorts and fish cannot be aged
with reasonable certainty; yet because it is simple and
tractable, this method is used anyway.
Several size-based abundance estimation methods do
not employ data auxiliary to catches (Jones 1974 and
1981, Brethes and Desrosiers 1981, Lai and Gallucci
1988). Instead of using fishing effort or relative abun-
dance samples to overcome the determination problem,
they assume that the size-frequency of the catch, and
thus of the stock (and recruitment magnitudes), is
constant (in steady state). That assumption greatly
restricts the usefulness of these methods.
Three items seem important when considering stock-
abundance estimators. First, the data an estimator re-
quires often may preclude its use if such data is not
usually available. Next, since the likelihood procedure
requires one, often a sampling distribution for an ob-
served statistic is assumed even though support for the
assumption cannot be offered. The resulting estimator
thus might be entirely based on an inappropriate prob-
ability expression. Last, the statistical properties of
an estimator are of concern. An estimator may be too
imprecise to be useful unless sample sizes are unrealis-
tically large, or its bias may be too large to ignore dur-
ing estimation.
Since the method of least squares is not based on
probability theory, statistical characteristics of such
estimators are very imcertain. The likelihood procedure
tends to generate estimators with superior statistical
characteristics, but success is not guaranteed. Com-
monly, estimators of parameters of nonlinear models
are problematic. They cannot be written in closed form
so their expectations, which lead to bias and variance
expressions, cannot be derived analytically. Since the
estimator's performance characteristics cannot be
predicted, they must be established from Monte Carlo
studies. If such studies do not exist, the estimator's
usefulness is unknown.
The first size-based procedure, a least-squares esti-
mator, was developed (Beddington and Cooke 1981)
and applied to sperm whales (Anon. 1981a, Cooke and
Beddington 1982, Cooke et al. 1983b, Shirakihara and
Tanaka 1983, de la Mare and Cooke 1984) to assess the
northwestern Pacific stock (Beddington et al. 1983,
Cooke and de la Mare 1983b, Shirakihara and Tanaka
1983). It is based entirely on size-specific catches and
assumes a known adult-progeny ratio instead of using
fishing effort or other auxiliary data. The statistical
characteristics of the estimator were established with
extensive Monte Carlo studies (Cooke et al. 1983a,
Cooke and de la Mare 1983a, Shirakihara and Tanaka
1984, de la Mare and Cooke 1985 and 1987, Shirakihara
et al. 1985, de la Mare 1988).
The method of Fournier and Doonan (1987) was
derived by the likelihood method by assuming that
catch and effort are each lognormal random variables
and that the first four moments of length-frequencies
are normal random variables. Monte Carlo tests
established the estimator's ability to predict optimal
long-term fishing effort, but the errors of the stock-
abundance estimates are not described. The maximum-
likelihood method of Schnute et al. (1989) assumes that
the annual ratio of total yield to total effort is a nor-
mal random variable. The statistical characteristics of
the estimator are not described.
The method of Sullivan et al. (1990) is a least-squares
estimator based on catches, but Kalman filter method-
ology also may be used to obtain estimates (Sullivan
1989). The method does not require data other than
catches even though it is well known that, in the case
of age-based (VPA) methods, the system of catch equa-
tions without auxiliary data is not determined (Agger
304
Fishery Bulletin 90(2). 1992
et al. 1971, Doubleday 1975, Ulltang 1977, Pope and
Shepherd 1982). Sullivan et al. (1990) suggest expand-
ing the number of terms in the sum of squares to in-
clude effort and abundance indices if meaningful
weights for these auxiliary data can be found (guidance
for finding such weights is not provided). The statistical
characteristics of the estimator are not yet described.
The lack of Monte Carlo tests of the performance of
these estimators is a particular concern because,
without knowledge of their statistical behavior, little
certainty can be placed on the resulting estimates.
Some of the estimators were developed by the likeli-
hood method, but the justification for assuming the
chosen sampling distributions often seems weak or
lacking. The usefulness of those estimators that require
total fishing effort seem limited, since that statistic is
often estimated from catch and effort samples rather
than enumerated. Most of the methods estimate the
parameters of individual growth as part of the solution
vector. This seems questionable in view of findings in
a study of the separation of central moments of indi-
vidual distributions from distribution mixtures (Has-
selbald 1966), studies of the magnitude of correlation
between estimates of growth-equation parameters
(Gallucci and Quinn 1979), and of the performance of
methods that estimate growth parameters from size
distributions (Hampton and Majkowski 1987, Rosen-
berg and Beddington 1987, Basson et al. 1988). Also,
most of the methods are based on elaborate population
models, a characteristic that leads to two problems.
First, such models often include deterministic stock-
recruitment functions, and such functions are regarded
as unrealistic representations of the dynamics of fish
stocks. Second, since the population model is extensive,
it includes a large number of parameters that must be
estimated. It is well known that an exact representa-
tion of a real- world system is not possible; hence, a
suitably parsimonious model that is a useful approx-
imation with an informative structure is superior (Box
1979). The most germane variables are the current size-
specific abundances since they will determine stock pro-
duction in the immediate future.
The object of this study was to develop an abundance
estimator that would be appropriate in almost all cases,
whether or not the population is composed of cohorts,
or whether or not age data is available. Effort was
taken to write the estimation model as parsimonious
as possible, to base estimation on data commonly
collected from most fisheries, and to insure that the
correct sampling distribution was used in the likeli-
hood procedure. The bulk of the study was directed at
describing the statistical characteristics of the esti-
mator over a broad range of conditions from Monte
Carlo simulations.
Methods
Abundance estimator
An abimdance estimator was developed that uses a
model of individual growth, size-specific catches and
catch dates, and size-specific abundance observations
(sighting data, research cruise catch-per-tow, etc.). The
estimator makes three assumptions:
(1) Unobserved phenomena that change stock abun-
dance (immigration, emigration, unrecorded catch,
predation, and disease) are a (continuous) Poisson
process with combined rate z,
(2) the size of an individual on a date is a known deter-
ministic function of size on another date, and
(3) the sample average of relative abundance obser-
vations is a normally-distributed random variable
with an expectation equal to a portion of absolute
abundance.
The estimator uses a growth model to relate sizes and
dates and an abundance model to project abundance
from observed catches scaled to relative abundance
observations.
Consider T time-periods, not necessarily of equal
duration, so that 0<t<T. Within period t, relative
abundance was observed on date yt , then a catch oc-
curred on date q . The number of fish caught on date
Ct was Ct. Abimdance on the date of the relative abun-
dance observation (date yt) is of interest; let this abun-
dance (numbers of fish) be Nj. From assumption (1),
Nt+i = [Nt e-^'tlc.-y.) _ Q] e-^(i'fi-<H).
Abundance on the date of the final abundance sample
(i.e., Nt) is of most interest because stock production
in the immediate future depends on it. Writing the
above equation in terms of Nj as a time-series gives
a simple forward projection of abundance on each
relative abundance sampling date:
XT XT ^''kb'k^i-yk) -r- ^ Zk(Ck-yk)+^2i(yi»i-yi)
Nt = N^e'" -^ 2. Cue
k = t
If the unobserved change rate is assumed temporally
invariant, this simplifies to
T-l
Nt = Nt e'^'^T-yt) -i- 2! Ck e^<'^k-y,).
k = t
Each catch is subtracted separately; catching is not
assumed to occur continuously at a constant rate.
Abundance changes due to unobserved events are,
however, assumed to occur continuously at a constant
rate.
Parrack: Estimating stock abundance from size data
305
The model suggested by Chapman (1961) and Rich-
ards (1959) may be used to include growth. Letting A,
m, b, and k be parameters, s the size, and t the time
from birth, the general model
1
St = (Ai-™ - b e-"^ t)i^
is the "logistic" function of Verhulst if m = 2, the Brody
(monomolecular, von Bertalanffy) model if m = 0, and
it approaches the Gompertz function as m approaches
unity. Using the rationale of Fabens (1965) where Si
is the size at time tj and S2 is the size at time t2 , the
above growth model leads to
1
S2 = (A1-™ - (Ai-'"-Sii-"') e-Mt2-ti))T^. (1)
This satisfies assumption (3), without reference to the
actual age of individuals, by expressing size as a con-
tinuous function of time, but if growth is intermittent
or has changed, a specialized model is most appropriate.
From (1), or a more suitable model, let
I ' = the size of a fish on date yx that was size s on
date yt,
u' = the size of a fish on date yx that was size s -t- 1
on date yt,
a' = the size of a fish on date c^ that was size s on
date yt, and
b' = the size of a fish on date Cj, that was size s -t- 1
on date yt,
where V , u', a', and b' fall in size-classes /, u, a, and
b. Including size in the abundance equation gives
Nt,s =
j Nx.wdwe^fyT-y.) -I- X f Ck.wdw
(2)
;z(<^k-yt).
If size-classes are suitably narrow, the frequency of
size within size-classes tends to be proportional to size.
The frequency of size within a class is therefore approx-
imated by a trapezoid (i.e., trapezoidal integral approx-
imation). The number of fish within the size class is
s+l
F, = r f^dw = V2(s + l-s)(f, + fs^i)
s
= V2(f3-fs,l),
where s is a size class, fg is the frequency at size s,
and Fs is the number within size-class s. Let the
largest fish fall in class S:
fg = VeFs because fg+i = 0.
Rewriting gives fs = 2 Fg.
Proceeding to smaller sizes,
Fs-i = V2(fs_i + fs) =V2(fs_i + 2Fs)
sofs_i = 2(Fs_i-Fs).
Fs-2 = V2(fs_2 + fs-i) =V2(fs_2 + 2Fs_i + 2Fs)
sofs_2 = 2(Fs_2-Fs-i + Fs).
F, = V2(f3 + fs.l)
= V2(f, + 2F,,i-2F3,2+---±2Fs).
Rearrangement gives the general expression for the
frequency at size-class bounds:
fs = 2(Fs-Fs,i + F,,2-Fs.3 + • • • ± Fs).
The frequency of any size, s, within class s is also
required:
fs = f s + ^^^ (S'-S) = f s + (S'-S) (fs.i-fs).
S-l-l-S
The approximate integrals for equation (2) are thus:
\i\x = I: C
Nt,w dw =
V2(u'-0(l/ + (^'-0(lui-'7;) + (u'-0('7/+i -»];)■
or
if u > ^: I Ntw dw =
I
y2(l+l-r){r]i + il' -l)ir]ui-r]l) + m*l) +•■•
. . .V2(u'-U)(2)1u-H(u'-U)(»1u.^i-I7u))
u-l
+ I Nx.i,
306
Fishery Bulletin 90(2|. 1992
and
if b = a: I Cb „ dw =
;
V2(b'-a')fo+(a'-a)(^a+i-fe) + «b
or
;
if b < a: Ck w dw =
V2(a+l-a')(<a + (a'-a)(ca+i-«a) + ^a+i)+ • • •
. . .V2(b'-b)(2ft + (b'-b)Ub+i-<b)),
where rjs = 2(Nt,s-Nt,s+i + Nt,s+2-Nt,s+3+ • • • ±Nt,s.
and Cs = 2(Ck,s-Ck,s+i + Ck.s+2-Ck.s+3+ • • • ±Ck.s-
On a sampling date, r measures are recorded and
sample mean calculated for each size class:
Yt,s = I
Yt,s,k
According to assumption (3), the expectation of relative
abundance is
E[Yt.s] =/[/3|Y,C] = qsNt.s,
where p contains the sampling-gear efficiency coeffi-
cients (the Qs), the unobserved change rate (z), and the
abundance of each size-class on date yx (the Nt,s)-
Nt,s is as defined by (2), Y indicates a matrix of
relative abundance observations and C catches. Since
it is a mean, clearly
Yt.s~N /[/}|Y,C],
o2[Yt,3]\
(assumption 3). This implies the likelihood.
L(/5) = n (2")"''^ o[Yt,s]-' e-v^(Y,,-/[PIY,ci)'^+<,^[Y,,i_
t,s
where n is the product of the number of size-classes
and sampling dates. Maximizing its logarithm (constant
terms ignored),
lI(Y,s-/[/?|Y,C])2^oii[Yt,3]
(4)
with respect to p yields maximum-likelihood estimates
of the Qs, the Nx.s, and z. Maximization was achieved
by minimizing the negative of (4) by the "Marquardt"
method (Morrison 1960, Marquardt 1963, Conway et
al. 1970, Gallant 1975, Press et al. 1986).
This estimator is equivalent to common least-squares
if size and date variances are equal, but that restric-
tion seems unlikely. Since
o'-[Y,J = Nt,3- Var[qJ,
(5)
abundance is the dominant term. Abundance is depen-
dent on reproductive success and a mortality history.
Both are time-variant, so an assumption of equal
variances is inappropriate.
This abundance estimator possesses few restrictions.
Relative-abundance measures and catches can occur on
any date. Any number of catches, or none at all, can
occur between relative abundance samples or visa
versa. The period of data collection may be short; the
time-series may be brief. Individual growth can follow
any form. Most important, recruitment to the exploited
stock can occur continuously so that breeding (spawn-
ing) and birth (hatching) need not happen just once dur-
ing each period. Reproduction may be continuous so
age-specific cohorts need not exist. This estimator is
not a cohort analysis, but it uses similar data.
Monte Carlo tests
Each test was designed to collect a history of estimator
performance over many applications of the method in
similar circumstances. Each test was composed of
several trials. On each trial, a new exploited popula-
tion was simulated, followed by relative abundance
sampling, growth rate estimation, and catch estima-
tion. Next, p was estimated by (4) from the data col-
lected in the second step. Last, estimation error for
each element of p was calculated. The familiar mea-
sure of error, e = (ji-P), where /? is the vector of
population parameters estimated by p, was not ap-
propriate because p changed from one simulation to the
next. Error was measured by the sufficient statistic
E. = p^p. The bias of each element of p was estimated
as the average € over the n trials (Monte Carlo sam-
ples). If a particular estimate was unbiased, then
/i[€] = 1 for that parameter. The estimated error vari-
ance of each parameter, s2[E], was also calculated.
Parrack: Estimating stock abundance from size data
307
A significance level for bias larger than 10% was
found by computing the probability of the standard
normal random variable as follows:
significance level (««:;[g«;^) =
z
J fpdp, z = (G-o.9)-[s(e)/\^].
— oo
significance level {«0:;:[|H 5;}) =
oo Z
J" fp dp = 1.0 - J U dp. =
YEARS
Figure 1
Frequency of recruitment dates from one trial of uniform
recruitment simulations.
(G-1.1) - [S(e)/^].
The results between tests were statistically compared
by placing confidence intervals on the difference
between the biases (Law and Kelton 1982:319) and
using the variance ratio test {F test) to compare error
variances.
In each Monte Carlo test, the intent was to complete
trials until the estimate of bias was within a given
bound with a prescribed probabihty (Law and Kelton
1982). Several parameters were estimated, so several
biases were involved. It was too costly to confirm that
all bias estimates were trustworthy and many param-
eters were not of primary interest, so the error of last-
period total stock size (E[N(T.)]) was used as the
reference statistic. Trials were completed until
1.962 . s2(£[N(T.)]) - n < <D2,
where <t> was usually small. The 95% confidence bound
half-lengths for all parameters were computed to in-
dicate how well bias was estimated for each parameter.
The method of Schrage (1979) was used to generate
uniform random variables because it is portable and
knowTi to perform well (Law and Kelton 1982:227-
228). Normal random variables were generated by the
polar method (Law and Kelton 1982:259). The method
of Scheuer and Stoller (1962) was used to generate cor-
related bivariate normal random numbers.
In most trials, the lives of 20,000 fish were individual-
ly simulated over 20 time-periods. A history of abun-
dance and catch was created, then abundance sampling,
catch estimation, and growth parameter estimation
was simulated. Each fish possessed a unique growth
pattern and recruitment date and independently en-
countered unobserved events and fishing death. The
result of these encounters, growth rates, and recruit-
ment dates were tabulated into size-class and date-
specific matrices of numerical abundance and catch.
The sequence of events of the population simulation is
diagrammed in Appendix 1. A detailed description of
the simulation and justification of control variable
levels is given by Parrack (1990).
Von Bertalanffy growth was simulated by fixing m
of equation (1) null. For each fish, A and k of (1) were
drawn as normal random variables. The expectations
were set near those estimated for many stocks, in-
cluding Pacific cod (N.J.C. Parrack 1986), and their
coefficients of variation (cv) were set as high or higher
than common in other studies (<0.4).
Two kinds of recruitment were considered, uniform
and seasonal. The uniform pattern (Fig. 1) simulated
continuous recruitment of constant magnitude. The
date each fish recruited to the minimum size category
was drawn as a U(l,20) random variable. Seasonal
recruitment dates were drawn from normal distribu-
tions so that recruitment magnitudes varied U(l,20)
between periods and so that a typical "pulse" of yoimg
fish recruited once each period, with some recruitment
occurring continuously. The recruitment peak was
simulated to occur randomly during April, May, and
June by drawing the expected recruitment date for
each period U(0.25,0.50). Protracted and contracted
seasonal recruitment patterns were considered. Sea-
sonal protracted recruitment was simulated by draw-
ing the standard deviation of recruitment dates
U(0.20,0.33) so that 80% of recruitment occurred ran-
domly within ±3-5 months of the peak (Fig. 2). Sea-
sonal contracted recruitment was simulated by draw-
ing the standard deviation U(0.13,0.26) so that 80%
308
Fishery Bulletin 90(2), 1992
Figure 2
Frequency of seasonal, protracted recruitment dates from one
trial. Peak recruitment occurs 1 April-1 June, 80% occurs
within 3-5 months of the peak, and recruitment levels vary.
Figure 3
Frequency of seasonal, contracted recruitment dates from one
trial. Peak recruitment occurs 1 April-1 July, 80% occurs
within 2-4 months of the peak, and recruitment levels vary
20-fold between time periods.
of recruitment occurred randomly within ± 2-4 months
of the peak (Fig. 3).
The unobserved change rate was simulated both tem-
porally invariant and variant. If variant, then Zt~
U(zi,Z2) on each new trial; Zj and Zo were simulation
control variables. Catching was simulated either as a
single event that occurred once each midperiod or as
a continuous event in each period. Fishing mortality
was not imposed until period 6 so that the stock would
accumulate as soon as possible after simulation in-
itialization. In most tests the fishing mortality rate was
drawn U(Fi,F2) on each new trial; Fj and F2 were
control variables. Period-specific rates were set con-
stant over all trials in two tests to guarantee a stock
depletion caused by a rapid increase in fishing levels.
Sampling simulation included the generation of catch
estimates, growth parameter estimates, and relative
abundance measures. Populations and catches were
generated over 20 time-periods. Relative abundance
samples and catch estimates were simulated in the last
four periods only, but catches were considered to be
removed after the date of abundance samples so catch
in the last period was irrelevant to estimation (and thus
was not computed).
Size-class and date-specific catch estimates were
drawn from a Gaussian distribution with catches from
the simulator as the expectations and with variances
specified by a cv. The estimator of catches was thus
unbiased, and estimation errors (estimator variances)
were proportional to catches.
A complete simulation of growth sampling and esti-
mation was deemed too costly, so a reasonable proxy
of unbiased estimation was used. Let Aj be a growth
parameter of fish i such that Ai~N(A, a'[A]. Defining
the uniqueness in growth of fish i as tj = Aj- A, o-[A]
= ZTi2/N. Let Aj be unbiasedly measured by aj with
normal error so that ai = Ai-i-e|, ei~N(0, o-[e]). Since
ei = ai-Ai, o2[e] = lei2/N, o-[ai] = E{a,-E[ai]}2 = o2[A]
-i-o2[e]-i-2o[T,e] where the last term is null because
T and e are independent. Using the sample mean of g
fish to estimate A, a-[A] = (a~[A] + a'^[e])lg, so that
growth-parameter estimation variance is separated
into two parts, that of inherent variability from fish
to fish and that of growth measurement error. CV's
were used as simulation control input instead of vari-
ances, so growth parameter estimates were N(A,A2
(cv[A]2-i-cv[e]2)/g) random variables. In reality, all
growth parameters are estimated simultaneously. As
a rule, growth parameter estimates are highly nega-
tively correlated (Gallucci and Quinn 1979, Knight
1968, Burr 1988) with correlation coefficients often
-0.90 or less. Estimates of k and A of (1) were drawn
as normal random correlated variables (Rubinstein
1981:86) with a correlation coefficient of -0.95. The
Parrack: Estimating stock abundance from size data
309
number of fish sampled for growth (g) was specified
in the simulation indirectly as a probability level and
limit of a confidence bound. For a 1 - a level confidence
interval on A of bound length 2<t>A,
0A = Z(l-a/2)o[A]
so
= Z(l-a/2) \/A2(cv[A]2 + cv[e]2)/g,
g = Z2 I 1 - - 1 (cv[A]2 + cv[e]2) - (D2.
On each sampling date r, relative abundance samples,
l<k<r, were simulated as
Yt,s = Z Yt.s,k - r = X qs.k Nt,s - r,
k=l k=l
qs.k ~ N (q^, cv[q]2 q^^),
process variables (z, growth parameters and recruit-
ment magnitude, duration, and timing), but time trends
were not. A different unobserved change rate was
drawn for each time-period (zt~NO^, o2[z])), but the
expectation and variance were constant over time and
size. Different growth parameters were drawn for each
fish, but the expectations and variances were the same
for all fish. Recruitment magnitudes for each time-
period were drawn from a uniform distribution so time
trends were not simulated. A different recruitment
peak (i.e., the expectation of recruitment date) was
drawn for each time-period from a common expecta-
tion and variance. A duration of recruitment (i.e.,
variance of recruitment date) was drawn for each time-
period, but with the same expectation and variance.
Random variation was simulated in sampling variables
(catch estimates, growth parameter estimates, and the
Qs), but biased estimates were not simulated. Al-
though a different vector of sampling efficiencies (the
Qs) were drawn for each sampling date, the expecta-
tion and variance for each size was temporally constant.
where cv[c[] and the q^ were simulation constants. The
qs were 0.025, 0.05, 0.175, 0.225, 0.2425, and 0.25
(smallest to largest size-class) in most simulations.
Several other variations were tried, and it was found
that these constants did not affect results at all. Cv's
of q were 0.4 or less. The observation and its variance
were calculated as the maximum likelihood estimates,
Yt,s = 1 Y,,,,k/r
k
s2[Y,„3] = I (Yt,s,k-Yt,s)^/(r2-r).
k
The sample size (r) was fixed indirectly by two control
variables, the probability level and confidence-bound
width for Yt,s where a2[Yt,s] is as (5). If a 1 -a level
confidence interval on Yj g was to be of bound length
2»<J)«Nt,s*qs then:
<»Nt,sqs = 1 - Ho[Yt,s],
r = Z2(l-a/2)cv[qj2 - (1)2.
Sampling error entered the simulation as variation in
qs, not as variation in the Ytsi the variance of the
abundance index was not an input.
These simulations encompassed many possibilities,
but not all. Random variation was simulated in all
Results
The Monte Carlo tests fall into two categories: those
that investigate the influence of population process
variability on estimation errors, and those that test the
effects of sampling and data estimation. Process vari-
ability includes recruitment phenomena, growth rates,
and unobserved change due to emigration, immigra-
tion, natural death, and unrecorded catch. Sampling
variation and data estimation includes four topics:
catch estimation error, unrecorded dates of catch,
growth parameter estimation error, and variability in
sampling-gear efficiency coefficients, and thus in the
abundance indices. Each of these items were studied
separately in 14 tests.
Population process variability
For these tests, catches, dates of catch, and popula-
tion growth parameters were considered to be known,
and sampling gear efficiencies (the q's) invariant so that
all sampling variation was absent. Catches were taken
at midperiod. The probability of death due to catching
in each period was an 11(0.05,0.2) random variable, and
asymptotic size was 11.95 units (i.e., 119.5cm, with
10 cm intervals).
Recruitment patterns Uniform recruitment test re-
sults (Table 1) show little bias and high precision in
estimates of abundance and q's. Significance levels for
the hypothesis of bias = 1.0 (unbiased) versus biasi^l.O
(biased) were <0. 00005 in almost every case, but bias
310
Fishery Bulletin 90(2). 1992
Table
I
Monte Carlo tests of populations processes. Catches, dates of catch,
and population growth parameters
were assumed known. Sam- |
pling gear efficiencies (q^
were invariant. The probability of death due to catching in each period was a U(0.05,0.2) random
variable.
Catches occurred at midperiod.
Asymptotic size (i.e.,
ti[A]) was 11.95.
Recruitment patterns
Variable
growth
t 1
Variable, rapid
growth - Test 2
0.34
Variabl
f iiinw
^'W
Uniform
Protracted
Contracted
z variable
Tes
growth - Test 3
0.085
0.17
0.17
0.17
0.17
0.17
cv[A & k]
0.00
0.00
0.00
0.00
0.40
0.40
0.40
Loss rate z
0.10
0.10
0.10
U(0. 1,0.4)
0.10
0.10
0.10
Recruit levels
constant
U(l,20)
U(l,20)
U(l,20)
U(l,20)
U(l,20)
U(l,20)
Recruit dates
U(l,20)
N(ii,o-)
N(^
o')
N(t<
.o')
N(p
on
N(^
.0')
NOi
"')
M(t)
U (0.25, 0.5)
11(0.25,0.5)
U(0.25,0.5)
U (0.247, 0.5)
U(0.247,0.5)
U(0.247,0.5) 1
o(t)
U(0.2,0.33)
U(.13,0.26)
U(0.2,0.33)
U (0.2, 0.33)
U (0.2, 0.33)
U(0.2
0.33)
95% CI of biasof N(T.)
Vz width achieved 0.0024
0.0114
0.0160
0.0139
0.0434
0.0457
0.0498
Number of trials
1010
210
775
105
109
327
32
Variable Bias
s^[e]
Bias
s^e]
Bias
s'[e]
Bias
s'[e]
Bias
s^ei
Bias
s^e]
Bias
s^[€]
N(T, 3)
0.9711
0.0032
1.1319
0.0333
1.3138
0.2460
1.1569
1.0508
0.7575
0.0204
0.8594
0.0319
0.8542
0.0456
N(T, 4)
0.9923
0.0152
0.9140
0.2824
0.9780
1.5205
0.9833
0.0955
1.0697
0.0588
0.7410
0.0152
0.8815
0.0379
N(T, 5)
0.9986
0.0055
0.9710
0.0527
0.9802
0.1454
1.0191
0.0364
0.9596
0.0134
0.9087
0.0168
0.9745
0.0044
N(T, 6)
0.9941
0.0010
1.0271
0.0078
1.0632
0.0165
1.0511
0.0341
0.9964
0.0010
1.0168
0.0101
1.0051
0.0036
N(T, 7)
0.9989
0.0016
0.9651
0.0072
0.9383
0.0133
0.9742
0.0046
0.9850
0.0010
0.9892
0.0041
0.9989
0.0036
N(T, 8)
1.0049
0.0005
1.0044
0.0012
1.0018
0.0041
0.9925
0.0017
1.0041
0.0003
0.9899
0.0013
1.0364
0.0043
N(T, 9)
0.9912
0.0008
0.9933
0.0024
0.9940
0.0036
1.0003
0.0024
0.9924
0.0004
0.9911
0.0004
1.0219
0.0100
N(T,10)
1.0084
0.0009
0.9916
0.0036
0.9905
0.0087
0.9735
0.0122
1.0200
0.0007
0.9991
0.0004
1.0836
0.0241
N(T,11)
0.9752
0.2291
0.6789
1.5190
0.4585
2.1818
0.2999
5.7229
1.0590
0.4399
1.2595
0.1852
1.9661
1.5290
N(T,12)
2.3825
2.7796
3.2020
1.3185
3.7363
5.9230
N(T,13)
2.9063
5.4646
3.8776
2.0041
5.5877
13.3879
N(T,14)
3.3177
8.3135
4.8876
1.9756
9.0415
77.7329
N(T,15)
3.9400
19.2576
6.3783
2.9742
11.3780
166.9735
N(T,16)
4.4074
45.0368
8.3752
6.7119
14.2365
403.7108
N(T,17)
5.5771
120.2284
11.3020
27.2042
14.9405
905.7713
N(T,18)
5.1845
203.8775
15.4323
50.7470
18.1915
903.7181
N(T,19)
9.6012
870.1467
23.3307
358.9141
26.6017
1253.4483
N(T,20)
1.3244
222.7073
26.9926
486.5381
11.0799
640.3803
N(T,21)
4.1712
169.5628
38.8813
2180.3053
24.4425
2117.6233
N(T,22)
-1.5598
53.1271
40.2471
3481.6977
2.3077
9.6225
N(T,23)
1.9878
41.6990
48.5059
4438.4966
N(T,24)
-8.5990
2.7742
33.4447
6841.2892
N(T,25)
32.0736
4537.5995
N(T,26)
1.6141
5790.4488
N(T.)
0.9908
0.0015
1.0064
0.0071
1.0491
0.0516
1.0245
0.0053
1.0988
0.0535
1.8714
0.1781
1.1147
0.0206
q( 3)
1.0331
0.0039
0.9010
0.0209
0.8103
0.0410
0.8891
0.0268
1.3544
0.0543
1.2201
0.0676
1.4094
0.1195
q( 4)
1.0178
0.0036
0.8844
0.0085
0.7299
0.0185
0.8728
0.0091
1.1304
0.0161
1.3772
0.0487
1.1669
0.0297
q( 5)
0.9977
0.0030
0.9787
0.0067
0.9292
0.0181
0.9581
0.0058
1.0987
0.0118
1.2252
0.0325
1.0611
0.0064
q( 6)
1.0090
0.0012
1.0212
0.0033
1.0177
0.0091
1.0161
0.0043
1.0478
0.0031
1.1014
0.0104
1.0094
0.0043
q( ^)
0.9870
0.0011
1.0758
0.0062
1.1436
0.0137
1.0855
0.0073
1.0110
0.0023
1.0738
0.0063
0.9641
0.0025
q( 8)
1.0133
0.0015
0.9290
0.0046
0.8422
0.0100
0.8780
0.0157
0.9706
0.0056
1.0444
0.0033
0.9198
0.0114
q( 9)
1.0311
0.0039
1.0462
0.0199
1.0381
0.0275
0.9848
0.0437
0.9363
0.0100
1.0057
0.0044
0.9531
0.0258
q(10)
0.9717
0.0190
0.9655
0.1925
0.9282
0.2356
0.7786
0.3357
0.8790
0.0230
0.9282
0.0058
0.8257
0.0304
q(ll)
0.7365
0.1902
0.6016
1.5307
0.8387
1.7031
0.1939
2.5069
0.6449
0.3805
0.4157
0.0087
0.4976
0.0643
q(12)
0.4371
0.0403
0.3132
0.0056
0.3373
0.0272
q(13)
0.3349
0.0263
0.2554
0.0047
0.2187
0.0129
q(14)
0.2587
0.0174
0.1940
0.0588
0.1427
0.0084
q(15)
0.1777
0.0124
0.1510
0.0221
0.0983
0.0038
q(16)
0.1242
0.0051
0.1171
0.0049
0.0654
0.0019
q(n)
0.0887
0.0036
0.0822
0.0029
0.0437
0.0015
q(18)
0.0589
0.0020
0.0580
0.0012
-0.2827
2.9493
q(19)
0.0394
0.0022
0.0365
0.0009
0.7259
13.8313
q(20)
-0.0925
1.0754
0.0247
0.0003
-0.1340
0.1934
q(21)
0.2614
2.9476
0.0165
0.0001
-0.0174
0.0165
q(22)
-0.1324
0.5450
0.0200
0.0473
-9.6131
278.5656
q(23)
-0.3803
3.0233
0.5313
53.3122
q(24)
0.0163
0.0000
0.0122
0.2028
q(25)
0.0177
0.0001
-0.0096
0.0421
q(26)
-2.1688
167.9745
z
1.5847
0.0715
1.6436
0.1442
1.8264
0.2772
1.9012
0.3023
2.7344
0.5514
0.9394
0.1280
exp(z)
1.0608
0.0008
1.0676
0.0017
1.0877
0.0034
1.0959
0.0036
1.1926
0.0075
0.9946
0.0013
Parrack: Estimating stock abundance from size data
31 1
was not large; the significance levels for the hypotheses
of bias < 10% were >0.99995 for all estimates. Preci-
sion was not a problem, although error variances were
not zero.
Since sampling variation was zero and the estima-
tion model encompassed all of the population char-
acteristics simulated, the estimation bias and impreci-
sion were unexpected. The only possible source of that
error are the integral approximations required in
estimation.
The estimate of the unobserved change rate (z) was
biased high by about 60% and its error variance was
large. It entered estimation in an exponent, so the term
in the model was the exponent of z (the reciprocal of
"survival" from unobserved change), not z. The error
term was again computed on the exponent of the
estimate of z instead of z. The estimated bias was ten
times lower and the error variance was several orders
of magnitude less. This result proved consistent in all
tests of population processes.
Partial correlation coefficients between parameter
estimates did not exhibit meaningful trends. Although
some adjoining abundance estimates were correlated
(probably because the abundances were), evidence of
other correlations were absent. Estimates of z were not
correlated with the estimates of the q's or abundances;
estimates of the q's were not correlated with abun-
dance estimates. This result proved consistent. The cor-
relation matrices for this and following tests are not
shovra for the sake of brevity but are presented in Par-
rack (1990).
The two seasonal recruitment tests (protracted and
contracted patterns) show increased bias and impreci-
sion. As the recruitment frequency contracted, bias and
error variance-of-abundance estimates of the smallest
and largest size-classes increased. This problem was
worst for the largest size-class. Estimates of the q's
also degraded.
Unobserved change rate The estimator assumes
that the rate of change due to phenomena that cannot
be observed (natural death, migration, unrecorded
catch) is constant over periods. Since the assumption
is undoubtedly false, estimation errors resulting from
assigning a U(0.1,0.4) random variable to z for each
period were investigated. Other simulation character-
istics were as in the seasonal protracted recruitment
test. The 95% confidence intervals on the difference
of abundance estimation bias between this test and the
protracted recruitment test included zero for size-
classes 3 and 4, most others, and total abundance.
Error variances were likely equal for size-classes 3, 4,
and total abundance (SL 0.005, SL<0.000, SL<0.000).
Correlations between estimates were low. A fourfold
random variability in z did not affect estimation at all.
Growth Three tests consider highly variable growth.
The cv's of asymptotic size and k were 0.4. Test 1
simulated the same growth parameters as the pro-
tracted recruitment test (k 0.17), test 2 considered
growth twice as rapid (k 0.34), and test 3 growth twice
as slow (k 0.085). All other simulation control variables
are the same as the protracted recruitment test, so the
results are comparable.
The results of all three tests were very similar. All
reflected the high variation of asymptotic size: the
parameter vector included size-classes larger than the
asymptote. Abundance and q estimates of these classes
(12 and larger) were worthless; huge bias and impreci-
sion occurred. Abundances of smaller size-classes in all
three tests were more precise than in the protracted
recruitment test where growth was not variable. Biases
and error variances of abundance and q estimates for
size-class 1 1 and smaller were very similar in the three
tests; performance seemed unaffected by growth rates.
The exponent of z was again estimated much better
than z in all three tests; estimates were precise al-
though significant bias was present in the case of rapid,
variable growth. Evidence of correlated estimates was
absent. The introduction of an extremely high level of
variation on individual growth parameters did not
negatively affect estimates.
Data estimation and sampling
Errors attributable to sampling and the compilation of
various input statistics were studied in seven tests.
Catches are rarely censused as assumed by the esti-
mator; estimates are usually the available statistics.
The estimator models the dates of each catch, yet catch
statistics are usually summed over an interval of dates.
Growth rates are assumed to be known, but that is
never possible; growth parameters must be estimated.
Last, the variability in sampling-gear efficiency coef-
ficients, and thus in the abundance indices, is also a
source of uncertainty.
Most of the simulation control variables in these
seven tests were the same as in the protracted recruit-
ment test. Asymptotic size was 11.95, growth k was
0.17, the unobserved loss rate (z) was fixed at 0.1, and
the seasonal, protracted recruitment pattern was
employed; thus recruitment levels varied 20-fold be-
tween periods. Catching was simulated differently than
in the protracted recruitment test. Catching was con-
tinuous (see Appendix 1, step 4) instead of a single sub-
traction at midperiod, and the fishing mortality rate
(F) was a U(0. 1,0.4) random variable.
Catch dates A single scenario was used to investi-
gate the importance of recording each catch date and
modeling each catch separately. The summed catch
over each period was assumed to be known, but not
312
Fishery Bulletin 90(2). 1992
Table 2
Monte Carlo tests for the effects of samp
ing variation.
^[A] = 11.95, M[k]
= 0.17, Z
= 0.1, and seasona
, protracted recruitment was 1
simulated. Catching was simulated as a continuously occurring event. The instantaneous rate of fishing mortality was a U(0
1,0.4)
random variable.
Unknown
Catch
Growth parameter
measurement error
Relative abundance
with process
cv[A&k]
catch date
estimation error
40% Error
15% Error
variance
cv[q] 0.4, r3
cv[q]0.2, rl6
0.00
0.00
0.00
0.00
0.00
0.20
0.00
Catch estimation
catch dates
absent
absent
absent
absent
absent
absent
absent
cv[C(t,s)]
0.00
0.40
0.00
0.00
0.00
0.00
0.00
Growth estimation
cv [error]
0.00
0.00
0.40
0.15
0.15
0.00
0.00 1
precision level
-
-
-
-
0.02
-
probability level
-
—
0.95
-
fish sampled, g
]
1
1
1
601
]
]
Sampling efficiency
cv[q(s)]
0.00
0.00
0.00
0.00
0.00
0.40
0.10
precision level
-
-
—
—
0.50
0.05
probability level
-
-
-
—
0.95
0.95
sample size, r
]
1
1
]
1
3
16
95% CI of bias of N(T.)
Vzwidth achieved
0.0155
0.0197
0.4789
0.0210
0.0198
0.0354
0.0198
Number of trials
101
84
92
198
79
200
52
Variable
6
sn€]
s
s=[6]
s
s'le]
6
sne]
6
s^e]
6
s^e]
E
snej
N(T, 3)
1.1081
0.0295
1.1817
0.0646
1.8710
217.1297
1.1439
0.1424
0.9947
0.0355
1.3381
5.2758
1.0972
0.0768
N(T, 4)
0.9710
0.3532
1.0314
0.2057
1.0034
3.4342
0.9228
1.0249
1.0184
0.0767
0.9247
3.0564
0.9288
0.0823
N(T, 5)
1.0009
0.0253
0.9350
0.0398
1.1353
3.4290
1.0028
0.1066
1.0026
0.0108
0.8733
0.2110
0.9128
0.0471
N(T, 6)
1.0162
0.0052
1.0197
0.0034
1.3096
6.9835
1.0375
0.0511
0.9902
0.0011
1.0156
0.0289
1.0092
0.0056
N(T. 7)
0.9740
0.0051
0.9660
0.0055
1.3544
12.7702
0.9675
0.0170
0.9953
0.0009
0.9530
0.0430
0.9191
0.0119
N(T, 8)
1.0000
0.0010
0.9956
0.0079
1.6209
17.9277
1.0437
0.0598
1.0019
0.0010
1.0243
0.0194
1.0239
0.0045
N(T, 9)
1.0003
0.0007
0.9939
0.0061
1.5861
11.9433
1.0458
0.0952
0.9959
0.0008
0.9703
0.0079
0.9688
0.0024
N(T,10)
0.9914
0.0045
0.9973
0.0012
2.6909
26.6738
1.3480
1.1729
1.0096
0.0019
1.0077
0.0112
1.0021
0.0004
N(T.ll)
0.8191
3.9963
0.7914
1.2824
3.4334
111.2210
1.2774
5.3882
1.1619
0.3091
1.1734
1.1168
0.9568
0.1516
N(T,12)
2.3955
1.6866
N(T,13)
3.4498
5.9767
N(T,14)
3.2299
16.5785
N(T,15)
3.6964
41.8706
N(T,16)
0.4291
92.2468
N(T,17)
6.5117 174.0005
N(T,18)
1.5430
10.9191
N(T,19)
18.1902 768.6451
N(T.)
1.0242
0.0063
1.0259
0.0086
1.4105
11.4628
1.0416
0.0226
1.0542
0.0081
1.0345
0.0653
0.9799
0.0053
q(3)
0.9180
0.0200
0.8754
0.0291
1.3725
9.3538
0.9143
0.0715
1.0346
0.0290
0.9656
0.1063
0.9661
0.0460
q( 4)
0.8578
0.0062
0.8729
0.0126
1.5114
25.7585
0.8825
0.0451
1.0391
0.0157
0.8933
0.0776
0.9397
0.0124
q( 5)
0.9715
0.0051
0.9725
0.0103
2.0644
87.9064
1.0305
0.3386
1.0472
0.0098
1.0162
0.0709
1.0402
0.0091
q( 6)
1.0046
0.0043
1.0141
0.0065
1.9815
47.7841
1.0593
0.1689
1.0315
0.0021
1.0467
0.0553
1.0629
0.0045
q( 7)
1.0783
0.0059
1.0679
0.0065
2.0977
54.6183
1.0887
0.0294
1.0112
0.0022
1.1524
0.0767
1.1242
0.0159
q( 8)
0.9093
0.0058
0.9342
0.0061
1.5326
20.5395
0.9387
0.0249
0.9900
0.0058
0.9596
0.0590
0.9590
0.0075
q( 9)
1.0159
0.0190
1.0447
0.0239
1.7836
46.8489
1.0023
0.0620
0.9735
0.0082
1.1949
0.1016
1.1.582
0.0194
q(10)
1.0151
0.1572
1.0147
0.0816
1.4402
25.2990
0.8819
0.1195
0.9388
0.0293
1.0083
0.1004
0.9874
0.0182
q(ll)
0.5695
0.8583
0.8320
1.4223
1.4677
34.9181
0.9667
0.5829
0.6416
0.0701
0.9200
0.6084
0.8700
0.1365
q(12)
0.3774
0.0303
q(13)
0.2138
0.0091
q(14)
0.1017
0.0042
q(15)
0.0488
0.0012
q(16)
-3.1478 425.1308
q(17)
0.0866
0.0670
q(18)
-0.0707
0.0967
q(19)
1.2251
4.4242
z
0.8137
0.1022
1.6881
0.2587
0.2301
25.5499
1.0396
0.0219
0.1835
0.0000
0.8525
1.0626
0.8542
0.2896
exp(2)
0.9820
0.0010
1.0736
0.0031
0.9828
0.0416
1.0527
0.0067
1.0898
0.0022
0.9913
0.0104
0.9869
0.0029
Parrack: Estimating stock abundance from size data
313
the dates of the catches. The accumulated catch each
period was assigned to the midpoint of each period for
estimation. The results (Table 2) were almost identical
with those of the protracted recruitment test (Table
1). The 95% confidence interval (Welch 1938) on the
difference between total-abundance estimation bias of
the protracted recruitment test and this test included
zero (-0.0014 to 0.0372). The error variances were
very similar (0.0072 and 0.0063). Estimates were not
correlated. The absence of exact catch dates did not
affect estimation.
Catch estimation error The effects of estimating
catches rather than enumerating them were investi-
gated by drawing size-class-specific catch estimates as
normal random variables with expectation C(t, s) and
variance (cv[C] ■ C(t,s))2. This simulated unbiased
catch estimation and estimation error proportional to
catches. A large degree of catch estimation uncertainty
was imposed (cv[C] = 0.40). Simulation control variables
were the same as in the catch date test and the pro-
tracted recruitment test. Results were also similar. The
bias of total abundance estimates was about the same
for all three tests and the error variances were nearly
so. Correlated estimates were not evident. Confidence
intervals (95%) on the difference in bias between this
test and the protracted recruitment test included zero
for all size-classes and total abundance. The error
variance for size-class 3 was different (SL<0.0005) and
might have been different for size-class 4 (SL 0.052),
but probably not for total abundance (SL 0.142) and
all others. Imprecise catch estimates did not impact
bias or error variance.
Growth parameter estimation error The effect of
imprecise growth parameter estimates was also con-
sidered. Estimates of growth parameters were simu-
lated as normal correlated random variables with
expectations equal to those of the population. As ex-
plained in the Monte Carlo methods section, the vari-
ance of a growth parameter estimate is composed of
two parts: process variation due to variant individual
growth, and growth measurement error. Simulation
control constants were therefore the cv of A and of k,
the growth measurement error cv, the two constants
required to compute the sample size used to estimate
the growth parameters, and the correlation coefficient
between estimates (-0.95). Simulation constants were
as in the catch date test except those related to growth
parameter estimation.
Three tests were carried out, two without process
variation. First, the effect of two measurement error
cv's was studied in the absence of growth variability.
The sample size was set at one fish in these two tests
so affects due to measurement error would be magni-
fied. Then, the combined effect of process variation and
estimation error was considered.
In the first test with extremely imprecise growth
parameter estimates (cv 0.4), Monte Carlo trials were
carried out until it became obvious that little more in-
formation would be gained with further computations.
Error variances were huge (Table 2). Only the expo-
nent of z was reasonably estimated. Many estimates
were correlated, particularly those of z with those of
sampling-gear efficiency coefficients. Even without in-
dividually variant growth rates (an unlikely prospect),
large growth-parameter measurement error created
significant uncertainty.
The second test simulated 15% measurement error.
A 95% confidence interval on the difference between
the bias of total abundance estimates between this and
the protracted recruitment test included zero, but the
error variances were probably different (SL< 0.0001);
most error variances were higher. Bias was imaffected
although error variance approximately doubled. The
estimates did not seem correlated. The introduction
of a 15% growth measurement error increased error
variances but did not affect bias.
The third test simulated both process error (cv 0.2)
and 15% growth measurement error, but with a sam-
ple size such that 95% confidence intervals on the
estimate of the expectation of growth parameters were
with precision ±2% (g = 601 fish). The 95% confidence
interval on the difference in bias of total abimdance
estimates between this test and the protracted recmit-
ment test included zero ( - 0.0220 to 0.0238) although
error variances perhaps differed (SL=:0.05). Estimates
were not correlated. Apparently 15% (or less) measure-
ment error, even with natural growth variation, min-
imally affects estimation.
Gear efficiency variability The estimator is derived
from the density function of relative abundance obser-
vations (Y), but the effect of Y variability on estima-
tion error was not of large interest. The variance of
Y is o^lYt s] = Nt s2cv[q]". The dominant term is the
square of abundance, so as abundance increases, o-
[Yts] increases. This may be dampened a bit by an in-
crease in q with size, but the dominant factor in the
variance expression for the observations is abundance.
Abundance levels cannot be controlled or anticipated
beforehand, so knowledge of the effect of Y variabil-
ity is of little value. Knowledge of the effect of q vari-
ability is useful, however, since care may be taken in
the selection and design of sampling gear.
Studies that document the statistics necessary to
calculate the variability of relative abundance sam-
pling-gear efficiencies are not common. Studies of
commerical fishery statistics offer different but useful
information. Yield is a portion of biomass; the pro-
314
Fishery Bulletin 90(2), 1992
portion is the product of fishing effort and q for the
fishing method. Since yield is the product of q, effort,
and biomass, then yield-per-effort equals the product
of q and biomass and q is yield-per-effort divided by
biomass. It then follows that the cv's of q and yield-
per-effort are equal. The cv of yield-per-effort of the
Pacific halibut longline fishery is estimated to be 0.02
(Quinn et al. 1982), and that of Newfoundland flounder
trawlers on the Grand Bank (Smith 1980) is estimated
to be about the same. The levels used in these simula-
tions (0.4 and 0.2) are about an order of magnitude
higher than those.
Effects of the variability in q on estimation errors
were investigated in three tests. All simulation con-
stants were as in the protracted recruitment test ex-
cept those related to abundance sampling. Simulation
control constants were cv[q] and the two constants re-
quired to compute the sample size. Although they were
probably unrealistically large, a cv[q] of 0.4 was used
in the first test and 0.2 was used in the second.
First, the impact of extreme variability (cv 0.4) and
extremely light sampling was tested. The sample size
(r 3) was such that a 95% confidence interval on relative
abundance was within ± 50% of the expectation. The
extremely high cv[q] and low relative-abundance sam-
ple size were not reflected in error variances as much
as expected (Table 2), but error variances were higher
than those of the protracted recruitment test. Most
abundance estimates were biased by less than 10%.
Estimates were not correlated.
Next, the cv[q] was reduced to 0.1 and the sample
size was increased so that a 95% confidence interval
on relative abundance was within ± 5% of the expec-
tation (r 16). The result was very similar to those of
the first test except error variances were much lower.
Biases of abundance estimates were ± 10% or less and
estimates were not correlated.
There was no evidence that high variation in the
qs biased abundance estimates even if sample sizes
were insufficient, but error variances were affected.
Error variance was considerably reduced with reason-
able sample sizes.
Bias
The results of these experiments (Tables 1 and 2) show
that abundances and gear efficiencies (q's) of the
smallest and largest size-classes were often biased. Bias
did not occur with uniform, constant recruitment and
no sampling variation, but as process and sampling
variation increased, bias in estimates of the smallest
and largest sizes became pronounced.
Each expected value is a proportion of calculated
abundance. The abundance calculation sums future
catches (data), last-period abundance (estimates), and
an amount for unobserved changes (estimate). Future
catches and terminal abundance are thus the major
components of each projection. Both catch and final
abundance must be integrated over size. The integra-
tion of catch over size at each catch date following the
date of the expected value is required. The integration
of abundance over size on the date of the final relative
abundance sample is also necessary. All integrals are
approximated, so these calculations are the source of
the bias. The amount of error incurred at each integra-
tion depends on how well the trapezoidal rule approx-
imates the size distribution. Since the size frequency
within a size-class is never smooth, the approximation
will be in error with the amount depending on the
degree of smoothness within the size-class. If growth
is variable or the number of fish is small, clumps in
size frequencies can result from chance alone, but the
major factor is the growth and recruitment pattern
combination.
Narrowing the size-classes eliminates this problem.
If they are narrowed enough to eliminate clumping
caused by the particular recruitment frequency con-
traction, the size frequency within size-classes will be
smooth and the trapezoidal approximation will be ac-
curate. The seasonal contracted pattern of recruitment
test 3 was again used to demonstrate this. An asymp-
totic size of 120 cm was simulated with recruitment
occurring at 20 cm. First, it was assumed that the data
were collected in 20 cm intervals so that the asjonptotic
size was 6 and the recruitment size was class 1. In the
second case, it was assumed that data were collected
in 2 cm groups so that the asymptotic size was 60 and
the recruitment size was class 10. The unobserved
change rate was set at 0.2 in both tests, and all other
simulation control variables were as in the contracted
recruitment test.
Ninety-two trials were required to obtain a 95% con-
fidence interval half-length of 0.05 on the bias of total
abundance in bias test 1 with 20 cm interval data.
Estimates of the smallest and largest size-class abun-
dances were biased and the error variances were very
large (Table 3), particularly for the largest size-class.
The estimate of the survival from unobserved change
(z) was, however, reasonably accurate and precise.
Only 16 trials were required to obtain a 95% con-
fidence interval half-length of 0.03 on the bias of total
abundance in bias test 2 with two-unit size-interval data
because the error variances were very low. Estimates
of the first three size groups were probably biased by
10% or more, but the rest were not. Only six of the
47 estimates were probably biased at all (0.95 level).
The estimate of the exponent of z was also not biased.
Although the matrix was too large to be included (194
rows and columns), there was no evidence that esti-
mates were correlated.
Parrack: Estimating stock abundance from size data
315
Table 3
Biases for 20 cm size
-class width data (bias test 1,
92 trials) versus
Diases for 2
cm size
class width data (bias test 2,
16 trials).
Bias
Significance levels
Bias
Significance levels
Estimates
Estimates
Tj;
95% CI
V2-width
HO:Bias^0.9
HA:Bias<0.9
HO:Bias<l.l
HA;Bias>l.l
Bias
test Variable
95% CI
HORiasSO 9
HOBias<l 1
DldS
test Variable
Bias
s=[e]
Bias
sMG]
V2-width
HA:Bias<0.9
HA:Bias>l.l
1
N (20-39)
1.2941
0.1332
0.0746
1.0000
0.0000
1
q (20-39)
0.8296
0.0207
0.0294
0.0000
1.0000
2
N(20-21)
1.8093
0.5273
0.3558
1.0000
0.0000
2
N (22-23)
1.5883
0.2529
0.2464
1.0000
0.0001
2
q(20-21)
0.6373
0.0560
0.1160
0.0000
1.0000
2
N (24-25)
1.5190
0.3576
0.2930
1.0000
0.0025
2
q (22-23)
0.6924
0.0504
0.1101
0.0001
1.0000
2
N (26-27)
1.1204
0.0790
0.1377
0.9991
0.3856
2
q(24-25)
0.7464
0.0621
0.1221
0.0068
1.0000
2
N (28-29)
0.9746
0.0578
0.1178
0.8927
0.9815
2
q (26-27)
0.9410
0.0465
0.1057
0.7766
0.9984
2
N (30-31)
0.8908
0.0124
0.0545
0.3699
1.0000
2
q (28-29)
1.0818
0.0616
0.1216
0.9983
0.6151
2
N (32-33)
1.0826
0.0440
0.1028
0.9997
0.6303
2
q(30-31)
1.1408
0.0234
0.0749
1.0000
0.1431
2
N (34-35)
0.9963
0.0002
0.0073
1.0000
1.0000
2
q (32-33)
0.9545
0.0300
0.0849
0.8959
0.9996
2
N (36-37)
1.0142
0.0032
0.0276
1.0000
1.0000
2
q(34-35)
0.7751
0.0463
0.1055
0.0101
1.0000
2
N (38-39)
0.9924
0.0005
0.0109
1.0000
1.0000
2
q(36-37)
0.6608
0.0531
0.1129
0.0000
1.0000
2
q(38-39)
0.7494
0.0379
0.0954
0.0010
1.0000
1
N (40-59)
1.0071
0.0234
0.0313
1.0000
1.0000
1
q (40-59)
1.0050
0.0125
0.0228
1.0000
1.0000
2
N (40-41)
1.0052
0.0003
0.0089
1.0000
1.0000
2
N(42-43)
1.0009
0.0000
0.0032
1.0000
1.0000
2
q (40-41)
0.9352
0.0099
0.0487
0.9216
1.0000
2
N (44-45)
1.0013
0.0001
0.0041
1.0000
1.0000
2
q (42-43)
1.0218
0.0205
0.0701
0.9997
0.9856
2
N (46-47)
0.9977
0.0001
0.0042
1.0000
1.0000
2
q (44-45)
1.0357
0.0093
0.0473
1.0000
0.9961
2
N (48-49)
1.0060
0.0008
0.0136
1.0000
1.0000
2
q (46-47)
0.8792
0.0154
0.0608
0.2513
1.0000
2
N(60-51)
0.9966
0.0004
0.0097
1.0000
1.0000
2
q (48-49)
0.7148
0.0542
0.1141
0.0007
1.0000
2
N (52-53)
1.0022
0.0000
0.0030
1.0000
1.0000
2
q(50-51)
0.7373
0.0333
0.0894
0.0002
1.0000
2
N (54-55)
0.9985
0.0001
0.0046
1.0000
1.0000
2
q (52-53)
0.8691
0.0090
0.0464
0.0960
1.0000
2
N (56-57)
0.9998
0.0000
0.0013
1.0000
1.0000
2
q (54-55)
1.0389
0.0438
0.1025
0.9961
0.8786
2
N (58-59)
0.9969
0.0001
0.0047
1.0000
1.0000
2
q (56-57)
1.0358
0.0056
0.0365
1.0000
0.9997
2
q (58-59)
0.8082
0.0187
0.0670
0.0036
1.0000
1
N (60-79)
0.9741
0.0288
0.0347
1.0000
1.0000
1
q (60-79)
0.9049
0.0193'
0.0284
0.6333
1.0000
2
N (60-61)
1.0046
0.0001
0.0045
1.0000
1.0000
2
N (62-63)
0.9975
0.0002
0.0077
1.0000
1.0000
2
q(60-61)
0.6377
0.0760
0.1351
0.0001
1.0000
2
N (64-65)
0.9998
0.0000
0.0018
1.0000
1.0000
2
q (62-63)
0.8143
0.0401
0.0982
0.0434
1.0000
2
N (66-67)
0.9998
0.0001
0.0059
1.0000
1.0000
2
q (64-65)
0.9890
0.0204
0.0700
0.9936
0.9991
2
N (68-69)
1.0026
0.0002
0.0062
1.0000
1.0000
2
q (66-67)
1.0381
0.0257
0.0785
0.9997
0.9388
2
N(70-71)
0.9959
0.0002
0.0071
1.0000
1.0000
2
q (68-69)
0.8908
0.0447
0.1036
0.4306
1.0000
2
N (72-73)
1.0034
0.0001
0.0060
1.0000
1.0000
2
q(70-71)
0.7287
0.0583
0.1183
0.0023
1.0000
2
N (74-75)
1.0025
0.0001
0.0050
1.0000
1.0000
2
q (72-73)
1.0168
0.0152
0.0604
0.9999
0.9965
2
N (76-77)
1.0039
0.0003
0.0081
1.0000
1.0000
2
q (74-75)
1.0449
0.0163
0.0626
1.0000
0.9577
2
N (78-79)
1.0004
0.0001
0.0049
1.0000
1.0000
2
q (76-77)
0,8313
0.0415
0.0998
0.0886
1.0000
2
q (78-79)
0.7050
0.0727
0.1321
0.0019
1.0000
1
N (80-99)
1.0517
0.0745
0.0558
1.0000
0.9553
1
q (80-99)
0.7632
0.2257
0.0971
0.0029
1.0000
2
N(80-81)
0.9982
0.0003
0.0080
1.0000
1.0000
2
N (82-83)
1.0009
0.0001
0.0045
1.0000
1.0000
2
q (80-81)
0.9684
0.0392
0.0970
0.9167
0.9961
2
N (84-85)
1.0037
0.0002
0.0067
1.0000
1.0000
2
q (82-83)
1.0334
0.0214
0.0717
0.9999
0.9657
2
N (86-87)
1.0013
0.0000
0.0010
1.0000
1.0000
2
q (84-85)
0.6644
0.0478
0.1071
0.0000
1.0000
2
N (88-89)
0.9995
0.0000
0.0016
1.0000
1.0000
2
q (86-87)
0.9742
0.0679
0.1277
0.8727
0.9732
2
N (90-91)
0.9992
0.0001
0.0054
1.0000
1.0000
2
q (88-89)
0.7991
0.0914
0.1481
0.0909
1.0000
2
N (92-93)
0.9994
0.0000
0.0009
1.0000
1.0000
2
q (90-91)
0.7715
0.0667
0.1266
0.0233
1.0000
2
N (94-95)
0.9963
0.0001
0.0041
1.0000
1.0000
2
q(92-93)
0.9841
0.0515
0.1148
0.9244
0.9761
2
N (96-97)
0.9989
0.0000
0.0017
1.0000
1.0000
2
q (94-95)
0.7314
0.1194
0.1749
0.0294
1.0000
2
N (98-99)
0.9974
0.0000
0.0022
1.0000
1.0000
2
q (96-97)
0.8420
0.0390
0.1000
0.1279
1.0000
2
q (98-99)
0.9202
0.1227
0.1835
0.5856
0.9726
1
N(100-119)
6.4942
254.5641
3.2603
0.9996
0.0006
1
q(100-119)
0.1117
0.5453
0.1509
0.0000
1.0000
2
N(lOO-lOl)
1.0003
0.0000
0.0007
1.0000
1.0000
2
N(102-103)
0.9997
0.0003
0.0091
1.0000
1.0000
2
q(lOO-lOl)
0.7954
0.1388
0.1952
0.1467
0.9989
2
N(104-105)
0.9979
0.0003
0.0105
1.0000
1.0000
2
q(102-103)
0.9848
1.5099
0.6680
0.5983
0.6323
2
N(106-107)
0.9997
0.0000
0.0005
1.0000
1.0000
2
q(104-105)
1.0579
0.0507
0.1224
0.9943
0.7497
2
N(108-109)
1.0009
0.0000
0.0005
1.0000
1.0000
2
q(106-107)
0.8840
0.0867
0.1923
0.4350
0.9862
2
N(llO-lll)
0.9981
0.0000
0.0035
1.0000
1.0000
2
q(108-109)
0.8992
0.0936
0.2682
0.4976
0.9289
2
N(112-113)
1.0000
0.0000
0.0005
1.0000
1.0000
2
q(llO-lll)
1.0052
0.1531
0.3835
0.7047
0.6860
2
q(112-113)
1.0121
0.0004
0.0238
1.0000
1.0000
1
N(T.)
1.1808
0.0592
0.0497
1.0000
0.0007
2
N(T.)
1.0388
0.0049
0.0342
1.0000
0.9998
1
exp(z)
0.9716
0.0135
0.0237
1.0000
1.0000
2
exp(z)
0.9783
0.0169
0.0637
0.9920
0.9999
316
Fishery Bulletin 90(2). 1992
Estimates of the first three size-classes were both
biased and imprecise. Poor estimates of the smallest
few size-classes were expected. These classes lacked
a catch history at the time of the last sample, so these
estimates of abundance (at the time of the last relative
abundance sample) were a function of the last-period
relative abundance sample only.
Examples
Two tests were used to discover what might be ex-
pected when assessing populations with no periodicity
in recruitment at all; recruitment dates were complete-
ly protracted imiformly through time (Fig. 1). Most con-
trol variables were the same in the two tests. Data were
assumed to be available in two-unit size intervals. A
120-unit asymptotic size fell in size-class 60, and a
30-unit recruitment size in class 15. The growth param-
eter k was left at 0.17. Continuous fishing was simu-
lated; the fishing mortality rate (F) for each period was
drawn from a 11(0.3,0.8) distribution. The expectations
of sampling efficiencies (q^) were arbitrarily chosen so
that their regression on size was sigmoid, reaching an
asymptote at size-class 30 (0.028, 0.031, 0.033, 0.038,
0.044, 0.053, 0.069, 0.101, 0.153, 0.190, 0.218, 0.234,
0.242, 0.247, 0.249, and 0.250). Catch estimates were
simulated to be imprecise (cv 0.4). A 10% growth mea-
surement error was simulated. Sampling intensities
were the same in both tests; sample sizes for growth
parameter estimates and for relative abundance obser-
vations were such that a 95% CI was of width ±5%.
Although the levels of population processes were the
same in both tests, the amoimt of process variability
was much higher in test 2. Normal growth variability
was simulated in test 1 and extreme variability in test
2. The rate of unobserved change in test 1 was con-
stant, but varied three-fold in test 2. The variance of
sampling efficiencies was set one order of magnitude
larger than that observed for commercial fishing gear
in test 1 and twice that in test 2.
Error variances-of-abundance estimates were very
low in the case of normal process variability (Table 4).
Estimates of all but the smallest six size-classes were
biased by 10% or less, if at all, and were precise. Bias
(more than 10%) and imprecision of the smallest six
size-class estimates was expected because the smaller
fish were barely represented in the catch and appeared
in the relative abundance samples just once. Estimates
of the sampling efficiencies (the qs) tended to be im-
precise. Some were biased from 20% to 30% and a few
even more. The estimate of survival from unobserved
change was biased low (about 15%), yet precise. Esti-
mates did not tend to be correlated. The correlations
between the estimate of the unobserved change rate
(z) and other estimates (Table 5), particularly of the
C[s , were of interest because other studies found corre-
Table 4
Monte Carlo test results for two
examples of constant, uniform recruitment.
Loss rate z
Example
1, Normal process variability
Examp
e 2, High process variability
0.20
U (0.2, 0.6)
Growth cv[A & k]
0.20
0.40
Growth estimation
cv [error]
0.10
0.10
precision level
0.05
0.05
probability level
0.95
0.95
fish sampled, g
77
262
Sampling efficiency
cv[q(s))
0.20
0.40
precision level
0.05
0.05
probability level
0.95
0.95
sample size, r
62
246
95% CI of bias of N(T.)
Vz width achieved
0.01«
0.024
Number of trials
97
64
Bias
Significance levels
Significance levels
N(T,15)
95% CI HO:Bias»0.9
'/2 width HA:Bias<0.9
HO:Bias<l.l
HA:Bias>l.l
95% CI HO:Bias>0.9 HO:Bias<l.l
1/2 width HA:Bias<0.9 HA:Bias>l.l
Bias
Variance
Bias
Variance
0.7669
0.0368
0.0382
0.0000
1.0000
0.5605
0.0273
0.0405 0.0000 1.0000
N(T.16)
0.7556
0.0520
0.0454
0.0000
1.0000
0.6660
0.0633
0.0616 0.0000 1.0000
N(T,17)
0.8022
0.0718
0.0533
0.0002
1.0000
0.7294
0.0567
0,0583 0.0000 1.0000
N(T,18)
0.9012
0.0960
0.0617
0.5157
1.0000
0.7382
0.0765
0.0678 0.0000 1.0000
N(T,19)
0.8423
0.0689
0.0522
0.0152
1.0000
0.7538
0.0754
0.0673 0.0000 1.0000
Parrack: Estimating stock abundance from size data
317
Table 4
(continued)
Bias
Significance levels
Bias
Significance levels
N(T.20)
Estimates
95% CI
V2 width
HO:Bias»0.9
HA:Bias<0.9
HO:Bias<l.l
HA:Bias>l.l
Estimates
95% CI
V2 width
HO:Bias»0.9
HA:Bias<0.9
HO:Bias<l.l
HA:Bias>l.l
Bias
Variance
Bias
Variance
0.8227
0.0662
0.0512
0.0015
1.0000
0.7826
0.0722
0.0658
0.0002
1.0000
N(T,21)
0.9475
0.0440
0.0418
0.9871
1.0000
0.8975
0.0738
0.0666
0.4702
1.0000
N(T,22)
1.0006
0.0001
0.0019
1.0000
1.0000
0.9982
0.0003
0.0042
1.0000
1.0000
N(T,23)
1.0007
0.0001
0.0020
1.0000
1.0000
1.0012
0.0004
0.0051
1.0000
1.0000
N(T,24)
0.9973
0.0001
0.0024
1.0000
1.0000
0.9993
0.0005
0.0053
1.0000
1.0000
N(T,25)
0.9992
0.0001
0.0019
1.0000
1.0000
0.9979
0.0002
0.0037
1.0000
1.0000
N(T,26)
0.9995
0.0001
0.0020
1.0000
1.0000
0.9946
0.0003
0.0040
1.0000
1.0000
N(T,27)
0.9979
0.0000
0.0012
1.0000
1.0000
0.9938
0.0002
0.0038
1.0000
1.0000
N(T,28)
0.9984
0.0000
0.0013
1.0000
1.0000
0.9970
0.0001
0.0022
1.0000
1.0000
N(T,29)
0.9992
0.0000
0.0012
1.0000
1.0000
0.9944
0.0004
0.0050
1.0000
1.0000
N(T,30)
0.9973
0.0000
0.0013
1.0000
1.0000
0.9944
0.0002
0.0032
1.0000
1.0000
N(T,31)
1.0008
0.0001
0.0016
1.0000
1.0000
0.9948
0.0002
0.0038
1.0000
1.0000
N(T,32)
0.9965
0.0001
0.0016
1.0000
1.0000
0.9956
0.0001
0.0028
1.0000
1.0000
N(T,33)
0.9984
0.0000
0.0010
1.0000
1.0000
0.9954
0.0001
0.0030
1.0000
1.0000
N(T,34)
0.9973
0.0000
0.0014
1.0000
1.0000
0.9951
0.0005
0.0055
1.0000
1.0000
N(T,35)
0.9987
0.0001
0.0015
1.0000
1.0000
0.9948
0.0002
0.0034
1.0000
1.0000
N(T,36)
0.9950
0.0002
0.0026
1.0000
1.0000
0.9944
0.0004
0.0051
1.0000
1.0000
N(T,37)
0.9974
0.0001
0.0017
1.0000
1.0000
0.9960
0.0001
0.0029
1.0000
1.0000
N(T,38)
0.9977
0.0001
0.0020
1.0000
1.0000
0.9980
0.0002
0.0034
1.0000
1.0000
N(T,39)
0.9984
0.0000
0.0011
1.0000
1.0000
0.9989
0.0001
0.0029
1.0000
1.0000
N(T,40)
0.9984
0.0001
0.0014
1.0000
1.0000
0.9946
0.0002
0.0035
1.0000
1.0000
N(T.41)
0.9980
0.0001
0.0016
1.0000
1.0000
0.9964
0.0001
0.0029
1.0000
1.0000
N(T,42)
0.9975
0.0001
0.0016
1.0000
1.0000
0.9983
0.0001
0.0019
1.0000
1.0000
N(T,43)
0.9988
0.0001
0.0015
1.0000
1.0000
0.9991
0.0000
0.0017
1.0000
1.0000
N(T,44)
0.9985
0.0001
0.0016
1.0000
1.0000
0.9971
0.0001
0.0026
1.0000
1.0000
N(T,45)
0.9996
0.0000
0.0006
1.0000
1.0000
0.9977
0.0001
0.0026
1.0000
1.0000
N(T,46)
0.9997
0.0000
0.0006
1.0000
1.0000
0.9995
0.0000
0.0009
1.0000
1.0000
N(T,47)
0.9997
0.0000
0.0014
1.0000
1.0000
1.0041
0.0016
0.0113
1.0000
1.0000
N(T,48)
1.0000
0.0000
0.0000
1.0000
1.0000
1.0475
0.0902
0.0931
0.9990
0.8655
N(T,49)
0.9984
0.0001
0.0022
1.0000
1.0000
0.9992
0.0000
0.0016
1.0000
1.0000
N(T,50)
1.0245
0.0480
0.0579
1.0000
0.9947
1.2121
1.1789
0.3952
0.9392
0.2892
N(T,51)
0.9994
0.0000
0.0011
1.0000
1.0000
1.0000
0.0000
0.0000
1.0000
1.0000
N(T,52)
0.9744
0.0256
0.0503
0.9981
1.0000
1.0000
0.0000
0.0000
1.0000
1.0000
N(T,53)
0.9152
0.2376
0.1663
0.5709
0.9853
1.2839
0.9669
0.5151
0.9280
0.2420
N(T,54)
1.0519
0.0840
0.1093
0.9968
0.8060
0.9500
0.0250
0.0980
0.8413
0.9987
N(T,55)
1.0000
0.0000
0.0000
1.0000
1.0000
0.8944
0.1003
0.2069
0.4790
0.9743
N(T,56)
1.0000
0.0000
0.0000
1.0000
1.0000
0.8500
0.1350
0.2940
0.3694
0.9522
N(T,57)
1.0000
0.0000
0.0000
1.0000
1.0000
1.2500
0.2500
0.4900
0.9192
0.2743
N(T,58)
0.9000
0.0400
0.1960
0.5000
0.9772
2.0500
2.2050
2.0580
0.8633
0.1828
N(T,59)
0.6500
0.3675
0.6860
0.2375
0.9007
1.4500
0.0050
0.0980
1.0000
0.0000
N(T,60)
0.1000
0.6050
1.0780
0.0729
0.9655
N(T.)
0.9023
0.0056
0.0149
0.6165
1.0000
0.7959
0.0096
0.0240
0.0000
1.000
q(15)
1.3860
0.1273
0.0710
1.0000
0.0000
1.9296
0.3080
0.1360
1.0000
0.0000
q{16)
1.4553
0.2337
0.0962
1.0000
0.0000
1.7558
0.6789
0.2019
1.0000
0.0000
q(n)
1.4210
0.3804
0.1227
1.0000
0.0000
1.4925
0.1753
0.1026
1.0000
0.0000
q(18)
1.2625
0.2663
0.1027
1.0000
0.0010
1.6064
0.6341
0.1951
1.0000
0.0000
q(19)
1.3163
0.2459
0.0987
1.0000
0.0000
1.5429
0.4501
0.1644
1.0000
0.0000
q(20)
1.3521
0.2433
0.0982
1.0000
0.0000
1.4564
0.4088
0.1566
1.0000
0.0000
q(21)
1.1274
0.0701
0.0527
1.0000
0.1544
1.2697
0.1354
0.0901
1.0000
0.0001
q(22)
1.0136
0.0295
0.0342
1.0000
1.0000
1.0701
0.0400
0.0490
1.0000
0.8842
q(23)
1.0494
0.0236
0.0306
1.0000
0.9994
1.0805
0.0313
0.0434
1.0000
0.8113
q(24)
1.0770
0.0349
0.0372
1.0000
0.8875
1.0460
0.0401
0.0491
1.0000
0.9846
q(25)
1.0515
0.0266
0.0324
1.0000
0.9983
1.0328
0.0437
0.0512
1.0000
0.9949
q(26)
1.0264
0.0288
0.0338
1.0000
1.0000
0.9658
0.0560
0.0580
0.9870
1.0000
q(27)
0.9990
0.0386
0.0391
1.0000
1.0000
1.0400
0.0476
0.0534
1.0000
0.9861
q(28)
1.0209
0.0333
0.0363
1.0000
1.0000
0.9478
0.0739
0.0666
0.9202
1.0000
q(29)
0.9869
0.0450
0.0422
1.0000
1.0000
0.9101
0.0623
0.0611
0.6264
1.0000
q(30)
0.9408
0.0390
0.0393
0.9790
1.0000
0.9330
0.0487
0.0540
0.8843
1.0000
q(31)
0.9596
0.0434
0.0415
0.9976
1.0000
0.8261
0.0839
0.0710
0.0206
1.0000
q(32)
0.9150
0.0425
0.0410
0.7631
1.0000
0.9050
0.0622
0.0611
0.5634
1.0000
q(33)
0.9620
0.0444
0.0419
0.9981
1.0000
0.8922
0.0814
0.0699
0.4138
1.0000
q(34)
0.9061
0.0598
0.0486
0.5808
1.0000
0.8964
0.0935
0.0749
0.4623
1.0000
q(35)
0.9179
0.0619
0.0495
0.7603
1.0000
0.8245
0.0933
0.0748
0.0240
1.0000
q(36)
0.8685
0.0924
0.0605
0.1537
1.0000
0.7950
0.1000
0.0775
0.0039
1.0000
q(37)
0.8611
0.0802
0.0563
0.0883
1.0000
0.7620
0.0863
0.0720
0.0001
1.0000
q(38)
0.8240
0.0971
0.0620
0.0082
1.0000
0.7804
0.1357
0.0903
0.0047
1.0000
q(39)
0.7925
0.0791
0.0560
0.0001
1.0000
0.7502
0.1244
0.0871
0.0004
1.0000
318
Fishery Bulletin 90(2), 1992
Table 4 (continued)
q(40)
Estimates
Bias
95% CI
Vz width
Significance levels
Estimates
Bias
95% CI
V2 width
Significance levels
HO:Bias>0.9
HA:Bias<0.9
HO:Bias<l.l
HA:Bias>l.l
HO:Bias»0.9
HA:Bias<0.9
HO:Bias<l.l
HA:Bias>l.l
Bias
Variance
Bias
Variance
0.8476
0.1304
0.0722
0.0775
1.0000
0.7426
0.0887
0.0747
0.0000
1.0000
q(41)
0.7956
0.0730
0.0546
0.0001
1.0000
0.7352
0.4793
0.1737
0.0315
1.0000
q(42)
0.6863
1.5698
0.2546
0.0500
0.9993
0.7424
0.1202
0.0885
0.0002
1.0000
q(43)
0.8525
0.1842
0.0882
0.1455
1.0000
0.8617
0.1037
0.0829
0.1827
1.0000
q(44)
0.8133
0.0973
0.0648
0.0044
1.0000
0.8777
0.1561
0.1044
0.3375
1.0000
q(45)
1.0254
2.2033
0.3119
0.7846
0.6804
0.8420
0.0913
0.0799
0.0774
1.0000
q(46)
0.8358
0.1309
0.0774
0.0519
1.0000
1.0564
0.9370
0.2606
0.8803
0.6284
q(47)
0.3470
25.6967
1.1397
0.1708
0.9023
63.2003
<99999.99
121.9043
0.8417
0.1590
q(48)
0.8576
0.6898
0.1974
0.3370
0.9919
1.0744
0.3067
0.1716
0.9768
0.6152
q(49)
0.7220
0.4331
0.1638
0.0166
1.0000
0.0008
75.0753
3.0502
0.2817
0.7600
q(50)
0.9710
3.3934
0.4868
0.6124
0.6983
2.7969
115.6154
3.9135
0.8290
0.1977
q(51)
-0.3572
73.2005
2.4998
0.1621
0.8734
0.7113
2.1593
0.6141
0.2735
0.8926
q(52)
-3.7106
632.0744
12.6793
0.2380
0.7715
0.8225
0.6106
0.3610
0.3370
0.9340
q(53)
1.0422
0.3956
0.2146
0.9031
0.7011
1.7977
6.0280
1.2861
0.9144
0.1438
q(54)
1.0498
0.3516
0.2237
0.9054
0.6700
2.2403
8.9080
1.8499
0.9222
0.1135
q(55)
0.0243
7.1499
1.2023
0.0767
0.9602
3.6459
67.4526
5.3658
0.8421
0.1762
q(56)
4.1165
89.6428
6.1858
0.8459
0.1696
3.2914
12.9452
2.8790
0.9482
0.0679
q(57)
0.8978
0.0700
0.2116
0.4919
0.9694
0.9321
0.0669
0.2536
0.5980
0.9028
q(58)
-3.1313
74.4663
8.4568
0.1751
0.8366
0.6640
0.2346
0.6713
0.2454
0.8985
q(59)
-5.6513
172.1638
14.8479
0.1936
0.8136
0.7438
0.0005
0.0294
0.0000
1.0000
q{60)
-0.3459
7.3840
3.7661
0.2584
0.7741
exp(z)
0.6422
0.0084
0.0182
0.0000
1.0000
lations (Paloheimo 1980, Collie and
Sissenwine 1983). These estimates do
not seem highly correlated.
The unobserved change rate was a
random variable in the second test, so
its estimation error was not computed.
Error characteristics-of-abundance
estimates were extremely similar to
those of example one; apparently high
process variability does not adversely
affect estimation even in the presence
of sampling variance.
The contracted seasonal recruitment
pattern (Fig. 3), conventionally inter-
preted as age-specific cohorts, was
used in the last two examples. Growth
parameters were the same as the two
previous examples and growth varia-
tion was moderate (cv 0.1). Sampling
efficiencies were also unchanged and
their variability set at that of example
1 (cv[q] = 0.2). The unobserved change
rate randomly varied five-fold (zt~
U(0.05,0.25)). Catching was continu-
ous so each period's catch was as-
signed to midperiod for estimation. Overfishing was
simulated by rapidly increasing exploitation enough to
decrease stock abundance 36% during the four periods
of sampling (last four). The fishing mortality rates for
periods 6-19 were: 0.05, 0.1, 0.15, 0.2, 0.25, 0.8, 0.6,
Table 5
Correlation coefficients between estimates of the unobserved change rate (z) and |
all other estimates
Estimate
Rho
Estimate
Rho
Estimate
Rho
Estimate
Rho
N(T,15)
0.46
N(T,38)
0.08
q(15)
-0.53
q(38)
-0.18
N(T,16)
0.34
N(T,39)
0.11
q(16)
-0.30
q(39)
0.09
N(T,17)
0.45
N(T,40)
0.21
q(17)
-0.36
q(40)
-0.03
N(T.18)
0.20
N(T.41)
0.22
q(18)
-0.24
q(41)
0.02
N(T,19)
0.27
N(T,42)
0.22
q(19)
-0.18
q(42)
-0.09
N(T,20)
0.23
N(T,43)
0.15
q(20)
-0.18
q(43)
-0.14
N(T,21)
0.23
N(T,44)
0.08
q(21)
-0.22
q(44)
-0.01
N(T,22)
0.11
N(T,45)
0.31
q(22)
-0.22
q(45)
-0.07
N(T.23)
0.27
N(T,46)
0.20
q(23)
-0.16
q(46)
0.03
N(T.24)
0.22
N(T.47)
0.04
q(24)
-0.15
q(47)
0.11
N(T,25)
0.21
N(T,48)
0.19
q(25)
-0.27
q(48)
0.11
N(T,26)
0.10
N(T,49)
0.13
q(26)
-0.31
q(49)
0.08
N(T,27)
0.19
N(T,50)
0.26
q(27)
-0.26
q(50)
0.12
N(T.28)
0.22
N(T,51)
0.22
q(28)
-0.10
q(51)
0.02
N(T.29)
0.29
N(T,52)
0.11
q(29)
-0.35
q(52)
-0.11
N(T.30)
0.09
N(T,53)
0.16
q(30)
-0.10
q(53)
0.15
N(T,31)
0.17
N(T,54)
0.12
q(31)
-0.43
q(54)
0.15
N(T,32)
0.19
N(T,55)
0.21
q(32)
0.09
q(55)
0.02
N(T,33)
0.21
N(T,56)
0.16
q(33)
-0.17
q(56)
-0.05
N(T,34)
0.25
N(T.57)
0.06
q(34)
-0.19
q(57)
0.15
N(T,35)
0.18
N(T.58)
0.03
q(35)
-0.02
q(58)
-0.11
N(T.36)
0.08
N(T.59)
0.06
q(36)
0.04
q(59)
-0.11
N(T,37)
0.23
N(T,60)
-0.03
q(37)
-0.01
q(60)
-0.06
0.4, 0.5, 0.8, 0.6, 0.8, 1.0, and 1.2. Example 3 simulated
very low sampling levels and example 4, high levels.
It was of interest to find if abundance would be cor-
rectly estimated during overfishing under either sam-
pling condition.
Parrack: Estimating stock abundance from size data
319
Example 3 was the limited-data case. The growth
measurement error was large (cv 0.20) and the sam-
ple size for growth parameter estimation was moderate
(95% CI of width ± 5%, 77 fish). The precision of catch
estimates was low (cv[C] = 0.4) and relative abundance
sampling was meager (95% CI of width ±30%, two
samples each period).
Error variances of the smallest seven size class abun-
dance estimates were very large (Table 6), but error
variances were low for size-classes 25 and larger.
Usefully narrow confidence intervals on the bias of
these estimates were obtained with few trials. Signif-
icance levels suggested that abundance estimates of
size-classes 17-21 might not have been biased and un-
biased estimation seemed likely for size-classes 22 and
larger. Estimates of sampling gear efficiencies (q(s))
also seemed accurate although error variances were
high.
Example four simulated sufficient sampling. A
growth parameter measurement error (cv 0.05) and
sample size (99% CI of width ±1%, 829 fish) more
characteristic of databases for heavily sampled fisheries
were used. Catches were precisely estimated (cv[C]
= 0.2) and relative abundance sampling was at a very
sufficient level (99% CI of width ±3%, 295 samples
each period).
Biases (Table 6) were very similar to those of exam-
ple 3. Abundance estimates for the smaller size-classes
that appeared in relative abundance samples just once
were probably biased by more than 10%, but the rest
were not. Estimates of q for the smallest 10 size-classes
were biased by more than 10% and the rest were prob-
ably not. Most error variances for stocksize estimates
were several times smaller than those of example 3,
and some were an order of magnitude smaller. Like-
wise, the error variance of q estimates was also smaller.
As may be expected, sufficient sampling levels in-
creased precision but did not affect bias. Abundance
estimates of sizes that appeared in abundance samples
more than once were estimated accurately when over-
fishing occurred, whether or not sampling levels were
sufficient or not.
Estimates of historical stock sizes are usually used
to find out if stock abtmdance is increasing or decreas-
ing. Errors of virtual population analysis back-calcu-
lations of cohort- specific abundances converge as dates
decrease (Agger et al. 1971, Pope 1972, Jones 1981).
Conventional wisdom is thus that abundance estimates
for the last period are extremely uncertain, but due to
the convergence, estimated abundance trends are
reliable. For this size-based estimator, (2) provides
abundance calculations before date y(T) from the
estimates available at the solution of (4).
Error characteristics of historical abundance esti-
mates (Table 7) were unexpected. Bias and error
variance increased as dates decreased. Last-period
Table 6
Examples
for seasonal, contracted recruitment and overfishing.
Catch estimation
Example 3,
Limited sampling
Example 4, Sufficient sampling
catch dates
absent
absent
cv[C(t,s)
0.40
0.20
Growth estimation
cv [error]
0.20
0.05
precision level
0.05
0.01
probability level
0.95
0.99
fish sampled, g
77
829
Sampling efficiency
cv(q(s)]
0.20
0.20
precision level
0.30
0.03
probability level
0.95
0.99
sample size, r
2
295
Number of trials
83
128
Bias
Significance levels
Bias
Significance levels
Estimates
N(T,15)
95% CI HO:Bias>0.9
V2 width HA:Bias<0.9
HO:Bias<l.l
HA:Bias>l.l
95% CI
Vz width
HO:Bias>0.9
HA:Bias<0.9
HO:Bias<l.l
HA:Bias>l.l
Bias
Variance
Bias
Variance
9.3261
759.7793
5.9301
0.9973
0.0033
12.8103
1297.3179
6.2399
0.9999
0.0001
N(T,16)
8.3623
546.7384
5.0304
0.9982
0.0023
8.3910
128.6375
1.9649
1.0000
0.0000
N(T,17)
2.0437
320.3240
3.8505
0.7198
0.3155
3.4678
49.7123
1.2215
1.0000
0.0001
N(T,18)
1.8321
42.6741
1.4054
0.9032
0.1536
1.2202
1.3293
0.1997
0.9992
0.1191
N(T,19)
0.9011
1.3723
0.2520
0.5033
0.9391
0.7992
0.3075
0.0961
0.0199
1.0000
N(T,20)
1.1948
1.6789
0.2788
0.9809
0.2525
0.8677
0.3403
0.1011
0.2657
1.0000
320
Fishery Bulletin 90(2). 1992
Table 6 (continued)
N(T,21)
N(T,22)
N(T,23)
N(T,24)
N(T,25)
Estimates
Bias
95% CI
Vz width
Significance levels
Estimates
Bias
95% CI
Vz width
Significance levels
HO:Bias>0.9
HA:Bias<0.9
HO:Bias<l.l
HA:Bias>l.l
HO:BiasS=0.9
HA;Bias<0.9
HO:Bias<l.l
HA:Bias>l.l
Bias
Variance
Bias
Variance
1.3401
0.9265
0.9335
0.9017
0.9988
5.0248
0.1610
0.4053
0.7664
0.0606
0.4823
0.0863
0.1370
0.1895
0.0533
0.9632
0.7262
0.6840
0.5072
0.9999
0.1646
1.0000
0.9914
0.9799
0.9999
1.7942
0.8043
0.8082
0.9673
1.0007
1.3107
1.2599
0.7830
0.1843
0.0106
0.1983
0.1945
0.1533
0.0744
0.0179
1.0000
0.1673
0.1203
0.9619
1.0000
0.0000
0.9986
0.9999
0.9998
1.0000
N(T,26)
N(T,27)
N(T.28)
0.9540
0.9890
0.9865
0.0476
0.0186
0.0129
0.0472
0.0295
0.0246
0.9875
1.0000
1.0000
1.0000
1.0000
1.0000
1.0021
0.9898
0.9855
0.0032
0.0055
0.0109
0.0098
0.0129
0.0181
1.0000
1.0000
1.0000
1.0000
1.0000
1.0000
N(T,29)
N(T,30)
0.9596
1.0184
0.0341
0.0600
0.0400
0.0530
0.9983
1.0000
1.0000
0.9987
0.9795
0.9979
0.0174
0.0235
0.0230
0.0269
1.0000
1.0000
1.0000
1.0000
N(T,31)
N(T,32)
0.9606
0.9922
0.0254
0.0073
0.0349
0.0188
0.9997
1.0000
1.0000
1.0000
0.9916
0.9945
0.0084
0.0029
0.0161
0.0094
1.0000
1.0000
1.0000
1.0000
N(T.33)
N(T,34)
N(T,35)
1.0158
0.9851
0.9700
0.0266
0.0117
0.0138
0.0359
0.0240
0.0260
1.0000
1.0000
1.0000
1.0000
1.0000
1.0000
0.9861
0.9824
0.9846
0.0022
0.0040
0.0025
0.0082
0.0112
0.0090
1.0000
1.0000
1.0000
1.0000
1.0000
1.0000
N(T.36)
N(T.37)
0.9908
0.9852
0.0339
0.0071
0.0411
0.0195
1.0000
1.0000
1.0000
1.0000
0.9921
0.9988
0.0008
0.0004
0.0052
0.0038
1.0000
1.0000
1.0000
1.0000
N(T.38)
N(T,39)
0.9807
0.9874
0.0027
0.0116
0.0126
0.0278
1.0000
1.0000
1.0000
1.0000
0.9992
0.9957
0.0006
0.0003
0.0048
0.0037
1.0000
1.0000
1.0000
1.0000
N(T.40)
0.9841
0.0024
0.0132
1.0000
1.0000
0.9986
0.0004
0.0042
1.0000
1.0000
N(T.41)
0.9992
0.0010
0.0090
1.0000
1.0000
1.0001
0.0001
0.0028
1.0000
1.0000
N(T,42)
N(T.43)
1.0102
0.9924
0.0314
0.0012
0.0549
0.0115
1.0000
1.0000
0.9993
1.0000
1.0002
0.9987
0.0000
0.0001
0.0013
0.0025
1.0000
1.0000
1.0000
1.0000
N(T,44)
0.9798
0.0101
0.0452
0.9997
1.0000
0.9986
0.0001
0.0027
1.0000
1.0000
N(T,45)
0.9833
0.0012
0.0158
1.0000
1.0000
1.0000
0.0000
0.0000
1.0000
1.0000
N(T.46)
N(T.47)
0.9854
1.0000
0.0006
0.0010
0.0141
0.0253
1.0000
1.0000
1.0000
1.0000
1.0000
1.0000
0.0000
0.0000
0.0000
0.0000
1.0000
1.0000
1.0000
1.0000
N(T,48)
0.8000
0.0200
0.1960
0.1587
0.9987
1.0000
0.0000
0.0000
1.0000
1.0000
N(T,49)
N(T.)
1.1369
32.0777
1.2185
0.6484
0.4763
1.0000
1.1493
0.0000
29.8091
0.0000
0.9459
1.0000
0.6973
1.0000
0.4593
q(15)
0.3282
0.1186
0.0741
0.0000
1.0000
0.2474
0.0501
0.0388
0.0000
1.0000
q(16)
0.3998
0.2084
0.0982
0.0000
1.0000
0.2984
0.0519
0.0395
0.0000
1.0000
q(n)
q(18)
0.7850
1.3854
0.7019
0.9594
0.1802
0.2107
0.1055
1.0000
0.9997
0.0040
0.6889
1.2926
0.2424
0.6908
0.0853
0.1440
0.0000
1.0000
1.0000
0.0044
q(19)
1.9397
2.7953
0.3597
1.0000
0.0000
1.6785
0.6126
0.1356
1.0000
0.0000
q(20)
1.3499
0.6184
0.1692
1.0000
0.0019
1.4341
0.3226
0.0984
1.0000
0.0000
q(21)
0.9199
0.4187
0.1392
0.6104
0.9944
0.6845
0.0844
0.0503
0.0000
1.0000
q(22)
0.6958
0.2367
0.1047
0.0001
1.0000
0.5667
0.0730
0.0468
0.0000
1.0000
q(23)
q(24)
0.7353
0.9467
0.2491
0.5200
0.1074
0.1561
0.0013
0.7213
1.0000
0.9729
0.6515
0.7885
0.0753
0.0992
0.0475
0.0546
0.0000
0.0000
1.0000
1.0000
q(25)
1.0306
0.3424
0.1266
0.9784
0.8584
0.9317
0.0882
0.0515
0.8861
1.0000
q{26)
1.0431
0.2997
0.1185
0.9910
0.8266
1.0073
0.0575
0.0415
1.0000
1.0000
q(27)
1.0743
0.3260
0.1236
0.9971
0.6583
0.9851
0.0475
0.0377
1.0000
1.0000
q(28)
1.0729
0.4116
0.1389
0.9927
0.6489
0.9155
0.0586
0.0419
0.7658
1.0000
q(29)
1.0112
0.3553
0.1290
0.9545
0.9113
0.9105
0.0767
0.0482
0.6652
1.0000
q(30)
q(31)
0.9345
1.0777
0.4927
0.5119
0.1519
0.1568
0.6719
0.9869
0.9836
0.6096
0.9002
0.8955
0.0721
0.0796
0.0471
0.0495
0.5027
0.4297
1.0000
1.0000
q(32)
1.0668
0.3685
0.1339
0.9927
0.6864
0.9643
0.1123
0.0587
0.9841
1.0000
q(33)
1.0548
0.2669
0.1139
0.9961
0.7818
1.0293
0.1086
0.0578
1.0000
0.9918
q(34)
1.0891
0.4220
0.1442
0.9949
0.5588
0.9230
0.1037
0.0569
0.7862
1.0000
q(35)
1.2517
1.4470
0.2670
0.9951
0.1327
0.9907
0.1106
0.0597
0.9985
0.9998
q(36)
1.0885
0.5550
0.1664
0.9868
0.5537
1.0126
0.1046
0.0591
0.9999
0.9981
q(37)
1.0643
0.3099
0.1286
0.9939
0.7066
1.0658
0.1899
0.0814
1.0000
0.7950
q(38)
1.2725
1.2458
0.2713
0.9964
0.1064
0.9944
0.0895
0.0589
0.9992
0.9998
q(39)
0.9949
0.7569
0.2239
0.7970
0.8211
1.0445
0.2018
0.0923
0.9989
0.8806
q(40)
1.1630
0.5160
0.1934
0.9962
0.2617
0.9748
0.1302
0.0806
0.9656
0.9988
q(41)
1.3898
1.4057
0.3320
0.9981
0.0435
0.9666
0.2227
0.1130
0.8761
0.9896
q(42)
1.4005
2.9713
0.5342
0.9668
0.1351
1.0596
0.2525
0.1316
0.9913
0.7265
q(43)
1.0215
0.2530
0.1643
0.9263
0.8256
-0.7903
123.8881
3.4933
0.1715
0.8556
q(44)
0.9185
0.2248
0.2132
0.5674
0.9524
1.1289
0.8725
0.3051
0.9293
0.4263
q(45)
q(46)
q(47)
q(48)
4.0163
1.2540
1.2032
1.1792
137.2726
0.3985
0.2502
0.2007
5.4127
0.3572
0.4002
0.6210
0.8704
0.9740
0.9312
0.8109
0.1455
0.1990
0.3067
0.4013
1.4194
1.0488
1.9698
1.0484
2.0954
0.0964
5.6933
0.0029
0.5674
0.1396
1.4101
0.0473
0.9636
0.9816
0.9315
1.0000
0.1350
0.7639
0.1133
0.9839
q(49)
0.9818
0.0718
0.2626
0.7291
0.8113
Parrack: Estimating stock abundance from size data
321
Table 7
Error statistics of historical abundance estimates from
example 4.
Error
Bias
Error
Bias
Error
Bias
Variable
n
variance
estimate
Variable
n
variance
estimate
Variable
n
variance
estimate
N(l,15)
128
115.5114
5.9596
N(l,27)
128
0.1871
0.9173
N(l,39)
91
0.4743
0.7745
N(2,15)
128
86.8395
6.0680
N(2,27)
128
0.1899
0.8982
N(2,39)
91
0.5662
0.9049
N(3.15)
128
119.8275
5.3517
N(3,27)
128
0.4459
1.0290
N(3,39)
91
0.4302
1.0320
N(4,15)
128
1296.6914
12.8388
N(4,27)
128
0.0056
0.9896
N(4,39)
91
0.0003
0.9957
N(l,16)
128
30.4340
4.2588
N(l,28)
128
0.4354
0.9534
N(l,40)
77
1.1408
0.9088
N(2,16)
128
184.7422
7.4660
N(2,28)
128
0.2830
1.0502
N(2,40)
77
0.7205
1.0607
N(3,16)
128
68.0011
6.3408
N(3,28)
128
0.4060
1.1279
N(3,40)
77
0.4176
0.9065
N(4.16)
128
128.6176
8.3893
N(4,28)
128
0.0109
0.9854
N(4,40)
77
0.0004
0.9986
N(1.17)
128
10.8289
1.7620
N(l,29)
127
0.3800
0.9685
N(l,41)
67
0.8746
1.0508
N(2.17)
128
50.7792
2.9860
N(2,29)
127
0.6222
1.2014
N(2,41)
67
0.7359
0.9693
N(3.17)
128
6.3308
2.2878
N(3,29)
127
0.6897
1.1322
N(3,41)
67
0.6803
1.1584
N(4,17)
128
49.7113
3.4675
N(4.29)
127
0.0174
0.9794
N(4,41)
67
0.0001
1.0001
N(l,18)
128
0.5225
0.8387
N(l,30)
125
1.3161
1.1217
N(l,42)
56
0.2236
0.5534
N(2,18)
128
1.8065
1.0217
N(2,30)
125
0.6497
1.1361
N(2,42)
56
0.9773
0.8919
N(3,18)
128
1.3119
1.1786
N(3,30)
125
0.6635
1.1661
N(3,42)
56
0.8370
1.0048
N(4,18)
128
1.3292
1.2202
N(4,30)
125
0.0234
0.9971
N(4,42)
56
0.0000
1.0002
N(l,19)
128
0.2076
0.5674
N(1.31)
125
0.4402
0.9570
N{1,43)
39
0.3770
0.7205
N(2,19)
128
0.6367
0.7745
N(2.31)
125
0.7042
1.1035
N(2.43)
39
0.7222
1.0594
N(3,19)
128
0.2891
0.8107
N(3,31)
125
1.2571
1.1914
N(3,43)
39
0.5246
0.9171
N(4,19)
128
0.3074
0.7989
N(4,31)
125
0.0084
0.9916
N(4.43)
39
0.0001
0.9987
N(1.20)
128
0.3131
0.6707
N(l,32)
125
0.4139
0.8962
N(1.44)
36
0.3870
0.8145
N(2,20)
128
0.7200
0.8515
N(2,32)
125
0.4712
1.0943
N(2,44)
36
0.3686
0.8018
N(3,20)
128
0.4237
0.9559
N(3,32)
125
0.6876
1.1207
N(3,44)
36
0.2355
0.8025
N(4,20)
128
0.3411
0.8679
N(4,32)
125
0.0029
0.9944
N(4.44)
36
0.0001
0.9986
N(l,21)
128
1.3938
1.2717
N(l,33)
125
0.2451
0.7301
N(1.45)
25
0.4571
0.8148
N(2,21)
128
0.9207
1.3560
N(2,33)
125
0.4089
1.0021
N(2,45)
25
0.3210
0.8143
N(3.21)
128
1.5664
1.6328
N(3,33)
125
0.6057
1.1316
N(3,45)
25
0.1735
0.7110
N(4,21)
128
1.3111
1.7941
N(4,33)
125
0.0022
0.9864
N(4,45)
25
0.0000
1.0000
N(l,22)
128
3.4192
1.7054
N(l,34)
123
0.7671
0.8896
N(l,46)
19
0.3338
0.6204
N(2,22)
128
6.9873
2.0668
N{2,34)
123
0.5906
0.9968
N(2,46)
19
0.3380
0.5746
N(3.22)
128
7.5242
2.6186
N(3,34)
123
0.9354
1.1637
N(3,46)
19
0.1914
0.8296
N(4,22)
128
1.2599
0.8042
N(4,34)
123
0.0040
0.9824
N(4,46)
19
0.0000
1.0000
N(l,23)
128
1.1626
1.6450
N(l,35)
119
0.2867
0.8088
N(1.47)
11
0.1290
0.4227
N(2,23)
128
1.6512
1.5052
N(2,35)
119
0.6020
1.0950
N(2,47)
11
0.2434
0.4377
N(3,23)
128
8.1272
2.3806
N(3,36)
119
0.7955
1.1538
N(3,47)
11
0.0694
0.9398
N(4,23)
128
0.7830
0.8082
N(4,35)
119
0.0025
0.9846
N(4,47)
11
0.0000
1.0000
N(l,24)
128
1.2296
1.3307
N(l,36)
115
0.2635
0.6926
N(1.48)
5
0.1338
0.4633
N(2.24)
128
2.6137
1.3469
N(2,36)
115
0.4223
0.9411
N(2.48)
5
0.0132
1.0800
N(3.24)
128
23.1672
1.9363
N(3,36)
115
0.7491
1.0997
N(3.48)
5
0.1087
0.7000
N(4,24)
128
0.1842
0.9673
N(4,36)
115
0.0008
0.9921
N(4,48)
5
0.0000
1.0000
N(l,25)
128
0.5672
0.9923
N(l,37)
110
0.2637
0.6720
N(l,49)
4
0.1558
0.3750
N(2,25)
128
0.7867
0.9758
N(2,37)
110
0.5350
0.8555
N(2,49)
4
0.4950
1.1000
N(3,25)
128
5.6842
1.4490
N(3,37)
110
0.6735
1.0410
N(3.49)
4
0.0050
0.9500
N(4,25)
128
0.0106
1.0005
N(4,37)
no
0.0004
0.9986
N(4.49)
4
0.0000
1.0000
N(l,26)
128
0.2515
0.8747
N(l,38)
99
0.8277
0.8260
N(2,26)
128
0.8233
0.9564
N(2,38)
99
0.2681
0.7480
N(3,26)
128
3.1000
1.2203
N(3,38)
99
0.8225
1.2851
N(4,26)
128
0.0032
1.0024
N(4,38)
99
0.0006
0.9989
abundance estimates were accurate, but those of pre-
ceding periods were not. Error variances of estimates
before the last period tended to be one to two orders
of magnitude higher than those of the last period. This
implies that abundance trends estimated in this man-
ner probably will be wrong. In this example, the prob-
lem is large enough to mask much of the 36% decrease
in abundance; the downward abundance trend would
not be clear in calculations of historical stock sizes.
Discussion
These Monte Carlo tests show that size-based methods
can be accurate and precise estimators of stock abun-
dance. Population characteristics need not conform
to the restrictive assumptions of traditional VPA
methods. Any sort of fish stock can be successfully ad-
dressed with size-based techniques, an important
aspect when assessing populations where ageing is im-
possible or where recruitment is not periodic.
322
Fishery Bulletin 90(2). 1992
These results imply that little will be gained from the
extensive age sampling programs that are the founda-
tion of VPA-based methods. They are not needed if
size-based methods are used; only size samples and the
rate of growth are required. Light, periodic growth
sampling is sufficient to monitor possible growth rate
changes through time. Also, since growth rates are re-
quired instead of ages, mark-recapture methods can be
used to obtain growth measures if hardpart interpreta-
tions (age is not observed on hardparts; instead, char-
acteristic marks are interpreted as annular occur-
rences) are difficult or expensive to obtain.
The method of abundance estimation developed in
this study, a meticulous bookkeeper of size data as is
the method of Beddingtc. and Cook (1981), is primitive
compared with other size-based methods (Foumier and
Doonan 1987, Schnute et al. 1989, Sullivan 1989, Sul-
livan et al. 1990). Its degree of success in estimating
abundance suggests that complete population-model
structures are unnecessary. Estimates were usefully
accurate and precise even with very high process vari-
ability. Very pronoimced individual growth variation
did not cause estimation problems. These results show
that precise, accurate abundance estimates are possible
with any recruitment pattern imaginable. It was a par-
ticular surprise to find that temporally variant (four-
fold) unobserved change rates ("natural mortality" of
Ricker (1948) but including migration and unrecorded
catch) did not affect estimation at all. That result is
reassuring, since the rate is probably extremely vari-
able in nature.
Sampling problems did not seem to degrade estima-
tion either. The level of catch estimation error proved
unimportant and there was no indication that exact
catch dates need to be recorded. Highly variable sam-
pling efficiencies (qg) did not cause estimation prob-
lems, particularly when sample sizes were adequate.
Highly variable individual growth rates (20%) and
significant growth measurement error (15%) did not
adversely affect abundance estimation when sampling
was sufficient. Very large growth-parameter measure-
ment error (40%) and small sample size destroyed per-
formance; although bias was not a problem, extreme
error variances and correlated estimates were.
It is of particular interest that this was the only test
where estimates of the unobserved change rate (z) and
sampling efficiencies (q) were highly correlated. The
lack of a pronounced correlation between sampling
gear efficiencies and the unobserved change rate in all
other tests except this one was unexpected; similar
studies of VPA-based methods (Paloheimo 1980, Collie
and Sissenwine 1983) found such correlation a major
characteristic. It thus seems possible that ageing
errors, or the violation of a connected VPA assump-
tion, contributed to correlation in those studies.
Abundances of most size-classes were estimated
precisely with little or no bias, but biased and imprecise
abundance estimates occurred in three circumstances.
First, abundances of very small fish that were recruited
between the next-to-last and last relative abundance
sample were estimated poorly. A recruitment group
had to be present in the relative abundance samples
twice to be estimated with a useful degree of certain-
ty. In practice, this problem is easily fixed if obtaining
certain estimates of recent recruitment of small fish
is important enough to justify the cost of additional
samples during the last period. Since the estimator is
not based on equal time units, only dates, additional
sample(s) vdll monitor the size-classes of interest
several times instead of just once. Second, wide size-
classes caused bias and imprecision, particularly for
larger sizes. This bias was easUy eliminated by narrow-
ing size-classes. Last, calculations of historical abun-
dances were in large error. It is well known that VPA
calculations are poor for the most recent period of data
and improve as dates decrease. Though they are not
germane to current production levels, estimates of the
oldest stock sizes are the most certain ones in VPA.
The exact opposite is true for this size-based method.
Estimates of historical abundances obtained in the solu-
tion calculation should not be used; error variances of
these computations are very large. Since the estimates
of the final-period abundances are accurate and precise,
this is probably not a problem even if historical stock-
size estimates are needed. Although the procedure was
not tested, these estimates might be obtained by start-
ing with the initial four periods of data, estimating the
fourth period abundance vector, and then progressing
forward one period at a time. Abundance in the first
three periods cannot be estimated but subsequent abun-
dances can. The relation between the number of periods
in the data and estimation errors was not investigated,
but the authors- experience with VPA-based methods
indicates little, if any, would be gained with a longer
time-series.
This study shows that a priori knowledge of the un-
observed change rate (z) is not required to accurately
and precisely estimate abundance with this size-based
method, yet it is well known (Paloheimo 1980, Collie
and Sissenwine 1983, Deriso 1985, Pope and Shepherd
1985) that such knowledge is necessary when apply-
ing VPA-based procedures.
This study suggests that the unobserved change rate
(z) will often be estimated with bias, yet z should be
included in the vector of estimates anyway. Monte
Carlo tests of the Beddington and Cook model estab-
lished that simultaneous estimation of a natural mor-
tality schedule (analogous to the unobserved change
rate in this study) is necessary to avoid biased abun-
dance estimates (de la Mare 1988). If z is fixed instead
Parrack: Estimating stock abundance from size data
323
of estimated, abundance estimation bias is assured
because stock size is a function of that rate. It thus
seems prudent to include the rate in the vector of
estimates to avoid abundance estimation bias even if
it is not useful. When necessary, Monte Carlo methods
can be used to establish interval estimates on e"^. This
study indicates that estimates of e"^ are often biased,
yet precise. The estimate of error variance over the 97
trials of example 1 was 0.0084, so the 95% CI width
is ±1.96\/(0.0084^97) or ±0.0182, and the bias ad-
justment is 0.6422.
Acknowledgments
I express very sincere appreciation to Douglas G. Chap-
man for his knowledgeable and diligent guidance of this
research and William G. Clark whose critical sugges-
tions and encouragement significantly broadened the
scope of this study. I am indebted to Bradford E.
Brown, of the Southeast Fisheries Center of the Na-
tional Marine Fisheries Service, who provided com-
puter resources and other critical support. I especially
thank Nancie J. Parrack for many helpful suggestions
as the study progressed and Stephen B. Mathews for
technical recommendations as well as a critical review
of the original manuscript.
Citations
Agger, P., I. Boetius, and H. Lassen
1971 On errors in virtual population analysis. Int. Counc. Ex-
plor. Sea Doc. CM 1971/H:16, 10 p.
Anonymous
1981a Report of the sub-committee on sperm whales (Appen-
dix II). Rep. Int. Whaling Comm. 31:100-101.
1981b Report of the ad hoc working group on the use of ef-
fort data in assessments. Int. Counc. Explor. Sea Doc. CM
1981/0:5, 65 p.
1983 Report of the working group on methods of fish stock
assessment. Int. Counc. Explor. Sea Doc. CM 1983/Assess:17,
73 p.
1984 Report of the working group on methods of fish stock
assessments. Int. Counc. Explor. Sea Doc. CM 1984/Assess:
19, 56 p.
1986 Report of the ad hoc multispecies assessment working
group. Int. Counc. Explor. Sea Doc. CM 1986, 10 p.
Bartoo, N.W., and K.R. Parker
1982 Stochastic age-frequency estimation using the von Ber-
talanffy growth equation. Fish. Bull., U.S. 81:91-96.
Basson, M., A.A. Rosenberg, and J.R. Bedding^on
1988 The accuracy and reliability of two new methods for
estimating growth parameters from length-frequency data. J.
Cons. Cons. Int. Explor. Mer 44:277-285.
Beamish, R.J., and G.A. McFarlane
1983 The forgotten requirement for age validation in fisheries
biology. Trans. Am. Fish. Soc. 112:735-743.
Beddington, J.R., and J.G. Cooke
1981 Development of an assessment technique for male sperm
whales based on the use of length data from the catches, with
special reference to the North-west Pacific stock. Rep. Int.
Whaling Comm. 31:747-760.
Beddington, J.R., J.G. Cooke, and W.K. de la Mare
1983 Estimations of the western North Pacific sperm whale
population using the length distribution of catches. Rep. Int.
Whaling Comm. 33:747-749.
Box, G.E.P.
1979 Robustness in the strategy of scientific model building.
In Launer, R.L., and O.N. Wilkinson (eds.), Robustness in
statistics, p. 201-236. Academic Press, NY.
Brethes, J.F., and G. Desrosiers
1981 Estimation of potential catches of an unexploited stock
of soft-shell clam (Mya arenaria) from length composition
data. Can. J. Fish. Aquat. Sci. 38:371-374.
Burr, R.L.
1988 Inferring the distribution of the parameters of the von
Bertalanffy growth model from length moments. Can. J. Fish.
Aquat. Sci. 45:1779-1788.
Chapman, D.G.
1961 Statistical problems in dynamics of exploited fisheries
populations. In Proc, 4th Berkeley symp. math. stat. and
probability. Contrib. Biol, and Probl. Med. 4:153-168. Univ.
Calif. Press.
Clark. W.G.
1981 Restricted least-squares estimates of age composition
from length composition. Can J. Fish. Aquat. Sci. 38:297-307.
Collie, J.S., and M.P. Sissenwine
1983 Estimating population size from relative abundance data
measured with error. Can. J. Fish. Aquat. Sci. 40:1871-1879.
Conway, G.R., N.R. Glass, and J. C. Wilcox
1970 Fitting nonlinear models to biological data by Marquardt's
algorithm. Ecology 51(3):503-507.
Cooke, J.G., and J.R. Beddington
1982 Further development of an assessment technique for male
sperm whales based on length data from the catches. Rep.
Int. Whaling Comm. 32:239-241.
Cooke. J.G.. and W.K. de la Mare
1983a Description of and simulation studies on the length-
specific sperm whale assessment technique. Rep. Int. Whal-
ing Comm. 33:741-745.
1983b Further analyses for the location of the possible stock
boundary for the North-West Pacific sperm whale based on
length distributions of males. Rep. Int. Whaling Comm. 33:
275-277.
Cooke. J.G.. W.K. de la Mare, and J.R. Beddington
1983a Some aspects of the reliability of the length data for the
western North Pacific stock of sperm whales. Rep. Int. Whal-
ing Comm. 33:265-267.
1983b An extension of the sperm whale model for the simula-
tion of the male population by length and age. Rep. Int. Whal-
ing Comm. 33:731-733.
de la Mare. W.K.
1988 Extended sensitivity trials on the length-specific method
for estimating the size of exploited sperm whale popiilations.
Rep. Int. Whaling Comm. 38:243-248.
de la Mare, W.K., and J.G. Cooke
1984 A new model for the female sperm whale incorporating
aspects of the reproductive cycle. Rep. Int. Whaling Comm.
34:251-254.
1985 Analyses of the sensitivity of the length-specific estima-
tion procedure to some departures from underlying assiunp-
tions. Rep. Int. Whaling Comm. 35:193-197.
324
Fishery Bulletin 90(2). 1992
1987 Further sensitivity trials in the length specific method
for estimating the size of exploited sperm whale popula-
tions. Rep. Int. Whaling Comm. 37:219-222.
Deriso, R.B.
1985 Catch-age analysis with auxiliary information. Can. J.
Fish. Aquat. Sci. 42:815-824.
Doubleday, W.G.
1975 A least squares approach to analyzing catch at age data.
Int. Comm. Northwest Atl. Fish. Res. Doc. 75/35, 22 p.
Fabens, A.J.
1965 Properties and fitting the von Bertalanffy growth curve.
Growth 29:265-289.
Fournier, D.A., and I.J. Doonan
1987 A length-based assessment method utilizing a generalized
delay-difference model. Can. J. Fish. Aquat. Sci. 44:422-437.
Fournier, D.A., J.R. Sibert, J. Majkowski, and J. Hampton
1990 MULTIFAN: A likelihood-based method for estimating
growth parameters and age composition from multiple length-
frequency data sets illustrated using data for southern bluefin
tuna. Can. J. Fish. Aquat. Sci. 47:301-317.
Fry, F.E.J.
1949 Statistics of a lake trout fishery. Biometrics 5:27-67.
Gallant, A.R.
1975 Nonlinear regression. Am. Stat. 29(2):73-81.
Gallucci, V.F.. and T.J. Quinn II
1979 Reparameterization, fitting, and testing a simple growth
model. Trans. Am. Fish. Soc. 108:14-25.
GuUand, J.A.
1965 Estimation of mortality rates. Annex to Rep. Arctic
Fish. Working Group, Int. Counc. Explor. Sea Doc. CM 1965(3),
9 p.
Hampton, J., and J. Majkowski
1987 An examination of the accuracy of the Elefan computer
programs for length based stock assessment. In Pauly, D.,
and G.P. Morgan (eds.), Length-based methods in fisheries
research, p. 203-216. ICLARM (Int. Cent. Living Aquat.
Resour. Manage.) Conf. Proc. 13, Manila.
Hasselblad, V.
1966 Estimation of parameters for a mixture of normal dis-
tributions. Technometrics 8:431-444.
Hoenig, J.M., and D.M. Heisey
1987 Use of a log-linear model with the EM algorithm to cor-
rect estimates of stock composition and to convert length to
age. Trans. Am. Fish. Soc. 116:232-243.
Jones, R.
1961 The assessment of long-term effects of changes in gear
selectivity and fishing effort. Mar. Res. (Scotl.) 1961(2):1-19.
1974 Assessing the long term effects of changes in fishing ef-
fort and mesh size from length composition data. Int. Counc.
Explor. Sea Doc. CM 1974/F:33, 13 p.
1981 The use of length composition data in fish stock assess-
ments (with notes on VPA and cohort analysis). FAO Fish.
Circ. 734, 60 p.
Kimura, K.K., and S. Chikuni
1987 Mixtures of empirical distributions: An iterative applica-
tion of the age-length key. Biometrics 43:23-35.
Knight, W.
1968 Asymptotic growth: An example of nonsense disguised
as mathematics. J. Fish. Res. Board Can. 25:1303-1307.
Kreuz, K.F., A.V. Tyler, and G.H. Kruse
1982 Variation in growth of Dover soles and English soles as
related to upwelling. Trans. Am. Fish. Soc. 111:180-192.
Lai, H.L., and V.F. Gallucci
1988 Effects of parameter variability on length-cohort
analysis. J. Cons. Cons. Int. Explor. Mer 45:82-92.
Laurec, A., and F.X. Bard
1980 Analyse multicohorte sur deux classes d'age. Int. Com.
Cons. Atl. Tunas Collect. Vol. Sci. Pap XV (SCRS-1979):l-20.
Law, A.M., and W.D. Kelton
1982 Simulation modeling and nalysis. McGraw-Hill, NY,
400 p.
Macdonald, P.D.M., and T.J. Pitcher
1979 Age groups from size frequency data: A versatile and
efficient method of analyzing distributing mixtures. J. Fish.
Res. Board Can. 36:987-1001.
Marquardt, D.W.
1963 An algorithm for least-squares estimation of non-linear
parameters. J. Soc. Ind. Appl. Math 11(2):431-441.
Mendelssohn, R.
1988 Some problems in estimating population sizes from catch-
at-age data. Fish. Bull., U.S. 86:617-630.
Morrison, D.D.
1960 Methods for nonlinear least squares problems and con-
vergence proofs, tracking programs, and orbit determina-
tion. Proc. Jet Propul. Lab. Sem. (1960): 1-9.
Murphy, G.I.
1965 A solution of the catch equation. J. Fish. Res. Board
Can. 9:450-491.
Paloheimo, J.E.
1980 Estimationof mortality rates in fish populations. Trans.
Am. Fish. Soc. 109:378-386.
Parrack, M.L.
1 98 1 An assessment of the Atlantic bluefin tuna resource. Int.
Com. Cons. Atl. Tunas Collect. Vol. Sci. Pap. XV:259-272.
1986 A method of analyzing catches and abundance indices
from a fishery. Int. Com. Cons. Atl. Tunas Collect. Vol. Sci.
Pap. XXIV:209-221.
1990 Estimating stock abundance from size data. Ph.D. diss.,
School Fish., Univ. Wash., Seattle, 116 p.
Parrack, N.J.C.
1986 Aspects of Pacific cod biology and population dynamics
in the North Pacific. M.S. thesis, Univ. Wash., Seattle, 157 p.
Pauly, D.
1982 Studying single species dynamics in a tropical multi-
species context. In Pauly, D., and G.I. Murphy (eds.), Theory
and management of tropical fisheries, p. 33-70. ICLARM (Int.
Cent. Living Aquat. Resour. Manage.) Conf. Proc. 9, Manila.
Pauly, D., M.L. Palomares, and F.C. Gayanilo
1987 VPA estimates of monthly population length composi-
tion, recruitment, mortality, biomass and related statistics of
Peruvian anchoveta, 1953 to 1981. In Pauly, D., and I. Tsu-
kayama (eds.). The Peruvian anchoveta and its upwelling
ecosystem: Three decades of change, p. 142-166. ICLARM
(Int. Cent. Living Aquat. Resour. Manage.) Stud. Rev. 15,
Manila.
Pikitch, E.K., and R.L. Demory
1988 Assessment of scales as a means of aging Dover sole.
Trans. Am. Fish. Soc. 117:345-349.
Pope, J.G.
1972 An investigation of the accuracy of virtual population
analysis using cohort analysis. Int. Comm. Northwest Atl.
Fish. Res. Bull. 9:65-74.
Pope, J.G., and J.G. Shepherd
1982 A simple method for the consistent interpretation of
catch-at-age data. J. Cons. Cons. Int. Explor. Mer 40:176-184.
1985 A comparison of the performance of various methods for
tuning VPAs using effort data. J. Cons. Cons. Int. Explor.
Mer 42:129-151.
Press, W.H., B.P. Flannery, S.A. Teukolsky, and W.T. Vetterling
1986 Numerical recipes, the art of scientific computing. Cam-
bridge Univ. Press, Cambridge, 818 p.
Parrack: Estimating stock abundance from size data
325
Quinn, T.J.. S.H. Hoag, and G.M. Southward
1982 Comparison of two methods of combining catch-per-unit-
effort data from geographic regions. Can. J. Fish. Aquat. Sci.
39:837-846.
Richards, F.J.
1959 A flexible growth function for empirical use. J. Exp.
Botany 10(29) 290-300.
Ricker, W.E.
1948 Methods of estimating vital statistics of fish populations.
Indiana Univ. Publ. Sci. Ser. 15, 101 p.
1975 Computation and interpretation of biological statistics
offish populations. Bull. Fish. Res. Board Can. 191, 382 p.
Rosenberg, A. A., and J.R. Beddington
1987 Monte Carlo testing of two methods for estimating
growth from length frequency data, with general conditions
for their applicability. In Pauly, D., and G.P. Morgan (eds.),
Length-based methods in fisheries research, p. 283-289.
ICLARM (Int. Cent. Living Aquat. Resour. Manage.) Conf.
Proc. 13, Manila.
Rubinstein, R.Y.
1981 Simulation and the Monte Carlo method. John Wiley,
NY. 278 p.
Scheuer. E.M., and D.S. Stoller
1962 On the generation of normal random vectors. Tech-
nometrics 4:278-281.
Schnute, J.T., L.J. Richards, and A.J. Cass
1989 Fish survival and recruitment: Investigations based on
a size-structured model. Can. J. Fish. Aquat. Sci. 46:743-769.
Schrage, L.
1979 A more portable Fortran random number generator.
ACM (Assoc. Comput. Machines) Trans. Math. Software 5:
132-138.
Shepherd, J.G.
1985 Deconvolutionof length compositions. Int. Counc. Ex-
plor. Sea Methods Working Group, Working paper, 7 p.
Shirakihara, K., and S. Tanaka
1983 An alternative length-specific model and population
assessment for the western North Pacific sperm whales. Rep.
Int. Whaling Comm. 33:287-294.
1984 Simulation studies on length-specific population assess-
ment techniques for western North Pacific sperm whales.
Rep. Int. Whaling Comm. 34:259-264.
Shirakihara, K., S. Tanaka, and T. Nakano
1985 Further simulation studies on length-specific population
assessment techniques for western North Pacific sperm
whales. Rep. Int. Whaling Comm. 35:199-203.
Smith, S.J.
1980 Comparison of two methods of estimating the variance
of the estimate of catch per unit effort. Can. Fish. Aquat. Sci.
37:2346-2351.
Sullivan, P.J.
1989 Kalman filter approach to catch at length analysis. Ph.D.
diss., Univ. Wash., Seattle, 160 p.
Sullivan, P.J., H.L. Lai, and V.F. Gallucci
1990 A catch-at-length analysis that incorporates a stochastic
model of growth. Can. J. Fish. Aquat. Sci. 47:184-198.
Ulltang, 0.
1977 Sources of errors in and limitations of virtual population
analysis (cohort analysis). J. Cons. Cons. Int. Explor. Mer
37:249-260.
Welch, B.L.
1938 The significance of the difference between two means
when the population variances are unequal. Biometrika
25:350-362.
326 Fishery Bulletin 90(2). 1992
Appendix: Simulation steps of IVIonte Carlo tests
Control variables are F,, F2, z,, pt, nu pi2. o,, 02, /^k/ cv[l<], jua^ cv[AJ, cvfqj, cv[C]
Compute the following once each trial for Kt«T:
1 If the Ft are variable, Ft~U(Fi, Fg)
2 If the Zt are variable, Zt~U(zi, Z2)
3 If a single catch occurs once each period, then
A probability of unobserved events (z) during yt to Ct is Pr[z']t=l-e"^t(ct-yt),
B probability of being caught (on date Ct) is Pr[C]t = Ft,
C probability of unobserved events (z) during q to yt+j is Pr[z']t= l-e"^t(yt+i-Ct)^ or
4 If catching is continuous, then
A probability of death during y^ to yt+i is Pr[D]t=l-e"<^+'^t)(yt.i-yt), and
B Pr[z]t = Pr[D]tZt-(zt + Ft).
5 If recruitment is seasonal, then
A pt~U(l,20) where pt is the proportion recruited during period t,
t T-l
B <5t = ^ Pi ^ Pj = the accumulative frequency,
i=i j=i
C /Jt~U(p(i, M2). and
D ot~U(oi, 02).
Compute the following once for each fish:
6 Draw growth parameters k and A such that
A k~NU,(MkCv[k]2),
B A~NOiA, (ma cv[A]2).
7 Draw a recniitment data, tj , such that
A if recruitment is uniform, then ti~U(l,20), or
B if recruitment is seasonal,
(1) draw t with probability specified by 6,
(2) drawti~N(Mt, o2t).
8 If fishing is continuous, then
A for the time period of recruitment draw u where u~U(0,l).
(1) If u<Zti(l-e-<^ti+Ft,)(yt..i-t,)H-(zti + Fti), the fish exited of unobserved causes; STOP.
(2) If not, but if u<l-e-<^ti+F.i)(y„,,-t,)_ the fish was caught; go to step 8C(2)(a).
(3) If neither occurred, the fish lived through the time period of recruitment; continue.
B Add a fish to the abundance matrix.
(1) t = t+1.
(2) ti = yt.
(3) Si = the lower bound of the minimum size-class.
(4) Compute the size-class from equation (1).
(5) Nt.s2 = Nt,32+1.
(6) If t = T, STOP.
C Draw u where u~U(0,l).
(1) If u<Pr[z]t, the fish exited dur to unobserved events; STOP.
(2) If not, but if u<Pr[D]t, the fish was caught.
(a) Draw u where u~U(0,l).
(b) t2 = t-t-U.
(c) Compute the size-class equation (1).
(d) Add to the catch matrix: Ct,s = Ct,s + l; STOP.
Parrack. Estimating stock abundance from size data 327
(3) If neither occurred, the fish survived; go to step 8B.
9 If fishing occurs just once each period, then
A If ti<Cti, the fish recruited before the catch.
(1) Draw u where u~U(0,l).
(2) If u<l-e-^ti(<=ti-ti), the fish exited unobserved events before date q; STOP.
(3) If not, the fish survived to the catch date; go to step 9F.
B If ti = Cti , the fish recruited on the catch date; go to step 9F.
C If ti>Cti, the fish recruited after the catch.
(1) Draw u where u~U(0,l).
(2) If u< l-e'^ti(yti*i-ti), the fish exited due to unobserved events before the next abundance
sample (date yt+i); STOP.
(3) If not, the fish survived fishing and so was ahve on the next sampHng date: t = t+l.
D Add a fish to the abundance matrix.
(1) to = yt.
(2) Si = lower bound of the minimum size-class.
(3) Compute the size-class from equation (1).
(4) Nt,32 = Nt,32 + 1.
(5) Ift = T, STOP.
E Draw u where u~U(0, 1). If u<Pr[z']t, the fish exited due to unobserved events before the date of
catch; STOP.
F Draw u where u~U(0,l). If u<Pr[C]t, the fish was caught on date Ct.
(1) t2 = q.
(2) Sj = lower bound of the minimum size-class.
(3) Compute the size-class when caught from equation (1).
(4) Ct,,2 = Ct,s2+1-
(5) STOP.
G Draw u where u~U(0,l).
(1) If u<Pr[z']t, the fish exited due to unobserved events before the next abundance sampling
date; STOP.
(2) If not, the fish survived to the next relative abundance sample date.
(a) t = t-i-l.
(b) Go to step 9D.
Collect samples once each trial:
10 Draw an extimate of the growth parameters such that
p[A,k] = -0.95,
k ~ NOik. Mk(cv[k]^-t-cv[ek]2)divg, and
A '^ NO^A, MA(cv[A]2 + cv[eA]2)±g.
11 Draw an extimate of catch for all t and s where Ct,s'^N(Ct g, Ct^ cv[C]2).
12 For each t and s draw qt,s,k~N(qs, (qg cv[q]2) for l<k<r.
13 Calculate Yt, sand s2[Yt,s].
14 Determine the largest sampled size-class.
Abstract. -Feeding ecology, age
and growth, length-weight relation-
ships, and reproductive biology of
two species of tonguefishes, Cyno-
glossus arel and C. lida, from Porto
Novo, southeast coast of India, were
studied during October 1981-Sep-
tember 1982. These tonguefishes are
benthophagus; adults feed primarily
on polychaetes, while juveniles more
often consume smaller prey such as
hyperiid amphipods and copepods. A
negative correlation between spawn-
ing activity and gastrosomatic index/
hepatosomatic index was noted for
C. arel. In C. lida, a higher percent-
age of empty stomachs was observed
in males than in females.
Age and growth of these tongue-
fishes were determined by three
methods, viz, (1) Petersen method,
(2) probability plot, and (3) von Ber-
talanffy's equation. Rate of growth
from the time of hatching through
the first year is higher than that
of older year-classes. Both species
reach commercial size during their
2d and 3d year, and have a life-span
of 3-4 years. Value of L„ (theoret-
ical maximum attainable length) is
570 mm for male and 615 mm for
female C. arel, and 335 mm for male
and 340 mm for female C. lida.
Analyses of the length-weight rela-
tionship showed a significant differ-
ence in length-weight slopes of male
and female C. arel. Due to gonad
development, mature female C. lida
deviated significantly from the 'cube
law.'
Cynoglossus arel and C. lida have
prolonged spawning periods of 10
months, with a spawning peak in
January and September, respective-
ly. Individuals spawn only once dur-
ing each season. Both sexes of both
species attain first sexual maturity
during the 2d year. In male C. lida,
higher values of the gonadosomatic
index (GSI) in September indicate
the occurrence of fully-mature speci-
mens during this period. A rise in Kn
values (relative condition factor)
corresponds with a rise in gonadal
activity in female C. arel. The cor-
relation coefficient shows that fecun-
dity in C arel is correlated with total
length, total weight, ovary length,
and ovary weight, whereas in C. lida
it is correlated only with ovary
length and ovary weight.
Manuscript accepted 4 March 1992.
Fishery Bulletin, U.S. 90:328-367 (1992).
Biology of two co-occurring
tonguefisFies, Cynoglossus arel
and C lida (Pleuronectiformes:
Cynoglossidae), from Indian waters
Arjuna Rajaguru
Systematics Laboratory. National Marine Fisheries Service. NOAA
National Museum of Natural History, Washington, DC 20560
Out of 77 species of flatfishes occur-
ring along the east and west coasts
of India (Rajaguru 1987), only one
species, viz, the Malabar sole Cyno-
glossus macrostomus, constitutes an
important fishery along the Malabar
coast (west coast of India) (Bal and
Rao 1984). The Indian halibut Pset-
todes erumei, because of its larger
size and delicious flesh, fetches a high
value in fish markets (Pradhan 1969);
however, it does not comprise a high
value fishery. Other species of flat-
fishes which contribute to fisheries
along the Indian coasts are: Cyno-
glossus viacrolepidotus, C. arel, C.
dubius, C. lida, C. puncticeps, C.
bilineatus, C. lingua, Paraplagusia
spp., Solea spp., and Pseudorhombus
spp. (Seshappa 1973, Ramanathan
1977, Rajaguru 1987). However,
none of these species comprises a
single-species fishery. Separate sta-
tistics are not reported for these
species; all flatfish species are joint-
ly reported as 'soles' (CMFRI 1969,
Fischer and Bianchi 1984). Average
landings of flatfishes along the Indian
coast is about 2% of the total marine
fish catches (Ramanathan 1977, Ra-
jaguru 1987). Most of these flatfish
species became prominent in the
landings only after the introduction
of trawlers (Devadoss and Pillai
1973). These species, except the
malabar sole Cynoglossus macrosto-
mus, are generally not the target
species, but are taken incidentally in
the penaeid shrimp fishery. Along
the Porto Novo Coast, of the 47 flat-
fish species (Rajaguru 1987), only
Psettodes erumei, Pseudorhomims ar-
sius, Cynoglossus arel, and C. lida
occur throughout the year. The latter
two species are taken in a fishery
throughout the year, even during
the northeast monsoon period, when
other marine fish are generally
absent.
The biology of these two tongue-
fishes is poorly known, except for
work on age and growth of 138 C.
lida from the west coast of India
(Seshappa 1978). The present study
examines various aspects of biol-
ogy, including feeding ecology, age,
growth, length-weight relationships,
and reproductive biology of C. arel
and C. lida in Porto Novo coastal
waters.
Objectives of the study on the feed-
ing ecology of these two species of
tonguefishes are to determine (a) the
diet of juveniles and adults, (b) dif-
ferences in diet between seasons, and
(c) relationships between feeding
morphology, digestive morphology,
and diet. An age and growth study
was also undertaken to (a) evaluate
differences in growth patterns be-
tween males and females, (b) deter-
mine age of recruitment to the Porto
Novo fishery, and (c) determine lon-
gevity of these two tonguefishes. The
objective of the studies on length-
weight relationships is to determine
if there is a significant deviation from
the cube law of length-weight rela-
tionship related to ontogeny and go-
nadal development. The final aspect
328
Rajaguru: Biology of Cynoglossus arel and C lida from Indian waters
329
79'
FATHd^t LINE
10 FATHOM LINE
^
11 30' N
BAY OF BENGAL
w-
Figure 1
Map showing 5- and 10-fathom lines off Porto Novo coast, India.
of the study is reproductive biology. The objectives are
to (1) determine the spawning season, spawning
periodicity, age and size at first maturity, and (2) ex-
amine relationships between fecundity and total length,
total weight, ovary length and ovary weight.
Materials and methods
Samples of large-scaled tonguefish Cynoglossus arel
(Bloch and Schneider 1801) and shoulder-spot tongiie-
fish C. lida (Bleeker 1851) were collected twice week-
ly (a total of 96 collections) from commercial fish
catches landed in Porto Novo, southeast coast of India
(11°29'N, 79°46'E; Fig. 1), from October 1981 to Sep-
tember 1982. Fishing operations were confined to the
upper continental shelf, to a depth of 18-22 m, up to
4 km from the coast.
A total of 1220 specimens of C. arel (627 males, 569
females, 24 juveniles) and 1382 specimens of C. lida
(718 males, 640 females, 24 juveniles) were collected
for stomach analyses. For the age and growth study,
a total of 1203 specimens of C. arel (634 males and 569
females), and 1374 specimens of C. lida (724 males and
650 females) were utilized; since juveniles were avail-
able only for 4 months, they were not included in the
age and growth study. Length-weight equations were
computed using data of 1281 spe-
cimens of C. arel (655 males, 599
females, and 27 juveniles) and
1519 specimens of C. lida (768
males, 723 females, and 28 juve-
niles). A total of 1196 specimens
of C. arel (627 males and 569
females) and 1358 specimens of
C. lida (718 males and 640 fe-
males) were examined for the re-
productive biology studies. Some
specimens were used for all four
studies. Size range of the speci-
mens was as follows: C. arel
(males 95-360 mm TL, females
99-435 mm TL, juveniles 83-128
mmTL) and C. lida (males 97-
248mmTL, females 98-242 mm
TL, juveniles 81-125mmTL).
Total length (TL) of each fish was
measured to the nearest 1mm;
total weight (TW) was recorded
to the nearest O.lg. Sex, matur-
ity stages, TL, and TW were
noted in fresh-caught fish. Size
of monthly samples utilized for
various analyses is given in
Appendix.
Feeding ecology
Stomachs were removed and preserved in 5% formalin.
Some empty stomachs were shrunken and contained
mucus, while others were expanded but completely
empty; the latter type is believed to occur in fish which
have recently regurgitated (Daan 1973). Regurgitated
stomachs, as well as fish with food remains in their
mouths, were discarded.
Gastrosomatic index (GI) and hepatosomatic index
(HI) were calculated to examine monthly variations in
feeding intensity and to correlate these variations with
breeding cycles, using the following formulae:
Gastrosomatic index =
Weight of gut (including contents) x 100
Hepatosomatic index =
Weight of fish
Weight of liver x 100
Weight of fish
For stomach analysis, the Index of Relative Impor-
tance (IRI) (Pinkas et al. 1971) was used. It incor-
porates percentage by number (N), volume (V), and
frequency of occurrence (F) in the formula
330
Fishery Bulletin 90(2). 1992
IRI = (%N + %V) X %F.
The percentage IRI was calculated for the entire data
set of juveniles, males, and females, and by length
intervals (45 mm interval for C. arel, and 21mm for
C. lida).
Stomach contents were sorted, identified to the
lowest possible taxa, and enumerated. Appendages and
remains of the unidentified crustaceans are classified
as "crustacean fragments." Volume of each taxonomic
group of prey was measured by water displacement.
To determine ontogenetic variations in feeding habits,
stomach contents of juveniles were analyzed separately
from those of adults.
Age and growth
Length measurement data were grouped into size-
classes at intervals of 15 mm for C. arel and 7 mm for
C. lida. Percentage frequencies were calculated by
month. Sexes were treated separately, to determine
whether there were differences in growth patterns be-
tween males and females. The Petersen method, prob-
ability plot method, and von Bertalanffy's equation
were used to determine age and growth. Attempts to
detect growth layers in hard parts (scales, otoliths,
opercular bones, and supraoccipital crests) were not
successful.
Petersen method This method of growrth analysis is
based on the assumption that the lengths of individuals
of the same age in a population are distributed normal-
ly. When there are distinct intra-annual spawning
periods, the length-frequency distribution may be
multimodal, representing successive age-groups. The
rate of growth slows with age (Ford 1933), and as a
result the modes overlap, making interpretation dif-
ficult. In the case of fishes, such as tonguefishes, which
have a prolonged spawning period, various broods
entering the fishery overlap. In this case, it is necessary
to trace a size-group for as many months as possible
after it enters the commercial fishery and to find the
average monthly growth rate for different size-classes.
Approximate values of average size at different ages
may then be calculated.
Probability plot method Plots of cumulative per-
centages of length distribution on probability paper
provide estimates of the length ranges of fish in each
age-group (Harding 1949, Cassie 1954). Hence fish
lengths were used to obtain an approximation of the
length-at-age structure. One difficulty in this method
is the uncertainty surrounding whether the deviations
represent virtual inflexion points of the lines. Another
difficulty is locating each inflexion point, since any
bend in the line is considered an inflexion point. Follow-
ing this procedure, the line was divided into separate
parts and for each (Cassie 1954), partial straight lines
were drawn from which, a mean length was calculated
for each age-group.
von Bertalanffy's equation The most widely ac-
cepted growrth model is that of von Bertalanffy (1938),
U = L„(l-e-Mt-t„))
where Lt = length at age t,
L^ = theoretical maximum attainable (asymp-
totic) length,
k = a constant, expressing the rate of change
in length increments with respect to t,
to = hypothetical age at zero length, and
e = base of Naparian or natural logarithm.
The value of to was calculated as follows:
-to = 1/k [loge (L^) - loge (L^-U)] - t.
Walford's (1946) procedure was used to substitute
Lf -H 1 for Lt . The equation now can be written as
Lt + 1 = L^(l-e-'^) + L,e-K
Length-weight relationship
Length-weight curves were obtained by using the equa-
tion W = aL''. The least-squares regression of the log-
arithmic transformation,
logic W = logioa + b logioL,
where logioW = Y, logioa = a, logioL = X, b = n, was
used for estimating the values of a and b (Snedecor
1956). This linear equation was fitted separately for
males, females, and unsexed juveniles of C. arel and
C. lida from monthly data.
To determine whether increased weight at a given
length was caused by increased gonad weight in mature
fish, the length-weight relationship was compared be-
tween different stages of maturity. Adults of both
sexes of C. arel and C. lida were classified into three
stages (Rajaguru 1987):
Immature (Stage I for both sexes): n 56 male and 47
female C. arel; 105 male and 54 female C. lida;
Maturing (Stage II for males. Stages II-III pooled
for females): n 221 male and 224 female C. arel; 259
male and 342 female C. lida; and
Mature (Stage III for males. Stages IV-VI pooled for
females): n 359 male and 292 female C. arel; 363
Rajaguru: Biology of Cynoglossus are! and C Itda from Indian waters
331
male and 254 female C. lida. (Refer to section on
Reproductive biology, for Stages I-VI.)
The significance of variation in the estimate of b,
from the expected value B ( = 3) for an ideal fish was
tested by the i-test in both sexes of C. arel and C. lida
(James 1967):
Sb
where B = hypothetical b ( = 3), and Sb = standard error
of b.
Analysis of covariance (Snedecor 1956) was used for
all comparisons.
Reproductive biology
Tonguefishes have no secondary sexual characters to
distinguish the sexes. In females with gonads in ad-
vanced stages of maturity, ovaries can be seen easily
through the body wall when the fish is held against
light. In earlier stages of maturity, sexes are distin-
guishable only after dissection. Extension of gonads
into body cavity, and their color, shape, and size, were
noted after dissection. Ovary length was measured to
the nearest mm, while weight of testis/ovary was
recorded to the nearest mg. Ovaries were fixed in
modified Gilson's fluid (Simpson 1951) for ova diameter
studies.
To investigate the distribution pattern of ova in dif-
ferent regions of the ovary, ova were taken from an-
terior, middle, and posterior regions of eyed-side and
blind-side lobes of ovaries in different stages of matur-
ity (Clark 1934, Hickling and Rutenberg 1936, de Jong
1940). Ova diameter measurements in each part were
noted separately. Results showed a uniform distribu-
tion of ovum size in different parts of both ovarion
lobes. Hence to study development of ova, random
samples of '^^500 ova per ovary were measured from
ovaries representing Stages I-VI (a total of 108
ovaries, at 18 ovaries/stage in C. arel, and a total of
168 ovaries, at 28 ovaries/stage in C. lida), using an
ocular micrometer at a magnification which gave a
value of 12.5^ (0.0125mm) to each micrometer division
(m.d.). Ova diameter-frequency polygons were drawn
after grouping the ova into 3 m.d. (0.04 mm) class-
intervals.
Spawning seasons in both species were determined
from percentage occurrence of different maturity
stages during various months of the year.
Generally, gonad weight depends on size and stage
of gonadal development. To account for effects of dif-
ferential body size on gonad size, gonad weight was
expressed as a percentage of body weight (Nikolsky
1963). This ratio,
Weight of gonad x 100
Weight of fish
is termed gonadosomatic index (GSI). To determine
the spawning season, GSIs for various months were
calculated.
Relative condition factor (Kn) was calculated for in-
dividual fish of both sexes from the formula (Le Cren
1951),
Kn = W/W
where, W_ = observed weight, and W = calculated
weight (W = a-i-bx). Monthly mean values of Kn
were also calculated to confirm the spawning season.
To determine minimum length-at-first-maturity (i.e.,
Lm or L5o = length at which 50% of fish are mature),
specimens of C. arel and C. lida were grouped into
15mm and 7mm class-intervals, respectively. Sexes
were treated separately. Percentage occurrence of im-
mature and mature fish of various length-groups was
determined, and then percentage occurrence of mature
fish was plotted for both sexes.
Fecundity was determined by the gravimetric meth-
od. For this study, 26 ovaries of C. arel (from specimens
200-439mmTL) and 19 of C. lida (161-201 mm TL)
were used. Since some ova might already have been
shed, ovaries with oozing ova were not used. Ovaries
were removed, measured to the nearest mm, and
weighed to the nearest mg. From each ovary, three
subsamples (each ~50mg) were taken and weighed
after removing excess moisture, and fixed in modified
Gilson's fluid. From each of these subsamples, yolked
ova were separated and counted. Mean number of ova
from three subsamples was multiplied by the ratio of
subsample weight: ovary weight to obtain an estima-
tion of the total number of mature cva in the ovary.
Numbers of ova per mm body length, per g body
weight, per mm ovary length, and per mg ovary weight
were also calculated. Fecundity of C. arel and C. lida
was related to total length (TL), total weight (TW),
ovary length (OL), and ovary weight (OW) using linear
regression. Statistical comparisons (Snedecor 1956) of
fecundity to TL, TW, OL, and OW were made.
To determine the differential distribution of sexes
during the spawning migration, as well as during ag-
gregation, the sex ratio was calculated for each month.
To test the homogeneity in distribution of males and
females, the chi-square formula was used.
Classification of maturity stages
Maturity stages were indexed for both sexes of C. arel
and C. lida, following the ICES scale (Lovern and
Wood 1937), with the following modifications. Color,
shape, and extension of the ovary into the body cavity.
332
Fishery Bulletin 90(2). 1992
6.40%
6,90%
!.20%
12 40%
1 1 .40%
Figure 2
Percentage contribution of food items to the
diet of juvenile, male, and female Cynoglossiis
arel and C. lida caught commercially off
Porto Novo, India, October 1981-September
1982. Only values >5% IRI are individually
shown; values < 5% IRI are clumped together
into a single category, the unshaded wedge
of the pie chart. PO = polychaetes, PR =
prawns; CF = crustacean fragments, FS =
fish scales, AM = amphipods, CO = cope-
pods, TN = tintinnids, FI = fishes, MI =
miscellaneous.
as well as color and shape of ova,
were considered to define stage
of maturity in females. Degree of
transparency of the ovary was
also used as a criterion, since it
is one of its characteristic fea-
tures during early as well as
fully-mature phases. Color and
size of testis were used to deter-
mine the stage of maturity in
males. In both species, testes
were divided into four stages,
and ovaries into seven stages as
follows:
Males:
Stage I (Immature) Testis
minute, pale white.
Stage II (Maturing) Testis
slightly enlarged, sac-like,
creamy white; no milt oozes out
on pressure.
Stage III (Mature) Testis
enlarged, sac-like, creamy white;
whitish milt running from vent
on slight pressure.
Stage IV (Spent) Not found
during the present study.
Females:
Stage I (Immature) Ovary
very small, thread-like and trans-
parent; under microscope, yolk-
less and transparent ova seen
with prominent nuclei in the cen-
ter; ova invisible to naked eye.
Stage II (Virgin maturing) Ovary slightly thicker,
translucent and yellowish; occupying 1/3 to 1/2 of body
cavity; ova invisible to naked eye; under a microscope,
translucent ova seen with yolk granules around
nucleus.
CYNOGLOSSUSAREL CYNOGLOSSUS LIDA
JUVENILE (N 24) JUVENILE (n 24)
6.90%
5.10%
2.40%
5.10%
7.80%
59.20%
12.90%
, 44 70%
MALE (N 627)
5.40%
13.20%
MALE (N 718)
6.80%
5.00%
15,50%
48.40%
53.40%
1 9.30%
18.90%
FEMALE (N: 569)
7.10%
FEMALE (N: 640)
7.40%
12.00%
16.80%
12.30%
19.60%
■
PO
Q
CO
B
PR
D
TN
O
CF
Q
FI
n
■
FS
AM
O
Ml
65 go-
stage III (Maturing) Ovary yellowish, granular,
extending to more than 1/2 the length of body cavity,
with vascularization; ova small; under microscope,
opaque; nucleus hidden by yolk.
Stage IV (Mature) Ovary creamy yellow, with
Rajaguru: Biology of Cynoglossus arel and C lida from Indian waters
333
Table I
Percent IRI of various food items of male Cynoglossus arel caught commercially off Porto Novo, India, October 1981-
September 1982. 1
N = number of stomachs analyzed;
Crustacean fr.
= crustacean fragments
UI = unidentified. (Data presented to one decima
point;
0.0 denotes value of <0.05
and dash denotes absence of food item.)
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sep.
Oct.
Nov.
Dec.
Total
Food items A''
53
24
54
46
56
35
69
44
56
75
86
29
627
%
Polychaetes
28.0
0.3
62.2
52.4
25.2
60.4
57.9
75.6
70.1
73.2
56.1
19.1
580.5
(48.4)
Prawns
19.1
21.0
12.5
12.5
58.7
25.4
23.9
13.4
22.0
11.2
5.2
1.8
226.7
(18.9)
Crustacean fr.
23.5
15.6
14.0
25.2
10.9
9.6
8.1
3.6
0.6
4.3
4.2
70.9
190.5
(15.9)
Fish scales
26.6
53.1
10.4
2.8
1.4
2.7
2.3
1.8
4.9
3.7
26.7
0.3
136.7
(11.4)
Amphipods
0.4
0.2
0.7
1.0
3.3
0.6
7.5
5.1
0.0
2.4
2.2
3.4
26.9
(2.2)
Fish bone
1.3
—
—
—
—
—
—
_
—
—
5.2
0.6
7.1
(0.6)
Fish spine
—
5.8
—
—
0.0
—
0.0
—
—
—
—
_
5.8
(0.5)
Bivalves
0.1
0.3
0.1
0.1
0.0
0.3
0.1
0.1
0.0
4.2
0.3
—
5.7
(0.5)
Lingula sp.
—
—
—
5.4
0.0
—
—
—
—
—
—
0.0
5.4
(0.5)
Fishes
—
3.1
0.1
—
0.4
—
0.1
—
0.1
—
0.0
0.3
4.1
(0.3)
Crabs
0.9
0.4
0.0
0.4
0.1
0.8
0.1
0.1
0.4
0.2
0.1
—
3.6
(0.3)
Isopods
—
0.1
0.0
—
—
—
—
0.0
1.7
0.4
0.0
0.0
2.3
(0.2)
Algae
—
—
—
—
—
—
—
—
0.0
0.0
0.0
2.2
2.2
(0.2)
Fish eggs
—
—
0.0
—
—
—
—
—
0.0
0.1
—
1.3
1.4
(0.1)
Copepods
0.1
—
0.0
—
0.0
0.0
0.0
0.3
—
0.0
—
_
0.4
(0.0)
Gastropods
—
—
0.0
—
—
—
—
0.0
0.0
0.2
0.0
0.1
0.3
(0.0)
Squilla sp.
—
0.1
—
—
0.0
0.1
—
—
—
—
0.0
—
0.2
(0.0)
Cosdnodiscus
—
—
—
—
—
—
—
—
0.2
—
—
_
0.2
(0.0)
Brittle star
—
—
—
0.1
—
—
0.0
—
—
—
—
—
0.1
(0.0)
Medusae
—
—
—
—
—
—
—
—
0.0
0.0
0.0
—
0.0
(0.0)
Egg mass (UI)
—
—
—
—
—
—
—
—
—
—
—
0.0
0.0
(0.0)
Nematode
0.0
—
0.0
—
—
—
0.0
0.0
—
—
—
_
0.0
(0.0)
Sand dollar
—
—
—
—
—
—
—
_
_
—
0.0
—
0.0
(0.0)
Echinoderm (UI)
—
—
—
—
—
—
—
—
0.0
—
—
—
0.0
(0.0)
prominent blood vessels; occupying 2/3 of body cavity;
ova visible to naked eye and rich with yolk.
Stage V (Ripe) Ovary resembling Stage IV, but oc-
cupying more than 2/3 of the body cavity; under a
microscope, ova slightly translucent with yolk granules;
ova not running out of genital aperture on application
of gentle pressure.
Stage VI (Oozing) Ovary yellowish and transpar-
ent, occupying entire length of body cavity; ripe ova
running out through genital aperture on application of
gentle external pressure on ovary; imder a microscope,
ova transparent.
Stage VII (Spent) Not found during the study.
Results
Feeding ecology
Food composition In Cynoglossus arel, polychaetes
made up the bulk (44.5-48.4% IRI) of the diet of adults
(Tables 1-2, Fig. 2). At least 11 species of polychaetes,
viz, Nephtys polybranchia, N. oligobranchia, Clymene
annandalei, Phyllodoce sp., Ancistrosyllis constricta,
Nereis chilkaensis, Diopatra sp., Onuphis sp., Eunice
sp., Terebellides stroemi, and Stemaspis sp., were con-
sumed. The next most important prey items were
prawns (18.9-19.6% IRI), crustacean fragments (15.9-
16.8% IRI), and fish scales (11.4-12.0% IRI). The
prey species which were consumed in smaller quantities
included bivalves (represented by Amussium sp..
Placenta sp.. Area sp., and Pinna sp.), gastropods
(by Umbonium sp., Turritella sp., and Dentalium sp.),
and fishes (by gobiids and Cynoglossus monopus)
(Tables 1-2).
In adult C. lida, polychaetes (same species as in
C. arel) dominated (53.4-65.9% IRI) (Tables 3-4, Fig.
2), while crustacean fragments (14.4-19.3% IRI) and
prawns (12.3-15.5% IRI) ranked next in importance.
The prey species which were consumed in smaller quan-
tities included bivalves (represented by Placenta sp.),
gastropods (by Umbonium sp. and Turritella sp.),
and fishes (by gobiids and Cynoglossus monopus)
(Tables 3-4).
Food of juveniles and adults Differences can be
seen in stomach contents between juveniles and adults
of both species (Tables 1-5, Fig. 2). Larger tongue-
fishes ate larger individuals of food species than did
334
Fishery Bulletin 90(2), 1992
Table 2
Percent IRI of various food items of female Cynoglossus
irel caught commercially
off Porto Novo,
India,
October 1981-September |
1982. See Table 1 for abbreviations.
Food items N
Jan.
39
Feb.
45
Mar.
49
Apr.
55
May
54
June
40
July
56
Aug.
54
Sep.
35
Oct.
54
Nov.
54
Dec.
34
Total
%
569
Polychaetes
4.6
10.7
25.2
74.4
19.6
36.4
60.7
71.9
71.6
72.9
37.4
48.3
533.7
(44.5)
Prawns
30.8
17.9
26.4
9.5
41.0
41.1
19.0
3.3
10.8
14.3
12.7
8.7
235.5
(19.6)
Crustacean fr.
28.0
35.1
25.4
11.8
33.6
17.6
5.1
2.1
4.6
2.4
7.6
28.7
202.0
(16.8)
Fish scales
21.6
23.2
14.3
2.0
0.7
1.5
8.0
20.9
1.1
9.0
40.5
0.7
143.5
(12.0)
Amphipods
0.2
0.7
1.3
1.4
2.1
1.8
5.4
0.9
—
0.3
0.8
3.2
18.1
(1.5)
Fish bone
13.9
—
—
—
—
—
—
—
—
0.2
0.0
—
14.1
(1.2)
Fish spine
—
7.9
2.6
—
—
—
0.0
0.0
-
-
-
-
10.5
(0.9)
Crabs
0.7
0.0
1.6
0.2
2.6
1.4
0.1
0.0
0.7
0.1
0.4
2.6
10.4
(0.9)
Isopods
0.1
—
0.0
—
—
0.0
0.0
0.0
7.3
0.0
—
0.0
7.4
(0.6)
Fish eggs
—
0.0
—
—
—
—
—
—
—
0.0
0.0
6.1
6.1
(0.5)
Bivalves
0.1
1.6
1.0
0.1
0.0
0.0
0.0
0.0
1.7
0.1
0.2
0.5
5.3
(0.4)
Fishes
_
2.3
0.4
—
—
0.1
0.4
0.9
0.7
0.1
0.2
—
5.1
(0.4)
Lingula sp.
—
—
1.8
0.6
—
—
—
0.0
-
-
-
-
2.4
(0.2)
Gastropods
0.0
—
0.0
0.0
0.0
—
—
—
1.2
0.0
0.1
0.7
2.0
(0.2)
Lucifer
—
—
—
—
—
—
1.2
—
0.0
—
—
0.1
1.3
(0.1)
Copepods
—
0.6
—
—
0.4
—
0.0
0.0
—
0.0
—
0.0
1.0
(0.1)
CiHates (UI)
—
—
—
—
—
—
—
—
—
0.5
—
—
0.5
(0.0)
Algae
—
—
—
—
—
—
—
—
0.0
0.0
0.0
0.4
0.4
(0.0)
Coscinodiscus
—
—
—
—
—
—
—
—
0.3
—
—
—
0.3
(0.0)
Squilla sp.
—
—
—
0.0
0.0
—
0.0
—
—
—
0.1
0.0
0.1
(0.0)
Brittle star
—
—
—
—
—
—
—
0.0
—
0.1
—
—
0.1
(0.0)
Medusae
—
—
—
—
—
—
0.1
—
0.0
—
—
—
0.1
(0.0)
Sand dollar
—
—
—
—
—
0.1
—
—
—
—
—
—
0.1
(0.0)
Nematode
0.0
0.0
0.0
—
0.0
—
—
0.0
—
—
—
0.0
0.0
(0.0)
Egg mass (UI)
—
—
—
0.0
—
—
—
0.0
—
—
0.0
0.0
0.0
(0.0)
Tube-like worm
—
—
—
—
—
—
—
—
0.0
—
—
—
0.0
(0.0)
Jelly fish
—
—
—
—
—
—
0.0
—
—
—
—
—
0.0
(0.0)
Sepia
—
—
—
—
—
0.0
—
—
—
—
—
—
0.0
(0.0)
smaller tonguefishes. In C. arel, amphipods (59.2% IRI)
dominated diets of juveniles, followed by tintinnids
(12.4% IRI). Smaller-sized prawns (8.2% IRI), copepods
(6.9% IRI), and polychaetes (6.4% IRI) were next in
importance. Fish remains, isopods, smaller crabs, and
nematodes were found in decreasing order of impor-
tance and never composed more than 5% of the IRI.
Breadth of the diet is much smaller in juveniles than
adults (compare Tables 1-4 and 5). Only 10 types of
food items occurred in stomachs of relatively few
juveniles examined, whereas 29 different types of prey
were noted in stomachs of adult C. arel (Tables 1-2,
5). In adult stomachs, fewer amphipods and more
polychaetes were found than in juvenile stomachs.
Prawns were the third most important prey in the diet
of the juveniles, whereas in adults they were the sec-
ond most important. Algal filaments were found only
in stomachs of adults, while tintinnids were found only
in stomachs of juveniles.
Juvenile C. lida fed on only 10 types of prey items
and usually smaller sizes, whereas adults consumed 24
types of relatively large-sized prey items (Tables 3-5).
Copepods (44.7% IRI) were preyed upon predominantly
by juveniles of C. lida (Fig. 2), while polychaetes were
dominant in the diet of adults. Hyperiid amphipods
(13.2% IRI), which were of secondary importance and
abundant in the diet of juveniles, occurred in smaller
quantities in adult stomachs. Crustacean fragments
were the second most important food item for adults.
Other food items of juveniles are listed in Table 5.
However, the sample sizes for the juveniles of both
tonguefishes are quite smaller than those of the adults.
Therefore, the differences in number of prey in adults
and juveniles may reflect differences in sample sizes.
Food of males and females A total of 76% of males
and females of C. arel, and 65% of males and 73%
of females of C. lida, had identifiable prey in their
stomachs. In C. lida, females consumed 19 types of
food items and males consumed 24 types (Tables 3-4).
Polychaetes were relatively more abundant (Fig. 2) in
the diet of females than males (65.9% vs. 53.4% IRI).
Rajaguru: Biology of Cynoglossus are! and C lida from Indian waters
335
Table 3
Percent IRI of various food items of male Cynoglossus lida caught
commercially off Porto Novo, India, October 1981-September 1982. |
See Table 1 for abbreviations.
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sep.
Oct.
Nov.
Dec.
Total
Food items N
32
59
64
84
38
56
25
43
46
83
160
28
718
%
Polychaetes
5.4
42.3
31.2
89.4
21.4
67.6
27.8
93.8
87.4
83.0
89.2
2.8
641.3
(53.4)
Crustacean fr.
64.5
12.6
26.4
1.0
16.5
4.9
18.2
0.1
2.5
0.7
1.6
83.1
232.1
(19.3)
Pravms
17.5
9.6
26.6
3.8
57.3
21.5
23.3
3.6
8.0
6.2
5.1
3.8
186.3
(15.5)
Fish scales
6.4
20.4
12.7
0.8
2.7
0.2
3.8
0.2
0.5
6.5
0.6
5.2
60.0
(5.0)
Amphipods
0.6
2.4
0.2
3.2
1.9
5.7
24.7
2.0
_
3.4
3.3
0.2
47.6
(4.0)
Algae
5.2
4.3
—
—
—
—
—
0.0
0.0
—
_
4.4
13.9
(1.2)
Medusae
—
8.3
—
—
—
—
—
—
—
—
0.0
_
8.3
(0.7)
Lucifer
0.0
—
—
—
0.1
—
2.1
—
0.1
—
—
—
2.3
(0.2)
Copepods
0.0
—
2.0
0.0
—
—
—
0.0
0.0
—
—
0.1
2.1
(0.2)
Lingula sp.
—
0.0
0.2
1.6
—
—
—
—
—
_
0.0
—
1.8
(0.2)
Crabs
—
0.1
0.4
0.1
0.1
0.1
—
0.1
0.0
—
0.0
0.1
1.0
(0.1)
Isopods
0.3
0.0
0.3
—
—
—
—
—
0.3
—
—
0.0
0.9
(0.1)
Fishes
—
—
—
0.1
—
—
—
—
0.7
—
—
—
0.8
(0.1)
Fish eggs
0.1
0.1
—
—
—
—
—
—
—
—
—
0.3
0.5
(0.0)
Gastropods
—
0.0
—
—
—
—
—
0.1
0.2
—
0.1
—
0.4
(0.0)
Cosdruxiiscus
—
—
—
—
—
—
—
—
0.3
—
—
_
0.3
(0.0)
Bivalves
—
0.0
—
—
—
—
—
0.1
0.0
0.2
0.1
0.4
(0.0)
Nematode
0.0
0.0
—
—
—
—
—
_
0.0
—
—
0.0
0.0
(0.0)
Brittle star
—
—
—
0.0
—
—
—
—
—
_
—
—
0.0
(0.0)
Squilla sp.
-
0.0
-
-
—
—
—
—
—
—
0.0
—
0.0
(0.0)
Fish spine
—
—
—
0.0
—
—
—
—
—
—
0.0
—
0.0
(0.0)
Octopus sp.
—
0.0
—
—
—
—
—
—
—
—
—
_
0.0
(0.0)
Tape worm
—
—
—
0.0
—
—
—
—
—
—
—
—
0.0
(0.0)
Egg mass (UI)
—
0.0
—
-
-
-
-
-
-
-
-
-
0.0
(0.0)
Table 4
Percent IRI of various food items of female Cynoglossus lida caught commercially off Porto Novo, India, October 1981-September
1982. See Table 1 for abbreviations.
Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Total
Food items A^ 42 16 18 51 31 59 25 62 25 76 207 28
Polychaetes
Crustacean fr.
Prawns
Amphipods
Fish scales
Fish eggs
Lingula sp.
Algae
Crabs
Isopods
Bivalves
Squilla sp.
Nematode
Copepods
Gastropods
Coscinodiscus
Medusae
Egg mass (UI)
Fish spine
17.7 58.8 58.1 84.5 60.8 73.4 63.3 95.0 93.4 87.9 94.2
48.5
26.5
0.3
4.4
0.4
1.3
0.1
0.0
0.3
0.1
0.3
0.0
0.0
0.1
0.0
12.5
13.6
12.0
0.1
1.2
1.0
0.1
0.5
0.1
0.1
7.7
17.3
8.2
1.2
0.3
6.7
0.2
0.2
0.1
2.0
7.6
4.1
0.0
0.0
5.2
31.3
2.4
0.2
0.1
0.0
5.0
20.3
0.9
0.4
0.0
0.0
0.0
0.0
10.3
16.7
7.9
1.4
0.6
1.8
2.6
0.0
0.0
0.0
0.3
0.1
1.2
2.9
0.9
0.9
0.3
0.0
0.3
0.1
0.8
4.2
0.9
6.0
0.1
0.1
0.0
0.0
0.3
2.5
2.8
0.0
0.0
0.0
0.2
0.0
0.0
0.0
4.0
78.7
2.5
0.0
2.6
7.5
4.6
0.1
0.0
640
0.0
791.1
172.8
147.2
42.1
16.2
9.2
8.5
7.1
1.3
1.3
1.2
0.6
0.4
0.3
0.4
0.2
0.1
0.0
0.0
(65.9)
(14.4)
(12.3)
(3.5)
(1.4)
(0.8)
(0.7)
(0.6)
(0.1)
(0.1)
(0.1)
(0.1)
(0.0)
(0.0)
(0.0)
(0.0)
(0.0)
(0.0)
(0.0)
336
Fishery Bulletin 90(2). 1992
Table 5
Percent IRI of various
, food items of juvenile Cynoglossus
arel and C
. lida caught
commei
-cially off Porto Novo
, India,
October
1981-September
1982.
See Table 1 for abbreviations. (Data presented
to one decimal point; dash denotes absence of food item.)
C.
arel
C. lida
Mar.
June
Oct.
Nov.
Total
Apr.
Nov.
Dec.
Total
Food items
N
6
5
7
6
24
%
8
9
7
24
%
Amphipods
50.0
83.0
60.4
43.5
236.9
(59.2)
3.0
32.8
3.9
39.7
(13.2)
Copepods
7.6
—
0.1
19.9
27.6
(06.9)
75.9
1.7
56.4
134.0
(44.7)
Tintinnids
11.1
—
16.4
21.9
49.4
(12.4)
—
12.0
3.3
15.3
(5.1)
Polychaetes
17.1
8.5
—
—
25.6
(6.4)
13.3
18.7
6.7
38.7
(12.9)
Prawns
—
1.3
20.4
10.9
32.6
(8.2)
2.5
12.7
—
15.2
(5.1)
Fish scales
1.0
1.6
1.2
3.0
6.8
(1.7)
1.7
18.4
6.3
26.4
(8.8)
Fishes
—
—
—
—
—
—
—
—
23.4
23.4
(7.8)
Crustacean fr.
11.1
5.1
—
—
16.2
(4.1)
—
0.8
—
0.8
(0.3)
Isopods
2.2
—
0.8
0.8
3.8
(1.0)
—
—
—
—
—
Ciliates (UI)
—
—
—
—
—
—
3.6
—
—
3.6
(1.2)
Nematode
—
0.3
0.1
0.1
0.5
(0.1)
—
3.0
—
3.0
(1.0)
Crabs
—
—
0.6
—
0.6
(0.2)
"
"
"
In C. arel, males consumed 24 types of food and
females had 28 types of prey (Tables 1-2). Polychaetes
were slightly more important in diets of males (48.4%
IRI) than in females (44.5% IRI) (Tables 1-2, Fig. 2).
The difference in numbers of identified prey in males
and females is due to rare species occurring in some
individuals.
Table 6
Percent frequency of occurrence (%F), percent of total number (%N), percent of total volume (%V), and index of relative importance |
(IRI) for food items of male and female Cynoglossus arel caught off Porto Novo
India, October 1981
-September 1982. Size
groups:
95-139mmTL(w 10 o-,
n 15 9), and 140-184 mm TL (
n 122 CT, n 96 9) combined.
n = number of stomachs ana
lyzed; Crustacean fr.
= crustacean fragments
UI = unidentified.
(Data presented to om
> decimal point;
0.0 denotes value of < 0.05, and dash denotes absence |
of food item.)
Food items
Male {n
132)
Female (n
Ill)
%F
%V
%N
IRI
%IRI
%F
%V
%N
IRI
%IRI
Polychaetes
27.7
62.8
30.7
2590.0
66.3
24.8
52.7
27.8
1996.4
62.7
Crustacean fr.
15.8
12.5
24.4
583.0
15.0
16.1
15.0
31.5
748.7
23.5
Prawns
14.9
15.9
10.4
391.9
10.0
11.2
10.7
7.5
203.8
6.4
Fish scales
10.9
1.7
17.0
203.8
5.2
9.9
0.4
3.8
41.6
1.3
Amphipods
12.4
0.9
8.6
117.8
3.1
11.8
0.7
7.5
96.8
3.0
Bivalves
2.0
1.5
1.4
5.8
0.2
3.8
2.2
2.3
17.1
0.5
Gastropods
2.0
0.1
0.6
1.4
0.0
2.5
0.2
1.5
4.3
0.1
Isopods
1.5
0.5
0.5
1.5
0.0
1.2
0.4
0.4
1.0
0.0
Copepods
1.5
0.0
0.6
0.9
0.0
2.5
0.0
1.3
3.3
0.1
Crabs
1.5
0.6
0.5
1.7
0.0
3.2
2.1
1.5
11.5
0.4
Fish
0.5
2.0
0.2
1.1
0.0
1.2
11.9
1.0
15.5
0.5
Fish bone
0.5
0.0
0.2
0.1
0.0
—
—
—
—
—
Fish spine
1.0
0.0
0.5
0.5
0.0
—
—
—
—
—
Fish egg
1.9
0.1
1.1
2.3
0.1
3.8
0.4
7.7
30.8
1.0
Squilla
—
_
_
_
—
0.6
0.0
0.2
0.1
0.0
Linffula sp.
1.0
0.5
1.1
1.6
0.0
1.2
1.5
2.9
5.3
0.2
Nematode
0.5
0.0
0.2
0.1
0.0
1.2
0.1
0.8
1.1
0.0
Algae
2.4
0.1
1.1
2.9
0.1
3.1
0.1
1.3
4.3
0.1
Brittle star
1.0
0.4
0.5
0.9
0.0
—
—
—
—
—
Echinoderm (UI)
0.5
0.2
0.2
0.2
0.0
—
—
—
—
—
Egg mass (UI)
0.5
0.2
0.2
0.2
0.0
—
—
—
—
—
Lucifer
—
—
—
—
1.9
1.6
1.0
4.9
0.2
Rajaguru: Biology of Cynoglossus arel and C lida from Indian waters
337
Table 7
Percent frequency of occurrence (%F), percent of total number (%N), percent of total volume (%V), and index of relative importance |
(IRI) for food items of male and female Cynoglossus
arel caught ofl
Porto Novo,
India, October 1981-September 1982. Size
group:
185-229 mm TL (n 338 o-
n 229 9)
See Table 6 for abbreviations.
Food items
Male (n
338)
Female (n
229)
%F
%V
%N
IRI
%IRI
%F
%V
%N
IRI
%IRI
Polychaetes
28.2
57.5
30.5
2481.6
60.5
22.2
58.8
33.6
2051.3
58.8
Prawns
19.3
25.6
18.1
843,4
20.5
20.7
22.1
16.8
805.2
23.1
Crustacean fr.
14.7
8.6
18.2
394.0
9.6
14.2
7.9
18.0
367.8
10.5
Fish scales
13.6
1.7
18.1
269.3
6.6
12.4
0.6
7.1
95.5
2.8
Amphipods
11.6
0.7
7.3
92.8
2.3
12.4
0.8
9.3
125.2
3.6
Bivalves
3.3
1.4
1.5
9.6
0.2
2.3
3.5
4.0
17.3
0.5
Gastropods
0.4
0.0
0.1
0.0
0.0
1.2
0.3
2.6
3.5
0.1
Isopods
1.5
0.5
0.4
1.4
0.0
1.6
0.5
0.6
1.8
0.1
Copepods
1.3
0.0
0.4
0.5
0.0
1.2
0.0
0.3
0.4
0.0
Crabs
3.1
1.5
1.0
7.8
0.2
3.5
1.8
1.3
10.9
0.3
Fish
0.2
0.6
0.1
0.1
0.0
0.4
1.2
0.1
0.5
0.0
Fish egg
0.7
0.0
0.4
0.3
0.0
1.4
0.1
3.3
4.8
0.1
Squilla
—
—
—
—
—
0.4
0.2
0.1
0.1
0.0
Lingula sp.
0.6
1.6
3.4
3.0
0.1
0.8
0.2
0.5
0.5
0.0
Nematode
0.2
0.0
0.1
0.0
0.0
0.4
0.0
0.1
0.0
0.0
Algae
0.7
0.0
0.2
0.1
0.0
1.0
0.0
0.6
0.5
0.0
Egg mass (UI)
—
—
—
—
—
0.4
0.1
0.1
0.1
0.0
Lucifer
—
—
—
—
—
1.9
0.9
0.7
3.0
0.1
Sepia
—
—
—
—
—
0.2
0.1
0.1
0.0
0.0
Sand dollar
—
—
—
—
—
0.2
0.5
0.1
0.1
0.0
Tube-like worm
0.6
0.3
0.2
0.3
0.0
0.8
0.2
0.3
0.4
0.0
Jelly fish
—
—
—
—
—
0.2
0.1
0.1
0.0
0.0
CoscinodisciLS
—
—
—
—
—
0.2
0.0
0.3
0.1
0.0
Food vs. fish size In C. arel, the dominant size-group
in both sexes is 185-229 mm TL (54% of males, and
40% of females). Females in this size-group had eaten
23 types of prey, while males consumed only 16 types
(Table 7). In the remaining size-groups of both sexes,
there is no obvious difference in the number of prey
types consumed (Tables 6, 8-9). In both sexes of C. arel,
fish <275mmTL preyed predominantly on polychaetes
(55.1-66.3% IRI in males, 53.4-62.7% IRI in females),
whereas in fish >275mmTL the polychaetes were of
lesser importance (10.5% IRI in females, <5.0% IRI
in males), with fish remains being the most abundant
(54.5% IRI in males, and 48.9% IRI in females) (Tables
6-9, and Fig. 3).
In both sexes of C. lida, fewer prey types were
consumed by fish >200mmTL (8-9 prey types) and
by fish <136mmTL (10-13 types), compared with
fish 137-199mmTL (16-19 types). Among fish
<200mmTL, polychaetes were the most abundant
prey in both sexes (67.2-89.0% IRI in females,
61.0-81.5% IRI in males) (Tables 10-13, Fig. 4).
Among the fish >200mmTL, polychaetes were the
most abundant prey only in females (90.2% IRI),
whereas polychaetes were the second-most important
prey in males (28.2% IRI) and prawns the most abun-
dant prey (52.6% IRI) (Table 14, Fig. 4).
Seasonal variations in diet composition In male C.
arel, polychaetes were dominant, except in February,
May, and December (Table 1). During these 3 months,
other prey items, viz, prawns (in May), crustacean
fragments (in December), and fish remains (in
February) were more important in the diet. In females,
polychaetes also formed the primary food during 6
months (April, July, August, September, October, and
December). In other months, prawns (January, March,
May, and June), crustacean fragments (February), and
fish remains (November) were the primary food con-
sumed (Table 2).
In male C. arel, prawns were the secondary prey item
for 6 months (February, June, July, August, Septem-
ber, and October), with polychaetes in May and De-
cember, crustacean fragments in March and April, and
fish remains in January and November. In females,
crustacean fragments were the secondary prey item
for 5 months (January, March, April, May, and
December), prawns for 3 months (July, September, and
October), polychaetes for 2 months (June and
November), and fish remains for 2 months (February
and August). The tertiary food group in the diet of
338
Fishery Bulletin 90(2). 1992
Table 8
Percent frequency of occurrence (%F), percent of total number (%N), percent of total volume (%V), and index of relative importance |
(IRI) for food items of male and female Cynoglossus
arel caught off
Porto Novo,
India, October 1981-September 1982. Size
group:
230-274 mm TL (n 136 or
n 182 9). See Table 6 for abbreviations.
Food items
Male (n
136)
Female (re
182)
%F
%V
%N
IRI
%IRI
%F
%V
%N
IRI
%IRI
Polychaetes
23.2
46.8
21.7
1589.2
55.1
23.7
50.1
25.8
1798.8
53.4
Prawns
18.8
20.0
12.4
609.1
21.1
18.0
24.9
17.0
756.0
22.4
Crustacean fr.
14.8
10.0
18.6
423.3
14.7
17.3
11.9
24.5
629.7
18.7
Fish scales
10.3
0.5
3.6
42.2
1.5
10.7
0.5
4.9
57.8
1.7
Amphipods
11.8
1.1
6.3
87.3
3.0
10.5
0.4
3.8
44.1
1.3
Bivalves
3.3
3.0
2.8
19.1
0.7
2.7
1.0
1.0
5.4
0.2
Gastropods
2.6
0.1
0.6
1.8
0.1
1.6
0.0
0.3
0.5
0.0
Isopods
1.8
1.3
1.2
4.5
0.2
1.8
0.8
0.8
2.9
0.1
Copepods
1.5
0.0
0.3
0.5
0.0
—
—
—
—
—
Crabs
3.7
2.2
1.3
13.0
0.4
5.9
5.5
3.8
54.9
1.6
Fish
1.5
3.9
0.3
6.3
0.2
0.9
2.0
0.2
2.0
0.1
Fish bone
2.2
9.6
29.7
86.5
3.0
0.2
0.0
0.1
0.0
0.0
Fish spine
—
—
—
—
—
0.9
0.1
0.5
0.5
0.0
Fish egg
0.7
0.0
0.2
0.1
0.0
0.9
0.1
1.4
1.4
0.0
Squilla
1.1
0.6
0.2
0.9
0.0
0.9
0.5
0.3
0.7
0.0
Lingula sp.
0.4
0.0
0.1
0.0
0.0
0.9
1.3
3.0
3.9
0.1
Nematode
0.4
0.0
0.1
0.0
0.0
0.5
0.0
0.1
0.1
0.0
Algae
1.1
0.0
0.4
0.0
0.0
0.5
0.1
0.2
0.2
0.0
Egg mass (UI)
—
—
—
—
—
0.9
0.1
11.8
10.7
0.4
Lucifer
—
—
—
—
—
0.5
0.1
0.1
0.1
0.0
Sand dollar
0.4
0.8
0.1
0.4
0.0
—
—
—
—
—
Tube-like worm
0.4
0.1
0.1
0.1
0.0
—
—
—
—
—
Coscinodiscus
—
—
—
—
—
0.2
0.0
0.1
0.0
0.0
Brittle star
—
—
—
—
—
0.5
0.6
0.3
0.5
0.0
Table 9
Percent frequency of occurrence
(%F), percent of total number (%N), percent of total volume (%V), and index of relative importance 1
(IRI) for food items of male and female Cynoglossus
arel caught off Porto Novo,
India, October 1981
-September
1982. Size
groups:
275-319mmTL(ra 12 ct,
«24 9)
320-364
mmTL («
9 0-, n 8 9);
365-409 mmTL (re 0 cr, re
9 9), and 410-454 mmTL (n 0 o-
n6 9)
combined. See Table 6 for abbreviations.
Food items
Male (m
21)
Female (re 47)
%F
%V
%N
IRI
%IRI
%F
%V
%N
IRI
%IRI
Fish scales
13.9
31.2
66.3
1355.3
54.5
15.3
23.4
60.8
1288.3
48.9
Prawns
20.8
19.2
2.7
455.5
18.3
18.5
16.9
2.9
366.3
13.9
Crustacean fr.
25.0
11.6
4.9
412.5
16.6
16.1
19.6
10.2
479.8
18.2
Polychaetes
11.1
8.4
0.9
103.2
4.2
14.5
16.9
2.2
277.0
10.5
Amphipods
5.6
4.5
0.2
26.3
1.1
4.0
0.1
0.2
1.2
0.0
Bivalves
—
—
—
—
—
3.3
0.5
0.1
2.0
0.1
Gastropods
—
—
—
—
—
1.6
0.0
0.0
0.0
0.0
Isopods
4.2
0.7
0.2
3.8
0.2
0.8
0.1
0.0
0.1
0.0
Copepods
—
—
—
—
—
1.6
0.1
3.0
5.0
0.2
Crabs
5.5
1.5
0.2
9.4
0.0
6.5
2.0
0.3
15.0
0.6
Fish
4.2
10.1
0.2
43.3
1.8
7.3
10.8
0.3
81.0
3.1
Fish bone
1.4
5.0
10.8
22.1
1.0
2.4
0.1
0.1
0.5
0.0
Fish spine
2.7
7.5
12.7
54.5
2.2
4.0
9.5
19.8
117.2
4.5
Fish egg
—
—
—
—
—
0.8
0.0
0.0
0.0
0.0
Squilla
1.4
0.3
0.0
0.4
0.0
—
—
—
—
—
Nematode
1.4
0.0
0.0
0.0
0.0
3.3
0.0
0.1
0.3
0.0
Algae
1.4
0.0
0.0
0.0
0.0
—
—
—
—
—
Coscinodiscus
1.4
0.0
0.9
1.3
0.1
—
—
—
—
—
Rajaguru: Biology of Cynoglossus arel and C lida from Indian waters
339
CYNOGLOSSUS AREL
MALE-1 (N 132)
5.2% 3.5%
10%
15%
66.3%
MALE-2 (N 338)
6.6% 2.8%
9.6%
20.5%
MALE-3 (
60.5%
N: 136)
9.1%
14.7%
21 1%
MALE-4 (
55.1%
N: 21)
10.6%
16.6%
54.5%
FEMALE-1 (
N: 111)
7.4%
6.4%
23.5%
FEMALE-2
62.7%
(N: 229)
7.6%
10.5%
23.1%
58.8%
FEMALE-3 (n i82)
5.5%
18.7% yi
22.4%
53.4%
FEMALE-4 (n 24)
8.5%
10.5% ^ —
7/>^
■ PO
yy^K
S PR
□ CF
13.9% [ ^
V/yx// 48.9%
0 FS
n Ml
\ .r^^ J' ^ A
t^^ V \ \ \
T^ / / ^ • y-
\^^ /■ y /■ /■
^
1 8.3%
Figure 3
Percentage contribution of food items to the diet of various size groups of male and female Cynoglossus
arel caught commercially off Porto Novo, India, October 1981-September 1982. Male-1 {n 132) and
Female-1 (» 111) = size group 95-184 mm TL; Male-2 (n 338) and Female-2 {n 229) = 185-229mmTL;
Male-3 (n 136) and Female-3 (n 182) = 230-274mmTL; lVIale-4 (n 21) = 275-364 mm TL; and Female-4
{n 47) = 275-454 mm TL. Only values >5% IRI are individually shown; values <5% IRI are clumped
together into a single category, the unshaded wedge of the pie chart. PO = polychaetes, PR = prawns,
CF = crustacean fragments, FS = fish scales, MI = miscellaneous.
340
Fishery Bulletin 90(2). 1992
Table
10
Percent frequency of occurrence (%F), percent of total number (%N), percent of total volume (%V), and index of relative importance |
(IRI) for food items of male and female Cynoglossus
lida caught off
Porto Novo,
India, October 1981
-Septembei
• 1982. Size
groups;
95-115mmTL (« 6 o-, re
13 9),
and 116-136 mm TL (re 85 a, re
56 9) combined.
re = number of stomachs analyzed; Crustacean fr. 1
= crustacean fragments
(Data presented
to one decimal point;
0.0 denotes value of <0.05
and dash denotes absence of food item.)
Food items
Male (n
91)
Female (re
69)
%F
%V
%N
IRI
%IRI
%F
%V
%N
IRI
%IRI
Polychaetes
29.0
73.4
42.6
3364.0
81.5
30.4
66.2
35.4
3088.6
70.2
Prawns
16.9
10.9
8.5
327.9
8.0
17.4
13.8
9.9
412.4
9.4
Crustacean fr.
8.9
8.4
19.6
249.2
6.1
13.0
16.2
34.5
659.1
15.0
Fish scales
13.0
0.8
9.0
22.1
0.5
10.9
0.4
4.3
51.2
1.2
Amphipods
13.0
0.6
6.7
94.9
2.3
17.4
0.8
8.1
154.9
3.5
Bivalves
—
—
—
—
—
1.1
0.3
0.3
0.7
0.0
Isopods
1.6
0.5
0.5
1.6
0.0
1.1
0.3
0.3
0.7
0.0
Copepods
1.6
0.0
0.5
0.8
0.0
—
—
—
—
—
Crabs
2.4
1.7
1.3
7.2
0.2
—
—
—
—
—
Fish egg
2.4
0.1
1.3
3.4
0.1
1.1
0.0
0.9
1.0
0.0
Lingxda sp.
4.0
2.2
5.0
28.8
0.7
3.3
1.9
4.1
19.8
0.5
Algae
5.6
0.2
4.2
24.6
0.6
4.3
0.1
2.2
9.9
0.2
Lucifer
0.8
0.3
0.3
0.5
0.0
—
—
—
—
—
Brittle star
0.8
0.9
0.5
1.1
0.0
—
—
—
—
—
Table
11
Percent frequency
of occurrence
(%F), percent of total number (%N), percent of total volume (%V),
and index of relative importance 1
(IRI) for food items of male and female Cynoglossiis lida caught off Porto Novo,
India, October 1981-September 1982. Size
group:
137-157mmTL(n
211 cr, re 136 9). See Table 10 for abbreviations
Food items
Male (n
211)
Female (re
136)
%F
%V
%N
IRI
%IRI
%F
%V
%N
IRI
%IRI
Polychaetes
25.5
61.7
28.4
2297.6
61.0
30.0
62.0
29.9
2757.0
67.2
Crustacean fr.
15.3
16.7
30.7
725.2
19.2
15.5
16.5
31.8
748.7
18.3
Prawns
15.6
13.6
8.3
341.6
9.1
14.5
17.0
10.9
404.6
9.9
Fish scales
16.3
1.4
12.8
231.5
6.1
11.4
0.5
5.2
65.0
1.6
Amphipods
12.3
1.1
10.5
142.7
3.9
13.6
0.6
5.5
83.0
2.0
Bivalves
1.8
0.8
0.7
2.7
0.1
1.8
0.7
0.7
2.5
0.1
Gastropods
0.3
0.0
0.1
0.0
0.0
0.9
0.0
0.2
0.2
0.0
Isopods
0.9
0.4
0.4
0.7
0.0
1.4
0.4
0.4
1.1
0.0
Copepods
0.9
0.0
0.6
0.5
0.0
0.9
0.0
0.3
0.3
0.0
Crabs
1.3
0.7
0.4
1.4
0.0
0.9
0.7
0.5
1.1
0.0
Fish
0.3
1.1
0.1
0.4
0.0
—
—
—
—
_
Fish spine
0.3
0.0
0.2
0.1
0.0
_
_
—
—
—
Fish egg
0.6
0.0
0.2
0.1
0.0
2.3
0.5
9.7
23.5
0.6
Squilla
—
—
—
—
—
0.4
0.2
0.1
0.1
0.0
Lingula sp.
0.9
1.0
2.0
2.7
0.1
1.4
0.7
1.3
2.8
0.1
Nematode
0.3
0.0
0.1
0.0
0.0
1.4
0.0
0.5
0.7
0.0
Algae
5.3
0.2
3.3
18.6
0.5
3.2
0.2
2.9
9.9
0.2
Lucifer
0.9
0.4
0.2
0.5
0.0
Coelenterate
0.9
0.7
0.9
1.4
0.0
Cosnnodis<ms
—
—
—
—
—
0.4
0.0
0.1
0.0
0.0
Tape worm
0.3
0.2
0.1
0.1
0.0
—
—
—
—
—
Rajaguru: Biology of Cynoglossus arel and C lida from Indian waters
341
Table
12
Percent frequency of occurrence
(%F), percent of total number (%N), percent of total volume (%V), and index of relative importance 1
(IRI) for food items of male and female Cynoglossus lida caught off Porto Novo,
India, October 1981-September 1982. Size
group:
158-178 mm TL (re 284
Of, n 260
9). See Table 10 for abbreviations
Food items
Male in
284)
Female (re
260)
%F
%V
%N
IRI
%IRI
%F
%V
%N
IRI
%IRI
Polychaetes
29.0
61.2
28.4
2598.4
67.4
32.2
75.9
47.4
3970.3
81.3
Crustacean fr.
11.2
16.2
30.1
518.6
13.5
10.7
8.2
20.4
306.0
6.3
Prawns
16.2
14.9
9.2
390.4
10.1
18.3
12.5
10.4
419.1
8.6
Fish scales
14.9
1.4
12.7
210.1
5.6
8.9
0.4
4.4
42.7
0.8
Amphipods
9.7
0.8
7.1
76.6
2.0
14.7
0.6
8.1
127.9
2.6
Bivalves
0.5
0.4
0.4
0.4
0.0
2.5
0.7
0.9
4.0
0.1
Gastropods
2.6
0.2
1.2
3.6
0.1
1.4
0.1
0.5
0.8
0.0
Isopods
—
—
—
—
—
0.5
0.1
0,1
0.1
0.0
Copepods
0.8
0.0
0.2
0.2
0.0
0.8
0.0
0.3
0.2
0.0
Crabs
1.3
0.6
0.4
1.3
0.0
1.3
0.3
0.3
0.8
0.0
Fish spine
0.3
0.0
0.2
0.1
0.0
—
—
—
—
—
Fish egg
1.3
0.0
0.5
0.7
0.0
1.4
0.1
3.1
4.5
0.1
Squilla
0.5
0.5
0.2
0.4
0.0
1.0
0.4
0.3
0.7
0.0
Lingula sp.
0.5
0.1
0.2
0.2
0.0
0.8
0.4
1.0
1.1
0.0
Nematode
0.8
0.0
0.2
0.2
0.0
1.0
0.0
0.5
0.5
0.0
Algae
7.2
0.3
4.7
36.0
0.9
3.6
0.1
2.0
7.6
0.2
Egg mass (UI)
0.3
0.1
0.1
0.1
0.0
0.3
0.1
0.1
0.1
0.0
Coelenterate
2.1
3.2
3.9
14,9
0.4
0.3
0.1
0.1
0.1
0.0
Coscinadiseus
0.5
0.0
0.2
0.1
0.0
0.3
0.0
0.1
0.0
0.0
Octopus sp.
0.3
0.1
0.1
0.1
0.0
—
—
—
—
—
Table
13
Percent frequency
of occurrence
(%F), percent of total number (%N), percent of total volume (%V), and index of relative importance 1
(IRI) for food items of male and female Cynoglossus lida caught of
Porto Novo,
India, October 1981-September 1982. Size
group:
179-199 mm TL (re
111
Cf, re 144
9), See Table 10 for abbreviations
Food items
Male (re
111)
Female (n
144)
%F
%V
%N
IRI
%IRI
%F
%V
%N
IRI
%IRI
Polychaetes
31.8
60,0
33.7
2979.7
70,8
35.7
82.4
56.5
4958.7
89.0
Prawns
20.6
16.9
12.6
607.7
14.4
14.9
9.1
8.3
259.3
4.7
Crustacean fr.
11.2
12,8
28.8
465.9
11.1
10.0
5.4
14.9
203.0
3.6
Fish scales
10.6
0.5
6.0
68.9
1.6
10.0
0.3
3.5
38.0
0.7
Amphipods
10.0
0.4
4.7
51.0
1.2
13.1
0.5
6.6
93.0
1.7
Bivalves
1.2
0.4
0.5
1.1
0.2
2.3
0.5
0.6
2.5
0.0
Gastropods
0.6
0.0
0.2
0.1
0.0
1.4
0.2
1.5
2.4
0.0
Isopods
1.2
0.3
0.3
0.7
0.0
3.2
0.5
0.7
3.8
0.1
Copepods
1.2
0.0
0.3
0.4
0.0
0.4
0.0
0.1
0.0
0.0
Crabs
1.2
0.4
0.3
0.8
0.0
1.4
0.3
0.3
0.8
0.0
Fish
0.5
1.8
0.2
1.0
0.0
—
—
—
—
—
Fish spine
—
—
—
—
—
0.4
0.0
0.2
0.1
0.0
Fish egg
1.8
0.0
0.5
0.9
0.0
2.3
0.1
2.8
6.7
0.1
Lingula sp.
—
—
—
—
—
0.4
0.6
1.8
1.0
0.0
Nematode
1.2
0.0
0.3
0.4
0.0
0.9
0.0
0.2
0.2
0.0
Algae
4.7
0.1
2.8
13.6
0.3
2.7
0.1
1.6
4.6
0.1
Lucifer
0.5
0.4
0.3
0.4
0.0
—
—
—
—
—
Coelenterate
1.2
6.0
8.3
17.2
0.4
—
_
_
—
—
Coscirwdiscus
0.5
0.0
0.2
0.1
0.0
0.9
0.0
0.4
0.4
0.0
342
Fishery Bulletin 90(2), 1992
CYNOGLOSSUS LIDA
MALE-1 (N: 91)
6.1% 4.4%
FEMALE-1 (N 69)
5.4%
9.4%
15%
81.5%
MALE-2 (N: 211)
6.1% 4.6%
702%
FEMALE-2 (N: 136)
9.9%
4.6%
9.1%
1 9.2% V
18.3%
61%
MALE-3 (N: 284)
5.6% 3.4%
67.2%
FEMALE-3 (N: 260)
63%. 3.8%
10.1%
13.5%
8.6%
67.4%
MALE-4 (N: 1 1 1)
11,1% 3.7%
81.3%
FEMALE-4 (N: 144)
11%
14.4%
1 1 .6%
70.8%
MALE-5 (N; 21)
7.6%
52.6%
89%
FEMALE-5 (N 31)
7.3% 2.5%
28.2%
■ PO
m PR
□ CF
a Fs
n Ml
90.2%
Figure 4
Percentage contribution of food items to the diet of various size groups of male and female
Cynoglossvs lida caught commercially off Porto Novo, India, October 1981 -September
1982. Male-1 (n 91) and Female-1 (w 69) = size group 95-136 mm TL; Male-2 {n 211)
and Female-2 (n 136) = 137-157 mm TL; Male-3 (n 284) and Female-3 (n 260) =
158-178 mm TL; Male-4 (n 1 11) and Female-4 (w 144) = 179-199mmTL; and Male-5 (w
21) and Female-5 (n 31) = 200-262 mm TL. Only values >5% IRI are individually shown;
values <5% IRI are clumped together into a single category, the unshaded wedge of
the pie chart. PO = polychaetes, PR = prawns, CF = crustacean fragments, FS =
fish scales, MI = miscellaneous.
male and female C arel is shown
in Figure 5. All other food items
occurred sporadically (Tables
1-2).
In male C. lida, polychaetes
were the dominant prey for 9
months (Table 3). In the remain-
ing months, crustacean frag-
ments (January and December)
and prawns (May) dominated. In
females, polychaetes were the
primary food item for every
month, except in January and
December when crustacean frag-
ments were the most important
prey item (Table 4).
Prawns were next in impor-
tance in both sexes of C. lida. In
males, prawns formed the sec-
ondary prey item except in Feb-
ruary, May, July, October, and
December. During these 5
months, fish remains (February,
October, and December), poly-
chaetes (May), and amphipods
(July) were the secondary prey.
In females, prawns were the
secondary food, except in August
and October-December. During
these 4 months, amphipods (Aug-
ust and November), fish remains
(October), and fish eggs (Decem-
ber) were consumed by females.
The tertiary food group in the
diet of male and female C. lida
is shown in Figure 5. Organisms
of lesser importance are listed in
Tables 3-4.
Gastro- (Gl) and hepatosomatic
(HI) Indices and occurrence of
empty stomachs In relation to
spawning In male C arel, a
peak occurrence of empty stom-
achs (Fig. 6) occurred in January,
which is the peak spawning
period. Lowest gastro- and hepa-
tosomatic indices were also ob-
served in January (Fig. 6). How-
ever, over the rest of the year,
these factors did not appear to be
related. The gastro- and hepato-
somatic indices did not track the
percentage occurrence of empty
stomachs throughout the year.
Rajaguru: Biology of Cynoglossus arel and C lida from Indian waters
343
Table 14
Percent frequency of occurrence (%F), percent of total number (%N), percent of total volume (%V), and index of relative importance |
(IRI) for food items of male and female Cynoglossus
lida caught off
Porto Novo,
India, October 1981
-September
1982. Size
groups:
200-220mmTL(nl9o-
n 27 9);
221-241 mm TL (n 1 ct, n 3 9), and 245
-262mmTL(
M 1 o", M 1 9) combined. See Table 10 for abbrevations.
Food items
Male (n
21)
Female (» 31)
%F
%V
%N
IRI
%IRI
%F
%V
%N
IRI
%IRI
Polychaetes
25.0
25.7
16.4
1052.5
28.2
39.1
84.8
70.8
6084.0
90.2
Prawns
25.0
42.5
36.1
1965.0
52.6
21.7
10.7
11.9
490.4
7.3
Crustacean fr.
10.0
12.2
31.1
433.0
11.6
10.9
1.8
6.2
87.2
1.3
Fish scales
10.0
0.3
3.4
37.0
1.0
8.7
0.1
2.3
20.9
0.3
Amphipods
15.0
0.6
8.2
132.0
3.4
8.7
0.3
4.2
39.2
0.5
Bivalves
—
—
—
—
—
4.3
2.1
3.4
23.7
0.4
Isopods
5.0
1.3
1.6
14.5
0.4
—
—
—
—
—
Copepods
—
—
—
—
—
2.2
0.0
0.4
0.9
0.0
Crabs
5.0
2.0
1.6
18.0
0.5
_
—
—
—
—
Fish
5.0
15.4
1.6
85.0
2.3
_
—
_
_
_
Lingula sp.
—
—
—
—
—
2.2
0.2
0.4
1.3
0.0
Nematode
—
—
—
—
2.2
0.0
0.4
0.9
0.0
C. AREL: MALE C. AREL: FEMALE
CO
/k ^,
CO
j? n Q 0
s
/ ^^'''x ^
s
A / \
LU
/ \ '
LU
t 60-
fv p f — / \ '
t 60-
/ \ / \
Q
/ \i ■ '■ / \ 1
O
/ \ / \
O
."• / \ ' * / \ 1
o
/ \ / \
o
■' '• / \ ■' »/ \ '
o
/ \ / \ /^
u.
' / \ • / \ '
Li.
\ / \ /
c 40-
■■ •/ \ A \
IE ^°-
\ v"i V
o
'./ ;\ / ■, i\
o
'-■n / V A
-)
■' r •' \ / ' ' \
S' ^ / / 1 \ / \ ■' '• B
<
Hi' •' V ' • ' \
<
*" ^i 'A / « •• ^
= 20-
Ll.
-if\ ^\ X A / X
O
s?
Vfv^^ — "' v;\
O
5?
7 ^ V ^ •< •■ />- -"?■ -°''-
FT '* -'■■-. 'v.-' >B- •' ^
V •••.....,....--.:.><--v'*-«-v
0 -*— 1 — >-V— ■— 1 — T-T — >— j — •—[ — ' 1 ■ 1 — ' T ' 1 — '— 1 — ■ T 1 " 1 1 ■ 1
JFMAMJJASOND JFMAMJJASOND
MONTHS MONTHS
C. LIDA: MALE C. LIDA: FEMALE
w
r\.
CO
r^"^^.^
s
A / ^^*^~^r(''^
E
- 1 ^^"^ 1
UJ 80-
A / \ J
^ 80-
/\ / \ '
o
/ I B / \ ''
O
/ \ /"n, / \ 1
o
' / \ A / \ '
o
1 \/ l^ \ '
O so-
\ 1 \ O 1 \ 1 \ 1
O 60 -
P Q 1 '
il.
\ \ ■' \ 1 \'
Li.
/ \l
EC
\ V '■ \ V
CE
^ I \
o
\ Q / y Y \ / J
o
\ 1 \
T 40 -
^ /x 1 \ i\ \ 1 A
-3 40 -
\ 1 \
<
^ / ^\ / ' \ / '» \ / i\
<
I '\
s
/ 0 M/ \ 6 / 1
s
oA A-, ' \
u.
/\ /\ .' g b- -9 1 1
u.
A « ' * 1 \
O 20-
°IA': \ '■ f^ /""o
O 20-
Q "■-> .-o. / "■"-■o 1 \
s?
s?
ft' -■, / \ ' \
J FMAMJJASOND ..^TI ^ JFMAMJJASOND
MONTHS --0-- PR MONTHS
-■-■•-■■ FS
Figure 5
Seasonal variations in percentage contribution of major food items to the diet of male and female Cynoglossus arel and C. lida caught
commercially off Porto Novo, India, October 1981-September 1982. Only values >5% IRI value are graphed. (Refer to Appendix for
monthly sample size). PO = polychaetes, CF = crustacean fragments, PR = prawns, FS = fish scales.
344
Fishery Bulletin 90(2). 1992
MALE
EMPTY STOMACHS
— o — C AREL
•-•- C LIDA
If)
X
o
<
o
I-
01
a.
N; 50 (C. AREL) N: 129 (C. LIDA)
40-
30 -
20:
9 ' '/ V '
10 ■;
\y V^^.,,^^ /\A^^
0-
^ — f — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — r
J F
M A M J J A S
MONTHS
O N D
GASTRO-SOMATIC INDEX
N: 627 (C. AREL) N: 718 (C. LIDA)
T — I — I — I — I — I — I — r-"T — I — r
J FMAMJ JASOND
MONTHS
HEPATO-SOMATIC INDEX
N: 627 (C. AREL) N: 7 18 (C. LIDA)
0 95 -
FMAMJ JASOND
MONTHS
FEMALE
EMPTY STOMACHS
30
in
O 25
<
o 20
: N: 43 (C. AREL) N: 74 (C. LIDA)
15-
Q.
UJ
MONTHS
GASTRO-SOMATIC INDEX
N; 569 (C. AREL) N: 640 (C LIDA)
O N D
MONTHS
1.67
HEPATO-SOMATIC INDEX
N: 569 (C. AREL) N; 640 (C. LIDA)
J FMAMJ JAS
MONTHS
O N D
Figure 6
Seasonal variations in the percentage occurrence of empty stomachs, gastrosomatic index (GI), and hepa-
tosomatic index (HI) for male and female Cynoglossus arel and C. lida caught commercially off Porto Novo,
India, October 1981-September 1982. (Refer to Appendix for monthly sample size).
In female C. arel, occurrence of empty stomachs did
not correspond with spawning. However, the lowest
values of gastro- and hepatosomatic indices were
recorded (Fig. 6) only during the peak spavraing period
(in January).
In both sexes of C. lida, gastro-Zhepatosomatic in-
dices and the occurrence of empty stomachs did not
reveal any relationship (Fig. 6) with peak spawning ac-
tivities (in September) of this species.
Rajaguru: Biology of Cynoglossus are/ and C lida from Indian waters
345
>
o
z
UJ
o
>-
CJ
z
m
o
oc
CYNOGLOSSUS AREL
MARCH
25
20
15
10
N: 54 (M)
N: 51 (F)
ill iil
35
30
25
20
15
10
5 b
N: 24 (M)
N: 44 (F)
uuJL
30
25
>-
o
z
?n
UJ
3
o
1S
LU
DC
U.
10
m
N: 54 (M)
N: 39 (F)
Iil iil li i 11
lr)CDLfi0^r)OlJ^OLno^J^ou^OLnou^o^r)omom
CT»'-cMTjtnt^aDO'-rT^c£>r^cr)Oc\icnuotDcoai^oj
SIZE GROUPS (mm)
o
z
UJ
o
UJ
cr
ir)01pompLr)oi/^oir)omoinOLnoir)Oi/)oin
ai'-cM-^inr^coo-.-cT^(£)f^oiocvjcoir:c£)coaiT-c\j
'-'-»-'-»-»-c\jojc\jc\jc\jc\jojnmcoc^cr)cor)^^
SIZE GROUPS (mm)
FEBRUARY
30
25
JUNE
-
N: 35 (M)
N: 40 (F)
20
15
_
10
-
1
5
n
il
M
MALE
FEMALE
30
>. 25
U
^ 20
C3 15
iiJ
DC
^ 10
SIZE GROUPS (mm)
MAY
LoouooinoinoLnou^OLnoLoomoLnomoLO
CT)'-(M"^ior^ooO'-mtjtor-.oioc\jr)ir)CDCDcn'^c\j
'-'-^'-'-■--ojcNjcocMCMCNjojncocoronncO'^Tt
SIZE GROUPS (mm)
JANUARY
N: 56 (M)
N: 54 (F)
ifjOLnomoinoi/ioinomoinoLnomoirjOLr)
(7)'-c\j'^uir~-coO'^cO'^cDr~-aioc\jcoir)u>coa)»-csj
'-'-'-'-'-'-CMCOCMCSJCNJCMCMCOCOCOCOCOnCT^'^
SIZE GROUPS (mm)
>-
o
z
UJ
O
UJ
DC
25
APRIL
N: 48 (M)
20
-
N: 55 (F)
15
-
10
-
5
n
1
III
i/)Omoi/iomomomoiDomoLnoLnomoir)
O)'~cvj^ifir^c0O'— n^(Dr-^oioc\jc^i/)cDcocj)»— CM
'-'-■^'-•-'-CMCNJCMOJCNJCMCMCOCnncOCOCOCO'^'^
SIZE GROUPS (mm)
Figure 7A
Size-frequency histograms for
male (M) and female (F) Cynoglos-
sus arel (January-June) caught
commercially off Porto Novo,
India, October 1981-September
1982.
Age and growth
Petersen method Progression of modes in the length-
frequency data could be traced for both sexes of C. arel
and C. lida (Figs. 7,8).
In male C. arel (Fig. 7), the first mode was the
155-169 mm length-group in January. A progressive
shift during subsequent months until October, to the
290-304 mm length-group, indicated a growth of 135
mm in 9 months. Assuming the same rate of growth.
a fish would attain a length of 180 mm in the first year.
Beyond November, it was not possible to trace
length-groups.
In female C. arel (Fig. 7), the first mode was the
125-139mm length-group in September. A progressive
shift during subsequent months until March, to the
230-240 mm length-group, indicated a growrth of 105
mm in 6 months. Groups could not be traced beyond
April. At the same rate of growth, a fish would have
346
Fishery Bulletin 90(2). 1992
Figure 7B
Size-frequency histograms for
male (M) and female (F) Cyrwglos-
sus arel (July-December) caught
commercially off Porto Novo,
India, October 1981-September
1982.
>-
o
z
UJ
o
LXJ
>
o
z
UJ
o
LU
>-
o
z
LLI
o
CYNOGLOSSUS AREL
SEPTEMBER
30
25
20
N; 57 (M)
N: 35 (F)
>-
CJ
z
ai
o
UJ
25
20
15
^/lOl/^Olr)omOl/lOu^ou^Olnol/)Ou^ou^Ol/l
cTi,-c\j^mr-^a3o»~n^tDr^aioc\jcoincococ3i»-c\j
T-,-,-»-T-T-c\jcvjojrvjc\jc\jcsjcncnr)cncnnco^'j
SIZE GROUPS (mm) ■
AUGUST
B
35
30
25
20
15 -
10
5
lit!
N: 45 (M)
N: 54 (F)
.a. .6,8
MALE
FEMALE
25
O
z
m
o
UJ
(T
u.
inOLnoir)Oir)ouioir)oaioir)Oir)Oir)omoi/i
Oli-CNJ-^lDf^CDOT-CT^tDr^CJIOCNJCOLniDCDOlT-CNJ
SIZE GROUPS (mm)
JULY
DECEMBER
N; 29 (M)
N: 37 (F)
Li
inoinomoinoLnomoinoLnoirjOLnoLOOLn
cD'-CM'^inr-.coO'-cT^iDr-cnocvjcniotococn^cNj
■.-T-T-T-.-T-CMCNJOJCMCMCNJOJCOcnnnnCOCT^'^
SIZE GROUPS (mm)
NOVEMBER
20 -
15 -
N; 87 (M)
N. 54 (F)
i/iomoi/ioinoinoinoLnoinoLnomoirjOLO
ai'-oj'^ir)r~~-coo--cO"^tDr~^cnorMmr)tDcDai'-CNj
•—»—■•— ■^■•-■•— CMCMOJOJOJCMCvjcommcnrocnn-^"^
SIZE GROUPS (mm)
OCTOBER
30
25
20
15
N: 69 (M)
N: 56 (F)
noinoui
nr^cooT-
-■^■r-CNJC\J
ouioaioinoinoiDomoiD
co-^tDr^aiocNjroinu^oocn^CM
SIZE GROUPS (mm)
lr)Omou^omoino^no^i^Olr)OLnoLnolnoln
C7)^CM'4'ir)h-ajO'-cOTj-tDr^CT)OC\jcoir)<JDcoCT)T-c\i
'-•-'-'-■'-^CMCMCMCMCNJCMOjrxrjnncniDCT^'^
SIZE GROUPS (mm)
attained a length of 210 mm in the first year. There was
another mode at the 200-214 mm length-group during
February. A progressive shift of this mode during
subsequent months until November, to the 260-274 mm
length-group, indicated a growth of 60 mm in 9 months.
Groups could not be traced beyond December. Based
on this rate of growth, a fish would reach 290 mm at
the end of the second year. Because of poorer represen-
tation in older size-groups, later modes were not traced.
In male C. lida (Fig. 8), the first mode was the
102-108 mm size-group in November. This was traced
to the 179-185 mm size-group in May, 77 mm of growth
in 6 months. Length-groups could not be traced beyond
June. At this rate, a fish would be 154 mm at the end
of the first year. The mode at the 151-157mm size-
group in March was traced to the 179-185 mm size-
group in November, 28 mm growth in 8 months. At
this rate of growth, a fish at the end of the second
Rajaguru: Biology of Cynoglossus arel and C lida from Indian waters
347
>-
o
z
LU
o
LU
DC
o
CYNOGLOSSUS LIDA
MARCH
FEBRUARY
30
?S
>-
o
z
?n
LU
3
o
t;
LU
(E
10
J
N: 58 (M)
N: 16 (F)
JANUARY
20
15
LU
O 10 h
LU
IE
Li.
JUNE
30
\ J N: 56 (M)
§ 25
LU
3 20
O
10
ll
N: 59 (F)
0?
5
0
,M\
illlin
mCNjci(£>cooN.-*jT-comrua>tocnor^'<a'^cDuicM
^ -.- T- T- ^ ^ T- T- T- ^ ^ ^ r- T~ c\j eg eg c\j c\j c\j cvj
SIZE GROUPS (mm)
0
MALE
FEMALE
>
(J
z
LU
o
LU
(C
i/)C\iaicDcooh^^T-comega)cocoor-^»?^cOLDeg
OjoO'-egcnrO'^irjmtDr-^r-cDooo^rvjejr)-^
'-T-^»-^r-T-T-T-T-»-^T-^egr\jegcgc\jegeg
SIZE GROUPS (mm)
MAY
tnojOicDnor-'^T-coLncMCDcDcoor^^jT-ooinoj
o>oO'-CMcnoT^mmtDr~^h~coCT)oo»-c\]OjrO"^
'-^'-^'-■■~»-'-'-»---'^'-'-C\JC>JCVJC\JCMCMCO
SIZE GROUPS (mm)
lj^CMC71CDCOOr--^'-OOir)C\JOHO(T>Or--'^»-aOl/lC\J
a)OO^cNjno'^ir)LOUJr~-r-.cooioo»-c\j(Mco'^
'~T-T-^T-T-..-T-T-T-T-T-T-^C\JC\JC\JC\JCMC\J(M
SIZE GROUPS (mm)
APRIL
Jj
N: 34 (M)
N: 45 (F)
15
O
z
LU
o
LU
cc
N: 84 (M)
N; 51 (F)
JUL
inCMOitDcoot^ ■^'-coi/ic\jcncor)of^'<t'-coLnc\j
»-'t-^»-^T-..-T-..-r-^^»-T-CMC\JCMC\JC\JC\JC\J
SIZE GROUPS (mm)
i/iegcnocnor^^T-oomc\jai(£)cnor^*^»-a)i/)eg
oioOT-egocT^mi/)(£)r^r^ooo>oo'-egegc*?^
'-'-'-'-'-'-'-'-'-'-'-^'-^cvjegegegcvjcjeg
SIZE GROUPS (mm)
Figure 8A
Size-frequency histograms for
male (M) and female (F) Cynoglos-
sus lida (January-June) caught
commercially off Porto Novo,
India, October 1981-September
1982.
year would be 196 mm. Further modes could not be
traced.
In female C. lida (Fig. 8), the first mode was the
102-108 mm size-group in November. This was traced
to the 193-199mm size-group in June, 91mm growth
in 7 months. At this rate, a fish would be 156 mm in
the first year. The mode at the 151-157 mm size-group
in January was traced back to the 193- 199 mm size-
group in January, 42 mm growth in 12 months. A fish
would be 198 mm at the end of the second year. The
mode at the 193- 199 mm size-group in May was traced
to the 207-213 mm size-group in December, 14 mm
growth in 7 months. At this rate, a fish at the end of
the third year would be 222 mm.
The rate of growth from the time of hatching and
throughout the first year would be more rapid than that
of the older year-classes.
348
Fishery Bulletin 90(2). 1992
Figure 8B
Size-frequency histograms for male
(M) and female (F) Cynoglossus lida
(July-December) caught commer-
cially off Porto Novo, India, October
1981-September 1982.
CYNOGLOSSUS LIDA
SEPTEMBER
DECEMBER
1 N: 50 (M)
N: 28 (M)
>•
o
20
-
N; 26 (F)
>
o
25
~
N: 31 (F)
z
z
20
^
UJ
UJ
3
lb
"
D
3
o
1
O
IS
_
UJ
UJ
cc
in
-
cr
J
u.
1
u.
10
"
1
0?
5
_
I
s?
s
1 111
h
0
II i
h
0
jiiJuijJi
1 fl
mcgo>tDcoor^-<r^a3inc\jcjitDcoor--^^coir)C\j
ai o o t- CO CO CO ^ in Lf) (£) r-- r^ m oi o o »- eg c\j CO ^
SIZE GROUPS (mm)
SIZE GROUPS (mm)
AUGUST
NOVEMBER
20
N: 43 (M)
. N; 160 (M)
>-
o
z
Ul
O
Hi
u.
15
10
1
1 N 62 (F)
>-
o
z
UJ
O
Ul
K
U.
15
10
-
1
N: 207 (F)
s?
5
' inh 1
1
1 1
J«
5
0
',..,11. jj,
II...
u
ir)C\ja>ujcoor-'^'-coL
nc\jcrmDcoor--"^T-oomcM
D r-- r-- 00 oi o o T- cvj c\j CT<t
tf)CMOm3not^-^'^a3inc\ion£>coor^'^'^cDinc\j
oi o o *- c\i en cT^ u> inio r- r-- CD cr> o o ^ CM CO cT*
SIZE GROUPS (mm)
SIZE GROUPS (mm)
JULY
20
OCTOBER
Jb
N; 25 (M)
^ J N: 83 (M)
>
o
30
25
N: 26 (F)
>-
U
z
15
-
1 N 77 (F)
Ul
o
UJ
cc
u.
20
15
1
Ul
O
UJ
c
U.
10
#
s?
10
:
a«
5
1 1
yj
b
1
1
■ MAL
i.Mli. .
ni
0
mCMC3HOCOOh~-^T-00
cjioOT-cvjcoco-^inin
£>h-.h^«00>OO»-CSJC\JCT^
E
U-) c\j CT) u3 n o r-- ■* ^ CD ID csj o> to n o r-- -^ -- CO IT) O.J
fiiOo»-(NncT*»m/nDh^r^aoa)Oo--c\jc\jcT^
SIZE GROUPS (mm)
0 FEMALE SIZE GROUPS (mm)
Probability plot method Cumulative percentage
distribution of lengths was calculated for C. arel and
C. lida, and plotted against the midpoints of length-
groups on probability paper (Fig. 91, J). These points
formed approximately straight lines, but slight devia-
tions could be recognized. Based on the probability
plots, male C. arel attained 194 mm, 272 mm, and
333 mm in the 1st, 2d, and 3d years, respectively, while
females reached 201, 312, and 393 mm for these years.
In C. lida, males attained 151, 188, and 218mm, while
females reached 153, 188, and 216 mm, in the 1st, 2d,
and 3d years, respectively.
von Bertalanffy's equation Plots of Lfi- 1 against
Lt , showing a straight-line relationship for C arel and
C. lida, were drawn. A least-square line was then fitted
and an estimate of L^ was obtained (Fig. 9A,C, E,G).
By this Ford-Walford graph, L^ was 570 mm for male
Rajaguru: Biology of Cynoglossus arel and C lida from Indian waters
349
B
650
j/^
^
j/\
400
r
r
y^
550
/> 615 La
/
550
■
/\
r T
/>
360
/
-
yf^TQ Lo
320
« 450
y'
/
/
450
//
E
A ^
,^
320
/
1
E
■
280
1
£ 350
E
E
280
: /
350
■
^ /
E
240
: /
*- 250
/
^
. /
T
2 50
.
^ /
- /
/
240
. /
—I
.
/
_j
2 00
- /
150
/
^
200
:/
150
-
/
/
/
/
160
/
50
SO
-X
^
/" , , ,
100 300
500
-10
1 2 3
100 300
500
-to
1 2 3
Lt (mm)
Ag
e (years)
Lt (mm)
Age (years)
H
350
300
:
y^3A0 Lol
220
200
- /
^ /
350
300
^
y 220
yfiZi La 200
; /
2 50
:
//
180
- /
2 50
-
//' '80
■ /
1 200
■
//''
E 160
7
?
E
200
/
yi''' E 160
/ 1
\\
7 150
-
/ •
- 140
7
1]
150
/ /
- 140
_l
:/
100
- ^
120
f
100
- y
^ /
120
j
50
■
'-'
loa
-to
50
^
10Q
100 200 300
1 2 3
100
200 300 *'° 12 3
Lt (mm)
Ag
e (years)
Lt (mm) Age (years)
0.05 0.5 2 10 30 50 70 90 98 99
CUMULATIVE FREQUENCY
0.05 0.5 2 10 30 50 70 90 98 99 100
CUMULATIVE FREQUENCY
Figure 9
Age and growth oi Cynoglossus arel and C. lida caught commercially off Porto Novo, India, October 1981-September
1982. Ford-Walford plot for female (A) and male (C) C. arel. and for female (E) and male (G) C. lida; theoretical
growth curve for female (B) and male (D) C. arel, and female (F) and male (H) C. lida; probability plot for male and
female C. arel (I) and C. lida (J).
350
Fishery Bulletin 90(2). 1992
and 615mm for female C. arel;
335 mm for male and 340 mm for
female C. lida.
Based on values obtained, the
von Bertalanffy's equations are:
C. arel
Male Lt =
570 (1 - e " 0-2376 (t + 0.7753) \
Female Lt =
615(1 - e -0.3151(t + 0.2645))
C. lida
Male Lt =
335(1 -e -0-2326(1+1.6348))
Female Lt =
340(1 -e -0-2231 (t+ 1.8029))
Male C. arel reached 194, 272,
and 333 mm in the 1st, 2d, and 3d
years, respectively (Fig. 9D),
while females attained a length
of 201, 312, and 393mm for the
1st, 2d, and 3d years, respective-
ly (Fig. 9B). Male C. lida reached
151, 188, and 218mm (Fig. 9H),
while females attained lengths of
153, 188, and 216 mm in the 1st,
2d, and 3d years, respectively
(Fig. 9F).
Estimates of age and growth,
based on the three different meth-
ods, are presented in Table 15.
Length-weight
relationships
The linear relationships in
logarithmic values of length and
weight for males, females, and
juveniles are shown in Figure
lOA. These were typical length-
weight relationships in which
length increase is rapid initially,
but later slows down with a cor-
responding increase in weight.
Correlation coefficient values (r)
are highly significant as follows:
Figure lOA
Length-weight relationships for male,
female, and juvenile Cynoglossus arel and C.
lida caught commercially off Porto Novo,
India, October 1981-September 1982.
Table 1 5
Mean length (mm) attained in
different years of life
(per 3 methods) by Cynoglossus arel 1
and C. lida caught commercially off Porto Novo, India, October 1981-
-September 1982.
Species
Year
Method
Petersen
Probability plot
von Bertalanffy
Cynoglossus arel Male
180
194
194
II
—
272
272
III
—
333
333
Female
210
201
201
II
290
312
312
III
—
393
393
Cynoglossus lida Male
154
151
151
II
196
188
188
III
—
218
218
Female
156
153
153
II
198
188
188
III
222
216
216
CYNOGLOSSUS AREL
MALE
CYNOGLOSSUS LIDA
MALE
Ic 2 06
g
g 1 60
o
O 114-1
N 655
197 207 217 227 2 37 2 47 2 57
LOG LENGTH
FEMALE
2 08 2 18 2 28
LOG LENGTH
FEMALE
N
723
■^
£iss-
g
UJ
g 124-
^
o
2 093-
X
^
2,19 2 39 2.59
LOG LENGTH
JUVENILE
2 09 2 19 2 29
LOG LENGTH
JUVENILE
N 27
N
28 y^
X 0.88 -
y^
EIG
y^
g 0 70 -
y/^
o
Oo52-
0 34 -
/
1 1 , 1 1 1 .
191 1 96 201 206
LOG LENGTH
1.95 2 00 2 05
LOG LENGTH
Rajaguru: Biology of Cynoglossus arel and C lida from Indian waters
351
CYNOGLOSSUS LIDA
IMMATURE (F)
N 54 y^
1-
I
UJ
5
12-
1 0-
X
o
_l
08-
06-
1 1 1 1 1 1 1 1 1
X
C3
HI
o
o
IMMATURE (M&F)
1 8-
1 5-
N 159
1 2 -
' ..Ji^^^^^
09-
06-
^^
LOG LENGTH
MATURING (F)
2 07 2 17
LOG LENGTH
MATURING (M&F)
2 14 221
LOG LENGTH
MATURE
(F)
N
254
/
X
C3
UJ
5
1,7-
15-
J#
-/
o
-1
1 3-
1 1 -
/
I 1 7
g
UJ
5 15
o
O ,3
N 617
221 231
LOG LENGTH
2 21 2 31
LOG LENGTH
C. arel Male 0.9870 (P<0.001)
Female 0.9905 (P<0.001)
Juvenile 0.8747 (P<0.001)
C. lida Male 0.9782 (P<0.001)
Female 0.9756 (P< 0.001)
Juvenile 0.9409 (P< 0.001).
Calculated b, a, r values, and observed F values are
presented in Tables 16 and 17.
Linear equations were computed separately for
males and females of each month to examine variations
in growth patterns. In C. arel, significant differences
were observed in regression coefficients during Janu-
ary, February, May, and July, with no significant
Figure I OB
Length-weight relationships for immature,
maturing, and mature females (F), and
pooled sexes (M&F) of Cyrwglossas lida
caught commercially off Porto Novo, India,
October 1981-September 1982.
differences noted for the remain-
ing 8 months (Table 17). In C.
lida, no significant differences
were observed in regression coef-
ficients of length and weight be-
tween males and females.
Analysis of covariance was
employed to determine whether
growth patterns differed signifi-
cantly between stages of matur-
ity (immature, maturing, and ma-
ture) in males and females. No
significant differences were noted
for C. arel (P>0.05). Hence im-
mature, maturing, and mature
male and female C. arel were
combined irrespective of sexes.
As there was no significant dif-
ference in the regression of Y
and X between maturity stages
irrespective of sexes, the data for
male, female, and juvenile C. arel
were pooled for the entire year,
irrespective of months and
maturity stages, and the linear
equation was fitted for males,
females, and juveniles. Analysis
of covariance was again em-
ployed for the pooled data to test
whether growth patterns dif-
fered significantly between sexes
of C. arel. Significant differences
were obtained in the b value between male, female, and
juvenile C. arel. On comparing males and females,
males and juveniles, and females and juveniles, signifi-
cant values were obtained. Since the growth rates of
males, females, and juveniles differed significantly
from one another, three separate equations, relating
logW to logL, are presented for C. arel as follows,
C. arel Male logW = -5.9551 -i- 3.2665 logL
Female logW = -5.8231 + 3.2100 logL
Juvenile logW = -4.8615 + 2.7901 logL,
and the parabolic equations are
Male W = 0.0000011 L^-^ees
Female W = 0.0000015 L^^ioo
Juvenile W = 0.0000138 L^-^^OK
214 221
LOG LENGTH
MATURE (M&F)
352
Fishery Bulletin 90(2). 1992
Table 16
Results of linear regressions
af length-weight relations in Cynoglossus arel and C. lida \
caught off Porto Novo, India
October 1981-September
1982. r*
= all r values were |
significant at 0.1% level.
Sample
Cynog
ossiis arel
Cynog
ossus lida
N
r*
a
b
N
r*
a
b
By months
Male
January
52
0.9920
-5.6785
3.1411
24
0.9900
-6.4285
3.5199
February
34
0.9902
-4.9278
2.8233
48
0.9872
-5.8703
3.2715
March
51
0.9785
-5.7827
3.1860
85
0.9822
-6.4397
3.5300
April
46
0.9834
-5.5809
3.0970
86
0.9878
-5.4470
3.0715
May
57
0.9813
-6.3448
3.4326
62
0.9823
-5.6679
3.1747
June
44
0.9887
-6.3781
3.4432
54
0.9705
-5.6760
3.1793
July
70
0.9868
-5.8109
3.1991
37
0.9677
-6.0573
3.3573
August
54
0.9951
-6.0999
3.3202
45
0.9860
- 5.6693
3.1758
September
59
0.9913
-5.9119
3.2575
62
0.9809
-5.2038
2.9803
October
81
0.9907
-6.1854
3.3770
83
0.9678
-5.5118
3.1283
November
70
0.9922
-6.1151
3.3468
135
0.9823
-5.4126
3.0827
December
37
0.9930
-6.0433
3.3060
47
0.9879
-5.7740
3.2354
Total
655
0.9870
-5.9551
3.2665
768
0.9782
-5.7717
3.2315
Female
January
49
0.9959
-5.9501
3.2687
31
0.9899
-6.1744
3.4089
February
52
0.9913
-5.6840
3.1432
45
0.9935
-5.8431
3.2598
March
47
0.9926
-5.7543
3.1782
38
0.9866
-6.1015
3.3718
April
58
0.9825
-5.3251
2.9920
54
0.9893
-5.6817
3.1849
May
52
0.9812
-5.6617
3.1412
43
0.9703
-5.7400
3.2157
June
41
0.9831
-6.0536
3.3047
54
0.9624
-5.1232
2.9342
July
55
0.9807
-5.0773
2.8787
44
0.9601
-5.9178
3.2983
August
66
0.9969
-6.2004
3.3643
64
0.9886
-5.6065
3.1489
September
39
0.9955
-5.8507
3,2361
37
0.9533
-5.2175
2.9923
October
57
0.9928
-5.9232
3.2654
80
0.9739
-5.7639
3.2454
November
45
0.9870
-5.8994
3.2511
168
0.9680
-5.7278
3.2335
December
38
0.9921
-6.0110
3.2960
65
0.9898
-5.9707
3.3277
Total
599
0.9905
-5.8231
3.2100
723
0.9756
-5.9084
3.2987
Juvenile
All months
27
0.8747
-4.8615
2.7901
28
0.9409
-6.5983
3.6579
Male
By maturity stages
Immature
56
0.9524
-4.7589
2.7205
105
0.8264
-5.5077
3.1101
Maturing
221
0.8111
-5.4998
3.0714
259
0.9260
-5.8584
3.2713
Mature
359
0.8909
-5.4090
3.0404
363
0.9328
-5.5189
3.1182
Female
Immature
47
0.9266
-4.5468
2.6221
54
0.8867
-5.0851
2.9124
Maturing
224
0.9463
-4.8001
2.7575
342
0.9392
-6.2651
3.4629
Mature
292
0.9104
-4.9627
2.8516
254
0.8054
-5.5588
3.1443
Male and female
Immature
103
0.9391
-4.6496
2.6699
159
0.8542
-5.3162
3.0209
Maturing
445
0.8725
-5.0463
2.8687
601
0.9399
-6.2280
3.4439
Mature
651
0.9089
-5.1071
2.9122
617
0.8726
-5.6599
3.1843
ences. Hence logarithmic equa-
tions for immature, maturing,
and mature females, as well as
pooled sexes of C. lida, are pre-
sented as follows,
C. lida
Female alone
Immature logW
-5.0851 -(-2.
Maturing logW
- 6.2651 -f 3.
Mature logW
- 5.5588 -H 3.
Pooled sexes
Immature logW
-5.3162 -f-3.
Maturing logW
- 6.2280 -H 3.
Mature logW
- 5.6599 -H 3.
9124 logL
4629 logL
1443 logL
0209 logL
4439 logL
1843 logL.
In C. lida, the tests made to check the relationship
between length and weight during various stages of
maturity (immature, maturing, and mature) in males
and females showed significant differences between the
three maturity stages in females alone and in pooled
sexes, whereas males showed no significant differ-
for juveniles, are
C. lida
Male & female
Juveniles
The linear relationships in log-
arithmic values of length and
weight for immature, maturing,
and mature female, as well as
pooled sexes of C. lida, are
shown in Figure lOB.
Although female maturity ex-
hibited a significant effect on the
length-weight relationship, all
data for male, female, and juve-
nile C. lida were treated separ-
ately, irrespective of month and
maturity stage. Analysis of co-
variance was used to find vari-
ations in the growth patterns of
males, females, and juveniles.
Significant differences were ob-
served in regression coefficients
of males, females, and juveniles.
While comparing males and
juveniles, a significant difference
was noted; however, no signifi-
cant differences were observed
in comparing males and females,
and females and juveniles. Hence
two logarithmic equations, one
common equation for adults
(male and female) and another
presented for C. lida as follows,
logW = - 5.8643 -H 3.2761 logL
logW = - 6.5983 -H 3.6579 logL,
and the parabolic equations are
Rajaguru: Biology of Cynoglossus arel and C Ma from Indian waters
353
Table 17
Observed F values and their significance in length-weight rela-
tionships of Cynoglossits arel and C. lida caught commercial-
ly off Porto Novo, India, October
1981-September 1982. |
•P<0.05.
Samples
C. arel
C. lida
Comparison between male, female
, and juvenil
e
January Male x Female
8.0000'
1.8571
February Male x Female
12.2941*
0.0000
March Male x Female
0.0000
1.6923
April Male x Female
1.1667
1.8182
May Male x Female
5.5000*
12.0000
June Male x Female
1.1429
2.1765
July Male x Female
10.1667*
0.0000
August Male x Female
1.8333
9.0000
September Male x Female
8.0000
0.0000
October Male x Female
2.3125
1.1875
November Male x Female
1.0625
3.1333
December Male x Female
0.0000
1.1250
All months
Male X Female x Juvenile
4.8571*
3.7368*
Male X Female
4.2500*
3.3889
Male X Juvenile
6.8000*
5.5882*
Female x Juvenile
4.6522*
3.1905
Comparison between maturity stages
Male
Immature x Maturing x Mature
1.5915
1.2727
Immature x Maturing
2.1798
1.3333
Immature x Mature
3.5556
0.0000
Maturing x Mature
24.6667
2.1176
Female
Immature x Maturing x Mature
1.0303
4.3913*
Immature x Maturing
1.0625
11.6296*
Immature x Mature
1.2826
1.4038
Maturing x Mature
1.6098
4.6000
Male and female combined
Immature x Maturing x Mature
1.7971
7.2162*
Immature x Maturing
2.2131
13.6765*
Immature x Mature
3.4789
1.4000
Maturing x Mature
5.0714
8.6452*
Male and female W
Juvenile W
0.0000014 L3-2761
0.0000003 L3-6579
The t -test was employed, and the calculated b value
was foimd to differ significantly from the hypothetical
B value ( = 3), at 5% level, in male and female C. arel
and in adult and juvenile C. lida, whereas juvenile
C. arel showed no significant difference:
C. arel
C. lida
Male t = 12.8125
Female ( = 11.5385
Juvenile t = - 0.6788
Male & female t
Juvenile t
14.8441
2.5470
Hence it is clear that the cubic formula is not a proper
representation of length-weight relationship in male
and female C. arel and in adult and juvenile C. lida.
Reproductive biology
Seasonal occurrence of maturity stages Female
C. arel vdth Stage-I ovaries occurred throughout the
year, with a peak in September (Fig. 11). Stage-II
ovaries were also present during all months, with
higher percentages in April and October. Individuals
with Stage-Ill ovaries occurred throughout the year,
with higher proportions during March-May and July.
Specimens with Stage-IV (mature ovaries) were pres-
ent throughout the year, with a peak in November.
Stage-V (ripe ovaries) were also noted during all
months of the year, but maximum abundance was
observed in November and December. Specimens with
Stage-VI (oozing ovaries) were collected in all months
except April and May. High incidence of oozing ovaries
was observed in January and February. This indicates
that the spawning occurs for up to 10 months (June-
March). Occurrence of Stage-VI specimens, with a peak
in January, indicates that the maximum number of in-
dividuals may spawTi during January, which is the post-
(northeast) monsoon period in Porto Novo.
In male C. arel (Fig. 12), immature (Stage-I), matur-
ing (Stage-II), and mature (Stage-Ill) individuals
occurred throughout the year. High percentages of
individuals with Stage-I testes occurred in March and
October-December. Maturing specimens (Stage II)
were abundant from February to September. Occur-
rence of mature males (Stage III) showed a peak in
January. Occurrence of a higher percentage of fully-
mature specimens in January indicated that even
though the spawning probably occurred year-round, the
majority of individuals might spawn during the post-
(northeast) monsoon period (January) in Porto Novo.
Female C. lida (Fig. 11) with Stage-I (immature)
ovaries occurred for 9 months (absent in February,
March, and July), with a peak in April. Stage-II (virgin
maturing) individuals were present during all months
except February and March, with a peak in December.
Stage-Ill (maturing) ovaries were present throughout
the year, with abundance in January, March, June,
November, and December. Stage-IV (mature) speci-
mens were present during all months of the year, with
higher proportions during February, March, May, and
July. Stage-V (ripe) individuals occurred throughout the
year, except in December, with maximum abundance
in February and September-October. Specimens with
Stage-VI (oozing) ovaries were noted for 10 months
(absent in January and December), with a peak in
September. This indicates that the spawning period
lasts for 10 months (February-November), while a
354
Fishery Bulletin 90(2), 1992
>
o
z
111
o
111
IT
>
O
3
o
111
cc
>
o
z
111
3
o
111
cc
FEMALE
STAGE-I
JFMAMJJASOND
MONTHS
STAGE-MI
J FMAMJ JASOND
MONTHS
STAGE-V
N: 97 (C. AREL) N: 82 (C LIDA)
J FMAMJ JASOND
MONTHS
STAGE-II
>
o
z
LU
=)
o
UJ
IT
■ C AREL
0 C LIDA
J FMAMJ JASOND
MONTHS
STAGE-IV
FMAMJ JASOND
MONTHS
STAGE-VI
>-
o
z
UJ
3
o
liJ
cc
u.
N: 25 (C. AREL) N: 35 (C, LIDA)
M „ IJl Ji JIjI
FMAMJ JASOND
MONTHS
Figure 1 1
Monthly percentage occurrence of different ova maturity stages in female Cynoglossus arel and C. lida
caught commercially off Porto Novo, India, October 1981-September 1982. (Refer to Appendix for monthly
sample size).
maximum number of individuals spawn in September,
which is the pre- (northeast) monsoon period in Porto
Novo.
In male C. lida (Fig. 12), immature (Stage-I) speci-
mens occurred throughout the year, except July and
September, with a peak in January. Specimens at Stage
II (maturing) were observed throughout the year, with
high percentages at all months, except August and
September. Stage-Ill (mature) males were available
throughout the year, with a peak in September. This
indicates that spawning occurs throughout the year,
but a maximum number of males also seemed to spawn
during September. Maximum occurrence of matiu-e
males in September corresponds with maximum occur-
rence of oozing females in the same period, and sup-
ports this view.
Rajaguru: Biology of Cynoglossus are! and C lida from Indian waters
355
>
O
z
UJ
o
LU
z
UJ
o
LU
cr
MALE
STAGE-I
J FMAMJ JASOND
MONTHS
■ C. ARKL
B C. LIDA
STAGE-II
J FMAMJ JASOND
MONTHS
STAGE-HI
JFMAMJJASOND
MONTHS
Figure 12
Monthly percentage occurrence of different testis
maturity stages in male Cynoglossus arel and C. lida
caught commercially off Porto Novo, India, October
1981-September 1982. (Refer to Appendix for
monthly sample size).
during all stages of maturity, only ova >0.11mm were
taken into consideration, from Stage II onwards. Pro-
gressive maturation to spawning condition was evident
from increasing ova diameters of the most advanced
mode at each stage.
For C. arel (Fig. 13) in Stage I, maximum number
of ova measured 0.01-0.04 mm; however, a few rela-
tively larger ova (0.09-0.11 mm) were also recorded.
In Stage II, a mode was discernible with a stock of ova
(0.16-0. 19mm) separated from immature stock. In
Stage III, the previous mode (at 0.16-0. 19 mm) shifted
to 0.24-0.26 mm. In Stage IV, a mode made by opaque
ova was observed at 0.36-0.38 mm. In Stage V, two
modes were found, one with a peak at 0.43-0.45 mm,
and another at 0.50-0.53 mm. In Stage VI, the pre-
ceding two modes formed jointly a single mode, with
fully mature, transparent, and large-sized ova of
0.54-0.56 mm.
For C. lida (Fig. 13) in Stage I, immature ova mea-
sured 0.01-0.04 mm. In addition, a few relatively larger
ova (0.09-0.11 mm) were also seen. In Stage II, a mode
was discernible with a stock of ova at 0.09-0.11 mm,
which was separated clearly from immature ova. In
Stage III, the previous mode was shifted to 0.20-0.23
mm. In Stage IV, a mode of opaque ova was located
at 0.31-0.34 mm. In Stage V, a mode of ripe ova was
noted at 0.35-0.38 mm. In Stage VI, fully mature,
transparent, and large-sized ova formed a mode at
0.39-0.41 mm.
Results indicate that individuals of C arel and C. lida
spawn only once during each season. Further, mature
modes were wide-based (0.49-0.83 mm in C. arel, and
0.34-0.64 mm in C. lida); therefore, the spawning
period of these species must be extended.
Gonadosomatic Index In male C. arel (Fig. 14), the
highest GSI peak was in March, and the lowest GSI
value was observed in January. In female C. arel (Fig.
14), the highest GSI peak was observed in November,
and the lowest value was in January. In both sexes of
C. arel, the peak values of GSI did not correspond with
the observed spawning period in that year.
In male C. lida (Fig. 14), the highest GSI peak
occurred in September, and the lowest value was in
January, November, and December. In female C. lida
(Fig. 14), the highest GSI peak was observed in May,
while the lowest value was in December. Only in male
C. lida did the high GSI peak coincide with the observed
peak spawning period in September.
Ova diameter Ova diameter frequencies, from ova-
ries of Stage I-VI, are shown in Figure 13 for C. arel
and C. lida. Since immature, transparent, and micro-
scopic ova (< 0.11 mm) outnumbered the maturing ova
Relative condition factor (Kn) In male C. arel (Fig.
14), the highest Kn value peak was observed in Feb-
ruary, and the lowest value was in October. In female
C. arel (Fig. 14), the highest Kn peak was observed in
January, and the lowest value was in November. Only
356
Fishery Bulletin 90(2). 1992
STAGE-I STAGE-II
80 - ,
O 40-1
' A
% FREQUENCY
8 8 8 f
OVA DIAMETER in MD OVA DIAMETER in MD
Q — C AREL
--•- C LIDA
STAGE-MI STAGE-IV
% FREQUENCY
3 8 8 4
% FREQUENCY
3 8 8 4
f
u -^
OVA DIAMETER in MD
STAGE-V
5
OVA DIAMETER in MD
STAGE-VI
S
>
UJ
O 20-
UJ
DC
U.
10 -
f
>
UJ
3
O 20-
UJ
DC
U.
10 -
r \
1 ♦
0 i
Percents
commer
(Numbe
OVA DIAMETER In MD
Fig
ige frequency of ova diameter in various
cially off Porto Novo, India, October 1
■ of fish examined for each stage: 18 (
s
ure 13
maturity stages i
981-September
:. arel. 28 C. lid
-^'~92S2Rii(!S?;55fiS3!SSS(flSE
OVA DIAMETER in MD
n Cynoglossus arel and C. lida caught
1982. MD = micrometer division.
a).
S
in female C. arel did a rise in Kn value correspond with
a rise in gonadal activity, indicating the spawning
period (in January).
In male C. lida (Fig. 14), the highest Kn peak was
seen in January and the lowest value was in February.
In females of C. lida (Fig. 14), the highest Kn peaks
were seen in January, July, and December, and the
lowest value in May. In both sexes of C. lida, a rise
in Kn value did not indicate the spawning period.
SIze-at-flrst-maturlty Both sexes of C. arel began to
mature after the 140-154mm size-group (Fig. 15).
From the 155- 169 mm size-group onwards, percentage
occurrence of mature males and females increased
steadily. Maturity reached 100% in the 245-259mm
size-group in males, and in the 275-289 mm size-group
Rajaguru: Biology of Cynoglossus arel and C. lida from Indian waters
357
MALE: GSI
FEMALE: GSI
Ul
3 0.15 -
_l
<
>
« 0 10-
N: 627 {CARED /,
N: 718 (C. LIDA) ft / i
5 96 -
UJ
3 4.96 -
-1
<
> 3.96 -
«296-
1.96 -
N 569 (C. AREL)
J^ N: 640 (C. LIDA)
I >
( I
, 1 > / \
rf/ V-«
p 1 1 1 1 1 1 1 1 1 1 1
JFMAMJJASOND
MONTHS
MALE: Kn
J FMAMJ JASOND
MONTHS
FEMALE: Kn
1.0010 -
UJ
■=>
^ 1.0005 -
c
1.0000 ■
t N; 655 (C. AREL)
M N: 768 (C. LIDA)
1 0000 -
UJ
D
-1
^ 0.9997 -
C
0.9994 -
n N: 599 (C. AREL)
\ N: 723 (C. LIDAl
I ' V ' \l *< JO /
' / ' ' \ / \ /
» \ 1 \ / \ /
0 9995 -
H 1 1 1 1 1 1 r
JFMAMJJA
MONTHS
—I — 9 — 1 1-
S 0 N D
SE
J F M
X RATIO
AMJ JASOND
MONTHS
1.75 -
UJ
< 1.50:
1 .»■
< 1.00 :
UJ 0.75 -
U.
0.50-
H N: F 569/M 627 (C. AREL)
A N: F 640/M 718 (C. LIDA)
--•- C LIDA
Relative <
arel and (
Appendix
;ondition factor (Kn), gons
7. lida caught commercial
for montiily sample size'
JFMAMJJASOND
MONTHS
Figure 14
idosomatic index (GSI), and sex ratio
y off Porto Novo, India, October 198
(female/male) in Cynoglossus
1-September 1982. (Refer to
in females. The calculated Lm for C. arel was 217 mm
for males and 225 mm for females.
In C. lida, no specimen of either sex was mature until
the 137- 143 mm size-group, and the percentage occur-
rence of mature specimens increased gradually from
then on (Fig. 15). All male fish were mature at the
186-192mm size-group, and females at the 193-199
mm size-group. In C. lida, Lm ( = L5o) was calculated
as 167mm for males and 179mm for females. In both
species, males mature at a smaller size than females.
358
Fishery Bulletin 90(2), 1992
C. AREL: MALE
C. LIDA: MALE
UJ
Z)
I-
<
fK3-0-Q-0-D-0-D-e-9
N: 718
'I ' I' I 'M I' I 'I ' I' I "I ' I' I M ' I'
OJCNJCMOjmmtOn
SIZE GROUPS (TL in mm)
C. AREL: FEMALE
SIZE GROUPS (TL in mm)
C. LIDA: FEMALE
aD -
UJ
tL
m -
Z>
1-
<
s
«) -
s?
33 -
N: 569
I |i |i |i I i| i| i| i| ■!• |i|i I ■! i|i| I !■ |i |i
SIZE GROUPS (TL in mm)
loo ■- OJ n 1
1' I ' I' 1 ' I 'I ' Ml ' I •
i-^inintor-r^a>a>oo— cj<>J(^ ^
■ --'- ■-■- '-'- ----OJOJOJ(N<NJC\IOJ
SIZE GROUPS (TL in mm)
Figure 15
Size-at-first-maturity in male and female Cynoglossus arel and C. lida caught commercially off
Porto Novo, India, October 1981-September 1982.
Age-at-flrst-maturlty In their first year, males of
C. arel grow to 180 mm according to the Petersen
method, and to 194 mm according to the probability plot
method and von Bertalanffy's equation (Rajaguru
1987). Female C. arel reached 201 mm as per the prob-
ability plot method and von Bertalanffy's equation, and
210 mm according to the Petersen method, at the end
of their first year of life (Rajaguru 1987). Hence it
appears that 50% of male and female C. arel attain first
maturity at the beginning of their second year of life.
Male C. lida, at the end of their first year of life,
would grow to 151-154 mm based on all three methods
(Rajaguru 1987). Females of C. lida reach 153 mm
according to von Bertalanffy's equation and the prob-
ability plot method, and 156 mm according to the
Petersen method, at the end of the first year of their
life (Rajaguru 1987). Therefore, 50% of male and
female C. lida attain first sexual maturity during the
second year of their life.
Fecundity Fecundity varied from 14,972 to 127,001
eggs/ovary in C. arel, and 11,267 to 81,004 eggs/ovary
Rajaguru: Biology of Cynoglossus arel and C. lida from Indian waters
359
CYNOGLOSSUS AREL
TL X F
o
o
5.1 -
♦
N: 26
4.9-
♦
^<^
4.7-
♦ * *j-<^
■"""^
4.5-
43-
*
* ♦
*
•
2 5
LOG TL
TW X F
26
o
5.1 -
*
N: 26
4.9-
♦
•
...-<
4 7-
.
* ,* • f,--— ^
4.5-
4.3-
*
♦
♦
'
1
1.9 2 1 2.3
LOG TW
OL X F
OW X F
o
2.5
2.7
CYNOGLOSSUS LIDA
TL X F
2.20 2.22 2.24 2.26 2.28 2.30 2.32
LOG TL
TW X F
o
o
OL X F
o
o
OW X F
C3
O
29 3.1
LOG OW
Figure 16
Relationships between total length-fecundity (TL x F), total weight-fecundity (TW x F), ovary length-
fecundity (OL X F), and ovary weight-fecundity (OW x F), in Cynoglossus arel and C. lida caught
commercially off Porto Novo, India, October 1981-September 1982.
in C. lida. Number of ova/g body weight was 124-1096
(x 464) in C. arel, and 287-1664 {x 988) in C. lida.
Scatter diagrams of fecundity (F) against TL, TW, OL,
and OW are shown in Figure 16.
Fecundity was found to increase with TL (Fig. 16).
The calculated equation for F against TL is,
a arel logF = 1.0629 -t- 1.4459 logTL
C. lida logF = - 0.4923 -h 2.2006 logTL.
360
Fishery Bulletin 90(2). 1992
The correlation coefficient (r) for this relationship in
C. arel is 0.6043 (P< 0.001) and is statistically signif-
icant; in C. lida, it is 0.2579 (P>0.05) and is not
statistically significant.
Fecundity against TW showed a linear relationship
(Fig. 16), and equations for the transformed data are,
a arel logF = 3.5907 + 0.4995 logTW
C. lida logF = 3.5536 + 0.6108 logTW.
The correlation coefficient (r) for this relationship in
C. arel is 0.6345 and is highly significant (P<0.001);
in C. lida, the correlation coefficient of 0.2302 (P>0.05)
did not indicate a significant correlation between these
two variables.
Ovary length showed a straight-line relationship with
fecundity (Fig. 16). In logarithmic form, the relation-
ships between F and OL can be expressed as follows;
a arel logF = 1.8472 + 1.3693 logOL
C. lida logF = 0.3206 + 2.2630 logOL.
The correlation coefficient in C. arel is 0.6632 (P<
0.001) and is statistically significant; in C. lida, it is
0.4990 {P< 0.05), showing a high degree of correlation.
Fecundity plotted against OW showed a linear rela-
tionship (Fig. 16) and equations for these two vari-
ables are,
C. arel logF = 2.1858 + 0.7050 logOW
C. lida logF = 1.3909 + 1.0038 logOW.
The correlation coefficient in C. arel is 0.7729 (P<
0.001), indicating a high degree of correlation between
these two variables. In C. lida, the correlation coeffi-
cient of 0.8606 (P< 0.001) is highly significant.
In the present study, the exponential value (b) for
total length-fecundity was higher than for total weight-
fecundity. Similarly, the b value for ovary length-
fecimdity was higher than for ovary weight-fecundity.
Sex ratio The sex ratio was about 1:1 for both spe-
cies (Table 18, Fig. 14). However, the ratio varied in
monthly samples, and chi-square values showed a
significant deviation from the expected 1:1 ratio for
3 months (February, September, and November) in
C. arel, and during February-April, September, and
November in C lida. Since data were pooled for one
year, the chi-square value conformed to the expected
1:1 ratio in C arel; whereas in C. lida, it deviated
significantly from the expected 1:1 ratio. The devia-
tion may be due to multiple testing.
Discussion
Feeding ecology
Cynoglossus arel and C. lida feed predominantly on
polychaetes and crustaceans, followed by other phyla
such as molluscs, echinoderms, and coelenterates.
These similarities in diets indicate common feeding
strategies within the tonguefishes and soles (Seshappa
and Bhimachar 1955, Kuthalingam 1957, de Groot
1971, Braber and de Groot 1973ab, Stickney 1976,
Pearcy and Hancock 1978, Langton and Bowman 1981,
Wakabara et al. 1982, Langton 1983, Honda 1984).
Table 18
Sex ratio of Cynoglossus
arel and C. lida
caught commercially
off Porto Novo,
India, October 1981-September 1982. F = Probability.
Months
Cynoglossus arel
Cynoglossus lida
o-A^
9N
a %
9 %
ct;9
x'
F
aN
<}N
Of %
9 %
o-:9
^
F
Jan.
53
39
57.6
42.4
1.4:1.0
2.1304
>0.05
32
42
43.2
56.8
0.8:1.0
1.3514
>0.050
Feb.
24
45
34.8
65.2
0.5:1.0
6.3913
<0.05
59
16
78.7
21.3
3.7:1.0
24.6533
<0.001
Mar.
54
49
52.4
47.6
1.1:1.0
0.2427
>0.05
64
18
78.0
22.0
3.6:1.0
25.8049
<0.001
Apr.
46
55
45.5
54.5
0.8:1.0
0.8020
>0.05
84
51
62.2
37.8
1.7:1.0
8.0667
<0.010
May
56
45
55.4
44.6
1.2:1.0
1.1980
>0.05
38
31
55.1
44.9
1.2:1.0
0.7101
>0.050
June
35
40
46.7
53.3
0.9:1.0
0.3333
>0.05
56
59
48.7
51.3
0.9:1.0
0.0783
>0.050
July
69
56
55.2
44.8
1.2:1.0
1.3520
>0.05
25
25
50.0
50.0
1.0:1.0
0.0000
_
Aug.
44
54
44.9
55.1
0.8:1.0
1.0204
>0.05
43
62
41.0
59.0
0.7:1.0
3.4381
>0.050
Sep.
56
35
61.5
38.5
1.6:1.0
4.8462
<0.05
46
25
64.8
35.2
1.8:1.0
6.2113
<0.050
Oct.
75
54
58.1
41.9
1.4:1.0
3.4186
>0.05
83
76
52.2
47.8
1.1:1.0
0.3082
>0.050
Nov.
86
54
61.4
38.6
1.6:1.0
7.3143
<0.01
160
207
43.6
56.4
0.8:1.0
6.0191
< 0.050
Dec.
29
34
46.0
54.0
0.9:1.0
0.3968
>0.05
28
28
50.0
50.0
1.0:1.0
0.0000
-
Total
627
560
52.8
47.2
1.1:1.0
3.7818
>0.05
718
640
52.9
47.1
1.1:1.0
4.4801
< 0.050
Rajaguru; Biology of Cynoglossus arel and C lids from Indian waters
361
Diet of fishes is related not only to their feeding
behavior, but also to their digestive morphology and
mouth structure (Stickney et al. 1974). In C. arel and
C. lida, jaws are asymmetrical so that the mouth points
to the bottom when opened, aiding feeding upon ben-
thic prey. Flatfishes that feed on benthose usually have
asymmetrical jaws (Pearcy and Hancock 1978). Cyno-
glossus arel, C. lida, and other tonguefishes are, in
general, polychaete feeders. These fishes have small
stomachs (not highly demarcated) and long intestines,
and lack gillrakers and pyloric caecae.
Although there were similarities in food items, im-
portance of prey species differed between adults and
juveniles. Juveniles of C. arel and C. lida, probably
owing to their very small mouths, fed predominantly
on smaller prey such as amphipods and copepods, and
ingested fewer types (only 10) of food items. Adults of
both species, in contrast, had eaten 24 and 29 types,
respectively, of relatively large-sized prey, primarily
polychaetes, prawns, crustacean fragments, and fish
remains. Mouth size severely limits the size of prey
which can be ingested (Stickney 1976). According to
Honda (1984), the extent of food demand and ability
for food acquisition increase with growth and devel-
opment of fish. Lande's (1976) findings on the dab
Limanda limanda revealed that larger fish consumed
large-sized prey compared with smaller fish. Pearcy
and Hancock (1978) studied feeding habits of Dover
sole Microstomus pacificus, rex sole Glyptocephalits
zachirus, slender sole Lyopsetta exilis, and Pacific
sanddab Citharichthys sordidus off Oregon, and con-
cluded that the number and size of prey taxa gener-
ally increased with size in these flatfishes, due to the
ability of larger fish to consume a larger range of prey
sizes than smaller fish.
During the present investigation, fewer empty stom-
achs were noted in female than male (11.6% vs. 18.0%)
C. lida. A similar trend was observed by Langton
(1983) for yellow-tail flounder Limanda ferruginea off
the northeastern United States. In female and male
C. arel, the percentage occurrence of empty stomachs
was similar (7.7% vs. 8.0%).
Sexual differences in food and feeding habits of flat-
fishes have not been reported. In this study, there was
some indication of differences in prey between males
and females. The primary food group (polychaetes) was
the same in both sexes of C. lida; however, polychaetes
were somewhat more important in females (IRI 65.9%)
than males (IRI 53.4%). Moreover, the breadth of the
diet was somewhat less in females which fed upon only
19 food types, in contrast to males in which 24 prey
types were consumed.
The present analysis on feeding intensity reveals that
in males of C arel, the peak occurrence of empty
stomachs had a positive correlation with peak spawn-
ing activity (in January). Spavniing fish contained
the least amounts of prey, or had empty stomachs.
This result is consistent with the findings of Rama-
nathan and Natarajan (1980) on Indian halibut Pset-
todes erumei and floimder Pseudorhombtis arsius, and
with those of Langton (1983) on yellowtail flounder
Limanda ferruginea. However, in female C. arel and
both sexes of C. lida, the occurrence of empty stomachs
had no obvious relationship with spawning activities.
Seshappa and Bhimachar (1955) also reported that in
Malabar sole C. semifasciatus, feeding intensity was
not interrupted by increased reproductive activity.
Male and female C. arel showed an inverse relation-
ship between gastrosomatic/hepatosomatic indices and
breeding cycle, with the lowest values observed dur-
ing peak spawning (in January). This indicates that
gut/liver energy reserves may be used for gonadal
recrudescence. Such a correlation was observed by
Ramanathan (1977) for the Indian halibut Psettodes
erumei and flounder Pseudorhombus arsius. Wingfield
and Grimm (1977) found HI to be highest in the
prespawning season and lowest in the postspawning
period of the Irish Sea plaice Pleuronectes platessa.
However, C. lida did not show a relationship between
gastrosomatic/hepatosomatic indices and breeding
cycle.
Although the primary diet of these two demersal flat-
fishes consisted of benthic prey such as polychaetes,
prawns, echinoderms, and molluscs, it was surprising
to find that pelagic amphipods (<59.2% IRI) and
copepods (<44.7% IRI) were also relatively important
in their diets, especially in juveniles. Cynoglossus arel
and C. lida are demersal flatfishes that have never been
caught in the pelagic waters off Porto Novo, either
during day or night. These tonguefishes are not known
to undergo vertical feeding migrations. Based on the
present study, it is speculated that these tonguefishes
ingested pelagic prey such as hyperiid amphipods and
copepods when these prey organisms approached or
contacted the bottom during vertical migrations
through the water column. Hyperiid amphipods have
been reported to undertake extensive vertical migra-
tions (Bowman et al. 1982, Roe et al. 1984, Clark et
al. 1989). Isaacs and Schwartzlose (1965) and Pereyra
et al. (1969) have reported that in the eastern North
Pacific Ocean, demersal fishes feed on pelagic prey,
when such prey approach the bottom along the edge
of the continental shelf.
Polychaetes, prawns, amphipods, copepods, crusta-
ceans, and fishes were important prey for both Cyno-
glossus arel and C. lida. These tonguefishes shared
25 different food items as prey (out of 30 food types
in C. arel, and out of 26 in C. lida). High overlap in
diet may reflect abundant prey resources, reducing
competition. Lande (1976) observed such a high prey
362
Fishery Bulletin 90(2), 1992
abundance for Norwegian flatfishes. However, during
the present study, some individuals had full and gorged
stomachs, filled only with either polychaetes or prawns.
This might indicate either greater availability or patchy
distribution of the major food items. Seshappa and
Bhimachar (1955) reported for Malabar sole Cyno-
ghssus semifasciatiis, from the west coast of India, that
during certain months the guts were gorged with only
one prey, mostly polychaetes.
During the present investigation, most stomachs of
C. arel and C. lida were found to contain considerable
quantities of sediment (sand and mud). In some speci-
mens, the entire stomach was filled with sediment.
Algal filaments were also found in some stomachs.
Sediment and algal filaments were probably ingested
accidentally with bottom-living polychaetes and other
infauna. Since demersal fishes browse near the sea
bottom, some amount of sediment may frequently be
in their gut. This has been reported for other flat-
fishes, such as Malabar sole Cynoglosstis semifasciatics
(Seshappa and Bhimachar 1955) and C. lingua (Kutha-
lingam 1957), and for other demersal fishes (Sedberry
and Musick 1978). Stickney (1976) stated that the high
percentage occiu-rence of sand in the stomachs of
blackcheek tonguefish Symphurus -plagiusa might be
due to ingestion of a significant quantity of detrital
material in its feeding activities. It is unknown if sedi-
ment ingestion in C. arel and C. lida is accidental or
represents a deliberate feeding action. In situ or
aquarium studies on feeding habits would be required
to answer this question.
Nematodes, present in stomachs of several specimens
of C. arel and C. lida, were not attached to the stomach
wall but, rather, appeared to be free-living species.
age from growth to reproduction, so that the rate of
growth in males is reduced at an earlier age than in
females. Results of age and growth studies on yellow-
tail flounder Limandaferruginea from New England
(Lux and Nichy 1969), Limanda herzensteini from
Japan (Wada 1970a), Agulhas sole Aicstroglossus pec-
toralis from South Africa (Zoutendyk 1974a), and Solea
solea from Spain (Ramos 1982) are also consistent with
Pitt's view. In contrast to the above view, no signifi-
cant difference was observed between the growth
patterns in males and females of C. lida.
It is important to know at what age fishes are
recruited to the fishery. The present study reveals that
C. arel and C lida reach commercial size during their
2d and 3d year. Botha et al. (1971) stated that Agulhas
sole Austroglossiis pectoralis off South Africa reached
commercial size during their 3d-5th years, and at
certain times their 2d-4th years. Lux and Nichy (1969)
observed that yellowtail flounder Limandaferruginea
of the New England fishing grounds recruited to the
commercial fishery during their 3d and 4th years. Ac-
cording to Seshappa and Bhimachar (1955), the bulk
of commercial catches of Malabar sole Cynoglossus
semifasciatus consisted of 2d-year individuals.
Cynoglossus arel and C. lida have a life-span of a little
over 3 years in the southeast coast of India. The
longevity for C. lida from the west coast of India has
also been reported to be 3-4 years (Seshappa 1978).
Longevity in most tropical fish species is relatively
shorter and seldom exceeds 2-3 years (Qasim 1973b).
However, temperate flatfishes were reported to have
a longevity of 6-30 years (Devoid 1942, Arora 1951,
Pitt 1967, Lux and Nichy 1969, Lux 1970, Wada 1970a,
Zoutendyk 1974a, Smith and Daiber 1977).
Age and growth
In the present study, distinct annual markings were
not seen in scales, otoliths, opercular bones, and
supraoccipital crests of C. arel and C. lida. Struhsaker
and Uchiyama (1976) have stated that tropical and sub-
tropical fishes are difficult to age, because they general-
ly experience little seasonal and environmental changes
and so do not develop annual rings clearly.
It was observed in C. arel and C. lida that after very
rapid growth during the first year, there is a consid-
erable reduction in the growth rate during the years
when sexual maturity sets in; afterwards, the growth
rate decreases slightly with age. This observation is
consistent with the findings of Ford (1933) and Devoid
(1942).
Females of C. arel show faster growth, compared
with males, and also live longer. According to Pitt
(1966, 1967), since males mature earlier than females,
it seems likely that energy is diverted at an earlier
Length-weight relationships
During the present analyses, C. arel showed differ-
ences in characteristic length-weight slopes for males
and females. Similar observations were made by
Ketchen and Forrester (1966) and Powles (1967), while
analyzing the length-weight relationships of Petrale
sole Eopsetta jordani and American plaice Hippoglos-
soides platessoides, respectively. However, C. lida
showed no significant differences in characteristic
length-weight slopes for males and females. Zoutendyk
(1974b) on Agulhas sole Austroglossvs pectoralis, and
Smith and Daiber (1977) on summer flounder Para-
lichthys dentatus, did not report significant differences
in length-weight characteristics of males and females.
Cynoglossus arel and C. lida showed significant
regression coefficients (b-values), which differed
significantly from the hypothetical B value ( = 3). Webb
(1972) made similar observations for yellow-bellied
flounder Rhombosolea leporina.
Rajaguru: Biology of Cynoglossus arel and C. lida from Indian waters
363
Significant deviation from the 'cube law' was ob-
served in mature females of C. lida due to gonad devel-
opment. Similar findings were observed by Dawson
(1962) in hogchokers Trinectes maculatus, and by Lux
(1969) in yellowtail flounder Limanda ferruginea.
In male and female C. arel, and male, female, and
juvenile C. lida, the exponent values are >3, indicating
that the weight increase is more in relation to length.
But the exponent value for juveniles of C. arel is <3,
indicating that an increase in weight is less compared
with length.
During the present investigation, specimens <83
mmTL in C. arel and <81mmTL in C. lida were not
available from the continental shelf off Porto Novo.
Arora (1951) reported such an absence of juveniles in
the commercial catches for California sand dab Citha-
richthys sordidus. Non-avaUability of juveniles in com-
mercial catches might be due to the gears operated or
due to their occurrence in deeper waters, since spawn-
ing of cjmoglossines in inshore waters has not been
reported (Seshappa and Bhimachar 1955).
Reproductive biology
Spawning periods of C. arel and C. lida were pro-
longed, lasting for 10 months. The present study agrees
with Qasim's (1973a) view that in Indian waters many
fish species may be prolonged breeders. Seshappa and
Bhimachar (1955) reported that the spawning season
in the Malabar sole Cynoglossus semifasciatus off the
west coast of India was prolonged (8 months).
In C. arel and C. lida, ova in different maturity
stages taken from anterior, middle, and posterior re-
gions of both lobes of ovaries showed no variation in
their mean diameter. It is therefore concluded that the
development of ovarian eggs proceeds uniformly
throughout the ovary. Such a distribution of ova has
been reported for Indian halibut Psettodes erumei and
flounder Pseudorhombus arsius (Ramanathan and
Natarajan 1979).
Male C. arel and C. lida attained maturity earlier
than females. Pitt (1966) observed that males of
American plaice Hippoglossoides platessoides were
obviously smaller than females at first maturity.
Results obtained by Lux and Nichy (1969) for yellowtail
flounder Limanda ferruginea (New England), by Wada
(1970b) for Limanda herzensteini (Japan), by Zouten-
dyk (1974a) for Agulhas sole Austroglossus pectoralis
(South Africa), and by Ramos (1982) for Solea solea
(Spain) were similar to the present findings.
The GSI is used widely as an index of gonadal activ-
ity and as an index for spawning preparedness. In male
and female C. arel and in female C. lida, GSI did not
accurately reflect gonadal activity; the relation of
gonadal weight to body weight did not change with
stage of gonadal development, de Vlaming et al. (1982)
stated that the GSI is widely and consistently used for
gonadal size and activity without verification of its
validity. According to de Vlaming et al. (1982), the GSI
is not always the best way of expressing gonadal ac-
tivity, and so this index should not be applied without
validation. Chrzan (1951) concluded that the ratio of
gonad weight to body weight, which normally char-
acterizes sexual maturity, could not be determined ex-
actly. According to Delahunty and de Vlaming (1980),
the exponential relationship between ovarian weight
and body weight did not change with the phase of
oocyte development. However, in male C. lida, higher
values of GSI indicated the occurrence of fully-mature
specimens during this period, and a sudden fall in GSI
value after September appeared to be due to spawn-
ing. Such a relationship between GSI and gonadal ac-
tivity was reported for Indian halibut Psettodes erumei
and flounder Pseudorhombus arsiv^ (Ramanathan and
Natarajan 1979).
In male C. arel and male and female C. lida, a rise
in Kn value did not correspond with a rise in gonadal
activity. Webb (1973) observed no significant variation
in body condition, with onset of spawning, in sand
flounder Rhombosolea plebeia and yellow-bellied
flounder R. leporina off New Zealand. However, in
female C. arel, a rise in Kn value corresponded with
a rise in gonadal activity, and thus showed a positive
correlation.
A linear relationship between fecundity and other
variables (total length, total weight, ovary length, and
ovary weight) was observed in C. arel and C. lida. The
result agrees with those of Hoda (1976). The correla-
tion coefficient between fecundity and total length,
total weight, ovary length, and ovary weight showed
a high positive degree of correlation in C. arel. In
C. lida, the correlation coefficient between fecundity
and ovary length and ovary weight showed a high
positive degree of correlation; whereas the correlation
coefficient between fecundity and total length and total
weight did not show significant correlations. Hence in
C. lida, fecundity was dependent only on ovary length
and ovary weight.
In C. arel and C. lida, fecundity was better correlated
with total length and ovary length than with total
weight and ovary weight. According to Colman (1973),
in sand flounder Rhombosolea plebeia and yellow-bellied
flounder R. leporina off New Zealand, fecundity in-
creased at a rate greater than the cube of length, and
more than proportionately to weight; fecundity was
probably slightly less proportional to ovary weight. Col-
man (1973) suggested that this might be due to large
ovaries containing either great quantities of ovarian
fluid or connective tissue or a high proportion of
nondeveloping eggs.
364
Fishery Bulletin 90|2). 1992
In C. arel and C. Ma, age of fish had no effect on
the number of eggs. Among fish of the same length,
older ones did not contain more eggs than yoimger
ones. This resiilt is consistent with the findings of Simp-
son (1951) and Bagenal (1957).
More fecund C. lida (relative fecundity 287-1664,
X 988) laid smaller eggs (< 0.6250 mm d.m.), while
less fecund C. arel (relative fecundity 124-1096, x
464) laid larger eggs (<0.8125mm d.m.). Dahl (1918)
and Svardson (1949) also found that more fecund
species lay smaller eggs.
In C. arel and C. lida, males outnumbered females
and were relatively smaller in size than females. The
present finding is inconsistent with Qasim's (1966) view
that the sex which outnumbers the other attains a much
bigger size.
Chi-square values showed a significant deviation
from the expected 1 : 1 ratio for 3 months in C. arel and
for 5 months in C. lida. Such a deviation could be due
to a partial segregation of mature forms through
habitat preference (Reynolds 1974), due to migration
(Collignon 1960) or behavioral differences between
sexes (Polonsky and Tormosova 1969), thus rendering
one sex to be more easily caught than another.
During the present investigation, spent individuals
of C. arel and C. lida were not found throughout the
study period, since spawning of tonguefishes appears
to take place mainly in deeper waters, as observed by
Seshappa and Bhimachar (1955) for Malabar sole
Cynoglossus semifasciatus. This is a gap in the repro-
ductive biology of these tonguefishes. Deep-sea fishing
is needed to confirm this type of spawning behavior
by the tonguefishes.
For C. arel, the spawning peak was in January which
is the post-(northeast) monsoon period along the south-
east coast of India. Monsoonal floods end by this time,
and food resources (like copepods and amphipods,
which are essential food items of juveniles) are abun-
dant; this season would appear to be a favorable time
for spawning. The spawning peak of C. lida was in Sep-
tember, which is the pre-(northeast) monsoon period
along the southeast coast of India; this period coincides
with the most active southwest monsoon period along
the west coast of India. Most of the rivers originating
in the west receive floodwaters through the southwest
monsoon and empty them into the Bay of Bengal, which
thus gets rich primary food resources at this time. This
period would also appear to be a favorable time for
spawning, because of food abundance.
Thus C. arel and C. lida, though co-occurring sym-
patrically in the continental shelf waters off Porto
Novo, share available food resources and appear to
avoid competition for food and space for their juveniles
by exhibiting spawning peaks during different periods
(pre-/post-northeast monsoon).
Acknowledgments
I am indebted to Dr. Bruce B. Collette, Systematics
Laboratory, for his many helpful suggestions in the
modification of this manuscript for publication, and for
providing the facilities of the Systematics Laboratory.
My sincere thanks to Dr. Thomas Munroe for critical-
ly reviewing part (food and feeding) of the manuscript.
Comments made by the editor, and two anonymous
reviewers helped in modifying this manuscript. Special
thanks go to Ms. Ruth E. Gibbons, for "her suggestions
regarding computer graphics. Grateful acknowledge-
ment is extended to the University Grants Commission,
New Delhi, to Dr. R. Natarajan, and to the authorities
of Annamalai University, India, for offering financial
support and necessary facilities to carry out this study
as a part of my Ph.D. program. I thank Dr. K. Srira-
man, Tamil Nadu Fisheries, Porto Novo, for his help
in the statistics. My sister, Miss G. Shantha, helped me
in processing the data. Computer help extended by Mr.
H.A. Kurt Luginbyhl, Dr. Jeffrey Williams, Mr. Tom
Orrell, and Mr. Jeffrey Howe is greatly appreciated.
Citations
Arora, H.L.
1951 An investigation of the California sand dab, Citharkkthys
sordidus (Girard). Calif. Fish Game 37(l):3-42.
Bagenal, T.B.
1957 The breeding and fecundity of the long rough dab Hip-
poglossmdes platessoides (Fabr.) and the associated cycle in con-
dition. J. Mar. Biol. Assoc. U.K. 36(2):339-375.
Bal, D.V., and V. Rao
1984 Marine fisheries. Tata McGraw-Hill, New Delhi, 470 p.
Botha, L., D.K. Lucks, and D.S. Chalmers
1971 Mesh selectivity experiments on the east coast sole. S.
Afr. Ship. News & Fishing Ind. Rev. 26(10):50-57.
Bowman, T.E., A.C. Cohen, and M.M. McGuiness
1982 Vertical distribution of Themisto gaudichaudii (Am-
phipodaiHyperiidea) in deep-water dumpsite 106 off the mouth
of Delaware Bay. Smithson. Contrib. Zool. 351. 24 p.
Braber, L., and S.J. deGroot
1973a On the morphology of the alimentary tract of flatfishes
(Pleuronectiformes). J. Fish. Biol. 5(2):147-153.
1973b The food of five flatfish species (Pleuronectiformes) in
the southern North Sea. Neth. J. Sea Res. 6(1-2):163-172.
Cassie, R.M.
1954 Some uses of probability paper in the analysis of size
frequency distributions. Aust. J. Mar. Freshwater Res. 5(3):
513-522.
Chrzan, F.
1951 Studies on the biology of the cod in the Gulf of Gdansk.
Rep. Sea Fish. Inst. (Gdynia, Poland) 6:1-28.
Clark, F.N.
1934 Maturity of the California sardine {Sardina caenilea)
determined by ova diameter measurements. Calif. Dep. Fish.
Game Fish Bull. 42, 49 p.
Rajaguru: Biology of Cynoglossus arel and C lida from Indian waters
365
Clark, M.R., K.J. King, and P.J. McMillan
1989 The food and feeding relationships of black oreo, Allocyt-
tus niger, smooth oreo, Pseudocyttus maculatus, and eight
other fish species from the continentcd slope of the south-west
Chatham Rise, New Zealand. J. Fish. Biol. 35(4):465-484.
CMFRI
1969 Marine fish production in India, 1950-1968. Bull. Cent.
Mar. Fish. Res. Inst. 13, Cochin, India, 144 p.
CoUignon, J.
1960 Contribution a la connaissance des Otolithus des cotes
d'Afrique Equatoriale. Bull. Inst. Etudes Centrafracaines
Brazzaville, N.S. 19-20:55-84.
Colman, J. A.
1973 Spawning and fecundity of two flounder species in the
Hauraki Gulf, New Zealand. N.Z. J. Mar. Freshwater Res.
7(l&2):21-43.
Daan, N.
1973 A quantitative analysis of the food intake of North Sea
cod, Gadus morhua. Neth. J. Sea Res. 6(4):479-517.
Dahl, K.
1918 Studies of trout and trout waters in Norway. Salmon
Trout Mag. 17:58-79.
Dawson, C.E.
1962 Length-weight and standard length-total length relation-
ships of South Carolina hogchokers, Trinectes maculatus.
Trans. Am. Fish. Soc. 91(l):89-90.
deGroot, S.J.
1971 On the interrelationships between morphology of the
alimentary tract, food and feeding behaviour in flatfishes
(Pisces:Pleuronectiformes). Neth. J. Sea Res. .5(2):121-196.
de Jong, J.K.
1940 A preliminary investigation of the spawning habits of
some fishes of Java Sea. Treubia Buitenzorg 17(4):307-330.
Delahunty, G., and V.L. deVlaming
1980 Seasonal relationships of ovary weight, liver weight and
fat stores with body weight in the gold fish, Carassius auratus
(L.). J. Fish. Biol. 16(1):5-13.
Devadoss, P., and P.K.M. Pillai
1973 Observations on the food of juveniles otPsettodes erumei
(Bloch). Indian J. Fish. 20(2):664-667.
deVlaming, V., G. Grossman, and F. Chapman
1982 On the use of the Gonadosomatic Index. Comp. Bio-
chem. Physiol. 73A(l):31-39.
Devoid, F.
1942 Plaice investigations in Norwegian waters. Rep. Norw.
Fish. Mar. Invest. VII(3), 83 p.
Fischer, W., and G. Bianchi
1984 FAO species identification sheets for fishery purposes. II.
W. Indian Ocean-Fishing area 51. FAO, Rome.
Ford, E.
1933 An account of the herring investigations conducted at
Plymouth during the years from 1924-'33. J. Mar. Biol.
Assoc. U.K. 19(l):305-384.
Harding, J.P.
1949 The use of probability paper for the graphical analysis
of polymodal frequency distributions. J. Mar. Biol. Assoc.
U.K. 28(1):141-153.
Hickling, C.F., and E. Rutenberg
1936 The ovary as an indicator of the spawning period in
fishes. J. Mar. Biol. Assoc. U.K. 21(1):311-317.
Hoda, S.M.S.
1976 Reproductive biology and length-weight relationship of
Thryssa dussumieri (Val.) of the Pakistan coast. J. Mar. Biol.
Assoc. India 18(2):272-287.
Honda, H.
1984 Food acquisition patterns in some demersal teleosts.
Tohoku J. Agric. Res. 35(l):33-54.
Isaacs, J.D., and R.A. Schwartzlose
1965 Migrant sound scatterers: Interaction with the sea floor.
Science (Wash. DC) 150(3705):1810-1813.
James, P.S.B.R.
1967 The ribbonfishes of the family Trichiuridae of India.
Mem. Mar. Biol. Assoc. India 1, 226 p.
Ketchen, K.S., and C.R. Forrester
1966 Population dynamics of the Petrale sole, Eopsetta jor-
dani, in waters off western Canada. Bull. Fish. Res. Board
Can. 153, 195 p.
Kuthalingam, M.D.K.
1957 Life history and feeding habits of Cynoglossus lingua
(Ham. Buch.). J. Zool. Soc. India 9(2):208-215.
Lande, R.
1976 Food and feeding habits of the dab (Limanda limanda
[L.]) in Borgenfjorden, north Trondelag, Norway. Norw. J.
Zool. 24(3):225-230.
Langton, R.W.
1983 Food habits of yellow-tail flounder, Limanda ferruginea
(Storer), from off the northeastern United States. Fish. Bull.,
U.S. 81(l):15-22.
Langton, R.W., and R.E. Bowman
1981 Food of eight northwest Atlantic pleuronectiform fishes.
NOAA Tech. Rep. NMFS-SSRF 749,16 p.
Le Cren, E.D.
1951 The length-weight relationship and seasonal cycle in
gonad weight and condition in the perch (Percajluviatilis). J.
Anim. Ecol. 20(2):201-219.
Lovern, J. A., and H. Wood
1937 Variations in the chemical composition of herring. J.
Mar. Biol. Assoc. U.K. 22(l):281-293.
Lux, F.E.
1969 Length-weight relationships of six New England flat-
fishes. Trans. Am. Fish. Soc. 98(4):617-621.
1970 Note on growth of American plaice, Hippoglossoides
platessoides (Fabr.) in ICNAF Subarea 5. Int. Comm. North-
west Atl. Fish. Spec. Publ. 7:5-7.
Lux, F.E., and F.E. Nichy
1969 Growth of yellowtail flounder, Limanda ferruginea
(Storer), on three New England fishing grounds. Int. Comm.
Northwest Atl. Fish. Res. Bull. 6:5-25.
Nikolsky, G.V.
1963 The ecology of fishes. Academic Press, NY, 352 p.
Pearcy, W.G., and D. Hancock
1978 Feeding habits of dover sole, Microstomus padficus ; rex
sole, Glyptocephalus zachirus ; slender sole, Lyopsetta exilis ;
and Pacific sand dab, Citharichthys sordidus, in a region of
diverse sediments and bathymetry off Oregon. Fish. Bull.,
U.S. 76(3):641-651.
Pereyra, W.T., W.G. Pearcy, and F.E. Carvey Jr.
1969 Sebastesjlavidus, a shelf rockfish feeding on mesopelagic
fauna, viith consideration of the ecological implications. J.
Fish. Res. Board Can. 26:2211-2215.
Pinkas, L., M.S. Oliphant, and I.L.K. Iverson
1971 Food habits of albacore, bluefin tuna and bonito in Califor-
nian waters. Calif. Fish Game 152, 105 p.
Pitt, T.K.
1966 Sexual maturity and spawning of American plaice, Hip-
poglossoides platessoides (Fabricius), from Newfoundland and
Grand Bank areas. J. Fish. Res. Board Can. 23(5):651-672.
1967 Age and growth of American plaice, (Hippoglossoides
platessoides) in the Newfoundland area of the Northwest Atlan-
tic. J. Fish. Res. Board Can. 24(5):1077-1099.
366
Fishery Bulletin 90(2), 1992
Polonsky, A.S., and I.D. Tormosova
1969 The spawning of the jack-mackerel of the northeast Atlan-
tic and the distribution of its eggs and larvae. Trudy Atl.
Nauchno-Issled. Inst. Rybn. Khoz. Okeanogr. 23:27-48 [in
Russian].
Powles, P.M.
1967 Length-weight relationships for American plaice, witch,
and yellowtail, in ICNAF Subarea 4. Int. Comm. Northwest
Atl. Fish. Res. Bull. 4:121-123.
Pradhan, M.J.
1969 Fishery and biology of Psettodes eruTnei (Schneider) an
Indian Ocean flatfish. Bull. Natl. Inst. Sci. India 38, Part II.
Proc, Symp. Indian Ocean, Sect. IV, Fisheries, p. 885-894.
Qasim, S.Z.
1966 Sex ratio in fish population as a function of sexual dif-
ferences in growth rate. Curr. Sci. (Bangalore) 35(6):140-142.
1973a An appraisal of the studies on maturation and spawn-
ing in marine teleosts from Indian waters. Indian J. Fish.
20(1):166-181.
1973b Some implications of the problem of age and growth
in marine fishes from the Indian waters. Indian J. Fish.
20(2):351-371.
RajaguTU, A.
1987 Studies on taxonomy of Indian flatfishes (Teleostei:
Pleuronectiformes) and biology of two species of Cynoglossus
Ham. Buch., 1822 (Cynoglossidae) from Porto Novo. Unpubl.
Ph.D. diss., Annamalai Univ., Tamil Nadu, India, 301 p.
Ramanathan, N.
1977 Studies in flatfishes of Porto Novo waters (Teleostei:
Pleuronectiformes). Unpubl. Ph.D. diss., Annamalai Univ.,
Tamil Nadu, India. 250 p.
Ramanathan, N., and R. Natarajan
1979 Breeding biology oi Psettodes erumei (Bloch. and Schn.)
and Pseudorhomincs arsius (Ham. Buch.) (Pisces:Pleuronec-
tiformes), along Porto Novo coast (S. India). Aquaculture
18(3):269-282.
1980 Food and feeding habits of Psettodes erumei (Bloch and
Schn.) and Pseudorhominis arsius (Ham. Buch.). Matsya (BuU.
Indian Soc. Ichthyol.) 6:39-42.
Ramos, J.
1982 Estudio de la edad y crecimiento del lenguado, Solea solea
(Linneo, 1758) (Pisces:Soleidae). Invest. Pesq. 46 (l):15-28.
Reynolds, J.D.
1974 Biology of small pelagic fishes in the new Volta Lake in
Ghana. Part III: Sex and reproduction. Hydrobiologia 45(4):
489-508.
Roe, H.S.J., P.T. James, and M.H. Thurston
1984 The diel migrations and distributions within a mesopelagic
community in the north east Atlantic. 6. Medusae, cteno-
phores, amphipods and euphausiids. Prog. Oceanogr. 13:
425-460.
Sedberry, G.R., and J. A. Musick
1978 Feeding strategies of some demersal fishes of the con-
tinental slope and rise off the mid-Atlantic coast of the
USA. Mar. Biol. (Beri.) 44:357-375.
Seshappa, G.
1973 The flatfish resources of the west coast of India. Proc,
Symp. Living Resour. Seas around India, CMFRI (Cochin,
India), p. 470-482.
1978 Some observations on the size distribution and the oc-
currence of growth rings in the scales of three species of
Cynoglossus at Calicut. Indian J. Fish. 25 (1&2): 188-196.
Seshappa, G., and B.S. Bhimachar
1955 Studies on the fishery and biology of the Malabar sole,
Cynoglossiis semifasciatns Day. Indian J. Fish. 2(l):180-230.
Simpson, A.C.
1951 The fecundity of the plaice. Fish. Invest. Minist. Agrl.
Fish. U.K. Ser. II, 17(5), 27 p.
Smith, R.W., and F.C. Daiber
1977 Biology of the summer flounder, Paralichthys dentatus,
in Delaware Bay. Fish. Bull., U.S. 75(4):823-830.
Snedecor, G.W.
1956 Statistical methods, 5th ed. Iowa State Coll. Press,
Ames, 534 p.
Stickney, R.R.
1976 Food habits of Georgia estuarine fishes. II. Symphurus
plagiusa (Pleuronectiformes:Cynoglossidae). Trans. Am. Fish.
Soc. 105(2):202-207.
Stickney, R.R., G.L. Taylor, and R.W. Heard III
1974 Food habits of Georgia estuarine fishes. I. Four species
of flounders (Pleuronectiformes:Bothidae). Fish. Bull., U.S.
72:515-525.
Struhsaker, P., and J.H. Uchiyama
1976 Age and growth of the nehu, Stolephorus purpureus
(Pisces: Engraulidae), from the Hawaiian islands as indicated
by daily growth increments of sagittae. Fish. Bull., U.S.
74:9-17.
Svardson, G.
1949 Natural selection and egg number in fish. Rep. Inst.
Freshwater Res., Drottningholm 29:115-122.
von Bertalanffy, L.
1938 A quantitative theory of organic growth. Hum. Biol.
10(2):181-213.
Wada, K.
1970a Studies on the population biology of the flatfish, Lirmm-
da herzensteini Jordan et Snyder in Niigata Region. Bull. Jpn.
Sea Reg. Fish. Res. Lab. 22:31-43.
1970b Studies on the population biology of the flatfish, Liman-
da herzensteini Jordan et Snyder in Niigata Region. II. Matur-
ity and spawning. Bull. Jpn. Sea Reg. Fish. Res. Lab. 22:
45-57.
Wakabara, V., E.K. deRezende, and A.S. Tararam
1982 Amphipods as one of the main food components of three
pleuronectiforms from the continental shelf of South Brazil and
North Uruguay. Mar. Biol. (Beri.) 68:67-70.
Walford, L.A.
1946 A new graphic method of describing growth of animals.
Biol. Bull. (Woods Hole) 90(2):141-147.
Webb, B.F.
1972 Fish populations of the Avon-Heathcote estuary. 1.
General ecology, distribution, and length-frequency. N.Z. J.
Mar. Freshwater Res. 6(4):570-601.
1973 Fish populations of the Avon-Heathcote estuary. 2.
Breeding and gonad maturity. N.Z. J. Mar. Freshwater Res.
7(l&2):45-66.
Wingfield, J.C, and N.S. Grimm
1977 Seasonal changes in plasma Cortisol, testosterone and
oestradiol-17p in the plaice, Pleuronectes platessa L. Gen.
Comp. Endocrinol. 31:1-11.
Zoutendyk, P.
1974a The biology of the Agulhas sole, Austroglossus pec-
toralis, Part-2. Age and g^rowth. Trans. R. Soc. S. Afr.
41(1):33-41.
1974b The biology of the Agulhas sole, Austroglossvs pec-
toralis, Part-3. Length-weight relationships. Trans. R. Soc.
S. Afr. 41(2):99-108.
Rajaguru: Biology of Cynoglossus are/ and C. hda from Indian waters
367
Appendix
Monthly sample sizes
(n) for various
studies
on biology of Cynoglossus arel and C. lida
caught
commercially off Porto Novo,
India,
October 1981-September ]
982. GI
= gastrosomatic
index; HI = hepatosomatic index;
GSI =
gonadosomatic index; L-W = length- |
weight relationships;
Kn =
= relative condition factor.
Months
GI/HI/GSI/Sex ratio
L-W/Kn
Age and growth
C.
arel
C.
lida
C.
arel
C. lida
C
arel
a
lida
a
9
cr
9
Of
9
a
9
a
9
cr
9
January
53
39
32
42
52
49
24
31
54
39
34
45
February
24
45
59
16
34
52
48
45
24
44
58
16
March
54
49
64
18
51
47
85
38
54
51
64
18
April
46
55
84
51
46
58
86
54
48
55
84
51
May
56
54
38
31
57
52
62
43
56
54
39
32
June
35
40
56
59
44
41
54
54
35
40
56
59
July
69
56
25
25
70
55
37
44
69
56
25
26
August
44
54
43
62
54
66
45
64
45
54
43
62
September
56
35
46
25
59
39
62
37
57
35
50
26
October
75
54
83
76
81
57
83
80
76
55
83
77
November
86
54
160
207
70
45
135
168
87
54
160
207
December
29
34
28
28
37
38
47
65
29
37
28
31
Total
627
569
718
640
655
599
768
723
634
569
724
650
Abstract.- The inverse method
for mortality and growth estimation
(IMMAGE) is a new approach to
obtain unbiased estimates of mortal-
ity and growth parameters for lar-
val fishes from length-frequency
data biased by the size selectivity of
plankton nets. The performance of
IMMAGE is compared with methods
which attempt to eliminate selection
bias from sampled length-frequen-
cies. Using Monte Carlo simulations,
IMMAGE estimates growth and
mortality parameters that are more
accurate and precise than those pro-
duced by other methods.
Inverse method for mortality
and growth estimation: A new
method for larval fishes
David A. Somerton
Honolulu Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA
2570 Dole Street, Honolulu, Hawaii 96822-2396
Present address: Alaska Fisheries Science Center
National Marine Fisheries Service, NOAA
7600 Sand Point Way NE, Seattle, Washington 98 11 5-0070
Donald R. Kobayashi
Honolulu Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA
2570 Dole Street Honolulu, Hawaii 96822-2396
Manuscript accepted 9 March 1992.
Fishery Bulletin, U.S. 90:368-375 (1992).
Estimation of the growth and mor-
taHty rates of larval fishes is com-
plicated by the sampling biases that
can result from the size selectivity
of plankton nets. Size selectivity due
to net avoidance by larvae, for exam-
ple, results in an underestimation of
larval abundance that progressively
increases with increasing larval
length. This bias leads to an under-
estimation of mean length-at-age and
therefore growth rate, because
larger larvae in each age-class are
underrepresented relative to smaller
larvae. Such bias also leads to an
overestimation of mortality rate,
because older larvae are underrepre-
sented relative to younger larvae.
Size selectivity due to extrusion of
larvae results in an underestimation
of larval abundance that progressive-
ly decreases with increasing larval
length and likewise leads to bias in
estimates of growth and mortality
rates.
Grovirth studies rarely address such
size selection, and when they do, the
approach taken is usually to devise a
sampling procedure that provides un-
biased length-frequencies (Methot
and Kramer 1979, Yoklavich and
Bailey 1990). Mortality studies, by
comparison, almost always address
size selection and do so after the fact
by taking one of several approaches
to eliminate the selection bias from
the sampled length-frequencies. One
approach taken by mortality studies
is to divide the sampled length-fre-
quencies by length-specific estimates
of capture probability. Such capture
probabilities have been obtained from
field studies and estimated as (1)
catch ratios of large to small mesh
nets (Lenarz 1972, Leak and Houde
1987), (2) catch ratios of day to night
sampling (Houde 1977, Zweifel and
Smith 1981, Morse 1989, Somerton
and Kobayashi 1989), or (3) catch
ratios of plankton nets to purse
seines (Murphy and Clutter 1972,
Leak and Houde 1987). Capture
probabilities have also been based on
theoretical escapement models (Ware
and Lambert 1985). A second ap-
proach taken by mortality studies is
to simply eliminate the biased por-
tions of the length distribution. Such
elimination may exclude only small
(Morse 1989) or large larvae (Houde
1977, Methot and Kramer 1979) or
both (Essig and Cole 1986). Ehmina-
tion of biased length-frequencies also
has been combined with the use of
capture probabilities (Houde 1977,
Morse 1989).
368
Somerton and Kobayashi: Inverse method for mortality and growth estimation
369
Although the various approaches may differ in the
specifics of their application, all are based on the
premise that length-frequency data must first be cor-
rected for selection bias before they can be utilized to
estimate growth and mortality. Here we introduce a
new approach which reorders and joins the processes
of data correction and parameter estimation. This
approach, which is based on a stock assessment tech-
nique known as synthesis modeling (Methot 1989,
1990), will herein be referred to as the inverse method
for mortality and growth estimation (IMMAGE). The
use of IMMAGE is examined to estimate growth and
mortality rates from biased length-frequency data.
Additionally, the performance of IMMAGE, using
Monte Carlo simulation, is compared with approaches
used to correct selection bias in length-frequency data
prior to parameter estimation.
Materials and methods
IMMAGE vs. bias correction
To understand how IMMAGE works and why it is an
inverse method for obtaining estimates of growth and
mortality parameters, the bias-correction approach
should first be examined (Fig. la-d). One variant of
the bias-correction approach might include (1) esti-
mating the unbiased length-frequency distribution
(Fig. lb) by dividing the observed length-frequency
distribution (Fig. la) by length-specific estimates of
capture probability; (2) converting the unbiased length-
frequency distribution to an age-frequency distribution
(Fig. Ic) using age and length information; and (3)
estimating the instantaneous mortality rate (M) as the
slope of a straight line fit to the logarithms of numbers
at age (Fig. Id).
The IMMAGE approach, if applied to the same data,
would include (1) choosing initial values for M and the
number of day-0 larvae (Ng); (2) estimating an un-
biased age-frequency distribution (Fig. le) based on the
values of M and Nq; (3) estimating the unbiased
length-frequency distribution (Fig. If) from the age-
frequency distribution using age and length informa-
tion; (4) estimating the observed (i.e., biased) length-
frequency distribution (Fig. Ig) by multiplying the
unbiased length-frequency distribution by estimates of
capture probability; and (5) iteratively varying M and
Nq, and repeating steps 2-4, until the best fit is
achieved between the estimated and observed length-
frequency distributions.
Thus both approaches estimate M by fitting a mor-
tality model. However, in the bias correction approach,
the model is fit to numbers-at-age derived from the
observed length-frequencies; while in the IMMAGE
approach, the model is fit to the observed length-
frequencies themselves. Growth parameters are esti-
mated by IMMAGE in a similar manner, except a
growth model rather than a mortality model is fit to
the length-frequencies.
To estimate the observed length distribution,
IMMAGE requires specification of a process model and
ancillary data. The process model contains parameters
that are iteratively varied to achieve the best fit to the
observed length-frequency distribution; the ancillary
data are parameters assumed to be known. For growth
estimation, the process model consists of a growth func-
tion describing the mean length-at-age and a variance
function describing the variance in length-at-age. An-
cillary data include estimates of the capture probabil-
ity at each length. For mortality estimation, the pro-
cess model consists of a mortality function describing
the instantaneous mortality rate at age or length. An-
cillary data include the mean and variance in length-
at-age, and the capture probability at each length.
Growth and mortality process models are not restricted
to any particular form and may include linear or
nonlinear functions.
The performance of IMMAGE and several of the bias
correction approaches to parameter estimation was
examined by using a Monte Carlo simulation model.
For growth parameter estimation, bias correction ap-
proaches were not examined because no application to
larval fishes could be found in the literature. For mor-
tality parameter estimation, three bias correction ap-
proaches were examined: (1) elimination of the biased
portions of the observed length-frequency distribution,
(2) division of the observed length-frequency distribu-
tion by estimates of capture probability (correction),
and (3) elimination of the biased ages from a corrected
age distribution.
Monte Carlo model
The Monte Carlo simulation model is designed to mimic
the sequence of steps typically used in growth and mor-
tality studies. A central feature of this model is the
simulated collection of three types of data: length-
frequency samples, selection samples, and ageing
samples.
Length-frequency samples are either unbiased, rep-
resenting random samples drawn from a larval fish
population, or biased, representing samples collected
with a plankton net. Selection samples are two indepen-
dent length-frequency samples, one biased and the
other unbiased, used to estimate length-specific cap-
ture probabilities. Such samples represent those that
might be produced by an experiment to estimate the
length-selection characteristics of a plankton net (i.e.,
day to night catch comparisons). Ageing samples are
length-frequency samples in which each length mea-
370
Fishery Bulletin 90(2). 1992
Figure I
Comparison of the bias-
correction and IMMAGE
approaches to estimating
instantaneous mortality
rate (M) from selection-
biased length-frequency
data. Bias correction Oeft
column) begins by dividing
(a) the observed length-
frequency distribution by
estimates of capture prob-
ability to estimate (b) the
unbiased length distribu-
tion. The unbiased length
distribution is then con-
verted to (c) an age distribu-
tion, and M is estimated
with (d) linear regression.
IMMAGE (right column)
begins by creating (e) an
unbiased age distribution
using initial estimates of M
and the number of day-0
larvae, N„. The unbiased
age distribution is con-
verted to (f) an unbiased
length distribution using
the ageing sample, (g) The
unbiased length distribution
is multiplied by the capture
probabilities to estimate the
sampled length distribution
(solid line), then mortality
estimates are varied itera-
tively to minimize the resid-
ual sum of squares between
the observed (histogram)
and the estimated length
distributions.
BIAS CORRECTION
IMMAGE
150
12S
£100
75
, I r?',lU
flHr., ,,,Ti,
4 6 0 10 12 14 16 ie 20 22 24 26 26 30
Langth (mm)
4 6 B 10 12 14 16 16 20 22 24 26 28 30
L*ngth (mm)
•| 10
4 6 9 10 12 14 16 16 20 22 24 26 26 30
Length (mm)
4 6 0 10 12 14 16 162022 24 26263032 34
Length (mm)
"300
I 200
c
r
p— 1 — 1 — I-
I 200
i
I
wm
Hf^^nnr
012345676 9 10 1112 13 14 15
Age (days)
0123456769 101112131415
Age (days)
MORTALITY
ESTIMATE
0123456769 101112131415
Age (days)
surement is associated with an age. Ageing samples
are considered biased when used in growth parameter
estimation but are considered unbiased when used in
mortality parameter estimation. This distinction is made
because mortality parameters can be influenced by
selection bias in ageing samples as well as by selection
bias in the length-frequency samples. To simplify inter-
pretation of the results and avoid compounding the ef-
fects of the two sources of bias, bias in the ageing
samples has been ignored in the mortality estimation.
Somerton and Kobayashi: Inverse method for mortality and growth estimation
371
- 1 00
200
-0.80
ISO
//V \ ?
-0.60
100-
IM^ \ '
0.40
1 260
yy/^^^^^^^ ^
-0.20
b
z
A
200-
/A
■
160-
//v
■
100-
jm.
•
60-
IJJ^^
■
0 5 10 16 20 25 30 35 40
Langth (mm)
Figure 2
(a) Unbiased length distribution and relative abundance of each
age-class in the simulated population of larvae, and the as-
sumed values of capture probability (solid line) and the length
of 95% selection (dotted line). With the length elimination
method of bias corrector, all lengths to the right of the 95%
selector line are eliminated, (b) Length distribution and
relative abundance of each age-class within a sample collected
with a size-selective plankton net. This distribution is created
by multiplying the length distribution shown in (a) by the cap-
ture probabilities.
cumulatively summing across all lengths. Individual
lengths within a sample were chosen by determining
which category in the cumulative length probability
distribution just exceeded the value of a generated
uniform random number.
Biased length-frequency samples were generated by
simulating the sampling of the model population by
using a plankton net, which allowed zero extrusion and
produced capture probabilities (Pc.i) described by an
inverse logistic function:
Pc.l = 1
1 -H 9.00 X 10^ e -2-611
(3)
where 1 is length (in mm) [Fig. 2a; parameters in Eq.
(1-3) were chosen arbitrarily and were not intended
to represent any particular species or sampling gear].
Samples were drawn by using the same procedure as
used for unbiased samples except the population length
distribution was multiplied by the capture probabOities
(Fig. 2b).
Ageing samples were generated similar to length-
frequency samples, but after each length was drawn,
an associated age was also drawn by using the cum-
ulative age probability distribution at each length and
an additional uniform random number. The sample
sizes used in the simulations [1000 length-frequency
samples, 300 ageing samples, 600 selection samples,
with the biased sample size set equal to the unbiased
sample size x the average probability of capture deter-
mined from Equation (3)] were arbitrarily chosen but
were similar to those used in Somerton and Kobayashi
(1989, unpubl. data).
Unbiased length-frequency samples for these three
types of data were generated by simulating the ran-
dom sampling of a larval fish population (Fig. 2a) with
a constant daily recruitment, a constant instantaneous
daily mortality rate (M) of 0.20, and a length distribu-
tion at each age conforming to a normal probability
distribution. Mean (It) and variance (Var(lt)) of
length-at-age were chosen, for simplicity, to be linear
functions of age:
It = 10.00 H- 1.50t, and
Var(lt) = 2.50 + 0.2&t,
(1)
(2)
where t is age (in days) and It is length (in milli-
meters). Samples were drawn from the cumulative
length probability distribution of this population that
was constructed by dividing each of the population
length-frequencies by the total sample size, then
Growth simulations
Growth simulations examining the performance of
IMMAGE consisted of 1000 repetitions of the follow-
ing sequence. First, a biased ageing sample and a selec-
tion sample were generated. Second, capture prob-
abilities were estimated from the selection sample by
fitting an inverse logistic function of length, using
nonlinear regression, to the ratios of the biased to the
unbiased length-frequencies. Third, initial parameter
estimates for the growth process model [Eq. (1) and
(2)] were obtained from the ageing sample by fitting
straight lines to length-at-age and variance of length-
at-age. Fourth, the unbiased length distribution of each
age-class was estimated as a normal distribution with
mean and variance predicted from Eq. (1) and (2)
evaluated at the current parameter estimates. Fifth,
the biased length-frequency distribution of each age-
class was estimated by multiplying the unbiased length-
frequency distribution by the estimated capture prob-
abilities. Sixth, parameter estimates for Eq. (1) and (2)
372
Fishery Bulletin 90(2), 1992
were iteratively varied, and steps 4 and 5 were re-
peated, until the minimum residual sum of squares
(RSS) was achieved. The RSS was defined as
I I (Fu - h)'
(4)
where Fjj and Fjj are the observed and estimated
frequency within the ith length interval and the jth
age-class.
Mortality simulations
Mortality simulations examined the performance of
IMMAGE and three bias-correction approaches: length
elimination, division by capture probabilities, and age
elimination. Each of the 1000 repetitions of a simula-
tion began by generating a biased length-frequency
sample, an unbiased ageing sample, and a selection
sample. For all simulations, except those examining
IMMAGE, mortality estimation began with an attempt
to derive an unbiased age distribution. If length elim-
ination was used, this was accomplished in two stages.
First, an unbiased length-frequency distribution was
estimated by eliminating all length categories with a
capture probability of <0.95 [based on capture prob-
abilities defined by Equation (3), length-classes 3-18
were retained; Fig. 2a]; a probability of 0.95 was used
instead of 1.00 because it is better defined. Second, the
age-frequency distribution was estimated from the un-
biased length-frequency distribution by using the age-
ing sample and a procedure known as age-slicing
(Mesnil and Shepherd 1990). To do this, lengths from
the ageing sample were regressed on the ages, and the
fitted linear regression equation was evaluated to
determine the length corresponding to each age bound-
ary (i.e., 0.5, 1.5, 2.5 days, and so on). Age-frequencies
were then estimated by summing length-frequencies
between age boundaries.
If division by capture probabilities was used, the age-
frequency distribution was estimated by first dividing
the observed length-frequency distribution by esti-
mates of capture probability, then converting the
length-frequencies to age-frequencies using age-slicing.
If age elimination was used, age-classes with a capture
probability of <0.95 at the mean length also were
eliminated from the age-frequency distribution (age-
classes 0-5 days were retained; Fig. 2a). For all three
cases, instantaneous mortality rate was then estimated
as the slope of an unweighted linear regression to the
natural logarithm of numbers-at-age (Ricker 1975).
For simulations examining IMMAGE, mortality esti-
mation proceeded as follows. First, initial values of M
and No were obtained from the ageing sample by fit-
ting a straight line to the logarithm of numbers-at-
age. Second, values of It and Var(lt) were estimated
from the ageing sample by fitting straight lines to
length-at-age and Var(lt)-at-age, using linear regres-
sion, and then evaluating the fitted regression equa-
tions at each t. Third, capture probabilities were
estimated from the selection sample. Fourth, the un-
biased age distribution was estimated as Nt=No e '^^.
Fifth, the unbiased length distribution of each age-class
was estimated as Nt times a normal probability distri-
bution with a mean equal to It and a standard devia-
tion equal to the square root of Var(lt). The unbiased
length distribution for the population was then esti-
mated by summing the age-specific length distributions
over all age-classes. Sixth, the biased (observed) length
distribution was estimated by multiplying the unbiased
population length distribution by the estimated capture
probabilities. Seventh, Nq and M were iteratively
varied, and steps 4-6 repeated, untO the minimum RSS
was achieved. The RSS was defined as
(Fj - Fj)2
(5)
where Fj and Fj are the observed and estimated fre-
quencies within the jth length interval.
The IMMAGE application used in the simulations
(i.e., one that assumes linear growth and constant mor-
tality) is available from the authors as a stand-alone
program (IMMAGE, written in Microsoft QuickBasic)
designed to run on IBM-compatible microcomputers.
Results and discussion
Growth
The type of size selection examined in the simulations
(i.e., a decrease in the probability of capture with in-
creasing larval length) complicates the estimation of
growth and mortality in slightly different ways. For
growth estimation, the primary effect is that the
largest larvae in each age-class are undersampled
relative to the smallest larvae, and the mean lengths-
at-age are therefore underestimated (Fig. 2b). Since
the bias progressively increases with age, plots of mean
length against age are curvilinear and falsely indicate
a declining growth rate (Fig. 3). Such curvilinear or
even asymptotic growth patterns are often reported
in studies of wild-caught larvae (Bailey 1982, Laroche
et al. 1982, Lough et al. 1982, Thorrold 1988, Warlen
1988). Although there may be biological reasons to ex-
pect such a pattern, especially for species with pro-
nounced ontogenetic changes in body form, length-
selective sampling may be a contributing factor.
IMMAGE estimates of the slope (1.499 ±0.005; x
1000 replicates ±2 SE) and intercept (10.002 + 0.012)
Somerton and Kobayashi: Inverse method for mortality and growth estimation
373
Figure 3
Mean length-at-age in the simulated population (•), mean
length-at-age estimated without correction for size selection
(■), and mean length-at-age (±2 SE) estimated with IM-
MAGE (vertical bars).
of the linear growth function [cf. Eq. (1)] were both
unbiased, as were all of the estimates of mean length-
at-age over the entire age range (Fig. 3). This indicates
that, at least for linear growth, IMMAGE provides un-
biased estimates of growth parameters from biased
length and age samples.
Mortality
For mortality estimation, the primary effect of the
decrease in capture probability with increasing larval
length is that relative abundance is progressively
underestimated with increasing age (Fig. 2b). If ig-
nored, such a progressive underestimation would result
in positively biased mortality estimates. In the sim-
ulated population, for example, mortality estimates
obtained from the observed length-frequency samples
(M = 0.450 ± 0.004) had a highly significant positive bias
of 125% (Fig. 4a).
Elimination of the biased length-frequencies was only
partially effective in reducing the bias in estimated
mortality rates, because the mortality estimates (M =
0.364 ± 0.007) still had a highly significant positive bias
of 80% (Fig. 4b). In practice, length elimination is likely
to be even less effective than it appears here, because
it is usually applied to cases where the capture prob-
abilities are crudely known, whereas exact knowledge
is assumed in the simulations.
Figure 4 (right)
Frequency distribution of 1000 simulated estimates of instantaneous
mortality rate computed from (a) observed length-frequency data, (b)
observed length-frequency data after eliminating all length-frequencies
with a capture probability of < 0.95, (c) observed length-frequency data
divided by estimates of capture probability, (d) corrected age-frequency
data after eliminating all age-frequencies with a capture probability of
<0.95, and (e) observed length-frequency data using IMMAGE.
36
30-
26
20
16
10
6
0 ' 1 1 1 1 1 u N 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 n 11 1 ^lW
36
30
28
20
IS
10
6'
0 " I I I I I I I I I I I I I I I I I I F I I I I I I ifl
ImTlttHiii
9
m
Si
o
a
2
35
30
26
20
16
10
6
0-wmM«wW4V
44m
IUM^T
36
30
25-
20
16
10
6
0 1 1 1 1 1 1 1 1 1 1 1 1 n M 1 1 1 ij 1 1 1 iT^MMHt
36
so
26-
20-
16-
10
6
fti 1 1 I I I 1 1 I I I 1 1 I I I I I 1 1 I r
0 1 1 1 I I I I 1 1 I I 1 1 M M I I I I I I I
-0.4 -0.2 0.0 0.2 0.4 0.8 0.8
MortaJity rate
374
Fishery Bulletin 90(2). 1992
Dividing the observed length-frequency distribution
by estimates of capture probability was as ineffective
because the mortality estimates (-0.035 + 0.030) had
a highly significant negative bias of 83% (Fig. 4c). The
negative bias and the strong negative skew in the fre-
quency distribution are due to the infrequent genera-
tion of large larvae. When such larvae occur in a
sample, their relative abundance is greatly magnified
by their extremely small capture probabilities. The
overestimation of large length-classes results in a cor-
responding overestimation of old age-classes. Because
the overestimated age-frequencies are at the extreme
of the age range, they exert considerable influence on
the slope of the mortality regression and thereby result
in the underestimation of M.
Elimination of the biased age-frequencies nearly
eliminates this problem and results in mortality esti-
mates (0.192 ± 0.002) with a significant but small nega-
tive bias of 4% and a considerable reduction in variance
(Fig. 4d). The apparently greater effectiveness of age
elimination compared with length elimination is a func-
tion of the variance in length-at-age within the larval
fish population. For example, when length elimination
is applied to the simulated population, essentially all
of the undersampled lengths are removed. However,
this creates a new bias in the age distribution, because
some age-classes experience greater elimination than
others (Fig. 2a). Clearly, if no variation exists in length-
at-age, length elimination will be identical to age
elimination. To our knowledge, the use of age elimina-
tion has not been reported in the literature.
Mortality estimates produced by IMMAGE (0.201
+ 0.002) are unbiased and, along with age elimination,
have the smallest variance (Fig. 4e). This superior per-
formance is achieved because, unlike length or age
elimination, IMMAGE uses all of the sampled length-
frequency data and, unlike correction, uses capture
probability multiplicatively and therefore avoids magni-
fying the sampling error.
Practical application of IMMAGE
The application of IMMAGE using the Monte Carlo
simulation has been chosen because of its simplicity.
Although linear growth, constant mortality, and mono-
tonically increasing size selection may not always be
suitable for a particular application, the IMMAGE
procedure is extremely adaptable in the way growth,
mortality, and size selection can be specified. For
example, growth could be specified as an exponential
or a Laird-Gompertz function, and size selection could
be specified as a double logistic function (Somerton and
Kobayashi 1989) describing extrusion and avoidance
simultaneously. Mortality could be specified as either
a stage-specific function, where the mortality rates of
yolksac and feeding larvae differ, or an inverse func-
tion of age (Lo 1986). More importantly perhaps, mor-
tality could also be specified as a function of length.
Although mortality rates of larvae likely decline with
length for many species (Pepin 1991), length-dependent
mortality rates are difficult to estimate because such
mortality induces a progressive bias in mean length-
at-age (if the largest larvae in an age-class survive
better than the smallest, the apparent growth rate is
positively biased; Methot and Kramer 1979). This
problem can be circumvented with IMMAGE by
estimating the size-selected length-frequency distribu-
tion using a length-based population model (Somerton
and Kobayashi 1990) which mimics the growth and sur-
vival of individual members of an age-class over time.
Using such an approach, the likelihood of size-depen-
dent mortality could be tested against constant or age-
dependent mortality based on goodness-of-fit to the
observed length-frequencies.
Several variations on the application of IMMAGE
described herein may be more appropriate in other
cases. First, the estimation of growth and mortality
parameters could be accomplished simultaneously
rather than separately by allowing the mortality pro-
cess model to include growth parameters as variables
rather than as known quantities. Second, the objective
function used for parameter estimation could be speci-
fied as a likelihood function rather than a sum-of-
squares function. This would be especially appropriate
in cases where the errors about the observed length-
frequencies are not normally distributed. Third, prior
estimates of some parameters could be included in the
growth and mortality process models rather than
estimating all parameters directly from the three
samples (i.e., ageing, length-frequency, and selection
samples).
We believe the best way of estimating parameter
variances for an IMMAGE application is to use a
sample reuse technique known as boot-strapping
(Efron and Tibshirani 1991), because all sources of
sampling variability can be included. Boot-strapping
IMMAGE, however, is computationally intensive and
potentially time-consuming. To facilitate variance
estimation on slow computers, we have therefore in-
cluded in the IMMAGE program an approximate tech-
nique that is based on the inverse of the information
matrix (Ratkowsky 1983).
When obtaining larval fish samples free of selection
bias is difficult, IMMAGE can still obtain unbiased
estimates of growth and mortality parameters. Not
only does IMMAGE provide estimates that are more
accurate and precise than other approaches, its greater
flexibility in form allows estimation of length-depen-
dent mortality rates that are perhaps biologically more
realistic than the constant rates now estimated.
Somerton and Kobayashi: Inverse method for mortality and growth estimation
375
Acknowledgments
We thank George Boehlert, Ed DeMartini, Chyan Huei
Lo, Sue Picquelle, and Tim Ragen for reviewing the
manuscript and offering helpful suggestions. In addi-
tion, we thank Rick Methot who gave us the original
idea of the inverse estimation procedure.
Citations
Bailey, K.M.
1982 The early life history of Pacific hake Merlucdus produc-
tus. Fish Bull, U.S. 80:.589-598.
Efron, B., and R. Tibshirani
1991 Statistical analysis in the computer age. Science (Wash.
DC) 253:390-395.
Essig, R.J., and C.F. Cole
1986 Methods of estimating larval fish mortahty from daily
increments in otoliths. Trans. Am. Fish. Soc. 115:34-40.
Houde, E.D.
1977 Abundance and potential yield of the round herring,
Et/nim-eus teres, and aspects of its early life history in the
eastern Gulf of Mexico. Fish. Bull., U.S. 75:61-89.
Laroche, J.L., S.L. Richardson, and A. A. Rosenberg
1982 Age and growth of a pleuronectid, Parophrys vetulus,
during the pelagic larval period in Oregon coastal waters.
Fish. Bull., U.S. 80:93-104.
Leak, J.C, and E.D. Houde
1987 Cohort growth and survival of bay anchovy A rec/ioa mit-
chilli larvae in Biscayne Bay, Florida. Mar. Ecol. Prog. Ser.
37:109-122.
Lenarz, W.
1972 Mesh retention of larvae of Sardinops eaerulea and
Engraulis mordax by plankton nets. Fish. Bull., U.S. 70:
789-798.
Lo, N.
1986 Modeling life-stage-specific instantaneous mortality rates,
an application to northern anchovy Engraulis mordax eggs and
larvae. Fish. Bull, U.S. 84:395-407.
Lough, R.G., M. Pennington, G.R. Bolz, and A.A. Rosenberg
1982 Age and growth of larval Atlantic herring, Clupea
harengiis, in the Gulf of Maine-Georges Bank region based on
otolith growth increments. Fish. Bull., U.S. 80:187-199,
Mesnil, B., and J.G. Shepherd
1990 A hybrid age- and length-structured model for assess-
ing regulatory measures in multiple-species, multiple-fleet
fisheries. J. Cons. Cons. Int. Explor. Mer 47:115-132.
Methot, R.D.
1989 Synthetic estimates of historical abundance and mortality
for northern anchovy. In Edwards, E.F., and B. Megrey
(eds.), Mathematical analysis offish stock djTiamics, p. 66-82,
Am, Fish, Soc. Symp. 6, Bethesda.
1990 Synthesis model: An adaptable framework for analysis
of diverse stock assessment data. In Low, L.L. (ed.). Pro-
ceedings, Symposium on application of stock assessment tech-
niques to gadids, p. 259-277. Int. North Pac. Fish. Comm.
Bull. 50.
Methot, R.D., and D. Kramer
1979 Growth of northern anchovy larvae, Engraulis mordax,
in the sea. Fish. Bull., U.S. 77:413-423,
Morse, W.W.
1989 Catchability, growth and mortality of larval fish. Fish.
Bull., U.S. 87:417-446.
Murphy, G.L, and R.L Clutter
1972 Sampling anchovy larvae with a plankton purse seine.
Fish. Bull., U.S. 70(3S):789-798.
Pepin, P.
1991 Effect of temperature and size on development mortal-
ity and survival of marine fish. Can. J. Fish. Aquat. Sci. 48:
503-518.
Ratkowsky, D.A.
1983 Nonlinear regression modeling. Marcel Dekker, NY,
276 p.
Ricker, W.E.
1975 Computation and interpretation of biological statistics
of fish populations. Bull. Fish. Res. Board Can. 191, 382 p.
Somerton, D.A., and D.R. Kobayashi
1989 A method for correcting catches of fish larvae for the
size selection of plankton nets. Fish. Bull., U.S. 87:447-455.
1990 Robustness of the Wetherall length-based method to
population disequilibrium. Fish. Bull., U.S. 89:307-314.
Thorrold. S.
1988 Estimating some early life history parameters in a
tropical clupeid, Herklotsichthys castelnaui, from daily growth
increments in otoliths. Fish. Bull., U.S. 87:73-83.
Ware, D.M., and T.C. Lambert
1985 Early life history of Atlantic mackerel (Scomber scom-
brus) in the southern Gulf of St. Lawrence. Can. J. Fish.
Aquat. Sci. 42:577-592.
Warlen, S.M.
1988 Age and growth of larval gulf menhaden, Brevoortia
patronsis. in the northern Gulf of Me.xico. Fish. Bull., U.S.
86:77-90.
Yoklavich, M.M., and K.M. Bailey
1990 Hatching period, growth and survival of young walleye
pollock Theragra chaleogramma as determined from otolith
analysis. Mar, Ecol. Prog. Ser. 64:12-23.
Zweifel, J., and P. Smith
1981 Estimates of abundance and mortality of larval anchovies
(1951-75): Application of a new method. Rapp. P.-V. R^un.
Cons. Int. Explor. Mer 178:248-259.
Abstract. -The seasonal distri-
bution and relative abundance of
river herring ^iosa pseudoharengiis
and A. aestivalis off Nova Scotia is
examined using Canadian Depart-
ment of Fisheries and Oceans data
from bottom-trawl surveys (1970-89)
and the International Observer Pro-
gram (1980-89). River herring oc-
curred throughout the year in re-
gions characterized by strong tidal
mixing and upwelling in the Bay of
Fundy and off southwestern Nova
Scotia. During spring, river herring
were most abundant in the warmer,
deeper waters of the central Scotian
Shelf, particularly between Emerald
and Western Banks, and in areas of
warm slope water intrusion along
the Scotian Slope, the western and
southern edges of Georges Bank,
and in the eastern Gulf of Maine-Bay
of Fundy. Most catches occurred
at bottom temperatures of 7-1 1°C
offshore at mid-depths in spring
(101-183m), in shallower nearshore
waters in summer (46-82 m) and
in deeper offshore waters in fall
(1 19-192 m). Diel variation in catch
occurred during summer and fall
but not during spring, with largest
catches during daylight. Seasonal
distribution patterns of small (< 19
cmFL) and large (>19cmFL) river
herring overlapped geographically.
Small fish preferred shallow regions
(<93 m) during spring and fall, while
large fish occurred in deeper areas
(>93 m) during all seasons. The
temporal and spatial distribution of
river herring off the coast of Nova
Scotia is hkely influenced by zoo-
plankton concentrations and occur-
rence of bottom temperatures >5°C.
The pattern of seasonal movement is
generally inshore and northward
during spring, and offshore and
southward in the fall.
Seasonal distribution of river
Inerring Alosa pseudoharengus
and A. aestivalis off the Atlantic
coast of IMova Scotia
Heath H. Stone
Brian M. Jessop
Department of Fisheries and Oceans, Biological Science Branch
P.O. Box 550, Halifax, Nova Scotia B3J 2S7, Canada
Manuscript accepted 6 March 1992.
Fishery Bulletin, U.S. 90:376-389 (1992).
River herring (a collective term for
the alewife Alosa pseudoharengus
and the blueback herring A. aesti-
valis) are anadromous clupeids native
to the Atlantic coast of North Ameri-
ca. These closely-related species are
remarkably similar in morphology
and life history and differ only slight-
ly in terms of meristics, morphomet-
ries, growth parameters and time of
spawning (Bigelow and Schroeder
1953, Leim and Scott 1966, Messieh
1977, Loesch 1987). Alewives and
blueback herring are sympatric; ale-
wife range from Newrfoundland to
North Carolina, and blueback her-
ring from the Gulf of St. Lawrence
to Florida (Bigelow and Schroeder
1953). They are fished commercially
in the Maritime provinces of Canada
and Atlantic coastal United States
during their spring spawning migra-
tions and are often marketed togeth-
er as alewife, gaspereau, or river her-
ring, depending on local convention.
Both species often co-occur in
freshwater (Loesch et al. 1982, Jes-
sop and Anderson 1989), estuarine
(Stone and Daborn 1987), and marine
(Neves 1981) habitats. While the
freshwater life histories of alewives
and blueback herring have been well
documented (reviewed by Loesch
1987), less information is available
on their distribution and movements
at sea, particularly in Canadian coast-
al areas. Off the Atlantic coast of
Nova Scotia, bottom trawls have
caught river herring on the western
Scotian Shelf, with alewives collected
during spring and fall (Vladykov
1936, Neves 1981, Vinogradov 1984)
and blueback herring in spring and
summer (Netzel and Stanek 1966,
Neves 1981). In the inner Bay of
Fundy, river herring aggregate dur-
ing summer in the turbid estuarine
waters of Minas Basin (Rulifson
1984, Stone and Daborn 1987) and
Cumberland Basin (Dadswell et al.
1984). In the outer Bay of Fundy,
alewives and blueback herring have
been captured during summer and
autumn in herring weirs, purse
seines (Jessop 1986), and during
bottom-trawl surveys (Neves 1981).
Neves (1981) examined the marine
distribution and seasonal movements
of river herring along the continen-
tal shelf from Cape Hatteras, North
Carolina, to southwestern Nova
Scotia. In this paper, we describe the
seasonal marine distribution and
relative abundance of river herring
off Nova Scotia based on 20 years of
combined alewife and blueback
herring catch data from Canadian
research vessel surveys, thereby ex-
tending northward the analysis by
Neves (1981) of their distribution off
the Atlantic coast of the United
States. Offshore distributions are in-
terpreted in relation to water depth
and temperature within the survey
area.
Although marine exploitation of
alewives and blueback herring in
the Maritime provinces by incidental
376
Stone and Jessop. Seasonal distribution of Alosa pseudoharengus and A aestivalis
377
58 °N
Figure 1
Topographical map of Cana-
dian groundfish survey areas
(1970-89) and geographic fea-
tures mentioned in text. Spring
(1979-85), summer (1970-89),
and fall (1978-84) survey cov-
erage of the Scotian Shelf and
Bay of Fundy (- - -) is shaded
light-gray. Spring (1986-89)
survey coverage of the eastern
shelf and Georges Bank (...)
is shaded dark-gray. Offshore
banks are delineated by the
100 m depth contour; basins
and the outer edge of the con-
tinental shelf are delineated by
the 200 m depth contour.
catches in offshore bottom-trawl fisheries from 1984
to 1989 has averaged l,400t, which is <17% of the
freshwater exploitation (Statistics Branch, Dep. Fish.
Oceans, P.O. Box 550, Halifax, Nova Scotia B3J 2S7),
only a limited amount of information on marine dis-
tribution is available from commercial fishing opera-
tions. More comprehensive information comes from the
bycatch of the annual bottom trawl surveys of the Sco-
tian Shelf-Bay of Fundy region conducted by the Cana-
dian Department of Fisheries and Oceans (DFO) to
monitor temporal changes in the abundance of commer-
cially exploited groundfish species and associated
environmental conditions (Halliday and Koeller 1981).
Additional information on river herring distribution
was obtained from 10 years of bycatch data from
foreign and domestic fishing operations compiled by
the DFO International Observer Program.
Description of survey srea
The survey area, which includes the Scotian Shelf,
eastern Gulf of Maine, Bay of Fundy, and, recently,
Georges Bank, is topographically and hydrographical-
ly complex (Fig. 1). The oceanographic and biological
characteristics of this area influence the distribution
of other species (Sinclair 1988) and are assumed to do
so for river herring.
The Scotian Shelf is characterized by deep central
basins (>200m) and shallow offshore banks (<50m).
Over the deeper parts of the shelf, three distinct ver-
tical layers occur: a surface mixed layer with seasonal
temperature changes, an intermediate layer with tem-
peratures <5°C regardless of season, and a warm
bottom layer derived from cross-shelf intrusions of
slope water (Hatchey 1942, Smith et al. 1978). The
coldest water at any depth occurs in the northeastern
Shelf and the warmest in the bottom waters of the cen-
tral Shelf (also termed the Scotian Gulf) where intru-
sion of warm slope water to adjacent Emerald and
LaHave Basins is a persistent feature (McLellan 1954).
Nutrient-rich upwellings along the shelf-slope interface
sustain high levels of biological productivity (Fournier
et al. 1977). Consequently, pelagic fish production on
the shelf-slope is much higher than on the shelf (Mills
and Fournier 1979).
The Bay of Fundy and eastern Gulf of Maine regions
are vertically well mixed due to the action of strong
tidal currents (Greenburg 1984). Strong, persistent
summertime fronts in sea-surface temperature occur
near the mouth of the Bay and off southwestern Nova
Scotia (Loder and Greenburg 1985). The upwelling of
nutrient-rich deep water from the Gulf of Maine and
Scotian Shelf supports high biological productivity
during spring and summer (Fournier et al. 1984). Sec-
ondary production in the outer Bay of Fundy occurs
378
Fishery Bulletin 90(2), 1992
Table 1
Summary of river herring (Alosa spp.) catch (in numbers) and effort for spring, summer, and fall groundfish
surveys conducted off Nova Scotia by the Canadian Department of Fisheries and Oceans, 1970-89. Paren-
theses enclose percentages of total numbers of sets containing catch. SD= sample standard deviation.
Season
Year
No. of
surveys
No. of
sets
Catch-per-set
X
SD
Total no.
of fish
Sets with
catch
Spring (Feb-Apr)
Summer (June-July)
Fall (Sept-Dec)
Total
1979-89
1970-89
1978-84
1970-89
11
20
7
38
1231
1892
982
4105
6.6
2.0
1.6
3.3
73.85
18.05
9.47
42.53
8117
3703
1537
13357
268 (21.8)
214 (11.3)
120 (12.3)
602 (14.7)
predominantly within the water column and provides
forage to pelagic fish species (Emerson et al. 1986).
On Georges Bank, frontal regions generated by tidal
mixing along the northern and southern edges are
highly productive due to advection of nutrients from
deeper waters on both sides of the bank (Cohen et al.
1982). Intrusions of warm, saline slope water occur into
the southern Gulf of Maine through the Great South
Channel (Mountain et al. 1989) and the Northeast
Channel (Ramp and Wright 1979).
Materials and methods
Survey design and sampling
River herring catch and length-frequency data were
obtained from 38 bottom-trawl surveys conducted
between 1970 and 1989 (Table 1). The main study area
(Scotian Shelf, eastern Gulf of Maine, and Bay of
Fundy; Fig. 1) was surveyed at least once annually
during this period. A single summer (June-July) sur-
vey was conducted annually between 1970 and 1977;
spring (February-April), summer, and fall (Septem-
ber-December) surveys were made from 1978 to 1984;
and spring and summer surveys from 1985 to 1989.
Changing research requirements shifted spring survey
effort to Georges Bank and the eastern Scotian Shelf
in 1986, thereby excluding the western shelf region and
the Bay of Fundy (Fig. 1). Recent (1987-89) summer
surveys also included some coverage of northeastern
Georges Bank.
All surveys used a stratified random design with
trawl stations allocated to depth strata in proportion
to stratum area and randomly positioned within strata.
Stratum depth ranges were 0-92 m (0-50 fm), 93- 183 m
(51-lOOfm), and 184-366 m (101-200 fm). Before 1981,
summer cruises used a No. 36 Yankee bottom trawl
with a 10mm stretched-mesh liner in the cod end; all
other cruises used a Western IIA trawl with a 10- or
20 mm stretched-mesh liner. Tows at each sampling
station were for 30 minutes. Halliday and Koeller
(1981) and Smith (1988) give further details of Cana-
dian groundfish survey methods.
Total number and weight (kg) of river herring in the
catch were recorded for each tow as was bottom-water
temperature (°C), tow deployment time (Atlantic Stan-
dard Time), latitude, longitude, and bottom depth (m).
Fork lengths (FL) were measui-ed (to nearest cm) for
all fish in catches of < 250 fish; otherwise, catches were
subsampled (no fixed procedure) for length. Ale wife
and blueback herring were not differentiated and sex
was not determined.
Catches of river herring in Canadian surveys prob-
ably consist mainly of alewives since both species tend
to be vertically separated by depth. Blueback herring
frequent upper levels; alewives frequent mid-depths
(Neves 1981) where they are more available to capture
in bottom-trawl gear. All of the river herring catch
from 1990 spring and summer groundfish surveys
(n 1048) were confirmed by examination to be ale-
wives. However, because river herring were not iden-
tified to species during the 1970-89 surveys, upon
which our analysis is based, we cannot exclude the
possibility that the catch included a small proportion
of blueback herring captured incidentally when setting
or hauling the gear.
Data analysis
Catch-per-set was standardized to a tow length of
1.75nm with no adjustment for differences between
gear types. Only cruises with catches of one or more
fish were analyzed; 17 cruises on the eastern Scotian
Shelf with no catches were omitted. River herring
>33cmFL were excluded from the data set (<0.5% of
all fish measured), since they exceed the maximum fork
lengths recorded for alewife and blueback herring
(Loesch 1987) and are believed to be incorrectly iden-
tified American shad Alosa sapidissima.
Stone and Jessop: Seasonal distribution of Alos3 pseudoharengus and A. aestivalis
379
Seasonal distributions of relative abundance were
obtained from plots of average catch-per-set data (in-
cluding zero catches) aggregated by 20-minute rec-
tangles of latitude and longitude. Seasonal, rather than
monthly, distributions gave a more complete picture
of offshore distribution patterns. The locations of 2, 5,
and 10°C isotherms, generated from plots of the mean
bottom-water temperature in 20-minute rectangles,
were superimposed on seasonal distribution plots. Dif-
ferences in the seasonal distribution of two size-groups
of river herring (i.e., <19cm and >19cmFL) were ex-
amined by plotting the capture locations of each group.
The size-at-first-spawning of Saint John River blueback
herring (Jessop et al. 1982), which mature at a smaller
size than alewife, was used as the separation criterion.
Approximate randomization tests, with 1000 per-
mutations (Edgington 1987) were used to examine the
following relationships: (a) effects of season (spring,
summer, fall) and depth (<93m, 93-183m, >183m) on
mean catch-per-set (including sets with no catches), (b)
diel variability in seasonal catch-per-set (day and night,
based on gear deployment times in relation to monthly
morning and evening civil twilight times for the ap-
propriate geographic location), and (c) mean fork length
by season and depth. All comparisons used catch or
fork-length data from a 6-year time-series (1979-84)
in which annual sampling occurred during spring, sum-
mer, and fall. Each dependent comparison (i.e., depth
effect within season, season effect within depth), used
a Bonferroni significance level (a 0.05, divided by the
number of dependent comparisons) (Day and Quinn
1989).
Randomization procedures were used for the analysis
because statistically significant heteroscedasticity in
the variances of (a) transformed (In X -i- 1) catch per set
data by season and depth (Cochran's C = 0.372, P<
0.0001, df 289, 9), (b) transformed On X -h 1) diel catch-
per-set by season (Cochran's C = 0.365, P<0.0001, df
359, 6), and (c) fork length by season and depth
(Cochran's C = 0.293, P<0.0001, df 637, 9) violates an
assumption of parametric statistics. This violation is
compounded by unequal sample sizes.
Bycatch data
Set locations and fork lengths of river herring bycatch
in foreign and domestic commercial groundfish opera-
tions (1980-89) obtained from the DFO International
Observer Program database were compared with re-
search survey data. Catch locations were plotted for
spring (February-May) and summer (June-August),
when fishing effort and bycatch were highest. Fork
length distributions were truncated at 33 cm due to
assumed misclassification (4% of fish measured ex-
ceeded this length). Annual landings of alewife in
metric tons (t) for the Scotian Shelf-Bay of Fundy
region (4VWX) from 1970 to 1989 were obtained from
Northwest Atlantic Fisheries Organization (NAFO)
Statistical Bulletins and correlated (Spearman rank)
with catch indices from research vessel surveys. The
analysis was conducted to determine if survey indices
(i.e., mean number and mean weight per set ■ season "^
• year'i) were consistent with bycatch landings as an
indicator of relative abundance.
Results
Seasonal distribution and abundance
A total of 13,357 river herring were captured in 602
(15%) of 4105 bottom-trawl sets conducted between
1970 and 1989 (Table 1). Spring survey catches were
the highest and most variable, followed by summer.
The proportion of sets with river herring from spring
surveys was nearly double that from the other two
seasons. A maximum catch of 2292 river herring oc-
curred on 17 March 1980 in the Scotian Gulf region,
south of Emerald Basin.
During spring surveys, river herring dominated in
three regions: the Scotian Gulf, southern Gulf of
Maine, and off southwestern Nova Scotia from the
Northeast Channel north to the central Bay of Fundy
(Fig. 2). Catches also occurred along the southern edge
of Georges Bank and in the canyon between Ban-
quereau and Sable Island Banks. Relative abundance
was highest in the Scotian Gulf between Emerald and
Western Banks, and on the southern slope of Georges
Bank. Most catches of river herring occurred where
bottom temperatures exceeded 5°C (Fig. 2), although
in the Bay of Fundy, captures occurred at lower
temperatures.
Summer distributions of river herring were less ex-
tensive than in spring and were limited mainly to the
eastern Gulf of Maine (off southwestern Nova Scotia)
and the Bay of Fundy, with a few occurrences near-
shore in the central Shelf region (Fig. 2). Catches were
highest along the northern shore of the Bay of Fundy,
with very few fish captured in the Scotian Gulf and on
the eastern Scotian Shelf. Bottom temperatures ex-
ceeded 5°C at all capture locations.
Fall distributions of river herring were more exten-
sive than in summer (Fig. 2). Moderate to large catches
were obtained from southwestern Nova Scotia to the
Bay of Fundy, the central Scotian Shelf, and Sydney
Bight. As in the case of spring and summer surveys,
very few fish were captured on the eastern half of the
Scotian Shelf. All catches occurred at bottom tem-
peratures exceeding 5°C.
Bycatch of river herring from foreign fishing fleets
(1980-89) occurred mainly during spring in a narrow
380
Fishery Bulletin 90(2), 1992
43-1
41
47
45
43
V
Summer
41
4
Fall
:,Ai
..,.^•3^''^'
70
68
66
1 T"
64
1—^
Abundance
#100-1000
• 10-100
. 1-10
0.1-1
62
60
58°N
Figure 2
Distribution and relative abundance (x catch per 20-min
square) of river herring from spring (1979-89), summer
(1970-89), and fall (1978-84) groundfish surveys conducted
off the Atlantic coast of Nova Scotia. Area sampled is in-
dicated by shading, and mean catch (n/tow) for each 20-min
square is represented by a solid circle, scaled to catch level.
Isotherms for 2, 5, and 10°C of bottom temperature are also
indicated.
band along the edge of the Scotian Shelf from Emerald
Bank east to Sable Island Bank (Fig. 3). This distribu-
tion reflects the location of the Soviet and Cuban silver
47
45
43
41
47
45
43
41 -
°W
.rriTi'
Spring
S €^t
t--- ^-^ U*, " ' *• *'-*^w
^'
V
Summer
^^^
'X>^'
70
T — I r
68
66
T 1 1 r
62
64
60
58''N
Figure 3
Spring and summer set locations of river herring bycatch
from foreign and domestic fishing operations on the Sco-
tian Shelf recorded by Canadian International Observers,
1980-89.
hake fishery (M. Showell, Bedford Inst. Oceanogr.,
Dartmouth, N.S. B2Y 4A2, pers. commun., Jan. 1991)
and indicates a more widespread occurrence of river
herring along the shelf break than was apparent from
bottom-trawl surveys. Bycatch from domestic fishing
operations during spring occurred on the edge of
Western and Emerald Banks and corresponds with
catch locations from spring groundfish surveys.
Summer bycatch locations were similar to spring but
extended further south and west along the shelf edge,
with clusters of catches on the northern edge of
Georges Bank, off southwestern Nova Scotia, and in
the mouth of the Bay of Fundy (Fig. 3). River herring
distribution along the shelf break, as indicated from
spring and summer bycatches in the silver hake fishery,
reflects the spatial distribution of that fishery. The
number of summer observations from a reduced
foreign fishing effort (n 86) was much less than spring
(n 460). In both seasons, most bycatches occurred in
Stone and Jessop: Seasonal distribution of Alosa pseudoharengus and A. aestivalis
381
*- 1
Q.
O
c5
o
c
10 n
8
6
4
2
0
Spring
70
60
50
40 -
30
20
10-
— 1 0
12
27
302
Summer
12
Fall
10
f- ' I
-r-T
302
Bottom temperature (C)
Depth (m)
Figure 4
Mean catch-per-set of river herring
within 1°C intervals of bottom tem-
perature and 18.2m (lOfm) depth in-
tervals from spring, summer, and fall
groundfish surveys, 1979-84, con-
ducted off the Atlantic coast of Nova
Scotia. Vertical bars represent ± 1
SE. Sample sizes are >10 sets for
each temperature and depth interval.
regions where bottom temperatures from groundfish
surveys exceeded 5°C.
Temperature, depth, and
diel effects on catch
The relationship between catch, bottom temperature,
and depth was examined by plotting mean catch-per-
set by season for intervals of 1°C and 18.3m (lOfm)
(Fig. 4). Despite much variability in the data, most
catches occurred within the 7-1 1°C range regardless
of season, with maximum catches within 9-1 1°C from
spring through fall. An exception is the moderate
catches at bottom temperatures <2°C during spring
surveys in the Bay of Fundy (Figs. 2 and 4). The depth
distribution of catches was more variable; spring
catches occurred mainly at intermediate depths of
101-174m, summer catches in shallow areas (46-82m),
and fall catches at mixed depths (i.e., 46m, 119-192m).
Mean catch-per-set of river herring for annual
(1979-84) spring, summer, and fall surveys varied
significantly by season and depth (Table 2, Fig. 5).
Season effects within depth strata were significant for
depths of <93m and 93-183m, but not for depths
>183m where all catches were low. Catches were
highest within the <93m strata during summer and
within the 93-183 m depth strata during spring. Depth-
within-season interaction was significant only during
382
Fishery Bulletin 90(2). 1992
Table 2
Summary of results from approximate
randomization tests used to
examine
the following relation- 1
ships: (a) effects of season and depth on mean catch-per
set, (b) diel variability in seasonal catch-per-
set, and (c), mean fork length (FL) by season and depth.
All comparisons used catch and fork-length
data from Canadian groundfish
surveys
conducted 1979
-84. Interaction effects with an asterisk are
significant at the adjusted Bonferroni
significance level of P<0.017.
Catch by season
Catch by
season
FLby
season
Source
and depth
and time
and
depth
n
P
n
P
n
P
Main effects
Season
2157
0.002
2157
0.006
5739
0.001
Depth
2157
0.010
—
—
5739
0.001
Time
-
-
2157
0.80
-
-
Interactions
Season effect within depth
<93m
800
0.008*
—
—
1564
0.001*
93- 183 m
836
0.001*
—
—
3769
0.001*
>183m
521
0.027
-
-
406
0.001*
Depth effect within season
spring
788
0.002*
—
—
3807
0.001*
summer
565
0.028
—
—
1114
0.001*
fall
804
0.060
-
-
818
0.001*
Season effect within time
day
—
—
1231
0.228
—
—
night
-
-
926
0.033
-
-
Time effect within season
spring
—
—
788
0.043
—
—
summer
—
—
565
0.001*
—
—
fall
—
—
804
0.017*
—
—
spring when the range in catches was greatest;
summer and fall catches were similar at all depths
(Table 2).
Mean catch-per-set of river herring varied signif-
icantly by season but not by time-period (Table 2,
Fig. 6). Interactions of season effect within time-
period were nonsignificant for day and night; interac-
tions of time effect within season were significant for
summer and fall but not for spring. For summer and
fall surveys, catches from sets conducted during
daylight were significantly higher than those conducted
at night. While night catches from spring surveys
tended to be higher than day catches (but not sig-
nificantly so), they were extremely variable (x 13.4,
sample SD 131.32). The proportion of sets with catch
Figure 5
Mean catch-per-set of river herring by depth strata from spring,
summer, and fall groundfish surveys, 1979-84, conducted off the
Atlantic coast of Nova Scotia. Vertical bars represent -t- 1 SE.
Number of tows is indicated above error bars.
during day and night showed a seasonal pattern similar
to that for mean catch, i.e., similar in spring (20% vs.
24%) and higher during the day in summer (15% vs.
5%) and fall (16% vs. 9%).
30 n
300
*- 25-
n spring
S5 20-
Q.
§ 15-
m
o
c 10-
(0
<D
^ 5-
205
1
D summer
■ fall
299
296
'j^ Fo^:-
0 ■*
<93
93-183 >183
Depth range (m)
Stone and Jessop: Seasonal distribution of Alosa pseudoharengus and A. aestivalis
383
spring
summer
Season
Figure 6
Mean catch-per-set of river herring in day and night collec-
tions from spring, summer, and fall groundfish surveys,
1979-84, conducted off the Atlantic coeist of Nova Scotia. Ver-
ticaJ bars represent + 1 SE. Number of tows is indicated above
error bars.
Survey catch indices and NAFO landings
Seasonal catch indices (annual mean number and
weight of river herring per set) from research vessel
surveys were uncorrelated with NAFO bycatch data
for the Scotian Shelf-Bay of Fundy area (4VWX)
(Table 3). However, landings from the central and
eastern shelf (4 VW) were significantly correlated with
spring (1979-89) survey indices. The relationship be-
tween mean weight index and 4VW NAFO landings
(Fig. 7) gave the highest Spearman rank correlation
coefficient (r^ 0.74).
Seasonal length-composition
and distribution
River herring from bottom-trawl collections
measured 5-33cmFL (x 23.7cm; Table 4).
Mean fork length in spring catches was less
than in summer or fall. Fork lengths of river
herring from foreign and domestic fisheries
were 22-33cm (i 29.6cm) and were of
comparable size in spring and summer sam-
ple collections. The larger size of bycatch
fish compared with those from bottom-trawl
surveys is probably due to the larger cod end
mesh size (6 cm stretched) of commercial
gear permitting escapement of smaller fish.
In all seasons, large river herring (>19
cmFL) were more abundant than small
«19cm) fish (Fig. 8). Polymodal length-
frequency distributions indicated the
Table 3
Spearman rank correlation coefficients for river herring (Alosa
spp.) catch indices from spring (1979-89), summer (1970-89),
and fall (1978-84) groundfish surveys vs. NAFO landings from
the Scotian Shelf-Bay of Fundy (4VWX) and the central and
eastern Scotian Shelf (4VW). (n = number of years; * signif-
icant at P<0.05; "significant at P<0.01).
NAFO areas
Spring
(nil)
Summer
(n20)
Fall
(n7)
Mean catch/set ■ yr '
4VWX -0.23 -0.15 -0.07
4VW 0.60* -0.01 -0.18
Mean weight/set • yr"'
4VWX -0.19 -0.17 -0.29
4VW 0.74*' -0.05 -0.09
12001 [-10.0
><
^ 1000-
. • — NAFO landings
i3
0)
E
j\ — .B-- survey index
-7.5 ^
- 800-
/ \
cn
cn
/ \
^-^
■6 600 ■
/ ■' \
•5.0 -g,
CO
^ 400-
/ \\ A
0)
>
/ \\ y/ \ y\
■2.5 §
-J 200-
"> \ ^."^ -•• \/ \
0)
'•- ^^..--"' ''-y . \
^
0-
\.-'-'--)
■0.0
1980 1982 1984 1986 1988
Year
Figure 7
Annual landings in metric tons (t) reported to the Northwest
Atlantic Fisheries Organization for the central and eastern
Scotian Shelf (4VW), and spring survey catch index (kg/set-
yr-')for 1979-89.
Table 4
River herring (Alosa spp.) fork-length statistics by season from ground-
fish surveys conducted by the Canadian Department of Fisheries and Oceans
and from foreign and domestic bycatches recorded by Canadian Interna-
tional Observers. All fork lengths were truncated at 33cm (see 'Materials
and methods' for explanation). SD = sample standard deviation.
Fork length (cm)
Season
Years
SD
range
Groundfish survey data
Spring (Feb-Apr) 1979-89 7543 22.7 5.08 5.0-33.0
Summer (June-July) 1970-89 3167 25.4 3.77 9.0-33.0
Fall (Sept-Dec) 1978-84 1512 24.8 4.92 5.0-33.0
Combined 1970-89 12213 23.7 4.91 5.0-33.0
Bycatch data
Spring (Feb-May) 1980-89 1754 29.4 2.14 20.0-33.0
Summer (June- Aug) 1980-89 249 30.4 1.45 26.0-33.0
Combined 1980-89 2032 29.6 2.13 20.0-33.0
384
Fishery Bulletin 90(2), 1992
12.
9.
spring
n=7,534
1
6.
nf]
n
3.
n n
n
0.
,j,nnr-inmn 1 Ijiji 1
nnn„
5 10 15 20 25 30
>• 12
c
CD
t 6.
8
55 3.
Q.
summer
n.3,167
n
., nnfll
Rfi
0.
IL
5 10 15 20 25 30
12.
fall 1
9.
iipi
6.
n
fl
3.
0^
nHn nl
fln
5 10 15 20 2
5 30
Fork length (cm)
Figure 8
Length-frequency distributions of river herring from spring
(1979-89), summer (1970-89), and fall (1978-84) groundfish
surveys conducted off the Atlantic coast of Nova Scotia.
presence of at least three year-classes in the fall and
four in the spring. Smaller individuals (<19cmFL)
represented a larger proportion of the overall length
composition in spring and fall than in summer.
River herring > 1 9 cm FL were more densely and ex-
tensively distributed in all seasons than smaller fish.
During summer and fall, most river herring <19cm
long were captured within the Bay of Fundy, although
a few occurred in the central Shelf region and in
Sydney Bight (fall only) (Fig. 9). The greatest overlap
in the spatial distribution of both size-groups was dur-
ing spring, especially in offshore regions (i.e., central
Scotian Shelf, southern Gulf of Maine).
The mean fork length of trawl-caught (1979-84) river
herring varied significantly by season of capture and
depth (Table 2, Fig. 10). Interactions of season effect
within depth strata and depth effect within season were
all significant. Smaller fish were caught more frequent-
ly at depths <93m during spring and fall. Larger fish
occurred at all depths during summer (more numerous-
ly at < 93 m depth) and at depths ^93m during spring
and fall. In all three depth strata, river herring from
spring surveys averaged smaller in length than those
captured in summer and fall because of the greater oc-
currence of fish <19cm throughout the survey area
during spring (Fig. 9).
Discussion
Canadian groundfish survey data indicate persistent
patterns in the temporal and spatial distribution of
river herring off the Atlantic coast of Nova Scotia
which appear to be greatly influenced by oceanographic
features. In spring, river herring were most abundant
in the warmer, deeper waters of the Scotian Gulf, par-
ticularly along the edges of Emerald and Western
Banks and within the channel separating them, and in
regions of warm slope-water intrusion along the Sco-
tian Slope, the western and southern edges of Georges
Bank, and the eastern Gulf of Maine. In all seasons,
river herring occurred in the Bay of Fundy and off
southwestern Nova Scotia, regions characterized by
strong tidal mixing and upwelling, but were rarely
present on the eastern Scotian Shelf.
Water temperature evidently influences temporal
and spatial patterns in river herring depth distribution.
In all seasons, most catches occurred within the 7-1 1°C
range but shifted from mid-depths offshore in spring
(101-183 m) to shallower, nearshore waters in summer
(46-82 m), and to deeper offshore waters in fall (119-
192 m). River herring were not present in colder
regions on the eastern and western Scotian Shelf.
Catches of river herring along the U.S. continental
shelf were most frequent at bottom temperatures of
4-7°C and depths <92m (Neves 1981). Spring catches
of river herring at bottom temperatures <5°C in the
Bay of Fundy indicate some flexibility in thermal selec-
tion, as might be expected of a migratory anadromous
fish. American shad, which are closely related to
alewives and blueback herring, can remain for ex-
tended periods in temperatures outside their usual
range (7-13°C) and migrate rapidly between areas with
different temperature regimes (Dadswell et al. 1987).
Seasonal shifts in zooplankton abundance, which are
influenced by local oceanographic features, may also
influence river herring distribution patterns off Nova
Scotia. Both alewives and blueback herring are zoo-
Stone and Jessop. Seasonal distribution of Alosa pseudoharengus and A aestivalis
385
4H
y Fall
i 19 cm
-^■"
..':»:3>'
T I I I I I I I I I I I
70 68 66 64 62 60 58
V / Fall
,y'--^^'' > 1 9 cm
— I — I — I — I — I — I — I — I — r
68 66 64 62 60 58°N
Figure 9
Set locations for catches of river herring < 19 cm FL and >19cmFLfrom spring (1979-89), summer (1970-89),
and fall (1978-84) groundfish surveys conducted off the Atlantic coast of Nova Scotia.
planktivores. Stomach contents of river herring col-
lected on Georges Bank and the Scotian Shelf consisted
mainly of euphausiids and calanoid copepods (Vino-
gradov 1984). These prey items are concentrated dur-
ing winter and spring in deep basins of the Scotian
Shelf (Sameoto and Herman 1989, Herman et al. 1991)
as well as the outer Shelf and Shelf Slope (Sameoto
1982). Calanoid copepods dominate zooplankton pro-
duction on Georges Bank during spring (Sherman et
al. 1987) and are abundant during summer and fall in
the eastern Gulf of Maine-outer Bay of Fundy region
(Emerson et al. 1986) along with large populations of
euphausiids (Kulka et al. 1982). Aggregations of river
herring during spring on the central Shelf, Shelf Slope,
and Georges Bank, during summer in the Bay of
Fundy-eastern Gulf of Maine, and during fall in the
386
Fishery Bulletin 90(2), 1992
30 1
E
o
26-
r
m
c
22-
_o
^
1-
o
18-
c
CO
14-
---•— fall
10
<93 93-183 >183
Depth range (m)
Figure 10
Mean fork length (cm) of river herring by depth strata from
spring, summer, and fall groundfish surveys, 1979-84, con-
ducted off the Atlantic coast of Nova Scotia. Vertical bars
represent 95% CI. The number offish in each category is ad-
jacent to symbols.
outer Bay of Fundy and central Scotian Shelf coincide
with high secondary productivity and an abundance of
prey. River herring may move inshore in summer and
offshore in winter in order to exploit seasonally avail-
able food resources.
The diurnal pattern of vertical migration by river her-
ring accounts for the higher catches and proportion of
sets with catches during daylight hours in summer and
fall. Alewives and blueback herring that are closer to
the bottom during the day are more susceptible to cap-
ture in bottom-trawling gear (Neves 1981, Loesch et
al. 1982). Diel migrations, involving an upward move-
ment at dusk followed by a downward movement at
dawn, occur in landlocked adult alewives (Janssen and
Brandt 1980), as well as anadromous juvenile (Jessop
1990) and adult river herring (Neves 1981). While
spring catches did not follow this pattern, the presence
of a cold (< 5°C) intermediate water mass over warmer,
deeper waters on the Scotian Shelf (Hatchey 1942),
where the largest catches occurred, may have re-
stricted the extent of vertical migration resulting in
more captures at night. Our study indicates that few
river herring are captured in areas where bottom
temperatures are <5°C during spring; therefore, ver-
tical migrations may be confined by a water temper-
ature inversion.
Catch indices by number and weight from spring
groundfish surveys (when relative abundance is great-
est) both appear to be useful indicators of river her-
ring bycatch in foreign and domestic trawl fisheries on
the eastern and central shelf regions (4VW). Poor cor-
relations, obtained when the Bay of Fundy-eastern
Gulf of Maine bycatch landings (NAFO area 4X) were
included in the analysis, were puzzling. Survey data in-
dicate the presence of river herring in this area from
spring through fall. Underreporting of river herring
bycatches in domestic fishing operations (i.e., bottom
trawl, gillnet, purse seine) may explain the poor cor-
relations when landings from 4X and 4VW were com-
bined. Catches from the central and eastern shelf
region (4VW) are largely bycatches from the silver
hake fishery. The frequent presence of DFO observers
aboard these vessels may reduce the incidence of
misreporting. Another possibility is the seasonal im-
migration of American-origin river herring into the Bay
of Fundy-eastern Gulf of Maine region (Rulifson et al.
1987) which would increase the variabOity in the NAFO
landings for area 4X.
The small proportion of fish < 19 cm in summer col-
lections, relative to spring and fall collections, may
reflect their movement outside the survey area into
coastal embayments and estuarine habitats in Maine
and the inner Bay of Fundy. These areas serve as im-
portant summer feeding areas for river herring (Stone
and Daborn 1987) and other anadromous fish species
(Haedrich and Hall 1976). Summer resident river her-
ring generally leave the inner Bay of Fundy in autumn
when secondary production declines (Stone 1985).
Both large (> 19 cm FL) and small (< 19 cm FL) river
herring occurred nearshore from spring through fall,
but were widely distributed offshore during spring (i.e.,
southern Gulf of Maine, Scotian Gulf). Most river her-
ring <19cmFL are sexually immature while those
>19cmFL are generally mature fish which have
spawned previously or are maturing to spawn for the
first time. Smaller, immature river herring evidently
migrate offshore seasonally as do larger, mature fish.
Size-related differences in depth distribution were such
that small river herring occurred in shallow regions
(<93m) during spring and fall, while larger fish oc-
curred in deeper areas (> 93 m) in all seasons. Janssen
and Brandt (1980) reported that the nocturnal depth
distribution of adult landlocked alewife differed by size-
class, with the smaller fish at shallower depths.
Both Canadian and American marine survey data
provide evidence of distinct seasonally and geograph-
ically separate aggregations of river herring. Off the
Atlantic coast of the United States, the Middle Atlan-
tic Bight is an important overwintering area for river
herring, while in summer they concentrate further
north in the Nantucket Shoals and on Georges Bank
(Neves 1981). Aggregations of river herring in spring
and fall on the central Scotian Shelf and in the eastern
Gulf of Maine-Bay of Fundy suggest that these areas
are important overwintering sites off Nova Scotia,
Stone and Jessop: Seasonal distribution of Alosa pseudoharengus and A. aestivalis
387
although the Scotian Shelf has not been surveyed dur-
ing winter months (i.e., November-February). The
main summer concentration extends northward from
southwestern Nova Scotia into the Bay of Fundy.
Other members of the clupeid family also exhibit
spatial and temporal discontinuity in marine distribu-
tion patterns. American shad winter off Florida, the
mid-Atlantic Bight, and in the Scotian Shelf-Bay of
Fundy region (Neves and Depres 1979, Dadswell et al.
1987), while summer concentrations occur off New-
foundland and Labrador (Hare and Murphy 1974), the
inner Gulf of St. Lawrence (Dadswell et al. 1987), the
Gulf of Maine (Neves and Depres 1979), and in the in-
ner Bay of Fundy (Dadswell et al. 1983). Atlantic her-
ring populations in the Gulf of Maine-Scotian Shelf
region have several geographically separate areas for
summer feeding (southwest Nova Scotia, Georges
Bank, Bay of Fundy) and overwintering (Long Island
Sound, Chedabucto Bay) (Sinclair and Isles 1985).
Seasonal movement patterns of river herring infer-
red from American and Canadian survey data involve
a north-south progression and an inshore-offshore
movement similar to that described for American shad
populations along the Atlantic coast of North America
(Neves and Depres 1979, Dadswell et al. 1987). Dur-
ing spring, river herring from the Middle Atlantic Bight
move north as far as the Nantucket Shoals, Georges
Bank, coastal Gulf of Maine and even the inner Bay
of Fundy for the summer, then return south to the mid-
Atlantic coast in winter and early spring (Neves 1981,
Rulifson et al. 1987). The spring aggregation of mature
river herring observed in Canadian survey catches from
the southern Gulf of Maine likely consists of fish which
will move inshore to spawn in rivers along the eastern
seaboard of the United States, although some may
enter Canadian rivers. A large component of the over-
wintering population on the Scotian Shelf moves in-
shore during spring to spawn in rivers along the Atlan-
tic coast of Nova Scotia, the Bay of Fundy, and perhaps
the Gulf of Maine. American shad tagged in rivers in
Nova Scotia (Melvin et al. 1986) and in Quebec (Vlady-
kov 1956) were recaptured on the Scotian Shelf in
winter. The large aggregation of river herring in the
eastern Gulf of Maine, apparent during spring sur-
veys, may include fish in transit from overwintering
areas on the Shelf to spawning rivers along the Bay
of Fundy-Gulf of Maine coast. Considering their pref-
erence for water temperatures above 5°C, the migra-
tion route would occur along the Shelf Slope and into
the eastern Gulf of Maine through the Northeast Chan-
nel. Postspawning river herring probably feed during
summer in the Bay of Fundy-eastern Gulf of Maine
before returning offshore to the central Shelf in the
fall to overwinter. Some may move offshore soon after
spawning, as indicated by the presence of large fish
(x FL 30.4cm) in the summer bycatch from the silver
hake fishery along the shelf slope. Another component
of the Shelf overwintering population may move north
around Cape Breton to the Gulf of St. Lawrence in
spring to spawn in natal rivers, returning in autumn
to the Scotian Shelf to overwinter. This hypothesis is
supported by the fall concentration of river herring in
the Sydney Bight area and movement of an alewdfe ac-
cidently tagged in the Sydney Bight fall fishery for
Atlantic herring, to the Margaree River (southern Gulf
of St. Lawrence) where it was recaptured the follow-
ing spring (Jessop, unpubl. data).
Most river herring overwintering on the Scotian
Shelf probably originate in the Canadian Maritime
Provinces and U.S. Gulf of Maine region. Some river
herring of Canadian Maritime origin evidently migrate
south to overwinter off the Middle Atlantic Bight as
do American shad (Melvin et al. 1986, Dadswell et al.
1987). The tagging of over 50,000 river herring in the
Saint John River, New Brunswick, produced two recap-
tures off North Carolina the following spring and other
recoveries along the intervening coast (Jessop, unpubl.
data). In another study, most recaptures from over
19,000 river herring tagged during summer and fall in
the upper Bay of Fundy occurred in spring fisher-
ies in Nova Scotia rivers, but one occurred off Mas-
sachusetts and several came from North Carolina
(Rulifson et al. 1987). Summer aggregations of river
herring in the Bay of Fundy-eastern Gulf of Maine
may therefore consist of a mixture of stocks from the
entire Atlantic coast, as do similar aggregations of
American shad (Dadswell et al. 1987).
An understanding of the seasonal movements, stock
composition, and exploitation of river herring popula-
tions which overwinter on the Scotian Shelf may help
fishery managers to explain high variability in the
returns of spawning fish regionally and to particular
river systems. Stock composition and migratory routes
remain to be examined.
Acknowledgments
We thank M. Dadswell and K. Frank for critically
reviewing earlier drafts of the manuscript. We also
wish to thank J. McMillan and C. Harvie for their
assistance in data analysis and G. Black for helping
with the offshore distribution plots. The constructive
comments of the anonymous reviewers are also
appreciated.
388
Fishery Bulletin 90(2). 1992
Citations
Bigelow, H.B., and W.C. Schroeder
1953 Fishes of the Gulf of Maine. BuU. U.S. Fish. WUdl. Serv.
74, 577 p.
Cohen, E.B., M.D. Grosslein, and M.P. Sissenwine
1982 Energy budget of Georges Bank. In Mercer, M.C. (ed.),
Multispecies approaches to fisheries management advice, p.
95-107. Can. Spec. Publ. Fish. Aquat. Sci. 59.
Dadswell, M.J., G.D. Melvin, and P.J. Williams
1983 Effect of turbidity on the temporal and spatial utiliza-
tion of the inner Bay of Fundy by American shad {AU>sa
safridissima) (Pisces: Clupeidae) and its relationship to local
fisheries. Can. J. Fish. Aquat, Sci. 40(Suppl. l):322-330.
Dadswell, M.J., R. Bradford, A.H. Leim, D.J. Scarratt,
G.D. Melvin, and R.G. Appy
1984 A review of research on fishes and fisheries in the Bay
of Fundy between 1976 and 1983 with particular reference to
its upper reaches. In Gordon, B.C., and M.J. Dadswell (eds.).
Update on the marine environmental consequences of tidal
power development in the upper reaches of the Bay of Fundy,
p. 163-294. Can. Tech. Rep. Fish. Aquat. Sci. 1256, 686 p.
Dadswell, M.J., G.D. Melvin, P.J. Williams, and D.E. Themelis
1987 Influences of origin, life history and chance on the Atlan-
tic coast migration of American shad. In Dadswell, M.J., et
al. (eds.), Common strategies of anadromous and catadromous
fishes, p. 313-330. Am. Fish. Soc, Bethesda.
Day, R.W., and G.P. Quinn
1989 Comparisons of treatments after an analysis of variance
in ecology. Ecol. Monogr. 59:433-463.
Edgington, E.S.
1987 Randomization tests. Marcel Dekker, NY.
Emerson, C.W., J.C. Roff, and D.J. Wildish
1986 Pelagic-benthic energy coupling at the mouth of the Bay
of Fundy. Ophelia 26:165-180.
Fournier, R.O., J. Marra, R. Bohrer, and M. Van Dei
1977 Plankton dynamics and nutrient enrichment of the
Scotian Shelf. J. Fish. Res. Board. Can. 34:1001-1018.
Fournier, R.O., M. Van Det, N.B. Hargraves, J.S. Wilson
1984 Physical factors controlling summer distribution of
chlorophyll a off southwestern Nova Scotia. Limnol.
Oceanogr. 29:517-526.
Greenburg, D.A.
1984 A review of the physical oceanography of the Bay of
Fundy. In Gordon, D.C., and M.J. Dadswell (eds.). Update
on the marine environmental consequences of tidal power
development in the upper reaches of the Bay of Fundy, p.
9-30. Can. Tech. Rep. Fish. Aquat. Sci. 1256, 686 p.
Haedrich, R.L., and C.A.S. Hall
1976 Fishes and estuaries. Oceanus 19:55-63.
Halliday, R.G., and P.A. Koeller
1981 A history of Canadian groundfish trawling surveys and
data usage in ICNAF divisions 4TVWX. In Doubleday, W.G.,
and D. Rivard (eds.). Bottom trawl surveys. Can. Spec. Publ.
Fish. Aquat. Sci. 58:27-41.
Hare, G.M., and H.P. Murphy
1974 First record of American shad (Alosa sappidissima) from
Labrador waters. J. Fish. Res. Board Can. 31:1536-1537.
Hatchey, H.B.
1942 The waters of the Scotian Shelf. J. Fish. Res. Board
Can. 5:377-397.
Herman, A.W., D.D. Sameoto, C. Shunnian, M.R. Mitchell,
B. Petrie, and N. Cochrane
199 1 Sources of zooplankton on the Nova Scotia Shelf and their
aggregations within deep shelf basins. Continental Shelf Res.
11:211-238.
Janssen, J., and S.W. Brandt
1980 Feeding ecology and vertical migration of adult alewives
{Alosa pseudoharmgus) in Lake Michigan. Can. J. Fish.
Aquat. Sci. 37:177-184.
Jessop, B.M.
1986 Alewife and blueback herring in the Gulf of Maine area.
Can. Atl. Fish. Sci. Adv. Comm. Res. Doc. 67, 7 p.
1990 Diel variation in density, length composition, and feeding
activity of juvenile alewife, Alosa pseudoharmgus Wilson, and
blueback herring, A. aestivalis Mitchill, at 'near-surface depths
in a hydroelectric dam impoundment. J. Fish. Biol. 37:
813-822.
Jessop, B.M., and W.E. Anderson
1989 Effects of heterogeneity in the spatial and temporal pat-
tern of juvenile alewife (Alosa pseudoharengus) and blueback
herring (^4. aestivalis) density on estimation of an index of abun-
dance. Can. J. Fish. Aquat. Sci. 46:1564-1574.
Jessop, B.M, A.H. Vromans, and W.E. Anderson
1982 Life-history data on alewife and blueback herring, Mac-
taquac Dam, 1975-81. Can. Data Rep. Fish. Aquat. Sci. 367,
43 p.
Kulka, D.W., S. Corey, and T.D. Isles
1982 Community structure and biomass of euphausiids in the
Bay of Fundy. Can. J. Fish. Aquat. Sci. 39:326-334.
Leim, A.H., and W.B. Scott
1966 Fishes of the Atlantic Coast of Canada. Bull. Fish. Res.
Board Can. 155, 485 p.
Loder, J.W., and D.A. Greenburg
1985 Predicted positions of tidal fronts in the Gulf of Maine.
Continental Shelf Res. 4:397-414.
Loesch, J.G.
1987 Overview of life history aspects of anadromous alewife
and blueback herring in freshwater habitats. In Dadswell,
M.J., et al. (eds.). Common strategies of anadromous and
catadromous fishes, p. 89-103. Am. Fish, Soc, Bethesda.
Loesch, J.G., W.H. Kriete, and E.J. Foell
1982 Effects of light intensity on the catchability of juvenile
anadromous A /osa species. Trans. Am. Fish. Soc. 111:41-44.
McLellan, H.J.
1954 Bottom temperatures on the Scotian Shelf. J. Fish. Res.
Board Can. 11:404-418.
Melvin, G.D., M.J. Dadswell, and J.D. Martin
1986 Fidelity of American shad, Alosa sapidissima (Clupeidae),
to its river of previous spawning. Can. J. Fish. Aquat. Sci.
43:640-646.
Messieh, S.N.
1977 Population structure and biology of alewives (Alosa
pseudoharengus) and blueback herring (A. aestivalis) in the
Saint John River, New Brunswick. Environ. Biol. Fishes
2:195-210.
Mills, E.L., and R.O. Fournier
1979 Fish production and the marine ecosystems of the Scotian
Shelf, eastern Canada. Mar. Biol. 54:101-108.
Mountain, D.G., M. Pastuszak, and D. A. Busch
1989 Slope water intrusion to the Great South Channel dur-
ing autumn, 1977-85. J. Northwest Atl. Fish. Sci. 9:97-102.
Stone and Jessop: Seasonal distribution of Alosa pseudoharengus and A. aestivalis
389
Netzel J., and E. Stanek
19G6 Some biological characteristics of blueback, Pomolbiis
aestivalis (Mitchell) and alewife, Pomolbus pseudoharengics
(Wilson), from Georges Bank, July and October, 1964. Int.
Comm. North Atl. Fish. Bull. 3:106-110.
Neves, R.J.
1981 Offshore distribution of alewife, Alosa pseudoharengus
and blueback herring, A. aestivalis, along the Atlantic coast.
Fish. Bull., U.S. 79:473-485.
Neves, R.J., and L. Depres
1979 The oceanic migration of American shad, Alosa sapi-
dissima, along the Atlantic coast. Fish. Bull., U.S. 77:
199-212.
Ramp., S.R., and W.R. Wright
1979 Northeast Channel flow: The view after one year's mea-
surements. Int. Counc. Explor. Sea CM 1979/C:54.
Rulifson, R.A.
1984 Tagging studies of river herring (Alosa pseudoharengus
and A. aestivalis) in Bay of Fundy, Nova Scotia. N.C. Dep.
Nat. Resour. Comm. Develop., Div. Mar. Fish., Completion
Rep. AFC-22, E. Carolina Univ., Greenville, 26 p.
Rulifson, R.A., S.A. McKenna, and M.L. Gallagher
1987 Tagging studies of striped bass and river herring in upper
Bay of Fundy, Nova Scotia. N.C. Dep. Nat. Resour. Comm.
Develop., Div. Mar. Fish., Completion Rep. AFC-28-1, E.
Carolina Univ., Greenville, 175 p.
Sameoto, D.D.
1982 Zooplankton and micronekton abundance in acoustic scat-
tering layers on the Nova Scotian slope. Can. J. Fish. Aquat.
Sci. 39:760-777.
Sameoto, D.D., and A.W. Herman
1989 Life cycle and distribution of C alarms finmarchicus in
deep basins on the Nova Scotia shelf and seasonal changes in
Calanus spp. Mar. Ecol. Prog. Ser. 66:225-237.
Sherman, K., W.G. Smith, J.R. Green, E.B. Cohen, M.S. Berman,
K.A. Marti, and J.R. Goulet
1987 Zooplankton production and the fisheries of the north-
eastern shelf. In Backus, R.H., and D.W. Bourne (eds.),
Georges Bank, p. 268-282. MIT Press, Cambridge.
Sinclair, M.
1988 Marine populations: An essay on population regulation
and speciation. Univ. Wash. Press, Seattle, 252 p.
Sinclair, M., and T.D. Isles
1985 Atlantic herring (Clupea harengus) distributions in the
Gulf of Maine-Scotian Shelf area in relation to oceanographic
features. Can. J. Fish. Aquat. Sci. 42:880-887.
Smith. P.C, B. Petrie, and C.R. Mann
1978 Circulation, variability, and dynamics of the Scotian Shelf
and Slope. J. Fish. Res. Board Can. 35:1067-1083.
Smith, S.J.
1988 Abundance indices from research survey data. In
Rivard, D. (ed.). Collected papers on stock assessment methods,
p. 16-43. Can. Atl. Fish. Sci. Adv. Comm. Res. Doc. 61.
Stone, H.H.
1985 Composition, morphom'etric characteristics and feeding
ecology of alewives (Alosa pseudoharengus) and blueback her-
ring (A. aestivalis) (Pisces: Clupeidae) in Minas Basin. MSc.
thesis, Acadia Univ., Wolfville, N.S., 191 p.
Stone, H.H., and G.R. Daborn
1987 Diet of alewives, Alosa pseudoharengus and blueback her-
ring, A. aestivalis (Pisces: Clupeidae) in Minas Basin, Nova
Scotia, a turbid, macrotidal estuary. Environ. Biol. Fishes
19:55-67.
Vinogradov, V.I.
1984 Food of silver hake, red hake and other fishes on (jcorges
Bank and adjacent waters, 1968-1974. NAFO (Northwest Atl.
Fish. Organ.) Sci. Counc. Stud. 7:87-94.
Vladykov, V.D.
1936 Occurrence of three species of anadromous fishes on the
Nova Scotian banks during 1935 and 1936. Copeia 1936:168.
1956 Distant recaptures of shad tagged in Quebec. Natural-
iste Canadien (Quebec) 83:235-248.
Long-term coded wire tag retention
in juvenile Sciaenops ocellatus
Britt W. Bumguardner
Robert L. Colura
Texas Parks and Wildlife Department, Perry R, Bass Marine Fisheries Research Station
HC 2, Box 385, Palaclos, Texas 77465
Gary C. Matlock
Texas Parks and Wildlife Department, Fisheries and Wildlife Division
4200 Smith School Road, Austin, Texas 78744
Red drum Sciaenops ocellatus, a
popular sport fish in the Gulf of
Mexico and associated estuarine
systems, have been subjected to in-
creasing fishing pressure in recent
years which has led to declining
population size in Texas (Matlock
1982) and poor annual survival in
Texas bays (Green et al. 1985).
Commercial harvest of both inshore
stocks of red drum in Texas prior
to 1981 (Matlock 1982) and offshore
stocks in the Gulf of Mexico prior to
1987 (Goodyear 1987) contributed
to the apparent population decline
in red drum. Documented commer-
cial landings in the Gulf of Mexico
were less than 50% of estimated
recreational harvest prior to 1984.
However, documented commercial
landings increased to more than
double the estimated recreational
harvest from 1984 to 1986, primar-
ily due to expansion of an oceanic
purse seine fishery which began in
1978 (Goodyear 1987).
In Texas, reported commercial
landings of red drum were more
than double estimated recreational
landings for 1976-77, then declined
to slightly more than recreational
landings for 1978-80. Estimated
recreational landings were relative-
ly stable, with a general downward
trend, during 1976-80 (Matlock
1982). The sale of red drum har-
vested from Texas public waters
was prohibited by legislative action
as of 1 September 1981 (Maddux et
al. 1989), while the purse seine fish-
ery for offshore stocks of adult red
drum was closed by the Gulf of Mex-
ico Fisheries Management Council
in 1986. Increasing sportfishing
pressure and catastrophic freezes,
which caused extensive fish kills in
bays along the northern Gtilf of
Mexico (Maddux et al. 1989), have
also contributed to imposition of in-
creasingly restrictive sport bag and
size limits for red drum in Texas.
Development of controlled spawn-
ing and pond culture techniques for
red drum has allowed large-scale
production and stocking of red
drum fingerlings to enhance declin-
ing populations (Colura et al. 1976,
Arnold et al. 1977, McCarty et al.
1986). Over 68 million red drum
fingerlings have been stocked for
population enhancement in Texas
coastal waters since 1975, with
the majority of fingerlings stocked
since 1983 (Dailey 1990). Develop-
ment of a reliable method for iden-
tifying stocked fish would allow
evaluation of this stocking program.
The fish, which are typically <50
mm total length (TL) when stocked
(Dailey 1990), are frequently re-
leased in spring and summer when
no small red drum (<100mmTL)
occur naturally in bays (McEachron
and Green 1986), as red drum
spawn in the fall (Comyns et al.
1991). Survival of fish stocked in
spring and summer can be moni-
tored by analysis of length-frequen-
cies for about 9 months, at which
time variation in growth masks the
initial length differences. Fish
stocked in fall cannot be monitored
by length-frequency methods due to
onset of the spawning season and
resultant confusion of stocked and
wild fish of similar size (Dailey and
McEachron 1986, Matlock et al.
1986).
For stocking to be considered suc-
cessful, hatchery fish must survive
long enough to be recruited to the
fishery and then to offshore schools
of mature red drum. When evalua-
tion of stocking success is based on
recapttire of tagged fish which must
grow large enough to enter the
fishery, determination of long-term
tag retention and detection rates is
necessary for accurate evaluation
of fingerling stocking success. Ap-
preciable tag loss or nondetection
would result in underestimation of
the proportion of hatchery fish in
the population (Heimbach et al.
1990).
Tagging of hatchery fish has had
little success (Matlock et al. 1984
and 1986, Gibbard and Colura 1980,
Bumguardner et al. 1990). Only 10
of 5942 hatchery-reared red drum
{x 452mmTL) tagged with monel
jaw tags on the opercula were re-
captured within 8 months of release
(Matlock et al. 1984). Three fish
from over 38,000 fingerlings (40-
120mmTL) tagged in the snout
with coded wire microtags and
released in St. Charles Bay, Texas,
were recaptured (Matlock et al.
1986). The low recapture rate of
microtagged fish was probably due
to tag loss. Gibbard and Colura
(1980) reported 27% retention of
coded wire tags placed in the nose
of red drum fingerlings (50 mm
mean TL) after 1 year. Bumguard-
ner et al. (1990) conducted a short-
term study (114 days) of red drum
fingerlings (x 52mmTL) tagged in
the adductor mandibularis (cheek
Manuscript accepted 9 March 1992.
Fishery Bulletin, U.S. 90:390-394 (1992).
390
NOTE Bumguardner et a\ : Long-term tag retention in juvenile Saaenops ocellatus
391
muscle) with coded wire micro-
tags. Loss of coded wire tags was
initially high (32.7% after 24
hours), but the rate of tag loss
declined substantially 23 days
post-tagging.
Tag retention by the same
group of fish initially tagged by
Bumguardner et al. (1990) was
monitored 115-464 days post-
tagging to determine if addition-
al tag loss occurred. Tag detec-
tion rates using two methods
of tag detection— a Northwest
Marine Technology Field Sam-
pling Device, and examination of
X-ray negatives— were also de-
termined and contrasted with
tag detection rates reported for
the two methods by Bumguard-
ner et al. (1990). Our primary ob-
jective was to determine if tag
retention rates declined between
114 and 464 days post-tagging,
and to what extent tag loss and
nondetection affected estimates
of tag retention rates. While Bumguardner et al. (1990)
considered mortality a component of tag loss and
reported differential mortality between tagged and un-
tagged fish, we limited the scope of this project to
investigation of tag loss and nondetection rates. We
did not consider mortality a component of tag loss
because the facilities to maintain a group of control fish
were not available.
Materials and methods
Coded-wire microtag retention for red drum was moni-
tored from tagging to 464 days post-tagging. About
2100 red drum fingerlings (x 52mmTL) were tagged
with coded wire microtags on 13 July 1987. Tags (1.07
X 0.25 mm) were inserted horizontally in the cheek mus-
cle using a Northwest Marine Technology (NMT) Model
MK2A tagging unit (Northwest Mar. Technol., Shaw
I., WA) equipped with a plastic side mold to orient fish
for consistent tag placement. An NMT Quality Control
Device was used to magnetize tags and separate tagged
from untagged fish.
Tagged fish were held in a 3.0 x 0.6 x 0.6m tank for
24 hours, stocked in three 0.1 -ha ponds at 500 fish/
pond for 23 days, then transferred to three 0.2-ha
ponds for 91 days (Table 1). Surviving fish harvested
from each 0.1 -ha pond were restocked as a group in
separate 0.2-ha ponds. Fish were fed a commercially-
Table 1
Coded wire microtag retention for red drum Sciaerwps ocellatus through 464 days post-
tagging, determined with the NMT Field Sampling Device.
No. fish
Tag
Cumulative
Interval
No. fish
retaining
retention
tag retention
Activity
No. fish
(days)
examined
tags
(%)
(%)
Tagged
2124
0-1
220"
148
67.3
67.3
Stocked in
1500
2-23
844"
397
69.8"^
47.0=
0.1-ha ponds
±31.2
±20.2
Restocked in
599
24-114
238"
108
96.6'
45.4=
0.2-ha ponds
±25.8
±12.1
Held in tank''
52
115-285
33"
31
93.9
42.6
(93.9)''
Stocked in
32
286-464
31"
26
89.3
38.0
0.4-ha pond
(83.9)'
'Fish selected randomly from the total number of fish tagged.
"Number of fish surviving at the end of the interval.
"Reported as weighted average for three ponds with standard error.
''First 52 fish encountered while monitoring tag retention were overwintered in indoor
tanks.
" Cumulative percent tag retention for days 1 15-285 used to calculate percent tag retention
for 286-464 day interval.
' Cumulative percent tag retention fordaysll5-464 used to calculate percent tag retention
for 286-464 day interval.
prepared trout feed daily while in ponds. Tag reten-
tion was determined at 24 hours (prestocking), 23 days
(harvest from 0.1-ha ponds), and 114 days (harvest
from 0.2-ha ponds) post-tagging with an NMT Field
Sampling Device (FSD) (Bumguardner et al. 1990).
Fish were harvested from 0.2-ha ponds 114 days post-
tagging, and 52 fish (x 220mmTL) confirmed by the
FSD as retaining tags were placed in a 4200 L circular
fiberglass tank on 11 October 1987 for overwintering.
As available tank space was limited, overwintering was
restricted to 52 fish confirmed as retaining tags. Ex-
perience has shown indoor overwintering is required
to insure survival of red drum in hatcheries during
episodic freezing conditions on the Texas coast. Fish
were fed 300 g chopped fish and shrimp daily. Fish were
treated with a 0.25 mg/L Cu* * bath on four occasions
for a protozoan parasite infestation tentatively iden-
tified as Amyloodinium sp. Fish were immersed in a
20 mg/L oxytetracycline HCl bath, and about lOmL of
injectable oxytetracycline solution (50 mg oxytetra-
cycline HCl/mL solution) was placed in chopped shrimp
and fish offered as feed to combat a bacterial infection.
Surviving fish {n 33) were removed from the tank on
22 April 1988 (285 days after tagging), measured and
checked for tag presence with the FSD.
The 33 surviving fish (x 352mmTL) were placed in
a 0.4-ha pond, with the exception of one fish that had
lost the caudal fin, presumably as the result of a bac-
terial infection. These fish were fed a a 35% protein
392
Fishery Bulletin 90(2), 1992
floating fish ration (Texas Farm Products,
Nacogdoches, TX), 0.45kg/day, 5 days/week,
as a supplement to natural forage available in
the pond. Fish were harvested on 11 October
1988, 464 days post-tagging. Microtag reten-
tion was determined with the FSD, fish were
measured (x 473mmTL), and 10 of 31 sur-
viving fish were selected at random and pre-
served in 50% formalin for X-ray analysis of
tag retention. X-ray negatives of the preserved
fish were visually inspected to confirm the
presence or absence of tags as determined by
the FSD.
Tag retention was determined for each in-
terval, and overall or cumulative tag retention
was determined at the end of each interval. As
mortality was not considered tag loss in this
study, cumulative tag retention reflects only
the percentage of tag losses from shedding and
nondetection of tags. A problem encountered
in the course of this program was the calcula-
tion of tag retention rates when fish which had
shed tags were not removed from the group
at the end of the interval (2-23 days, 24-114
days, and 286-464 days). The percent decrease in
cumulative tag retention was selected as an estimate
of the percentage of fish losing tags in these intervals.
Conversely, when fish that had lost tags were removed
from the group, determination of tag retention for that
interval (days 115-285) was simple (no. fish with tags/
total no. fish examined), but cumulative tag retention
had to be calculated. Tag retention for the interval in
question was multiplied by cumulative tag retention
from the previous interval to determine cumulative tag
retention for the interval. The relationship used in
these calculations was
0 50 100 150 200 250 300 350 400 450 500
Day
Figure 1
Observed values (X) of cumulative tag retention for microtagged red
drum Sciaenops oceUatiis fmgerlings through 464 days post-tagging.
TR; =
CTRj
CTRi.
X 100,
where TR; is percent tag retention for interval i, CTRj
is percent cumulative tag retention for interval i, and
CTRj. 1 is percent cumulative tag retention for the in-
terval prior to interval i. Percent tag retention and per-
cent cumulative tag retention for 1-23 and 24-114 day
intervals for fish from individual ponds were used to
calculate weighted means reported in Table 1. The
weighting factor used was the number offish harvested
from each pond.
Results and discussion
Tag retention for surviving fish at 115-464 days post-
tagging was 83.9%. Tag retention was 93.9% at 115-
285 days post-tagging, and 89.3% at 286-464 days
post-tagging (Table 1). Cumulative retention of coded
wire microtags for red drum was 38.0% at 464 days
post-tagging (Table 1, Fig. 1). Lack of replication at
all intervals prevented statistical comparison of tag
retention for different intervals. However, tag reten-
tion values of 96.6% for 24-114 days, 93.9% for 115-
285 days, and 89.3% for 286-464 days post-tagging in-
dicate cumulative tag retention decreased in the inter-
val 24-464 days post-tagging, although at a slower rate
than for the period 0-23 days (Table 1). Numerous
authors (Gibbard and Colura 1980, Klar and Parker
1986, Fletcher et al. 1987, Williamson 1987, Bum-
guardner et al. 1990, and Dunning et al. 1990) have
reported that the majority of coded-wire tag losses
occur within a relatively short period (14-90 days) post-
tagging. Our results agree with this generalization, but
indicate tag losses may continue at a much reduced rate
for extended periods after tagging. While our results
are based on a small unreplicated sample {n 31 fish at
study end), we believe they indicate long-term tag loss
may be important when estimating the contribution of
hatchery fish to a population. Accounting for this con-
tinued tag loss would prevent underestimation of the
proportion of tagged fish occurring in the population
(Heimbach et al. 1990).
Although Bumguardner et al. (1990) reported the
FSD failed to detect tags present in 9% of live fish 114
days after tagging as determined by examination of
X-ray negatives (n 186), no difference in tag detection
between the FSD and X-ray negatives was found in this
NOTE Bumguardner et al.: Long-term tag retention in juvenile Sciaenops ocellatus
393
study. Both X-ray negatives and the FSD indicated that
3 of 10 preserved fish lost tags. The criteria used to
select fish for this study, i.e., confirmation of tag
presence by the FSD, may have biased the comparison
by eliminating fish with weakly magnetized tags.
Inserting coded wire tags horizontally in the cheek
musculature of red drum fingerlings resulted in low tag
retention. The site of tag insertion and tag orientation
may affect tag retention. Tags implanted in striped
bass Morone saxatilis and largemouth bass Microp-
terus salmoides cheek musculature resulted in higher
retention rates than tags placed in snouts of striped
bass and largemouth bass (Klar and Parker 1986, Flet-
cher et al. 1987, Williamson 1987). Changing the plane
of tag insertion in the cheek muscle may increase tag
retention. Dunning et al. (1990) reported coded-wire
microtag retention in striped bass (65-100 mm TL) was
greater when tags were inserted vertically rather than
horizontally in the cheek muscle. A possible explana-
tion of poor retention and high initial loss of wire
microtags implanted horizontally in the cheek muscle
of small fish may be the small margin of error in depth
placement of the tag, due to size and thickness of the
target area. Tags may be implanted too deeply, pene-
trate the muscle, and lodge in the buccal cavity. Anes-
thetized fish could retain the tag in the buccal cavity
while passing through the Quality Control Device which
magnetizes the tag and confirms tag presence, but then
eject the tag after regaining equilibrium in the recovery
tank. Changing tag orientation in the cheek muscle
from horizontal to vertical would provide a thicker
target for tag insertion and may be responsible for
higher reported retention of microtags inserted ver-
tically rather than horizontally in the cheek muscle of
small fish.
Stocked red dnjm fingerlings are typically harvested
at about 25mm TL. Attempts to tag red drum of that
size with wire microtags have resulted in high mortality
(Gene McCarty, Texas Parks Wildl. Dep., Austin, un-
publ. data). Tagging larger fish might improve reten-
tion rates and would reduce tagging mortality, but the
fish would not be representative of the size of fish
normally stocked. These factors would complicate any
attempt to evaluate the effectiveness of stocking
hatchery-reared red drum fingerlings using fish tagged
with coded wire microtags.
Acknowledgments
We would like to acknowledge the assistance of Max-
ine Kubecka, D.V.M., Linda Kocurek, and the staff of
the Palacios Veterinary Clinic in obtaining X-ray nega-
tives. We also thank Paul Hammerschmidt for review-
ing the manuscript. This study was conducted with par-
tial funding from the U.S. Department of the Interior,
Fish and Wildlife Service, under DJ 15.605 (Project
F-36-R).
Citations
Arnold, C.R., T.D. Williams, W.H. Bailey, and J.L. Lasswell
1977 Laboratory spawning and rearing of red drum and south-
ern flounder. Proc, Annu. Conf., Southeast. Assoc. Fish
Wildl. Agencies 31:437-440.
Bumguardner, B.W., R.L. Colura, A.F. Maciorowski, and
G.C. Matlock
1990 Tag retention, survival and growth of red drum finger-
lings marked with coded wire tags. Am. Fish. Soc. Symp.
7:286-292.
Colura, R.L., B.T. Hysmith, and R.E. Stevens
1976 Fingerling production of striped bass (Morone saxatilis),
spotted seatrout {Cynoscion jiebulosus), and red drum (Sciae-
nops ocellatus) in saltwater ponds. Proc. World Maricult. Soc.
7:79-92.
Comyns, B.H., J. Lyczkowski-Schultz, D.L. Nieland, and
C.A. Wilson
1991 Reproduction of red drum, Sciaenops ocellatus, in the
north central Gulf of Mexico: Seasonality and spawner bio-
mass. In Hoyt, R. (ed.), Larval fish recruitment and research
in the Americas, p. 17-26. NOAA Tech. Rep. NMFS 95.
Dailey, J. A.
1990 Fish stocking in Texas Bays: 1975-1989. Manage. Data
Ser. 49, Texas Parks Wildl. Dep., Fish. Div., Coastal Fish. Br.,
Austin, 32 p.
Dailey, J. A., and L.W. McEachron
1986 Survival of unmarked red drum stocked into two Texas
bays. Manage. Data Ser. 116, Texas Parks Wildl. Dep.,
Coastal Fish. Br., Austin, 8 p.
Dunning, D.J., Q.E. Ross, B.R. Friedman, and K.L. Marcellus
1990 Coded vdre tag retention by, and tagging mortality of,
striped bass reared at the Hudson River hatchery. Am. Fish.
Soc. Symp. 7:262-266.
Fletcher, D.H., F. Haw, and P.K. Bergman
1987 Retention of coded wire tags implanted into cheek mus-
culature of largemouth bass. N. Am. J. Fish. Manage. 7:
436-439.
Gibbard, G.L., and R.L. Colura
1980 Retention and movement of magnetic nose tags in juve-
nile red drum. Proc, Texas Chap. Am. Fish. Soc. 3:22-29.
Goodyear. C.P.
1987 Status of the red drum stocks of the Gulf of Mexico.
Contr. CRD 86/87-34, NMFS Southeast Fish. Sci. Cent., Miami,
121 p.
Green, A.W., H.R. Osburn, G.C. Matlock, and H.E. Hegen
1985 Estimated survival rates for immature red drum in north-
west Gulf of Mexico bays. Fish. Res. (Amst.) 3:263-277.
Heimbach, D.G., D.J. Dunning, H. Wilson, and Q.E. Ross
1990 Sample-size determination for mark-recapture experi-
ments: Hudson River case study. Am. Fish. Soc. Symp. 7:
684-690.
Klar. G.T., and N.C. Parker
1986 Marking fingerling striped bass and blue tilapia with
coded wire tags and microtaggants. N. Am. J. Fish. Manage.
6:439-444.
394
Fishery Bulletin 90(2). 1992
Maddux, H.R., H.R. Osburn, D.L. Trimm, and K.W. Spiller
1989 Trends in finfish landings by sport-boat fishermen in
Texas marine waters, May 1974-May 1988. Manage. Data
Ser. 8, Texas Parks Wildl. Dep., Fish Div., Coastal Fish. Br.,
Austin, 496 p.
Matlock, G.C.
1982 The conflict between user groups of red drum and spotted
seatrout in Texas. In Stroud, R. (ed.), Proc, 7th Annu. Mar.
Rec. Fish. Symp. p. 101-109. Sport Fish. Inst., Wash., DC.
Matlock, G.C, B.T. Hysmith, and R.L. Colura
1984 Returns of tagged red drum stocked into Matagorda Bay,
Texas. Manage. Data Ser. 63, Texas Parks Wildl. Dep.,
Coastal Fish. Br., Austin, 6 p.
Matlock, G.C, R.J. Kemp Jr., and T.J. Heffernan
1986 Stocking as a management tool for a red drum fishery,
a preliminary evaluation. Manage. Data Ser. 75, Texas Parks
Wildl. Dep., Coastal Fish. Br., Austin, 27 p.
McCarty, C.E., J.E. Geiger, L.N. Sturmer, B.A. Gregg, and
W.P. Rutledge
1986 Marine finfish culture in Texas: A model for the future.
In Stroud, R. (ed.). Fish culture in fishery management,
p. 249-262. Am. Fish. Soc, Fish Cult. Fish. Manage. Sec,
Bethesda.
McEachron, L.W., and A.W. Green
1986 Trends in relative abundance and size of selected finfish
in Texas bays: November 1975-June 1984. Manage. Data
Ser. 91, Texas Parks Wildl. Dep., Coastal Fish. Br., Austin,
271 p.
Williamson, J.H.
1987 Evaluation of wire nosetags for marking largemouth
bass. Prog. Fish-Cult. 49:156-158.
Growth and mortality of
Lutjanus vittus (Quoy and Gaimard)
from the North West Shelf of Australia
Tim L.O. Davis
Grant J. West
CSIRO Division of Fisheries, Marine Laboratories
GPO Box 1538, Hobart, Tasmania 7001, Australia
The tropical waters of the North
West Shelf of Australia are highly
productive (Tranter 1962) and sup-
port a diverse fish fauna (Sainsbury
et al. 1985). A significant multispe-
cies traw^l fishery has developed in
the region, its total catch peaking
in 1973 at 37,000 t, although this
had decreased to 2700 t in 1989
(Jernakoff and Sainsbury 1990).
Lutjamis vittus is an important and
highly valued fish in this trawl fish-
ery, comprising about 4% of the
total catch (Jernakoff and Sains-
bury 1990).
Assessment of fish yields of the
North West demersal trawl fishery
is based mainly on a Beverton and
Holt dynamic pool model (Sainsbury
1987), which requires estimates of
mortality and growrth parameters
for each species. In this paper we in-
vestigate the age, growth, popula-
tion structure, and mortality of L.
vittus collected from random trawl
surveys of the North West Shelf
during 1982-83.
Materials and methods
Field collection
Material was obtained from the
CSIRO North West Shelf program
(Young and Sainsbury 1985), which
surveyed the shelf waters within
latitudes 116-119°E about every
two months between August 1982
and October 1983 (Fig. 1). Fish
were caught with a Frank and
Bryce trawl (30.5 m foot rope and a
20 mm cod-end hner) towed at 3.5-
4.5 knots for 30 minutes during the
day. Demersal tows were made at
computer-generated random posi-
tions in 13 strata defined by water
depth (10-50 m, 50-120m, and 120-
210 m), sediment type (nominally
shelly sand, sand, sandy silt, and
silt), and two geographical zones in
which different fishing regimes
were planned in the future (Table
1). Sixty-two trawl positions were
produced for each sampling survey,
with effort being allocated accord-
ing to the mean and variance of
catches determined by preliminary
surveys and the area of each stra-
tum. On average, 58 trawls were
completed each survey.
At each random station the total
weight of L. vittus and the fork
length by 10 mm classes of each L.
vittus was recorded. A subsample of
20-40 fish, approximately repre-
senting the size/frequency composi-
tion of the total catch (Kimura 1977)
was then selected from each station
for further analysis. Fork length
was measured to the nearest mm
and total weight to the nearest g,
and sex were recorded. Sagittal
otoliths and urohyals were collected
for age determination.
Manuscript accepted 18 February 1992.
Fishery Bulletin, U.S. 90:395-404 (1992).
18°S
19°
20°
115°E 116°
120°
Figure 1
Distribution of 407 random stations sampled during seven
cruises on the North West Shelf, September 1982-October
1983. The 20, 50, 120, and 200m depth contours used to
stratify sampling are shown.
Table 1
Stratified random trawl survey on the North West Shelf. The
13 strata sampled during each survey based on depth, geo-
graphical zones, and sediment type. Area (km') of each stratum
and number of random trawls made in each stratum (in paren-
thesis) on each survey are shown.
Sand
Shelly-
sand
Silt
Sandy-
silt
20-49 m
116°E-117°30'E
117°30'E-119°E
50-119m
116°E-117°30'E
117°30'E-119°E
1 20-200 m
116°E-117°30'E
117°30'E-119°E
3278(6)
6309(6)
6123(7)
9679(7)
3123(4)
1732(4)
2381(4)
2381(4)
5505(5)
1423(5)
4577(4)
4886(4) 2412(2)
395
396
Fishery Bulletin 90(2). 1992
1
^
^^
i
^^» «*»
^
Dorsal
.9
11
fM
r
"^
^■■■■■f
■»""'l
■f
l£] '^^^H
Figure 2
Urohyal bone oi Lutjanus vittus with one check (129 mmFL) and three checks (242mmFL).
measurement.
A-B taken as hne of
Age determination
A preliminary examination of the sagittal otoliths, uro-
hyals, scales, and vertebrae from 60 L. vittus indicated
that checks were more clearly defined in otoliths and
urohyals than in other hard parts. Due to their thick-
ness and opacity, otoliths in older fish required section-
ing because inner checks were obscured. As urohyals
required little preparation before reading, they were
chosen as the primary ageing structure; their only
disadvantage being that checks in older fish were
represented by a cluster of bands, so that determining
the point at which the check was formed was somewhat
subjective. Otoliths were referred to only when inter-
pretation of urohyals was difficult. Urohyals were
frozen and the flesh later removed by dipping in boil-
ing water for 5 minutes, scrubbing, rinsing, and air dry-
ing before long-term storage.
Urohyals were examined dry on a black surface
under incident light using a dissecting microscope.
Checks under this lighting appeared as dark (hyaline)
bands (Fig. 2). The distance from the origin to each
check and the outer margin of the urohyal was mea-
sured along the axis indicated in Figure 2. The period-
icity of check formation was determined from analysis
of the temporal pattern of marginal increment devel-
opment (distance from the outermost check to the
outer margin of the urohyal) calculated as the index
of completion (C) using the formula of Tanaka et al.
(1981):
WJW,
n-l.
(1)
where Wn = marginal increment, and Wn_i = pre-
vious complete increment. Analysis of variance was
NOTE Davis and West: Growth and mortality of Lutjanus vittus
397
used to test for significant differences in this index with
time of year after arcsine square-root transformation.
Growth analysis
Two forms of length-at-age data were available: lengths
were back-calculated to the last annulus (Whitney and
Carlander 1956, Carlander 1981) to provide length-at-
age data unconfounded by differences in the time of
year of sampling. Absolute age-at-observed-length was
also assigned, using an artificial January 1 birthdate.
The von Bertalanffy growth curve parameters were
fitted to both sets of length-at-age data by direct non-
linear least-squares estimation. The null hypothesis that
there was no difference between males and females in
the three growth-curve parameters was tested using
the extra sum-of-squares principle (Draper and Smith
1981, Ratkowsky 1983). The mean lengths at the last
annulus of fish aged 1-6 years were also compared be-
tween sexes using analysis of variance.
Population structure and mortality
Sex-specific length-frequency distributions and sex-
specific age-length keys were obtained from the sub-
samples from each random station, pooled for each
sampling period. It was assumed that neither the sex
ratio nor the sex-specific growth rate varied in some
systematic way between the different strata. The log-
likelihood ratio X" was used to test for departures
from a 1:1 sex ratio.
For each sampling period in 1983, the length fre-
quency of the total population was determined using
the following equation (K.J. Sainsbury, CSIRO Div.
Fish., pers. commun. 1991):
j = 13
Fi = I fi,A^/n„
(2)
where F; is the relative frequency of size-class i in the
population, fjj is the frequency of size-class i in stra-
tum j, Aj is the area of stratum j, and nj is the number
of trawls in stratum j. These length frequencies were
then broken down by sex, using sex-specific length-
frequency distribution determined for each sampling
period in 1983. The sex-specific age structure at each
sampling period was then calculated using the sex-
specific age-length keys determined for each sampling
period (Kicker 1975, Kimura 1977). A catch curve for
each sex was then constructed (Gulland 1969) and total
instantaneous mortality estimated by least-squares
linear regression of the descending right-hand of the
catch curve. Equality of mortality rates between the
sexes was determined by analysis of covariance.
-,
80-
•
_ 60-
O)
•
1 40-
CC
O
•
•
20-
• * •
0-
(
3 50 100 150 200
Depth (m)
Figure 3
Catch per trawl of Lutjanus vittus by water depth on the
North West Shelf of Australia, 1982-83.
Results
Depth distribution
Lutjanus vittus were caught in depths from 20 m (the
shallowest depth sampled) to 120 m, with the largest
catches being at 30-70 m (Fig. 3). There was a positive
correlation between individual fish lengths and depth
(r 0.337, t 24.7, df 4754, P<0.001). While almost the
full size-range was encountered at most depths, there
was a marked absence of fish < 200 mm at depths >90m
(Fig. 4).
Length/weight relationship
In the regression of log weight/log length, the test for
homogeneity of slopes between sexes was found to be
not significant (ANCOVA, F 0.318, df 1, 2604, P 0.57)
and, assuming a common slope, there was no signifi-
cant difference in the intercepts for the two sexes
(ANCOVA, F 1.76, df 1, 2605, P 0.19). Both sexes and
juveniles whose sex could not be determined were then
combined and a general relation between length (L in
mm) and weight (W in g) for L. vittus was determined:
W = 9.99x10-6 L3086
{F 367248, df 1, 2797, P<0.001).
398
Fishery Bulletin 90(2). 1992
400-
12
29
8J
^
5226
34239514 12
Length (mm)
I\J CO
O O
O O
1 1 1 1
2
54
33
H
1-
h^
V H
h ^
-li*
100-
1
0-
c
) 20 40 60 80 100 120
Depth (m)
Mean lengt
range, and
intervals.
Figure 4
h, 95% confidence limits (vertical bars), length
sample size of Lutjanus vittus by 10 m depth
Annulus formation
Evidence that checks are formed annually was obtained
by examining the index of completion at about 2-month
intervals throughout one season. The index of comple-
tion is a measure of the amount of bone growth since
the last check was formed, expressed as a proportion
of the previous growth increment. The indices of com-
pletion for fish aged 1-6 years were combined after
each age-group was observed to follow the same
seasonal changes in the index (Fig. 5). There were
significant (P< 0.001) differences in this index with
time of year for urohyals having one, two, three, four,
five and six checks (ANOVA, F 31.4, df 4, 172; F 80.1,
df 6, 246; F 100.7, df 6, 263; F 40.0, df 6, 141; F 15.5,
df 6, 69; F 88.8, df 5, 233, respectively). While there
was considerable variation in this index at any one
sampling period, there was a steady increase in the
mean index from October to August, followed by a
marked drop between August and October. It appears
that checks are laid down some time between August
and October.
Bacl< -calculation
Lengths were back-calculated to the last annulus, using
a proportional method based on the regression of fish
length on urohyal length— the body proportional
hypothesis (BPH) of Francis (1990). A quadratic equa-
tion best described the relationship between body
ONDJ FMAMJ
1982
1983
Figure 5
Seasonal change in the index of completion (amount of bone
growth since the last check was formed) in Lutjamis inttus
urohyals with 1-6 checks. Individual indices are plotted against
day of sampling, and the line links the mean index for each
sampling period.
4UU-
.*■ :* ^^
„ 300-
E
E,
^'
o) 200-
0)
■^
1:-
to
"^ 100^
y
0-
1
10 15 20
Urohyal length (mm)
25
Figure 6
Curvilinear relationship between urohyal length and fish
length in Lutjanus I'itttis.
length (L in mm) and urohyal length (U in mm) (Fig. 6):
L= - 25.48 -H20.485U-0.193U2 (r2 0.95, df 1102).
The mean absolute difference between using BPH
and SPH (regression of fish length on urohyal length)
was 1.6mm; BPH back-calculated smaller lengths in
fish <150mm and larger lengths in fish >200mm.
Growth
Von Bertalanffy growth curves were fitted to length-
at-age data for each sex separately. Fish whose sex
NOTE Davis and West Growth and mortality of Lutjanus vittus
399
Table 2
Mean lengths back-calculated to the last annulus, |
95% confidence limits, and
sample number (n)
at each age (years) for male and female Lutjanus |
vitttis.
Age
Mean length
(yr)
(mm)
95% CL
n
Males
1
90
87-93
92
2
164
160-167
104
3
214
210-217
122
4
261
256-266
66
5
295
288-302
52
6
313
304-322
22
7
338
330-347
14
Females
1
89
87-92
92
2
167
163-170
148
3
216
214-219
159
4
251
247-253
115
5
276
271-282
45
6
291
284-298
17
could not be identified (104 juveniles) presented a
problem, because excluding them created a bias since
fish that could be sexed in age-class 1 (40 males and
40 females) were larger animals. There were significant
differences (P< 0.001) in back-calculated and observed
lengths between age-class 1 males, females, and juve-
niles (one-way ANOVA, F 50.7, df 2, 181; F 89.0, df 2,
181, respectively). Multiple comparison by the Tukey
test indicated that age-class 1 fish that could be sexed
were significantly larger than juveniles by about 17 mm
for back-calculated lengths and 33 mm for observed
lengths. To eliminate this bias, juveniles were ranked
by size. The smallest was randomly assigned a sex, and
then each juvenile in order was assigned to alternate
sexes. The mean lengths back-calculated to the last
annulus at each age for male and female (including
assigned sexes in age-class 1) are presented in Table
2 and Figure 7.
Back-calculated length-at-age data minimize the ef-
fects of seasonal growth but do not completely elim-
inate it, because the time of check formation ranges
over several months (Fig. 5). Assigning an absolute age
using an arbitrary birthdate will only compensate for
growth differences between fishes caught at different
times of the year when there is little seasonal varia-
tion in growth rate. However, assigning absolute ages
does enable age-class 0 data to be used in determining
growth curves (Fig. 8). While back-calculated lengths
cannot use age-class 0 data, they do enable a more
realistic time-scale parameter (tg) to be estimated.
400
300 -
E
E
^200
c
100 -
52 22
14 2
Males
-1 — I — I — I — I — I — I — I — I — I
400
300 -
E
E
o) 200
c
CD
100
45
17 1
Females
— I 1 1 1 r 1 1 1 1 1
01 23456789 10
Age (years)
Figure 7
Growth of male and female Lutjanus vittus. Mean back-
calculated lengths to the last annulus, 95% confidence limits
(vertical bars), range, and sample size have been plotted.
Von Bertalanffy growth curves were fitted to individual
observations.
The least-squares estimates of the von Bertalanffy
growth curve parameters are quite different between
the sexes for both forms of length-at-age data (Table
3). Independent of any assumed growth curve, there
were significant differences in mean back-calculated
lengths between sexes for age-classes 4-6 years
(ANOVA, F 42.1, df 1, 311, P<0.001) but age-classes
1-3 were not significantly different (F0.32, df 1,
605, P 0.569). Only fish whose sex was determined
400
Fishery Bulletin 90(2), 1992
E
E
C
400
300
200
100-
• ^.^
'^!^it^^
• *
■i^
Mj^I
•
Orrl
■M
~5h
W *
1 1 1—
Males
— 1 1 1 1 1 r
1
400
300-
E
E
en
c
200
100
Females
— I — \ — I — I — I — I — I — I — I — I
01 23456789 10
Age (years)
Figure 8
Growth of male and female Lutjanus vittus assigned an ab-
solute age assuming a 1 January birthdate.
o
n)
E
o
■c
o
a.
1.0
0.8 -
0.6
0.4
0.2
0.0
t£) CVJ [^ Ifi^
100
200
Length (mm)
300
400
Figure 9
Proportion of males by 10 mm length-class. Means, 95%
binomial confidence limits (vertical lines), and sample size are
plotted.
were used in this analysis. Both males and females
grow at the same rate for the first three years, after
which females grow at a markedly slower rate than
males.
Sex ratio
There was a marked departure from a 1:1 sex ratio
(Fig. 9; likelihood ratio x' 152.1, df 29, P<0.001)
which can be attributed to different growth rates be-
tween the sexes. Below 300 mm, sex ratios did not dif-
fer from 1:1 (likelihood ratio x" 23.3, df 18, P 0.18) but
in all larger size groups there was a predominance of
males.
Table 3
Estimated parameters ( ± SE) of the von
Bertalanffy growth curve for Lutjanus vittus.
Growth curve parameters
F test of parameter estimates
F df P
df L„ (mm) K
to(yr)
Back-calculated length-at-age
Males 473 403(10.4) 0.26(0.01)
Females 578 323(6.6) 0.39(0.02)
0.02(0.05)
0.17(0.04)
21.7
3, 1045 <0.001
Length at absolute age
Males 486 422(15.9) 0.22(0.02)
Females 586 325(7.7) 0.37(0.03)
-0.56(0.09)
-0.23(0.08)
19.9
3, 1066 <0.001
Length-frequency
distributions
Length-frequency distributions of
the population were determined
separately for males and females
(Fig. 10). Each length-class was
separated into age-classes based
on the urohyal data. There was
a jump in age-class between sam-
ples taken in August and October
because a new check was formed
in the intervening period. While
NOTE Davis and West Growth and mortality of Lutjanus vittus
401
Females
100 150 200 250 300 350 400
Length (mm)
50 100 150 200 250
Length (mm)
300 350 400
Figure 10
Total population length-frequency distribution of male and female Lutjanus inttus at each sampling period in 1983. Hatching within
each distribution indicates the age-class structure determined from urohyal ageing. All age-groups increment by 1 year, August-
September, due to check formation.
402
Fishery Bulletin 90(2), 1992
separation of sexes would have reduced variance in the
length-frequency distribution of the older age-classes
due to growth differences between sexes, there was
still considerable size overlap of age-classes and diffi-
culty in identifying modes after age 2 (Fig. 10). There
was a clear progression in the length of age-classes 1
and 2 through the year, although age-class-2 fish of
both sexes were somewhat larger than expected in
February based on the progression in length of this age-
class in subsequent months.
Mortality
The relative abundance of each age-class by sex was
determined for the five periods sampled in 1983 (Fig.
11). A line was fitted by least squares to the descend-
ing limb of the catch curve. Fish not considered to be
fully recruited to the sampling gear (circled points)
were excluded. There was no significant difference in
the slopes of the lines for males and females (ANCOVA,
F 0.85, df 1, 47, P 0.36) and no significant sex effect
(ANCOVA, F 1.23, df 1, 47, P 0.27) so a catch curve
was fitted to the combined data. The instantaneous rate
of annual mortality (Z) for males and females was
estimated to be 0.98 (SE 0.076).
Discussion
Lutjanus vittus was caught at depths of 20 m (the
shallowest depth sampled) to 120 m, with larger fish
tending to inhabit deeper waters. This tendency has
also been observed in other shallow-water lutjanids
such as L. aya (Moseley 1966), L. griseus (Starck 1971),
and L. hohar (Wright et al. 1986).
Most ageing studies on lutjanids have relied on oto-
liths as the principal structure (see review by Manooch
1987).However, a few authors (i.e., Reshetnikov and
Claro 1976, Pozo and Espinosa 1982, Claro 1983,
Palaz6n and Gonzalez 1986) have used urohyals.
Reshetnikov and Claro (1976) had difficulty determin-
ing the boundaries of the annual increment after the
second or third annulus in urohyals because the annuli
were made up of multiple bands. It was our experience
that, despite this problem, increments were still easier
to measure on urohyals than on whole otoliths, and
preparation was far less time-consuming.
Our preliminary investigation indicated that the same
number of checks were formed on a variety of hard
structures, including urohyals. Data on marginal in-
crements in urohyals showed a seasonal pattern with
one check being formed each year, consistent with most
other studies on lutjanids in tropical waters. However,
studies on two lutjanid species from Cuban shelf waters
have suggested that checks are formed twice a year
01 23456789
®v
y=-1.07x-f
17.3
14 -
@ • \
® a
X". ■
?r
\
• Jan 1983
c
x
■ Apr 1983
3
cr
®
12 -
<
° \
0 Jun 1983
o Aug 1983
cu
* Oct 1983
o
10 -
8 _
©
Females
0 \
■
D \
0123456789
Age (years)
Figure 1 1
Logj frequency against absolute age for male and female
Lutjanus vittus. Data for fish <2 years old (circled) were ex-
cluded from the regressions, as they were incompletely
recruited age-classes.
(Espinosa and Pozo 1982, Pozo and Espinosa 1982).
The growth of male and female L. vittus was sig-
nificantly different after 3 years of age, with females
growing markedly slower than males. There are few
documented cases of growth rates differing between
sexes in lutjanids. However, female L. vittus in New
Caledonia were found to grow at a slower rate, and
slight growth differences were found in L. amabilis
in New Caledonia (Loubens 1980) and L. synagris
in Trinidad (Manickchand-Dass 1987). All mature
females observed were 3 years of age or older (unpubl.
data), and it seems likely that females grow more
slowly than males at this stage because they expend
proportionally more energy on gamete production than
do males. Stunting in females from a sexually preco-
cious population oiLates calcarifer was also attributed
NOTE Davis and West: Growth and mortality of Lutjanus vittus
403
to channeling energy into gonadal growth at the ex-
pense of somatic growth at a relatively early age (Davis
1984).
Length-frequency distributions did not show the
modal structure one would expect knowing the age
structure of the population. The length-frequencies
showed three modes, whereas direct ageing suggested
there should be at least six. Length-based methods of
ageing work best with fish that spawn over a short
period of time, have short life spans, and are fast grow-
ing; characteristics not typical of lutjanids (Manooch
1987).
A preponderance of females at larger sizes has been
reported in studies of other lutjanids, e.g., L. synagris
(Rodriguez Pino 1962, Erhardt 1977), Etelis carbun-
culus (Everson 1984), E. coruscans, Aprion viriscens
(Everson et al. 1989), and Rhomboplites aurorubens
(Grimes and Huntsman 1980). The latter authors at-
tributed the preponderance to differential mortality
and longevity. L. vittus goes against this trend: males
predominate the larger size-classes, as is the case for
Lutjanus mnabilis (Loubens 1980) and Lutjanus buc-
canella (Thompson and Munro 1983). The preponder-
ance of males at larger sizes appears to be due largely
to a reduction in growth rates of mature females.
No significant differences were found in the instan-
taneous rate of annual mortality (Z) between male and
female L. vittus. One of the assumptions of estimating
mortality using the catch curve method of Gulland
(1969) is that the mortality rate is constant for all years
used in the estimation. This may not be the case for
female L. vittus after 6 years of age. However, the data
points in the oldest age-groups are based on smaller
sample sizes, so the mortality curve at this stage should
be interpreted with caution. Using the relationship be-
tween natural mortality (M) and the growth coefficient
(K) for snappers and groupers determined by Ralston
(1987) from published data provides us with estimates
for M of 0.59 for males and 0.92 for females. The value
for males seems reasonable, but that for females is
unlikely if total mortality is about 0.98. Clearly, regres-
sion methods to produce estimates of M such as those
used by Pauly (1980) and Ralston (1987) should be ap-
plied with caution.
Acknowledgments
This paper is dedicated to the memory of Mr. Otto
Augustine, a technician with the CSIRO Division of
Fisheries. He was responsible for the ageing of many
fish species in the Division's programs from the late
sixties up until his death in 1990. He determined the
age and marginal increment data used in this paper.
We wish to thank W. Thomas for laboratory assistance
and all people who assisted in the fieldwork on the
Northwest Shelf Program. We are grateful to K.J.
Sainsbury for providing length-frequency and catch
data from his research program, and K. Haskard for
statistical advice. S. Blaber and J.S. Gunn reviewed
the manuscript.
Citations
Carlander, K.D.
1981 Caution on the use of the regression method of back-
calculating lengths from scale measurements. Fisheries
(Bethesda) 6:2-4.
Claro, R.
1983 Ecologia y ciclo de vida del caballerote, Lutjanus grisevs
(Linnaeus), en la plataforma Cubana: 2. Edad y crecimiento,
estructura de las poblaciones, pesquerias. Rep. Invest. Inst.
Oceanol. Acad. Cienc. Cuba 8, 26 p.
Davis, T.L.O.
1984 A population of sexually precocious barramundi, Lates
calcarifer, in the Gulf of Carpentaria, Australia. Copeia 1984:
144-149.
Draper, N.R., and H. Smith
1981 Applied regression analysis, 2d ed. Wiley, NY, 407 p.
Erhardt, H.
1977 Beitrage zur biologie von Lutjanus synagris (Linnaeus
1758) an der Kolumbianischen Atlantikkuste. Int. Rev.
Gesamten Hydrobiol, 62:161-171.
Espinosa, L., and E. Pozo
1982 Edad y crecimento del sesi {Lutjanus buccanella Cuvier,
1828) en la plataforma suroriental de Cuba. Rev. Cubana
Invest. Pesq. 7(1):80-100.
Everson, A.R.
1984 Spawning and gonadal maturation of the ehu, Etelis car-
bunculus, in the Northwestern Hawaiian Islands. In Grigg,
R.W., and K.Y. Tanoue (eds.), Proc, Second symposium on
resource investigations in the northwestern Hawaiian Islands,
vol. 2, p. 128-148. UNIHI-SEAGRANT-MR-84-01, Univ.
Hawaii. Honolulu.
Everson, A.R., H.A. Williams, and B.M. Ito
1989 Maturation and reproduction in two Hawaiian eteline
snappers, uku, Aprion virescens, and onaga, Etelis coruscans.
Fish. Bull., U.S. 87:877-888.
Francis, R.I.C.C.
1990 Backcalculation of fish length: A critical review. J. Fish
Biol. 36:883-902.
Grimes, C.B., and G.R. Huntsman
1980 Reproductive biology of the vermilion snapper, Rhom-
boplites aurorubens, (Cuvier) from North and South Carolina
waters. Fish. Bull. U.S. 78:137-146.
Gulland, J.A.
1969 Manual of methods for fish stock assessment. Part 1.
Fish population analysis. FAO Man. Fish. Sci. 4:1-154.
Jernakoff, P.. and K.J. Sainsbury
1990 CSIRO's northern demersal finfish stock assessments:
1980 to 1990. Bur. Rural Resour., Dep. Primary Ind. Energy
Aust. Inf. Pap. IP/6/90, Canberra, A.C.T.. Aust., 169 p.
Kimura, D.K.
1977 Statistical assessment of the age-length key. J. Fish.
Res. Board Can. 34:317-324.
404
Fishery Bulletin 90(2). 1992
Loubens, G.
1980 Biologie de quelques especes de poissons du lagon n^o-
caldonien. II. Sexuality et reproduction. Cah. Indo-Pac.
II(l):41-72.
Manickchand-Dass, S.
1987 Reproduction, age and growth of the lane snapper, Lut-
janus synagris (Linnaeus), in Trinidad, West Indies. Bull.
Mar. Sci. 40 (l):22-28.
Manooch, C.S. Ill
1987 Mortality rates of snappers and groupers. In Polovina,
J. J., and S. Ralston (eds.), Tropical snappers and groupers:
Biology and fisheries management, p. 329-373. Westview
Press, Boulder.
Mori, K.
1984 Early life history ofLutjanus intta (Lutjanidae) in Yuya
Bay, the Sea of Japan. Jpn. J. Ichthyol. 30(4):374-392.
Moseley, F.N.
1966 Biology of the red snapper, Lutjanus aya Bloch, of the
Northwestern Gulf of Mexico. Publ. Inst. Mar.Sci. Univ. Tex.
11:90-101.
Pauly, D.
1980 On the interrelationships between natural mortality,
growth parameters, and mean environmental temperature in
17.5 fish stocks. J. Cons. Perm. Int. Explor. Mer 39:175-192.
Palazon, J.L., and L.W. Gonzalez
1986 Edad y crecimiento del pargo cebal, Lutjanus analis
(Cuvier, 1828) (Teleostei; Lutjanidae) en la isla de Margarita
y alrededores, Venezuela. Invest. Pesq. 50(2):151-165.
Pozo, E., and L. Espinosa
1982 Estudio de la edad y el crecimento del pargo del alto {Lut-
janus vivanus Cuvier, 1828) en la plataforma suroriental de
Cuba. Rev. Cubana Invest. Pesq. 7(2):l-23.
Ralston, S.V.D.
1987 Mortality rates of snappers and groupers. In Polovina,
J.J., and S. Ralston (eds.). Tropical snappers and groupers:
Biology and fisheries management, p. 375-404. Westview
Press, Boulder.
Ratkowsky, D.A.
1983 Nonlinear regression modeling: A unified practical ap-
proach. Marcel Dekker, NY, 276 p.
Reshetnikov, Y.S., and R.M. Claro
1976 Cycles of biological processes in tropical fishes with
reference to Lutjanus synagris. J. Ichthyol. 16:711-723.
Ricker, W.E.
1975 Computation and interpretation of biological statistics
of fish populations. Bull. Fish. Res. Board Can. 191, 382 p.
Rodriguez Pino, Z.
1962 Estudios estadlsticos y bioWgicos sobre la biajaiba (Lut-
janus synagris). Cent. Invest. Pesq., Notas Sobre Invest.
4:1-91.
Sainsbury, K.J.
1987 Assessment and management of the demersal fishery on
the continental shelf of northwestern Australia. In Polovina,
J.J., and S. Ralston (eds.). Tropical snappers and groupers:
Biology and fisheries management, p. 465-503. Westview
Press, Boulder.
Sainsbury, K.J., P.J. Kailola, and G.G. Leyland
1985 Continental shelf fishes of northern and northwestern
Australia. Clouston & Hall and Peter Pownall Fish. Inf. Serv.,
Canberra, Aust., 375 p.
Starck, W.A. II
1971 Biology of the gray snapper, Lutjanus griseus. In Starck,
W.A. II, and R.E. Schroeder (eds.), Investigations on the gray
snapper, Lutjanus griseus. p. 1-150. Stud. Trop. Oceanogr.
(Miami) 10, Univ. Miami Press.
Tanaka, K., Y. Mugiya, and J. Yamada
1981 Effects of photoperiod and feeding on the daily growth
patterns in otoliths of juvenile Tilapia nilotica. Fish. Bull.,
U.S. 79:459-465.
Thompson, R., and J.L. Munro
1983 The biology, ecology and bionomics of the snappers, Lut-
janidae. In Munro, J.L. (ed.), Caribbean coral reef fishery
resources, p. 94-109. ICLARM Stud. Rev. 7, Int. Cent. Liv-
ing Aquat. Resour. Manage., Manila.
Tranter, D.J.
1962 Zooplankton abundance in Australian waters. Aust. J.
Mar. Freshwater Res. 13:106-129.
Whitney, R.R., and K.D. Carlander
1956 Interpretation of body-scale regression for computing
body length of fish. J. Wildl. Manage. 20:21-27.
Wright, A., P.J. Dalzell. and A.H. Richards
1986 Some aspects of the biology of the red bass, Lutjanus
bohar (Forsskal), from the Tigak Islands, Papua New Guinea.
J. Fish Biol. 28:533-544.
Young, P.C, and K.J. Sainsbury
1985 CSIRO's North West Shelf program indicates changes
in fish populations. Aust. Fish. 44(3): 16-20.
Correlation of winter temperature
and landings of pink shrimp
Penaeus duorarum in IMorth CavoWna
William F. Hettler
Beaufort Laboratory, Southeast Fisheries Science Center
National Marine Fisheries Service, NOAA, Beaufort, North Carolina 28516-9722
In habitats where low water temper-
ature is not a Hmiting factor, pink
shrimp Penaeus duorarum produc-
tion has been related to rainfall
and surface-water inflow (Browder
1985, Sheridan 1991). In contrast,
North Carolina landings of pink
shrimp were correlated with water
temperature during the previous
winter, but not to rainfall (Hettler
and Chester 1982). In that study,
the average water temperature of
the two coldest consecutive weeks
of each year recorded at a single
temperature station located at the
Beaufort Laboratory was a predic-
tor of spring landings (through July)
for the entire North Carolina
fishery. Fifteen years of tempera-
ture records and landings were used
to determine this relationship. Since
the last year reported (1981), 10 ad-
ditional years of temperature and
landings data have become avail-
able. This note presents these new
data and uses the resulting 25-year
time-series to report that average
minimum winter water tempera-
ture remains a reliable basis for
forecasting landings of this species.
The temperature/landings relation-
ship previously published (Hettler
and Chester 1982) was recalculated
after adding the 1982-91 tempera-
ture and landings data (Table 1, Fig.
1). No evidence of curvilinearity in
the relationship could be found by
fitting higher-order polynomial
models. A time-series model was
not appropriate because pink
shrimp are 'annuals' and their an-
nual population levels generally
show low autocorrelation as sug-
gested by the 1962-91 North Caro-
lina pink shrimp heads-off landings
data (Fig. 1). Thus the simple linear
Table I
Actual and predicted landings (heads-off) of pink shrimp Penaeus duorarum in the
North Carolina spring fishery, February-July, based on average water temperature
of the two coldest consecutive weeks of the preceding winter.
Year
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
Temp.
°C
Landings (kg)
Actual
Predicted
Percent over ( + )
or under ( - )
5.0
8.8
5.9
5.4
6,9
8.3
6.1
8.1
3.7
10.0
197,630
451,163
184,380
126,797
307,514
551,521
433,125
639,166
66,853
592,.381
173,527
491,765
248,899
207,025
332,646
449,892
265,648
433,142
64,656
592,262
+ 13.9
-8.3
-25.9
-38.7
-7.6
-H22.6
-1-63.0
-I- 47.5
+ 3.4
< + 0.1
model was retained. The new re-
gression to determine spring pink
shrimp landings in North Carolina
was
Landings (kg) = 83747(T)- 245208,
where T was the average tempera-
ture of the two coldest consecutive
weeks (°C). The relationship was
significant (P<0.001, r- 0.803).
The more general relationships of
average winter water temperature
(Dec-Mar) or average midwinter
water temperature (Jan-Feb) did
not correlate with landings over the
25-year time-series.
Predicted landings of pink shrimp
were calculated and averaged with-
in 25% of the actual landings for the
recent 10-year period. Landings in
1991 were within >0.1% of the pre-
diction. Possible causes of the rela-
tively large deviations in some
years' landings from the predicted
are discussed in Hettler and Ches-
ter (1982) and include errors in the
process of estimating landings,
year-to-year changes in fishing ef-
fort, and, in addition, possible local
thermal anomalies.
These new data continue to sup-
port the hypothesis that reduced
pink shrimp landings in North Caro-
lina are probably a result of cold kill
of overwintering shrimp caused by
cold water temperatures. In the
coldest years (1963, 1977, 1978, and
1990) when spring landings were
less than 100,000kg, lethal cold
water probably penetrated all but
the most highly protected over-
wintering estuarine habitat. North
Carolina is the northern limit in the
range of pink shrimp, thus this
species is more likely to encounter
low temperature stress in this loca-
tion than in more southerly loca-
tions. The linearity of the model is
perhaps a consequence of these
shrimp's inherent vulnerability to
cold water temperatures interact-
Manuscript accepted 20 March 1992.
Fishery Bulletin, U.S. 90:405-406 (1992).
405
406
Fishery Bulletin 90|2|. 1992
ex
y:
o
o
o
o
o
a
o
z
<
Figure I
(Upper) Regression and 95% confidence limits for 25 years
of temperatures and landings (heads-off) of pink shrimp
Penaetis diiorarum. Open circles represent years before 1982
reported by Hettler and Chester (1982); closed circles repre-
sent the years 1982-91. (Bottom) Spring landings (heads-off)
of pink shrimp in North Carolina since 1962.
ing with the geographical/spatial distribution and
availability of habitats that respond differently to drop-
ping temperatures. The safest habitats would include
favorable sediments for deep burrowing, deep water,
and physiologically isosmotic salinity.
Acknowledgment
Landings data for 1982-91 were provided by the North
Carolina Division of Marine Fisheries, NC/NMFS Co-
operative Regional Statistics Program.
Citations
Browder, J. A.
1985 Relationship between pink shrimp production on the Tor-
tugas grounds and water flow patterns in the Florida Ever-
glades. Bull. Mar. Sci. 37:839-856.
Hettler, W.F., and A.J. Chester
1982 The relationship of winter temperature and spring land-
ings of pink shrimp, Penaeus dvxirarum, in North Carolina.
Fish. Bull., U.S. 80:761-768.
Sheridan, P.F.
1991 Tortugas pink shrimp forecast. In Ba.xter, K.N. (ed.),
Shrimp resource review, briefing book, p. 39. NMFS South-
east Fish. Sci. Cent., Galveston, TX.
Growth of five fishes
in Texas bays in the 1960s
Gary C. Matlock
Texas Parks and Wildlife Department
4200 Smith School Road, Austin, Texas 78744
The estuarine sport and commercial
fish fisheries in Texas have histor-
ically relied upon five species: black
drum Pogonias cromis, red drum
Sciaenops ocellatus, sheepshead Ar-
chosargus probatocephalus, south-
ern flounder Paralichthys lethostig-
ma, and spotted seatrout Cynoscion
nebulosus. Regulation of these fish-
eries dramatically increased as hu-
man demand for fish generally in-
creased through the 1980s. For
example, the sale of red drum and
spotted seatrout caught in Texas
was prohibited in 1981, use of nets
in coastal waters was prohibited in
1988, and size, bag, and possession
limits were imposed for each species
by 1988. Growth information was
used in selecting appropriate regu-
lations for optimizing yield and sus-
taining recruitment. However, com-
prehensive, coastwide growth rates
were available only for red drum,
black drum, and spotted seatrout
caught in the late 1970s and 1980s
when exploitation was extremely
high (Doerzbacher et al. 1988,
Green et al. 1990). Potential yields
may be underestimated when based
on growth rates obtained when
fishing mortality is high. Tagging
Figure I
Location of Texas
bay systems.
data from which growth param-
eters could be estimated for those
species had been collected sporad-
ically from the late 1950s through
the early 1970s (Green 1986) when
fishing effort was presumably lower
than in the 1980s, but these data
have not been examined. The objec-
tive of this study was to describe
quantitatively the growth of black
drum, red drum, sheepshead, south-
ern flounder, and spotted seatrout
tagged in the 1960s.
Methods
Data on total length (TL, mm) at
tagging and recapture, and the
number of days free until recapture
for five fishes— black drum, red
drum, sheepshead, southern floun-
der, and spotted seatrout— tagged
by the Texas Parks and Wildlife
Department (TPWD) in Texas bays
(Fig. 1) and recaptured during the
period 1950-75 were obtained from
Green (1986). No length data were
available for fish tagged in the
Matagorda Bay system, however.
Data resulted from a variety of pro-
jects designed to obtain life history
information on fishes, mainly red
drum and spotted seatrout. Fish for
tagging were obtained using rod
and reel, trotlines, and trammel and
gill nets. Monel strap tags and in-
ternal abdominal tags were primar-
ily used. The release of tagged fish
and requests for information con-
cerning recaptured fish were adver-
tised through the news media and
posters placed in areas frequented
by fishermen. Non-monetary re-
wards of various types were usual-
ly offered for returned tags. Addi-
tional details are contained in Green
(1986). The mean daily growth rate
(G) was used to examine the suit-
ability of the von Bertalanffy model
for describing growth of each spe-
cies. The growth rate was calcu-
lated as follows:
Manuscript accepted 9 March 1992.
Fishery Bulletin, U.S. 90:407-411 (1992).
407
408
Fishery Bulletin 90(2). 1992
(Ir
,)/d,
where If = TL at recapture,
Im = TL at tagging, and
d = time in days between tagging and
recapture.
A plot of mean daOy growth rate versus TL at tagging
for each species suggested asymptotic growth, since
growth rate generally declined as size-at-tagging in-
creased. Therefore, the von Bertalanffy growth model
was chosen as an empirically-based description of
growth (Moreau 1987) to which these tagging data
were fit. Of the currently available estimating pro-
cedures for using tag data to describe growth follow-
ing the von Bertalanffy growth equation, Fabens'
(1965) method provides the most accurate estimates
(Sundberg 1984). Data were analyzed using the Fishery
Science Application System (Saila et al. 1988) and
Fabens' (1965) iterated least-squares method for
estimating K and L^ in the von Bertalanffy growth
equation,
Ir = Im + (L^-U)[l-exp(-Kd)]
where 1^, Im . and d are defined as above, and
L^ = the average TL in a population of fish
allowed to grow indefinitely following
the von Bertalanffy growth function, and
K = Brody's growth coefficient (per day).
Before analysis, data were screened following pro-
cedures of Doerzbacher et al. (1988) to eliminate
outliers. Fish with growth rates >3 mm/day or <-3
mm/day were eliminated from the data set. The mean
± 3 SD for the remaining data were then calculated, and
fish with growth rates outside this range were also
eliminated from the data set. Sufficient data were
avaUable to analyze tagged red drum separately by bay
system (except for Sabine Lake and Matagorda Bay).
Data for each of the other species were analyzed for
all tagging locations combined.
The measure of effectiveness (P) used by Phares
(1980) which is similar to the multiple correlation coef-
ficient of linear regression (i?^) was used to determine
how well the von Bertalanffy model fit the data:
P = (SSL-SSE)/SSL,
where SSL is the sum of squares of (1^ -!„,), and SSE
is the residual sum of squares of the model,
SSE = (1/-1,)2,
where Ir' is the model's predicted length-at-recapture,
and n is the number of recaptured tagged fish (after
data screening). The value of 1 r' for each tagged fish
was calculated following Parrack (1979):
; = L.
(Loo-lm)e-K(d)).
Standard errors of each estimated K and L^ were
estimated using 10-fold cross-validation technique (a
form of jackknife resampling) described by Verbyla and
Litvaitis (1989). For each data set, the original data
were randomly partitioned into ten subsamples, nine
of which each contained 10% of the data, and one which
contained the remainder. The first subsample was ex-
cluded from the data set, and K and L^ were reesti-
mated. The first subsample was recombined with the
data set, and the second subsample was excluded, and
so on, until all 10 subsamples had been excluded. The
standard error of each parameter of the original data
set is approximated by the standard deviation of the
mean of the 10 separate estimates made after remov-
ing each subsample.
Results and discussion
Most of the data reported for recaptured tagged fish
during the 1960s were included in the analyzed data
set (i.e., few outliers were found). Of 1630 recaptured
fish, only 72 (4.4%) fish were excluded from the anal-
yses (Table 1). Red drum from the lower Laguna Madre
had the greatest proportion of outliers (13 of 69 fish).
However, the size range at tagging of the remaining
56 fish was comparable to the range of red drum tagged
in other bays. These results are similar to those of
Doerzbacher et al. (1988) for red drum and black drum,
and are supported by Ferguson et al. (1984) who
demonstrated that red drum lengths reported by sport-
fishermen were accurate.
Mean daily growth rates of tagged fish during the
time between release and recapture were about 0.2
mm/day for all species, except red drum which aver-
aged about 0.4-0.7 mm/day (Table 1). These means
mainly represent the growth of smaller fish within each
range because the size data were skewed toward small
fish. For example, of 254 recaptured black drum, over
250 were <300mmTL at tagging and recapture.
However, the estimates of daily growth for black drum,
red drum, sheepshead, and spotted seatrout in this
study were within the ranges of those reported by
Colura et al. (1984), Cornelius (1984), Beckman et al.
(1988, 1990, 1991), Doerzbacher et al. (1988), Murphy
and Taylor (1989), Matlock (1990), and Green et al.
(1990).
The estimated L^ for black drum, red drum, south-
oo ' "
ern flounder, and spotted seatrout tagged in Texas
NOTE Matlock: Growth of five fishes tagged in Texas bays in the 1960s
409
Table 1
Size, time free, and growth rate of five fishes tagged and released
n Texas bays and
recaptured by sport and commercial fishermen 1
during the period 1950-75. Outliers were removed (screened) before anaylsis following the procedures described by Doerzbacher et |
al. (1988).
TL (mm) at
TL (mm) at
Time free
Growth rate
No.
No. in
No.
release
recapture
(days)
(mm/day)
Mean
Mean
Mean
Mean
Species
Bay system
tagged
analysis
screened
Range
(SD)
Range
(SD)
Range
(SD)
Range
(SD)
Black drum
All bays
28,423
254
6
160-750
317
(101)
175-965
373
(116)
4-4143
273
(467)
-1.167-1.438
0.187
(0.125)
Red drum
Galveston
1370
73
2
155-620
342
(104)
241-762
453
(123)
2-1079
204
(200)
-0.500-1.667
0.624
(0.395)
San Antonio
1272
101
4
220-720
397
(96)
220-915
506
(120)
11-2432
204
(259)
-0.679-1.847
0.569
(0.343)
Aransas
3061
435
7
175-615
360
(85)
230-838
473
(107)
2-784
206
(169)
-0.378-1.729
0.565
(0.365)
Corpus Christi
835
58
4
185-520
322
(93)
280-762
462
(123)
3-692
199
(142)
0-1.686
0.733
(0.396)
Upper
2857
147
5
133-693
426
203-774
544
6-831
250
0.600-1.526
0.416
Lagima Madre
(121)
(110)
(177)
(0.316)
Lower
2202
56
13
151-685
326
171-1016
440
11-5078
412
-0.274-0.938
0.395
Laguna Madre
(122)
(146)
(824)
(0.267)
Sheepshead
All bays
6530
56
6
200-555
313
(74)
210-555
336
(75)
1-630
148
(119)
-0.085-0.779
0.167
(0.209)
Southern
All bays
3176
21
0
255-505
337
250-560
394
1-546
197
0-0.647
0.223
flounder
(78)
(84)
(169)
(0.192)
Spotted
All bays
20,517
357
25
192-762
373
192-762
406
1-1315
173
-0.786-1.220
0.171
seatrout
(90)
(98)
(196)
(0.276)
bays was about 840-950 mm TL, whereas the sheeps-
head estimate was about 470 mm (Table 2). Daily
growth coefficients (K) were about 0.0005 (0.183
annualized) for black drum, southern flounder, and
spotted seatrout, and about 0.001 (0.365 annualized)
for red drum and sheepshead (Table 1). The 1960s
estimates of L^ for black drum, red drum, sheepshead,
and spotted seatrout in Texas were generally higher
than comparable estimates made in the 1980s. Red
drum L^ in the 1960s ranged from 879 mm in the up-
per Laguna Madre to 1177 mm in the Aransas Bay
system; L^^ was 918mm in the 1980s (Doerzbacher et
al. 1988). Values for black drum, sheepshead, and
spotted seatrout were as follows (1960s vs. 1980s): 844
mm vs. 798mm (Doerzbacher et al. 1988); fork length
(FL) 478 mm vs. 419 mm (males) and 447 mm (females)
(Beckman et al. 1991); and 836mm vs. 691mm (Green
et al. 1990), respectively. No estimates were available
for southern flounder in the 1980s.
Red drum growth varied among bays. Estimates of
L^ for red drum in each bay system approximated
930 mm, except in Aransas Bay where L^ was 1177
mm, and K (annualized) varied between 0.3 and 0.5.
Reasons for the interbay variation in L^ and K for red
drum in the 1960s are unknown. However, factors
affecting growth (e.g., fishing mortality, food supply,
red drum density, and environmental conditions like
salinity and temperature) varied among bays (Matlock
1984).
The estimated values of L^^ for black drum and red
drum from fish tagged in the 1960s and 1980s appear
to be underestimates because the data include few adult
fish which reside mostly in the Gulf of Mexico (Matlock
1987, 1991). The addition of older adults would prob-
ably increase L^ and reduce K for both species, but
the change in parameter estimates would depend on
the average maximum age and size actually reached
relative to the largest fish included in the analysis.
Parameter estimates (standard error) for the von Ber-
talanffy model for black drum (0-58 years old) growth
in Florida were 1172mm (±9mm) and 0.124mm
( ± 0.0003 mm), respectively. When the von Bertalanffy
growth equation was fit to length and age (from
otoliths) data for adults off Louisiana, the estimate for
L^ was lOOOmmFL (Beckman et al. 1991); recall,
L^ for Texas black drum was 844mmTL. However,
Beckman et al. (1991) questioned the biological signif-
icance of their L^ estimates because an asymptotic
410
Fishery Bulletin 90(2). 1992
Table 2
Estimates of parameters (daily K and L„) in the von Bertalanffy growth equation for five fishes tagged
in Texas bays during the period 1950-75 (A^ = number of fish used in analysis). Approximate standard errors
(SE) were estimated using ten-fold validation (Verbyla and Litvaitis 1989). Annualized K and associated
SE were estimated by multiplying daily K and daily SE by 365 days. Measure of effectiveness (P) reflects
how well the von Bertalanffy model fit the data (Phares 1980).
Species
Bay system
N
K(±l SE)
Loo (mm)
(±1 SE)
P
(%)
Daily
Annual
Black drum
All bays
254
0.00048
(0.000039)
0.175
(0.014)
844
(40)
77.7
Red drum
Galveston
73
0.00116
(0.000118)
0.423
(0.043)
900
(61)
91.7
San Antonio
101
0.00112
(0.000119)
0.409
(0.043)
978
(77)
88.3
Aransas
435
0.00075
(0.000036)
0.274
(0.013)
1177
(33)
90.2
Corpus Christi
58
0.00138
(0.000210)
0.504
(0.077)
940
(82)
90.0
Upper Laguna Madre
147
0.00127
(0.000085)
0.464
(0.031)
879
(27)
90.0
Lower Laguna Madre
56
0.00075
(0.000221)
0.274
(0.810)
957
(88)
87.0
Sheepshead
All bays
56
0.00098
(0.000289)
0.358
(0.105)
478
(36)
43.9
Southern flounder
All bays'
21
0.00063
(0.000066)
0.230
(0.024)
848
(32)
80.4
Spotted seatrout
All bays 357
a (successive estimates differ by
0.00045
(0.000040)
<2xl0"'^) was
0.164 836
(0.015) (36)
not met after 25 iterations.
63.0
* Convergence criten
size was not attained within the size range sampled and
growth was practically linear beyond age 5. Further,
neither the von Bertalanffy nor power model accurately
described the growth of black drums younger than age
5. A similar result was found for red drum when the
von Bertalanffy model was fit to data from fish from
the Gulf of Mexico (Beckman et al. 1989). The estimates
for L^ were 909mm FL for males and lOlSmmFL for
females, but estimates for K (0.137 for males and 0.088
for females) were smaller than published estimates
based primarily on young fish (Beckman et al. 1989).
They suggested that separate models may be necessary
to describe growth of young red drum from estuarine
areas or old fish from offshore.
The estimates for spotted seatrout growth are prob-
ably more accurate than those for the other four spe-
cies. The sample size is large, and fish of all sizes are
well represented in the data set, including adult spotted
seatrout which generally reside in the bays (Perret et
al. 1980). Estimates for L^ and K using published
length-at-age data collected from spotted seatrout in
the Gulf of Mexico sporadically during 1929-84 were
655 mm and 0.2mm, respecitvely (Condrey et al. 1985).
The L^ estimate for southern flounder (848mmTL)
may be an overestimate, whereas L^ for sheepshead
(478mmTL) may be an underestimate. State records
for southern flounder and sheepshead caught in Texas
salt waters are 711mm and 641mm, respectively
(Anonymous 1989). Reasons for the apparent bias are
unknown but may be related to the few recaptures of
tagged southern flounder (21 fish) and the few large
sheepshead recaptured. Only one sheepshead was
>500mmTL.
Citations
Anonymous
1989 Texas state fish records. Tex. Parks Wildl. Dep. PWD-
L-9000-5-3/89, Austin, 1 p.
Beckman, D.W., A.L. Stanley. J.H. Render, and C.A. Wilson
1990 Age and growth of black drum in Louisiana waters of
the Gulf of Mexico. Trans. Am. Fish. Soc. 119:537-544.
1991 Age and growth-rate estimation of sheepshead Archo-
sar(fus probatoeephalus in Louisiana waters using otoliths.
Fish. Bull., U.S. 89:1-8.
NOTE Matlock: Growth of five fishes tagged in Texas bays in the 1960s
411
Beckman, D.W., G.R. Fitzhugh, and C.A. Wilson
1988 Growth rates and validation of age estimates of red drum,
Sciaenops ocellatus, in a Louisiana salt marsh impoundment.
Contrib. Mar. Sci. (Suppl.) 30:93-98.
Beckman, D.W., C.A. Wilson, and A.L. Stanley
1989 Age and growth of red drum, Sciaenops ocellatus, from
offshore waters of the northern Gulf of Mexico. Fish. Bull.,
U.S. 87:17-28.
Colura, R.L., C.W. Porter, and A.F. Maciorowski
1984 Preliminary evaluation of the scale method for describ-
ing age and growth of spotted seatrout (Cynoscion nebulosus)
in the Matagorda Bay system, Texas. Manage. Data Ser. 57,
Tex. Parks Wildl. Dep., Coastal Fish. Br., Austin, 17 p.
Condrey, R.E., G. Adkins, and M.W. Wascom
1985 Yield-per-recruit of spotted seatrout. Gulf Res. Rep.
8(l):63-67.
Cornelius, S.E.
1984 Contribution to the life history of black drum and analysis
of the commercial fishery of Baffin Bay, Vol. II. Tech. Rep.
6, Caesar Kleberg Wildl. Res. Inst., Kingsville, TX, 241 p.
Doerzbacher, J.F., A.W. Green, and G.C. Matlock
1988 A temperture compensated von Bertalanffy growth
model for tagged red drum and black drum in Texas bays.
Fish. Res. (Amst.) 6:135-152.
Fabens, A.J.
1965 Properties and fitting of the von Bertalanffy growth
curve. Growth 25:265-289.
Ferguson, M.O.. A.W. Green, and G.C. Matlock
1984 Evaluation of the accuracy and precision of volunteered
size data from tagged red drum returns. N. Am. J. Fish.
Manage. 4:181-185.
Green, A.W., L.W. McEachron, G.C. Matlock, and H.E. Hegen
1990 Use of abdominal streamer tags and maximum-likelihood
techniques to estimate spotted seatrout survival and growth.
Am. Fish. Soc. Symp. 7:286-292.
Green, L.
1986 Fish tagging on the Texas coast, 1950-1975. Manage.
Data Ser. 99, Tex. Parks Wildl. Dep., Coastal Fish. Br., Austin,
206 p.
Matlock, G.C.
1984 A basis for the development of a management plan for
red drum in Texas. Ph.D. diss., Texas A&M Univ., College
Station, 287 p.
1987 Maximum total length and age of red drum off Texas.
Northeast Gulf Sci. 9:49-52.
1990 The life history of red drum. In Chamberlain, G.W., R.J.
Miget, and M.G. Haby (compilers). Red drum aquaculture, p.
1-22. TAMU-SG-90-603, Texas A&M Univ. Sea Grant Coll.
Prog., College Station.
1991 Maximum total length and age of black drum off Texas.
Northeast Gulf Sci. 11:171-174.
Moreau, J.
1987 Mathematical and biological expression of growth in
fishes: Recent trends and further developments. In Sum-
merfelt, R.C., and G.E. Hall (eds.), The age and grovrth offish,
p. 81-113. Iowa State Univ. Press, Ames.
Murphy, M.D., and R.G. Taylor
1989 Reporduction and growth of black drum, Pogonias cromis,
in northeast Florida. Northeast Gulf Sci. 10:127-137.
Parrack, M.L.
1979 Aspects of brown shrimp, Penaeus azteeus, growth in the
northern Gulf of Mexico. Fish. Bull., U.S. 76:827-836.
Ferret, W.S., J.E. Weaver, R.O. Williams, P.L. Johansen,
T.D. Mcllwain, R.C. Raulerson, and W.M. latum
1980 Fishery profiles of red drum and spotted seatrout. Gulf
States Mar. Fish. Comm. 6, Ocean Springs, MS, 60 p.
Phares, P.L.
1980 Temperature associated grovrth of white shrimp in Loui-
siana. NOAA Tech. Memo. NMFS-SEFC-56, Southeast Fish.
Sci. Cent., Galveston, TX, 19 p.
Saila, S.B., C.W. Recksiek, and M.H. Prager
1988 Basic fishery science programs, a compendium of micro-
computer programs and maual of operation. Dev. Aquacult.
Fish. Sci. 18, Elsevier Sci. Publ. Co., NY, 178 p.
Sundberg, P.
1984 A Monte Carlo study of three methods for estimating the
parameters in the von Bertalanffy growth equation. J. Cons.
Cons. Int. Explor. Mer 41:248-258.
Verbyla, D.L. and J. A. Litvaitis
1989 Resampling methods for evaluating classfication accuracy
of wildlife habitat models. Environ. Manage. 13:783-787.
A mortality model for a population In
which harvested Individuals do not
necessarily die: The stone crab
Victor R. Restrepo
University of Miami, Rosenstiel School of Marine and Atmospheric Science
Cooperative Institute for Marine and Atmospheric Studies
4600 Rickenbacker Causeway, Miami, Florida 33149
Stone crabs Menippe mercenaria
support a valuable commercial fish-
ery in the Gulf of Mexico, with most
of the catch occurring near south-
west Florida. Florida landings in-
creased from about 400,000 lbs per
fishing season (15 October- 15 May)
in the early 1960s to an average
2.7 million lbs since 1988. The 1990
landings were valued at over $15
million.
The stone crab fishery is unique
in that only the crabs' claws can
be harvested, provided that the
claws are of legal size (70 mm in
propodus length); declawed crabs
must be returned to the ocean.
Stone crabs can regenerate their
massive claws which contain much
of the crabs' edible meat (a large
claw can weigh 250 g). In a sense,
stone crabs are a "reusable re-
source" because claw regeneration
by previously declawed crabs ac-
counts for 1-10% of the annual
landings (Savage et al. 1975, Ehr-
hardt and Restrepo 1989).
The main difficulty associated
with estimating mortality rates in
this unique fishery is that existing
models are not applicable to the
crabs' population dynamics. Tradi-
tional fisheries models are usually
based on the equation (see Beverton
and Holt 1957),
dN
dt
(F + M)Nt,
where N is the population size, t is
time, and F and M are the instan-
taneous fishing and natural mortal-
ity rates. An implication of this
model is that all harvested animals
die. Up to 50% of harvested stone
crabs may survive, depending on
fishing practices such as the amount
of time animals are exposed to air
and on the extent of the injury
caused by declawing (Davis et al.
1979). Therefore, the above model
is not appropriate for this fishery or
others like it. In this paper I develop
a mortality model that accounts for
the possibility that harvested indi-
viduals may survive. The model can
be used to estimate fishing mortal-
ity rates for stone crabs.
The model
Consider a closed population of large-
sized individuals (large enough to
lose both claws to fishing upon cap-
ture), in which catches are moni-
tored for a short period of time.
This time-period should be suffi-
ciently short to ensure that de-
clawed crabs will not have time to
regenerate their claws. Claw regen-
eration in large stone crabs takes
one year or more (Restrepo 1990),
so this should not be a major con-
straint. The population dynamics
during this time-period can be
modeled by subdividing the pop-
ulation into harvestable and unhar-
vestable crabs (those with and with-
out legal-sized claws, respectively).
Harvestable crabs may become un-
harvestable if they survive fishing;
unharvestable crabs may not be-
come harvestable within the time-
interval, since it is assumed that
claw regeneration does not occur.
Let
^'N, "N = population sizes (in
numbers) of harvest-
able and unharvestable
crabs,
F = rate of capture (assum-
ed to be the same for
both types of crabs),
^M, "M = natural mortality rates
for harvestable and un-
harvestable crabs, and,
S = fraction of harvestable
crabs that survive claw
removal and release (0
<S<1).
For simplicity, assume that har-
vest and natural mortality rates re-
main fixed during the time-period.
Note also that unharvestable crabs
are immediately returned to the
water upon capture so that their
mortality due to capture is negli-
gible. The differential equations
describing the two-compartment
model are
dt
- (hM-HFS-HF(l-S))hNt
= - (hM-(-F)hN,, and (la)
d"Nt
= - "M"Nt-HFShNt. (lb)
Equation (la) is the standard mor-
tality model and simply shows that
crabs disappear from the population
due to fishing and natural mortali-
ty. Losses due to fishing are F'^Nt.
Of these, a fraction (1 - S) actually
die, and a fraction (S) become part
of the unharvestable population
(Eq. lb). Thus, F is a true fishing
mortality only when S = 0.
Equation (la) has the general
solution
^Nt = hNoe-(f-'M)t_
(2)
Manuscript accepted 11 March 1992.
Fishery Bulletin, U.S. 90:412-416 (1992).
412
NOTE Restrepo: Estimating fishing mortality rates for Menippe mercenana
413
where 'iNq is the population size at the beginning of
the time-period (t = 0). This solution can be substituted
into Equation (lb) to solve it since, without claw re-
generation, ''Nt is independent of "Nt (i.e., by assump-
tion, there is no transfer from the nonharvestable into
the harvestable population):
d"Nt
dt
"M"Nt-HFShNo e-<F *'")'.
(3)
This is a first-order linear differential equation that
can be solved with the integrating factor
g/"Mdt _ g"Mt
Multiplying (3) throughout by this factor gives
dt
= FS^Nn e-<F+'''^-"'^".
Integrating and letting "Nt = "No at t = 0 gives the
solution to Equation (3):
"Nt
(4)
FSi^No (i_e-(f'+''M-"M)t) e-"Mt
F-t-hM-"M
-t- "No e -"■«'.
The djrnamics explained by Equations (2) and (4) de-
pend on six parameters (^Nq, "Nq, ''M, "M, S, and F).
The main usefulness of these two equations lies in
simulation modeling (e.g., for yield-per-recruit anal-
yses) in which parameters are given as inputs rather
than estimated from fitting the equations to data.
However, as shown below, the number of parameters
can be reduced to three by taking the ratio R = "Nt/
^'Nt . Note that the ratio of the two population types
is largely independent of the level of sampling inten-
sity, provided that the availabilities of harvestable
and unharvestable crabs to the sampling gear do not
change. Obtaining estimates of either "Nt or ''Nf alone
would be a more difficult task which could involve tag-
ging or detailed survey statistics (see Seber 1982 for
a discussion on the estimation of abundance). Dividing
Equation (4) by Equation (2) gives
Figure 1
Expected trends in the ratio R (number of nonharvestable:
harvestable stone crabs Menippe mercenaHa) for three values
of b in Eq. (5). Lower curve: b is negative and the trend ex-
hibits convex curvature. Middle curve: b approaches zero and
the trend is a straight line. Upper curve: b is positive and the
trend is concave-upwards.
b = F-i-'^M-"M, and
FS
c =
F-i-hM-"M
Rt
a e
bt
- c,
(5)
Equation (5) shows that in a closed population and
in the absence of claw regeneration, the ratio of non-
harvestable to harvestable crabs should change expo-
nentially with a concave, convex, or straight trend
depending on the value of b. Consider a hypothetical
population in which hNo = 1000, "No = 0.0, S = 0.5, F =
0.2 per month, and 'iM = 0.2 per month. When "M =
0.6, b= -0.2 , and R increases with convex curvature
(Fig. 1, filled squares). When "M approaches 0.4, b
approaches zero, and R increases linearly (Fig. 1,
crosses). When "M = 0.2 ("M = 'iM), b is positive and the
trend in R is a concave-upwards curve (Fig. 1, aster-
isks). In practice, some of the model's assumptions may
not be always met. For instance, if "M = 0.2 per month
and 250 crabs recruit to the harvestable stock at the
beginning of months 5, 6, and 7, then the trend in R
decreases starting in month 5, while fishing is still
ongoing (Fig. 2).
where R^ =
''Nt'
a =
FS
"No
F-i-hM-"M hN '
Application of tfie model to a data set
No studies have been carried out in which the data
necessary for the model have been collected. For this
reason the estimates presented below are meant to
414
Fishery Bulletin 90(2). 1992
Figure 2
Trend in the ratio R (number of nonharvestable:harvestable
stone crabs Menippe mercenaris) when the assumption of a
closed population is not met. The value of b is positive, as in
the upper curve in Fig. 1, but recruitment of harvestable crabs
occurs in months 5, 6, and 7.
illustrate how the model can be applied. The data set
I used (Sullivan 1979) was collected during 1975 and
1976 in an area where the fishery has been traditionally
most intense. This data set contains detailed informa-
tion on every individual captured, including carapace
size, claw sizes, and claw status (presence/absence,
regeneration stage, etc.).
The first step in analyzing the data is to define ex-
actly how to categorize crabs in order to meet the
assumption that unharvestable crabs do not become
harvestable due to claw regeneration during the study
period. One way to do so is as follows (Restrepo 1990):
"Harvestable" crabs are those with two normal, legal-
sized claws (normal claws are defined as those that have
no signs of regeneration); "nonharvestable" are those
without claws.
With this definition, all harvestable crabs that are
caught will likely lose both claws and hence become
part of the nonharvestable population if they survive.
Conversely, crabs without any claws will not quickly
become part of the harvestable population because it
would take several regenerative molts (years) before
their claws looked normal. Note that the definition
above excludes from the analysis all crabs that have
either one or two sublegal claws which could, through
molting, become part of the harvestable population. In
terms of meeting the model's assumptions, the above
definition still poses a problem in that crabs with only
one claw of legal size (which are relatively uncommon),
whether normal or not, may become part of the non-
harvestable stock upon declawing.
Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct
Figure 3
Observed values of the ratio of male stone crabs Menippe
mercenaria without claws to those with two normal (not
regenerated) claws, by size-group. Data from Sullivan (1979).
Figure 3 shows the observed trend in the ratio R [0
claws: 2 normal claws] of male stone crabs from Sul-
livan's (1979) data, for several size-groups. (Female
crabs are also harvested, but they are excluded from
this analysis because few of them reach sizes at which
both claws are of legal size.) Note that R is relatively
constant and near zero for crabs < 90 mm in carapace
width (CW) (Fig. 3). Based on claw size— carapace
width relationships (Restrepo 1990), the smaller of the
crabs' claws (the "pincer") becomes harvestable only
when the carapace reaches 90 mm in width. Thus males
with two normal claws are not expected to lose both
claws to fishing at sizes <90mmCW, a fact which is
corroborated by Figure 3. Otherwise, R values for the
smaller crabs would show larger deviations from the
zero line in Figure 3. For this reason, the analyses were
conducted with crabs >90mmCW (Fig. 4).
The trend in the observed R values (Fig. 4, filled
squares) is reminiscent of that in Figure 2: it appears
to increase concavely upwards from November to Feb-
ruary, suggesting that "M<''M-t-F, and it then de-
creases starting in March. This decline is possibly a
consequence of recruitment of large crabs with normal
claws into the fishing grounds, suggesting a failure of
the closed-population assumption. Empirical evidence
for a similar recruitment peak of large males in the
spring was found by Ehrhardt et al. (1990) in Ever-
glades National Park. In addition to recruitment, the
decline in R after February could also be attributed to
declawed crabs being removed from the study site in
greater numbers after this month (some vessels may
remove the claws at the end of the day as they travel
from the fishing grounds to port). Because of these
NOTE Restrepo: Estimating fishing mortality rates for Menippe mercenarid
415
Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct
Month (1975 1976)
Figure 4
Grouped values from Fig. 3 for crabs >90mmCW (symbols).
Solid line: fit of Eq. (5) t» the first four observations (estimates
of fitted parameters are given in Table 1).
Table 1
Parameter estimates and correlation matrix obtained by fit-
ting Equation (5) to Sullivan's (1979) data on stone crab
Menippe mercenaria.
Parameter Estimate
SE
Correlation with
a b c
a 0.191
b 0.618
c 0.175
0.175
0.252
0.229
1.000
-0.994 1.000
0.972 -0.949 1.000
problems, only the first four data points in the series
were used to estimate the parameters in Equation (5).
The least-squares parameter estimates, standard
errors and correlation matrix are given in Table 1 (fit
shown in Fig. 4). Note that the errors and correlations
are extremely high, owing to the small amount of data
used (4 data points to estimate 3 parameters). From
these estimates, the following are obtained:
"Nn
= a - c = 0.016
hNo
FS = c b = 0.109
F + hM - nM = b = 0.618.
Thus by fitting Equation (5) to the data, the relevant
parameters that are estimated are the initial ratio
(which could be measured directly, anyway) and the
product FS. Due to the indeterminacy in the esti-
mates of FS and F-i-*'M-"M, auxiliary information is
required before the harvest rate (F) can be deduced.
One possibility is to assume that the natural mortalities
of clawed and clawless crabs are identical. If so, then
F = b and S = c. This would result in F = 0.618/month
and S = 17.6%; instantaneous mortality due to fishing
would be F(l-S) = 0.51/month. If "M>hM as sug-
gested by Bert et al. (1978), then the value of F will
be less than that estimated by b. Note that F = b =
0.618/month translates into about 4.3/year (during a
7-month fishing season), which appears to be unrea-
sonably high, giving indirect support to Bert et al.'s
(1978) contention.
Conclusions
An application of the model to estimate current ex-
ploitation rates has not been carried out. As shown in
the previous section, the data required to do so are
relatively simple (crab size, number of claws, and type
of claws) but cannot be obtained from the fishery
landing statistics. Therefore, a research sampling pro-
gram would have to be set up to monitor the popula-
tion and obtain an estimate using the model. Such a
sampling program should give consideration to the
following requirements:
1 The areal coverage should be large enough to en-
sure that declawed crabs are not removed from the
study site by the fishing vessels.
2 Time-periods when recruitment, immigration, or
emigration take place should not be included in the
analyses.
3 Counts of both harvestable and nonharvestable in-
dividuals should be made periodically (e.g., weekly) so
the counts represent the number of individuals at a par-
ticular time, rather than the average number of in-
dividuals during a long time-period as was done in the
preceding section.
4 To avoid imprecision and parameter correlations
such as those in Table 1, a large number of R values
should be available for parameter estimation.
Acknowledgments
I am grateful to Nelson Ehrhardt, David Die, and Clay
Porch for their discussions and to an anonymous
reviewer for helpful comments. Special thanks are due
to the Marine Research Lab of the Florida Department
of Natural Resources for allowing me the use of their
data. Financial support for this study was provided by
416
Fishery Bulletin 90(2), 1992
the Florida Sea Grant College Program under project
No. R/LR-b-24, and through the Cooperative Institute
for Marine and Atmospheric Studies by the National
Oceanic and Atmospheric Administration Cooperative
Agreement No. NA90-RAH-0075.
Citations
Bert, T.M., R.E. Warner, and L.D. Kessler
1978 The biology and Florida fishery of the stone crab Menippe
mercenaria (Say) with emphasis on southwest Florida. Fla.
Sea Grant Tech. Rep. 9, Univ. Fla. Sea Grant Prog., Gaines-
ville, 82 p.
Beverton, R.J.H., and S. Holt
1957 On the dynamics of exploited fish populations. Fish. In-
vest. Minist. Agric. Fish. Food U.K. (Ser. 2), 19, 533 p.
Davis, G.E., D.S. Baughman, J.D. Chapman, D. MacArthur, and
A.C. Pierce
1979 Mortality associated with declawing stone crabs, Menippe
mercenaria. Rep. SFRC T-552, Natl. Park Serv., South Fla.
Res. Cent., Homestead, 23 p.
Ehrhardt, N.M., and V.R. Restrepo
1989 The Florida stone crab fishery: A reusable resource? In
Caddy, J.F. (ed.), Marine invertebrate fisheries: Their assess-
ment and management, p. 225-240. Wiley, NY.
Ehrhardt, N.M.. D.J. Die, and V.R. Restrepo
1990 Abundance and impact of fishing on a stone crab (Menippe
■mercenaria) population in Everglades National Park, Florida.
Bull. Mar. Sci. 46:311-323.
Restrepo, V.R.
1990 Population dynamics and yield-per-recruit assessment of
southwest Florida stone crabs, Menippe mercenaria. Ph.D.
thesis, Univ. Miami, Coral Gables, 225 p.
Savage, T., J.R. Sullivan, and C.E. Kalman
1975 An analysis of stone crab (Menippe mercenaria) landings
on Florida's west coast, with a brief synopsis of the fishery.
Fla. Dep. Nat. Resour. Mar. Res. Lab. Publ. 13, St. Petersburg,
37 p.
Seber, G.A.F.
1982 The estimation of animal abundance, 2d ed. Charles
Griffin, London.
Sullivan, J.R.
1979 The stone crab, Menippe mercenaria, in the southwest
Florida fishery. Fla. Dep. Nat. Resour. Mar. Res. Publ. 36,
St. Petersburg, 37 p.
Optimal course by dolphins
for detection avoidance
Carlos A.M. Salvadd
Pierre Kleiber
Andrew E. Dizon
Southwest Fisheries Science Center, National Marine Fisheries Service, NOAA
P.O. Box 271, La Jolla, California 92038-0271
One of the assumptions of line tran-
sect sampling is that movement of
animals being counted is not in re-
sponse to the approaching vessel
before the animals are detected
(Burnham et al. 1980). By observ-
ing from a helicopter the reaction of
dolphins to an approaching survey
vessel, Au and Ferryman (1982) and
Hewitt (1985) demonstrated that
dolphin schools can detect the ap-
proach and maneuver to attempt
to avoid detection. Because it may
be that dolphin exhibit forms of
optimal behavior (Au and Weihs
1980), it is of interest to determine
whether there is a direction the
dolphin should take that maximizes
their distance to the vessel at the
point of closest approach and, if
there is such a direction, to deter-
mine whether dolphin use it. If this
is so, this may be the way of deter-
mining through aerial means when
dolphin first react to an approach-
ing vessel and whether it is after
they are detected by a shipboard
observer.
Since the advent of purse-seine
fishing in the eastern tropical Pacif-
ic in 1959, dolphin that associate
with yellowfin tuna (i.e., primarily
Stenella attenuata, S. longirostris,
and Delphinus delphis) are chased.
track of vessel
Figure 1
The vessel traveling at speed Vg and the dolphin traveling
at V[) as seen in the stationary frame of reference. The direc-
tion a is the one chosen by the dolphin. Distance f is the perpen-
dicular distance between the initial position of the dolphin and
the projected path of the vessel, while a + f is the distance
between vessel and dolphin when the dolphin is abeam.
caught in nets, and sometimes
drowned (Perrin 1968, 1969).
Stuntz and Perrin (1979) reported
that these species of dolphin are
more difficult to capture in areas
where purse-seine-vessel fishing
effort has been greatest, imply-
ing that evasive behavior may be
learned. It has also been reported
by Au and Ferryman (1982) that
evasive maneuvers by dolphin upon
approach of a vessel sometimes
begin at a distance that is approx-
imately the shipboard observer's
horizon. Because the visual horizon
of even a leaping dolphin is shorter
than that of a shipboard observer,
it is likely that they are reacting to
the vessel sound. It is therefore
plausible that by experiencing re-
peatedly the approach of such ves-
sels, dolphin not only have learned
to evade but do so optimally by
choosing through trial and error the
direction of escape, if it exists, in
which the noise amplitude increases
the least. Because the attenuation
of sound is proportional to the dis-
tance transversed by it, escaping
from a sound source in the direction
where the amplitude increases the
least is the same direction that max-
imizes the distance between a uni-
formly-moving source and receiver
at the point of closest approach.
Here, we formulate the following
problem: Upon detecting the ap-
proach of a vessel, a dolphin at-
tempts to avoid detection by re-
treating. If the velocity of the vessel
is Vb and that of the dolphin is Vp,
is there a direction in which a dol-
phin can escape to maximize its
distance from the vessel at the point
of closest approach (Fig. 1)? And if
so, what direction is it? We will
show that there is such a direction:
If a is the angle between Vb and
Vd, the angle a = arccos(VDA'^B).
where Vg and Vd are, respectively,
the speeds of the vessel and the
dolphin, will maximize the distance
Manuscript accepted 5 February 1992.
Fishery Bulletin, U.S. 90:417-420 (1992).
417
418
Fishery Bulletin 90(2). 1992
between vessel and dolphin at the point of closest
approach.
Finally, to ease the task of data analysis by whoever
makes the necessary observations, we have derived the
expressions that relate dolphin speed and direction to
their range and bearing from the vessel. In a cartesian
coordinate system let V^ and Vy be, respectively, the
X and y components of the dolphin velocity minus,
respectively, the x and y components of the vessel
velocity. Both Vx and Vy are constructed using range
and bearing measurements of the dolphin from the
vessel. Then VD = [(Vx + VB)2 + Vy2]i'2 and a = arctan
[Vy/(V, + VB)].
Problem solution
As stated in the formulation above, the problem makes
sense only for the case Vd<Vb. With reference to
Figure 2, to maximize the distance between vessel and
dolphin at the point of closest approach, we must find
the maximimi value of r^^j^ with respect to angle 0 <
r» X
'. actual track
'. of dolphin
Figure 2
The apparent motion of the dolphin as seen in the moving
frame of reference of the vessel. The apparent velocity V of
the dolphin is the resultant of the vector addition of -Vg
and Vp . Distance f is the perpendicular distance between the
initial position of the dolphin and the projected path of the
vessel, while a + f is the distance between vessel and dolphin
when the dolphin is abeam.
For the purpose of the following exposition, we
define initial position to be that position of the vessel
(or the dolphin) at the time when the dolphin detects
the approaching vessel and begins evasion. Initial time
is the time corresponding to the initial position.
In Figure 2, at the point of closest approach the
distance between vessel and dolphin is given by
Tmin = (a + f)COS/J,
(1)
with respect to a simultaneously).
In Figure 1, c is the distance along the projected path
of the vessel between the vessel's initial position and
the point that is abeam of the dolphin's initial position.
Let t(. be the time it takes the vessel to transverse
distance c, and t that time from the initial time until
the vessel has the dolphin abeam. From the applica-
tion of the Pythagorean Theorem we can deduce
where 0</3<n/2, f is the perpendicular distance be-
tween the initial position of the dolphin and the pro-
jected path of the vessel, and (a-t-f) is the distance
between vessel and dolphin when the dolphin is abeam.
The vector diagram of Figure 2 shows that the ap-
parent track of the dolphin as seen from the vessel is
a function of a, the direction of escape of the dolphin.
Therefore, to solve the problem as posed, we must find
the extreme value of Eq. (1) with respect to angle a.
By computing the derivative with respect to a of Eq.
(1), we find that r^i„ is rendered an extreme value
when
dfmin „^^ „ da
= cos p —
da da
(a-hf)sin/?^
d«
(2)
vanishes. Depending on the functional dependence of
a and ft on a, an equation of the form of Eq. (2), could
vanish either term by term or by cancellation of the
terms. For the former case, each term could vanish
trivially (i.e., a and p are independent of a), or non-
trivially (i.e., both a and p are rendered extreme values
Because
(VDt)2 = a2 + [VB(t-t,)]^
a = Vpt sin a,
(3)
(4)
it can then be shown by substitution of Eq. (4) into Eq.
(3) that
*" (5)
t =
l-(VD/VB)cosa
By substituting Eq. (5) into Eq. (4), we find that as a
function of a, a is given by
a =
Vote sin a
1-(Vd/Vb)cosc»"
(6)
Because at least a is a function of a, we can conclude
that in general Eq. (2) does not vanish trivially. How-
ever, p is also a function of a as can be deduced by the
application of the Law of Sines to Figure 2:
NOTE Salvado et al.: Dolphin detection avoidance
419
arctan
Vd\ sin a
Vb l-(VDA^B)cosa
(7)
Next we investigate whether Eq. (2) vanishes non-
trivially. Computing the derivative with respect to a
of Eq. (6), we get
da ^ VpteLcos g-CVp/VB)]
da [l-(VD/VB)cosa]2
which vanishes only if
/Vol
arccos — = a a.
Vb
(8)
(9)
Noting the similarity between a and p given respec-
tively in Eq. (6) and (7), we can easily and simply ex-
press one in terms of the other. Writing
p = arctan
V^t
B^-c
(10
the derivative of p with respect to a is given by
dp cos^ p da
da Vstc da
(11)
Substituting into Eq. (2) the equivalence of Eq. (11),
we find that
dr
da
= cos p
1 -
(a-i-f) sin p cos p
da
d^'
(12)
which allows us to conclude that r^jn is extreme at the
point a = ao because, as we found in Eq. (8), a is ex-
treme at that point. However, as can be appreciated
in Eq. (12), there may be other points where rn^^ is ex-
treme. We now investigate whether r^^jn is extreme
for values of a other than ag.
Other extreme values of r^jn may be attained if
or
cos /? = 0
(a-i-f) sin p cos p = VBte
(13)
(14)
are physically realizable. Eq. (13) is satisfied for p =
nn/2 (n= 1,3,5,...), and of these only p = n/2 concerns
us. It corresponds to the upper limit of the physically
realizable range of 0^p<n/2, and represents the un-
interesting case Vd = Vb (i.e., ao = 0) which is the
trivial special case of this problem: r^^^ is rendered
constant when Vd = Vb (i.e., the dolphin swims with
the same velocity as the vessel), so the vanishing of Eq.
(2) is trivially satisfied.
In Eq. (14) let f=0. This limit will not diminish the
generality of the solution. The direction a dolphin
should take to maximize its distance to the vessel at
the point of closest approach will not depend on how
close the dolphin is initially to the projected path of the
vessel. So without loss of generality we investigate if
there is a physically realizable angle p such that
a sin p cos p = VBte
(15)
is satisfied. With the result from Eq. (10) and the iden-
tity tan p = sin p cos" ' p, Eq. (15) can be expressed as
the condition sin /3 = ± 1 which is satisfied by the same
uninteresting case that satisfies Eq. (13). Therefore,
we can conclude that the nontrivial extreme value
attained by rmin at a = ao is unique in the interval
0<a<7i.
We only have left to show that the extreme value at-
tained by r^in at a = ao is a maximum. Let ao and pQ
be the respective values of a and ^ at a = ao . The sec-
ond derivative of r^jn with respect to a evaluated at
a = ao is given by
d^r,
min
da^
"0
(16)
cos Po
1 -
ao sin /?o cos ^o
VBt
d^a
da2
"o"
Because the second derivative of a with respect to a
evaluated at a = ao is
d^a
da2
-Vnt
Dl-c
[1-(Vd/Vb)2]3/2
<o,
(17)
to determine whether Eq. (16) is negative we must
show that
ao sin p cos Pq < VBte
(18)
We have shown already that Eq. (15) can only be sat-
isfied for an angle /J = n/2. Then Eq. (18) is satisfied for
0</}<n/2. We have seen already that /? = n/2 when a = 0,
so da/da = 0 at that point also. Therefore, we can also
conclude that the nontrivial extreme value achieved by
r^in at a = ao, unique in the interval 0<a<n, is a max-
imum. Because Eq. (2) vanishes term by term, the same
result is achieved by finding the extreme value with
respect to a of either /3 or a.
In conclusion, a dolphin escaping at speed Vp at an
angle a relative to the velocity Vb of an approaching
vessel will maximize its distance to the vessel at the
point of closest approach if a = arccos (Vd/Vb).
420
Fishery Bulletin 90(2). 1992
Determination of dolphin velocity
from range and bearing measurements
In this section we relate the dolphin velocity to prac-
tical in situ measurements. We will derive a relation-
ship between dolphin speed and direction to its range
0 < r < oo and bearing 0 < 6< 27i from the vessel that trig-
gers the dolphin to flight.
For the times {tji i = l,2,...,n} we perform the cor-
responding measurements {ri,Gi. i = l,2,...,n}. It is not
necessary that the measurements be made from the
vessel, but with reference to it (e.g., aerial measure-
ments). However, it is necessary that there be no other
vessel in the vicinity that perturbs the measurements
by reaction of the dolphin to its presence.
The measurements of range and bearing of the
dolphin from the vessel are equivalent to the cylindrical
coordinates of the dolphin with respect to the moving
frame of reference of the vessel. The cartesian coor-
dinates -oo<x<oo and -oo<y<oo with respect to the
same frame of reference are determined from
X = r cos 9 and y = r sin 9.
(19)
Let Ax, Ay, and At be, respectively, the increments of
the variables x, y, and t. For each of the (n-1) con-
secutive intervals, we can compute the average speeds
in the x and y directions by
Ax , ^, Ay
V^ = — and Vv = — .
At
At
(20)
These are the components of the dolphins' apparent
velocity V in the frame of the moving vessel (Fig. 2).
We can express the dolphin velocity in the moving
frame as a function of its speed and direction in the
stationary frame by
Vx = Vd cos a - Vb and Vy = Vq sin a, (21)
which are a system of two coupled, nonlinear equations
with Vd and a as unknowns. The solution to this set
of equations is given by
With these results we can compare the (n-1) time-
intervals the direction a taken by the dolphin given Eq.
(23), with the optimal direction given in Eq. (9) com-
puted from the result given Eq. (22).
Acl<nowledgments
We gratefully acknowledge the valuable suggestions
by Dave Au and Bill Perrin, and a correction of an
erroneous definition by Tim Gerrodette.
Citations
Au, D., and W. Ferryman
1982 Movement and speed of dolphin schools responding to
an approaching ship. Fish. Bull., U.S. 80:371-379.
Au, D., and D. Weihs
1980 At high speeds dolphins save energy by leaping. Nature
284(5756):548-560.
Burnham, K.P., D.R. Anderson, and J.L. Laake
1980 Estimation of density from line transect sampling of
biological populations. Wildl. Monogr. 72, 202 p.
Hewitt, R.P.
1985 Reaction of dolphins to a survey vessel: Effects on cen-
sus data. Fish. Bull., U.S. 83:187-193.
Perrin, W.F.
1968 The porpoise and the tuna. Sea Frontiers 14:166-174.
1969 Using porpoise to catch tuna. World Fishing 18(6):
42-45.
Stuntz, W.E.. and W.F Perrin
1979 Learned evasive behavior by dolphins involved in the
eastern tropical Pacific tuna purse seine fishery. In Abstracts
from presentations at the Third Biennial Conference of the
Biology of Marine Mammals, Seattle, Oct. 7-11, 1979, p.
58. [Avail. Library, Natl. Mar. Mammal Lab., Seattle 98115.]
Vd = VV, + Vb + V/, (22)
where we have chosen the positive root, and
V„
a = arctan
Vx + Vb
(23)
Effects of microprobe precision
on hypotheses related to
otolith Sr:Ca ratios
Christopher L. Toole
Department of Fisheries and Wildlife. Oregon State University
104 Nash Hall. Corvallis, Oregon 97331-3803
Present address: Environmental and Technical Services Division
National Marine Fisheries Service. NOAA
911 NE 1 1th Street. Suite 620. Portland. Oregon 97232
Roger L. Nielsen
College of Oceanography. Oregon State University
Ocean Administration 104. Corvallis. Oregon 97331-5503
Several recent studies have used
the electron microprobe to infer
environmental temperature at the
time of otolith formation from the
concentration ratio of strontium
and calcium. Sr/Ca ratios of otoliths
from fish held at constant tempera-
ture or collected at known temper-
ature were examined using atomic
absorption spectrophotometry
(Radtke 1984, 1989) or wavelength
dispersive electron microprobe an-
alysis (Townsend et al. 1989, Kalish
1989, Radtke et al. 1990). These
studies, with the exception of Kalish
(1989), concluded that there is a
negative linear relation between en-
vironmental temperature and oto-
lith Sr/Ca ratio. This relationship,
coupled with assignment of age to
each microprobe sample site, has
been used to infer the relative tem-
perature histories of wild-caught
fish (Radtke 1984,1987,1989;
Radtke and Targett 1984; Radtke
and Morales-Nin 1989; Townsend et
al. 1989; Radtke et al. 1990). The
most ambitious application of the
method used otolith Sr/Ca ratios to
contrast the calculated temperature
histories of different subpopula-
tions of larval herring in the Gulf of
Maine (Townsend et al. 1989).
Using the electron microprobe to
calculate individual fish tempera-
ture histories from otolith Sr/Ca
ratios is potentially a useful tech-
nique for fisheries biologists. How-
ever, precision of back-calcuJated
temperature estimates should be
examined in greater detail. Previ-
ous studies do not explicitly state
confidence limits for mean re-
sponses or prediction limits for new
observations. The scatter of points
in Radtke (1989), Townsend et al.
(1989), and Radtke et al. (1990)
suggest that widths of 95% pre-
diction limits may be on the order
of one to several °C for most levels
of Sr/Ca examined. While this
might be acceptable for studies of
fish which are exposed to wide
variations in environmental tem-
perature, it is of less use for species
which experience more subtle tem-
perature changes.
Future validation experiments
may improve the predictive capabil-
ities of the Sr/Ca vs. temperature
relationship by examining effects
of other variables. For instance,
the regression model might be ex-
panded to include growth rate (Ka-
lish 1989) and some measure of
physiological stress (Townsend et
al. 1989), since these also appear to
influence the Sr/Ca ratio.
However, one component of the
variation not likely to change in fu-
ture experiments employing the
electron microprobe is the model er-
ror term associated with measure-
ment. Usually measurement error
is considered insignificant in rela-
tion to other sources of variation
and is incorporated into the total
error term:
Y = a + b*X + exotai
where Gxotal = ^Measurement + ^Other ■
Measurement error can be thought
of as a lower bound to the varia-
tion associated with the regression
model when other sources of error
are minimized.
We suspect that measurement
error may be nontrivial when de-
riving Sr/Ca vs. temperature rela-
tionships. Sr/Ca ratios associated
with a 1°C change in environmen-
tal temperature were approximate-
ly 0.00013-0.00036 in previous
studies (Table 1). It is difficult to
evaluate the significance of these
small values without more informa-
tion on the analytical precision of Ca
and Sr detection in fish otoliths
using the electron microprobe. Of
the studies cited above, only Kalish
(1989) reported analytical precision
for representative values of Sr and
Ca. In that study, measurement er-
ror associated with Sr was 3.5% and
that associated with Ca was 0.5%
for an Sr/Ca ratio of 0.002.
One purpose of the present study
was to examine the precision asso-
ciated with measuring Sr/Ca ratios
in fish otoliths, and to demonstrate
how this error affects temperature
estimates derived from published
regressions. Our approach was to
intensively sample one otolith from
one fish at three beam-power den-
sities and four counting times. By
using one otolith, between-fish ef-
fects could be ignored. Within-fish
Sr/Ca effects were minimized by
referencing samples to the same
growth zones, leaving the different
analytical techniques as the primary
source of variation.
Manuscript accepted 15 January 1992.
Fishery Bulletin, U.S. 90:421-427 (1992).
421
422
Fishery Bulletin 90(2), 1992
Published Sr/Ca
Table 1
vs. temperature relationships.
Source and
species examined
Equation
Radtke et al. (1990)
Clupea harengus
Atlantic herring
T = 19.172 - 2.955 • (Sr/Ca* 1000)
1°C = 0.000338 (Sr/Ca)
Sr/Ca range 0.002-0.0045
Townsend et al. (1989)
Clupea harengus
Atlantic herring
T = 12.6 - 2.81 (Sr/Ca* 1000)
1°C = 0.000356 (Sr/Ca)
Sr/Ca range 0.001-0.0045
Radtke (1989)
Fundulu^ lieteroclitus
Mummichog
(Sr/Ca* 1000) = 16.371 - 0.219 * T
1°C = 0.000219 (Sr/Ca)"
Sr/Ca range 0.009-0.013
Radtke (1984)
Gadus morhua
Atlantic cod
(Sr/Ca* 1000) = 4.19 - 0.13 * T"
rC = 0.000130 (Sr/Ca)"
Sr/Ca range 0.0028-0.0038
r/Ca on temperature (T) will also predict T on Sr/Ca.
Radtke (1984), but later reported in Kalish (1989).
"Assumes that slope of S
■^'Not stated explicitly in
the otolith formed under natural con-
ditions (inside the OTC band) were
analyzed.
An otolith from a randomly selected
fish was mounted on a slide with a tolu-
ene-based medium. It was ground using
600-grit paper along the saggital plane
to a level near the central primordium.
The mounting medium was then melted,
the otolith was removed, washed, and
remounted on its opposite side with
heat-setting epoxy. The second side was
then ground to the central primordium
and polished with a series of diamond
and alumina grits, ending with 0.05^m
alumina. The specimen was cleaned
ultrasonically in detergent and water
between grit changes and given final
rinses in water and methanol. Prior to
microprobe analysis, the specimen was
carbon coated.
A second purpose was to determine the effect of
beam exposure on the constancy of Sr/Ca ratios. This
was necessary because analytical techniques, such as
increasing the counting time, will improve the preci-
sion of an analysis but may reduce its accuracy through
beam damage to the specimen (e.g.. Smith 1986, Potts
1987). This problem is encountered in the analysis of
other carbonates, but is particularly severe for otoliths,
which contain organic material in addition to CaCOs
(Degens et al. 1969). CO2 is lost during electron beam
exposure and, because it is not actually measured by
the microprobe but assumed to occur on a 1:1 basis
with cations such as Sr and Ca, concentrations of
those elements wall increase with increasing beam
damage. However, if Sr and Ca are not fractionated
from one another by beam damage, their ratio should
remain unchanged. Absence of change would indicate
that methods which improve precision can be imple-
mented without affecting the accuracy of Sr/Ca ratio
determinations.
Methods
Dover sole Microstomus pacificus is a common Pacific
coast flatfish. Juvenile Dover sole 54-104 mmSL were
captured by trawling off the Oregon coast on 17 March
1990 and immediately injected with oxytetracycline
(OTC). Within 12 hours, fish were transferred to aqua-
ria in Corvallis, Oregon, where they were held for up
to 48 days. The OTC produced a fluorescent band which
delineated growth prior to capture from subsequent
grovrth under laboratory conditions. Only portions of
Beam power density and precision
Wavelength-dispersive electron microprobe analysis
was performed with a Cameca SX-50 microprobe with
a 40° beam angle. Three levels of beam-power density
were obtained by varjang the beam diameter while
holding accelerating voltage and beam current constant
at 15 KV and 20 nA, respectively. These voltage and
current settings are common to most of the previous
studies (R. Radtke, Hawaii Inst. Geophys., Univ.
Hawaii, Honolulu 96822, pers. commun. 1990), al-
though Kalish (1989) used a lOnA current. Defocused
beam diameters of 5, 7, and lO^^m resulted in beam-
power densities of 1.019, 0.520, and 0.255 nA/fim. The
most common beam diameter used in previous studies
was 5^im (R. Radtke, pers. commun. 1990), although
Kalish (1989) rastered a 12.5fimx 12.5jim square.
Sr and Ca concentrations were calculated as nor-
malized mole fractions (equivalent to the atomic ratios
of Kalish 1989). Mole fractions are more informative
than weight percentages for examination of Sr/Ca
ratios, since the substitution of Sr for Ca in otolith
aragonite theoretically occurs on a per-atom basis (e.g.,
Radtke 1989). Normalization also reduces effects of
beam damage on concentrations.
Precision of elemental measurements was deter-
mined as the coefficient of variation (CV) (Williams
1987),
CV = Ok.ratio^k-ratio
where the k-ratio is the ratio of x-ray counts from the
otolith to those of the standard (i.e., the calibrated frac-
tion of that element in the otolith) and o^.ratio is the
NOTE Toole and Nielsen: Microprobe precision associated with SrCa ratios
423
standard deviation of that measurement. For a single microprobe analysis,
this is calculated as
CV =
i=i n
n \ 2
0.5
\ n /
Np +
N,
(Np-Nb)
Otolith
Standard
0.5
where n = number of samples taken on the standard,
Nj = x-ray count (corrected for background count) from ith sample
on the standard,
Np = x-ray count for peak wavelength of element in sample,
Nb = x-ray counts from background wavelengths of element in sample,
tp = peak wavelength counting time, and
tb = background wavelength counting time.
Approximate 95% confidence limits for each element measured in each
sample were considered ± 2 * CV, since the Poisson distribution underlying
these calculations approximates a normal distribution when sample size (the
number of x-rays detected by the spectrometer during an analysis) is high
(Williams 1987). X-ray counts in this experiment were on the order of
102-103 for Sr and lO^-lO^ for
Ca. Confidence limits for the
Sr/Ca ratio were also calculated
as ± 2 * CV, but in this case the
standard deviation of the k-ratio
was calculated as
OSr/Ca =
OSr
k-ratiosr
OCa
k-ratioca
0.5
Sr and Ca were analysed using
the TAP (Sr L-a) and PET (Ca K-
a) crystals. Background counts
were taken at ±(0.005* sin 6)
(where G is the angle of the spec-
trometer crystal when it is detec-
ting peak counts) for the same
length of time as the peak count.
Due to interference with a sec-
ond-order Ca K-a peak, only one
background count was made for
Sr. Strontianite (NMNH R10065)
and calcite (USNM 136321) were
used as standards.
30 sec
10,5,10,7nm
20 sec
5,7, 10 Mn
100 urn
m
Figure I
Photomicrograph of otolith from 65.7 mm SL juvenile Dover sole Microstomus pacificus,
showing location of 12 microprobe transects used for analysis. Each circular area
represents one analysis. Note hyaline area near central primordium at inner end of
transects and more opaque area towards outer end. The 13th transect was an acciden-
tal repetition of the lOfim, 30-sec transect. Bar indicates lOOjim.
Counting time and precision
Counting time refers to the
length of time a spectrometer is
collecting counts of character-
istic x-rays for an element during
one analysis. Counting times of
10, 20, 30, and 40 sec were com-
pared for each beam-power den-
sity. The most commonly used
counting time for both elements
in previous studies was 20 sec (R.
Radtke, pers. commun. 1990), al-
though Kalish (1989) analyzed Sr
at 100 sec and Ca at 20 sec. Pre-
cision was determined as with
beam-power density.
Transects of twelve analyses
each were made for the 12 com-
binations of beam power density
(4) and counting time (3) (com-
bined Af= [12*3*4] =144). These
transects passed from an area
near the central primordium to
an area just inside the discontinu-
ity created by accessory primor-
dia (Fig. 1). This discontinuity
424
Fishery Bulletin 90|2). 1992
1.0
0.8
0.6
0.4
0.2
y—
o
o
X
99.8
7-
99.6
o
99.4
1-
( )
99.2
<r
tr
99.0
u.
98.8
Ml
_i
O
s
1.0
0.8
0.6
0.4
0.2
CALCIUM
Sr/Ca RATIO
2 4 6 8 10 12
TRANSECT POSITION
Figure 2
Example of changes in Sr and Ca concentration and the Sr/Ca
ratio along one of 12 transects made on a Dover sole otolith.
A Tfjm beam and 20-sec counting time were used. Position
1 is the point closest to the central primordium, and position
12 is closest to the otolith edge.
was >100nm inside the OTC mark. Starting and end-
ing points for all transects were referenced to specific
growth areas identified by dark continuous bands, and
the remaining points were evenly spaced between these
two points. Locations at the start of the transects were
in a translucent area of the otolith assumed to have
little organic material (Dannevig 1955), while the end
points were in a more opaque area, which probably con-
tained more organic material.
Exposure time and accuracy
Counting time and exposure time were distinguished
in this experiment. Counting time is the minimum time
the specimen is exposed to the electron beam, while
exposure time also includes the time necessary to col-
lect background counts and counts of other elements.
Six sequential analyses were made at each of six loca-
tions (combined N 36) to determine changes in elemen-
tal concentration. The locations were the start and end-
points of each 20-sec transect used for the precision
analysis. Sequential analyses at each location were
45
4
3.5
3
25
2
1,5
0 02 04 0,6 08 1 1.2 1.4
Sr MOLE FRACTION (K 0 01)
Figure 3
Relation of coefficient of variation (measurement error) to
elemental concentration for transects from one Dover sole
otolith. Each regression represents three transects with 12
points each (A^ 36). Regression equations are presented in
Table 2.
Table 2
Relationship between Sr concentration and coefficient of varia- |
tion (CV) for different counting
times, based
on microprobe
transects along
the saggital plane of a Dover
sole Microsto-
7nus pacifictLS
otolith. Each equation was
derived from
three transects
of twelve points
each {N 36). Equations are
in the form: CV = exp(A + (B* S
r mole fraction)). Standard
errors in parentheses.
Seconds
A
B
R'
10
1.546
-59.261
0.940
(0.012)
(2.558)
20
1.150
-47.156
0.938
(0.011)
(2.076)
30
0.941
-44.454
0.947
(0.010)
(1.806)
40
0,789
-42.135
0.938
(0.010)
(1.863)
made in increments of 20-sec counting times, which cor-
responded to exposure times of 65, 130, 195, 260, 325,
and 390 sec. These exposure times were approximate-
ly twice as long as those which would result from an
analysis of Sr and Ca alone, because S was also ana-
lyzed (results not reported).
Statistical analyses
The effect of elemental concentration on Sr precision
(CV's) was examined with linear and nonlinear regres-
sions. To determine if beam-power density affected Sr
precision, multiple regressions containing normalized
concentration and "dummy variables" corresponding
to beam size were analyzed with partial-F tests (Neter
et al. 1989:364-370). Each of the four counting times
was analyzed separately.
NOTE Toole and Nielsen: Microprobe precision associated with Sr:Ca ratios
425
Widths of 95% confidence intervals associated with
Sr/Ca ratios were determined with linear and nonlinear
regressions for each counting time. The ratios and
widths of confidence intervals were then converted to
temperatures using the four previously published Sr/Ca
vs. temperature regressions in Table 1.
The effect of exposure time on Sr/Ca constancy was
analyzed with a multiple regression containing ex-
posure time and each location (coded as O's and I's) as
independent variables. Locations were included to re-
move possible effects of initial Sr/Ca concentrations,
which varied between sites. After determining that in-
teractions and nonlinear terms did not improve a model
with parallel straight lines, the common slope was com-
pared with a slope of 0 using a <-test.
Whenever the null hypothesis could not be rejected
at a = 0.05, statistical power (1-/3) of the test was
calculated as in (Neter et al. 1989:74-75). The power
of a test was considered acceptable if (l-/3)>0.80
(Peter man 1990).
Results
The twelve transects made under different beam con-
ditions on the otolith of a 65.7 mm SL juvenile Dover
sole exhibited consistent patterns of strontium and
calcium concentrations. Sr concentrations were highest
at the two innermost positions and lowest at the two
outermost positions in all transects. Sr/Ca ratios mir-
rored the pattern of Sr. Ca concentrations were ap-
proximately 100-500 times higher than Sr concentra-
tions. An example of one of the 12 transects (7/im beam
at 20-sec counting time) is presented in Figure 2.
Relative error of Sr measurements decreased as
counting time and elemental concentrations increased,
and this was best described by an exponential regres-
sion model (Fig. 3, Table 2). The coefficient of varia-
tion was 1.4-4.2% for Sr concentrations of 0.2-1.2%.
When the effect of elemental concentration was re-
moved, Sr CV's increased with decreasing beam-power
density (Table 3); however, this effect was small com-
pared with those of elemental concentration and count-
ing time. Differences in Sr CV's attributable to beam-
power density was 0.012-0.076%.
The coefficient of variation associated with Ca mea-
surements was 0.5% for 10- and 20-sec counts and 0.4%
for 30- and 40-sec counts, regardless of Ca concentra-
tion and beam-power density.
Regressions of the widths of 95% confidence inter-
vals for Sr/Ca determinations against measured Sr/Ca
ratios are presented in Figure 4 and Table 4. These
regressions include only the effects of elemental con-
centration and counting time; the effect of beam-power
density is omitted. Although relative error decreases
Table 3
Relationship between Sr coefficient of variation (CV) and
beam power density, holding Sr concentration as a nonlinear
covariate, based upon microprobe transects along the saggital
plane of a Dover sole otolith. Each regression represents three
transects with 12 points each (A^ 36). Equations are in the
form; CV = A; -i- AoZj + AjZ, + B,* (Sr mole fraction) + B,« (Sr
mole fraction)^ where A is the intercept for the 10 ^m beam
(0.255 nAJiim density), A + AoZ, is the intercept for the T^m
beam (0.520 nA/jim density), and A + A3Z2 is the intercept for
the 5^im beam (1.019 nAJ^im density); Zj and Z, are dummy
variables for the 7 and 5fjm beams; and B, and B, are fitted
slope parameters. Partial-F tests indicate the significance of
beam power density effects in the model. Standard errors in
parentheses.
Counts
Parameter
10-sec
20-sec
30-sec
40-sec
A,
' (adj.)
(0 05.2,31)
4.990
(0.101)
-0.013
(0.027)
-0.076
(0.027)
-389.80
(40.88)
17940.3
(3996.70)
0.954
4.479
0.020
3.394
(0.043)
-0.019
(0.018)
-0.055
(0.018)
-219.41
(13.49)
7746.5
(921.47)
0.969
4.833
0.015
2.767
(0.051)
-0.027
(0.017)
-0.052
(0.017)
-169.07
(15.73)
5386.3
(959.47)
0.959
4.531
0.019
2.420
(0.041)
-0.039
(0.014)
-0.053
(0.014)
-151.79
(12.81)
5229.6
(796.05)
0.962
8.443
0.001
_l
< 0 12
10 SEC y 20 SEC /
DC
/ ^/
UJ
/ /^
^ 0 10
/t] yu 30 SEC
LU
/ y^ y"^
0
/ / u^"^ y'^'^
Z 0.08
m r^ y^^\
111
i§ft n^ ^y .jc^
Q _
X7 .-v^ r>^y^^^
U. 0
2 d 0.06
m htP^ y\y^^^ ^^ ^^^
0 ^
vW ^^ y^C-^
0
m^^^^^yC^
S? 0.04
^ _^^^^^
m
/ jjy^^^SWffilJF'
05
/ Jot^^''^
LL
/ y/\y^^^
0 0.02
- /yyy^
I
y^
H
Q
§
1 ... 1 ... 1 ... 1 ... 1 ... 1 ... r ... 1 ... 1
0 0 2 0 4 0 6 0 8 10 12 1.4 1.6
(X0 01)
Sr/Ca RATIO
Figure 4
Relationship between Sr/Ca ratios and 95% confidence inter-
vals of measurements at different counting times, based upon
microprobe transects along the saggital plane of one Dover
sole otolith. Each regression represents three transects of 12
points each (N 36). Regression equations are presented in
Table 4.
426
Fishery Bulletin 90(2). 1992
Table 4
Relationship between Sr/Ca ratios and 95% confidence interval
of measurements at different counting times, based upon
microprobe transects along the saggital plane of a Dover sole
otohth. Each equation was derived from three transects of
12 points each (AT 36). Form of the relationship is: 95%
CI = A + (B* Sr/Ca ratio). Standard errors in parentheses.
Seconds
B
R^
10
20
30
40
1.791E-4
0.1024
0.9861
(9.829E-6)
(0.0021)
1.551E-4
0.0669
0.9823
(7.990E-6)
(0.0015)
1.477E-4
0.0508
0.9817
(6.579E-6)
(0.0012)
1.166E-4
0.0482
0.9888
(4.843E-6)
(0.0009)
with increasing Sr/Ca, the actual width of the confi-
dence interval increases. Conversion of Sr/Ca ratios
and 95% confidence limits to temperatures, using the
20-sec regression and previously published temperature
vs. Sr/Ca ratios, is presented in Figure 5. Confidence
limits associated with calculated temperatures were
0.6-4.7°C, depending upon species, study, and tem-
perature level.
The model which best fit the six multiple exposures
is presented in Figure 6. The common slope of - 1.3*
10 '' was not different from a slope of 0 (fo.05,29 =
0.176, P = 0.86). This experiment could have detected
a change as small as 1.86*10-'' Sr/Ca per sec in-
creased exposure (or 1.21*10''' for each 65-sec treat-
ment) at 0 = 0.05 and (1 -/3) = 0.90, had such an effect
existed.
Discussion
Our results confirm that measurement error asso-
ciated with Sr/Ca determinations is nontrivial. At
the standard counting time of 20 sec, measurement
error associated with Sr/Ca determinations (expressed
as 95% confidence intervals) was equal to or greater
than the Sr/Ca increment representative of a 1°C tem-
perature change in three of the four previously-pub-
lished studies. Even in Townsend et al. (1989), at tem-
peratures <4°C, measurement error was >1°C. The
highest measurement error in the studies was repre-
sentative of a 4.7°C temperature change. Inferred
temperature differences between otolith regions or be-
tween fish should be considered in light of these values.
Statistical error in the Sr/Ca vs. temperature regres-
sions will add to the measurement error associated with
temperature calculations.
<
>
LU
O
LU
Q
o
LO
CD
O
I
I-
9
5
RADTKE (1989)
MUMMICHOG
RADTKE (1984)
ATLANTIC COD
RADTKE ETAL (1990)
ATLANTIC HERRING
TOWNSEND ET AL (1989)
ATLANTIC HERRING
TEMPERATURE (C)
Figure 5
Relationship between back-calculated temperature estimates
and 95% confidence intervals (for measurement error only)
surrounding those estimates. Temperature vs. Sr/Ca conver-
sions are from Table 1. Confidence intervals are converted
from Sr/Ca confidence intervals for 20-sec counts in Table 4
and Figure 4.
400
EXPOSURE TIME (SEC)
Figure 6
Relationship between Sr/Ca level and exposure time, based
upon microprobe samples at six sites on one Dover sole otolith.
Six sequential analyses were made at each site (N 36). The
six exposure times corresponded to counting times of 20, 40,
60, 80, 100, and 120 sec. The equation describing the rela-
tionship is: Sr/Ca level = 0.0023 -f (0.0054*Z, ) -^ (0.00026*2, )
-^(0.0114•Z.^) + (6.6xl0-*•ZJ■^(0.012•Z5)-(1.30xl0-'•
exposure time), where Z,-Z5 are dummy variables for loca-
tions. Adjusted fi- = 0.991, P<0.0001.
NOTE Toole and Nielsen: Microprobe precision associated with Sr Ca ratios
427
At least 40-sec counts would be necessary to detect
a 1°C change in temperature experienced by herring
(Townsend et al. 1989, Radtke et al. 1990) at all tem-
perature levels examined in those studies. Detection
of a 1°C temperature change in cod (Radtke 1984) and
Fundulus (Radtke 1989) would require much longer
counting times, beyond the range examined in this
experiment.
This experiment documents the improvement in
precision which is possible when otoliths are analyzed
at longer counting times and higher beam power den-
sities. Neither treatment appeared to affect the level
of Sr/Ca accuracy under the range of conditions ex-
amined. Obvious burns on the otolith (Fig. 1) indicate
that beam damage did occur in all of our experimental
treatments, and we suspect that it also occurred in
other studies using similar analytical conditions. How-
ever, whatever effect this may have had on the ac-
curacy of the molecular weight percent concentrations
for Ca and Sr, the ratio of the two elements remained
constant, indicating no observable fractionation.
One implication of these results is that Sr/Ca preci-
sion can be increased, with no apparent loss of ac-
curacy, when analyses are conducted for 40-sec rather
than 20-sec counting times, and at 5/.im rather than
10 (im beam sizes, at an accelerating voltage of 20 nA.
The 5pim beam allows greater temporal resolution,
which is helpful when matching the sample location to
structures such as daily growth increments. These may
be as small as 0.1-0.2^m, depending upon species,
growth rate, and age (Campagna and Neilson 1985).
Because the level of precision may affect conclusions
of studies relating otolith Sr/Ca levels to environmen-
tal temperature, it is important to know the analytical
conditions under which each study is conducted.
Minimal information required includes beam current
and voltage, beam size, counting time for each element,
standards used, and precision of measurements. This
information has not been reported in sufficient detail
in some of the previous studies, making interpretation
difficult. The methods described in this experiment are
proposed as a means of defining measurement preci-
sion in future studies of Sr/Ca ratios in fish otoliths.
Acknowledgments
We are grateful to Douglas Markle for his advice and
support. Comments by John Kalish on an earlier draft
greatly improved the final version. This work was
funded in part by the Oregon Sea Grant Program,
Project No. NA85AA-D-SG095; Pacific Outer Con-
tinental Shelf Region of the Minerals Management Ser-
vice, Department of Interior Contract No. 14-12-0001-
30429; and the Oregon State University Research Of-
fice. We also thank Capt. Terry Thompson of the FV
Olympic for donation of a portion of the ship time.
Citations
Campagna, S.E., and J. Neilson
1985 Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci.
42:1014-1032,
Dannevig, E.H.
1955 Chemical composition of the zones in cod otoliths. J.
Cons. Cons. Int. Explor. Mer 21:156-159.
Degens, E.T., W. Deuser. and R. Haedrich
1969 Molecular structure and composition of fish otoliths.
Mar. Biol. (Berl.) 2:105-113.
Kalish, J.M.
1989 Otolith microchemistry: Validation of the effects of phys-
iology, age, and environment on otolith composition. J. Exp.
Mar. Biol. Ecoi. 132:151-178.
Neter, J., W. Wasserman, and M. Kutner
1989 Applied linear regression models. Irwin, Boston, 667 p.
Peterman, R.M.
1990 Statistical power analysis can improve fisheries research
and management. Can. J. Fish. Aquat. Sci. 47:2-15.
Potts, P.J.
1987 A handbook of silicate rock analysis. Chapman & Hall,
NY, 622 p.
Radtke, R.L.
1984 Cod fish otoliths: Information storage structures. In
The propagation of cod Gadus morhua L. Flodevigen Rapp.
1:273-298.
1987 Age and growth information available from the otoliths
of the Hawaiian snapper, PristipoTnoidesfilamentosus. Coral
Reefs 6:19-25.
1989 Strontium-calcium concentration ratios in fish otoliths
as environmental indicators. Comp. Biochem. Physiol. 92A:
189-193.
Radtke, R.L., and B. Morales-Nin
1989 Mediterranean juvenile bluefin tuna: Life history pat-
terns. J. Fish Biol. 35:485-496.
Radtke, R.L., and T. Targett
1984 Rythmic structural and chemical patterns in otoliths of
the Antarctic fish Notothemia larsoni : Their application to age
determination. Polar Biol. 3:203-210.
Radtke, R.L., D. Townsend, S. Folsom, and M. Morrison
1990 Strontiuin:calcium concentration ratios in otoliths of her-
ring larvae as indicators of environmental histories. Environ.
Biol. Fish. 27:51-61.
Smith, M.P.
1986 Silver coating inhibits electron microprobe beam damage
of carbonates. J. Sedimentary Petrol. 56:107-108.
Townsend, D.W., R. Radtke, M. Morrison, and S. Folsom
1989 Recruitment implications of larval herring overwinter-
ing distributions in the Gulf of Maine, inferred using a new
otolith technique. Mar. Ecol. Prog. Ser. 55:1-13.
Williams, K.L.
1987 An introduction to X-ray spectrometry: X-ray fluores-
cence and electron microprobe analysis. Allen & Unwin, Lon-
don, 370 p.
Order Processing Code:
*5178
Charge your order.
It's Easy!
Superintendent of Documents Subscriptions Order Form
n YES, enter my subscripUon as follows: To fax your orders (202) 512-2233
subscriptions to FISHERY BULLETIN (FB) for $24.00 per year ($30.00 foreign).
Masrertoraj
The total cost of my order is $.
_. Price includes regular domestic postage and handling and is subject to change.
(Company or Personal Name)
(Please type or print)
(Additional address/attention line)
Please Choose Method of Payment:
I I Check Payable to the Superintendent of Documents
□ GPO Deposit Account | | | | | | | l-fl
(Street address)
I I VISA or MasterCard Account
(City, State, ZIP Code)
(Daytime phone including area c»de)
1 hunk you Jur
(Credit card expiration date) , ,
^ ' ' your order!
(Purchase Order No.) ^es no
May we make your name/address avaflable to other mailers? r~| | |
(Authorizing Signature) "
Mail To: New Orders, Superintendent of Documents
PO. Box 371954, Pittsburgh, PA 15250-7954
The National Marine Fisheries Service (NMFS) does not approve, recommend or en-
dorse any proprietary product or proprietary material mentioned in this publication.
No reference shall be made to NMFS, or to this publication furnished by NMFS, in
any advertising or sales promotion which would indicate or imply that NMFS ap-
proves, recommends or endorses any proprietary product or proprietary material
mentioned herein, or which has as its purpose an intent to cause directly or indirect-
ly the advertised product to be us<»d or purchased because of this NMFS publication.
5330 -SS
Fishery Bulletin
Guidelines for Contributors
Form of manuscript
The document should be in the following
sequence: Title Page, Abstract (not re-
quired for Note), Tfext, Acknowledgments,
Citations, Text footnotes. Appendices,
Tables, Figure legends, and Figures.
Title page should include authors' full
names and mailing addresses and the
senior author's telephone and FAX
numbers.
Abstract Not to exceed one double-
spaced typed page. Should include a
sentence or two explaining to the gen-
eral reader why the research was under-
taken and why the results should be
viewed as important. Abstract should
convey the main point of the paper and
outline the results or conclusions. No
footnotes or references.
Text A brief introduction should por-
tray the broad significance of the paper.
The entire text should be intelligible to
readers from different disciplines. All
technical terms should be defined, as
well as all abbreviations, acronyms, and
symbols in text, equations, or formulae.
Abbreviate units of measure only when
used with numerals or in tables and
figures to conserve space. Measure-
ments should be expressed in metric
units, with other equivalent units given
in parentheses. Follow the U.S. Govern-
ment Printing Office Style Manual,
1984 ed., and the CBE Style Manual,
5th ed. Fishery and invertebrate no-
menclature should follow the American
Fisheries Society Special Publication 12
(for fishes), 16 (for moUusks), and i7 (for
decapod crustaceans).
Text footnotes should be numbered
in Arabic numerals and typed on a sep-
arate sheet from the text. Footnotes are
not used for reference material or per-
sonal communications, but rather to
explain or define terms in the text and
for contribution numbers on the title
page.
Informal sources Personal communi-
cations, unpublished data, and untitled
manuscripts in preparation are noted
parenthetically in the text (full name.
affiliation, brief address including zip
code, and month and year when appro-
priate).
Acknowledgments Gather all ac-
knowledgments into a brief statement
at the end of the text. Give credit only
for exceptional contributions and not to
those whose contributions are part of
their normal duties.
Citations All titled sources should be
listed in the Citations section, including
unpublished and processed material. In
text, cite as Smith and Jones (1977) or
(Smith and Jones 1977); if more than
one citation, list chronologically (Smith
1936, Jones 1975, Doe In press). All
sources cited in the text should be listed
alphabetically by the senior authors'
surnames under the heading CITA-
TIONS. Abbreviations of periodicals
and serials should conform to Serial
Sources for the BIOSIS Data Base™.
Indicate whether sources are in a langu-
age other than English. For informal
literature, include address of author or
publisher. Authors are responsible for
the accuracy of all citations.
Tables should supplement, not dupli-
cate, the text. Each table should be
numbered and cited consecutively, with
headings short but amply descriptive so
that the reader need not refer to the
text. For values less than one, zeros
should precede all decimal points. In-
dicate units of measure in column head-
ings; do not deviate from the unit of
measure within a column. Table foot-
notes should be noted consecutively in
Roman letters across the page from left
to right and then down. Since all tables
are typeset, they need not be submitted
camera-ready.
Figures Photographs and line draw-
ings should be of professional quality-
clear and concise— and reducible to 42
picas for full-page width or to 20 picas
for a single-column width, and to a max-
imum 55 picas high. All graphic ele-
ments in illustrations must be propor-
tioned to insure legibility when reduced
to fit the page format. Line weight and
lettering should be sharp and even. Let-
tering should be upper and lower case.
and vertical lettering should be avoided
whenever possible (except for vertical,
y, axis). Zeros should precede all deci-
mal points for values less than one. Re-
productions of line artwork are accepted
in the form of high-quality photographic
prints from negatives or photomechan-
ical transfer (PMT). Halftones should be
sharply focused with good contrast.
Micron rules should be inserted on elec-
tron micrographs, even when magnifi-
cation is included in the figure legend.
There should be clear distinction be-
tween identifying letters (press-on or
overlay) and background of photograph.
Label each figure in pencil on the back.
Send only xerox copies of figures to the
Scientific Editor; originals or photo-
graphic prints will be requested later
when the manuscript is accepted for
publication.
Copyright Government publications
are in the public domain, i.e., they are
not protected by copyright.
Submission of manuscript
Disks Authors are encouraged to re-
tain manuscripts on word-processing
storage media (diskettes, floppy disks)
and submit a double-spaced hardcopy
run from the storage media. Submit
disks as MS-DOS "print" or "non-
document" files (often called "ASCII
files"). If a disk cannot be converted to
an ASCII file, the author should indicate
on the disk the source computer and
software language along with the file
name. Either 5V4-inch or SVs-inch disks
from IBM-compatible or Apple/Macin-
tosh systems (non-graphics only) can be
submitted, double-sided/double-density
or high-density, limiting each file to 300
kilobytes. All 8-inch word-processing
disks (e.g., Wang or NBI) must be con-
verted onto 51/4- or 3V2-inch MS-DOS
print disks.
Send original hardcopy and two dupli-
cated copies to:
Dr. Linda L. Jones, Scientific Editor
National Marine Mammal Laboratory
F/AKC3
National Marine Fisheries Service,
NOAA
7600 Sand Point Way NE
Seattle, WA 98115-0070
Copies of published articles and notes
The senior author and his/her organiza-
tion each receive 50 separates free-of-
charge. Additional copies may be pur-
chased in lots of 100.
MBL WHO! LIBRARY
lilH nuN S
kgg>gg&^ ^d/€^ ^^■^^*^<^ ^^ €^ i<2^>^or^ S^iy^^
U.S. Department
of Commerce
Volume 90
Number 3
July 1992
Fisher
Bulleti
Marine Biological Laboratory
LIBRARY
DEC 91992
Woods Hole, Mass.
U.S. Department
of Commerce
Barbara Hackman Franklin
Secretary
National Oceanic
and Atmospheric
Administration
John A. Knauss
Under Secretary for
Oceans and Atmosphere
National Marine
Fisheries Service
William W. Fox Jr.
Assistant Administrator
for Fisheries
<°^%.
©raDDoGam
Scientific Editor
Dr. Linda L. Jones
Northwest Fisheries Science Center
National Marine Fisheries Service, NOAA
2725 Montlake Boulevard East
Seattle, Washington 981 12-2097
Editorial Committee
Dr. Andrew E. Dizon National Marine Fisheries Service
Dr. Charles W. Fowler National Marine Fisheries Service
Dr. Richard D. Methot National Marine Fisheries Service
Dr. Theodore W. Pietsch University of Washington
Dr. Joseph E. Powers National Marine Fisheries Service
Dr. Tim D. Smith National Marine Fisheries Service
Dr. Mia J. Tegner Scripps Institution of Oceanography
The Fishery Bulletin (ISSN 0090-0656)
is published quarterly by the Scientific
Publications Office, National Marine
Fisheries Service, NOAA, 7600 Sand
Point Way NE, BIN C15700, Seattle, WA
98115-0070. Second class postage is paid
in Seattle, Wash., and additional offices.
POSTMASTER send address changes for
subscriptions to Fishery Bulletin, Super-
intendent of Documents, Attn: Chief,
Mail List Branch, Mail Stop SSOM,
Washington, DC 20402-9373.
Although the contents have not been
copyrighted and may be reprinted entire-
ly, reference to source is appreciated.
The Secretary of Commerce has deter-
mined that the publication of this period-
ical is necessary in the transaction of the
public business required by law of this
Department. Use of funds for printing of
this periodical has been approved by the
Director of the Office of Management and
Budget.
For sale by the Superintendent of
Documents, U.S. Government Printing
Office, Washington, DC 20402. Subscrip-
tion price per year: $24.00 domestic and
$30.00 foreign. Cost per single issue:
$12.00 domestic and $15.00 foreign. See
back page for order form.
Managing Editor
Nancy Peacock
National Marine Fisheries Service
Scientific Publications Office
7600 Sand Point Way NE, BIN CI 5700
Seattle, Washington 981 15-0070
The Fishery Bulletin carries original research reports and technical notes on investiga-
tions in fishery science, engineering, and economics. The Bulletin of the United States
Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries
in 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates
were issued as documents through volume 46; the last document was No. 1 103. Begin-
ning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate
appeared as a numbered bulletin. A new system began in 1963 with volume 63 in
which papers are bound together in a single issue of the bulletin. Beginning with
volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued
quarterly. In this form, it is available by subscription from the Superintendent of
Documents, U.S. Government Printing Office, Washington, DC 20402. It is also
available free in limited numbers to libraries, research institutions, State and Federal
agencies, and in exchange for other scientific publications.
U.S. Department
of Commerce
Seattle, Washington
Volume 90
Number 3
July 1992
Fishery
Bulletin
Contents
Errata
III
429
439
454
469
476
483
494
505
Marine Biological Laooratofy
LIBRARY
DEC 91992
ass.
Baker, C. Scott, Janice
Anjanette Perry
Population characteristics of individually identified humpback whales
in southeastern Alaska: Summer and fall 1986
Bradford, Michael J.
Precision of recruitment predictions from early life stages of
marine fishes
Feeney, Richard F.
Post-yolksac larval development of two southern California sculpins,
Clinocottus analis and Orthonoptas triacis (Pisces: Cottidae)
Graves, John E., Jan R. McDo\vell, and
M. Lisa Jones
A genetic analysis of weakfish Cynoscion regalis stock structure
along the mid-Atlantic coast
Moffitt, Robert B., and Franl< A. Parrish
An assessment of the exploitable biomass of Heterocarpus laevigatas
in the main Hawaiian Islands. Part 2: Observations from a submersible
Polovina, Jeffrey J., and Gary T. Mitchum
Variability in spiny lobster Panulirus marginatus recruitment and
sea level in the Northwestern Hawaiian Islands
Ralston, Stephen, and Darryl T. Tagami
An assessment of the exploitable biomass of Heterocarpus laevigatas
in the main Hawaiian Islands. Part 1 : Trapping surveys, depletion
experiment, and length structure
Reilly, Carol A., Tina Wyllie Echeverria, and
Stephen Ralston
Interannual variation and overlap in the diets of pelagic juvenile
rockfish (Genus: Sebastes] off central California
Fishery Bulletin 90(3), 1992
516 Sadovy, Yvonne, Miguel Figuerola, and Ana Roman
Age and growth of red hind Epinephelus guttatus in Puerto Rico and St. Thomas
529 Sogard, Susan M., Kenneth W. Able, and Michael P. Fahay
Early life history of the tautog Tautoga onitis in the Mid-Atlantic Bight
540 Stein, David L., Brian l\l. Tissot, Mark A. Hixon, and William Barss
Fish-habitat associations on a deep reef at the edge of the Oregon continental shelf
552 Thompson, Grant G.
Management advice from a simple dynamic pool model
561 Thompson, Grant G.
A Bayesian approach to management advice when stock-recruitment parameters are uncertain
574 Walker, H J. Jr., and Keith W. Radford
Eastern Pacific species of the genus Umbrina (Pisces: Sciaenidae) with a description of a new species
588 Weinrich, Mason T., Richard H. Lambertson, Cynthia R. Belt, Mark R. Schilling,
Heidi J. Iken, and Stephen E. Syrjala
Behavioral reactions of humpback whales Megaptera novaeangliae to biopsy procedures
599 Wigley, Susen E., and Fredric M. Serchuk
Spatial and temporal distribution of juvenile Atlantic cod Oadus morhua in the Georges Bank-Southern
New England region
607 Zuhiga, Humberto N., and Enzo S. Acuha
Larval development of two sympatric flounders, Paralichthys adspersus (Steindachner, 1867) and
Paralichthys microps (Gunther, 1881). from the Bay of Coquimbo, Chile
Notes
621 Campbell, R. Page, Terry J. Cody, C.E. Bryan, Gary C. Matlock, Maury F. Osborn,
and Albert W. Green
An estimate of the tag-reporting rate of commercial shrimpers in two Texas bays
625 Edwards, Elizabeth F., and Peter C. Perkins
Power to detect linear trends in dolphin abundance: Estimates from tuna-vessel observer data, 1975-89
ERRATA Fishery Bulletin 90(3)
Errata
(1)
Markle. Douglas F., Phillip M. Harris, and
Christopher L. Toole
Metamorphosis and an overview of early-life-history
stages in Dover sole Microstomus pacificus.
Fish. Bull., U.S. 90(2):285-301.
Table 4 (p. 295) should have reference to A'^ (no. of
trawls) deleted; the text on p. 297-298 correctly refers
to Table 1 for the correct number of trawls each month.
The line art which appears in Figiire 2 (p. 288) should
be replaced by the following (right column):
(2)
Stone, Heath H., and Brian M. Jessop
Seasonal distribution of river herring Alosa pseudo-
harengus and A. aestivalis off the Atlantic coast of
Nova Scotia.
Fish. Bull., U.S. 90(2):376-389.
Figure 1 (p. 377) should be corrected as follows:
Depths should read 100 and 200 m, as the caption
states, rather than 50 and 100m.
(3)
Authorship of the following article should be corrected
to read as follows:
Bartley, Devin, Boyd Bentley, Jon Brodziak,
Richard Gomulkiewicz, Marc Mangel, and
Graham A.E. Gall
Geographic variation in population genetic structure
of chinook salmon from California and Oregon.
Fish. Bull., U.S. 90(1):77-100.
Standard Length (mm)
Standard Length (mm)
Figure 2
Relationship between body depth at anus and standard length
in Dover sole Microstomus pacificus : (A) scatterplot of data
points. Stages 1-5; (B) polygons circumscribing areas bounded
by specimens in Stages 1-5.
Abstract. — In the summer and
fall of 1986 a total of 257 humpback
whales Megaptera novaeangliae were
individually identified during nonsys-
tematic vessel surveys of southeast-
ern Alaska. The majority of adult
animals (n 130. 54.6%) identified in
1986 had been identified previously
in southeastern Alaska during the
years 1979-85. Capture-recapture
estimates suggested that this region-
al subpopulation increased in abun-
dance from 1979 to 1986, and in-
cluded 547 individual whales (95%
CL: 504-590) at the time of the 1986
surveys. An average reproduction
rate of 0.36 calves/mature female-
year-' (95% CL: 0.28-0.43) was es-
timated for this regional subpopula-
tion using individual identification
records collected during 1980-86. In
the Frederick Sound-Stephens Pas-
sage area, the largest number of
whales was found during August and
their predominant prey appeared to
be euphausiids. In the Glacier Bay-
Icy Strait area, the relative abun-
dance of whales was greatest in June
and July and their predominant prey
appeared to be schooling fish. Low
levels of interchange between sur-
veyed areas for much of the summer
season indicated strong preferences
for local habitats among individual
whales. The documented presence of
some individual whales for at least
6 months is evidence that southeast-
ern Alaska is the primary feeding
ground for many of the whales iden-
tified in these surveys.
Population characteristics of
individually identified humpback
whales in southeastern Alaska:
Summer and fall 1986
C. Scott Baker
University of Hawaii, Pacific Biomedical Research Center
Kewalo Marine Laboratory, 41 Ahui Street, Honolulu, Hawaii 96813
Janice M. Straley
Institute of Marine Sciences, University of Alaska, Fairbanks, Alaska 99775
Anjanette Perry
University of Hawaii, Pacific Biomedical Researcfi Center
Kewalo Marine Laboratory, 41 Ahui Street. Honolulu, Hawaii 96813
Manuscript accepted 18 May 1992.
Fishery Bulletin, U.S. 90:429-437 (1992).
Humpback whales Megaptera novae-
angliae in the central and eastern
North Pacific, like those in the west-
em North Atlantic (Katona and Beard
1990), appear to form several geo-
graphically-isolated subpopulations
during the summer and fall feeding
season (Baker et al. 1986, Perry et al.
1990). Following their yearly migra-
tion south, individuals from these
feeding herds intermingle in the
waters of either Hawaii or Mexico
during the winter breeding season
(Darling and Jurasz 1983, Baker et
al. 1985, Darhng and McSweeney
1985, Baker et al. 1986).
The coastal waters of southeastern
Alaska (56-59°N lat.) seem to encom-
pass the primary feeding ground of
a single 'herd' or regional subpopula-
tion estimated to number between
327 and 421 individual whales as of
1983 (Baker et al. 1986). Although
the exact geographic boundaries of
each herd are unknown, whales from
southeastern Alaska appear to re-
main segregated from those that
summer to the west in the Gulf of
Alaska, including Prince William
Sound, and those which summer to
the south along the coast of central
California (Baker et al. 1986, Perry
et al. 1990). Fidelity to a particular
feeding ground appears to be mater-
nally directed (Martin et al. 1984,
Baker et al. 1987, Clapham and Mayo
1987) and may persist across many
generations, as suggested by geo-
graphic segregation of mitochondrial
DNA haplotypes (Baker et al. 1990).
Within southeastern Alaska, how-
ever, the distribution of whales is not
homogeneous and intermingling of
individuals is not random (Baker
1985a, Baker et al. 1985). Some
whales return vdth considerable fidel-
ity to specific areas or 'neighbor-
hoods' such as Glacier Bay, Sitka
Sound or Frederick Sound and, at
least during part of the feeding sea-
son, may establish restricted local
ranges (Jurasz and Palmer 1981,
Perry et al. 1985, Baker et al. 1988,
Straley 1990). Changes in distribu-
tion and local movement v«thin a sea-
son appear to reflect changes in prey
availability. The relatively early ar-
rival of whales into the Glacier Bay
area indicates that this may be an im-
portant area for early-summer feed-
ing on schooling fish, including cape-
lin Mallotus villosus, sand lance
Ammodytes hexapterus, and Pacific
herring Clupea harengus (Wing and
Krieger 1983, Krieger and Wing
1984 and 1986, Perry et al. 1985).
429
430
Fishery Bulletin 90(3). 1992
By late summer, whales typically congregate in
Frederick Sound and Stephens Passage where large
swarms of euphausiids, primarily Thysonoessa raschii
and Euphausii pacifica, are common (Krieger and
Wing 1984, 1986). Some whales feed throughout fall
and early winter in areas such as Seymour Canal and
Sitka Sound where euphausiids and schooling herring
appear to remain available (Baker et al. 1985, Straley
1990).
Here we summarize the results of nonsystematic
surveys of individually identified humpback whales in
southeastern Alaska during the summer and through
late fall of 1986. The 1986 surveys were designed to
overlap in geographic range and seasonal timing with
previous coverage during the years 1979-85 (Baker et
al. 1985, Baker 1985b). In keeping with recommended
management plans (Anonymous 1984), our surveys
documented regional abundance and distribution of
humpback whales in areas that may be impacted direct-
ly or indirectly by vessel activity in Glacier Bay Na-
tional Park. More specifically, we sought to evaluate
trends in the abundance, reproductive rates, and
primary prey of humpback whales in southeastern
Alaska across the years 1979-86. Documentation of
long-term trends in these population characteristics are
valuable for assessing the influences of human activ-
ity, such as mining, logging, or petroleum exploration
and development, or natural environmental fluctua-
tions such as El Niiio events, on the habitat use and'
recovery of this endangered species (National Marine
Fisheries Service 1991).
Methods
Vessel surveys
Humpback whales were observed and individually iden-
tified primarily in two areas or subregions of south-
eastern Alaska (Fig. 1); Glacier Bay and the adjacent
waters of Icy Strait (referred to collectively as Glacier
Bay); and the contiguous waters of Stephens Passage
and Frederick Sound, including Seymour Canal (re-
ferred to collectively as Frederick Sound). Photographs
of whales were also collected in Chatham Strait and
Sitka Sound on an opportunistic basis throughout the
summer and fall.
Whales in Glacier Bay were censused by one of us
(CSB) from 22 May to 10 September under the auspices
of the National Park Service. A total of 42 one-day
surveys were conducted aboard a 17-foot fiberglass
boat powered by a 50-hp outboard motor. The lower
and middle bay (i.e., from Bartlett Cove to the mouths
of Muir Inlet and the West Arm) were surveyed not
less than twice and not more than three times a week.
The mouth of Glacier Bay and the adjacent waters of
Figure 1
Southeastern Alaska region and primary survey areas
(shaded).
Icy Strait were surveyed at least once and not more
than twice a week. Study period and survey coverage
were designed to overlap and extend previous coverage
during the summers of 1982-85 (Baker et al. 1985,
Baker 1985b).
Whales in Frederick Sound were censused during
three summer surveys: 31 July-3 August; 29 August-
1 September; and 12 September- 15 September. These
survey cruises were conducted aboard the RV Sashin,
a 22-foot stern-drive vessel provided by the Auke Bay
Laboratory, National Marine Fisheries Service. Each
cruise originated and ended in Juneau and surveyed
the length of Stephens Passage and Frederick Sound
south to Cape Fanshaw and west to Pybus Bay (see
shaded area, Fig. 1). A fourth survey of Frederick
Sound was conducted from 29 November to 9 Decem-
ber aboard the MV Fairweather, a 43-foot, diesel-
powered cabin cruiser. This cruise originated and ended
in Sitka, Alaska, and surveyed the southern half of
Chatham Strait and Frederick Sound, north to Sey-
mour Canal. The dates and geographic coverage of
Frederick Sound surveys were chosen to coincide with
those of similar previous surveys during the summers
of 1984-85 (Krieger and Wing 1986, CSB unpubl.
data), the fall or winters of 1979-85 (Straley 1990),
Baker et al,: Population characteristics of Megaptera novaeangliae in southeastern Alaska
431
and with field efforts during the summers of 1981-82
(Baker et al. 1985).
Prey assessment
Humpback whale prey species were assessed in Glacier
Bay with a Ross Fineline 250C recording fathometer
equipped with a 22° beam, 105-kHz transducer. In
Frederick Sound, prey were assessed with a Lowrance
recording fathometer equipped with a 250-kHz trans-
ducer. Putative identification of primary prey species
type (e.g., euphausiids vs. schooling fish) was based on
qualitative differences in target strength, as judged
from the relative intensity of fathometer recordings,
and the size, shape, and depth of prey schools. These
interpretations were based on reference to previous
documentation of humpback whale prey using quan-
titative hydroacoustics and net sampling (Wing and
Krieger 1983, Krieger and Wing 1984 and 1986). On
occasion, observations of feces from feeding whales or
the presence of prey species at the surface provided
direct confirmation of primary prey species type.
Individual identification
We attempted to individually identify all humpback
whales encountered by collecting photographs of the
ventral surface of the whales' flukes. The uniqueness
of the coloration, shape, and scarring pattern of the
flukes' ventral side allowed for the reliable identifica-
tion of individual whales (Katona et al. 1979). Because
our primary objective was to collect individual iden-
tification photographs for use in capture-recapture
analyses and the estimation of long-term reproductive
rates, we did not attempt to count unidentified whales
along the survey tracks. Consequently, all references
to 'sightings' or 'observations' of whales are based only
on photographs of unique individuals.
Methods for processing and comparison of fluke
photographs followed that described by Perry et al.
(1988). Photographs of whales were taken with a 35 mm
single-lens reflex camera equipped with a motor drive
and a 300mm telephoto or 70-210mm zoom lens. High-
speed (ASA 400-1600) black-and-white film was used.
From each observation of a whale or group of whales,
the best photograph of each individual's flukes was
printed and assigned a "fluke observation" or iden-
tification number. Information on the location, date,
and social affiliation of each fluke identification was
stored in a data retrieval file at the University of
Hawaii Computing Center. During the matching of
fluke photographs, a whale that was identified on more
than one occasion was also assigned an "animal"
number. This number allowed us to reference all fluke
observations, or identifications, of that individual. All
Table 1
Between-years
reidentification and the Petersen population |
estimate using
Bailey's
correction (in parentheses) of hump-
back whales Megaptera
novaeangliae in southeastern Alaska.
Year
Identifiec
no.
[newly]
1:
Reidentification year
1980
1981 1982 1983 1984 1985 1986
1979
83
32
41 48 11 36 33 40
[83]
(307) (294) (310) (353) (435) (498) (484) |
1980
121
—
58 53 11 48 53 51
[89]
(306) (410) (514) (479) (457) (556)
1981
148
—
- 85 26 73 66 63
[71]
(315) (280) (388) (451) (553)
1982
182
—
- 31 80 81 79
[66]
(290) (436) (453) (544)
1983
50
—
_ _ _ 35 29 23
[12]
(269) (340) (498)
1984
193
[76]
(486) (599)
1985
203
[74]
79
(606)
1986
238
[108]
Sum
579
fluke photographs were judged to be of either good,
fair, or poor quality. Good- and fair-quality photographs
showed at least 50% of both flukes at an angle suffi-
ciently vertical to distinguish the shape of the flukes'
trailing edges. For this study, poor-quality photographs
were deleted from the data set.
Results
Abundance and regional fidelity
A total of 257 humpback whales, including 19 calves,
were individually identified in southeastern Alaska dur-
ing 1986. This total includes 29 adults identified only
in Glacier Bay, 183 identified only in Frederick Sound,
16 identified only in Sitka Sound or Chatham Strait,
and 10 adults common to more than one subregion. The
majority (n 130, 54.6%) of the 238 adults identified in
1986 had been identified in southeastern Alaska pre-
viously, based on comparison with photographs col-
lected by University of Hawaii researchers and asso-
ciates during the years 1979-85 (Perry et al. 1988). The
addition of the 108 newly identified individuals to the
existing catalogue of photographs resulted in a cum-
ulative total of 579 adult whales identified in south-
eastern Alaska across the 8 study years (Table 1).
To determine the fidelity of humpback whales to
regional feeding grounds, photographs collected from
southeastern Alaska during 1986 were compared with
432
Fishery Bulletin 90(3), 1992
photographs of 95 individuals from the western Gulf
of Alaska (von Ziegesar and Matkin 1989), 18 from cen-
tral California collected during 1977-85 (Perry et al.
1988), and 225 individuals from central California iden-
tified during 1987-88 (Calambokidis et al. 1990). This
comparison provided no evidence of movement by in-
dividual whales between these three feeding regions.
Two whales previously identified in both southeastern
Alaska and Prince William Sound (Baker et al. 1986)
were not reidentified in southeastern Alaska in 1986,
suggesting that their immigration to southeastern
Alaska may have been temporary.
The identification and reidentification of individual
animals across years lends itself to the estimation of
abundance using capture-recapture formulae (e.g.,
Hammond 1986). Table 1 summarizes abundance esti-
mates of the southeastern Alaska feeding herd from
a pair-wise comparison of all yearly samples using the
Petersen estimate with Bailey's correction (Caughley
1977). The yearly estimates range from a low of 269
(1983-84) to a high of 606 (1985-86). The weighted
mean of the Petersen estimate (i.e., the Schnabel
estimate; Seber 1982) across the 8-year study indicated
Table 2
Observed and expected frequency of yearly identifications for
579 adult humpback whales Megaptera novaeangliae in
southeastern Alaska during the years 1979-86.
Identification
frequency (years)
1
2
3
4
5
6
7 +
Observed
Expected
330
216
81
186
56
108
46
47
32
16
24
6
10
Note: Expected frequencies were calculated from the zero-
truncated Poisson distribution according to the methods
described by Caughley (1977).
Table 3
Calving rates of mature female humpback whales Megaptera novaeangliae in southe;istem
Alaska, based on reproductive histories of 41 individuals identified in two or more sum-
mer seasons (see Baker et al. 1987).
Identification year
1980 1981 1982 1983 1984 1985 1986 Sum
Females identified
Total calves
Calves/female
2
0.25
33
9
0.27
33
15
0.45
12
3
0.25
31
15
0.48
21
5
0.24
'Includes one calf thought to have died during the summer. See text for details.
that this regional subpopulation has included 547
animals (95% CL: 504-590).
Possible inequalities of individual reidentification
probabilities were examined by calculating the iden-
tification frequencies for individual whales across the
8 study years (Table 2). The observed frequency
distribution showed fewer 2- or 3-year reidentification
records and more single identifications and reidentifica-
tion records of extreme frequencies than expected
when compared with a zero-truncated Poisson distribu-
tion calculated according to Caughley (1977). The
significant departure of the observed from the expected
distribution (x.^ [4] 291, p<0.001) suggests that all in-
dividual whales were not equally available for reiden-
tification during the study period. Possible causes of
this unequal 'catchability' include births, deaths, and
permanent emigration across the 8-year study, as well
as heterogeneity of reidentification probabilities due
to local habitat preferences and the limited range of
surveys.
Reproductive rates
Among the 238 adults individually identified in 1986,
there were 32 cows accompanied by calves assumed to
be less than a year old. Using this census information
we estimated the crude birth rate in 1986 to be 0.125,
calculated as the total number of identified cows (n 32)
divided by the total number of identified whales of all
classes (n 257, including only identified calves). This
estimate, however, may have been biased by the
greater visibility of cow/calf pairs and by additional
effort directed towards individually identifying mem-
bers of this age/sex class.
An alternate estimate of annual reproductive rates
was calculated using the identification histories of in-
dividual females known to be reproductively mature
prior to the 1986 surveys (Baker et al. 1987). Of the
41 mature females previously
identified by Baker et al. (1987),
24 were reidentified during the
1986 surveys and 9 were accom-
panied by a calf, yielding an esti-
mate of 0.375 calves/mature fe-
male-year ^ The addition of
the 1986 identifications provides
an updated estimate of the long-
term calving rates for 41 females
previously discussed by Baker et
al. (1987) (Table 3). Between
1980 and 1986, these 41 females
were observed with 58 individual
calves across 162 seasonal iden-
tifications. Although annual calv-
ing rates appeared to alternate
24
9*
0.38
162
58
0.36
Baker et al : Population characteristics of Megaptera novaeangliae in southeastern Alaska
433
Table 4
Within-year
l)etween-survey)
reidentification and the Peter-
sen population estimates with Bailey's correction (in paren- |
theses) of adult humpback whales Megaptera
novaeangliae in
Glacier Bay,
1986.
Identified:
no.
Survey month [newly]
No. reidentified
June July
Aug Sept
June
27
17
12 7
[27]
(42)
(39) (37)
July
27
— —
12 9
[10]
(39) (30)
August
18
— —
10
[3]
(18)
September
10
[0]
— —
— —
Sum
40
Table 5
Within-year (between-survey) reidentification and the Peter- |
sen population estimates with Bailey
s correction (in paren-
theses) of adult humpback whales Megaptera novaeangliae in |
Frederick Sound,
1986.
Survey dates
Identified:
No. reidentified
[newly] 1
2
3 4
31 July-3 Aug
72
22
19 13
[72]
(247)
(234) (283)
29 Aug-1 Sept
78
—
23 9
[56]
(211) (429)
12-15 Sept
64 -
—
- 4
[30]
(704)
29 Nov-7 Dec
54 -
[36]
—
—
Sum
194
between high and low years from 1981 to 1986, a Test
of Independence indicated that these year-to-year dif-
ferences were not significant (x" [6] 6.88, p 0.332).
Average calving rate across the 7-year study was 0.36
calves/year (95% binomial CL: 0.284-0.432), similar to
the previously reported rate of 0.37 for the years
1980-85 (Baker et al. 1987).
Local abundance and interchange
Capture-recapture estimates of seasonal abundance for
the Glacier Bay and Frederick Sound subregions were
calculated using the Petersen formula with Bailey's cor-
rection and treating each survey or survey period as
a sample (Tables 4 and 5). In Glacier Bay, the number
of individual whales identified was greatest during June
and July and declined through August and September
Table 6
Local movement of humpback
whales Megaptera novaeangliae
between Glacier Bay (GB) and Frederick Sound (FS) during |
the summer and fall of 1986.
Animal
From
To
Interval
no.
Last ident. date
First ident. date
(days)
#117
GB, 25 July
FS, 30 Aug
36
#161
GB, 22 July
FS, 2 Aug
8
#155
GB, 22 July
FS, 30 Aug
39
#196
FS, 31 July
GB, 8 Aug
8
#221
GB, 29 July
FS, 31 Aug
33
#350
GB, 21 July
FS, 30 Aug
40
#564
GB, 22 July
FS, 26 Aug
35
#566
GB, 11 July
FS, 30 Aug
50
#587
GB, 14 Aug
FS, 4 Dec
112
#616
GB, 16 July
FS, 31 July
15
#616
FS, 31 July
GB, 14 Aug
14
#616
GB, 14 Aug
FS, 30 Aug
16
(Table 4). The percentage of newly-identified whales
declined rapidly through the summer, suggesting that
the census of identified individuals approached a com-
plete count of the whales in this subregion. Capture-
recapture estimates based on monthly censuses ranged
from 18 to 42 and agreed closely with the total number
of 40 adults identified in this subregion.
In Frederick Sound, the number of individual whales
identified during each survey remained constant from
late July to mid-September and declined by late fall
(Table 5). The percentage of newly-identified whales
decreased through the three summer surveys but in-
creased in the late-fall survey. The Frederick Sound
capture-recapture estimates from the three summer
surveys ranged from 211 to 247, exceeding the total
of 158 individuals identified during this period but not
approaching the between-year estimates of regional
abundance (see Table 1). Capture-recapture estimates
increased considerably when summer surveys were
compared with the fall surveys. Ranging from 283 to
704, the fall estimates agreed more closely with across-
year estimates for the entire southeastern Alaska
region. The larger capture-recapture estimates from
the fall survey and the increase in percentage of new-
ly identified whales suggested the dissolution of popula-
tion stratification observed during the summer months
or the arrival of individuals from unsurveyed areas of
southeastern Alaska.
Documented interchange between the southeastern
Alaska subregions was limited to 12 transits by 10 in-
dividual whales (Table 6). Eight one-way transits were
from Glacier Bay to Frederick Sound, and a single one-
way transit was from Frederick Sound to Glacier Bay.
One individual, animal #616, traveled from Glacier
434
Fishery Bulletin 90(3), 1992
Bay to Frederick Sound and back between 16 July and
14 August. Animal #616 was last identified in Fred-
erick Sound on 30 August.
Regional occupancy
The interval between the first and last identification
of an individual whale provided a minimum estimate
of its occupancy in southeastern Alaska (Baker et al.
1985). Although it was not possible to document con-
tinuous residency of individual whales in either of the
primary study areas (i.e., Glacier Bay or Frederick
Sound), there was no evidence that individuals mi-
grated to other known feeding regions between surveys
(see 'Abundance and regional fidelity'). The longest
documented regional occupancy was 192 days for
animal #587. This individual was first identified on 1
June in Glacier Bay and last identified on 9 December
in Frederick Sound. Animal #587's identification rec-
ord, discussed by Baker et al. (:!987), showed that she
lost a calf sometime during the summer of 1986. Three
other adults and one calf had documented occupancies
of nearly equivalent length: 191 days for #616, an
animal of unknown age-sex class (see also Table 6); 183
days for #350, an animal of unknown age-sex class; and
186 days for #161 and her calf.
Foraging behavior
During summer surveys, whales in Frederick Sound
tended to occur in aggregations of 20 to 80 animals
often clustered along submerged ridges and mounts,
as determined by reference to fathometer recordings
and navigational charts. Observations of whale feces
and fathometer recordings of dense scattering layers
below feeding whales indicated that euphausiids were
the primary prey for these aggregations. During the
late-fall survey, we were unable to collect fathometer
recordings or to observe whale feces in order to con-
firm the primary prey species. However, the surface-
movement and diving patterns of whales and the loca-
tion of feeding aggregations were similar to that
observed during summer surveys, suggesting that
euphausiids were again the primary prey.
The predominant prey of humpback whales in Glacier
Bay was schooling fish, as evidenced by fathometer
recordings and observations of schooling fish at the sur-
face. Within the Bay, whales fed singly or in pairs on
dense schools of capelin and sandlance. Outside the
Bay, in the adjacent waters of Icy Strait, the predomi-
nant prey of humpback whales appeared to be herring
as demonstrated in previous years using hydroacoustic
techniques and net tows (Wing and Krieger 1983,
Krieger and Wing 1984 and 1986). As in previous years
(Baker 1985a), whales near Icy Strait formed a social-
ly cohesive pod of 7 to 9 individuals that appeared to
cooperate in foraging on schools of herring.
Discussion
Population characteristics
The number of individual whales photographically iden-
tified during the 1986 surveys, 238 adults and 19
calves, can be considered a minimum estimate of abun-
dance for the southeastern Alaska feeding herd.
Capture-recapture analyses of across-year identifica-
tion records, however, provide estimates of this
regional population that are two or three times larger
than that based on the 1986 census alone. Although
these analyses are more likely than simple counts to
provide realistic estimates of regional abundance, they
should be interpreted with caution since the behavior
of whales seldom conforms strictly to the theoretical
assumptions underlying these models (e.g., Hammond
1986, Perry et al. 1990). Violation of the assumption
of equal catchability among southeastern Alaska
whales, for example, is indicated by the analysis of
reidentification frequencies across the 8-year study
period. Births, deaths and permanent emigration ob-
viously contribute to this unequal catchability (i.e.,
reidentification inequality). Another probable source of
unequal catchability is heterogeneity due to local
habitat preference by individual whales and the vari-
able and limited geographic coverage of the surveys.
While births and deaths cause a positive bias in the
Petersen estimate of abundance, reidentification het-
erogeneity causes a negative bias (Hammond 1986 and
1990).
Assuming, however, that adult mortality among
humpback whales is low (e.g., Buckland 1990) and that
permanent emigration to other feeding regions is in-
frequent (e.g.. Perry et al. 1990), the weighted Peter-
sen estimate of 547 whales (95% CL: 503-590) may be
our most robust for the southeastern Alaska subpop-
ulation in 1986. By using the cumulative reidentifica-
tion records of individuals across years and weighting
the final estimate by the largest sample year, the
weighted Petersen should be less biased than the
between-year Petersen estimates by heterogeneity due
to local habitat preferences or variation in survey ef-
fort. Births during the study period are included in the
cumulative population estimate when the calves mature
sufficiently to become available for individual identifica-
tion. The weighted Petersen is also consistent with
other estimates derived from the individual identifica-
tion records. The upper confidence interval of this
estimate overlaps with the total count of 579 whales
identified during the 8-year study and agrees closely
with the unbiased Petersen estimates from the pair-
Baker et al Population characteristics of Meqaptera novaeanghae in southeastern Alaska
435
wise comparisons of years with the largest sample of
identified whales, 1984-86 and 1985-86 (see Table 1).
Regardless of the exact number of individuals inhab-
iting this region, the individual identification surveys
and mark-recapture estimates suggest that the south-
eastern Alaska herd increased from 1979 to 1986. In
Frederick Sound, overall survey effort decreased since
1981-82 but, with the exception of 1983 when only a
single 4-day survey was conducted, the number of iden-
tified whales increased. As confirmed by photographic
documentation, a general increase in the number of
whales in the Glacier Bay area during the last few years
was the result of the continued return of past residents
and the recruitment of their offspring (Baker et al.
1988). In terms of overall regional abundance, the
mark-recapture estimates from pair-wise comparisons
of 1986 to previous years suggest an increase from 484
to 606 across 1979-86, while estimates from contiguous
years suggest an increase from 307 to 606 (Table 1).
Requiring an annual rate of increase from 3.4 to 10.4%,
these trends in estimated abundance are within the
range reported for population growth of other unex-
ploited baleen whales based on individual identification
data (e.g., Hammond 1990, Best and Underbill 1990,
Bannister 1990). More accurate estimates of the cur-
rent abundance and the true rate of increase in the
southeastern Alaska subpopulation will require further
detailed analyses of survival rates and the biases in-
troduced by heterogeneity of identification records.
Although apparently sufficient to sustain some de-
gree of population growth, the observed reproduction
rate of humpback whales in southeastern Alaska seemed
low in comparison with other studied populations and
to the maximum reproductive potential of 0.50, or even
1.00 (calves/mature female • year M as observed in
some individually identified females (Darling 1983,
Glockner-Ferrari and Ferrari 1984 and 1990, Baker et
al. 1987, Clapham and Mayo 1987 and 1990, Straley
1989). The estimated calving rate of 0.36 (calves/mature
female- year- 1) across the 1980-86 study suggests
that females from this region give birth to a calf that
survives its first migration from the wintering grounds
about once every 2.8 years. In the Gulf of Maine,
Clapham and Mayo (1990) report an average reproduc-
tion rate of 0.41 (calves/mature female • year" ' ) and an
average calving interval of 2.35 years for the period
1979-87, using individual identification methods sim-
ilar to those used here. Pregnancy rates from exploited
populations, as summarized by Baker et al. (1987), all
exceed the estimated calving rate for southeastern
Alaska, although this historical comparison is con-
founded by differences in methodology.
Seasonal trends and foraging strategies
The number of whales identified in Glacier Bay and Icy
Strait was greatest during late June and early July,
and declined through August and September. Since
survey effort in Glacier Bay was high relative to total
number of whales identified, and constant throughout
the study period, we believe that trends in the month-
ly censuses or counts of individuals reflected changes
in seasonal abundance for this subregion. Although
surveys of Frederick Sound were not frequent enough
to track the seasonal increase in whales during early
summer, the greatest numbers of whales were found
during late July and August, approximately 1 month
after the local peak in Glacier Bay.
We could not determine if these seasonal trends
reflect primarily changes in the timing of migratory
arrival on the feeding grounds or the pattern of local
movement among subregions of southeastern Alaska.
Within the geographic limits of our surveys, seasonal
changes in influx were accompanied by some local
movement between subregions; the decline in numbers
of whales in Glacier Bay was, in part, the result of their
relocation to Frederick Sound. Studies in previous
years also demonstrated that local movement between
these subregions tends to be one-directional, resulting
in the whales congregating in Frederick Sound during
late summer and fall (Baker 1984, Perry et al. 1985,
Krieger and Wing 1986). Large areas of available
habitat in southeastern Alaska remain entirely un-
surveyed (see Fig. 1), including the outer coast of
Baranof Island and the inside passage to the south of
Frederick Sound. The increase in percentage of newly-
identified whales during the late-fall survey of 1986
suggests local movement from these unsurveyed
areas.
Local movement may be an attempt to take advan-
tage of seasonal changes in prey availability. Hump-
back whales in Frederick Sound fed almost entirely on
euphausiids while those in Glacier Bay fed almost en-
tirely on schooling fish. Movement from Glacier Bay
to Frederick Sound was presumably accompanied by
a shift in primary prey species. Similar contrasts in the
primary prey species of whales in these two subregions
have been documented in previous years (Krieger and
Wing 1984 and 1986). Some whales, however, showed
a strong preference for particular prey species or local
habitat throughout the summer. This was indicated by
the persistence of certain individual whales feeding on
herring in Icy Strait late through the summer, when
other whales had moved to feed on euphausiids in
Frederick Sound.
436
Fishery Bulletin 90(3), 1992
Stock identity and management
The summer and late-season surveys of 1986 and
previous years (Baker et al. 1985) demonstrated that
many whales remained to feed in southeastern Alaska
for much of the summer and into late fall. Intervals be-
tween first identification and last reidentification of
some individual whales indicated seasonal occupancies
of at least 6 months. Since no surveys were conducted
from 15 September to 29 November, it was not pos-
sible to document continuous residency of individual
whales in either of the primary study areas (i.e., Glacier
Bay or Frederick Sound). However, comparisons of in-
dividual identification photographs collected in the cen-
tral and western Gulf of Alaska, including Prince
William Sound, and along the coast of central Califor-
nia indicate that whales which summer in southeastern
Alaska seldom migrate to alternate feeding grounds
within seasons or across years (Baker et al. 1986, Perry
et al. 1990). These observations are strong evidence
that southeastern Alaska is the migratory terminus and
primary feeding ground for a distinct herd or seasonal
subpopulation of humpback whales.
Comparisons of individual identification photographs
and analysis of mitochondrial DNA haplotypes demon-
strate that many members of the southeastern Alaska
feeding herd migrate to wintering grounds near the
islands of Hawaii (Darling and Jurasz 1983, Baker et
al. 1986, Perry et al. 1990, Baker et al. 1990). The
migratory connection between these primary seasonal
habitats provides a unique opportunity to study and
protect a population of humpback whales that spends
the majority of its time within U.S. coastal waters (Na-
tional Marine Fisheries Service 1991).
Acknowledgments
Surveys of humpback whales in southeastern Alaska
during 1986 were made possible by funding from the
U.S. Marine Mammal Commission (contract number
MM3309822-5) and the cooperation of personnel from
Glacier Bay National Park and the Auke Bay Labor-
atory, National Marine Fisheries Service. C.S. Baker
was supported by employment to Glacier Bay National
Park and Preserve during these surveys and by a
postdoctoral fellowship from the Smithsonian Institu-
tion during the preparation of this manuscript. The
views and opinions expressed in this paper do not
necessarily reflect those of these agencies. Photographs
collected during 1979 and 1980 were made available
by the National Marine Mammal Laboratory, Seattle,
courtesy of W.S. Lawton. Photographs collected dur-
ing 1981, 1982, and 1984 were made available courtesy
of L.M. Herman, University of Hawaii. We thank the
following people for their assistance in the field: Gary
Vequist, Glacier Bay National Park; Ken Krieger,
George Snyder, and Bruce Wing, Auke Bay Labora-
tory, NMFS; Carol and Jim Greenough, and Chuck
Johnstone, Sitka, Alaska. The manuscript benefited
from a thorough review by S. Swartz, Marine Mam-
mal Commission, L. Jones and H. Braham, National
Marine Mammal Laboratory, and the comments of two
anonymous reviewers.
Citations
Anon3Tnous
1984 Glacier Biy National Park and Preserve, Alaska; Pro-
tection of humpoack whales; Proposed rules. Federal Register
49;15482-15494.
Baker, C.S.
1984 Population characteristics and feeding behavior of hump-
back whales in southeastern Alaska during 1984. Unpubl. rep.
to NMFS Auke Bay Lab., Auke Bay, AK, 44 p.
1985a The population structure and social organization of the
humpback whale (Megaptera novaeangliaf) in the central and
eastern North Pacific. Ph.D. diss., Univ. Hawaii. Honolulu,
307 p.
1985b The humpback whales of Glacier Bay and adjacent
waters: Summer 1985. Unpubl. rep. to Glacier Bay Natl.
Park, Gustavus, AK, 33 p.
Baker, C.S., L.M. Herman, A. Perry, W.S. Lawlon, J.M. Straley,
and J.H. Straley
1985 Population characteristics and migration of humpback
whales in southeastern Alaska. Mar. Mammal Sci. 1:304-323.
Baker, C.S., L.M. Herman, A. Perry, W.S. LavMon, J.M. Straley,
A.A. Wolman. G.D. Kaufman, H.E. Winn. J.D. HaH.
J.M. Reinke, and J. Ostman
1986 Migratory movement and population structure of hump-
back whales {Megaptera novaeangliae) in the central and
eastern North Pacific. Mar. Ecol. Prog. Ser. 31:105-119.
Baker, C.S., A. Perry, and L.M. Herman
1987 Reproductive histories of female humpback whales
(Megaptera novaeangliae) in the North Pacific. Mar. Ecol.
Prog. Ser. 41:103-114.
Baker, C.S., A. Perry, and G. Vequist
1988 Humpback whales of Glacier Bay. Alaska. Whale-
watcher 22(3): 13-17.
Baker, C.S.. S.R. Palumbi. R.H. Lambertsen, M.T. Weinrich,
J. Calambokidis, and S.J. O'Brien
1990 The influence of seasonal migration on the geographic
distribution of mitochondrial DNA haplotypes in humpback
whales. Nature (Lond.) 344:238-240.
Bannister. J.
1990 Southern right whales off western Australia. In Ham-
mond, P.S., etal. (eds.). Individual recognition of cetaceans:
Use of photo-identification and other techniques to estimate
population parameters, p. 279-288. Rep. Int. Whaling Comm.
Spec. Issue 12.
Best, P.B., and L.G. Underbill
1990 Estimating population size in southern right whales
(Eubalaena auxtralw) using naturally marked animals. In
Hammond, P.S., et al. (eds.). Individual recognition of ceta-
ceans: Use of photo-identification and other techniques to
estimate population parameters, p. 183-190. Rep. Int. Whal-
ing Comm. Spec. Issue 12.
Baker et al Population characteristics of Megaptera novaeangliae in southeastern Alaska
437
Buckland. S.T.
1990 Estimation of survival rates from sightings of individually
identified whales. In Hammond, P.S., et al. (eds.), Individual
recognition of cetaceans; Use of photo-identification and other
techniques to estimate population parameters, p. 149-154.
Rep. Int. Whaling Comm. Spec. Issue 12.
Calambokidis, J., J.C. Cubbage, G.H. Steiger. K.C. Balcomb, and
P. Bloedel
1990 Population estimates of humpback whales in the Gulf of
the Farallones. In Hammond, P.S., et al. (eds.), Individual
recognition of cetaceans: Use of photo-identification and other
techniques to estimate population parameters, p. 325-334.
Rep. Int. WTialing Comm. Spec. Issue 12.
Caughley. G.
1977 Analysis of vertebrate populations. Wiley, NY, 288 p.
Clapham, P.J., and C.A. Mayo
1987 Reproduction and recruitment of individually-identified
humpback whales, Megaptera novaeangliae, in Massachusetts
Bay: 1979-1985. Can. J. Zool. 65:2853-2863.
1990 Reproduction of humpback whales {Megaptera novae-
angliae) observed in the Gulf of Maine. In Hammond, P.S.,
et al. (eds.). Individual recognition of cetaceans: Use of photo-
identification and other techniques to estimate population
parameters, p. 171-176. Rep. Int. Whaling Comm. Spec. Issue
12.
Darling. J.D.
1983 Migration, abundance and behavior of Hawaiian hump-
back whales, Megaptera novaeangliae (Borowski). Ph.D. diss.,
Univ. Calif., Santa Cruz, 147 p.
Darling, J.D., and CM. Jurasz
1983 Migratory destinations of North Pacific humpback whales
{Megaptera novaeangliae). In Payne, R. (ed.). Communication
and behavior of whales, p. 359-368. Westview Press, Boulder.
Darling, J.D., and D.J. McSweeney
1985 Observations on the migrations of North Pacific hump-
back whales {Megaptera novaeangliae). Can. J. Zool. 63:
308-314.
Glockner-Ferrari, D., and M. Ferrari
1984 Reproduction in humpback whales, Megaptera novae-
angliae, in Hawaiian waters. In Perrin, W.P., et al. (eds.).
Reproduction in whales, dolphins, and porpoises, p. 237-242.
Rep. Int. Whaling Comm. Spec. Issue 6.
1990 Reproduction in the humpback whales {Megaptera novae-
angliae) in Hawaiian waters, 1975-1988: The life history,
reproductive rates, and behavior of known individuals iden-
tified through surface and underwater photography. In Ham-
mond, P.S., et al. (eds.). Individual recognition of cetaceans:
Use of photo-identification and other techniques to estimate
population parameters, p. 161-169. Rep. Int. Whaling Comm.
Rep. Spec. Issue 12.
Hammond, P.S.
1986 Estimating the size of naturally marked whale popula-
tions using capture-recapture techniques. In Donovan, G.P.,
(ed.). Behavior of whales in relation to management, p.
253-282. Rep. Int. Whaling Comm. Spec. Issue 8.
1990 Heterogeneity in the Gulf of Maine? Estimating hump-
back whale population size when capture probabilities are not
equal. In Hammond, P.S., et al. (eds.). Individual recognition
of cetaceans: Use of photo-identification and other techniques
to estimate population parameters, p. 135-140. Rep. Int.
Whaling Comm. Spec. Issue 12.
Jurasz, CM., and V.P. Palmer
1981 Censusing and establishing age composition of humpback
whales {Megaptera novaeangliae) employing photodocumenta-
tion in Glacier Bay National Monument, Alaska. Unpubl. rep.
to Natl. Park Serv., Alaska Reg. Office, Anchorage, 44 p.
Katona, S.K., and J. A. Beard
1990 Population size, migrations and feeding aggregations of
the humpback whale {Megaptera novaeangliae) in the western
North Atlantic Ocean. In Hammond, P.S., et al. (eds.). In-
dividual recognition of cetaceans: Use of photo-identification
and other techniques to estimate population parameters, p.
295-307. Rep. Int. Whaling Comm. Spec. Issue 12.
Katona, S., B. Baxter, O, Brazier, S. Kraus, J. Perkins, and
H. Whitehead
1979 Identification of humpback whales by fluke photographs.
In Winn, H.E., and B.L. Olla (eds.). Behavior of marine
animals, vol. 3, p. 33-44. Plenum Press, NY.
Krieger, K., and B.L. Wing
1984 Hydroacoustic surveys and identification of humpback
whale forage in Glacier Bay, Stephens Passage and Frederick
Sound, southeastern Alaska, Summer 1983. NMFS Auke Bay
Lab., Juneau, 60 p.
1986 Hydroacoustic monitoring of prey to determine hump-
back whale movements. NOAA Tech. Memo. NMFS F/NWC-
66, NMFS Auke Bay Lab., Juneau, 62 p.
Martin, A.R., S.K. Katona, D. Matilla, D. Hembree, and
T.D. Waters
1984 Migration of humpback whales between the Caribbean
and Iceland. J. Mammal. 65:330-333.
National Marine Fisheries Service
1991 Recovery plan for the humpback whale {Megaptera
novaeangliae). Prepared by the Humpback Whale Recovery
Team for the Natl. Mar. Fish. Serv., Silver Spring, MD, 105 p.
Perry, A., C.S. Baker, and L.M. Herman
1985 The natural history of humpback whales in Glacier Bay,
Alaska. Final rep. to Natl. Park Serv., Alaska Reg. Office,
Anchorage, 22 p.
1990 Population characteristics of individually identified hump-
back whales in the central and eastern North Pacific: A sum-
mary and critique. In Hammond, P.S., et al. (eds.), Individual
recognition of cetaceans: LTse of photo-identification and other
techniques to estimate population parameters, p. 307-318.
Rep. Int. Whaling Comm. Spec. Issue 12.
Perry, A., J.R. Mobley, C.S. Baker, and L.M. Herman
1988 Humpback whales of the central and eastern North
Pacific: A catalog of individual identification photographs.
Sea Grant Misc. Rep. UNIHI-SEAGRANT-MR-88-02, Univ.
Hawaii, Honolulu, 37 p.
Seber, G.A.F.
1982 The estimation of animal abundance and related param-
eters. Macmillan, NY, 6.54 p.
Straley, J.M.
1989 Population characteristics of humpback whales {Megap-
tera novaeangliae) in Glacier Bay and adjacent waters: 1989.
Unpubl. rep. to Glacier Bay Natl. Park, Gustavus, AK, 33 p.
1990 Fall and winter occurrences of humpback whales {Megap-
tera novaeangliae) in southeastern Alaska. In Hammond, P.S.,
et al. (eds.), Individual recognition of cetaceans: Use of photo-
identification and other techniques to estimate population
parameters, p. 319-324. Rep. Int. Whaling Comm. Spec. Issue
12.
Von Ziegesar, 0., and CO. Matkin
1989 A catalogue of Prince Williams Sound humpback whales
identified by fluke photographs between the years 1977 and
1988. North Gulf Oceanic Soc, P.O. Box 15244, Homer, AK,
28 p.
Wing, B.L., and K. Krieger
1983 Humpback whale prey studies in southeastern Alaska,
summer 1982. Unpubl. rep. to NMFS Auke Bay Lab., Juneau,
60 p.
Abstract. - To test the hypoth-
esis that year-class strength in ma-
rine fishes is determined in the early-
larval stages, and that these stages
can be used to predict recruitment,
I modeled the recruitment process
using a modified form of key-factor
analysis. Using data compiled from
the fish literature, I found a signifi-
cant relationship {K" 0.90, P« 0.001,
n 97) between the mean and interan-
nual variance of stage-specific mor-
tality rates that provided variance
estimates for the model. The R'~
values for the true correlation be-
tween abundances of small larvae
and subsequent recruitment for four
example species of marine fish were
predicted to lie between 0.10 and
0.57, depending on the assumptions
of the model. I therefore suggest
that recruitment levels are 'ixed
after the early-larval period. How-
ever, the precision of sample correla-
tions are too low (10-yr data series)
to empirically test whether abun-
dances or mortality rates of early lar-
vae are in reality strongly or poorly
correlated with recruitment. After
metamorphosis, the strength of the
true relationship and the precision of
sample correlations increase suffi-
ciently to permit precise forecasting
of recruitment. Recruitment is a
complex process in which variation
in all life stages contributes substan-
tially to the variability in final abun-
dance; therefore, researchers should
recognize the importance of the later
prerecruit stages and the interac-
tions among all stages.
Precision of recruitment
predictions from early
life stages of marine fishes
Michael J. Bradford
Department of Biology, McGill University
1205 Ave. Dr Penfield, Montreal, Quebec H3A-1BI, Canada
Present address: Department of Fisheries and Oceans, West Vancouver Laboratory
4160 Marine Drive, West Vancouver, BC V7V IN6, Canada
Manuscript accepted 6 May 1992.
Fishery Bulletin, U.S. 90:439-453 (1992).
A major problem in the management
of marine fisheries is the unpredict-
able fluctuations in stock size result-
ing from variable recruitment. Hjort
(1913) first recognized this recruit-
ment variability and proposed a num-
ber of hypotheses that linked the
survival of small first-feeding larvae
and subsequent year-class strength
(reviewed by Wooster and Bailey
1989). These hypotheses have formed
the basis of much research on the
early life history of fishes, research
which has been largely focused on the
'critical period' during the transition
from endogenous to exogenous
modes of feeding. While much has
been learned about the biology of
larval fish, the evidence for a critical
period of increased mortality and a
link between the larval stage and
recruitment remains equivocal (May
1974, Ware and Lambert 1985,
Peterman et al. 1988, Campana et al.
1989). Explanations for this failure
have ranged from sampling and
technical difficulties, such as inap-
propriate scales of sampling (Leggett
1986, Taggart and Leggett 1987,
McGurk 1989), to the suggestion that
no such critical period exists, and
that all prerecruit stages contribute
to some degree to variability in year-
class strength (Sissenwine 1984, An-
derson 1988, Peterman et al. 1988).
There have been few attempts to
model the recruitment process to
assess the likelihood that early-larval
mortality is a dominant feature of
year-class variability. Manipulations
of life tables have shown that small
changes in larval mortality have the
potential to cause great variation
in recruitment (Smith 1985, Houde
1987 and 1989, Pepin and Myers
1991); however, the influence of the
larval stages in a fully dynamic model
incorporating variability in all stages
has not been investigated. In par-
ticular, the role of postlarval mortal-
ity in causing recruitment variability
is unclear and has been the cause of
some controversy (Sissenwine 1984,
Peterman et al. 1988, Taggart and
Frank 1990, Wooster and Bailey
1989).
An often-stated justification for
research on the early life history of
fish is to provide short-term forecasts
of recruitment, thereby allowing
managers to adjust fishing regula-
tions in response to changes in stock
size (Gulland 1989). While it is obvi-
ous that the stages very close to re-
cruitment will give the most accurate
predictions, sampling these stages is
often difficult and expensive (Smith
1985), unless they are caught inciden-
tally in other fisheries. Rather, ef-
forts have usually been concentrated
on finding a predictive relationship
between recruitment and the abun-
dance or some measure of survival of
larvae and recruits, on the working
assumption that Hjort's hypothesis of
year-class determination at this early
stage is valid (Peterman et al. 1988,
Gushing 1990).
The utility of short-term (i.e., an-
nual) predictors of recruitment in the
439
440
Fishery Bulletin 90(3). 1992
management of fish stocks has recently been chal-
lenged by Walters and Collie (1988) and Walters (1989).
In simulated management examples, Walters (1989)
finds that only extremely accurate forecasts of recruit-
ment can offer significant improvements over using the
long-term mean recruitment in stock assessment
models. Thus, while studies of the early stages of
marine fish may reveal insights into their ecology, it
is unclear whether sufficiently accurate forecasts of
recruitment wall ever be possible from these early
stages.
In this paper I first pose the question, "How strong
are the correlations between abundances or mortality
rates of the early life stages and recruitment likely to
be?" I develop a simple analytical model based on key-
factor analysis (Varley and Gradwell 1960, Manly
1977). I use parameter estimates compiled from a
literature survey to calculate the expected correlations
between life stages for the prediction of fish recruit-
ment. I suggest that the assertion that year-class
strength is fixed in the early-larval stages is not gen-
eral, and, furthermore, under likely field conditions it
will be difficult to quantitatively test this hypothesis.
The model
I developed a simple model to simulate the variability
in population numbers and the strength of correlations
between life stages. In brief, the model generated an-
nual abundances and mortality rates over a specified
number of years from which correlations between
early-life-history stages and recruitment were calcu-
lated. This process was repeated in a Monte Carlo
fashion to estimate the sampling distribution of the
correlation coefficients.
I divided the egg-recruit period into four intervals:
(1) egg-yolksac larvae, (2) early-feeding larvae, (3) late-
feeding larvae, and (4) juveniles from metamorphosis
to age 1, which I assumed to be the age of recruitment.
I assumed that populations would be sampled at five
distinct times that divide the egg-recruit period into
four intervals. Sampling points were: eggs spawned
(Ng), first-feeding larvae (Nf), young larvae (N|),meta-
morphs (N^), and recruits (Nr). First-feeding larvae
were operationally defined as larvae that have just
begun to feed, while young larvae were defined as
having an age of 10 days after the onset of feeding.
In any year, the number of recruits is the product
of the number of eggs spawned and the survival rates
of the prerecruit stages:
Recruitment = Eggs • Sys • Sei • Su • Sj ,
where the subscripts refer to the egg-yolksac, early-
larval, late-larval, and juvenile periods outlined above.
Expressing survival rates as instantaneous mortalities,
M = - ln(S), and taking logs of the abundances give the
usual equation of key-factor analysis (Varley and Grad-
well 1960):
N, = N, - M,, - M,, - Mn - M
*ys
^J'
(1)
where Nr and Ng are log abundances of recruits and
eggs of a particular cohort, and the M; values are
interval-specific instantaneous mortalities for four in-
tervals defined above. I assume, following Hennemuth
et al. (1980) and Peterman (1981), that log abundances
and instantaneous mortality rates are normally dis-
tributed with stage-specific variances described below.
This multiplicative process results in lognormally dis-
tributed recruitment, consistent with empirical results
(Hennemuth et al. 1980). All subsequent references to
abundance made in this paper are to log-transformed
values.
Since I am interested in short-term forecasting, I
assumed that stock size and, therefore, mean egg pro-
duction are stationary in time and that variation in egg
production is independent of recruitment. Thus, in the
absence of density-dependent processes, recruitment
is linearly related to egg production.
To introduce stochastic variation in the model, the
abundance of eggs, Ng, and the interval-specific mor-
tality rates were simulated as normal random variables.
As the time-series of egg production was stationary and
my interest is in correlations rather than abundances,
the abundance of eggs and the mortality rates all had
a mean of 0.
To start the sequence of calculations in a given model
year, the initial abundance of eggs was randomly
chosen. In the simplest version of the model, which
assumes mortality in each interval is independent of
the others and is density-independent, the following
equation was then used to calculate the numbers of
each subsequent stage:
Nk+i = Nk-mk,
(1)
where N^ is the abundance of stage k, and m is a nor-
mal random deviate that simulates random interannual
variability in mortality of interval k.
The complete independence of mortality of one stage
with that of a subsequent stage is probably an un-
realistic assumption because, for example, years which
are good for yolksac larval survival may also be good
for the survival of older larvae. This can be modeled
by introducing covariances between the interval-
specific mortality rates (Gerrodette et al. 1984). Co-
variation between interval-specific mortality rates was
modeled by assuming that there was a positive corre-
Bradford Recruitment predictions from early life stages of marine fishes
441
lation between the mortality rate of adjacent intervals
across years. The mortality of a given interval in any
model year then depends partially on the mortality of
the previous period in that same year. With p equal to
the correlation between adjacent interval-specific mor-
tality rates, I used the following equation to calculate
the mortality rate of successive intervals:
M
k+l
^ |SD(M,,i)]
Mk + (l-p2)"=Mk.i.
In this equation, the actual mortality for stage k-i- 1 is
a linear combination of the random variables simulating
the variability in stages k and k-i-1. The correlation
coefficient determines how much mortality in stage
k-H 1 is similar to that of stage k. The ratio of standard
deviations in the first term scales the contribution of
the mortality of the previous interval to the appropriate
variance. To simplify, I assumed throughout this paper
that there was no covariance between the number of
eggs spawned and mortality in subsequent intervals.
Finally, density-dependent mortality was incorpor-
ated in some versions of the model. Density-dependent
mortality was added to the juvenile period, following
suggestions of Houde (1987) and Smith (1985) that this
is the most likely interval for density effects. While a
number of formulations are possible, I chose a power
function (Peterman 1982):
iX\
where in this case X and Y are the abundances of
juveniles and recruits, respectively. For density-
independent mortality, b=l; b is <1 for density-
dependent cases. The parameter a is thus the density-
independent survival rate. After taking logs, the log
of the abundance of recruits is now a function of
the log of the number of metamorphs, N^ , and the
density-independent mortality, Mj:
N,
bN„
M,.
(2)
In the stochastic simulations, this equation was used
to calculate recruitment with a random normal deviate
substituted for Mj.
The full model was run for 1000 10-yr trials in SAS
(1987), and a matrix of abundances and mortality rates
for each stage was built up. For each 10-yr trial, cor-
relation coefficients were calculated between the
various predictors of recruitment (i.e., abundances and
mortality rates of each of the prerecruit stages), and
the numbers of recruits and summary statistics of the
distributions of correlation coefficients were derived.
Table 1
Daily mortality rates (M) and interval durations (t, in days)
for four species used as e.xamples in the analysis. Egg mor-
tality includes the yolksac period up to first feeding; larval
periods explained in text. Values were adapted from Houde
(1987; cod Gadus morhua, and herring Clupea hareiigus).
Smith (1985; anchovy EngrauHs mordax), and Zijlstra and
Witte (1985; plaice Pleuronectes platessa).
Egg
Early
larvae
Species
M
M
t
Late
larvae
M t
Juveniles
M
Cod 0.061 18 0.160 10 0.063 46 0.010 291
Herring 0.050 21 0.080 10 0.034 70 0.015 264
Anchovy 0.250 7 0.160 10 0.050 79 0.012 269
Plaice " 0.068 .38 0.104 10 0.045 77 0.008 245
Model parameters
To generalize the results, I used four fish species as
examples (Table 1). These were not chosen to be rep-
resentative of a specific stock or situation, but rather
to indicate the effect of different life histories on our
ability to forecast recruitment. To parameterize the
model for a specific species, the interannual variance
of the number of eggs laid and the mortality of each
prerecruit stage was required.
I obtained estimates of the variance in the number
of eggs spawned from published reconstructions of
stock abundances (Table 2). Except for cod, I used the
residuals of linear regressions of log(eggs) on time to
estimate the variance, since time trends existed for
some stocks.
Estimates of the variability in mortality rates for all
prerecruit stages are unavailable; I therefore sought
a predictive relationship between interannual variance
and the mean of daily mortality rates. This allowed
estimation of the variances of mortality rates of the
early life stages from mean daily rates. I surveyed the
literature for papers containing 2 or more years of
estimates of age- or stage-specific mortality for the
same population or stock. All stages from egg to adult
were used, for marine, freshwater, and anadramous
fish species. No screening of the data was done except
for estimates from adult fish, where only estimates
using methods independent of catch-data analysis were
used (i.e., tagging). Most adult estimates were from
lightly or unfished stocks. In some cases I estimated
mortality from annual estimates of abundance or from
regressions of log abundance on time. All estimates
were converted to daily values using annual estimates
of stage duration if available, or the long-term average
stage duration. Daily mortalities were then averaged
442
Fishery Bulletin 90(3), 1992
over the number of years of data available, and the
variance calculated. Both variates were log-trans-
formed, and a least-squares regression was fitted to
the data.
I used the variance-mean relationship to calculate the
interannual variance in mortality from mean daily mor-
tality rates extracted from published life tables (Table
1). I split the larval period and defined the first 10 days
of feeding as the early stage. This period corresponds
to the usual definition of the 'critical period' for first-
feeding larvae (Leggett 1986): few marine larvae can
survive more than 10 days without feeding (Miller et
al. 1988). Except for anchovy, where values were taken
directly from Smith (1985), the daily mortality rate for
the early period was set at twice the average rate for
the whole larval period. Mortality rates for the late
period were adjusted so that the mortality for the total
larval period matched the published life tables. The
result of these calculations was that the daily mortal-
ity rates of the early-larval interval were about 2.5
times those for the late-larval period. It is difficult to
assess whether this decline is realistic, because there
is considerable variability in the decline in mortality
over time in empirical studies; in many cases mortal-
ity has been found to be nearly constant over much of
the larval period (Dahlberg 1979), while there are other
cases where significant declines have been observed
(i.e., Savoy and Crecco 1988). Declines in mortality wath
larval age may be accentuated by a possible bias due
to sampling interval (Taggart and Frank 1990). The
variance in mortality over the duration of a particular
interval was then calculated as the product of the
square of the interval's duration (in days), and the
variance of the daily mortality rate predicted from the
variance-mean relationship.
Covariation in mortality rates
Two scenarios were developed concerning the effects
of covariation between mortality rates. In the inde-
pendent case, all mortality rates were varied in-
dependently of one another, while for the 'covariance'
version, mortality rates of adjacent stages were as-
sumed to be correlated across years. Few data are
available to estimate the strength of these correlations,
so I assumed p values for the correlations between ad-
jacent M|( based on the likelihood of common agents
of mortality. I assigned a relatively low p value of
0.25 for the correlation between the egg/yolksac period
and the early-larval mortality because early-larval mor-
tality is thought to be strongly affected by feeding suc-
cess, which does not affect egg survival. Nonetheless,
predation pressures are probably similar for both
stages, causing some covariation in mortality rates.
A p value of 0.5 was used between the early- and late-
Table 2
Interannual variability in log-transformed egg production and
recruitment, compiled from literature values. All egg esti-
mates are residuals from linear regressions of log abundance
on time, except for cod where an intermediate value between
herring and plaice was used.
Species
Var(NJ
Var(N,)
Cod
Herring
Anchovy
Plaice
0.075
0.081
0.282
0.055
0.40
1,92
1.91
0.14
Data sources
Cod: mean of 5 northwest Atlantic stocks in Koslow et al,
(1987).
Herring: mean of 7 northwest Atlantic stocks in Winters and
Wheeler (1987),
Anchovy: eggs— Peterman et al. (1988), recruitment— Methot
(1989).
Plaice: Bannister (1978).
larval intervals because of the similarity of habitat
between these two periods. For the pelagic species,
anchovy and herring, p = 0.25 was used for the correla-
tion between the late-larval and juvenile intervals,
while for the demersal species, cod and plaice, I set
p = 0, reflecting the major habitat shifts associated wath
metamorphosis.
Density-dependence
To explore the effects of density-dependence on cor-
relations, I ran the model with b= 1.0, the density-
independent case, or b = 0.7, simulating moderately
strong density-dependent mortality. The variance of
juvenile mortality predicted from Figure 1 is in fact
the sum of both the density-independent and density-
dependent sources of mortality. To estimate the
density -independent component of mortality (Mj) re-
quired for Eq. (2), I had to remove the density-
dependent mortality from the total juvenile mortality
predicted by Figure 1. Rearranging Eq. (2) and solv-
ing for the total juvenile mortality (Mjto,) yields
M
■jtot
N„,-N, = (l-b)N„, + Mj.
In the models without covariances between mortality
rates, and in the covariance model for cod and plaice
where there is no covariation in mortality across meta-
morphosis, taking variances yields
Var(Mj) = Var(Mju„) - (1 - b)^Var(N„,).
In these cases, to find Var(Mj) I ran the stochastic
Bradford: Recruitment predictions from early life stages of marine fishes
443
model up to the metamorph stage and calculated the
median Var(Nn,). Var(Mj) was then found by subtrac-
tion using Var(Mjtot) predicted from Figure 1 (Table
3). For herring and anchovy in the covariance model,
the equation above should include a term for the covar-
iance between Mjtot and N,,,. In these cases, Var(Mj)
was found by trial by running the model with different
values of Var(Mj) and matching the median Mjtot with
the value predicted from the regression equation of
Figure 1.
To provide objective criteria for evaluating recruit-
ment hypotheses, I defined two performance criteria
for the correlations with recruitment. Recruitment
research is commonly cast as a search for the stage
when "year-class strength is determined" or "recruit-
ment is fixed." I define such a stage as having an
i?2>0.50 with recruitment, i.e., being able to account
for at least half of the variability in year-class strength.
A more rigorous standard of i?->0.80 was set for cor-
relations to be used for management purposes (Walters
1989).
(D
O
c
m
'\—
CO
>
O
-10
-15
-
o^**
■
^x^O^
o- %K
♦ ">
D oa
1 , 1
1
-20
-8 -6 -4 -2 0
Log Daily Mortality
Figure 1
Relationship between interannual variance in daily mortal-
ity rates and mean daily mortality from published values. Sym-
bols indicate eggs (•). larvae (<>). juveniles (■), and adults
(D). Equation of the line: ln{Var(M)} = 2.231 ln(M) - 1.893
(7?" 0.90, P<0.0001). Regression uses the square root of
number of years comprising each data point as weights.
Table 3
Variances of juvenile mortality rates V(Mj) used in the four versions of
the model and the variance of log recruitment, V(N,), generated by the
model. Model versions include density-dependent (DD) or independent (DI)
juvenile mortality and, in some cases, covariance between stage-specific
mortality rates (COV). Variances for M^ in the DD models are for the
density-independent component only, and were found by simulation.
DI
DI-COV
DD
DD-COV
Species V(Mj) V(N,) YiM,) V(N,) V(Mj) V(N,) V{M-,) V(N,)
Cod 0.45
Herring 0.91
Anchovy 0.58
Plaice 0.22
1.49
1.49
2.54
1.64
0.45
0.91
0.58
0.22
2.04
2.08
3.91
2.46
0.35
0.85
0.40
0.09
0.87
1.14
1.36
0.79
0.34
0.92
0.45
0.09
Results
Variance-mean relationship
There was a highly significant relationship {R~ 0.90.
P<0.0001, n 97) between the log of mean daily stage-
specific mortality and the log of the interannual vari-
ance in the daily mortality rate (Fig. 1). The variance
in mortality rate was independent of the number of
years of data comprising each point (multiple regres-
sion with mean mortality, P 0.81 for sample size). The
square root of sample size was used as a weight in all
analyses. There was no significant effect of life history
(freshwater, marine, or anadramous) on the variance-
mean relationship (ANCOVAR; for slopes and adjusted
means, all P>0.20). There was no difference in the rela-
tionship between the variance and mean of mortality
among the egg, juvenile, and adult stages (P>0.5), but
the slope for the larval stage was significantly differ-
ent from the other three stages (intercept P 0.10, slope
P 0.010). Because there were a number of studies on
the same species, I also averaged the data across both
species and stage to decrease the non-independence of
the data due to common phylogeny. The variance-mean
regression for this averaged dataset was almost iden-
tical to the full set (P2o.92, P<0.0001,
n 53); the regression parameters differed
by <2%. In this case, the regression for the
larvae was not different than for the other
three stages (intercept P 0.28, slope P 0.11),
suggesting the significant effect found for
the full dataset may have been due to the
overrepresentation of some species. I there-
fore used the overall regression (Fig. 1) to
predict the variance of mortality of all
stages, rather than using a separate regres-
sion for larvae. This is a conservative pro-
cedure for rejecting Hjort's hypothesis,
because the single regression predicts a
more variable mortality for the early-larval
stage than does the separate larval re-
gression; the single regression produces
stronger correlations between abundance of
1.15
1.59
2.16
1.04
444
Fishery Bulletin 90(3). 1992
early larvae and recruitment than the separate larval
regression.
Correlations between
early life history and recruitment
Correlations between abundances at early life stages
and recruitment increased in strength as the interval
between the two stages decreased (Figs. 2, 3). Overall,
covariances in mortality rates across stages increased
i?2 values between early abundances and recruitment
by 0.01-0.25, while density-dependent juvenile mortal-
ity had only a small and usually negative effect onR^
values.
Correlations between egg or first-feeding larvae and
recruitment were weak; the average R^ over all spe-
cies and models was 0.05 for eggs and 0.20 for first-
feeding larvae. None of the values exceeded 0.50, in-
dicating that these early stages have little predictive
as
CO
400
1
0.8
0.6
0.4
0.2
Ih*i^
^^z::^^^rf^-
Density Independent
Covariance
100 200
Age (Days)
300
400
Figure 2
Predicted R~ values for correlations between recruitment
and early life stages for cod (O), anchovy (■), plaice (»), and
herring (D). For each species, symbols represent, from left
to right: abundance of eggs (at t = 0), first-feeding larvae, 10-d
larvae, and metaniorphs. Dotted line indicates the strength
of correlations required for recruitment prediction (Walters
1989). Both examples include density-independent juvenile
mortality; lower panel also incorporates covariance between
interval-specific mortality rates.
capability. At the end of the early-larval stage, R^
values increased; and in 4 of 16 cases in Figures 2 and
3 the R~ values exceeded 0.50. However, no values
exceeded 0.80, the suggested requirement for recruit-
ment forecasting to be beneficial for management
(Walters 1989).
In nearly all cases, the majority {R^>0.50) of re-
cruitment variation was predictable at the age of meta-
morphosis. The exception was the herring example,
which gave low correlations because of high variabil-
ity in the juvenile mortality rate. Half the correlations
met the forecasting requirement of i?">0.80 by the
age of metamorphosis; these cases occurred in species
with the lowest juvenile mortality rates.
The success of larval mortality rates in predicting
recruitment was lower than for larval abundance esti-
mates. The correlation between the mortality rate of
the early-larval period and recruitment was strongly
affected by the presence of covariation between stage-
specific mortality rates; without these covariances the
average R- was 0.12; the largest value was 0.18.
When the covariances were incorporated, these correla-
tions are increased, although none exceed 0.5 (Fig. 4).
Ten of 16 R- values exceeded 0.50 for the much
400
Density Dependent
Covariance
1 00 200 300
Age (Days)
400
Figure 3
As in Figure 2, except both versions include density-dependent
juvenile mortality; lower panel also incorporates mortality
covariances.
Bradford: Recruitment predictions from early life stages of marine fishes
445
longer late-larval period. The recruitment-forecasting
threshold of 0.8 was never reached for correlations be-
tween recruitment and any larval mortality rate.
A wide range of R'^ values can result from a short
time-series. For example, the 95% range of R- values
for the correlation of early cod larvae with recruitment
(10-yr time-series) in the density-independent model
that includes covariance in mortality rates extended
from 0.07 to 0.86 (Fig. 5). The 95% range decreases
if the true relationship between the variables is
stronger; for cod metamorphs the conclusion that this
stage can be used to describe the majority of recruit-
ment variation will nearly always be reached (Fig. 5).
Sensitivity analysis
Two sensitivity analyses were conducted to assess
dependence of the results on input parameters. First,
stage-specific variances in mortality were recalculated
with the slope of the variance-mean regression set at
DI-COV
DD-COV
CO
1
0.8
-
-f -— '
t
0.6
r* *' '^
'■-7^
-0
0.4
-^~^^^~^-=-~^^-^^^^ ^y
~--~B
0.2
Dl
DD DI-COV
Model Version
DD-COV
Figure 4
Predicted R'^ values for correlations between recruitment
and early- and late-larval mortality rates for cod (O). anchovy
(■), plaice (•), and herring (□). Axis labels refer to four ver-
sions of the model, incorporating density-independent (DI) or
-dependent (DD) juvenile mortality and covariances between
mortalities (GOV).
its 95% confidence limits; the intercept was derived by
constraining the line through the mean of both vari-
ables. For the cod-DI model, increasing the slope to the
upper confidence limit increased the R~ for the cor-
relations between the abundance of recruits and early
larvae or metamorphs by about 0.05; decreasing the
slope lowered R^ values by similar amounts. There
was little effect on correlations involving the egg or
first-feeding stages. I also recalculated the correlations
with the intercept of the variance-mean regression at
its 95% confidence limits. With the intercept at its
lower limit, R'~ values increased by 0.01-0.04, and at
the upper limit the correlations decreased by a similar
amount. Thus, the overall results are not particularly
sensitive to the sampling error associated with the data
in Figure 1.
I also varied the length of the early-larval period. In
the life tables (Table 2), I fixed the early-larval period
at 10 days and set the daily mortality rate at twice the
average for the whole larval period. In sensitivity runs
I varied this period from 5 to 15 days; duration and
mortality rate of the late-larval period were recalcu-
lated to keep the total mortality for the larval period
constant. The duration of this period of high larval mor-
tality had a strong effect on the strength of the cor-
relation between abundance of larvae sampled at the
end of the early period and recruitment. When the
early-larval period was increased by 5 days, the i?- in
the cod-DI model increased by 0.21 (Fig. 6).
cr
1.0
0.6
0.4
0.2
-0.2
A
1
Egg F-feeding E-larvae Metamorphs
Stage Sampled
Figure 5
Variability in sample R'^ values (10-yr series) for the correla-
tion between recruitment and abundances of early stages for
cod in the model, with density-independent juvenile mortal-
ity and mortality covariances. Shown are the median (bar),
interquartile (rectangle), and 95% ranges (line). Data are from
1000 runs; note that the criterion for significance {R->0.
a 0.05) is 0.40.
446
Fishery Bulletin 90(3). 1992
40
Age (Days)
Figure 6
Effects of altering length of the early-larval stage on
recruitment correlation for the cod DI-COV model.
Early stage is defined as having twice the daily mor-
tality rate of the total larval period. Shown are results
with the early stage set at 5 (■). 10 (♦). and 15 (D)
days. Symbols represent, from left to right: egg, yolk-
sac larvae, early larvae, and metamorph stages.
Discussion
My results indicate that only predictions of recruitment
based on abundances of postmetamorphic fish are likely
to be useful for the management of marine fishes. The
contribution to recruitment variation made by egg
number (and, therefore, stock biomass) is very small,
a prediction confirmed by most stock-recruit data (Par-
rish 1973). Correlations involving the abundance of
early larvae are stronger, but are still too weak for
forecasting. The model R'-^ values for correlations in-
volving early larvae are similar to the range, extending
from 0.01 to 0.66, for published values compiled by
Peterman et al. (1988). Accurate recruitment fore-
casting may be possible by sampling during the late-
larval period (Graham and Sherman 1987). However,
this is highly dependent on parameters and the dy-
namics of the particular species; only for cases with low
variability in juvenile mortality or with mortality rates
correlated across stages are the abundances of late lar-
vae likely to be useful for recruitment forecasting.
Research on recruitment variability has been ori-
ented to the early-larval stages largely as the result
of Hjort's (1913) hypotheses and the observation that
most of the individuals of a year-class die during the
first few weeks of life (Wooster and Bailey 1989). In
my four example species, the average cumulative mor-
tality on the cohort to the 10-d larval stage is 93%
(Table 1), yet the variation in abundance of these lar-
vae explains more than 50% of recruitment variability
in less than half of the cases. Variability in the late-
larval and juvenile stages is still large enough to in-
fluence the strength of correlations of recruitment with
larval abundances. My results suggest that the stage
'when year-class strength is determined' (defined here
as i?^>0.5) occurs after this early critical period.
However, the sensitivity analysis indicates that the
strength of the correlation between the abundance of
larvae and recruitment will depend strongly on the rate
at which mortality declines during the larval period,
and at what age the larvae are being sampled.
Strong linkages in mortality rates across intervals
also render the definition of a 'critical period' less con-
cise. Correlations between early life stages and recruit-
ment were stronger when there were linkages, because
survival to the age of sampling will be correlated to
survival in the future. An estimate of mortality or abun-
dance in one stage will be an index of mortality in all
early life stages. This may be especially true for the
early-larval stages (the classical 'critical period') be-
cause small larvae are probably subjected to a similar
source of mortality as older larvae, especially if spawn-
ing occurs over a protracted period, mixing larvae of
different ages together in the same body of water. In
addition, environmental conditions during an early
stage may affect survival of the cohort in the future.
Poor feeding conditions of early larvae, for example,
may have a long-term effect on growth and survival
(Frank and McRuer 1989). In these cases, recruitment
will be somewhat predictable from the early-larval
stages, but this is not support for a strict interpreta-
tion of Hjort's hypothesis that an early critical period
determines recruitment because mortality is correlated
across all prerecruit stages.
The difficulty and expense of obtaining accurate
estimates of abundances of eggs and larval fish have
led to increased interest in finding indirect estimates
of mortality rates that may be simpler to collect and
could provide an index of year-class strength. Such
measures include estimates of growth (Houde 1987),
condition, lipid content (Theilacker 1986), and RNA/
DNA ratios (Buckley and Lough 1987) as well as ocean-
ographic variables such as upwelling and wind events
(Peterman and Bradford 1987). My results show that
such mortality estimates made on small larvae are not
likely to be strongly correlated with recruitment
(Fig. 4). Mortality estimates on older larvae will have
stronger correlations, potentially closer to a value of
0.50. Note that the correlations in Figure 4 are for
direct estimates of mortality, indirect indices will be
more poorly related to recruitment. A combination of
larval abundances and mortality rate estimates may
allow more precise prediction of recruitment (Graham
and Sherman 1987, Frank and McRuer 1989); if esti-
mates are accurate and are based on older larvae,
Bradford Recruitment predictions from early life stages of marine fishes
447
correlations nearly as strong as those predicted for
metamorphs might be possible.
The correlations will be weaker if sampling errors
are included in the estimates of abundance. Preliminary
simulations with random sampling errors with a coef-
ficient of variation of 50% (untransformed abundances)
decreased R- values in Figures 2 and 3 by 0.10-0.15
(Bradford unpubl.). Biased estimates, e.g., due to gear
avoidance (Lo et al. 1989), will not affect correlations
between an early stage and recruitment, unless the
magnitude of the bias is correlated with the estimate.
Precision, through the use of consistent technique
across years, is more important for the purposes of
forecasting. Large-scale surveys of abundance of late
larvae or juveniles may be sufficiently accurate for the
forecasting of recruitment (Lo et al. 1989) if the stage
sampled is likely to be strongly correlated v, ith recruit-
ment (Fig. 5).
An implicit, though infrequently stated, assumption
of research on early-life-history influences on recruit-
ment is that the mean and the interannual variance of
mortality rates are correlated. High mortality alone will
not cause recruitment variation; it must also be coupled
with high interannual variability. The data compiled in
Figure 1 provide evidence that this is generally true,
and that the interannual variability in the larval period
is proportionately no greater than that found for other
stages. In addition, my sensitivity analysis suggests
that the general conclusions of this paper are robust
to the sampling variability of this relationship. How-
ever, detailed investigation of the recruitment dynam-
ics of an individual species will require estimation of
the variance of stage-specific mortality rates, because
the predictive power of Figure 1 is still relatively low
for any particular case, and the biology of an individual
species may not result in rates that follow the overall
average pattern. Examples are provided by species
which spawn during periods of extreme climatic events
such as wind storms (e.g., capelin Mallotus villosus or
red drum Sciaenops ocellatus, reviewed by Taggart and
Frank 1990). In these cases, interannual variability in
the mortality of the earliest stages is probably larger
than predicted by the regression of Figure 1, and the
correlation between the early stages and recruitment
is likely to be stronger than I have predicted. In con-
trast, for the North Sea plaice a relationship {R^ 0.7)
was found between egg abundance and recruitment
(Zijlstra and Witte 1985), which is higher than my
model predicts for this species (although just within
the 95% range). This species has relatively low recruit-
ment variation, suggesting that larval and juvenile sur-
vival rates are not as variable as predicted by Figure
1, or that density-dependent mortality might be impor-
tant in regulating recruitment (Zijlstra and Witte
1985).
The recruitment variances generated by various ver-
sions of the model tend to be higher than published
values (Tables 2, 3). These literature estimates will like-
ly be underestimates of the true variability in recruit-
ment, because errors in catch sampling and ageing can
greatly reduce recruitment variability estimated from
sequential population analysis (Rivard 1989, Bradford
1991). Alternatively, my recruitment variances could
be too high because I have either overestimated the
variances of mortality rates or underestimated the
severity of density-dependent mortality. Since the data
in Figure 1 include sampling error, all of the variances
in Tables 2 and 3 will be somewhat inflated. If sam-
pling error is proportional to the rate of mortality, the
sensitivity analysis suggests that removing sampling
error (i.e., lowering the intercept of Fig. 1) will have
only a slight effect on recruitment correlations.
One additional source of variability not explicitly con-
sidered in my analysis is the effect of varying stage
duration, due to interannual variability in growth rates.
Houde (1987, 1989) has demonstrated through life-table
manipulation that small variations in larval growth may
have large effects on the number of metamorphs pro-
duced. The effect on recruitment will be buffered
somewhat as shortening the larval period will increase
the length and, therefore, the total mortality of the
juvenile stage. However, if the variation in growth
rates is due to temperature, Pepin (1991) suggests that
the offsetting effects of temperature on development
and mortality will result in no net effect of temperature
variation on cumulative mortality over the egg and lar-
val stages. In this case, by not including variation in
growth rates I will have overestimated the variability
in larval mortality. However, to some extent the ef-
fects of gi'owth-rate variation are already included in
my model because many of the estimates in Figure 1
are based on total stage length and will, therefore, in-
clude the effects of varying stage duration caused by
variation in growth rates in the calculation of the
average daily mortality rate.
The sampling variability of correlations from short
datasets makes it difficult to draw inferences about the
causes of recruitment variability. This low precision
suggests that confidence limits around the sample
estimates, r or R~, should always be supplied, much
in the same way that standard errors are given for sam-
ple means. A population correlation from Figures 2 and
3 is a value that would be obtained from a very long
time-series of data, and is a true measure (in the con-
text of the model) of the contribution of an early life
stage to recruitment variation. However, there is a
good chance (e.g., >30% for early larvae in Fig. 5) that
a sample correlation between the abundance of an early
stage and recruitment may not be significantly dif-
ferent from 0. Even if the correlation is significant.
448
Fishery Bulletin 90(3), 1992
statements about whether the relationship is, in real-
ity, weak or strong cannot be made because the 95%
confidence limits around the sample correlation are
wide (Fig. 5). Published correlations between early life
stages and recruitment vary greatly in strength (e.g.,
Peterman et al. 1988, Stevenson et al. 1989, Gushing
1990); unfortunately with short data series, true dif-
ferences in the biology of these species cannot be
distinguished from sampling error. The correlations
between early larval abundances and recruitment com-
piled by Peterman et al. (1988) also illustrate this point:
in only 1 of 7 cases do the 95% confidence limits around
R^ not include both 0.2 and 0.8. This problem of low
precision is less serious when the true correlation is
likely to be fairly high (Fig. 5, metamorphs).
The precision of correlations is also relevant to anal-
yses involving oceanographic or climatic variables and
recruitment. These studies usually invoke hypotheses
that the environmental variables are agents of larval
mortality, either through transport or their effects on
the production or concentration of larval food (Shep-
herd et al. 1984, Hollowed and Bailey 1989); therefore,
their true correlations with recruitment can be no
stronger than the correlations for mortality rates
directly (Fig. 4). Yet the sampling variability of R'^ for
a short series of data suggests that there will be a good
chance of finding at least one strong sample correla-
tion among a group of 4-5 predictors that may be, in
reality, only weakly related to recruitment. Adding
more data will result in the sample correlation declin-
ing towards p ; this frequently results in the sample cor-
relation becoming nonsignificant (Koslow et al. 1987,
Walters and Collie 1988, Prager and Hoenig 1989). My
model results suggest environmental variables will be
strongly correlated with recruitment only if the en-
vironmental factor is related to mortality across all
prerecruit stages (e.g.. Fig. 4; covariance models).
In summary, my analysis indicates that it is unlikely
that estimates of abundance of survival rates of the egg
and early-larval stages of marine fish will lead to useful
predictions of recruitment. Although mortality in the
earliest life stages is a major source of recruitment
variability, the late-larval and juvenile periods are also
important. Peterman et al. (1988), Fritz et al. (1990),
and Pepin and Myers (1991) argue for the need for coor-
dinated research on all prerecruit stages, rather than
focusing only on the early stages, and my results sup-
port this view. The modeling approach I have developed
here can be easily modified for any particular species
to estimate a priori the likelihood of success of pro-
posed recruitment research and to suggest particular-
ly fruitful avenues of investigation.
Acknowledgments
This analysis would not have been possible without
the efforts of many scientists in estimating the vital
rates of fish populations in multiyear studies. I thank
G. Gabana for discussions of variability and correlations
and for help collecting data. This paper has been im-
proved through the comments of G. Gabana. M. La-
Pointe, B. McKenzie, P.M. Peterman and D. Roff, and
two reviewers. Partial support was provided by post-
graduate fellowships from the National Science and
Engineering Research Gouncil of Ganada (NSERG) and
the Max Bell Foundation, and NSERG operating
grants to D. Roff.
Citations
Ahlstrom. E.H.
1954 Distribution and abundance of egg and larval populations
of the Pacific sardine. Fish. Bull.. U.S. 56:83-140.
Alexander, G.R., and D.S. Shatter
1969 Trout production and angling success from matched
plantings of brook trout and rainbow trout in East Fish Lake,
Michigan. J. Wildl. Manage. 33:682-692.
Al-Hossaini, M., Q. Lui, and T.J. Pitcher
1989 Otolith microstructure indicating growth and mortality
among plaice, Pleurmiectes platessa, post-larval sub-cohorts. J.
Fish. Biol. 35(Suppl. A):81-90.
Allen, R.L.
1977 Status of the upper Columbia river salmon and steelhead
runs. In Schwiebert, E. (ed.), Columbia river salmon and
steelhead, p. 23-30. Am. Fish. Soc. Spec. Ptibl. 10, Bethesda.
Anderson, J.T.
1984 Early life history of redfish {SebaMes spp.) on Flemish
Cap. Can. J. Fish. Aquat. Sci. 41:1106-1116.
1988 A review of size dependent survival during pre-recruit
stages of fishes in relation to recruitment. J. Northwest Atl.
Fish. Sci. 8:55-66.
Bannister, R.C.A.
1978 Changes in plaice stocks and plaice fisheries in the North
Sea. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 172:86-101.
Bannister, R.C.A. , D. Harding, and S.J. Lockwood
1974 Larval mortality and subsequent year-class strength in
the plaice {Pleunmfdes platessa). In Blaxter, J.H.S. (ed.), The
early life history of fish, p. 21-37. Springer-Verlag. NY.
Barnaby, J.T.
1944 Fluctuations in abundance of red salmon, Oncorh i/nch us
nerka. of the Karluk River, Alaska. Fish. Bull., U.S. 50:
237-295.
Bradford. M.J.
1991 Effects of ageing errors on recruitment time series esti-
mated from sequential population analysis. Can. J. Fish.
Aquat. Sci. 48:555-558.
Buckley, L.J., and R.G. Lough
1987 Recent growth, biochemical composition, and prey field
on larval haddock {Melanogrammus aeglefinus) and Atlantic
cod {Gadus morhua) of Georges Bank. Can. J. Fish. Aquat.
Sci. 44:14-25.
Bradford: Recruitment predictions from early life stages of marine fisfies
449
Campana, S.E., K.T. Frank. P.C.F. Hurley, P.A. Koeller,
F.H. Page, and P.C. Smith
1989 Survival and abundance of young Atlantic cod {Gadus
morhua) and haddock (Melanogrammus aeglejinus) as indi-
cators of year-class strength. Can. J. Fish. Aquat. Sci. 46
(Suppl. 1):171-182.
Chadwick, E.M.P.
1982 Stock-recruitment relationship for Atlantic salmon {Sahno
salar) in Newfoundland rivers. Can. J. Fish. Aquat. Sci. 39:
1496-1501.
1987 Causes of variable recruitment in a small Atlantic salmon
stock. In Dadswell, M.J., et al. (eds.), Common strategies of
anadromous and catadromous fishes, p. 390-401. Am. Fish.
Soc. Symp. 1, Bethesda.
Clady. M.D.
1975 Early survival and recruitment of smallmouth bass in
northern Michigan. J. Wildl. Manage. 39:194-200.
Gushing, D.H.
1990 Plankton production and year-class strength in fish
populations: An update of the match/mismatch hypothesis.
Adv. Mar. Biol. 26:2.50-294.
Dahlberg, M,D,
1979 A review of survival rates of fish eggs and larvae in rela-
tion to impact assessments. Mar. Fish. Rev. 41(3):1-12.
Dey, W.P.
1981 Mortality and growth of young-of-year striped bass in the
Hudson river estuary. Trans. Am. Fish. Soc. 110:151-157.
Dragesund, O., and O. Nakken
1973 Relationship between parent stock size and year class
strength in Norwegian spring spawning herring. Rapp.
P.-V. Reun. Cons. Int. Explor. Mer 164:15-29.
Farris, D.A.
1960 Abundance and distribution of eggs and larvae and sur-
vival of larvae of jack mackerel (Trachurus symmetricus).
Fish. Bull., U.S. 61:247-279.
Foerster, R.E.
1968 The sockeye salmon {Oncorhynchiis nerka). Fish. Res.
Board Can. Bull. 162, 422 p.
Forney, J.L.
1971 Development of dominant year classes in a yellow perch
population. Trans. Am. Fish. Soc. 100:739-749.
Frank. K.T.. and J.K. McRuer
1989 Nutritional status of field-collected haddock (Melanogram-
mus aeglefinus) larvae from southwestern Nova Scotia: An
assessment based on morphometric and vertical distribution
data. Can. J. Fish. Aquat. Sci. 46(Suppl. 1):125-133.
Franklin, D.R., and L.L. Smith
1963 Early life history of the northern pike with special ref-
erence to the factors influencing the numerical strength of year
classes. Trans. Am. Fish. Soc. 92:91-110.
Freeburg, M.H., W.W, Taylor, and R.W. Brown
1990 Effect of egg and larval survival on year-class strength
of lake whitefish in Grand Traverse Bay, Lake Michigan.
Trans. Am. Fish. Soc. 119:92-100.
Fritz, E.S., L.B. Crowder. and R.C. Francis
1990 The National Oceanic and Atmospheric Administration
plan for recruitment fisheries oceanography research. Fish-
eries (Bethesda) 15:25-31.
Gerrodette, T., D. Goodman, and J. Barlow
1984 Confidence limits for population projections when vital
rates vary randomly. Fish. Bull, U.S. 83:207-215.
Graham. J.J.. and K.M. Sherman
1987 Recruitment mechanisms of 2-year-old atlantic herring
to the Maine fishery: The larval stage. In Hoyt. R.D. (ed.),
10th annual larval fish conference, p. 61-68. Am. Fish. Soc.
Symp. 2. Bethesda.
GuUand, J.A.
1989 Fish populations and their management. J. Fish. Biol.
35A:l-9.
Harding, D.. J.H. Nichols, and D.S. Tungate
1978 The spawning of plaice [Pleurovectes platessa) in the
southern North Sea and English Channel. Rapp. P.-V. Reun.
Cons. Int. Explor. Mer 172:102-113.
Hatch, J.T., and J.C. Underhill
1988 Abundance, distribution, and mortality of larval lake
herring in western Lake Superior. In Hoyt, R.D. (ed.), 11th
annual larval fish conference, p. 95-103. Am. Fish. Soc.
Symp. 5, Bethesda.
Heessen, H.J.L., and A.D. Rijnsdorp
1989 Investigations on egg production and mortality of cod
(Gadus morhua) and plaice (Pleuronectes platessa) in the
southern and eastern North Sea in 1987 and 1988. Rapp.
P.-V. Reun. Cons. Int. Explor. Mer 191:15-20.
Hennemuth, R.C, J.E. Palmer, and B.E. Brown
1980 A statistical description of recruitment in eighteen
selected fish stocks. J. Northwest Atl. Fish. Sci. 1:101-111.
Hjort, J.
1913 Fluctuations in the great fisheries of northern Europe
viewed in the light of biological research. Rapp. P.-V. Reun.
Cons. Int. Explor. Mer 19:1-228.
Hollowed, A.B., and K.M. Bailey
1989 New perspective on the relationship between recruitment
of Pacific hake (Merlucius productus) and the ocean environ-
ment. In Beamish, R.J., and G.A. McFarlane (eds.), Effects
of ocean variability on recruitment and an evaluation of
parameters used in stock assessment models, p. 207-220.
Can. Spec. Publ. Fish. Aquat. Sci. 108.
Houde, E.D.
1987 Fish early life dynamics and recruitment variability.
In Hoyt, R.D. (ed.), 10th annual larval fish conference, p.
17-29. Am. Fish. Soc. Symp. 2, Bethesda.
1989 Subtleties and episodes in the early life of fishes. J. Fish.
Biol. .3.5A:29-38.
Howe, A.B., P.G. Coates, and D.E. Pierce
1976 Winter flounder estuarine year-class abundance, mortal-
ity, and recruitment. Trans. Am. Fish. Soc. 105:647-657.
Johannessen, A.
1986 Recruitment studies of herring in Lindaaspollene,
western Norway, 1-3. Fiskeridir. Skr. Ser. Havunders. 18:
139-240.
Jonsson, B., S. Skulason, S.S. Snorrason, O.T. Sandlund,
H.J. Malmquist, P.M. Jonasson, R. Gydemo, and T. Lindem
1988 Life history variation of polymorphic Arctic charr (Salve-
linus alpinus) in Thingvallavatn, Iceland. Can. J. Fish. Aquat.
Sci. 45:1537-1547.
Kawanabe, H.
1969 The significance of social structure in production of the
ayu, Plectoglossus altivelis. In Northcote, T.G. (ed.). Sym-
posium on salmon and trout in streams, p. 243-265. H.R.
MacMillan Lect. Fish., Univ. Brit. Col., Vancouver.
Koslow, J. A., K.R. Thompson, and W. Silvert
1987 Recruitment to northwest Atlantic cod (Gadus mordua)
and haddock (Melanogrammus aeglefinus) stocks: The influ-
ence of stock size and cHmate. Can. J. Fish. Aquat. Sci.
44:22-39.
Kramer, R.H., and L.L. Smith
1962 Formation of year classes in largemouth bass. Trans.
Am. Fish. Soc. 91:29-41.
450
Fishery Bulletin 90(3), 1992
Leggett, W.C.
1986 The dependence of fish larval survival on food and pred-
ator densities. In Skreslet, S. (ed.). The role of freshwater
outflows in coastal marine systems, p. 117-137. Springer-
Verlag, Berlin.
Lo, N.C.H., J.R. Hunter, and R.P. Hewitt
1989 Precision and bias in estimates of larval mortality. Fish.
Bull., U.S. 87:399-416.
Lockwood, S.J.
1980 Density-dependent mortality in 0-group plaice (Pleuro-
rwctes platessa) populations. J. Cons. Cons. Int. Explor. Mer
39:148-153.
Manly, B.J.F.
1977 The determination of key factors from life table data.
Oecologia 31:111-117.
Manooch, C.S., and G.R. Huntsman
1977 Age, growth, and mortality of the reg porgy, Pagrus
pargi~us. Trans. Am. Fish. Soc. 106:26-35.
Mansfield, P.J., and J.D. Jude
1986 Alewife (Alosa pseudoharengiis) survival during the first
grovrth season in southeastern Lake Michigan. Can. J. Fish.
Aquat. Sci. 43:1318-1326.
Mathews, S.B.
1984 Variability of marine survival of Pacific salmonids: A
review. In Pearcy, W.G. (ed.). The influence of ocean condi-
tions on the production of salmonids in the north Pacific, p.
161-182. Greg. State Univ., Corvallis.
Mathews. S.B., and F.W. Olson
1980 Factors affecting Puget Sound coho salmon {Oncorhyn-
chus kisutch) runs. Can. J. Fish. Aquat. Sci. 37:1373-1378.
May, R.C.
1974 Larval mortality in marine fishes and the critical period
concept. In Blaxter, J.H.S. (ed.). The early life history of fish,
p. 3-19. Springer-Verlag, Berlin.
McDonald. J., and J.M. Hume
1984 Babine Lake sockeye salmon (Orworhynchus nerka)
enhancement program: Testing some of the major assump-
tions. Can. J. Fish. Aquat. Sci. 41:70-92.
McGurk, M.D.
1989 Advection, diffusion and mortality of Pacific herring lar-
vae Clupea harengus pallasi in Bamfield Inlet, British Colum-
bia. Mar. Ecol. Prog. Ser. 51:1-18.
McNeil, W.J.
1969 Survival of pink and chum salmon eggs and alevins. In
Northcote, T.G. (ed.). Symposium on salmon and trout in
streams, p. 101-117. H.R. MacMillan Lect. Fish., Univ. Brit.
Col., Vancouver.
Methot, R.D.
1989 Synthetic estimates of historical abundance and mortality
for the northern anchovj'. In Edwards, E.F., and B.A. Megrey
(eds.), Mathematical analysis of fish stock dynamics, p.
66-82. Am. Fish. Soc. Symp. 6, Bethesda.
Miller, T.J., L.B. Crowder, J. A. Rice, and E.A. Marshall
1988 Larval size and recruitment mechanisms in fishes:
Towards a conceptual framework. Can. J. Fish. Aquat. Sci.
45:1657-1670.
Nickelson. T.E.
1986 Influences of upwelling, ocean temperature, and smolt
abundance on marine survival of coho salmon (Oncorhynchus
kisutch) in the Oregon production area. Can. J. Fish. Aquat.
Sci. 43:527-535.
Nielsen, L.A.
1980 Effect of walleye {Stizostedion vitreum vitreum) preda-
tion on juvenile mortality and recruitment of yellow perch
(Perca Jlavescens) in Oneida Lake, New York. Can. J. Fish.
Aquat. Sci. 37:11-19.
Palomera, I., and J. Lleonart
1989 Field mortality estimates of anchovy larvae, Engraulis
encrasicolus, in the western Mediterranean. J. Fish. Biol.
35(Suppl. A):133-138.
Parker, R.R.
1968 Marine mortality schedules of pink salmon of the Bella
Coola River, central British Columbia. J. Fish. Res. Board
Can. 25:757-794.
Parrish, B.B. (editor)
1973 Fish stocks and recruitment. Rapp. P.-V. Reun. Cons.
Int. Explor. Mer 164.
Pearcy, W.G.
1962 Ecology of an estuarine population of winter flounder,
Psevdopleuronectes americanus. III. Distribution, abundance,
growth, and production of juveniles; survival of larvae and
juveniles. Bull. Bingham Oceanogr. Collect. Yale L'niv. 18:
39-64.
Pepin, P.
1991 Effect of temperature and size on development, mortality
and survival rates of the pelagic early life stages of marine
fish. Can. J. Fish. Aquat. Sci. 48:503-518.
Pepin, P.. and R.A. Myers
1991 Significance of egg and larval size to recruitment vari-
ability of temperate marine fish. Can. J. Fish. Aquat. Sci.
48:1820-1828.
Peterman. R.M.
1981 Form of random variation in salmon smolt-adult relations
and its influence on production estimates. Can. J. Rsh. Aquat.
Sci. .38:1113-1119.
1982 Nonlinear relation between smolts and adults in Babine
Lake sockeye salmon {Oncorhynchus nerka) and implications
for other salmon populations. Can. J. Fish. Aquat. Sci. 39:
904-913.
Peterman, R.M., and M.J. Bradford
1987 Wind speed and mortality rate of a marine fish, the
northern anchovy, Engraulis morda^. Science (Wash. DC)
235:354-3.56.
Peterman, R.M., M.J. Bradford, N.C.H. Lo, and R.D. Methot
1988 Contribution of early life stages to interannual variability
in recruitment of northern anchovj' (Engraidis nwrdax). Can.
J. Fish. Aquat. Sci. 45:8-16.
Poole, J.C.
1969 A study of vnnter flounder mortality rates in Great South
Bay, New York. Trans. Am. Fish. Soc. 98:611-616.
Prager, M.H., and J.H. Hoenig
1989 Superposed epoch analysis: K randomization test of en-
vironmental effects on recruitment with application to chub
mackerel. Trans. Am. Fish. Soc. 118:608-618.
Pritchard, A.L.
1948 Efficiency of natural propagation of the pink salmon
(Oncorhynchus gorbuscha) in McClinton Creek, B.C. J. Fish.
Res. Board Can. 7:224-236.
Rivard. D.
1989 Overview of the systematic, structural, and sampling
errors in cohort analysis. In Edwards, E.F., and B.A. Megrey
(eds.). Mathematical analysis of fish stock dynamics, p.
49-65. Am. Fish. Soc. Symp. 6, Bethesda.
Roberson, K., and R.H. Holder
1987 Development and evaluation of a stream.side sockeye
salmon (Oncorhynchus nerka) incubation facility, Gulkana
River, Alaska. In Smith, H.D., L, Margolis, and C.C. Wood
(eds.), Sockeye salmon (Oncorhynchus nerka) population biology
and future management, p. 191-197. Can. Spec. Publ. Aquat.
Sci. 96.
Bradford, Recruitment predictions from early life stages of marine fisfies
451
SAS
1987 SAS/STAT guide for personal computers, version 6.
SAS Inst. Inc.. Gary. NC.
Savoy. T.F.. and V.A. Crecco
1988 The timing and significance of density-dependent and
density-independent mortality of American shad, Alosa sapi-
dissima. Fish, Bull,, U,S, 86:467-488.
Seelbach, P.W.
1987 Effect of winter severity on steelhead smolt jaeld in
Michigan: An example of the importance of environmental
factors in determining smolt yield. In Dadswell, M.J., et al.
(eds.), Common strategies of anadromous and catadromous
fishes, p. 441-4.50. Am. Fish. Soc. Symp. 1, Bethesda.
Shepherd, J.G., J.G. Pope, and R,D, Cousens
1984 Variations in fish stocks and hypotheses concerning their
links with climate, Rapp. P.-V. Reun. Cons. Int. Explor. Mer
185:255-267.
Shetter, D.S.
1961 Survival of brook trout from egg to fingerling stage in
two Michigan trout streams. Trans. Am. Fish. Soc. 90:
252-258.
Sissenwine, M.P.
1984 Why do fish populations vary? In May, R.M,, (ed.), Ex-
ploitation of marine communities, p. 59-94. Springer- Verlag,
Berlin,
Smith, P,E.
1985 Year-class strength and survival of 0-group clupeoids.
Can, J. Fish. Aquat. Sci. 't2(Suppl. l):69-82.
Stevenson, D.K., K.M. Sherman, and J.J. Graham
1989 Abundance and population dynamics of the 1986 year
class of herring along the Maine coast. Rapp. P.-V. Reun.
Cons. Int. Explor. Mer 191:345-350.
Sunby, S., H. Bjorke, A.V. Soldal. and S. Olsen
1989 Mortality rates during the early life stages and year-class
strength of northeast Arctic cod {Gadus morhua). Rapp.
P.-V. Reun. Cons. Int. Explor. Mer 191:351-358.
Taggart, C.T., and K.T. Frank
1990 Perspectives on larval fish ecology and recruitment pro-
cesses: Probing the scales of relationships, /n Sherman, K.,
CM. Alexander, and B.D. Gold (eds.). Large ecosystems: Pat-
terns, processes and yields, p. 151-164, Am, Assoc, Adv, Sci,,
Wash. DC.
Taggart, C.T., and W.C. Leggett
1987 Short-term mortality in post-emergent larval capelin
Mallotus villosus. I. Analysis of multiple i" situ estimates.
Mar. Ecol. Prog. Ser. 41:205-217.
Taylor, W.W., M,A. Smale, and M.H. Freeberg
1987 Biotic and abiotic determinants of lake whitefish (Core-
gonus clupeaformis) recruitment in northeastern Lake Michi-
gan, Can, J, Fish, Aquat, Sci, 44(Suppl, 2):313-323.
Theilacker, G.H.
1986 Starvation-induced mortality of young sea-caught jack
mackerel, Trachurus symmetricus. determined with histo-
logical and morphological methods. Fish. Bull., U.S. 84:1-17.
Thome, R.E., and J.J. Ames
1987 A note on variability of marine survival of sockeye salmon
{Oncorhynchus nerka) and effects of flooding on spawning suc-
cess. Can. J. Fish. Aquat. Sci. 44:1791-1795.
Turner, J.L., and H.K. Chadwick
1972 Distribution and abundance of young-of-year striped bass,
Morone saxatilis, in relation to river flow in the Sacramento-
San Joaquin estuary. Trans. Am. Fish. Soc. 101:442-452.
Uphoff, J.H.
1989 Environmental effects on survival of eggs, larvae, and
juveniles of striped bass in the Choptank River, Maryland.
Trans. Am. Fish. Soc. 118:251-263.
Van Woert, W.F.
1980 Exploitation, natural mortality, and survival of small-
mouth bass and largemouth bass in Shasta Lake. California.
Calif. Fish Game 66:163-171.
Varley. G.C., and G.R. Gradwell
1960 Key factors in population studies. J. Anim. Ecol. 29:
399-40L
Walters, C.J.
1989 Value of short-term forecasts of recruitment variation
for harvest management. Can. J. Fish. Aquat. Sci. 46:
1969-1976.
Walters, C.J., and J.S. Collie
1988 Is research on environmental factors useful to fisheries
management? Can. J. Fish. Aquat. Sci. 45:1848-1854.
Walton, C.J.
1987 Parent-progeny relationships for an established popula-
tion of anadramous alewives in a Maine lake. In Dadswell,
M.J.. et al. (eds.). Common strategies of anadromous and
catadromous fishes, p. 451-454. Am. Fish. Soc. Symp. 1,
Bethesda.
Ward. B.R.. and P.A. Slaney
1988 Life history and smolt-to-adult survival of Keogh River
steelhead trout {Salmo gairdneri) and the relationship to smolt
size. Can. J. Fish. Aquat. Sci. 45:1110-1122.
Ware, D.M., and T.C. Lambert
1985 Early life history of Atlantic mackerel {Scomber scoin-
brus) in the southern Gulf of St. Lawrence. Can. J. Fish.
Aquat. Sci. 42:577-592.
Watanabe, Y., and N.C.H. Lo
1989 Larval production and mortality of Pacific saury, Colo-
labis saira, in the northwestern Pacific Ocean. Fish. Bull.,
U.S. 78:601-613.
Winters, G.H., and J. P. Wheeler
1987 Recruitment dynamics of spring-spawning herring in the
Northwest Atlantic. Can. J. Fish. Aquat. Sci. 44:882-900.
Wooster, W.W., and K.M. Bailey
1989 Recruitment of marine fishes revisited. In Beamish.
R.J.. and G.A. McFarlane (eds.), Effects of ocean variability
on recruitment and an evaluation of parameters used in stock
assessment models, p, 153-159. Can. Spec. Publ. Fish. Aquat.
Sci. 108.
Zijlstra, J.J., and J. IJ. Witte
1985 On the recruitment of 0-group plaice in the North Sea.
Neth. J. Zool. 35:360-376.
Zijlstra, J.J., R. Dapper, and J, IJ. Witte
1982 Settlement, growth and mortality of post-larval plaice
{Plewronectes platessa) in the western Wadden Sea. Neth. J.
Sea. Res. 15:250-272.
452
Fishery Bulletin 90(3). 1992
Appendix
Daily mortality rates and interannual
variances plotted in Figure 1. Stage refers to
egg (E), larval (L), juvenile (J), or
adult (A) periods. For anadromous salmon, fry-smolt and smolt-
adult mortality
were
zlassified as juvenile and adult mor-
tality, respectively. N is the number of years
of data in each
study.
Species
Stage
M
var(M)
N
Study
Alma pseudoharengus
L
0.0490
4.31E-04
9
Mansfield and Jude 1986
L
0.1980
1.07E-04
8
Walton 1987
A. sapidissima
L
0.0524
2.96E-04
8
Savoy and Crecco 1988
L
0.0978
1.28E-03
8
L
0.2290
3.50E-03
8
J
0.0027
1.30E-07
8
J
0.0185
5.08E-06
8
Clupea harengus
E
0.3135
2.80E-02
2
Dragesund and Nakken 1973
L
0.2060
l.lOE-02
2
L
0.0275
3.38E-04
11
Graham and Sherman 1987
L
0.1168
1.83E-03
4
Johannessen 1986
L
0.0900
8.98E-03
2
McGurk 1989
Cololabis saira
L
0.0726
9.19E-04
14
Watanabe and Lo 1989
Coregonus artedii
L
0.5000
1.63E-02
2
Hatch and Underbill 1988
C. dupeaformis
E
0.0299
6.40E-05
2
Taylor et al. 1987
L
0.0143
9.30E-05
2
L
0.0216
2.10E-04
2
Freeberg et al. 1990
EngrauHs encrasicolis
L
0.2000
l.llE-02
3
Palomera and Lleonart 1989
E. mordax
E
0.3600
1.53E-02
13
Peterman et al. 1988
L
0.1860
5.52E-04
13
J
0.0180
9.77E-06
13
Esox lucius
L
0.1650
3.60E-03
3
Franklin and Smith 1963
Gadus morhua
E
0.1890
4.84E-03
3
Campana et al. 1989
L
0.1940
2.37E-03
3
E
0.2304
1.48E-02
7
Heessen and Rijnsdorp 1989
L
0.1010
5.63E-04
8
Sunby et al. 1989
J
0.0312
8.90E-05
7
Melanogrammus aeglefinus
E
0.1270
4.43E-03
3
Campana et al. 1989
L
0.2630
8.56E-03
3
Micropterous dulomieui
E
0.0890
7.12E-05
3
Clady 1975
L
0.0310
7.74E-05
3
A
0.0007
6.00E-09
2
Van Woert 1980
M. sahnoides
L
0.0224
9.90E-05
3
Kramer and Smith 1962
J
0.0078
1.80E-06
4
Morons saxatilis
L
0.1770
4.50E-04
2
Dey 1981
J
0.0040
2.00E-06
2
L
0.1550
4.30E-03
6
Uphoff 1989
J
0.0530
2.80E-04
10
Turner and Chadwick 1972
Oncorhynchus gorbuscha
E
0.0097
5.32E-06
6
Pritchard 1948
J
0.0260
7.80E-05
3
Parker 1968
A
0.0080
1.65E-06
3
0. kisutch
A
0.0028
1.37E-07
8
Mathews 1984
A
0.0078
5.15E-07
17
Mathews and Olson 1980
A
0.0049
3.51E-07
22
Nickeison 1986
0. mykiss
J
0.0036
1.85E-05
3
Seelbach 1987
A
0.0020
1.02E-07
4
Allen 1977
A
0.0022
3.08E-07
7
Ward and Slaney 1988
Bradford; Recruitment predictions from early life stages of marine fisfnes
453
Appendix
(continued)
Species
Stage
M
var(M)
N
Study
Oncorhynchus nerka
Sixmile Creek
E
0.0103
3.15E-06
2
Foerster 1968
Scully Creek
E
0.0116
8.82E-07
6
Williams Creek
E
O.OUl
5.91E-06
3
Chilko Creek
E
0.0123
2.24E-06
7
Tally Creek
E
0.0146
1.53E-05
11
Port John Lake
E
0.0123
1.20E-05
9
Karyinai Spring
E
0.0120
7.23E-06
8
Chilko Lake
J
0.0018
3.96E-08
7
Karymai Spring
J
0.0056
1.31E-06
8
Cultus Lake
J
0.0053
1.47E-06
10
Port John Lake
J
0.0028
1.77E-06
8
Babine Lake
J
0.0052
1.17E-06
10
J
0.0032
1.72E-06
14
McDonald and Hume 1984
A
0.0026
2.98E-07
7
Foerster 1968
A
0.0021
3.26E-07
24
Peterman 1982
Karluk River
A
0.0017
7.86E-09
6
Barnaby 1944
Summit Lake
A
0.0036
1.73E-08
3
Roberson and Holder 1987
Ten Mile Lake
A
0.0032
2.88E-07
6
Gulkana Hatchery
A
0.0031
7.65E-08
9
Lake Washington
A
0.0034
4.42E-07
11
Thorne and Ames 1987
Oncorhynchus spp.
E
0.0147
2.90E-05
23
McNeil 1969
Pagnts pagrus
A
0.0012
7.76E-09
3
Manooch and Huntsman 1977
Perra flavescans
J
0.0146
2.36E-05
6
Forney 1971
J
0.0320
1.53E-04
8
A
0.0079
8.90E-06
12
Nielson 1980
Plecoglossus altivelis
J
0.0097
6.70E-06
3
Kawanabe 1969
Pleuronectes platessa
E
0.0783
1.24E-03
11
Harding et al. 1978
E
0.1165
4.70E-03
2
Heessen and Rijnsdorp 1989
L
0.0525
7.24E-04
4
Bannister et al. 1974
L
0.0112
6.46E-05
7
Zijlstra et al. 1982
J
0.0336
7.69E-04
2
Al-Hossaini et al. 1989
Firemore
J
0.0303
1.07E-04
4
Lockwood 1980
Filey Bay
J
0.0245
1.18E-04
4
J
0.0030
1.08E-06
5
Zijlstra et al. 1982
Pseudopleuronectes americanus
J
0.0389
1.09E-04
3
Howe et al. 1976
J
0.0058
3.04E-07
2
Pearcy 1962
A
0.0019
1.18E-07
5
Poole 1969
Salmo salar
J
0.0019
7.07E-07
7
Chadwick 1982
J
0.0027
1.38E-07
10
Chadwick 1987
A
0.0058
2.36E-06
9
Salcelinus alpinus
A
0.0015
1.53E-06
2
Jonsson et al. 1988
S. frontinalis
J
0.0082
6.47E-03
6
Shelter 1961
J
0.0110
8.76E-07
9
J
0.0106
1.55E-07
3
S. salvelmus
A
0.0042
4.59E-06
5
Alexander and Shelter 1969
Sardiiiops caerulea
L
0.0831
3.32E-04
2
Ahlstrom 1954
Scomber scomber
E
0.5260
2.50E-02
4
Ware and Lambert 1985
L
0.5110
2.12E-02
3
Sebastes spp.
L
0.0680
1.25E-03
2
Anderson 1984
Trachurus symmetrichus
L
0.1387
4.62E-04
3
Farris 1960
AbStreiCt. — Complete series of
field-collected larvae were used to
describe the post-yolksac develop-
ment of two common southern Cali-
fornia marine sculpins, Clinocottus
analis and Orthmiopias triads. Char-
acters diagnostic of C. analis include
nape pigment, dorsal head pigment,
heavy rows of dorsal gut melano-
phores, 18-33 postanal ventral mela-
nophores (PAVM). Postflexion lar-
vae develop multiple preopercular
spines (9-12) and several post-tem-
poral/supracleithral spines, and later
stages also acquire a W-shaped patch
of pigment on the body under the
second dorsal fin. Characters diag-
nostic of Orthonopias triads include
a heavy cap of dorsoposterior gut
pigment, 26-55 PAVM, occasional-
ly one or two dorsocranial melano-
phores, and, rarely, one melanophore
at the nape; postflexion 0. triads
develop four preopercular spines.
Comparison with other cottid species
is included.
Field collection data (1978-85) in-
dicate C. analis and 0. triads larvae
both occur in greatest densities off
rocky habitats along the 15 m iso-
bath. A key is provided for known
preflexion marine sculpin larvae
found in southern California.
Post-yolksac larval development
of t\A/o southern California
sculpins, Clinocottus analis and
Orttionopias triads (Pisces: Cottidae)
Richard F. Feeney
Section of Fishes, Natural History Museum of Los Angeles County
900 Exposition Boulevard, Los Angeles, California 90007
Manuscript accepted 6 May 1992.
Fishery Bulletin, U.S. 90:4.54-468 (1992).
Clinocottus analis and Orthonopias
triads axe two common marine sctil-
pins (Pisces: Cottidae) of the rocky
intertidal and subtidal areas of south-
ern California (Miller and Lea 1972,
Eschmeyer et ai. 1983). The range of
C. analis extends from Cape Men-
docino, northern California, to Asun-
cion Pt., Baja California Sur; 0. tri-
ads extends from Monterey, central
California, to San Geronimo I., cen-
tral Baja California (Fig. 1).
A description of the embryology
and larval development of Clinocot-
tus analis was first attempted by
Eigenmann (1892) who gave a pre-
liminary description of the eggs and
yolksac larvae of C. analis from
reared eggs obtained in San Diego
Bay CA, and subsequently by Budd
(1940) from eggs obtained in Monte-
rey Bay CA. In both studies the lar-
vae died at the end of the yolksac
stage. Bolin (1941) described the em-
bryology and yolksac development of
reared Orthonopias triads.
Hubbs (1966) described many char-
acteristics of C. analis embryology,
especially in response to tempera-
ture, but gave no description of the
larvae. Washington (1986) presented
a description of a limited series of
postflexion C. analis larvae and juve-
niles identified on the basis of meris-
tic and morphological characters. A
7.0mm 0. triads was previously il-
lustrated (Washington et al. 1984).
No description, however, of a com-
plete larval series of either species
exists, despite the common occur-
rence of adults in California coastal
waters and the existence of several
partial descriptions of their larval
development in the literature.
The following is a description of
larval series for both C. analis and 0.
triads based on field-collected spe-
cimens from southern California and
Baja California, Mexico. Comparison
with other cottid species and occur-
rence is discussed. A key to known
southern California preflexion cottid
larvae is included to summarize early-
life-history information from many
sources including Richardson and
Washington (1980), Richardson (1981),
Washington et al. (1984), Washing-
ton (1986), Feeney (1987), and Mata-
rese et al. (1989). This work is in-
tended to aid in identification and
hopefully stimulate further research
on the development of related species.
Materials and methods
A total of 145 larvae and 9 juveniles
of Clinocottus analis and 322 larvae
and 4 juveniles of Orthonopias tria-
ds were studied. Specimens were ex-
amined from the Scripps Vertebrate
Collection (SIO), the Southwest Fish-
eries Science Center (SWFSC), the
Ctilifomia Academy of Sciences (CAS),
and the Natural History Museum
of Los Angeles County, Section of
Fishes (LACM).
The SIO specimens (21) are pre-
served in 50% isopropanol and were
collected in Baja California at Bahia
Todos Santos (SIO H5M9B); the lot
454
Feeney: Post-yolksac development of two southern California sculpins
455
Orlhoiwpiits tnitcis - I |
Isia San Geroriimo
Puma Asuncion
Pacific Ocean
Figure 1
Geographic range of Clinocottus analis and Orthonopias triads.
contained an excellent series of both C. analis and
0. triads postflexion larvae, the discovery of which
became the impetus for the present study.
The SWFSC material (10 specimens) is preserved in
5% formalin; some specimens (6) were collected in Baja
(6607-AX-110.32; 6806-JD-110.32; SWFSC/SIO H51-
106), the remainder were collected in California. Two
C. analis specimens (SWFSC/SIO H46-63) and two
0. triads (SWFSC 6607-AX-l 10.32) were cleared and
stained.
CAS material included one lot of postflexion C. analis
(SU 68789, 70% ethanol), collected in Monterey Bay,
California.
LACM specimens (fixed in 5% formalin and pre-
served in either 5% formalin or 70% ethanol) were col-
lected in coastal waters (<75m depth) of the Southern
California Bight between Pt. Conception and the Mex-
ican border. Most specimens were collected during the
Coastal Resources and 316b phases of the Ichthyo-
plankton Coastal and Harbor Studies (ICHS) Program
and during the Bightwide Program; methods and
station locations can be found in Brewer et al.
(1981), Brewer and Smith (1982), and Lavenberg
et al. (1986). Also, five postflexion C. analis
specimens (LACM 45404-1, 45414-1-45417-1; in
ethanol) were collected at the Catalina Island
Marine Science Center (Ninos 1984). Six addi-
tional C. analis juveniles from the general LACM
collection were used: four collected at Santa Bar-
bara Island (LACM 31546-4), one at Catalina
Island (LACM 35695-1), and one at Palos Verdes
Peninsula (LACM 1993).
Morphometric data, including preanal length,
body depth, pectoral length, head length, and eye
diameter were measured from 50 C. analis and
54 0. triads specimens. Data were entered into
an "Excel" spreadsheet program on a Macintosh
Ilci. Means and standard deviation of morpho-
metric measurements were computed using
"SYSTAT." Frequency plots of melanophores vs.
length were made using "SYGRAPH" and the
"LOWESS" (locally-weighted least squares) scat-
terplot smoothing method (Wilkinson 1989).
Specimens were illustrated using a camera
lucida attached to a Wild M3 stereomicroscope.
Occurrence data are based on specimens taken
during 1978-85 on ICHS and Bightwide cruises
using a variety of sampling gears. During Coastal
Resources collections (ICHS cruises, 1978-79)
oblique bongo samples and discrete depth samples
were taken monthly along a grid of 10 transects,
each with 4 stations. The transects were evenly
spaced along the coast from Point Conception to
San Diego. The stations corresponded to bottom
depths of 8, 15, 22, and 36m. Additionally, 8
stations (4 sites each) were located in Los Angeles-
Long Beach Harbor and San Diego Bay. Integrated
water-column samples were collected by fishing a 70 cm
bongo sampler from the bottom to the surface. Discrete
depth samples were collected at the surface (manta
sampler), at the mid-depth of the water column (70 cm
bongo sampler) and at the bottom (70 cm bongo sampler
equipped with wheels). All samplers had nets of 335 ^i
mesh Nitex, and attached flowmeters gave estimates
of the volume of water filtered. During the 316b phase
(ICHS cruises, 1979-80) the number of transects was
increased to 20 and the number of stations was reduced
to 2 (8 and 22 m) except for 4 "expanded" transects
(Ormond Beach, Playa del Rey, Seal Beach, San
Onofre), which retained 4 stations (8, 15, 22, and 36m).
For epibenthic sampling, the benthic bongo sampler
was replaced by a larger "Auriga" sampler. Collections
were taken monthly during the 316b phase. Samples
were sporadically taken in 1981, but no data from them
are used here.
456
Fishery Bulletin 90(3), 1992
Figure 2
Field-collected Clinocottus anaiis larvae: (A) 3.9mm (LACM KH #22), (B) 5.6mm (LACM KH #22), (C) 5.6mm (LACM
018SO-36-AU-01), (D) 8.6 mm (SIO H51-19B).
Feeney: Post-yolksac development of two southern California sculpins
457
The Bightwide program began in 1982 and samples
were taken bimonthly at the four "expanded" 316b
transects. During the Bightwide program, a fifth sta-
tion (75 m) was added to each transect. Only oblique
bongo samples were taken during the Bightwide phase.
Additional details are provided in Lavenberg et al.
(1986).
Estimates of larval abundance (n/lOm- of sea sur-
face) for each taxon were estimated (for methods, see
Smith and Richardson 1977). These abundances were
plotted against variables, such as transect, station
depth, gear type and date, to determine patterns of
local occurrence.
Identification
Yolksac and small post-yolksac larvae of Clinocottus
analis and Orthonopias triads were identified by com-
parison with descriptions of reared larvae (Eigenmann
1892, Budd 1940, Bolin 1941). Larger preflexion and
flexion larvae were associated to postflexion larvae and
juveniles using pigment characters, number of preoper-
cular spines, length of gut, and location of the anus.
Washington (1986) was helpful in linking postflexion
C. analis individuals to juveniles using melanophore
patterns and meristics. For definition of terms, see
Feeney (1987).
Results
Description of Clinocottus analis larvae
Distinguishing characters Distinguishing characters
of Clinocottus analis preflexion larvae include heavy
dorsoposterior gut pigment, nape pigment (usually with
a nape bubble), 18-25 postanal ventral melanophores
(PAVM), and melanophores on the head over the mid-
brain. Late preflexion larvae may develop up to 33
PAVM. Larger flexion and postflexion larvae develop
multiple preopercular spines (9-12) similar to other
Clinocottus and Oligocottus species (Washington 1986).
Transforming larvae develop a W-shaped patch of pig-
ment under the 2d dorsal and have an advanced anus.
In juveniles, the preopercular spines coalesce to one
bifurcate spine; small, prickly scales begin to develop
under the 2d dorsal fin. The anus advances about
halfway to pelvic fin origin.
Morphology Clinocottus analis yolksac larvae hatch
at lengths of 3.7-4.5 mm (Eigenmann 1892, Budd
1940); preserved field-collected larvae are found as
small as 8.1 mm (due to shrinkage during preservation).
Larvae are robust with fully pigmented eyes at hatch-
ing. Dorsal gut diverticulae (wings) as seen in some
Artedius (Washington 1986) are absent; however.
Table 1
Morphometries
3f larvae and juveniles
of Clinocottus analis
and Orthonopiat
triads, represented as a mean percentage 1
of standard length ± the standard deviation, with range in |
parentheses.
Measurement
stage
Clinocottus analis
Orthonopias triads
Preanal length
Preflexion
46.0 + 3.4(40.0-52.2)
38.8±3.0(31.5-44.8)
Flexion
47.1 ±2.2(44.6-48.5)
41.7±2.8(37.8-47.2)
Postflexion
.50.5 ±2. 1(46.9-54.5)
43.5 ±3.0(39.1-48.3)
Juvenile
47.3± 1.1(46.0-48.4)
43.5 + 2.7(39.9-46.3)
Body depth
Preflexion
24.8±2.2(19.7-29.7)
24.313.1(19.8-33.2)
Flexion
23.8±2.2(21.3-25.3)
24.312.9(19.8-28.5)
Postflexion
28.7 ±2.0(25.6-32.7)
25.912.4(22.1-29.5)
.Juvenile
26.4 ±2.7(24.5-30.4)
23.611.6(21.7-25.3)
Pectoral length
Preflexion
8.5±1.3 (6.6-11.3)
8.111.2 (6.1-11.3)
Flexion
8.9±2.0 (7.3-11.2)
10.212.6 (6.3-16.0)
Postflexion
27.8±4.3(16.9-32.5)
18.215.0(10.8-25.4)
Juvenile
35.411.8(33.0-36.8)
34.912.0(32.4-37.1)
Head length
Preflexion
21.8±2.1(18.6-26.1)
21.211.7(18.0-24.4)
Fle.xion
23.3 ±0.7(22.5-23.8)
24.112.7(19.3-29.1)
Postflexion
30.2 ±1.8(25.6-32.8)
28.012.3(25.0-31.4)
Juvenile
36.6 ±4.4(33.8-43.2)
34.511.3(33.1-35.7)
Eye diameter
Preflexion
10.7±1.1 (8.3-12.5)
10.010.8 (8.5-12.2)
Flexion
10.1±0.2 (9.9-10.2)
9.210.9 (7.8-10.7)
Postflexion
8.910.6 (8.0-10.5)
9.111.1 (8.1-10.8)
Juvenile
10.0±0.6 (9.4-10.7)
10.9±1.1 (9.9-12.4)
sometimes a bump can be seen in that area.
The preanal length averages 46% of notochordal
length (NL), which is closer to Eigenmann's illustra-
tion (est. 44% ) than to Budd's illustration (est. 33%);
the minimum preanal length from field-collected speci-
mens was 40% NL (Table 1). During flexion the preanal
length increases slightly to an average of 47%. In
postflexion larvae, preanal length increases to an aver-
age 51.5% standard length (SL). In juveniles, the pec-
toral fin and head lengthen to an average 35 and 37%
SL, respectively (Table 1).
In postflexion larvae, the anus is slightly advanced
of the anal fin origin. In transforming postflexion lar-
vae, the anus advances from the anal fin to about one-
third the distance to the pelvic fin origin. In juveniles,
the anus advances almost halfway to the pelvic fins.
At 9.8mm a cirrus appears on each dorsal orbit
(Fig. 3B).
Fin development In postflexion larvae, fin elements
start to form; the caudal rays become segmented.
Pelvic fins appear as buds (Table 2). At 9.8mm, fin
rays, including the pelvics, are well-formed.
458
Fishery Bulletin 90(3). 1992
Figure 3
Field-collected airwcoMw-s analis larvae and juveniles: (A) 9.7 mm (SIO H51-19B), (B) 9.8mm (SIO H5M9B), (C) 10.6mm
(LACM 008-88-22-MA-Ol), (D) 13.3mm (LACM 45404-1).
Feeney: Post-yolksac development of two southern California sculpins
459
Spination Preopercular spines begin to develop in the
late preflexion stage at ~5.5mm NL; the 5.6 mm spe-
cimen (in Fig. 2B) has developed 2 spines. During flex-
ion, the number of preopercular spines increases to 5
(Table 2, Fig. 2C).
In postflexion larvae, the preopercular spines number
6-12 (Table 2); the upper spine is elongated. A post-tem-
poral/supracleithral spine appears at 8 mm (Fig. 2D).
By 9.7 mm, a pair of nasal spines appear (Fig. 3A).
The dorsalmost preopercular spine elongates to about
twice the length of other spines. The number of post-
temporal/supracleithral spines increases to 3. At 9.8
mm, a small spine (not illustrated) may be present
where the sensory canal forms over the parietal,
anterior to the nape; the spine persists in specimens
up to 11mm SL (CAS SU 68789).
In juveniles, multiple preopercle spines (about 10)
coalesce to 1 elongate, bifurcated uppermost spine and
2 convex undulations ventrally where the other spines
had been. Larval post-temporal spines form the anter-
iormost part of the lateral line which later becomes
decorated with a series of multispined scales. Smaller
spines (prickles) form laterally below the 2d dorsal and
lateral line.
Pigmentation In yolksac Clinocottus analis, about
140 dense melanophores in 6-7 rows line the dorso-
posterior gut (peritoneal) membrane (Eigenmann 1892,
Budd 1940). Nape melanophores number 11-15 with
several extending onto a bubble of skin that is usually
present at the anterior nape. A stellate melanophore
can usually be found on the head over one or both sides
of the midbrain. A row of 18-25 PAVM is present from
about the 6th postanal myomere to the caudal area; the
last 2-3 melanophores usually extend down into the
finfold.
Post-yolksac larvae retain much of the appearance
of the yolksac larvae (Fig. 2A). The number of PAVM
may increase to 33, but usually ranges in the mid-20s,
generally decreasing in larger larvae (Fig. 4).
Late preflexion larvae develop numerous head
melanophores over the midbrain (Fig. 2B). One 4.6 mm
specimen had 19 midbrain melanophores and one
forebrain melanophore; however, the melanophores
over the midbrain usually number 10-15 with no fore-
brain pigment. Melanophores sometimes form at the
anus in this stage; however, these usually form in the
postflexion stage. One 5.2mm specimen had 5-6
melanophores in a circle around the anus.
By 9.7mm, the number of PAVM has decreased to
less than 23 (Fig. 4). In a 9.8mm specimen (Fig. 3B),
melanophores begin to form below the nape and lateral-
ly below the second dorsal fin.
Table 2
Meristics of larvae and juveniles of Clinocottus analis (speci-
mens inside the two dashed lines are undergoing flexion of
the notochord).
Size
(mm) D,
D,
A
P
V
PS
PCV CV TV
M
PAVM
3.1 -
0
— —
35
29
4.2 -
—
_
—
—
0
—
— —
33
28
5.6 -
-
-
-
-
2
-
- -
34
24
5.2 -
_
?
34
25
5.6 -
-
-
-
-
5?
-
- -
34
26
8.4 IX?
17?
13
15? buds 6-8
32
21
9.7 IX
16
14
15
1,3
10
—
— —
33
14
10.5 IX
17
14
15
1,3
10
—
— —
31?
17?
10.9* IX
17
13
15
1,3
11?
11
22 33
—
—
11.1 IX
17
13
15
1.3
9
—
— —
32
9
11.4* IX
17
14
15
1.3
11
11
22 33
—
—
13.3" IX
16
13
15
1.3
1
11
21 32
—
11
15.1 IX
16
13
15
1,3
It
—
— -
—
14
15.8" IX
16
13
15
1.3
1
11
21 32
—
8
21.0 IX
16
13 15
spines
1,3
It -
PCV
- - - 8
precaudal vertebrae
D, dorsal fin
D,, dorsal
rav
5
cv
caudal vertebrae
A anal fin rays
TV
total vertebrae
P pectoral fin rays
M
myomeres
V pelvic
rays
PAVM
postana
ven
tral
PS preopercle
spi
nes
melanophores 1
* cleared and
Stained larvae
* * x-rayed
tThe one
preopercle
spme
has a
double point
40
30
o
Clinocottus analis
°aD°
OD O
20
%
10
oroNv o
O ^^^-^ 0
8 ^"'^^^
0
o -
-10
(
D 10 20 3
0
LENGTH
Figure 4
Frequency of postanal ventral melanophores (PAVM) vs.
length (mm) with a LOWESS regression line at F = 0.5 (half
the points included in a running window) for Clinocottus analis
larvae and juveniles.
460
Fishery Bulletin 90(3). 1992
Transforming postflexion larvae develop a wide band
of pigment under the second dorsal that is typically
W-shaped and extends ventrally almost to the anal fin
(Fig. 3C). Another band of dense pigment forms under
the first dorsal fin and extends down across and onto
the pectoral fin base. The head becomes heavily pig-
mented; about 15 large stellate melanophores (along
with numerous small ones) extend across the pre-
opercle and below the eye. Two or three melanophores
appear on the posterior maxillary. Melanophores sur-
round the nasal openings and spine. A band of pigment
runs across the anterior upper lip (premaxillary). The
lower jaw and chin also have pigment. The ventral
gut is not pigmented. Several of the caudal rays are
pigmented.
Juvenile C. analis continue to add pigment dorso-
laterally while still retaining some of the larval pigmen-
tation (Fig. 3D). The W-shaped patch is still present
under the second dorsal, as well as a band of pigment
under the first dorsal and across the pectoral fin base.
The number of PAVM continue to decrease (Fig. 4).
Two new patches of melanophores appear on the caudal
peduncle and over the hypural plates. Melanophores
appear in the dorsal, pectoral and caudal fins.
Meristics Clinocottus analis postflexion larvae have
6 branchiostegal rays and twelve (6 + 6) principal caudal
rays which are consistent with adult counts. Other
meristics are given in Table 2. Numbers of fin and
vertebral elements match well with modes given by
Howe and Richardson (1978).
Comparison with other species
Clinocottus analis larvae have no anterior gut pigment
like C. recalvus larvae (Morris, 1951). Clinocottus acu-
ticeps also has forebrain pigment and a longer trailing
gut than C. analis, no early head pigment, fewer
PAVM, and hindgut diverticulae (Washington 1986).
Clinocottus embryum has fewer nape and PAVM.
Clinocottus globiceps has anterior gut pigment and only
four or five PAVM.
Preflexion Oligocottus maculosus have shorter guts
(preanal averages 39.1% SL) than C. analis (Washing-
ton 1986). Oligocottus snyderi has no head pigment and
few PAVM (~6). Larvae of 0. rubellio (rosy sculpin)
and 0. rimensis (saddleback sculpin) have not been
described. A 15.6mm juvenile 0. rubellio (LACM
42918-1) differs from C analis juveniles by having
more cirri on the head, no W-shaped pigment patch
laterally, and no banding anywhere, just a fine cover-
ing of light melanophores. Oligocottus rimensis differs
by having an elongate body and a high number of dor-
sal soft rays (16-19). A 17.1mm 0. rimensis (LACM
943) is developing saddles of pigment typical of adults
but lacks the W-shaped patch of C. analis. Oligocottus
rimensis has a single large preopercular spine (single
pointed) and 3 smaller spines, similar to the "Myoxo-
cephalus" group (Washington et al. 1984), i.e., 4 pre-
opercle spines throughout their early development; the
dorsal spine elongates in juveniles. Oliocottus rimen-
sis juveniles also have no head cirri and the first pelvic
ray appears double (split in two).
Clinocottus analis differs from some Artedius (A.
fenestralis, A. lateralis, A. spp.) by having no large gut
diverticulae (wings). Species oi Artedius without wings
{A. creaseri) differ by having anterior gut pigment and
fewer PAVM (~10) (See Appendix 1).
Occurrence
Oblique bongo samples from coastal waters (8, 15, 22,
36, and 75 m depths) of the Southern California Bight
taken during the period 1978-84 (see Lavenberg et al.
(1986) for methods) indicate C. analis larvae (mostly
preflexion) were captured at the greatest densities
along the 15 m isobath off rocky tidepool areas in
southern California, especially off Palos Verdes Penin-
sula and Gaviota in 1979-80. Larvae occurred during
all months of the year, with peak abundance in July.
Wells (1986) found that C. analis spawn throughout the
year, with a peak in September-November in 1971-72,
based on gonosomatic index values and the appearance
of juveniles in the tidepools.
In discrete depth (neuston, middepth, epibenthic)
samples taken in the Southern California Bight in
October 1978 and June 1979-July 1980, 100% of the
C. analis larvae (almost exclusively notochordal and
flexion sizes) were caught in epibenthic samplers (ben-
thic bongo or auriga nets) indicating the smaller lar-
vae are near the bottom. Large postflexion individuals
were common in neuston tows (manta nets) taken dur-
ing the Coastal Resources Program (1978-79 except
October; not fully sorted to date) at Coho Bay (Pt. Con-
ception), and Playa del Rey and Seal Beach (stations
on each side of Palos Verdes; no station at Palos
Verdes) indicating larger postflexion, metamorphosing
larvae are located near the surface. Ninos (1984) col-
lected many larger postflexion larvae (~10mm) dur-
ing surface night-lighting at Catalina Island. At Palos
Verdes, juveniles (<25mm) are found back in the in-
tertidal in small pools, separated from the larger adults
(Wells 1986).
Description of Orthonopias triads larvae
Distinguishing cliaracters Distinguishing characters
for Orthonopias truwis larvae include a heavy cap of
pigment on the dorsoposterior gut, 26-55 PAVM, nape
melanophores usually absent, no wings, short gut
Feeney Post-yolksac development of two southern California sculpins
461
(preanal length 31.5-44.8% SL in preflexion larvae),
and 4 preopercular spines in late-flexion and postflex-
ion larvae. Postflexion larvae and juveniles have an
anus advanced from the anal fin. Juveniles develop
rows of spiny scales between the dorsal fin and lateral
line.
Morphology Orthonopias triads larvae hatch at
2. 9-3. 8mm (Bolin 1941); field-collected larvae are as
small as 2.6mm (after preservation). At 4.3 mm, the
caudal fin anlage is forming (Fig. 5B). Flexion occurs
in larvae between 4.2 and 7.2 mm (Table 3).
In preflexion larvae, the preanal length averages
39% SL (Table 1). During flexion the preanal length in-
creases to an average of 42% SL. Postflexion preanal
length averages 43.5% SL. Small juveniles (13.2 mm,
LACM W67-153, not illustrated) also have an average
preanal distance of 43.5% SL.
In postflexion larvae, the anus starts to advance
anteriorly from the developing anal fin. In larger post-
flexion larvae (Fig. 6C), a cirrus forms on the orbit.
Small juveniles (13.2mm, LACM W67-153, not illus-
trated) have a cirrus on the orbit and one in the inter-
orbital space; they also have lateral line scales and scale
bands under the dorsal fin.
Larger juveniles (Fig. 6D) have numerous cirri and
spines on the head; a smaller cirrus forms on the max-
illary and cirri develop between the preopercular
spines. The anus is located about halfway to pelvic
origin.
Fin development In postflexion larvae, complete
rays are formed by 7.2mm in all fins except the pelvics,
which are present as buds (Table 3).
Spination Preopercular spines start to form in
Orthonopias triads during flexion at 4.2-5.8 mm
(Fig. 5C, Table 3). Postflexion larvae typically have
4 preopercular spines of about equal size and equally
spaced (Fig. 6B). Sometimes an accessory preoper-
cular spine is present; a 7.3 mm larva possesses a
smaller spine adjacent to the 2 large spine from the
top.
Small juveniles (13.2mm, LACM W67-153, not illus-
trated) still retain the 4 preopercular spines (Table 3).
In larger juveniles (Fig. 6D, Table 3), preopercular
spines are reduced to 3 and a bump where the ventral-
most one used to be; the dorsalmost spine becomes
bifurcate.
In large postflexion larvae (Fig. 6C), nasal spines are
present. Three post-temporal/supracleithral spines
appear above the opercular flap near the point where
the lateral line will start to form. A small foramen is
present on the parietals where a sensory canal forms.
Table 3
Meristics of larvae and juveni
es of Orthonopi
as triads (sped- 1
mens
mside the two dashed lines
ire undergoing flexion of |
the nc
tochord).
Size
(mm)
D:
D, A P
V
PS
PCV CV TV
M
PAVM
2.6
_
0
34
40
3.4
—
_ _ _
—
0
—
—
—
36
51
4.3
-
- - -
-
0
-
-
-
36
30
4.2
_
_ _ _
1
37?
43
4.9
_
_ _ _
—
2
—
—
—
34
35
5.5
—
— — —
—
2
—
—
—
35
35
6.5
—
_ _ _
—
4
—
—
—
35
27
7.2
-
- - -
-
4
-
-
-
34
28
6.8
IX? 16? 11? 14? buds 5*
34
25
7.3
IX? 16? 12? 14? buds 4
—
—
—
35
33
8.2*
IX
16 12 14
1,3
4
11
24
35
—
—
9.2
IX
16 12 14
1,3
4
—
—
—
35
7?
13.2
IX
17 12 15
1,3
4
—
—
—
—
1
17.2
IX
17 12 14
1,3
4
—
—
—
—
0
23.0t
IX
17 12 14
1,3 3tT
) preoperc
11 24 35 - 8
le spine has smaller spine
* Second (from dorsun-
next to it.
* * cleared
and stained
arvae
t x-rayed
tt Dorsal preopercle sp
ne has a double point.
Pigmentation Yolksac Orthonopias triads have
pigmented eyes at hatching, a cap of dense pigment
on the dorsoposterior gut, and about 35 PAVM that
start on the 3d or 4th postanal myomere. One or a pair
of head melanophores is sometimes present (Bolin
1941).
The dorsoposterior gut pigment in field-collected
larvae is composed of ~80-90 melanophores in a cir-
cular pattern (Fig. 5 A). Small larvae (<4mm) have
32-55 PAVM (Fig. 7); preflexion larvae >4mm have
26-43 PAVM. One or two head melanophores over the
midbrain occur in about 33% of preflexion larvae. Nape
pigment is usually absent; one punctate melanophore
occurs at the nape in about 25% of preflexion larvae.
Flexion larvae have similar pigment as above (Fig.
5D, 6A). The first few PAVM are formed as dashes of
pigment at the start of the anal fin base (Fig. 6A). A
7.0mm specimen (Washington et al. 1984) has at least
3 head melanophores and 1 nape melanophore, and
seems to be just completing flexion. A 5.8 mm specimen
(that may have shrunk to a greater extent because it
was ETOH-preserved) completing flexion has 12 head
melanophores and 2 nape melanophores (LACM 009-
80-36-BB-Ol).
462
Fishery Bulletin 90(3), 1992
Figure 5
Field-collected Orthonopias triads larvae: (A) 3.3mm (LACM 026-PV-22-OB-02p), (B) 4.3mm (LACM 026-PV-15-OB-01p),
(C) 5.8mm (SIO H51-19B), (D) 5.5mm (LACM 012-80-08-BB-Ol).
Feeney Post-yolksac development of two southern California sculpins
463
/;g2:22^^,_
Figure 6
Field-collected Orthonopias triads larvae and juveniles: (A) 5.9mm (LACM 026-PV-15-OB-0:p), (B) 7.2mm (LACM
012-88-36-BB-Ol). (C) 9.2mm (SIO H5M9B), (D) 23.0mm (LACM 9423-8).
464
Fishery Bulletin 90(3). 1992
Postflexion PAVM pigment takes the form of dashes
at the base of each anal ray. At least 3 melanophores
can be found on the head over the midbrain. In larger
postflexion larvae, the number of PAVM is greatly
reduced (Fig. 7).
Small juveniles (13.2 mm, LACM W67-153, not illus-
trated) have numerous melanophores over the mid-
brain. Few or no PAVM may be present (Fig. 7). In
the 13.2mm juvenile, a dark patch of melanophores on
the pectoral fin base extends to and around the pelvic
girdle and meets at the ventral midline. Bands of pig-
ment extend down from the dorsum and stop just ven-
tral to lateral line.
In larger juveniles (Fig. 6D) a patch of melanophores
is present on the pectoral fin base, but may not be con-
tinuous across the pelvic girdle as it is in the 13.2 mm
juvenile. Light circles appear in the dense pigment
below the lateral line.
Meristics Meristics for 0. triads (Table 3) are com-
parable to published accounts. Modes for the fin ele-
ments matched those given in Howe and Richardson
(1978). Vertebrae (35) were 1 greater than the mode
(34) in Howe and Richardson. Branchiostegal rays (BR)
form during flexion; a 5.6mmFL larva had 5 visible
BR. In postlarvae and juveniles, branchiostegal rays
= 6, PCR = 6-h6.
Comparison with otiner species
Orthonopias triads are similar to Artedius meanyi
larvae (Washington 1986) by possession of 4 preoper-
cular spines, a short compact gut, and an eye cirrus;
A. meanyi postflexion larvae and juveniles also develop
small, prickly scales on the head and below the dorsal
fin. Artedius meanyi differ in having far fewer PAVM
(<13), pigment in the dorsal finfold, anterior gut
melanophores, and in undergoing flexion at a larger
size (6.2-9.4 mm). Artedius meanyi and 0. triads were
put in the "Myoxocephalus" group by Washington et
al. (1984) due to the presence of 4 preopercular spines.
Orthonopias triads larvae are similar to others
within the "Myoxocephalus" group, including Icelinus
and Chitonotus, in having no heavy nape pigment and
a high number of PAVM; Icelinus quadriseriatus has
25-63 PAVM (Feeney 1987) and Chitonotus has 24-45
PAVM (Richardson and Washington 1980). Orthono-
pias triads lacks ventral gut pigment (see Appendix 1).
Orthonopias triads can not be assigned to the "Ar-
tedius/Clinocottus/Oligocottus" group, as tentatively
suggested by Richardson (1981), because it lacks the
multiple preopercular spine pattern, gut diverticulae,
and trailing gut. Clinocottus analis postflexion larvae
(this paper) are similar to 0. triads because of the
advanced anus and presence of head pigment, nasal
>
Figure 7
Frequency of postanal ventral melanophores (PA\'TM) vs.
length (mm) with a LOWESS regression line at F = 0.5 (half
the points included in a running window) for Orthonopias
triads larvae and juveniles.
spines, post-temporal/supracleithral spines, cirri over
the eye, and similar meristics. Clinocottus analis dif-
fers in having multiple (>5) preopercular spines, a 'W'
shaped pigment patch on the side of the body, and a
longer gut (preanal = 46.0-54.5% SL vs. 39.i-48.3%
SL in 0. triads).
Orthonopias triads larvae initially have more PAVM
than C. analis; however, the number of PAVM de-
creases with length more quickly than C. analis (Figs.
6, 7); linear regressions (not shown) of 0. triads
PAVM have a greater negative slope (-2.413 vs.
- 1.161) than C. analis. Linear regression lines were
not used, however, in the final plots (Figs. 6, 7) because
LOWESS smoothing (Wilkinson 1989) indicates that
the relationship between PAVM and length may be
nonlinear, especially in 0. triacis. Additional large
postflexion and juvenile specimens need to be exam-
ined to verify this relationship.
Occurrence
During 1978-84, 0. triacis larvae (like C analis) were
collected in highest densities off Palos Verdes and other
rocky areas, at the 15 m isobath during the entire year,
peaking in spring and summer. Approximately 72% of
the larvae in discrete depth tows were collected in
epibenthic tows and none in neuston tows. Flexion lar-
vae were rarely collected. Postflexion individuals have
not been found in the 1978-79 neuston tows as were
Feeney: Post-yolksac development of two southern California sculpins
465
C. analis. Postflexion/metamorphosing individuals
apparently do not exhibit neustonic behavior like C.
analis. Juvenile 0. triads have been collected sub-
tidally on reefs and off rocky areas (LACM collection
data).
Conclusions
Clinocottus analis larvae can be grouped with the
"Artedius" group of cottid larvae based on the high
number of preopercular spines (9-12) (Washington et
al. 1984). The advanced anus of postflexion larvae is
typical of Clinocottus. Swank (1988) showed that Cli-
nocottus analis is more closely related to other species
within the genus rather than to Oligocottus ynaculosiis;
C. analis was found to be the most divergent in the
genus. Larval characters presented here lend support
to her conclusions. Clinocottus analis larvae share
many characters with other Clinocottus, but still have
some significant differences, i.e., a high PAVM count
and development of prickly spines on the body.
Orthonopias triads larvae can be grouped, along
with A. creaseri, A. meanyi, Chitonotus, and Icelinus,
in the "Myoxocephalu^" group (Washington et al. 1984)
because of the presence of four preopercular spines.
Body morphology of 0. triads is most similar to A.
meanyi.
Orthonopias triads and C. analis preflexion larvae
co-occur in the same areas (15 m isobath near rocky
habitats), but can be easily distinguished using pigment
and morphological characters. Larger postflexion lar-
vae can be distinguished by the number of preopercular
spines.
Acknowledgments
I thank the following for their help: David Ambrose,
Daniel Cohen, Pamela and Lissette Feeney, Javier
Gago, Robert Lavenberg, Gerald E. McGowen, Geof-
frey Moser, Margaret Neighbors, Debra Oda, Brenda
and Jim Rounds, Jeremyn Schmitz, Helga Schwarz,
Jeffrey Seigel, Camm C. Swift, H.J. Walker, Brian
White, William Watson, and the Natural History Mu-
seum of Los Angeles County.
Citations
Bolin, R.L.
1941 Embryonic and early larva! stages of the cottid fish Or-
thonopias triads Starks and Mann. Stanford Ichthyol. Bull.
2:73-82.
Brewer, G.D.. and P.E. Smith
1982 Northern anchovy and Pacific sardine spawning off
southern California during 1978-80: Preliminary observations
of the importance of the nearshore coastal region. Calif. Coop.
Oceanic Fish. Invest. Rep. 23:160-171.
Brewer, G.D., R.J. Lavenberg, and G.E. McGowen
1981 Abundance and vertical distribution of fish eggs and
larvae in the Southern California Bight: June and October
1978. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 178:165-167.
Budd, P.L.
1940 Development of the eggs and early larvae of six Califor-
nia fishes. Calif. Div. Fish Game, Fish Bull. 56. 50 p.
Eigenmann, C.H.
1892 The fishes of San Diego, California. Proc. U.S. Natl.
Mus. 15:123-178.
Eschmeyer, W.N., E. Herald, and H. Hammann
1983 A field guide to Pacific Coast fishes of North America.
Peterson Field Guide Ser. 28, Houghton Miffin, Boston, 336 p.
Feeney, R.F.
1987 Development of the eggs and larvae of the yellowchin
sculpin, Icelinus qiutdriseriatus (Pisces: Cottidae). Fish. Bull.,
U.S. 85:201-212.
Howe, K.M., and S.L. Richardson
1978 Taxonomic review and meristic variation in marine
sculpins (Osteichthyes: Cottidae) of the northeast Pacific
Ocean. Final rep. NOAA-NMFS Contract 03-78-M02-120.
School Oceanogr., Oreg. State Univ., Corvallis, 142 p.
Hubbs, C.
1966 Fertilization, initiation of cleavage, and developmental
temperature tolerance of the cottid fish, Clinocottus analis.
Copeia 1966:29-42.
Lavenberg, R.J.. G.E. McGowen, A.E. Jahn, and J.H. Petersen
1986 Abundance of Southern California nearshore ichthyo-
plankton: 1978-1984. Calif. Coop. Oceanic Fish. Invest. Rep.
27:53-64.
Matarese, A.C., A.W. Kendall Jr., D.M. Blood, and B.M. Vinter
1989 Laboratory guide to early life history stages of northeast
Pacific fishes. NOAA Tech. Rep. NMFS 80, 652 p.
Miller, D.J.. and R.N. Lea
1972 Guide to coastal marine fishes of California. Calif. Dep.
Fish Game, Fish Bull. 157, 249 p.
Morris. R.W.
1951 Early development of the cottid fish, Clinocottiis recalvus
(Greeley). Calif. Fish Game 37:281-300.
Ninos, M.
1984 Settlement and metamorphosis in Hypsoblennius (Pisces,
Blenniidae). Ph.D. diss., Univ. South. Calif., Los Angeles,
181 p.
Richardson, S.L.
1981 Current knowledge of larvae of sculpins (Pisces: Cottidae
and allies) in northeast Pacific genera with notes on relation-
ships within the family. Fish. Bull., U.S. 79:103-121.
Richardson, S.L., and B.B. Washington
1980 Guide to the identification of some sculpin (Cottidae) lar-
vae from marine and brackish waters off Oregon and adjacent
areas of the northeast Pacific. NOAA Tech. Rep. NMFS
Circ-430, 56 p.
Smith, P.E., and S.L. Richardson
1977 Standard techniques for pelagic fish egg and larva
surveys. FAO Fish. Tech. Pap. 175, 100 p.
Swank, E.S.
1988 Biochemical systematics of the genus Clinocotttis (Cot-
tidae). Bull. South. Calif. Acad. Sci. 87(2):57-66.
466 Fishery Bulletin 90(3). 1992
Washington, B.B. Wells, A.W.
1986 Systematic relationships and ontogeny of the sculpins 1986 Aspectsof ecology andlife history of the woolly sculpin,
Artedius, Clinocottiis, and Oligocottus (Cottidae: Scorpaeni- Clinocottus analis, from Southern California. Calif. Fish
formes). Proc. Calif. Acad. Sci. 44:157-223. Game 72:213-226.
Washington, B.B., W.N. Eschmeyer, and K.M. Howe Wilkinson, L.
1984 ScorpaenLformes: Relationships. /« Moser, H.G.. et al. 1989 Sygraph: The system for graphics. Systat, Inc., Evans-
(eds.), Ontogeny and systematicsof fishes, p. 438-447. Spec. ton, 600 p.
Publ. 1, Am. Soc. Ichthyol. Herpetol. Allen Press, Lawrence,
KS.
Appendix 1 : Key to known southern California sculpin larvae (preflexion stage)
Comments: The following key is provided as a guide to identifying known Southern California sculpin larvae.
Some larvae may not key out exactly due to variation in pigment or because they are a species that is not de-
scribed yet (see table at end of key). Types (e.g., Artedius type 16) are named as they are labeled in the LACM
collection. Equivalent types in literature are noted.
1 Wings (gut diverticulae) present 2
Wings absent (or as bumps only) 4
2(1) Postanal ventral melanophores 3-12 3
Postanal ventral melanophores 22-32 Artedius lateralis
3(2) Nape pigment present; myomeres 32-35 Artedius type 16
(= Artedius type 3 (Washington 1986); may be either A. corallinus or A. notospilosus)
Nape pigment absent; myomeres 36-37 Artedius type A
(= undescribed; has wings like other Artedius, no nape pigment, and only 3-8 PAVM)
4(1) Body covered with melanophores 5
Body melanophores restricted to ventral midline, gut, nape or head region 9
5(4) ~26-28 myomeres; stubby body (hatches at 6-7 mm) Rhamphocottus richardsonii*
36-41 myomeres; elongate body 6
6(5) Elongate pectoral fin (>20% of the body); pigment extending into dorsal and anal finfold
(hatches at ~7-9mm) Nautichthys oculofasciatus*
Pectoral fins not elongate (<15% of body); no pigment extending into fins 7
7(6) Lateral body relatively unpigmented; preanal length 36-42% of notochordal
length Hemilepidotus spinosus
Lateral body covered with melanophores; preanal length ~44-46% of notochordal length 8
8(7) Series of elongate melanophores present along the lateral midline; pointed snout Radulinus sp.
No distinct series of elongate melanophores along the lateral midline; blunt
snout Scorpaenichthys marmoratus
9(4) Otic capsule pigment present (may be present in Paricelinus hopliticus) 10
Otic capsule pigment absent 11
Feeney: Post-yolksac development of two southern California sculpins 467
10(9) Nape pigment present (several melanophores); dorsal gut pigment not in
bands Oligocottus/Clinocottus type D
(= C. recalvus (Morris 1951) or C. globiceps or 0. maculosus (Washington 1986). These
specimens are most similar to larger C. recalvus from same locality; however, they share
characters with all three species (e.g., otic capsule pigment) or with one species (e.g., nape
bubble like 0. maculosus).
Nape pigment absent (or 1 melanophore occipitally); dorsal gut pigment in
bands Leptocottus armatus
11(9) Nape pigment present 12
Nape pigment absent (rarely present in Orthonopias triacis and flexon Chitonotus pugetensis) ... 19
12(11) Anterior gut pigment present; ventral gut pigment present 13
Anterior gut pigment absent; ventral gut pigment absent 14
13(12) Myomeres 40-42; postanal ventral melanophores ~37-38; pigment on
snout Paricelinus hopliticus*
Myomeres 27-30; postanal ventral melanophores <15; pigment on snout
absent Enophrys taurinal
(= undescribed; only one specimen collected, similar to E. bison description (Richardson and
Washington 1980) except for scattered ventral gut pigment)
14(12) Postanal ventral melanophores 5-14 15
Postanal ventral melanophores 15-33 16
15(14) Preanal myomeres 7-9; head pigment usually absent; gut pigment light and
scattered Oligocottus synderi
Preanal myomeres 10-12; head pigment present; gut pigment
heavy Oligocottus/Clinocottus type B
(= undescribed; looks much like C. analis but has only 10-14 PAVM)
16(14) Nape melanophores <6 17
Nape melanophores >6 18
17(16) Postanal ventral melanophores 15-21; dorsal gut pigment light to
moderate Clinocottus embryum *
Postanal ventral melanophores 21-33; dorsal gut pigment heavy Artedius harringtoni
18(16) Head pigment absent; dorsal gut pigment moderate; nape melanophores
~18 Oligocottus/Clinocottus type C
(= undescribed; no head pigment, 17-31 PAVM; lighter pigment than C. analis, may be
0. maculosus)
Head pigment present (except in smallest larvae (<3.0mm)); dorsal gut pigment heavy;
nape melanophores < 17 Clinocottus analis
19(11) Postanal ventral melanophores 7-18 20
Postanal ventral melanophores ^24 21
20(19) Ventral gut pigment absent; preanal myomeres 8-10 Artedius creaseri
Ventral gut pigment present; preanal myomeres 11-12 Cottus asper
468 Fishery Bulletin 90(3). 1992
21(19) Ventral gut pigment present 22
Ventral gut pigment absent Orthonopias triads
22(21) ~5 parallel lines (striations) of pigment oriented horizontally on posterior gut .... Synchirus gilli*
No parallel lines of pigment on posterior gut 23
23(22) Two or more prominent anterior gut melanophores present 24
Anterior gut melanophores absent (or one only) 26
24(23) Dorsal head pigment absent; jaw angle pigment present 25
Dorsal head pigment present (except in larvae ~3.5mm or smaller); jaw angle pigment
usually absent Chitonotus pugetensis
25(24) Anterior gut melanophores ~2; myomeres 33-37 IcelinusI Chitonotus
( = not a type, but a category for ambiguous or damaged specimens that may be either Icelinus
or Chitonotus)
Anterior gut melanophores ~10; myomeres 38-40 Icelinus type A
(= undescribed; probably /. tenuis based on high myomere counts (38-40))
26(23) Jaw angle (quadrate) pigment present Icelinus quadriseriatus
Jaw angle pigment absent IcelinusI Chitonotus
* I have not examined larvae of this type; characters were taken from the literature.
Appendix table
The following larvae have not been described and are probably unknown, which should be taken into consideration when using the
key, especially if the larva(e) do not key out exactly.
Taxa Comments
Artedius corallinus 15-16 dorsal rays; A. type 3?; intertidal (see Washington 1986).
Artedius notospikitus A. type 3? (Washington 1986).
Icelinxis burchami fiiscescens Rare; found in deep water (126-549m); 16-18 dorsal rays.
Icelinus cavifrons X-XI dorsal spines (>98%); IX dorsal spines (<2%).
Icelinus filamentosus 15-18 dorsal rays.
Icelinus Jimbriatus Rare; found at moderate depths (60-265 m); X-XI dorsal spines.
Icelimis oculatus Rare; found in deep water (109-274 m); X-XI dorsal spines; 37 vertebrae.
Icelinus sp. nov. X dorsal spines; rare?
Leiocottus hirundo 16-17 dorsal rays; recent occurrence on reefs in Santa Barbara area and Santa Cruz I., California.
Oligocottus rimensis 16-19 dorsal rays; juveniles lack 'W'-shaped pigment on side; intertidal.
Oligocottus rubellio Juveniles lack 'W'-shaped pigment on side; intertidal.
Psychrolutes phrictus Rare, found in deep water (839-2800 m), 22-26 pectoral rays; may = "cottoid A" (Richardson and
Washington 1980).
Radulinus vinculus Rare; southern range limit is Anacapa I.
Zesticelus profundorum Rare; 25-26 vertebrae; deep water (88-2580 m).
Abstract. - To investigate the
genetic basis of stock structure of the
weakfish Cynoscion regalis, a total
of 370 individuals was collected from
four geographic sites along the mid-
Atlantic coast of the United States
over a period of 4 years. Restriction
fragment length polymorphism
(RFLP) analysis of weakfish mito-
chondrial DNA, employing either 6
or 13 restriction endonucleases, dem-
onstrated a low level of intraspecific
mtDNA variation, with a mean nu-
cleotide sequence divergence of 0.13%
for the pooled samples. The common
mtDNA genotype occurred at a fre-
quency of 0.91-0.96 in all samples,
and no significant heterogeneity was
found among samples in the occur-
rence of the common mtDNA geno-
type or rare variants. The lack of
spatial partitioning of rare mtDNA
genotypes among collection sites sug-
gests considerable gene flow along
the mid-Atlantic coast. Together
these data are consistent with the
hypothesis that weakfish comprise a
single gene pool, and indicate that
the fishery should be managed as a
single, interdependent unit.
A genetic analysis of \A/eal<fish
Cynoscion regalis stock structure
along the mid-Atlantic Coast*
John E. Graves
Jan R. McDowell
Virginia Institute of Marine Science, T-chool of Marine Science
College of William and Mary, Gloucester Point, Virginia 23062
M. Lisa Jones
Department of Biology, College of William and Mary, Williamsburg, Virginia 23185
The weakfish Cynoscion regalis is
broadly distributed along the Atlan-
tic coast of the United States. It is
common from Long Island NY to
Cape Canaveral FL, and has occa-
sionally been reported from as far
north as Nova Scotia and as far south
as the Gulf of Mexico (Bigelow and
Schroeder 1953, Weinstein and Yer-
ger 1976). Weakfish abtmdance varies
considerably on both a spatial and
temporal basis, especially in the
northern part of the species' range
(Bigelow and Schroeder 1953).
The life history of the weakfish has
been well studied (reviewed in Wilk
1979). Spawning occurs in estuarine
and coastal waters from late spring
through early fall, with a peak of ac-
tivity in late May and early June.
There appears to be little offshore
transport of the early-life-history
stages, and young-of-the-year remain
in shallow estuarine and coastal
waters during their first summer.
Like many fishes along the mid-
Atlantic coast, weakfish move off-
shore to overwinter as coastal waters
cool during the fall, returning in the
spring when inshore temperatures
increase.
The seasonal inshore and offshore
movements of weakfish could lead to
significant mixing of fish from differ-
Manuscript accepted 4 June 1992.
Fishery Bulletin, U.S. 90:469-475 (1992).
'Contribution 1749 of the Virginia Institute
of Marine Science and School of Marine Sci-
ence, College of William and Mary.
ent coastal areas. Tagging studies by
Nesbit (1954) showed that a large
proportion of weakfish tend to return
to the same coastal region in which
they were tagged, although many
fish were recaptured in areas distant
from the original tagging location.
The differential size distribution of
weakfish along the mid- Atlantic coast
is consistent with the hypothesis that
mixing of weakfish from different
coastal areas occurs. Larger (older)
weakfish are more predominant in
northern waters during the summer,
and the mean size of weakfish tends
to decrease as one moves down the
Atlantic coast (Wilk and Silverman
1976). Whether this represents an
ontogenetic change in seasonal move-
ments or differential survival or
growth of fish from different coastal
areas is not known.
Weakfish support an important
commercial and recreational fishery.
Commercial landings over the past
110 years have undergone dramatic
fluctuations. Combined commercial
and recreational landings were at a
recent peak during 1980 at 36,400
metric tons (t) and subsequently
dropped to about 19.lt over a period
of 2 years (Vaughan et al. 1991). The
total catch has remained fairly con-
stant for the last 8 years, although a
significant decline in landings from
northern waters has been noted over
the period (Vaughan et al. 1991). For
example, the combined commercial
469
470
Fishery Bulletin 90(3). 1992
and recreational catch in New York dropped from 840 1
in 1982, to 224 1 in 1986, to 10 1 in 1990 (NMFS Cur-
rent Fishery Statistics Series). The recreational catch
typically represents a sizable fraction of the total land-
ings, and at times surpasses the commercial catch (Wilk
1979).
Many weakfish are lost from the fishery as inciden-
tal by catch in shrimp trawling operations. The inciden-
tal weakfish bycatch, which is greatest in the southern
part of the species' range, consists mostly of young-
of-the-year fish. It is difficult to determine the magni-
tude of the weakfish bycatch, but it is estimated that
it exceeds the combined recreational and commercial
catch in the southern states (South Carolina, Georgia
and Florida) and may approach 30% of the total coastal
fishery for adults (Keiser 1976, Mercer 1983, Vaughan
et al. 1991).
The Fishery Management Plan for Weakfish was
adopted in 1985 by the Atlantic States Marine Fish-
eries Commission (Mercer 1985). At that time, the
genetic basis of weakfish stock structure was not well
understood, and most states have independently
managed their vveakfish fisheries. As a result, different
gear restrictions and minimum sizes are enforced along
the mid-Atlantic coast. For example, Florida, Georgia,
South Carolina, North Carolina, New Jersey, and Con-
necticut have no recreational minimum size limit, but
a 9-inch size limit is enforced in Virginia, 10 inches in
Maryland and Delaware, and 12 inches in New York
and Rhode Island.
A thorough understanding of weakfish stock struc-
ture is essential for effective management of the fish-
ery. Several management decisions require knowledge
of the interdependence of fishery resources from dif-
ferent coastal areas. The recent decline in landings
from northern waters has coincided with increased
catches of large weakfish in the North Carolina winter
offshore (sinknet) fishery (Vaughan et al. 1991), but it
is not known if the two fisheries operate on the same
stock of fish. On a larger geographic scale, the rela-
tionship between bycatch mortality of young weakfish
in southern waters and landings of older weakfish in
northern waters in subsequent years has not been
determined. A detailed genetic analysis of weakfish
stock structure would provide information required to
test hypotheses of the independence of weakfish from
different coastal areas.
Several studies have investigated weakfish stock
structure employing a variety of ecological and mor-
phological techniques including mark and recapture
data (Nesbit 1954), scale circuli patterns (Perlmutter
et al. 1956), morphological characters (Seguin 1960,
Scoles 1990), and relative growth rates and reproduc-
tive characters (Shepherd and Grimes 1983, 1984).
Most of these studies concluded that weakfish comprise
two or more stocks; however, the inability to distin-
guish between ecophenotypic and genetic character
variation in these studies has confounded interpreta-
tion of the results.
There are few studies on the biochemical genetics of
the weakfish. Crawford et al. (1989) analyzed water-
soluble protein variation using starch gel electro-
phoresis. They found no significant genetic differen-
tiation between weakfish collected from Buzzards Bay
MA to Cape Hatteras NC, and so were unable to
disprove the null hypothesis that weakfish along the
mid- Atlantic coast share a common gene pool. Of the
25 protein-encoding loci surveyed in the Crawford et
al. (1989) study, only two were polymorphic within the
pooled sample, and the mean heterozygosity was low
relative to other marine fishes.
Studies of protein variation have been extremely
useful in demonstrating the intraspecific genetic struc-
ture of many marine fishes (reviewed in Ryman and
Utter 1987). For those species which display little in-
traspecific variation, like the weakfish, sample sizes
must be very large to detect significant differentiation
between putative stocks, if it exists. In such cases,
analysis of a more rapidly evolving set of molecular
characters may provide a better estimate of population
structure with a more manageable number of samples.
Restriction fragment length polymorphism (RFLP)
analysis of mitochondrial DNA (mtDNA) has provided
such a tool, and the technique has been useful in resolv-
ing stock structure within species which exhibit little
protein variation (reviewed by Ovenden 1990).
This paper reports the results of an RFLP analysis
of weakfish mtDNA to determine if fish along the mid-
Atlantic coast share a common gene pool. The study
began as a spatial and temporal investigation of a large
number of individuals from a few collection sites along
the central mid-Atlantic coast with 6 restriction endo-
nucleases. Because there was a high degree of genetic
homogeneity among these samples, we expanded the
investigation to include an intensive analysis of weak-
fish from the northern and southern ends of the species'
range with 13 restriction endonucleases.
Materials and methods
For our sampling protocol we assumed that if separate
genetic stocks of weakfish exist, they should be separ-
ated at the time of spawning. We therefore restricted
our collections to female weakfish that were ready to
spawn as evidenced by high gonadosomatic indices
(GSI). For example, the mean GSI of the New York
1988 sample was 7.7% + 3.1 SD.
Ripe, female weakfish were obtained from commer-
cial fishermen, sportfishing tournaments, and the
Graves et al : Genetic analysis of Cynosaon regalis stock structure
471
National Marine Fisheries Service In-
shore Trawl Survey. Dates and locations
of capture, and the size composition of
the collections are presented in Table 1.
Freshly-caught weakfish were measured
for standard length and then dissected.
Ovaries were removed and quickly frozen
at -20°C.
Mitochondrial DNA was obtained by
the rapid isolation procedure of Chapman
and Powers (1984) for the initial survey
of mtDNA genetic heterogeneity. After
ethanol precipitation, the mtDNA-en-
riched DNA pellet was rehydrated in 75
^L of distilled water, and the yield from
~7g of ovarian tissue was usually suffi-
cient for at least 7 restriction digestions
visualized with ethidium bromide.
The 1988 and 1989 samples were sur-
veyed with the following six restriction
endonucleases: Aval, BamUl, Bgll,
Hindlll, Neil, and Pvull. Restriction
fragments were separated by horizontal
gel electrophoresis on 0.8-1.5% agarose
gels run at 2 v/cm for 16 h. Gels of restric-
tion digestions of isolations containing
high yields of mtDNA were visualized
after ethidium bromide staining with
ultraviolet light illumination (Maniatis et
al. 1982) and photographed with a Pola-
roid CU-5 camera using a Wratten #5
filter. For those samples in which there
was not sufficient mtDNA for direct
visualization, restriction digestions were
endlabeled before electrophoresis with
the appropriate ^^S nucleotide triphos-
phate using the Klenow fragment (Maniatis et al. 1982).
After electrophoresis, the gels were rinsed with a scin-
tillation enhancer, dried, and autoradiographed at
-70°C for 5d.
To compare the genetic relationship of weakfish from
the northern and southern ends of their range in great-
er detail, mtDNA was purified from ovarian tissues
using the CsCl density-gradient centrifugation protocol
of Lansman et al. (1981). The mtDNA from these
samples was surveyed with the 6 restriction endonu-
cleases listed above and the following 7 enzymes: Apal,
Avail, Banl, Bell, EeoRV, Hindi, and Hhal. The
restriction digestions were endlabeled, separated on
agarose gels, and autoradiographed as described above.
The different fragment patterns produced by each
restriction endonuclease were each assigned a letter.
A composite mtDNA genotype, consisting of the letters
representing the fragment patterns generated by each
restriction endonuclease, was then constructed for each
Table 1
Weakfish Cynoscion regalis
collection data.
Standard length
Sample
Location
Date
N
Mean, range (mm)
NY88
Long Island NY
6/88
58
560, 254-750
NY89
Long Island NY
5/89
65
619, 216-750
DE88
Lewes DE
5/88
74
597, 256-761
DE89
Lewes DE
5/89
51
521, 290-744
DE91
Lewes DE
6/91
25
522, 390-670
NC88
Hatteras NC
6/88
72
278, 221-342
S091
SC, GA. FL
.S/91
25
201, 174-273
Table 2
Distribution of weakfish Cy
wscion
•egalis m
DNA genotypes
based on 6 re- 1
striction endonucleases among the different collection
s. The order of restric- |
tion enzyme
morphs
represented from left to
right, is
mytdlU
, Pi'ii II
Bgll.
BamHl, Neil, and Aval. Fragment sizes for each restriction pattei
■n are
available from the senior author upon request.
Genotype
Sample
Total
NY88
NY89
DE88
DE89
DE91
NC88
S091
AAAAAA
55
60
67
48
22
69
24
345
AAAAAC
0
2
2
0
1
1
0
6
AAAAAB
0
0
2
0
1
0
1
4
AAAAFA
1
0
0
1
0
1
0
3
AAAADA
0
2
0
1
0
0
0
3
ABAAAA
1
1
0
0
0
0
0
2
AAAABA
1
0
1
0
0
0
0
9
AAAACA
0
0
2
0
0
0
0
2
DAAAAB
0
1
0
0
0
0
0
1
AAAAEA
0
0
0
0
0
1
0
1
CAAAAA
0
0
0
0
0
1
0
1
Total
58
66
74
50
24
73
25
370
individual. The nucleon diversity (Nei 1987) was cal-
culated for each sample and for the pooled samples.
The percent nucleotide sequence divergence between
mtDNA genotypes was estimated by the site approach
of Nei and Li (1979) and the percent mean nucleotide
sequence divergences within and among weakfish
samples were calculated following the method of Nei
(1987), with the latter value being corrected for within-
group heterogeneity. The distribution of genotypic
frequencies was evaluated for homogeneity between
collections using the G-test (Sokal and Rohlf 1981).
Results
Weakfish mtDNA demonstrated very little variation.
Of the 370 weakfish surveyed with 6 restriction en-
donucleases, 345 shared a common mtDNA genotype
(Table 2). Ten variant genotypes were encountered in
472
Fishery Bulletin 90(3). 1992
the analysis, distributed among the remaining 25 in-
dividuals. In each case, variant genotypes were close-
ly related to the common genotype, differing by no
more than two restriction site changes. Restriction site
gains or losses were inferred from completely additive
changes in fragment patterns. No length polymor-
phisms or heteroplasmy were observed. A total of 43
restriction sites were detected in the 6-enzyme survey,
and the average individual was scored for 29 sites,
representing about 1.0% of the weakfish mtDNA
genome.
The common mtDNA genotype, AAAAAA, occurred
in the great majority of fish in all samples, ranging in
frequency from 0.905 (DE88) to 0.960 (DE89, S091),
with a value of 0.932 for the pooled sample (Table 2).
The next most-common genotype, AAAAAC, occurred
in the pooled sample at a frequency of 0.017, and was
present in 4 of the 7 samples. Because of the predom-
inance of a single genotype in all samples, nucleon
diversities were relatively low (Table 3), ranging from
0.079 (DE89) to 0.180 (DE88), with a value of 0.130
for the pooled samples. As all variant mtDNA geno-
types were related to the common genotype by no more
than 2 restriction site changes, the percent mean
nucleotide diversity within each sample was also quite
low (Table 3), ranging from 0.06 (DE89) to 0.18 (NY89).
Little genetic differentiation was detected among
weakfish samples collected at the same location over
2 or more years (Table 4). Among samples collected in
Delaware during 1988, 1989, and 1991, and in New
York during 1988 and 1989, the percent mean nucleo-
tide sequence divergences, corrected for within-sample
variation (Nei 1987), ranged from 0.00 (DE88/DE91,
NY88/NY89) to 0.01 (DE89/DE91).
Little genetic differentiation was encountered among
samples of weakfish collected along the mid-Atlantic
coast. The nucleotide sequence divergences among
samples collected at geographically distant sites dur-
ing the same spawning season ranged from 0.00
(NY88/NC88) to 0.03 (NY89/DE89). These values are
of the same magnitude as those found among samples
of weakfish collected at the same site over 2 or more
years, indicating a lack of spatial genetic structuring.
An analysis of the distribution of mtDNA genotypes
also revealed no significant heterogeneity among tem-
poral or spatial collections. To avoid a bias caused by
including expected values <1, we initially pooled all
alternate genotypes for an analysis of heterogeneity.
The results of a G-test (Sokal and Rohlf 1981) revealed
no significant spatial or temporal differences in the
distribution of the common and pooled rare genotypes
among the 7 samples (Gh 2.88, 0.75>p>0.50). Ex-
panding the analysis to include the common genotype
and all 10 rare genotypes separately (each with ex-
pected values < 1 in one or more collections) did not
Table 3
Genotypic diversity and percent mean nucleotide sequence
diversities for 7 weakfish Cynoscion regalis collections. Values
for both 6 and 13 restriction enzyme surveys are listed for
the Delaware (DE) 1991 and southern (SO) 1991 samples.
Percent mean
Sample
Nucleon
diversity
nucleotide sequence
diversity
NY88
0.102
0.09
NY89
0.171
0.18
DE88
0.180
0.15
DE89
0.079
0.06
DE91 (6)
0.163
0.19
(13)
0.239
0.10
NC88
0.107
0.11
S091 (6)
0.080
0.09
(13)
0.080
0.04
Total
0.130
0.13
Table 4
Percent
mean nucleotide
sequence divergences
between weak- |
fish Cynoscion
regalis collections. Values are
sased on
results
from 6 restriction endonucleases and have been corrected for |
within-group sequence diversity (N
ei 1987).
NY89
DE88
DE89
DE91
NC88
S091
NY88
0.00
0.00
0.00
0.01
0.00
0.00
NY89
—
0.00
0.03
0.00
0.00
0.00
DE88
—
—
0.00
0.00
0.00
0.00
DE89
—
—
—
0.01
0.02
0.00
DE91
—
—
—
—
0.00
0.00
NC88
—
—
—
—
—
0.00
significantly change the outcome. Once again, the null
hypothesis of homogeneity could not be disproved
(Gh 51.62, 0.50>p>0.25).
The low level of variation detected in our analysis
of weakfish mtDNA could be the result of many fac-
tors. After reviewing the 1988 and 1989 results, we
felt that we might have biased our estimates of mean
nucleotide sequence divergence by using 6 restriction
endonucleases that, by chance, were not variable within
weakfish. To test this hypothesis, we analyzed the
DE91 and S091 weakfish collections with an additional
7 restriction endonucleases. The average individual in
the 13-enzyme analysis was scored for 65 restriction
sites, or approximately 2.4% of the weakfish mtDNA
molecule. Of the 49 fish in the two samples surveyed,
only one variant mtDNA genotype was found (one fish
from the S091 sample with the common 6-enzyme
mtDNA genotype exhibited a site gain relative to the
common pattern for the enzyme Bell). As a result.
Graves et a\ Genetic analysis of Cynoscion regalis stock structure
473
the nucleon diversity, which is sensitive to the number
of enzymes employed, increased for only one of the two
samples. The within-sample percent mean nucleotide
sequence diversity, which is not as sensitive to the
number of restriction sites surveyed, was slightly lower
for both samples (Table 4), and the corrected values
of percent mean nucleotide sequence divergence be-
tween the DE91 and S091 samples was essentially the
same whether based on 6 or 13 informative restriction
enzymes.
Discussion
The presence of alleles unique to samples from par-
ticular geographic locations has been used to infer in-
traspecific genetic structuring and to determine levels
of gene flow among collection sites (Slatkin 1989). This
model assumes that increased frequencies of "private"
alleles are a direct result of limited gene flow. An in-
spection of the above data reveals several genotypes
that are present at very low frequencies (Table 2);
however, it is interesting to note that most genotypes
represented by more than a single individual occur in
two or more geographic samples. For example, the
genotype AAAAFA was encountered three times in
the analysis of 370 weakfish, occurring in the NY88,
NC88, and DE89 samples. The lack of spatial partition-
ing of rare alleles is strongly suggestive of a high rate
of gene flow among collection locations.
The level of mtDNA variation found within the weak-
fish is among the lowest reported for any species of
fish. While it is difficult to compare nucleon diversities
from different studies because the value is dependent
upon the number of restriction sites surveyed, relative
levels of variability can be determined from com-
parisons of studies involving about the same numbers
and types of restriction endonucleases. The nucleon
diversity of the 1991 weakfish samples surveyed with
13 enzymes was 0.157, a value that falls well below the
range of 0.473-0.998 reported by Avise et al. (1989)
for other fishes analyzed with about the same number
of enzymes. The nucleon diversity of the weakfish was
also substantially below the mean value of 0.943 re-
ported for the red drum Sciaenops ocellatus (Gold and
Richardson 1991). Relatively low values of nucleon
diversity have been found for the black drum Pogonias
cromis (0.584) and the spotted seatrout Cynoscion
nehulosus (0.531), a congener of the weakfish (C. Fur-
man and J.R. Gold, Texas A&M Univ., College Station,
pers. commun., Aug. 1991). However, these values are
still substantially larger than those we found for the
weakfish. Comparisons of nucleotide sequence diver-
sities among these species also indicate that the
weakfish is relatively depauperate in terms of mtDNA
variation. The mean percent nucleotide sequence diver-
sities within 7 black drum samples (0.142) and 5 spotted
seatrout samples (0.222) are substantially higher than
that within the 7 weakfish samples (0.10) surveyed in
this study.
The finding of relatively low levels of mtDNA varia-
tion within the weakfish is consistent with the lack of
allozyme variation reported by Crawford et al. (1989).
Low levels of mtDNA variation have generally been
attributed to small effective population sizes of females,
resulting in relatively rapid sorting of gene trees (Nei
1987, Avise et el. 1988, Chapman 1990, Bowen and
Avise 1990). Variations in weakfish abundance over the
last 110 years have been reflected in commercial
catches, which have fluctuated from a high of 44.5
million pounds in 1908 to a low of 3.1 million pounds
in 1967 (Vaughan et al. 1991), but it is unlikely that
such variations over recent history have drastically
reduced the effective population size of female weak-
fish. Population bottlenecks on a larger time-scale (e.g.,
glaciation events) or cyclical fluctuations in population
size may have resulted in the reduced genetic diver-
sity within the weakfish, but such explanations are
merely speculative and do not necessarily agree with
the observation that other sciaenids with similar
distributions and life histories do not exhibit such low
levels of mtDNA diversity.
Reductions in effective population size can also occur
due to differential reproductive contribution, resulting
from skewed sex ratios, limited mating opportunities,
or varying of survival among progeny. While little is
known of weakfish spawning behavior or differential
recruitment success, the sex ratio tends to be very close
to 1.0 (Wilk 1979). Thus, the cause or causes con-
tributing to the low genetic variation observed among
weakfish relative to other fishes is problematic.
In addition to low levels of within-sample variation,
we detected little temporal or spatial genetic differen-
tiation among weakfish samples. Because there were
few variant mtDNA genotypes, and almost all of the
rare variant genotypes occurred in more than one
population, the uncorrected mean nucleotide sequence
divergences among weakfish samples were of the same
magnitude as mean nucleotide diversities found within
samples. Thus, the mean difference among mtDNA
genotypes randomly drawn from within a single sample
was equivalent to the mean difference among mtDNA
genotypes drawn from different samples.
Low levels of within-group mtDNA variation do not
preclude the occurrence of significant between-group
differentiation. Bowen and Avise (1990) recently re-
ported low values of mtDNA diversity within samples
of Atlantic and Gulf of Mexico black sea bass Cen-
tropristis striata (within-sample percent nucleotide
sequence diversity of 0.03), yet their study revealed
474
Fishery Bulletin 90(3). 1992
significant differentiation between the two populations
(an uncorrected percent mean sequence divergence of
0.75). The lack of significant population structuring
within the weakfish relative to the black sea bass is
evidenced in a comparison of the ratio of between-
group to within-group sequence divergences: For the
black sea bass the ratio is 24, while for the weakfish
it is ~1.
The results of our investigation suggest that weak-
fish comprise a single genetic stock throughout the
species' range. No significant genetic differentiation
was found among geographic samples or among
samples taken at the same site over several years. Con-
sequently, at the level of genetic resolution we em-
ployed, we cannot disprove the null hypothesis that
weakfish share a common gene pool. The inference that
gene flow occurs throughout the species' range is sup-
ported by the homogeneous distribution of rare mtDNA
genotypes.
The genetic homogeneity found within the weakfish
in this study and in the allozyme analysis of Crawford
et al. (1989) contrast with the geographical variation
of morphological and life-history characters reported
in other studies (Perlmutter et al. 1956, Seguin 1960,
Shepherd and Grimes 1983 and 1984, Scoles 1990). The
degree of plasticity of weakfish morphological and life-
history characters to different environmental condi-
tions has not been determined, but in light of research
on other fishes (Barlow 1961), it would not be surpris-
ing if much of the geographic variation previously
described among weakfish is ecophenotypic.
Our inference that there is sufficient gene flow
among weakfish along the mid-Atlantic coast to pre-
vent even minor genetic differentiation from occurring
has several management implications. There is clear-
ly some interdependence among areas, a conclusion
also supported by the tagging data of Nesbit (1954).
To obtain a meaningful estimate of the magnitude of
the interdependence between these areas would require
a tagging study much more extensive than that of
Nesbit (1954), which would involve considerable time
and expense. Until such information is available, it
would be best to manage the weakfish resource con-
servatively, as a single interdependent stock.
Acknowledgments
This study would not have been completed without the
assistance of several individuals along the mid- Atlantic
coast. The perseverance and efforts of all those who
collected weakfish samples are gratefully acknowl-
edged, especially Alice Webber, New York Department
of Environmental Conservation; Rich Seagraves,
Delaware Division of Fish and Wildlife; and Charlie
Wenner, South Carolina Wildlife and Marine Re-
sources. A portion of this project was initiated by Herb
Austin and Brian Meehan, Virginia Institute of Marine
Science. Ana Beardsley provided technical assistance.
Carol Furman and John Gold kindly shared unpublished
data. Dan Scoles, Herb Austin, and Mark Chittenden
critically read the manuscript. Funding for this study
was provided by the Atlantic States Marine Fisheries
Commission (88-lWSID), the U.S. Fish and Wildlife
Service (F-60-R), and the Commonwealth of Virginia.
M. Lisa Jones was supported by a Howard Hughes
Summer Fellowship in Molecular Biology.
Citations
Avise, J.C, R.M. Ball, and J. Arnold
1988 Current versus historical population sizes in vertebrate
species with high gene flow: A comparison based on mitochon-
drial DNA lineages and inbreeding theory for neutral muta-
tions. Mol. Biol. Evol. 5:331-344.
Avise, J.C, B.W. Bowen, and T. Lamb
1989 DNA fingerprints from hypervariable mitochondrial
genotypes. Mol. Biol. Evol. 6:258-269.
Barlow, G.W.
1961 Causes and significance of morphological variation in
fishes. Syst. Zool. 10:105-117.
Bigelow, H., and W. Schroeder
1953 Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish.
Bull. .53, 577 p.
Bowen, B.W., and J.C. Avise
1990 Genetic structure of Atlantic and Gulf of Mexico popula-
tions of sea bass, menhaden, and sturgeon; Influence of zoo-
geographic factors and life-history patterns. Mar. Biol. (Berl.)
107:371-381.
Chapman, R.W.
1990 Mitochondrial DNA analysis of striped bass populations
in Chesapeake Bay. Copeia 1990:355-366.
Chapman, R.W., and D.A. Powers
1984 A method for the rapid isolation of mitochondrial DNA
from fishes. Tech. Rep. UM-SG-TS-84-01, Maryland Sea
Grant Prog., Univ. Md., College Park, 11 p.
Crawford, M.K., C.B. Grimes, and N.W. Buroker
1989 Stock identification of weakfish, Cynoscion regalia, in the
middle Atlantic region. Fish. Bull., U.S. 87:205-211.
Gold, J.R., and L.R. Richardson
1991 Genetic studies in marine fishes. IV. An analysis of
population structure in the red drum {Sciae7wps ocellatux) using
mitochondrial DNA. Fish. Res. (Amst.) 12:213-241.
Reiser, R.K. Jr.
1976 Species composition, magnitude and utilization of the in-
cidental catch of the South Carolina shrimp fishery. Tech.
Rep. 16, S.C. Mar. Resour. Cent., Charleston. 94 p.
Lansman, R.A., R.O. Shade, C.F. Shapira. and J.C. Avise
1981 The use of restriction endonucleases to measure mito-
chondrial DNA sequence relatedness in natural populations.
III. Techniques and potential applications. J. Mol. Evol.
17:214-226.
Maniatis, T., I.F. Fritsch, and J. Sambrook
1982 Molecular cloning: A laboratory manual. Cold Spring
Harbor Lab., Cold Spring Harbor NY, 545 p.
Graves et al : Genetic analysis of Cynoscion regalis stock structure
475
Mercer, L.P.
1983 A biological and fisheries profile of weakfish, Cynoscion
regalis. Spec. Sci. Rep. 39, N.C. Dep. Nat. Resour. & Com-
munity Dev., Div. Mar. Fish., 107 p.
1985 Fishery management plan for the weakfish (Cynoscion
regalis) fishery. Spec. Sci. Rep. 46, N.C. Dep. Nat. Resour.
& Community Dev., Div. Mar. Fish. Fish. Manage. Rep. 7,
Atl. States Mar. Fish. Comm., 129 p.
Nei, M.
1987 Molecular evolutionary genetics. Columbia Univ. Press,
NY. 512 p.
Nei. M., and W-H. Li
1979 Mathematical model for studying genetic variation in
terms of restriction endonucleases. Proc. Natl. Acad. Sci.
76:5269-5273.
Nesbit. R.A.
1954 Weakfish migration in relation to its conservation. U.S.
Fish Wildl. Serv., Spec. Sci. Rep. Fish. 115, 81 p.
Ovenden, J.R.
1990 Mitochondrial DNA and marine stock assessment: A
review. Aust. J. Mar. Freshwater Res. 41:835-53.
Perlmutter, A., S.W. Miller, and J.C. Poole
1956 The weakfish {Cynoscion regalis) in New York waters.
N.Y. Fish Game 3:1-43.
Ryman, N., and F. Utter (editors)
1987 Population genetics and fishery management. Univ.
Wash. Press, Seattle, 420 p.
Scoles, D.
1990 Stock identification of weakfish. Cynoscion regalis, by
discriminant function analysis of morphometric character-
istics. Masters thesis, College of William & Mary, Williams-
burg, 51 p.
Seguin, R.
1960 Variation in the Middle Atlantic coast population of the
grey squeteague, Cynoscion regalis (Bloch and Schneider),
1801. Ph.D. diss., Univ. Delaware, Newark, 70 p.
Shepherd, G.R., and C.B. Grimes
1983 Geographic and historic variations in growth of weakfish,
Cynoscion regalis. in the middle Atlantic Bight. Fish. Bull.,
U.S. 81:803-813.
1984 Reproduction of weakfish, Cynoscion regalis, in the New
York Bight and evidence for geographically specific life history
characteristics. Fish. Bull., U.S. 82:501-511.
Slatkin, M.
1989 Gene flow and the geographic structure of natural popula-
tions. Science (Wash. DC) 236:787-792.
Sokal, R.R., and F.J. Rohlf
1981 Biometry. W.H. Freeman, NY.
Vaughan, D.S., R.J. Seagraves, and K. West
1991 An assessment of the Atlantic weakfish stock, 1982-1988.
Atl. States Mar. Fish. Comm. Spec. Rep. 21, Wash. DC. 29 p.
-I- tables.
Weinstein, M.P., and R.W. Yerger
1976 Protein taxonomy of the Gulf of Mexico and Atlantic
Ocean seatrouts, genus Cynoscion. Fish. Bull., U.S. 74:
599-607.
Wilk, S.J.
1979 Biological and fisheries data of weakfish, Cynoscion
regalis (Bloch and Schneider). NOAA Tech. Ser. Rep. 21,
NMFS Sandy Hook Lab., Highlands NJ, 49 p.
Wilk, S.J., and M.J. Silverman
1976 Fish and hydrographic collections made by the research
vessels Dolphin a.nd Delaware II during 1968-1972 from New
York to Florida. NOAA Tech. Rep. NMFS SSRF-697, 159 p.
Abstract. - Five submersible
dives were conducted to evaluate the
behavior of deepwater shrimp and
the relationship of their density to
bottom type and trap yield. Differ-
ences in behavior of two species of
Heterocarpus were observed: H. en-
sifer tended to group around large
anemones and other benthic relief
over otherwise flat, sandy bottom
and were very active in the presence
of a baited container; whereas H. lae-
vigatus were solitary and showed
little activity around a baited con-
tainer. Greater densities of H. laevi-
gatus were observed on volcanic than
on coralline substrate, indicating a
possible association wdth this bottom
type. Trap catches were regressed
against observed H. laeingatus den-
sities yielding an estimate of the
catchability coefficient. This coeffi-
cient differed from that obtained
from a previously conducted Leslie
model depletion study. Factors con-
tributing to this difference may in-
clude comparing estimates of catch-
ability based on data from different
areas, bias in the estimate of catch-
ability based on observed density,
and bias in the estimate of catchabil-
ity from the depletion study. A com-
bined fishing and visual census study
is suggested as the best assessment
technique.
An assessment of the exploitable
biomass of Heterocarpus laevigatus
in the main Hawaiian Islands.
Part 2: Observations from
a submersible
Robert B. Moffitt
Frank A. Parrlsh
Honolulu Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service. NOAA
2570 Dole Street, Honolulu, Hawaii 96822-2396
Manuscript accepted 26 May 1992.
Fishery Bulletin, U.S. 90:476-482(1992).
Tropical deepwater pandalid shrimp
have potential for commercial har-
vesting in many areas of the Pacific
(Struhsaker and Aasted 1974, Wilder
1977, Moffitt 1983, King 1984, Taga-
mi and Barrows 1988). These shrimp
are readily trapped but not easily
trawled (Struhsaker and Aasted
1974). The largest and most commer-
cially desirable species in Hawaii is
Heterocarpus laevigatus (Tagami and
Barrows 1988); the smaller H. ensifer
has less commercial appeal but is also
abundant (Struhsaker and Aasted
1974). In the early 1980s, several
boats initiated a trap fishery target-
ing H. laevigatus, and landings rose
to a high of 159 metric tons (t) in 1984
(HDLNR 1986). By 1985, most ves-
sels left the fishery, and the annual
landings dropped to <6t/yr (West.
Pac. Fish. Inf. Network, NMFS
Honolulu Lab., unpubl. data).
Early predictions of maximum sus-
tainable yield for Hawaiian shrimp,
based on little or no direct data, were
as much as 1000-2000t/yr (Struh-
saker and Aasted 1974, HDLNR
1979). Recent research on population
dynamics combined with systematic
trapping surveys has resulted in
more refined estimates of exploitable
biomass and maximum sustainable
yield for H. la£vigatus in various
island locations (Dailey and Ralston
1986, Ralston 1986, Moffitt and Polo-
vina 1987, Ralston and Tagami 1992).
The most recent estimate of exploit-
able biomass for the main Hawaiian
Islands, 271-1050 1, is based on an
estimate of the catchability coeffi-
cient (q) obtained through a Leslie
model depletion study, coupled with
catch-per-unit-effort (CPUE) values
and habitat area estimates obtained
through systematic trapping (Ralston
and Tagami 1992).
The relationship of observed tar-
get-species density to fishing-gear
CPUE has been used to estimate
stock biomass and catchability (Rals-
ton et al. 1986, Kulbicki 1988). Esti-
mates of abimdance obtained through
visual census techniques are general-
ly higher than those based on catches
of fishing gear, and the relative reli-
ability of the various assessment
methods must be analyzed on a case-
by-case basis (Uzmann et al. 1977,
Powles and Barans 1980, Kulbicki
and Wantiez 1990, Matlock et al.
1991).
In the present study, we conducted
submersible dives at several sites in
the Hawaiian Islands to observe
shrimp behavior both away from and
in the vicinity of a baited container
and to record density and substrate
associations of //. laevigatus. Obser-
vations of shrimp behavior and sub-
strate associations have applications
to commercial fishermen in terms of
476
Moffirt and Parrish: Assessment of exploitable biomass of Heterocarpus laevigatas in Hawaiian Is , Part 2
477
trapping technique and site selection. The mean of
observed densities recorded during submersible dives
is regressed against yields of a trap set at the dive sites
to obtain an estimate of q, and this value is compared
with that reported by Ralston and Tagami (1992). An
accurate estimate of catchability is important in order
to better estimate exploitable biomass for management
purposes.
Methods
A total of five submersible dives were conducted in the
main Hawaiian Islands, at two sites off leeward Oahu
in February 1988 and three sites off the Kona coast
of the Island of Hawaii in August 1988 (Table 1). The
Oahu sites were selected for their proximity to port and
because of previously-observed concentrations in the
area of unidentified red shrimp at appropriate depths
(400-900 m) for Heterocarpus. The three Kona sites
were selected as extremes in H. laetngatus yield for the
area during a trapping survey using pyramid traps con-
ducted in March 1988 (see Tagami and Barrows (1988)
and Ralston and Tagami (1992) for trap description and
trapping methods). Catches of H. laevigatus were
lowest for the Kona area (< 1 kg/trap-night) at one of
the sites and highest (> 10 kg/trap-night) at the remain-
ing two sites.
Visual censuses
All dives used the Pisces V, a three-man submersible
that allowed simultaneous observations by two re-
searchers through separate view ports with non-
overlapping fields of view directed diagonally forward
and down. A video camera continuously recorded the
bottom throughout each dive as well. The same two
researchers estimated shrimp abundance on all dives
and independently reviewed the dive video tapes as a
check on observer bias. On each survey, the submers-
ible descended to depths of 480-920 m, then traveled
to an arbitrary starting location at ~600-750m depth.
At this point, a baited container was placed on the
bottom and observations of shrimp behavior in the
presence of bait were recorded. After observing shrimp
behavior for ~15min, the submersible traveled a hap-
hazard, rectangular track at a speed of ~2 knots along
the contours within the zone of maximum shrimp abun-
dance (defined below), returning to the baited container
for retrieval at the end of the dive. At preselected time-
intervals (5 or 10 min), the submersible settled to the
bottom and counts of shrimp were taken by each
observer in an independent quadrant filling the field
of view. The estimated area of each quadrant was
10 m-, which was calibrated by underwater observa-
Table I
Locations of five study sites off the islands of Oahu and Hawaii
and 1988 sampling dates.
Site
no.
Location
Dive
date
Trap
date
Oahu
1
9
2ri9.3'N, 158°10.1'W
2r31.0'N, 158°16.8'W
8 Feb.
9 Feb.
13 Mar.
13 Mar.
Kona
1
2
3
19°14.0'N, 155°54.9'W
19°20.7'N, 155°54.2'W
19°47.5'N, 156°07.8'W
23 Aug.
24 Aug.
25 Aug.
18 Mar.
18 Mar.
15 Mar.
tion of known dimensions with the submersible at-
tached to its launch-and-retrieval vehicle. The minimum
distance between observation sites was ~100m. The
number of observations of shrimp density varied be-
tween dives, for bottom time was dependent on bat-
tery power.
Bottom depth, temperature, and substrate type were
recorded with each shrimp count. The substrate within
each quadrant was categorized by composition and par-
ticle size of the major component. Substrate composi-
tion included coralline, volcanic, and mixed; particle
size included sand, rock both small (~<15cm diameter)
and large (>15cm diameter), and pavement.
A x^ goodness-of-fit test using a Poisson distribution
for the expected frequencies was conducted to deter-
mine whether H. laeingatus were concentrated or even-
ly distributed over the bottom at each dive site. Mean
H. laevigatus density and 95% CI was calculated for
each dive site based on a Poisson distribution,
CL = D,i, + (1.96)
f(D(i,
where D(i) and n(,) are the mean density and number
of observations for each dive site. Expected density
values (De) for each site were calculated using trap
landings for the site and the normalized catchability
coefficient (q) reported by Ralston and Tagami (1992),
using the following formula:
De(i, =
CPUE,i)
Confidence limits for the expected density values for
each site could not be calculated, since variance can-
not be computed for CPUE, based as it is on the catch
of a single trap.
An analysis of variance (ANOVA) was performed to
determine whether higher mean densities of H. laevi-
478
Fishery Bulletin 90(3). 1992
gatus could be attributed to different dive sites or
bottom types and to determine whether there was
observer bias. Independent variables for the ANOVA
were dive site, substrate material, substrate particle
size, and observer.
Comparisons of trap landings
and density estimates
Trap catch rates were obtained in March 1988 from one
pyramid shrimp trap set at each of the five dive sites
and allowed to soak overnight (see Ralston and Tagami
(1992) for details). Trap catches ofH. laevigatus were
regressed against mean densities obtained from visual
counts for the five study sites, fitting a linear model
with a zero intercept. The slope of this regression is
an estimate of q, which was compared with that
reported by Ralston and Tagami (1992).
Results
A total of 923 shrimp were captured in the 5 pyramid
traps set at the study sites. Of these, 705 (76%) were
H. laevigatus (Table 2), 217 wereH. ensifer, and 1 was
Acanthephyra eximia. During the scheduled observa-
tion periods on the submersible dives, a total of 494
shrimp were observed (194 total quadrant observa-
tions). Of these, only 95 (19%) were if. laevigatus, and
the remainder consisted primarily of Plesionika sp.,
tentatively identified as P. ensis, and a few individuals
of P. alcocki, H. ensifer, A. eximia, and Gnathophausia
longispina. All H. laevigatus observed from the sub-
mersible appeared well within the size range of those
captured in the traps, indicating that both stock assess-
ment methods sample the same population.
Visual censuses
During our dives, the shrimp showed little reaction to
the presence of the submersible or its lights. When the
submersible came within a few inches of the shrimp,
they swam a short distance avoiding collision. When
the photoflash was used, the shrimp within a few feet
of the submersible started, darting a distance of 1-4
cm. No other reactions to the submersible or its lights
were observed. Several behavioral differences were
noted between the various species observed.
A total of 94% (89 of 95) of the H. laevigatus ob-
served during census periods were seen between 550
and 675 m, the depth range herein defined as the zone
of maximum abundance. Individuals of H. laevigatus
were observed at each dive site, but not necessarily dur-
ing the scheduled census periods conducted within the
zone of maximum abundance. Only counts taken within
Table 2
Trap catches, predicted densities, and mear
1 observed densities
with 95% CL of Heterocarpus laevigatus
at five study sites
off the islands of Oahu and Hawaii in 1988
N = the number
of observations.
Dive
site
Trap
catch
(ra/trap)
Density (re/ha)
Predicted
Observed
CL N
Oahu
1
5
0.53
0
34
9
26
2.7
0
8
Kona
1
0
0
200
(-680)-1080 10
2
376
40
1360
(-570-3290 14
3
298
31
890
220-1560 76
this depth range were used in the analysis.
Heterocarpus laevigatus were observed as solitary
individuals on the bottom, usually stationary but oc-
casionally walking, and rarely swimming near the
bottom. They showed little activity in the presence of
a baited container and were not observed crawling
over or entering it. Conversely, H. ensifer were found
in groups near relief features (e.g., large sea anemones)
at shallower (450-550 m) depths, either stationary on
the bottom or swimming about 1 m above the bottom.
They were very active in the presence of a baited con-
tainer, aggregating quickly and crawling over and
entering the container through the mesh and other
holes. Plesionika alcocki usually were seen on the
bottom, whereas P. ensis generally were seen hang-
ing motionless in the water column ~l-2m off the
bottom. Each showed some activity around the baited
container. Acanthephyra eximia and G. longispina
were observed swimming 1-2 m off the bottom, but
were not seen at the baited container.
Bottom temperature varied during the dives from a
low of 3.9°C at 920 m to a high of 6.0°C at 480 m. The
temperature range within the zone of maximum H.
laevigatus abundance was 4.8-5.9°C.
Bottom type varied considerably among the sites.
The bottom at the two Oahu sites was classified as
coralline sand making up an even, featureless plain
with a gradual (<20°) slope. The bottom at the three
Kona coast sites was much steeper, generally about a
35-45° slope, but with some sections near vertical or
even slightly undercut at the Kona site 3. At Kona site
1, the bottom was nearly uniformly composed of small
(5-10 cm diameter), sharp-edged volcanic rocks and
very little coralline material. Kona site 2 differed from
site 1 in that the small volcanic rocks were more
weathered and the substrate had a greater coralline
component. Kona site 3 had many sandy areas, at
Moffitt and Parrish: Assessment of exploitable biomass of Heterocarpus laevigatas in Hawaiian Is , Part 2
479
Table 3
Analysis of variance for Heterocarpus laeingatus density by
dive, bottom type, and observer.
Source
df
SS
MS
F
P
Dive
4
31.25
7.81
6.87
0.0001
Substrate
material
particle size
2
3
24.53
5.96
12.26
2.98
10.79
2.62
0.0001
0.0765
Observer
1
3.40
3.40
2.99
0.0859
Error
132
150.08
1.14
—
-
Corrected
total
141
215.22
-
-
-
times covering the entire 10 m^ quadrant, as well as
areas of exposed limestone often forming undercut
cliffs, and areas of small weathered volcanic rocks.
The distribution ofH. laevigatus over the bottom was
evaluated in several ways. No significant values were
found for the x^ goodness-of-fit tests (P>0.10) of the
density observations recorded at each dive site, sug-
gesting that the shrimp were randomly distributed (in
a Poisson manner) instead of clumped at each dive site.
The mean and confidence intervals of H. laevigatus
density observed at each dive site are presented in
Table 2. Pooling data from all dive sites for the indepen-
dent variables (dive site, substrate material, substrate
particle size, and observer), only dive site and substrate
material were significantly correlated with shrimp den-
sity (ANOVA, P<0.05; Table 3). The distribution of
residuals did not differ significantly from a normal
distribution. Independent review of the dive video by
each observer yielded complete agreement on H. laevi-
gatus counts and substrate classification, indicating a
lack of observer bias in shrimp density estimation and
substrate associations.
Comparisons of trap landings
and density estimates
Mean observed densities were regressed on trap
catches for each dive site (Fig. 1) (r^ 0.97, P 0.0003).
The least-squares regression equation is
CPUE, = 0.2896 (Dj), (SE 0.02364).
An estimate of catchability is obtained directly from
the value of the slope (0.2896ha/trap-night) with
a confidence interval calculated as 0.2144-0.3648/
trap-night. This estimate of catchability is <y3oth
that reported by Ralston and Tagami (1992) and is
reflected in the differences between observed and
Q-
tuu -
/m
300 ■
• /
200 ■
/
100 -
- /
./
i
i^ •
1
— 1
400 800 1200
Density (No. /ha)
1600
Figure 1
Regression of trap catch per unit effort (CPUE;
number per trap-night) and observed density of
Heterocarpus laerigatus (number/ hectare) at
five sites in the Hawaiian Islands.
expected H. laevigatus densities for Kona sites 2 and
3 (Table 2).
Discussion
Differences in behavior between H. laevigatus and
H. ensifer, the two species with greatest commercial
potential, may lead to some practical applications for
fishermen. The high activity level noted for /f. ensifer
in the presence of a baited container has also been
reported by Gooding et al. (1988) and Saunders and
Hastie (1989). The rapid attraction and entry of this
species into traps, even during daylight hours, indicate
that a short soak time may be adequate for commer-
cial harvesting. The lower activity level of H. laeviga-
tus observed in our study and reported by Saunders
and Hastie (1989) may indicate that a longer soak time
is more appropriate for this species. If so, a small vessel
with a limited number of traps could maximize total
catch by making two short sets during daylight hours
on H. ensifer grounds, followed by an overnight set on
H. laevigatus grounds, assuming that suitable concen-
trations of both species are present within a reasonable
proximity.
480
Fishery Bulletin 90(3), 1992
Previous observations of H. ensifer from a submer-
sible found higher densities on flat, silty, sandy areas
than over low-relief, rocky outcroppings (Gooding et
al. 1988). Although statistical analysis of substrate
associations were not conducted for H. ensifer in this
study, we did observe a similar substrate association.
Heterocarpus ensifer were abundant at the Oahu dive
sites at depths (500-800 m), although this is deeper than
their reported optimum range of 300-600 m (Gooding
1984). The substrate on these two dives was flat, cor-
alline sand with few isolated, low-profile features (e.g.,
sea anemones, small rocks) around which the shrimp
appeared to concentrate. No rocky outcroppings were
observed on these dives. Very few H. ensifer were
observed at the three Kona dive sites, where the bot-
tom was steep and composed largely of rocky rubble
with few sandy patches, though the dive depths again
were deeper than the optimum range for this species.
The substrate associations of H. laevigatiis appeared
to differ from those of H. ensifer. Although the differ-
ences in substrate particle size were not significant, the
ANOVA test revealed significantly higher densities on
volcanic compared with coralline substrates (Table 3)
with data from all dive sites pooled. The significant
results for substrate type, however, must be viewed
with caution because of the significance of dive site to
H. laevigatus density and the unbalanced sample de-
sign. Not only were all substrate types not present on
a single dive, but those types present were not found
in equal proportions on any dive. Therefore, differences
in density attributed to substrate type may actually be
a reflection of differences related by some other factor
to dive site. In particular, the Kona dive sites were
largely volcanic, and the majority of the H. laevigatus
observed were from Kona sites 2 and 3. Although Kona
site 1 also was largely volcanic, the volcanic rocks
differed from those observed at sites 2 and 3, in that
the appearance was of a more recent rock slide (sharper
edges vs. weathered). This apparent instability may
be responsible for the low shrimp density observed at
site 1. Other aspects of the bottom, such as slope,
substrate complexity, stability, and current patterns
may be of considerable importance and should be in-
vestigated in future work on the substrate associations
of H. laevigatus.
With visual censusing techniques, there is always a
concern regarding the reliability of abundance esti-
mates. Various factors, including sampling techniques,
species behavior, and physical conditions, can bias
results (Colton and Alevizon 1981, Sale and Douglas
1981, Brock 1982, Ralston et al. 1986, Matlock et al.
1991). Some authors believe that density estimates
based on direct visual surveys, though often much
higher, are more reliable than those estimated from
fishing gear catches (Uzmann et al. 1977, Powles and
Barans 1980, Kulbicki and Wantiez 1990). Individuals
of the target species, H. laevigatus, were easily counted
because they were in the open and reacted almost with
indifference to the presence of the submersible, and
because the low, uncomplicated relief at the study sites
offered little opportunity for their concealment. Avoid-
ance of the submersible by the shrimp seems unlikely.
Observed densities were much greater than expected,
yet these would be underestimates if avoidance oc-
curred. We cannot discount the possibility of bias in
our density estimates caused by attraction of shrimp
to the baited container placed at the beginning of our
dives. However, we observed no increased density gra-
dient in the vicinity of the container, and density obser-
vations were taken well away from the container site
( > 100 m), presumably outside the drawing range of the
bait, leading us to believe that bias due to this source
was small.
Recalculation of exploitable biomass for the main
Hawaiian Islands using Ralston and Tagami (1992) data
and methods, but substituting the q value obtained in
this study, would lead to a 33-fold increase in the esti-
mate of exploitable biomass (~9000t instead of 271 1).
Just as Ralston and Tagami (1992) suggest that their
estimate may be too low, we suggest that 9000 1 may
be unreasonably high, considering the preliminary
nature of this estimate and the failure of the Hawaiian
fishery that was at least partly due to drops in catch
rates at annual yields of <200t (Tagami and Barrows
1988). The acceptance of either of these estimates
would drastically affect management decisions, and
careful evaluation of these two values must be made.
Contributions to the difference between the two esti-
mates may be from three sources: actual differences
in catchability for the two studies related to differences
in time and study locations, error in our estimate of
q, and error in the Ralston and Tagami (1992) estimate.
The estimation of q can be influenced by a variety
of factors including currents, water turbidity and tem-
perature, type of bait, soak time of fishing gear, and
density of the target species (Morgan 1974, Richards
and Schnute 1986, Miller 1990). For the two studies
involved in this discussion, many of the potential
sources of error were standardized. Both studies used
the same traps, same bait, and same soak times. They
did not, however, conduct studies at the same location
or time, and the range of catch rates encountered dif-
fered for the two studies. In both studies it is assumed
that catchability is constant for all catch rates and den-
sities involved, but this may not be true, particularly
between studies. Unfortunately, we are unable to
evaluate the extent of the error involved from these
sources.
The estimate of q presented in this study could also
be biased. Sources of potential bias include lack of
Moffitt and Parrish- Assessment of exploitable biomass of Heterocarpus laevigatas in Hawaiian Is . Part 2
481
representative catch rates for study sites, error in den-
sity estimation, incompatibility of CPUE and density
estimates collected at different points in time, and the
few data pairs involved. Not only were shrimp catches
for each dive site based on a single trap-night of ef-
fort, but also traps generally were set at depths greater
than the observed range of maximum abundance deter-
mined from our submersible observations (~750m vs.
550-675 m). This results in no estimate of error for
CPUE estimates at each site and no way to determine
whether the traps were set within the range of max-
imum shrimp abundance at the time of trapping. If the
traps were not set within this zone, yields at our sites
may underrepresent relative shrimp abundance,
leading to a lower-than-actual estimate of q. Another
potential source of error in our q estimate is the ac-
curacy of our density estimate. Confidence limits on
our density estimates are quite broad, allowing for a
fair degree of error. At many dive sites this problem
is related to the few observations of density taken
within the zone of maximum abundance. Additional
problems in density estimation associated with the
presence of bait in the water during the dives have
already been addressed. The question of compatibility
between data pairs of density estimates and CPUE
values obtained at different points in time stems from
possible changes in density or in q over time. Although
trap catch rates in the Mariana Archipelago did not
vary significantly on a seasonal basis (Polovina et al.
1985), suggesting that q does not vary seasonally,
H. laevigatus may undergo temporal changes in depth
range on either a diurnal or seasonal basis (King 1984,
Dailey and Ralston 1986). If such movements do occur
(the evidence is not strong) and depth range expands
or contracts during these changes, densities observed
during midday periods in February and August may
differ from those occurring during trapping in March.
Finally, although the fit is quite good, our estimate of
q is based on only five data pairs covering limited values
of CPUE and density.
The potential error in the Ralston and Tagami (1992)
q estimate depends on the appropriateness of using
their habitat area estimate in normalizing q and the
validity of the assumption of constant catchability for
all members of the population. These error sources are
not necessarily greater than those discussed above, but
are much easier to quantify. Even when an accurate
estimate of biomass is obtained for a study site, calcula-
tion of a normalized q is dependent on the estimated
habitat area of the study site. Estimated habitat area
is apt to be larger when depth range is estimated from
trap catches, as opposed to visual surveys, because of
the ability of the traps to draw shrimp outside of their
normal depth range. Recalculating the habitat area of
the Ralston and Tagami study site using the observed
depth range (550-675 m) instead of the reported range
(420-640 m) results in a reduction in area to 63% of the
original value (748 ha instead of 1187 ha). Normalizing
q with this reduced study-site area estimate gives an
adjusted q value of 5.999 ha/trap-night (CI 2.6709-
9.3271 ha/trap-night. The ratio of this adjusted value
to the q obtained in the present study is 20.7 instead
of the original 32.7. Ralston and Tagami (1992)
discuss the effect on their q value of a large portion
of the population not being susceptible to trap capture
for the duration of the study period. They supply
evidence that their original estimate of catchability may
have been four times too high, resulting in a four-fold
underestimation of exploitable biomass. Other authors
have reported similar overestimates of catchability
resulting from depletion studies (Morgan 1974, Mor-
rissy 1975, Miller 1990). A further four-fold reduction
of the Ralston and Tagami q value results in a ratio
of 5.9 relative to our q value and only 1.8 for the ex-
tremes of the 95% CL (the minimum value for the
Ralston and Tagami confidence interval compared with
our maximum value). Coupling these quantifiable fac-
tors with the non-quantifiable factors discussed above
could bring the two estimates of catchability into
agreement.
Because of the importance of accurate estimates of
exploitable biomass to the management process, it
would be desirable to conduct a combined technique
survey of the H. laevigatus resource. This should in-
clude direct visual density estimation with a trapping
study conducted at the same time and place to obtain
a reliable estimate of catchability and thereby exploit-
able biomass. Until that time, the expanded exploitable
biomass estimate (1050 1) for the main Hawaiian Islands
as presented by Ralston and Tagami (1992) should be
accepted for management purposes as a reasonable,
conservative approximation.
Acknowledgments
We would like to thank the staff of the Hawaii Under-
sea Research Laboratory for their support during the
field portion of this study. We have also appreciated
the comments and suggestions made by various
reviewers, including M.G. King, S. Ralston, W.B.
Saunders, M.P. Seki, and D.T. Tagami, which have
helped to form this paper.
Citations
Brock, R.E.
1982 A critique of the visual census method for assessing cor-
al reef fish populations. Bull. Mar. Sci. 32:269-276.
482
Fishery Bulletin 90(3). 1992
Colton, D.E., and W.S. Alevizon
1981 Diurnal variability in a fish assemblage of a Bahamian
coral reef. Environ. Biol. Fishes 6:341-345.
Dailey. M.D., and S. Ralston
1986 Aspects of the reproductive biology, spatial distribution,
growth, and mortality of the deepwater caridean shrimp,
Heterocarpus laevigatus in Hawaii. Fish. Bull., U.S. 84:
915-925.
Gooding, R.M.
1984 Trapping surveys for the deepwater caridean shrimps,
Heterocarpus laevigattis and H. ensifer, in the Northwestern
Hawaiian Islands. Mar. Fish. Rev. 46(2): 18-26.
Gooding, R.M., J.J. Polovina, and M.D. Dailey
1988 Observations of deepwater shrimp, Heterocarpus ensifer,
from a submersible off the island of Hawaii. Mar. Fish. Rev.
50(l):32-39.
HDLNR (Hawaii Department of Land and Natural Resources)
1979 Hawaii fisheries development plan. Dep Land Nat.
Resour., State of Hawaii. Honolulu, 297 p.
1986 Hawaii fisheries plan, 1985. Div. Aquat. Resour., Dep.
Land Nat. Resour., State of Hawaii, Honolulu, 163 p.
King, M.G.
1984 The species and depth distribution of deepwater caridean
shrimps (Decapoda, Caridea) near some Southwest Pacific
islands. Crustaceana 47:174-191.
Kulbicki, M.
1988 Correlation between catch data from bottom longlines
and fish censures in the SW lagoon of New Caledonia. In
Choat, J.H., et al. (eds.), Proc. 6th Int. Coral Reef Symp.,
Townsville, Aust., 2:305-312.
Kulbicki, M., and L. Wantiez
1990 Comparison between fish bycatch from shrimp trawlnet
and visual censuses in St. Vincent Bay, New Caledonia. Fish.
Bull., U.S. 88:667-675.
Matlock, G.C.. W.R. Nelson, R.S. Jones. A.W. Green, T.J. Cody,
E. Gutherz, and J. Doerzbacher
1991 Comparison of two techniques for estimating tilefish,
yellowedge grouper, and other deepwater fish populations.
Fish. Bull., U.S. 89:91-99.
Miller. R.J.
1990 Effectiveness of crab and lobster traps. Can. J. Fish.
Aquat. Sci. 47:1228-1251.
Moffitt, R.B.
1983 Heterocarpus longirostris MacGilchrist from the North-
ern Mariana Islands. Fish. Bull., U.S. 81:434-436.
Moffitt, R.B., and J.J. Polovina
1987 Distribution and yield of the deepwater shrimp Heterocar-
pus resource in the Marianas. Fish. Bull., U.S. 85:339-349.
Morgan, G.R.
1974 Aspects of the population dynamics of the western rock
lobster, Panulirus cygnus George. II Seasonal changes in the
catchability coefficient. Aust. J. Mar. Freshwater Res.
25:249-259.
Morrissy, N.M.
1975 The influence of sampling intensity on the 'catchability'
of marron, Cherax teyiuimanus (Smith) (Decapoda: Parasta-
cidae). Aust. J. Mar. Freshwater Res. 26:47-73.
Polovina, J.J., R.B. Moffitt. S. Ralston, P.M. Shiota. and
H.A. Williams
1985 Fisheries resource assessment of the Mariana Archi-
pelago, 1982-85. Mar. Fish. Rev. 47(4):19-25.
Powles, H., and C.A. Barans
1980 Groundfish monitoring in sponge-coral areas off the
southeastern United States. Mar. Fish. Rev. 42(5):21-35.
Ralston, S.
1986 An intensive fishing experiment for the caridean shrimp,
Heterocarpus laevigatus. at Alamagan Island in the Mariana
Archipelago. Fish. Bull.. U.S. 84:927-934.
Ralston, S., and D.T. Tagami
1992 An assessment of the exploitable biomass of Heterocar-
pus laeirigatus in the main Hawaiian Islands. Part 1: Trapping
surveys, depletion experiment, and length structure. Fish.
Bull., U.S. 90:494-504.
Ralston, S., R.M. Gooding, and G.M. Ludwig
1986 An ecological survey and comparison of bottom fish
resource assessments (submersible versus handline fishing) at
.Johnston Atoll. Fish. Bull., U.S. 84:141-155.
Richards, L.J., and J.T. Schnute
1986 An experimental and statistical approach to the ques-
tion: Is CPUE and index of abundance? Can. J. Fish. Aquat.
Sci. 43:1214-1227.
Sale, P.F., and W.A. Douglas
1981 Precision and accuracy of visual census technique for fish
assemblages on coral patch reefs. Environ. Biol. Fish. 6:
333-339.
Saunders, W.B., and L.C. Hastie
1989 Deep-water shrimp survey and feasibility study. Republic
of Palau, Western Caroline Islands. Pac. Fish. Dev. Found.,
Final Rep., Proj. 63A, Honolulu.
Struhsaker, P., and D.C. Aasted
1974 Deepwater shrimp trapping in the Hawaiian Islands.
Mar. Fish. Rev. 36(10):24-30.
Tagami, D.T., and S. Barrows
1988 Deep-sea shrimp trapping for Heterocarpus laevigatus
in the Hawaiian Archipelago by a commercial fishing vessel.
NOAA Tech. Memo. NMFS-SWFC-103. NMFS Honolulu Lab.,
14 p.
Uzmann, J.R., R.A. Cooper, R.B. Theroux, and R.L. Wigley
1977 Synoptic comparison of three sampling techniques for
estimating abundance and distribution of selected mega-
fauna: Submersible vs camera sled vs otter trawl. Mar. Fish.
Rev. 39(12):11-19.
Wilder. M.J.
1977 Biological aspects and fisheries potential of two deep-
water shrimps Heterocarpus ensifer and Heterocarpus laevi-
gatus in waters surrounding Guam. M.S. thesis, Univ. Guam,
Agana, 79 p.
Abstract.- Research and com-
mercial trapping data show variation
in recruitment to the fishery for
spiny lobster Panulirus marginahis
at Maro Reef, relative to Necker
Island which is 670 km to the south-
east. Recruitment to the fishery at
Maro Reef is shown to be highly cor-
related with the difference in sea
level 4 years earlier between French
Frigate Shoals and Midway Islands.
Geosat altimeter data indicate that
the relative sea level between French
Frigate Shoals and Midway is an in-
dicator of the strength of the Sub-
tropical Counter Current. Mechan-
isms linking the Subtropical Counter
Current with larval advection and
survival are discussed. The sea level
index provides a forecast of recruit-
ment 4 years later to the fishery at
Maro Reef.
Variability In spiny lobster
Panulirus marginatus recruitment
and sea level In the Northwestern
Ha\A/allan Islands*
Jeffrey J. Polovina
Honolulu Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service. NOAA
2570 Dole Street, Honolulu, Hawaii 96822-2396
Joint Institute of Marine and Atmospheric Research (JIMAR)
University of Hawaii, Honolulu. Hawaii 96822
Department of Oceanography. School of Ocean and Earth Science and Technology
University of Hawaii, Honolulu, Hawaii 96822
Gary T. Mitchum
Joint Institute of Marine and Atmospheric Research (JIMAR)
University of Hawaii, Honolulu, Hawaii 96822
Department of Oceanography, School of Ocean and Earth Science and Technology
University of Hawaii, Honolulu, Hawaii 96822
Significant correlations between com-
mercial landings or recruitment esti-
mates and one or more environmen-
tal indices are commonly reported in
the fisheries literature, but few have
served as accurate predictors of fu-
ture population levels (Drinkwater
and Myers 1987). However, such cor-
relations can lead to the formulation
or support of hypotheses regarding
the factors responsible for population
changes. For example, an inverse
correlation between the survival of
Pacific mackerel Scomber japonicus
to age 1 and the strength of the Cali-
fornia Current, and the lack of corre-
lation between survival and plankton
biomass, have been offered as evi-
dence that advection, rather than
starvation, controlled survival of the
planktonic stages of this species
(Sinclair et al. 1985).
Correlative studies on lobsters sug-
gest that population size results from
changes in survival and advection at
the larval stage, but in at least one
Manuscript accepted 20 May 1992.
Fishery Bulletin, U.S. 90:483-493 (1992).
' Contribution 2820 of the School of Ocean and
Earth Science and Technology. University of
Hawaii; JIMAR Contribution 91-0243.
instance, density-dependent mechan-
isms after postsettlement may damp-
en this variation (Pollock 1986). Fluc-
tuations in sea-surface temperature
appear to result in changes in larval
survival and catches 6 years later for
the clawed lobster Homarus ameri-
canus in Maine (Fogarty 1988). Vari-
ation in the strength of the Leeuwin
Current, which may be linked to El
Nino Southern Oscillation (ENSO)
events, is suggested as a cause of
variation in the number of larvae re-
turned to the coast and subsequent
recruitment to the fishery for the
western rock lobster Panulirus
argus (Pearce and Phillips 1988).
Changes in recruitment levels of the
California spiny lobster P. interrup-
tus to the northern portion of its
habitat may be episodic, influenced
by large-scale, interannual El Nino
events (Pringle 1986). Variation in
postlarval recruitment in the South
African rock lobster Jasus lalandii
is thought to arise from changes in
the paths and velocities of extensive
offshore currents, which eventually
return larvae to the coast. However,
density-dependent phenomena influ-
483
484
Fishery Bulletin 90(3). 1992
encing juvenile and adult stages may substantially
dampen this variation and produce fairly stable recruit-
ment to the fishery (Pollock 1986).
In the Northwestern Hawaiian Islands (NWHI), a
substantial drop in catches and catch-per-unit-effort
(CPUE) of spiny lobster P. marginatus Quoy and
Gaimard 1825 was recently documented (Polovina
1991). This study examines whether these declines in
catches and CPUE are due to overfishing or to ocean-
ographic factors which impact spiny lobster population
dynamics.
NWHI lobster fishery
The NWHI region is an isolated range of islands, islets,
banks, and reefs that extend 2775km northwest from
Nihoa Island to Kure Atoll (Fig. 1). In 1977 after re-
search cruises documented a substantial lobster pop-
ulation in the NWHI, a commercial trap fishery was
initiated. The fishery targeted two species: the endemic
spiny lobster P. marginatus and the slipper lobster
Scyllarides squammosus Mike-Edwards 1837. A fish-
ery management plan implemented in 1983 mandated
that vessels submit logbooks recording daily catch and
number of traps set (effort); the plan also established
a minimum harvest size for spiny lobster and prohibited
the harvest of egg-bearing females. Subsequent amend-
ments to this plan added a minimum legal size for slip-
per lobster and required that traps have escape vents.
In 1990, low catches and CPUE prompted a 6-month
closure of the fishery (May-November 1991).
-« — N.|£quatorlol < ■ Si
Equatorial Counter » —n
Kure I.
^) , Mi<i»0, I.
French Frigole
Shoals
Figure 1
Pacific Ocean and major cur-
rents with an inset of the
Hawaiian Archipelago, includ-
ing the Northwestern Hawai-
ian Is.
Polovina and Mitchum: Recruitment of Panulirus marginatus relative to sea level in NW Hawaiian Is
485
Since 1983, the lobster fleet
has been composed of 9-14 ves-
sels (20-30 m long), each averag-
ing 3 trips per year. The vessels
set about 800 traps per day and
remain at sea almost 2 months
per trip. Landings in recent
years have averaged almost 1
million lobsters, valued at about
US$6 million ex-vessel. Because
of heavy fishing since 1986, the
population has been fished down
to the point that 3-year-old re-
cruits comprised most of the
fishery catches (Polovina 1991).
Since 1988, about 80% of land-
ings have been spiny lobster
(Table 1). Two banks-Necker I.
at the southeast end of the
NWHI, and Maro Reef which is
670km northwest of Necker I.— account for over 60%
of the fishery's catches. There is no recreational lobster
fishery in the NWHI.
Spiny lobster spawn over a broad spring, summer,
and fall period. After hatching, the eggs are planktonic;
the planktonic period for the larvae is estimated at 12
months based on spawning season and larval tow data
(NMFS Honolulu Lab., unpubl. data). Further, the
larval tow data suggest that mid- to late-stage spiny
lobster larvae are close to the surface at night and
move down to ~100m during the day (Polovina, pers.
observ.). Based on growth curves estimated from both
tagging (MacDonald 1984) and length-based methods
(Polovina and Moffitt 1989), spiny lobster reach the
minimum legal size (which is slightly larger than the
size at onset of sexual maturity) approximately 3 years
after they settle onto benthic habitat. After settlement,
the lobster probably do not move between banks since
interbank depths exceed 1000 m.
Regional oceanography
The Hawaiian Archipelago lies within the subtropical
gyre formed by the Kuroshio Current to the west and
the north, the California Current to the east, and the
North Equatorial Current to the south (Fig. 1). The
speed of the gyre in the vicinity of the archipelago is
slow (<5cm/s; Roden 1991). An eastward-flowing cur-
rent within the subtropical gyre, named the Subtropical
Counter Current (SCC), was predicted by Yoshida and
Kidokoro (1967) and subsequently confirmed by Robin-
son (1969) and Uda and Hasunuma (1969) (Fig. 1). More
recent work has shown that, in at least the western por-
tion, the interior of the subtropical gyre is composed
Table 1
Annual landings of spiny (Panulirus marginatus) and slipper
lobsters, trapping effort, and percentage of spiny lobster in
{Scyllarides squammostis)
the landings, 1983-90.'
Year
Lobster
landings (10')
Trap hauls
(10^)
CPUE
% spiny
lobster
Spiny
Slipper
Total
1983'
158
18
176
64
2.75
90
1984
677
207
884
371
2.38
78
1985
1022
900
1922
1041
1.83
53
1986
843
851
1694
1293
1.31
50
1987
393
352
745
806
0.92
53
1988
888
174
lOr.2
840
1.26
84
1989
944
222
1166
1069
1.09
81
1990
591
187 778
he NMFS Honolulu Lab
he W. Pac. Reg. Fish.
1182 0.66 76
, as required by the Crustacean Fishery
Vlanage. Counc, Honolulu.
' Data were provided to t
Management Plan of t
-April-December 1983.
of a quasi-stationary banded structure of easterly- and
westerly-flowing currents (White and Hasunuma 1982).
The SCC consists of two bands of eastward flow at 23°
and 28°N, with mean annual speeds of 8 and 6cm/s,
respectively (White and Hasunuma 1982).
In addition to these large-scale features, the meso-
scale oceanography around the Hawaiian Archipelago
is a complex system of fronts and eddies resulting from
both interactions between alternating east and west
currents and interactions between current and the
topography of the archipelago.
Data and analysis
Research data
Standardized trapping surveys, using the same traps
set at the same sites, were conducted at Necker I. and
Maro Reef during June and July of 1986-88 and 1990.
The size-frequency data were converted to age-fre-
quency data with a von Bertalanffy growth curve
(MacDonald 1984). The age-frequency distribution was
standardized for the number of traps deployed to
estimate the relative age-frequency distribution of the
population.
Fishery data
Although detailed catch and effort data were not avaO-
able until after the logbook regulations were estab-
lished in 1988, catch and effort were generally light
and were concentrated around Necker I. from the in-
ception of the fishery until 1984 (Fig. 2). The combined
CPUE for slipper and spiny lobsters in 1983-90
generally declined from 2.8 to ~0.7 lobster per trap-
486
Fishery Bulletin 90(3), 1992
CT>
C
■o
c
Year
Figure 2
Total slipper (Scyllarides squammosics) and spiny (Panulirus marginatus) lobster land
ings and CPUE from the Northwestern Hawaiian Is., 1977-90.
haul (Fig. 2), based on catch and effort data reported
in the logbooks. Catch data in the logbooks are checked
against landings by enforcement agents, so misreport-
ing is not a problem. Common assessment approaches,
such as length-based cohort analysis, are not applicable
to this fishery, given the relatively short time-series
of catch and effort data, the difficulty in routinely
ageing lobsters, and the lack of information on the size-
frequency from the landings and the nature of a stock-
recruitment relationship. While a dynamic surplus pro-
duction model has been applied to the data, an implicit
assumption about the form of the stock recruitment
relationship is required (Polovina 1991).
A more general approach is to begin with a model
which expresses Nj as the number of exploitable lob-
sters at time t as a function of Nt_ i , Z as the total in-
stantaneous mortality from time t - 1 to t, and r as the
number which recruit and survive from t - 1 to t as
Nt = r -I- Nt^i e-^
Using the relationship that the product of catchability
(q) and N(t) is CPUE(t), this model becomes
CPUEt = q*r -H CPUEt.i e-M-qf,
where M and f are annual instantaneous natural mor-
tality and fishing effort, respectively, during the period
t- 1 to t. This CPUE model, a simple version of a size-
structured model developed by Schnute et al. (1989),
was used to estimate population parameters and to
evaluate the extent that fishing
effort explains the observed vari-
ation in CPUE. This model as-
sumes constant catchability and
recruitment; hence, the differ-
ences between predicted and
observed CPUE are interpreted
as variation in recnoitment, catch-
ability, or both.
The commercial data do not
indicate whether effort was di-
rected at slipper or spiny lobster.
However, the catches can be
grouped into two periods based
on the proportion of spiny to slip-
per lobsters. In period 1 (1983-
84 and 1988-90), ~80% of the
landings were spiny lobster; in
period 2 (1985-87), ^56% of the
landings were spiny lobster
(Table 1). The change in propor-
tion of spiny lobster catches is
likely due to changes in targeting
and abundance. The CPUE
model is modified so that a catchability coefficient can
be estimated for each period. Our modified CPUE
model regresses the CPUE of spiny lobster above the
minimum size in month t (CPUEt) on the CPUE of the
same month in the previous year:
(M + Q,f.)
CPUEt = K*Qt e
+ (CPUEt_i2)(e-M-Q,f,)
with
' Qt
Qt-12
Qt = qiii.t + q2i
2,t
where qj is the catchability of spiny lobster during
period 1, q2 is the catchability during period 2, M is
the annual instantaneous natural mortality, R is the
annual recruitment, f is the cumulative fishing effort
during the period (t- 12, t- 1), and Ij , (i= 1,2) is the
indicator or set function which takes the value 1 if t
is within period i or otherwise takes the value 0. Esti-
mates of R, q] , qo , and M were obtained by minimiz-
ing the sum of squares of the difference between the
square root of the observed and predicted CPUE with
a simplex algorithm.
Sea level data
To examine the relationship between lobster recruit-
ment variation at Maro Reef and physical factors such
Polovina and Mitchum: Recruitment of Panulirus rmrginstus relative to sea level in NW Hawaiian Is 487
as variation in tlie SCC, we fo-
cused on the analysis of sea level
data from the NWHI. Our choice
of sea level was primarily a prac-
tical one. In comparison to cur-
rent or upper-layer temperature
records, the sea level records are
of long duration, and the data are
measured continuously and are
available in nearly real-time. An
additional advantage is that sea-
surface height data from the
Geosat satellite altimeter are
available to provide a spatial de-
scription that complements the
temporal description available
from the sea level stations.
Data on the difference in sea
level between the gauges at
French Frigate Shoals (FFS) and
at Midway Is. have beer avail-
able since 1976 (Figs. 1,3). This
sea level difference (denoted as
FFS-Midway sea level) serves as
an index of the geostrophic cur-
rent anomalies across the NWHI in the region of Maro
Reef. For example, an increase in the sea level height
at FFS relative to Midway Is., measured from tide
gauges, indicates the strengthening of a current that
is across the gradient between the two locations and
is flowing from the southwest to the northeast.
To interpret these flow anomalies as a manifestation
of the variations in SCC strength, the spatial structure
of the sea-surface height variation was examined by
mapping the variability observed by the Geosat altim-
eter during November 1986-November 1988. These 2
years were selected because more accurate orbit
estimates were available during this time-period and
would result in more accurate sea-surface height
fields. The Geosat geophysical data records were ob-
tained from NOAA (Cheney et al. 1987) and were pro-
cessed with software developed at the University of
Hawaii.
Averages of the Geosat data over November 1986-
November 1987 were subtracted from the averages
over November 1987-November 1988. Before using
the Geosat data, we checked that the resulting sea level
differences from the altimeter were consistent with the
corresponding sea level differences from tide gauges
at FFS and Midway (not shown). Choosing these time-
periods also allowed us to contrast conditions during
the ENSO period of 1986-87, when the FFS-Midway
sea level was low (~520mm), with conditions during
the normal period of 1987-88, when the FFS-Midway
sea level was higher (~600mm).
150
, ENSO events
"g 100
- 1976/77 / \ 1982/83 1986/87
^ 50
" / \ '' / 1
a
Si 0
/ \ / \
o
-1 -50
'2:
1
- / \ f\-
u. -'00
U-
\
-150
N/
1976 1978 1980 1982 1984 1986 1988 1990
Year
Figure 3
Annual French Frigate Shoals (FFS)-Midway sea level differences from tide gauges,
1976-90 (ENSO = El Nino Southern Oscillation).
Puerulus settlement
During the last planktonic stage (i.e., postlarval or
puerulus stage), spiny lobster acquire the benthic mor-
phological features of adults and become active swim-
mers seeking benthic habitat. MacDonald (1986)
studied puerulus settlement in the Hawaiian Archi-
pelago with traps known as Witham Collectors at Kure
Atoll (north of Midway Is.) in 1979-83 and at FFS in
1981-85. He computed mean catch per collector over
12-month periods (June-May) at Kure Atoll and FFS.
These data will be compared with the FFS-Midway sea
level data.
Results
The fit of the model to the commercial CPUE data and
the resulting residuals indicate the model fits the trend
in CPUE, but considerable unexplained variation ex-
ists in CPUE within and between years (Fig. 4). For
example, given the fishing effort, CPUE was greater
than expected in 1988 but declined more than expected
in 1990. Since the model assumes both constant recruit-
ment and constant catchability, the residuals may
reflect variation in these factors. From the fit of the
model, R = 1.2 xlO^ adult lobsters/yr, M = 0.71/yr, qi =
1.2x10-6, and q2 = 0.6xl0-6. Thus 1.2 million lob-
sters recruit to the fishery annually; with an M of
0.71/yr, only 50% of the 3-year-olds survive 1 year (in
488
Fishery Bulletin 90(3), 1992
the absence of fishing). Further, a CPUE of 1.2 spiny
lobster/trap-haul means the exploitable population is
1 million spiny lobster. An independent estimate of M
2.4
t
2.2
2
- l\j\
O 14
: ^
I J? «
■g 0 =
~ 0-6
^ 0.4
IpnA. ^ iJ "^ t s ^^
Q- 02
CO
" ,,/^^^SA^l^^ AVV v
-02
- VMA/ V^/V ^ ^Vr
-0.4
' y
-0.6
1984 1985 1966 1967 1988 1 969 1990
3 glilsli 9I3 9I3 9I3I9I3I9I
6 12 6 12 6 12 6 12 6 12 6 12 6 12
Month
Figure 4
Fit of the CPUE model ( + ) to monthly CPUE ( D) for spiny lobster PanuUrus
marginatiis in the Northwestern Hawaiian Is., and residuals from the fit
(♦).
1986
1987
— .
3
N=969
3
ra
N = 1410
ti.s
S1.O
0
mO.S
i1.5
n
1 ^ °
O
Q.
0
[ ,
a.
o
1 1
Mil
1
2 3 4 5 6
1 2 3 4 S 6
Year class
Year class
1988
1990
ii.s
0
mO.S
N = 1635
3
Sl.O
ifl
XI
o
m 0.5
N^170
=>
a.
0
1 1
O
' — ^ 1 ', — ,
1
2 3 4 5 6
12 3 4 5 6
Year class
Year class
Figure 5
Age-frequ
ency distributions of spiny lobster Panuli
rus marginatus, based on research
sampling
n 1986-88 and 1990 at Maro Reef (N =
no. of spiny lobster in the sample).
from tagging at FFS is 0.5/yr (MacDonald 1984).
Commercial trapping effort since 1985 has averaged
about 1 million trap-hauls (Table 1); using the qj esti-
mate as catchability, annual fishing mor-
tality (F) is estimated as 1.2/yr or 1.7 xM.
With these figures and the estimates of
growth and age at onset of sexual matur-
ity, the Beverton-Holt yield equation esti-
mates the spawning-stock biomass per
recruit, when effort is 1 million trap-hauls,
is 40% of what it would be in the absence
of fishing (Polovina 1991). Thus, the ratio
of F to M and the relative spawning-stock
biomass calculations suggest that the
spawning biomass in 1985-86 was not
fished down to a level that would cause the
poor recruitment to the fishery 4 years
later (1989-90).
Much of the variation in residuals from
the CPUE model is due to variation in re-
cruitment at Maro Reef. For example, for
the entire NWHI in 1990, trapping effort
increased 11% from the previous year
while the catch declined 33%, resulting in
a 39% decline in CPUE. However, the
decline in CPUE was most striking at
Maro Reef, where CPUE declined 42%
even though effort decreased by
37%. At Necker I., CPUE also
declined (40%) but effort in-
creased 35%.
The estimated age-frequency
distributions based on research
cruises at Maro Reef show a
strong 3-year-old class in 1988
and a striking absence of all age-
classes in 1990 (Fig. 5). This is
consistent wdth the hypothesis
that recruitment of the 3-year-
olds to the fishery was weak in
1990 and subsequent fishing
reduced all older age-classes.
Necker I. had many more 2-year-
olds in the samples since some
trapping sites include nursery
habitat; but between years, the
abundance of 2-year-olds was
relatively constant, whereas
older lobsters declined in 1990,
likely because of the increase in
fishing effort (Fig. 6).
The NWHI lobster fleet is very
mobile and shifts its trapping
locations according to abundance
of lobsters. By 1985, both Maro
Polovina and Mitchum: Recruitment of Psnulirus marginatus relative to sea level in NW Hawaiian Is
489
Reef and Necker I. had gone through a period of fishing
down the pre-exploitation population; the relative
change in catches between the
two banks may reflect changes in
their relative recruitment. Since
both banks are not always fish-
ed each month, we pooled the
catches by quarter. A 3-quarter
moving average of the ratio of
quarterly catches at Maro Reef
to the combined quarterly catches
at Necker I. and Maro Reef shows
considerable variation (Fig. 7).
For example, catches from Maro
in 1985 and 1988 represented
almost 80% of the catches from
the two banks, but in 1990 they
represented less than 20%. A
3-quarter moving average of the
residuals from the CPUE model
shows the same trend as the ratio
of catches from Maro Reef rela-
tive to Necker I. and Maro Reef
combined (Fig. 7). This suggests
that the variation in recruitment,
catchability, or both at Maro
Reef is responsible for most of
the variation not explained by
fishing effort observed for the
entire NWHI.
height of the sea level ridge stretching across the
Pacific. The height and location of this sea level ridge
1986
N = 858
1
1 1 11 !
1988
N = 16S7
i 1 1 1 , ,
Year class
1987
N = 1248 '
! 1 1 ! : ,
1990
N=767
[ 1 . 1 1 1
3 4
Year class
Figure 6
Age-frequency distributions of spiny lobster Panulirus marginatus, based on research
sampling in 1986-88 and 1990 at Necker I. (N = no. of spiny lobster in the sample).
Variation between
sea level and the SCC
Differences in sea level over the Pacific, be-
tween a year when the FFS-Midway sea level
was high and a year when it was low, appear
as a ridge of positive values, extending from
southwest to northeast, that parallels a
trough of negative values to the northwest
(Fig. 8). Midway lies in the trough, Honolulu
is on the ridge, and FFS lies on the gradient,
which corresponds to the region of the most
energetic geostrophic flow anomalies. This
ridge and trough indicate that the change in
the FFS-Midway sea level from low to high
reflects the increase in a ridge extending
across the western Pacific. The increase in
the ridge and trough pattern represents an
increase in the current flow along the gradi-
ent of this ridge. The path of this gradient or
flow across the Pacific is consistent with the
general path of the SCC. Thus FFS-Midway
sea level measures a large-scale oceano-
graphic feature which is represented by the
Figure 7
Three-quarter moving average of the ratio of quarterly landings of spiny
lobster Panulirus marginatus at Maro Reef to quarterly landings at Maro
Reef and Necker I. (D), and 3-quarter moving average of the residuals
from the fit of the CPUE model to monthly spiny lobster CPUE (♦).
490
Fishery Bulletin 90(3). 1992
20 N SjS:^-'
160E
140-W
Figure 8
Differences between mean sea level during a non-ENSO (El Nino Southern Oscillation) period (Nov. 1987-Sept.
1988) and an ENSO period (Nov. 1986-Sept. 1987) from Geosat data. Negative values denoted by shaded areas.
01
"D
O
E
Ol
(_)
e
o
o
T3
m
a:
Figure 9
Overlay of 3-month moving average of French fVigate Shoals (FFS)-Midway sea level
advanced by 4 years (D) with a 3-month moving average of residuals from the fit of
theCPUE model (-i-).
correspond to the SCC strength
and position, respectively.
Relationship between sea
level and lobster abundance
Lagged cross-correlations between
FFS-Midway sea level and the
variables (i.e., the ratio of catches
at Maro Reef to the combined
catches at Maro Reef and Necker
I., and the residuals from the
CPUE model) have their strong-
est correlations (r 0.82 and 0.68,
respectively) with sea level
lagged by exactly 4 years. When
sea level is lagged by 4 years and
overlayed with these time-series,
there is good agreement (Figs.
9, 10). Based on research samples
pooled over 1986-88, the mean
estimated age of lobsters caught
by the fishery is 3.8 years (after
settlement).
Polovina and Mitchum: Recruitment of Panulirus marginatus relative to sea level in NW Hawaiian Is
491
Based on the comparison of FFS-Midway sea level
with the available puerulus settlement data from
MacDonald (1986) in the same year, the FFS-Midway
sea level correlates positively with mean puerulus
catches at Kure Atoll (r 0.78, P 0.11) and shows no
significant correlation with mean
puerulus catches at FFS (r
-0.37, P>0.25) (Fig. 11).
ment differs between Necker I. and Maro Reef. At
Maro Reef, large-scale oceanographic features appear
to control the abundance of late-stage larvae, which in
turn results in interannual variation in recruitment to
the fishery.
Discussion
The research and commercial
catch and effort data presented
here show that the recruitment
of 3-year-old spiny lobster to the
fishery has varied considerably
at Maro Reef but has remained
stable at Necker I., 670km to the
southeast. Fishing effort is not
considered sufficiently heavy to
explain a decline in recruitment,
especially a decline at one bank
and not the other. The relation-
ship between recruitment to the
fishery at Maro Reef and the
FFS-Midway sea level advanced
by 4 years suggests that environ-
mental factors impacting the lar-
val stage are responsible for the
recruitment variation. The Geo-
sat data suggest that the FFS-
Midway sea level measures the
sec. Hence the SCC strength or
location dictates recruitment
strength to the fishery 4 years
later. Consistent with this hy-
pothesis is the mean age of lob-
sters in the commercial catches
as well as the correlation be-
tween puerulus settlement at
Kure Atoll and FFS-Midway sea
level. The lack of correlation be-
tween puerulus settlement at
FFS and sea level is consistent
with the observation that recruit-
ment at the lower end of the
NWHI is not linked to the same
pattern of variation as Maro
Reef. Annual variation in both
SCC strength and position has
been observed in the western
Pacific (White and Hasunuma
1982). In summary, the temporal
pattern of spiny lobster recruit-
D
o
?
■g
I
CO
123412341234123412341234
Quarter of year
Figure 10
Overlay of 3-quarter moving average of French Frigate Shoals (FFS)-Midway sea level
advanced by 4 years (D), with a 3-quarter moving average of the ratio of Maro Reef
to Maro Reef plus Necker I., spiny lobster landings ( + ).
-
770
1.4
_o
u
0, 1 2
o
U M
1-
<^
CL '
_0 09
o
U
08
c
o
01 07
\/'U /''■
760
750 E
740 _£
730 _
720 5;
710 _0
700
690 0)
680 ">
670 >■
660 5
650 ;2
640 2!
630 1
620 )^
610 LL
VV>^!
0.6
1 1 1 1 1
600
590
79-80 80-81 81 -62 B2-83 83-84 84-85
Year (June — May)
Figure 1 1
Mean annual puerulus settlement from traps at Kure Atoll ( + ) and French Frigate Shoals
(FFS) (0) and FFS-Midway sea level (D), all computed on a June-May year.
492
Fishery Bulletin 90(3). 1992
The underlying mechanism Unking the correlation
between the SCC and subsequent recruitment at Maro
Reef is not known. It is possible that the SCC returns
larvae, which have been advected west of the archi-
pelago, back to Maro Reef. The SCC has been hypoth-
esized to transport Acropora coral from Johnston Atoll
(lat. 16°45'N., long. 169°31'W) to FFS (Grigg 1982).
In addition, larvae of a spiny lobster species not re-
corded as an adult in Hawaii have been transported
from the Marshall Is. to the Hawaiian Archipelago
(Phillips and McWilliam 1989).
However, it may be that the SCC impacts not advec-
tion but larval survival. Laboratory studies have shown
that spiny lobster larvae suffer a high level of mortal-
ity when water temperatures drop below 20°C (T.
Kazama, NMFS Honolulu Lab., pers. commun., Sept.
1991). In the years we estimated that the SCC was
weak, water temperatures <20°C in the winter have
been observed at Maro Reef but not Necker I. If little
larval mixing occurs between Maro Reef and Necker
I., larval mortality at Maro Reef resulting from low
winter temperatures could account for the observed
recruitment variation.
A third hypothesis is that when the SCC has a par-
ticular speed and location, it produces fronts which re-
tain larvae near Maro Reef. When the SCC is weak or
shifts, these fronts are not formed near Maro Reef.
Preliminary evidence from the drifter buoys and lar-
val sampling in our study suggests fronts north of Maro
Reef and south of Necker I. may be important for
lobster larvae (Polovina, pers. observ.)
o
"5
"5
x>
_o
c
a.
1234123412341234123412341234123412341234
Quarter of year
Figure 12
Three-quarter moving average of the ratio of quarterly landings of spiny lobster PanuLirus
marginatus at Maro Reef to quarterly landings at Maro Reef and Necker I. (bold line)
overlayed with the 3-quarter moving average of French Frigate Shoals (FFS)-Midway
sea level deviation advanced by 4 years (thin line).
One potential management application of the lagged
relationship between FFS-Midway sea level and recruit-
ment is that it provides up to a 4-year forecast of re-
cruitment to the fishery at Maro Reef. A 3-quarter
moving average of the FFS-Midway sea level shifted
forward by 4 years forecasts poor recruitment in 1991,
followed by an improvement beginning in late 1992 (Fig.
12). During January-May 1991 before the fishery was
closed for 6 months, recruitment at Maro Reef clearly
had not recovered, as only 1052 spiny lobster were
harvested from Maro Reef while 34,746 spiny lobster
were harvested from Necker I. Recall that the relative
catches between banks provide an index of relative abun-
dance, since the fleet moves to maximize the CPUE.
The FFS-Midway sea level data forecast that catches
at Maro Reef will improve beginning in late 1992 (Fig.
12). Data from larval tows are consistent with this
forecast. Standardized larval tows, taken in June and
November 1989 over a grid of stations from the 200 m
isobath out to 56km around both Necker I. and Maro
Reef, caught 3802 and 3342 late-stage phyllosomes,
respectively (J. Polovina, unpubl. data). A f-test, based
on a lognormal distribution, finds no significant differ-
ence in the mean abundance of larvae between Maro
Reef and Necker I. If we assume that larval abundance
was high around Necker I. in 1989, then good larval
recruitment apparently has returned to Maro Reef.
This is consistent with the observed higher sea-level
values in 1989 (shown as 1993 values in Fig. 12, since
the sea level has been advanced by 4 years) and sug-
gests that catches will be high at Maro Reef in 1993.
The FFS-Midway sea level time
series from 1976 to 1990 (Fig. 3)
shows that ENSO events may re-
sult in poor recruitment to the
fishery 4 years later, but the series
also shows a long-term decline.
Reasons for the low FFS-Midway
sea level during ENSO events are
not known, but may be related to
a decrease in surface water sup-
plied to the SCC in the western
Pacific. Such a change could be
associated with the circulation
disruptions observed in the trop-
ical Pacific during ENSO events
(Meyers and Donguy 1984). The
long-term decline in sea level
from 1976 to 1990 suggests there
is a low-frequency component in
the variation in SCC strength
and, hence, lobster recruitment.
Thus, it may be some time before
recruitment to the fishery is at
the early 1980s' level.
800
750
700
650
E
E
01
>
0)
a
0)
o
T3
I
in
Polovina and Mitchum Recruitment of Panuhrus margtnatus relative to sea level in NW Hawaiian Is,
493
Acknowledgments
Partial support for this study was provided by the Na-
tional Science Foundation under grant OCE8911163
by the TOGA Sea Level Center through NOAA Coop-
erative Agreement NA90RAH00074 to the Joint In-
stitute for Marine and Atmospheric Research (JIMAR),
University of Hawaii; and by NASA through the Jet
Propulsion Laboratory as part of the TOPEX Altim-
etry Research in Ocean Circulation Mission.
Citations
Cheney, R.E.. B.C. Douglas, R.W. Agree, L.L. Miller, and
D.L. Porter
1987 Geosat altimeter geophysical data record (GDR) user
handbook. NOAA Tech. Memo. NOS NGS-46, Natl. Geod.
Surv., Rockville, MD. 29 p.
Drinkwater, K.F., and R.A. Myers
1987 Testing predictions of marine fish and shellfish landings
from environmental variables. Can. J. Fish. Aquat. Sci. 44:
1568-1573.
Fogarty, M.J.
1988 Time series models of the Maine lobster fishery: The ef-
fect of temperature. Can. J. Fish. Aquat. Sci. 45:1145-1153.
Grigg, R.W.
1981 Acropora in Hawaii. Part 2. Zoogeography. Pac. Sci.
35(l):15-24.
MacDonald, CD.
1984 Studies on recruitment in the Hawaiian spiny lobster,
Panulinis marginatus. In Proc. Res. Invest. NWHI, p. 199-
220. UNmi-SEAGRANT-MR-84-01, Univ. Hawaii Sea Grant
Coll. Prog., Honolulu.
1986 Recruitment of the puerulus of the spiny lobster, Panu-
lirus marginatus, in Hawaii. Can. J. Fish. Aquat. Sci. 43:
211-2125.
Meyers, G., and J.R. Donguy
1984 The North Equatorial Counter Current and heat storage
in the western Pacific Ocean during 1982-83. Nature (Lond.)
5991 (312)258-260.
Fearce, A.F., and B.F. Phillips
1988 ENSO events, the Leeuwin Current, and larval recruit-
ment of the western rock lobster. J. Cons. Cons. Int. Explor.
Mer 45:13-21.
Phillips. B.F., and P.S. McWilliam
1989 Phyllosoma larvae and the ocean currents off the Ha-
waiian Islands. Pac. Sci. 43(4):352-361.
Pollock, D.E.
1986 Review of the fishery for and biology of the Cape rock
lobster, Jasus Mandii, with notes on larval recruitment. Can.
J. Fish. Aquat. Sci. 43:2107-2117.
Polovina, J.J.
1991 Status of lobster stocks in the Northwestern Hawaiian
Islands, 1990. Admin. Rep. H-91-04, Honolulu Lab., NMFS
Southwest Fish. Sci. Cent., Honolulu, 15 p.
Polovina, J.J., and R.B. Moffitt
1989 Status of lobster stocks in the NWHI. 1988. Admin.
Rep. H-89-3, Honolulu Lab., NMFS Southwest Fish. Sci. Cent.,
Honolulu, 10 p.
Pringle, J,D.
1986 California spiny lobster (Panulirus interruptus) larval
retention and recruitment: A review and synthesis. Can. J.
Fish. Aquat. Sci. 43:2142-2152.
Robinson, M.K.
1969 Theoretical predictions of Subtropical Countercurrent
confirmed by bathythermograph (BT) data. Bull. Jpn. Soc.
Fish. Oceanogr. Spec. (Prof. Uda'sCommem. Pap.):115-121.
Roden, G.I.
1991 Effects of the Hawaiian Ridge upon oceanic flow and ther-
mohaline structure. Deep-Sea Res. (Suppl. 1) 38:S623-S654.
Schnute, J.T., J. Richards, and A.J. Cass
1989 Fish survival and recruitment: Investigations based on
a size-structured model. Can. J. Fish. Aquat. Sci. 46:743-767.
Sinclair, M., M.J. Tremblay, and P. Bernal
1985 El Nino events and variability in a Pacific mackerel (Scom-
ber japonicus) survival index: Support for Hjort's second
hypothesis. Can. J. Fish. Aquat. Sci. 43:602-608.
Uda, M., and K. Hasunuma
1969 The eastward Subtropical Countercurrent in the Western
North Pacific Ocean. J. Oceanogr. Soc. Jpn. 25:201-210.
White, W.B., and K. Hasunuma
1982 Quasi-stationary banded structure in the mean zonal
geostrophic current regimes of the western North Pacific. J.
Mar. Res. 40(4):1035-1046.
Yoshida, K., and T, Kidokoro
1967 Subtropical countercurrent in the North Pacific— An
eastward flow near the Subtropical Convergence. J.
Oceanogr. Soc. Jpn. 23:88-91.
Abstract. - A deepwater trap-
ping sun'ey for Heterocarpits laeviga-
tus was conducted around the main
islands of the Hawaiian Archipelago
to estimate exploitable biomass and
potential yield. Stratified sampling
by depth zone and island was con-
ducted over a 3-year period to evalu-
ate shrimp catch rates. Catchability
of the traps was estimated from a
12-day intensive fishing experiment
performed at a small, isolated site in
the Kaulakahi Channel; habitat areas
were determined by digitizing nauti-
cal charts.
Results from a Leslie analysis of
the depletion experiment showed that
H. laevigatus is very susceptible to
capture by traps (i.e., catchability q
= 9. 48 ha/trap-night). There was no
ewdence of a change in size structure
through the course of the experiment.
Shrimp catch rates varied greatly
by island and depth of capture. Ex-
ploitable biomass was greatest in the
460-640 m depth range; negligible
amounts of shrimp occurred shallow-
er than 350 m and deeper than 830 m.
Catch rates were highest at Niihau
and lowest at Oahu. The total ex-
ploitable biomass of shrimp in the
main Hawaiian Is. was estimated to
be 271 MT, a figure substantially less
than previously believed.
Analysis of multiple size-frequency
distributions for each sex showed no
evidence of modal size progression.
Assuming equilibrium conditions,
application of the Wetherall et al.
(1987) method to these data resulted
in estimates of M/K =1.01 for female
shrimp and 0.74 for males. From
these results and estimates of L„
we calculate that F(i.i/M = 0.75 for
females and 0.86 for males.
An assessment of the exploitable
biomass of Heterocarpus laevigatus
in the main Ha\A/aiian Islands.
Part 1: Trapping surveys, depletion
experiment, and length structure
Stephen Ralston
Tiburon Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA
3150 Paradise Drive, Tiburon, California 94920
Darryl T. Tagami
Honolulu Laboratory, Southwest Fisheries Science Center
National IVIanne Fisheries Service. NOAA
2570 Dole Street, Honolulu, Hawaii 96822-2396
The caridean shrimp, Heterocarpus
laevigatus Bate 1888, is an abundant
deepwater inhabitant of the Hawaiian
Is., where it has been fished sporad-
ically since 1970 (Anon. 1979). Two
early research surveys (Clarke 1972,
Struhsaker and Aasted 1974) showed
that large catches of this species, and
its smaller congener H. ensifer, could
readily be taken in baited traps in
water depths of 365-825 m (200-450
fm).* Based on catch rates from around
the island of Oahu, the deepwater
shrimp resource seemed sufficiently
abundant to support a commercial
fishery (2. 6 kg/trap-night fori/, laein-
gatus and 6. 6 kg/trap-night fori/, en-
sifer; Struhsaker and Aasted 1974).
To date, however, attempts to har-
vest the resource have met with lim-
ited success, even though 75 MT of ii.
laevigatus were landed by one vessel
during a 14-month period (Tagami
and Barrows 1988). At its peak in
1983-84, a short-lived fishery devel-
oped, involving as many as 7 me-
dium-sized (23-40 m) boats. At that
time over 190 MT of H. laevigatus
Manuscript accepted 27 May 1992.
Fishery Bulletin, U.S. 90:494-504 (1992).
'Depths are given in meters with fathom
equivalents (1.0fm= 1.83 m). although con-
touring and stratification were based on
nautical charts in fathoms.
were landed, with an ex-vessel value
of $1.5 million ($7.85/kg; Hawaii Dep.
Land & Nat. Resour., Div. Aquat.
Resour.). The fishery soon failed,
however, primarily due to problems
with gear loss, product processing,
and localized stock depletions. Even
so, the product was well received by
the public, and rejuvenating the
shrimp fishery remains a fishery
development goal of the State of
Hawaii (Anon. 1984).
With this interest in developing the
Hawaiian fishery, major gaps in our
knowledge of local H. laevigatus
shrimp stocks have become apparent
(see Gooding 1984, Dailey and Ral-
ston 1986), although elsewhere in the
Pacific more extensive data are avail-
able (e.g.. Wilder 1977, King 1984
and 1986, Ralston 1986, Moffitt and
Polovina 1987). Particularly lacking
are estimates of the absolute abun-
dance of the H. laevigatus stock in
Hawaii and its ability to withstand
sustained fishing pressure.
The primary objective of the work
presented here was to estimate the
exploitable biomass of H. laevigatus
in the main Hawaiian Is. A secondary
objective was to estimate growth
and mortality rates through analysis
of length-frequency data. Together
494
Ra Iston and Tagami: Exploitable biomass of Heterocarpus laevigatas in Hawaiian Is , Part 1
495
these findings could form a basis for estimating the
potential yield of the shrimp resource.
To determine the exploitable biomass of shrimp at
spatially discrete locations in Hawaii, the formula
(Ricker 1975) was used:
CPUE = q
where CPUE is the catch-per-unit-effort, q is the catch-
ability coefficient of the fishing gear, B is exploitable
biomass, and A is the area occupied by the population.
This relationship is based on the explicit assumption
that catch rate is strictly proportional to shrimp den-
sity (B/A), and that q is the proportionality constant
equating these quantities. By rearrangement we have,
B =
CPUE • A
Thus, to estimate the exploitable shrimp biomass at a
locality we need (1) an unbiased estimate of catch rate,
(2) a measure of the habitat area over which the catch
rate prevails, and (3) knowledge of the sampling gear's
efficiency (i.e., an estimate of q).
To accomplish these three objectives, the study was
divided into two phases. First, a depletion experiment
was conducted to estimate the catchability coefficient
(q). This was followed by a depth-stratified sampling
program for H. laevigatus around each of the main
islands of the Hawaiian archipelago (i.e., Hawaii, Maui,
Kahoolawe, Lanai, Molokai, Oahu, Kauai, and Niihau).
Methods
Table I
Summary of shrimp-trapping cruise dates
and locations.
No. of
Date
Location
traps hauled
April 1985
Molokai
3
Lanai
1
May 1986
Kaulakahi Channel
105
July 1986
Kaulakahi Channel
10
Sept. 1986
Kaulakahi Channel
10
Nov. 1986
Kaulakahi Channel
12
Sept./Oct. 1987
Niihau
68
Kauai
128
Kaulakahi Channel
11
Feb./ilarch 1988
Kaulakahi Channel
10
Oahu
8
Hawaii
31
Oct. 1988
Kaulakahi Channel
17
Molokai
20
Lanai
18
Maui
38
Kahoolawe
8
March 1989
Oahu
67
Molokai
32
face, each trap was emptied and the contents were
sorted to species, counted, and weighed to the nearest
0.01kg. Random subsamples of ~200 H. laevigatus
were routinely collected, from which carapace length
(CL) was measured to the nearest 0.1mm using dial
calipers. For all measured shrimp, sex was determined
by examining the endopodite of the first pleopod
(spatulate in males, pinnate in females; see King and
Moffitt 1984). In addition, the ovigerous condition of
females was recorded.
Shrimp trapping was conducted during a series of nine
cruises of the NOAA ship Townsend Cromwell (Table
1). During each cruise, standard fishing gear was util-
ized to gather CPUE statistics at specific geographical
locations. The gear employed was a top-loading pyra-
midal shrimp trap, identical in construction to those
used commercially in Hawaii from 1983 to 1984. Each
trap was made of welded steel reinforcement bars, had
a 1.83 m- base, an overall volume of 1.84m^, and was
covered by 1.27x2.54 cm mesh hardware cloth. A full
description and illustration of the gear is given in
Tagami and Barrows (1988).
Typically, 6-10 solitary traps were set daily and
allowed to soak overnight. Traps were generally de-
ployed in the afternoon and hauled the following morn-
ing, being in the water for a period of 16-20 hours. All
traps were baited with approximately 3 kg of chopped
mackerel Scomber japonicus. After hauling to the sur-
Depletion experiment
To estimate q, an intensive fishing experiment was con-
ducted (see also Ralston 1986). Depletion experiments,
including the Leslie method used here (Ricker 1975),
have three restrictive assumptions. First, all individuals
in the exploitable portion of the population are equally
likely to be captured with the fishing gear. Second, the
fished population is closed, or else additions exactly
balance removals other than those due to fishing. Third,
fishing removals account for all changes in stock bio-
mass, such that natural mortality, growth, and recruit-
ment have negligible effects during the period of
fishing. Thus, the best site for a depletion experiment
is a naturally isolated, small area so that removals can
be carried out over as short a time-interval as possible.
A small rise midway in the Kaulakahi Channel (21°
54.5'N, 159°56.5'W) separating Kauai and Niihau was
496
Fishery Bulletin 90(3), 1992
chosen for the work. This nearly circular rise (Fig. 1)
with a crest at 421m (230fm), has an area of 1187ha
(horizontal planar area <640m or 350 fm) and is
isolated from the Islands of Kauai and Niihau by depths
>730m (400 fm; Fig. 2). The site lies in the required
depth range for H. laevigatus and has relatively high
densities of the target species.
The intensive fishing experiment was conducted
13-24 May 1986. During each of the 12 days of the
experiment, 6-14 pyramidal shrimp traps were set be-
tween depths of 421 and 695 m (230 and 380 fm).
Following the Leslie method (Ricker 1975, Seber 1982),
catchability was estimated directly from the slope of
the linear regression of CPUE on corrected cumulative
catch. That is,
ca
57
56
55
to
0)
■+->
54
53 I I I I I I I I I I I I I I I I I I I I I I I I I
59 58 57 56 55 54
Longitude
(minutes W of 159°)
Figure 1
Contour map of the Kaulakahi Channel study site. Depths are in fathoms,
and circles represent trap-set locations.
22°30'
CPUEi = q • Bj
= q(Bo - K,)
= qBo - qKi,
where CPUE; is the catch-per-unit-effort
on day i (kg/ trap-night), q is the catchabil-
ity coefficient/trap-night of the pyramid
traps, Bj is the average biomass (kg) pres-
ent on day i, Bq is the biomass (kg) of
shrimp present at the start of the experi-
ment, and Kj is the corrected cumulative
removals for day i, defined as
i-l
K,
2 n = l
0)
^ 22 00'
CO
-J
160° 30'
WO'OO'
159'30'
Longitude (W)
Figure 2
Contour map of the Kauai-Niihau area depicting the 100, 200, 300, 400, 500, 600, and
lOOOfm isobaths. Note the depletion study site, i.e., the small rise located mid-channel.
where Cn(n = l,2 i) is the catch (kg)
taken on each day of the experiment. Note
that the estimate of catchability (q) pertains
strictly to the interaction between the traps
we used and the stock resident in the study
area, which is normalized to unit
area after multiplying by 1187ha
(i.e., the area of the study site).
Determination of
habitat areas
The distribution of H. laevigatus
is strongly dependent on bottom
depth. Little or no shrimp occur
in waters outside the 366-915 m
(200-500 fm) range (Struhsaker
and Aasted 1974, Gooding 1984,
Dailey and Ralston 1986). In the
main portion of the Hawaiian
Archipelago (Hawaii, Maui, La-
nai, Kahoolawe, Molokai, Oahu,
Kauai, and Niihau) a number of
islands share a common 915m
(500 fm) depth contour (e.g, Kau-
ai and Niihau; Fig. 2). Even so,
in our study each island was
treated as a separate stock for
Ralston and Tagami: Exploitable biomass of Heterocarpus laevigatus in Hawaiian Is , Part
497
purposes of geographically stratifying the analysis. An
exception was made for the islands of Maui, Lanai,
Kahoolawe, and Molokai (MLKM), which share, in ad-
dition to the 915 m (500 fm) contour, a common 366 m
(200 fm) isobath. These four islands were, therefore,
treated as a single geographic locality.
Estimates of the amount of suitable shrimp habitat,
in hectares (lha = 0.01km"), were obtained by deter-
mining the horizontal planar area lying between
charted depth contours. A large digitizing tablet was
used to calculate all area estimates directly from nau-
tical charts (NOAA charts 19016, 19019, and 19022,
and Defense Mapping Agency bottom contour charts).
These charts included 915m (500 fm) isobaths, but we
manually contoured all of the 366 m (200 fm) isobaths
using the sounding data provided on each chart. In
addition, good detailed bathymetry was available for
the islands of Kauai and Niihau, and at these sites the
458 m (250 fm), 549 m (300 fm), 640 m (350 fm), 732 m
(400 fm), and 824 m (450 fm) isobaths were contoured
and digitized as well.
Each contour was digitized three times by each
author, providing a minimum estimate of measurement
error in our calculation of habitat areas. These errors
were typically small (median CV 0.5%, range 0.1-
1.9%). A potentially more serious type of error con-
cerns discrepancies between the actual locations of con-
tours and their representations on charts. However,
we had no information concerning the magnitude of
this type of error and, given that measurement errors
were negligible, we assumed that our estimates of
habitat area were accurate and precise.
These data were then used to calculate habitat areas
for each 92 m (50 fm) depth interval between 366 and
915 m (200-500 fm). First, the relative distribution of
habitat was calculated from the Kauai and Niihau data.
To estimate depth-specific habitat areas for the three
remaining sites (Oahu, MLKM, and Hawaii), the com-
bined relative proportions of habitat for each depth
interval obtained at Kauai-Niihau were applied to
the estimates of total habitat area between 366 and
915m (200-500fm). In support of this procedure,
results in Mark and Moore (1987) indicate that slope-
depth relationships among the main islands of the
archipelago are, in general, similar.
Depth-stratified sampling
For the second phase of the assessment, each of the
island areas was targeted for comprehensive trapping
surveys to determine abundance patterns (i.e., catch
rate) with depth and to estimate standing stocks (Table
1). A depth-stratified sampling approach was used.
From preliminary data gathered at Kauai and Niihau
during the September 1987 cruise, the mean and vari-
ance in CPUE were calculated for each of the six 92m
(50fm) depth intervals lying in the 366-915m (200-
500 fm) range. Based on the results of this vertical
distribution survey, sampling effort was optimally par-
titioned into depth strata by Neyman allocation (Coch-
ran 1977), i.e., trap allocations to each depth interval
were based on the product of abundance (CPUE • habi-
tat area) and the standard deviation of CPUE at that
depth. As each cruise progressed, CPUE means and
variances \\fere recalculated daily and the trap alloca-
tion schedule was updated.
From the results of the surveys, exploitable biomass
was estimated (Eq. 1) for each depth interval at each
site visited. This calculation assumes that the catch-
ability estimate, which was determined at the deple-
tion experiment study site, can be extended to all other
localities sampled. An estimate of the variance of the
biomass for each stratum was obtained from Eq. 1
using the delta method (Seber 1982), resulting in
A2 A^-CPTJF^
VAR[B] = — VAR[CPUE] + VAR[q]
q2 q4
if all covariance terms are zero (a reasonable first
assumption) and VAR[A] is negligible (see above). Con-
fidence intervals were then calculated using the distri-
bution of standard normal scores (a = 0.05, Z = 1.96).
Length-frequency analysis
The Kaulakahi Channel experimental depletion site was
visited on seven separate occasions during May 1986-
March 1988 (Table 1). During each visit a length-
frequency sample of H. laevigatus was obtained, with
the ultimate goal of analyzing the progression of size
modes over time (egg bearing is strongly seasonal;
Dailey and Ralston 1986, Moffitt and Polovina 1987).
Additional length-frequency samples were obtained
during the course of the depth-stratified sampling at
each of the island sites.
Mortality and growth parameters were estimated
from length-frequency distributions using the regres-
sion method of Wetherall et al. (1987). This technique
requires an equilibrium population size-structure, an
undesirable and restrictive assumption. Even so, data
are available to support its use. Dailey and Ralston
(1986) present length-frequency data for male and
female shrimp sampled during the earliest stages of the
fishery (1983-84). The data are very similar to those
presented here, suggesting that exploitation has yet
to seriously affect size composition. Additionally, the
time-invariance of the size-frequency data we collected
at the Kaulakahi study site (see below) indicates
equilibrium conditions.
498
Fishery Bulletin 90(3), 1992
n
o
a
Pi
05
u
65 99
55 ■' ''~-----' " ---
*^' 1 Jl-n-Jl 1 I CL -□
25 _ ^ _! ^ ,
S ^ -
55 ' - . _
25 ,-.__^ _. ^ ,._
15 — . — . — . — . . . — . . — . . — , , —
1 2 3 4 5 6 7 B 9 10 11 12
Day of Experiment
Figure 3
Distributions of carapace length for male and female
Heterocarpus laevigatiLS during the depletion experi-
ment. Open squares are means bracketed by ± 1 SD;
dashed lines represent maximum and minimum ob-
served sizes.
For each sex, the von Bertalanffy asymptotic length
(CL^) and the ratio (0) of the total mortality rate
(Z/yr) to the von Bertalanffy growth coefficient (K/yr)
were estimated from a weighted regression of mean
lengths (CL^i) on full vulnerability cutoff lengths
(CLei). In the analysis, the CLd were incremented in
1.0 mm steps above CL^o, the minimum size of full
vulnerability to the gear, and the CLj^j were recalcu-
lated at each step. Using the morphometric functional
regressions presented in Dailey and Ralston (1986),
CLco was set equal to 30 mm (the length of H. laeviga-
^Ms ■■Anth carapace width = 12.7mm), corresponding to
the least dimension of the wire mesh covering the
traps. This approach to estimating CLpo deviates from
that used in previous applications of the regression
method to H. laevigatus stocks. Dailey and Ralston
(1986) and Moffitt and Polovina (1987) both assumed
that shrimp were not fully selected by baited traps until
reaching a CL greater than the modal size of length-
frequency distributions from trap catches.
Because the size-structure we observed was similar
to the unexploited stock (see above), it follows that
0 = M/K, where M is natural mortality/yr. Likewise,
using the length-weight regressions of Dailey and
Ralston (1986), we estimated the asymptotic weights
ta
o
f 40
0
a
to
o o
<^ 30
OB
' I
.M
g-____8° ° 9 ° o
20
V
° ^ ^ ~ — -S.-^ 0 8 ° 0 °
to
^ 10
fi o " o 0 r^ &-~.^
^ ° ° 8 o ° § T^T9~
X!
S 0 .
O Bo 0 o o 0
■4->
O O O 0
5 o\
0 ° n
0 500 1000 1500
Corrected Cumulative Removals (kg)
Figure 4
Leslie model applied to the experimental depletion oiHetero-
carpus laevigatus at the Kaulakahi Channel study site. Each
point represents one valid trap-night of fishing.
(W^) for each sex as the predicted weight at CL =
CL^. The average size at entry to the fishery (CLp
sensu Beverton and Holt 1957) was obtained by averag-
ing the minimum size caught and CLj-^,. Given esti-
mates of M/K, W^, and CLp, we used the tables
presented in Beverton and Holt (1966) to determine
sex-specific values of yield-per-recruit (Y/R) at various
levels of exploitation (F/M). From these data we com-
puted values of Fq j/M, the exploitation level at which
the marginal increase in Y/R declines to 10% of its
value at the origin (Gulland and Boerema 1973, Gulland
1983).
Results
Depletion experiment
During the depletion experiment, 123 pyramid shrimp
traps were set at the Kaulakahi Channel study site. Of
these, 19 were lost, resulting in 104 effective trap-
nights of standard fishing effort and a gear loss rate
of 15%. A total of 45,482 H. laevigatus were caught,
which collectively weighed 1499 kg. The average size
of each shrimp was therefore 33. Og. During the 12-day
course of the experiment, no change occurred in the
daily mean size of shrimp caught (Fig. 3; r<j = 0.32,
df 10; ro.= -0.28, df 10).
Individual trap catches were regressed on values of
corrected cumulative removals to date (Fig. 4). Traps
that did not fish properly (e.g., the funnel entrance was
ajar upon retrieval) were not included, although cumu-
lative removals (Kj) included all shrimp caught in the
study area (<695m or 380 fm). Therefore, each point
represents an observation of CPUE from one valid
Ralston and Tagami: Exploitable biomass of Heterocarpus laevigatas in Hawaiian Is , Part
499
Depth (m)
400
500
600
700
BOO
goo
Kauai
KSl Niihau
^M Average
200 250 300 350 400 450 500
Depth Interval (fathom)
Figure 5
Distribution of habitat by depth at Kauai and Niihau. The com-
posite average distribution is based on poohng habitat at these
two sites.
400
Depth (m)
500 BOO 700
eoo
900
Niihau
- ■ — ■ Kauai
Hawaii
MLKM
Oahu
200 250 300 350 400 450
Depth Interval (fathom)
500
Figure 6
Average CPUE oi Heterocarpiis laengatus relative to depth
at each site sampled.
trap-night of fishing. Also presented is the ordinary
least-squares regression equation relating these
variables. The equation of the line is
CPUEj = 22.84 - 0.007988[K,],
with standard errors of the slope and intercept equal
to 0.001997 and 1.9368, respectively. The regression
is highly significant (Fi,sy= 16.00, P 0.0001). The re-
siduals show no obvious departure from linearity, an
indication of constant catchability.
Under the Leslie model, the exploitable biomass
at the start of the experiment is defined by the x-
intercept, i.e., 2859kg. Because the study site covered
1187 ha, this amounts to an initial density of 2.4087
kg/ha, which produced an initial catch rate (CPUEo)
equal to 22.84 kg/trap-night (i.e., the y-intercept). Then,
q expressed on a hectare basis, rather than defined in
terms of the study site, is estimated to be 9.4798ha/
trap-night. In real terms, one overnight soak of a
single, large pyramidal shrimp trap is estimated to have
captured ~0.8% of the shrimp in the entire study area,
which is equivalent to all the shrimp in ~9.5ha, were
they randomly dispersed. The standard error of q/ha
is the product of A (1187 ha) and the standard error
of q measured over the study site (0.001997) (Seber
1982).
It is instructive to note that an initial density of
2.4087 kg/ha is equivalent to an average of 73 shrimp/
ha (see statistics for mean shrimp weight above). Based
on this density, the average utilization of habitat by
each shrimp was 137 m^, a remarkable figure given
the relatively high catch rate encountered at the begin-
ning of the depletion experiment (22.83 kg/trap-night).
Depth-stratified sampling
Habitat areas by 92m (50fm) depth intervals were
estimated for the Oahu, Hawaii, and MLKM sites by
assuming that the proportionate distribution of habitat
between 366 and 915 m (200-500 fm) at these sites was
the same as the composite distribution obtained at
Kauai and Niihau. While there were some differences
in the distribution of habitat with depth between Kauai
and Niihau (Fig. 5, Table 2), they were relatively minor.
At both sites the amount of habitat in the 458-549 m
(250-300 fm) depth interval was slightly less than in the
366-458 m (200-250 fm) interval; but with increasing
depth below that, the amount of habitat per 92 m (50 fm)
depth interval increased steadily.
There were, however, marked differences in CPUE
with depth among the five localities sampled (Fig. 6,
Table 2). Catch rates at Niihau were particularly high
relative to the other areas, especially at 458-549 m
(250-300 fm). Catch rates at Oahu and MLKM were
much lower. The modes of the distributions at Kauai
and Niihau were shifted to the shallow end of the depth
range, whereas at Hawaii it was shifted deeper.
Results presented in Table 2 provide estimates of the
exploitable biomass (B), as well as variance estimates,
for the depth intervals sampled at each site. Although
catch rates at Niihau are quite high, the reduced
amount of habitat at this island (69,530ha) is sufficient
to support only a small stock of shrimp (35.7 MT). Oahu,
with its much lower catch rates, has a larger exploit-
500
Fishery Bulletin 90(3). 1992
Table 2
Depth-stratified sampling
results for
Heterocarpus laevigat
us in the main
Hawaiian Is
Depth
range
No. traps
Mean CPUE
Habitat area
B
(m)
(fm)
set
kg/trap
VAR[CPUE]
(ha)
(kg)
VAR[B]
Kauai
366-458
200-250
29
1.566
0.894
12,760
2,108
1,897,271
458-549
250-300
37
7.587
1.093
12,070
9,663
7,608,356
549-640
300-350
25
3.607
0.322
14,410
5,481
2,621.353
640-732
350-400
18
1.324
0.436
16,940
2,367
1,742,930
732-824
400-450
6
0.953
0.203
17,770
1,786
912,472
824-915
450-500
7
0.299
0.038
20,200
637
197.960
Totals
122
94,150
22,042
14,980,342
Niihau
366-458
200-250
9
0.788
0.586
10,670
887
791,140
458-549
250-300
18
12.214
7.253
10,570
13,611
20.586,349
549-640
300-350
19
8.719
3.335
11,010
10,127
10,907,759
640-732
350-400
10
5.533
2.737
11,730
6,847
7,120,644
732-824
400-450
4
2.903
0.906
12,210
3,739
2,377,052
824-915
450-500
4
0.380
0.094
13,340
535
204,092
Totals
64
69,530
35.746
41,987.037
Oahu
366-458
200-250
6
0.122
0.017
41,570
535
342,588
458-549
250-300
20
3.482
3.239
40,100
14,726
71,490,061
549-640
300-350
33
3.115
0.480
45,070
14,810
24,568,526
640-732
350-400
13
1.594
0.698
50,800
8,542
24,613,230
732-824
400-450
2
1.990
0.303
53,300
11,187
17,379,858
824-915
450-500
0
0.000
0.000
59.510
0
0
Totals
74
290,350
49,800
138,394,264
Hawaii
366-458
200-250
1
3.270
2.152
38,720
13,357
47,049,206
458-549
250-300
4
3.750
2.223
37,320
14,767
48,097,991
549-640
300-350
11
5.008
3.882
41,980
22,171
106,796,660
640-732
350-400
11
6.325
3.036
47,300
31,560
137,841,468
732-824
400-450
4
0.858
0.889
49,630
4,491
25,603,225
824-915
450-500
0
0.000
0.000
55,430
0
0
Totals
31
270,350
86,345
365,388,550
Maui-Lanai-Kahoolaw«
-Molokai
366-458
200-250
15
0.500
0.110
69,420
3.662
6,716,452
458-549
250-300
43
3.338
0.399
66,920
23,567
54,580.681
549-640
300-350
42
2.823
0.168
75,250
22,409
41.983.921
640-732
350-400
19
2.795
0.653
84.790
25,004
91,353,003
732-824
400-450
1
0.300
0.079
88,970
2,815
7,443,949
824-915
450-500
0
0.000
0.000
99,370
0
0
Totals
120
484.730
77,457
202,078,006
able biomass oiH. laevigatus than does Niihau. In ag-
gregate, we estimate the exploitable stock at all islands
to be Bitot = 271.4 MT (p[217.3<Bii„t<325.5] =0.95,
SE 27.6MT, CV 10.18%).
Analysis of length-frequency data
Although the Kaulakahi Channel study site was sampled
on seven different occasions over a 29-month period
(May 1986-October 1988), during which over 6800
female and 1 1 ,800 male shrimp were sexed and mea-
sured, there was little evidence of progression in size
modes (Fig. 7). The CL frequency distributions of male
shrimp were particularly stagjiant, and those of fe-
males showed no coherent pattern that could be attrib-
uted to the influx of year-classes into the exploitable
population.
Due to the apparent stationary behavior of these
Ralston and Tagami: Exploitable biomass of Heterocarpus laevigatas in Hawaiian Is , Part 1
501
99
cTcT
UnT 14-25, 19Be N - Z.448
July 11.1
■ 16 20 26 30 36 40 46 60 65 15 20 26 30 35 40 46 60 65 80
Carapace Length (mm)
Figure 7
Length-frequency distributions of female and male Heterocar-
pus laevigatus at the Kaulakahi Channel study site.
o
fl
4)
cr
■•J
a
0)
u
u
u
a,
12.0
10.0
B.O
6.0
4.0
2.0
0.0
$9 N- 16.970
Cfcf N-Z4.0B4
FuU
Recruitment
16 20 26 30 36 40 46 60 66 60
Carapace Length (nun)
Figure 8
Composite length-frequency distributions for all female and
male Heterocarpus lamgatiis sampled.
--' 7.0-
^ 6.0
b 5.0
y^'
(S 4.0
/ Fo..
fe 3.0 ]
13 1.0-
>- 0.0-
0.
/ Optimum
1 w
.... cT
— 9
00 0.50 1.00 1.50
2.00 2.50
F/M
Figure 9
Yield-per-recruit oi Heterocarpus laevigatus
and estimates of
optimum exploitation level (Fo,/M).
distributions, all the length data were pooled (Fig. 8).
It is evident from the figure that female shrimp reach
substantially larger sizes than do males, in agreement
with previously published work (Dailey and Ralston
1986, Moffitt and Polovina 1987). Superimposed on the
combined length-frequency distributions of males and
females is the estimated size of//, lamngatus when fully
vulnerable to the traps (i.e., CLco = 30mmCL). The
carapace width of shrimp this size is equal to 1.27 cm
(0.5"), the minimum mesh size of the traps.
We applied the Wetherall et al. (1987) regression
method to these length-frequency data and estimated
that 0, = 1.01 ± 0.052 and ©o- = 0.74 ± 0.075. For fe-
males we estimated that CL^ = 58 mm (SE 0.37) and
for males CL^ = 50mm (SE 0.44), corresponding to
W^ = 80g for females and 55g for males.
The smallest shrimp we captured in the pyramid
traps was 16mmCL. Below this size, selectivity of the
gear was zero. By equating carapace width and mesh
size, we determined that CLeo = 30mm. Thus, our
estimate of the size at 50% recruitment to the fishery
(CL, ) is 23 mm. Given this result, and sex-specific
estimates of 0 and W^ , we calculate that for females
Fo.i/M = 0.75 and for males Fo.i/M = 0.86 (Fig. 9).
Discussion
In this assessment, projections of exploitable biomass
depend greatly on the estimate of catchability obtained
from the Leslie depletion experiment. Due to its major
role in the calculations, sources of bias in its estima-
tion must be considered carefully.
502
Fishery Bulletin 90(3). 1992
There is evidence to show that estimates of crusta-
cean population size obtained through survey removal
methods like the Leslie and DeLury methods (Ricker
1975) may severely misrepresent the actual size of the
population. In his study, Morrissy (1975) found that
DeLury estimates of population number of Cherax
tenuimanus were anywhere from 39 to 53% of those
based on a complete count of the population. Similar-
ly, DeLury estimates of the population density of
Panulirus cygnus were 25% of those estimated from
diver counts (Morgan 1974a).
The most likely origin of bias in these situations is
that not all individuals in the exploitable portion of the
population are equally vulnerable (sensu Morrissy 1973)
to the gear. For example, when sampling with baited
drop nets, the catch of C. tenuimanus in intermolt, ex-
pressed as a known fraction of the actual population,
was much higher than the catch in a premolt condition;
individuals in molt stages immediately preceding and
following ecdysis were not caught at all (Morrissy
1975).
It is possible that a similar bias was operating dur-
ing the depletion study at the Kaulakahi Channel study
site. The presence of shrimp in the exploitable portion
of the population, which were less susceptible to trap-
ping, would result in overestimation of catchability and
underestimation of biomass. Factors such as molt stage
(Morgan 1974b, Morrissy 1975), sex (Morrissy 1973),
and feeding history (Sainte-Marie 1987) are known to
cause variation in vulnerability to trapping. Although
social dominance, mediated through differences in size,
can affect catchability (Morrissy 1973, Chittleborough
1974), it is unlikely to have substantially biased our
results because the catch size-structure remained un-
changed during the experiment (Fig. 3). Also, Gooding
et al. (1988) did not see "any overt aggressive behav-
ior" among H. ensifer that were active around baits.
Lastly, seasonal and interannual alterations in catch-
ability due to temperature (Chittleborough 1970,
Morgan 1974b), salinity (Morgan 1974b), and food
availability (Chittleborough 1970) operate over longer
time-scales than the depletion experiment.
Two other lines of evidence support the premise that
shrimp biomass was underestimated. Although results
presented in Figure 4 indicate that 12 days of trapping
dropped the catch rate to 48% of its starting value
(10.87 kg/trap-night), it had risen to 19.73 kg/trap-night
when resampled 47 days later (data from 9-11 July
1986 cruise; see Table 1). If catchability is estimated
from the decline in catch rate that occurred between
the beginning of the depletion study (22.84 kg/trap-
night) to the time the site was resampled 2 months
later, and we assume the decline was due only to trap
removals (1499.00kg), we obtain q = 2.4624 ha/trap-
night. This represents a 74% reduction in the estimate,
which in turn would inflate biomass estimates by a fac-
tor of 3.85 (i.e., Bitot = 1050 MT). This relative bias is
similar to that reported by Morgan (1974a) for Panu-
lirus cygnus (see above).
From submersible observations of H. laevigatus den-
sity, Moffitt and Parrish (1992) obtained q = 0.2895
ha/trap-night for the same traps we used, amounting
to a 33-fold difference relative to our Leslie analysis.
They, too, expressed concerns about bias in catchability
estimates derived from depletion experiments due to
variable susceptibility to the gear. Conversely, their
estimate of catchability was based on comparing site-
specific March 1988 trap catches with submersible
observations made during August, even though H. lae-
vigatus undergoes seasonal vertical migrations (King
1983, Dailey and Ralston 1986). In addition, at the start
of each dive, they deployed a baited trap in the area
of the submersible. Both factors could lead to under-
estimation of catchability.
It is clear that biased estimates of q will result if the
probability of capture is not uniform among shrimp.
In an attempt to solve this problem, Quinn (1987)
developed a depletion model that explicitly incorpor-
ated a term for non-constant catchability. Application
of his model to Pacific halibut effectively accounted for
short-term trends in q, but auxiliary estimates of fish-
ing and natural mortality were required, data that are
unavailable here.
The primary objective of this study was to determine
the exploitable biomass of H. laevigatus in the main
Hawaiian Is. (MHI). Even if shrimp biomass were as
high 1050 MT, rather than 271 MT (see above), our
results indicate that the MHI stock is much smaller
than previously believed, and that prior estimates of
maximum sustainable yield (MSY) are much too high.
For example, Struhsaker and Aasted (1974), using
figures from a fishery for H. reedi off the coast of ChUe,
speculated that H. ensifer in Hawaii could sustain a
level of production equal to 10-20 kg/ha ■ yr-^. If
H. laevigatus were assumed to be equally productive
(e.g.. Anon. 1979), then, given there are ~350,000ha
of prime habitat at 458-640 m (250-350 fm) in the MHI
alone (Table 2), the resulting estimate of MSY exceeds
stock biomass many times over. Moreover, catch rates
of//, laevigatus in the distant Northwestern Hawaiian
Is. (Nihoa to Kure), which represent a similar amount
of shrimp habitat as the MHI, are no more than half
those observed in the MHI (Gooding 1984, Tagami and
Barrows 1988, Tagami and Ralston 1988).
Our estimates of 0, = 1.01 and ©(^ = 0.74 are much
lower than those given in Dailey and Ralston (1986),
who reported 0g =2.9 and Q^, = 4.3. By requiring CL^
to be greater than the modal size, they effectively con-
strained 0 to values much greater than unity. A similar
requirement was imposed by Moffitt and Polovina
Ralston and Tagami: Exploitable biomass of Heterocarpus laevigatas in Hawaiian Is , Part
503
(1987), who estimated 0g = 1.9 and ©o- = 2.1 for H. lae-
vigatus in the Mariana Is. Additionally, use of the
"mode" criterion to establish the minimum size at full
vulnerability results in substantially greater sensitiv-
ity of 0 to input estimates of CLeo- For example, if
CLfo is 30 mm as we suggest (Fig. 8), ±2 mm pertur-
bations in CLco result in - 9% and -i- 12% changes in
estimates of Q„. However, if the mode of the size-
frequency distribution is used instead (CLeo = 44 mm),
the same perturbations alter estimates of Q„ by
- 16% and -i- 46%. Similar sensitivity was observed in
estimates of 0, . In summary, it is our belief that in-
dependent estimates of CL^o are superior to those ob-
tained from the size data analyzed, particularly when
there is no reason to suspect that agonistic interactions
affect the catch size-structure.
Lower values of 0 indicate a reduction in the instan-
taneous mortality rate (Z), an increase in the von Ber-
talanffy grovvi;h coefficient (K), or both. We favor the
first hypothesis, largely because of the cold (4-6°C),
trophically-impoverished habitat in which H. laeviga-
tus reside. In many respects these shrimp represent
a crustacean analog to Pacific ocean perch Sebastes
alutus, which grow very slowly, exhibit extreme
longevity, and display low rates of natural mortality
(Leaman and Beamish 1984). Consequences of this life-
history pattern are that (1) under pristine conditions,
individuals accumulate in the largest size-categories
(Fig. 8), (2) the ratio of production to biomass is low,
and (3) stocks are very susceptible to overfishing.
Acknowledgments
We would like to thank the scientists and crew of the
NOAA RV Townsend Cromwell for their assistance in
helping us complete this study. This paper benefited
from reviews by G.W. Boehlert, A.D. MacCall, R.B.
Moffitt, and J.J. Polovina.
Citations
Anonymous
1979 Hawaii fisheries development plan. Hawaii Dep. Land
and Nat. Resour., Honolulu, 297 p.
1984 Status of fisheries assessment of development and man-
agement needs for selected crustacean species in the western
Pacific region. West. Pac. Reg. Fish. Manage. Counc. Hono-
lulu. 60 p.
Beverton, R.J.H., and S.J. Holt
1957 On the dynamics of exploited fish populations. U.K.
Minist. Agric. Fish., Fish. Invest. (Ser. 2) 19, .533 p.
1966 Manual of methods for fish stock assessment. Part. 2.
Tables of yield functions. FAO Fish. Tech. Pap. 38, Rev. 1,
67 p.
Chittleborough, R.G.
1970 Studies on recruitment in the western Australian rock
lobster Panulinis hnffipes cyg7ius George; Density and natural
mortality of juveniles. Aust. J. Mar. Freshwater Res. 21:
131-148.
1974 Home range, homing and dominance in juvenile western
rock lobsters. Aust. J. Mar. Freshwater Res. 25:227-234.
Clarke, T.A.
1972 Collections and submarine observations of deep benthic
fishes and decapod Crustacea in Hawaii. Pac. Sci. 26:310-317.
Cochran, W.G.
1977 Sampling techniques. Wiley, NY, 428 p.
Dailey, M.D., and S. Ralston
1986 Aspects of the reproductive biology, spatial distribution,
growth, and mortality of the deepwater caridean shrimp,
Heterocarpus laevigatas, in Hawaii. Fish. Bull., U.S. 84:
915-925.
Gooding, R.M.
1984 Trapping surveys for the deepwater caridean shrimps.
Heterocarpus laevigatus and H. ensifer in the Northwestern
Hawaiian Islands. Mar. Fish. Rev. 46(2):18-26.
Gooding, R.M., J.J. Polovina, and M.D. Dailey
1988 Observations of deepwater shrimp, Heterocarpus ensifer,
from a submersible off the island of Hawaii. Mar. Fish. Rev.
50(l):32-39.
Gulland, J.A.
1983 Fish stock assessment. A manual of basic methods.
Wiley, NY, 223 p.
Gulland, J. A., and L.K. Boerema
1973 Scientific advice on catch levels. Fish. Bull., U.S. 71:
325-335.
King. M.G.
1983 The ecology of deepwater caridean shrimps (Crustacea:
Decapoda: Caridea) near tropical Pacific Islands with particular
emphasis on the relationship of life history patterns to depth.
Ph.D. thesis. Univ. South Pacific. Suva, Fiji, 258 p.
1984 The species and depth distributions of deepwater cari-
dean shrimps (Decapoda, Caridea) near some southwest Pacific
Islands. Crustaceana 47:174-191.
1986 The fishery resources of Pacific island countries. Part
1. Deep-water shrimps. FAO Fish. Tech. Pap. (272.1). 45 p.
King, M.G.. and R.B. Moffitt
1984 The sexuality of tropical deepwater shrimps (Decapoda:
Pandalidae). J. Crustacean Biol. 4:567-571.
Leaman, B.M., and R.J. Beamish
1984 Ecological and management implications of longevity in
some northeast Pacific groundfishes. Int. N. Pac. Fish.
Comm. Bull. 42:85-97.
Mark, R.K., and J.G. Moore
1987 Slopes of the Hawaiian ridge. In Decker, R.W. (ed.),
Volcanism in Hawaii, p. 101-107. U.S. Geol. Surv. Prof. Pap.
1350.
Moffitt, R.B., and F.A. Parrish
1992 An assessment of the exploitable biomass oi Heterocar-
pus laevigatus in the main Hawaiian Islands. Part 2: Obser-
vations from a submersible. Fish. Bull. U.S. 90:476-482.
Moffitt, R.B., and J.J. Polovina
1987 Distribution and yield of the deepwater shrimp Heterocar-
pus resource in the Marianas. Fish. Bull., U.S. 85:339-349.
Morgan, G.R.
1974a Aspects of the population dynamics of the western rock
lobster, Panulirus cygn.us George. I. Estimation of population
density. Aust. J. Mar. Freshwater Res. 25:235-248.
504
Fishery Bulletin 90(3). 1992
1974b Aspects of the population dynamics of the western rock
lobster, Panulirus cygnus George. II. Seasonal changes in the
catchability coefficient. Aust. J. Mar. Freshwater Res. 25:
249-259.
Morrissy, N.M.
1973 Normal (Gaussian) response of juvenile marron, Cherax
tenuimanus (Smith) (Decapoda: Parastacidae), to capture by
baited sampling units. Aust. J. Mar. Freshwater Res. 24:
183-195.
1975 The influence of sampling intensity on the 'catchability'
of marron, Cherax tenuimanus (Smith) (Decapoda:
Parastacidae). Aust. J. Mar. Freshwater Res. 26:47-73.
Quinn, T.J. Ill
1987 Standardization of catch-per-unit-effort for short-term
trends in catchability. Nat. Resour. Model. l(2):279-296.
Ralston, S.
1986 An intensive fishing experiment for the caridean shrimp,
Heterocarpus laevigatus, at Alamagan Island in the Mariana
Archipelago. Fish. Bull., U.S. 84:927-934.
Ricker, W.E.
1975 Computation and interpretation of biological statistics
offish populations. Fish. Res. Board Can. Bull. 191, 382 p.
Saint-Marie, B.
1987 Meal size and feeding rate of the shallow- water lysianassid
Anonyx sarsi (Crustacea: Amphipoda). Mar. Ecol. Prog. Ser.
40:209-219.
Seber, G.A.F.
1982 The estimation of animal abundance, 2d ed. Griffin, Lon-
don, 506 p.
Struhsaker, P., and D.C. Aasted
1974 Deepwater shrimp trapping in the Hawaiian Islands.
Mar. Fish. Rev. 36(10):24-30.
Tagami, D.T., and S. Barrows
1988 Deep-sea shrimp trapping Heterocarpus laevigatus in the
Hawaiian Archipelago by a commercial fishing vessel. NOAA
Tech. Memo. NOAA-TM-NMFS-SWFC-103, NMFS Honolulu
Lab., 14 p.
Tagami, D.T., and S. Ralston
1988 An assessment of exploitable biomass and projection of
maximum sustainable yield for Heterocarpus laevigatus in the
Hawaiian Islands. Admin. Rep. H-88-14, Honolulu Lab.,
Southwest Fish. Sci. Cent., 22 p.
Wetherall, J. A., J.J. Polovina, and S. Ralston
1987 Estimating growth and mortality in steady-state fish
stocks from length-frequency data. In Pauly, D., and G.R.
Morgan (eds.). Length-based methods in fisheries research, p.
53-74. ICLARM Conf. Proc. 13., Int. Cent. Living Aquat.
Resour. Manage., Manila, and Kuwait Inst. Sci. Res., Safat,
Kuwait.
Wilder, M.J.
1977 Biological aspects and fisheries potential of two deep-
water shrimps Heterocarpus ensifer and Heterocarpus laeviga-
tus in waters surrounding Guam. M.S. thesis, Univ. Guam,
Agana, 79 p.
Abstract . - The diets of pelagic
juveniles of widow rockfish Sebastes
entom.elas, yellowtail rockfish S.. /la -
iridus, chilipepper S. goodei, short-
belly rockfish S. jordani, and bocac-
cio S. paucispinis were compared
using samples collected during 1984-
87. All five species co-occur as pelag-
ic juveniles off central California.
Frequency of occurrence, percent by
number, and a ranking index of prey
items were determined from 1088
stomachs. Major prey of pelagic juve-
nile rockfish were the various life
stages of calanoid copepods and sub-
adult euphausiids (including eggs).
For each year, dietary overlap was
quantified between interspecific
pairs using the Colwell and Futuyma
(1971) index. Amount of overlap
varied from year to year. Long-term
intraspecific dietary overlap, based
on the 4 years of data, was general-
ly less than interspecific overlap
within years. Year-to-year variation
in the diets of these species was gen-
erally greater than within-year varia-
tion among them, suggesting that, as
a group, pelagic juvenile rockfishes
are opportunistic feeders. Also, if in-
terannual variation in the distribu-
tion and abundance of foods has a
major impact on recruitment, the
high dietary overlaps of these co-oc-
curring species would suggest paral-
lel survival and year-class success.
Multivariate analysis of variance
was used to examine the effects of
latitude, depth, and fish size on food
consumption. Alterations in diet
were related to latitude, depth, and
a latitude-depth interaction for three
species in 1987 and, also, for short-
belly rockfish in 1984-86. Diet was
apparently unrelated to fish size.
Interannual variation and overlap
in tlie diets of pelagic juvenile
rockfish (Genus: Sebastes)
off central California
Carol A. Reilly
Tiburon Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA, 3 1 50 Paradise Drive. Tiburon, California 94920
Tina Wyllie Echeverria
Tiburon Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA, 3 1 50 Paradise Drive, Tiburon, California 94920
Present address: School of Fisheries and Ocean Sciences
University of Alaska, Fairbanks, Alaska 99775-1080
Stephen Ralston
Tiburon Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA, 3 1 50 Paradise Drive, Tiburon. California 94920
Manuscript accepted 20 May 1992.
Fishery Bulletin, U.S. 90:505-515 (1992).
Rockfishes of the genus Sebastes are
a major component of the west coast
groundfish fishery (Gunderson and
Sample 1980), yet little is known of
their early life history. Kendall and
Lenarz (1987) noted a particular lack
of information on the biology of the
pelagic juvenile life-stage. To date
most work on pelagic juveniles has
addressed problems in identification
(e.g., Moser et al. 1977, Laroche and
Richardson 1980 and 1981, Matarese
et al. 1989), growth (Boehlert 1981a,
Boehlert and Yoklavich 1983, Penney
and Evans 1985, Laidig et al. 1991,
Woodbury and Ralston 1991), and
vertical distribution (Boehlert 1977
and 1981b, Moser and Alhstrom
1978, Moser and Boehlert 1991,
Lenarz et al. 1991).
Female rockfishes undergo inter-
nal fertilization and the eggs develop
within the ovary for a 40-50 day pe-
riod (Kendall and Lenarz 1987). Lar-
vae hatch internally, are extruded
approximately 1 week later, and be-
gin feeding. Larvae grow and trans-
form into juveniles, a developmental
stage characterized by the attain-
ment of full meristic characters.
Many rockfishes have a pelagic juve-
nile stage. Pelagic juveniles ranging
in size from 15-100 mm SL are abun-
dant off central California from April
to June, although distributional pat-
terns vary markedly among species
and years (Wyllie Echeverria et al.
1990). The pelagic juvenile stage ends
with settlement into demersal or
nearshore habitats.
Evidence strongly indicates that
the recruitment of marine fishes is
heavily influenced by events that oc-
cur early in the life history (Blaxter
1974). A frequently proposed expla-
nation is that the availability and
abundance of foods appropriate for
first-feeding larval and later juvenile
stages are critical to adequate sur-
vival and growth. A reduction in the
fine-scale density of suitable prey
items, whether due to an absolute
decrease in prey abundance (Hjort
1914) or to a randomized dispersion
of what formerly was a patchy prey
resource (Lasker 1975), can have a
negative impact on survival. Reduced
prey densities can affect survivorship
directly through starvation, or in-
directly by reducing growth rates
and thereby prolonging exposure to
other size-specific mortality factors
505
506
Fishery Bulletin 90(3). 1992
(e.g., predation, advection, etc.).
Regardless of the mechanism,
variation in the availability of
food can have a major effect on
year-class strength (Lasker
1981). The study of food utiliza-
tion patterns and diet overlap is,
therefore, useful in understand-
ing survival mechanisms during
the pelagic juvenile life stage.
Moreover, annual variation in
the extent of interspecific dietary
overlap may indicate changes in
the distribution and abundance
of prey (e.g., Zaret and Rand
1971). Since this may well be
critical in determining the suc-
cess of a year-class, similarity in
food habits among pelagic juve-
nOe rockfish may result in similar
recruitment dynamics.
Previous published dietary
studies of juvenile rockfish have
been limited to (1) experimental
work on food ration and growth
in black rockfish (S. melanops,
Boehlert and Yoklavich 1983); (2)
a description of the diet of new-
ly settled Pacific ocean perch
(S. alutus, Carlson and Haight
1976); (3) a comparison of the
food habits of seven Sebastes spp. in a nearshore kelp-
forest habitat (Singer 1985); and (4) predation on bar-
nacle larvae by a mixed assemblage of settled kelp resi-
dent juvenile rockfishes (Gaines and Roughgarden
1987). The purpose of this study was to examine the
feeding ecology of several co-occurring young-of-the-
year pelagic juvenile rockfishes, including widow rock-
fish S. entomelas, yellowtail rockfish S.Jlavidus, chili-
pepper iS. goodei, shortbelly rockfish S. jordani, and
bocaccio S. paucispinis. Specific goals of this study
were to (1) identify the food habits of these five species
during the pelagic juvenile stage, (2) determine the ex-
tent of dietary overlap among the five species, and (3)
determine the degree of interannual variation in pat-
terns of prey utilization.
Materials and methods
Juvenile rockfish used in this study were obtained from
midwater trawl samples made during a series of an-
nual pelagic juvenile rockfish surveys conducted off
central California during 1984-87. Details of these
surveys are described in Wyllie Echeverria et al. (1990).
\^ STATIONS FOR
i^Jeoner RQCKFISH STOMACH
\ SAMPLES. 1984-87
• •• • • •|\ V/i/\
38-
• • /A-?' Reyes! J^
« \BoJinas? (^
• ^T \
•"° ^ San Fcanc&cqV^^
• • •••• • , 1 v., \
SHalfmoorrv
37=20 \ ^'•''
• • • • •
( Pescadero
37-
— \ Davenport _
• • • • • • X^^.^..-..
• \.
•• •.•/
#S Monterey
1 Pt. Sur
1 1 1 \ \
125*W 124' 123' 122"
Figure 1
Map of the central California coast showing locations of midwater trawl stations where
Sebastes stomach samples were obtained, 1984-87.
The primary purpose of the surveys was to estimate
the distribution and abundance of the pelagic-stage
juveniles of age-0 rockfishes. Survey areas and dates
differed somewhat from year to year (Fig. 1, Table 1).
The surveys were conducted during June, except in
1987 when the survey extended from late-May to June.
In 1984 and 1985, the survey area extended from Point
Sur Oat. 36°18'N) to Point Cabrillo Oat. 39°20'N). Bot-
tom depths at each trawl station ranged from <50m
at nearshore localities to > 3700 m beyond the continen-
tal shelf. The sampling plan was revised in 1986; seven
transects composed of 36 stations were selected based
on previous records of rockfish abundance and the
availability of ship time. These stations were sampled
repetitively during three consecutive sweeps of the
area. After 1985, the survey area extended from
Cypress Point Oat. 36°35'N) to Point Reyes Oat. 38°
OO'N), with station depths ranging from <50 to 1000m.
Collections were made from the RV David Starr Jor-
dan with a modified Cobb midwater trawl net having
a 24.4m head rope and 0.76cm mesh liner in the cod-
end. The standard depth sampled was 30 m. However,
at shallow stations Osottom depth < 100 m) the net was
set at 5- 10 m. At some deep stations samples were
Reilly et al : Diets of pelagic juvenile Sebastes off central California
507
Table 1
Number of juvenile Sebastes stomachs examined from
juvenile rockfish
surveys.
1984-87.
No. of
Range
Year Survey dates
Species
stomachs
(mm SL)
1984 8-24 June
Widow rockfish
15
40-63
Yellowtail rockfish
40
36-56
Chilipepper
20
38-55
Shortbelly rockfish
120
30-65
Bocaccio
50
21-77
1985 5-30 June
Widow rockfish
75
43-63
Yellowtail rockfish
30
39-48
Shortbelly rockfish
85
49-75
1986 3-25 June
Yellowtail rockfish
10
35-47
Shortbelly rockfish
168
15-47
Bocaccio
25
18-40
1987 23 May-21 June
Widow rockfish
105
48-80
Yellowtail rockfish
17
39-52
Chilipepper
125
41-76
Shortbelly rockfish
150
17-78
Bocaccio
53
22-86
Total stomachs examined
1088
also collected at 100 m. Nets were fished for 15 min at
depth during the night, ~30 min after simset, or before
sunrise.
Five specimens of each species were randomly sub-
sampled from each haul for dietary analysis. General-
ly, no samples were taken if fewer than five individuals
were taken in a haul. Specimens were tentatively iden-
tified to species and preserved whole in 10% buffered
formalin, usually within 1 hour of collection. Identifica-
tions were later verified ashore with meristics keys
(Matarese et al. 1989, Moreland and Reilly 1991);
samples were transferred to 70% isopropyl alcohol
within 1 month of collection. Standard length (SL) was
later measured to the nearest 0.1 mm. Stomachs were
removed and stored in 70% isopropyl alcohol until
examined.
Stomach contents were examined with a dissecting
microscope. Empty stomachs were noted and the
digestive state of each prey item was coded on a scale
of 1-3, with 3 representing digestion too advanced for
identification. All prey types were identified to the
lowest possible taxonomic level and counted. When
possible, a subsample of all prey types was measured
along the longest axis with an ocular micrometer.
Heads or eyes were used to obtain total counts when
food items were fragmented. For each rockfish species,
the proportion of prey types in the diet was calculated
as the percentage of total prey numbers consumed in
a year, summed over all the individuals examined for
stomach contents.
A ranking index, modified from
Hobson (1974), was calculated for the
major food items. The index (Ir) is the
product of proportional frequency of
occurrence and percent by number,
calculated for all specimens of a spe-
cies in a year. To quantify dietary
overlap among species, the index of
Colwell and Futuyma (1971) was used,
that is,
■'ih
1.0 - 0.5
Pij - Phj
j = i
where pij and Phi are the numerical
proportions of prey j = 1 . . . N found in
the diets of species i and h, respective-
ly. The index has a minimum value of
zero, when no overlap occurs, and a
maximum value of one, when all prey
are shared in equal proportions by the
two species.
Multivariate analysis of variance
(MANOVA) was used to examine relationships among
latitude, bottom depth, and the diets of chilipepper,
shortbelly, and widow rockfish (Green 1978, SAS 1985).
Although only in 1987 were there sufficient data to
analyze the diets of all three species, adequate samples
of shortbelly rockfish were obtained during all years
(1984-87). Thus, examination of overall variation in
diet through time, vis-a-vis latitude and depth, was
limited to shortbelly rockfish. Analyses were confined
to the three prey types of highest frequency of occur-
rence during the year examined, which varied among
the different species and years. The numerical propor-
tions of the three prey types (the dependent variables)
were arcsine-transformed (Sokal and Rohlf 1981) prior
to MANOVA testing. Latitude, depth, and a latitude-
depth interaction term were the independent variables.
Station latitude was classified as either north or south
of lat. 37°20'N. Similarly, station depth was divided
into deep (>100m) or shallow (<100m) categories.
Data for chilipepper, shortbelly, and widow rockfish
sampled in 1987 were also divided into large (>1987
median SL) and small (<1987 median SL) size-classes
to examine diet variation as a function of fish size.
Shortbelly rockfish were sufficiently numerous during
all years to analyze diet variation as a function of
predator size. Prey types for this analysis were again
limited to the three prey categories with the highest
overall frequencies of occurrence in a year, and the
dependent variables were the arcsine-transformed
numerical proportions in the diet.
508
Fishery Bulletin 90(3). 1992
Table 2
Summarj' of stomach contents for five species of pelagic juvenile Sebastes,
1984. FO
= frequency of occurrence; % = percent by number.
Widow rockfish
Yellowtail rockfish
Chilipepper
Shortbelly rockfish
Bocaccio
Prey category
(n
15)
in
40)
(n
20)
(n 120)
(n
50)
FO
%
FO
%
FO
%
FO
%
FO
%
EUPHAUSIACEA
Furcilia
33.3
10.1
50.0
12.3
30.0
6.8
54.2
35.1
36.0
24.1
Calyptopis
2.5
0.3
1.0
0.1
2.0
0.2
Juveniles
20.0
3.2
15.0
3.4
15.0
2.6
14.1
10.2
38.0
25.3
AMPHIPODA
Hyperiid juveniles
1.0
0.1
4.0
0.3
CUMACEA
1.7
0.1
DECAPODA
Natantia juveniles
6.7
0.5
1.0
0.2
COPEPODA
Calanus spp.
46.7
41.0
52.5
28.7
40.0
27.2
40.8
15.1
34.0
31.6
Candacia sp.
5.0
0.4
6.7
0.5
2.0
0.2
Copepods (unidentified)
26.7
17.6
47.5
14.7
15.0
10.5
36.7
12.0
30.0
15.8
Juveniles
20.0
27.7
17.5
40.2
40.0
52.9
10.8
26.8
2.0
0.9
OSTEICHTHYES
Fish larvae (unidentified)
20.0
1.7
Results
Frequency of occurrence
and percent number
Stomachs from 1088 pelagic juvenile
rockfish collected from midwater
trawls during the four survey years
(Table 1) were examined. Frequency of
occurrence and percent number for
specific prey types of each rockfish
species varied considerably from year
to year (Tables 2-5). In 1984, bocac-
cio differed from all other rockfish
species in the frequency of occurrence
of fish larvae as a prey type (Table 2).
Euphausiid eggs occurred in the
stomachs of all three species in 1985,
although there is a disparity in the per-
cent number (Table 3). Euphausiid
eggs were much less frequent in the
diets of the three rockfish species in
1986 (Table 4), whereas juvenile eu-
phausiids occurred more frequently.
Overall, data from 1984-87 show that
prey items having a high frequency of
occurrence generally had a high per-
centage by number. Euphausiid eggs
and juveniles and unidentified cope-
pods often had high percentages by
number relative to their frequencies of
Table 3
Summary of stomach contents for three
species
of pelagic juvenile
Sebastes
, 1985.
FO = frequency of occurrence;
% = percent by
number.
Widow
Yellowtail
Shortbelly
rockfish
rockfish
rockfish
Prey category
(n
75)
(n
30)
(n
85)
FO
%
FO
%
FO
%
EUPHAUSIACEA
Furcilia
22.7
1.7
30.0
1.2
14.1
0.4
Calyptopis
2.7
0.1
6.7
1.5
1.2
0.1
Juveniles
5.3
0.7
10.0
0.2
16.5
0.4
Euphausiid eggs
18.7
8.2
36.7
59.1
48.2
59.3
AMPHIPODA
Hyperiid juveniles
6.7
0.1
3.3
0.1
7.1
0.1
LARVACEA
6.7
4.1
7.1
2.4
CHAETOGNATHA
1.3
0.1
DECAPODA
Natantia juveniles
3.3
0.1
1.2
O.I
COPEPODA
Calanus spp.
22.7
1.7
16.7
1.2
25.9
1.2
Candacia sp.
2.7
0.1
3.3
0.1
1.2
0.1
Copepods (unidentified)
13.3
3.7
17.6
2.9
Eucalanus sp.
8.0
1.8
3.5
0.1
Euchirella sp.
2.7
0.2
3.5
0.1
Juveniles
44.0
77.6
56.7
36.5
57.6
33.3
Metridia sp.
3.3
0.1
1.2
0.1
OSTEICHTHYES
Fish larvae (unidentified)
1.3
0.1
Reilly et al : Diets of pelagic juvenile Sebastes off central California
509
Table 4
Summary of stomach contents for three species
of pelagic juvenile Sebastes
1986.
FO = frequency of occurrence
; % = percent by
number
Yellowtail
Shortbelly
rockfish
rockfish
Bocaccio
Prey category
(nlO)
(n
168)
(m25)
FO %
FO
%
FO
%
EUPHAUSIACEA
Furcilia
10.0 1.6
13.2
0.9
4.0
0.2
Calyptopis
10.0 0.6
3.6
0.1
Juveniles
70.0 7.1
59.9
10.0
52.0
22.0
Euphausiid eggs
4.8
1.1
CUMACEA
10.0 1.0
1.2
0.1
DECAPODA
Natantia juveniles
20.0 1.0
0.6
0.1
COPEPODA
Calanus spp.
38.3
6.6
64.0
30.5
Copepods (unidentified)
34.1
3.9
28.0
5.9
Epilabidocera sp.
40.0 8.7
Juveniles
50.0 80.1
71.9
77.3
44.0
41.4
occurrence. Euphausiid eggs and juve-
nile copepods were the smallest signifi-
cant prey of pelagic juvenile rockfish.
It is therefore not surprising that these
categories often display high percent-
ages by number. Likewise, the cate-
gory 'unidentified copepods' typically
was based on counts of small items
(e.g., head fragments).
Ranking index
Prey having both a high frequency of
occurrence and percentage by number
are the most important items in the
diet (Tables 2-5). Calanus spp. cope-
pods were particularly important in
1984 when all rockfish species con-
sumed substantial numbers of this
Table 5
Summary of stomach contents for five species of pelagic juvenile Sebastes,
1987. F0 =
= frequency of occurrence; % = percent by number.
Widow rockfish
Yellowtail rockfish
Chilipepper
Shortbelly rockfish
Bocaccio
Prey category
(n
105)
n\l)
(n 125)
(w 150)
(«
53)
FO
%
FO
%
FO
%
FO
%
FO
%
EUPHAUSIACEA
Furcilia
6.6
0.20
11.8
1.1
8.0
1.30
40.7
4.00
1.9
0.1
Calyptopis
8.5
0.30
5.9
0.4
2.4
0.40
6.7
1.60
1.9
0.1
Juveniles
39.6
9.00
76.5
31.1
19.8
10.30
6.7
0.60
30.2
10.7
Euphausiid eggs
27.4
52.00
5.9
7.7
10.4
35.80
28.7
65.80
1.9
10.5
Euphausiids (unidentified)
8.5
0.30
5.9
0.2
8.0
0.60
8.0
0.20
Euphausia sp.
3.8
0.2
Thysanoessa sp.
1.0
0.01
AMPHIPODA
Hyperiids
1.0
0.01
1.6
0.10
1.0
0.01
1.9
0.1
CUMACEA
1.0
0.04
DECAPODA
Natantia juveniles
1.0
0.04
1.0
0.04
Crab megalopa
3.2
0.20
Crab zoea
1.6
0.10
1.0
0.01
CIRRIPEDIA
Cypris larva
1.0
0.10
COPEPODA
Calanus spp.
33.0
19.80
47.1
13.9
28.0
22.80
44.0
20.30
32.1
11.5
Candacia sp.
1.0
0.10
1.9
0.2
Copepods (unidentified)
22.6
7.30
5.9
1.1
8.0
4.10
4.0
1.00
13.2
45.8
Eucalanits sp.
1.9
0.10
1.9
0.1
Euchaeta sp.
1.9
0.8
Rhincalanus sp.
1.0
0.01
1.0
0.01
Juveniles
24.5
10.00
17.7
44.5
14.4
24.30
16.0
6.30
5.7
19.3
TEUTHOIDEA
Squid larva
1.9
0.1
OSTEICHTHYES
Fish larvae
15.1
0.6
Eggs
3.8
1.00
2.0
0.40
510
Fishery Bulletin 90(3), 1992
Q
Z
o
z
2
46
40
2
Q
Z
o
z
z
SHORTBELLY
ROCKFISH PREY
M2
X
X
X
X
X
X
X
X
X
X
X
X
X
X
BOCACCIO PREY
CHILIPEPPER PREY
IM.
fc,r^
^ CALANUS SPP.
g^ COPEPOD JUV.
T77\ UNID. COPEPODS
^ EUPH. EGGS
{^J EUPH. JUV.
Figure 2
Histograms of the prey ranking index across years for five species of pelagic juvenile Sebastes. Calanus spp.
= CalaniLs spp. copopods; copepod juv. = copopod juveniles; unid. copepods = unidentified copepods; euph.
eggs = euphausiid eggs; euph. juv. = euphausiid juveniles.
prey. Likewise, 1985 was a year in which copepod
juveniles and euphausiid eggs dominated the diets of
the three species examined (widow, yellowtail, and
shortbeily rockfish) and copepod juveniles were again
important to all species in 1986.
Interspecific prey utilization patterns were less ob-
vious. There is some indication that widow and chili-
pepper rockfish consumed more copepod juveniles and
Calanus spp. copepods than did the other species.
Similarly, shortbeily rockfish appeared to consume
more euphausiid eggs, while bocaccio consumed more
euphausiid juveniles. Likewise, there was some sugges-
tion that bocaccio fed on larger prey than the other
species (e.g., euphausiid adults and fish larvae).
Nonetheless, no distinctive separation in primary prey
species was evident among the five species examined.
Based on the Ir ranking index, it is apparent that
the diet of pelagic juveniles is typically dominated by
Reilly et al : Diets of pelagic juvenile Sebastes off central California
51 1
a single prey type each year, followed by several prey
types with indices at much lower values (Fig. 2). There
were few instances in which the two most important
prey types were similar in ranking index, e.g., yellow-
tail rockfish consuming euphausiid eggs (21.7) and
juvenile copepods (20.7) in 1985. This result suggests
that each year these species, to a large extent, spe-
cialize on foods that are intermittently abundant. Also,
the ranking index data, together with information on
frequency of occurrence and percent by number, in-
dicate the major prey items of pelagic juvenile rockfish
were various life stages of copepods and subadult
euphausiids.
Dietary overlap
The extent of interspecific dietary similarity was quan-
tified by comparing dietary overlaps among ali possible
pairs of species within each year (1984-87). Ten species
pairs were possible, but not all pairs were observed
each year since all five species were not always col-
lected (Table 6). Overlap indices are sensitive to the tax-
onomic level to which prey items are categorized; thus,
statistical tests of significance concerning the data are
arbitrary. Therefore, the convention established by
Langton (1982) and Brodeur and Pearcy (1984) was in-
voked. Overlap index values of 0.00-0.29 were con-
sidered low, values of 0.30-0.60 were considered
medium, and values >0.60 were considered high.
Using these criteria, annual comparisons of the dis-
tribution of overlap indices for 1984 indicate that 60%
of all comparisons were classified as medium and 40%
were classified as high. Results from 1985 and 1986
indicate that 67% of the scores were medium and 33%
were high. In contrast, index values during 1987
generally had the lowest amount of overlap: 10% low,
70% medium, and 20% high. Based on these findings
we conclude that, although overall patterns of dietary
overlap do vary from one year to the next, variations
are relatively modest (only in 1987 was any low overlap
observed). Moreover, in this study most within-year
species pairings showed >30% overlap. The principal
exception to this generalization was for yellowtail and
shortbelly rockfish sampled in 1987. Their diets were
quite dissimilar.
Overlap indices were also calculated for all possible
interannual intraspecific combinations. These calcula-
tions allow an assessment of the temporal stability of
the diet relative to the amount of interspecific dietary
overlap displayed during a given year. The frequency
distribution of dietary overlap values derived from self-
pairing of rockfish species from different years is
shifted well to the left (toward zero) of the distribution
of interspecific scores obtained within a year (Fig. 3).
These findings show that in any particular year the dif-
Table 6
Diet overlap indices for individual pairings
Df pelagic juvenile |
Sebastes (1984-87). Wid =
widow rockfish
;Yel =
yellowtail
rockfish; Chi =
= chilipepper
Sho = shortbelly rockfish; Boc |
= bocaccio.
Species pair
Year
1984
1985
1986
1987
Wid-Yel
0.84
0.53
—
0.39
Wid-Chi
0.58
—
—
0.69
Wid-Sho
0.52
0.46
—
0.77
Wid-Boc
0.57
—
—
0.50
Yel-Chi
0.61
—
—
0.52
Yel-Sho
0.58
0.90
0.91
0.22
Yel-Boc
0.64
—
0.46
0.48
Chi-Sho
0.38
—
—
0.56
Chi-Boc
0.41
—
—
0.49
Sho-Boc
0.61
—
0.51
0.30
20
15-
o
c
0)
NWN Interspecific (within year)
^H Interannual (within species)
10- I
- I al ll
11
0.00 0.20 0.40 0.60 0.80
Dietary Overlap
1.00
Figure 3
Frequency of dietary overlap indices among all interspecific
Sebastes pairs within years, compared with frequency of
overlap indices calculated for each Sebastes species self-paired
across years.
ferent species of rockfish are opportunistic feeders that
utilize relatively similar prey items, but substantial
dietary change can occur from year to year.
Latitude and depth effects
Dietary variation with respect to station latitude (north
or south of lat. 37°20'N), station depth (deeper or
shallower than 100 m), and the interaction of these
variables were analyzed using MANOVA (Table 7). The
statistical significance of each analysis depended on the
particular combination of species and year examined.
In 1987, highly significant (P< 0.001) diet variations oc-
curred with depth for shortbelly and widow rockfish.
512
Fishery Bulletin 90(3). 1992
Table 7
Results of MANOVA of depth,
latitude, and depth by latitude effects on three principal prey types of pelagic juvenile Sebastes. Cal 1
= Calaniis spp
; CoJv =
copepod juveniles:
EJv =
euphausiid juveniles; EuEg =
= euphausiid
eggs; ELv =
euphausiid larvae; UnCo |
= unidentified
copepods
Fur
= furcilia.
Species
Year
Prey type
MANOVA model effects
Depth
Latitude
Depth"
Wilks' A
latitude
P
I
II
III
Wilks' X
P
Wilks' X
P
Chilipepper
87
Cal
CoJv
EJv
0.9019
0.0580''
0.9186
0.1044
0.9715
0.5525
Widow
87
Cal
EuEg
EJv
0.6638
0.0001**
0.9536
0.2813
0.9805
0.6632
Shortbelly
87
Cal
EuEg
ELv
0.7649
0.0001**
0.9805
0.5059
0.9243
0.0252*
Shortbelly
86
Cal
CoJv
EJv
0.8917
0.0005**
0.7522
0.0001**
0.9404
0.0234*
Shortbelly
85
Cal
EuEg
CoJv
0.8934
0.0472*
0.9293
0.1600
0.9185
0.1121
Shortbelly
84
els:
Cal
UnCo
Fur
0.9647
0.3110
0.9213
0.0429*
0.9095
0.0240*
Significance lev
'' borderline
*P<0.05
*»P<0.01
In that year, latitude had no discernible influence on
the diet of these two species, although for shortbelly
rockfish a significant interaction between depth and
latitude was evident. These findings strongly suggest
that spatial variability in the environment (i.e., latitude
and depth of the water column) can influence, to some
extent, the diets of pelagic juvenile rockfish in a species-
specific manner.
Results for the full time-series of shortbelly rockfish
data (1984-87) also show that spatial patterns change
over time. Although depth had a highly significant ef-
fect on diet in 1986 and 1987, it was not significant in
1984 or 1985. Similarly, latitude had no appreciable
relationship to the diet of shortbelly rockfish in 1985
and 1987, but it had a highly significant effect in 1986.
Importantly, whenever a latitude correlation with diet
was present, the interaction term (depth * latitude) was
significant as well. We believe that the erratic influence
of spatial structure on the shortbelly diet is likely due
to the dynamic nature of the nearshore pelagic/neritic
physical environment.
Using the 1987 data, we examined the least-squares
means (Searle et al. 1980) of the transformed numerical
proportions of the individual prey types to learn exactly
how dietary composition varied when statistically-
significant model effects occurred. In that year, the
diet of chilipepper showed borderline significance with
depth (P = 0.058); the least-squares means revealed that
chilipepper consumed more Calanus spp. copepods in
shallow water, and more juvenile copepods and juvenile
euphausiids at bottom depths >100m. Likewise, all
three prey types {Calanus spp., euphausiid eggs, and
juvenile euphausiids) of widow rockfish were consumed
in greater proportion in deep water, especially euphau-
siid eggs. For shortbelly rockfish, which displayed a
significant interaction term, consumption of Calanus
spp. copepods was noticeably depressed at shallow
southern stations. Euphausiid eggs were consumed in
much greater quantities at deep stations, both north
and south, while fewer larval euphausiids (furcilia and
calyptopis) were found in fish from northern deep
stations.
Predator size
Results were inconsistent when these same data (i.e.,
numerical proportions in the diet of the three most fre-
quently occurring prey items for 1984-87 shortbelly
rockfish, 1987 chilipepper, and 1987 widow rockfish)
were also explored with MANOVA to assess the effect
of fish size on composition of the diet. In each instance,
fish were assigned to either small or large size-classes,
based on whether standard lengths were smaller or
larger than the annual median of that species.
Of the six cases examined (Table 8) two yielded
significant (P<0.05) results. Large shortbelly rockfish
sampled in 1986 tended to eat a higher proportion of
Calanus spp. copepods, whereas small fish had a higher
fraction of juvenile euphausiids and juvenile copepods
in their diet. Results from that year, therefore, sup-
port the view that large fish tend to consume large
prey. Even so, a significant size effect was demon-
strated for 1985 shortbelly rockfish, which was exact-
ly the opposite of 1986; large fish consumed fewer
Calanus spp. copepods and a greater percentage of
euphausiid eggs than did small fish. Sample size was
not adequate to statistically analyze fish length jointly
with distributional patterns. However, in 1985, 34 of
Reilly et al : Diets of pelagic juvenile Sebsstes off central California
513
Table 8
Results of MANOVA of fish size on three principal prey
types of pelagic juvenile 1
Sebastes. Cal =
Cala
nus spp.;
CoJv = copepod juveniles; EJv =
euphausiid
juveniles; EuEg
= euphausiid eggs; ELv = euphausiid larvae; UnCo
= uniden-
tified copepods;
Fur =
furcilia.
Species
Year
Prey type
Fish
size
I
II
III
Wilks' X
P
Chilipepper
87
Cal
CoJv
EJv
0.9927
0.9088
Widow
87
Cal
EuEg
EJv
0.9678
0.4407
Shortbelly
87
Cal
EuEg
ELv
0.9634
0.2132
Shortbelly
86
Cal
CoJv
EJv
0.9362
0.0161'
Shortbelly
85
Cal
EuEg
CoJv
0.8801
0.0261*
Shortbelly
84
Cal
UnCo
Fur
0.9928
0.8654
•P<0.05
the 42 shortbelly rockfish that were classified as small
came from deep stations. Results presented earlier
(Table 7) showed that the diet of shortbelly rockfish
varied significantly with depth in 1985 (i.e., fewer
euphausiid eggs and copepod juveniles at deep sta-
tions). Thus, the conclusion that small fish consumed
large prey in 1985 is, to some degree, confounded with
this spatial effect.
Discussion
The five species of pelagic juvenile rockfish examined
in this study consumed pelagic zooplankton almost
exclusively. Relatively few prey types made up the
major portion of the diet each year. Various life history
stages of calanoid copepods and euphausiids dominated.
Carlson and Haight (1976) reported that copepods and
euphausiids were important in the diet of pelagic juve-
nile Pacific ocean perch S. alutics. Singer (1985) recent-
ly reported that settled juveniles of several rockfish
species consumed copepods and zoea larvae in a cen-
tral California kelp forest. Other studies (Robb and
Hislop 1980, Bowman 1981, Conway 1980) have also
demonstrated that calanoid copepods and euphausiids
are extremely important foods to pelagic juvenile fishes
in the northeastern Pacific Ocean. These studies
demonstrate that the diets of pelagic juvenile rockfishes
are similar to those of other species possessing pelagic
juvenile life stages.
A significant finding of this study is that Sebastes
spp. juveniles periodically forage heavily on euphau-
siid eggs. Euphausiid eggs have not been previously
reported as a prey item of pelagic juvenile rockfish and
yet they were a very important dietary component both
in 1985 and 1987. During those years, euphausiid eggs
averaged over 37% of the prey items
consumed by the five species studied.
However, euphausiid eggs were ab-
sent from samples collected in 1984
and were a minor component in 1986.
Some species of euphausiids brood
their eggs prior to hatching (e.g., Nyc-
tiphanes spp.), whereas other species
release eggs upon fertilization (e.g.,
Euphausia pacifica and Thysanoessa
spinifera). Since adult euphausiids
were not found in any stomachs in
1985 (Table 3), and since only the lat-
ter genera were encountered in large
swarms in the study area in 1987
(Smith and Adams 1988), rockfish
must have consumed eggs after re-
lease. It was not expected that a non-
motile prey would constitute such an
important food resource to pelagic juvenile rockfish.
The appearance of eggs in clumped masses in guts sug-
gests that eggs were not individually picked from the
plankton.
Another interesting finding was the consumption of
fish larvae by bocaccio juveniles. A total 15-20% of all
bocaccio sampled in 1984 and 1987 contained larval
fish. In our surveys, bocaccio grow faster and reach
larger sizes as pelagic juveniles ( > 1 00 mm SL) than do
other species (Woodbury and Ralston 1991). They are
also distributed at shallower depths (Lenarz et al.
1991).
We used the I^ statistic to rank the importance of in-
dividual prey items in the diet. This statistic differs
from a similar statistic used by Hobson (1974) in that
it is the product of proportional frequency of occur-
rence and percent by number, rather than percent by
volume. Use of this statistic allowed us to characterize
the prey types consumed by Sebastes in each of the 4
years studied. No obvious species-specific patterns
emerged in the absence of a temporal component.
Our results indicate that pelagic juvenile Sebastes
tend to respond similarly to environmental fluctuations
in their food base, suggesting an opportunistic feeding
strategy. Intraspecific dietary overlap between interan-
nual pairings was much lower than were interspecific
interannual pairings. On a relative basis, interannual
differences in diet were tracked similarly among the
five species we examined. Annual changes in diet are
likely to reflect annual differences in the composition,
availability, and abundance of prey.
It was not possible to infer from our results whether
or not food is limiting to pelagic juvenile rockfishes,
given the relatively large interannual variation in the
diet among these species and the likelihood that varia-
tion in the availability of prey is likely responsible.
514
Fishery Bulletin 90(3). 1992
Even so, high dietary overlap observed among co-
occurring pelagic juvenile rockfishes suggests that
similar recruitment dynamics must exist if the distribu-
tion and abundance of foods has a major impact on
recruitment.
Intraspecific spatial variation was observed (Table
7), even though substantial interspecific overlap exists
in patterns of food utilization. In some instances,
parallel spatial differences were observed for different
species. For example, in 1987 both widow and short-
belly rockfish fed on euphausiid eggs to a much greater
extent in deep water (>100m) than in shallow water.
In other cases, however, species-specific differences in
diet due to depth were reversed. In 1987, for exam-
ple, the consumption of Calanus spp. copepods by
chilipepper was higher in shallow water, while con-
sumption by widow rockfish was higher in deep water.
With the exception of the predator-size MANOVA
discussed previously (i.e., shortbelly rockfish in 1985),
sample sizes for each treatment combination in all
MANOVA tests were reasonably well balanced. There-
fore, it is unlikely that our conclusions were compro-
mised by our choice of statistical tests.
The spatial incongruity of within-year dietary pat-
terns among species also extended to interannual
within-species comparisons. For example, shortbelly
rockfish sampled in 1984 and 1987 consumed substan-
tially fewer Calanus spp. copepods in the shallow
southern quadrant than anywhere else. However, in
1986 consumption of this prey was greatest in fish
taken in this region.
These interspecific (within-year) and interannual
(within-species) comparisons demonstrate a lack of
stability in the specifics of how spatial dietary effects
are expressed. It is likely that the complex nearshore
pattern of circulation that characterizes the study area
(frontal structures, mesoscale eddies, turbulent jets,
and upwelling plumes are common recurrent features;
Mooers and Robinson 1984, Flament et al. 1985, Njoku
et al. 1985, Schwing et al. 1990) defines the spatial
distribution of the zooplanktonic animals upon which
these rockfish feed. Thus, the dynamic nature of the
physical environment off central California generates
spatial instabilities in the distribution and abundance
of prey.
Acknowledgments
We are grateful to all personnel of the Tiburon Lab-
oratory, Southwest Fisheries Science Center, National
Marine Fisheries Service, particularly the staff of the
Groundfish Analysis Investigation, who participated in
the juvenile rockfish recruitment surveys in 1984-87.
We thank Master Milt Roll and the officers and crew
of the RV David Starr Jordan for their valuable help
in collecting juvenile rockfish during these research
cruises. Sharon Moreland (currently of the Army Corps
of Engineers) taught us identification techniques for
young-of-the-year pelagic juvenile rockfish. Tony Chess
(Tiburon Laboratory) assisted with identification of
copepods, and Margaret Knight (Scripps Institute of
Oceanography) confirmed identification of the euphau-
siid eggs. Jim Bence (Tiburon Laboratory) generously
assisted us with statistical analyses. This paper was
greatly improved by thoughtful, constructive reviews
by Pete Adams, Jim Bence, George Boehlert, Ted Hob-
son, Ralph Larson, Bill Lenarz, Jeannette Whipple, and
three anonymous reviewers.
Citations
Blaxter, J.H.S. (editor)
1974 The early life history of fish. Springer-Verlag, NY,
765 p.
Boehlert. G.W.
1977 Timing of the surface-to-benthic migration in juvenile
rockfish, Sebastes diploproa, off southern California. Fish.
Bull., U.S. 75:887-890.
1981a The effects of photoperiod and temperature on labora-
tory growth of juvenile Sebastes diphproa and a comparison
with growth in the field. Fish. Bull., U.S. 79:789-794.
1981b The role of temperature and photoperiod in the onto-
genetic migration of prejuvenile Sebastes diploproa (Pisces:
Scorpaenidae). Calif. Fish. Game 67:164-175.
Boehlert, G.W., and M.M. Yoklavich
1983 Effects of temperature, ration, and fish size on growth
of juvenile black rockfish, Sebastes melanops. Environ. Biol.
Fish. 8:17-28.
Bowman, R.E.
1981 Food of 10 species of northwest Atlantic juvenile ground-
fish. Fish. Bull., U.S. 79:200-206.
Brodeur, R.D., and W.G. Pearcy
1984 Food habits and dietary overlap of some shelf rockfishes
(Genus Sebastes] from the northeastern Pacific Ocean. Fish.
Bull., U.S. 82:269-293.
Carlson. H.R., and R.E. Haight
1976 Juvenile life of Pacific Ocean perch, Sebastes alutus. in
coastal fiords of southeastern Alaska: Their environment,
growth, food habits, and schooling behavior. Trans. Am. Fish.
Soc. 105:191-201.
Colwell, R.K., and D.J. Futuyma
1971 On the measurement of niche breadth and overlap.
Ecology 52:567-576.
Conway. D.V.P.
1980 The food of larval blue whiting, Mieromesistius poutasscni
(Risso), in the Rockall area. J. Fish Biol. 16:709-723.
Flament, P., L. Armi, and L. Washburn
1985 The evolving structure of an upwelling filament. J.
Geophys. Res. 90(C6):11, 765-11, 778.
Gaines, S.D., and J. Roughgarden
1987 Fish in offshore kelp forests affect recruitment to inter-
tidal barnacle populations. Science (Wash. DC) 235:479-481.
Green. P.E.
1978 Analyzing multivariate data. Dryden Press, Hinsdale,
IL, 519 p.
Reilly et al,: Diets of pelagic juvenile Sebastes off central California
515
Gunderson, D.R., and T.M. Sample
1980 Distribution and abundance of rockfish off Washington,
Oregon, and California during 1977. Mar. Fish. Rev. 42(3-4):
2-16.
Hjort, J.
1914 Fluctuations in the great fisheries of northern Europe
viewed in the light of biological research. Rapp. P.-V. Reun.
Cons. Perm. Int. Explor. Mer 20:1-228.
Hobson. E.S.
1974 Feeding relationships of teleostean fishes on coral reefs
in Kona, Hawaii. Fish. Bull., U.S. 72:915-1031.
Kendall, A.W. Jr.. and W.H. Lenarz
1987 Status of early life history studies of northeast Pacific
rockfishes. In Proc. Int. Rockfish Symp., Oct. 1986, An-
chorage, Alaska, p. 99-128. Alaska Sea Grant Rep. 87-2, Univ
Alaska, Fairbanks.
Laidig, T.E., S. Ralston, and J.R. Bence
1991 Dynamics of growth in the early life history of shortbelly
rockfish, Sebastes jordani. Fish. Bull.. U.S. 89:611-621.
Langton, R.W.
1982 Diet overlap between Atlantic cod, Gadus morhua, silver
hake, Merluccius bilinearis. and fifteen other northwest Atlan-
tic finfish. Fish. Bull., U.S. 80:745-759.
Laroche, W.A., and S.L. Richardson
1980 Development and occurrence of larvae and juveniles of
the rockfishes Sebastes Jlaindiis and Sebastes melanops (Scor-
paenidae) off Oregon. Fish. Bull., U.S. 77:901-924.
1981 Development of larvae and juveniles of the rockfishes
Sebastes entomelas and S. za-c^ntrus (Family Scorpaenidae) and
occurrence off Oregon, with notes on head spines of S. mys-
tinus, S. Jlavidus, and S. melanops. Fish. Bull., U.S. 79:
231-258.
Lasker, R.
1975 Field criteria for survival of anchovy larvae: The rela-
tion between inshore chlorophyll maximum layers and suc-
cessful first feeding. Fish. Bull., U.S. 73:453-462.
1981 The role of a stable ocean in larval fish survival and subse-
quent recruitment. In Lasker, R. (ed.). Marine fish larvae —
morphology, ecology, and relation to fisheries, p. 80-87. Univ.
Wash. Press, Seattle.
Lenarz, W.H., R.J. Larson, and S. Ralston
1991 Depth distributions of late larvae and pelagic juveniles
of some fishes of the California current. Calif. Coop. Oceanic
Fish. Inst. Rep. 32:41-46.
Matarese, A.C., A.W. Kendall Jr., D.M. Blood, and B.M. Vinter
1989 Laboratory guide to early life history stages of northeast
Pacific fishes. NOAA Tech. Rep. NMFS 80, 652 p.
Mooers, C.N.K., and A.R. Robinson
1984 Turbulent jets and eddies in the California current and
inferred cross-shore transports. Science (Wash. DC) 223:
51-53.
Moreland, S.L., and C.A. Reilly
1991 Key to the juvenile rockfishes of central California. In
Laidig, T.E., and P.B. Adams (eds.). Methods used to identify
pelagic juvenile and larval rockfish (genus Sebastes) occurring
along the coast of central Califomia, p. 59-180. NOAA Tech.
Memo. NOAA-TM-NMFS-SWFC-166, NMFS Tiburon Lab.
Moser, H.G., and E.H. Ahlstrom
1978 Larvae and pelagic juveniles of blackgill rockfish, Sebastes
mAanostomus. taken in mid-water trawls off southern Califor-
nia and Baja Califomia. J. Fish. Res. Board Can. 35:981-996.
Moser, H.G., E.H. Ahlstrom, and E.M. Sandknop
1977 Guide to the identification of scorpionfish larvae (Fam-
ily Scorpaenidae) in the eastern Pacific with comparative notes
on species of Sebastes and Helioeotenus from other oceans.
NOAA Tech. Rep. NMFS Circ. 402, 71 p.
Moser, H.G., and G.W. Boehlert
1991 Ecology of pelagic larvae and juveniles of the Genus
Sebastes. Environ. Biol. Fish. 30:203-224.
Njoku, E.G., T.P. Barnett, R.M. Laurs, and A.C. Vastano
1985 Advances in satellite sea surface temperature measure-
ment and oceanographic applications. J. Geophys. Res.
90(C6):11, 573-11, 586.
Penney, R.W., and G.T. Evans
1985 Growth histories of larval redfish {Sebastes spp.) on an
offshore Atlantic fishing bank determined by otolith increment
analysis. Can. J. Fish. Aquat. Sci. 42:1452-1464.
Robb, A.P., and J.R.G. Hislop
1980 The food of five gadoid species during the pelagic 0-group
phase in the northern North Sea. J. Fish Biol. 16:199-217.
SAS Institute Inc.
1985 SAS User's guide: Statistics, version 5 edition. SAS In-
stitute Inc.. Cary, NC, 956 p.
Schwing, F.B., S. Ralston, D.M. Husby, and W.H. Lenarz
1990 The nearshore physical oceanography off the central
California coast during May-June, 1989: A summary of CTD
data from juvenile rockfish surveys. NOAA Tech. Memo.
NOAA-TM-NMFS-SWFSC-153, NMFS Tiburon Lab., 142 p.
Searle, S.R., F.M. Speed, and G.A. Milliken
1980 Population marginal means in the linear model: An alter-
native to least squares means. Am. Stat. 34(4):216-221.
Singer, M.M.
1985 Food habits of juvenile rockfishes {Sebastes} in a central
California kelp forest. Fish. Bull., U.S. 83:531-541.
Smith, S.E., and P.B. Adams
1988 Daytime surface swarms of Thysanoessa spinifera (Eu-
phausiacea) in the Gulf of the Farallones, California. Bull.
Mar. Sci. 42(l):76-84.
Sokal. R.R., and F.J. Rohlf
1981 Biometry. W.H. Freeman. San Francisco, 859 p.
Woodbury. D., and S. Ralston
1991 Interannual variation in growth rates and back-calculated
birthdate distributions of pelagic juvenile rockfishes {Sebastes
spp.) off the central California coast. Fish. Bull., U.S. 89:
523-533.
Wyllie Echeverria, T., W.H. Lenarz, and C. Reilly
1990 Survey of the abundance and distribution of pelagic
young-of-the-year rockfishes off central California. NOAA
Tech. Memo. NOAA-TM-NMFS-SWFC-147, NMFS Tiburon
Lab., 125 p.
Zaret, T.M., and A.S. Rand
1971 Competition in tropical stream fishes: Support for the
competitive exclusion principle. Ecology 52:336-342.
Abstract.- The red hind Ejnne-
phelus guttatics, a grouper of com-
mercial importance in the central
western Atlantic, is believed to be
overexploited in a number of areas.
Red hind taken by fish trap and
hook-and-line in western Puerto Rico
and the U.S. Virgin Island of St.
Thomas were aged using sectioned
otoliths (sagittae). Ages were vali-
dated by marginal increment analy-
sis for fish ages 1-10 yr, and by a
field study involving oxytetracycline
injection for fish ages 1-4; a single
opaque and translucent zone (viewed
under transmitted light) is deposited
annually. For Puerto Rico, the von
Bertalanffy growth function (VBGF)
was Lt=514.5 (i_e-»ioi(t+2.94))
Back-calculated mean fork lengths
ranged from 163 mm at age-1 yr, to
448mm at maximum age-17. For St.
Thomas, the VBGF was L, =601.0
(1-
-0 071(t+4.69)
). Back-calculated
mean fork lengths ranged from 194
mm at age-1, to 470 mm at maximum
age-18. Sex and stage of sexual
maturation were determined for a
subsample of aged fish from Puerto
Rico. Fifty percent of females had at-
tained sexual maturity by age 3 yr.
Ages of females were 1-9 yr; males,
2-17 yr, and individuals undergoing
sexual transition from female to
male, 3-7 yr. The male to female sex
ratio was 1:2.6. The occurrence of
sexually-transitional individuals, as
well as significant differences be-
tween the sexes in both size and age,
confirm protogynous hermaphrodit-
ism for fish from Puerto Rico.
Age and growth of red hind
Epinephelus guttatus in
Puerto Rico and St. Thomas
Yvonne Sadovy
Miguel Figuerola*
Ana Roman
Fisheries Research Laboratory, Department of Natural Resources
P O Box 3665. MayagiJez, Puerto Rico
The red hind Epinephelus guttatus is
a serranid of considerable commer-
cial importance throughout the
Caribbean, the Bahamas, and Ber-
muda (Burnett-Herkes 1975, Mahon
1987). In Puerto Rico and the U.S.
Virgin Islands, this species is one of
the most-frequently reported group-
ers in commercial landings. It is
taken by hook-and-line, fish trap, and
speargun, over the insular shelf to a
depth of about 80 m.
Grouper are relatively long-lived
and slow-grovdng fishes. These char-
acteristics, combined with the proto-
gynous sexual pattern (female to
male sex change) reported for many
grouper, and intensive fishing over
short-term traditional spawning ag-
gregations, render grouper species
especially vulnerable to overexploita-
tion (Bannerot et al. 1987, Manooch
1987, Ralston 1987, Shapiro 1987,
Bohnsack 1989).
There are indications that red hind
resources of Puerto Rico and the U.S.
Virgin Islands are being overex-
ploited. Commercial grouper land-
ings reported in Puerto Rico have
declined consistently and substantial-
ly over the last decade, from 386 mt
in 1978 to 47 mt (of which 38% were
red hind) in 1990 (Matos and Sadovy
1989, Sadovy In press, Sadovy and
Figuerola 1992). Yield-per-recruit
analyses indicate growth overfish-
ing (harvesting at too small a size to
maximize potential yield) in western
Manuscript accepted 27 May 1992.
Fishery Bulletin, U.S. 90:516-528(1992).
• Reprint requests should be addressed to this
author.
Puerto Rico (Stevenson 1978, Sadovy
and Figuerola 1992). All known an-
nual spawning aggregations in both
Puerto Rico and St. Thomas are
heavily exploited. In addition, recent
length-frequency data from commer-
cial catches in St. Thomas indicate
that mean length declined substan-
tially between 1984 and 1988 (Beets
and Friedlander 1992), although it is
not clear to what extent this decline
is attributable to overfishing, or is
related to annual variation in recruit-
ment (Appeldoorn et al. 1992), or a
combination of the two.
Little is known of the life history
of the red hind. Previous studies on
age and growth in this species have
been conducted in Bermuda using
whole otoliths (Burnett-Herkes
1975), and in Jamaica using length-
frequency analysis (Thompson and
Munro 1974). However, neither
study is recent and neither validated
the ageing techniques. The sexual
pattern is reported to be protogyny
in Bermuda (Smith 1959, Burnett-
Herkes 1975), and protogyny is also
indicated for Puerto Rico stocks
(Shapiro et al. unpubl. data). The ob-
jectives of this study were to deter-
mine age and growth of the red hind
in two heavily-exploited areas-
western Puerto Rico and St. Thomas,
U.S. Virgin Islands— and to confirm
sexual pattern. This information is
necessary to allow stock assessments
to be made for this species, and to
permit the development of a manage-
ment policy for the red hind in the
region.
516
Sadovy et al,: Age and growth of Epinephelus guttatus in Puerto Rico and St Thomas
517
Methods
Samples of Epinephelus guttatus were obtained from
local fishermen and from Fisheries Research Labora-
tory (FRL) research programs using hook-and-iine and
arrowhead fish traps (3.2cm (1.25 in.) galvanized mesh).
Monthly collections were made between September
1987 and January 1989 with a minimum of 80 fish for
most months from Puerto Rico. Smaller monthly
samples from February 1988-January 1989 were
received from St. Thomas, which lies on the same
geological platform. Fish from St. Thomas were taken
by hook-and-line and by fish trap (3.81cm (1.5 in.)
galvanized mesh).
For each fish, the weight (whole weight to nearest
gm) and length (fork length (FL) and standard length
(SL) to nearest mm) were measured. Otoliths (sagit-
tae) were extracted, washed, and stored dry prior to
processing. Preliminary work determined these cal-
careous structures to be more suitable than other
calcareous structures for ageing purposes: dorsal
spines exhibited growth lines but the central portion
was often eroded resulting in an incomplete growth
history, and scale markings were irregular and thus
considered unreliable for ageing. Gonads in good con-
dition were removed whenever possible, fixed in David-
son's fixative (Yevich and Barszcz 1981), embedded in
paraffin, sectioned at S^m, and stained with hematoxy-
lin and eosin.
Examination of whole otoliths under transmitted
light revealed alternating opaque and translucent zones
(terminology follows that of Wilson et al. 1983). To
count the zones, however, otoliths had to be section-
ed. Preliminary sectioning in two planes (frontal and
transverse; N=20 otoliths in each plane) established
that transverse sections most clearly revealed growth
zones. For sectioning, otoliths were mounted with glue,
using a hot glue gun, on small cards, and sectioned
through the focus with a single 7.2cm (3 in.) diameter,
low-concentration diamond blade on a Buehler Isomet
low-speed saw. From each otolith, three sections of
0.36-0.43 mm were mounted on glass slides using Flo-
Texx mounting medium.
Otolith width (OW) of a sample of unsectioned oto-
liths from a wide size-range of fish was measured to
describe the OW/FL relationship. Measurements of
otolith sections were made from the point where the
sulcus meets the focus to the dorsal margin of the
otolith (the region of most rapid growth) and to the
proximal edge of each opaque zone. Measurements
were also made from the distal edge of the outermost
(opaque) zone to the dorsal margin for marginal incre-
ment analysis. Total number of opaque zones was
noted. Measurements were made with an ocular
micrometer, to the nearest micrometer unit (where
1mm = 32 micrometer units). Each otolith was read
twice. When readings disagreed by more than one
opaque zone, the otolith was eliminated. Readings of
a subsample of otoliths from Puerto Rico, of a wide
range of ages and size-groups, were also made by an
independent researcher.
To validate the temporal significance of opaque
zones, a field study was undertaken. Individuals were
captured by hook-and-line baited with squid from a
30 X 30 m area on a shallow (7 m) relatively-unfished reef
known as "El Negro," 6km off western Puerto Rico.
The study site was visited over a 15-month period be-
tween April 1988 and June 1989. Individuals were
tagged with a numbered FLOY anchor tag inserted
into the dorsal musculature, and/or with FLOY ab-
dominal tags for identification. Each fish was measured
(FL) and injected with a dosage of lOOmg/kg body
weight of Terramycin 100 (Pfizer) (ImL contains
100 mg oxytetracycline-OTC-hydrochloride), and
released. The dosage necessary to produce a visible
mark under longwave ultraviolet light was established
by preliminary tests (50 and lOOmg/kg body weight
were tested; 50mg/kg body weight did not consistent-
ly leave OTC marks) and the correct dosage determined
on-site from a weight/FL relationship. Fish recaptured
were measured and the otoliths examined for opaque
zone formation following deposition of the OTC time
marker.
Data were analyzed separately for Puerto Rico and
St. Thomas using Lotus 1-2-3 and Basic Fishery
Science Programs (Saila et al. 1988). The Kolmogorov-
Smirnov two-sample test and the t -test were used to
compare size-frequency distributions and mean size,
respectively (Sokal and Rohlf 1981). Weight (W) on FL
regressions were calculated using the relationship
W = aFL''. The SL:FL and OW:FL regressions were
determined. The Lee method (Carlander 1981) of back-
calculating body length from prior annuli was used:
Li = a -1- [(Le - a) (0,/OR)],
where: L; = length at time of ith annulus formation
a = intercept
Lc = length at time of capture (FL)
0; = otolith radius at time of ith annulus
formation
OR = otolith radius at time of capture.
This method requires knowledge of the relationship
between OR along the line of measurement and FL.
The constant a is obtained from this relationship and
used in Lee's formula.
Growth was assumed to conform to the von Ber-
talanffy growth function (VBGF) (Ricker 1975). This
518
Fishery Bulletin 90|3). 1992
was calculated from the Saila et al. (1988) statistical
package (FISHPARM program using Marquadt's non-
linear least-squares method) and fitted to mean back-
calculated lengths-at-age. The VBGF is
Lt = L^ (1 -
-K(t-
'"'),
where
Lt
Loc
K
t
to
= length at age t
= asymptote of the growth-in-length
curve
= Brody growth coefficient
= age of the fish
= the theoretical origin of the growth
curve, i.e., age at which fish would have
zero length if it had always grown in a
manner described by the equation.
To establish chronological age at the time of first
opaque zone formation, data on the early growth of
juvenile red hind were assembled from field collections
and observations taken over 9-month periods follow-
ing spawning in January of 1985 and 1987 (Sadovy,
unpubl. data). Since spawning occurs over a limited
period, during, at most, 2 months each year (Erdman
1976, Beets and Friedlander 1992, Shapiro et al. In
press), and settlement may be assumed to occur be-
tween 3 weeks to 2 months after spawning (Colin et
al. 1987), growth rates of individuals in the months
following settlement could be estimated.
160 zoo 240 280 320 360 400 440 480 520
FORK LENGTH lmm|
Figure I
Size-frequency distributions of all Epinephelus guttatus col-
lected (stippled) from Puerto Rico and St. Thomas, and sub-
samples from which otoliths were analyzed (solid).
Results
Samples
Of 1684 Epinephelus guttatus collected from Puerto
Rico, otoliths were sectioned from 1098. Opaque and
translucent zones were detectable in almost all otolith
sections. When zones lacked sufficient definition for
focus-to-ring measurements, the otoliths were dis-
carded as unsuitable for use in calculating growth
parameters, although some were used to assign ages
to sexed fish by counts of opaque zones. A total of 624
(63%) otoliths were used to count growth zones and for
focus-to-ring measurements. Of these, a subsample of
73 otoliths was read by an independent researcher; only
one was rejected because of a discrepancy of more than
one zone compared with our readings. Of the 501 St.
Thomas samples, otoliths were sectioned from 490, and
162 (33%) were judged to be sufficiently clear for
analysis; it is not known why otolith legibility was so
low for St. Thomas samples.
Size-frequency distributions of all fish collected in
Puerto Rico and in St. Thomas, and the subsamples us-
ed for analysis of otoliths from each location, are shown
in Figure 1. For Puerto Rico, size-frequency distribu-
tions of individuals and subsamples used for age deter-
mination did not differ significantly (Kolmogorov-
Smirnov: D = 0.043, NS). This confirmed our impres-
sion that illegible otoliths occurred at all fish sizes and
ages, and affirmed that their elimination introduced no
bias to the calculation of growth parameters. For St.
Thomas, however, the distributions differed signifi-
cantly (D = 0.150, p<0.05). Therefore, growth param-
eters derived for St. Thomas should be treated with
caution.
Frequency distributions of the distance from the
focus to each opaque zone in Puerto Rico collections
are shown in Figure 2. Relationships of SL:FL and
W:FL were established for each location.
Puerto Rico:
FL = 3.86-H 1.2044 SL (r2 0.99; Af 227)
LogW =-5.21-^3.1422 Log FL (r^ 0.97; A^ 1619)
St. Thomas:
FL = 24.49-H 1.1101 SL (r2 0.98; Af 494)
LogW =-4.68-1-2.9402 Log FL (r2 0.92; iV 493)
Sadovy et al Age and growth of Epinephelus gunatus in Puerto Rico and St Thomas 5^9
lOO-
80 -
Epinephelus guttatus PUERTO RICO ^^
>
z
UJ
g .0.
LU
GC
Li.
Z
m 40 -
O
cc
LLI
13.
4mA
20 -
/ / ]( A^iNA/i\n\l\X\
-
1 //YJ^&l^Mst:^
10 30 SO 70 90 "0 130
FOCUS to Ring 1 micrometer units 1
Figure 2
Frequency distributions of the distance from focus to prox-
imal edge of each opaque zone (1-17) in sectioned sag^ittae of
Epinephelus guttatus from Puerto Rico.
Validation
Marginal increment analysis Mean marginal incre-
ments for Puerto Rico fish were plotted on a monthly
basis for annuli I-VI individually, and combined for
annuli VII-X because of low monthly sample sizes for
these older age-groups (Fig. 3). These data indicate
that, at least for annuli II-V, the opaque zone begins
to form between about April and May and is a true
annulus. For annulus VI, the zone is laid down later,
between May and July. For annuli VII-X combined,
opaque zones are apparently deposited annually be-
tween May and July.
Since sample sizes from St. Thomas were too low for
marginal increment analysis by individual age-group,
otoliths from all age-groups were combined and plotted
on a monthly basis by the percent of sections that
lacked a marginal increment. These data indicate that
the time of opaque zone formation (i.e., spring/summer)
is similar for both Puerto Rico and St. Thomas (Fig.
4). A possible 'pseudoannulus' was detected in 74 (12%)
of sectioned otoliths. This was a wide, weakly -discern-
ible band always located between the focus and the first
annulus. It occurred sporadically in fish of all sizes.
Field validation study Of a total of 139 fish tagged,
injected, and released, 8 females from age-classes 1-4
were recaptured 5-16 months following tagging. Mean
monthly growth rates for recaptured individuals from
Puerto Rico measured 1-8 mm (Table 1). One individual
(#00038) was recaptured in February 1991, 18km from
the tagging site on an offshore bank, a known spawn-
9-1
2 ,
ANNULUS
6-
3
^ ^"
1 4-
£ 2
0)
E
o
1 '
„ 2-
c
1 1-
0)
i
» 3-
c
S 2-
10
5 1-
3-
2-
1
\;_J/
1 5
19 e
3 13^^*^--^!°^^*^
\6
ANNULUS
II
4
2 2 ^
2 2
\ ™
ANNULUS
IV o -o
9
t-^
9 "
9 8 6
6
ANNULUS
V
2
6
I f ^ 7 f
8 1
Q \l3,
11 VJ
ANNULUS
VII-X o o
11 7 10/
S 0 N D J F IM
Month
A M
J J A
Figure 3
Marginal increment analyses for annuli I-X for Epinephelus
guttatus from Puerto Rico (1 micrometer unit = 0.031mm).
Numbers above plots are sample sizes. Dashed lines connect
samples separated by more than one month.
a.
100
80
60
40
20
\
■ v
li \
\
^\
■ ^■~'~-->*__
'■
■^ ? . ¥~-Z~
-^ , r- . -f
, *-
SONDJ FMAMJ JA
Month
Figure 4
Percentage of otolith sections lacking a marginal in-
crement for Epinephelus guttatus from Puerto Rico
(solid line) and St. Thomas (dashed line) for age groups
1-10.
520
Fishery Bulletin 90(3), 1992
Temporal
fish. OR
Table 1
significance of opaque growth zones in otoliths oiEpinephelus guttatus from Puerto Rico, based
= otolith radius; MI = marginal increment; OTC = oxytetracycline mark.
on OTC
-marked/recaptured
Tag
no.
Date (D/MAO
Capture
Recapture
FL (mm)
Capture
Recapture
Days in
field
Measurements (32 micrometer units
= 1 mn:
1)
No. of
opaque
zones
Focus to ring
OR
MI
1
2
3 4
OTC
00094
01/06/88
31/10/88
163
203
152
37*
-
- -
'46
55
4
1
00078
25/05/88
04/05/89
127
151
344
45*
59
— —
■52
63
0
2
00076
01/06/88
14/02/89
202
231
258
40
52
61 -
^65
72
4
3
00468
13/05/88
20/03/89
210
220
311
38
54
60 -
=68
73
5
3
00418
02/06/88
31/10/88
214
220
151
34
51
62* -
'64
71
0
3
00385
19/08/88
10/02/89
218
248
175
39
50
69
^60
70
0
3
^00038
11/10/89
14/02/91
261
289
491
43
61
75 -
^72
85
3
3
00426
08/06/88
07/02/89
287 244 45
317
zone indicated by asterisk,
opaque and translucent zones.
;ent zone.
)ecember 1988; inner OTC mark 52'^'.
60 73 87 =90 95 4
Retagged 11/10/89 (only retag data presented).
4
'OTC mark lies within opaque
^OTC mark lies at junction of
^OTC mark lies within transluc
■•Fish originally tagged June-E
Table 2
Mean back-calculated and observed fork lengths (mm) at time of opaque zone formation (yr)
for age-groups 1-
-17 of 624 Epinephelus \
guttatus from Puerto Rico.
Age-
Mean length
group
N
at capture
SD
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15 16 17
1
19
184
47
161
2
97
247
21
168
213
3
90
256
23
162
203
234
4
52
272
26
159
195
226
253
5
87
306
30
163
205
237
263
286
6
126
321
33
165
207
237
262
283
303
7
52
342
42
163
205
236
262
284
305
325
8
29
353
34
169
209
236
260
283
304
320
336
9
26
374
36
160
209
241
269
290
310
327
340
357
10
27
393
32
170
213
244
271
293
312
329
346
361
374
11
5
413
41
163
211
247
270
298
315
333
349
366
382
399
12
4
416
30
160
194
226
257
283
302
321
335
352
366
382
394
13
2
422
17
168
219
250
275
296
313
334
347
360
372
384
399
410
14
5
448
31
172
212
247
278
304
323
340
354
368
383
409
417
432
15
0
—
—
16
1
448
—
155
213
238
262
277
289
308
332
353
372
387
393
405
414
424 436
17 2 458 46
Back-calculated mean lengths
159
216
247
271
293
312
323
336
348
362
380
395
406
416
431 440 448
164
206
236
262
286
306
326
342
359
375
390
401
412
426
429 439 448
(weighted)
Growth increments
42
30
26
24
20
20
16
16
16
15
11
11
14
3 10 9
Sadovy et al Age and growth of Epinephelus guttatus in Puerto Rico and St Thomas
521
ing aggregation area. To reach the bank, this fish must
have crossed water of at least 194 m depth, a substan-
tial depth for similar-sized individuals of this species
(Sadovy et al., unpubl. data).
During the tagging study, significant data loss oc-
curred; in approximately 60% of tagged fish resighted,
the identifying number of the dorsal tag had detached,
leaving behind a monofilament anchor partially em-
bedded in dorsal musculature. On the other hand,
resightings of fish marked with abdominal tags in-
dicated that all had retained both the numbered tag
anchor and the attached color streamer.
Data for age-groups of recaptured fish indicated that
no more than one opaque zone is deposited annually,
although sample size was limited. Opaque zone forma-
tion had begun in or after February, had terminated
prior to August, and occurred somewhat later in the
year in older age-groups.
Recaptures were initiated as early as 5 months after
tagging because individuals typically disappeared from
the immediate study site within a few months of cap-
ture. Results covering less than a 12-month field period
should be treated with caution, although all results
were consistent with the marginal increment analysis
in terms of both the temporal nature of opaque zones
and the time of their annual deposition.
Age and growth
For Puerto Rico, the FL/OR relationship is
FL = 33.2180 -H 3.0743 OR (r2 0.76; N 624).
GROWTH CURVES for Epinephelus gultalus
PUERTO RICO
Nt624
• BACK- CALCULATED
O OBSERVED
- THEOBETICAL
J STANDARD DEVIATION
|1S.dl
AGE I years I
Figure 5
Empirical, back-calculated, and theoretical (von Berta-
lanffy) growth curves for Epinepheliis guttatus from
Puerto Rico.
Table 2 shows the mean back-calculated lengths for
ages 1-17 years from 624 fish. The following growth
parameter estimates were obtained from the von
Bertalanffy growth function (with asymptotic SE in
parentheses):
L^ = 514.5mmFL (6.29)
K = 0.1013 (0.003765)
to = -2.944 (0.1357).
parentheses):
L^ = eOl.OmmFL (32.82)
K = 0.0705 (0.009954)
to = -4.690 (0.5920).
Figure 6 shows the empirical mean lengths and their
standard deviations, as well as back-calculated and
theoretical growth curves, for St. Thomas.
Figure 5 shows empirical mean lengths and their
standard deviations, as well as back-calculated and
theoretical (VBGF) growth curves, for Puerto Rico.
For St. Thomas the FL/OR relationship is
FL = 94.7206-1-2.4757 OR (r^- 0.68; A^ 162).
Table 3 shows the mean back-calculated lengths for
age-groups 1-18 from 162 fish. The following growth
parameter estimates were obtained from the von
Bertalanffy growth function (with asymptotic SE in
For Puerto Rico, the OW/FL relationship is
OW = 1.4205 + 0.0108 FL (r2 0.93; N 315; Fig. 7).
For St. Thomas, the OW/FL relationship is
OW = 0.5591 -H 0.0049 FL (r^ 0.93; A^ 101).
When mean back-calculated fork lengths for annuli
I-V for age-groups 1-14 from Puerto Rico and St.
Thomas are plotted (Fig. 8), two points are worthy of
note. If regressions for each annulus are calculated for
all available ages up to age-group 14, all are statistically
522
Fishery Bulletin 90(3). 1992
Table 3
Mean back-calculated and observed fork lengths (
mm) at time of opaque
zone
formation (yr) for age-groups 1-18 of 162 Epinepkelus
guttatui
Age-
fron
1 St. Thomas.
Mean length
group
N
at capture
SD
1
2
3
4
5
6
7
8
9
10
11
12
13
14 15 16 17 18
1
7
249
12
204
2
10
260
17
195
232
3
39
280
25
190
229
259
4
10
284
28
185
213
243
265
5
16
319
28
195
229
260
286
308
6
27
335
27
195
228
255
278
298
316
7
11
348
29
194
227
255
278
297
317
333
8
13
362
29
195
230
258
282
303
320
333
349
9
9
370
55
195
226
252
275
296
315
331
343
358
10
5
356
28
184
212
238
261
279
296
312
324
333
345
11
5
391
45
195
228
253
279
300
317
332
345
355
367
378
12
3
424
51
203
242
273
294
311
329
346
358
375
388
401
411
13
1
460
—
206
234
270
290
332
348
365
382
396
410
424
438
452
14
3
442
22
222
246
269
291
318
339
355
368
378
391
399
412
420
430
15
2
492
12
215
263
290
316
343
362
378
389
405
421
434
450
460
469 478
16
0
—
—
17
0
—
—
18 1 475
Back-calculated mean lengths
192
217
248
273
299
324
350
368
386
401
414
427
432
439 452 460 4G5 470
194
228
256
279
301
318
335
349
362
377
400
423
438
445 469 460 465 470
(weighted)
Growth increments
34
28
23
22
17
17
14
13
15
23
23
38
7 24 -9 5 5
GROWTH
CURVES for Epinephelus guttatus
St. THOMAS
500 -
• ^
400 -
L<^
r^'
'iHl
f
tK
300 -
.;
V
X
200 -
/
•
N=T62
• BACK-CALCULATED
O OBSERVED
100 -
- THEORETICAL
T STANDARD DEVIATION
AGE lyearsi
Figure 6
Empirical, back-calculated, and theoretical (von Bertalanffy)
growth curves for Epinepheliis guttatus from St. Thomas.
significant (jD<0.05), with the exception of annuli I and
II for Puerto Rico. Also of note are two distinct depres-
sions in back-calculated mean lengths for each geo-
graphic location. For Puerto Rico these are indicated
at age-groups 4 and 12, and for St. Thomas at age-
groups 4 and 10. These points will be addressed in the
discussion.
Data on growth during the months following spawn-
ing indicate rapid growth from the time of settlement
in February or March at ~40mmFL (A^29), to 60
mmFL in April/May (A'' 3), 108mm (A'' 2) in August, and
115 mm {N 4) in October.
Age, sex, and size
A total of 186 individuals from Puerto Rico were sexed,
by histological examination of gonads, and aged. Of
these, 131 were female and 50 were male; 5 were con-
sidered to be in sexual transition between female and
male, with gonads consisting of degenerating sexually-
mature (i.e., vitellogenic) ovarian tissue and scattered
areas of spermatogenic tissue exhibiting various stages
of spermatogenesis (Sadovy and Shapiro 1987). All
males exhibited testes with an ovarian-like configura-
tion and lumen (Sadovy and Shapiro 1987).
Mean observed FL, standard deviation, and sample
size by sex and age-class from Puerto Rico are shown
Sadovy et al : Age and growth of Epinephelus guttatus in Puerto Rico and Si Thomas
523
■ "
130-
. .
■ V#" ""■■
F
110-
.■«
*'flC-C^"
(0
-1
■
■
■ Jl
U&O"
■D
to
90-
■;5
OC
- *^
^F ^
r
■ WW
■
o
70-
--In
O
50-
Jw
■ ■
1(
)0
200
300
400 5C
Fork Length (mm)
100 200 300 400
Fork Length (mm)
500
Figure 7
(upper) Relationship between otolith radius and fork length
for Epinephelus guttatus from Puerto Rico (1 micrometer
unit = 0.0312 mm), (lower) Relationship between un-
sectioned otolith width and fork length for Epinephelus
guttatus from Puerto Rico.
PUERTO RICO
4 6 8 10
Age Group
ST. THOMAS
^^ Annulus I -•- Annulus II Annulus III
-*- Annulus IV — — Annulus V
Figure 8
(upper) Mean back-calculated fork length of annuli I-V
for Epinephelus guttatus from Puerto Rico. Asterisks
denote significance at p<0.05; NS = nonsignificant.
Oower) Mean back-calculated fork length of anmoli I-V
for Epinephelus guttatus from St. Thomas. Asterisks
denote significance at p<0.05.
in Table 4. Fork length at time of capture (i.e., observed
FL) did not differ significantly between the sexes for
age-groups 2-8 (Table 4). However, both age- and size-
frequency distributions from both Puerto Rico and St.
Thomas differed significantly by sex (age: D = 0.593,
p<0.01; size: D = 0.576, p<0.01) (Fig. 9).
Females were found at ages 1-9, males at ages 2-17,
and individuals undergoing sexual transition at ages
3-7 (Table 4). No sexually-mature individuals were
detected below age 2 years. Fifty percent of females
had attained sexual maturity by age 3.
Discussion
Validation and formation of opaque zones
Opaque zones incrementally deposited in the sagittae
of the red hind Epinephelus guttatus were validated
as annual in this study, both by marginal increment
analysis and by a field study involving the marking of
otoliths by OTC. Although the possibility of bi- or multi-
annual growth lines (Deelder 1981, Lee et al. 1983) for
fish aged 11 years and over could not be discarded,
these age-groups constituted a small percentage of
sampled individuals; hence, the validation may be ap-
plied confidently to the exploited segment of red hind
stocks in the region.
It is not known what causes the formation of opaque
and translucent zones in this species. However, since
zones are found in both adults and juveniles, they are
clearly not caused exclusively by spawning activity
(Nekrasov 1980). Opaque zone formation has been pro-
posed to be associated with low somatic growth rates,
and translucent zones with high growth rates, in the
white grunt Ha^mulon plumieri (Sadovy and Severin
1992). A similar relationship is proposed for the red
524
Fishery Bulletin 90(3). 1992
PUERTO RICO
30
Frequency
n
10
n
,
1.
ll
IiDI I 1 ..
0 2 4 6 8 10 12 14 16
Age
20
Ifl
Frequency
L
5
p
.n
lll.lll...
165 210 255 300 345 390 435 480
Fork Length (mm)
hind (Sadovy and Severin, unpubl. data). Since time of opaque zone
formation was earlier in the year in younger than in older fish, zone
formation is unlikely to be caused by a simple environmental factor
acting directly and equally on all individuals. Time of opaque zone
formation is February-July, which is similar to that reported for four
other groupers of the genus Epinephelus from the western Atlantic
(Table 5). The pattern of earlier annual opaque zone formation in
younger individuals noted in our study was also reported in otoliths
from E. morio (Moe 1969) and Mycteroperca microlepis (Collins et
al. 1987), and in pike Esox lucius (aged using cleithra; Casselman
1983).
Data on growth of red hind in the months following settlement
indicate rapid growth from the time of settlement at ~40mmFL
(N 29), to 115 mm PL the following October. These data indicate that
the first opaque zone, which is laid down between March and April
at a back-calculated 164mmFL (SD 18mm) in Puerto Rico and
194mmFL (SD 17mm) in St. Thomas, represents an age 1+ year
fish (13-15 months old, depending on month of spawning).
Growth parameters and longevity
Red hind in Puerto Rico and St. Thomas are long-lived and attain
their maximum size slowly, following fast growth during the first
year. Thompson and Munro (1974), using length-frequency analysis,
calculated L^ = 520mmFL for Jamaica-Pedro Bank fish, and L^ =
500mmFL for Jamaica-Port Royal fish. Burnett-Herkes (1975),
Figure 9
(upper) Age-frequency distributions of female
(stippled) and male (solid) Epinephelus guttatus
from Puerto Rico, (lower) Size-fre-
quency distributions of female (dotted)
and male (solid) Epinephelus guttatus
from Puerto Rico.
Table 4
Mean
observed fork length (mm) and standard deviation
for male and female red
hind.
Epinephelus guttaUis, by age-group. Fish from Puerto Rico collected September 1987-Septem-
ber 1988 (•p<0.05). NS = nonsignificant.
Observed fork length (mm)
Female
Male
Student's
t
fish
Age
X
SD
N
X
SD
N
FL
N
1
196.3
16.3
13
0
2
238.7
20.6
23
253.7
34.2
3
1.1076 NS
0
3
251.5
18.6
37
301.0
—
1
0.1410 NS
275
1
4
267.2
28.0
23
280.7
10.5
3
0.8150 NS
278
1
5
307.2
28.4
10
288.0
7.3
5
1.4621 NS
0
6
344.9
35.8
12
322.2
30.4
15
1.7823 NS
315
1
7
340.6
25.9
8
336.6
49.4
10
0.2066 NS
258; 380
2
8
362.5
17.5
2
368.7
50.4
3
0.2536 NS
0
9
326.7
12.5
3
—
—
0
—
0
10
—
—
—
418.0
14.9
4
—
0
11
—
—
—
—
—
0
_
0
12
—
—
—
440.0
10.0
2
—
0
13
—
—
—
—
—
0
—
0
14
—
—
—
470.0
10.0
2
—
0
15
—
—
—
—
—
0
—
0
16
—
—
—
448.0
—
1
_
0
17
-
-
-
490.0
-
1
-
0
Total
268.0
50.0
131
342.0
65.0
50
8.1597*
301
5
Sadovy et al : Age and growth of Epinephelus guttatus in Puerto Rico and St Thomas
525
Table 5
Caribbean and western Atlantic Epinephelus spp. aged by
whole or
sectioned otoliths.
Grovrth parameters
♦
Time of opaque
Max. age
L„
to
Species
(mm FL)
K
(yr)
zone formation
(yr)
Source
£. nigritus
2394
0.054
-3.616
April-May
41
Manooch and Mason 1987
E. niveatus
1320
0.087
-1.012
May-July
17
Moore and Labisky 1984
E. niveatus
1255
0.074
-1.920
May-July
17
Matheson and Huntsman 1984
E. drummondhayi
967
0.130
-1.010
April-June
15
Matheson and Huntsman 1984
E. morio
928
0.113
0.091
—
14 +
Melo 1975 (cited in Manooch 1987)
E. morio
792
0.179
-0.449
March-July
25 +
Moe 1969
E. guttatus
601
0.071
-4.690
March-June?
18 +
This study; St. Thomas
E. guttatus
515
0.101
-2.944
February-July
17 +
This study; Puerto Rico
E. guttatus
507
0.180
-0.440
—
17 +
Burnett-Herkes 1975; also used
the growth-in
-length curve
K = Brody growth coefficient;
length-frequency data
t(, = theoretical origin of von Bertalanffy growth
* L„ = asymptote of
curve (Ricker 1975)
using whole otoliths and length-frequency analyses
for ageing, reported L^=507mmFL in Bermuda.
The largest fish sampled in the present study were
490 mm FL in Puerto Rico and 504mmFL in St.
Thomas. In several years of intensive sampling of
thousands of red hind from local commercial landings,
Fisheries Research Laboratory (FRL) data recorded
<2% of individuals >500mmFL (FRL, unpubl. data).
These data reflect asymptotic lengths established in
the present study. Randall (1983) reported the largest
West Indian specimen collected to be 673mmFL, and
Smith (1971) reported the largest fish he examined to
be SlOmmSL (618mmFL using the above FL/SL
relationship).
Growth parameters obtained using otoliths as the
ageing structure for western Atlantic species of the
genus Epinephelus are shown in Table 5. The data for
E. guttatus in Puerto Rico and St. Thomas fall within
the range of values of L^, K, and maximum age
reported for western Atlantic grouper. A maximum of
17 growth zones in Puerto Rico and 18 growth zones
in St. Thomas were recorded. We consider the esti-
mated longevity of 17 + and 18+ to be reasonable for
commercially-taken red hind in Puerto Rico and St.
Thomas, respectively. Luckhurst et al. (1992) recorded
22 ( ± 1) opaque zones in an unusually-large 720mmFL
individual from Bermuda.
Lengths-at-capture were consistently higher than
back-calculated lengths for each age-group (Figs. 5, 6).
The higher observed mean fork lengths generally
reflect additional growrth between previous ring forma-
tion and time of capture. However, the notably high
observed mean FL of ages-1 and -2 fish for both Puerto
Rico and St. Thomas may be due, in part, to selection
by the fishery of the largest fish in these younger age-
groups. Similar selection was also reported for the first
two age-classes in E. morio (Moe 1969) and may be
especially common in longer-lived, slower-growing
species, such as grouper (Bannerot 1984). When sam-
pling is biased towards larger individuals of young year-
classes, there may be artificial depression of K in the
VBGF (Ricker 1975). A downward bias would generally
produce conservative management advice in terms of
justifying imposition of minimum size regulations based
on future returns to the fishery (Bannerot 1984).
The relationship between FL and OR for both loca-
tions is somewhat weaker than in other studies of fish
age and growth. Since the OW/FL relationships for
otoliths from both locations are strong, this indicates
that variability is introduced by the position on the sec-
tioned otolith selected for measurement and counting
of opaque zones, rather than by a poor relationship
between body length and otolith size (Fig. 7).
The regressions for mean back-calculated fork
lengths of annuli I-V of age-groups 1-14 for Puerto
Rico and St. Thomas are statistically significant, with
two exceptions (Fig. 8). Such a trend could suggest a
reverse "Rosa-Lee" phenomenon, indicating enhanced
survivorship of fast-growing fish (Ricker 1975). If this
were true, it would result in an upward bias of the
parameter K in the VBGF. However, sample sizes for
individuals above age-group 10 are very low, and re-
gressions for age-groups 5-10 are not significant, with
the exception of annulus II for Puerto Rico. Since these
age-groups comprise the bulk of commercial landings,
we believe that bias to estimates of growth parameters
derived in the present study is negligible. Furthermore,
since a similar increase is apparent in most age-groups
526
Fishery Bulletin 90(3|. 1992
below age-group 4, there is clearly no consistent rela-
tionship between growth rate and mortality.
The depressions in back-calculated mean lengths
(Fig. 8) apply to years 1983-84 (age-group 4) and
1975-76 (age-group 12) for Puerto Rico, and 1984 (age-
group 4) and 1978 (age-group 10) for St. Thomas. In
the case of older age-classes, small sample sizes could
have produced sampling errors. However, we believe
that in the case of 1984, this pattern is unlikely to be
the result of sampling artifacts. Possible explanations
for lower mean back-calculated lengths in both loca-
tions include environmental factors, such as unusually
low temperatures or reduced food availability, or
alterations in fishing effort and associated demographic
changes. We know of no changes in fishing effort or
gear during the early 1980s at either location. Reduced
environmental temperatures or food availability may
have caused age-1 and -2 fish to experience a decrease
in growth rate that carried over into later years. In-
terestingly, catch curves developed from the same data
set indicate a particularly low recruitment into the
fishery of age-4 fish in 1984 (Sadovy and Figuerola
1992). This trend in the catch curves is strikingly
similar for both locations, strongly suggesting a region-
wide phenomenon. Poor recruitment into the fishery
of a young age-class could result from slow growth
early in life of individuals of that cohort.
Examination of temperature records for the region
indicates that the winter of 1984 was the first since
that of 1975 in which the mean minimum temperature
dropped below 26°C (lat. 14.7°-18.2°) (Atwood and
Hendee In press). In summary, we suggest that lower
temperatures may have retarded growth in young in-
dividuals and that this reduction in size-at-age early in
life was carried through the growth history of the
animal.
The red hind and fishery management
The condition of growth overfishing in the red hind
(Sadovy and Figuerola 1992), and the general vulner-
ability of grouper species to fishing pressure, indicate
the urgent need for management, stock monitoring,
and assessment throughout its geographic range. In
particular, given the apparent importance of spawn-
ing aggregations for annual reproductive output in the
red hind (Bohnsack 1989; Shapiro 1987), and the in-
tensity with which these are exploited locally (Sadovy,
unpubl.), the possibility of recruitment overfishing
needs to be addressed.
Acknowledgments
We thank the Caribbean Fishery Management Coun-
cil, NMFS/NOAA, which is largely responsible for fun-
ding this research, and Omar Munoz-Roure and Miguel
Rolon for support in the early phases of developing this
study. We are particularly grateful to Richard Appel-
doorn, Jim Burnett-Herkes, Mark Collins, George
Dennis III, Doug DeVries, Allyn Johnson, Angle
McGehee, Charles Manooch III, and George Mitcheson
for information, help, and advice. Charles Manooch
kindly evaluated a subsample of otoliths. Stephen
Ralston drew our attention to implications in the data
we have discussed and represented as Figure 8. Thanks
are due to Jim Beets who provided the otolith samples
from St. Thomas. The cooperation and support of the
Exploration Team of the Fisheries Research Labora-
tory and of local fishermen, especially Santiago V6lez
Ocasio and Wilfredo Velez Ocasio, were much appre-
ciated. We thank Bonny Bower-Dennis who prepared
some of the figures.
Age and sex
Among sexed individuals, the majority (80%) of females
were ages 1-5, and the males ages 2-10. Empirical
mean lengths for age-groups 2-8 did not differ by sex.
Moe (1969) also found the empirical growth curves of
the sexes oiE. morio to be similar, indicating that there
are no marked differences in growth between the sexes
or sexual phases of an individual. The combination of
histological data, especially the presence of transi-
tionals, size/age frequency distributions, and a female-
biased sex ratio, confirm protogyny for this species in
Puerto Rico. However, the presence of males as young
as the youngest mature female indicates that at least
some males may develop directly from a juvenile phase
without passing through an initial functional female
phase.
Citations
Appeldoorn, R., J. Beets, J. Bohnsack, S. Bolden, D. Matos,
S. Meyers, A. Rosario, Y. Sadovy, and W. Tobias
1992 Shallow water reef fish stock assessment for the U.S.
Caribbean. NOAA Tech. Memo. NMFS-SEFSC-304, NMFS
Southeast Fish. Sci. Cent., Miami, 70 p.
Atwood, D.K., and J.C. Hendee
In press An assessment of global warming stress on Carib-
bean coral reef ecosystems. Proc. Gulf Caribb. Fish. Inst. 44.
Banneret, S.P.
1984 The dynamics of exploited groupers (Serranidae): An in-
vestigation of the protogynous hermaphroditic reproductive
strategy. Ph.D. diss., Univ. Miami, Coral Gables, 393 p.
Bannerot, S.P.. W.W. Fox Jr., and J.E. Powers
1987 Reproductive strategies and the management of snap-
pers and groupers in the Gulf of Mexico and Caribbean. In
Polovina, J.J., and S. Ralston (eds.), Tropical snappers and
groupers: Biology and fisheries management, p. 561-603.
Westview Press, Boulder.
Sadovy et al Age and growth of Epinephelus guttatus in Puerto Rico and St Thomas
527
Beets, J., and A. Friedlander
1992 Stock analysis and management strategies for red hind,
Epinephelus guttatus, in the U.S. Virgin Islands. Proc. Gulf
Caribb. Fish. Inst. 42:66-80.
Bohnsack, J. A.
1989 Protection of grouper spawning aggregations. Coastal
Resour. Div. Contrib. CRD 88-89-06, NMFS Southeast Fish.
Sci. Cent., Miami, 8 p.
Burnett-Herkes, J.
1975 Contribution to the biology of the red hind, Epinephelus
guttatus, a commercially important serranid fish from the
tropical western Atlantic. Ph.D. diss., Univ. Miami, Coral
Gables. 1.54 p.
Carlander. K.D.
1981 Caution on the use of the regression method of back-
calculating lengths from scale measurements. Fisheries
(Bethesda) 6(l):2-4.
Casselman. J.M.
1983 Age and growth assessment of fish from their calcified
structures— techniques and tools. In Prince, E.D., and L.M.
Pulos (eds.), Proc, Int. workshop on age determination of
oceanic pelagic fishes, p. 1-17. NOAA Tech. Rep. NMFS 8.
Colin, P.L., D.Y. Shapiro, and D. Weiler
1987 Aspects of the reproduction of two groupers, Epimphehis
guttatus and E. striatus in the West Indies. Bull. Mar. Sci.
40(2):220-230.
Collins, M.R., C.W. Waltz, W.A. Rouirillat, and D.L. Stubbs
1987 Contribution to the life history and reproductive biology
of gag, Mycteroperca microlepis (Serranidae), in the south
Atlantic Bight. Fish. Bull., U.S. 85:648-653
Deelder, C.L.
1981 On the age and growth of cultured eels, Anguilla anguilla
(Linnaeus, 1758). Aquaculture 26:13-22.
Erdman, D.S.
1976 Spawning patterns of fishes from the northeastern Carib-
bean. Contrib. Agric. Pesq., Dep. Agric. Commonwealth of
Puerto Rico 8(2): 1-37.
Lee, D.W.. E.D. Prince, and M.E. Crow
1983 Interpretation of growth bands on vertebrae and otoliths
of Atlantic bluefin tuna, Thunnus thunnus. In Prince, E.D.,
and L.M. Pulos (eds.), Proc, Int. workshop on age determina-
tion of oceanic pelagic fishes, p. 61-69. NOAA Tech. Rep.
NMFS 8.
Luckhurst, B., J. Barnes, and Y. Sadovy
1992 Record of an unusually large red hind, Epirwpkelus gut-
tatus (Serranidae), from Bermuda with comments on its age.
Bull. Mar. Sci. 51(2):267-270.
Mahon, R. (editor)
1987 Report and proceedings of the expert consultation on
shared fishery resources of the Lesser Antilles region,
Mayagiiez, Puerto Rico, 8-12 Sept. 1986. FAO Fish. Rep. 383,
278 p.
Manooch, C.S. Ill
1987 Age and growth in snappers and groupers. In Polovina,
J.J., and S. Ralston (eds.). Biology and management of snap-
pers and groupers, p. 329-373. Westview Press, Boulder.
Manooch, C.S. Ill, and D.L. Mason
1987 Age and growth of the Warsaw grouper and black
grouper from the southeast region of the United States.
Northeast Gulf Sci. 9(2):65-75.
Matheson, R.H. Ill, and G.R. Huntsman
1984 Growth, mortality, and yield-per-recruit models for
speckled hind and snowy grouper from the United States South
Atlantic Bight. Trans. Am. Fish. Soc. 113:607-616.
Matos, D., and Y. Sadovy
1989 CODREMAR/NMFS cooperative statistics program for
1988-1989, Annual report July 1989. Coop. Stat. Prog.,
NMFS, St. Petersburg FL, and Puerto Rico Dep. Nat. Resour.,
56 p.
Melo, A.M.
1975 Aspectos biol6gicos pesqueros de Epinephelus morio
(Val.). Unpubl. thesis, Univ. Nac. Aut6n. Mexico, Mexico D.F,
68 p.
Moe, M.A.
1969 Biology of the red grouper Epinephelus morio (Valen-
ciennes), from the eastern Gulf of Mexico. Fla. Dep. Nat.
Resour. Mar. Res. Lab. Prof. Pap. Ser. 10, 95 p.
Moore, CM., and R.L. Labisky
1984 Population parameters of a relatively unexploited stock
of snowy grouper in the lower Florida Keys. Trans. Am. Fish.
Soc. 113:322-329.
Nekrasov, V.V.
1980 The causes of annulus formation in tropical fishes.
Hydrobiol. J. (Engl, transl. Gidrobiol. Zh.) 2:35-39.
Ralston, S.
1987 Mortality rates of snappers and groupers. In Polovina,
J.J., and S. Ralston (eds.). Tropical snappers and groupers:
Biology and fisheries management, p. 375-404. Westview
Press, Boulder.
Randall, J.E.
1983 Caribbean reef fishes, 2d ed., rev. T.F.H. Publ., Nep-
tune, NJ, 350 p.
Ricker, W.E.
1975 Computation and interpretation of biological statistics
of fish populations. Bull. Fish. Res. Board Can. 191, 382 p.
Sadovy, Y.
In press Grouper stocks of the western central Atlantic: The
need for management and management needs. Proc. Gulf.
Caribb. Fish. Inst. 43.
Sadovy, Y., and M. Figuerola
1992 The status of the red hind fishery in Puerto Rico and St.
Thomas, as determined by jield-per-recruit analysis. Proc.
Gulf Caribb. Fish. Inst. 42:23-38.
Sadovy, Y., and K.P. Severin
1992 Trace elements in biogenic aragonite: Correlation of body
growth rate and strontium levels in the otoliths of the white
grunt, Haemulon plumieri (Pisces: Haemulidae). Bull. Mar.
Sci. 50(2):237-257.
Sadovy, Y., and D.Y. Shapiro
1987 Criteria for the diagnosis of hermaphroditism in fishes.
Copeia 1987 (1):136-156.
Saila, S.B.. C.W. Recksiek, and M.H. Prager
1988 Basic fishery science programs: A compendium of micro-
computer programs and manual of operation. Dev. Aquacult.
Fish. Sci. 18. Elsevier Press, Amsterdam, 230 p.
Shapiro, D.Y.
1987 Reproduction in groupers. In Polovina, J.J., and S.
Ralston (eds.). Tropical snappers and groupers: Biology and
fisheries management, p. 295-327. Westview Press, Boulder.
Shapiro, D.Y., Y. Sadovy, and M.A. McGehee
In press Size, composition, and spatial structure of the annual
spawning aggregation of the red hind, Epinephelus guttatus
(Pisces: Serranidae). Copeia.
Smith, C.L.
1959 Hermaphroditism in some serranid fishes from Bermuda.
Pap. Mich. Acad. Sci. Arts Lett. 44:111-119.
1971 A revision of the American groupers: Epinephelus and
allied genera. Bull. Am. Mus. Nat. Hist. 146:69-241.
528
Fishery Bulletin 90(3). 1992
Sokal, R.R., and F.J. Rohlf
1981 Biometry, 2d ed. W.H. Freeman, San Francisco, 859 p.
Stevenson, D.K.
1978 Management of a tropical fish pot fishery for maximum
sustainable yield. Proc. Gulf. Caribb. Fish. Inst. 30:95-115.
Thompson, R., and J.L. Munro
1974 The biology, ecology and bionomics of Caribbean reef
fishes: Serranidae (hinds and groupers). Res. Rep. Zool. Dep.,
Univ. West Indies, Jamaica, 3(5b), 82 p.
Wilson, C.A., E.B. Brothers, J.M. Casselman, C.L. Smith, and
A. Wild
1983 Glossary. In Prince, E.D., and L.M, Pulos (eds.), Proc,
Int. workshop on age determination of oceanic pelagic fishes,
p. 207-208. NOAA Tech. Rep. NMFS 8.
Yevich, P.R., and C.A. Barszcz
1981 Preparation of aquatic animals for histopathological ex-
amination. Aquat. Biol. Sect., Biol. Methods Br., Environ.
Monit. & Support Lab., U.S. Environ. Protect. Agcy, Cincin-
nati, 81 p.
Abstract.- Spawning patterns,
larval distribution, and juvenile growth
characteristics were examined for
tautog Tautoga onitis in New Jersey
and the Mid- Atlantic Bight. We ana-
lyzed data from plankton surveys
(1972-1990) over the continental
shelf and in the Great Bay-Mullica
River estuarine system. Data on size
and abundance of juveniles were
derived from throw trap and trawl
collections in New Jersey estuaries
(1988-89). In addition, we validated
the daily deposition of otolith incre-
ments and used increment counts to
estimate juvenile age and growth
patterns. Extensive egg and larval
collections indicated that spawning
occurs from April through Septem-
ber, with a peak in June and July.
Spawning over the continental shelf
is concentrated off Long Island and
Rhode Island. Based on validated
daily increments in sagittal otoliths
and the formation of a well-defined
settlement mark, tautog larvae spend
about 3 weeks in the plankton. Both
spawning and settlement occur over
a prolonged period, based on otolith
back-calculations. Three methods of
estimating young-of-the-year growth
rates, including length-frequency
progressions, otolith age/fish-size
comparisons, and direct measure-
ment of growth in caging experi-
ments, indicated an average growth
rate of about 0.5 mm/day during the
peak midsummer growing season.
Length-frequency distributions sug-
gested tautog reach a modal size of
about 75 mm SL after their first sum-
mer, and 155 mm by the end of their
second summer.
Early life history of the
tautog Tautoga onitis
in the Mid-Atlantic Bight*
Susan M. Sogard
Marine Field Station, Institute of Marine and Coastal Sciences
Rutgers University, Great Bay Boulevard, Tuckerton, New Jersey 08087
Present address: Hatfield Marine Science Center
Oregon State University, Newpori:, Oregon 97365
Kenneth W. Able
Marine Field Station, Institute of Marine and Coastal Sciences
Rutgers University, Great Bay Boulevard, Tuckerton, New Jersey 08087
Michael P. Fahay
Sandy Hook Laboratory, Nori:heast Fisheries Science Center
National Marine Fisheries Service, NOAA, Highlands, New Jersey 07732
The tautog Tautoga onitis is one of
two labrid wrasses common along the
northeast coast of the United States
(the other is the cunner Tautogola-
brus adspersus). Tautog occur in
coastal areas from Nova Scotia to
South Carolina, but are abundant
only from Cape Cod to the Delaware
Capes (Bigelow and Schroeder 1953).
Adult tautog form a minor compo-
nent of local commercial fisheries and
a major component of the recrea-
tional catch. They reach a maximum
size of about 90 cm and 10 kg (Bige-
low and Schroeder 1953), and an age
of 34 years (Cooper 1967). Large
juveniles and adults depend on young
mussels Mytilus edulis for food (011a
et al. 1974), and the diet of recently-
settled juveniles consists primarily of
copepods and amphipods (Grover
1982). Spawning takes place from
May to August, with a peak in June
(Kuntz and Radcliffe 1918, Colton et
al. 1979, Eklund and Targett 1990).
Egg and larval development are de-
scribed in detail by Kuntz and Rad-
cliffe (1918) and Williams (1967); ad-
ditional information on life history is
Manuscript accepted 20 May 1992.
Fishery Bulletin, U.S. 90:529-539 (1992).
' Contribution 92-07 of the Rutgers University
Institute of Marine and Coastal Sciences.
presented in Auster (1989).
Both juvenile and adult tautog are
dependent on habitats with structure
or cover, which presumably aids in
protection from predators (011a et al.
1974 and 1979, 011a and Studholme
1975). Tautogs typically become qui-
escent at night, resting in association
with some type of shelter (011a et al.
1974). Smaller fish (subadults <25
cm) may range only a few meters
from that shelter during daytime ac-
tivity, while larger individuals (adults
>30cm) cover a broader area for
foraging, returning to the same gen-
eral shelter area at night (011a et al.
1974).
Declining water temperatures in
the fall trigger an offshore migration
of adults (age 4-1-). An increase in
schooling behavior and night activity
also occurs (011a et al. 1980), perhaps
related to migratory activity. Labor-
atory studies indicate that adults at-
tain a dormant state at temperatures
<5°C. Juveniles (age 2-3) also be-
come torpid in winter, but they re-
main inshore, either partially buried
or in close proximity to structure
(011a et al. 1974). In spring and sum-
mer adults return to inshore habitats.
On hard-bottom reefs off Maryland
529
530
Fishery Bulletin 90(3). 1992
Table 1
Collection sources for Tautoga onitis eggs, larvae
, and juveniles used in analyses of seasonal distribution and growth comparisons.
Years
No. of tows
Stage
Sampling location
sampled
Sampling frequency
or samples
Data source
Eggs
Mullica River, Great Bay,
Little Egg Inlet
1972-75
monthly or bimonthly
462
Milstein and Thomas 1977
Larvae
Nova Scotia to North Carolina
1977-87
monthly or bimonthly
11,438
MARMAP
Great Bay, Little Sheepshead
1989-90
weekly
913
Rutgers Marine Field
Creek
Station Plankton Survey
Early juveniles
Great Bay, Little Egg Harbor
1988-89
biweekly, May-September
436
Sogard and Able 1991
Great Bay; artificial seagrass
1988
weekly, June-September
54
Sogard 1990
Late juveniles
Great Bay, Little Egg Harbor
1988-89
monthly
808
Rutgers Marine Field
Station Trawl Survey
and Virginia (25-35 m in depth) fish-trap catches of
tautog are lowest in summer (Eklund and Targett
1991), perhaps due to inshore migrations of winter
residents. Tagging studies conducted by Cooper (1966)
suggest relatively discrete populations of tautog, with
adults returning to the same spawning location follow-
ing their winter residence offshore.
These prior studies of habitat requirements, behav-
ior, and growth have focused primarily on fishes older
than 1 year. In this paper we concentrate on life-history
aspects for young-of-the-year individuals, particularly
larvae and juveniles that have just recently metamor-
phosed and settled from the plankton. We present in-
formation on spatial and temporal distribution of eggs
and larvae, larval stage duration, juvenUe habitat, daily
growth, and otolith-size/fish-size relationships.
Materials and methods
Reproductive seasonality and
larval distribution
Information on timing of spawning and spatial distribu-
tion of tautog larvae was obtained from three sources
(Table 1). Egg abundances were assessed in plankton
collections during December 1972-December 1975 in
the Mullica River-Great Bay estuary and adjacent
ocean off Little Egg Inlet, New Jersey (Fig. 1). Sam-
pling was conducted with 0.5m and 1.0m diameter
plankton nets and 20 cm and 36 cm diameter bongo
samplers with 0.5 mm mesh. Surface, midwater, and
bottom tows were made with the plankton nets; the
bongos were fished obliquely. Data used in this study
were reanalyzed from a report by Milstein and Thomas
(1977).
Larval occurrences over the continental shelf were
determined from collections made on Marine Resources
Monitoring, Assessment and Prediction (MARMAP)
surveys (Sherman 1988) by the National Marine Fish-
Figure 1
Map of Mullica River-Great Bay estuary system in New
Jersey, and location of plankton sampling site in Little Sheeps-
head Creek.
eries Service (NMFS). Surveys were conducted from
Cape Hatteras, North Carolina, to Cape Sable, Nova
Scotia. Sampling stations are shown in Figure 2; sam-
pling methods are described in Sibunka and Silverman
(1989).
Our third source of plankton data was obtained from
a sampling program at the Rutgers University Marine
Field Station. Aim diameter, 1mm mesh net was
fished at the surface and just above the bottom on night
flood tides in Little Sheepshead Creek, adjacent to
Great Bay (Fig. 1).
Juvenile length-frequency comparisons
Monthly patterns in length-frequency distributions
were assessed from collections across several sites in
the Little Egg Harbor and Great Bay estuaries (Fig.
1) using throw-trap and otter trawl sampling (Table 1).
A throw trap is a 1 m- open box that is thrown onto
Sogard et al.: Early life history of Tautogs onitis in the Mid-Atlantic Bight
531
»^6B ■' ■ ^/6
N 44
76 74
Tautoga onitis
July
Larvae / 10 m'
0
.1-10
11-100
101-1000
//KK
\ 44
: -f 44.
Figure 2
Average monthly distributions of Tautoga onitis larvae based on 11 years (1977-87) of sampling in the Mid-Atlantic Bight. Density
calculations are based on collections within 1 km' blocks.
the desired substrate, with all animals subsequently
removed with aim wide net scraped across the bot-
tom substrate. All throw-trap sampling was conducted
at low tide in shallow (<0.5m at low tide) vegetated
and unvegetated habitats. Further details on the throw-
trap method and sampling schedule are presented in
Sogard and Able (1991). Additional length data were
obtained from tautog collected with throw traps from
artificial seagrass habitats on a shallow sandflat in
Great Bay (Sogard 1990).
Tautog in deeper waters (1-8 m) of the Great Bay-
Little Egg Harbor estuarine system were collected
532
Fishery Bulletin 90(3). 1992
during the Rutgers Marine Field
Station trawl survey. A 4.9m ot-
ter trawl (6.3 mm mesh cod end,
19 mm mesh wings) was towed
for 2 minutes at a total of 14 sta-
tions, which were representative
of a variety of habitat types.
Four replicate trawls were taken
at each station.
\
Otolith increment analysis
More detailed information on
planktonic stage duration, settle-
ment patterns, and growth of
young-of-the-year tautog was
derived from analysis of otolith
increments. To validate a daily
rate of increment formation, the
number of increments following
a tetracycline-induced fluores-
cent mark on the sagitta was
compared with the actual num-
ber of days elapsed. Juvenile
tautog (19-63 mm SL) were im-
mersed for 24 h in a 500mg/L
solution of oxytetracycline dihyd-
rate in natural seawater (20-25
ppt) diluted with distilled water
to about 17 ppt. They were then
held in laboratory aquaria, fed
daily with Artemia, and pre-
served in 95% ethanol after 6-30
days. The sagittae of these indi-
viduals were removed, embedded
in Spurr resin, and polished in
the sagittal plane to the central
primordium on both sides, using
a series of 400-1500 grit sand-
paper and alumina powder (0.3
fim), following the methods of
Secor et al. (1991). The number
of increments following the tetracycline mark was
counted with UV microscopy at 400-1000 x magnifi-
cation.
The degree of correspondence between otolith size
and fish size was determined for 55 juvenile tautog by
comparing radial measurements of the rostrum, post-
rostrum, and antirostrum (Fig. 3) with standard
lengths. Radial measurements were made with an
image analysis system attached to an Olympus BH-2
microscope, using a magnification on the monitor of
160 X or 410 X, depending on the size of the otolith.
The relationship between otolith radial measurements
and length (SL) of the fish was determined by regres-
PR
AR
/
\
Figure 3
Ground and polished left sagitta of a 38.7 mm SL juvenile Tautoga onitis. (upper) Photo-
graphed at lOOx magnification; scale bar = 200fim. R = rostrum. PR = postrostrum.
AR = antirostrum. Tip of the rostrum was partially destroyed during polishing, (lower)
Closeup view of central region (400 x ). Arrow points to transition between pelagic and
demersal stage increments. Scale bar = 50/jm.
sion analysis. Because preliminary analysis of a
matched set of sagittae foimd no significant differences
between left and right radius measurements (paired
comparisons ^tests: n 8, P>0.10 for all three radii),
either sagitta was used in subsequent analyses.
Increments were counted for a series of tautog (n 37,
7.6-62.8 mm SL) collected from early- July through late-
September in 1988. Larval and juvenile increments
were distinguished on the basis of an apparent settle-
ment mark in the sagittae (see below). Increments were
counted independently on three different dates by the
same reader, and the results were averaged. Prelim-
inary counts of matched sagittae found no difference
Sogard et al Early life history of Tautogs onitis in the Mid- Atlantic Bight
533
between left and right (paired comparisons (-test:
P>0.10), with the two sides differing by <2%. Thus,
either sagitta could be used for increment counts.
The mean duration of the planktonic stage was
estimated by the mean number of increments preceding
the settlement mark. Birth and settlement dates were
estimated by subtracting the number of total incre-
ments and juvenile increments, respectively, from the
date of capture. Assuming that initial increment for-
mation occurred at about the time of hatching, as in
other wrasses (Victor 1982), our estimates of birthdates
should correspond within a few days to the date of
hatching.
Juvenile growth rates
We used three independent methods to estimate
growth rates of young-of-the-year juveniles during the
summer. The relationship between otolith age (total in-
crements) and standard length was fit to a linear equa-
tion, using the slope as an estimate of daily growth.
We also examined the progression of mean lengths for
tautog collected on a weekly basis in 1988 (primarily
from artificial seagrass experiments). Weekly mean
lengths were determined and regressed on time, with
the slope of the resulting equation used as a second
estimate of daily growth. Growth rates based on these
two indirect estimates were compared with a third,
direct measurement of individual tautog growth in field
caging experiments by Sogard (In press).
Results
Reproductive seasonality
and larval distribution
In the Great Bay-Mullica River
estuarine system, tautog eggs oc-
curred in plankton collections
from April through August, with
peak abundances in June and
July (Table 2). Initial occurrence
and peak abundance of eggs
were earlier in the Mullica River
than in the bay and adjacent in-
let, suggesting that spawning
began earlier in the season in the
upper part of the estuary, and
continued later in the summer in
the lower estuary and offshore
waters. Tautog larvae in weekly
plankton collections in Great Bay
(Table 1) occurred in July and
August of 1989 (n 12) and July of
1990 (w 9). Larvae were collected
in the offshore MARMAP surveys from May through
October, with a peak in July (Table 3, Fig. 2).
Based on geographic distribution of larvae, spawn-
ing was concentrated in southern New England waters
(Fig. 2). Spawning activity in continental shelf waters
appeared to follow a northward progression through
the summer, beginning as early as May in the southern
part of the region (Table 3).
Daily increment validation
Results of the validc.tion tests indicated that increments
on sagittae of juvenile tautog were deposited on a daily
basis. The slope of the regression comparing the actual
number of days elapsed with the number of increments
following tetracycline marks did not differ from 1
(P>0.05, r2 0.86, Fig. 4). Comparison of sagittal
Table 2
Monthly mean densities (no./lOOO m^ ) of Tautoga onitis eggs |
in Mullica
River, Great Bay, and adjacent Atlantic Ocean off
Little Egg Inlet, New Jersey, December 1972
-December 1975.
Plankton
sampling was conducted throughout the year, but
eggs were
not collected in months not appearing in the table.
Mullica River Great Bay
Atlantic Ocean
April
66 3
0
May
116 726
853
June
169 165
1259
July
22 2221
1984
August
0 13
20
Table 3
Abundance of Tautoga onitis larvae ( x no
cruises, 1977-87, by subarea and month.
./100m') collected during MARMAP survey
Mean abundance is followed by number of
occurrences (2d line) and total number of stations
sampled (3d hne).
Subarea
May
June
July
Aug
Sept
Oct
Georges Bank
0
0.01
1
0
0
0
0
332
152
213
312
144
396
Southern New England
0
0
1.75
26
0.19
18
0.08
4
0.01
1
225
131
231
191
103
224
New Jersey
—
0.07
0.06
0.03
—
—
0
4
9
4
0
0
209
139
176
174
120
143
Delmarva Peninsula
—
0.16
0.02
0.01
—
—
0
6
3
1
0
0
163
82
104
140
126
45
Virginia and North Carolina
0.08
7
0.09
3
0.01
1
0.01
1
0
0
135
66
76
122
103
37
534
Fishery Bulletin 90(3). 1992
30 H
/
• /
' 1
^
o
25 -
2/
b
yi
0)
20 -
^ / 1
M—
10
/
in
15 -
Y
-*-*
45 /
c
T^ /
OJ
/
E
10 ■
1-
/
<u
l-
/•
o
c
b -
/
c
) 5 10 15 20 25
Days after marking
30
Figure 4
Validation of daily deposition of otolith increments
in Taidoga onitis, comparing the number of incre-
ments outside a tetracycline-induced mark with the
number of days since marking. Numbers above data
points are numbers of fish tested; error bars are
SD's. Resulting regression line does not deviate
significantly (P>0.05) from a line of one-to-one
correspondence.
Postrostrum
1000 -|
PR =
-269.7+138.4.SQRT(SL)
0.94 /♦
13
BOO -
o
600 -
SI
400 -
m^*
"o
o
200 -
0 -
■^
700
600
500 -
400
300
200 -
100
0
Antirostrum
AR = -1 1 1.9 + 80. 6*SQRT(SL)
r^ = 0.94
75
800
700
600 -
500
400
300
200
100 -
0
Rostrum
-154.7-t-95.4.SQRT(SL)
r = 0.92
0 25 50 75
Standard length
Figure 5
Regressions comparing otolith radial measurements (see Fig. 3) with standard length
of juvenile Tautoga onitis. Displayed curves fit the square-root equations derived in regres-
sion analysis.
radius measurements with standard length demonstrated
a strong correspondence for all three radii (Fig. 5). For
all three cases, a square-root equation provided the best
fit.
Settlement marks and larval
stage duration
An obvious transition in the ap-
pearance of increments occurred
in the sagittae (Fig. 3). Inner in-
crements were generally more
distinct because they were higher
in contrast, darker in appear-
ance, and more circular than in-
crements outside the transitional
area. In the sagittal plane, outer
increments diverged in morphol-
ogy, with increased deposition
along the eventual axes of ros-
trum, postrostrum, and antiros-
trum. Sagittae of larval tautog
(n 5) were comprised of only the
darker, inner increments. Thus,
we believe this transition in in-
crement contrast and shape
takes place at or near the time of
settlement, when the individual
has completed transformation
and moved from a planktonic to
epibenthic lifestyle. Settlement
marks are a common feature of
labrid otoliths, allowing ready
distinction of larval and juvenile
increments (Victor 1986).
The total number of incre-
ments (separated into larval and
juvenile stages) was counted for
37 individuals collected in the
Great Bay and Little Egg Har-
bor sampling. The number of in-
crements deposited during the
25
Sogard et al Early life history of Tautoga onitis in the Mid-Atlantic Bight
535
pelagic larval stage was remarkably similar, with a
mean of 20.4 (SD 2.7). Assuming that the first incre-
ment is deposited at about the time of hatching, tautog
spend 3 weeks in the plankton before settling to the
benthos. Subtraction of total increments from the date
of collection resulted in a wide spread of estimated birth
(hatch) dates, with a mean of 4 June and a range of
17 April-22 July. These dates are consistent with the
general timing of the collection of eggs and larvae
(Tables 2, 3). Settlement dates, estimated by subtract-
ing only the juvenile-stage incre-
ments from the date of capture,
were correspondingly wide-
spread, with a mean of 25 June
and a range of 6 May-13 August.
creased in number in August collections. By Septem-
ber, the young-of-the-year dominated the trawl samples
while decreasing in throw-trap sampling.
Modal progression of length-frequency distributions
demonstrated relatively rapid growth for both young-
of-the-year and 1 -year-old tautog during the summer
months. In contrast, comparison of young-of-the-year
sizes in October with 1 -year-old lengths in June in-
dicated only minor growth during the fall, winter, and
spring. Juvenile tautog attained a size of 40-100 mm
Juvenile habitat, size
composition, and growth
In throw-trap samples collected
in the shallow waters of Great
Bay and Little Egg Harbor, juve-
nile tautog were collected only on
vegetated substrates, and were
more abundant in sea lettuce
(Ulva lactuca, n 19) than in eel-
grass (Zostera marina, n 2) (So-
gard and Able 1991). Juveniles
<40mm in length were rare in
the deeper waters sampled by
trawls, but the larger young-of-
the-year and 1 -year-old tautog
collected by trawling were most
abundant in eelgrass beds. Of 14
sampling stations throughout
Great Bay and Little Egg Har-
bor, two were in eelgrass habi-
tats. These two stations ac-
counted for 69% of the 235 tau-
tog collected by trawling in 1988
and 1989.
Combined length-frequency
data from throw-trap sampling
and trawling efforts suggested
that most tautog in the Great
Bay-Little Egg Harbor system,
based on these sampling techni-
ques, belonged to one of two
year-classes (Fig. 6). Young-of-
the-year first appeared in July,
primarily in the shallower (< 1 m)
habitats sampled by throw trap-
ping. In the deeper areas (> 1 m)
sampled by otter trawl, larger
young-of-the-year fishes in-
60'
40
20
60
40
20
60
W
_i
< 40-
Q
> 20-
Q
Z
fe 60
EC
m 40
m
^ 20
60
40
20 H
NOVEMBER,
MARCH, APRIL, MAY
■ PLANKTON NET
n THROW TRAP
D TRAWL
JUNE
^FHr
JULY
rln^-,- , r-n^EPqnrii— ir-,r— I
II
FJHr
AUGUST
B.
SEPTEMBER
H^p
gf , ^7^i=pi:pr~i
a
60
40-
20
OCTOBER
^^ — ^ ■ F^^ F^ ^
5 25 45 65 85 105 125 145 165 185 205 225 245 265
STANDARD LENGTH (mm)
Figure 6
Length-frequency distributions of Tautoga onitis collected in the Little Egg Harbor-
Great Bay estuarine system. Plankton samples were collected on a weekly basis
throughout the year. Throw-trap samples were collected from shallow habitats (< 1 m
at low tide), May-October; trawl samples were collected from deeper habitats (1-8 m)
monthly, with no tautog collected during December-February.
536
Fishery Bulletin 90(3). 1992
65 -|
SL = -30.49 -1- 7.20 • SORT(Age) ,
60 -
r^ = 0.75
55 -
• • ,
• . ^/
50 ■
-C
y^
■^ 45 ■
m y^
cn
• X
C 40 -
' »/•
(U
.• X-
35 -
• */^ • •
XI
/ •
o ^°-
"a
c " -
/
B 20-
/•. *
C/1
15 -
/
p •
r*
10 -
5 -
/
0 10 20 30 40 50 60 70 80 90 100110120130140
Total increments
Figure 7
Total number of increments vs. standard length for 37
juvenile Tautoga onitis. The plotted square-root equa-
tion provided the best fit in regression analyses.
SL in their first growing season, with a modal size of
75 mm in October (Fig. 6). One-year-old fish reached
a size of 110-170 mm SL by the end of their second sum-
mer, with a modal size in September of 155 mm.
Comparison of otolith ages (total increment counts)
with standard lengths provided a general estimate of
juvenile growth. The resulting relationship was best
described by a square-root equation (Fig. 7), indicating
a slight decline in absolute growth rate with age. If the
data are fit to a linear equation, a slope of 0.47 results,
thus estimating an average rate of 0.47 mm/day dur-
ing the early juvenile stage. Substantial variability was
evident, especially among older individuals (Fig. 7).
To obtain an estimate of growth based on length-
frequency distributions, we compared length with the
date of capture for 236 juveniles collected only by throw
traps in 1988. When the mean length each week was
regressed on time (Julian date), the resulting slope pro-
vided an estimated growth rate of 0.52 mm/day (Fig. 8).
Discussion
Spawning patterns
Based on the seasonal occurrence of eggs and larvae,
the peak spawming period for tautog in the Mid-Atlantic
Bight and inshore New Jersey waters is during the
summer. Spawning appears to follow a geographical
progression, beginning earlier in the southern part of
the region. Consistent with this pattern, Eklund and
Targett (1990) report that gonosomatic indices of adult
D
-o
C
D
JUL 10 JUL 30 AUG 19 SEP 8 SEP 28
Date of collection
Figure 8
Mean and range in standard length of juvenile
Tautoga onitis collected on a weekly basis in 1988.
All fish were collected with throw traps, primarily
from artificial seagrass substrates. The regression line
was fit to means for each sampling date (n 13).
Numbers above ranges are the number of tautog col-
lected each week.
tautog off Maryland and northern Virginia are highest
in May.
The egg collections in New Jersey and high egg and
larval abundances in areas such as Narragansett Bay
(Bourne and Govoni 1988) demonstrate that tautog
spawn primarily inside estuaries or nearshore waters.
The MARMAP collections indicate that spawning
activity involves offshore continental shelf waters as
well, since all of the tautog larvae obtained during
MARMAP surveys were preflexion stage.
Otolith deposition patterns
Otolith increments of juvenile tautog can be reliably
used to obtain valuable age and growth information.
The strong correspondence of otolith size (based on
radial measurements) with fish size suggests that ac-
curate back-calculation of size-at-age is possible. In-
crements on the sagittae are deposited on a daily basis
and can be readily separated into planktonic and
demersal stages, due to the distinct contrast in micro-
structure at the time of settlement. We did not, how-
ever, test increment deposition rates under conditions
of poor or negative growth. These conditions have
resulted in less than daily increments in other species
(Geffen 1982, Lough et al. 1982, McGurk 1984, Alhos-
saini and Pitcher 1988, Siegfried and Weinstein 1989,
Sogard et al : Early life history of Tautoga onitis in the Mid-Atlantic Bight
537
Sogard 1991, Szedlmayer and Able In press), and we
caution that this may also be the case for tautog
otoliths.
Settlement
Increment coimts preceding the settlement mark aver-
aged 20.4, suggesting larvae spend approximately 3
weeks in the plankton before settling to the benthos.
This estimate of larval stage duration is similar to that
derived by Victor (1986) for a sample of five tautog
(x 25.4). The planktonic stage for tautog is relatively
short compared with other labrids; Victor (1986) esti-
mated average larval durations of 17-104 days for
other wrasse species.
The earliest estimated date of settlement, based on
otolith increments, was earlier than the first collections
of juveniles with throw traps, suggesting that tautog
were not available to the collecting gear during and im-
mediately after settlement. The smallest juveniles for
which we have otolith information were 7. 6-13. 2mm
SL and had 11-28 increments (i 16.3) deposited after
the settlement mark. Victor (1983) reported that new-
ly settled wrasses of the species Halichoeres bivitta-
tus bury in sediments immediately following settlement
from the plankton and remain buried for an average
of 5 days. We do not know if a similar behavior occurs
in Tautoga onitis.
Juvenile habitat utilization
Our collection of juvenile tautog primarily in vegeta-
tion is in accord with prior studies, which demonstrated
an association with structured habitats (011a et al. 1974,
011a et al. 1979). Several studies comparing eelgrass
vs. unvegetated substrates noted significantly higher
densities of tautog in grassbeds, with few or no tautogs
collected on bare substrates (Briggs and O'Connor
1971, Orth and Heck 1980, Weinstein and Brooks 1983,
Heck et al. 1989). The importance of sea lettuce as a
nursery habitat has received only limited attention,
although Nichols and Breder (1926) mentioned its at-
traction to small juvenile tautog. In a separate study
that also quantitatively compared sea lettuce and
eelgrass habitats in New Jersey, using suction sam-
pling. Able et al. (1989) also noted higher abundances
of early juvenile tautog in sea lettuce patches than in
eelgrass, although the total catch was relatively small.
Larger juveniles make extensive use of rocky reef
habitats (011a et al. 1979). The importance of hard
substrates for newly settled tautog has not been
examined.
All of the smaller juvenile tautog (<35mmSL) that
we collected were from sea lettuce patches or artificial
seagrass plots. These individuals were a brilliant green
in color. As noted by Nichols and Breder (1926), this
color closely matches that of Ulva lactuca, but pre-
sumably would be conspicuous on a bare sand sub-
strate. In our sampling, these early juveniles were ab-
sent from eelgrass beds. The larger juveniles collected
during trawl sampling (in eelgrass and other habitats)
had a dark, mottled coloration similar to that of the
adults as depicted by Bigelow and Schroeder (1953).
Over the course of our summer sampling, we ob-
served a shift in concentration of young-of-the-year
from the shallow areas sampled by throw traps to
deeper waters sampled by otter trawl (Fig. 6). This
shift suggests that newly-settled juveniles concentrate
in shallow waters, moving to deeper sections of the
estuary with growth.
Although trawling was conducted year round, in-
dividuals of age 1 or older were common only from June
through September. This pattern could result from in-
accessibility to the gear. Some individuals may move
out of shallow habitats in the fall to deeper areas of
the estuary with more stable sheltering refuges.
Behavioral responses displayed by tautog in cold
temperatures, i.e., dormancy and remaining in close
contact with sheltering structure (011a et al. 1974) or
burying in sediments (011a et al. 1979), would also
reduce capture rates in winter. In addition, some in-
dividuals may leave the estuary to winter offshore,
although 011a et al. (1974) suggest that most tautogs
less than 4 years old remain inshore.
Tautogs older than 1 year may be more abundant in
the estuary than trawl catches would indicate. Larger
individuals inhabit holes and crevices of eroding salt
marsh banks, and other physical structures such as
pilings and rock jetties, where they would not be
available to trawling gear.
Juvenile growth
Our estimates of sizes attained by juvenile tautog at
the end of the first and second summers are larger than
the mean lengths (TL) at ages 1 and 2 calculated for
Rhode Island tautog from opercular bone annuli
(Cooper 1967). Warmer temperatures in New Jersey
may support faster mean growth rates than in Rhode
Island. In addition, more southern estuaries in the
tautog's range have an extended summer season,
allowing both earlier spawning in the spring and con-
tinued rapid growth prior to declining water temper-
atures in the fall.
Analysis of length progressions and otolith ages
resulted in two similar estimates of natural growth
rates for juvenile tautog (0.52 mm/day and 0.47mm/
day). Individual growth rates of juvenile tautog were
also measured in the field in caging experiments
(Sogard In press). To summarize results of Sogard
538
Fishery Bulletin 90(3), 1992
(In press), growth rates varied significantly, depending
on location in the estuary and habitat type (vegetated
or unvegetated). Across four experiments and a total
of 141 tautog, growth averaged 0.18 mm/day, with a
range of -0.47 to +0.84 mm. At the site (Great Bay
1) and habitat (sea lettuce) supporting the fastest
growth, the mean rate was 0.45 mm/day. Thus, length-
frequency patterns and otolith ages reported in this
study provided growth estimates that were higher than
the overall average in caging experiments but com-
parable to rates for tautog caged in the best habitats.
Based on throw-trapping results, juvenile tautog were
rare at those sites and habitats where growth in cages
was poor (Sogard In press). Thus, for the areas where
tautog were likely to be common, the directly measured
growth rate in cages was comparable to rates indirectly
calculated for unrestrained fishes. The three methods
together estimated a rapid growth rate of about 0.5
mm/day for southern New Jersey estuaries during the
first summer.
Acknowledgments
We thank Dan Roelke and additional Rutgers Marine
Field Station personnel for their assistance in field
sampling. Dave Witting and Steve Szedlmayer coor-
dinated the plankton sampling and trawling collections,
respectively, in Great Bay. Susan Kaiser provided
valuable help in data compilation and analysis. We are
especially grateful to Rich McBride for his assistance
with the otolith validation studies. Financial support
was provided through New Jersey Sea Grant NA89AA-
D-SG057 (NJSG-92-258).
Citations
Able, K.W, K.A. Wilson, and K.L. Heck Jr.
1989 Fishes of vegetated habitats in New Jersey estuaries:
Composition, distribution, and abundance based on quantitative
sampling. Publ. 1041, Cent. Coastal & Environ. Stud.,
Rutgers Univ., New Brunswick, 38 p.
Alhossaini, M., and T.J. Pitcher
1988 The relation between daily rings, body growth and en-
vironmental factors in plaice, Pkuromctes platessa L., juvenile
otoliths. J. Fish Biol. 33:409-418.
Auster, P.J.
1989 Species profiles: Life histories and environmental re-
quirements of coastal fishes and invertebrates (North Atlan-
tic and Mid-Atlantic)-tautog and cunner. U.S. Fish Wildl.
Serv. Biol. Rep. 82, 13 p.
Bigelow, H.B., and W.C. Schroeder
1953 Fishes of the Gulf of Maine. Fish. Bull., U.S. 53, 577 p.
Bourne, D.W., and J.J. Govoni
1988 Distribution of fish eggs and larvae and patterns of water
circulation in Narragansett Bay, 1972-1973. In Weinstein,
M.P. (ed.), Larval fish and shellfish transport through inlets,
p. 132-148. Am. Fish. Soc. Symp. 3, Bethesda.
Briggs, P.T., and J.S. O'Connor
1971 Comparison of shore-zone fishes over naturally vegetated
and sand-filled bottoms in Great South Bay. N.Y. Fish Game
J. 18:25-41.
Colton, J.B., W.G. Smith, A.W. Kendall Jr.. P.L. Berrien, and
M.P. Fahay
1979 Principle spawning areas and times of marine fishes. Cape
Sable to Cape Hatteras. Fish. Bull., U.S. 76:911-915.
Cooper, R.A.
1966 Migration and population estimation of the tautog, Tau-
toga onitis (Linnaeus), from Rhode Island. Trans. Am. Fish.
Soc. 95:239-247.
1967 Age and growth of tautog, Tautoga onitis (Linnaeus),
from Rhode Island. Trans. Am. Fish. Soc. 96:134-142.
Eklund. A.M., and T.E. Targett
1990 Reproductive seasonality of fishes inhabiting hard bot-
tom areas in the Middle Atlantic Bight. Copeia 1990:
1180-1184.
1991 Seasonality of fish catch rates and species composition
from the hard bottom trap fishery in the Middle Atlantic Bight
(U.S. east coast). Fish. Res. (Amst.) 12:1-22.
Geffen, A.J.
1982 Otolith ring deposition in relation to growth rate in her-
ring (Clupea harengus) and turbot {Scophthalmus maxirmis)
larvae. Mar. Biol. (Berl.) 71:317-326.
Grover, J.J.
1982 The comparative feeding ecology of five inshore, marine
fishes off Long Island, New York. Ph.D. diss., Rutgers Univ.,
New Brunswick, 197 p.
Heck, K.L. Jr., K.W. Able, M.P. Fahay, and C.T. Roman
1989 Fishes and decapod crustaceans of Cape Cod eelgrass
meadows: Species composition and seasonal abundance pat-
terns. Estuaries 12:59-65.
Kuntz, A., and L. Radcliffe
1918 Notes on the embryology and larval development of
twelve teleostean fishes. Bull. U.S. Bur. Fish. 35:87-134.
Lough, R.G., M. Pennington, G.E. Bolz, and A. A. Rosenberg
1982 Age and growth of larval Atlantic herring, Clupea haren-
gus L., in the Gulf of Maine— Georges Bank region based on
otolith growth increments. Fish. Bull., U.S. 80:187-199.
McGurk, M.D.
1984 Ring deposition in the otoliths of larval Pacific herring
Clupea harengus pallasi. Fish. Bull., U.S. 82:113-120.
Milstein, C.B., and D.L. Thomas
1977 Summary of ecological studies for 1972-1975 in the bays
and other waterways near Little Egg Inlet and in the ocean
in the vicinity of the proposed site for the Atlantic Generating
Station, New Jersey. Bull. 18, Ichthyol. Assoc, Ithaca. NY,
757 p.
Nichols. J.T., and CM. Breder Jr.
1926 The marine fishes of New York and southern New Eng-
land. Zoologica (NY) 9:1-192.
011a, B.L., and A.L. Studholme
1975 The effect of temperature on the behavior of young
tautog, Tautoga onitis (L.). In Barnes, H. (ed.), Proc, Ninth
Euro. Mar. Biol. Symp., p. 75-93. Aberdeen Univ. Press,
Aberdeen.
Sogard et aL: Early life history of Tautogs onitis in the Mid-Atlantic Bight
539
011a, B.L.. A.J. Bejda, and A.D. Martin
1974 Daily activity, movements, feeding, and seasonal occur-
rence in the tautog, Tautoga onitis. Fish. Bull.. U.S. 72:27-35.
1979 Seasonal dispersal and habitat selection of cunner,
Tautogolabrus adspersus, and young tautog, Tautoga onitis,
in Fire Island inlet, Long Island, New York. Fish. Bull., U.S.
77:255-261.
011a, B.L., A.L. Studholme, A.J. Bejda, and C. Samet
1980 Role of temperature in triggering migratory behavior of
the adult tautog Tautoga onitis under laboratory condi-
tions. Mar. Biol. (Berl.) 59:23-30.
Orth, R.J., and K.L. Heck Jr.
1980 Structural components of eelgrass (Zostera marina)
meadows in the lower Chesapeake Bay— Fishes. Estuaries
3:278-288.
Secor, D.H., J.M. Dean, and E.H. Laban
1991 Manual for otolith removal and preparation for micro-
structural examination. Tech. Rep. BI1191-1, Belle Baruch
Inst.. Columbia, SC, 85 p.
Sherman, K.
1988 Ichthyoplankton surveys: A strategy for monitoring
fisheries change in a large marine ecosystem. In Smith, W G.
(ed.). An analysis and evaluation of ichthyoplankton survey data
from the northeast continental shelf ecosystem, p. 3-9. NCAA
Tech. Memo. NMFS-F/NEC-57, NMFS Northeast Fish. Sci.
Cent., Woods Hole.
Sibunka, J., and M. Silverman
1989 MARMAP surveys of the continental shelf from Cape
Hatteras, North Carolina to Cape Sable, Nova Scotia (1984-
1987). Atlas No. 3. Summary of operations. NOAA Tech.
Memo. NMFS-F/NEC-68, NMFS Northeast Fish. Sci. Cent.,
Woods Hole, 197 p.
Siegfried, R.C. Ill, and M.P. Weinstein
1989 Validation of increment deposition in the otoliths of spot
(Leiostomus xanthurus). Estuaries 12:180-185.
Sogard, S.M.
1990 Parameters of habitat quality for epibenthic fishes and
decapod crustaceans in New Jersey estuaries. Ph.D. diss.,
Rutgers Univ., New Brunswick, 195 p.
1991 Intei-pretation of otohth microstructiire in juvenile winter
flounder {Pseudopleuronectes americanus): Ontogenetic
development, daily increment validation, and somatic growth
relationships. Can. J. Fish. Aquat. Sci. 48:1862-1871.
In press Variability in growth rates of juvenile fishes in dif-
ferent estuarine habitats. Mar. Ecol. Prog. Ser.
Sogard, S.M., and K.W. Able
1991 A comparison of eelgrass, sea lettuce macroalgae, and
marsh creeks as habitat for epibenthic fishes and decapods.
Estuarine Coastal Shelf Sci. 33:501-519.
Szedlmayer, S.T., and K.W. Able
In press Daily increment formation in larval and juvenile sum-
mer flounder, Paralichthys dentatas. Can. J. Fish. Aquat. Sci.
Victor, B.C.
1982 Daily otolith increments and recruitment in two coral-
reef wrasses, Thalassorna. bifasciatum and Halichoeres bivit-
tatus. Mar, Biol. (Beri.) 71:203-208.
1983 Settlement and larval metamorphosis produce distinct
marks on the otoliths of the slippery dick, Halichoeres bivit-
tatus. In Reaka, M.L. (ed.). Results of workshop on coral reef
ecology. Philadelphia, 1983, sponsored by Am. Soc.
Zool. Symp. Ser. Undersea Res., NOAA Undersea Res. Prog.
l(l):47-52.
1986 Duration of the planktonic larval stage of one hundred
species of Pacific and Atlantic wrasses (family Labridae). Mar.
Biol. (Beri.) 90:317-326.
Weinstein, M.P., and H.A. Brooks
1983 Comparative ecology of nekton residing in a tidal creek
and adjacent seagrass meadow: Community composition and
structure. Mar. Ecol. Prog. Ser. 12:15-27.
Williams, G.C.
1967 Identification and seasonal size changes of eggs of the
labrid fishes, Tautogolabrus adspersus and Tautoga onitis, of
Long Island Sound. Copeia 1967:452-453.
Abstract. - Heceta Bank is a
large reef on the edge of the central
Oregon continental shelf that sup-
ports a wide variety of commercial
fisheries. Using the research sub-
mersible Delta, we studied fish abun-
dances on Heceta Bank and the rela-
tionship between species composition
of fish assemblages and bottom
types. Cluster analysis indicated that
fish assemblages were most unique
on mud, boulder, rock ridge, mud
and cobble, and mud and boulder
substrates. Rockfishes, particularly
pygmy Sebastes vnlsoni, sharpchin S.
zacentrus, rosethom S. helvomacula-
tus, and yellowtail S.flavidus, were
the most abundant fishes and dom-
inated all substrates except mud,
where Dover sole Microstomics paci-
ficus and zoarcids Lycodes pacificus
were most abundant.
Principal component analysis
(PCA) and canonical correlation anal-
ysis (CCA) were used to determine
the sources of variation within the
data. PCA demonstrated that habi-
tat variability was a fundamental
cause of heterogeneity among fish
assemblages. In contrast, CCA
showed how species occurrences
were related to specific substrates.
Ontogenetic shifts in behavior and
substrate preference occurred in
pygmy rockfish. Small juveniles
often formed dense schools above the
bank's shallower rocky ridges.
Larger individuals occurred in non-
polarized assemblages on the bottom
in cobble and boulder fields.
Fish-habitat associations on a
deep reef at the edge of the
Oregon continental shelf
David L. Stein
Department of Oceanography, Oregon State University, Corvallis, Oregon 9733 1 -29 1 4
Brian N. Tissot
Department of Zoology, Oregon State University, Corvallis, Oregon 97331-2914
Present address: Department of Biology, College of Arts and Sciences
University of Hawaii, Hilo, Hawaii 96720
Marl< A. Hixon
Departments of Oceanographiy and Zoology, Oregon State University
Corvallis, Oregon 97331-2914
William Barss
Oregon Department of Fish and Wildlife. Newport, Oregon 97365
Manuscript accepted 11 May 1992.
Fishery Bulletin, U.S. 90:540-551 (1992).
Heceta Bank is a major commercial
fishery zone off central Oregon. It
supports a wide variety of fisheries:
a demersal trawl fishery for many
species of flatfishes; a longline fish-
ery for halibut Hippoglossus stenole-
pis; midwater trawl and vertical long-
line fisheries for rockfishes (Sebastes
spp.); a midwater trawl fishery for
hake Merluccius productus; and dur-
ing upwelling, a troll fishery for
salmon (Oncorhynchus spp). Despite
its importance to commercial fisher-
ies, little was known about Heceta
Bank prior to our 1987 submersible
studies (Pearcy et al. 1989). From
those exploratory dives we learned
that (1) the bank is composed of
diverse substrates, each supporting
fish assemblages differing in species
composition and relative abimdances;
(2) shallow areas of the bank act as
a nursery for juvenile rockfishes; and
(3) commercially valuable species of
rockfish are associated with the shal-
low bank top in untrawlable areas,
which thus serve as refugia from
most commercial fishing.
Our 1987 studies focused on initial
exploration and description of the
Bank. Here we report results from
our 1988 submersible-based surveys
in which we again studied the fishes
occurring on the Bank, concentrating
specifically on their associations with
various bottom types. We selected
sampling stations that represented
the range of habitats described by
Pearcy et al. (1989) (Fig. 1). Our ob-
jectives were to (1) further develop
methods of collecting and analyzing
data that could be gathered from a
submersible to study rocky banks;
(2) identify the species occurring on
Heceta Bank and estimate their rela-
tive and absolute abundances; (3)
obtain detailed information about the
variability of bottom types occurring
within each station; and (4) assess the
composition of fish assemblages in
relation to different bottom types.
Methods
Data collection
We used the submersible Delta to
make 18 dives at six stations on
Heceta Bank in September 1988
(Fig. 1). These stations represented
all substrates and depths within
range of the submersible (to 366 m).
540
Stem et al Fish-habitat associations at edge of Oregon continental shelf
541
44° 10' N
44 N
At each station we made three
dayhght dives, each by a differ-
ent observer (DS, MH, WB).
Dives began and ended at least
an hour after dawn and an hour
before sunset, respectively, min-
imizing the possible effects of
diurnal migration by fishes. Al-
most all dives at each station
were made on the same day.
Our methods basically follow
those developed for use by scuba
divers working on shallow reefs
(Brock 1954, Ebeling 1982). Each
observer made two 30-minute
visual belt transects during each
dive, yielding 6 transects per sta-
tion (i.e., a total of 36 transects,
12 by each observer). To deter-
mine if there were any discern-
ible effects from lights or motor
noise of the submersible on the
fishes, a 10-minute rest was
taken with all lights and machin-
ery off between each pair of tran-
sects. To minimize variability
caused by within-transect sub-
strate changes, all transects
within a station started as close-
ly as possible at the same posi-
tion, as determined by Loran C.
However, due to limits in the ac-
curacy of Loran C and variabil-
ity in current speed and direc-
tion, transects within stations
were usually 100-300 m apart.
The observer in the submer-
sible viewed the bottom through
a single bow port which limited
observation to about a 90° view.
Submersible altitude above bot-
tom (at height of observers' eyes
from the bottom) was held as
closely as possible to 2 m, as mea-
sured by an altimeter on the vehicle and by a chain
suspended from the submersible (see below). Widths
of the viewing path at altitudes of 0.5-2.0 m were deter-
mined empirically by "flying" the vehicle at right
angles across a decimeter-striped 3 m pole placed on
the bottom and noting the length of the pole visible to
the observer between two fixed points on the submer-
sible. At 2m altitude, the transect width was 2.3m.
Thus, the density of fishes (no./m^) was calculated as
the number of fishes seen along a transect divided by
2.3 times the transect length in meters. To aid in
124 50' W
124° 40' W
Bathymetric chart
sampled.
Figure 1
of Heceta Bank, Oregon, indicating locations of the six stations
estimating fish length and maintaining vehicle altitude,
an ~0.4m long fiberglass rod, striped in alternating
black and white decimeters, was hung by chain from
the vehicle within the observer's view. Chain length
was adjusted so that when the rod was just above the
bottom, the observer's altitude was 2m.
The goal of the observer during a dive was to iden-
tify, count, and estimate the lengths (to the nearest
decimeter) of all fishes seen along the transect. Fishes
were categorized into "schooling" when five or more
individuals formed a polarized group (i.e., all fish
542
Fishery Bulletin 90(3). 1992
moving syncronously in the same direction). Non-polar-
ized aggregations or solitary individuals were consid-
ered "non-schooling." Data w^ere collected by contin-
uous audio tape recordings of the observer during
transects, continuous video records (also including
audio, time, and date), and 35 mm still photographs
automatically triggered every 30 seconds. We used a
PhotoSea 1000 35 mm still camera and a PhotoSea 2000
video camera, both on fixed mounts outside the vehi-
cle. The video camera was mounted on the starboard
bow of the submersible and recorded a field of view that
partially overlapped that of the observer within the
submersible. The audio track of the videotape recorded
the observers comments which allowed real-time inte-
gration of fish observations and bottom-type descrip-
tions (see below). Visibility always extended at least
to the limits illuminated by the lights (i.e., ~3m or more
except where limited by topography). Immediately
following each dive, data were entered by computer
into a relational database system and verified against
the audio tapes.
We tried to minimize inherent biases of submersible
studies as suggested by Ralston et al. (1986), such as
fishes not seen or unidentified, diurnal variability, and
effects of vehicle on fishes. Through a detailed analysis
of fish and bottom-type observations recorded in the
continuous video coverage of each transect, several
observer-related factors affecting data collection were
discovered. First, the diving observer usually noted
fishes first, then bottom type. When fishes were pres-
ent coincidentally with a substrate change, fish records
were frequently correlated with the wrong bottom
type. Second, observers tended to record substrate
types based upon larger (high-relief) features rather
than small (low-relief) ones, even when the smaller ones
were preponderant. Apparently, boulders impressed
observers more than cobble or mud, even when the
latter were most abundant. Neither of these sources
of error was intuitively obvious or suspected. If left
uncorrected, these errors would have changed the
apparent fish-substrate associations.
Due to these inherent biases, we extracted data on
bottom types from the videotape record of each tran-
sect. In order to standardize any bias in the evaluation
of bottom types, a single observer (BT) reviewed all
videotapes. Dominant substrates were categorized
using a two-code combination of seven possible cate-
gories: mud (code M), sand (S), pebble (P, diameter
<6.5cm), cobble (C, >6.5 and <25.5cm), boulders (B,
>25.5cm), flat rock (F, low vertical relief), or rock
ridges (R, high vertical relief). Substrate was noted as
either "primary" if it covered at least 50% of the area
viewed (the first code), or "secondary" if it covered
more than 20% of the area viewed (the second code).
For example, a mud-boulder bottom type (code MB)
consisted of at least 50% cover by mud with at least
20% cover by boulders. In contrast, a mud bottom (MM)
consisted of >80% cover by mud.
We defined each transect segment of uniform bot-
tom type as a "habitat patch." Transects within sta-
tions were therefore represented by a series of habitat
patches defined by the frequency of substratum change
along a transect. As a result, the size of habitat pat-
ches varied both within and among transects in con-
junction with the area of uniform bottom types. The
average habitat patch measured 150.8 m^ (SE 15.4 m^,
n 524).
Data analysis
Although data were collected on all observed fish, data
analysis focused on the distribution and abundance of
non-schooling fishes rather than schooling fishes,
because data for the former were more reliable. First,
due to the lack of a manipulator on the submersible,
we were unable to collect schooling fishes, which were
typically small and unidentifiable, to obtain voucher
specimens for positive identification. Second, school-
ing species were generally more abundant above the
bottom in midwater and were not common in the tran-
sect path.
We tested for statistical differences among stations
and observers in non-schooling fish abundance using
a nested two-factor analysis of variance (ANOVA).
Thirty-minute transect segments served as nested
replicates. Sample variances were examined for
homogeneity using Bartlett's test (Sokal and Rohlf
1981) prior to using the ANOVA. Because the raw data
were heteroscedastic, the analysis examined the log-
transformed total abundance of non-schooling fish
per m^.
To examine the variation in fish assemblages among
transects, data were analyzed using principal compo-
nent analysis (PCA). The PCA was an R-mode analysis
of the variance-covariance matrix based on the log-
transformed abundance of non-schooling fish per m^.
By definition, the axes examined in PCA are statistical-
ly independent of on another (Pimentel 1979). Rare
species were eliminated from analysis by selecting only
species present on at least 10 of the possible 36
transects. A total of 30 taxa met this criterion and were
used in the analysis.
To examine the overall similarity of fish assemblages
occurring on different substrates, data were analyzed
using hierarchical cluster analysis. The analysis was
limited to 21 species which were present on at least
12 of the possible 36 transects. The data in this analysis
were the log-transformed total number of individuals
per m^ of each species that occurred on each substrate
combination. A dendrogram was constructed using
Stem et al Fish-habitat associations at edge of Oregon continental shelf
543
Euclidean distance as a measure of similarity and the
group-average clustering method (Pimentel 1979).
To examine specific associations between fish abun-
dance and bottom-type characteristics, data were ex-
amined using canonical correlation analysis (CCA).
CCA maximizes correlations among two sets of vari-
ables while it minimizes correlations within sets
(Pimentel 1979). We used CCA to quantify associations
between abundances of non-schooling fish species (data
set 1) and bottom types (data set 2). Our primary goal
was to extract meaningful, natural associations be-
tween fishes and habitat factors potentially influenc-
ing their distribution and abundance. CCA estimates
these associations using four metrics (Pimentel 1979).
First, the canonical correlation measures the overall
association between the two data sets. Second, the
redundancy coefficient measures the amount of overall
variation in one data set as predicted by the other.
While the canonical correlation coefficient describes the
goodness-of-fit of the two data sets, which can be in-
fluenced by a single high correlation between one
variable in each data set, the redundancy coefficient
measures the extent of overlap in the variation of the
two data sets. Third, the variable loadings indicate
which variables are correlated on a particular axis. The
fourth metric, canonical variate scores, measures the
contribution of each sampHng unit (in this analysis, the
habitat patch) to the fish-habitat pattern depicted on
each axis. Canonical variate scores are derived for each
data set: scores for the habitat data indicate the rela-
tive cover of specific bottom types on each axis, while
scores for the fish data indicate the relative abundance
of specific fish on each axis. Canonical variate scores
derived from CCA represent a powerful way to mea-
sure the abundance of fish in reference to habitat type.
In essence, the method controls for the effects of sam-
pling across a range of different habitats, and thus in-
creased our power to detect meaningful spatial varia-
tion in fish abundance.
Data for CCA were derived using observations of
habitat patches, which were discrete segments of uni-
form bottom type within each transect {n 524 segments
for all transects). For each habitat patch, the abun-
dances of 21 fish species were tabulated relative to the
summed total area (in m-) comprised by the habitat.
For mixed bottom types, the total patch area was ap-
portioned 80% to the primary substrate and 20% to the
secondary substrate.
Results
The six stations represented a wide variety of sub-
strates, ranging from shallow rocky ridges separated
by sand, to intermediate-depth cobble and boulder
fields, to deep mud and pebble bottoms (Figs. 1,2). Sta-
tions 1 and 3 (shallow bank tops) were rocky ridges at
60-80 m depth separated by sand and boulder- filled
valleys; station 2 (bank saddle) was primarily mud with
interspersed cobble at 150-200 m; station 4 was mud,
ridge, and cobble at 145-175m; station 5 was mud at
250-340 m; and station 6 was boulder and cobble
grading into mud at 200-270 m. Because transects
were always run into the current to insure control-
ability of the vehicle, distance traveled along the
transects was not standardized, but was in the range
467-2367m {x length 1357m, SE 460m).
Nested two-way ANO VA of transects and observers,
based on the relative abundances of all non-schooling
fish species summed, indicated significant differences
among stations (F 6.22, df5,18, P<0.01), but not
among observers (F 1.39, df 2,18, P>0.05), or in inter-
actions among observer transects and stations (F 0.48,
df 10,18, F>0.05). A Student-Newman-Keuls multiple-
range test separated the mean number of non-schooling
fish at stations into two subgroups: station 4, where
fish were most abundant at 2.09 fish/m-, and all other
stations, which ranged between 1.84 fish/m^ (station
6) and 0.31 fish/m^ (station 1).
Species identified: Number and size
We identified 38 taxa to species in our 1988 dives. This
represents a 23% increase over the 31 species identified
in 1987 (Pearcy et al. 1989). The increase was due
primarily to species that were uncommon, suggesting
that we identified most or all of the numerically im-
portant species on Heceta Bank. There were distinct
differences in taxonomic composition and abundances
between non-schooling and schooling fishes. About 89%
of the non-schooling fishes seen were identified to
species; fewer than 2% were not identified to family
or genus. All schooling fishes seen were Sebastes. Of
these, only 49% were identified to species; the re-
mainder were identified to genus only (Table 1). Most
of the schooling fish were small or juvenile fish that
we could not identify without voucher specimens.
We counted 10,102 non-schooling fish, ranging from
3829 individuals of pygmy rockfish Sebastes wilsoni to
one individual in each of ten species (Table 1). School-
ing fishes comprised 22,470 individuals, over 50% of
which (12,820) were unidentified small Sebastes. The
most abundant identifiable schooling species was again
the pygmy rockfish, with 8390 individuals. The least-
abundant schooling species was widow rockfish
Sebastes entomelas with 20 counted. The total number
of fish schools seen (all species) was 145, ranging from
70 pygmy rockfish schools to one school of widow
rockfish. The number of individuals per school ranged
from about 10 to 330.
544
Fishery Bulletin 90(3), 1992
>
o
o
-I—'
c
o
(U
100 J Station 1. n
75-
101
^
_KfS_^^_
-EfSL-
BC
MP
100 1
r Station 2, n = 46
75-
F:r«=q
50-
^
25-
0-
1 1 1 1 1 —
^ , ^f^
^
1
MC
100 -|
rSt
ation 3, n = 1 73
75-
50-
25-
=:^
n-
^
L^ — 1_
— h s=1== 1^ f— 1
BS CB
CM
SC
S0--
Station 4, n = 1 34
^
_s;sL.
_KJSl_
F* p^l ■
SB
BC
BS
CB CM
SC
100-,
[-Station 5, n = 20
— ^
75-
■
^
50-
■
^
25-
^
oJ
1 1 1 1 1 1 1 1 1 1
^
1 1
75
50--
Station 6, n = 50
^ F^ I ^ E^S
BB
BC
BS CB
Bottom Type
Figure 2
Percent cover of the ten most abundant bottom-type combinations at Heceta
Bank, Oregon. RR = rocl< ridge; BB = boulder; BC = boulder-cobble; BS
= boulder-sand; CB = cobble-boulder; CM = cobble-mud; SC = sand-cobble;
MC = mud-cobble; MP = mud-pebble; MM = mud (see text for a descrip-
tion of bottom-type codes), n = number of habitat patches sampled per
station.
Among non-schooling species, the estimated total
length ranged from 105 cm (dogfish shark Squalus
acanthias) to a mean of 12 cm (unidentified small Sebas-
tes) (Table 1). Many smaller fishes were seen, but could
be identified only to genus {Sebastes juveniles) or family
(Gobiidae). Among schooling species, average length
varied from 42 cm (yellowrtaO rockfish Sebas-
tes flavidus) to 11cm (Sebastes juveniles).
Ontogenetic habitat changes
Several rockfish species occurred both in
schools and singly, including pygmy, yel-
lowtail, sharpchin S. zacentrus, redstripe
S. proringer, and canary S. pinniger. Pyg-
mies were the single-most abundant spe-
cies identified in either category. Schools
of pygmy rockfish consisted of significant-
ly smaller individuals, averaging 16.1cm,
whereas non-schooling aggregations and
solitary individuals averaged 19.4cmTL
(ANOVA, F 18.0, df 1,699, P<0.01). Fur-
thermore, schools were usually associated
with ridge tops shallower than 100 m, while
non-schooling fish were on cobble and
boulder bottoms at depths of 100-150 m.
A similar analysis of the data for yellow-
tail, sharpchin, redstripe, and canary rock-
fish showed no significant difference be-
tween sizes of individuals in and out of
schools (ANOVA, all P>0.05).
Differences among stations:
Fish assemblages
Results of the PCA indicated striking dif-
ferences among stations in both the com-
position of fish assemblages and the
similarity of individual transects within
stations (Table 2, Fig. 3). Bartlett's spher-
icity test indicated that the first two axes
described significant non-random patterns
of variation among species, and accounted
for 70% of the total variation. The first
axis, which accounted for 49% of the varia-
tion, primarily contrasted transects at sta-
tion 4 (intermixed mud and rocky ridges)
vs. transects at station 5 (mud) (Fig. 3).
Transects from stations 1 and 3 (shallow
rocky bank tops) formed intermediate
homogeneous groups, while transects from
stations 2 and 6 (medium-depth boulder
and cobble fields) formed intermediate
heterogenous groups. Variable loadings in-
dicated that this axis primarily contrasted
variation in the relative abundance of rosethorn,
pygmy, canary, sharpchin, yellowtail, and greenstriped
rockfish, which were abundant at stations 1, 3, and 4
with the relative abundance of thornyhead rockfish, rex
sole, sablefish, poachers, and zoarcids, which were
abundant at station 5 (Table 2).
Stem et al : Fish-habitat associations at edge of Oregon continental shelf
545
Table 1
Numbers of individuals, average total lengths, and standard errors of lengths for schooling and non-schooling fish species identified |
on Heceta Bank, Oregon, September 1988.
Mean
Mean
No.
length
No.
length
Species
seen
(cm)
SE
Species
seen
(cm)
SE
Schooling
Non-schooling (continued)
Unident. small rockfish {Sebastes spp.)
12820
11.1
1.02
Yelloweye rockfish (S. ruberrimus)
27
45.4
1.78
Pygmy rockfish (S. wilsoni)
8390
16.1
0.72
Petrale sole (Eopsetta jordani)
26
27.3
0.55
Yellowtail rockfish (S. Jlavidus)
590
42.3
1.94
Threadfin sculpin
26
23.5
0.52
Sharpchin rockfish (S. zacentrus)
320
24.1
2.38
(Icelinus filamentostis)
Redstripe rockfish (S. proriger)
220
25.0
4.73
Longnose skate (Raja rhina)
23
72.8
40.6
Canary rockfish (S. pinniger)
110
41.7
4.70
Sandpaper skate (R. kincaidii)
22
30.0
2.77
Widow rockfish {Sebastes entomelas)
20
35.0
♦
Lingcod (Ophiodon elongatus)
20
46.5
9.96
Non-schooling
Blacktail snailfish
15
30.3
0.69
Pygmy rockfish (S. wilsoni)
3829
19.4
0.23
(Careproctus melanurus)
Sharpchin rockfish (S. zacentrus)
2030
23.7
0.28
English sole (Parophrys vetulus)
15
32.3
1.65
Rosethorn rockfish
931
21.6
0.10
Splitnose rockfish (S. diploproa)
13
31.9
2.03
(S. helvomaculatus)
Black eelpout (Lycodes diapterus
13
28.1
1.10
Dover sole {Microstomus pacificus)
436
30.3
0.37
Redbanded rockfish (S. babcocki)
11
28.6
3.78
Unident. sculpins (Cottidae)
319
15.8
0.09
Eared blacksmelt
11
15.0
*
Shortspine thornyhead
310
21.7
0.32
(Bathylagus oc.hotensis)
(Sebastolobiis alascanus)
Big skate (R. binoculata)
10
68.0
32.4
Blackbelly eelpout (Lycodes pacificus)
307
25.4
0.29
Hagfish (Eptatretus sp.)
10
34.0
3.83
Greenstriped rockfish
288
26.5
0.33
Unident. blennies (Blenniidae)
6
15.0
*
(Sebastes elongatus)
Bocaccio (S. paucispinis)
5
53.0
3.13
Unident. poachers (Agonidae)
248
22.5
0.15
Sculpin (Icelinus sp.)
4
27.5
1.25
Slender sole (Lyopsetta exilis)
240
22.5
0.16
Tiger rockfish (S.nigrocinctus)
4
40.0
1.67
Unident. small rockfish {Sebastes spp.)
130
11.6
1.35
Sanddab (Citharichthys sp.)
3
15.0
*
Yellowtail rockfish (5. jlavidus)
126
43.6
0.62
Unident. blotched rockfish
3
25.0
*
Rex sole {Glyptocephalus zachirus)
120
27.3
0.49
(Sebastes spp.)
Unident. flatfish (Pleuronectidae)
93
21.3
1.21
Eelpout (Lycodapus spp.)
2
15.0
*
Canary rockfish (S. pinniger)
78
46.9
0.98
Arrowtooth flounder
1
45.0
—
Thornback sculpin
.57
14.8
0.05
(Atheresthes stomias)
(Paricelinus hoplitieus)
Darkblotched rockfish (S. crameri)
35.0
—
Spotted ratfish (Hydrolagtis colliei)
54
44.1
0.53
Spiny dogfish (Squalus acanthias)
105.0
—
Unident. fish
47
18.2
1.30
Unident. smelt (Osmeridae)
15.0
—
Unident. large rockfish (Sebastes spp.)
47
23.1
0.94
Pacific Ocean perch (S. alutus)
25.0
—
Redstripe rockfish (S. proriger)
38
29.0
1.14
Pacific cod (Gadus macrocephalus)
55.0
—
Ronquils (Bathymasteridae)
33
14.7
0.16
Sand sole (Psettichthys melanostictus)
35.0
—
Sablefish (Anoplopoma fimbria)
33
54.7
3.53
Silvergray rockfish (S. bremspinis)
55.0
—
Kelp greenling
29
32.2
0.64
Unident. eelpout (Zoarcidae)
35.0
—
(Hexagrammos decagrammus)
ed to be
same length.
Unident. skate (Rajidae)
15.0
*No observed variation; all fish estimat
The second axis, which described 20% of the varia-
tion of relative fish abundance, presented an additional
independent pattern of variation among stations (Table
2, Fig. 3). This axis primarily contrasted transects at
stations 1 and 3 (shallow rock ridge) with transects
from all other stations. As in the first axis, transects
within stations along the second axis were heterogen-
eous; that is, the relative abundances of the fishes seen
varied among transects. Station 1 transects were
relatively homogeneous compared with transects at
stations 2, 3, and 6, which varied considerably. Variable
loadings indicated that this axis represented variation
in the relative abundance of kelp greenlings and
lingcod, which were abundant on bank tops, versus
thornyhead and sharpchin rockfish, zoarcids, thread-
fin sculpin, and dover, rex, and slender sole, which were
abundant at other stations.
Fish-habitat associations
Of the 49 possible combinations of bottom type (7x7
types), 27 were encountered. Cluster analysis indicated
that habitat types had varying degrees of similarity in
fish assemblages (Fig. 4). Mud had the most distinct
546
Fishery Bulletin 90(3). 1992
Table 2
Results of principal component analysis. Underlined bold characters indicate high variable
loadings, with positive and negative loadings being inversely correlated along each axis.
Thus, PCl depicts a gradient from soft-bottom species (negat
ve loadings) to hard-bottom
species (positive loadings). PC2 depicts a secondary gradient from hard-bottom
species
(negative loadings) to soft-bottom species (positive loadings).
Eigenvalue
PCI
PC2
3.056
1.266
Percent of total variation
49.3
20.4
Chi-square
1778
1557
Degrees of freedom
434
405
Variable loadings
Canary rockfish {Sebastes pinniger)
0.826
0.135
Yelloweye rockfish (S. ruberrimus)
0.477
-0.230
Yellowtail rockfish (S. flavidus)
0.657
-0.092
Redbanded rockfish (S. babcocki)
-0.552
0.338
Redstripe rockfish (S. proriger)
0.411
0.180
Sculpins (Cottidae)
0.748
0.036
Threadfin sculpin {Icelinus filamentosus)
0.339
0.447
Kelp greenling {Hexagrammos decagrammus)
0.189
-0.621
Lingcod {Ophiodon elongatus)
0.203
-0.525
Rosethorn rockfish (S. helvomaculatiis)
0.888
-0.296
Pygmy rockfish (S. wilsoni)
0.892
0.372
Sharpchin rockfish (S. zacentrus)
0.714
0.602
Splitnose rockfish (S. diploproa)
-0.595
0.266
Greenstriped rockfish (S. elongatus)
0.670
0.286
Shortspine thornyhead (Sebastolobus alascanus)
-0.690
0.446
Eelpouts (Zoarcidae)
-0.701
0.606
Hagfish {Eptatretus sp.)
-0.222
0.385
Spotted raffish (Hydrolagus colliei)
-0.326
0.461
Poachers (Agonidae)
-0.751
0.519
Sablefish (Anoplopoma fimbria)
-0.668
0.400
Searchers (Bathymasteridae)
0.441
-0.141
Blacktail snailfish (Careproctus melanurus)
-0.641
0.198
Dover sole {Microstormis pacificus)
-0.503
0.761
English sole (Parophrys vetulus)
-0.066
0.140
Petrale sole (Eopsetta jordani)
-0.005
0.192
Rex sole {Glyptocephalus zachirus)
-0.644
0.448
Slender sole {Lyopsetta exilis)
-0.180
0.582
Big skate {Raja binoculata)
-0.179
0.115
Longnose skate (R. rhina)
0.311
0.021
Sandpaper skate (R. kincaidii)
-0.429
0.2.38
T,
CM
O
CL
-1 ■
-3
DA
o Station 1
° Station 2
• Station 3
'^ Station 4
■ Station 5
* Station 6
5
3-11
PCI
fish assemblage, followed by
boulder, rocky ridge, mud and
cobble, and mud and boulder
habitats. In contrast, habitats in-
volving combinations of boulder,
mud, sand, and cobble had com-
paratively similar fish assem-
blages.
The results of the cluster anal-
ysis provide information relevant
to the interpretation of the PCA
results (Figs. 3,4). Stations that
displayed little among-transect
variability in fish assemblages
were composed primarily of rocky
ridge (stations 1 and 3), and mud
(station 5): habitats that had
relatively distinct fish assem-
blages. In contrast, stations with
high among-transect variability
(primarily stations 2 and 6) were
composed of mixtures of mud,
cobble, and boulders: habitats
sharing relatively similar fish
assemblages.
There were additional habitat
patterns evident in the distribu-
tion of the most abundant rock-
fish species (Table 3). Comparing
abundances of the four most
abtmdant species within subhabi-
tats, pygmy rockfish dominated
all except mud, mud and cobble,
and flat rock. Sharpchin domin-
ated mud and cobble; rosethorn,
the flat rock (Table 3). Compar-
ing abundances for each subhabi-
tat within species, it is clear that
each species, even though it
might not be numerically domi-
nant overall, was most abundant
in a particular habitat. Thus,
pygmy rockfish were most abun-
dant on mud and boulder; sharp
chin and greenstriped rockfish on mud and cob-
ble; rosethorn rockfish on boulder; and yellowtail
rockfish on rock ridges.
Figure 3
Ordination of first and second principal component
scores for 36 transects sampled at six stations on
Heceta Bank. The analysis is based on the relative
abundances of 30 fish taxa observed (see Table 2
for species list).
Stem et al.. Fish-habitat associations at edge of Oregon continental shelf
547
Hud & Sand -
Sand & Ridge -
Cobble i Ridge -
Boulder & Flat Rock -
Boulder & Hud -
Boulder & Ridge -
Ridge & Cobble -
Cobble & Sand -
Sand & Mud -
Pebble i Cobble -
Ridge & Boulder -
Ridge & Sand -
Boulder S, Sand -
Sand & Cobble -
Sand & Boulder
Sand -
Cobble & Boulder -
Mud i Pebble -
Boulder & Cobble -
Cobble i Mud -
Flat Rock
Cobble ■
Hud & Boulder
Hud i Cobble ■
Ridge -
Boulder
Hud
0.25
Euclidean distance
0.50
Figure 4
Cluster analysis of all observed bottom types based on the relative abundances of 21
fish taxa (see Table 4 for species list).
In general, the degree of bot-
tom-type relief varied inversely
with depth (Fig. 5). High relief
substrates such as rock ridges,
boulder, and cobble occurred at
relatively shallow 80-100 m depths,
while low-relief muddy substrates,
such as mud and boulder, mud
and cobble, and pure mud, oc-
curred relatively deeper, at 160-
240 m depths (Fig. 5).
CCA described associations be-
tween species abundance and the
distribution of specific habitat
types (Table 4, Fig. 6). Bartlett's
test indicated that the first three
axes represented significant
canonical correlations. The low
values of the redundancy coeffi-
cient for these axes (0.10-0.03,
measuring variability in fish abun-
dance explained by habitat varia-
tion) demonstrated strong corre-
lations between several species
and habitats rather than general
associations among all species
and all habitats.
Table 3
Average number of fish per
hectare (10''m-) on the seven
most distinct habitat types, as
determined by cluster analysis (see Fig. 4). 1
Only the 21 most abundant taxa are listed
these taxa used
in the canonical correlation analysis. Most-abundant taxon in
each category
underlined in bold characters. Species absent from a
specific habitat are
indicated with dashes.
Species
Mud Mud & cobble
Mud & boulder
Cobble
Boulder
Flat rock
Rock ridge
Agonidae
186
464
1122
—
25
—
18
Bathymasteridae
7
7
-
-
-
-
15
Big skate
7
—
51
-
-
-
-
Canary rockfish
—
14
102
—
—
158
82
Cottidae
24
79
51
67
—
158
73
Dover sole
499
343
2295
—
—
—
15
Greenstriped rockfish
64
364
204
266
25
-
79
Kelp greenling
-
-
-
67
76
316
27
Lingcod
—
—
-
67
-
-
30
Longnose skate
7
14
51
-
-
—
6
Pygmy rockfish
21
2129
8926
999
2772
—
1785
Redstripe rockfish
—
7
-
-
-
—
43
Rex sole
107
57
1887
—
—
—
—
Rosethorn rockfish
26
343
408
933
161
474
675
Sharpchin rockfish
60
2930
2754
133
—
—
277
Shortspine thornyhead
239
443
2193
-
-
—
-
Slender sole
76
107
408
—
—
—
—
Spotted raffish
26
14
510
—
—
—
—
Yelloweye rockfish
—
7
-
-
25
—
27
Yellowtail rockfish
—
29
—
67
176
—
191
Zoarcidae
282
279
1887
—
50
—
18
548
Fishery Bulletin 90(3). 1992
240 n
I
O
^
1
o
1 6
-C
160-
^
-)— '
Q_
cu
"O
o
o 9
80
o
c
D
QJ
^
RR
FF BB CC MB MC
Bottom type
MM
Table 4
Results of canonical correlation analysis
Variables with high loac
ings are indicated in 1
underlined boldface characters. High negative loadings on the first canonical
ixis, CCl,
indicate fish that were abundant in mud habitats.
Similarily, high loadings or
1 CC2 and
CCS indicate fish that were abundant on
cobble-boulder bottoms and rock ridge-sand |
valley bottoms, respectively.
Canonical correlation
CCl
CC2
CC3
0.849
0.786
0.489
Chi-square
1197
656
249
Degrees of freedom
147
120
95
Pjgj, Canonical variate loadings
Canary rockfish (S. pinniger)
0.032
-0.278
0.090
Yelloweye rockfish (S. ruberrimus)
0.035
-0.382
-0.139
Yellowtail rockfish {S. flavidus)
0.048
-0.021
0.434
Redstripe rockfish (S. proriger)
0.015
-0.218
-0.098
Cottidae
0.029
-0.200
0.360
Kelp greenling {Hexagrammos decagrammus)
0.048
-0.140
0.102
Lingcod (Ophiodon elongatus)
0.017
-0.134
0.367
Rosethom rockfish (S. helvomaeulatus)
0.026
-0.842
0.295
Pygmy rockfish (S. wilsoni)
-0.007
-0.571
0.012
Sharpchin rockfish (S. zacentrus)
-0.075
-0.837
-0.052
Greenstriped rockfish (S. elongatus)
-0.155
-0.037
0.618
Shortspine thornyhead {Sebastolobus alascaniis)
-0.461
0.066
-0.104
Zoarcidae
-0.696
0.077
-0.080
Spotted ratfish (Hydrolagus colliei)
-0.337
-0.504
-0.172
Agonidae
-0.678
0.079
-0.020
Bathymasteridae
0.027
0.043
0.080
Dover sole {Microstomias pacifiais)
-0.951
0.028
-0.005
Rex sole (GlyptocephaliLS zachirus)
-0.665
0.065
-0.086
Slender sole (Lyopsetta exilis)
-0.088
0.047
-0.081
Big skate {Raja binoculata)
-0.324
0.037
-0.059
Longnose skate {Raja rhina)
-0.048
0.069
-0.129
Variance extracted
0.132
0.113
0.051
Redundancy
0.095
0.070
0.012
Habitat
Mud
-0.998
0.052
-0.017
Sand
0.037
0.041
0.665
Pebble
-0.013
0.070
-0.023
Cobble
-0.041
-0.514
0.269
Boulders
0.026
-0.925
-0.159
Flat Rock
0.020
0.026
-0.060
Rocky Ridge
0.095
-0.007
0.529
Variance extracted
0.144
0.161
0.118
Redundancy
0.104
0.100
0.028
Figure 5
Average depth (±1 SE) of the eight most distinct
bottom type combinations on Heceta Bank, Oregon.
Bottom codes and sample sizes are as follows: RR
= rock ridge {n 109); FF = flat rock (n 4); BB =
boulder (n 29); CC = cobble (n 8); MB = mud-
boulder {n 11); MC = mud-cobble (n 45); MP =
mud-pebble {n 26); MM = mud (n 55) (see text for
a description of bottom type codes), n = number
of habitat patches sampled per station.
Because bottom-type changes were highly correlated
with changes in depth (Fig. 5), the CCA did not con-
found species associations with bottom types from dif-
ferent depths. Each axis measured associations occur-
ring within the depth range of
the habitat indicated by the vari-
able loadings on each axis.
The first axis described varia-
tion in fish abundance associated
with mud habitats (160-240m).
Variable loadings indicate that
thornyheads, zoarcids, poachers,
and rex and Dover sole common-
ly occur on mud (Table 4). Can-
onical variate scores on this axis
were significantly different
among stations in both the habi-
tat and fish scores: station 5, the
only pure mud station (Fig. 2),
was significantly different from
all others (Kruskal-Wallis, p<
0.01) (Fig. 6A).
The second axis contrasted an
additional independent fish-habi-
tat association. Variables load-
ings indicated that ratfish, and
rosethorn, sharpchin, yelloweye,
canary, and pygmy rockfish were
associated with boulder and
cobble fields at 75- 100 m depths.
Canonical scores for the second
axis also differed significantly
among stations on both habitat
and fish scores: station 6, the
station with the highest cover of
boulder-cobble (Fig. 2), was
significantly different from all
other stations (Kruskal-Wallis,
p<0.01) (Fig. 6B).
The third axis indicated an ad-
ditional association between fish
and habitat. Variables loadings
indicated that greenstriped and
yellowtail rockfish, lingcod, and
cottids were associated with sand
Stem et al,: Fish-habitat associations at edge of Oregon continental shelf
549
and ridge habitats at 75-lOOm depths. Canonical
variate scores among stations differed on the third axis
with respect to habitat scores, but not on fish scores.
With respect to habitat, stations 1 and 3, which had
the highest amount of sand and ridge cover (Fig. 2),
were significantly different from stations 2, 4, and 6
(Kruskal-Wallis, jd<0.01) (Fig. 6C). In contrast, with
respect to fish abundance, stations were highly vari-
able, and not significantly different among stations
(Kruskal-Wallis, p>0.05) (Fig. 6C).
Discussion
The principal objective of our study was to develop
methods to estimate spatial variation in fish abundance
on Heceta Bank. However, the high variability of bot-
tom types encountered required that we understand
the effects of bottom-type variation on fish abundance
and distribution.
Fish-habitat associations
The principal components analysis showed that stations
with the least variability in fish abundance among
replicate transects were those at stations composed of
rock ridge, such as the bank tops (stations 1 and 3),
and of mud (station 5). In contrast, high variability in
fish abundance among replicate transects occurred at
stations having combinations of mud, cobble, and
boulders (stations 2, 4 and 6). Moreover, canonical cor-
relation analysis indicated that fish assemblages asso-
ciated with these habitats were unique. Mud, cobble-
boulder, and ridge-sand habitats displayed different
species composition and relative abundance.
In most sampling situations, such as use of a bottom
trawl or bottom-set gillnet, analysis would be limited
to a fish-only PCA-type appproach. Within-station
variability, such as that documented here, would be
largely unaccounted for without detailed information
about bottom type. In the present study, canonical cor-
relation analysis of fish abundance relative to bottom
type provided key information on a major source of
within-station variability.
The ability to estimate bottom-type composition and
determine the relationships of species with each sub-
strate is a critical advantage of submersible studies.
In shallow water, this has been done using scuba (e.g.,
Hixon 1980, Larson 1980, Hallacher and Roberts 1985).
However, there are few such studies below scuba
depths. In the northeastern Pacific, Carlson and Straty
(1981) and Straty (1987) used a submersible to study
habitat and nursery areas for rockfishes in southeast-
ern Alaska. Straty (1987) concentrated on species of
juvenile rockfishes and their occurrence in relation to
0.5
A
IP
-n.5
-£^
o
1
X
-1.5
,_
( )
o
-2.5
-2.5 -1.5 -0.5
CCl Fish
O
I
to
o
o
c
1 1
l-A-
---
° Station
1
D Station
2
• Station
3
' Station
4
■ Station
5
» Station
6
-0.8 -0.3
CC3 Fish
XI
o
X
o
o
B ^g--*-
I
. A .
-2.0 -1.0 0.0 1.0
CC2 Fish
Figure 6
Average canonical variate scores for each station (± 1 SE) on
three canonical correlation axes (see Table 4). Canonical scores
for habitat indicate the relative cover of specific bottom types
on each axis, while scores for the fish data indicate the relative
abundance of specific fish on each axis.
substrate type and relief; he did not attempt to quan-
tify abundances. Richards (1986) similarly investigated
distributions of deep rockfishes at 21-140 m and related
their occurrence to bottom type. Although she was able
to show substrate associations for three species {Sebas-
tes elongatus, S. maliger, and S. ruberrimus), only
three substrate categories were used, and abundances
of fishes were determined on the basis of distance of
maximum visibility. Nevertheless, she recognized the
importance of such studies and developed initial
methods for obtaining data on this subject.
Abundances of species
We know of no comparable data to that presented here
for fish abundances on Heceta Bank. However, similar
studies have been done on inshore reefs in central
California. Miller and Geibel (1973), using scuba tech-
550
Fishery Bulletin 90(3). 1992
niques similar to our submersible methods, estimated
abundances of juvenile and adult rockfishes along
transects on an inshore reef supporting an extensive
kelp forest. Their estimates of juvenile abundances are
much higher than ours: more than 46,000 fish/ha com-
pared with our maximum of 15,039 fish/ha (station 3).
Comparing abundances of adults at the same stations,
they estimated 3133-5046 fish/ha vs. 558-1724 fish/ha
at our station 3. The maximum estimated number of
adult fishes at any of our stations was 9635 fish/ha (sta-
tion 4). However, our estimates at the shallow bank top
stations (1 and 3) are low because they did not include
most of the schooling fishes, which were above the
submersible. We have no accurate estimates of the
abundances of these fishes, primarily yellowtail and
widow rockfish. They occurred in schools of thousands.
Availability of comparative data
There are few data sets comparable with ours that were
obtained by other methods. Rough bottoms are un-
trawlable, reducing usable gear to longlines or set nets.
Even where these are used, if substrate varies over the
length of the longline or net, they "integrate" the fishes
over those different bottom types, preventing associa-
tion of species with specific substrate types. Using otter
trawls with foot rope rollers, Barss et al. (1982) studied
fish assemblages associated with "rough" (rocky bot-
tom fishable with nets using rollers) and "smooth" bot-
tom on the west (offshore) side of Heceta Bank. The
areas with most relief were unfishable (including our
stations 1 and 3). They found distinct differences in
catches between the two types of areas, but admitted
that their results were biased by the type of gear they
were forced to use in order to trawl on rough bottom.
Recently, Matthews and Richards (In press) used
gillnets to compare fish assemblages on trawlable and
untrawlable bottoms west of Vancouver Island. Their
goal was to determine whether, as commercial fisher-
men believe, untrawlable bottom west of Vancouver
Island provides refuges for commercially exploited
fishes (primarily Pacific Ocean perch) caught nearby.
They concluded there were no reservoir populations.
However, given the mesh size of their bottom-set gill-
nets, they were unlikely to sample either juveniles or
semipelagic species such as yellowtail and widow
rockfish. Thus, given our current and previous obser-
vations of juvenile and yellowtail rockfish associated
with shallow, high-relief rocky ridges (Pearcy et al.
1989), we suggest that these unfished areas could still
provide refuges for fishes in either of those categories.
Habitat shifts
Ontogenetic habitat shifts, such as those desribed here
for pygmy rockfish, are common among rockfish
species. Westrheim (1970), Carlson and Haight (1976),
and Straty (1987) found that juvenile Pacific Ocean
perch Sebastes alutus were usually shallower than
adults. Carr (1983) described the growth-related migra-
tion of juvenile S. atrovirens, S. camatm, S. chryso-
melas, and S. caurinus to the bottom in a central
California kelp forest. However, Hallacher (1977),
studying adults and juveniles of 5. rnystinus and S. ser-
ranoides in Monterey Bay at depths of '^'25m or less,
found that abundances of both increased with depth!
maxima occurring at the greatest depths sampled. This
difference could be related to degree of association with
the bottom as adults. The species Carr (1983) studied
were benthic as adults, whereas the latter two species
occur in the water column.
Conclusions
The results presented here show the utility of using a
submersible rather than bottom set nets, traps, or long-
lines to study fish-substrate associations in deep water
areas where substrate is heterogeneous. Other meth-
ods, such as Remote Operated Vehicles, rely on video
and still camera images, which are not as adequate for
accurate identification as the human eye. Moreover,
other types of gear do not allow detailed characteriza-
tion of the substrate sampled, but rather integrate the
catch from a variety of habitats. We believe that the
methods presented here, in addition to describing basic
fish-habitat associations, allowed us to control the ef-
fects of sampling across a range of different habitats,
and increased our ability to detect meaningful spatial
variation in fish abundance.
Acknowledgments
We thank the crews of the RV William A. McGaw and
RS Delta for help in obtaining the data; their coopera-
tion and expertise are appreciated and gratefully
acknowledged. Comments by L. E. Hallacher and
L. Richards greatly improved the manuscript. The
research was supported by the Minerals Management
Service, Department of the Interior, under MMS
Agreement No. 14-12-0001-30445, and the National
Undersea Research Program of NOAA.
Stei n et al Fish-habitat associations at edge of Oregon continental shelf
551
Citations
Barss. W.H., S.L. Johnson, and R.L. Demory
1982 Biological studies on rockfish and associated species from
Heceta Bank off Oregon, 1980-81. Oreg. Dep. Fish Wildl.
Completion Rep. Proj. M51-R2, Comm. Fish. Res. Dev. Act.,
Newport, 28 p.
Brock, V.E.
1954 A preliminary report on a method of estimating reef fish
populations. J. Wildl. Manage. 18:297-308.
Carlson, H.R., and R.E. Haight
1976 .Juvenile life of Pacific Ocean perch. Sebastes alutus, in
coastal fiords of southeastern Alaska; Their environment,
growth, food habits, and schooling behavior. Trans. Am. Fish.
Soc. 10.5(2):191-201.
Carlson, H.R., and R.R. Straty
1981 Habitat and nursery grounds of Pacific rockfish, Sebastes
spp., in rocky coastal areas of southeastern Alaska. Mar. Fish.
Rev. 43(7):13-19.
Carr. M.
1983 Spatial and temporal patterns of recruitment of young-
of-the-year rockfishes (genus Sebastes) into a central Califor-
nia kelp forest. M.S. thesis, San Francisco State Univ., 104 p.
Ebeling, A.W.
1982 Scuba diver observations and statistical analysis. In
DeWees, CM. (ed.), Guidelines for marine ecological surveys,
nekton, p. 6-9. Calif. Sea Grant Coll. Mar. Adv. Prog., Univ.
Calif., Davis.
Hallacher, L.
1977 Patterns of space and food use by inshore rockfishes
(Scorpaenidae: Sebastes) of Carmel Bay, California. Ph.D.
thesis, Univ. Calif., Berkeley, 115 p.
Hallacher, L.E., and D.A. Roberts
1985 Differential utilization of space and food by the inshore
rockfishes (Scorpaenidae: Sebastes) of Carmel Bay, California.
Environ. Biol. Fish. 12:91-110.
Hixon, M.A.
1980 Competitive interactions between California reef fishes
of the genus Embiotoca. Ecology 61:918-931.
Larson, R.J.
1980 Competition, habitat selection, and the bathymetric
segregation of two rockfish (Sebastes) species. Ecol. Monogr.
50:221-239.
Matthews, K.R., and L.J. Richards
In press Rockfish (Scorpaeidae) assemblages of trawlable and
untrawlable habitats of Vancouver Island, British Colum-
bia. N. Am. J. Fish. Manage.
Miller, D.J., and J.J. Geibel
1973 Summary of blue rockfish and lingcod life histories; a reef
ecology study; and giant kelp, Macrocystis pyrifera. experi-
ments in Monterey Bay, California. Calif. Dep. Fish Game
Fish Bull. 158:1-137. "
Pearcy, W.G., D.L. Stein, M.A. Hixon, E. Pikitch. W. Barss, and
R. Starr
1989 Submersible observations of deep-reef fishes of Heceta
Bank, Oregon. Fish. Bull., U.S. 87:955-965.
Pimentel. R.A.
1979 Morphometries, the multivariate analysis of biological
data. Kendall-Hunt. Dubuque, 276 p.
Ralston, S., R.M. Gooding, and G.M. Ludwig
1986 An ecological survey and comparison of bottom fish
resource assessments (submersible versus handline fishing) at
Johnston Atoll. Fish. Bull., U.S. 84(1);141-155.
Richards, L.J.
1986 Depth and habitat distributions of three species of rock-
fish (Sebastes) in British Columbia: Observations from the
submersible PISCES IV. Environ. Biol. Fishes 17(1);13-21.
Sokal, R.R. and F.J. Rohlf
1981 Biometry: The principals and practice of statistics in
biological research, 2d ed. W.H. Freeman, NY, 859 p.
Straty, R.R.
1987 Habitat and behavior of juvenile Pacific rockfish (Sebastes
spp. and Sebastolobus alascanus) off southeastern Alaska. In
Cooper, R.A., and A.N. Shepard (eds.). Scientific applications
of current diving techniques on the U.S. continental shelf, p.
109-123. Symp. Ser. Undersea Res., NOAA Undersea Res.
Prog. 2(2), Wash. DC.
Westrheim, S.J.
1970 Survey of rockfishes, especially Pacific Ocean perch, in
the northeast Pacific Ocean, 1963-66. J. Fish. Res. Board
Can. 27:1781-1809.
Abstract. -This paper is devoted
to a theoretical examination of two
rules of thumb commonly used in
fishery management: (I) the fishing
mortality rate associated with max-
imum sustainable yield (Fm^y ) equals
the natural mortality rate, and (II)
the equilibrium stock biomass at
maximum sustainable yield equals
one-half the pristine stock biomass.
Taken together, these rules of thumb
are shown to be inconsistent with
any simple dynamic pool model in
which three conditions hold; (1) the
first derivative of the stock-recruit-
ment relationship is uniformly non-
negative, (2) the second derivative of
the stock-recruitment relationship is
uniformly nonpositive, and (3) the
first derivative of the weight-at-age
relationship is uniformly positive. An
example of such a model is presented
and the equilibrium solution derived
analytically. In this model, Fmsy can
be either greater than or less than
the natural mortality rate, while the
equilibrium stock biomass at max-
imum sustainable yield is consistent-
ly less than one-half the pristine
stock biomass. To illustrate the util-
ity of the theoretical framework de-
veloped, the model is applied to the
eastern Bering Sea stock of rock sole
Pleuronectes bilineatus.
Management advice from
a simple dynamic pool model
Grant G. Thompson
Resource Ecology and Fisheries Management Division
Alaska Fisheries Science Center, National Marine Fisheries Service, NOAA
7600 Sand Point Way NE, Seattle, Washington 981 15-0070
Two rules of thumb
Despite its acknowledged shortcom-
ings (e.g., Larkin 1977), management
for maximum sustainable yield (MSY)
remains a common strategy among
fisheries professionals. Under a con-
stant harvest rate policy, this strate-
gy is implemented by exploiting the
stock at the fishing mortality rate
corresponding to MSY (Fmsy)- Alter-
natively, this strategy could be imple-
mented by exploiting the stock so as
to maintain its biomass at the level
corresponding to MSY, B(Fmsy)- To
estimate Fmsy and B(Fmsy). fishery
scientists and managers employ a
variety of approaches, ranging from
highly sophisticated simulation
models to simple "rules of thumb."
Frequently used examples of the
latter can be found in the form of two
hypotheses employed by Alverson
and Pereyra (1969) in their analysis
of the potential yield of certain fish
stocks. These hypotheses (hereafter
referred to as Rules I and II) are
and
Fmsy
M
B(Fmsy)
B(0)
0.5,
(I)
(11)
Manuscript accepted 4 June 1992.
Fishery Bulletin, U.S. 90:552-560 (1992).
where F is the instantaneous rate of
fishing mortality per year, Fmsy is
the value of F that produces MSY in
equilibrium, M is the instantaneous
rate of natural mortality per year,
B(F) is the equilibrium stock biomass
corresponding to a fishing mortality
rate of F, B(Fmsy) is the equilibrium
stock biomass when F = Fmsy. and
B(0) is the pristine stock biomass
(i.e., equilibrium stock biomass when
F = 0).
Alverson and Pereyra (1969) pre-
sented a sketchy derivation of Rules
I and II, leaving open the question of
which models might be capable of
leading to the hypothesized relation-
ships. A number of authors have sub-
sequently examined specific models
in this context and shown them to be
inconsistent with Rules I and II.
Gulland (1971) and Beddington and
Cooke (1983) cast doubt on the
robustness of Rules I and II in terms
of their appHcability to the "simple"
model of Beverton and Holt (1957),
but did not generalize their conclu-
sions beyond that particular model.
Likewise, Francis (1974) demon-
strated inconsistencies between
Rules I and II and a set of assump-
tions derived from the Schaefer
(1954) model, although his argument
was weakened somewhat by com-
puting MSY in terms of numbers, not
biomass. Deriso (1982) showed that
the discrete fishing mortality rate
generated by his delay-difference
model at MSY was consistently
higher than the discrete natural mor-
tality rate when recruitment was
constant, while under several other
stock-recruitment assumptions the
relationship was reversed. Shepherd
(1982) also demonstrated that Rules
I and II did not adequately describe
the behavior of a particular surplus
production model.
Since none of these authors ad-
dressed the possibility that other
552
Thompson" Management advice from a simple dynamic pool model
553
models might support Rules I and II, it remains to be
seen whether these rules are inconsistent only for
isolated special cases, or are actually incompatible with
a major class of models.
Review of simple dynamic pool models
One place to start in the search for models that might
be compatible with Rules I and II is within the family
of simple dynamic pool models. As distinguished from
surplus production models such as those of Schaefer
(1954) and Pella and Tomlinson (1969), dynamic pool
models describe stock dynamics in terms of the indi-
vidual processes of recruitment, growth, and mortal-
ity, and incorporate age structure at least implicitly
(e.g., Pitcher and Hart 1982). Within the broad class
of dynamic pool models, a model will be referred to here
as "simple" if it reflects the following assumptions: (A)
Cohort dynamics are of continuous- time form, (B) vital
rates are constant with respect to time and age, (C)
fish mature and recruit to the fishery continuously and
at the same invariant ("knife-edge") age, (D) mean
body weight-at-age is determined by age alone, (E) the
stock (or population) consists of the pool of recruited
individuals, (F) maximum age is infinite, (G) the stock
is in an equilibrium state determined by the fishing
mortality rate, and (H) recruitment is determined by
stock biomass alone. Within the framework provided
by these assumptions, particular models are distin-
guished by the forms assigned to the weight-at-age and
stock-recruitment functions.
Assumptions (A-C) imply that simple dynamic pool
models conform to the following pair of equations:
dn(F, a)
da
= -n(F, a)Z,
and
n(F, a) = n(F, a^) e-Z(a-a,),
(1)
(2)
where a = age, n(F, a) is the stationary population
distribution (in numbers) by ages a when the stock is
exploited at a fishing mortality rate of F, Z is the in-
stantaneous rate of total mortality (F -i- M), and a^ is
the age of recruitment.
Equation (1) gives the instantaneous rate of change,
by age, of the distribution n. When integrated with Z
constant (Assumption B), Equation (1) gives Equation
(2), numbers as a function of age. Assumption (D) im-
plies that Equation (2) can be cast in terms of biomass
by multiplying both sides of the equation by the weight-
at-age function w(a):
b(F, a) = W(a) n(F, a^) e-^^'^-'^l
(3)
where b(F, a) is the stationary population distribution
(in biomass) by ages a when the stock is exploited at
a fishing mortality rate of F.
Assumptions (B), (C), (E), and (F) imply that total
equilibrium stock numbers can be obtained by inte-
grating Equation (2) from a = ar to a = oo, giving
N(F)
n(F, a^)
(4)
where N(F) represents total equilibrium numbers when
the stock is exploited at a fishing mortality rate of F.
Likewise, equilibrium stock biomass is obtained by
integrating Equation (3) from a = a,. to a = °°:
X
B(F) = n(F, a,.) f w(a) e -Zf^'-ar) da.
(5)
In the case where a = ar, Assumptions (G) and (H)
imply that the left-hand side of Equation (3), recruit-
ment biomass, is a deterministic fimction of equilibrium
stock biomass r(B(F)):
b(F. a,) = r(B(F)).
(6)
Average weight of individuals in the stock W(F) can
be written
;:
w(a) e-^'^^-^r) da
W(F)
= Z
e-Z(a-a,) da
w(a) e"^*^"'''' da.
(7)
Equation (5) can then be rewritten
B(F)
W(F) n(F, a,)
(8)
For the case of a pristine stock (F = 0), Equations (4)
and (8) imply that equilibrium stock size (in terms of
554
Fishery Bulletin 90(3). 1992
numbers and biomass, respectively) is given by
N(0) =
n(0, a^)
(9)
and
B(0)
W(0) n(0, ar
M
(10)
Inconsistency of Rules I and II
Tlie argument of Francis (1974) can be generalized to
address more fully the compatibility of Rules I and II.
The method to be used is as follows: First, it will be
shown that if Rules I and II were to hold simultaneously
with the properties of simple dynamic pool models,
these rules would imply a particular result. It will then
be shown that this result is incompatible with a major
subset of the family of simple dynamic pool models,
thus proving that Rules I and II are also incompatible
with this subset.
Rule II and Equation (10) imply
B(Fmsy) =
W(0) n(0, a,)
2M
(11)
b(FMSY. ar)
b(0, a,)
r(B(FMSY))
r(B(0))
(15)
Now let the discussion be restricted to models in
which the first derivative of the stock-recruitment rela-
tionship is uniformly nonnegative. In such cases. Equa-
tion (15) indicates that the left-hand side of Equation
(14) is less than or equal to 1 if equilibrium stock
biomass decreases as a function of F [i.e., if B(Fmsy)
<B(0), then r(B(FMSY))<i'(B(0))]. To examine the
conditions under which this occurs, let Equation (8) be
rewritten
B(F)
W(F) r(B(F))
w(ar) Z
Equation (16) can be differentiated as follows:
(16)
dB(F)
dF
r(B(F))
/dW(F)',
Zi-^] - W(F)
w(ar) Z-W(F)
dr(F(F))
dB(F)
(17)
Equation (8) implies that B(Fmsy) must also con-
form to
B(Fmsy) =
W(FMSY)n(FMSY,a,)
Fmsy+M
(12)
Solving Equations (11) and (12) for F^sy gives
The numerator in Equation (17) is negative whenever
dW(F)/dF<0, which is easily shown to be true when-
ever w(a) is monotone increasing, a characteristic
typical of all commonly used growth functions (Schnute
1981).
Thus, it follows that dB(F)/dF will likewise be nega-
tive whenever the denominator in Equation (17) is
positive; that is, whenever
f2W(FMsv) n(F„.., a,) _ >
W(0) n(0, a,)
w(a,)Z ^ dr(F(F))
W(F)
dB(F)
(18)
Next, Rule I and Equation (13) imply
By Equation (16), the left-hand side of (18) can be
rewritten as the ratio of r(B(F)) to B(F), giving
n(FMSY. ar)
W(0)
(14)
n(0, a,) W(Fmsy)
The left-hand side of Equation (14) can be rewritten
n(FMSY, ar) w(ar) n(FMSY. ar)
n(0, ar)
w(ar) n(0, ar)
r(B(F)) ^ dr(B(F))
B(F) dB(F)
(19)
Given that the discussion has been restricted to
models with stock-recruitment relationships that are
nondecreasing (nonnegative first derivative), a suffi-
cient condition for Equation (19) to hold is for dr(B
(F))/dB(F) to be nonincreasing (nonpositive second
Thompson: Management advice from a simple dynamic pool model
555
derivative). Therefore, for all simple dynamic pool
models in which r(B(F)) is nondecreasing and dr(B
(F))/dB(F) is nonincreasing, the left-hand side of Equa-
tion (14) is less than or equal to 1.0.
Turning to the right-hand side of Equation (14), note
that this expression is necessarily greater than 1.0
whenever dW(F)/dF<0, a condition which has already
been noted to hold whenever w(a) is monotone
increasing.
Summarizing the argument, then, it has been shown
that Rules I and II cannot hold simultaneously for any
simple dynamic pool model in which the first derivative
of the stock-recruitment relationship is uniformly non-
negative, the second derivative of the stock-recruit-
ment relationship is uniformly nonpositive, and the first
derivative of the weight-at-age relationship is uniformly
positive.
Example of a simple dynamic
pool model
Growth, bJomass, recruitment, and yield
As an alternative to Rules I and II, it is possible to
examine the behavior of Fmsy/M and B(Fmsy)''B(0)
explicitly for a particular model. The model to be ex-
amined here incorporates a linear weight-at-age func-
tion (Schnute 1981). Let
w(a) = w(ar
a — a0
a,- — a* I
(20)
where ao represents the age intercept.
Biomass at age is then
b(F,a) =
b(F, a^)(a-ao)e-z(a-ar)
a,, -ao
(21)
The stock-recruitment relationship of Gushing (1971)
will be used to complete the model, giving recruitment
as a power function of stock size:
b(F, a,) = pB(F)q,
(23)
where p and q are constants, and 0<q<l. In the
limiting case of q = 0, recruitment is constant, while in
the other limiting case of q= 1, recruitment is propor-
tional to biomass.
The Gushing stock-recruitment relationship has the
advantage of rendering Equation (22) explicitly solv-
able. Substituting Equation (23) into Equation (22) and
rearranging terms gives the following equation for
equilibrium stock biomass:
B(F) =
1 +
1
Z(a,. -ao)/J
1
(24)
Multiplying both sides of Equation (24) by F then
gives the equation for yield Y(F) shown below:
Y(F)
[P
1 +
1
Z(ar-ao)/-
l-q
(25)
A partitioning of stocl< production
From this point on, it will prove helpful to make use
of a new parameter K", defined as follows:
K" =
1
M(ar-ao)
(26)
The parameter K" has a special biological interpreta-
tion in the context of the present model. To develop
this interpretation, first multiply Equation (22) through
by Z, yielding:
For a given value of b(F, a^), biomass at age can be
integrated from a = ar to a = °° to obtain the correspon-
ding equilibrium stock size. Equation (21) can be in-
tegrated by parts, giving the following expression for
equilibrium stock biomass (Hulme et al. 1947):
B(F)
(a-ao) e-^'^'-^'r) da
fb(F, a,)\
1 -1-
1
Z (a^ - ao )
(22)
Z B(F) = b(F, a,) 1 -h
Z (a^ - ao ) I
(27)
Assuming no immigration or emigration, stock losses
due to mortality must equal stock gains due to recruit-
ment and growth at equilibrium (Russell 1981). Since
the left-hand side of Equation (27) represents losses due
to mortality, the right-hand side must equal the sum
of equilibrium recruitment and growth. Therefore,
Equation (27) can be rearranged to define equilibrium
stock growth G(F) as follows:
G(F) = Z B(F)
b(F, a^)
b(F,ar) = — -• (28)
Z(ar -ao)
556
Fishery Bulletin 90(3). 1992
Dividing Equation (28) through by b(F, ar) gives the
ratio between the two components of stock production,
i.e., growth and recruitment:
1
G(F) ^
b(F, a,) Z(ar-ao)'
(29)
In the case of a pristine stock, Equation (29) reduces
to
G(0)
b(0, ar) M(ar-ao)
= K".
(30)
In other words, K" is simply the pristine ratio of
growth to recruitment. At values of K">1 pristine pro-
duction is dominated by growth, while at K" = 1 the two
components of pristine production are equal, and at
values of K"<1 pristine production is dominated by
recruitment.
" MSY
\ \
4
- \ \
3
\k"=i \k"=o
2
1
K"=infinity ^~~^^^^^^^^^^^^i=^^-__
0 0 0 2 0 4 0 6 0 8 1.0
Recruitment Parameter q
Figure 1
Ratio of Fj,,Y to M. For a given K' value, the ratio decreases
toward zero as q approaches 1.0. Higher K" values result in
lower values of the ratio, reaching a lower limit as K" ap-
proaches infinity.
Fishing mortality at maximum sustainable yield
Differentiating Equation (25) with respect to F and setting the resulting expression equal to zero gives the follow-
ing equation for F^sy •
q+1
. ar - ao ,
\+M + ]
q+1 - (6q-2)M
-I- + M-
^T — ao/ \ ar-ao /
MSY
M.
2q
(31)
Using F' to denote the ratio F/M, Equation (31) can be simplified via Equation (26) to
- (q-t-1) K" -h 1 -^ ^(q-Hl)^ K"2 -t- (6q-2) K" -h 1
F'mSY = T 1-
2q
(32)
Figure 1 illustrates the behavior of F'msy as a func-
tion of q for four values of K" (0, 1, 3, and °°). Note that
F'msy can deviate substantially from the value of 1.0
suggested by Rule I. The locus of parameter values for
which Rule I holds under Equation (32) is
q =
1
K"-(-2
(33)
implying that q must be less than 0.5 in order for Rule
I to hold.
When q=l. Equation (32) falls to zero. As q ap-
proaches zero, Equation (32) approaches an upper limit
F'max defined by
K" + l
K"-l'
(34)
The limits of Equation (32) as K" approaches zero and
infinity are, respectively.
and
1-q
hm Fmsy =
K"-(i q
1-q
hm F MSY = •
K^<» 1-i-q
(35)
(36)
Thompson: Management advice from a simple dynamic pool model
557
When pristine growth and recruitment are exactly
balanced (K" = 1), Equation (32) reduces to
F'
MSY
(37)
In the case of Equation (34), Rule I is always an
underestimate (i.e., F'^sy is always greater than 1.0).
In the case of Equation (35), Rule I is an underestimate
whenever q<0.5 and an overestimate whenever q>0.5.
In the case of Equation (36), Rule I is always an over-
estimate, except in the limiting case where q = 0. In the
case of Equation (37), Rule I is an underestimate
whenever q< 1/3 and an overestimate whenever q> 1/3.
Biomass at MSY relative to pristine biomass
Substituting M + Fmsy for Z in Equation (24) gives
B(Fmsy)- Likewise, substituting M for Z in Equation
(24) gives B(0). Forming a ratio from these two bio-
masses gives
B(Fmsy)
B(0)
K" + F'
MSY
1
1
1-q
(K" + 1)(F'msy + 1)'
(38)
where F'msy is given by Equation (32).
Equation (38) is illustrated in Figure 2. Note that all
of the curves in Figure 2 exhibit the same upper bound
(1/e, about 0.37), which is reached in the limit as q ap-
proaches 1.0. Thus, Rule II always overestimates Equa-
tion (38) by a minimum of about 36%. At values of
q>0.5, the biomass ratio is always greater than 0.25,
but at lower values of q the ratio can be much smaller.
Multiplying Equations (32) and (38) gives the ratio
MSY/MB(0), which is plotted in Figure 3. This ratio
describes a stock's maximum sustainable fishery-
induced losses as a proportion of its pristine losses.
Alverson and Pereyra (1969) suggested that the MSY/
MB(0) ratio should equal 0.5, a figure obtained by
multiplying Rules I and II together. Note that this sug-
gestion errs on the high side whenever K" exceeds 1.0,
as well as whenever q exceeds ~0.23.
Applying the model to rock sole
As an illustration of the approach suggested above, the
model can be applied to the eastern Bering Sea stock
of rock sole Pleuronectes bilineatus. This stock is ex-
ploited by a multispecies flatfish fishery, and is also the
target of an important roe fishery (Walters and Wilder-
buer 1988).
B(Fmsy)/B(0)
0.4 06
Recruitment Parameter q
Figure 2
Ratio of B(Fj,sy) to B(0). For any given K' value, the ratio
increases toward a value of 1/e as q approaches 1.0. Higher
K" values result in higher values of the ratio, reaching an upper
limit as K" approaches infinity.
MSY/MB(0)
0.4 0.6
Recruitment Parameter q
Figure 3
Ratio of MSY to the product of M and B(0). For a given K"
value, the ratio decreases toward a value of zero as q ap-
proaches 1.0. Higher K" values result in lower values of the
ratio, reaching a lower limit as K" approaches infinity.
The parameters to be estimated are q and K". The
parameter q can be determined from data on stock
biomass and recruitment. Trawl survey estimates of
rock sole stock biomass are available for the years
1979-88 (Walters and Wilderbuer 1988). In addition,
age composition of the stock has been determined for
the years 1979-87. In order to obtain an estimate of
age composition for 1988, the "iterated age-length
key" approach of Kimura and Chikuni (1987) was ap-
plied to the 1986 age-length key and the 1988 length-
558
Fishery Bulletin 90(3). 1992
frequency distribution. Assuming that rock sole recruit
at age 3 (Walters and Wilderbuer 1988), these data
provide seven years of information on the stock-
recruitment relationship. Fitting Eq. (23) to these
seven points (assuming lognormal error, Fig. 4) gives
q = 0.235.
The composite parameter K" can be estimated from
its constituent parameters SLy, ao, and M (Eq. 26).
Walters and Wilderbuer (1988) set a^ = 3 and M = 0.2.
The parameter ay can be derived by regressing a line
through the mean weights-at-age, as shown in Figure
5 (R'~ =0.904). This gives an ao value of 1.475 years,
implying a K" value of 3.279.
With these parameter values, Equation (32) gives
F'msy = 0.880, or Fmsy =0-176. This estimate of Fmsy
compares favorably with the value of 0.155 that
Walters and Wilderbuer (1988) derived from a surplus
production model. It is relatively close to (within 12%
of) the value indicated by Rule I.
However, Rule II does not fare so well in this ex-
ample. Equation (38) estimates the ratio between
B(Fmsy) and B(0) at a value of 0.245, 51% below the
value predicted by Rule II.
Discussion
The topic of this paper, management advice from a
simple dynamic pool model, has been considered from
the perspective of how two commonly used rules of
thumb compare with simple dynamic pool models in
general, and how they compare with one such model
in particular.
Choice of functional forms
Within the family of simple dynamic pool models, a
particular model is defined by its stock-recruitment
and growth functions. As Paulik (1973) and Ricker
(1979) state, the choice of functional form for these
two processes is largely a matter of convenience. The
linear growth and Gushing stock-recruitment functions
have been chosen for the proposed model, in part
because of the tractability they confer. For example,
their use permits explicit specification of Fmsy (im-
possible in other known examples of simple dynamic
pool models, except in the special case where Fmsy =
F,„ax). Another advantage is economy of parametriza-
tion: only two parameters (K" and q) are required. The
main disadvantage is the possibility that the simplicity
of these functional forms might ignore critical
behaviors.
The linear growth assumption is probably the more
controversial of the two choices. The primary criticism
of the linear growth equation is that it implies a con-
Recruitment Biomass (thousands of t)
00 0.1 0.2 0.3 04 O.S 06 07 0.8 0.9 1.0
stock Biomass (millions of t)
Figure 4
Stock-recruitment data and curve for eastern Bering Sea rock
sole Pleuronectes bilineatus. Age-3 biomass (lagged 3 yr) is
plotted against stock biomass for the years 1979-88.
Weight (kg)
1.0
^ . ^^
08
^^
06
.^/f
0.4
■y^
0.2
/^
00 5.0 lO.O 15.0 20.0 25 0
Age (years)
Figure 5
Weight-at-age data and relationship for eastern Bering Sea
rock sole Pleuronectes bilineatu.s. Data are from the 1986 trawl
survey conducted by the Alaska Fisheries Science Center.
stant growth rate, whereas other commonly used func-
tions exhibit decreasing growth rates at upper ages
(Beverton and Holt 1957), usually manifested in the
form of an upper asymptote. In practice, however, the
absence of an asymptote may be inconsequential or
even preferable (Knight 1968, Ricker 1979) for two
reasons: (1) In exploited populations, individuals may
only rarely survive to reach the portion of the growth
curve where a marked decrease in growth rate would
be most discernible; and (2) in functional forms that in-
corporate an asymptote, this parameter is often poor-
ly estimated, being highly correlated with at least one
other parameter in the equation.
Thompson Management advice from a simple dynamic pool model
559
Robustness of the rules of thumb
Neither Rule I nor Rule II is particularly robust when
applied to simple dynamic pool models in general or
to the model developed here in particular. Rule I can
drastically over- or underestimate the true relationship
between Fmsy and M. When q exceeds 0.5, Rule I con-
sistently overestimates the ratio between Fmsy and M,
whereas when q is less than 0.5, the ratio can range
both well above and well below the value suggested by
Rule I.
Although these results do not provide much theoret-
ical support for Rule I, it is still possible that Rule I
holds as an empirical generalization (it turned out to
be fairly close in the case of eastern Bering Sea rock
sole, for example). If Rule I does hold as an empirical
generalization, Equation (33) indicates that this implies
an inverse relationship between the relative importance
of growth in pristine production (K") and the degree
of density-dependence in the stock-recruitment rela-
tionship (q). Further work is necessary to see if such
an inverse relationship is supported on the basis of life
history or other theory.
Rule II consistently overestimates the ratio between
B(Fmsy) and B(0) in the model presented here (Eq.
38, Fig. 2). In the case of eastern Bering Sea rock sole.
Rule II was off by 51%. The problem with Rule II stems
from the "diminishing returns" nature of the relation-
ship between F and B(F), wherein successive increases
in F result in less and less of an impact on biomass.
Rule II, on the other hand, was inspired by the Schaefer
(1954) model, in which the relationship between F and
B(F) is linear (i.e., it exhibits constant returns to scale).
Interestingly, the upper asymptote displayed in Fig-
ure 2 corresponds exactly to the asymptote observed
in a pair of surplus production models proposed by Pella
and Tomlinson (1969, reparametrized by Fletcher 1978)
and Fowler (1981), models that are conceptually very
different from the one presented here. Mathematical-
ly, the isomorphism stems from the fact that all three
models involve functions that raise a parameter x to
an exponent of the form 1/(1 -x). The fact that this
result can be obtained from both surplus production
and dynamic pool models indicates that it may be
worthy of further investigation.
Since the rule of thumb setting MSY/MB(0) equal
to 0.5 was derived by multiplying Rules I and II, it is
affected by the upward bias inherent in Rule II. This
is reflected in the eastern Bering Sea rock sole ex-
ample, where the estimated value for the MSY/MB(0)
ratio was only 0.216. It appears that the "MSY/MB(0)
rule" can be a good approximation only when Rule I
results in a major underestimate, which in the context
of the model developed here requires two things: (1)
Recruitment must be relatively independent of stock
size, and (2) pristine production must be relatively
dependent on recruitment (Fig. 3). Another conse-
quence of this relationship is that Rule I can never hold
when MSY/MB(0) = 0.5, and vice-versa. This conclusion
stands in stark contrast to the traditional view which
holds that the MSY/MB(0) rule derives from Rule I.
Instead, it seems more likely that the two are mutual-
ly exclusive, at least in the context of simple dynamic
pool models.
Acknowledgments
I would like to thank James Balsiger, Nicholas Bax,
Roderick Hobbs, Daniel Kimura, Richard Methot, and
Thomas Wilderbuer of the Alaska Fisheries Science
Center for reviewing all or portions of this paper in
various stages of development. Comments provided by
Ian Fletcher of the Great Salt Bay Experimental Sta-
tion were especially helpful. Three anonymous review-
ers also supplied constructive suggestions.
Citations
Alverson. D.L., and W.T. Pereyra
1969 Demersal fish explorations in the northeastern Pacific
Ocean— an evaluation of exploratory fishing methods and
analytical approaches to stock size and yield forecasts. J. Fish.
Res' Board Can. 26:1985-2001.
Beddington, J.R., and J.G. Cooke
1983 The potential yield of fish stocks. FAO Fish. Tech. Pap.
242, 47 p.
Beverton. R.J.H.. and S.J. Holt
1957 On the dynamics of exploited fish populations. Fish
Invest. Minist. Agric, Fish. Food (G.B.) Ser. 2, 19, 533 p.
Gushing, D.H.
1971 The dependence of recruitment on parent stock in dif-
ferent groups of fishes. J. Cons. Cons. Int. Explor. Mer 33:
340-362.
Deriso, R.B.
1982 Relationship of fishing mortality to natural mortality and
growth at the level of maximum sustainable yield. Can. J.
Fish. Aquat. Sci. .39:10.54-1058.
Fletcher, R.I.
1978 Time-dependent solutions and efficient parameters for
stock-production models. Fish. Bull.. U.S. 76:377-388.
Fowler. C.W.
1981 Density dependence as related to life history strategy.
Ecology 62:602-610.
Francis. R.C.
1974 Relationships of fishing mortality to natural mortality
at the level of maximum sustainable yield under the logistic
stock production model. J. Fish. Res. Board Can. 31:
1539-1.542.
Gulland, J. A.
1971 The fish resources of the ocean. Fishing News (Books)
Ltd.. Surrey. England, 255 p.
Hulme, H.R., R.J.H. Beverton. and S.J. Holt
1947 Population studies in fisheries biology. Nature (Lond.)
159:714-715.
560
Fishery Bulletin 90(3). 1992
Kimura, D.K., and S. Chikuni
1987 Mixtures of empirical distributions: An iterative applica-
tion of the age-length key. Biometrics 43:23-35.
Knight, W.
1968 Asymptotic growth: An example of nonsense disguised
as mathematics. J. Fish. Res. Board Can. 25:1303-1307.
Larkin, P.A.
1977 An epitaph for the concept of maximum sustainable
yield. Trans. Am, Fish. Soc. 106:1-11.
Paulik, G.J.
1973 Studies of the possible form of the stock-recruitment
curve. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer 164:
302-315.
Pella, J.J., and P.K. Tomlinson
1969 A generalized stock production model. Bull. Inter-Am.
Trop. Tuna Comm. 13:419-496.
Pitcher, T.J., and P.J.B. Hart
1982 Fisheries ecology. AVI Publ. Co., Westport, CT, 414p.
Ricker. W.E.
1979 Growth rates and models, /n Hoar, W.S., D.J. Randall,
and J.R. Brett (eds.), Fish physiology, vol. 8, p. 677-743.
Academic Press, NY.
Russell, E.S.
1931 Some theoretical considerations on the 'overfishing' prob-
lem. .1. Cons. Cons. Int. Explor. Mer 6:3-20.
Schaefer, M.B.
1954 Some aspects of the dynamics of populations important
to the management of the commercial marine fisheries. Bull.
Inter-Am. Trop. Tuna Comm. 1:27-56.
Schnute, J.
1981 A versatile growth model with statistically stable param-
eters. Can. J. Fish. Aquat. Sci. 38:1128-1140.
Shepherd, J.G.
1982 A family of general production curves for exploited popu-
lations. Math. Biosci. 59:77-93.
Walters, G.E., and T.K. Wilderbuer
1988 Rock sole. In Condition of groundfish resources of the
eastern Bering Sea and Aleutian Islands region in 1988, p.
106-117. Unpubl. manuscr. submitted to Int. N. Pac. Fish.
Comm., Sept. 1988. NMFS Alaska Fish. Sci. Cent., Seattle,
219 p.
Abstract. - a simple dynamic
pool model is used to examine the
problem of stock-recruitment param-
eter uncertainty from a Bayesian
perspective. Probabilities associated
with different parameter values are
used to weight the losses (i.e., oppor-
tunity costs to society) associated
with any given fishing mortality
rate. By choosing appropriate forms
for the loss and probability density
functions, the model is shown to re-
sult in an analytic solution. Because
this solution gives the fishing mor-
tality rate that maximizes the ex-
pected value of the logarithm of sus-
tainable yield, it is denoted Fmelsy-
The solution is a monotone-decreas-
ing function of parameter uncertain-
ty, converging on the fishing mortal-
ity rate corresponding to maximum
sustainable yield as the degree of
uncertainty approaches zero. As an
empirical illustration, the model is
applied to the eastern Bering Sea
stock of rock sole Pleuronectes bi-
lineatus.
A Bayesian approach to management
advice when stocl<-recruitment
parameters are uncertain
Grant G. Thompson
Resource Ecology and Fisheries Management Division
Alaska Fisheries Science Center, National Marine Fisheries Service, NOAA
7600 Sand Point Way NE, Seattle, Washington 981 15-0070
Manuscript accepted 4 June 1992.
Fishery Bulletin, U.S. 90:561-573 (1992).
Exploiting a stock at the fishing mor-
tality rate (F) associated with max-
imum sustainable yield (MSY) is a
common fishery management strate-
gy. For the most part, three simple
propositions are sufficient to justify
this strategy: (1) The stock exhibits
a sustainable yield determined by the
fishing mortality rate, (2) more sus-
tainable yield is always preferable to
less, and (3) the parameters underly-
ing the stock's dynamics are known
with certainty. However, parameters
governing stock dynamics are typi-
cally not known with certainty, and
in such cases it is possible to demon-
strate that the appropriate F value
may be less than the value corre-
sponding to MSY (Fmsy)-
The approach to be used in this
demonstration is taken from Baye-
sian decision theory (e.g., Raiffa
1968, DeGroot 1970). Early applica-
tions of Bayesian theory to fisheries
problems were presented by Roths-
child (1972), Lord (1973, 1976),
Walters (1975), and Walters and Hil-
bom (1976). Of the many more recent
applications, those presented by Lud-
wig and Walters (1982), Clark et al.
(1985), and Walters and Ludwig
(1987) bear most closely on the pres-
ent study.
For simplicity, it will be assumed
here that stock dynamics are deter-
ministic but governed by parameters
which may be imprecisely estimated.
This approach is distinct from the
more common one of assuming that
stock dynamics are the product of a
deterministic system (with param-
eter values given and fixed) modified
by a random error term. Important
early examples of the latter approach
include Ricker (1958), Larkin and
Ricker (1964), and Tautz et al. (1969).
Ludwig and Walters (1982) and
Mangel and Clark (1983) incorporate
both approaches in a systematic fash-
ion which makes the distinction espe-
cially clear.
The basic model
Thompson (1992) developed a simple
dynamic pool model which can be
solved explicitly for Fmsy- Ii terms
of biomass per recruit, the model is
basically that of Hulme et al. (1947);
thus, body weight is taken to be a
linear function of age, with intercept
ao . The main departure from Hulme
et al. is that biomass at recruitment
age ar is taken to be proportional to
stock biomass raised to a power q
(Cushing 1971). With these specifica-
tions, sustainable jdeld Y(F) can be
written
Y(F)
(1)
l-HK"-fF'
(1 + F')2
1
l-q
where M is the instantaneous rate of
natural mortality, F' = F/M, p is the
proportionality term in the Cushing
stock-recruitment relationship, and
K" = l/[M(ar - ao)] (which can be in-
terpreted in this model as the pristine
ratio of growth to recruitment). The
561
562
Fishery Bulletin 90(3), 1992
Gushing exponent q is constrained to fall between 0
and 1. In the limiting case of q = 0, recruitment is con-
stant, while in the other limiting case of q= 1, recruit-
ment is proportional to biomass.
Differentiating Equation (1) with respect to F and
setting the resulting expression equal to zero gives the
following equation for Fmsy •
F'
MSY =
(q+ 1) K" + 1 + V(q+1)2K"2 + (6q- 2)K" + 1
2q
(2)
1,
where F'msy = Fmsy/M.
A common rule of thumb is that F'msy should equal
1. The locus of parameter values for which this rule
holds precisely is given by
1
K" = - - 2.
q
(3)
where L[z(F, q)] represents the losses resulting from
selection of a particular value of F given a particular
value of q, and E(L[z(F, q)]} is the expected value of
L[z(F, q)] (the "risk," DeGroot 1970). The minimum
value of E (L [z(F, q)] } is referred to as the "Bayes risk"
(DeGroot 1970). The integral is taken over the inter-
val 0 to 1 because the Gushing stock-recruitment rela-
tionship constrains q to that range.
The Bayes decision can be derived by differentiat-
ing E {L[z(F, q)]} with respect to F and solving for the
value that sets the derivative equal to zero. The valid-
ity of this procedure requires that all parameter values,
including those describing P(c[), remain constant into
the future. The solution corresponding to such an
assumption is sometimes known as a "myopic Bayes"
solution (Ludwig and Walters 1982, Mangel and Glark
1983, Mangel and Plant 1985, Parma 1990). A more
general alternative is to allow for the possibility that
parameter estimates will be updated in the future,
but this approach is vastly more difficult (Glark et al.
1985, Mangel and Plant 1985, Walters and Ludwig
1987).
Analyzing the model
in a Bayesian framework
Parameter estimates in any model are by definition
associated with some degree of uncertainty. For ex-
ample, parameters governing the stock-recruitment
relationship are particularly difficult to estimate
precisely (Larkin 1973, Paulik 1973, Ludwig and
Walters 1981, Walters and Ludwig 1981 and 1987,
Shepherd 1982, Glark 1985, Glark et al. 1985, Roths-
child and Mullen 1985, Shepherd and Gushing 1990).
In the presence of such uncertainty, a Bayesian ap-
proach would use the probabilities associated with
different parameter values to weight the losses (i.e.,
opportunity costs to society) associated with choosing
a particular fishing mortality rate. Following similar
studies by Ludwig and Walters (1982), Glark et al.
(1985), and Walters and Ludwig (1987), the present
analysis will focus on the uncertainty surrounding a
single parameter, in this case the stock-recruitment
exponent q. This uncertainty takes the form of a prob-
ability density function (pd^ P(q) which describes the
relative credibility of alternative q values.
To simplify notation, define z(F, q) as the ratio of
Y(F) to MSY for an arbitrary value of q drawn from
P(q). Then, the "Bayes decision" (DeGroot 1970) is the
value of F that minimizes
E{L[z(F, q)]}
X'
L[z(F, q)l P(q) dq, (4)
Minimizing risl< under
a logarithmic loss function
Of course, specification of the functions L and P is
crucial to this problem. Following Lord (1976) and Lud-
wig and Walters (1982), one possible choice is to assume
that L is a linear function of z(L(z)=l-z). Another
common form is the quadratic L(z) = (l -z)-. which has
been used in the fisheries literature by Walters (1975),
Hightower and Grossman (1987), and Gharles (1988).
One of the oldest alternatives is the logarithmic loss
function, L(z)= -ln(z), dating back to the work of Ber-
noulli in 1738 (transl. 1954). Logarithmic loss (or, con-
versely, utility) seems first to have been used in the
fisheries literature by Gleit (1978), followed by Lewis
(1981, 1982), Mendelssohn (1982), Opaluch and Bock-
stael (1984), Ruppert et al. (1984, 1985), Deriso (1985),
Walters (1987), Walters and Ludwig (1987), Getz and
Haight (1989), Hightower and Lenarz (1989), High-
tower (1990), Parma (1990), and Parma and Deriso
(1990).
Linear, quadratic, and logarithmic loss functions are
compared in Figure 1. As Figure 1 indicates, the
logarithmic loss function corresponds to a "preserva-
tionist" viewpoint, in which extinction of the stock is
absolutely unacceptable (i.e., the loss corresponding to
extinction is infinite). Because the logarithmic loss func-
tion is clearly identifiable as a risk-averse alternative
function (see Discussion), it is a good candidate for il-
lustrating how a Bayesian approach can differ from
more traditional approaches which do not incorporate
uncertainty in an explicit fashion.
Thompson: Bayesian approach to management advice from a simple dynamic pool model 563
To incorporate the logarithmic loss concept into the model, first note that Equation (1) allows z(F, q) to be written
1
r/D\ /i+K"+F'\ii-<5
F
Y(F) L\M/\ (1-F')^ /J F' /1 + F'msy\M 1 + K" + F' \i-q
z(F, q) = ^-^ = —^ = 3 I 1^ I r-zr. 1 • (5)
MSY
MSY
fp \/1 + K" + F'msy\
,M (1 + F'msy)^
j_ F'
l-q
MSY
1 + F'
1 + K" + F'
MSY,
For an arbitrary value of q, the (logarithmic) loss associated with a given choice of F is thus
,,,^ ,, , ,!.' ^ 2 1n(l + FMSY)- ln(l + K" + FMSY) ,,„,, 2 1n(l + F') - ln(l + K" + F-)
L[z(F, q)] = In(FMSY) ln(F ) + : • (6)
l-q
l-q
Substituting Equation (6) into Equation (4), the risk can be written
i.,Tr.T. M, r'i>. ^(w^>' ^ 2 1n(l + F'MSY) - ln(l + K" + F-msy)\ ,
E{L[z(F, q)]} = I P(q) ln(F msy) - dq
Jo l-q /
/,
P(q) In(F')
2 ln(l + F') - ln(l + K" + F')\
l-q
dq.
(7)
From Equation (2), it is clear that F'msy involves only K" and q. Thus, regardless of the form of P(q), the first
integral on the right-hand side of Equation (7) is independent of F. Therefore, the problem of finding the Bayes
decision is equivalent to minimizing the second integral on the right-hand side of Equation (7). Remembering that
the integral (taken over the interval 0 to 1) of a cons-
tant multiplied by P(q) is equal to the constant itself,
the following proxy objective function is obtained:
Llz(F,q)l
logarithmic
Figure 1
Three possible loss functions. Loss, or relative utility foregone,
is plotted against the ratio of Y(F)/MSY for quadratic, linear,
and logarithmic loss functions.
Ei{L[z(F, q)]} = -ln(F') +
' P(q)
[21n(l-t-F') - ln(l-hK"-HF')] I dq. (8)
J 0 q
Incorporating a beta
probability density function
The next step in determining the Bayes decision is to
select a form for the pdf P(q). Bayesian decision theory
frequently makes use of the beta family of pdfs (e.g.,
DeGroot 1970, Holloway 1979). The beta distribution
would seem to be a natural candidate for P(q), since
it constrains q to the necessary (0,1) range. In its
standard form, the beta distribution can be written
564 Fishery Bulletin 90(3). 1992
/ r(o+/3) \ „ ,
\ r(a) r{p) j
where a and p are positive constants and r(-) is the gamma function, which, except for r(l)= 1, can be described
in terms of the recursion formula
r(a) = (a-l)r(a-l). (10)
By Equations (9) and (10), then, the integral in Equation (8) can be evaluated as follows:
■ip(q), / r(a + /3) \ r' w, .«_, . / r(a + /3) Wr(a)r(/3-l)\
r ^dq=P^^ f q-(l-q)^-dq
J ^, 1-q \r(a)r(/?)|J„
r(a)r(/j)/\ r(a+/3-i)
fr(/?-l)W r(a + P) \ a + /J-l
(11)
\ r(/?) /\r(a + /5-l)/ /3-1
Substituting Equation (11) into Equation (8) then gives
[2 1n(l + F') - ln(l + K" + F')](a + /?-l)
Ei{L[z(F, q)]} = - In(F') + ^ ^^^ -. (12)
^-1
Differentiating Equation (12) with respect to F' and setting the resulting expression equal to zero yields the
quadratic expression
aF'2 + [K"(2a + /3-l) + a - /? + 1] F' - (/3-l)(K"+l) = 0. (13)
Before solving Equation (13), it would be helpful to cast the solution in terms of parameters which are more
intuitive than a and ft, for example the mean and variance of P(q). The beta distribution has mean m and variance
V as follows:
a a/?
m = and v = . (14) and (15)
a + p {a + P)~{a + p+l)
Conversely, Equations (14) and (15) can be solved simultaneously to describe a and /? in terms of m and v:
/m(l-m) \ /m(l-m) \
a = [^ ^-1 m and P = \— ^-1 (1-m). (16) and (17)
Unlike the normal distribution, the variance of the beta distribution exhibits a maximum possible value for a
given mean. Remembering that a and p are constrained to be positive, the maximum possible value of v can be
derived from either Equation (16) or Equation (17) by setting the left-hand side equal to 0 and solving for v. This
exercise results in a maximum v equal to m(l - m). Thus, a and p can be written in terms of the mean and a scaled
variance v | = 1 as follows:
m(l-m)
Thompson, Bayesian approach to management advice from a simple dynamic pool model
565
1 m
and
P = (— - ll(l-m)
(18) and (19)
For a given set of K", m, and v' values, Figure 2 shows the risk (depicted by the area under a particular curve)
associated with three possible F' values.
Fishing mortality at maximum expected log-sustainable yield
Substituting Equations (18) and (19) into Equation (13) and solving for F' gives the value that minimizes risk.
Because of the form used for the loss function, this process is equivalent to finding the level of F' that maximizes
the expected value of the logarithm of sustainable yield. It is thus convenient to refer to this value as F'melsy
(for "maximum expected log sustainable yield"), which for this particular model can be written
[(m + 2) K"-2] v' - (m+1) K" + 1 + Vks v'^ - k, v' + ko
F'melsy = r 7, r - 1-
2 m(l-v)
(20)
where k. = (m + 2)2 K"~ + (12m -8) K" + 4,
ki = (2m2 + 6m + 4) K"2 + (18m -8) K" + 4, and
ko = (m+l)2 K"- + (6m-2) K" + 1.
Figure 3 illustrates how F'melsy varies with K", m, and v'. A few special cases are of particular interest. For
example, when q is known with certainty, i.e., m = q and v' = 0. Equation (20) reduces to Equation (2). Equation
(2) is thus the "certainty equivalent" solution (Ludwig and Walters 1982). The ratio between Fmelsy ^nd Fmsy
is illustrated in Figure 4. Differences in K" tend to have less influence on this ratio than differences in either
m or v'.
Other important special cases of Equation (20) include the limits as K" approaches zero and infinity, which are
shown respectively below:
lim F'melsy
K"-0
l-m(l-v') - 2v'
m(l-v')
l-m(l-v') - 2v'
and hm F melsy = :; " "•
K'^oo l + m(l-v ) - 2v
(21) and (22)
Llz(F,q)lP(q)
0.25
^ A
0 20
/ \
0.15
0.10
f\\/:^..
0.05
K\\/x^^^^^^V
f Wv/.^-^-?^^^^^^^^
0 0 0 1 0 2 0 3 0.4 0.5 0 6 0.7 0 8 0 9 10
Recruitment parameter q
Figure 2
Risk under different F levels. The area under a curve is the
risk associated with the F level that defines the particular
curve. Parameter values used to generate these curves were
K' = 2.5, m = 0.2. and v' = l/ll.
3.0
Fmelsy
\ \
2.5
\ \
' \k =1, v =1/3
2.0
\ \k'=1. v=0
1.5
1.0
0.5
^^^^^
K =3, V =1/3*^^
.^ K"=3rv^5''^==============>._.___^
0.0 0.2 0.4 0.6 0.8 10
Expected value of q (m)
Figure 3
Values of F'j,j.lsy resulting from different combinations of
parameter levels. F'melsy tends to decrease as K", m, or v'
increases.
566
Fishery Bulletin 90(3). 1992
Fmelsy/Fmsy
1 0 V
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Scaled variance v
Figure 4
Ratio of F'jiELSY ^° F'^jy under different combinations of
parameter levels. The ratio tends to decrease as K" decreases
or as m or v' increases.
Equation (20) also implies that F'melsy falls to zero
whenever v' reaches a critical value v'o defined as
Scaled variance v
00 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Expected value of q (m)
Figure 5
Limiting values of v'. The solid curve shows v'„ , the locus at
which F'j,E.Lsv=0- The dashed curve shows v', , the locus
limiting the parameter subspace for which F'melsy ^^^ ^^~
ceed 1. For (m, v' ) combinations below the v', curve, F'j,elsy
can take any value, depending on K'. For (m, v' ) combinations
between the two curves, F'^^lsy can range between 0 and 1.
again depending on K". For (m, v' ) combinations on or above
1-m
Vo =
(23)
■m
By Equation (19), v'o corresponds to a /3 value of 1.
Whenever p< 1, the right-hand tail of the beta distribu-
tion fails to reach zero, implying a non-zero probabil-
ity that q= 1. When q= 1, any positive F value causes
the stock to go extinct. Given the preservationist at-
titude implicit in the logarithmic loss function, any
possibility of extinction is unacceptable, so F'melsy
drops to zero in this case. Note that F'melsy is never
positive for values of v' greater than 0.5.
Just as Equation (2) could be solved to determine the
locus of parameter values under which F'msy takes on
the special value of 1 (Eq. 3), Equation (20) can be
solved to determine the following locus of parameter
values under which F'melsy = 1-
K"
l-2v'
m(l-v')
(24)
In the certainty equivalent case, Equation (24)
reduces to Equation (3). As K" approaches zero, Equa-
tion (24) defines an upper limit on v' (v'l) for the
special case where F'melsy = 1:
Vi =
l-2m
2-2m'
(25)
Under Equation (3), F'msy could exceed 1 only if q
were less than 0.5. While Equation (25) implies essen-
tially the same property (replacing F'msy with F'melsy
and q with m), it adds a similar restriction on v', namely
that F'melsy can exceed 1 only if v' is less than 0.5.
[Note that this is a weaker version of the restriction
implied by Equation (23). Equations (23) and (25) are
compared in Figure 5.]
Biomass at MSY compared
with biomass at MELSY
Dividing Equation (1) through by F gives equilibrium
stock biomass. By substituting Equations (20) and (2)
into this expression and setting q = m, the ratio of stock
biomass at MSY to stock biomass at MELSY is given
by
B(Fmsy)
B(Fmelsy)
(26)
/F'
MELSY
+ V
K"-!- F'msy + 1
F'
with limits
MSY
-1-1
iK"-hF'
MELSY
r + l
1
1-m
Thompson Bayesian approach to management advice from a simple dynamic pool model 567
1 1
/ B(Fmsy) \ /l-2v'\i-'" / B(Fmsy) \ I (l + m)(l-2v') \^"^
lim — ^ ^^^' = and lim — ^ = \— ~ ^ . (27) and (28)
K--0 \B(FMELsy)/ \ 1-v' / K-oc \B(Fmelsy)/ \(l + m)(l-v') - v'/
Equations (26-28) decline from a value of 1 at v' = 0 to a minimum at v' = v'o. The minimum value depends on
K" and m, but is never greater than 1/e.
Estimating the parameters of the beta distribution
To fit the Gushing stock-recruitment curve to a set of n stock-recruitment data points, it seems reasonable to
assume the following model:
Yi = P -H qxi -I- £i, (29)
where x; represents the natural logarithm of the ith stock biomass datum, y; represents the natural logarithm
of the ith recruitment datum (lagged according to the age of recruitment), p = ln(p), and £, is an independent error
term distributed as N(0,o-).
Press (1989) presented a Bayesian approach to estimating the parameters of the pdf of q using Equation (29)
as the underlying model. The following paragraphs summarize this presentation, which begins by rephrasing the
problem in the form of Bayes' theorem:
h(q,p,o I X, y) oc
|~| f(yi I Xi,q, p,o)
i = l
gi(q) g2(p) g3(o), (30)
where x is the vector (xj, . . ., Xn)'; y is the vector (yi, . . ., yn)'; h(q, p,o | x, y) represents the posterior pdf
of the parameters q, p, and o; f(yi | Xi,q, p,o) represents the conditional pdf of yj given the observed value of X;
and any particular values of q, p, and o; and gj() represents the prior pdf of the jth parameter.
Given the assumptions implicit in Equation (29), f(yi | Xi,q,p,a) can be written
/- (yi-p-qx,)-\
exp
2o2
f(yi I x,,q,p,o) = ^ —^ '-. (31)
A special case of interest is the one in which the gj(-) are all "vague" (also called noninformative or indifference)
priors. These are pdfs which reflect indifference regarding the probability of alternative parameter values. Press
(1989) treated gi(q) and g2(p) as constants, implying that all values on the real line are equally likely in the prior
distribution. Since a is constrained to be positive, however. Press set g3(o) = l/o, reflecting a uniform prior
distribution for ln(o).
Using Equation (31) and the priors specified by Press (1989), Eq. (30) gives a straightforward solution. The
classical least-squares estimates of q and p (q and p, respectively) obtain as the maximum-likelihood estimates.
In their posterior pdf, q and p jointly follow a bivariate Student's t distribution, so that marginally the posterior
pdf of q, hi(q | x, y), follows a univariate 3-parameter t distribution with n-2 degrees of freedom:
n-l\
hi(q I x,y) = , (32)
n(n- 2) si 1 -i-
M (n-2)s2
568
Fishery Bulletin 90(3). 1992
where s% is the estimated variance of q given by
r-'
i = l
^.r
''q -
n
(n-2) ^ (X,-
i = l
-X)2
(33)
For the present application, the solution given by
Press (1989) needs to be modified in only one respect.
His suggested form for g] (q) implies a uniform distri-
bution over the entire real line, whereas here P(q) has
been specified a priori to be zero for all values less than
0 or greater than 1. Given Equations (30) and (31), this
implies that the suggested uniform shape for gi(c|)
should be truncated outside the range 0 to 1. This in
turn implies that hi(q | x,y) should also be truncated
outside the range 0 to 1 (and rescaled appropriately).
Strictly speaking, then, P(q) follows a truncated t
distribution in this approach, rather than the hypothe-
sized beta. However, a beta distribution can be made
to approximate the truncated t by solving for m and
v as follows:
/:
q hi(q I x,y) dq
m =
;
(34)
hi(q I x,y) dq
and
;:
(q-m)2 hi(q | x,y) dq
V =
r hi(q I x,y) dq
0
(35)
Applying the model to rock sole
As an illustration of the approach suggested above, the
model can be applied to the eastern Bering Sea stock
of rock sole Pleuronectes bilineatus. This stock is ex-
ploited by a multispecies flatfish fishery, and is also the
target of an important roe fishery (Walters and Wilder-
buer 1988).
The parameters to be estimated are K", m, and v'.
Thompson (1992) estimated K" for this stock at a value
of 3.279, and described a set of stock and recruitment
data (n = 7) which can be used to estimate m and v'. Fit-
ting Equation (29) to these data gives q = 0.235 and
s"q = 0.114 (Fig. 6). Substituting these parameters into
Equations (34) and (35) gives m = 0.369 and v = 0.057,
with v' = 0.243. The relationship between the truncated
t distribution defined by these values and the beta ap-
proximation is shown in Figure 7 (i?- = 0.97).
With parameter values K" = 3.279, m = 0.369, and
v' = 0.243, Equation (20) gives F'melsy = 0.365. Multi-
plying through by M (set at 0.2 by Walters and Wilder-
buer 1988) gives F'melsy =0.073. Substituting m for
q in Equation (2) yields F'msy =0.607, or Fmsy =0.121.
This value of Fmsy differs somewhat from the value
of 0.176 given by Thompson (1992), which was based
on the least-squares estimate of q (q) instead of the
Recruitment biomass (thousands of t)
0 0 0 1 0 2 0,3 0-4 0.5 0 6 0 7 0 8 0 9 10
stock biomass (millions of t)
Figure 6
Stock-recruitment data and curve for eastern Bering Sea rock
sole Pleuronectes bilineatus. Age-3 biomass (lagged 3 yr) is
plotted against stock biomass for the years 1979-88. The curve
is the least-squares fit.
Probability density
t distribution
0 0 0.1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0.9 1.0
Stock-recruitment exponent q
Figure 7
Comparison of truncated t and beta pdfs for the stock-
recruitment exponent q in the eastern Bering Sea rock sole
PleuroTwctes bilineatus example.
Thompson Bayesian approach to management advice from a simple dynamic pool model
569
Bayesian estimate (m). These two Fmsy values bracket
the value of 0.155 which Walters and Wilderbuer (1988)
derived from a surplus production model. Regardless
of which Fmsy value is chosen, however, it exceeds
Fmelsy by a significant amount.
Discussion
Evaluation of assumptions
The approach described here consists of three main
components: the basic model represented by Equation
(1), the logarithmic loss function, and the beta form for
P(q). These components were chosen in part because
they are tractable, making possible the analytic solu-
tion for F'melsy given by Equation (20). In addition,
each has some degree of theoretical support, as de-
scribed below.
The basic model The basic model was evaluated by
Thompson (1992). In brief, the model includes terms
for all of the requisite features of dynamic pool models
(recruitment, growth, natural mortality, fishing mor-
tality). The distinguishing features of the model (linear
growth and a Gushing stock-recruitment relationship)
satisfy the principal theoretical requirements for
growth and stock-recruitment functions given by
Schnute (1981) and Ricker (1975), respectively. Al-
though the basic model is a simple one, it approximates
more complicated models fairly well under a wide range
of parameter values.
Logarithmic loss function The logarithmic loss func-
tion may require a bit more discussion. As mentioned
earlier, this loss function is only one of several pos-
sibilities, two of the other most-common being the
linear and quadratic forms. The principal argument
against the linear loss function is that it implies strict
risk neutrality, whereas most individuals tend to be at
least somewhat risk-averse. Thus, if fishery managers
tend to be risk-averse, a linear loss function would be
inappropriate, except over a narrow range of yield
values.
In contrast, the quadratic loss function implies a
degree of risk aversion. In addition, the quadratic form
has properties which prove convenient for a number
of statistical applications. However, it has also been the
subject of substantial criticism (Pratt 1964, Samuelson
1967, Box and Tiao 1973). Although the quadratic loss
function does fall into the "risk-averse" category, this
functional form manifests its risk aversion somewhat
perversely by exhibiting increasing absolute risk aver-
sion (Pratt 1964). In other words, a fishery manager
using a quadratic loss function would be less willing to
take risks as yields became higher.
The logarithmic loss function is another risk-averse
alternative. It can be described as a special case of the
isoelastic marginal loss function defined by L(z) =
(l-z'^)/<t>, where ^>0 (the logarithmic case being ob-
tained in the limit as <t> approaches zero). Unlike the
quadratic loss function, isoelastic marginal loss func-
tions exhibit decreasing absolute risk aversion (Pratt
1964). Isoelastic marginal loss functions also display
the conveniens property of constant relative risk aver-
sion R(z), defined as -zL"(z)/L'(z) (Pratt 1964). Spe-
cifically, R(z) = l-it> for the isoelastic marginal loss
family. The logarithmic case, where R(z)=l, thus
represents a clear risk-averse alternative to the risk-
neutral linear loss function, where ij> = l and R(z) = 0.
The fact that the logarithmic loss function tends
toward negative infinity as the resource approaches ex-
tinction may be viewed as problematic by some. On the
other hand. Smith (1985) views this behavior as a re-
quisite characteristic for any loss function to be used
in the context of renewable resources, arguing that it
"introduces a useful conservation motive into the deci-
sion making process." Opaluch and Bockstael (1984)
go even further, stating, "It is well known that the log
function exhibits the best properties of the simple func-
tional forms. ..."
Beta probability density function The principal
justification for using the beta pdf to describe P(c|) is
that the beta is a natural choice for the pdf of any con-
tinuous variable which is constrained to fall within the
0 to 1 range. The fact that it allows for an explicit solu-
tion to Equation (7) is another argument in its favor.
Unfortunately, the method presented here for esti-
mating the parameters of P(q) is based on a model
(Press 1989) which yields a truncated t distribution,
not a beta distribution. If this model is accepted as a
true description of reality, then the beta form for P(c[)
is only an approximation. Of course, most functional
forms used in modeling are only approximations, so the
question is whether the advantages of increased tract-
ability provided by the beta distribution outweigh any
attendant losses of accuracy. Holloway (1979) argues
in the affirmative after noting the difficulty of identi-
fying natural processes which yield the beta distribu-
tion as a formal result.
In general, the effectiveness of Bayes decisions is
relatively insensitive to small changes in the assumed
pdf (DeGroot 1970). This being the case, the question
really is whether the difference between the truncated
t distribution and the beta approximation is typically
small. To assess the magnitude of this difference, the
goodness-of-fit between the truncated t and beta dis-
tributions was examined for a wide range of n, q,
and s^q values (Fig. 8). Note that R^>0.95 for a wide
range of parameter values, indicating that the loss of
570
Fishery Bulletin 90(3). 1992
accuracy resulting from the beta approximation is often
small.
Another fact to keep in mind is that the model pre-
sented by Press (1989) is only one possibility. Despite
the pessimism conveyed by Holloway (1979), it is con-
Estimated variance of q
0 2 0 3 0-4 0 5 0,6 0 7 0,8 0 9 10
Least-squares estimate of q
Figure 8
Loci of parameter values under which a beta approximation
to the truncated t distribution gives an R- value of 0.95. if '
was calculated by comparing the two distributions at q values
of 0.01, 0.02 0.99. For n =5, parameter combinations
lying to the interior of the two curves correspond to i?-
values <0.95. For n = 10, R' values <0.95 correspond to
parameter combinations lying above the curve.
ceivabie that other models could yield the beta distribu-
tion as an exact result.
Comparison with previous studies
Of the many previous applications of Bayesian decision
theory to fisheries, the studies by Ludwig and Walters
(1982), Clark et al. (1985), and Walters and Ludwig
(1987) are most closely related to the present work. The
various features of the four approaches are outlined
in Table 1. The three previous studies exhibit certain
common features which distinguish them from the
present study, namely: (1) use of a discrete time scale;
(2) inclusion of an explicit adaptive management strate-
gy; (3) inclusion of environmental stochasticity as well
as parameter uncertainty; (4) inclusion of a positive dis-
count rate in the objective function; (5) assumption of
a normal form for the pdf of the uncertain parameter;
and (6) inability to derive an exact analytic solution,
even in the myopic case (except for one special instance
considered by Clark et al.). The present study is also
the only one of the group which includes both a
biomass-based model and a risk-averse loss function.
Ludwig and Walters (1982) found that the deter-
ministic optimum escapement level can be less than half
the value of the Bayesian solution. Although the con-
tinuous form of the model used in the present study
makes it difficult to talk about escapement per se,
equilibrium stock size might serve as a suitable proxy
Table 1
Comparison of four studies describing Bayesian approaches to
fishery management.
Ludwig and Walters
Clark et al.
Walters and Ludwig
Feature
(1982)
(1985)
(1987)
This study
Time scale
discrete
discrete
discrete
continuous
Yield metric
numbers
biomass
numbers
biomass
Adaptive strategy included
yes
yes
yes
no
Age structure included
no (discrete generations)
yes
no (discrete generations)
yes
Discounting included
yes
yes
yes
no
Harvesting costs included
no
yes
no
no
Stochasticity included
yes
yes
yes
no
Loss function
linear
linear
logarithmic
logarithmic
Growth function
none
isometric von Bertalanffy
none
linear
Stock-recruitment function
Ricker (1954)
a) Cushing
b) stock-independent
c) linear-threshold
Cushing
Cushing
Uncertain parameter
Ricker exponent
a) In (Cushing multiplier)
b) mean In (recruitment)
c) mean In (recruitment)'
Cushing exponent (q)
q
Pdf
normal
normal
normal
beta
Analytic solution obtained
no case (b) (myopic only) approximate- (myopic only)
1 stock sizes above the threshold were used to calculate the mean.
yes
' Only recruitment data fron
-Approximate solution valid
only for pdfs with variance
<0.01.
Thompson: Bayesian approach to management advice from a simple dynamic pool model
571
for comparison with the results of Ludwig and Walters.
As Equations (26-28) indicate, a variety of parameter
combinations allow for B(Fmsy) to be less than half of
B(Fmelsy)- Since the results presented by Ludwig
and Walters (1982) were derived from a numbers-based
model, Equation (27) is particularly relevant. Under
this equation, a v' value greater than 1/3 is sufficient
to guarantee that the stock size at MSY will be less
than half the stock size at MELSY, regardless of the
value of m. At values of m>0.5, a v' value of 0.227 is
sufficient.
Clark et al. (1985) found that the relationship be-
tween the myopic Bayes and certainty-equivalent solu-
tions depended on the model used. In the special case
where recruitment is independent of stock size, for
example, they found that the myopic Bayes solution
always exceeded the certainty equivalent solution. For
the same model, the authors also found that the myopic
Bayes solution always increased with the level of uncer-
tainty. These results are precisely the opposite of those
obtained in the present study, where F^elsy is always
less than Fmsy arid decreases monotonically with v'.
In their "full cohort model" with a stock-recruitment
relationship, however, Clark et al. (1985) obtained
results similar to those of the present study. In one
example, the myopic Bayes solution prescribed a
30-50% reduction in F relative to the certainty-
equivalent solution. Using yet another model, Walters
and Ludwig (1987) also found that the myopic Bayes
solution was a monotone-decreasing function of
uncertainty.
Conclusion
This paper describes an approach for treating the prob-
lem of parameter uncertainty in a systematic fashion.
Although fisheries are often managed as though stock
parameters are known with certainty, it would be
preferable to develop a management approach more
consistent with the fact that such certainty is the ex-
ception rather than the rule. Such an approach was
developed here in the context of Bayesian decision
theory. When applied to the particular model pre-
sented, this approach indicates that the optimal fishing
mortality rate Fmelsy (Eq. 20) is always less than
Fmsy (Eq. 2) except in the limiting case where q is
known with certainty (Fig. 4).
This result provides formal support for the intuitive
conclusion (e.g., Kimura 1988) that fishing mortality
should be strongly constrained when the stock-recruit-
ment relationship is uncertain. Similarly, Equation (25)
indicates that if recruitment is highly dependent on
stock size (specifically, if m exceeds 0.5), Fmelsy will
always be less than the natural mortality rate.
The rock sole example illustrates the basic con-
servatism of the Fmelsy approach. In this example,
Fmelsy was less than Fmsy by about 40%. Given that
neither the Fmsy value (0.121) nor the fit from the
stock-recruitment regression (Fig. 6) was atypical of
groundfish stocks, the ratio between Fmelsy and Fmsy
in this example provides a practical illustration of the
extent to which an explicit accounting for uncertainty
can influence management strategy. The magnitude of
the effort reduction prescribed in this example is
similar to results described by Ludwig and Walters
(1982) and Clark et al. (1985). The confirmatory nature
of these studies may suggest that the conventional
wisdom regarding optimal exploitation rates should be
reexamined. At the very least, the Fmelsy approach
provides a low-end estimate of the maximum accept-
able harvest rate and a warning against taking Fmsy
estimates too seriously.
A great deal of the conservatism resulting from the
Fmelsy approach as developed here stems from the
assumption that all values of q are logically possible,
despite the fact that a q value of 1 results in extinction
under any level of fishing. One alternative might be to
examine q in the context of life-history theory, to deter-
mine if it is possible to justify some other upper limit
on the logically permissible range. A related alternative
would be to use a nonuniform prior in estimating P(q).
The assumption of a uniform prior may be overly
pessimistic, since fishery biologists often have an in-
tuitive feel for stock-recruitment parameters, even in
the absence of data for a particular stock. Such infor-
mation could be used to define an alternate prior pdf.
Another possibility would be to establish an empirical
prior based on the results of other stock-recruitment
studies, but this would likely require a fairly elaborate
weighting scheme so that stock-recruitment param-
eters from the most dissimilar stocks or environments
would have the least influence on the form of the
resulting pdf.
An additional factor which may add to the conser-
vatism of the Fmelsy strategy as developed here is
the use of the myopic Bayes solution rather than an
actively adaptive solution. An actively adaptive solu-
tion would attempt to anticipate and make use of
changes in available information resulting from alter-
native management actions (e.g., Walters and Hilborn
1976, Smith and Walters 1981, Ludwig and Walters
1982, Ludwig and Hilborn 1983, Clark et al. 1985,
Walters 1986, Milliman et al. 1987, Walters and Lud-
wig 1987, Parma 1990, Parma and Deriso 1990). How-
ever, myopic Bayes (or similar) solutions often perform
nearly as well as their actively adaptive counterparts
(Mendelssohn 1980, Walters and Ludwig 1987, Parma
1990, Parma and Deriso 1990), and if the myopic Bayes
solution is reestimated each year, the result is a
572
Fishery Bulletin 90(3). 1992
passively adaptive strategy which is asymptotically
optimal over time (Walters 1987). Most important for
the purposes of the present study, though, is the fact
that the myopic Bayes solution is computationally much
simpler than the actively adaptive solution.
In conclusion, it should be stressed that while the ap-
proach suggested here was developed in the context
of a particular model and particular loss and probability
density functions, this development was meant primar-
ily to illustrate the approach, not to limit it. More
sophisticated applications— utilizing alternative as-
sumptions, functional forms, and solution techniques-
are certainly to be encouraged. In particular, future
research might incorporate recruitment stochasticity,
positive discount and cost rates, additional objective
function components (e.g., yield variability), and uncer-
tainty in other parameters and variables (e.g., the
natural mortality rate, growth rate, and stock size).
Acknowledgments
I would like to thank James Balsiger, Nicholas Bax,
Russell Kappenman, Daniel Kimura, Richard Methot,
and Thomas Wilderbuer of the Alaska Fisheries Sci-
ence Center for reviewing all or portions of this paper
in various stages of development. Three anonymous
reviewers also provided helpful suggestions. In addi-
tion, I would like to thank Robert Burr and Loveday
Conquest of the University of Washmgton's Center for
Quantitative Science for their assistance.
Citations
Bernoulli, D.
1954 Exposition of a new theory on the measurement of risk.
Econometrica 22:23-36.
Box, G.E.P., and G.C. Tiao
1973 Bayesian inference in statistical analysis. Addison-Wes-
ley, Reading, MA, 588 p.
Charles, A.T.
1988 In-season fishery management: A Bayesian model. Nat.
Resour. Model. 2:599-629.
Clark, C.W.
1985 Bioeconomic modelling and fisheries management. John
Wiley, NY, 291 p.
Clark, C.W., A.T. Charles, J.R. Beddingrton, and M. Mangel
1985 Optimal capacity decisions in a developing fishery. Mar.
Resour. Econ. 2:25-53.
Cushing, D.H.
1971 The dependence of recruitment on parent stock in dif-
ferent groups of fishes. J. Cons. Cons. Int. Explor. Mer 33:
340-362.
DeGroot, M.H.
1970 Optimal statistical decisions. McGraw-Hill, NY, 489 p.
Deriso, R.B.
1985 Risk adverse harvesting strategies. Lect. Notes Bio-
math. 61:65-73.
Getz, W.M., and R.G. Haight
1989 Population harvesting: Demographic models of fish,
forest, and animal resources. Princeton Univ. Press, Prince-
ton, 391 p.
Gleit, A.
1978 Optimal harvesting in continuous time with stochastic
growth. Math. Biosci. 41:111-123.
Hightower, J.E.
1990 Multispecies harvesting policies for Washington-
Oregon-California rockfish trawl fisheries. Fish. Bull., U.S.
88:645-656.
Hightower, J.E., and G.D. Grossman
1987 Optimal policies for rehabilitation of overexploited fish
stocks using a deterministic model. Can. J. Fish. Aquat. Sci.
44:803-810.
Hightower, J.E., and W.H. Lenarz
1989 Optimal harvesting policies for the widow rockfish fish-
ery. In Edwards, E.F., and B.A. Megrey (eds.), Catch-at-age,
bioenergetics, system, and sampling models for microcomputer
analyses for fishery dynamics, p. 83-91. Am. Fish. Soc. Symp.
6. Bethesda.
Holloway, C.A.
1979 Decision making under uncertainty: Models and choices.
Prentice-Hall, Englewood Cliffs, NJ, 522 p.
Hulme, H.R., R.J.H. Beverton, and S.J. Holt
1947 Population studies in fisheries biology. Nature (Lond.)
159:714-715.
Kimura, D.K.
1988 Stock-recruitment curves as used in the stock-reduction
analysis model. J. Cons. Cons. Int. Explor. Mer 44:253-258.
Larkin, P.A.
1973 Some observations on models of stock and recruitment
relationships for fishes. Rapp. P.-V. Reun. Cons. Int. Explor.
Mer 164:316-324.
Larkin, P.A., and W.E. Ricker
1964 Further information on sustained yields from fluctuating
environments. J. Fish. Res. Board Can. 21:1-7.
Lewis, T.R.
1981 Exploitation of a renewable resource under uncertainty.
Can. J. Econ. 14:422-439.
1982 Stochastic modeling of ocean fisheries resource manage-
ment. Univ. Wash. Press, Seattle, 109 p.
Lord, G.E.
1973 Characterization of the optimum data acquisition and
management of a salmon fishery as a stochastic dynamic pro-
gram. Fish. Bull., U.S. 71:1029-1037.
1976 Decision theory applied to the simulated data acquisition
and management of a salmon fishery. Fish. Bull., U.S. 74:
837-846.
Ludwig. D., and R. Hilborn
1983 Adaptive probing strategies for age-structured fish
stocks. Can. J. Fish. Aquat. Sci. 40:559-569.
Ludwig, D., and C.J. Walters
1981 Measurement errors and uncertainty in parameter esti-
mates for stock and recruitment. Can. J. Fish. Aquat. Sci.
38:711-720.
1982 Optimal harvesting with imprecise parameter estimates.
Ecol. Model. 14:273-292.
Mangel, M., and C.W. Clark
1983 Uncertainty, search, and information in fisheries. J.
Cons. Cons. Int. Explor. Mer 41:93-103.
Mangel, M., and R.E. Plant
1985 Regulatory mechanisms and information processing in
uncertain fisheries. Mar. Resour. Econ. 1:389-418.
Thompson: Bayesian approach to management advice from a simple dynamic pool model
573
Mendelssohn, R.
1980 Using Markov decision models and related techniques for
purposes other than simple optimization: Analyzing the con-
sequences of policy alternatives on the management of salmon
runs. Fish. Bull., U.S. 78:35-50.
1982 Discount factors and risk aversion in managing random
fish populations. Can. J. Fish. Aquat. Sci. 39:1252-1257.
Milliman, S.R., A.P. Grima, and C.J. Walters
1987 Policy making within an adaptive management frame-
work, with an application to lake trout {Salvelinus namaycush)
management. Can. J. Fish. Aquat. Sci. 44(Suppl. 2):425-430.
Opaluch, J.J., and N.E. Bockstael
1984 Behavioral modeling and fisheries management. Mar.
Resour. Econ. 1:105-115.
Parma, A.M.
1990 Optimal harvesting of fish populations with non-station-
ary stock-recruitment relationships. Nat. Resour. Model.
4:39-76.
Parma, A.M., and R.B. Deriso
1990 Experimental harvesting of cyclic stocks in the face of
alternative recruitment hypotheses. Can. J. Fish. Aquat. Sci.
47:595-610.
Paulik, G.J.
1973 Studies of the possible form of the stock-recruitment
curve. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 164:316-324.
Pratt, J.W.
1964 Risk aversion in the small and in the large. Econo-
metrica 32:122-136.
Press, S.J.
1989 Bayesian statistics: Principles, models, and applications.
John Wiley, NY, 237 p.
Raiffa, H.
1968 Decision analysis: Introductory lectures on choices under
uncertainty. Addison- Wesley, Reading, MA, 309 p.
Ricker, W.E.
1954 Stock and recruitment. J. Fish. Res. Board Can. 11:
559-623.
1958 Maximum sustained yields from fluctuating environments
and mi.\ed stocks. J. Fish. Res. Board Can. 15:991-1006.
1975 Computation and interpretation of biological statistics
for fish populations. Fish. Res. Board Can. Bull. 191, 382 p.
Rothschild, B.J.
1972 An exposition on the definition of fishing effort. Fish.
Bull, U.S. 70:671-679.
Rothschild, B.J., and A.J. Mullen
1985 The information content of stock-and-recruitment data
and its non-parametric classification. J. Cons. Cons. Int.
Explor. Mer 42:116-124.
Ruppert, D., R.L. Reish, R.B. Deriso, and R.J. Carroll
1984 Optimization using stochastic approximation and Monte
Carlo simulation (with application to harvesting of Atlantic
menhaden). Biometrics 40:535-545.
1985 A stochastic population model for managing the Atlan-
tic menhaden (Brevoortia tyranmis) fishery and assessing
managerial risks. Can. J. Fish. Aquat. Sci. 42:1371-1379.
Samuelson, P.A.
1967 General proof that diversification pays. J. Financial &
Quant. Analysis 2:1-13.
Schnute, J.
1981 A versatile growth model with statistically stable param-
eters. Can. J. Fish. Aquat. Sci. 38:1128-1140.
Shepherd, J.G.
1982 A versatile new stock-recruitment relationship for fish-
eries, and the construction of sustainable yield curves. J. Cons.
Cons. Int. E.xplor. Mer 40:67-75.
Shepherd, J.G., and D.H. Gushing
1990 Regulation in fish populations: Myth or mirage? Philos.
Trans. R. Soc. Lond. B 330:151-164.
Smith, A.D.M.. and C.J. Walters
1981 Adaptive management of stock-recruitment systems.
Can. J. Fish. Aquat. Sci. 38:690-703.
Smith, J.B.
1985 A discrete model of replenishable resource management
under uncertainty. Mar. Resour. Econ. 1:283-308.
Tautz, A., P.A. Larkin, and W.E. Ricker
1969 Some effects of simulated long-term environmental fluc-
tuations on maximum sustained yield. J. Fish. Res. Board
Can. 26:2715-2726.
Thompson, G.G.
1992 Management advice from a simple dynamic pool model
incorporating constant marginal cost and discount rates. Fish.
Bull., U.S. 90:552-560.
Walters, C.J.
1975 Optimal harvest strategies for salmon in relation to
environmental variability and uncertainty in production
parameters. J. Fish. Res. Board Can. 32:1777-1784.
1986 Adaptive management of renewable resources. MacMil-
lan, NY, 374 p.
1987 Approaches to adaptive policy design for harvest manage-
ment. Lect. Notes Biomath. 72:114-122.
Walters. C.J., and R. Hilborn
1976 Adaptive control of fishing systems. J. Fish. Res. Board
Can. 33:145-159.
Walters, C.J., and D. Ludwig
1981 Effects of measurement errors on the assessment of
stock-recruitment relationships. Can. J. Fish. Aquat. Sci.
38:704-710.
1987 Adaptive management of harvest rates in the presence
of a risk averse utility function. Nat. Resour. Model.
1:321-337.
Walters, G.E., T.K. Wilderbuer
1988 Rock sole. In Condition of groundfish resources of the
eastern Bering Sea and Aleutian Islands region in 1988, p.
106-117. Unpubl. manuscr. submitted to Int. N. Pac. Fish.
Comm., Sept. 1988. NMFS Alaska Fish. Sci. Cent., Seattle,
219 p.
Abstract.— The only cosmopoli-
tan sciaenid genus, Umbrina. is rep-
resented in the eastern Pacific Ocean
by eight species: U. analis, U. bus-
singi, U. dorsalis, U. galapagorum,
U. reedi, U. roncador, U. winter-
steeni n. sp., and U. xayiti. Umbrina
analis is removed from the synon-
ymy of U. xayiti. Lectotjrpes are
designated for U. dorsalis, U. gala-
pagorum and U. xanti. Umbrina
wintersteeni, which usually occurs in
shallow, protected waters of the
southern Gulf of California and the
west coast of southern Baja Califor-
nia Sur, apparently is morphological-
ly intermediate between two major
groups of eastern Pacific species.
Distinguishing characters of U. win-
tersteeni include peritoneum and in-
side gill cover with little or no pig-
ment; barbel relatively short and
stout; anal fin with six soft rays; anal
fin darkly pigmented to dusky; pelvic
fins usually dusky; second anal spine
of moderate length.
Eastern Pacific species of the genus
Umbrina (Pisces: Sciaenldae) vj'Mh
a description of a new species
H.J. Walker Jr.
Scripps Institution of Oceanography
University of California at San Diego, La Jolla, California 92093-0208
Keith W. Radford
Department of Biology, Mesa College
7250 Mesa College Drive, San Diego. California 921 1 I
Manuscript accepted 6 August 1992.
Fishery Bulletin, U.S. 90:574-587 (1992).
Of the more than 70 genera in the
percoid family Sciaenidae, only Um-
brina has a worldwide distribution
(Chao 1986a). The approximately 15
species that constitute Umbrina oc-
cur in tropical to temperate waters
over the continental shelf to the
upper slope. In the New World, Um-
brina comprises four species in the
Atlantic (Gilbert 1966, Miller 1971)
and eight in the Pacific (this study).
Most eastern Pacific species are col-
lected with beach seines over sand or
sand-mud bottoms, along open coasts
or in bays, and probably support ar-
tisanal or sportfisheries wherever
they are found. In southern Califor-
nia the yellowfin croaker U. roncador
and spotfin croaker Roncador steam-
sii) together make up about 10 per-
cent of the surf fisherman's catch
(Frey 1971).
No review of the eastern Pacific
species of Umbrina has been pub-
lished, although McPhail (1958)
wrote extensive keys to all known
eastern Pacific sciaenids, and Lopez
S. (1980) described a new species of
Umbrina from this area. The pur-
poses of this paper are to review the
eastern Pacific species of Umbrina,
provide a key and characters useful
in their identification, and describe a
new species.
Materials and methods
Counts and measurements generally
follow those of Hubbs and Lagler
(1958). Gill raker counts include rudi-
ments. Unless otherwise stated, stan-
dard length (SL) is used throughout.
Vertebral and procurrent caudal ray
counts were made from radiographs.
A short, stout barbel is defined as one
whose length roughly equals its width
at midlength (seen in side view); an
elongate barbel is at least twice as
long as wide. Mean percentages of
certain morphometries used in spe-
cies diagnoses were calculated usual-
ly from 30 specimens, occasionally
~20 (when available), selected from
the entire size range of the species.
Standard errors associated with
these means were calculated strictly
to show relative variation for a par-
ticular proportion and were always
0.6% (once) or less. All pigmentation
notes were made from alcohol-
preserved specimens. Institutional
abbreviations follow Leviton et al.
(1985). There have been many in-
stances where eastern Pacific species
have been ascribed to Umbrina (e.g.,
U. panamensis = Menticirrhus pana-
mensis; U. imberbis = Sciaena, prob-
ably callaensis), and these are beyond
the scope of this paper. Type material
574
Walker and Radford: Eastern Pacific species of the genus Umbrina
575
of all synonyms listed in the species accounts was ex-
amined by the authors and/or C.L. Hubbs (deceased).
Systematics
Genus Umbrina Cuvier
Synonymy
Umbrina Cuvier 1816:297 (type species Sciaena cir-
rosa Linnaeus, by monotypy, see Opinion 988, Bull.
Zool. Nom. (1972):123).
Attilus Gistel 1848:109 (type species Sciaena cirrosa
Linnaeus, by monotypy).
Asperina Ostroumoff 1896:30 (type species A. impro-
viso ( = U. cirrosa) Ostroumoff, by monotypy).
Diagnosis Deep-bodied to moderately elongate, com-
pressed sciaenid fishes with a single mental barbel,
usually with an apical pore; swim bladder single-
chambered, usually carrot-shaped, with no diverticula,
located entirely abdominally; preopercular margin with
bony serrations; two anal spines, the second long and
thick.
Description As in Gilbert (1966), Trewavas (1977),
and Chao (1978, 1986a, b), with some additions: back
slightly arched; ventral profile nearly straight; head
oblong; snout thick and protuberant with 5-7 rostral
and 5 marginal pores; chin with two pairs of lateral
pores surrounding the short barbel; mouth small, in-
ferior, horizontal or nearly so; teeth small, villiform,
set in bands in both jaws, outer row of teeth in upper
jaw may be slightly enlarged; sagitta (largest otolith)
thick, oval, with smooth inner surface and crested or
nodular outer surface; Cauda of sulcus bent sharply and
not reaching ventral edge of sagitta, ostium reaching
anterior edge; gill rakers short; caudal fin truncate to
slightly emarginate or pointed; scales ctenoid; verte-
brae 10-11 + 14-16 = 25-26; dorsal fin rays IX-X +
1,21-33; anal rays 11,5-10; pectoral rays 14-20; over-
all background coloration white to silver or yellow to
brown; usually with dark-brown stripes: oblique dor-
solaterally, more longitudinal midlaterally and on
peduncle area, becoming faint or absent ventrally,
usually faint or absent on head.
Reiationships The genera Sciaena and Umbrina are
the only representatives of the tribe Sciaenini (Chao
1986a). Characters of the swim bladder are the most
important factors in assessing the phylogenetic rela-
tionships among suprageneric groups of sciaenids and
the single-chambered swim bladder, lacking append-
ages, characteristic of the sciaenines, is the most
primitive (plesiomorphic) form (Chao 1986a). The genus
Sciaena (species have no barbels) is a polyphyletic
assemblage containing numerous species and is in need
of revision (Chao and Miller 1975, Chao 1986a). Al-
though apparently monophyletic, we presently can only
define Umbrina with synplesiomorphies or homoplas-
tic apomorphies (e.g., pored mental barbel) (L.N. Chao,
Bio-Amazonica Conserv. Int., Brazil, pers. commun.,
Sept. 1991).
Key to the eastern Pacific species of Umbrina
lA Inside gill cover dark to black, particularly in
area of pseudobranch 2
IB Inside gill cover pale or lightly punctate 4
2A Dorsal fin with 21-23 soft rays; no stripes on
body U. bussingi
2B Dorsal fin usually with 26-30 soft rays; dark
to dusky horizontal or oblique stripes on
body 3
3A Anal fin normally with 7 soft rays; peritoneum
dark U. roncador
SB Anal fin normally with 6 soft rays; peritoneum
light ventrally (may be dark dorsally)
U. xanti
4A Anal fin with 9(8) soft rays; dorsal fin with
IX 4- 1 spines U. reedi
4B Anal fin with 6-7(8) soft rays; dorsal fin with
X -I- 1 spines 5
5A Dorsal fin usually with 30-33 soft rays; snout
length less than eye diameter U. dorsalis
5B Dorsal fin usually with 24-29 soft rays; snout
length greater than eye diameter (adults) .... 6
6A Body stripes distinct; pectoral fin rays 17 or
fewer; dorsal fin soft rays 27 or fewer 7
6B Body stripes indistinct or lacking; pectoral fin
rays usually 18 or more; dorsal fin rays 27
or more U. galapagorum
7A Second anal spine ~1.5 in head; pelvic fins
with little or no pigment U. analis
7B Second anal spine ~2.0 in head; pelvic fins
usually dusky to dark U. wintersteeni
Umbrina bussingi Lopez S.
Figure 1
Synonymy
Umbrina bussingi Lopez S. 1980:203-208 (original
description: holotype LACM 38715-1; Costa Rica).
Diagnosis A small species of Umbrina (max. length
252 mm) characterized by the following combination of
characters: inside gill cover dark to black; no dark-
brown stripes; caudal fin pointed; barbel compressed,
576
Fishery Bulletin 90(3). 1992
.<s=^
^
S^
1
fi
1
i
? .
?fe
T^^^^
^^^^
^^^^v
s.^
M'
^N^
f ''ij- '
■^
^1
b^.
„,■"
^^.,„^
>J
D^^
ki
Jk
W'--
—^
-i^.
::"!!
iliii^
y
9i
V
»C-
Figure 1
Umbrina bussingi, 230nimSL, LACM 9099-28.
with anterior, slit-like (vertical) pore (large adults);
peritoneum dark to dusky ventrally and laterally,
lighter dorsally; soft dorsal fin rays 21-23; soft anal
rays 7; pectoral rays usually 18-19; procurrent caudal
rays 7-8 + 7-8(9); dorsal spines X + I; gill rakers usu-
ally 19-20; vertebrae 10 -i- 15; length of second anal
spine, X 17% SL; body depth, x32%SL; eye length,
i9%SL; upper-jaw length, xl3%SL; pectoral fin
length, 5 26%SL.
Description Counts and measurements are given in
Tables 1-6. Soft dorsal and anal fin rays as in Diag-
nosis. Pectoral rays 17-19; gill rakers 17-22; barbel not
fully developed (usually bulbous) on specimens >154
mm; pigment inside operculum appears externally as
large, dark spot; no dark-brown stripes at any size;
background color uniformly light- (in young, to ~155
mm) to medium-brown; second anal spine relatively
long and thick; body fairly deep; eye large, relatively
smaller in larger specimens, but length always greater
than snout length; head and upper jaw relatively long;
pectoral fins extremely long, proportionately shorter
at larger sizes; lateral line scales 47-49, x 47.95, SE
0.18; spinous dorsal fin dark to dusky, lighter in smaller
specimens; soft dorsal, pelvic, and caudal fins light to
dusky, becoming darker with increasing size, most pig-
ment on pelvic and caudal fins appearing on distal two-
thirds; pectoral fins essentially unpigmented; anterior,
proximal portion of anal fin darkly pigmented at most
sizes, dark pigment on most of fin at larger sizes.
Distribution Southern Gulf of California, south of
Los Frailes to Golfo de Chiriqui, Panama (Fig. 2). A
relatively deep-living species, taken in depths of 32 m
to >183m (Lopez S. 1980).
Umbrina roncador Jordan and Gilbert
Figure 3
Synonymy
Umbrina roncador Jordan and Gilbert 1882:277-278
(original description: holotype USNM 29371, Bahia
Pequena, Baja California Sur).
Sciaena thompsoni Hubbs 1921:1-3 -i- pi. (original de-
scription: holotype UMMZ 55053, Santa Catalina I.,
CA).
Diagnosis An intermediate-sized species of Umbrina
(reported to ~381mm) characterized by the following:
inside gill cover dark to black, particularly in area of
pseudobranch; peritoneum usually dark; soft dorsal fin
rays usually 26-29; soft anal rays usually 7; pectoral
rays usually 17-18; procurrent caudal rays usually 9-10
-I- 8-9; dorsal spines X -i- 1; gill rakers usually 18-20;
vertebrae 10 -i- 15; barbel relatively elongate, slender,
more robust at sizes greater than ~200mm; length of
second anal spine, x 12% SL; body depth, x 29%; eye
length, x 6%; upper jaw length, x 11%; pectoral fin
length, X 17%.
Description Counts and measurements are given in
Tables 1-6. Soft rays of second dorsal fin 24-31; soft
anal rays 6-7; pectoral rays 15-20; gill rakers 15-22;
snout length greater than eye diameter (adults); pig-
ment inside operculum usually appears externally as
large, dark spot, dark pigment around pseudobranch
Walker and Radford: Eastern Pacific species of the genus Umbnna
577
Figure 2
Geographic distribution of Umbrina bussingi (O), U. dorsalis (•), U. analis (•), and U.
wintersteeni n. sp. (A).
usually present by 50-65 mm;
barbel fully developed by ~70
mm, indistinguishable or flat
at sizes <30mm, bulbous be-
tween ~30 and 70 mm; lateral
line scales 49-54, x 50.85, SE
0.04; virtually all brown
stripes present by 70 mm; 5-8
dark, vertical bars occasional-
ly on specimens 25-180 mm,
usually faint when >80mm;
dorsal fin dusky at most sizes;
pectoral, pelvic, anal, and
caudal fins usually with little
or no pigment, caudal occa-
sionally dusky.
Distribution Point Concep-
tion to Bahia Magdalena and
disjunctly in the northern Gulf
of California, north of 27°N
(Fig. 4). Taken in bays and the
surf zone to ~45m (Miller and
Lea 1972).
Umbrina xanti Gill
Figure 5
Synonymy
Umbrina xanti Gill 1862:257-
258 (original description:
syntypes USNM 7156,
USNM 2996, USNM 3693,
USNM 3694, Cabo San
Lucas, Baja California Sur;
lectotype USNM 7156 (79
Figure 3
Umbrina roncador, 217mmSL, SIO 64-65.
578
Fishery Bulletin 90|3). 1992
Figure 4
Geographic distribution of Umbrina roncador (O), U. xanti ( • ) plus one specimen from Chile
at 19°36'S, 70°13'W), and U. galapagorum (A).
mm), herein designated; pa-
ralectotypes USNM 316653,
[removed from USNM 7156]
USNM 2996, USNM 3693,
USNM 3694, MCZ 35976
[removed from USNM
3693]).
Umhrina sinaloae Scofield
1896:220-221 (original de-
scription: syntypes CAS-
SU 1632, Mazatlan, Sinaloa,
Mexico).
Diagnosis A small species of
Umbrina (max. length 295
mm) characterized by the
following: gill cover dark to
black, particularly in area of
pseudobranch; peritoneum
usually light ventrally, punc-
tate or occasionally dark dor-
sally; soft dorsal fin rays
usually 27-29; soft anal rays
usually 6; pectoral rays usual-
ly 17-18; procurrent caudal
rays usually 8-9 -t- 8-9; dorsal
spines X -i- 1; gill rakers usual-
ly 17-19; vertebrae 10 + 15;
barbel relatively elongate,
slender; length of second anal
spine, X 12% SL; body depth
X 29%; eye length, x 7%; up-
per jaw length, x 10%; pec-
toral fin length, x 16%.
Figure 5
Vmbrina xanti, 188mmSL, SIO 63-522.
Walker and Radford: Eastern Pacific species of the genus Umbrina
579
Description Counts and measurements are given in
Tables 1-6. Soft rays of second dorsal fin 26-30; soft
anal rays 6-7; pectoral rays 14-19; gill rakers 16-21;
snout length greater than eye diameter (adults); pig-
ment inside operculum generally appears externally as
large, dark spot; dark pigment around pseudobranch
evident by 35-40 mm; barbel fully developed by 45 mm,
usually knobby between 30 and 45 mm; lateral line
scales 48-52, x 50.02, SE 0.05; nearly all brown
stripes present by 75-80 mm; dark, vertical bars 6-9,
occasionally and only on specimens < 100 mm; dorsal
fin dusky; pectoral, pelvic, anal and caudal fins usual-
ly with little or no pigment, caudal occasionally dusky.
Distribution Bahia Magdalena and the southern Gulf
of California (south of 27°N) to northern Peru (incl. Isla
del Coco, Fig. 4), one specimen from Pisagua, Chile,
19°36'S, 70°13'W. Found in tide pools (juveniles) and
along beaches, to 36m.
Umbrina reedi Gun t her
Figure 6
Synonymy
Umbrina reedi Giinther 1880:25 (original description:
holotype BMNH 1919.5.14.283; I. Juan Fernandez,
Chile).
Diagnosis A large species of Umbrina (max. length
650 mm) characterized by the following: inside gill
cover pale to lightly punctate; peritoneum dark ven-
trally and laterally, lighter dorsally; large, dark spot
in axil of pectoral fin; soft dorsal fin rays usually 24-25;
soft anal rays usually 9; pectoral rays usually 18; pro-
current caudal rays usually 9-10 -t- 8-9; dorsal spines
IX -I- 1; gill rakers usually 19-20; vertebrae 10 -t- 15;
barbel relatively short, stout; length of second anal
spine, X 14% SL; body depth, x 35%; eye length, x 7%;
upper jaw length, i 13%; pectoral fin length, x 23%.
Description Counts and measurements are given in
Tables 1-6. Soft rays of second dorsal fin 21-26; soft
anal rays 8-10; pectoral rays 17-19; gill rakers 18-22;
body and caudal peduncle deep; snout length greater
than eye diameter at sizes greater than ~150mm; pec-
toral fins relatively long; upper jaw long; barbel bulbous
at 28mm, not fully developed until ~120mm; lateral
line scales 49-52, x 49.90, SE 0.17; caudal peduncle
and midlateral brown stripes evident by 120 mm; dor-
solateral stripes undifferentiated on most specimens,
appearing as irregular dashes of pigment; dorsal and
caudal fins dusky; pectoral, pelvic and anal fins dark
(pectorals frequently dusky), lighter on specimens less
than ~ 130 mm.
Distribution Islands off Chile: Islas Juan Fernandez
(~33°40'S, 78°55'W) and San Felix (~26°17'S, 80°
06' W). From the surf zone to 30 m.
•'^
'SKBBSSi-!-
\,
J
*
<^
'V
Figure 6
Umbrina reedi, 268mmSL. SIO 65-657.
580
Fishery Bulletin 90(3). 1992
Figure 7
Umbrina dorsalis. 236mmSL, SID 62-705.
J^^"
Figure 8
Umbrina galapagorum. 273minSL, SIO 79-51.
Umbrina dorsalis Gill
Figure 7
Synonymy
Umbrina dorsalis Gill 1862:257 (original description:
syntypes USNM 3696, Cabo San Lucas, Baja Califor-
nia Sur; lectotype USNM 3696 (77 mm), herein
designated; paralectotypes USNM 316654 [removed
from USNM 3696]).
Diagnosis An intermediate-sized species of Umbrina
(max. length 332 mm) characterized by the following:
inside gill cover and peritoneum pale to lightly punc-
tate; soft dorsal fin rays usually 30-32; soft anal rays
usually 7; pectoral rays usually 16-17; procurrent
caudal rays usually 7-8 + 6-7; dorsal spines X + l; gill
rakers usually 19-21; vertebrae 10 + 16; barbel rela-
tively long and thick; length of second anal spine,
X 15% SL; body depth, i 38%; eye length, x 8%; upper
jaw length, x 13%; pectoral fin length, x 21%.
Description Counts and measurements are given in
Tables 1-6. Soft rays of second dorsal fin 28-33; soft
anal rays 7-8; pectoral rays 15-17; gill rakers 18-25;
eye relatively large, proportionately smaller in larger
specimens, generally greater than snout length; upper
jaw long; body and caudal peduncle deep; barbel fully
developed by 75 mm, bulbous between 30 and 65 mm;
lateral line scales 49-54, x 51.73, SE 0.15; virtually
all brown stripes present by 75mm, no stripes on speci-
Walker and Radford: Eastern Pacific species of the genus Umbnna
581
mens <65mm; 5-8 wide, dark bars (saddles) on speci-
mens 30-65 mm; dorsal and pelvic (posterior) fins dusky
to dark; pectoral and caudal fins light to dusky (pos-
terior pectoral fin dark on specimens < 45 mm); anal fin
dark to dusky.
Distribution South of Bahia Magdalena and in the
southern Gulf of California to Equador (Fig. 2). Found
in tidepools (juveniles) to ~5m.
Umbrina galapagorum Steindachner
Figure 8
Synonymy
Umbrina galapagorum Steindachner 1878:20-21 (orig-
inal description: syntypes MCZ 8601, USNM
120437, James I., Galapagos Is.; lectotype MCZ 8601
(94 mm), herein designated; paralectotype USNM
120437).
Diagnosis An intermediate-sized species of Umbrina
(max. length 413 mm) characterized by the following:
inside gill cover lightly punctate or dusky, usually little
or no pigment in area of pseudobranch; brown stripes
indistinct or absent; peritoneum pale or lightly punc-
tate; soft dorsal fin rays usually 27-29; soft anal rays
usually 6-7; pectoral rays usually 17-19; procurrent
caudal rays usually 9-10 -i- 8-9; dorsal spines X -i- 1; gill
rakers usually 18-19; vertebrae 10 -i- 15; barbel rela-
tively elongate, thin; length of second anal spine, x
12% SL; body depth, i 29%; eye length, x 6%; upper
jaw length, x 10%; pectoral fin length, x 18%.
Description Counts and measurements are given in
Tables 1-6. Soft rays of second dorsal fin 26-30; soft
anal rays 5-7; pectoral rays 16-20; gill rakers 16-21;
snout length greater than eye length; pigment inside
operculum occasionally appearing as muted, dark spot;
barbel fully developed by 60 mm, bulbous between 30
and ~50mm; lateral line scales 47-52, 5:49.73, SE
0.08; faint brown stripes visible at 40-50 mm, occa-
sionally evident to ~100mm; ~9-10 dark, vertical bars
frequently on specimens between ~30 and 95 mm; dor-
sal and caudal fins usually dusky, caudal occasionally
lighter; pectoral, pelvic and anal fins unpigmented to
lightly punctate.
Distribution Endemic to the Galapagos Is. (Fig. 4),
found near beaches to 18 m.
Umbrina analis Giinther
Figure 9
Synonymy
Umbrina analis Gunther 1869(1866):426-427 (original
description: holotype BMNH 1867.9.23.18, Panama).
Umbrina tumacoensis Wilson 1916:67 (original descrip-
tion: holotype, presumed lost (pers. commun.: M.E.
Anderson, CAS, 2 March 1988; B. Chernoff, FMNH,
22 March 1988), paratypes CAS 62852, FMNH
56840).
Diagnosis A small species of Umbrina (max. length
231 mm) characterized by the following (based on nine
specimens): peritoneum and inside gill cover pale to
lightly punctate; soft dorsal fin rays 24-26; soft anal
rays 6; pectoral rays usually 17; procurrent caudal rays
8-9 -I- 7-8; dorsal spines X -i- 1; gill rakers usually 17-18;
-"-—■ iiiiir-niiMin^
Figure 9
Umbrina analis, 229mmSL, LACM 33822-32.
582
Fishery Bulletin 90(3). 1992
vertebrae 10 + 15; barbel relatively short, stout; length
of second anal spine, x 19% SL; body depth, i32%;
eye length, x 8%; upper jaw length, x 11%; pectoral
fin length, x2Q%.
Description Counts and measurements are given in
Tables 1-6. Soft dorsal and anal fin rays as in Diag-
nosis. Pectoral rays 16-18; gill rakers 16-21; barbel
bulbous at 51mm, more or less fully developed at 82
mm; no brown stripes on a 51 mm specimen, all stripes
evident by 82 mm; second anal spine extremely long
and thick; caudal peduncle deep; pectoral fins fairly
long; lateral line scales 46-50, x 47.88, SE 0.44; dor-
sal and anal fins dusky or dark; caudal fin usually
dusky; pectoral and pelvic fins with little or no pigment.
Distribution Known from five collections (one was
split), ranging from Costa Rica to Colombia (Fig. 2).
No depths recorded.
Umbrina wintersteeni n. sp.
Figure 10
HoJotype SIO 60-366, 193mm; Bahia Almejas, Baja
California Sur, F.H. Berry and party, 25 August 1960.
Paratypes 146 specimens (49-298 mm) from 21 col-
lections, all from Mexico: Baja California-west coast:
SIO 60-366 (29); SIO 62-126 (10); SIO 64-84 (49); CAS
35536 (4); USNM 316655 (5); AMNH 5514a (1). Gulf
of California: UCLA W50-27 (2); UCLA W52-48 (5);
UCLA W52-49 (2); UCLA W52-50 (1); UCLA W53-84
(6); UCLA W53-95 (8); UCLA W57-34 (1); UCLA
W57-36 (4); UCLA W57-42 (4); SIO 65-281 (3); SIO
76-275 (2); LACM 38104-26 (4); CAS-SU 375 (1);
CAS-SU 2855 (3); CAS-SU 47933 (1); AMNH 5498 (1).
Diagnosis A relatively small species of Umbrina
(max. size 298mm) characterized by the following:
inside gill cover pale to lightly punctate, little or no
pigment in area of pseudobranch; peritoneum pale or
lightly punctate; soft dorsal fin rays usually 25-27; soft
anal rays usually 6; pectoral rays usually 16-18; pro-
current caudal rays usually 7-8 + 6-7; dorsal spines
X -I- 1; gill rakers usually 17-20; vertebrae 10 + 15;
barbel relatively short, stout; length of second anal
spine, x 14% SL; body depth, x 29%; eye length, x 7%;
upper jaw length, i 10%; pectoral fin length, x 17%.
Description Counts and measurements are given in
Tables 1-6. Soft rays of second dorsal fin 23-28; soft
anal rays 5-6; pectoral rays 15-18; gill rakers 14-23;
snout length greater than eye length (adults); barbel
bulbous at 49mm, fully developed by ~100mm; lateral
line scales 48-51, x 49.14, SE 0.06; faint, brown
stripes at 49 mm, nearly all present by 80-90 mm;
~8-ll dark, vertical bars on many juveniles, never on
specimens > 100 mm; dorsal, pelvic, and anal fins usual-
ly dusky to dark (Gulf of California specimens occa-
sionally with light-to-dusky pelvic and anal fins); caudal
fin dusky; pectoral fins lightly pigmented; background
coloration tan to gold dorsally, grading to white or
silver ventrally; dark-brown stripes: 15-20 oblique
dorsolaterally, 8-10 more or less horizontal laterally
in abdominal area, 4-5 horizontal on lateral caudal
peduncle.
Figure 10
Umbrina wintersteeni, holotype, 193mmSL, SIO 60-366.
Walker and Radford Eastern Pacific species of the genus Umbnna
583
Etymology Named for the late John Wintersteen,
longtime researcher in the taxonomy of eastern Pacific
sciaenids.
Distribution Just north of Bahia Magdalena (25°23'
N, 112°06'W), and the southern Gulf of California from
~27°N to Mazatlan (Fig. 1). Usually collected in bays;
thus far, only recorded in depths to ~2m.
Relationships of New World
Umbrina species
We did not attempt a phylogenetic analysis because we
have not seen all Umbrina. species. In addition, the
status and limits of the only recognized sister genus,
Sciaena, are uncertain. However, morphological sim-
ilarity allows certain groups of the New World species
to be distinguished. Umbrina bussingi and U. milliae.
(Atlantic) share several characters: compressed barbel,
with anterior, slit-like opening; extremely long pectoral
fins; very large eyes; caudal rays longest in middle of
fin. They also lack stripes and differ slightly in relative
body depth and length of second anal spine; we
hypothesize that these are geminate species.
Umbrina analis, U. broussonnettii (Atlantic), U.
canosai (Atlantic), U. coroides (Atlantic), U. dorsalis,
and U. reedi share the following characters: relatively
deep body; relatively long pectoral fins; dusky to dark
anal fin. Umbrina analis, U. canosai, and U. reedi
have fairly short, stout barbels and 6, 8, and 9 anal rays,
respectively. Umbrina analis also has an extremely
long, second anal spine. The other three species in this
group have relatively slender, elongate barbels and 6
or 7 (U. dorsalis) anal rays. Umbrina dorsalis also has
a relatively large eye and high number of dorsal rays;
U. broussonnettii and U. coroides have low gill raker
counts and differ slightly in certain scale and fin covmts
and pigmentation.
Umbrina galapagorum, U. roncador, U. winter-
steeni, and U. xanti share the following characters:
relatively elongate body; relatively short pectoral fins;
little or no pigment on anal fin (exception: U. winter-
steeni). Except for U. winter steeni, these species have
a somewhat elongate, slender barbel. The inside gill
cover of U. roncador and U. xanti is dark to black and
the anal ray counts are 7 and 6, respectively. Umbrina
galapagorum, which usually lacks stripes (can be faint),
and U. wintersteeni, which has a short, stout barbel,
have 6 anal rays. Although we have no information for
some species, juvenile characters (e.g., extreme dif-
ferences in body depth, pectoral fin pigmentation, dor-
solateral pigmentation; Fig. 11) corroborate the latter
two major groups (six and four species), with U.
wintersteeni possibly an intermediate form.
Distribution
As with the species of Porichthys, which also are
associated with soft bottom (Walker and Rosenblatt
1988), the distributional limits of eastern Pacific species
of Umbrina generally coincide with zoogeographic
boundaries established for rocky shore fishes and other
fauna (Springer 1958, Rosenblatt 1967, Briggs 1974,
and others). In our area of concern, these boundaries
are Point Conception, Bahia Magdalena area. La Paz
for the western Gulf of California, and between
Guaymas and Mazatlan for the eastern Gulf, Golfo de
Tehuantepec area, and northern Peru. Umbrina ron-
cador occurs in the northern Gulf of California (north
of 27 °N) and from Point Conception to the Bahia
Magdalena area, which is also the northernmost Baja
California (west coast) limit for U. xanti, U. winter-
steeni, and U. dorsalis. Both U. xanti and U. dorsalis
are wide-ranging, also occurring from the southern
Gulf of California to northern Peru and Equador,
respectively. Umbrina wintersteeni also is found in the
southern Gulf as far south as Mazatlan. Umbrina bus-
singi and U. analis are each known from five collec-
tions. Umbrina bussingi occurs nearly throughout the
eastern tropical Pacific (southern Gulf of California to
Panama), while U. analis apparently is confined to the
south (Costa Rica to southern Colombia).
Additional materials examined
Umbrina bussingi 140 specimens (56-257mm) from 4 col-
lections. Mexico: SIO 62-51 (130); SIO 70-160 (1); CAS
36615 (4). Panama: LACM 9099-28 (5).
Umbrina roncador 367 specimens (23-338 mm) from 88
collections. California: SIO H45-130 (17); SIO H45-162 (3);
SIO H46-94 (6); SIO H47-160 (2); SIO H48-101 (1); SIO
H51-235 (3); SIO H49-90 (5); SIO 86-63 (2); SIO 88-91 (17);
CAS 18797 (8); CAS-SU 12666 (1); CAS-SU 19311 (2);
CAS-SU 9913 (4); CAS 19515 (1); CAS 19672 (1); CAS 18532
(1); CAS 18527 (1); CAS 18347 (2); CAS 12984 (2); CAS 18272
(2); UCLA W57-208 (8); UMMZ 162170 (1); UMMZ 177364-5
(1); UMMZ 177457 (1); LACM W.58-77 (2); LACM W50-126
(2); USNM 132385 (5); USNM 31316 (1); USNM 31317 (1);
USNM 31270 (1); USNM 26872 (4); USNM 26758 (4); USNM
5299 (1); USNM 52978 (1); USNM 59496 (3); USNM 54332
(1); USNM 34781 (1); USNM 132394 (1); USNM 124991 (15);
USNM 26849 (1). Mexico, Baja-west coast: SIOH46-215A
(7); SIO H48-56 (1); SIO H48-55 (1); SIO H52-160 (8); SIO
62-729 (2); SIO H48-48 (12); SIO 62-113 (1); SIO H48-88 (1);
SIO H52-137 (8); SIO H52-149 (8); SIO H52-135 (9); SIO
60-364 (5); SIO 60-364 (5); SIO 60-367 (1); SIO H48-91 (1);
SIO 62-217 (3); UCLA W61-107 (2); UCLA W52-93 (3); UCLA
W51-221 (1); UCLA W52-236 (3); LACM W52-248 (7); LACM
W52-270 (3); LACM W51-234 (1); CAS-SU 58622 (8); CAS-
SU 47932 (4); CAS W52-245 (8); CAS 11713 (2); CAS
W52-183 (1); CAS W52-85 (2); CAS W52-93 (2); CAS W52-101
(2); USNM 54514 (1); USNM 132406 (1); USNM 46730 (2).
Mexico, Gulf of California: UAZ 57 (1); UAZ 156 (71);
584
Fishery Bulletin 90(3). 1992
Figure 1 1
(upper) Umbrina xanti, 34mmSL, SIO 61-232; (lower) U. dorsalis, 33mmSL, SIO 61-232.
SIO 62-217 (3); SIO 63-532 (4); SIO 70-70 (1); CAS-SU 16585
(3); CAS-SU 6327 (1); CAS-SU 187 (1); UCLA W56-76 (3).
Umbrina xanti 397 specimens (20-296 mm) from 63 col-
lections. Mexico, Baja-west coast: SIO 62-705 (10); SIO
62-706 (136); SIO 62-707 (2); SIO 62-708 (1). Mexico, Gulf
of California: UCLA W51-22 (21); UCLA W51-24 (7); UCLA
W51-56 (4); UCLA W51-57 (17); UCLA W56-118 (2); UCLA
W59-248 (3); LACM W51-41 (1); LACM W51-49 (1); LACM
W52-50; LACM W58-46 (1); CAS 2570 (1); CAS 2571 (1); CAS
4489 (1); CAS W51-53 (7); UBC BC 59-216 (1); UBC BC
59-236 (1); USNM 2996 (2); USNM 47443 (4); USNM 47463
(3); USNM 122659 (1); SIO 61-232 (4); SIO 65-349 (1).
'Mexico, southern: MCZ 485 (1); SIO 62-23 (31); SIO 62-28
(2); SIO 62-47 (1); LACM W58-12 (1); UBC BC 57-78 (2); UBC
BC 57-85 (1); UBC BC 57-94 (1); UBC BC 57-98 (1);UBC BC
57-108 (1); UBC BC 57-129 (1); UBC BC 60-12 (1); UBC BC
60-13(1). Guatemala: UCR 355-6 (2). Nicaragua: UCR
379-13 (1). Costa Rica, incl. 1. Cocos: UCR 218-39 (1);
UCR 259-3 (1); UCR 137-6 (3); LACM 35473-1 (40); SIO 77-89
(38). Panama: UCLA W53-283 (3); UCLA W53-285 (3);
USNM 82233 (1). Equador: USNM 88744 (2). Peru:
CAS-SU 29850 (2); CAS-SU 37491 (3); UCLA W60-34 (1);
UBC BC 56-145 (1); UBC BC 56-149 (1); UBC BC 56-159 (1);
UBC BC 56-162 (1); UBC BC 56-165 (1); UBC BC 56-234 (1);
USNM 107150 (1); USNM 128009 (8); USNM 128010 (1).
Chile: Univ. Antofagasta. uncat. (1).
Umbrina reedi 86 specimens (9-650 mm) from 9 collec-
tions. SIO 65-623 (1); SIO 65-625 (1); SIO 65-626 (2); SIO
65-655 (2); SIO 65-657 (34); UCLA W66-50 (2); UCLA W66-56
(29); USNM 176411 (7); USNM 88745 (8).
Umhrina dorsalis 93 specimens (28-327 mm) from 25 col-
lections. Mexico: SIO 61-232 (12); SIO 61-236 (1); SIO
61-251 (7); SIO 62-23 (3); SIO 62-705 (1); UBC BC 57-100 (1):
LACM W55-120 (1); LACM 9044-28 (34); LACM 9045-23 (2):
LACM 9051-23 (2); UCLA W51-57 (1); UCLA W53-185 (1):
UCLA W59-248 (1). Costa Rica: USNM 94000 (1); UCLA
W54-168 (2); UCLA W54-172 (1). Panama: SIO 90-30
(1); UCLA W53-283 (1); UCLA W53-285 (1); UBC BC 60-
117 (1); USNM 144680 (1); USNM 81213 (1); CAS-SU 8113
(2). Colombia: UMML Argosy 28, uncat. (1). Equador:
UMML Argosy 44, uncat. (13).
Umbrina galapagorum 253 specimens (29-395 mm) from
27 collections. MCZ 8597, 8602 (38); MCZ 40448 (1); USNM
153626 (1); CAS 2311 (1); CAS 2313-2318 (6); CAS 2373 (1);
CAS 2374-2379 (6); CAS 2384 (1); CAS 2385 (1); CAS 4486
(1); CAS 4657 (1); CAS 4952 (1); CAS 6326 (1); CAS 62979
Walker and Radford Eastern Pacific species of thie genus Umbnna
585
(2); CAS-SU 24402 (2); CAS-SU 24413 (3); LACM W53-21
(5); LACM W53-28 (9); SIO 62-641 (151); SIO 79-51 (3); UBC
BC 56-429 (2); UBC BC 56-437 (1); UCLA W50-219 (2); UCLA
W53-144 (1); UCLA W55-314 (1); UCLA W56-325 (1).
Umbrina analis 8 specimens (51-229mm) from 6 collec-
tions. Colombia: CAS 62852 (2); CAS 62853 (1); FMNH
56840(2). Panama: USNM 81212(1). Costa Rica: LACM
33822-32 (1); USNM 94612 (1).
Umbrina wintersteeni 9 specimens (56-276 mm) from 7 col-
lections. Mexico: SIO 60-365 (3); UCLA W51-18 (1); CAS-
SU 4827 (1); CAS W53-94 (1); USNM 29430 (1); USNM 46956
(1); USNM 47463 (1).
Acknowledgments
We are especially grateful to Carl L. Hubbs (deceased)
for instigating this study and for providing data on
many specimens. We wish to acknowledge the guid-
ance, and data on holotypes, given by Labbish N. Chao.
We thank the following individuals for loan material
and/or correspondence relating to this study: M.E.
Anderson, W. Baldwin, J.C. Briggs, W.A. Bussing,
D. Buth, D. Catania, B. Chernoff, L.J. Dempster, A.W.
Ebeling, W. Eschmeyer, R. Feeney, N. Feinberg, W.L
Follett, K. Fujita, K.E. Hartel, S. Jewett, L.W. Knapp,
L Kong U., R.J. Lavenberg, H.C. Loesch, J.D. McPhail,
W.J. Rainboth, S. Raredon, C.R. Robins, L.P. Schultz,
J.A. Siegel, P. Sonoda, M. Stehmann, C.C. Swift, D.A.
Thomson, E. Trewavas, B. Walker, and W.J. Wilimov-
sky. We also appreciate the assistance of D. Gibson
and C. Klepadlo at SIO. This paper is dedicated to
S. Walker, T. Walker, J. Walker, and L. Walker for
inspiration and support. Special thanks go to R.H.
Rosenblatt and W. Watson for critically reviewing the
manuscript.
Citations
Briggs, J.C.
1974 Marine zoogeography. McGraw Hill, NY, 475 p.
Chao, L.N.
1978 A basis for classifying western Atlantic Sciaenidae
(Teieostei: Perciformes). NOAA Tech. Rep., Circ. 415, 64 p.
1986a A synopsis on zoogeography of the Sciaenidae. In
Uyeno, T., et al. (eds.), Indo-Pacific fish biology: Proc, 2d Int.
Conf. on Indo-Paclfic fishes, p. 570-589. Ichthyol. Soc. Jpn.,
Tokyo.
1986b Sciaenidae. In Whitehead, P.J.P.. et al. (eds.), Fishes
of the northeastern Atlantic and the Mediterranean, vol. II,
p. 865-874. UNESCO, Paris.
Chao, L.N., and R.V. Miller
1975 Two new species of sciaenid fishes (Tribe: Sciaenini) from
the Caribbean Sea and adjacent waters. Bull. Mar. Sci. 25(2):
259-271.
Cuvier, G.
1816 Le regne animal, ed. 1, vol. 2. Paris.
Frey, H.W. (editor)
1971 California's living marine resouces and their utiliza-
tion. Calif. Dep. Fish & Game, Mar. Resour. Agcy, Sacra-
mento. 148 p.
Gilbert. C.R.
1966 Western Atlantic sciaenid fishes of the genus Umbrina.
Bull. Mar. Sci. 16(2):230-258.
Gill, T.
1862 Catalogue of the fishes of lower California, in the Smithso-
nian Institution, collected by Mr. J. Xantus. Proc. Acad. Nat.
Sci. Phila. 13:249-262.
Gistel. J.
1848 Naturgeschichte des Thierreichs. Fiir hohere Schulen.
Scheitlin & Kiais, Stuttgart, 216 p.
Giinther, A.
1869 An account of the fishes of the states of Central America,
based on collections made by Capt. J.M. Dow, F. Godman, Esq.,
and 0. Salvin, Esq. Trans. Zool. Soc. Lond. 6:377-494.
1880 Report on the shore fishes. In Report on the scientific
results of the voyage of H. M.S. Challenger during the years
1873-1876. Zool. Challenger Exped. Pt. 6, London, 82 p.
Hubbs, C.L.
1921 Description of a new sciaenid fish from Santa Catalina
Island, California. Occas. Pap. Mus. Zool. Univ. Mich. 92, 3 p.
Hubbs, C.L., and K.F. Lagler
1958 Fishes of the Great Lakes region (rev. ed.). Cranbrook
Inst. Sci. Bull. 26, 186 p.
Jordan. D.S.. and C.H. Gilbert
1882 List of fishes collected by Lieut. Henry E. Nichols,
U.S.N. , in the Gulf of California and on the west coast of lower
California, with descriptions of four new species. Proc. U.S.
Natl. Mus. 4:273-279.
Leviton, A.E., R.H. Gibbs Jr., E. Heal, and C.E. Dawson
1985 Standards in herpetology and ichthyology: Part I. Stan-
dard symbolic codes for institutional resource collections in
herpetology and ichthyology. Copeia 1985:802-832.
Linnaeus, C.
1758 Systema naturae. Regnum animate. Tome I, 10th ed..
Holmiae, 824 p.
Lopez S., M.I.
1980 Umbrina biissingi, a new sciaenid fish from the tropical
eastern Pacific Ocean. Rev. Biol. Trop. 28(l):203-208.
McPhail, J.D.
1958 Key to the croakers (Sciaenidae) of the eastern Pacific.
Univ. Brit. Col., Inst. Fish., Mus. Contrib. 2, Vancouver, 17 p.
Miller, D.J., and R.N. Lea
1972 Guide to the coastal marine fishes of California. Calif.
Dep. Fish & Game Bull. 157, 235 p.
Miller, R.V.
1971 A new sciaenid fish (Pisces: Umbrinini) with a single
mental barbel, from the southern Caribbean. Copeia 1971:
300-306.
Ostroumoff, A. A.
1896 Zwei neue Relicten-Gattungen im Azow'schen Meere.
Zool. Abz. V. 19(493):30.
Rosenblatt, R.H.
1967 The zoogeographic relationships of the marine shore
fishes of tropical America. Stud. Trop. Oceanogr. 5:579-592.
Scofield, N.B.
1896 Family Sciaenidae. In Jordan, D.S., Notes on fishes,
little known or new to science, p. 220-22 L Proc. Calif. Acad.
Sci., 2d ser., 6:201-244.
Springer, V.G.
1958 Systematics and zoogeography of the clinid fishes of the
subtribe Labrisomini Hubbs. Publ. Inst. Mar. Sci., Univ. Tex.
5:417-492.
586
Fishery Bulletin 90(3), 1992
Steindachner, F.
1878 Uber zwei neue Fischarten von den Galapagos-Insulen.
In Ichthyologische Beitrage (VII), p. 19-21. Akad. Wiss. Wien
Sibzber, Abth. I. 1878 (1879):l-24.
Trewavas, E.
1977 The sciaenid fishes (croakers or drums) of the Indo- West-
Pacific. Trans. Zool. See. Lond. 33:253-541.
Walker, H.J. Jr., and R.H. Rosenblatt
1988 Pacific toadishes of the genus Porichthys (Batrachoididae)
with descriptions of three new species. Copeia 1988:887-904.
Wilson, C.
1916 Some marine fishes from Colombia and Ecuador. Ann.
Carnegie Mus. 10(l-2):57-76.
Table 1
Number of soft dorsal fin
rays
in eastern Pacific species of Umbrina.
(Asterisks indicate count of
primary
type.)
21
22
23
24
25
26
27
28
29
30
31
32
33
X
SE
U. analis
6
1
2*
24.56
0.29
U. bussingi
U. dorsalis
5
19*
4
1
3
15
21
8
2*
21.96
30.76
0.11
0.14
U. galapagorum
U. reedi
1
1
4
22*
21
9
1
59*
89
58
24
28.12
24.28
0.07
0.12
U. roncador
4
11
72
110
105*
46
6
1
27.32
0.06
U. vnntersteeni
1
3
25
37*
23
3
25.95
0.10
U. xanti
13
77
168*
114
30
4
1
28.03
0.04
1
Table 2
Number of soft anal fin rays in eastern
Pacific species of 1
Umbrina. (Asterisks indicate count of
primary
type.)
5
6
7
8
9
10
X SE
U. analis
9*
6,00 -
U. bussingi
30*
7.00 -
U. dorsalis
45*
5
7.10 0.04
U. galapagorjim
1
164*
73
6.30 0.03
U. reedi
2
46"
2
9.00 0.04
U. roncador
17
339*
6.95 0.01
U. wintersteeni
2
89*
5.98 0.02
U. xanti
402*
6
6.01 0.01
Table 3
Number of pectoral fin rays in eastern Pacific species of |
Umbrina. (Asterisks indicate count of primary
type.)
14 15
16
17 18 19
20
X SE
U. analis
2
14* 2
17.00 O.U
U. bussingi
3 30* 19
18.31 0.08
U. dorsalis 1
65*
33
16.32 0.05
U. galapagorum
1
57 323* 78*
4
18.06 0.03
U. reedi
12 69 19*
18,07 0,06
U. roncador 1
11
175* 422 48
1
17,77 0.02
U. wintersteeni 2
16
147* 17
16.98 0.04
U. xanti 1 1
20
532 192* 8
17.24 0.03
Table 4
Number of total gill rakers
in eastern Pacific species
of Umbrina.
14
15
16
17
18
19
20
21
22
23
24
25
X
SE
U. analis
1
5
3
1
17.60
0.43
U. bussingi
1
4
6
11
3
2
19.63
0.23
U. dorsalis
8
35
25
17
7
4
I
1
20.01
0.14
U. galapagorum
1
38
248
126
48
4
18.42
0.04
U. reedi
1
14
11
3
1
19.63
0.16
U. roncador
12
26
48
121
160
120
23
1
18.66
0.06
U. wintersteeni
2
5
12
21
46
43
17
5
4
2
18.32
0.13
U. xanti
10
91
475
90
21
1
18.03
0.03
Walker and Radford: Eastern Pacific species of the genus Umbnna
587
Table 5
Selected morphometries of eastern Pacific species of Umbrina. HL = head length; SN = snout length; INT = interorbital width;
SUB = suborbital width: UJ = upper jaw length; FED = caudal peduncle depth; PEL = caudal peduncle length; PA = preanal length;
PF = pectoral fin length; PRD = predorsal fin length; DEP = greatest depth; ASP = second anal spine length.
SL
Range (10
-^SL)
(mm) HL
EYE
SN
INT
SUB
UJ
FED
PEL
PA
PF
PRD
DEP
ASP
U. analis
52-231 292-332
67-96
76-103
61-87
50-63
91-120
100-117
220-263
670-721
183-204
362-386
302-342
170-214
U. bussingi
59-229 338-401
87-108
62-92
70-94
41-57
111-135
76-99
233-265
663-742
239-287
332-417
298-344
154-186
U. dorsalis
49-327 282-364
57-116
81-103
80-125
37-59
115-173
92-117
235-278
607-652
185-265
344-369
315-378
117-181
U. galapagorum
30-394 269-354
43-90
86-110
69-96
44-68
87-132
83-103
240-292
610-676
168-190
341-376
249-322
92-172
U. reedi
85-650 310-360
40-88
83-113
74-101
41-62
110-147
98-122
209-263
634-685
216-247
378-409
306-396
82-159
U. roncador
22-339 267-342
38-88
71-99
68-102
31-49
93-152
77-116
219-282
641-685
155-195
333-361
247-329
91-146
U. wintersteeni
30-271 279-344
57-97
83-105
66-100
51-60
92-134
88-112
218-255
648-686
161-199
334-365
261-315
112-168
U. xanti
20-287 267-316
49-105
69-107
66-99
31-36
85-152
80-113
239-316
605-642
150-172
328-358
235-330
97-157
Table 6
Number of procurrent
caudal fin rays
in eastern Pacific
species
of Umb
rina.
Dorsa
Ventral
6 7
8
9
10
X
SE
6
7
8
9
10
X
SE
U. analis
4
2
8.33
0.21
2
4
7.66
0.21
U. biissingi
6
12
7.67
0.11
12
5
1
7.39
0.14
U. dorsalis
1 17
16
4
7.61
0.12
10
22
1
6.73
0.09
U. galapagorum
2
40
12
9.19
0.03
4
33
13
8.18
0.08
U. reedi
1
19
9
9.28
0.05
15
10
2
8.52
0.12
U. roncador
5
29
15
9.20
0.05
1
27
13
8.29
0.08
U. wintersteeni
12
30
2
7.77
0.08
7
27
4
6.92
0.05
U. xanti
16
33
5
8.79
0.08
8
30
12
8.08
0.08
Abstract. -Accurate and precise
descriptions of behavioral indicators
of human activities which disturb
cetaceans are required to better con-
trol adverse human impacts on these
animals. We hypothesize that the
application of a technique used to
remove a small piece of innervated
tissue, a biopsy darting procedure, is
likely to result in the display of such
behavioral indicators. In order to
describe such displays, we recorded
behavior of 22 humpback whales
Megaptera novaeangliae before and
after biopsy procedures in the south-
ern Gulf of Maine. Reactions varied
considerably among animals. Al-
though respiratory responses were
not consistent, biopsied whales gen-
erally decreased their ratio of sur-
face to dive time and their net move-
ment rate. Hard tail flicks occurred
as an immediate reaction in approx-
imately half the cases. Although 31
behaviors were tested for variation,
only hard tail flicks significantly
increased in either the number of
animals that displayed them or the
overall frequency of occurrence dur-
ing postbiopsy reaction periods.
While not statistically significant,
some increase was noted in the fre-
quency of trumpet blows and tail
slashes, while slow swimming and
apparent investigative behavior
were noted to decrease. The strong-
est reactions, observed in two cases,
occurred when the dart and retrieval
line briefly snagged the whale's
flukes. These findings complement
and extend other studies on the re-
sponse of baleen whales to human ac-
tivity at sea.
Behavioral reactions of liumpback
wfiales Megaptera novaeangliae
to biopsy procedures
Mason T. Welnrich
Cetacean Research Unit, PO Box 159, Gloucester, Massachusetts 01930
Richard H. Lambertson
Department of Physiological Sciences, College of Veterinary Medicine
University of Florida, Box J- 144, JHMHC. Gainesville, Florida 32610
Cynthia R. Belt
Mark R. Schilling
Heidi J. Iken
Cetacean Research Unit. P O Box 159. Gloucester. Massachusetts 01930
Stephen E. Syrjala
Resource Assessment and Conservation Engineering, Alaska Fisheries Science Center
National Marine Fisheries Service, NOAA
7600 Sand Point Way NE, Seattle, Washington 981 15-0070
Manuscript accepted 1 June 1992.
Fishery Bulletin, U.S. 90:588-598 (1992).
The humpback whale Megaptera no-
vaeangliae is an endangered species
that has been protected from com-
mercial catches since the mid-1960s.
Protection from himting in the North
Atlantic portion of its range extends
back to 1955. The most recent popu-
lation estimates suggest close to
10,000 animals remain worldwide,
with 5500 in the western North At-
lantic (Johnson and Wolman 1985).
The endangered status of this spe-
cies as well as its affinity for near-
shore habits has brought increasing
concern that the collective effects of
industrial development, resource ex-
ploitation, and rapid increase in the
whale watching industry could result
in displacement, habitat degradation,
and behavior modification. It thus
has become important to determine
whether human activity that is not
directly lethal to individual whales
could still have deleterious effects on
the recovery of this species.
To assess potential deleterious
effects of artificial stimuli on the
normal behavior of a whale, defini-
tions of disturbed behavior must be
clarified. Disturbed behavior can be
defined as behavior that results from
a noxious stimulus that would not
otherwise have occurred. Previous
observations have been made under
conditions of potential disturbance,
such as the presence of boats or
divers or the production of under-
water noise (Baker and Herman
1982, Malme et al. 1983 and 1985,
Bauer and Herman 1985, Richardson
et al. 1985). However, a cause-and-
effect relationship between the stim-
ulus and a whale's response has been
difficult or impossible to achieve,
since baseline data on behavioral
reactions to clearly noxious stimuli
are almost entirely lacking.
Since 1979, humpback whales have
been studied intensively in the south-
ern Gulf of Maine to evaluate the
demographics, behavior, and ecology
of a group of annually-returning in-
dividuals (Mayo et al. 1985, Weinrich
1985 and 1986, Clapham and Mayo
1987). In 1983, the University of
Florida began studies to determine
the genetic characteristics and sex of
known individuals in this group of
588
Weinrich et al : Behavior of Megaptera novaeangliae during biopsy
589
whales. To obtain the tissue samples for this project,
a projectile biopsy dart was used (Lambertsen 1987,
Lambertsen and Duf field 1987, Lambertsen et al.
1988).
In an attempt to better understand the disturbance
response of large whales, the present study was under-
taken to assess the behavioral reaction of humpback
whales to the biopsy procedures previously described
(Lambertsen 1987). We reason that since the biopsy
dart removes a small piece of innervated tissue, it is
likely to be perceived by the whale as a noxious stim-
ulus and should cause some observable response. Our
results compare the behaviors of the whales before and
after exposure to this relatively short-term, moderate-
level stressful stimulus.
Materials and methods
Cruises to collect biopsy samples took place each year
from 1983 to 1985 on Jeffrey's Ledge or Stellwagen
Bank in the Gulf of Maine. In 1983 and 1984 various
types of small power vessels (< 12.8 m) were used, in-
cluding an 11.6m sportfishing vessel equipped with one
Detroit Diesel 671 engine, and a 12.8m pilot boat with
two Detroit Diesel 671 engines. In 1985, a 6.1m run-
about with a 175hp outboard engine and an 11m sail-
ing sloop were used simultaneously. The use of two
vessels allowed one (the 6.1 m runabout) to be dedicated
to collection of behavioral data in the methodology
described below. The immediate response of whales to
biopsy darting was recorded on two days in 1983, two
in 1984, and six in 1985.
The biopsy apparatus used in this study consisted of
a tethered retrievable biopsy dart, aimed at the flank
below the dorsal fin, fired from a 68 kg crossbow
(Lambertsen 1987). A small biopsy punch fitted with
internal prongs and attached to the tip of the dart shaft
removed the tissue from the animal. A small tissue
sample, including both epidermis and dermis, was thus
obtained by a cutting action on penetration, and tear-
ing on rebound. Upon penetration of the dart into the
whale, a rebound was forced by a 2.5cm diameter
flange set 2 cm back from the tip of the biopsy punch.
In the first 2 years of the study, emphasis was placed
on collecting as many biopsies as possible during brief
periods at sea. Behavioral observations were collected
opportunistically to provide qualitatively classified data
on immediate reactions. For comparative purposes,
these observations were ranked in a manner similar to
that used by Mathews (1986), who studied the reactions
of eight gray whales Eschrichtius robustus to a similar
biopsy procedure. Categories used in this initial anal-
ysis included:
No reaction The whale continued its prebiopsy be-
havior with no detectable change.
Low-level reaction The animal modified its behavior,
but displayed none of the overtly forceful behaviors
listed as moderate or strong reactions (e.g., imme-
diate dive).
Moderate reaction The animal modified its behavior
in a more forceful manner (trumpet blows, hard tail
flicks), but gave no prolonged evidence of behavioral
disturbance.
Strong reaction The animal modified its behavior to
a succession of forceful activities (continuous surges,
tail slashes, numerous trumpet blows).
To test statistical differences in reaction levels among
age-classes, non- and low-level reaction frequencies
were combined, as were moderate and strong reac-
tions. This was necessitated by the low expected values
of the frequencies in a chi-square table based on the
data in Table 1.
All data from 1983-85 were used to categorize im-
mediate response levels; in 1985, a 30-min prebiopsy
control period and a 30-min postbiopsy response
period were defined to standardize a paired data set
of respiratory and other surface behaviors. This ap-
proach used the vessel dedicated to behavioral data col-
lection to institute a focal sampling technique (Altmann
1974) to allow quantitative comparison of the pre- and
postbiopsy focal periods as "paired samples." Upon
sighting a group of whales, each individual was dis-
tinguished through 7 X 50 binoculars using distinctive
natural markings on the dorsal fin or the ventral sur-
face of the tail flukes (Katona and Whitehead 1981).
Because of the necessity of identifying individual
respirations and behaviors within the group, dorsal
fin shape was used whenever possible during focal
samples; permanent identification came from fluke
photographs taken during the approach for the biopsy
strike. Once individuals were distinguished from one
another, a 30-min "control" (i.e., pretreatment) focal
sample was then initiated. During this period the
engines of the observation vessel were shut down to
eliminate engine noise. No approaches closer than
100 m were made prior to the onset of, or during, the
30-min control period. If the whale moved farther than
1000m from the research vessel, making data collec-
tion difficult, the engine was started and approach was
made slowly to within ~300m of the whale. At the con-
clusion of the 30-min "control" focal sample the whale
was then approached at close range (3-40 m) for the
biopsy attempt. The same protocol was followed after
the biopsy for a comparable "experimental" (i.e., post-
treatment) data set, which started at the moment of
impact by the dart.
During focal samples, data were collected on four
respiratory variables: (1) number of respirations
("blows") during a given surface interval, (2) time
between each respiration ("blow interval"), (3) time
590
Fishery Bulletin 90(3). 1992
the animal spent at the surface ("surface interval"),
and (4) time spent below the surface during each dive
("dive time"), defined as the period of submergence
following typical sounding behavior (i.e., a prominent,
high arching of the back and tail, often followed by
bringing the tail flukes above the water surface). The
surface interval during which the biopsy strike took
place was excluded from our analyses of both the pre-
and postbiopsy respiratory variables. We determined
(using chi-square goodness-of-fit tests) that the distribu-
tion of the observed data was not significantly different
from a normal distribution, thus differences between
the pre- and postbiopsy values were compared using
two-tailed paired (-tests (Zar 1984). Surface-interval
to dive-time ratios were calculated for each whale
during pre-biopsy and postbiopsy focal samples, and
compared using a Wilcoxon sign rank test (Zar 1984).
The surface-interval/dive-time ratio integrates several
respiratory values into a single measure of the res-
piratory "strategy" for each individual.
LORAN-C positions, which have an error of ~30m
in the study area (Day 1983), were used in estimating
the net rate of movement of a whale as defined by its
surfacings. LORAN positions were recorded at the
start and end of each focal sample. With each LORAN-
C reading, the bearing to the whale to the nearest 5
degrees and the visually-estimated distance from the
vessel to the whale were also recorded. This informa-
tion was used to correct the LORAN-C data if the
whale was >30m from the vessel, which was likely
given the limitations of vessel movement during focal
samples described above. To estimate the net move-
ment rate, the distance between the first and last cal-
culated whale positions within each focal sample period
was divided by the elapsed time (30 minutes), yielding
a net movement rate in knots. Actual swimming speed
could not be determined due to uncertainty about the
direction of a whale's movement underwater or the
linearity of its track. Results between pre- and post-
biopsy periods were compared using a Wilcoxon sign
rank test (Zar 1984).
Photographs of the dorsal fin and tail flukes of in-
dividual whales were taken upon approach for biopsy.
Each whale was identified using the catalog of Gulf of
Maine humpback whales kept at the Cetacean Research
Unit, where individuals are assigned a two-letter, one-
number file code. If the animal could not be identified
in the catalog, it was assigned a three-digit code. When
possible, each animal was assigned to one of the follow-
ing age groups: juvenile (1-3 yr), adolescent (4-6 yr),
or adult (>6 yr). Age classifications were based on
previous and/or subsequent repeated annual observa-
tions of the same individuals by the authors from calf
year (used for juveniles and adolescents), sightings of
the individual as an initially small and subsequently
larger animal (juveniles and adolescents), or annually
repeated sightings of an individual with no appreciable
growth over several years (adults). If an animal was
sighted only during the year in which it was biopsied,
it was not classified by age-class.
During focal samples collected in 1985, a total of 30
behavior types were observed and analyzed. Behavior
types were defined using an ethogram for humpback
whales developed by the Cetacean Research Unit prior
to this study (unpubl. data). The probability that any
given behavior was displayed by more or fewer animals
in the pre- vs. postbiopsy focal samples was tested
using the binomial distribution (with the probability of
each period containing an occurrence of the behavior
assumed as 0.5), while the change in frequency of each
behavior in individuals, given that the behavior was
observed at all, was compared using Wilcoxon signed
rank tests (Zar 1984). All 30 behaviors were tested for
variation. Many of the behaviors did not show any vari-
ation between control and response periods, and there-
fore are not described in detail. These were belly-up
rolls, breaches, bubble clouds (bubble clouds followed
by obvious surface feeding), bubble cloud behaviors
(bubble clouds not followed by obvious surface feeding),
defecations, flipper flares, flipper flicks, flipper in air,
flukes, half flukes, hangs, high flukes, high head-ups
("spyhops"), lobtails, logging, low flukes, quarter rolls,
single bubbles, snakes (a twisting of the body), surges,
tail breaches, and passing under a boat. Definitions of
those behaviors which either varied significantly in
frequency or showed some notable variation in the fre-
quency of display following the biopsy procedure are
the following:
Back rise Animal breaks surface while swimming,
with no accompanying exhalation.
Belly-up lobtail Animal, ventral side up, elevates tail
into the air, then slaps the water surface with the
dorsal surface of its flukes.
Hard tail flick Animal rapidly and forcefully flexes
tail up and down one time during otherwise normal
swimming behavior; much spray can be thrown;
flukes clear surface. The hard tail flick is faster and
presents a less regular arching movement of the tail
than a lob-tail.
Low head-up Animal lifts head into air at 30-45°
angle to surface.
Sounding dive Animal arches its back in a typical div-
ing posture but does not bring its tail flukes above
the surface.
Tail rise Animal slowly straightens its caudal pedun-
cle at the surface during normal swimming.
Tail slash Animal moves tail forcefully from side to
side, flukes at or just below the surface; creates white
water frothing.
Weinrich et al Behavior of Megaptera novaeangliae during biopsy
591
Trumpet blow Loud, broad-band, wheeze-like sound
made during exhalation at the surface.
In addition, the following behaviors showed a variation
in frequency during the special case of SI4's reaction
(discussed below):
Belly-up Animal rolls so that the whale has its ven-
tral surface exposed above the surface (often for
longer than a second).
Half fluke Animal rolls on its side exposing one fluke
above and perpendicular to the surface.
Results
Immediate behavioral reactions, or the absence thereof,
to the biopsy procedure were recorded for 71 biopsy
strikes during the period 1983-85. Of these, 22 (re-
corded in 1985) are paired samples including a 3U-min
prebiopsy and 30-min postbiopsy focal sample. Two
cases contained clearly unusual reactions, including one
from the 1985 paired samples. These cases are dis-
cussed separately. This leaves 21 paired samples of
behavioral data; however, in five cases some respira-
tions could not be accurately assigned to a whale within
the focal group, leaving 16 paired samples of complete
respiratory data for analysis.
Immediate behavioral response
Of the 71 total biopsy attempts for which immediate
behavioral reactions were recorded, 7.0% involved no
behavioral reaction, 26.8% involved a low-level re-
action, 60.6% involved a moderate reaction, and 5.6%
involved a strong reaction (Table 1). All the strong reac-
tions involved snagging of the flukes by the mono-
filament line attached to the biopsy dart.
Immediate dives were the most common response to
the biopsy dart striking the animal, observed in 35
(49.2%) cases, hard tail flicks were present in 34
(47.8%), and trumpet blows were observed in 31
(43.6%) cases. Less than 20% of all reactions involved
immediate surges or visually detectable increases in
swimming speed.
Although an immediate dive was a frequently ob-
served response, this may have been due to the time
it took to approach the whale for the biopsy strike, i.e.,
the whale would have taken a dive at that point
regardless of the biopsy attempt. However, the mean
number of blows (4.89) during the surfacing interval
in which the biopsy dart was fired was significantly
lower than in the accompanying complete surfacings
immediately prior to the biopsy attempt (7.17) (paired
i-test: t -2.76, 15 df,jO 0.015),suggesting those dives
which occurred immediately after the strike of the
Table
1
Qualitative ranking of intensity of behavioral responses in
humpback whales Megaptera novaeangliae to biopsy proced-
ures. NR = No reaction.
NR
Low
Moderate
Strong
Total
Juveniles
2
3
10
3
18
(1-3 yr)
Adolescents
0
4
6
0
10
(4-6 yr)
Adults
3
12
23
1
39
(>6yr)
Unclassified
0
0
4
0
4
Total
5
19
43
4
71
biopsy dart were initiated as a response to the biopsy
procedure.
Study animals could be categorized by age-class in
68 of the 71 trials. There was no significant difference
in the intensity of reactions by age-class (x" 2.88, 3 df,
p 0.41) (Table 1). However, 3 of the 4 reactions we
ranked as strong were from juveniles. Also, strong
reactions were always associated with a snagging of
the retrieval line on the animals' flukes.
Respiratory and dive variables
There were no significant differences between pre- vs.
postbiopsy focal samples for any of the four respiratory
variables (paired i-tests (15 df): blow interval t 0.82,
p 0.42; number of blows/surfacing interval t -0.93,
j3 0.36; surfacing interval ^1.65, jaO.ll; dive time
t 0.61, p 0.55). There was a significant decrease in
the surface-interval/dive-time ratios during postbiopsy
focal samples (Wilcoxon signed rank test, Z -2.11,
p 0.03).
Substantial individual variation was found in
respiratory variables. Seven animals (43.8%) showed
a decrease in their mean blow interval following the
biopsy procedure, and eight (50.0%) showed an increase
(Table 2). Eleven individuals (68.8%) showed a decrease
in the number of blows per surfacing, while in only four
(25.0%) did it increase (Table 3). Similarly, eleven
whales (69.0%) reduced their surface interval in the
postbiopsy period, while in five (32.0%) this variable
increased (Table 4). Finally, eight of the 16 individuals
(50.0%) were found to decrease their dive times dur-
ing the postbiopsy period, while in the other eight
(50.0%) it increased (Table 5). The surface-interval/
dive-time ratio also showed a decrease in 9 of the 16
animals (57.0%), while in 5 (32.0%) there was no change
and in 2 (13.0%) there was an increase (Table 6). Based
on binomial distribution, any case with 9 or more, or
2 or less, animals showing a change in a particular
592
Fishery Bulletin 90(3), 1992
direction would be significant at p 0.07 (based on 11
samples, ignoring the 5 that showed no change), in-
dicating a significant decrease in the surface-interval/
dive-time ratio in the sample.
To determine whether the immediate behavioral
response to the biopsy dart affected subsequent
Table 2
Mean blow intervals (s) and standard deviations in humpbacl< |
whales Megaptera novaeangl
iae during focal samples before
and after biopsy procedures.
Each individual whale is repre-
sented by a two-letter one-number code
or a three-number |
code.
Animal
j'rebiopsy
Postbiopsy
AT 5 SD
Difference
N
X
SD
007
18
27.9
35.9
17
22.0
14.8
-5.9
SE6
11
49.2
43.2
15
46.3
45.9
-2,9
ZEl
19
36.4
23.7
23
26.8
23.4
-9.6
547
23
34.7
27.2
18
38.2
18.5
3.5
SWl
26
23.3
24,0
21
32.8
16.5
9.5
TH6
22
28.3
15.8
16
28.9
22.5
0.6
LAS
17
22.3
12.1
25
22.3
15.5
0.0
TR2
23
19.3
20,5
26
24.5
14.4
5.2
KE2
12
83.7
74.3
11
51.7
37.1
-32.0
MEl
22
14.5
2.6
19
16.0
2.0
1.5
C09
22
16.3
5.3
21
15.9
4.0
-0.4
STl
19
34.6
23.8
30
33.3
26.4
-1.3
OCl
38
30.4
20.4
36
32.6
27.1
2.2
SMI
32
27.8
22.8
21
33.7
27.1
5.9
CIl
23
14.3
5.1
17
18.3
13.3
4.0
RA6
12
31.8
20.3
10
19.3
7.1
-12.5
Sample
30.9
16.8
28.9
9.5
-2.0
Table I
1
Mean number of blows per
surface interval
n humpback
whales Megaptera novaeangliae during focal sa
mples before
and after the
jiopsy procedure
. Each individual whale
is repre-
sented bj
a t
wo-letter one-number code
or a three
number
code.
Prebiopsy
Postbiop
sy
Animal
N
X
SD
N
X
SD
Difference
C07
4
4.4
1.1
4
8,6
4.2
4.2
SE6
2
8.3
6.0
5
3,7
1.9
-4.6
ZEl
3
10.0
6.0
5
5.0
3.7
-5,0
547
S
4.7
2.9
5
4.0
2.0
-0.7
SWl
3
5.8
1.0
4
5.8
1.7
0.0
TH6
5
6.5
2.4
6
6.0
2.9
-0.5
LAS
8
4.2
2.8
5
2.8
2.2
-1,4
TR2
3
8.7
3.1
4
7.3
3.0
-1,4
KE2
4
4.0
2.S
5
3.2
2.3
-0.8
MEl
7
4.1
1.9
2
8.0
2.8
3.9
COB
S
5.4
2,5
11
2.9
1.6
-2,5
STl
7
3.9
l.S
5
7.0
6.6
3.1
OCl
4
10.5
6.4
5
8.2
6.8
-2.3
SMI
6
6.3
4.3
5
5.2
4.8
-1.1
CIl
6
4.4
1,8
2
2.3
1.5
-2.1
RA6
4
2.3
1,7
4
3.5
1.3
1,2
Sample
5.8
2,4
S.2
2.1
-0.6
Table 4
Mean surface interval length (s) in humpback whales Megap-
tera novaeangliae during focal samples before and after the
biopsy procedure. Each individual whale
is represented by a
two-letter one-number code
or a
three-number code.
Animal
Prebiopsy
Postbiopsy
Difference
N
X
SD
N
X
SD
C07
4
117,7
135.2
4
196.4
123.2
78.7
SE6
2
399,0
291.9
5
140.8
156.1
-258.2
ZEl
3
369.3
251.4
5
110.3
119.0
-259,0
547
5
158.0
115,7
5
121.0
58.5
-37,0
SWl
3
115.5
49,3
4
199.5
87.0
84.0
TH6
5
206.6
96,2
6
135.8
94.8
-70.8
LA5
8
62.5
71,5
5
46.3
21.6
-16.2
TR2
3
154.0
104.5
4
168.0
106.6
14,0
KE2
4
256.8
177.2
5
135.4
106.5
-121,4
MEl
7
61.1
47.2
2
122.0
42.4
60,9
C09
5
76.0
41.4
11
29.6
19.6
-46,4
STl
7
113.9
104.6
5
222,4
244.1
108.5
OCl
4
390,3
292.3
5
264.4
282.4
-125.9
SMI
6
216,8
157.3
5
138.6
163.8
-78.2
CIl
6
57.5
35.8
2
36.1
49.9
-21.4
RA6
4
64.5
21.5
4
32.0
14.4
-32.5
Sample
176.2
120.1
131.1
70.0
-45.1
Table 5
Mean dive interval length (s) in humpback wha.\es Megaptera
novaeangliae during focal samples before and after the biopsy
procedure. Each individual whale is represented by a Lwo-
letter one-number code or a three-number code.
Animal
Prebiopsy
Postbiopsy
Difference
N
X
SD
N
X
SD
C07
4
92.2
84.6
4
194,8
112.7
102.6
SE6
2
171.3
94.7
5
177,2
79.8
5.9
ZEl
3
181.0
23.5
5
145,4
87.0
-35.6
547
5
146,7
113.6
5
314,0
76.7
167.3
SWl
3
394.7
149.8
4
145,8
121.8
-248.9
TH6
5
127.5
22.9
6
200.0
88.3
72.5
LA5
8
254.5
35.4
5
216.5
48.9
-38.0
TR2
3
412.5
9.2
4
391.3
253.4
-21.2
KE2
4
228.5
90.6
5
125.0
90.6
- 103.5
MEl
7
250.5
125.5
2
565.5
38.9
315.0
C09
5
161.0
129.9
11
120.3
62.9
-40.7
STl
7
122.0
107,1
5
204.6
251.3
82.6
OCl
4
153.8
75,7
5
124.3
127.8
-29.5
SMI
6
96.7
57,7
5
199.0
109.2
102.3
CIl
6
170.3
35,0
2
113.6
76.9
-56,7
RA6
4
324.8
115,4
4
355.0
167.5
30,3
Sample
205.5
98.9
224.5
18.4
19,0
Weinrich et al : Behavior of Megaptera novaeangliae during biopsy
593
responses in respiratory variables, we examined sep-
arately those animals that reacted to the biopsy strike
with an immediate hard tail flick (n 9), the most ob-
viously forceful immediate response to the biopsy
strike. This subset would therefore eliminate those
animals who may have not been affected by the biopsy
strike. However, variation among individuals during
the postbiopsy period was not appreciably different
from that portion of the sample where no hard tail flick
was observed (binomial test). Hence, the occurrence of
an immediate forceful response to the biopsy procedure
does not appear to be associated with subsequent
changes in respiratory variables.
Net movement rate
For 11 of the 21 animals, LORAN-C fixes allowed a
calculation of the animal's net movement rate in pre-
and postbiopsy focal samples. During the prebiopsy
sample, only two animals showed values > 1 kn. Dur-
ing the postbiopsy period, the average rate did not
increase significantly (Wilcoxon signed rank test,
Z -1.82, p 0.07). However, only three animals had
rates <lkn, and a generally increasing trend was
recorded (Table 7).
Other behavioral responses
To consider changes in behavior elicited by the biopsy
procedure, the possibilities of introducing new behav-
Table 6
Mean surface-interval/dive-time ratio in humpback whales |
Megaptera
novaeangliae during focal samples before and after
the biopsy procedure.
Each individual whale is represented
by a two-letter one-number code or a three-number code.
Animal
Prebiopsy Postbiopsy
Difference
C07
1.3
1.0
-0.3
SE6
2.3
0.8
-1.5
ZEl
2.0
0.8
-1.2
547
1.1
0.4
-0.7
SWl
0.3
1.4
1.1
TH6
1.6
0.7
-0.9
LAS
0.2
0.2
0.0
TR2
0.4
0.4
0.0
KE2
1.1
1.1
0.0
MEl
0.2
0.2
0.0
C09
0.5
0.2
-0.3
STl
0.9
1.1
0.2
OCl
2.5
2.1
-0.4
SMI
2.2
0.7
-1.5
CIl
0.3
0.3
0.0
RA6
0.2
0.1
-0.1
Sample
1.1
0.7
-0.4
iors or altering display rates of regularly observed
behaviors were both considered. The former was exam-
ined using the number of pre- and postbiopsy focal
samples during which each behavior type was observed,
while the latter was examined using the direction and
magnitude of changes in observed behavior types
within individual paired samples (Tables 8, 9). Only one
of the 30 tested behavior types showed significant dif-
ferences between the pre- and postbiopsy period.
Eleven of the 21 (52.3%) postbiopsy focal samples
contained a hard tail flick, while the behavior was not
observed in the prebiopsy focal samples (binomial
distribution, p<0.001). Only once was a hard tail flick
observed more than one time after a biopsy strike. This
also was the only case in which the hard tail flick was
not an immediate response to the biopsy dart. The
percentage of biopsy strikes where the reaction in-
cluded a hard tail flick among paired samples was not
significantly different from that of the larger 1983-84
sample, where 34 of 50 animals displayed the hard tail
flick (x" 1.54, 1 df, p 0.21).
As was the case in the number of 30-min samples in
which a behavior was displayed, only hard tail flicks
showed a significant increase in frequency during
the postbiopsy period (binomial distribution, p 0.001).
While results were not significant, one or more animals
also showed notable increases in the numbers of
trumpet blows, tail slashes, and belly-up lobtails follow-
ing the biopsy procedure; similar nonsignificant but
notable decreases were seen in back rises, tail rises,
and low head-ups (Tables 8, 9). The latter three
behaviors are associated with slow, unhurried travel,
resting, or interest in nonessential environmental
stimuli (e.g., boats, seaweed).
Table 7
Net movement rate (kn) in humpback whales Megaptera novae- |
angliae
during focal samples before and after the biopsy pro- |
cedure.
Each individual whale
s represented by
a two-letter
one-number code or a three-number code.
Animal
Prebiopsy
Postbiopsy
Difference
C07
0.5
1.7
1.2
SWl
0.7
1.5
0.8
TR2
0.8
2.0
1.2
MEl
0.8
4.5
3.7
STl
0.5
0.6
0.1
OCl
0.9
0.6
-0.3
SMI
0.8
1.3
0.5
RA6
3.8
1.5
-2.3
BIl
1.5
3.8
2.3
PE4
0.5
1.1
0.6
SI4
0.5
0.7
-0.2
Sample
1.0 (SD0.9)
1.7 (SD 1.3)
0.7
594
Fishery Bulletin 90(3|. 1992
Table 8
Frequency of various behavior types
individual whale is represented by a
n humpback
two-letter (
whales Megaptera not
5ne-number code or a
aeangliae before the biopsy procedure ("control period"),
three-number code.
Each
C07 SE6 ZEl 547 SWl TH6 LAS TR2 KE2 MEl C09 STl OCl SMI CIl RA6 BIl CRl SI4 FL2 PE4 Total
Half fluke
2
1
1
1
5
Quarter roll
Bubble cloud
1
2
1
2
Single bubble
Breach
Back rise
9
3
11 2
1
1
1
2
1
7
1
3
1
1
0
2
42
Tail breach
2
2
Belly up
Belly-up lobtail
Cloud behavior
1
2
8
1
0
10
Defecation
1
1
Flipper flick
Flipper in air
Flipper flare
Fluke
3
2 2
2
1
3 2 4
4
2
3
2
1
4
3
3
4
5
7
0
2
0
55
Hang
High fluke
High head-up
Hard tail flick
1
1
1
0
1
2
0
Low fluke
1
1
1
1
1
3
1
9
Low head-up
3
4
1
1
9
Log (min)
Lobtail
9.3
9.5
1
1
1
19
3
Snake
1
1
Sounding dive
Surge
Tail rise
4
2
2
2
3
2
2
3
1 2 1
1 3 1
3
1
2
3
1
3
2
2
1
1
1
1
11
1
1
7
1
2
29
4
44
Trumpet blow
Tail slash
4
1
1
2
1
2
1
2
3
2
1
2
4
5
3
4
7
38
3
Under boat
1
1
2
Feeding behavior was observed with equal frequency
in both the pre- and postbiopsy samples. Those animals
engaged in feeding activity showed virtually no reac-
tion to the biopsy attempt. A hard tail flick was never
observed from an animal engaged in feeding activity,
although it was observed during all other prebiopsy
behavioral modes. Logging (resting) behavior was also
displayed equally in both sample periods; however,
whales logging when biopsied were observed to tem-
porarily interrupt their logging period immediately
following the biopsy.
Special cases
Two special cases of behavior modification were noted
in conjunction with the biopsy procedure. Both involved
the monofilament retrieval line becoming briefly
snagged around one of the flukes of the whale. These
represent the most vigorous and prolonged reactions
to the biopsy procedure we observed.
In one case, for a period of time after the biopsy
strike (~ 16 min) the line remained looped around the
tip of one fluke of the tail and the animal behaved
abnormally, swimming at elevated speeds (6-7 knots)
in a roughly S-figured course. Although visually esti-
mated, this speed appears higher when compared with
values reported above. Another whale accompanied
this animal in its vigorous swimming.
SI4 exhibited another unusual reaction after a biopsy
(at a different time than the reaction reported for the
same individual in Table 9). This whale had been
associated with CRl during the day of the biopsy ef-
fort; 40 min prior to the first strike of SI4, CRl was
sampled with little reaction. When SI4 was first struck
by the biopsy dart its reaction was also minimal, but
a tissue sample was not obtained. The next shot (29
min later) missed the whale, but involved a momentary
snag of the line on the animal's tail stock. In response,
the animal started to trumpet blow with increasing fre-
quency but remained stationary and was easily ap-
proached. A third firing of the biopsy dart 1 1 min later
was successful in obtaining a tissue sample.
Weinrich et a\. Behavior of Megaptera novaeanghae during biopsy
595
Table 9
Frequency of various
behavior types
in humpbact
. whales
Megaptera
novaeanglia£
following
the biopsy procedure
'("
-eaction period"). |
Each individual whale is
represented by a two-letter
one
-number code or
a three-number code
C07 SE6 ZEl
547
SWl
TH6 LA5 TR2 KE2 MEl C09 STl OCl
SMI CIl
RA6 BIl
CRl
SI4 FL2 PE4 Total
Half fluke
3
1
2
1
1
1
9
Quarter roll
1
1
1
3
Bubble cloud
2
2
Single bubble
2
2
Breach
0
Back rise
1
1
1
1
3
1
1
3
1
3
4
20
Tail breach
2
2
Belly up
1
1
Belly-up lobtail
20
20
Cloud behavior
7
7
Defecation
0
Flipper flick
1
1
Flipper in air
0
Flipper flare
1
1
Fluke
4
2
2
1
2
4
1
2
3
2
7
2
2
2
3
2
1
2
4
7
55
Hang
1
1
High fluke
1
1
1
1
1
5
High head-up
0
Hard tail flick
1
1
1
1
1
1
1
2
1
1
1
12
Low fluke
1
1
1
1
1
1
2
1
2
1
2
14
Low head-up
1
1
2
Log (min)
3
3
13
19
Lobtail
4
1
1
1
7
Snake
0
Sounding dive
4
6
2
3
1
5
2
2
1
4
5
1
3
5
44
Surge
1
1
1
3
Tail rise
3
2
2
3
2
1
1
1
1
1
4
3
2
2
2
1
31
Trumpet blow
1
1
3
2
8
1
1
1
1
1
4
9
5
4
5
10
57
Tail slash
1
2
1
1
1
11
Under boat
1
1
o
Following the final biopsy attempt, SI4 started a
series of stereotypic actions. Every 45-60 s, the animal
would trumpet blow loudly, then tail slash or low-lobtail
(a quick, low version of lob- tailing behavior), surge for-
ward, and roll sideways with great force, often rolling
ventral-side-up and spiraling underwater. Periods of
submergence were <30s in all cases. The swimming
path was erratic, but the animal was never >100m
from the vessel. It passed immediately below the vessel
twice, repeatedly surfacing on alternating sides of the
boat. Swimming speed appeared greater than normal,
although net movement in any one direction was
minimal. During the same period CRl appeared placid,
although it did trumpet blow three times. After 14 min,
the vigorous behavior of SI4 suddenly ended, and both
animals started logging side by side. At this point, they
were within 25 m of the vessel. Logging behavior con-
tinued for at least 15 min at which point the observa-
tion was terminated.
In order to compare the intensity of SI4's reaction
with the sample analyzed above, we compared the rate
at which it displayed various behavior types in the post-
biopsy focal sample with the larger paired sample
(n 21). To obtain a mean number of occurrences of each
behavior type in the postbiopsy period, the total num-
ber of observations of each behavior type was divided
by the number of paired samples (Table 10). From these
data, it is clear that unusually high numbers of tail
rises, trumpet blows, half flukes, belly-ups, lobtails, tail
flicks, and tail slashes occurred in SI4's response.
Discussion
The results of this study indicate that behavioral reac-
tions of individual whales to the biopsy procedure are
detectable but do not appear to be severe. Immediate
reactions (hard tail flicks) took place in >50% of 71
biopsy strikes, which is especially noteworthy given the
rarity of this behavior in any other context. However,
no significant difference was seen in most of the 30
observed behaviors in 30-min pre- and postbiopsy
596
Fishery Bulletin 90(3). 1992
Table
10
Frequency of various behavior types observed
n a humpback
whale Megaptera
novaeangliae
(animal SI4) subjected to
repeated strikes of the biopsy dart compared
with mean for
the entire study population (not
including SI4)
Values given
represent average
over the 30-min postbiopsy focal sample.
N of study population = 21.
Behavior
Study population
SI4
Half fluke
0.42
19
Quarter roll
0.14
2
Bubble cloud
0.09
0
Single bubble
0.09
1
Breach
0.00
0
Back rise
0.95
7
Tail breach
0.08
1
Belly up
0.04
12
Belly-up lobtail
0.95
0
Cloud behavior
0.33
0
Defecation
0.00
0
Flipper flick
0.04
1
Flipper in air
0.00
3
Flipper flare
0.04
2
Fluke
2.61
5
Hang
0.04
0
High fluke
0.23
0
High head-up
0.00
1
Low fluke
0.65
0
Low head-up
0.09
4
Log (min)
0.90
0
Lobtail
0.33
26
Snake
0.00
0
Sound
2.09
0
Surge
0.14
6
Tail flick
0.57
11
Tail rise
1.47
13
Trumpet blow
2.71
29
Tail slash
0.52
11
Under boat
0.09
2
behavioral focal samples. A significant decrease in
the ratio of surface interval to dive time followed
the biopsy procedure. Although not statistically sig-
nificant, increases in trumpet blows and, to a lesser
extent tail slashes and sounding dives, were noted
following biopsy strikes, as were decreases in the
amount of slow swimming and some nonessential
behaviors.
Two of the behavior types that were noted to in-
crease, trumpet blows and tail slashes, have been
previously suggested to be agonistic (Baker and Her-
man 1984, Watkins and Wartzok 1985). A tail slash
may be used by a humpback whale as a means of
aggression against another whale in what has been
interpreted as courtship battles (Baker and Herman
1984). Norris and Reeves (1977) identify "tail
swishing" (our "tail slashing") as one of the more com-
mon behavioral responses to harassment.
The behaviors elicited by the biopsy procedure in
most cases are not intrinsically different from those
behaviors which occur naturally in this species. Thus
we emphasize that it may be the change in frequency
of behaviors that should be viewed as indicative of
"affected" behavior, rather than the occurrence of such
displays per se. The one notable exception is the hard
tail flick, which rarely has been observed other than
in response to the biopsy procedure.
The possibility exists that the hard tail flick reaction
we observed is a reflex response. This reaction typically
occurred at the instant of dart impact and thereafter
was rarely repeated. Moreover, in some individuals a
single hard tail flick at the time of the biopsy was
followed by a period during which no other behavioral
change was observed. A reflex response is consistent
with our finding of no correlation of respiratory varia-
tion with the occurrence of this reaction.
While some of the hard tail flicks may have been
purely reflexive, the same behavior was seen once in
response to an extremely close vessel approach when
no physical contact was made. Further, a similar reac-
tion was reported by Watkins (1981), who labeled it a
"startle response." Hence it is uncertain whether this
behavior is reflexive or cognitive. It may have both
components.
In other studies of whale disturbance in response to
noxious stimuli, both Watkins (1981) and Mathews
(1986) mention the approach of the vessel as con-
tributing to the reaction of the animal. We made every
effort to diminish vessel effects. Both previous studies
were conducted from power-driven vessels approaching
at moderate to rapid speeds. In over half of our paired
samples, data were collected from the relatively silent
approach of a sailboat. Those approaches made under
power in paired samples were done at slow speed. Fur-
ther, we limited movement of the research vessels near
whales, except in the brief approach for the biopsy, to
lessen effects of vessels. While the effect of the vessels
was minimized, this approach is a necessary part of the
biopsy procedure and need not be considered separately
in an analysis of responses.
Our results are comparable with those found by
Mathews (1986), who examined the response of eight
gray whales to a similar biopsy procedure. Both studies
established great variability in the reaction of whales
to biopsy procedures. One clear difference is that the
blow interval of gray whales showed a significant in-
crease in the postbiopsy period, while that of the hump-
back whales we studied did not. Even so, four of the
eight gray whales studied by Mathews (1986) showed
a reduction in their surface-interval/dive-time ratios,
as did 57.0% of the larger sample; only one gray
whale showed an increased surface-interval/dive-time
ratio.
Weinrich et al . Behavior of Megaptera novaeangliae during biopsy
597
There have been other studies of the response of
humpback whales to human-induced stimuli. In Alaska,
17 humpback whales exhibiting "affected" behavior
associated with the proximity of vessels increased their
mean and maximum dive intervals, while their mean
blow interval decreased (Baker and Herman 1982). In
comparison, although the whales in our study did not
consistently increase the length of their dives follow-
ing the biopsy, blow intervals decreased slightly. In
both studies, whales decreased surface-interval/dive-
time ratios on average. The whales in our study and
in that of Baker and Herman (1982) also responded
with an increased rate of net movement.
Our results generally agree with other studies of the
reactions of baleen whales to a variety of human-
induced stimuli. Richardson et al. (1985) found that
bowhead whales Balaena mysticetus respond to a vari-
ety of man-made stimuli (drillships, vessels, aircraft)
by reducing their surface-interval/dive-time ratios.
Swimming speeds increased in response to vessel traf-
fic. Migrating gray whales, by comparison, have been
reported to slow down as their migration route took
them toward simulated offshore industrial activity
(Malme et al. 1983, 1985). Bauer and Herman (1985)
found that humpback whales on Hawaiian breeding
grounds reduced their surface interval as vessels ap-
proached closely. The blow interval decreased as either
the proximity or the number of vessels increased.
Similarly, pod speed increased as vessels approached.
Hence, the net effect in all these studies was the same
as we have found; namely, that the animal avoids the
source of the stimulus.
It is important to note that the reactions we describe
in most cases were elicited by a noxious stimulus of
brief duration and low-to-moderate amplitude. On this
basis, our findings likely underestimate the effects of
a more prolonged noxious stimulus, or one of greater
force. For example, extreme responses, including
escape, hard tail flicks, and immediate submergence,
has been documented in harpooned right whales Euba-
laena glacialis (Scammon 1874) and fin whales Balae-
noptera physalus (Lambertsen and Moore 1983).
In the context of current management problems, the
response of a whale to a prolonged sublethal noxious
stimulus is a critical issue, as habitat intrusion may
establish conditions of continuing, if not constant,
exposure to diverse noxious stimuli. Recognizing this,
Bauer and Herman (1985) considered the relationship
of stimulus amplitude and duration (expressed as the
number of whale-watching vessels and the length of
time a whale group was in close proximity to whale-
watching vessels) to elicited responses in their study
of the effects of vessel traffic on humpback whales.
In both cases, their data indicate a graded response
in strenuous episodes of breaching, lobtailing, and
flippering behavior and in movement away from the
path of vessels.
Although our present study was not designed to
evaluate the effects of increasing stimulus duration,
including that approximated by stimulus repetition, the
special case of SI4 is illuminating. Its progressively
increasing reaction to repeated biopsy strikes was
dramatic. After the first strike, the whale seemed
unperturbed. After the second, it appeared, from its
trumpet blowing and stationary position, to be annoyed
but passive. After the third, it reacted with great in-
tensity and subsequently appeared exhausted.
Based on these observations we conclude that
adverse responses to rapidly repeated or prolonged
noxious stimuli in whales may be incorrectly modeled
as a linear function. Given the lack of any detectable
response to the biopsy procedure in some animals,
there seems to be a threshold for stressor amplitude
below which no response will occur. Further, this
threshold of tolerance may be dependent upon the
specific activity in which the animal is engaged imme-
diately prior to the time the stressor is applied; e.g.,
in our study animals engaged in feeding were unlikely
to react to the strike of a biopsy dart. There likely are
also individual differences in this threshold, as sug-
gested by the wide variation in reactions observed.
Moreover, although one evidently can expect a
graded response in the disturbance of the animal above
its tolerance threshold, such gradation might be better
modeled as an exponentially increasing stimulus-
response function. As such, continuous or rapidly
repeated moderate-level noxious stimulation could
potentially lead to a general somatic alarm reaction,
with endocrinologic consequences (Selye 1936, 1946).
Thus, one of the important implications of this study
for current management strategies to promote the
recovery of endangered whale populations is that un-
controlled increases in the level or frequency of nox-
ious intrusion into cetacean habitat may, suddenly and
unexpectedly, have serious deleterious effects.
Acknowledgments
We are grateful to M. Gassel and P. Raid of the Ceta-
cean Research Unit, and S. Frohock of the Atlantic
Cetacean Research Center, who helped collect the data
presented in this study. S. Sears, T. Leland, C. Gun-
son, S. Larkin, Dr. R. Schaper, and Dr. D. Senior
provided great assistance at sea and in logistical
arrangements. D. Beach, T. McKenzie, and two anon-
ymous reviewers provided helpful comments on an
early draft of the manuscript. Funding for the study
was provided by the National Marine Fisheries Service
(PO 40EANF-501-0396) and Contract 50EANF-00094,
598
Fishery Bulletin 90(3). 1992
with additional support to R.H.L. in the form of a
fellowship from the Committee on the Challenges of
Modern Society, Division of Science and Environment,
North Atlantic Treaty Organization. The study by R.H.
Lambertsen and M.J. Moore referred to in the discus-
sion was undertaken with the cooperation of commer-
cial whalers at the request of the Humane Killing Sub-
committee and the Secretariat of the International
Whaling Commission (IWC). Biopsy sampling was con-
ducted under Marine Mammal Research Permit 393
issued by the National Marine Fisheries Service, Na-
tional Oceanic and Atmospheric Administration, U.S.
Department of Commerce.
Citations
Altmann, J.
1974 Observational study of behavior: Sampling methods.
Behaviour 49(3.4):227-267.
Baker, C.S.. and L.M. Herman
1982 The impact of vessel traffic on the behavior of humpback
whales in southeast Alaska. Unpubl. rep. to NMFS Natl. Mar.
Mammal Lab., Contract 81-ABC-00114, Seattle, 39 p.
1984 Aggressive behavior between humpback whales winter-
ing in Hawaiian waters. Can. J. Zool. 62:1922-1937.
Bauer, G.B., and L.M. Herman
1985 Effects of vessel traffic on the behavior of humpback
whales in Hawaii. Unpubl. rep. to NMFS Honolulu Lab.,
163 p.
Clapham, P.J., and C.A. Mayo
1987 Reproduction and recruitment of individually identified
humpback whales, Megaptera novaeangliae. observed in Massa-
chusetts Bay, 1979-1985. Can. J. Zool. 65:2853-2863.
Day, B.
1983 Operator's guide for the SITEX-757 LORAN-C receiver.
SITEX Co., Baltimore. 98 p.
Johnson, J.H., and A. A. Wolman
1985 The humpback whale. Mar. Fish. Rev. 46(4):30-37.
Katona, S.K., and H.P. Whitehead
1981 Identifying humpback whales using their natural mark-
ings. Polar Rec. 20:439-444.
Lambertsen, R.H.
1987 A biopsy system for large whales and its use for cyto-
genetics. J. Mammal 68:443-445.
Lambertsen, R.H., and D.A. Duffield
1987 Biopsy studies of the humpback whale, Megaptera novae-
angliae. NTIS Rep., Springfield, VA, 151 p.
Lambertsen, R.H., and M.J. Moore
1983 Behavioral and post mortem observations on fin whales
killed with explosive harpoons with preliminary conclusions
concerning killing efficiency. Rep. IWC/TC/36/HK3 to Int.
Whaling Comm. Dep. Physiol. Sci., Coll. Vet. Med., Univ.
Florida, Gainesville, 23 p.
Lambertsen, R.H., C.S. Baker, D.A. Duffield. and
J. Chamberlin-Lea
1988 Cytogenetic determination of sex among individually
identified humpback whales. Can. J. Zool. 66:1243-1248.
Malme, C.L. P.R. Miles, C.W. Clark, P. Tyack, and J.E. Bird
1983 Investigations of the potential effects of underwater noise
from petroleum industry activities on migrating gray whale
behavior. Unpubl. rep. 5366 by Bolt, Beranek, & Newman.
Inc., Cambridge, MA 02238, to Minerals Manage. Serv., 183 p.
1 985 Investigations of the potential effects of underwater noise
from petroleum industry activities on feeding humpback whale
behavior. Unpubl. rep. 5851 by Bolt, Beranek, & Newman,
Inc.. Cambridge, MA 02238, to Minerals Manage. Serv., 128 p.
Mathews, E.A.
1986 Multiple uses of skin biopsies collected from free-ranging
gray whales: Sex chromatin analysis, collection and process-
ing for cell culture, microbiological analysis of associated
organisms, behavioral responses of whales to biopsying, and
future prospects for using biopsies in genetic and biochemical
studies. M.S. thesis, Univ. Calif.. Santa Cruz, 118 p.
Mayo, C, C. Carlson, P. Clapham, and D. Mattila
1985 Humpback whales of the Southern Gulf of Maine. Prov-
incetown Cent. Coastal Stud., Provincetown, MA. 62 p.
Norris, K.S., and R.R. Reeves
1977 Report on a workshop related to humpback whales in
Hawaii. NTIS Rep. PB-280 794, Springfield. VA, 90 p.
Richardson, W.J., M.A. Fraker, B. Wursig. and R.S. Wells
1985 Behavior of bowhead whales summering in the Beaufort
Sea: Reactions to industrial activities. Biol. Conserv. 32:
195-230.
Scammon, C.A.
1874 The marine mammals of the northwest coast of North
America, John H. Carmany, San Francisco, 319 p.
Selye, H.
1936 A syndrome induced by diverse nocuous agents. Nature
(Lond.) 62:32-33.
1946 The general adaptation syndrome and the diseases of
adaptation. J. Clin. Endocrinol. 6(2):117-240.
Watkins. W.A.
1981 Reaction of three species of baleen whales Balaenoptera
physalus, Megaptera novaeangliae, and Balaenoptera edeni to
implanted radio tags. Deep-Sea Res. 28(A):589-599.
Watkins. W.A., and D. Wartzok
1985 Sensory biophysics of marine mammals. Mar. Mammal
Sci. 1:219-260.
Weinrich, M.T. (editor)
1985 Humpback whales of the Southern Gulf of Maine: Re-
cent findings on habitat use, social behavior, and feeding pat-
terns. Unpubl. spec. rep. 2, Cetacean Res. Unit, Gloucester,
42 p.
Weinrich. M.T.
1986 Observations: The humpback whales of Stellwagen Bank,
2d ed. Whale Res. Press, Gloucester, 162 p.
Zar, J.H.
1984 Biostatistical analysis. Prentice-Hall, Englewood Cliffs,
718 p.
Abstract.— Commercial landings
data and research-vessel survey data
collected by the Northeast Fisheries
Science Center during 1982-86 were
analyzed to identify spatial and tem-
poral patterns as well as possible
mechanisms associated with juve-
nile cod Gadus morhua distribution.
Analysis of survey data indicated
that cod ages 1-2, age 3, and age 4 -i-
were distributed at different depths
during the spring; however, during
the autumn, age-3 fish co-occurred
with age 1-2 fish.
Analysis of commercial landings
data revealed the following patterns
of distribution for age-2 cod: In
quarter 1, concentrations appeared
in the Nantucket Shoals region and
the central portion of Georges Bank;
in quarter 2, the concentration was
northeast of Nantucket Shoals and
also remained on Georges Bank; in
quarter 3, both aggregations moved
northeastward into deeper waters,
along the 100 m contour of the Great
South Channel and the Northern
Edge, respectively; and in quarter 4,
the Nantucket Shoals concentration
had moved southwestward to shal-
lower water, resuming locations iden-
tified in quarter 1 , while the Georges
Bank concentration remained as in
quarter 3.
While intraseasonal spatial distri-
butions did not appear to be defined
by temperature, seasonal shifts in
concentration of juvenile cod were
most likely associated with temper-
ature.
Spatial and temporal distribution of
juvenile Atlantic cod Gadus morhua
in the Georges Bank-Southern
l\le\A/ England region
Susan E. Wigley
Fredric M. Serchuk
U/oods Hole Laboratory, Northeast Fisheries Science Center
National Marine Fisheries Service. NOAA
166 Water Street. Woods Hole. Massachusetts 02543-1097
Manuscript accepted 20 May 1992.
Fishery Bulletin, U.S. 90:599-606 (1992).
The Atlantic cod Gadus rnoriiua has
accotinted for more catch, by weight,
than any other species in the U.S.
Atlantic coast groundfish fishery dur-
ing the past two decades (Serchuk
and Wigley In press). Recent declines
in annual landings of cod from the
Georges Bank-Southern New Eng-
land region (N. Atl. Fish. Org. Div.
5Z) have generated concern for the
fishery. Total nominal catches (U.S.
and Canadian commercial landings,
plus U.S. recreational catch) from
this area dropped from a high of
64,000 metric tons (t) live weight in
1982, to 27,900 t in 1986. Although
catches increased to 33,700 1 in 1987,
64% of the catch in numbers and 36%
in weight consisted of age-2 fish from
the strong 1985 year-class (NEFSC
1988). Of the various commercial
market categories of cod, 'scrod'
generally represents the smallest size
grouping of cod landed. Scrod land-
ings paralleled the general decline of
Georges Bank cod landings, decreas-
ing from 8100 t in 1982 to 3400 t in
1986. In addition, Northeast Fish-
eries Science Center (NEFSC) re-
search-vessel survey abundance in-
dices for spring and autumn 1987
were among the lowest observed for
cod in the 25-year survey time-series
(Serchuk and Wigley In press).
During the period December 1986-
March 1987, anomalously high dis-
card rates of juvenile cod below the
legal minimum landed size of 19 inches
(48.3cm) TL were associated with
commercial trawling operations using
small mesh in the Nantucket Shoals
area. This high discard level led to an
emergency action extending large
mesh regulation (5.5-inch mesh in
codend) to this region during 23 Feb-
ruary-31 March 1988 to "reduce
fishing effort and mortality on juve-
nile Atlantic cod stocks found in high
concentrations in this area at this
time" (Federal Register, 50 CFR
Part 651, 26 Feb. 1988). Large mesh
regulation was permanently extended
to the Nantucket Shoals area in Jan-
uary 1989 (Federal Register, 50 CFR
Part 651, 31 Jan. 1989) (Fig. 1). In
addition, the "Nantucket Shoals
scrod slaughter" may have prompted
the development and implementation
of the Flexible Area Action System
(FAAS) by the New England Fishery
Management Council (NEFMC) and
the National Marine Fisheries Ser-
vice (NMFS). Under this plan, the
Regional Director, NMFS Northeast
Region, cotild close an area to fishing,
impose mesh size restrictions, or es-
tablish catch limits for a period of 3
weeks to 6 months to minimize dis-
cards of juvenile fish. This represents
a significant departure from the tra-
ditional uses of seasonal or areal clos-
ures, such as protection of adult had-
dock Melanogrammus aeglefinics dur-
ing spawning (Halliday 1987).
For such a plan to be effective,
knowledge of fish distribution pat-
599
600
Fishery Bulletin 90(3), 1992
.iL
BEGULATEO MESH ABEA
SEASONAL REGULATED MESH ABEA
Figure 1
Georges Bank-Southern New England region of the North-
west Atlantic Ocean, showing geographic features and reg-
ulated mesh (5.5-inch) areas under the Northeast Multispecies
Fishery Management Plan of the New England Fishery
Management Council.
terns is necessary. In addition to documenting geo-
graphic, seasonal, and age-specific aspects of distri-
bution, studies must examine mechanisms (such as
temperature, depth, spawning, or feeding behavior)
underlying these observed patterns. Numerous tag-
ging studies have been conducted for Atlantic cod in
the Northwest Atlantic: McKenzie (1934, 1956) and
McCracken (1956) described cod movements in Cana-
dian waters based on tagging experiments, and Smith
(1902), Schroeder (1928, 1930), and Wise (1958, 1962)
tagged cod from Woods Hole, Nantucket Shoals, and
Nova Scotia to New Jersey, respectively. More recent-
ly, cod distributions delineated by bottom-trawl survey
data from research vessels have been presented by
Scott (1988) for Canadian waters and Grosslein and
Azarovitz (1982) and Almeida et al. (1984) for U.S.
waters. None of these studies considered fish size in
the analyses. Overholtz (1984) found age-specific pat-
terns of distribution for another gadoid species, had-
dock, in the Georges Bank region. Wigley and Gabriel
(1991) and Bowman et al. (1987) examined distributions
of several juvenile fishes, including cod, using NEFSC
bottom-trawl survey data collected from Cape Hat-
teras, NC to Nova Scotia, Canada.
In our study, NEFSC commercial landings data and
research-vessel survey data collected during 1982-86
were analyzed to identify spatial and temporal patterns
as well as possible mechanisms associated with distribu-
tions of juvenile cod. The study period corresponds
roughly to the duration of the NEFMC's Interim Fish-
ery Management Plan for Atlantic Groundfish (31
March 1982 to 18 September 1986; NEFSC 1987). Dur-
ing this period, fishing practices were reasonably un-
changed, although (1) an increase in minimum mesh
size for the Georges Bank region from 5.125 inches
(130 mm) to 5.5 inches (140 mm) was implemented on
1 April 1983, and (2) the International Court of Justice
(ICJ) line dividing Georges Bank into U.S. and Cana-
dian portions (Fig. 1) was established in October 1984.
The study period also encompasses years of both strong
and weak recruitment as well as a 50% reduction in
spawning-stock biomass, events that allow evaluation
of resulting distributions over varying year-class
strengths and stock sizes.
Materials and methods
Distribution by temperature and depth
Temperature and depth analyses were based upon
data collected for Atlantic cod during NEFSC spring
and autumn bottom-trawl surveys during 1982-86 in
the Georges Bank- Southern New England region in
depths of 9-366 m. The stratified-random survey design
and the standardization of survey gear and methodol-
ogy are described in detail by Grosslein (1969) and
Azarovitz (1981). Water-column temperature profiles,
including bottom temperature (recorded to 0.1 °C),
were obtained on approximately half the survey sta-
tions via expendable bathythermographs; depths (m)
were recorded using research-vessel electronic depth-
sounding equipment. All cod were measured (FL to
nearest cm) and a subset sampled for age, growth, and
maturity information. Otoliths were processed and age
determinations obtained according to procedures
described by Penttila (1988); maturity staging was
performed using classification criteria outlined by
Burnett et al. (1989).
Estimates of mean temperature and depth (weighted
by number of fish in each tow), and associated stan-
dard errors and ranges, were calculated by age for each
season and tested for age-specific effects. Based upon
results from analyses of individual age-groups, ages 1
and 2 were combined as well as fish age 4 or greater.
These age-groups, as well as age-3 fish, were then re-
tested for age-group specific effects. Assumptions of
data normality were complicated by two factors. The
first is the inherent nature of survey catch data as
described by Pennington and Grosslein (1978), who
found that the two most likely models for the distribu-
tion of fish (i.e., heterogeneous Poisson and randomly-
Wigley and Serchuk: Spatial and temporal distribution of juvenile Gadus morhua
601
distributed clumps) both generated a negative binomial
distribution. The second factor is that cod are not
fully recruited to the survey sampling gear until age
3 or 4 (Serchuk and Wigley 1986). For these reasons,
a distribution-free analysis of variance (Kruskal and
Wallis 1952) was employed for statistical compari-
sons of temperature and depth distributions by age
groups.
Spatial and temporal distribution
Commercial landings data (see Burns et al. [1983] for
a detailed explanation of the commercial catch sam-
pling program in the northeastern United States) for
scrod cod collected by NEFSC during 1982-86 from
the Georges Bank-Southern New England region
(NEFSC Statistical Areas 521-526 and 537-539) were
examined for spatial and temporal aspects of juvenile
cod occurrences. Biases associated with the use of
landings data were assumed to be negligible for this
highly-directed fishery. Otter trawl catches account for
86-90% of the annual total cod landings (Serchuk and
Wigley 1986); hence other gear types were excluded
from subsequent analysis.
Table I
Mean lengths-at-age
, samp
e sizes, standard errors (SE), and |
range of observed values
for age 0-
10 cod Gadus morhua
collected during NEFSC spring and autumn bottom-trawl |
surveys,
1982-86.
in the
Georges
Bank-Southern New |
England
region.
Season
Age
N
Length (
=m)
X
SE
Range
Spring
1
106
24.2
0.565
13.0-46.0
2
458
43.8
0.253
26.0-58.0
3
349
59.5
0.398
34.0-75.0
4
212
69.0
0.570
46.0-89.0
5
170
77.1
0.666
51.0-96.0
6
69
84.7
1.013
64.0-108.0
7
41
93.2
1.338
76.0-110.0
8
24
102.5
2.235
79.0-124.0
9
15
104.9
2.439
92.0-127.0
10
5
102.8
5.678
87.0-119.0
Autumn
0
99
11.7
0.400
6.0-25.0
1
279
36.9
0.308
23.0-51.0
2
306
53.7
0.313
32.0-71.0
3
145
68.0
0.557
42.0-83.0
4
37
74.3
1.373
49.0-89.0
5
16
85.4
2.432
60.0-99.0
6
7
90.3
3.227
80.0-102.0
7
6
98.8
2.613
94.0-110.0
8
4
96.3
4.608
86.0-105.0
9
3
105.0
5.196
96.0-114.0
10
1
118.0
—
—
Fish in the scrod market category weigh 0.7-1.4kg,
measure 40-60cm TL, and are 1-3 yr of age according
to the growth function developed by Penttila and Gif-
ford (1976).
Geographic resolution within Statistical Areas (SAR)
was obtained by assigning the landings within each
SAR to 10-min squares of latitude and longitude (~100
nm-) based upon information from interviewed (i.e.,
dockside interviews of captains in which precise catch-
location information was obtained) trips landing scrod
cod. Scrod landings associated with interviewed trips
were prorated upward for each 10-min square by the
ratio, derived for each SAR, of total monthly landings
to monthly interviewed landings. Prorated scrod land-
ings were summarized quarterly for each year for each
10-min square.
Scrod cod landings were partitioned into age-groups
1, 2, 3 and 4-i- by constructing a catch-at-age matrix
using a technique described by Serchuk and Wigley
(1986). In this method, quarterly mean weights for
scrod cod were calculated by applying a length-weight
equation to quarterly scrod length-frequency data;
these means, in turn, were divided into scrod landings
for the corresponding quarter to generate numbers
landed. Age compositions for scrod landings were
derived by applying quarterly age/length keys to
numbers at length landed; resulting proportions at age
were then applied to prorated scrod landings for each
10-min square to obtain estimates of landings by weight
for each age-group.
Results
Temperature and depth distribution
Analyses of survey data were based upon 1455 and 904
cod collected during spring and autumn bottom-trawl
surveys, respectively, during 1982-86. Mean lengths-
at-age and associated statistics for cod ages 0-10 are
presented in Table 1. Geographically, juvenile cod (de-
fined as fish <37cm, the minimum size at first matur-
ity; Morse 1979) exhibited seasonal patterns of distribu-
tion. In spring, juveniles are dispersed throughout the
Georges Bank-Southern New England region (Fig.
2A), while in autumn they are concentrated along the
100 m contour west of the Great South Channel and
the Northern Edge and Northeast Peak of Georges
Bank (Fig. 2B).
Differences in distribution with respect to both mean
temperature and depth were noted for all age-groups
of cod (Table 2). Mean temperatures were approx-
imately 5.3°C in spring and 9.2°C in autumn for all
age-groups, despite the fact that there was consider-
able overlap in the temperatures for the two seasons
(Table 2). However, within seasons, differences in mean
602
Fishery Bulletin 90(3). 1992
temperature observed between age-groups were within
0.5°C, except for age-0 cod in the autumn (Table 2).
In general, distribution patterns were delineated more
by differences in depth than temperature.
No age-0 cod were captured during spring surveys
due to their pelagic larval existence at this time (early
April-early May). Age-1 and age-2 cod were found at
-V^f
NUMBER OF FISH
Figure 2
Geographic distribution (number/tow) of juvenile cod Gains
morhua <37cm in length collected during NEFSC spring(A)
and autumn (B) bottom-trawl surveys, 1968-86, in the Georges
Bank-Southern New England region.
significantly shallower mean depths than age-3 and age
4+ cod (57.0 and 58.0m vs. 68.4 and 86.3 m, respec-
tively; Table 2). The difference between mean depths
of age-3 and age 4 + cod was also significant (Kruskal-
Wallis ANOVA, p<0.01; Table 3). In the autumn sur-
veys (corresponding to late September-late October),
age-3 cod were observed at a mean depth similar to
those for ages 1-2 (85.8 m vs. 86.8 and 85.2 m, respec-
tively; Table 2) and significantly different (Kruskal-
Wallis ANOVA, p<0.01; Table 3) from that for age 4 -i-
cod (116.0m; Table 2). Additionally, age-0 cod were cap-
tured in the autumn and observed to be distinct from
all other age-groups with respect to both temperature
and depth (Table 2).
Spatial and temporal distribution
Statistical Area 521 accounted for 49% of total pro-
rated scrod landings during the sampling period
1982-86, from a high of 57% in 1983 to a low of 29%
in 1986, while SAR 523 and 522 contributed 14% and
13%, respectively; landings from Southern New Eng-
land (SAR 537-539) accounted for only about 2% of
total scrod landings (Table 4). Interviewed coverage
was obtained for a high percentage of scrod landings
Table 2
Mean temperatures
and depths, sample sizes, standard errors 1
(SE), and
range of observed values for
age-groups
0,1,2,3.
and 4 + of cod Gadus morhua collected during NEFSC spring
and autumn bottom-trawl surveys, 1982-86, in the Georges
Bank-Southern New England region.
Season
Age
N X
SE
Range
Spring
1
Temperature
(°C)
106 5.18
0.124
2.8-9.0
2
458 5.03
0.050
3.3-12.1
3
349 5.51
0.064
2.8-9.4
4 +
542 5.42
0.051
3.0-12.1
Autumn
0
99 10.02
0.233
5.8-14.3
1
279 8.75
0.123
5.4-14.3
2
306 9.02
0.108
5.4-14.3
3
145 9.33
0.195
5.2-19.2
4 +
75 9.33
0.292
5.1-15.3
Spring
1
Depth (m
106 56.98
2.001
25-122
2
458 57.97
1.307
24-237
3
349 68.36
1.966
24-210
4-H
542 86.34
2.346
25-307
Autumn
0
99 68.77
2.934
28-153
1
279 86.80
2.185
28-230
2
306 85.19
2.618
28-230
3
145 85.80
3.945
31-205
4-1-
75 116.00
7.793
26-328
Wigley and Serchuk Spatial and temporal distribution of juvenile Gadus morhua
603
Table 3
Results of Kruskal-Wallis analyses of variance
of age-specific cod Gadus morhua distribution
by temperature and depth by season, based on
data for cod collected during NEFSC bottom-
trawl surveys, 1982-86, in the Georges Bank-
Southern New England region. NS = not sig-
nificant, p>0.05; **highly significant, p<0.01.
Season
Age-group Temp. Depth
Spring
Autumn
3 vs. 4 -H
NS
1-2 vs. 3
**
1-2 vs. i +
* *
3 vs. 4-H
NS
1-2 vs. 3
NS
1-2 vs. 4-1-
NS
NS
Table 4
Scrod cod Gadus morhua landings (t, live weight) by Statistical Area and |
year, 1982-86, in the
Georges
Bank-Southern
New England region. |
Percentages of NEFSC commercial scrod landing
s for which in
ter\'iewed
coverage
> was obtained are given in
parentheses. (Statistical Areas
are shown
in Figs.
3A-3D.)
Area
1982
1983
1984
1985
1986
521
4564(76)
4350(76)
1396(81)
3033(93)
1022(87)
522
1025(71)
978(73)
415(75)
855(93)
476(84)
523
1215(83)
958(95)
766(91)
939(87)
365(96)
524
554(68)
411(92)
517(92)
568(92)
757(84)
525
240(66)
109(80)
150(77)
204(91)
118(83)
526
466(92)
665(76)
413(90)
743(95)
579(84)
537
76(68)
135(73)
103(76)
60(74)
168(50)
538
4(30)
7(59)
3(61)
5(20)
5(60)
539
10(28)
3(22)
1(30)
<1(27)
<1(14)
Total
8154(76)
7616(79)
3764(84)
6407(92)
3490(84)
(76-92% annually during the period; Table
4), suggesting that the proration procedure
employed in this study accurately depicts
the patterns of landings.
Age-2 cod dominated commercial landings
of scrod in all years except 1984 and 1986,
when age-3 fish comprised the majority
(Table 5); this exception is due to weak 1982
and 1984 year-classes (Serchuk and Wigley
In press). Age-1 fish are too small to be
caught by the commercial gear until
quarters 3 and 4; conversely, age 4-i- fish
grow out of the scrod market category and
into the next market category after quarter
2 (Table 5). Based upon these observations,
and the observation above from analysis of
survey data that age-3 cod are seasonally
segregated from ages 1-2, analysis of
juvenile cod distribution from commercial
data was confined to age-2 fish.
The following patterns of age-2 juvenile
cod distribution emerged. In quarter 1,
juvenile cod were concentrated in the Nan-
tucket Shoals region (south of the stepped
portion of the boundary between SAR 521
and 526) as well as being dispersed generally
across the shallower central portions of
Georges Bank, primarily in SAR 522 and
524 (Fig. 3A). By quarter 2, the area con-
centration was north of Nantucket Shoals
(in SAR 521), while on Georges Bank there
continued to be a dispersed distribution as
in quarter 1 (Fig. 3B). In quarter 3, the Nan-
tucket Shoals concentration had moved northeastward
within SAR 521 to deeper water along the 100 m con-
tour of the west slope of the Great South Channel;
similarly, juveniles on Georges Bank had formed con-
Table 5
Age ct
)mposition by weight (t, live
weight) of scrod cod Gadus
morhua
landings by quarter, 1982-86, in the
Georges
Bank-Southern New England |
region
Quarter
Scrod
total
Year
Age-group 1
2
3
4
1982
1 0
0
4
230
234
2 417
1140
3141
1836
6534
3 190
469
203
31
893
4-H 198
240
56
0
494
Total
805
1849
3404
2097
81.55
1983
1 0
0
16
61
77
2 199
976
1554
1260
3989
3 437
1170
965
502
3074
4+ 212
156
93
15
476
Total
848
2302
2628
1838
7616
1984
1 0
0
1
78
79
2 73
381
395
418
1266
3 611
672
395
102
1780
4-H 236
182
185
33
636
Total
920
1235
976
631
3762
1985
1 0
0
10
101
111
2 304
1002
1813
1826
4945
3 192
510
335
123
1160
4-H 96
46
43
6
191
Total
592
1558
2201
2056
6407
1986
1 0
0
0
120
120
2 162
467
193
249
1071
3 948
850
289
52
2139
4-H 148
4
8
0
160
Total
12.58
1321
490
421
3490
centrations along the 100 m contour in SAR 522 (the
east slope of the Great South Channel) and 523 (the
Northern Edge and Northeast Peak areas; Fig. 3C).
By quarter 4, the Nantucket Shoals concentration had
604
Fishery Bulletin 90(3). 1992
MEIRIC
IONS
-HV£
1 -
5
•
6 -
25
•
26 -
50
Figure 3
Distribution of age-2 cod Gadus morhua by 10-niin squares of latitude and longitude from NEFSC scrod landings data (t, live
weight), 1982-86, for quarters 1 (A), 2 (B). 3 (C), and 4 (D) in the Georges Bank-Southern New England region.
shifted south and southwestward within SAR 521 to
shallower water and into SAR 526, resuming loca-
tions identified in quarter 1 ; however, the concentra-
tion on Georges Bank in SAR 523 was still present (Fig.
3D).
Visual inspection of quarterly distribution plots for
each year suggested that these patterns of juvenile
cod concentrations persisted over varying year-class
strengths and stock sizes; however, in the interest of
space, only aggregate plots are presented (Figs. 3A-
3D). Age-2 cod from strong 1980 and 1983 year-classes
(Serchuk and Wigley In press) exhibited the same
seasonal movements described above, as did age-2 fish
from the relatively weak 1982 and 1984 year-classes.
Similarly, no changes in observed annual patterns were
evident from 1982 to 1986, during which time spawn-
ing-stock biomass diminished from over 80,000 t to
about 33,000 t (Serchuk and Wigley In press).
Wigley and Serchuk: Spatial and temporal distribution of juvenile Gadus morhua
605
Discussion
The observed seasonal variation in distribution of age-3
cod relative to age groups 1-2 and 4 + may be associ-
ated with a transitional period involving both matura-
tion and feeding habits. Age 3 encompasses a period
in which there is a mismatch in size-at-first-maturity
and the attainment of the adult diet. Median size and
age at sexual maturity for cod is about 50cm and 2.5
yr, respectively (Livingstone and Dery 1976), and dur-
ing spring some age-3 fish recruit to the spawning
population. Autumnal co-occurrence of age-3 fish with
ages 1-2 may be related to diet. Bowman and Michaels
(1984) presented data which indicate that cod < 66-70
cm have not assumed the adult diet dominated by fish;
the mean length of age-3 cod in autumn is 68.0cm
(Table 1).
In a mathematical evaluation of spatial distributions
of several North Sea species, Houghton (1987) sug-
gested that cod distributions were more complex and
less persistent than those observed for haddock or flat-
fish. In this study, however, the spatial and temporal
patterns observed for juvenile Atlantic cod from land-
ings data were remarkably uniform over the study
period, and did not seem to vary according to stock size
or year-class strength. The use of commercial data in
this study is somewhat constrained by management
regulations, fishing practices, and the distribution of
fishing effort, yet results from analysis of survey data
in this study seem to corroborate these conclusions.
The use of mean values in this study to define pat-
terns of temperature and depth distribution may better
reflect general tendencies rather than absolute pref-
erences; in actuality, cod of all ages except age 0 were
found at virtually all available temperatures and
depths. Yet the patterns that emerged in this study are,
for the most part, consistent with those identified in
other studies. Both Schroeder (1930) and Wise (1962)
noted the tendency for older cod to move into greater
depths. Scott (1988) found that cod in colder Canadian
waters were distributed at temperatures of 2-10° C,
with largest catches occurring at 4-6°C. The apparent
contradiction posed by the movement of cod to deeper,
warmer water in winter-spring observed by Scott
(1988), and the observations in this study of movement
to shallower water on Nantucket Shoals and Georges
Bank during this period, is an artifact of the different
temperature regimes for the two regions; in each case,
cod are changing depth locations to maintain preferred
temperatures. Schroeder (1930) reported cod occur-
rences within an annual range of 0-17 °C in the region
from Nantucket Shoals to North Carolina, and attrib-
uted the triggering of the autumn migration of Nan-
tucket Shoals adult cod westward to New Jersey for
winter spawning to falling bottom temperatures in Oc-
tober. Similarly, movements of juvenile cod from Nan-
tucket Shoals to deeper water off Chatham and the
Great South Channel in summer-early autumn were
thought to be in response to locally-available, cooler
temperatures (Schroeder 1930). Wise (1958, 1962)
determined from tagging studies that a resident pop-
ulation of cod inhabited the Nantucket Shoals-Great
South Channel area year-round, but that Nantucket
Shoals also represented the summer residence for the
population of cod that wintered off the coast of New
Jersey. Thus, the distribution patterns observed in this
study within SAR 521 and 526 would most likely reflect
seasonal movements of resident cod, although the
migratory population may partially contribute to land-
ings for quarters 2 and 3 in SAR 526.
Scott (1982) concluded from an analysis of fish dis-
tribution by bottom type that, although generally asso-
ciated with sand-gravel sediments, cod occurred over
all substrates and that observed patterns of distribu-
tion were more likely due to the bottom-type prefer-
ences of major prey items (e.g.. Cancer crabs, sand
lance, Ammodytes sp., etc). Although no quantitative
analysis of distribution by bottom sediment was under-
taken here, the seasonal shifts in concentration iden-
tified in this study do not suggest any major change
in substrate preference of cod. However, the Great
South Channel and the Northern Edge-Northeast Peak
regions, where concentrations of scrod cod occur in
quarters 3 and 4, are characterized by coarser sediment
types than those generally found elsewhere on Georges
Bank (Wigley 1961, Schlee 1973).
Based on the above analyses, there is evidence for
well-defined seasonal and geographic shifts in concen-
tration for juvenile Atlantic cod in the Georges Bank-
Southern New England region. Moreover, these pat-
terns of concentration appear to be associated pri-
marily with temperature. The high level of spatial and
temporal resolution possible, i.e., 10-min squares of
latitude and longitude and quarters, suggest that this
type of study may be useful in assisting fisheries
managers with decisions regarding seasonal and areal
closures under the Flexible Area Action System.
Acl<nowledgments
We express our thanks to S.L. Brunenmeister for her
assistance in the initial stage of this work, J. Burnett
and R.K. Mayo of the NEFSC and anonymous review-
ers for their helpful comments and review of this
manuscript, and N.G. Buxton of the NEFSC for
graphic assistance.
606
Fishery Bulletin 90(3). 1992
Citations
Almeida, P.P., T.R. Azarovitz, L. O'Brien, and E.M. Pritchard
1984 The distribution of major finfish and shellfish species col-
lected during NEFC bottom trawl surveys, 1965-1978. Ref
Doc. 84-21, NMFS Woods Hole Lab., 101 p.
Azarovitz, T.R.
1981 A brief historical review of the Woods Hole laboratory
trawl survey time series. In Doubleday, W.G., and D. Rivard
(eds.), Bottom trawl surveys, p. 62-67. Can. Spec. Publ. Fish.
Aquat. Sci. 58.
Bowman, R.E., and W.L. Michaels
1984 Food of seventeen species on northwest Atlantic fish.
NCAA Tech. Memo. NMFS-F/NEC-28. Northeast Fish. Sci.
Cent., Woods Hole, 183 p.
Bowman, R.E.. T.R. Azarovitz, E.S. Howard, and B.P. Hayden
1987 Food and distribution of juveniles of seventeen northwest
Atlantic fish species, 1973-1976. NCAA Tech. Memo. NMFS-
F/NEC-45, Northeast Fish. Sci. Cent., Woods Hole. 57 p.
Burnett, J.M., L. O'Brien, R. Mayo, J. Darde, and M. Bohan
1989 The history of maturity sampling and classification
schemes used during the NEFC bottom trawl survey program,
1963-1989. NCAA Tech. Memo. NMFS-F/NEC-76, Northeast
Fish. Sci. Cent., Woods Hole, 33 p.
Burns, T.S., R. Schultz. and B.E. Brown
1983 The commercial catch sampling program in the north-
eastern United States. In Doubleday, W.G., and D. Rivard
(eds.), Sampling commercial catches of marine fish and in-
vertebrates, p. 82-95. Can. Spec. Publ. Fish. Aquat. Sci. 66.
Grosslein, M.D.
1969 Groundfish survey program of BCF Woods Hole.
Commer. Fish. Rev. 31 (8-9);22-35.
Grosslein, M.D., and T.R. Azarovitz
1982 Fish distribution, MESA (Mar. Ecosyst. Anal.) N.Y.
Bight Atlas Monogr. 15, 182 p.
Halliday, R.G.
1987 Haddock spawning area closures in the northwest Atlan-
tic, 1970-87. NAFO (N. Atl. Fish. Org.) SCR Doc. 87/13, Ser.
N1291, 22 p.
Houghton, R.G.
1987 The consistency of the spatial distribution of young
gadoids with time. ICES CM 1987/D:15, 6 p.
Kruskal, W.H., and W.A. Wallis
1952 Use of ranks in one-criterion variance analysis. J. Am.
Stat. Assoc. 47:583-621.
Livingstone, R. Jr., and L. Dery
1976 An observation on the age and length at maturity of cod
on the Georges and Browns stock. ICNAF (Int. Comm. North-
west Atl. Fish.) Res. Doc. 76/VI/42, Ser. 3826, 2 p.
McCracken, P.P.
1956 Cod and Haddock tagging off Lockeport, Nova Scotia.
Fish. Res. Board Can., Prog. Rep. Atl. coast stns 64: 10-15.
McKenzie, R.A.
1934 Cod movements on the Canadian Atlantic coast. Con-
trib. Can. Biol. Fish., new ser. 8 (31) [Ser. A, General (46)]:
434-458.
1956 Atlantic cod tagging off the southern Canadian main-
land. Fish. Res. Board Can., Bull. 105, 93 p.
Morse, W.W.
1979 An analysis of maturity observations of 12 groundfish
species collected from Cape Hatteras, North Carolina to Nova
Scotia in 1977. Ref Doc. 79-32, NMFS Sandy Hook Lab., 20 p.
NEFSC
1987 Status of mixed species demersal finiish resources in New
England and scientific basis for management. Ref. Doc. 87-08,
NMFS Woods Hole Lab., 105 p.
1988 Status of the fishery resources off the northeastern
United States for 1988. NCAA Tech. Memo. NMFS-F/NEC-
50, Woods Hole, 132 p.
Overholtz, W.J.
1984 Seasonal and age specific distribution of the 1975 and
1978 year-classes of haddock on Georges Bank. NAFO (North-
west Atl. Fish. Organ.) Sci. Counc. Stud. 8:77-82.
Pennington, M.R., and M.D. Grosslein
1978 Accuracy of abundance indices based on stratified trawl
surveys. ICNAF (Int. Comm. Northwest Atl. Fish.) Res. Doc.
77, Ser. 5264, 42 p.
Penttila, J.
1988 Atlantic cod Gadus morhua. In Penttila, J., and L.M.
Dery (eds.). Age determination methods for Northwest Atlantic
species, p. 31-36. NOAA Tech. Rep. NMFS 72.
Penttila, J.A., and V.M. Gifford
1976 Growth and mortality rates of cod from the (Georges Bank
and Gulf of Maine areas. ICNAF (Int. Comm. Northwest Atl.
Fish.) Res. Bull. 12:29-36.
Schlee, J.S.
1973 Atlantic continental shelf and slope of the United
States— sediment texture of the northeastern part. U.S. Geol.
Surv. Prof. Pap. 529L, 64 p.
Schroeder, W.C.
1928 Cod studies. In Higgins, E. (ed). Progress in biological
inquiries 1926, including the proceeding of the divisional con-
ference, January 4-7, 1927. Rep. U.S. Comm. Fish., fiscal
year 1927. Doc. 1029, app. 7, p. 601-605.
1930 Migrations and other phases in the life history of the cod
off southern New England. Bull. U.S. Bur. Fish. 46, 136 p.
Scott, J.S.
1982 Selection of bottom type by groundfishes of the Scotian
Shelf Can. J. Fish. Aquat. Sci. 39:943-947.
1988 Seasonal spatial distributions of groundfish of the Scotian
shelf and Bay of Fundy, 1974-79 and 1980-84. Can. Tech.
Rep. Fish. Aquat. Sci. 1653, 75 p.
Serchuk, F.M., and S.E. Wigley
In press Assessment and management of the Georges Bank
cod fishery: An historical review and evaluation. J. North-
west Atl. Fish. Sci.
1986 Assessment and status of the Georges Bank and Gulf of
Maine Atlantic cod stocks- 1986. Ref Doc. 86-12, NMFS
Woods Hole Lab., 84 p.
Smith, H.M.
1902 Notes of the tagging of 4,000 adult cod at Woods Hole,
Mass. U.S. Comm. Fish Fish., Part 27, Rep. of the Comm.,
year ending June 30, 1901. p. 193-208.
Wigley, R.L.
1961 Bottom sediments of Georges Bank. J. Sediment Petrol.
31:16.5-188.
Wigley, S.E., and W.L. Gabriel
1991 Distribution of sexually immature components of ten
Northwest Atlantic groundfish species, based on Northeast
Fisheries Center Bottom Trawl Surveys, 1968-1986. NOAA
Tech. Memo. NMFS-F/NEC-80, Northeast Fish. Sci. Cent.,
Woods Hole, 17 p.
Wise. J. P.
1958 The world's southernmost indigenous cod. J. Cons.
Cons. Int. Explor. Mer 23:208-212.
1962 Cod groups in the New England. Fish Bull., U.S.
62:189-203.
Abstract.— Developmental series
of two sympatric flounders of the
genus Pamlichthys, found in the Bay
of Coquimbo, are illustrated and de-
scribed. The series consist of yolksac
to metamorphosed larvae of artifici-
ally-reared Paralichthys adsperstis
(1.7-13.0 mm SL) and P. microps
(1.5-ll.Omm SL). Field-collected
larvae correspond to the size ranges
found in reared larvae. Degree of
cephalic spination (in particular,
sphenotic spines), pigmentation pat-
tern, and number of elongated dor-
sal-fin rays are useful for identifica-
tion of yolksac-to-postflexion larvae
of both species.
During early metamorphosis the
most valuable characteristics for
identification are the number of
elongated dorsal-fin rays, although
after their reabsorption several mor-
phometric relationships have to be
used. Paralichthys adspcrsus pre-
flexion larvae have two sphenotic
spines and almost no pigmentation
in the dorsal finfold, while P. microps
larvae have only one sphenotic spine
and a well-pigmented dorsal finfold.
Beginning at notochordal flexion, the
number of elongated dorsal-fin rays,
six for P. microps and three for P.
adspersus, can be used to identify the
larvae. During late metamorphosis,
morphometric relationships of SnL/
HL, HL/SL, and BD/SL must be
used to identify the larvae. Flexion
is complete at 7.2mmSL and meta-
morphosis at ~11.0mmSL in P.
microps, and at 8.6mmSL and 13.0
mm SL in P. adspeisus, respectively.
Larval development of two sympatric
flounders, Paralichthys adspersus
(Stelndachner, 1867) and
Paralichthys microps (Gunther, 1881)
from the Bay of Coquimbo, Chile
Humberto N. Zuhiga
Enzo S. Acuha
Departemento Biologfa Manna, Universidad Cat6lica del Norte
Sede Coquimbo, Casilla 1 1 7, Coquimbo, Chile
Manuscript accepted 20 May 1992.
Fishery Bulletin, U.S. 90:607-620 (1992).
Paralichthys is one of the most im-
portant genera of flatfish on both
coasts of North and South America
(Ginsburg 1952), considering number
of species, geographic distribution,
and economic importance. Seven spe-
cies of the genus have been reported
in Chilean waters (Bahamonde and
Pequeno 1975), Paralichthys adsper-
sus (Steindachner 1867) and P. mm'ops
(Gunther 1881) being the most abun-
dant and most widely distributed.
The former is found from the coast
of Paita (Peru) to Lota (Chile) and
Juan Fernandez Island; the latter
from Huacho (Peru) to the austral tip
of South America (Chirichigno 1974).
Because these two morphologically-
similar species co-occur over most of
their distributional ranges, adult and
larval identifications have been dif-
ficult. Muiioz et al. (1988) described
larvae of P. microps, but recognized
the possibility that specimens of both
species were included in their sample.
They indeed have one P. adspersus
larva (3.2 mm, Fig. 2b). Silva (1988)
published photographs of the eggs
and some larvae of P. microps.
In this paper, taxonomic characters
which separate these two species dur-
ing early-life-history stages, from yolk-
sac larva to juvenile, are described.
Material and methods
Most of the material examined in the
present study was obtained from
several experiments, resulting from
artificial fertilization of eggs and
sperm from ripe specimens captured
in the Bay of Coquimbo (29°59'S).
Larvae were cultured in 200 L
conical tanks, with a daily 25% water
renewal. From hatching through
flexion, larvae were fed the rotifer
Brachionus plicatilis in concen-
trations of 5/mL, and from flexion
through metamorphosis were fed
Arternia salina nauplii in concentra-
tions of 10/mL. Temperature range
during the experiment was 13-17°C
(Silva 1988). Larvae, sampled with a
Bongo net (Im, bOOju mesh) and an
epibenthic trawl (500 /.< mesh) at sta-
tions in Coquimbo Bay and adjacent
coastal areas, were compared with
cultured larvae.
A total of 49 larvae of P. adsper-
sus and 46 of P. microps were used;
of these, 39 larvae of both species
were cleared and stained using Pott-
hoff s (1984) method to determine the
sequence of development of the axial
skeleton. Pterygiophores and rays
were counted when present, regard-
less of their state of development.
Larvae were anesthesized with MS-
222 before fixing in 5% formalin, and
were later preserved in 3% buffered
formalin.
Specimens were divided into devel-
opmental stages following the defi-
nitions of Ahlstrom et al. (1976).
607
608
Fishery Bulletin 90(3). 1992
Table 1
Morphometric relationships in Paralichlhys adspersus and P. microi.
.s larvae.
A' = number of specimens; measurements in mm; length |
= NL for preflexion-flexion stages, SL for postflexion-juvenile.
Measure
Preflexion
Flexion
X
SD
Range
X
SD
Range
Paralichthys adspersus
N
21
11
Length
5.52
1.12
(3.6-7.0)
7.80
0.55
(6.9-8.6)
PAL/SL
41.40
1.32
(40.5-43.5)
39.60
3.08
(31.9-42.0)
BD/SL
13.90
3.80
(13.0-18.1)
17*
22.80
5.08
(17.4-31.4)
BDt/SL
HL/SL
18.20
1.13
(16.7-20.0)
21.50
2.53
(18.6-26.7)
UJL/HL
37.30
4.52
(29.4-39.4)
17*
37.90
2.32
(33.3-41.7)
LJL/HL
46.30
3.08
(41.2-51.8)
19*
48.10
3.15
(44.7-54.7)
SnL/HL
22.10
2.22
(18.3-26.8)
21.10
1.02
(18.9-22.4) 10*
ED/HL
28.00
2.36
(24.8-33.3)
26.80
1.50
(24.7-28.9)
Para/icAf%s
microps
N
22
8
SL
4.17
0.95
(2.95-6.0)
6.86
0.35
(6.2-7.2)
PAL/SL
41.15
1.83
(38-44)
40.90
3.27
(35-45)
BD/SL
13.30
2.16
(10-18)
25.60
6.40
(19-37)
BDt/SL
HL/SL
18.10
1.48
(15-20)
21.60
3.70
(17-28)
UJL/HL
34.60
3.60
(29-37) 11*
44.50
4.20
(38-50)
LJL/HL
51.98
4.50
(43-59) 20*
57.10
3.86
(50-62)
SnL/HL
22.90
3.56
(16-31) 20*
23.90
2.88
(20-28)
ED/HL
30.10
4.10
(23-43) 2
L*
26.80
1.70
(24-29)
Measure
Postflexion
Metamorphosis
Juvenile
X
SD
Range
X
SD
Range
X
SD
Range
Paralichthys adspersus
N
7
6
4 1
SL
8.90
0.35
(8.4-9.4)
10.00
0.53
(9.2-10.2)
13.60
1.04
(12.3-15.0)
PAL/SL
40.80
2.76
(37.6-45.2)
37.60
3.17
(33.9-43.5)
34.40
0.68
(33.3-35.0)
BD/SL
35.80
1.77
(33.7-38.2)
39.50
1.42
(36.6-41.3)
37.90
0.72
(37.2-39.0)
BDt/SL
41.90
2.90
(36.4-45.8)
35.30
0.83
(34.4-36.2)
HL/SL
30.60
2.35
(27.6-33.7) 6*
32.00
1.69
(29.7-34.7)
35.10
0.38
(34.7-35.7)
UJL/HL
34.10
1.71
(31.5-37.3)
34.50
1.49
(31.7-36.5)
34.40
0.94
(33.3-35.6)
LJL/HL
45.60
2.61
(41.7-50.5)
45.90
1.35
(42.9-48.4)
46.80
1.50
(44.4-48.1)
SnL/HL
20.70
2.55
(17.5-25.8)
18.30
1.66
(15.6-20.3) 5*
14.30
1.76
(11.5-16.3)
ED/HL
25.50
1.07
(24.1-26.8) 6*
25.00
2.57
(21.4-28.6)
28.70
2.09
(26.3-32.0)
/'ara/icArti/i
microps
iV
4
8
4 1
SL
6.90
0.77
(6-7.8)
9.30
0.67
(8.10-10.6)
15.20
1.95
(13.0-18.0)
PAL/SL
42.20
3.40
(37-46)
39.40
2.47
(36.2-43.5)
35.30
0.83
(34.4-36.2)
BD/SL
44.70
6.20
(35-46)
40.00
1.39
(37.7-42.0)
37.40
0.86
(36.2-38.5)
BDt/SL
46.50
1.03
(45.4-48.9)
HL/SL
37.30
5.10
(32-46)
39.10
2.68
(35.8-43.0)
37.40
1.19
(36.3-39.4)
UJL/HL
38.90
2.10
(35-41)
37.80
2.09
(34.7-39.7) 6*
35.60
1.47
(34.3-38.0)
LJL/HL
50.70
4.30
(42-56)
47.90
2.70
(44.2-51.7) 6*
45.80
1.86
(43.1-46.9)
SnL/HL
23.20
1.90
(20-25)
23.30
1.73
(21.3-26.6) 6*
16.70
1.39
(15.5-19.0)
ED/HL
24.40
0.70
indicated
(23-25)
above.
25.10
1.32
(23.0-26.6)
28.80
1.99
(25.4-30.2)
*Af differs from number
BD Body depth
PAL
Preanal length
BDt Body depth measured at anus SL
Standard len
^h
ED Eye diameter
SnL
Snout length
HL Head length
UJL
Upper jaw length
LJL Lower
jaw length
Zuniga and Acuna' Development of Paralichthys spp larvae
609
Morphometric measurements follow the definitions of
Gutherz (1970) and were made with an ocular microm-
eter (to 0.01mm). Notochordal length (NL) was used
for yolksac larvae through flexion: from then on, stan-
dard length (SL) was utilized. In preflexion and flex-
ion larvae, body depth (BD) is defined as the vertical
distance across the body at the anus including the
dorsal-fin pterygiophores. After flexion, it is defined
as the vertical distance across the body at the pelvic
fin, from its base to the base of dorsal-fin rays. Head
length (HL) is defined as the distance from the snout
to the cleithrum, until and through flexion, and there-
after from snout to the opercle edge. The total number
of myomeres and vertebrae does not include the uro-
style. Drawings were made from a compound micro-
scope equipped with a camera lucida.
Linear regression models were fitted to six morpho-
metric relationships of the larvae, comparing the pre-
flexion stages with flexion, postflexion, and metamor-
phosis, to separate larvae of both species. An F test
(Neter and Wasserman 1974) was used to compare the
morphometric relationships of these two groups of lar-
vae within and between species.
Determination of Paralichthys adults was based on
Table 2
Osteological development sequence of fins
n Paralichthys
adspersiis
and P. microps larvae. Length = NL
preflexion, SL
postflexion.
Paralichthys adspersus
Paralichthys microps
Rays +
Rays +
SL
Pterygiophores
Dorsal Anal
Rays + Radials
Pelvic Pectoral
Rays
Vertebrae
Pterygiophores
Dorsal Anal
Rays +
Pelvic
Radials
Pectoral
Rays
Vertebrae
Caudal
Caudal
4.1
4.5
4.7
—
—
-
-
-
-
4.8
5.4
5.5
-
-
-
-
-
-
0+1
5.9
0 + 2
-
-
-
-
-
6.1
—
—
—
—
—
2 + 2
—
—
—
—
—
6.2
3 + 3
—
—
—
—
—
6.2
3 + 3
—
—
—
—
—
6.5
3 + 4
—
—
—
—
3
6.5
4 + 5
0 + 20
—
—
—
27 + -
6.7
2 + 3
—
—
—
—
—
6.9
3 + 3
—
—
—
—
—
7.0
9 + 60
24 + 49
—
6 + 7
29 + -
7.1
64 + 62
53 + 51
3 + 0
8 + 7
30 + -
7.2
69 + 68
? + 53
5 + 0
1+9+8+1
34
7.7
3 + 3
—
—
—
_
9
71 + 69
57 + 55
5 + 0
1+10+8+1
34
7.7
3 + 4
—
—
—
—
26
7.8
71 + 71
53 + 53
5 + 0
1+10+9+0
34
8.1
5 + 4
—
—
—
—
32
73 + 72
57 + 54
6 + 0
0+10+8+0
34
8,1
6 + 30
0 + 23
—
—
—
33
8.1
10 + 60
0 + 48
—
—
—
33
8.2
13 + 48
0 + 42
—
—
9 + 8
33
8.5
45 + 64
38 + 51
3 + 0
—
9 + 9
33
8.6
56 + 65
30 + 51
3 + 0
—
1+9+8+1
33
73 + 73
58 + 57
6 + 0
—
1+9+9+1
34
8.6
64 + 66
50 + 51
4 + 0
—
1+9+8+1
33
8.6
69 + 67
55 + 53
5 + 0
—
1+9+8+1
33
8.8
73 + 72
55 + 55
5 + 0
—
1+10+9+1
33
9.2
71+71
57 + 57
5 + 0
—
1+9+9+1
32
75 + 75
60 + 59
6 + 0
0 + 2
1+10+8+1
34
9.7
—
75 + 75
64 + 61
6 + 0
3 + 0
1+10+8+1
35
10.1
74 + 72
59 + 57
6 + 0
—
1+9+9+1
34
72 + 71
58 + 56
6 + 2
6 + 3
1+10+9+1
33
10.2
70 + 70
54 + 53
6 + 0
0 + 3
1+9+9+1
33
10.9
71 + 70
57 + 55
6 + 2
7 + 4
1+9+9+1
33
11.3
70 + 68
59 + 57
6 + 3
14 + 4
1+9+9+1
34
12.3
68 + 66
55 + 54
6 + 3
14 + 4
1+10+9+1
33
12.7
73 + 72
59 + 56
6 + 3
14 + 4
1+9+9+1
33
14.0
70 + 69
57 + 56
6 + 3
13 + 4
1+9+8+1
33
610
Fishery Bulletin 90(3). 1992
E
Figure 1
Paralwhthys adspersus larvae. (A) Early yolksac, 1 .7 mm NL; (B) late yolksac, 3.4 mm NL; (C)
early preflexion, 3.9 mm NL; (D) late preflexion, 6.8mmNL; (E) flexion, 7.7 mm NL. Draw-
ings at right are dorsal head views.
Zuhiga and Acuna: Development of Paralichthys spp larvae
61
the criteria of Ginsburg (1952), who observed that the
origin of the dorsal fin in P. microps was over the
center of the upper eye, while in P. adspersus it was
over the eye's anterior margin. Furthermore, the
number of gill rakers over the lower portion of the first
arch is larger in P. microps (18-23) than in P. adsper-
sus (15-19). An additional criterion found by Zuniga
(1988) referring to the size of the nostrils was also used.
Description
Paralichthys adspersus
Hatching occurs ~60 hours after fertilization. Larvae
are ~1.7mmNL; yolksac is more than half the body
length; a small oil globule (0.13mm) is present posterior
to the yolksac (Fig. lA).
Diagnosis The most important distinguishing fea-
tures of preflexion P. adsperstis larvae are the presence
of two sphenotic spines (Fig. 2) and the lack of pigmen-
tation in the dorsal finfold (Fig. 1). This last character
may be useful through postflexion. Starting at noto-
chord flexion, the presence of two groups of numerous
opercular and preopercular spines, as well as 2-3
elongated dorsal-fin rays, is diagnostic. This last fea-
ture is useful until metamorphosis. Beyond metamor-
phosis, diagnosis should be based mostly on mor-
phometric relationships.
Pigmentation Eyes of yolksac larvae are not pig-
mented. Few, relatively-large melanophores are found
on the head, trunk, and yolksac except at the ventral
margin (Fig. lA). A series of small melanophores is
present near the tip of the notochord. Pigment forms
Figure 2
Heads of Paralichthys adspersus (left) and P. microps (right) >arvae, showing number, size, and location
of sphenotic, preopercular, and opercular head spines. Bars = 1.0 mm.
612
Fishery Bulletin 90(3), 1992
^^^^"i^^lfimJ:-^
jtV' '■■- j^-i-J- ^-^•»-^^- "<"■?..■»--■; "•'.'-
B
Figure 3
Paralichthys adspersus larvae. (A) Early postflexion, 8.8 mm SL; (B) early metamor-
phosis, 10.3 mm SL; (C) juvenile, 14.3 mm SL. Bars = 1.0 mm.
on the medial region of the pelvic
fin. At the end of the stage (Fig.
IB), the eyes start to pigment.
During the preflexion stage
(Fig. IC, D), stellate melano-
phores are present on the head
and over the anterior 2/3 of the
body. Melanophores are absent
on the dorsal finfold. At 3.5mm
NL, a series of melanophores
forms on each side, slightly dor-
sal to the midline. At 4.0mmNL,
an embedded series of melano-
phores begins to develop dorsal
to the notochord. A series of
melanophores is present at the
ventral margin of the body, from
the gular region to the anus.
Head pigmentation consists of
melanophores over both jaws,
preopercle, opercle, and dorsal
and lateral brain region. At
about S.OmmNL, a melanophore
is found internally above the
palate, where it persists until
metamorphosis.
At the beginning of the flexion
stage (Fig. 3A), pigment intensi-
fies in the tail region. The dorsal
finfold generally remains unpig-
mented; however, a few melano-
phores appear in some specimens.
A paired series of melanophores
develops above the dorsum. Tail
melanophores concentrate in the
ventrolateral region, while the
paired series dorsal to the noto-
chord is less visible. The paired
series dorsal to the gut becomes
continuous with the gular-abdom-
inal series. Head pigmentation
increases The interradial mem-
brane of the elongated dorsal-fin
rays becomes pigmented. Melano-
phores near the tip of the noto-
chord persist but migrate as the
caudal fin develops.
During postflexion (Fig. 3B),
the melanistic pattern is similar
to the previous stage except the
paired dorsal series is more evi-
dent. The dorsal fin is pigmented,
particularly in the posterior half,
and the ventral region of the ab-
domen becomes pigmented.
Zuniga and Acuna' Development of Paralichthys spp larvae
613
During metamorphosis (Fig. 3C) pigmentation in-
creases, especially on the left side of the body. Dorsal-
fin pigmentation is concentrated in the posterior half
of the fin. Groups of melanophores are present on the
interradial membrane of the anal fin. The pelvic fin is
almost completely pigmented.
Fin development Dorsal-fin pterygiophores and rays
begin to form simultaneously at ~6.5mmNL, reach-
ing their full complements at 8.8mmSL (Table 2). The
anal-fin pterygiophores appear in advance of their cor-
responding fin rays, at 8.1-8.6mmNL. In both fins,
development proceeds posteriad. The pelvic fin appears
at 8.5mmNL, and all six rays are present at 10.1mm
SL. The hypural complex develops between 8.1 and
8.6mmNL. There are 18 caudal-fin rays, plus 2 pro-
current rays.
Morphology With absorption of the yolk, the yolksac
larva becomes slender; the gut, jaws, and pectoral fins
develop; and two sphenotic spines begin to develop on
each side of the head. At 3.5mmNL (4-5 days post-
hatching), the yolk is exhausted, the mouth is func-
tional, eyes are pigmented, and the pectoral fin is
formed.
Two sphenotic spines appear at ~3.0mmNL on each
side of the head (Fig. 2); initially the upper one is the
larger. Both spines are reabsorbed before development
of elongated dorsal-fin rays, near the end of this stage.
On some specimens, a third, smaller sphenotic spine
can be found below the first two.
At about 4.5mmNL, preopercular and opercular
spines appear; the former are located along the pos-
terior margin of the preoperculum and on the anterior
preopercular ridge. Opercular spines are located at the
upper portion of the bone and are more prominent than
preopercular spines.
At ~6.5mmNL, the elongated rays of the dorsal-fin
crest begin to appear. Three rays (corresponding to the
second, third, and fourth dorsal-fin rays of the adult)
form the initial crest. At 6.2mmNL the gut begins to
coil. During preflexion, body depth is moderate (13.9%
NL) and preanal distance is ~41.4%NL. These propor-
tions remain relatively constant during later develop-
ment (Table 1). There are 33 myomeres (11 preanal and
22 postanal) at the end of the stage.
The beginning of the flexion stage is characterized
by an increase in body depth (22.8% NL) and develop-
ment of the caudal fin. Preopercular spines are in two
series in the upper and lower margins of the bone.
Opercular spines are also in two groups: an upper group
on the body of the operculum and a lower one along
its margin. Elongated dorsal-fin rays remain, the mid-
dle one being the longest. At ~7.5mmNL, the pelvic
fins begin to form and, by the end of the stage, rays
and pterygiophores of dorsal, anal and caudal fins are
more evident.
Morphometric proportions are similar to those of the
previous stage, except body depth which increases. A
small increase in head length is also apparent (Table
1). There are 33 (9 preanal and 24 postanal) myomeres
at the end of the stage.
During postflexion, body depth increases to 35.8%
SL; head length reaches 30.6% SL (Table 1). Dorsal-
crest fin rays increase in relative length; the second
reaches 50% SL. The short dorsal-fin ray anterior to
the crest begins to develop.
Preopercular and opercular spination increases in
some specimens, but the spines begin to reabsorb at
the end of the stage. The interocular region begins to
change in preparation for eye migration. There are 33
(7-8 preanal and 25-26 postanal) myomeres at the end
of the stage.
During metamorphosis, body depth continues to in-
crease (39.0% SL), and snout length decreases (18.3%
HL); however, other body proportions do not change
substantially (Table 1). The second elongated dorsal-
fin ray reaches its maximum length (~53.9% SL) before
being reabsorbed. Migration of the right eye to the left
side begins. Pectoral-fin rays form. Preopercular and
opercular spines are lost, as are the elongated dorsal-
fin rays. Eye migration is completed at ~13.0mm. The
smallest juvenile was 12.3 mm SL (Fig. 3C). There are
33 (4-6 preanal and 27-29 postanal) myomeres at the
end of the stage.
Paralichthys microps
Hatching occurs 57-68 hours postfertilization; yolksac
larvae are ~1.5mmNL; one oil globule is present.
Yolksac development is similar to P. adspersus yolk-
sac larvae, except that melanophores form on the dor-
sal and anal finfold and a simple sphenotic spine begins
to develop at yolk exhaustion (~3.2mmNL, 4-5 days
after hatching).
Diagnosis Distinguishing features of preflexion P.
microps larvae are the presence of only one sphenotic
spine (Fig. 2) and the dorsal finfold with pigmentation.
After notochord flexion until metamorphosis, the most
distinguishing feature is the presence of more than 3,
and later 6, elongated dorsal-fin rays (Figs. 4E, 5).
After reabsorption of these elongated rays, diagnosis
is mostly based on morphometric relationships.
Pigmentation During the preflexion stage, the pig-
mentation pattern is similar to that of P. adspersus,
but P. microps larvae have a different arrangement of
body and finfold melanophores and less head pigment.
Melanophores are relatively sparse over the trunk and
614
Fishery Bulletin 90(3). 1992
Figure 4
Paralichthys microps larvae: (A) Early yolksac, l.SmmNL; (B) late yolksac, 3.0mm
NL; (C)early preflexion, 3.4mmNL; (D) middle preflexion, 5.3 mm NL; (E) late flex-
ion, 7.0mmNL. Drawings at right are dorsal head views. Bars = 1.0 mm.
anterior one-third of the tail. A
dense zone of melanophores devel-
ops on the middle one-third of the
tail and associated dorsal and ven-
tral iinfold regions (Fig. 4). Pigment
is less dense on more anterior re-
gions of the finfold and is absent on
the posterior one-third of the tail
and finfold. Head pigmentation is
restricted to the jaws, dorsal brain,
and opercle.
During the flexion stage, the
melanistic zone on the tail and asso-
ciated finfold region intensifies. The
paired series along the dorsum and
the epaxial region remains visible,
while the embedded series dorsal to
notochord is less visible due to the
development of musculature. The
number of ventral and ventrolateral
abdominal melanophores increases.
The ventral region of the gut has
small melanophores, while those on
the side of the gut are larger and
stellate. The series above the gut
and along its ventral midline are
less apparent.
The pattern of melanophores on
the head remains about the same,
with brain melanophores the most
conspicuous. Head pigmentation
consists of small melanophores on
the jaws, palate, preoperculum,
operculum, and gular region, and
larger and stellate melanophores in
the brain region. The small melano-
phores located at the ventral mar-
gin near the tip of the notochord
disappear with development of the
caudal fin. Melanophores increase
in number on the dorsal-fin crest
and on the interradial m^embranes
of the dorsal and anal fins during
the postflexion stage (Fig. 5A).
Finally, during the metamorpho-
sis stage (Fig. 5C) melanophores in-
crease in numbers on the head and
body, especially on the left side.
Body melanophores are associated
with myosepta. The paired dorsal
series remains visible. Melano-
phores on the dorsal and anal fins
are arranged in groups. The elong-
ated dorsal-fin rays are covered with
melanophores. The rear margins
Zuniga and Acuna. Development of Paralichthys spp, larvae
615
of the hypural plates become pig-
mented, as do the bases of the
caudal-fin rays.
Fin development Pterygiophores
and fin rays of the dorsal fin ap-
pear at 5.5mmNL; full com-
plements are present at 7.7mm
SL (Table 2). Anal-fin pterygio-
phores and rays appear at 6.5
mmNL and have full comple-
ments at 7.7mmSL. The pelvic-
fin rays begin to develop at 7.0
mmNL and all are present at 8.1
mmSL; pterygiophores begin to
develop at 10.1 mmSL and all are
present at ll.SmmSL. The first
pectoral-fin rays appear at ~9.5
mmSL; full complements are
present at ll.SmmSL. The hypu-
ral complex develops between
6.2 and 7.2mmNL. The number
of caudal-fin rays is 18, plus 2
procurrent rays.
Morphology The sphenotic spine
is more developed than in P.
adspersus preflexion larvae (Fig.
2) (max. length is 24% eye diam-
eter) and disappears with the ap-
pearance of the elongated dorsal-
fin rays. At ~5mmNL, up to 4
spines may be found at the pre-
opercular margin; up to 3 spines
are found in the upper region of
opercle. At this size, the gut be-
comes coiled and the larva is
moderately slender. Preanal dis-
tance is 41.2% NL; body depth
(BD) is 13.3% NL, and upper jaw
length (UJL) is 34.6% HL (Table
1)-
At '^'6mmNL, three elongated
rays appear on the dorsal fin
crest. They correspond to the
second, third, and fourth rays of
the adult fin. The middle ray of
the crest is the longest. There are
34 (11-12 preanal and 22-23
postanal) myomeres at the end of
the stage.
The flexion stage is character-
ized by development of the hypu-
ral elements of the caudal fin.
Body depth increases to 25.6%
B
Figure 5
Paralichthys microps larvae. (A) early postflexion, 7.4mmSL; (B) early to middle
metamorphosis, 8.7mmSL: (C) juvenile, 15.5mmSL. Bars = 1.0mm.
616
Fishery Bulletin 90(3). 1992
NL; upper and lower jaw lengths increase to 44.5% and
57.1%HL. Relative eye diameter decreases to 26.8%
HL (Table 1).
The number and location of preopercular and oper-
cular spines remain almost constant. The pelvic fin
starts to develop at ~7.0mmNL. Up to 6 elongated fin
rays develop in the dorsal crest. There are 34 (9-11
preanal and 23-25 postanal) myomeres at the end of
the stage.
The 6 (sometimes 7) fin rays of the dorsal crest con-
tinue to elongate during the postflexion stage. The
fourth and fifth rays are more than half the body
length; the first dorsal-fin ray is apparent but poorly
developed. Rays and pterygiophores of median fins
become more apparent.
By the end of this stage, preopercular and opercular
spines start to disappear and the interocular region
begins to deform in preparation for eye migration.
Body depth increases (up to 44.7% SL) as does head
length (up to 37% SL), with a corresponding decrease
in relative jaw length and eye diameter (Table 1). There
are 34 (7-9 preanal and 25-27 postanal) myomeres at
the end of the stage.
Finally, during metamorphosis as the right eye
migrates towards the left side of the body, the dorsal
crest is lost, the mouth changes form, and pectoral-fin
rays form (Fig. 5C).
s .
A
5 •
c
u ■
•
r °
i
» •%••
^
0. °8l»- •
C T
• -^ 1
o'oo* ' '°
r 3 ■
. ° - -
^!°*
•• 1-
• ^•«
X
1 • ^
o«c
l°'\
0.
' • _)
„ ooo
S 2 ■
1° • J '
»Ji
>
1 .
°: 1 f
ooo •• UJ
qW
1 2
3456789 10 11 123456789 10 1'
Standard Length ( mm ) Standard Length ( mm )
5 .
B
20.
D
L .
oo 16-
•
%o °
o ° • • •
~
. • •
0 • ,
^ 3 .
•
o • •
•
i •• -
o> .
X
z
0 1 7 -
r -"^
_uj 2
°t n -°«-
•
OcpP •
Q
<
1
5 ; u 04-
' o
o'^°
; 1
orf* 1
1 2
3 4 5 6 7 8 9 10 11 1 2 3 4 5 6 7 9 9 10 11
Standard I ength ( mm ) Standard Length ( mm )
Figure 6
Morphometric relationships (vs. SL) of Paralichthys adspers-ii.'i { • ) and P. microps (O) larvae. (A) Body depth; (B) head length; (C) |
preanal length;
(D) lower jaw length. Solid vertical line shows flexion of P. microps, and broken line flexion of P. adspersus.
Zuhiga and Acuiia Development of Parahchthys spp larvae
617
Morphometries
Six morphometric functional relationships are shown
in Figiires 6 and 7, and all linear regression models and
their r- values are summarized in Table 3 (abbrevia-
tions as in Table 1). In general, all morphometric rela-
tionships were adequately described by the linear
regression model, especially the preflexion P. adsper-
sus larvae which always had higher r^ values than
those for other stages of the same species and all stages
of P. microps (Table 3). The relationships are not so
clear in P. microps, because the preflexion SnL/HL,
PAL/SL, BD/SL, and HL/SL relationships had higher
r- values than those of the other stages, while the
relationships UJL/SL, LJL/SL, and ED/SL of other
stages had higher values of r^ than those from pre-
flexion (Table 3).
F tests showed that models for all preflexion rela-
tionships and the PAL/SL of "other stages" could be
considered statistically identical for both species, while
all others were significantly different (Table 4). Regres-
sion models for all morphometric relationships (except
SnL/HL and PAL/SL) between the two groups of
developmental stages of P. microps were significant-
ly different, while in P. adspersus all but SnL/HL,
PAL/SL, and UJL/SL were significantly different
(Table 5).
A summary of larval characters useful to identify lar-
vae of both species during the different larval stages,
including morphology, pigmentation, and morphomet-
ries, is shown in Table 6.
Discussion
Characteristics of larval development of P. adspersus
and P. microps are, in general, similar to those ob-
served in other species of the genus (P. dentatus Smith
and Fahay 1970; P. olivaceus Mito 1963, Okiyama 1967;
P. califomicus, Ahlstrom and Moser 1975). Important
common characteristics are: presence of only one oil
globule posteriad in the yolksac larvae; small size at
hatching, notochordal flexion, and metamorphosis;
presence of sphenotic spines; two groups of preoper-
cular spines; elongated anterior dorsal-fin rays; a deep
laterally-compressed body; and a large visceral mass.
Opercular spines present in the two species described
herein are uncommon in the family.
Length at hatching of Paralichthyid larvae varies
between 1.5 and 3.7mmNL with a mean of 2.2mm,
while the range described for the genus Paralichthys
is 2.0-2.8 mm (Ahlstrom et al. 1984). Thus, hatching
sizes (1.5-1. 7mm) of P. adspersus and P. microps lar-
vae are smaller than any known congener. Lengths at
flexion and metamorphosis of both species fall within
20 -
A
16 .
o
-^
0^
z
z:
o
w 12 -
• •
x:
•.•••
tc
o
z
••
LU
-"08 ■
<
o 1 \
-3
o M° ^
'^ 0^ ■
.o^f^'^'
3:
oao
123iS6 789 10 1i
Standard Length ( mm )
20 -
B
16 -
z
r
- 12 -
cc
LU
O • 0
£
n '^O
^ 08 .
° • .
en
O M
LU
! o o •• •
>-
LU
<Po '•;
Oi -
o'^*^'^'*^
123456789 10 n
Standard Length ( mm )
1 0
C
0
00
o
8 .
• o
o
•
Z 6.
° ° ..• .:
E
°
—
•
• •
X
1 • _
i '•■
|o ^,
LLJ
1
«
1,2.
0 ^^
I 1
C/5
0*
1
1
1
1
1
1 2 3 1. 5 6 7 8 9 10 11
Head Length ( mm )
Figure 7
Morph
Dmetric relationships (vs. SL and HL) of Paralichthys
adsper
5MS (•) and P. microps (O) larvae. (A) Upper jaw
length,
(B) ocular diameter, (C) snout length. Symbols as in
Fig. 6.
61!
Fishery Bulletin 90(3). 1992
Table 3
Linear regression equations and r- values of selected morphometric relationships for preflexion and "other stages" larvae of Para-
lichythys adspersus and P. microps. Abbreviations as in Table 1.
Relationship
PAL/SL
BD/SL
HL/SL
UJL/SL
LJL/SL
ED/SL
SnL/HL
P. adspersus
Preflexion
0.091 +
-0.311 +
r
-0.189 +
r
-0.284 +
r
-0.137 +
r
0.026 +
r
-0.031 +
r
0.397X
0.975
0.205X
0.933
0.218X
0.971
0.119X
0.956
O.llOX
0.974
0.046X
0.931
0.254X
0.896
Other stages
0.652 + 0
r^ 0
-5.446 + 0
r= 0,
-3.428 + 0
r- 0
-1.012 + 0
r' 0
-1.390 + 0
r2 0
-1.114 + 0
r- 0
0.082 + 0
r- 0
328X
557
944X
,873
,669X
,817
,215X
.701
.288X
,723
,199X
.911
.166X
.794
microps
Preflexion
0.088 + 0.388X
r- 0.957
0.328 + 0.214X
r- 0.901
-0.149 + 0.217X
r- 0.937
- 0.129 + 0.094X
r- 0.805
-0.034 + 0.104X
r^ 0.815
0.047 + 0.042X
r'~ 0.785
0.006 + 0.223X
r- 0.792
Other stages
0.534 +
r'
-2.940 +
r^
-3.806 +
r'
-1.060 +
r-
-1.165 +
r-
-0.839 +
r-
0.033 +
0.338X
0.772
0.711X
0.821
0.796X
0.905
0.258X
0.889
0.309X
0.860
0.187X
0.934
0.222X
0.958
Table 4
Values of F for two regression
models of morphometric
relationships
setween preflex-
ion and other developmental stages (flexion,
postflexion
, and
metamorphosis) between |
Paralichthys adspersus
and P.
microps. Abbreviations
as m
Table 1
Relationship
Preflexion
Other stages
F
df
F
df
PAL/SL
1.18
(2, 41)
NS
0.05
(2, 38) NS
BD/SL
0.66
(2, 38)
NS
11.24
(2, 38) *
HL/SL
2.04
(2, 38)
NS
21.82
(2, 38) •
UJL/SL
2.46
(2, 23)
NS
19.73
(2, 38) *
LJL/SL
1.28
(2, 36)
NS
17.79
(2, 35) •
ED/SL
0.67
(2, 38)
NS
34.97
(2, 33) •
SnL/HL
1.25 (2, 38) NS
•Significant (P< 0.001)
14.81
(2, 35) •
NS = Non-significant
Table 5
F tests for two regressior
1 models
of morphometric relationships between
preflexion and
other developmental stages (flexion, postflexion. and
metamorphosis) within two species |
of Paralichthys.
Abbreviations as in Table 1
Relationship
P.
adspersus
P.
microps
F
df
F
df
PAL/SL
0.79
(2, 43)
NS
1.48
(2, 36) NS
BD/SL
50.90
(2, 41)
*
17.18
(2, 35) *
HL/SL
22.90
(2, 41)
*
36.18
(2, 35) •
UJL/SL
3.32
(2, 34)
NS
10.52
(2, 25) •
LJL/SL
10.39
(2, 40)
*
18.03
(2. 35) ♦
ED/SL
70.53
(2, 37)
«■
51.96
(2, 34) •
SnL/HL
ficant
2.14 (2, 40) NS
•Significant (P< 0.001)
0.83
(2, 33) NS
NS = Non-sign
the known range of the genus. P.
adspersus is larger at metamor-
phosis than P. microps (9.6-13.0
mmSL vs. 8.0-ll.OmmSL, re-
spectively) and is comparable to
the 10.2-14.2mmSL range for
P. olivaceus (Okiyama 1967).
Early presence of elongated
anterior dorsal-fin rays in P.
adspersus and P. microps, at 6.5
and 6.0mmSL, respectively, is
common in Paralichthys and
related genera of paralichthyids
(sensu Ahlstrom et al. 1984). The
six elongated dorsal-fin rays ob-
served in P. microps fall within
the described range (4-8) for
Paralichthys, while P. adspe7-sus
only has 3, as in the related
genus Citharichthys (Ahlstrom
et al. 1984). The shape and size
of these dorsal-fin rays is char-
acteristic of the genus Paralich-
thys and not Citharichthys.
The melanistic pattern of the
larvae of both species is very
similar to that described for the
genera Paralichthys and Pseudo-
rhombus (sensu Ahlstrom et al.
1984); however, the pigment
series along the horizontal sep-
tum described for other species
Zuniga and Acuiia- Development of Pcsra/zc/if/iys spp larvae
619
Table 6
Summary of larval characters which distinguish larvae of Paral
ichihys adspersiis and P. microps during different larval stages. Ab- |
breviations as in Table 1
Developmental stage
P. adspersus
P. microps
Preflexion
Two sphenotic spines present before develop-
One sphenotic spine present before development
ment of dorsal-fin rays.
of dorsal-fin rays.
Dorsal finfold unpigmented.
Dorsal finfold pigmented.
LJL/HL = 46.3%
LJL/HL = 52.0%
Flexion
Dorsal finfold unpigmented.
Dorsal finfold pigmented.
LJL/HL = 48.1%
LJL/HL = 57.1%
2-3 elongated dorsal-fin rays.
3-6 elongated dorsal-fin rays.
Postflexion
Dorsal fin poorly pigmented.
Dorsal fin pigmented.
HL/SL = 30.6%; BD/SL = 35.8%
HL/SL = 37.3%; BD/SL = 44.7%
3 elongated dorsal-fin rays.
6 elongated dorsal-fin rays.
33 myomeres.
34 myomeres.
Metamorphosis
3 elongated dorsal-fin ray? (before reabsorption).
6 elongated dorsal-fin rays (before reabsorption).
SnL/HL = 18.3%; HL/SL = 32.0%;
SnL/HL = 23.3%; HL/SL = 39.1%;
BDT/SL = 41.9%
BDT/SL = 46.5%
33 myomeres.
34 myomeres.
Juvenile
33 vertebrae.
34 vertebrae.
Meristics (see Table 7).
IS.
Meristics (see Table 7).
BDT measured at the am
is not present in P. adspersus or
P. microps. Reared and field-
caught specimens of P. adsper-
sus lack pigmentation in the dor-
sal finfold during the first half
of their larval development, a
unique feature among described
paralichthyid larvae.
In flounders, the main change
in body shape occurs during flex-
ion, wfith an increase in body
depth and head length. This
stage is characterized by devel-
opment of skeletal structures
and by a change in swimming
and feeding (Balart 1984). After
flexion, the rate of growth of the
head, snout, and jaws is compar-
atively greater in P. microps, while the rate of increase
in body depth is greater in P. adspersus.
Preflexion larvae of P. adspersus and P. microps are
statistically indistinguishable using morphometries.
After flexion and loss of the elongated dorsal-fin rays,
separation is based mostly on morphometric character-
istics, especially SnL/HL and HL/SL. After metamor-
phosis, during the juvenile stage when all fin rays are
already developed, the adult range of meristic counts
can be used (Table 7).
Table 7
Meristics of adult Paralichthys adspersus and P. microps
of use in identifying juveniles.
Character
P. adspersus
Norman (1937)
Ginsburg (1952)
P. microps
Origin of
Between anterior margin of
Over or slightly anterior to
dorsal fin
eye and pupil (7-12 cm).
Over anterior margin of eye
or near it (20-39 cm).
center of eye.
Gill rakers
Upper
15-19
18-23
Lower
7-8 (x 7)
9-10 Chirichigno (1974)
Total
22-27 (i 25-26)
27-33
Fin rays
Dorsal
68-76
68-80
Anal
54-61
56-65
Pectoral
11-13
11-12
Acknowledgments
The authors are deeply indebted to Mr. Alfonso Silva
for allowing them to use reared larvae which were part
of a culturing experiment and to Mr. Alejandro Aron
for providing field-collected larvae. Also, comments and
editorial recommendations of two anonymous
reviewers are greatly appreciated. This research was
financed by D.G.L Universidad Catolica del Norte.
620
Fishery Bulletin 90(3), 1992
Citations
Ahlstrom, E.H., and H.G. Moser
1975 Distributional atlas of fish larvae in the California Cur-
rent region: Flatfishes, 1955 through 1960. Calif. Coop.
Oceanic Fish. Invest. Atlas 23, 207 p.
Ahlstrom, E.H., J.L. Butler, and B.Y. Sumida
1976 Pelagic stromateoid fishes (Pisces, Perciformes) of the
eastern Pacific: Kinds, distributions, and early life histories
and observations on five of these from the Northwest Atlan-
tic. Bull. Mar. Sci. 26:285-402.
Ahlstrom, E.H., K. Amaoka, D.A. Hensley. H.G. Moser, and
B.Y. Sumida
1984 Pleuronectiformes: Development. /nMoser, H.G., etal.
(eds.), Ontogeny and systematics of fishes, p. 640-670. Spec.
Publ. 1, Am. Soc. Ichthyol. Herpetol. Allen Press, Lawrence,
KS.
Bahamonde, N., and G. Pequeno
1975 Feces de Chile. Lista sistematica. Bol. Mus. Nac. Hist.
Nat. (Chile), Publ. Occas. 22:3-20.
Balart, E.
1984 Osteological development in two teleost fishes Engraulis
japoninis and Parahchthys olivaceus, and their relation to
swimming and feeding ftmction. M.Sc. thesis, Dep. Fish., Fac.
Agric, Kyoto Univ., Japan, 149 p.
Chirichigno, N.
1974 Clave para identificar los peces marinos del Peru. Inf.
Inst. Mar. Peru 44:1-387.
Ginsburg, I.
1952 Flounders of the genus Paralichthys and related genera
in America waters. Fish. Bull., U.S. 52:267-351.
Gutherz, E.J.
1970 Characteristics of some larval Bothid flatfish, and
development and distribution of larval spotfin flounder,
CycUypsettafimbriata (Bothidae). Fish. Bull, U.S. 68:261-283.
Mito, S.
1963 Pelagic fish eggs from Japanese waters. III. Percina.
Vlll. Cottina. IX. Echeneida and Pleuronectida. Jpn. J.
Ichthyol. 11:39-102.
Munoz, H.. G. Herrera. and H. Fuentes
1988 Desarrollo larval del lenguado de ojos chicos Paralichthys
microps. Rev. Biol. Mar. Valparaiso, 24(l):35-53.
Netter, J., and W, Wasserman
1974 Applied linear statistical models. Richard D. Irwin,
Homewood IL, 842 p.
Norman, J.R.
1937 Coast fishes. Part II. The Patagonian region. Discovery
Rep. 16(2):1-150.
Okiyama, M.
' 1967 Study on the early life history of a flounder, Paralichthys
olivaceus (Temminck et Schlegel). I. Descriptions of post-
larvae. Bull. Jpn. Sea Reg. Fish. Res. Lab. 17:1-12.
Potthof, T.
1984 Clearing and staining techniques. In Moser, H.G., et al.
(eds.). Ontogeny and systematics of fishes, p. 35-37. Spec.
Publ. 1, Am. Soc. Ichthyol. Herpetol. Allen Press, Lawrence,
KS.
Silva, A.
1988 Observaciones sobre el desarrollo del huevo y estadios
larvarios del lenguado, Paralichthys microps (Gunther,
1881). Rev. Latinoam. Acui. Lima (Peru) 35:19-44.
Smith, W.G.. and M.P. Fahay
1970 Description of eggs and larvae of the summer flounder,
Paralichthys dentatus. U.S. Fish. Wildl. Serv. Res. Rep. 75,
21 p.
Zuiiiga. H.
1988 Comparaci6n morfoWgica y dietaria de Paralichthys
adspersus (Steindachner, 1867) y Paralichthys microps (Gun-
ther. 1881) en Bahi'a de Coquimbo. Mar. biol. thesis, Univ.
cat6lica del Norte, Coquimbo, Chile, 144 p.
An estimate of the tag -reporting
rate of commercial slirlmpers
in two Texas bays
R. Page Campbell
Terry J. Cody
Texas Parks and Wildlife Department
100 Navigation Circle, Rockport, Texas 78382
C.E. Bryan
Gary C. Matlock
Maury F. Osborn
Albert W. Green
Texas Parks and Wildlife Department
4200 Smith Scfiool Road, Austin, Texas 78744
Tag return rates are used to esti-
mate exploitation rates for many
animal species including penaeid
shrimp. To avoid systematic under-
estimation of exploitation, the num-
ber of tagged animals recaptured
but not reported must be reliably
estimated (Paulik 1963, Youngs 1972,
Seber 1973). Some investigators
have offered rewards for tags to in-
crease the tag return rate, but have
incorrectly assumed that all or near-
ly all harvested tagged animals were
reported (Kutkuhn 1966) or the rate
of non-reporting remained the same
throughout the experiment (Klima
1974, Kutkuhn 1966). Studies to
measure the reporting rate of com-
mercially-caught shrimp were con-
ducted by Klima (1974) and Johnson
(1981). The numbers of shrimp placed
in both studies were small (n 71 and
20, respectively) and return rates
differed markedly (82% and 10%,
respectively). One drawback of these
studies is that tagged shrimp were
placed into the catch at shrimp
houses or in the final processing
stages, and not on the vessel dur-
ing shrimping operations. There-
fore, return rates during fishing
operations were not measured.
Accurate reporting rates for re-
covered tags are essential for the
determination of fishing mortality
rates. The objective of the present
study was to determine reporting
rates of tagged shrimp captured
during regular shrimping opera-
tions. To that end, tagged shrimp
were surreptitiously placed in un-
culled catches. The reporting rates
determined in this study are in-
tended for use in correcting fishing-
mortality estimates generated from
a tagging program conducted dur-
ing the same period.
Materials and methods
Texas Parks and Wildlife Depart-
ment (TPWD) personnel placed
tagged shrimp in the catch aboard
Galveston and Aransas Bays' com-
mercial bay and bait shrimp boats,
May-November 1984. TPWD per-
sonnel and game wardens boarded
shrimp vessels during bay shrimp-
ing operations. While a game war-
den distracted the crew by checking
licenses, other TPWD personnel
placed a single tagged shrimp in un-
culled catches (on deck) or in a live
bait box. To conceal surreptitious
placement of shrimp, 20 individuals
of the target species Penaeus aztec-
us or P. setiferus were measured to
the nearest mm total length (TL) on
each vessel. A total of 219 shrimp
(115 brown and 104 white) were
surreptitiously placed aboard ves-
sels in Aransas (n 125) and Galves-
ton (n 94) Bay systems during the
study period. No more than 12
shrimp were surreptitiously placed
in each bay system in any one week.
Each tag was a uniquely-num-
bered black vinyl streamer (95 mm
long X 4 mm wide) tapered at each
end (Klima et al. 1987). Each tag
was inserted between the second
and third abdominal segments of
the shrimp. Shrimp in Aransas Bay
were measured (TL) prior to place-
ment and after being returned by
the fisherman. Since lengths were
not required in the original study,
measurements were not recorded in
Galveston Bay. Lengths of shrimp
placed in the catches and lengths of
shrimp returned by fishermen were
compared using student's ^test.
Also, length frequencies of mea-
sured shrimp on commercial boats
(n 2402) and of shrimp surrep-
titiously placed (n 105) aboard
boats were compared visually using
length-frequency histograms.
As part of a larger bay shrimp-
tagging program conducted jointly
by the TPWD and the National
Marine Fisheries Service (NMFS),
rewards were offered for tag re-
turns. The program was promoted
by distribution of posters to area
shrimp dealers and through news-
paper articles. No information was
provided to the public concerning
the surreptitious tagging activity.
Reporting rates (n reported/n
placed, expressed as percent) were
estimated for each species and bay
system. Reporting rates and con-
fidence intervals were estimated for
the two bay systems combined.
Reporting rates between species
and between bay systems were
compared using a Chi-square test
(Sokal and Rohlf 1981).
Manuscript accepted 11 June 1992.
Fishery Bulletin. U.S. 90:621-624 (1992).
621
622
Fishery Bulletin 90(3), 1992
Results
Overall, 16% (95% CI, 11-21%) of 219 tagged shrimp
were returned (Table 1). The return rate of 21% for
brown shrimp was greater than the 11% for white
shrimp (x^ 4.415, Idf, P<0.05). Reporting rates did
not differ between bay systems for brown shrimp (x^
2.081, Idf, P>0.05) or white shrimp (x^ 1.059, Idf,
P>0.05). Reporting rates for the two species of shrimp
were similar in Galveston Bay (x^ 0.001, Idf, P>0.05);
in Aransas Bay, reporting rates were greater for brown
shrimp (x^ 6.890, Idf, P<0.05). Sixty-eight percent of
shrimp returned were reported found on the same day
as placement.
Mean lengths of placed and returned brown shrimp
from Aransas Bay were similar (t - 0.48, P>0.05) (Fig.
1), while the mean length of placed white shrimp was
smaller than the mean length of those returned
(t -4.01, P<0.05). Placed brown shrimp were similar
in length to those measured from the unculled catches
on commercial shrimp boats (Fig. 2) whereas placed
white shrimp were clearly smaller than those measured
on commercial boats.
Discussion
Tag reporting rates for bay-caught shrimp have been
reported by Klima (1974) and Johnson (1981). Tag
reporting rates presented in this study are more precise
because sample sizes were larger than in previous
studies. Moreover, tag return rates in this study are
more realistic because the tagged shrimp were placed
in the catch before any processing occurred, rather
than at dockside during the final processing stages.
Table 1
Number and percent
of tagged Penaeus shrimp surreptitiously 1
placed on shrimp boat decks which
were found
and returned
to TPWD.
Returned
Bay
No.
Species
system
tagged
n %
P. aztecus
Galveston
43
6 14
(brown shrimp)
Aransas
72
18 25
Total
115
24 21
P. setiferus
Galveston
51
7 14
(white shrimp)
Aransas
53
4 8
Total
104
11 11
Combined
Galveston
94
13 14
species
Aransas
125
22 18
Total
219
35 16
The detection rate, and thus the reporting rate, of
tagged shrimp in unculled catches may be influenced
by size of tagged shrimp relative to size of other shrimp
in the catch and by overall volume of catch being pro-
cessed. In the fall shrimping season (15 August-15
December), there are no restrictions on the amount of
shrimp that can be retained. During 15 August-31
October, when white shrimp dominate the catch, the
minimum shrimp count size is 50 (heads-on) per pound
in major Texas bays (State of Texas 1987-88). Thus,
commercial fishermen selectively retain larger shrimp
during this interval. Since the TPWD gear used to
collect white shrimp for tagging was relatively non-
selective, surreptitiously-placed shrimp in Aransas Bay
were smaller than those in the catch in which they were
placed. In contrast, brown shrimp dominated the catch
in summer when there was no count size restriction,
and thus placed shrimp were similar in size to those
in the commercial catch. Because the placed white
Brown Shrimp
Placed (N-56)
Returned (N>17}
65 70 75 80 85 90 95 100105110 115120125130135140145150
Total Length (mm)
White Shrimp
_ 8
6
Placed (N-4g)
Relumed (N>4)
65 70 75 BO [>5 aO aS) 100 105 110 115 120 125 130 135 140 145 160
Total Length (mm)
Figure 1
Length-frequency (TL) of surreptitiously-placed brown and
white shrimp, and of returned brown and white shrimp from
Aransas Bay, May-November 1984.
NOTE Campbell et al ' Tag-reporting rates of commercial shrimpers in Texas bays
623
Brown Shrimp
f?300
0) 200
50 60 70 80 90 100 110 120 130 140 150 160 170 180
Total Length (mm)
White Shrimp
50 60 70 80 90 100 110 120 130 140 150 160 170 180
Total Length (mm)
Figure 2
Length-frequency histograms of brown shrimp and white
shrimp measured in commercial catches, and of tagged shrimp
surreptitiously placed aboard boats in Aransas Bay, May-
November 1984.
to promote the return of tags. Rawstron (1971) deter-
mined that some reward tags in his fish-tagging study
were not returned, but beheved that this number was
negligible. Likewise, Kutkuhn (1966) assumed low non-
reporting rates for reward tags. Published estimates
of tag-return rates for fish generally have ranged
between 55 and 65%, with rewards. Green et al. (1983)
reported much lower return rates by saltwater recrea-
tional anglers (29%) than had previously been esti-
mated, and that rates differed among species and
areas. Therefore, even with rewards, complete or high
reporting rates cannot be assured.
Previous studies have relied on public-information
dissemination plans to achieve high reporting rates of
reward and non-reward tags. Matlock (1981) found that
83% (n 102) of the anglers not reporting tags in their
catch knew about TPWD tagging programs, and that
78% of these anglers failed to find the tag. This sug-
gests that public-information programs cannot assure
high reporting rates. Even if fishermen are aware of
tagging programs, they may not report recaptured tags
if these programs have continued over a long period.
The shrimp fishery in Texas had been subjected to fre-
quent tagging experiments during the previous 10
years, and the shrimp fishermen's enthusiasm for
reporting tags may have decreased. However, there
are no data to examine this possibility.
Tag-return rates can be affected by many factors.
Each tagging study that depends on volunteer tag
returns would be enhanced by a concurrent estimate
of non-reporting rates. This would improve estimates
developed from returned tags. For example, during the
tagging program conducted during the same period as
this study, there were 2% of 25,870 released tagged
shrimp returned (Peng Chai, TPWD, Austin, pers.
commun.). If the reporting rate had been assumed to
be 100% rather than the observed 19%, fishing mor-
tality would be overestimated about five times.
shrimp were smaller than those in the commercial
catch, they may have been more difficult to detect and
hence, were reported at a lower rate than if they had
been similar in size to those in the commercial catch.
The overall reporting rate with these white shrimp ex-
cluded was 19% (95% CI, 13-26%) which is similar to
the reporting rate for brown shrimp (21%). Tagged and
untagged brown shrimp sizes were similar.
Complete return of tags cannot be assumed even if
rewards are offered. All tags used in the present study
were potential reward tags ($50-500) inserted into the
shrimp and placed into unculled shrimp catches;
however, only 19% of these shrimp were reported. Past
studies have relied on the use of monetary incentives
Acknowledgments
We would like to thank each member of the TPWD
Harvest and Resource Monitoring programs, without
whose assistance this task could not have been accom-
plished. Thanks also go to the Texas Parks and Wildlife
Department Law Enforcement officers who accom-
panied personnel into the bays and assisted in place-
ment of tagged shrimp on the commercial boats, and
also to the National Marine Fisheries Service for their
support. This study was funded by the Texas Parks and
Wildlife Department and by NMFS under P.L. 88-309
(Proj. 2-400-R).
624
Fishery Bulletin 90(3), 1992
Citations
Green, A., G.C. Matlock, and J.E. Weaver
1983 A method for directly estimating the tag-reporting rate
of anglers. Trans. Am. Fish. Soc. 112:412-415.
Johnson, M.F.
1981 Shrimp mark-release investigations. Vol. II. In Jackson,
W.B., and E.P. Wilkens (eds.), Shrimp and redfish studies;
Bryan Mound brine disposal site off Freeport, Texas,
1979-1981. NOAA Tech. Memo. NMFS-SEFC-66, Southeast
Fish. Sci. Cent, Miami, 110 p.
Klima, E.F.
1974 A white shrimp mark-recapture study. Trans. Am. Fish.
Soc. 103:107-113.
Klima, E.F., Refugion Gmo. Castro Melandez, N. Baxter,
F.J. Patella, T.J. Cody, and L.F. Sullivan
1987 "MEXUS-Gulf shrimp research. 1978-1984." /« Rich-
ards, W.J., and R. Juhl (eds.), The Cooperative MEXUS-Gulf
research program: Summary reports for 1977-85, p.
21-30. Mar. Fish. Rev. 49(1).
Kutkuhn, J.H.
1966 Dynamics of a penaeid shrimp population and manage-
ment imphcations. Fish. Bull., U.S. 65:313-338.
Matlock, G.C.
1981 Nonreporting of recaptured tagged fish by saltwater
recreational boat anglers in Texas. Trans. Am. Fish. Soc.
110:90-92.
Paulik, G.J.
1963 Detection of incomplete reporting of tags. Int. Comm.
Northwest Atl. Fish. Spec. Publ. 4:238-247.
Rawstron, R.R.
1971 Nonreporting of tagged white catfish, largemouth bass,
and bluegills by anglers at Folsom Lake, California. Calif. Fish
Game 57:246-252.
Seber, G.A.F.
1973 The estimation of animal abundance and related param-
eters. Griffin, London, 654 p.
Sokal, R.R., and F.J. Rohlf
1981 Biometry, W.H. Freeman. San Franc, 859 p.
State of Texas
1987-88 Texas Parks and Wildlife Laws. West Publ. Co., St.
Paul. MN, 455 p.
Youngs, W.D.
1972 Estimation of natural fishing mortality rates from tag
recaptures. Trans. Am. Fish. Soc. 101:542-545.
Power to detect linear trends In
dolphin abundance: Estimates from
tuna-vessel observer data, 1975-89
Elizabeth F. Edwards
Peter C. Perkins
Southwest Fisheries Science Center, National Marine Fisheries Service, NOAA
P.O. Box 271, La Jolla, California 92038-0271
Trends in abundance of dolphin
stocks affected by the tuna purse-
seine fishery in the eastern tropical
Pacific Ocean (ETP) are of intense
interest to a number of organiza-
tions concerned about the stocks'
continued survival (Hammond and
Laake 1983, Gerrodette 1987, Holt
et al. 1987, Buckland and Anga-
nuzzi 1988, Anganuzzi and Buck-
land 1989, Anganuzzi et al. 1991).
The most straightforward method
for estimating such trends is linear-
regression analysis of relative abun-
dance indices across time (e.g.,
Anganuzzi and Buckland 1989).
Such abundance indices can be
derived from data collected by ob-
servers aboard the tuna vessels
(Buckland and Anganuzzi 1988,
Anganuzzi and Buckland 1989,
Anganuzzi et al. 1991). Linear
trends in abundance over successive
5-year periods have been reported
by Buckland and Anganuzzi (1988),
Anganuzzi and Buckland (1989),
and Anganuzzi et al. (1991).
Power analysis provides a method
to quantify the probability of not
detecting low rates of change in
abundance over a specified time-
period. It also provides a method, in
cases where no statistically-signifi-
cant trends are apparent, for deter-
mining the steepness of change
necessary for its statistical detec-
tion given observed variability in
the data, i.e., detectable trend (Ger-
rodette 1987, Peterman 1990). We
use power analysis here to assess
the efficacy of weighted linear-
regression analysis for estimating
linear trends in abundance of eight
stocks of ETP dolphins. While it is
instructive to evaluate the power of
conclusions about observed trends,
it is perhaps even more important
to determine the magnitude of
change required for detection of a
trend, given observed variability in
the dolphin abundance estimates.
Therefore, we also calculate detect-
able trends, in addition to power of
observed trends.
We present here estimates of
observed trend, power to detect
trends, and detectable trends for
eight stocks of ETP dolphins, over
time-series of 5, 8, and 10-years,
assuming a two-sided hypothesis
with a = 0.10, using the noncentral
^distribution for the alternative
hypothesis. Detectable trends were
estimated assuming Type I (a) and
Type H (/?) error levels equal 0.10.
We estimated power and detectable
trends for all three sets of time-
series to determine how much im-
provement might be expected by
increasing the number of years in-
cluded in the trend estimate. We did
not include longer time-series, as it
is unlikely that even a population
with reproductive and individual
growth rates as relatively slow as
ETP dolphins would follow a linear
trend for more than a decade, if that
long.
Methods
Relative abundance indices and
their associated bootstrap standard
errors for eight stocks of ETP
dolphins during the years 1975-89
(Table 1, Fig. 1) formed the data-
base for the regression analyses
presented here. Indices and stan-
dard errors for 1975-87 were taken
from Anganuzzi and Buckland
(1989), and for 1988 and 1989, from
Anganuzzi et al. (1991). The eight
dolphin stocks included northern
offshore and southern offshore
stocks of the pantropical spotted
dolphin SteneUa attenuata, the east-
ern spinner dolphin SteneUa longi-
rostris orientalis, northern and
southern stocks of whitebelly spin-
ner dolphin (hybrid/intergrades be-
tween SteneUa I. orientalis and Ste-
neUa I. longirostris [Perrin 1990]),
and northern, central and southern
stocks of the common dolphin
Delphinus delphis.
Observed trends
We estimated linear trends in rela-
tive abundance for each of the eight
stocks over sequential series of 5,
8, and 10 years, using standard
weighted least-squares regression
(Wilkinson 1989). The slope of the
regression (b) estimates the trend in
abundance. The estimated standard
error of the estimated trend (st,)
indicates the variability associated
with the trend estimate. Weights
were the reciprocal of the square of
the bootstrap standard errors
(Buckland and Anganuzzi 1988). We
eliminated the estimate for 1983
from all analyses, because the pres-
ence of a very strong El Niiio that
year caused biologically unreason-
able estimates of abundance for
many of the stocks, in particular for
northern offshore spotted dolphins,
the stock affected by the fishery in
greatest numbers (Buckland and
Anganuzzi 1988, Anganuzzi et al.
1991). In the absence of any objec-
tive criteria for choosing which
stocks were (or were not) affected
by the El Nino, we elected to treat
Manuscript accepted 1 June 1992.
Fishery Bulletin, U.S. 90:625-631 (1992).
625
626
Fishery Bulletin 90(3|. 1992
Table 1
Estimated abundances (first row) and bootstrapped standard |
errors (second row) for eight stocks of dolphins
affected by
the tuna purse
-seine
fishery
in the eastern tropical Pacific
Ocear
1. Values
are in thousands.
Data for 1975-87
from
Buckland and
Anganuzzi 1988;
lata for 1988-89
from
Anganuzzi and Buckland 1989. Dashes indicate no values
reported for that year. Asterisks indicate values from 1983
omitted as anomalous. SOPS boundaries (Anganuzzi and Buck-
land
1989) used in
estimating southern
offshore spotted |
dolph
ns. See text for scientific names.
Dolphin
stocks
1
2
3
4
5
6
7
8
1975
3949
599
490
122
404
385
996
—
197
181
—
51
138
227
1976
4253
574
535
1205
115
344
287
586
908
212
190
293
62
130
80
223
1977
3828
924
514
588
203
637
466
229
751
332
152
163
87
200
90
162
1978
3212
584
395
613
65
358
329
—
543
302
124
150
47
126
105
—
1979
2950
1040
428
366
—
650
644
—
559
394
202
183
-
248
287
-
1980
3335
260
447
342
124
512
251
230
582
168
112
99
86
191
85
139
1981
2536
199
255
694
124
513
111
435
443
83
165
287
80
330
35
142
1982
2550
591
202
416
100
—
232
103
557
180
91
132
47
-
101
90
1983
*
*
*
•
•
•
*
*
*
•
*
«
*
*
*
*
1984
2158
244
340
253
182
—
71
—
362
115
85
72
71
-
91
-
1985
2884
238
586
648
247
—
265
249
352
68
124
128
64
-
105
318
1986
3165
154
584
451
—
475
169
—
302
60
108
95
—
237
50
—
1987
2953
—
384
650
—
304
60
—
293
-
87
105
-
123
18
-
1988
2689
79
717
484
88
323
241
253
326
30
110
92
33
93
50
100
1989
2910
560
389
515
190
243
125
179
275
140
71 78 69
•e spotted dolphin
107
24
47
1 = northern offshoi
2 = southern offshore spotted dolphin
3 = eastern spinner
dolphin
4 = northern
whitebelly spinner d
olphin
5 = southern whitebelly spinner d
olphin
6 = northern
common dolph
in
7 = central common
dolphin
8 = southern
common dolph
in
all eight stocks similarly by eliminating the 1983
estimate.
For each series, we calculated regressions using as
many data points as existed for each species for that
number of years. In some cases, this resulted in as few
as three data points contributing to the regression.
Because we omitted data from 1983, the 10-year series
contained at most nine data points. Because some years
were omitted or were missing abundance indices, not
all year-series comprised strictly consecutive x-values
(year values).
We reexpressed the slope estimate of trend (b) in
terms of a change parameter r, where
r = b/Ai
and Ai (estimated abundance in first year of series) is
calculated from the estimated slope and intercept for
each year-series. For these linear regressions, the
parameter r expresses the annual rate of change as a
fraction of the estimated initial abundance (Gerrodette
1987). Linear regressions were calculated only for
series with at least three data points.
Power
We estimated power of statistical conclusions about the
significance of each slope by assuming a two-sided
alternative hypothesis and using the non-central t (net)
distribution. In all cases, we assumed error levels a =
P = 0.10. We used a two-sided hypothesis test to be con-
sistent with earlier estimates of 5-year trends in abun-
dance (Buckland and Anganuzzi 1988, Anganuzzi and
Buckland 1989, Anganuzzi et al. 1991).
To calculate power using the net distribution, we
utilized a series of programs (available upon request)
designed to return power estimates as a function of
three input variables:
K.df = normal t statistic given a level of a
and degrees of freedom,
IDF = degrees of freedom, and
6 = b/Sb.
Degrees of freedom were n-2 where n is the number
of years for which abundance estimates existed in a
series. Values for b and s^ were calculated from the
weighted linear regressions.
6 is the offset of the alternative distribution (the net)
standardized by the standard error of the offset. In all
cases, we assumed as the alternative distribution the
observed trend for a series. 6 is thus the distance, ex-
pressed as standard deviation units, between the mean
of the null distribution (taken here to be zero slope) and
the mean of the alternative distribution (the slope
estimated from regression of the data).
NOTE Edwards and Perkins: Detecting linear trends in dolphin abundance
627
Z
I
a.
_i
O
Q
LL
o
CO
<
O
Northern Offshore Spotted
1976 1978 1980 1982 1984 1986 1988
8.
Southern Offshore
Spotted
1976 1978 1980 1982 1984 1986 1988
Eastern Spinner
1976 1978 1980 1982 1984 1986 1988
Northern Whitebelly Spinner
1976 1978 1980 1982 1984 1986
Southern Whitebeliy/Spinner
/
\
1976 1978 1980 1982 1984 1986 1988
Northern Common
1976 1978 1980 1982 1984 1986 1988
Central Common
1976 1978 1980 1982 1984 1986 1988
Southern Common
\^
1976 1978 1980 1982 1984 1986 1988
YEAR
Figure 1
Estimated abundances (dashed line) of dolphins bounded by 1 bootstrap standard error (solid lines). Years for which
no estimates were reported are omitted; estimates for 1983 are omitted as anomalous.
628
Fishery Bulletin 90(3), 1992
CC
LU
o
a.
+ 5yr A8yr D lOyr
Figure 2
Power of conclusions about lack of statistical significance for estimated 5, 8, and 10-year trends in dolphin abun-
dance. Power calculated for two-tailed noncentral t, assuming q = 0.10. Solid lines indicate maximum (7df) and
minimum (Idf) power envelope, d is the noncentral t parameter.
NOTE Edwards and Perkins: Detecting linear trends in dolphin abundance
629
Detectable trends
We estimated detectable trends (rj) for all data-series.
All estimates of r^ assume error levels a = p = 0.lO.
Detectable trends were estimated by determining the
value of Delta (do.g) that returns a power value of 0.90
from the net algorithm. As before, input value for IDF
was 71 - 2, and for si, was the value estimated from the
weighted regression. Then the value of b generating
the desired power level (bnct) is
bnct = f^o.g/Sb
and the detectable initial trend per year is
Td = bnct/Ai.
Results
Observed trends
The majority (151/192; 79%) of the series showed no
significant trend (specific data available from the
authors). Of those that did, most showed decreases
prior to the mid-1980s and no consistent trends since.
Where population-abundance indices changed relatively
regularly over time, successively longer time-series re-
tained the same general patterns as found in shorter
series. For example, observed trends were significantly
negative for northern offshore spotted dolphin during
the 5-year series 75/79 and 77/81, the 8-year series
75/83, 76/83, and 77/84, and the 10-year series 75/84
and 76/85. Similarly, 5-year negative trends were also
reflected in 8- and 10-year series for southern off-
shore spotted dolphin, eastern spinner dolphin, north-
ern whitebelly spinner dolphin, and central common
dolphin.
Data were so sparse and variable for southern white-
belly spinner dolphin and southern common dolphin
that little can be said about trends in these stocks.
Northern common dolphin were the only species for
which trends may have switched during the period of
investigation (from negative during earlier years, to
positive more recently); but it is obvious that here, as
in the other series, the pattern in trend estimates is
simply a function of the length of the series selected
and its placement in time.
Power
Power to reject a false null hypothesis increases with
increases in either or both of series length (as degrees
of freedom increase) or 6 (offset) (Fig. 2), but for TVOD
the increases generally were not sufficient to be of
practical use. Where no significant trends (slopes) were
found, power to detect a false null hypothesis was low,
averaging 20-30% in most cases and never exceeding
60%. Power for each test was small because the alter-
native hypothesis for these power calculations was
taken to be the observed slope, which was usually fair-
ly small, and also because scatter around the regres-
sion line tended to be large. Therefore the null and
alternative distributions overlapped considerably. The
low power of these tests simply means that if the true
slope equaled the observed slope, the power to distin-
guish the true slope from a slope of zero (i.e., no change
in abundance) would be quite small in most cases.
Detectable trends
The range of detectable trends decreased rapidly with
increasing series-length in all cases (Fig. 3), as this in-
creases the degrees of freedom (number of data points).
The decrease is misleading in most cases, however.
Although the improvement in ability to detect smaller
trends with longer time-series appears dramatic, in
most cases even the smallest detectable trends are still
much too large to be of use.
Even with as many as 10 years of data in a series,
linear trends less than about 10% per year could
be detected consistently only for northern offshore
spotted dolphin. For all other stocks, trends of at least
15-20% per year would be required to produce a signifi-
cant result (Fig. 2). Series lengths would have to be
such that populations more than doubled or decreased
to zero in order for the change to be statistically detect-
able. This would require series lengths of at least 10
years.
In many cases, where significant trends were found,
these trends were of lesser magnitude than the esti-
mated detectable trend. This occurs because the esti-
mated detectable trend is the expected value of the
alternative distribution. Any trend value which falls
below this expected value, but which also falls above
the Type-I error limit for the null distribution, will be
assumed significantly different from the null even
though the trend could actually belong to either
distribution. For example, if the Type-I error limit for
the null distribution occurs at a trend value of 0.75 (i.e.,
if the cut-off point for values assumed to belong to the
null distribution is 0.75), and the expected value (i.e.,
the mean) for the alternative distribution falls at 0.85,
any trend value within the range 0.75-0.85 will be
assigned to the alternative distribution even though it
is smaller than the expected value of the alternative
distribution. In practice for the ETP data, this effect
is unimportant compared with the overall problem of
high variability obscuring the possibility of detecting
managerially-relevant trends in abundance (i.e., the
630
Fishery Bulletin 90(3), 1992
<n -
.
•
_
•
Northern Offshore Spotted
in
Southern Whitebelly Spinner
in
CM _
d
•
■
o
•
-i
in
°'
s .
o
■
1
•
■
!
■
1
■
o
lO
d
•
•
1
•
I
•
•
1
•
Syr
8 yr lOyr
Syr
8yr
lOy
en -
Southern Offshore Spotted
o
a
•
Northern Common
DC
>
og -
eo
d
to
d
■
■
•
•
N«^
.
1
■
Q
Z
LJU
DC
h-
LU
_l
CD
^
O
LU
1-
LU
Q
1
•
•
•
•
•
1 1
d
d
o
■
•
•
•
•
■
■
Syr
8 yr 10 yr
o
Syr
8yr
lOyr
O
CO
O
CD
o
d
o
•
•
1
•
•
•
•
•
Eastern Spinner
•
!
i
o
o
in
o
■
1
1
•
•
•
•
Central Common
•
1
1
1
Syr
Byr 10 yr
Syr
8yr
lOyr
■
<D -
■
in
o
•
•
Northern Whitebelly Spinner
in -
•
Southern Common
d
•
•
•
•
•
•
1 1
• •
CO -
OJ -
o -
•
•
•
•
•
•
•
:
t
1
•
1
Syr
8yr 10 yr 5yr
8yr
lOyr
SERIES LENGTH
Figure 3
Estimated detectable trends for successive 5, 8, and 10-year
series of tuna vessel observer data.
NOTE Edwards and Perkins Detecting linear trends in dolphin abundance
631
spread and subsequent overlap between null and alter-
native distributions is so great that differences between
the cut-off point in the null and the mean of the alter-
native is effectively inconsequential).
Discussion
Given the observed variability in TVOD, it does not ap-
pear that linear trends in dolphin abundance estimated
from these data with these techniques will be detect-
able at any levels practical for management purposes,
except possibly for the stock of northern spotted dol-
phin. Even for this stock, survey series at least a decade
in length are required.
Our estimates of 5-year trends agree with previous-
ly published estimates derived by the same methods,
where 1983 has been omitted from the analyses (An-
ganuzzi and Buckland 1989). Our results imply that for
time-periods as short as 5 years, considerably larger
trends than those observed would be necessary to pro-
duce estimates of significant change with any reason-
able power. Except for northern offshore spotted
dolphin, this applies also to all other stocks, even for
the maximum time-series tested (10 years in this study).
It appears that other methods must be found to deter-
mine whether trends truly exist in dolphin abundance
in the ETP. For management purposes, longer time-
series must be monitored, for which linearity cannot
be assumed. Other regression procedures making
greater use of the precision estimates (standard errors)
of the indices could have more power, but for linear
analyses, at least, it is uncertain whether the increase
in power could overcome the inherent variabiity and
probable nonlinearities in the data.
A more effective approach to estimating trends in
dolphin abundance is probably represented by the
sophisticated smoothing method applied recently to
these data by Buckland et al. (1992). The method
reduces the relatively-scattered abundance estimates
to smoothly-changing estimates of abundance, but with
the advantage of producing confidence limits about the
smoothed trend and generating a more biologically-
reasonable result (abundance of natural populations
rarely changes linearly). However, simulation experi-
ments will be required to determine the circumstances
under which the smoothed trends do, or do not, reflect
accurately the true underlying dynamics of the stocks.
Such simulations are currently underway, but results
are as yet unavailable (Alejandro Anganuzzi, Inter-Am.
Trop. Tuna Comm., La Jolla, pers. commun., July
1991).
Regardless of the results of the tests of various
smoothing methods, it appears fruitless, based on the
results presented here, to use linear-regression tech-
niques to estimate trends in abundance of dolphin
stocks in the ETP, even for periods as long as a decade.
The power to detect ecologically (or managerially) rele-
vant trends, given the observed variability in the data,
is simply not sufficient.
Future efforts should, as suggested by Buckland et
al. (1992), focus on developing or applying robust,
curvilinear smoothing techniques that are reasonably
responsive to the underlying processes or mechanisms
controlling actual changes in dolphin abundance.
Acknowledgments
This study has benefited greatly from technical assis-
tance by Cheryl Glick and extensive helpful discussions
with Tim Gerrodette and Doug DeMaster.
Citations
Anganuzzi, A. A., and S.T. Buckland
1989 Reducing bias in estimated trends from dolphin abun-
dance indices derived from tuna vessel data. Rep. Int. Whal-
ing Comm. 39:323-334.
Anganuzzi, A. A., S.T. Buckland. and K.L. Cattanach
1991 Relative abundance of dolphins associated with tuna in
the eastern tropical Pacific, estimated from tuna vessel
sightings data for 1988 and 1989. Rep. Int. Whaling Comm.
41:497-506.
Buckland. S.T., and A. A. Anganuzzi
1988 Estimated trends in abundance of dolphins associated
with tuna in the eastern tropical Pacific. Rep. Int. Whaling
Comm. 38:411-437.
Buckland, S.T., K.L. Cattanach, and A. A. Anganuzzi
1991 Estimating trends in abundance of dolphins associated
with tuna in the eastern tropical Pacific, using sighting data
collected on commercial tuna vessels. Fish. Bull.. U.S. 90:
1-12.
Gerrodette, T.
1987 A power analysis for detecting trends. Ecology 68:
1364-1372.
Hammond, P.S., and J.L. Laake
1983 Trends in abundance of dolphins involved in the purse-
seine fishery for tuna in the eastern tropical Pacific Ocean,
1977-1981. Rep. Int. Whaling Comm. 33:565-588.
Holt, R.S., T. Gerrodette. and J.B. Cologne
1987 Research vessel survey design for monitoring dolphin
abundance in the eastern tropical Pacific Ocean. Fish. Bull.,
U.S. 85:435-446.
Perrin, W.F.
1990 Subspecies oiStenella longirostris (Mammalia: Cetacea:
Delphinidae). Proc. Biol. Soc. Wash. 103(2):453-463.
Peterman, R.M.
1990 The importance of reporting statistical power: The forest
decline and acidic deposition example. Ecology 71:2024-2027.
Wilkinson, L.
1989 SYSTAT: The system for statistics. Systat, Inc.,
Evanston.
Order Processing Code:
*5178
Superintendent of Documents Subscriptions Order Form
Charge your order.
It's Easyl
n YES, enter my subscription as follows: T„ j^ ^^^ ^^^^^ (202) 512-2233
subscriptions to FISHERY BULLETIN (FB) for $24.00 per year ($30.00 foreign).
The total cost of my order is $_
_. Price includes regular domestic postage and handling and is subject to change.
(Company or Personal Name)
(Please type or print)
(Additional address/attention line)
Please Choose Method of Payment:
I I Check Payable to the Superintendent of Documents
n GPO Deposit Account | | I I I I I l-fl
(Street address)
I I VISA or MasterCard Account
(City, State, ZIP Code)
(Daytime phone including area code)
(Purchase Order No.) ^^ ^^
May we make your name/address available to other mailers? I I I 1
Thank you for
(Credit card expiration date) , ,
^ your order!
(Authorizing Signature) "'9'
Mail To: New Orders, Superintendent of Documents
RO. Box 371954, Pittsburgh, PA 15250-7954
The National Marine Fisheries Service (NMFS) does not approve, recommend or en-
dorse any proprietary product or proprietary material mentioned in this publication.
No reference shall be made to NMFS, or to this publication furnished by NMFS, in
any advertising or sales promotion which would indicate or imply that NMFS ap-
proves, recommends or endorses any proprietary product or proprietary material
mentioned herein, or which has as its purpose an intent to cause directly or indirect-
ly the advertised product to be us^-d or purchased because of this NMFS publication.
Fisheiy Bulletin
Guidelines for Contributors
Form of manuscript
The document should be in the following
sequence: Title Page, Abstract (not re-
quired for Note), Tfext, Acknowledgments,
Citations, Text footnotes. Appendices,
Tables, Figure legends, and Figures.
Title page should include authors' full
names and mailing addresses and the
senior author's telephone and FAX
numbers.
Abstract Not to exceed one double-
spaced typed page. Should indude a
sentence or two explaining to the gen-
eral reader why the research was under-
taken and why the results should be
viewed as important. Abstract should
convey the main point of the paper and
outline the results or conclusions. No
footnotes or references.
Text A brief introduction should por-
tray the broad significance of the paper.
The entire text should be intelligible to
readers from different disciplines. All
technical terms should be defined, as
well as all abbreviations, acronyms, and
symbols in text, equations, or formulae.
Abbreviate units of measure only when
used with numerals or in tables and
figures to conserve space. Measure-
ments should be expressed in metric
units, with other equivalent units given
in parentheses. Follow the U.S. Govern-
ment Printing Office Style Manual,
1984 ed., and the CBE Style Manual,
5th ed. Fishery and invertebrate no-
menclature should follow the American
Fisheries Society Special Publication 12
(for fishes), 16 (for moUusks), and i7 (for
decapod crustaceans).
Text footnotes should be numbered
in Arabic numerals and typed on a sep-
arate sheet from the text. Footnotes are
not used for reference material or per-
sonal communications, but rather to
explain or define terms in the text and
for contribution numbers on the title
page.
Informal sources Personal communi-
cations, unpubhshed data, and untitled
manuscripts in preparation are noted
parenthetically in the text (full name,
affiliation, brief address including zip
code, and month and year when appro-
priate).
Acknowledgments Gather all ac-
knowledgments into a brief statement
at the end of the text. Give credit only
for exceptional contributions and not to
those whose contributions are part of
their normal duties.
Citations All titled sources should be
listed in the Citations section, including
unpublished and processed material. In
text, cite as Smith and Jones (1977) or
(Smith and Jones 1977); if more than
one citation, list chronologically (Smith
1936, Jones 1975, Doe In press). All
sources cited in the text should be listed
alphabetically by the senior authors'
surnames under the heading CITA-
TIONS. Abbreviations of periodicals
and serials should conform to Serial
Sources for the BIOSIS Data Base'".
Indicate whether sources are in a langu-
age other than English. For informal
literature, include address of author or
publisher. Authors are responsible for
the accuracy of all citations.
Tables should supplement, not dupli-
cate, the text. Each table should be
numbered and cited consecutively, with
headings short but amply descriptive so
that the reader need not refer to the
text. For values less than one, zeros
should precede all decimal points. In-
dicate units of measure in column head-
ings; do not deviate from the unit of
measure within a column. Table foot-
notes should be noted consecutively in
Roman letters across the page from left
to right and then down. Since all tables
are typeset, they need not be submitted
camera-ready.
Figures Photographs and line draw-
ings should be of professional quality-
clear and concise— and reducible to 42
picas for full-page width or to 20 picas
for a single-column width, and to a max-
imum 55 picas high. All graphic ele-
ments in illustrations must be propor-
tioned to insure legibility when reduced
to fit the page format. Line weight and
lettering should be sharp and even. Let-
tering should be upper and lower case.
and vertical lettering should be avoided
whenever possible (except for vertical,
y, axis). Zeros should precede all deci-
mal points for values less than one. Re-
productions of line artwork are accepted
in the form of high-quality photographic
prints from negatives or photomechan-
ical transfer (PMT). Halftones should be
sharply focused with good contrast.
Micron rules should be inserted on elec-
tron micrographs, even when magnifi-
cation is included in the figure legend.
There should be clear distinction be-
tween identifying letters (press-on or
overlay) and background of photograph.
Label each figure in pencil on the back.
Send only xerox copies of figures to the
Scientific Editor; originals or photo-
graphic prints will be requested later
when the manuscript is accepted for
publication.
Copyright Government publications
are in the public domain, i.e., they are
not protected by copyright.
Submission of manuscript
Disks Authors are encouraged to re-
tain manuscripts on word-processing
storage media (diskettes, floppy disks)
and submit a double-spaced hardcopy
run from the storage media. Submit
disks as MS-DOS "print" or "non-
document" files (often called "ASCII
files"). If a disk cannot be converted to
an ASCII file, the author should indicate
on the disk the source computer and
software language along with the file
name. Either 5V4-inch or 3V2-inch disks
from IBM-compatible or Apple/Macin-
tosh systems (non-graphics only) can be
submitted, double-sided/double-density
or high-density, limiting each file to 300
kilobytes. All 8-inch word-processing
disks (e.g., Wang or NBI) must be con-
verted onto 5V4- or 3V2-inch MS-DOS
print disks.
Send original hardcopy and two dupli-
cated copies to:
Dr. Ronald W. Hardy, Scientific Editor
Northwest Fisheries Science Center
National Marine Fisheries Service,
NOAA
2725 Montlake Boulevard East
Seattle, WA 98112-2097
Copies of published articles and notes
The senior author and his/her organiza-
tion each receive 50 separates free-of-
charge. Additional copies may be pur-
chased in lots of 100.
U.S. Department
of Commerce
Volume 90
Number 4
October 1992
. Iviari.j biolc-gicdl La.-.jraiufy
MAR 1 2 1993
Bulletin
WOJU:, Hoi
a. ass.
U.S. Department
of Commerce
Barbara Hackman Franklin
Secretary
National Oceanic
and Atmosphieric
Administration
John A. Knauss
Under Secretary for
Oceans and Atmosphere
National Marine
Fisiieries Service
William W. Fox Jr.
Assistant Administrator
for Fisheries
^'-^reso*''"
The Fishery Bulletin (ISSN 0090-0656)
is published quarterly by the Scientific
Publications Office, National Marine
Fisheries Service, NOAA, 7600 Sand
Point Way NE, BIN C15700, Seattle, WA
98115-0070. Second class postage is paid
in Seattle, Wash., and additional offices.
POSTMASTER send address changes for
subscriptions to Fishery Bulletin, Super-
intendent of Documents, Attn: Chief,
Mail List Branch, Mail Stop SSOM,
Washington, DC 20402-9373.
Although the contents have not been
copyrighted and may be reprinted entire-
ly, reference to source is appreciated.
The Secretary of Commerce has deter-
mined that the publication of this period-
ical is necessary in the transaction of the
public business required by law of this
Department. Use of funds for printing of
this periodical has been approved by the
Director of the Office of Management and
Budget.
For sale by the Superintendent of
Documents, U.S. Government Printing
Office, Washington, DC 20402. Subscrip-
tion price per year: $24.00 domestic and
$30.00 foreign. Cost per single issue:
$12.00 domestic and $15.00 foreign. See
back page for order form.
Scientific Editor
Dr. Linda L. Jones
Northwest Fisheries Science Center
National Marine Fisheries Service, NOAA
2725 Montlake Boulevard East
Seattle, Washington 981 12-2097
Editorial Committee
Dr. Andrew E. Dizon National Marine Fisheries Service
Dr. Charles W. Fo\/vler National Marine Fisheries Service
Dr. Richard D. Methot National Marine Fisheries Service
Dr. Theodore W. Pietsch University of Washington
Dr. Joseph E. Powers National Marine Fisheries Service
Dr. Tim D. Smith National Marine Fisheries Service
Dr. Mia J. Tegner Scnpps Institution of Oceanography
Managing Editor
Nancy Peacocl<
National Marine Fisheries Service
Scientific Publications Office
7600 Sand Point Way NE, BIN CI 5700
Seattle, Washington 981 15-0070
The Fishery Bulletin carries original research report;s and technical notes on investiga-
tions in fishery science, engineering, and economics. The Bulletin of the United States
Fish Commission was begun in 1881; it became the Bulletin of the Bureau of Fisheries
in 1904 and the Fishery Bulletin of the Fish and Wildlife Service in 1941. Separates
were issued as documents through volume 46; the last document was No. 1103. Begin-
ning with volume 47 in 1931 and continuing through volume 62 in 1963, each separate
appeared as a numbered bulletin. A new system began in 1963 with volume 63 in
which papers are bound together in a single issue of the bulletin. Beginning with
volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued
quarterly. In this form, it is available by subscription from the Superintendent of
Documents, U.S. Government Printing Office, Washington, DC 20402. It is also
available free in limited numbers to libraries, research institutions, State and Federal
agencies, and in exchange for other scientific publications.
U.S. Department
of Commerce
Seattle, Washington
Volume 90
Number 4
October 1992
Fishery
Bulletin
j;i:ic Gic'logical Laborator
LIBRARY
MAR 1 2 199:
Hoia. Mass.
Contents
633 Allen, Larry G., and Michael P. Franklin
Abundance, distribution, and settlement of young-of-the-year white
seabass Atractosaon nobilis. in the Southern California Bight, 1 988-89
642 Boggs, Christofer H.
Depth, capture time, and hooked longevity of longline-caught
pelagic fish: Timing bites of fish with chips
659 Boothroyd, Frank A., and Gerald P. Ennis
Reproduction in American lobsters Homarus amencanus transplanted
northward to St. Michael's Bay, Labrador
668 Ditty, James G., and Richard F. Shaw
Larval development, distribution, and ecology of cobia Rachycentron
canadum (Family: Rachycentridae) in the northern Gulf of Mexico
678 Edwards, Elizabeth F.
Energetics of associated tunas and dolphins in the eastern tropical
Pacific Ocean: A basis for the bond
691 Eggleston, David B., Romuald N. Lipcius, and
David L. Miller
Artificial shelters and survival of juvenile Caribbean spiny lobster
Panulirus argus: Spatial, habitat, and lobster size effects
703 Graves, John E., Jan R. McDowell,
Ana M. Beardsley, and Daniel R. Scoles
Stock structure of the bluefish Pomatomus saltatnx along the
mid-Atlantic coast
71 1 Leffler, Deborah L., and Richard F. Shaw
Age validation, growth, and mortality of larval Atlantic bumper
(Carangidae: Chloroscombrus chrysurus] in the northern Gulf
of Mexico
720 Parrish, Frank A., and Thomas K. Kazama
Evaluation of ghost fishing in the Hawaiian lobster fishery
Fishery Bulletin 90(4), 1992
726 Pearcy, William G.
Movements of acoustically-tagged yellowtail rockfish Sebastes flavidus on Heceta Bank, Oregon
736 Restrepo, Victor R., John M. Hoenig, Joseph E. Powers, James W. Baird,
and Stephen C. Turner
A simple simulation approach to risk and cost analysis, with applications to swordfish and cod fisheries
749 Schilling, Mark R., Irene Seipt, Mason T. Weinrich, Steven E. Frohocl<,
Anne E. Kuhlberg, and Phillip J. Clapham
Behavior of individually-identified sei whales Balaenoptera borealis during an episodic influx into the
southern Gulf of Maine in 1986
756 Somerton, David A., and Bert S. Kil^kawa
Population dynamics of pelagic armorhead Pseudopentaceros wheelen on Southeast Hancock Seamount
770 Utter, Fred M., Robin S. Waples, and David J. Teel
Genetic isolation of previously indistinguishable chinook salmon populations of the Snake and Klamath
rivers: Limitations of negative data
Motes
778 Jensen, Gregory C, Helle B. Andersen, and David A. Armstrong
Differentiating Paralithodes larvae using telson spines. A tail of two species
784 Matthews, Kathleen R.
A telemetric study of the home ranges and homing routes of lingcod Ophiodon elongatus on shallow
rocky reefs off Vancouver Island, British Columbia
791 Miller, George W.
An investigation of dolphin Tursiops truncatus deaths in East Matagorda Bay, Texas, January 1990
798 Secor, David H.
Application of otolith microchemistry analysis to investigate anadromy in Chesapeake Bay striped bass
Morone saxatilis
807 Skillman, Robert A., and George H. Balazs
Leatherback turtle captured by ingestion of sguid bait on swordfish longline
809 Taylor, Ronald G., and Michael D. Murphy
Reproductive biology of the swordfish Xiphias gladius in the Straits of Florida and adjacent waters
817 Index, Volume 90
Abstract. - Commercial and
sport landings of white seabass have
declined, particularly in southern
California, and the populations now
appear to be severely impacted. To
provide information critical to the
management of this species, settle-
ment patterns of white seabass with-
in the Southern California Bight
were investigated for 1988-89. Data
were obtained from 16 stations sam-
pled along the southern California
coastline during June-October 1988,
and at 12 stations sampled along the
coasts of the mainland and four
Channel Islands May-August 1989.
At each station, four 5-minute tows
were taken with a 1.6 m beam trawl
at each of two depths, 5 and 10m.
Most young-of-the-year white sea-
bass were <10mmSL and had set-
tled within 2-3 weeks of capture.
Density estimates for white seabass
off southern California were low,
ranging from 0.3 to only 37.8 individ-
uals per hectare. In 1988, catch-per-
unit-effort (CPUE) peaked in July
(1.10/tow) with differences being sta-
tistically significant among months.
In 1989, CPUE peaked in June
(0.45/tow) with differences being
statistically significant among
distance blocks from the mainland.
CPUE was 15 times higher at the
mainland stations compared with the
island stations (0.59/tow vs. 0.04/
tow). Abundance was significantly
correlated with warm bottom-water
temperatures in 1988, although not
in 1989.
Multivariate analysis of the catches
with selected environmental vari-
ables indicated that distance from
the mainland and bottom tempera-
ture may have been important fac-
tors influencing settlement. However,
in combination, these two variables
accounted for only 5% of the total
variance {R'^ 0.05) in abundance.
This finding implies that other fac-
tors, most notably the availability of
premetamorphic larvae, probably
have an influence on white seabass
settlement and need to be considered
in future studies.
Abundance, distribution, and
settlement of young-of-the-year
whiite seabass Atractoscion nobilis
in tlie Southern California
Bight 1988-89*
Larry G. Allen
Department of Biology. California State University, NortfiFTdyL', Cdlifui i lia 9 1 J30 i .. ,
i luuiiiifi Biological Laboratory |
Michael P. Franklin i LIBRARY
i'
Department of Biology, California State University, Northridge, California 91330
Present address: Department of Biological Sciences j MAR 1 ^ 1993
University of California. Santa Barbara, California 9B106
The white sesbs.?,?, Atractoscion nobi-
lis is the largest croaker (family Sci-
aenidae) occurring off southern Cali-
fornia (Miller and Lea 1974), where
it is important in both commercial
and sport fisheries. Despite attempts
to improve the fishery (e.g., impos-
ing minimum size requirements
and limits on sport and commercial
catches; Frey 1971), landings con-
tinue to decline and the stocks appear
to be severely impacted (Vojkovich
and Reed 1983), particularly in
southern California waters.
Despite their impacted status and
economic importance, little was
known about the early-life-history
stages of white seabass until recent-
ly. Moser et al. (1983) described the
larval development from hatchery-
reared eggs. Field investigations of
early-life-history stages were limited
to reports of larval occtirrence wnthin
California Cooperative Fisheries In-
vestigations (CalCOFI) collections
from 1950 to 1978. For example,
Moser et al. (1983) found that only
15% of white seabass larvae were
taken in southern California waters.
Most were taken near Sebastian
Viscano and San Juanico Bays, Baja
California.
Manuscript accepted 8 July 1992.
Fishery Bulletin, U.S. 90:633-641 (1992).
' Contribution 67 of the Ocean Studies Insti-
tute, California State University.
A few studies have provided lim-
ited information on the yoimg-of-year
(YOY) stages of white seabass. Allen
and Franklin (1988) examined the
abundance and distribution of juve-
nile white seabass in the vicinity of
Long Beach harbor and developed a
model for locating YOY white sea-
bass in coastal waters. We observed
that YOY white seabass were cap-
tured over sandy bottoms in shallow
water near the breaker line, most
often with submerged aquatic vege-
tation (drift algae: green, brown,
and red), encrusting bryozoans, and
terrestrial debris. This area seems to
be the nursery grounds for white
seabass. The drift material may be an
important component of these nur-
sery areas because these fish appear
to be structure-oriented early in life
(Allen and Franklin 1988, Margulies
1989, Donohoe 1990). Donohoe
(1990), based on field collections,
found that young seabass were asso-
ciated with the drift, and also ob-
served that the larvae and juveniles
moved toward structures in labora-
tory experiments. A significant rela-
tionship was found between the mass
of drift algae and the occurrence of
YOY white seabass from Oceanside,
California to the Mexican border,
suggesting that the drift habitat may
influence the distribution patterns
633
634
Fishery Bulletin 90(4), 1992
of these young fish (Donohoe 1990). Margulies (1989)
concluded that the visual perception of YOY seabass
improves with age, and that young fish begin to avoid
predators by moving to the drift.
Our studies on white seabass settlement were under-
taken in southern California where the main fishery
for this species still exists. The specific objectives were
to (1) examine the patterns of abundance, distribution,
and settlement of YOY white seabass off the coast of
southern California between Point Conception and San
Mateo Point and along the coastlines of four of the
larger Channel Islands, and (2) identify environmen-
tal factors that may influence these patterns.
Materials and methods
YOY white seabass were captured during the summers
of 1988 and 1989 as part of the Ocean Resources
Enhancement and Hatchery Program (OREHP) of the
California Department of Fish and Game, which em-
phasized studies within the Southern California Bight.
Trawls were made over flat bottoms just offshore of
open sand beaches, using two 5.2 m whalers. At each
station four 5-minute replicate tows were made in the
shallow, potential nursery areas by two whalers
simultaneously sampling along each of two isobaths (5
and 10m) using a 1.6m beam trawl. The trawl was com-
prised of 4 mm mesh in the wings and 2 mm knotless
mesh in the codend. Calibration tows using a meter
wheel indicated that a 5-minute
tow covered an average of 183 m
of bottom, yielding a mean cover-
age of 293 m-. Bottom profiles
were monitored using depth
finders mounted in each whaler.
Temperature, salinity, dissolved
oxygen, and pH were monitored
at the surface and bottom at each
station at both isobaths, using
a Hydrolab Surveyor II Water
Quahty Measurement System.
Submerged aquatic vegetation
(drift algae) captured in each tow
was weighed (to nearest kg) at all
stations. White seabass were
measured to the nearest 0.1mm
standard length (SL).
Sixteen stations were estab-
lished along the coast of southern
California from Point Conception
to San Mateo Point in 1988 (Fig.
1). These stations were approx-
imately lOnmi apart and were
sampled from June through Oc-
tober. The sampling regime yielded 128 tow samples
(4 tows at 2 depths at each of 16 stations) over 5
months, for a total of 640 tow samples in 1988.
In 1989, sampling was designed to examine the set-
tlement patterns along the mainland and around the
four largest offshore islands. Four mainland stations
and eight island stations were sampled each month
from May through August (Fig. 1). The mainland sta-
tions were those with the greatest consistency of catch
of YOY white seabass in 1988: Stns. 6 (Ventura), 10
(Malibu), 13 (Belmont Shore), and 15 (Laguna Beach).
Two new stations were established at each of four
Channel Islands: Stns. 17 and 18 (Santa Cruz), 19 and
20 (Santa Rosa), 21 and 22 (Santa Catalina), and 23 and
24 (San Clemente). The twelve stations sampled in 1989
were divided subjectively into three groups of four sta-
tions based on relative distance from shore or distance
blocks (DSTBLK). Stns. 6, 10, 13, and 15 were desig-
nated as being at the mainland (MAINLAND); Stns.
17, 18, 21, and 22 as near-island stations (NEAR ISL);
and Stns. 19, 20, 23, and 24 as far-island stations (FAR
ISL). The 1989 sampling regime yielded 96 tow samples
(4 tows at 2 depths at each of 12 stations) for each
month except May, when only 80 tows were made
because poor weather conditions prevented sampling
the Santa Rosa Island stations, for a total of 368 tows
overall.
Analysis of variance, f-test, and correlation analyses
were completed using the CSS:Statistica for desktop
computers (Stat Soft, Inc., Tulsa). For 1988 data, a
i''
^ — " Mr 7
■■'»
-
Pt Conception ^^<V~^
Santa Barbara
; g^ Ventura
-
00'
Santa Rosa 1
Santa Cruz 1. -■-■-,-. f;^-A. Maiibu
20 -,2,^
Los Angeles ->>
Belmont Shore
21
'4 ;\Lagun3 Beach
-
Santa Calalinal ^
l»22 '^^
San Mateo Pi ^ _
0(C
^
■- \
San Clemente 1 X^
San Oiego ,f^
0 iOkm. I
1
"'•i»»' 1 >
1!0'|00' , , 11B*,00 1
_l '"'■'"' 1 i-J
Map of locations of
seabass A. nobilis
Figure 1
mainland and Channel Island stations sampled during the YOY white
survey, 1988-89.
Allen and Franklin: Settlement patterns of Atractosaon nobihs in the Southern California Bight
635
balanced design, three-way ANOVA was used to test
the effects on catch-per-unit-effort (CPUE) of combina-
tions of independent variables (station, month, and
depth). In 1989, a similar three-way ANOVA design
was used to test the effects of distance block (distance
from the mainland), month, and depth on CPUE.
CPUE was used in all parametric analyses in order to
minimize any negative impact that the large number
of zero-catch tows would have on the analysis. Since
replicates had to be combined, the three-way-inter-
action mean square was utilized as a conservative
estimate of sampling error for the 1988 ANOVA test.
In 1989, the ANOVA design was unbalanced due to two
missing stations in May at Santa Rosa Island. In this
case, cell means estimation was utilized to overcome
the imbalance. Since the three-way-interaction term
was originally found to be significant in the 1989
analysis, its mean square was pooled with the within-
sample error in order to partition out the effect of the
interaction on the main effects of distance block,
month, and depth individually. Correlations and canon-
ical correlation analysis were utilized to examine the
possible association of various environmental factors
with settlement of YOY white seabass.
Results
Length-frequencies
The YOY white seabass captured during the 1988 sur-
vey ranged from 4.2 to 78mmSL (.r9.8mmSL). In
1989, the range was 4.5-51. 7mmSL (x 12.1mmSL).
Most YOY, however, ranged from 5 to 20mmSL (Fig.
2). Newly settled fish (<10mmSL) were caught from
June through September. Individuals >20mmSL were
more common from July through October.
Newly settled fish (< lOmmSL) made up 75% of all
YOY seabass in 1988 and 1989. Fish <20mmSL com-
prised 93% of the total catch. The paucity of larger
YOY in the samples (Fig. 2) from July to September
indicates that our beam-trawl catch may be biased
toward smaller, less-mobile fish.
Studies utilizing the daily growth rings on white
seabass otoliths (Franklin and Allen, unpubl. data) in-
dicate that fish of 5-20 mm SL are ~37-104 days old.
Since settlement occurred consistently at ~5mmSL,
white seabass in this range settled at 0-68 days before
capture. The majority of those < 10 mm SL had settled
14-21 days before capture.
Abundance and distribution
Summer 1 988 Sampling along the mainland yielded
270 YOY white seabass. Most (58%) were captured at
five of the 16 stations: Stn. 2 (Refugio Beach, n 38),
Stn. 6 (Ventura, n 28), Stn. 10 (Malibu, n 30), Stn. 13
(Belmont Shore, n 31), and Stn 15 (Laguna Beach,
n 31) (Fig. 3). Mean catches (CPUE) were highest at
these five stations; furthermore, the variance of these
five means was also very high. This was especially true
at Stn. 2 (Refugio Beach) where 36 of the 38 YOY cap-
tured were taken during a single month (July).
The CPUE for all stations was low in June 1988 (0.15
individuals/tow), peaked in July (1.10/tow), and de-
clined to 0.08/tow in October (Fig. 4). Catches in July
accounted for 52% (141 individuals) of YOY white
seabass taken in 1988. In June, catch was low (19 in-
dividuals) and white seabass were collected only from
southern stations (12, 13, and 15) (Fig. 5). By July,
150
125
100
S 75
P
bteiiiijfe^
".'^ , — "^
10 20 30 40 50 60 70
LENGTH (2mmSL increments)
80
Figure 2
Length-frequencies of YOY white seabass /I. nobilis from all
samples combined, 1988-89. Length increments are 2 mm
(n 3.54).
2.0
1.8
1.6
5^1.4
+ 1.2
^ 1.0
zO.8
I 0.6
0.4
0.2
0.0
3 4 5 6 7 H y 1 u 11 U' I :_! 1 4 1 5 1 0
STATION
Figure 3
CPLIE of YOY' white seabass A. nobilis by station, summer
1988. Bars represent 2SE of the mean.
636
Fishery Bulletin 90(4), 1992
2.0
c/;
CM
1
+
£1.0
z
<
^0.5
-.- ^
JUN JIL Al G SEP OCT
Figure 4
CPUE of YOY white seabass A. nobilis by month, summer
1988. Bars represent 2SE of the mean.
relatively heavy settlement was observed throughout
the coastal area, as far north as Stn. 2. The greatest
numbers were taken off Stns. 2 (Refugio Beach), 6
(Ventura), 10 (Malibu), and 13 (Belmont Shore). None
were taken at Stn. 15 (Laguna Beach) where they were
most abundant a month earlier. In August, the number
of recently-settled white seabass had declined from the
July peak. Moderate numbers were captured at Stns.
7 (Hueneme), 11 (El Segundo), 13 (Belmont Shore),
and 15 (Laguna Beach). In September, young seabass
settled at the middle stations (stns. 6-11). By October,
no new settlement was detected at any of the study
sites (Fig. 5). The only YOY white seabass was an older
fish (78mmSL) taken at Stn. 16 (San Mateo Pt.).
Sixty-three percent (170 individuals) of all YOY were
taken at the 5m depth in 1988. YOY white seabass
were most numerous at the 5 m isobath at Stns. 2-4,
10-13, and 15, but were more abundant at 10 m at Stns.
5-9 (Fig. 6).
Analysis of variance of CPUE values in 1988 indi-
cated that only the observed monthly differences were
statistically significant (Table 1). Differences in CPUE
among stations and depths were not significant in 1988.
YOY white seabass densities ranged from a low of
0.3 individuals/ha in October 1988, to a high of 37.8 in
July 1988. In 1988, population estimates varied great-
ly along the approximately 300km of coastline (ex-
cluding the offshore islands) covered in 1988. Overall
density for the 5-month period yielded a population
estimate of 130,000 individuals in the area of southern
California covered by the sampling.
Summer 1989 Sampling from May to August 1989
along the coastlines of the mainland and the offshore
islands produced 85 YOY. The catch rate at the main-
Figure 5
Abundance of YOY white seabass A. nobilis by station over
the five months (June-October) of the 1988 survey.
Figure 6
CPUE of YOY white seabass A. nobilis by station and
depth for the 1988 survey.
land stations was 15 times higher than at island sta-
tions (CPUE 0.59/tow in 128 tows vs. 0.04/tow in 240
tows) (Fig. 7). Most (88%) of the YOY were captured
at three of the mainland stations: Stns. 6 (Ventura,
n 22), 10 (Malibu, n 14), and 15 (Laguna Beach, n 38).
Five YOY were captured at Stns. 17 and 18 on Santa
Cruz I., four were taken at Stn. 22 (White's Cove) on
Santa Catalina I., and none were taken at Santa Rosa
or San Clemente I.
The CPUE was low in May (0.14/tow), peaked in June
(0.45/tow), and declined through July-August (0.14/
tow) (Fig. 8). Settlement was restricted to the southern
mainland stations (13 and 15) in May (Fig. 9). By June,
settlement was observed as far north as Stn. 6 (Ven-
tura) with the greatest numbers occurring off Stns. 10
(Malibu) and 15 (Laguna Beach). In July and August
settlement was highly variable at the mainland sta-
tions. The five YOY white seabass taken at Santa Cruz
I. (Stns. 17 and 18) were captured during June, July,
Allen and Franklin: Settlement patterns of Atractoscion nobilis in the Southern California Bight
637
Table I
Summary of three-way ANOVA results for catches of young-of-year white seabass A. nobilis during the 1988 and 1989 coastal surveys. |
Dependent variable in all cases
was CPUE. DSTBLK =
distance block;
• p<0.05; *♦ p<0.001.
Test
Effect
df
MS
F
P
1988 Survey
Station x Month x Depth
Station
15
0.9832
1.0086
0.4590
Month
4
5.7207
5.8686
0.0005 •*
Depth
1
1.9141
1.9635
0.1663
S X M
60
1.0263
1.0529
0.4212
S X D
15
1.3441
1.3788
0.1875
M X D
4
0.9687
0.9938
0.4180
Error
S X M X D
60
0.9748
-
-
1989 Survey
DSTBLK X Month x Depth
DSTBLK (DB)
2
2.9712
11.8180
0.0000**
Month
3
0.6036
2.4008
0.0745
Depth
1
0.5481
2.1801
0.1440
DB X M
6
0.4784
1.9029
0.0915
DB X D
2
0.3110
1.2371
0.2962
M X D
3
0.7950
3.1623
0.0295*
Error
DB X M X D
74
0.2514
—
—
and August. Stn. 22 (White's Cove, Santa Catalina I.)
was the only other Channel Island station where YOY
seabass were taken (Fig. 9).
During 1989, 69% (n 59) of YOY white seabass were
taken at 5m, while 31% (w 26) were captured at 10m.
Most of the fish taken at the 10 m isobath were cap-
tured at Stns. 6 (Ventura) and 15 (Laguna Beach). Two
YOY white seabass were taken at 10m off Santa Cruz
I. (Stns. 17 and 18).
In 1989, catches (CPUE) were then examined accord-
ing to distance block, month, and depth of capture.
Analysis of variance revealed the significant effect of
distance block (Table 1; Fig. 10) which was highly
significant (p< 0.0001). Although month and depth
were not significant main effects, a significant month-
by-depth interaction was detected in the three-way
ANOVA. The month-by-depth interaction indicated
that depth distributions changed significantly over the
period of May- August. Catches of YOY white seabass
increased at the 10 m depth stratum and decreased at
5 m over the course of the summer (Fig. 11).
The population estimate for 1989 based on mean
density along ~600km of mainland and offshore
islands coastland was about 118,000 individuals over
the 4-month sampling.
2.0
1.8
1.6
!l.4
(
1.2
]
;i.oh
;0.8
I 0.6
0.4
0.2
0.0
MAINLAND
ISLAND
10 13 15
17 18 19
STATION
Figure 7
CPL'E of YOY white seabass .4. nobilis by station, summer
1989. Bars represent 2 SE of the mean. Stations are grouped
into mainland and island sites for comparison.
2.0
+
W
gl.O
2
<
^0.5
0.0
CPUE
1989. I
r
I 1
■J?^^^ 1
f ; : : .:;:ii
\1\1 n\ JUL AUG
Figure 8
of YOY white seabass A. nobilis by month, summer
?ars represent 2SE of the mean.
638
Fishery Bulletin 90(4). 1992
6 vo ^-^ ^" ISLAND
Figure 9
CPUE of YOY white seabass A. nobUis by station over the
four months (June-October) of the 1989 survey. Stations are
grouped into mainland and island sites for comparison. Stns.
19 and 20 (Santa Rosa I.) were not sampled in May due to
severe weather conditions.
M^'^'
Figure 10
CPUE of YOY white seabass A. nobilis within stations
grouped by distance block (MAINLAND = four mainland sta-
tions, NEAR ISL = four near-island stations. FAR ISL =
four far-island stations) over the four months (May-August)
of the 1989 survey.
Comparison of 1988 and 1989
at four mainland stations
Settlement success in 1988 was compared with that in
1989 by examining the catch at the four mainland sta-
tions sampled during both years (Fig. 12). Abundance
varied significantly among months over the two sum-
mers (one-way ANOVA; F 2.52; 8,27df; p<0.05), but
not among stations. Settlement among the four sta-
tions was consistent during 1988, but highly variable
in 1989. Also, CPUE was higher in 1988 (0.75/tow) than
in 1989 (0.59/tow) but the difference was not signifi-
cant (^test;< 1.23; 15df;p 0.24). Catches differed most
between years at Stn. 13 (Belmont Shore) where abun-
dance dropped from 0.78/tow in 1988 to 0.06/tow in
NV.\^'
Figure 1 1
Abundance of YOY white seabass .4. nobilis by depth over
the four months (June-October) of the 1989 survey.
^grsTVlR^
Figure 12
CPUE of YOY' white seabass .4. nobilis at each of four
mainland stations by month over the summers of 1988 and
1989. (Stn. 6 = VENTURA. Stn. 10 = MALIBU. Stn. 13 =
BELMONT. Stn. 15 = LACUNA).
1989, and at Stn. 15 (Laguna Beach) where abundance
increased from 0.78/tow to 1.21/tow. The lack of signifi-
cant differences in catches between 1988 and 1989 was
probably due to the high variability and low numbers
at individual stations within each year. The relatively
high catch at Laguna Beach in 1989 was directly op-
posed to the lower catches at the other three stations.
Influence of environmental factors
Distance from the mainland Of the environmental
variables examined, only the distance of the station
from the mainland was significantly correlated with
CPUE over both years (Table 2), and this correlation
was negative. This corroborates the ANOVA results
from 1989 where the effect of distance block was highly
Allen and Franklin: Settlement patterns of Atractoscion nobilis in the Southern California Bight
639
Table 2
Correlation coefficients among catch-per-unit-effort (CPUE) of young-of
year white seabass/l. nohilis and six environmental variables
during th
e 1988 and 1989 coastal surveys (* p<0.05, 219df).
CPUE DSTMN BTMP
BSAL
BDO BSLOP BALGA
CPUE
_
DSTMN
-0.2030*
BTMP
0.0993 -0.0115 -
BSAL
0.0513 -0.0349 -0.0279
—
EDO
-0.0084 -0.0425 0.0674
-0.0402
—
BSLOP
0.0967 -0.3761* -0.0718
-0.0365
0.1427*
BALGA
0.0357 -0.0204 -0.0139
-0.0118
-0.0305 -0.1063 -
DSTMN
Distance from mainland
BTMP
Bottom temperature
BSAL
Bottom salinity
BDO
Bottom dissoved oxygen
BSLOP
Bottom slope
BALGA
Biomass algae
significant, furtiier emphasizing the inshore-offshore
distribution pattern of white seabass settlement.
Temperature Bottom temperature ranked second
among environmental variables in its correlation to
CPUE over both years, although the correlation of
0.10 was not statistically significant (Table 2). The lack
of significance may be due to the fact that the rela-
tionship of catch to temperature differed noticeably
in the 2 years and that overall catches were lower in
1989.
In 1988, the heaviest and most widespread settle-
ment of YOY coincided with the striking rise in coastal
temperature during July in the study area, resulting
in a significant correlation between log-transformed
[logio (x-i- 1)] abundance of YOY and bottom temper-
ature (rO.25, P<0.05, 74df). In 1989, however, the
greatest and most widespread YOY abundance was en-
countered in June when temperatures were generally
depressed. Thus the peak settlement in 1989 occurred
1 month earlier than in 1988 and was apparently not
as closely related to a rise in sea temperature as it
seemed to be in 1988.
Biomass of drift algae Samples of submerged drift
algae ranged from trace amounts (<50g) to > 500 kg
per tow for each depth and station. No significant cor-
relation (r 0.036; Table 2) was found between the
weight of drift algae and the abundance of young white
seabass. However, only two fish (both >60mmSL)
were captured without drift algae in the nets. Thus,
drift algae and YOY white seabass may be related on
a presence/absence rather than a quantitative basis.
Correlations of catch with other physicochemical
variables were too low to warrant consideration.
Multivariate model A combination of three environ-
mental variables— distance from the mainland, bottom
temperature, and biomass of drift algae— produced a
significant canonical correlation with CPUE (Table 3).
Though significant, the correlation accounted for only
5% (R- 0.052) of the variation in CPUE. A significant
canonical correlation with distance and bottom tem-
perature alone accounted for slightly less variation in
CPUE (R- 0.051).
Discussion
Density estimates for white seabass off the coast of
southern California were low. Population estimates
based on these densities for the Southern California
Bight were only 130,000 and 118,000 individuals in
1988 and 1989, respectively. The lower value in 1989
is not surprising since catches at the island stations
were extremely low (a high of five YOY at Santa
Cruz I., and none at Santa Rosa and San Clemente Is.).
Even if these estimates are assumed to be within an
order of magnitude of the real population levels, it is
obvious that settlement of white seabass was poor in
southern California waters. Our data showing relative-
ly low numbers of YOY white seabass in southern
California for both sampling years present a similar pic-
ture to that presented in Moser et al. (1983) for larval
white seabass. The major settlement areas for this
species undoubtedly occur to the south in Mexican
waters.
Catches of YOY white seabass were highly variable
in space and time. Only a small portion of this variabil-
ity was explained by the environmental variables
measured. Monthly differences in catch were marked
640
Fishery Bulletin 90(4). 1992
Table 3
Results of canonical correlation runs with catch-per-unit-effort (CPUE) of young-of-year white seabass A. mbilis,
variable and selected environmental (independent) variables (* p<0.05, ** p<0.01).
1988-89,
as the dependent
Run
Variables
Successive canonical correlation runs
Canonical R Canonical R-
x'
df
P
1
2
DSTMNLD
BTMP
BALGA
DSTMNLD
BTMP
0.2274
0.2250
0.0517
0.0506
11.495
11.274
3
2
0.0093"
0.0036'*
Canonical weights within runs
Variables Run 1
Run 2
DSTMNLD
BTMP
BALGA
land
e
-0.8848
0.4287
0.1450
-0.8973
0.4312
DSTMNLD
BTMP
BALGA
Distance from main
Bottom temperatur
Biomass algae
in both years, due to the peaks in abundance of YOY
observed in both 1988 and 1989, although monthly
differences were significant only in 1988. In 1989, a
significant spatial pattern of catches was detected,
related to distance from the mainland and depth of cap-
ture over months.
Both the distance block ANOVA from 1989 data and
the overall correlation analyses strongly suggest that
the abundance of YOY white seabass decreases rapid-
ly with distance from the mainland. Other factors
are less important. Nonetheless, the combination of
distance, temperature, and biomass of drift algae pro-
duced a highly significant canonical correlation with
distance and temperature contributing most heavily to
the relationship.
The large amount of unexplained variation in the
multivariate model suggests that important factors
may be missing from the analysis. We believe that one
such factor is the initial availability of presettlement
larvae in the plankton. A dearth of premetamorphic
larvae at a potential settlement site results in low
settlement, no matter how favorable the environmen-
tal conditions. Population sizes off southern California
might be limited largely by number of settling larvae
rather than site-specific environmental factors or
density-dependent survival of YOY. Only when larval
input is constantly high, as we suspect is the case in
Mexican waters, could the influence of environmental
factors on settlement success be determined with any
precision.
Factors affecting larval availability are not well
known. Spawning of white seabass occurs in the sum-
mer and may be related to lunar periodicity (moon
phase) (Franklin and Allen, unpubl. data) early in the
reproductive period. Lunar periodicity of spawning ac-
tivity coupled with adult stock size, larval transport
mechanisms, and larval growth dynamics could all
ultimately influence the availability of white seabass
larvae.
Distance from the mainland, the strongest correlate
with YOY abundance, probably reflects larval avail-
ability which may decrease with distance from coastal
stocks occurring in both southern and Baja California.
Island populations of adults were either not repro-
ducing or their larvae were being carried away from
settlement sites. Long-term settlement success of
white seabass to islands may be sporadic and highly
variable. For example, Cowen (1985) found that Cali-
fornia sheephead (Semicossyphus pulcher) settled only
sporadically to the offshore islands. The pattern of
settlement success of sheephead over a 7-9 year period
in areas without larval sources "upstream" of typical
current direction was highly variable and dependent
on episodic events, such as the El Nino climatic anoma-
ly (Cowen 1985).
Warm water currents may be important to white
seabass settlement for two reasons: (1) Large num-
bers of larvae carried northward from more southern
waters by warm water currents may settle after meta-
morphosis and locate suitable habitat; and (2) the warm
water itself may induce locally spawned larvae to
settle. On a larger scale, major water movements such
as the California Current, gyral circulation ("eddies"),
and other mesoscale flows (e.g., internal waves) may
Allen and Franklin Settlement patterns of Atractoscton nobtlis in the Southern California Bight
641
control white seabass settlement in the Southern
California Bight. Parrish et al. (1981) demonstrated
that seasonal effects of the California Current and
upwelling in central California had a major effect on
the distribution patterns of marine fish. The spawning
activities of most fishes coincide with the onshore flow
which is characteristic of the late winter and early
spring months and transports eggs and larvae into
shallow waters. The effects of major hydrographic
events on the abundance and distribution of YOY white
seabass remain largely unknown.
The main geographic source of white seabass larvae
that settle successfully in southern California is also
unknown. Southern California populations of adults
may be reduced to the point that they may be only a
minor source of larvae. Since larvae remain in the
plankton as long as 4-5 weeks, population centers of
adults off northern Baja California may constitute the
major source of southern California YOY seabass.
Thus, successful settlement to southern California
waters may depend largely on the northward-flowing,
warm-water currents best developed in the summer
months. Satellite infrared-imagery data indicated that
such a large, warm-water mass moved north along
the southern California coastline in early July 1988
(Jan Svedkowsky, Ocean Imaging, San Diego, pers.
commun.). The resulting dramatic rise in surface and
bottom temperatures may have accounted for the
marked increase in settlement of white seabass be-
tween June and Juiy of that year if the water mass also
contained a sufficient number of premetamorphic
larvae. Studies of subpopulation structure utilizing
restriction fragment length polymorphism (RFLP)
analysis of nuclear DNA are currently underway in our
laboratory in an attempt to identify the source of
newly-settled white seabass in southern California
coastal waters. If the main parental population of these
fish is located in more southerly waters, joint U.S. and
Mexican management efforts may be necessary to pre-
vent the decline of these major breeding stocks in the
south.
Acknowledgments
A study of this magnitude could not have been ac-
complished without the support of many people. We
thank those who ably assisted with the demanding field
work, especially Jan Cordes, Monica Lara, Julia Sears-
Hartley, Phyllis Travers, and Lisa Wooninck. The crew
of the RV Yellowfin, Jim Cvltanovich, Danny Warren,
and Dennis Dunn, assisted greatly through their
capable handling of vessels in shallow and often tur-
bulent waters. Al Ebeling and three anonymous
reviewers greatly improved this paper through careful
readings and numerous, helpful comments. We also
gratefully acknowledge the assistance provided by
Steve Crooke and Paul Gregory of the California
Department of Fish and Game. This research was sup-
ported through contracts with the Ocean Resource
Enhancement and Hatchery Program (OREHP) and
the Bay, Estuarine, and Nearshore Ecosystem Studies
(BENES) program, administered by the California Fish
and Game.
Citations
Allen, L.G., and M.P. Franklin
1988 Distribution and abundance of young-of-the-year white
seabass {Atractoscion iiobilis) in the vicinity of Long Beach
harbor, California, 1984-1987. CaHf. Fish Game 74:245-248.
Cowen, R.K.
1985 Large scale pattern of recruitment by the labrid, Se7ni-
cossyphus pulcher: Causes and implications. J. Mar. Res.
43:719-742.
Donohoe, C.
1990 Distribution, abundance, food habits, age determination,
and growth, of late larval and early juvenile white seabass,
Atractoscion nobilis. off San Diego County. Calif. Unpubl.
master's thesis, CaHf. State Univ., San Diego. 95 p.
Frey, H.W. (editor)
1971 California's living marine resources and their utilization.
Calif. Dep. Fish Game, Sacramento, 148 p.
Margulies, D.
1989 Size-specific vulnerability to predation and sensory
system development of white seabass, Atractoscion nobilis,
larvae. Fish. Bull., U.S. 87:537-552.
Miller. D.J., and R.N. Lea
1974 Guide to coastal marine fishes of California. Calif. Fish
Game, Fish Bull. 157:1-249.
Moser, H.G., D.A. Ambrose, M.S. Busby, J.L. Butler,
E.H. Sandknop, B.Y. Sumida, and E.G. Stevens
1983 Description of early stages of white seabass, Atractoscion
nobilis. with notes on their distribution. Calif. Coop. Oceanic
Fish. Invest. Rep. 24:182-193.
Parrish, R.H., C.S. Nelson, and A. Bakun
1981 Transport mechanisms and reproductive success of fishes
in the California current. Biol. Oceanogr. 1:175-203.
Vojkovich, M., and R. Reed
1983 \\'hite seabass, Atractoscion nobilis, in California-
Mexican water: Status of the fishery. Calif. Coop. Oceanic
Fish. Invest. Rep. 24:79-83.
Abstract. -To resolve the uncer-
tainty in estimating capture depths
of fish on pelagic longline gear, elec-
tronic microchip hook timers were
attached to branch lines to record
when bites occurred, and time-depth
recorders (TDRs) were attached to
longline gear, off Hawaii in January
1989 and January-February 1990.
Hook timers indicated that 32% of
the striped marlin Tetrapturus au-
dax, 21% of the spearfish T. angus-
tirostris, and 12% of the bigeye tuna
Thunrvus ohesus were caught on sink-
ing or rising hooks, demonstrating
that capture time data are needed to
correctly estimate capture depth.
Recorded and predicted longline
depths differed greatly, indicating
that TDRs are essential for describ-
ing depth distributions of fish from
longline catches. Most (> 60%) of the
spearfish and striped marlin were
caught on settled hooks (not sinking
or rising) at depths of < 120 m, where-
as most bigeye tuna were caught at
depths of > 200 m. This suggests that
eliminating shallow hooks could
substantially reduce the bycatch of
spearfish, striped marlin, and other
recreationally important billfish
without reducing fishing efficiency
for bigeye tuna. Bigeye tuna and
striped marlin survived up to 6-9
hours after capture, and over 50% of
12 frequently-caught taxa were alive
when retrieved, suggesting that the
release of live fish can be an effec-
tive management option.
Depth, capture time, and hooked
longevity of longline-caught
pelagic fish: Timing bites
of fish \N\th chips
Christofer H. Boggs
Honolulu Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA
2570 Dole Street, Honolulu, Hawaii 96822-2396
Manuscript accepted 5 May 1992.
Fishery Bulletin, U.S. 90:642-658 (1992).
Targeting specific depths can improve
longline catches of desired species,
such as bigeye tuna Thunnus obesus
(Saito 1975, Hanamoto 1976, Suzuki
et al. 1977, Suzuki and Kume 1982),
and reduce bycatch of other species,
such as billfish (Suzuki 1989). How-
ever, considerable uncertainty exists
in estimating the fishing depth of
longline gear. Predicted longline depth
based on catenary geometry, line
length, and distance between floats
(Yoshihara 1954) differs from true
depth (Saito 1973, Hanamoto 1974,
Nishi 1990) because of currents and
other factors, yet depth is often in-
ferred rather than measured (Suzuki
et al. 1977, Suzuki and Kume 1982,
Hanamoto 1987, Grudinin 1989). Fur-
thermore, fish may be caught while
the hooks are sinking, during deploy-
ment of the gear, or rising during its
retrieval (Saito 1973), making cap-
ture depths impossible to estimate ac-
curately without known capttu-e times.
Accurate estimates of fishing depth
can be made if time-depth recorders
(TDRs) are attached to longline gear.
Longline studies using TDRs (Saito
et al. 1970, Saito 1973, Yamaguchi
1989, Nishi 1990) have also inter-
preted TDR depth fluctuations as
records of times of capture, but few
such measurements exist. Instead,
capture has been assumed to occur
when the gear is settled, so capture
depth has been estimated as settled
hook depth (Hanamoto 1976, Suzuki
and Kume 1982). Hook timers, de-
signed to indicate when each hook is
struck (Somerton et al. 1988), pro-
vide a way to measure capture times
and survival times of hooked fish.
Capture times, together with TDR
records, can be used to estimate cap-
ture depths accurately.
Billfish catch rates in recreational
fisheries may be negatively affected
by nearby longline fisheries (Squire
and Au 1990), and interest in finding
ways to reduce the longline take of
billfish without reducing fishing effi-
ciency for target species is increas-
ing (Rockefeller 1989). Information
on capture depth, capture time, and
hooked longevity can be used to
design fishing methods that reduce
billfish mortality. Data on the selec-
tivity and efficiency of longline gear
at various depths are also critical for
stock assessments (Suzuki 1989).
The present study improves meth-
ods for estimating capture depths of
fish on longline gear using electronic
timing devices, and describes the
depth distributions and capture times
of tunas, billfishes, sharks, and other
pelagic fishes in Hawaiian waters in
winter. Water temperature and dis-
solved oxygen (DO) were measured
to describe the physical habitat in the
study area, since these variables ap-
pear to cause geographic variation
in depth distributions of fish (Hana-
moto 1975, 1987). Relative fishing
efficiency and the bycatch of bill-
fish were predicted for several gear
configurations.
642
Boggs: Estimating capture depths of longline-caught pelagic fish
643
Table 1
Summary of longline fishing operations conducted by the NOAA ship Townsend Cromwell off Hawaii, January 1989 and January-
February 1990. giving averages for three set types (ranges in parentheses). Baskets were intervals of continuous main line between
floats with snap-on branch lines, not spliced units of gear. Shortening rate was the ratio between ship speed and thrower speed.
Depths do not include branch line length. Predicted depths were calculated from the shortening rate and the main line length per
basket, assuming a catenary shape. TDR = time depth recorder.
Year Set type
Sets
(no.)
Time
Begin set End retrieval
Hooks per
set
(no.)
Line per
basket
(m)
Shortening
rate
(ratio)
Predicted
depth
(m)
Deep TDR
depth
(m)
Middle TDR
depth
(m)
1989 Regular
6
8:09
15:24
199
795
0.80
222
111'
82^
Deep
6
(4:31-10:05)
8:55
(14:20-16:41)
16:36
(128-278)
257
(640-1103)
1085
(0.69-0.98)
0.59
(88-304)
415
(43-180)
260
(32-133)
191-
Very
4
(5:29-12:47)
8:23
(10:34-20:40)
19:10
(185-392)
405
(990-1146)
1117
(0.46-0.83)
0.54
(273-489)
447
(241-303)
367
(178-224)
270-
deep
(8:16-8:41)
(17:.58-20:40)
(356-474)
(1053-1146)
(0.45-0.71)
(349-496)
(329-400)
(243-295)
1990 Regular
5
6:50
19:08
456
809
0.78
243
142
104
Deep
13
(6:14-7:18)
7:18
(16:24-22:07)
19:38
(212-591)
474
(611-1068)
1069
(0.67-0.90)
0.60
(150-355)
409
(78-183)
249
(71-140)
180
Very
4
(6:12-10:04)
5:29
(17:57-20:42)
19:03
(173-594)
404
(798-1265)
1165
(0.40-0.70)
0.62
(298-499)
436
(193-318)
416
(122-232)
291
deep
(4:45-6:21) (15:30-21:05) (219-600)
led for only three sets,
the ratio (0.73) between middle and deep TDR
(937-1427)
depths of sets
(0.50-0.74) (303-592)
in which middle-position
(340-517) (251-381)
data were available.
'TDR data obtaii
-Calculated from
Materials and methods
Longline fishing was conducted on board the NOAA
ship Townsend Cromwell in January 1989 and Janu-
ary-February 1990. Sets were made between lat. 14°
and 20°N, long. 148° and 159°W, 20-500nmi from the
main Hawaiian Islands, and within an area typically
fished by Hawaii's domestic longline fishery. Gear was
usually deployed in the morning and retrieved in the
afternoon or evening (Table 1), or occasionally at mid-
day to permit a second set on the same day. No sets
were made at night. Except for the hook timers and
TDRs, the fishing gear and operations were similar to
commercial longline fishing methods for tuna in Hawaii
(Kawamoto et al. 1989) prior to the advent of night
fishing for swordfish Xiphias gladius. Both this study
and the contemporary commercial longline fishery used
a wide variety of fishing depths. Commercial fishermen
used more gear (~1000 hooks), let it stay in the water
longer (~12h), and retrieved it faster than in this study.
The fishing gear consisted of 3.5mm-diameter nylon
monofilament main line deployed with a line thrower
(Kawamoto et al. 1989). The main line was supported
at intervals by vertical, 18m lines with floats at the
ends. Snap-on branch lines made of 2.1mm-diameter
clear-blue nylon monofilament (20 m long in 1989 and
11m long in 1990) were baited with thawed saury Colo-
labis saira on curved tuna hooks (one hook/branch line)
and attached to the main line between float lines.
Hooks were size 3.6 (Japanese size is 10.9cm from eye
to point). Each portion of the longline between floats
and the attached branch lines constituted a "basket,"
a term taken from older gear in which the number of
branch lines is fixed. However, this study used vary-
ing numbers of snap-on branch lines (12, 14, 16, or
20/basket), depending on the length of main line per
basket.
Hook position was controlled by timing the attach-
ment of branch lines as the main line was thrown over-
board mechanically at a controlled speed. A computer
program was used to signal and record attachment
times. Deviations from the programmed instructions
were noted, providing a record of set times for each
hook. The total number of hooks in each set was
128-600, and the amount of main line deployed per set
was 9-44 km (Table 1). The amount of gear increased
with crew experience but also varied because of incle-
ment weather and equipment failures.
Set depths
Fishing depth was altered by varying the slack in the
main line and the length of line per basket (Table 1)
and by exogenous factors such as wind and currents.
Line slack was quantified as the shortening rate (Saito
1973), or sagging rate (Suzuki et al. 1977), equal to the
horizontal distance between floats divided by the length
of line per basket (a dimensionless ratio). At deploy-
644
Fishery Bulletin 90(4), 1992
ment, the shortening rate was the same as the ratio
of ship speed through the water to line-thrower speed:
0.40 (maximum slack) to 0.98 (no slack). The length of
main line per basket was 640-1427m. The predicted
maximum depth of the main line during each set was
calculated from the shortening rate and the main line
length per basket (Table 1), assuming a catenary shape
(Yoshihara 1954).
The depth of each set was recorded with electronic
TDRs (Wildlife Computers, models MKII and MKIII)
programmed to sample depth once per minute. The
TDRs were attached at the deep positions, defined as
the attachment points for the branch line midway be-
tween floats (e.g., position 10 or 11 of 20 between
floats). In 1990, TDRs were also attached at the middle
positions between the deep positions and the float line
(e.g., at position 5 or 15 of 20 between floats).
The time that the gear took to sink during deploy-
ment (0.5h) and to rise during recovery (0.5h) was
quantified from TDR records. Set depth was described
as the typical depth observed in records from the deep-
positioned TDRs during the period after sinking and
before rising. Recorded depth was examined after each
set and compared with predicted depth. Shortening
rate, the length of line per basket, or both were ad-
justed in the subsequent set to reach targeted depths.
Hook depths
The settled depth of each attachment point for the
branch line was estimated by interpolating between (1)
the TDR record for the deep and middle positions or
(2) the latter point and the shallowest depth of the main
line (assumed to equal the length of the float line).
Settled hook depth was calculated by adding the branch
line length to the interpolated depth of the branch line
snap. Not enough TDRs were available (2 in 1989, 10
in 1990) to put 1 TDR on every basket. When fish were
caught by baskets without TDRs, average TDR depths
for that set were used to interpolate settled hook
depths. For middle positions without TDRs in 1989,
depth was estimated from the mean ratio of the middle
position to deep-position TDR depths based on 1990
data.
Hook timers
Hook timers were made of a plastic resin cast around
a battery-powered microchip clock controlled by a
magnet (Somerton et al. 1988). They were attached to
the branch lines near the snap, bridging a bend in the
line (Fig. 1). A fish striking the hook pulled the magnet,
thus triggering the timer. In 1989, a rubber band held
the magnet in place against a test weight of about
l-2kg. In 1990, thread with a breaking strength of
Main line
Hook-
timer
Snap
Breaking
thread
trigger
Holding
tape
Holding y^
thread ^
Figure I
A hook timer and its trigger mechanism as
arranged in 1990, when thread triggers were
used. In 1989, rubber bands served as the
trigger. The slack loop in the branch line was
pulled taut when a fish struck the hook,
breaking the trigger and pulling the magnet
from its recess in the bottom of the hook
timer.
4-5 kg bridged the bend in the line, and the magnet was
held in place by a weaker thread until the bridging
thread was broken (Fig. 1). Some branch lines were set
without timers (14% in 1989, 35.5% in 1990) to pre-
clude interruptions in fishing when timers were tangled
or otherwise unavailable.
Hook timers indicated elapsed time in whole minutes
(e.g., Omin indicated 0-59 s). Timers were read as the
branch lines were recovered, or soon after, with cor-
rections made for delays. Timers were categorized as
being triggered (1) at recovery (< 1 min before remov-
ing the branch line snap), (2) while rising (>lmin-
0.5h before recovery), (3) while settled (>0.5-<1.0h,
l-<2, 2-<3h, and so on before recovery), (4) while
Boggs Estimating capture depths of longline-caught pelagic fish
645
sinking (<0.5h after gear deployment), (5) at deploy-
ment (< 2 min after deployment), and (6) before deploy-
ment (timer triggered before setting commenced).
Timers activated but without fish were categorized
similarly except all settled categories (>0.5-9.0h) were
combined. Untriggered hook timers with fish also were
tallied, and hooks with timers that were damaged,
broken loose, or tangled too badly to be triggered were
counted as hooks without timers.
The numbers of fish caught while the gear was sink-
ing, settled, and rising were summarized. The uncon-
firmed depth of capture of each fish was defined as the
settled depth of the hook. Capture depths were con-
sidered confirmed only if hook timers indicated the cap-
ture occurred within the period in which the gear was
settled.
Catch and effort
Live fish that were not needed as specimens were
tagged and released. Steel head "H" type dart tags
(Squire 1987) were applied using a 3m tagging pole
while the fish remained in the water. Billfish were also
injected with 5-20 mg oxytetracycline/kg of fish using
pole-mounted syringes (Foreman 1987) to mark hard
parts for validation of growth increments. Fish were
released by cutting the branch lines close to the hooks
with a tree-trimming pole. The condition (alive or dead)
of the retained fish was noted, and it was weighed to
the nearest 0.5kg or measured to the nearest 0.1cm.
For the five most-frequently-caught species of commer-
cial importance, catch, number of hooks, and number
of hooks with timers were stratified into 40 m strata
(40- < 80 m, 80- < 120 m, and so on) based on settled
hook depths. The catch-per-unit-effort (CPUE) in each
depth stratum was examined in two ways: (1) by con-
firmed capture depth (CPUEd in number of fish/1000
hooks with timers) representing the depth distribution
of fish; and (2) by settled hook depth (CPUEh in num-
ber of fish/ 1000 hooks), representing the total effec-
tiveness of hooks while sinking, settled, or rising.
The CPUEh for each depth was used to predict
catch rates of "standardized" types of gear to illustrate
the use of catch by hook position in estimating rela-
tive gear efficiency for different gear configurations.
Total CPUE for each standardized gear configuration
was estimated by calculating the weighted average
CPUEh , with weights corresponding to a given num-
ber of hooks per depth stratum for each configuration.
Total CPUE was calculated from 1989 and 1990 data
separately and averaged. Gear efficiency was calcu-
lated as the ratio of the predicted CPUE for each con-
figuration to that of the regular configuration.
Standardized regular and deep longline gear config-
urations were assumed to have 6 and 13 hooks/basket,
respectively. A shortening rate of 0.6 and the dimen-
sions in Suzuki et al. (1977; without adjustment for cur-
rents) indicated hook depths of about 95, 140, and
170 m (for regular gear) and 100, 145, 190, 230, 265,
290, and 300 m (for deep gear). These depths corre-
spond roughly to the midpoints of hook depth strata
in the present study (100, 140, 180, 220, 260, and
300 m).
In addition to regular and deep gear, CPUE values
for two hypothetical gear types were predicted: (1)
shallow gear for which hooks are limited to the first
three depth strata of this study; and (2) a proposed new
gear for which no hooks would be deployed in the first
three depth strata and the distribution of deeper hooks
would match that of deep gear. The shallow gear con-
figuration may be representative of that achieved by
Hawaii's longline fishermen in 1989 and 1990 when
they first began using monofilament longline and had
difficulty achieving the depths formerly fished with
traditional rope gear. With the rope gear, slack was
obtained by manually throwing the baskets with the
main line partially coiled. The [predicted] CPUE for the
new gear type was estimated to indicate the reduction
in bycatch of some species by the elimination of shallow
hooks.
To show CPUE as it would appear in a study of gear
configurations without hook position, capture depth,
or capture time information, CPUEs values were cal-
culated from catch and effort by set type. Sets were
categorized on the basis of depth (TDR depth plus
branch line length) into three groups: 60- < 200 m (reg-
ular), 200- < 330 m (deep), and 330-530 m (very deep).
The first two groups contained depth ranges roughly
comparable to those expected for regular and deep
longline gear types, assuming a variety of shortening
rates and variation due to ocean currents (Suzuki et
al. 1977).
Oceanography
Vertical temperature structure in the area of each set
was measured by expendable bathythermographs
(XBTs; 400 m depth) and conductivity-temperature-
depth casts (CTDs; 500- 1000 m depth, usually 500 m)
before or after each set. Water samples were taken
with Niskin bottles to measure DO and to calibrate DO
measurements made by CTDs.
Many of the TDRs were equipped with a second chan-
nel to record temperature. The TDRs were attached
to the CTD probe to calibrate depth and temperature
measurements. The TDR temperature data were used
to estimate set depths exceeding 400 m (the lower limit
for accurate range depth measurement from the
TDRs).
646
Fishery Bulletin 90(4), 1992
Table 2
Catch data for 14 frequently-caught
taxa in research longline sets off Hawaii.
January
1989 and January-February
1990
Some
weights were calculated from length
measurements
some fish (i
.e., those released) were
not weighed.
Weight (kg)
No. weighed
or
Species
No
caught
measured
Average
Range
Al
ve (%)
Bigeye tuna Thunnus obesus
76
32
31.5
2.5-69.5
83
Yellowfin tuna T. albacares
16
11
39.5
7.5-62.5
63
Skipjack tuna Katsuwonus pelamis
5
5
9.0
7.0-11.0
20
Wahoo Acanthocybium solandri
4
3
17.5
7.5-25.0
0
Striped marlin Tetraptur%is audax
67
20
18.0
9.5-37.0
71
Spearfish T. angustirostris
41
23
13.5
8.5-18.5
56
Mahimahi Coryphaena hippurus
90
60
6.5
2.5-16.0
88
Pomfrets (Bramidae)
17
15
5.5
2.0-10.0
86
Lancetfish Alepisaunis ferox
132
111
1.5
0,1-8.0
64
Ribbonfish Trachipterus ishikawae
4
2
8.0
7.0-8.0
75
Brown ray Dasyatis violacea
8
4
2.0
1.0-2.5
88
Whitetip shark Carckarhinus longimanus
26
—
—
—
85
Blue shark Prionace glauca
21
1
68.0
—
100
Thresher shark Alopias spp.
6
1
91.0
—
60
Results
A total of 16 longline sets caught 149 fish in 1989
and 22 sets caught 401 fish in 1990. Fishing ef-
fort totaled 14,410 hooks including 10,236 hooks
with timers. There were 14 taxa for which more
than 3 fish were caught (Table 2).
Achieving deep sets when intended was some-
times difficult. Backlash of the main line into the
hydraulic line thrower created problems at high
thrower speeds, and ship speed through the water
was sometimes underestimated, reducing the
shortening rate. Wind and currents reduced set
depth by dragging floats and parts of the line in
opposing directions. In particular, current shear
between the surface and the waters below the
thermocline, observed with an acoustic Doppler
current profiler, seemed to prevent deep sets.
Observed set depths were highly variable and
usually less than the predicted depths (Table 1,
Fig. 2). For example, at a predicted depth of
about 490 m, observed depths were 200-400 m
(Fig. 2). Sets averaged only 54% and 68% of the
predicted depths in 1989 and 1990, respectively.
For the first three sets in 1989, the TDRs failed,
so depth was estimated as a percentage of the
predicted depth based on the average percentage
(49.3%) obtained from the next three sets with
similar configurations.
Capture depths
Capture depths were confirmed for those fish caught
>0.5h after deployment and >0.5h before retrieval,
600
/
^^
/
/
B 500
y 0
/
^ — '
/
^
/
+j 400
/ a
a
" / D
0)
A
'^ 300
'' 0
/ a
X)
/o °/ ' I 0
<u
/ <-
> 200
/ . 0 * »
S~i
^ »0 D
0)
y 0
/
c«
/
^ 100
/ 0 □
O
/
/
0 100 200 300 400 500 600
Predicted depth (m)
Figure 2
Relationship between predicted and observed set depths in 1989 (G)
and 1990 (<>). Observed depths were measured with time-depth
recorders, and predicted depths were calculated from the shorten-
ing rate and the main-line length per basket, assuming a catenary
shape.
because the TDR records showed that the main line
usually took 0.5 h to sink to within about 90% of its
settled depth and about 0.5 h to rise to the surface dur-
ing retrieval (Fig. 3). Records of settled depth some-
times varied <100m for the deep sets (e.g., set 14;
Boggs: Estimating capture depths of longline-caught pelagic fish
647
0
1:
a
1 ;
: '^ ' ^
1 ~ ; 1 \
100
1 ;
1 > ■ 1 f -
A
r.
'\s ' > ^
■^
I-
l\\ I / :•/ ,
Oi
V^«<»w-:»-^.9«>!l*»«U^ '
0)
73
200
I
Q
300
-
■ l'-" .•■ ,•■'■ ■■'•• ••
H
k -■■•■■••■•'■' '""■' — Set 6
ft
4)
400
■~
"■■'"•'■•■'■ — Set 16
(U
Set 14
Q
500
,
8 9
10 11 12 13 14 15 16 17 18 19 20 21
Time of day (h)
Figure 3 (left)
Sample records from time-depth recorders (TDRs)
measuring the deep positions on three sets (each with
two TDRs) in 1989, illustrating the typical sinking
time (0.5 h), variation in settled depth, and typical ris-
ing time (0.5 h).
Figure 4 (below)
Hook depths for catches of 14 frequently-caught taxa
in a study off Hawaii, winter 1989 and 1990 (com-
bined). Settled hook depths are shown for all hooks
that caught fish (unconfirmed) and for those hooks
that caught fish while settled (i.e., not sinking or ris-
ing) as indicated by hook-timer data (confirmed).
Fig. 3) and <40m for the regular sets.
Also, the gear sometimes took more
than 0.5h to rise (e.g., the first TDR on
set 16; Fig. 3) or sink. Such deviations
contributed to the variation in estimated
capture depths. Capture depths of fish
caught with hook timers on baskets with
TDRs were based on the TDR depth
at the time of capture; however, most
catches were made by baskets without
TDRs.
A comparison between the uncon-
firmed depths of all hooks that caught
fish and those confirmed to have caught
fish while settled (Fig. 4) showed that
without hook-timer confirmation, many
fish appeared to be caught at greater
depths than they actually were. For ex-
ample, mahimahi Coryphaena hippurus
had unconfirmed capture depths of
<420m and confirmed capture depths
of <190m. Most confirmed capture
depths were <100m for mahimahi and
skipjack tuna Katsuwonus pelamis.
Striped marlin Tetrapturus audax,
whitetip shark Carcharhinus longima-
nus, blue shark Prionace glauca, and
wahoo Acanthocybium solandri had un-
confirmed depths of < 350-420 m and
confirmed depths of < 200-230 m. Spe-
cies having a preponderance of con-
firmed capture depths of < 150 m were
yellowfin tuna Thunnus albacares,
striped marlin, and spearfish Tetraptu-
rus angustirostris. Most confirmed cap-
ture depths were >200m for thresher
0
8 ^ 1 1 «
%
o i
* »
o
100
• ° » * 1 ^
%
o
o
9
« » S 8
° • J h
♦ i o
go*
\
%
8
•
o
■
; •
, — .
200
o
o
o
8
» 1
■
a
8 '^
t
O
8
i
1
300
\ "
o
o
z
1
o o
o
o
o
o
o
O i
t
>
«
400
%
>
• ;
^
*
o
o
•
■4->
o
500
o
>
o
600
0
Unconfirmed
o
1
o
1
o
1 t
M
o
o
100
- § o i 1 °
a
a
s
\
a
^
B a «
g
e
□
e
□
D
§
200
° a 0
a
0
D
1
3
1
-d
1
0)
300
-
0
■
B
^H
a
§
3
-|J
a
a
+J
□
Q)
400
-
a
m
500
600
Confirmed
1 1
"
' "
* >> ■:3 ci ^
a
O es
ja
^
^
XI
* xa
Cl (15 ^ ■" i.
u
° ti
m
u
tn
d to
§ ■- 2 t; 2
(fl
^ 3
''*-t
en
<*-)
%-t
3 C
■^ rt S 3 ^
^ s 3 a
(K £
ttf
a;
X3
a
o
■^ d
g, X
kipjac
Bro
Ma
riped
tiitetip
n
o
Xi
CO
u
a,
g? X
>>
648
Fishery Bulletin 90(4). 1992
shark Alopias spp., pomfrets (Bramidae; species in-
cluded Taractichthys steindachneri, Taractes rubes-
cens, and Eumegistus illustris), lancetfish AiepisaurMs
ferox, and bigeye tuna (Fig. 4).
Capture times
Most of the fish (except ribbonfish Trachipterus ishi-
kawae and brown ray Dasyatis violacea) were caught
o
e
u
c
o
3
o
so-
ls
16 -
14
12
10
8
6
4
2
0
45
40
35 -
30
25
20
15
10
5
0
11
10
9
B
7
6
5
4
3
2
1
0
E3 Skipjack tuna
m Yellowfin tuna
EZD Spearfish
EIZJ Striped marlin
HB Bigeye tuna
[X] Ribbonfish
E2 Wahoo
cm Pomfret
I — I Mabimahi
Kaa Lancetfish
hm
m.
rx~] Brown ray
rrn Thresher shark
CZZI Blue shark
^ ffhitetip shark
Hffl
omI
>-. 'S Si si A jsjaj:ij3j3ja
OS in — * fv
Tf in ^ r^ CO CT^
I I I I I I
rr) -t in *i3 r^ CO
£ '*
Figure 5
Hook-timer data for 14 taxa caught off Hawaii, winter 1989
and 1990 ((.■omhined). Height of each bar represents the sum
of frequencies for each taxa (stacked bars). Hook timers were
either not triggered, triggered while the gear was being set
or recovered, or triggered by fish caught while gear was sink-
ing, settled (0.5-9. Oh before recovery), or rising.
1/5
o
c
u
c
CI
3
cr
a
u
b
12
10-
8-
ESa Live bigeye tuna
■1 Dead bigeye tuna
CD Live striped marlin
■i Dead striped marlin
1
EZ2 Live spearfish
^ Dead spearfish
u
MJ
>> Xi
-H tA) no
IP ^ !~-
^ r\j n
Figure 6
Condition (alive or dead) of three important species in rela-
tion to the elapsed time between capture and recovery as in-
dicated by hook timers, during a study off Hawaii, winter 1989
and 1990 (combined).
Boggs: Estimating capture depths of longline-caught pelagic fish
649
while the gear was settled rather than whOe it was sink-
ing or rising (Fig. 5), probably because the gear spent
much more time in the settled position. However,
substantial numbers of mahimahi, billfish, and other
species were caught while the gear was rising (Fig. 5),
which explains how fish were caught on deep-positioned
hooks (unconfirmed depths) when their confirmed
depth distribution was shallow (confirmed depths;
Fig. 4).
For many species, catch-per-unit-time (CPUT) may
have been highest while the gear was rising. The CPUT
values at <3 and ^3h before recovery were not directly
comparable because short sets resulted in lower effort
(number of hooks with timers) >3h before recovery.
The catch in the 1-2 h and 2-3 h periods (Fig. 5) must
be divided by 2 for comparison with the CPUT in the
0.5h and 0.5-<1.0h periods. For periods of 0.5-1.0,
1-2, and 2-3h before recovery, CPUT values were less
than during the rising period for ribbonfish, pomfrets,
mahimahi, lancetfish, striped marlin, spearfish, brown
ray, and whitetip shark (Fig. 5). In contrast, the values
for yellowfin and skipjack tunas were not much dif-
ferent between settled and moving gear and were
highest for bigeye tuna 1-2 h and 3-4 h before recovery.
Blue shark CPUT values were highest l-2h before
recovery.
Relatively large numbers of fish were categorized as
caught at recovery (Fig. 5). However, estimates of the
delay between recovery and reading each timer were
not precise (± 1 minute). Thus some fish caught "at
recovery" actually had timers triggered after recovery.
Hook timers that did not catch fish were most often
triggered "at recovery" (Table 3), suggesting that
handling activated the timers. A similar lack of preci-
sion affected capture times "at deployment."
Hooks with timers triggered by small fish or without
catching fish may have resulted in false capture times
if larger fish were caught later on those hooks. For-
tunately, small (<10kg) fish, particularly lancetfish
(Table 2), were most frequently caught without trig-
gering the timers (Fig. 5). It was unusual for larger
(~10-90kg) fish, such as tunas, billfishes, or sharks
(Table 2, Fig. 5), to be caught without triggering the
timers. The increase in breaking strength of the trig-
gers in 1990 (Fig. 1) decreased the relative number of
small fish that triggered timers, and reduced the pro-
portion of timers triggered without catching fish from
18.5% in 1989 to 9.7% in 1990 (Table 3).
Survival and release
Over 56% of the fish other than wahoo and skipjack
tuna were alive when recovered, and for most species,
survival was higher than 70% (Table 2). Based on fish
with hook-timer data, over half of the bigeye tuna
recovered up to 9 h after capture were alive. None of
the 11 bigeye tuna recovered 1-2 h after capture were
dead, and the shortest period between capture and
recovery of dead bigeye tuna was 2-3 h (Fig. 6). Striped
marlin were less hardy, with over half recovered dead
> 3 h after capture; nevertheless, many were recovered
alive up to 6-8 h after capture (Fig. 6). Spearfish were
the least hardy: The longest survival time was 5-6 h,
and dead fish were recovered at < 1-2 h after capture
(Fig. 6).
Of the 29 bigeye tuna, 35 striped marlin, and 11
spearfish tagged during the study, 2 bigeye tuna and
1 striped marlin were recaptured 3-10 months later.
These three fish were tagged after having been on
branch lines for 3-6 h. The marlin had been injected
Table 3
Frequencies of activated hook timers on branch Mnes without fish (as percentage of total timers) categorized by elapsed time since
the timers were triggered (range of values from individual sets in parentheses).
Year
Elapsed time
Before retrieval
After deployment
At retrieval
(•Slmin)
Rising
(>l-30min)
Settled
(>30min)
Sinking
(•eSOmin)
At deployment
(<2min)
Activated
before
deployment
No.
timers
Branch lines
with timers
(%)
1989
1990
Combined
6.4
(-)*
3.8
(1.1-6.2)
4.7*
1.2
(0-3.2)
1.5
(0-4.8)
1.4
3.9
(0.8-7.1)
2.6
(0.8-6.5)
3.1
1.0
(0-5,7)
0.4
(0-1.1)
0.6
4.8
(1.6-7.1)
1.0
(0-2.6)
2.4
1.0
(0-3.3)
0.4
(0-2.4)
0.6
3744
(126-356)
6492
(167-418)
10236
86.0
(61-100)
64.5
(34-99)
71.0
'Number recorded only during the first set in 1989. To calculate the combined frequency (4.7%), frequency was assumed to be 6.4%
throughout 1989.
650
Fishery Bulletin 90(4), 1992
with 6mg/kg oxytetracycline, but no flourescent mark
was found in the otohth or vertebrae.
Abundance in relation to depth
For bigeye tuna, the depth distribution of CPUEq
values (number of fish/1000 hooks with timers) was
similar in 1989 and 1990, with CPUEd highest at
360-400 m and relatively high at 200-400 m (Fig. 7).
Hooks with timers triggered before or at deployment
could not be subsequently triggered; therefore, they
were counted as hooks without timers when CPUEd
was calculated (Table 4). The data from wider depth
ranges in both years were pooled to obtain sample sizes
(number of hooks with timers; Table 4) large enough
to determine whether significant differences existed
between depths (Fig. 8). For bigeye tuna, CPUEd
values were significantly higher at depths of >200m
than at depths of <200m (P<0.05, based on 95% CI
for the difference between proportions). Few (12%) of
(n
0)
o
o
A
o
o
o
6
a
20
16
12
8
4
0
4
3
2
1
0
8
6
4
2
0
8
6
4
2
0
16
12
8
4
0
Bigeye tuna
^n rki
Nm
Yellowfin tuna
li
Striped marlin
Ik
^
Spearfish
l^Jl
Mahimahi
O
CO
o
r\i
I
o
oo
o
I
o
(V
o
o
o
I I
o
o
o
o
CO
(M
I
o
o
o
I I
-H ^ (V (V
o
CO
o
o
o
I
o
ro
+
o
o
Hook depth (m)
Figure 7
Indices offish abundance vs. depth, calculated as the number
of fish caught/lOOO hooks with timers (CPUE,,) in ten depth
.strata. Indices for li)89 (open bars) and 1990 (crosshatched
bars) are based on fish captured while the gear was settled
(as confirmed by hook timers) (Table 4).
(V
in
o
o
o
o
o
d
Hook depth (m)
Figure 8
Indices of fish abundance vs. depth, calcu-
lated as the number of fish caught/ 1000
hooks with timers (CPUE,,) in four pooled
hook-depth strata. Indices for 1989 and 1990
combined are based on fish captured while
the gear was settled (as confirmed by hook
timers) (Table 4). Error bars indicate 95%
CI of the CPUEp values for each depth
category.
Boggs Estimating capture depths of longline-caught pelagic fish
651
the bigeye tuna with hook timers were caught while
the hoolts were moving (sinking or rising; Table 4). No
clear relationship existed between depth and the pro-
portion caught on moving hooks (Table 4).
Yellowfin tuna were not very abundant, which is
typical for the winter months off Hawaii. The CPUEd
for yellowfin tuna was not the same in both years (Fig.
7), but the number of fish caught with timers was very
small, particularly in 1989 (Table 4). Pooled CPUEd
was highest at 40-200 m (Fig. 8) although no signifi-
cant difference in CPUEo by depth was found. The
40 m end of the depth range did not indicate the
shallowest depths preferred by any species, since no
hooks fished depths of <40m.
Timer-confirmed catches by settled hooks indicated
the highest catch rates for striped marlin were at
40-120 m in both years (Fig. 7), and pooled CPUEd
was clearly the highest at this depth range (Fig. 8). The
overall proportion of striped marlin caught on moving
hooks was high (32%; Table 4) and increased with
depth. At > 120 m most striped marlin were caught by
moving hooks, and at >200m only one was caught by
a settled hook (Table 4).
For spearfish, the pattern of CPUEd vs. depth dif-
fered between years. In 1989, the highest CPUEp was
at 120- 160 m although several fish were caught as deep
as 280-360 m; however, in 1990 the highest CPUEq
was at 40-80 m, and no confirmed capture depths were
recorded at >200m (Fig. 7). Pooled data suggested that
spearfish were more abundant at <120m, but the
CPUEd at 40-120 m was not significantly different
from that at 120-200 m (Fig. 8). In 1989, a large pro-
portion (43%) of the spearfish were caught on moving
hooks, but none were caught on moving hooks in 1990
(Table 4). Furthermore, for each of the major species,
a higher proportion of fish were caught on moving
hooks in 1989 than in 1990 (Table 4). An early report
(Boggs 1990) on this research was based on 1990 data
(Table 4) wherein only 12% of the tuna and billfish
(combined) were caught on moving hooks.
Table 4
Catch of five commercially-important
species in
research longline
sets off Hawaii, 1989 and 1990, giving
fishing effort (number of 1
hooks an
i timers) by depth strata, number caught (N) on known ht
ok position, number confirmed by timers to be
caught on
moving
(M; i.e.. sinking or rising) and settled (S) hooks (in paren
;heses), anc
percentage of fish caught on moving hooks. Depth
ranges
include
branch line length. Timers do not include those
triggered a
, or before deployment. Catch totals are sometimes less than in
Table 2
and Figure 5 because hook number and depth were not known for a few fish.
Hook
depth
Hooks
in
Timers
in
Bigeye tuna
Moving
Yellowfin tuna
Striped marlin
Moving
Spearfish
Vlahimahi
Moving
Moving
VIoving
(m)
Year
stratum
stratum
N (M,S)
(%)
N (M,S)
(%)
N
(M,S)
(%)
N (M,S)
(%)
iV
(M,S)
(%)
40-80
1989
546
489
1 (0,0)
0
(0,0)
5
(0,2)
0
4(1,1)
50
3
(0,0)
1990
1214
822
0(0,0)
-
4
(0,2)
0
16
(1,5)
17
8(0,7)
0
41
(5,14)
26
80-120
1989
684
586
0 (0.0)
—
1
(0,1)
0
9
(1,3)
25
4(1,1)
50
2
(0,1)
0
1990
1612
1026
2(0,1)
0
0
(0,0)
-
15
(2,7)
22
5(0,4)
0
14
(1,7)
12
120-160
1989
658
558
2(0,2)
0
0
(0,0)
—
6
(2,1)
25
5(1,3)
25
1
(0,0)
—
1990
1611
1007
0 (0.0)
-
3
(0.2)
0
3
(1,1)
50
3(0,2)
0
5
(1,3)
25
160-200
1989
350
281
1 (0,1)
0
0
(0,0)
—
2
(1,1)
50
1 (1,0)
100
0
(0,0)
—
1990
1812
1151
6(1,2)
33
4
(0,1)
0
5
(1,0)
100
5(0,1)
0
6
(0.1)
0
200-240
1989
553
427
6(1.3)
25
1
(1,0)
100
2
(0,0)
—
1(0,1)
0
1
(1.0)
100
1990
1160
748
11(1.6)
14
1
(0,0)
-
1
(0,1)
0
0 (0,0)
-
4
(0,0)
-
240-280
1989
524
406
5(0,3)
0
1
(0,1)
0
0
(0,0)
—
0(0,0)
—
1
(0,0)
—
1990
1321
795
11 (0,6)
0
1
(0,0)
-
0
(0.0)
-
0(0,0)
-
6
(1.0)
100
280-320
1989
353
279
3 (0, 1)
0
0
(0,0)
—
0
(0,0)
—
1 (0,1)
0
1
(0,0)
—
1990
626
395
7(0,4)
0
0
(0.0)
-
0
(0,0)
-
0(0,0)
-
1
(0,0)
—
320-360
1989
384
288
7(1.4)
20
0
(0.0)
—
2
(1,0)
100
4(2,1)
67
0
(0,0)
—
1990
198
152
0 (0,0)
-
0
(0.0)
-
0
(0,0)
-
0(0,0)
-
0
(0,0)
-
360-400
1989
214
148
6(1,3)
25
0
(0.0)
—
0
(0,0)
—
0(0,0)
—
0
(0,0)
_
1990
108
84
2(0.1)
0
0
(0,0)
-
0
(0,0)
-
0(0,0)
-
1
(0,0)
—
400 +
1989
87
60
2(0,0)
0
0
(0,0)
—
0
(0,0)
—
0(0,0)
—
0
(0,0)
—
1990
276
225
2(0,0)
-
0
(0,0)
-
0
(0,0)
-
0(0,0)
-
2
(1,0)
100
Total
1989
4352
3522
33(3,17)
15
3
(1,2)
33
26
(5,7)
38
20 (6, 8)
43
9
(1,1)
50
1990
10,058
6402
41 (2,20)
9
13
(0,5)
0
40
(5,14)
26
21 (0,14)
0
80
(9,25)
26
Combined total
14,410
9924
74 (5,37)
12
16
(1,7)
12
66(10,21)
32
41 (6.22)
21
89(10,26)
28
652
Fishery Bulletin 90|4). 1992
Figure 9
Comparison between indices of fish abundance from different types
of longline sets, calculated as the number of fish caught/1000 hooks
without regard to capture depth of individual fish (CPUEg). Three
types of longline sets were categorized on the basis of the deepest
hooks, but every set contained some hooks as shallow as 40-80 m.
Data are combined for 1989 and 1990.
Although relatively few mahimahi were captured
with timer data, these data indicated maximum abun-
dance was at 40-80 m in 1990 (only one fish was caught
on a settled hook in 1989; Fig. 7, Table 4). Pooled data
clearly indicated that CPUEd was highest at 40-120m
(Fig. 8). At >200m, all mahimahi with timer data were
caught on moving hooks.
Examining the CPUEg data as if the only available
depth information were the set type (Fig. 9) made it
difficult to correctly qualify the relative abundance of
fish in relation to depth. For example, mahimahi ap-
peared almost as abundant in deep sets as in shallow
Regular
(60-200 m)
10 -
o
o
o
*> flj t- 2 *^
CO .3
o a.
a B
Deep
taOO-330 m;
Very Deep
(330-5300 m)
Aa
C = o jd ."J
G .,3 a « ?
o o.
&£
03
^
Latitude (°N)
sets, and spearfish appeared more abun-
dant in very deep than in deep sets, il-
lustrating that it is impossible to cor-
rectly describe fish depth distributions
without data on catch by hook position,
hook depth, and capture time.
Oceanographjc habitat
The temperature profile in 1990 (Fig.
10) was representative of the study area
in both years, except the bottom of the
thermocline (i.e., the 12°C isotherm)
was ~40m deeper in 1989. In both
years, the highest catch rate of bigeye
tuna with confirmed capture depths
occurred at lat. 17°-18°N at 360-400m
in temperatures of 8°-10°C (Fig. 10).
The oxycline in 1990 (Fig. 10) also was
similar to that in the previous year (i.e.,
the S.Omg/L isopleth was only 10-20 m
deeper in 1989). Most bigeye tuna were
caught at DO concentrations of 2-6
mg/L. In both years, the highest catch
rate was at 2-3 mg/L.
Figure 10
Temperature and dissolved oxygen profiles of the
study area (lat. 14°-20''N, long. 148°-159°W) in
1990 (similar to 1989). Confirmed capture depths
of bigeye tuna in 1989 (D) and 1990 (<>) are
indicated.
Boggs: Estimating capture depths of longline-caught pelagic fish
653
Table 5
Standardized distribution of hooks by depth stratum for four standardized gear types
and predicted catch-per
■unit-effort (CPUE) for 1
each gear type based on the
weighted average
observed CPUE by hook depth for five commercially
-important species in research
longline
sets off Hawaii, 198£
and 1990. For each species, relative gear efficiency is
given as
the ratio between the CPUE for each
gear type and for regular gear (too few yellowfin tuna (A^ 3)
were caught in
1989 to warrant calculating relative gear efficiency).
Bigeye
Yellowfin
Striped
Shortbill
Model
gear
Hook number
ay depth (m)
tuna
tuna
marlin
spearfish
CPUE
Mahimahi
40-
80- 120- 160-
200-
240-
280-
CPUE
CPUE
CPUE
CPUE
type
80
120 160 200
240
280
320
Year
CPUE
ratio
CPUE
ratio
CPUE
ratio
CPUE
ratio
CPUE
ratio
Regular
0
2 2 2
0
0
0
1989
1.93
1.00
0.47
9.30
1.00
5.37
1.00
1.43
1,00
1990
1.52
1.00
1.36
1.00
4.83
1.00
2.57
1.00
5.03
1.00
X
1.73
1.00
0.91
7.07
1.00
3.97
1.00
3.23
1.00
Deep
0
2 2 2
2
2
3
1989
5.95
3.08
0.78
4.85
0.52
3.50
0.65
1.88
1.31
1990
6.02
3.97
0.88
0.65
2.36
0.49
1.19
0.46
4.45
0.88
X
5.99
3.53
0.83
3.60
0.51
2.34
0.56
3.16
1.10
Shallow
2
2 2 0
0
0
0
1989
1.60
0.83
0.47
10.43
1.12
6.83
1.03
3.23
2.26
1990
0.41
0.27
1.72
1.27
8.12
1.68
3.85
1.50
15.18
3.02
X
1.01
0.55
1.09
9.28
1.40
5.34
1.39
9.21
2.64
New
0
0 0 2
2
2
3
1989
7.93
4.10
0.82
2.04
0.22
2.09
0.39
1.76
1.22
1990
8.42
5.55
0.85
0.63
0.93
0.19
0.61
0.24
3.81
0.76
X
8.18
4.82
0.84
1.49
0.20
1.35
0.32
2.78
0.99
The area (lat. 17°-18°N) of highest catch rates for
bigeye tuna was on the south edge of a northward tran-
sition to a deeper thermochne and oxycHne (Fig. 10).
The north-south pattern is typical of the central Pacific
Ocean at these latitudes, whereas the highly variable
pattern in the thermocline between lat. 19.4° and 20°N
was probably caused by the proximity to the lee side
of the island of Hawaii.
With regard to the other species, the thermal struc-
ture of the habitat (Fig. 10) and the confirmed depth
distribution of fish (Figs. 4, 7, and 8) suggested that
yellowfin tuna were most abundant in the mixed layer
(24°-25°C) and the steepest part of the thermocline
down to about 15°C. Striped marlin appeared to be
most abundant in the mixed layer and the top of the
thermocline to ~20°C. Spearfish appeared to occupy
a habitat between that of yellowfin tuna and striped
marlin, and mahimahi occupied the mixed layer.
Standardized gear efficiency
For bigeye tuna in 1989-90, the CPUE ranges for stan-
dardized deep gear and proposed new gear were about
3.1-4.0 and 4.1-5.6 times, respectively, as great as
those for regular gear (Table 5). Shallow gear on aver-
age was about half as efficient as regular gear in catch-
ing bigeye tuna, whereas it was about 40% more effi-
cient than regular gear in catching spearfish and
striped marlin. Deep gear was only about half as effi-
cient as regular gear in catching striped marlin and
spearfish, and the proposed new gear was only about
20% as efficient for striped marlin and about 30% as
efficient for spearfish.
The numbers of yellowfin tuna and mahimahi caught
in 1989 were much lower than in 1990, so the latter
year provided better data for calculating gear efficiency
for these species (Table 5). Shallow gear was about 3.0
times as efficient at catching mahimahi, and deep and
new gear reduced efficiency to about 90% and 75% in
comparison with regular gear. For yellowfin tuna,
shallow gear was about 25% more efficient than reg-
ular gear, whereas the deep and new gear types were
each about 65% as efficient.
Discussion
Habitat deptli
Hook timers are useful in confirming whether fish are
caught while longline hooks are sinking, settled, or ris-
ing. Combined with capture depths from TDRs, hook
timers offer a new method for establishing the habitat
depth of large pelagic fishes. Stock assessments (Suzuki
1989) depend on the estimation of effective effort,
defined as fishing effort corrected for differences in
efficiency due to gear and habitat depth (Suzuki et al.
1977). Improving the definition of tuna and billfish
habitats and the estimation of effective effort in those
habitats should lead to significant improvements in
assessing true abundance.
Comparisons of CPUE by two gear types provide
only qualitative information on habitat depth. For
654
Fishery Bulletin 90(4|, 1992
example, since deep gear is more efficient than regular
gear for bigeye tuna, this species must occupy a rela-
tively deep habitat (Suzuki et al. 1977). More specific
information on habitat depth is provided by catches
and CPUEh by hook position (Hanamoto 1979 and
1987, Hanamoto et al. 1982, Suzuki and Kume 1982),
especially when TDRs are used to record gear depth
(Saito 1973 and 1975, Hanamoto 1974, Nishi 1990).
Capture depth estimates without TDR records ignore
major variations in actual gear depth (Fig. 2; Nishi
1990), and those without hook timers are biased by the
inclusion of inappropriate hook depths.
A possible source of bias in the present study is the
inclusion of some falsely confirmed depths due to fish
being caught with timers already activated. The pro-
portion of false estimates should be similar to the fre-
quency of timers that were without fish and were trig-
gered while settled, which was only 3.9% in 1989 and
2.6% in 1990 (Table 3). Thus it is unlikely that >4%
of confirmed capture depths in this study are incorrect
because of false timer readings.
Many pelagic longline studies (Saito 1975, Hanamoto
1976, Yang and Gong 1988) assume that fish are caught
while hooks are at settled depths. Supporting this
assumption, Saito (1973) has shown that albacore Thun-
nus alalunga are caught almost exclusively by settled
hooks, based on capture times indicated by fluctuations
in TDR records. Using hook timers, the present study
adds new information: Almost 90% of bigeye and
yellowfin tuna also are caught while hooks are at
settled depths (Table 4). However, hook timers indicate
this generalization does not extend to striped marlin,
spearfish, mahimahi (Table 4), and most of the com-
mercially unimportant species (Fig. 5). Although most
of these fish are also caught on settled hooks, a substan-
tial fraction are not, and this must be considered when
quantifying their depth ranges (Fig. 4).
Besides the present study, little information exists
on longline capture depths for mahimahi, spearfish, and
striped marlin. In the study area, CPUEn values for
these species (Fig. 7) indicate maximum abundance at
depths in the mixed layer for mahimahi (<100m,
24°-25°C; Fig. 10), extending into the top of the ther-
mocline for striped marlin (120 m, 20°C) and into the
middle of the thermocline for spearfish. Striped marlin
are reported to be caught most frequently on longline
hooks closest to the surface (60-90 m) in the eastern
tropical Pacific and Indian Oceans, but they may be
more abundant above this depth (Hanamoto 1979,
Hanamoto et al. 1982). Mahimahi and spearfish may
also be more abundant above the uppermost stratum
(40-80 m) in the present study, since their catch rates
appear to increase towards the surface (Fig. 7).
Striped marlin are also reported caught on deep
longline hooks (~200m; Hanamoto et al. 1982) and at
the deep end of vertical longline gear (336 m; Saito
1973); but in the present study, their deepest confirmed
capture depth is 210m. Tracking data on striped marlin
off California indicate a shallow (< 60 m) depth distribu-
tion with most of the daytime spent within 10 m of the
surface (Holts and Bedford 1989).
The depth distribution (200-400 m) of bigeye tuna in
the present study is deeper than in many previous
reports (Hanamoto 1974, 50-160 m; Saito 1975, 207-
245 m; Suzuki and Kume 1982, 1 70-300 m; Yang and
Gong 1988, 260-300m; Nishi 1990, 140-180m), al-
though these studies have found bigeye tuna are most
abundant on the deepest hooks fished. Hanamoto
(1987) hypothesizes a habitat depth of 250-400 m for
the central Pacific Ocean at latitude 25°N, based on
the observed maximum longline CPUE at tempera-
tures of 10°-15°C. The highest CPUEd values in the
present study are at the cold, deep end of this range
(Fig. 7), deeper than most hooks used in commercial
fishing gear. However, the CPUEq value at 280-400
m is not significantly different from that at 200-400 m
(Fig. 8). Although these results may be specific to
January and February, perhaps commercial CPUE
could be improved by increasing fishing depth, at least
during winter months.
Seasonal and geographic variation in temperature
and DO profiles may affect the depth preferences of
pelagic fish. Hanamoto (1975, 1987) has hypothesized
that the deep end of bigeye tuna habitat is limited by
DO concentrations below ImL/L (1.4mg/L) and by
temperatures below 10°C. Results of the present study
suggest that bigeye tuna are seldom caught in waters
with a DO concentration of ~<2mg/L (Fig. 10). Oxy-
gen concentrations of ~2-3mg/L cause significant
reductions in bigeye tuna cardiac output (1.9-2. 6 mg/L)
and heart rate (2.7-3.5 mg/L), suggesting that bigeye
tuna cannot maintain a full range of activity at lower
DO concentrations (Bushnell et al. 1990).
Longline data to support the hypothesis of a 10°C
temperature limit independent of the DO limit are
sparse. Few hooks have been deployed in waters colder
than 9°-10°C with DO concentrations of >lmL/L
(Hanamoto 1975, 1987). In the present study, the only
area with DO values > 2 mg/L and temperatures <8°C
was at lat. 10°-20°N (Fig. 10). Currents prevented
hooks from reaching cold (6°-8°C) water in this area.
Sonic tracking of bigeye tuna around Hawaii indi-
cates a depth distribution slightly shallower than that
in longline studies (Hanamoto 1987, 250-400 m; pres-
ent study, 200-400m). Holland et al. (1990) have
reported that tracked bigeye tuna spend most of the
daytime at 200-240m in 14°-17°C water. This may be
due to the association of the tracked bigeye tuna with
fish aggregating devices or due to a size-related differ-
ence. The 72- to 74 cm bigeye tuna studied by Holland
Boggs Estimating capture depths of longline-caught pelagic fish
655
et al. (1990) weighed ~10-12kg, whereas longhne-
caught bigeye tuna in the present study averaged
>30kg.
Results of the present study apply predominantly to
daytime habitat depths, but an important difference ap-
parently exists between the daytime and nighttime
depth distributions of bigeye tuna (Holland et al. 1990).
At night, tracked bigeye tuna move upward to ~70-
90m at temperatures of 23°-25°C. Confirmation of
this nocturnal behavior comes from a new nighttime
longline swordfish fishery that has recently developed
in Hawaii using chemical light sticks. Although this
fishery deploys very shallow (generally <90m) gear,
the bycatch of bigeye tuna is surprisingly high (S.
Pooley, NMFS Honolulu Lab., pers. commun., April
1991), indicating that bigeye tuna have a shallow night-
time depth distribution.
The small number of yellowfin tuna caught in this
study makes estimated habitat depth (40-200 m) less
certain, but it does not differ much from the 90-230 m
depth found in Suzuki and Kume (1982) and Yang and
Gong (1988). Tracking studies (Carey and Olson 1982,
Holland et al. 1990) show yellowfin tuna spend most
of their time at depths <100m. Depths of the highest
longline CPUEd for yellowfin tuna in the present
study (40-80 m; Fig. 7) are similar to the depths (30-
80 m) at which tracked yellowfin tuna in Hawaii spend
over 50% of their time during the day (Holland et al.
1990), tending to confirm that yellowfin tuna habitat
is mostly in the mixed layer.
Methods for estimating habitat depths in the present
study could be improved by increasing the number of
TDRs deployed or by developing a model, calibrated
with TDRs, to predict gear depth based on wind and
current measurements, divergence or convergence of
floats, and stops and starts in deployment and retrieval.
Procedures to estimate the capture depths of fish
caught while hooks are sinking or rising could also be
developed, but would depend on very accurate time-
keeping, since the gear rises rapidly during retrieval
(Fig. 3).
Catch by moving hooks
The catch of shallow-swimming species on deep hooks
moving through shallower depths could reduce the
selectivity of gear designed to catch deep-swimming
species. The results show that moving longline hooks
are more effective (per unit time) than settled hooks
at catching billfish, mahimahi, some sharks, and most
other non-tuna species. However, the majority of these
fish are caught on settled hooks, because of the longer
time that hooks are settled (Fig. 5). The relative
amount of time hooks are moving vs. settled is the only
aspect of the commercial daytime tuna longline opera-
tions that differs much from the fishing method used
in this study. The gear is left in the water longer and
then retrieved more rapidly during commercial fishing,
so hooks spend less time moving and more time settled.
This may result in greater proportions of fish being
caught on settled hooks by commercial fishermen than
in the present study.
Eliminating shallow-settled hooks should greatly
reduce the catch of shallow-swimming species. For non-
tuna species, deploying and retrieving the gear less
often (as in commercial operations) should decrease the
CPUT (catch-per-unit-time), but would increase the
CPUE because the latter increases with set duration.
In contrast, bigeye tuna CPUT and CPUE should in-
crease with less frequent deployment and retrieval,
because CPUT is highest for settled hooks.
The mechanism for increased CPUT on moving hooks
for non-tuna species is unclear. Moving bait may be
more attractive than settled bait, but the low number
caught on sinking hooks (Fig. 5) suggests that gear
motion alone is not responsible for increased catch rate.
Perhaps a gradual aggregation of fish around the gear
(or the vessel) while the gear is settled contributes to
the catch rate by rising hooks.
Although hook timer data provide a reliable way to
confirm when fish are caught on settled hooks, such
data may be less reliable as a measure of fish caught
on moving hooks, because of the uncertainty regarding
fish with timers triggered at recovery (Fig. 5). These
fish are not included in the number captured on mov-
ing hooks (Table 4); their timer readings cannot be
distinguished from ones triggered after being brought
aboard. Therefore, the estimates of fish caught on mov-
ing hooks (Table 4) may be too low. Alternatively, if
these readings indicate a tendency for some fish to not
activate timers until they struggle during recovery,
then the estimates of fish caught on moving hooks may
be too high. In either case, only inferences regarding
CPUT on moving and non-moving hooks, and the
estimated proportions of fish caught on moving hooks,
are affected by this uncertainty. The estimates of
catches on non-moving hooks are conservative, and
confirmed capture depths are not affected.
The higher proportion of fish caught on moving hooks
in 1989 compared with 1990 (Table 4) could have been
caused by moving hooks being less visible in 1990, since
branch lines were more often recovered after dark
(Table 1). Sets also lasted longer in 1990 (Table 1); this
may have increased the relative proportion of catches
on settled vs. moving hooks. The CPUT in relation to
sinking, settled, and rising gear, and to the time of day,
should be explored further using the techniques devel-
oped in the present study.
A TDR attached to vertical and regular rope longline
gear sometimes records abrupt depth changes as a fish
656
Fishery Bulletin 90(4), 1992
is caught, making the TDR equivalent to a hook timer
if it is close to a branch line that catches a fish (Saito
et al. 1970, Saito 1973, Yamaguchi 1989). The records
of TDRs at positions close to fish caught with hook
timers in the present study were checked to see
whether they indicated the time of capture, but the
depth of the monofilament longline gear was much less
stable (Fig. 3) than in the depth records of Saito et al.
(1970), Saito (1973), and Yamaguchi (1989) using TDRs
on rope gear. On monofilament longline gear, frequent
depth changes resembling fish captures occur even
when no fish are caught, making TDRs unreliable as
substitutes for hook timers.
Viability of released fish
Before the present study, it was believed that fish
would survive only a few hours after capture on
longline gear (Grudinin 1989, Yamaguchi 1989) despite
large pelagic species being known to survive capture
and release from other types of gear (Foreman 1987,
Squire 1987, Holts and Bedford 1989). Commercial
longline fishermen in Hawaii speculated that much of
their catch was made as hooks were sinking or rising,
because most were alive or appeared long dead (F.
Amtsberg, Der Fischen Co., Honolulu, HI 96822, pers.
commun., March 1988). Based on TDR data from fish
on regular longline gear (Yamaguchi 1989), vertical
movements stop 1.0-1.5h after capture for yellowfin
tuna, 1.5-4. Oh for bigeye tuna, and ~0.5h for spear-
fish and shark. This cessation of vertical movement has
been interpreted as death (Yamaguchi 1989). Grudinin
(1989) has reported on the diurnal periodicity of bigeye
and yellowfin tuna catch rates based on the proportion
recovered alive, assuming that tuna survive <2h on
longline gear. However, hook-timer results (Fig. 6)
show that fish survive much longer than this, suggest-
ing that vertical movement is not a reliable indicator
of survival. Alternatively, the results of the present
study could be specific to monofilament gear, which
could have less resistance to moving through the water
than does rope gear.
Clearly the high proportion of live fish (Table 2) is
not primarily the result of capture during the 0.5 h
rising period. The viability of longline-caught fish is in-
dicated by their hooked longevity and the recovery of
tagged fish. As a management option, non-retention
of striped marlin and spearfish could reduce fishing
mortality due to longline fishing. The importance of the
reduction would depend on the length of the fishing
operation; but in the present study, longline fishing
mortality for striped marlin could have been reduced
by 70% (Table 2) if all live fish had been released and
had survived.
Gear efficiency and selectivity
Gear efficiency, defined as the dimensionless ratio of
the CPUE of one gear type (i.e., deep gear) divided by
the CPUE of the regular gear type, is the factor used
to calculate effective effort by gear fishing at different
depths (Suzuki et al. 1977). Total effective effort can
then be used to calculate indices of relative abundance
and to model stock production (Suzuki 1989). The most
thorough approach thus far has been to calculate gear
efficiency by area and season (Suzuki and Kume 1982).
A better understanding of the variables that alter
habitat depth would permit gear efficiency to be pre-
dicted as a function of environmental conditions, and
help account for variation in abundance indices caused
by environmental anomalies.
The relative efficiency of standardized deep gear
(Table 5) follows the pattern observed in previous
studies (Suzuki et al. 1977, Yang and Gong 1988) in
which deep gear is more efficient at catching bigeye
tuna and less efficient at catching yellowfin tuna and
istiophorid billfish. However, the estimated efficiency
of the standardized deep gear for bigeye tuna in the
present study is greater (ratio 3.1-4.0 over the 2 years;
Table 5) than that reported by Suzuki et al. (1977) for
the central and western equatorial Pacific (1.8) or by
Yang and Gong (1988) for the Atlantic (1.9). Suzuki and
Kume (1982) have presented graphs of deep and reg-
ular CPUE for bigeye tuna on a quarterly basis by area
throughout the Pacific, and these data indicate very
little difference between gear types in the central
Pacific north of lat. 15°N. The high efficiency estimated
for deep gear in the present study may partly result
from using measured depths rather than inferred
depths to define deep and regular gear types. Also, a
high relative efficiency for deep gear may be specific
to the Hawaii area in the winter season.
The relative efficiency of deep gear for yellowfin tuna
in the Atlantic (0.95, Yang and Gong 1988) is greater
than in the central and western equatorial Pacific (0.73,
Suzuki et al. 1977) and in the present study (0.65, Table
5). Relative efficiency of deep gear for striped marlin
in the central and western equatorial Pacific (0.28,
Suzuki et al. 1977) is much lower than in the central
Pacific north of Hawaii (0.74, Suzuki 1989), nicely
bracketing the estimate from the present studv (0 51
Table 5).
The model estimates of gear efficiency (Table 5) are
not meant to supplant earlier estimates based on much
larger data sets (Suzuki et al. 1977, Suzuki and Kume
1982, Yang and Gong 1988, Suzuki 1989), but rather
to show how catch by hook position can be used to
estimate CPUE by different gear configurations,
especially hypothetical configurations for which no real
data exist. Efficiency estimates (Table 5) suggest that
Boggs: Estimating capture depths of longline-caught pelagic fish
657
shallow sets of the type hypothesized to represent early
use of monofilament longline gear in Hawaii would be
expected to catch about 40% more billfish and 160%
more mahimahi than would regular longline gear.
Large increases in longline catches of these fish in
Hawaii have occurred in recent years (1989-90, Boggs
1991) as the expanding Hawaii fishery adopted a new
type of gear. The proposed new gear configuration
would be an effective way to reduce the catch of spear-
fish and striped marlin by ~70-80% below that of
regular gear.
Hook timers and TDRs are useful in documenting the
depth distribution and habitat of pelagic fish and in
showing how different configurations of longline gear
and the release of live fish can be effective means of
reducing fishing mortality for some species. Better
methods of identifying the habitats of pelagic fishes
should make it easier to estimate real changes in fish
abundance by accounting for changes in fishing
methods and the environment.
Acknowledgments
Several of Hawaii's longline fishermen, especially
F. Amtsberg, provided technical advice without which
few fish would have been caught. Many of the staff at
the Honolulu Laboratory participated in the cruises,
and their creativity and hard work contributed substan-
tially to this study, especially R.K.C. Chang, A.E.
Chun, R. Ito, L.A. Koch, R.A. Skillman, D. Therry,
J.H. Uchiyama, and S. Yano. L.A. Koch tabulated the
data and produced the figures, and B.S. Kikkawa pro-
vided invaluable help in acquiring and manufacturing
the hook timers. Volunteer assistance on the research
cruises was given by H. Dewer, P. Fields, C. Hayashi,
and A. Sesawa. The officers and crew of the NO A A
ship Townsend Cromwell also were very helpful, espe-
cially LT R. Brainard, LTC B. Dearbaugh, H. Lariosa,
and CDR R. Marriner.
Citations
Boggs, C.H.
1990 Longline research on tuna and marlin. In Relatorio, X
Semana das Pescas dos Acores, March 12-16 1990, p.
249-252. Sec. Reg. Agric. Pescas. Dir. Reg. Pescas, Horta,
Faial, Azores.
1991 A preliminary examination of catch rates in Hawaii's troll
and handline fisheries over a period of domestic longline fishery
expansion. Admin. Rep. H-91-05, NMFS Southwest Fish. Sci.
Cent., Honolulu, 62 p.
Bushnell. P.G.. R.W. Brill, and R.E. Bourke
1990 Cardiorespiratory responses of skipjack tuna (Katsu-
wonus pelamis), yellowfin tuna (Thunnus albacares), and bigeye
tuna {Thunnus obesus) to acute reductions of ambient oxygen.
Can. J. Zool. 68:1857-1865.
Carey, F.G., and R.J. Olson
1982 Sonic tracking experiments with tunas. Int. Comm. Con-
serv. Atl. Tunas Collect. Vol. Sci. Pap. 17:458-466.
Foreman, T.J.
1987 A method of simultaneously tagging large oceanic fish
and injecting them with tetracycline. Fish. Bull., U.S. 85:
645-647.
Grudinin, V.B.
1989 On the ecology of yellowfin tuna {Thunnus albacares) and
bigeye tuna {Thunnus obesu^). J. Ichthyol. 26(6):22-29.
Hanamoto, E.
1974 Fishery oceanography of bigeye tuna— I. Depth of cap-
ture by tuna longline gear in the eastern tropical Pacific
Ocean. La Mer (Bull. Soc. Franco-Japonaise Oceanogr.)
12(3):128-136.
1975 Fishery oceanography of bigeye tuna— H. Thermocline
and dissolved oxygen content in relation to tuna longline fishing
grounds in the eastern tropical Pacific Ocean. La Mer (Bull.
Soc. Franco-Japonaise Oceanogr.) 13(2):58-71.
1976 The swimming layer of bigeye tuna. Bull. Jpn. Soc. Fish.
Oceanogr. 29:41-44. [Engl, transl. no. 21 by Tamio Otsu, 1977,
7 p.: avail. NMFS Honolulu Lab., Honolulu 96822-2396.]
1979 Fishery oceanography of striped marlin— IV swimming
layer in the tuna longline fishing grounds. Bull. Jpn. Soc. Sci.
Fish. 45:687-690.
1987 Effect of oceanographic environment on bigeye tuna
distribution. Bull. Jpn. Soc. Fish. Oceanogr. 51:203-216.
Hanamoto, E., T. Shibata, and H. Okui
1982 Some considerations on the vertical distribution par-
ticularly of the striped marlin in the Indian Ocean and on the
fishing characteristics of the tuna longline gear. Kanagawa
Prefect. Fish. Exp. Stn. Res. Rep. 4:17-21.
Holland, K.N., R.W. Brill, and R.K.C. Chang
1990 Horizontal and vertical movements of yellowfin and big-
eye tuna associated with fish aggregating devices. Fish Bull.,
U.S. 88:493-507.
Holts, D., and D. Bedford
1989 Activity patterns of striped marlin in the Southern Cah-
fornia Bight. In Stroud, R.H. (ed.), Planning the future of
billfishes, research and management in the 90's and beyond.
Part 2: Contributed papers. Mar. Rec. Fish. 13: 81-93. Natl.
Coalition Mar. Conserv., Savannah, GA.
Kawamoto. K.E., R.Y. Ito. R.P. Clarke, and A.E. Chun
1989 Status of the tuna longline fishery in Hawaii, 1987-1988.
Admin. Rep. H-89-10, NMFS Southwest Fish. Sci. Cent.,
Honolulu, 34 p.
Nishi. T.
1990 The hourly variations of the depth of hooks and the hook-
ing depth of yellowrfin tuna {Thunnus albacares). and bigeye
tuna {Thunnus obesus), of tuna longline in the eastern region
of the Indian ocean. Mem. Fac, Fish. Kagoshima Univ. 39:
81-98.
Rockefeller, W.
1989 Looking to the fiature of billfishes. /« Stroud, R.H. (ed.).
Planning the future of billfishes, research and management
in the 90's and beyond. Parti: Fishery and stock synopses,
data needs and management. Mar. Rec. Fish. 13:11-14. Natl.
Coalition Mar. Conserv., Savannah, GA.
Saito. S.
1973 Studies on fishing of albacore, Thunnus alalunga (Bon-
naterre) by experimental deep-sea tuna long-line. Mem. Fac.
Fish. Hokkaido Univ. 21:107-185.
1975 On the depth of capture of bigeye tuna by further im-
proved vertical long-line in the tropical Pacific. Bull. Jpn. Soc.
Sci. Fish. 41:831-841.
658
Fishery Bulletin 90(4), 1992
Saito, S., K. Ishii, and K. Yoneta
1970 Swimming depths of large sized albacore in the South
Pacific Ocean — 1. Fishing of albacore by a newly designed
vertical longline. Bull. Jpn. Soc. Sci. Fish. 36:578-584.
Somerton, D.A., B.S. Kikkawa, and CD. Wilson
1988 Hook timers to measure the capture time of individual
fish. Mar, Fish. Rev. 50(2):l-5.
Squire, J.L.
1987 Striped marlin, Tetraptuturs audax, migration patterns
and rates in the northeast Pacific Ocean as determined by a
cooperative tagging program: Its relation to resource manage-
ment. Mar. Fish. Rev. 49(2):26-43.
Squire, J.L.. and D.W.K. Au
1990 Striped marlin in the northeast Pacific— a case for local
depletion and core area management. In Stroud, R.H. (ed.),
Planning the future of billfishes, research and management
in the 90's and beyond. Part 2: Contributed papers. Mar.
Rec. Fish. 13:199-214. Natl. Coalition Mar. Conserv., Savan-
nah, GA.
Suzuki, Z.
1989 Catch and fishing effort relationships for striped marlin,
blue marlin, and black marlin in the Pacific Ocean, 1952 to
1985. In Stroud, R.H, (ed.). Planning the future of billfishes,
research and management in the 90's and beyond. Part 1:
Fishery and stock synopses, data needs and management.
Mar. Rec, Fish, 13:165-177. Natl Coalition Mar. Conserv..
Savannah, GA.
Suzuki, Z., and S. Kume
1982 Fishing efficiency of deep longline for bigeye tuna in the
Atlantic as inferred from the operation in the Pacific and In-
dian Oceans. Int. Comm. Conserv, Atl. Tunas Collect. Vol,
Sci, Pap, 17:471-486.
Suzuki, Z.. Y. Warashina, and M. Kishida
1977 The comparison of catches by regular and deep tuna
longline gears in the western and central equatorial Pacific,
Bull, Far Seas Fish, Res, Lab. 15:51-83,
Yamaguchi, Y.
1989 Tuna long-line fishing IV: Fish ecology in the context
of tuna long-line fishing. Mar, Behav, Physiol, 15:45-73,
Yang. W.S., and Y. Gong
1988 The vertical distribution of tunas and billfishes, and fish-
ing efficiency between Korean regular and deep longlines in
the Atlantic Ocean. Bull. Natl. Fish. Res. Dev. Agency [Pusan,
Korea] 42:39-42.
Yoshihara, T.
1954 Distribution of catch of tuna longline-IV. On the rela-
tion between k and <p° with a table and diagram. Bull. Jpn.
Soc. Sci. Fish. 19:1012-1014.
Abstract. — In the summers of
1982, 1983, and 1985, almost 5000
commercial lobsters were trans-
planted from an area on the north-
east coast of Newfoundland to St.
Michael's Bay in southern Labrador,
about 200 km beyond their reported
northern limit of distribution, in an
attempt to establish a self-sustaining
population. Biological sampling of
these lobsters was carried out each
summer from 1986 to 1991. A con-
tinuous shift to larger sizes and a
generally high incidence of new-shell
animals indicated molting was a com-
mon occurrence in these lobsters. All
nonovigerous females had ovaries
developing for extrusion that sum-
mer, and their seminal receptacles
were full. In contrast, percentages
of ovigerous females were low and
most of these had extruded recently.
Many ripe females apparently failed
to extrude, and many that did ex-
trude lost the entire clutch before the
following summer. Exposure to tem-
peratures near or below 0°C from
mid-November to mid-May, and to
near-continuous darkness below a
layer of ice during most of this
period, may cause a high incidence
of resorption of ripe ovaries. The
incidence of ovigerous females with
recently-extruded eggs increased
substantially in the later years of the
study, indicating a degree of physio-
logical adjustment to the adverse en-
vironmental conditions. However,
loss of the entire clutch of eggs con-
tinued to be prevalent. Prolonged
low temperature certainly retarded
embryonic development for the fe-
males that extruded and retained
their eggs. Six of 17 ovigerous
females with old, eyed eggs had less
than half the yolk remaining. Only
one brood would have hatched by
early August, long enough in ad-
vance of autumn cooling for develop-
ment to Stage IV and settlement in
the area to be possible. Lobsters
transplanted to St. Michael's Bay
will not likely become a self-sustain-
ing population. Any recruitment that
might occur would certainly be too
little and too irregular to support a
fishery.
Reproduction in American
lobsters Homarus americanus
transplanted northward to
St. Michaers Bay, Labrador
Frank A. Boothroyd
Biology Department. Memorial University of Newfoundland
St John's, Newfoundland AIB 3X9, Canada
Gerald P. Ennis*
Department of Fisheries and Oceans, P O, Box 5667
St John's, Newfoundland AlC 5X1, Canada
The American lobster Homarus ameri-
canus occurs in the western Atlantic
Ocean from the Strait of Belle Isle
area of southern Labrador and the
northern tip of the island of New-
foundland south to North Carolina
(Cooper and Uzmann 1980). The spe-
cies supports commercial fisheries of
considerable economic importance
throughout most of its range. Its high
commercial value led to repeated at-
tempts to establish lobster popula-
tions on the Pacific coast of North
America, but none of the transplants
was successful (Conan 1986). In re-
cent years, the Provincial Govern-
ment of Newfotmdland and Labrador
transplanted commercial (mostly
adult) lobsters from an area on the
northeast coast of the island of New-
foundland to a location ~ 200 km be-
yond the reported northern limit of
distribution in St. Michael's Bay,
Labrador (Fig. 1). The bay extends
inland ~28km from the open coast
and contains numerous small islands,
features promoting a circulation pat-
tern that would aid retention and
eventual settlement in the area of
any larvae produced by the trans-
planted lobsters. This was an impor-
tant consideration, since the aim of
the transplant was to establish a self-
sustaining population that would
eventually support a fishery.
Manuscript accepted 15 July 1992.
Fishery Bulletin, U.S. 90:659-667 (1992).
* Reprint requests should be addressed to this
author.
Lobsters were transplanted to St.
Michael's Bay in the summers of
1982, 1983, and 1985. Biological sam-
pling was conducted each summer
from 1986 to 1991. Our purpose is
to present observations related to
various aspects of population biol-
ogy, in particular, molting, mating,
ovary development, spawning and
embryonic development, and con-
sider the possibility of this trans-
planted population being or becoming
self-sustaining.
Methods and materials
Lobsters transplanted to St. Michael's
Bay were caught during May-June
by commercial fishermen near Com-
fort Cove, Notre Dame Bay, on the
northeast coast of Newfoundland
(Fig. 1). They were purchased from
a local buyer by the Newfoundland
and Labrador Department of Fisher-
ies and transported directly to St.
Michael's Bay by float plane. Trans-
plants were made in 1982, 1983, and
1985 and totaled 2174 males, 81-114
mm carapace length (CL), and 2310
nonovigerous females, 81-112mm
CL. Lobsters were released once on-
ly at eight widely-separated sites
around the bay where the shallow-
water habitat appeared quite suitable
for lobsters.
The authors conducted biological
sampling annually in the summer-
659
660
Fishery Bulletin 90(4), 1992
64° 62° 60° 58° 56° 54° 52°
I I I I ) i I I I I I L_
ST. MICHAEL'S
^ BAr
COMFORT COVE
V
— I 1 \ 1 1 1 1 1 1 1 1 r-
64° 62° 60° 58° 56° 54° 52°
Figure 1
Locations of Comfort Cove, Newfoundland, and St. Michael's
Bay, Labrador, the donor and recipient sites, respectively, for
lobsters Homarus americanus transplanted to Labrador. In-
sert shows Goose L and Indian Arm, the main sites in St.
Michael's Bay where transplanted lobsters were sampled.
time from 1986 to 1990. In 1991, less detailed sampling
was carried out by the Department of Fisheries of the
Province of Newfoundland and Labrador. Sampling
focused on lobsters in Indian Arm (transplanted in
1982) and at Goose Island (transplanted in 1985) (Fig.
1). These two sites were selected to include lobsters
from the earliest and latest transplants. Also, in initial
trap sampling in 1986, lobsters were caught more
readily at these sites than at others. Lobsters were
caught in baited traps in 1986, 1987, and 1991 and by
scuba-diving from 1988 to 1990. Samples over the 6
years totaled 295 males and 392 females. Numbers
included in sampling for various purposes described
below are summarized by year and sampling site in
Table 1. Carapace length of each lobster was measured
to the nearest mm. Shell condition was determined by
external macroscopic examination in 1986 and 1988-
90. In over 90% of the lobsters examined for shell con-
dition, those that molted the previous summer (new
shell) could be readily distinguished from those that did
not (old shell) by general brightness of coloration,
sharpness of spines, and the degree of darkening due
to abrasion on the leading edges and undersides of the
claws. The others could be categorized with reasonable
confidence. Each summer from 1987 to 1990, sub-
samples of nonovigerous females totaling 111 for the
4 years were sacrificed. For each of these, ovary
color, ova diameter, contents of seminal receptacles,
Table 1
Summary of numbers included in sampling of lobsters Homarus americanus transplanted to St.
in 1982, Goose I. in 1985), 1986-91.
Michael
's Bay, Labrador (Indian Arm
Number and
carapace
lengths
Year
Shell
condition
Setal
development
Nonovigerous females
Ovigerous
females
Egg
counts
Seminal
receptacles
Cement
glands
New
eggs
Old
eggs
sampled Male
Female
Male
Female
Male
Female
Ovaries
of old eggs
Indian Arm
1986 87
109
87
74
—
—
_
—
_
—
21*
—
—
1987 10
25
—
—
—
—
7
7
7
1
1
2
1
1988 32
48
32
42
—
—
27
27
27
5
2
7
2
1989 24
21
24
21
24
20
8
8
8
7
6
—
6
1990 17
22
17
22
17
21
8
8
8
11
3
—
3
1991 4
9
-
-
-
-
-
-
-
4
3
-
3
Goose Island
1986 30
32
30
30
—
_
—
_
_
0
0
_
_
1987 7
33
—
_
_
—
23
23
23
0
0
—
—
1988 16
38
10
25
—
—
7
7
7
0
0
—
—
1989 29
20
29
20
29
20
15
15
16
2
2
—
2
1990 23
22
23
22
23
22
15
15
15
7
0
—
—
1991 15
13 - - - -
d new-egg ovigerous females were not distinguished.
10
0
* In 1986, old- ar
Boothroyd and Ennis: Reproduction iri Homarus amencanus transplanted northward
661
and the extent of pleopod cement gland development
(Aiken and Waddy 1982) were determined. In 1989 and
1990, pleopod setal development (Aiken 1973) was
determined for 47 nonovigerous females and 93 males.
Eighty-four ovigerous females were included among
the 392 females sampled over the 6 years. Egg
numbers were determined for nine collected in Indian
Arm in 1987 and 1988. Five of these were small egg
masses that were counted directly, the other four were
estimated by drying and weighing the entire mass, then
weighing and counting a subsample representing about
10% of the total. Yolk content of eggs (to the nearest
tenth, i.e., 0.1) was determined for 17 ovigerous
females carrying old, eyed eggs collected from 1986
to 1991. Perkins Eye Indices (PEI), which provide a
measure of embryonic development (Perkins 1972),
were determined for four of the latter.
From July 1986 to September 1988, a continuously
recording thermogi'aph was maintained in St. Michael's
Bay at 7 m within the depth range occupied by the
transplanted lobsters. Mean daily temperature was ob-
tained from the tapes. These were averaged for the
first and second half of each month, with data for the
different years combined. Temperature during most of
June each year was not obtained because the recording
tape expired in 1987 and the instrument malfunctioned
in 1988. The June portion of the annual temperature
regime was approximated by extrapolating from tem-
peratures which were rising in a near-linear fashion
before and after.
Results
Changes in size composition
Indian Arm Mean CL of all the lobsters transplanted
to St. Michael's Bay in the summer of 1982 was 89.1
mm (range 81-1 14 mm, N 987) for males and 88.9 mm
(range 81-1 12 mm, A^ 1001) for females (Fig. 2). We
assume that these are also representative of the
lobsters released at the Indian Arm site. When first
sampled in the summer of 1986, mean CL of lobsters
in Indian Arm had increased significantly (Tukey test;
P<0.001) to 107.2mm for males (A^ 87) and 97.5mm
for females (N 109). There were further shifts to larger
sizes each year, especially among males, and in the
summer of 1990, mean CL was significantly larger
(P<0.001) than in all previous years at 132.5mm (range
116-153mm, N 17) for males and 109.6mm (range
98-116mm, A^ 22) for females.
Goose Island Mean CL of all the lobsters trans-
planted to St. Michael's Bay in the summer of 1985 was
84.1mm (range 81-95mm, N 687) for males and 84.4
mm (range 81-92 mm, N 811) for females (Fig. 3),
MAIIS
JiL.
i>=?B7
1=89.1
FEMALIS
Jh
1982
11=1X1
^jIUii
ii=!7
r-Wl
Jl
1987
ipIO
1=1110
JIL
tf=32
i=ll*i
JL
m
it=ll)9
H7i
i
1987
W1I7.0
Jill
1988
n=«
!=ll)li
■III
1989
11=21
!=1MJ
jI.
1990
ipB
!=1015
130 140 150
80 90 IW
120 130 1« 150
CARAPACE l£NGTH (mm)
Figure 2
Size-frequencies for all male and female lobsters Homarus
americanus transplanted to St. Michael's Bay. Labrador, in
1982, and when sampled at the Indian Arm release site,
1986-90.
significantly smaller (^test; P< 0.001) than for those
transplanted in 1982. We assume that these means are
representative of the lobsters released at Goose I. At
Goose I., annual shifts to larger sizes were small until
1988, when mean CL had increased significantly
(Tukey test; P< 0.001) to 96.4 mm for males (N 16) and
92.6mm for females (N 38). By summer 1990, mean
CL had increased significantly again (P< 0.001) to
105.6mm (range 90-131 mm, N 23) for males and to
95.9mm (range 89-105mm, N 22) for females.
Incidence of new-shiell lobsters
The donor population at Comfort Cove is subjected to
a very intensive fishery each spring which removes
most of the commercial lobsters (nonovigerous and
>81mmCL). Following the summer molting period,
the vast majority of commercial-size animals in the
population had molted and grown from smaller sizes
662
Fishery Bulletin 90(4). 1992
60
50
40
30
20
10
0
60
50
40
30
20
10
0
60
50
r 30
Z 20
u «
iL To
10
0
60
50
40
30
20
10
0
60
50
40
30
20
10
MALES
1 1985
II. '
FEMALfS
la '9>^
II K^l
II
1 1986
■1 ~'^''
■_ 1986
II. ^
1987
III '^'^
198?
_ Iri3
III.
1988
III "
1 •&
1989
llll ™
I 1989
II H4J
1990
ipU
IhiIi .
1990
1 11=22
III. ""
Size-f
amer
1985.
60 90 100 110 120 130 IW 150 80 90 10O 110 120 130 140 150
CARAPACE LINGTH (mm)
Figure 3
requencies for all male and female lobsters Homarus
xanus transplanted to St. Michael's Bay, Labrador, in
and when sampled at the Goose I. release site, 1986-90.
Table 2
Percentage with new shells (molted previous summer) among
lobsters Homarus americamis transplanted to St. Michael's
Bay, Labrador (Indian Arm in 1982, Goose L in 1985) during
sampling, 1986 and 1988-90. Number sampled in parentheses.
Year
sampled
Indian Arm
Goose Island
Males
Females
Males
Females
1986
1988
1989
1990
40.2 (87)
53.1 (32)
62.5 (24)
52.9 (17)
21.6 (74)
42.9 (42)
47.6 (21)
31.8 (22)
3.3 (30)
80.0 (10)
75.9 (29)
52.2 (23)
6.7 (30)
92.0 (25)
75.0 (20)
63.6 (22)
Table 3
Pleopod setal development stages for male and nonovigerous
female lobsters Homarus americanus transplanted to St.
Michael's Bay, Labrador (Indian Arm in 1982, Croose I. in 1985)
during sampling, 1989-90.
Setal development stages* (%)
Males
Females
Year
sampled
N
0-2.0
2.5-5.0
AT
0-2.0
2.5-5.0
Indian Arm
1989
24
70.8
29.2
8
62.5
37.5
1990
17
47.1
52.9
8
0
100
Goose Island
1989
29
69.0
31.0
16
81.2
18.8
1990
23
39.1
60.9
15
0
100
* From Aiken (1973). Stages 0-2.0 indicate molting is unlike-
ly, and stages 2.5-5.0 that molting is probable in the cur-
rent summer.
during the summer. As a consequence, the incidence
of new shells among commercial sizes averaged 96%
and 94% for male and female lobsters, respectively, in
autumn samples from 1986 to 1990 (Ennis, unpubl.
data). These are much higher percentages than could
be expected for the transplanted lobsters in St.
Michael's Bay. There has been no lobster fishing in St.
Michael's Bay and, because of their larger sizes,
lobsters there are less likely to molt. Old-shell lobsters
are therefore likely to be much more prevalent.
Among Indian Arm lobsters, the incidence of new
shells (indicating lobsters molted the previous summer)
ranged from 40.2% In 1986 to 62.5% in 1989 for males
and from 21.6% in 1986 to 47.6% in 1989 for females
(Table 2). In 1986, the incidence of new shells was very
low among Goose I. lobsters (3.3% for males and 6.7%
for females) indicating few molted in summer 1985
when they were transplanted. However, in 1988 and
1989 the incidence of new shells was quite high
(75.0-92.0%) for males and females (Table 2).
Pleopod setal development was determined in 1989
and 1990 only. Percentages with advanced stages (2.5
and higher), which indicated molting would occur later
in the summer, varied between years and sexes at both
sampling sites. Among males, advanced stages in-
creased from 29.2% to 52.9% and from 31.0% to 60.9%
in Indian Arm and Goose I. samples, respectively,
between 1989 and 1990 (Table 3). Among females, it
increased from 37.5% in the Indian Arm sample and
18.8% in the Goose I. sample in 1989, to 100% in both
in 1990 (Table 3).
Reproductive condition of
nonovigerous females
Ovaries of 110 nonovigerous females from Indian Arm
and Goose Island combined were examined from 1987
to 1990 and all were found to be medium- to dark-green
Boothroyd and Ennis Reproduction in Homarus annericanus transplanted northward
663
Table 4
Ovary color and ova diameter among nonovigerous female
lobsters Hoviarus americanus transplanted to St. Michael's
Bay, Labrador (Indian Arm in 1982, Goose I. in 1985) during
sampling, 1987-90.
Ova diameter
Vmr
Ovary color
(mm)
sampled A^
Med. green
Dark green
X
Range
Indian Arm
1987-88 34
IV
23
1.4
0.9-1.5
1989 8
0
8^
1.0
0.9-1.3
1990 8
1
7
1.0
0.8-1.2
Goose Island
1987-88 30
21"
9
1.3
0.9-1.5
1989 15
1
14
1.0
0.8-1.3
1990 15
0
15
1.1
0.9-1.2
"The high proportion with medium-green ovaries in the
1987-88 sample is probably due to a different observer than
in 1989-90.
■^'Includes one specimen with yellow specks throughout the
ovary, indicating resorption underway.
Table 5
Pleopod cement gland stages for nonovigerous female lobsters
Homarus americantis transplanted to St. Michael's Bay,
Labrador (Indian Arm in 1982, Goose I. in 1985) during
sampling, 1987-90.
Year N
Cement gland stages*
1
2
3 4
Indian Arm
1987 7
1988 27
1989 8
1990 8
7
27
7
8
0
0
0
0
0 0
0 0
1 0
0 0
Goose Island
1987 23
1988 7
1989 16
1990 15
23 0 0 0
7 0 0 0
15 0 1 0
15 0 0 0
dy (1982). Stages 3 and 4 indicate ex-
the current summer.
* From Aiken and Wac
trusion will occur in
in color, with ova 0.8-1. 5mm in diameter (Table 4). All
these ovaries were developing for extrusion in the
summer they were sampled; one had yellow specks
characteristic of an ovary being resorbed. However, of
these nonovigerous females, only 2 of 24 sampled in
1989 (2 out of 111 overall) had sufficient pleopod
cement gland development (Stage 3) to indicate egg
extrusion would occur (Table 5). The seminal recep-
tacles of all of the nonovigerous females examined from
1987 to 1990 were full, which means each had mated
at the last molt and was capable of fertilizing a clutch
of eggs. Pleopod setal development indicated 25% of
24 and 100% of 23 nonovigerous females sampled in
1989 and 1990, respectively, would molt (Table 3).
While molting does not preclude egg extrusion in the
same summer, this is unlikely to occur among lobsters
in St. Michael's Bay. This phenomenon probably in-
volves only animals extruding for the first time and
<81mmCL (Ennis 1984a).
Incidence of ovigerous females
The reproductive cycle in female lobsters normally
covers 2 years. Molting and mating occur one summer,
egg extrusion one year later, followed by hatching of
eggs, molting, and mating again in the third summer
(Aiken and Waddy 1980a). Departures from this 2-year
cycle known to occur in the wild include molting,
mating, and extrusion in the same summer (mentioned
in the preceding section) and resorption of ripe ovaries
just before extrusion (Aiken and Waddy 1976 and
1980ab, Ennis 1984b). While variability in the incidence
of these phenomena may contribute somewhat, the in-
tensity of the commercial fishery and its timing in rela-
tion to the spawning season exert by far the greatest
impact on the incidence of ovigerous females in a
lobster population.
At Comfort Cove, most of the nonovigerous com-
mercial-size females in the population are removed by
the spring fishery just before the summer spawning
period. In autumn sampling from 1986 to 1990 an
average 6% of commercial-size females were ovi-
gerous, excluding those that molted and grew from
subcommercial sizes and extruded as well during the
summer (Ennis, unpubl. data). Not being fished and
being more likely to spawn as well because of their
larger size, the percentage of females ovigerous among
St. Michael's Bay lobsters by comparison should be
quite high.
In St. Michael's Bay, the percentages of females that
were ovigerous vWth old eggs each summer (i.e., ex-
truded previous summer) were very low, particularly
in 1986-88 samples, considering all nonovigerous
females examined the previous summer had ripe
ovaries. In the 1986 Indian Arm sample, 19.3% of the
females were ovigerous, including both new- and old-
egged females which were not distinguished at the
time. In subsequent years, the incidence of ovigerous
females with old eggs increased from 4.0% in 1987 to
33.3% in 1991 (Table 6). No ovigerous females were
included in Goose I. samples until 1989, when 10% of
the females carried old eggs. In 1990 and 1991 samples,
however, there were none with old eggs (Table 6).
664
Fishery Bulletin 90(4), 1992
Table 6
Percentage of females that were ovigerous among lobsters Homarus
americanus
transplanted to St. Michael'
s Bay, Labrador (Indian 1
Arm in 1982, Goose I. in
1985) during
sampling, 1986-
-91.
Indian Arm
Goose Island
No.
% new
% old
%
No.
% new
%old
%
Sampling dates
females
eggs
eggs
ovigerous
females
eggs
eggs
ovigerous
22 July-25 Aug. 1986
109
—
—
19.3*
32
0
0
0
18-25 July
26 Aug.-3 Sept. 1987
25
4.0
4.0
8.0
33
0
0
0
25 June-2 July 1988
48
10.4
4.2
14.6
38
0
0
0
19-25 July 1989
21
33.3
23.8
.57.1
20
10.0
10.0
20.0
13-20 Aug. 1990
22
50.0
13.6
63.6
22
31.8
0
31.8
14-19 Aug. 1991
9
gg ovigerous
44.4 33.3 77.8
females were not distinguished.
13
76.9
0
76.9
*In 1986, old- and new-e
The percentages of females that were ovigerous with
new eggs (i.e., extruded within the preceding 2 or 3
weeks) ranged from 4% in 1987 to 50% in 1990 in the
Indian Arm samples, and increased from 10% in 1989
to 32% in 1990 in the Goose I. samples (Table 6). Of
64 ovigerous females included in the combined 1987-91
samples, 73% had new eggs and, overall, there was a
substantially higher incidence of ovigerous females
among St. Michael's Bay lobsters in 1989-91 than
previously.
The low incidence of ovigerous females in samples
up to 1988 indicates most of the nonovigerous females,
all of which had ripe ovaries in summer, failed to ex-
trude eggs. In subsequent years, however, the propor-
tion of nonovigerous females extruding increased
substantially. There was no evidence of hatching of old
eggs prior to our sampling each summer, which might
have accounted for the scarcity of old-egged relative
to new-egged females. This indicates that many
females that extruded lost their entire clutch of eggs
sometime prior to the following summer.
Fecundity and egg development
Egg numbers were determined for 9 ovigerous females
included in the 1987 and 1988 samples from Indian
Arm. Four of these egg counts were <0.1% of expected
numbers as determined from a size-fecundity relation-
ship for a population on the south coast of New-
foundland, one had 2%, and the others had 30-105%
of the expected number of eggs (Table 7). The extreme-
ly low numbers in 5 of the 9 specimens cannot be at-
tributed to high variability generally associated with
such data, but rather indicate a high incidence of
massive egg loss. Although egg numbers were not
determined for any of 21 ovigerous females included
Table 7
Egg numbers for nine ovigerous female lobsters Homarus
americanus transplanted to St. Michael's Bay, Labrador, in-
cluded in 1987 and 1988 samples from Indian Arm.
Carapace
%of
length (mm)
No. of eggs
expected number*
97
15
<0.1
97
7
<0.1
101
122
<0.1
102
186
<0.1
94
334
2.0
103
6588
30.0
104
10122
45.0
102
17878
84.0
95
17854
105.0
* Expected number of eggs was calculated from a CL-feeun-
dity relationship (log,,, F = 3.0984 log,,, CL - 1.8963) for
a population in Placentia Bay on the south coast of New-
foundland (Ennis 1981).
in the 1989 and 1990 samples, cursory examination
indicated that most had what appeared to be full
clutches, although one had just a few hundred eggs.
The extent of embryonic development for 17 ovi-
gerous females with old, eyed eggs collected from 1986
to 1991 ranged from 0.8 to 0.3 yolk content (Table 8).
Only 6 had less than half the yolk remaining when ex-
amined. Perkins Eye Indices (PEI) were determined
for four specimens. One with 0.3 yolk content had a
mean PEI of 470 on 29 June 1988. At an assumed con-
stant developmental temperature of 10°C, it was
estimated, using Perkin's (1972) formula, that hatching
would have occurred by 4 August. Another with 0.4
yolk content had a mean PEI of 431 on 1 August 1986,
for which hatching by 21 September was estimated.
Boothroyd and Ennis Reproduction in Homarus amencanus transplanted northward
665
Table 8
Yolk content of eggs for 17 ovigerous females
with old. eyed eggs included in 1986-91 samples
of lobsters Homarus americaniis transplanted
to St. Michael's Bay, Labrador. Yolk content to
the nearest tenth.
Sampling
date
Carapace
length (mm)
Yolk
content
1 Aug.
1986
94
0.4
19 July
1987
92
0.7
29 June 1988
95
0.3
30 June 1988
97
0.8
20 July
1989
88
0.4
21 July
1989
110
0.6
21 July
1989
108
0.6
21 July
1989
103
0.7
21 July
1989
109
0.6
21 July
1989
106
0.5
22 July
1989
92
0.7
16 Aug.
1990
116
0.3
16 Aug.
1990
111
0.4
16 Aug.
1990
104
0.5
19 Aug.
1991
101
0.5
19 Aug.
1991
114
0.5
19 Aug.
1991
107
0.4
The other two PEIs were 108 and 127 (0.8 and 0.7 yolk
content, respectively). Even at 10°C, well above the
temperature in St. Michael's Bay beyond September,
it was estimated these broods would not be ready to
hatch until late December-early January.
Temperature
Bottom temperature at 7 m in St. Michael's Bay drops
below 0°C in late November, remains around -1°C
throughout the winter, and rises above 0°C in early
June (Fig. 4). Summer warming is rapid. Temperature
reaches 9°C between mid- and late- July and peaks at
just over 10°C in late August. By mid-September,
autumn cooling is underway and temperature starts to
drop rapidly around the end of September. At Com-
fort Cove, where the transplanted lobsters originated,
sub-zero temperatures prevail for only about 2 months
in late winter (Fig. 4). Summer warming begins in late
April-early May and reaches a slightly higher peak at
around 11°C in late August. Autumn cooling begins
somewhat later and proceeds more slowly. In St.
Michael's Bay, transplanted lobsters are exposed to a
similar range in temperature as at Comfort Cove, but
to a substantially lower mean temperature during most
of the year.
16-
ST MICHAEL'S BAY
12-
^ 10-
Z"^^'
/ H-..
UJ
'- 2-
J Y
0-
■2-
'■■•■■... ■/ \
JAN FEB MAR APR MAT JUN JUL AUG SEP OCT NOV DEC JAN
MONTH
Figure 4
Annual temperature (°C) regimes from continuous-recording
thermographs maintained on the bottom at 7 m in St. Michael's
Bay, Labrador, summer 1986-autumn 1988, and at 9m in
Comfort Cove. Newfoundland, during the same period.
Discussion
The very low incidence of new-shell lobsters in the 1986
Goose I. sample indicates few molted in summer 1985
when they were transplanted. This was most likely due
to molt inhibition caused by handling-induced stress,
possibly including wide temperature fluctuations, dur-
ing transplant. The high incidence of new-shell lobsters
at Goose I. in 1988, and the substantial shift to larger
sizes among the Indian Arm lobsters (transplanted in
1982) by 1986, indicate the transplanted lobsters ac-
climated over time and resumed molting despite the
lower temperatures in St. Michael's Bay.
All nonovigerous females examined during the study
had advanced ovaries developing for extrusion in the
summer they were sampled. All of these had full
seminal receptacles indicating they mated at the last
molt. However, only 1.8% (2 out of 24 examined in
1989) had advanced pleopod cement gland development
indicating extrusion was imminent. Some ovigerous
specimens with recently-laid eggs were observed each
year. This suggests most nonovigerous females that
were going to extrude each summer had done so by the
time our samples were collected. These amounted,
however, to only 23% (A'' 104) of the females (excluding
old-egg ovigerous) examined at Indian Arm from 1987
to 1990, and 22.5% (A^ 40) of those examined at Goose
Island in 1989 and 1990. In 1991, this percentage in-
creased to 66.7% {N 6) and 77% (A^ 13) in the Indian
Arm and Goose I. samples, respectively. Advanced
pleopod setal development among the ripe nonovi-
gerous females that were examined in 1989 and 1990
indicated 62% would soon molt.
666
Fishery Bulletin 90(4). 1992
Based on the foregoing observations, it appears that
up untO 1990 at least most of the mature, nonovigerous
females in St. Michael's Bay did not extrude. Rather,
they resorbed the lipovitellin accumulated in the ripen-
ing oocytes, and many then proceeded to molt. Resorp-
tion of the ripe ovary has been associated with un-
favorable holding conditions near the expected time of
extrusion (Templeman 1940) but appears to be common
in the wild as well (Ennis 1984b). Resorption occurs
when the molting and reproductive cycles are out of
phase, and molting is due to occur within 3 or 4 months
after egg extrusion (Aiken and Waddy 1976, 1980ab).
These cycles are normally synchronized by temperature
and photoperiod regimes to ensure that when a female
lobster extrudes eggs in one summer, it will not molt
until after the eggs have hatched sometime the follow-
ing summer.
In their experiments on the effects of winter tem-
perature and photoperiod on spawning, Aiken and
Waddy (1989) found a high incidence of resorption
among mature females held at high temperature
throughout the winter, particularly when a short-day
photoperiod was maintained throughout the summer
(55% resorbed). Onset of a long-day photoperiod in
spring was necessary to trigger spawning among
females held at high winter temperature. The incidence
of spawning was high ( >90%) among those held at low
winter temperature, even without onset of a long-day
photoperiod in spring, and only a few resorbed.
These results do not explain the high incidence of
resorption among female lobsters in St. Michael's Bay
where they are exposed to environmental conditions
well outside the foregoing experimental treatments.
From mid-November to mid-May, the bottom temper-
ature is near or below 0°C and the bay is frozen over
for most of this period. Visual acuity of lobsters at very
low light intensity has not been described. It seems
unlikely, however, that sufficient light would penetrate
a layer of snow-covered sea ice underlaid by low-salinity
water (from continuous river discharge and slow mix-
ing under the ice) for lobsters to detect a light:dark
cycle. This combination of very low temperature and
near continuous darkness for 5-6 months from late
autumn to spring is probably the main cause of the high
incidence of ovary resorption that has prevailed among
female lobsters in St. Michael's Bay.
Most of the ovigerous females collected in St.
Michael's Bay had extruded quite recently. The re-
mainder were old-egged females that spawned the
previous summer. Apparently, many females that ex-
truded lost the entire clutch of eggs before the follow-
ing summer. Massive but incomplete loss of eggs was
also observed. Of 9 ovigerous females collected in
Indian Arm in 1987 and 1988, 5 had numbers of eggs
ranging from 7 to 334. Only one of the 21 ovigerous
females collected in 1989 and 1990 had just a few
hundred eggs. The overall incidence of ovigerous
females, particularly those with new eggs, increased
substantially in 1989 and 1990 and especially in 1991.
At Goose I., no ovigerous females at all were included
in samples until 1989, 4 years after they were trans-
planted. This indicates a high degree of acclimation or
physiological adjustment to environmental conditions
in St. Michael's Bay on the part of females, resulting
in much less resorption of ripe ovaries and much more
extrusion in recent years. However, loss of the entire
clutch of eggs appeared to be still quite prevalent.
Ovigerous females sometimes lose their entire clutch
of eggs by molting (Ennis 1984b). This may have been
a more common occurrence in later years, which could
explain the near absence of tiny clutches compared with
the high incidence observed in 1987 and 1988, and
would also be consistent with a high incidence of new
shells and advanced setai development among females
in 1989 and 1990. Despite some physiological adjust-
ment resulting in more extrusion, for most females the
molting and reproductive cycles continued to be out of
phase.
Embryonic development in lobsters proceeds slowly
at temperatures below 6°C, and hatching can be
delayed by as long as 6 months if temperature remains
at 2-3°C throughout spring, summer, and autumn,
although it will eventually occur, even at that tem-
perature (Aiken and Waddy 1986). Perkins (1972)
found that advanced embryos will develop more slow-
ly than less advanced ones when held at the same
temperature. In one specimen with 29-week-old eggs
on 10 January held at the Boothbay Harbor Laboratory
under local water conditions, there was no measurable
development for an 18-week period starting in early
December. Over most of the period, temperature
ranged from 0.1°C to 1.5°C.
In St. Michael's Bay, bottom temperature drops
below 6°C by mid-October, below 3°C by mid-Novem-
ber, and below 0°C by late November. Of the old-egged
ovigerous specimens from there, which are presumed
to have extruded the previous summer, there was one
with 0.8 yolk remaining when examined at the end of
June, three with 0.7 remaining around 20 July, and
three with 0.5 remaining in mid-August. Possibly some
of these females carry their eggs through a second
winter and may represent the ones observed in our
samples with sufficiently advanced development for
hatching to occur before the end of summer (i.e.,
hatching 2 years after extrusion rather than the usual
1).
Only 6 of the 17 old-egged ovigerous females col-
lected in St. Michael's Bay from 1986 to 1991 had less
than half the yolk remaining in the eggs when exam-
ined. Four of these would likely have been ready for
Boothroyd and Ennis: Reproduction in Homarus amencanus transplanted northward
667
hatching by sometime in September, around the time
autumn cooling begins. The other, which had 0.3 yolk
remaining and a PEI of 470 at the end of June, would
probably have hatched by early August. This was the
only specimen for which hatching as much as 4 or 5
weeks in advance of autumn cooling was a possibility.
However, extensive plankton sampling (130 15-min
tows with aim diameter net) conducted near the sites
where lobsters were released in St. Michael's Bay, from
28 July to 22 August 1986, and from 18 July to 30
August 1987, failed to produce lobster larvae.
Bottom temperature at 7m in St. Michael's Bay dur-
ing August ranged from around 9°Ctol0.5°C. Assum-
ing temperature in the surface layer was around 15°C,
at which lobster larvae will reach Stage IV in 25 days
(Templeman 1936), it is possible that larvae hatching
in early August would be in Stage IV at the end of
August and ready to settle by mid-September. Larvae
hatching in late August would be exposed to temper-
atures below 10 °C well in advance of reaching Stage
IV. At 10°C, it takes 2 months from hatching to Stage
IV; and at 5°C, larvae generally die without reaching
Stage IV (Templeman 1936). Caddy's (1979) considera-
tion of the influence of seasonal temperature regime
on survival of lobster larvae indicates poor chances of
survival if larvae have not reached Stage IV by the end
of August. Any larvae hatching in St. Michael's Bay
in late August-early September appear to have little
chance of reaching Stage IV and settling in the area.
The 1988-90 diver-collected samples involved 12 dives
totaling about 30 diver hours searching at the Indian
Arm and Goose I. sites. Not a single small lobster was
found, suggesting there had been very little if any
recruitment since lobsters were transplanted in 1982.
The possibility that the lobsters transplanted to St.
Michael's Bay will become a self-sustaining population
is remote. While the portion of females with ripe
ovaries that actually extrude has increased substantial-
ly during this study, many of these lose all or most of
their clutch of eggs before the following summer. Of
those females that carry their eggs for at least a year,
very few have eggs ready for hatching sufficiently early
in the summer that any larvae are likely to settle in
the area. Any local recruitment that might occur would
be too little and too irregular to support a fishery for
lobsters in St. Michael's Bay.
Citations
Aiken. D.E.
1973 Proecdysis, setal development, and molt prediction in the
American lobster (Homarus amencanus). J. Fish. Res. Board
Can. 30:1337-1344.
Aiken. D.E., and S.L. Waddy
1976 Controlling growth and reproduction inthe American
lobster. In Avaulk, J.W. Jr. (ed.), Proc, 7th Annu. meet,
world maricult. soc, p. 415-430. Louisiana St. Univ. Press,
Baton Rouge.
1980a Reproductive biology. In Cobb, J.S., and B.F. Phillips
(eds.), The biology and management of lobsters, vol. 1, Phys-
iology and behavior, p. 215-276. Academic Press, NY.
1980b Maturity and reproduction in the American lobster. In
Anthony, V.C, and J.F. Caddy (eds.), Proc, Canada-U.S.
workshop on status of assessment science for N.W. Atlantic
lobster (Homariis americanus) stocks, p. 59-71. Can. Tech.
Rep. Fish. Aquat. Sci. 932.
1982 Cement gland development, ovary maturation, and re-
productive cycles in the American lobster, Homarus ameri-
canus. J. Crustacean Biol. 2:315-327.
1986 Environmental influence on recruitment of the American
lobster. Homariis americamis : A perspective. Can. J. Fish.
Aquat. Sci. 43:2258-2270.
1989 Interaction of temperature and photoperiod in the regula-
tion of spawning by American lobsters. Homarus americanus.
Can. J. Fish. Aquat. Sci. 46:145-148.
Caddy, J.F.
1979 The influence of variations in the seasonal temperature
regime on survival of larval stages of the American lobster
{Homarus americanus) in the southern Gulf of St.
Lawrence. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 175:
204-216.
Conan, G.Y.
1986 Summary of session 5: Recruitment enhancement. Can.
J. Fish. Aquat. Sci. 43:2384-2388.
Cooper, R.A., and J.R. Uzmann
1980 Ecology of juvenile and adult //oma7-(is. /reCobb, J.S.,
and B.F. Phillips (eds.). The biology and management of
lobsters, vol. II, p. 97-142. Academic Press, NY.
Ennis, G.P.
1981 Fecundity of the American lobster, Homarus americanus,
in Newfoundland waters. Fish. Bull., U.S. 79:796-800.
1984a Incidence of molting and spawning in the same season
in female lobsters, Homarus americanus. Fish. Bull, U.S.
82:529-530.
1984b Comparison of physiological and functional size-maturity
relationships in two Newfoundland populations of lobsters
Homarus americamis. Fish. Bull., U.S. 82:244-249.
Perkins, H.C.
1972 Development rates at various temperatures of embryos
of the northern lobster {Homarus americanus Milne-
Edwards). Fish. Bull., U.S. 70:95-99.
Templeman, W.
1936 The influence of temperature, salinity, light and food con-
ditions on the survival and growth of the larvae of the lobster
{Homarus americanus). J. Biol. Board Can. 2:485-497.
1940 The life history of the lobster. Newfoundland Dep. Nat.
Resour., Serv. Bull. (Fish.) 15, 42 p.
Abstract. - CoMa is a highly
prized recreational species of world-
wide distribution in tropical and sub-
tropical seas, but the development,
distribution, and ecology of its early
life stages are poorly known. Eggs
are spherical, average 1.24 mm in
diameter, and have a single oil glob-
ule (mean diameter 0.45 mm). The
perivitelline space is narrow and the
embryo heavily pigmented. Eggs
hatch in about 24h at 29°C based on
the relationship between egg diam-
eter and water temperature to pre-
dict development time in other ma-
rine fishes. Larvae hatch at about
2.5mmSL. Cobia spawn in both es-
tuarine and shelf waters during the
day, and eggs and larvae are usual-
ly collected in the upper meter of the
water column. Larvae are recog-
nized by the large supraorbital ridge
with a single spine, laterally swollen
pterotics, heavy body pigmentation,
minute epithelial spicules covering
the body integument, and a pair of
moderate-to-large, simple spines on
either side of the angle of the pos-
terior preoperculum. Only 70 larvae
<20mmSL were collected and iden-
tified from the Gulf of Mexico be-
tween 1967 and 1988; most occurred
between June and September at sur-
face temperatures >25°C, salinities
>2T7™,, and within the 100 m depth
contour. Similar patterns of head
spination provide evidence of a
sister-group relationship between
cobia and dolphinfish rather than
that previously hypothesized be-
tween cobia and remoras.
Larval development,
distribution, and ecology of
cobia Rachycentron canadum
(Family: Rachycentridae) in the
northern Gulf of Mexico*
James G. Ditty
Richard F. Shaw
Coastal Fisheries Institute, Wetland Resources Building
Louisiana State University, Baton Rouge, Louisiana 70803-7503
Cobia, in the monotypic family Ra-
chycentridae, is distributed world-
wide in tropical and subtropical seas
(Briggs 1960), except the eastern
Pacific, and is found seasonally in
temperate waters (Hassler and Rain-
ville 1975). A highly prized recrea-
tional species, most of the U.S. land-
ings are from Gulf of Mexico (Gulf)
waters; they are also caught inciden-
tally in commercial fisheries (Shaffer
and Nakamura 1989). Recreational
landings are not well documented,
but cobia are reportedly not abun-
dant and recruitment is considered
low (Gulf of Mexico & S. Atl. Fish.
Manage. Counc. 1985). Cobia are
migratory and usually absent from
commercial and recreational catches
of the northern Gulf during late fall
and winter at which time they are
caught off the Florida Keys. They
migrate north and west along the
Gulf coast during the spring (Shaffer
and Nakamura 1989) and reappear in
northern Gulf waters during March
and April (Springer and Pirson 1958).
Cobia are usually caught in shallow
coastal waters (Shaffer and Naka-
mura 1989), although they are often
taken offshore along the Louisiana
and Texas coasts in association with
oil and gas platforms and rafts of
Sargassum (RFS, pers. observ.).
Manuscript accepted 16 July 1992.
Fishery Bulletin, U.S. 90:668-677 (1992).
'Contribution LSU-CFI-91-6 of the Coastal
Fisheries Institute, Louisiana State Univer-
sity.
Despite the recreational value of
cobia, its ecology, distribution, and
morphological development during
early life stages are poorly known.
Only 23 specimens <20mmSL are
reported in the historical literature
for the Gulf (Dawson 1971, Finucane
et al. 1978ab, Houde et al. 1979).
Richards (1967) reviewed cobia gen-
eral life history, Shaffer and Naka-
mura (1989) compiled biological data,
and Hassler and Rainville (1975) de-
veloped techniques for hatching and
rearing cobia. In addition, Johnson
(1984) discussed the utility of cobia
early life stages for examining pre\i-
ous phylogenetic hypotheses and the
evolutionary interrelationships of ech-
eneoids (Rachycentridae-Corj'phae-
nidae-Echeneididae). Aspects of early
egg development have been described
(Ryder 1887, Joseph et al. 1964, Has-
sler and Rainville 1975) but not devel-
opment of early larvae <12.6mmSL.
Most larval illustrations and photo-
graphs are of poor quality (Ryder
1887, Dawson 1971, Hassler and
Rainville 1975, Finucane et al. 1978a,
Johnson 1984). The objectives of this
study are to describe cobia egg and
larval development, to provide data
on the seasonal occurrence, distribu-
tion, and ecology (i.e., relationship of
lai-vae to water temperature, salinity,
and station depth at time of capture)
of early life stages of cobia in the
Gulf, and to further examine the in-
terrelationships of the echeneoids.
668
Ditty and Shaw Early life stages of Rachycentron canadum in nortfiern Gulf of Mexico
669
RRCHYCENTRON CflNnOUM
Figure 1
Distribution of larval cobia Rachycentron canadum in the nortliern Gulf of Mexico. Plus signs are positive catch stations; diamond
is location of Crystal River estuary where cobia eggs and yolksac larvae were collected.
Methods
Eggs and larvae were obtained from museum collec-
tions along both the Gulf and Atlantic coasts. We ex-
amined 70 cobia eggs (all late-stage embryos) and 76
larvae 2.6-25 mm SL and determined their seasonal
occurrence, distribution, and ecology. Hydrographic
parameters were weighted by the total number of lar-
vae caught at each station to derive mean and median
values. All specimens were formalin-preserved except
those from Southeast Area Monitoring and Assessment
Program (SEAMAP) ichthyoplankton surveys of the
Gulf which were in ethyl alcohol. We considered lat.
26 ° 00 'N as the southern boundary of our survey area
(Fig. 1). Temperature and salinity data were from the
surface only.
Body measurements were made to the nearest 0.1
mm with an ocular micrometer in a dissecting scope
and follow Hubbs and Lagler (1958) and Richardson
and Laroche (1979). All lengths refer to standard
length (SL) unless otherwise noted. A compound scope
was used to examine the origin and location of epithelial
spicules. Representative specimens were illustrated
with the aid of a camera lucida. Specimens were not
cleared and stained because of the paucity of material.
Fin rays were counted when their pterygiophores were
visible; spines when they resembled formed structures
(Richardson and Laroche 1979). Myomeres were diffi-
cult to count in fish >6mm due to heavy larval pig-
mentation and opacity of the musculature, even under
polarized light, but all specimens <6mm had 24 myo-
meres. Cobia undergoing transition to the juvenile
stage were those with a full complement of formed rays
in all fins. Egg staging followed Moser and Ahlstrom
(1985).
Egg and larval development
Cobia eggs were spherical and measured 1.15-1. 3mm
(x 1.24, N 31), with a single oil globule 0.4-0.65 (i
0.45, A^ 13) in diameter. The oil globule was pigmented
and lay near the vegetal pole opposite the developing
embryo. The perivitelline space was narrow with the
embryo occupying about 85% of egg volume (range
70-92%, A'^ 13). The chorion was smooth and unorna-
mented. Cobia eggs hatch in about 24h at 29°C based
on Pauly and Pullin's (1988) relationship between egg
diameter and water temperature to predict develop-
ment time in marine fishes.
The embryo of late-stage Gulf cobia eggs was heav-
ily pigmented except for the caudal peduncle which was
unpigmented. Late-stage embryos from north of Cape
Hatteras, NC, had pigment lightly scattered over the
peduncle. Early yolksac larvae (2.6-3.2 mm) were
heavily pigmented externally and lacked a functional
mouth, eye pigment, and all fins. A single oil globule
with pigment was also present in the middle of the
yolksac. External pigment occurred over the snout and
in a band immediately behind the primordial eye. The
eye remained unpigmented until larvae were 3.5-4.0
mm (Fig. 2). At higher magnification, tiny epithelial
670
Fishery Bulletin 90(4). 1992
A
Figure 2
Eggs and larval development of cobia Rachycentron canadum. (A) Late-stage egg, diameter 1.24 mm; (B-C) yolksac larvae
2.6 and S.OmmSL; (D-H) larvae 4.5mm, 6.8mm. 10.0mm, 14.1mm, and 18.9mmSL.
Ditty and Shaw: Early life stages of Rachycentron canadum in northern Gulf of Mexico 671
Figure 2 (continued)
672
Fishery Bulletin 90(4), 1992
Body proportions of larval cobia Rachycentron
Table 1
canaAum from the Gulf of Mexico, expressed as % standard length (SL).
SL
N
Preanal
length
Head
length
Snout
length
Orbit
diameter
Upper jaw
length
Body depth
cleithrum
Predorsal
length
Prepelvic
length
Peduncle
length
2.6
3.2
4.0-4.9
1
1
2
61.5
62.5
64.4-65.0
27.8-31.2
6.7-7.5
10.0-11.2
10.0-13.8
18.9-21.2
5.0-5.9
6.0-6.9
3
3
68.0-68.6
63.2-67.2
31.4-34.0
27.9-31.7
7.6-10.0
7.3-8.3
11.0-12.7
8.8-10.8
13.6
10.3-12.7
20.3-20.6
19.8-23.3
33.9
30.9
7.0-7.9
9.8
10.0-10.9
3
1
3
64.1-65.3
64.3
57.1-64.1
26.9-29.5
30.6
27.5-29.5
7.0-8.7
8.7
7.0-8.1
9.0-10.0
9.2
9.0-10.0
10.9-13.3
12.8
11.5-13.3
17.9-20.0
21.4
19.0-20.0
52.6
51.0
50.0-56.3
30.8-33.3
33.7
30.0-36.9
12.0-12.8
12.8
12.1-13.3
11.0-11.9
12.0-12.9
14.0-14.9
3
2
2
57.3-60.9
63.2-63.7
56.6-58.7
27.7-29.9
28.2-28.8
26.2-27.3
6.7-7.8
7.2-8.0
6.9-7.0
8.4-8.7
8.8-8.9
8.0-8.3
11.3-12.0
11.2-11.3
11.0-11.2
18.3-19.2
16.1-18.4
17.2-17.5
49.6-52.2
49.2-50.4
49.0-50.3
31.1-34.2
32.0-32.2
29.4-30.3
11.5-12.6
12.5-12.8
11.0-11.2
16.0-16.9
19.5
3
1
58.4-59.9
57.4
26.5-26.9
27.2
7.2-7.8
7.7
7.8-8.4
7.7
10.2-10.5
10.2
15.0-16.8
15.4
48.2-49.4
46.2
28.9-30.1
29.7
12.0-13.6
12.8
21.0
25.0
1
1
57.1
56.0
24.8
24.0
6.7
6.0
7.1
7.2
9.5
8.8
14.3
14.9
47.6
46.8
27.1
26.8
12.8
12.0
spicules were also visible over the entire body integu-
ment, except the pupil of the eye, by 4 mm.
Body measurements were made on 30 cobia larvae
to examine developmental morphology (Table 1). Pre-
anal length was 61.5-62.5% SL in yolksac larvae and
increased slightly during preflexion, but decreased
steadily thereafter. A single intestinal loop, visible
through the body wall, gave the ventral visceral mass
a bulbous appearance by 7 mm. Body depth was about
20% SL in larvae < 10 mm but decreased to about 15%
by 25 mm. Likewise, orbit diameter decreased from
about 11-7%SL as larvae increased in length. Other
body proportions were relatively stable until larvae
were >10mm. Thereafter, a slow but steady decline
occurred in all proportions when compared with SL,
except caudal peduncle length which remained constant
(Table 1). The relationship between standard and total
lengths (TL), as defined by regression analysis, was
linear and highly correlated (SL = 0.73 TL-t- 1.44; N 29,
r2 0.998) at all sizes.
The supraorbital ridge and two largest preoper-
cular spines were visible by 4 mm, and the pterotics
were laterally swollen. The two preopercular spines
were located on either side of the angle of the pos-
terior preopercle. Three smaller spines, the largest of
which was inserted between the two posterior preoper-
cular spines, were also present along the anterior
preopercle. Spines were added along both the anterior
and posterior preopercle until a total of 4 anterior
and 4-5 posterior preopercular spines was reached
by 14-15 mm. A single spine occurring along the
supraorbital ridge of each frontal bone by 4.5mm
was directed posterolaterally by 6 mm. The supra-
orbital spine and swollen pterotics were best observed
when viewed dorsally (see Hardy 1978 for illustration).
A supracleithral spine also occurred about 10.5-11
mm, and two posttemporal spines (supratemporal of
Dawson 1971) originating from a common base were
visible by 12 mm. All head spination was simple and
unserrated.
In early larvae, internal pigment on the roof of the
mouth and ventral to the hindbrain and otic capsule
formed a mediolateral stripe through the head. Exter-
nally, melanophores were scattered over the snout,
fore- and midbrains, on the nape, and over the oper-
culum. Pigment also occurred immediately anterior to
the cleithral symphysis and along the isthmus. Both the
tip of the quadrate bone and dentary remained unpig-
mented until 7-7.5 mm. Head pigmentation increased
with length (Fig. 2). Minute epithelial spicules (Johnson
1984) covered the body by 4 mm, but were best ob-
served on the head and larval finfold. Spicules were
more easily observed on the integument as larvae in-
creased in length.
Along the dorsal midline, a bilateral row of melano-
phores extended posteriorly from the nape to above the
anus, behind which these rows coalesced to form a con-
tinuous band of postanal, dorsal pigment. By 4.5mm,
pigment occurred on the pectoral fin base, and larvae
were sparsely pigmented dorsolaterally but heavily
pigmented ventrolate rally. The caudal peduncle was
unpigmented in early larvae, but pigment extended
onto the peduncle by 5.5-6.5 mm and over the lower
hypural bones by 7mm. Pigment was also present on
the posterior third of the anal finfold and proximally
on the ventral caudal-fin rays by 7.5 mm. Body and
anal- and caudal-fin ray pigment increased with length.
Pigment occurred on the posterior dorsal-fin pteryio-
Ditty and Shaw: Early life stages of Rachycentron canadunn in northern Gulf of Mexico
673
Table 2
Fin ray counts
of larval cobia Rachycent
ron canadum from |
the Gulf of Mexico.
Size
(mm SL)
N
Dorsal*
Anal
Pectoral
Pelvic
9.8
1
30
25
9
10.0-10.9
3
27-30
22-25
12-13
11.0-11.9
3
28-31
24-26
13-17
12.0-12.9
2
30-33
24-26
14
14.0-14.9
2
31
25
17
3
16.0-16.9
3
28-32
1,24-26
17-19
1,5
19.5
1
31
1,25
20
1,5
21.0
1
31
1,25
21
1,5
25.0
1
29 1,25 20
ire virtually impossible to count on
1,5
uncleared
* Dorsal spines i
or unstained specimens.
fins. Dorsal- and anal-ray anlagen, however, began to
develop along the posterior fin base and proceeded
anteriorly. Development of anal rays consistently
preceeded that of dorsal rays. All anal fin elements
were countable by 17 mm. Dorsal spines were very dif-
ficult to count in specimens not cleared and stained
because spines were short and often covered by integu-
ment. One partially cleared 16.7 mm specimen had 7
poorly-formed, short dorsal spines, 11 precaudal and
14 caudal vertebrae (including urostyle), and 7 bran-
chiostegal rays. Pelvic buds were visible by 6 mm, with
the full complement of elements (1,5) present by
16.5-17mm. Pectoral rays were first visible at about
10 mm and the full complement (20-21 rays) was pres-
ent by 19-20 mm. A full complement of rays in all fins
(around 20 mm) marked the beginning of transition to
the juvenile stage (Table 2).
phores by about 10.5mm. Fin pigmentation pro-
gressed anteriorly along the dorsal base and extended
onto posterior dorsal rays by about 18-19 mm. Pelvic
rays were first pigmented at about 13 mm, but pectoral
fins remained unpigmented at all sizes examined
(Fig. 2).
A 49mm juvenile had a jet black caudal fin except
for the distal tips of the upper principal rays. Pelvic
fins were completely black, but pigment was present
basally on only the upper few rays of the pectoral fin.
Pigment also covered all of the posterior rays of both
the dorsal and anal fins. Fin pigment decreased ante-
riorly, however, such that only the basal portions of
the anterior dorsal and anal rays were pigmented. All
dorsal spines were visible and the integument was en-
tirely jet black.
Fin development
A continuous median finfold extended posteriorly along
the body from the nape to the cleithral symphysis of
early larvae. About 5 mm, a ventral thickening occurred
near the tip of the unflexed notochord. Anlagen began
to form obliquely downward in the caudal finfold
during flexion (~6.5-8mm). As the hypural complex
shifted to a terminal position, caudal ray development
proceeded both dorsally and ventrally until the adult
complement of 9-i-8 principal rays was present at
10.5-llmm. By 19.5mm, the caudal fin was distinctly
spatulate and heavily pigmented. Dorsal- and anal-fin-
base development coincided with notochord flexion.
These fin bases began to differentiate centrally, and
development proceeded both anteriorly and posteriorly
with all pteryiophores countable by 10-1 1mm in both
Distribution and ecology
The only confirmed collection of both cobia eggs and
yolksac larvae from the Gulf was from the Crystal
River estuary, Florida, during July 1984. These speci-
mens were collected from waters 28.1-29.7°C and
30.5-34. rVoo, except for a single 3.2 mm yolksac larva
from a power plant discharge canal at 36.7°C and
25.2"/u(.. All other eggs and early larvae were from
stations along the outer perimeter of the study area
at station depths of 3-6 m. No eggs or larvae were col-
lected at stations located over oyster reefs, in the salt
marsh, or in tidal creeks. Gulfwide, larvae were first
collected during late May, with most (98%) collected
June-September. Cobia larvae also primarily (85%)
occurred at 25-30°C (x 28.2°C, range 24.2-32.0°C),
at >27'7o,i (x 30.8"/nn, range 18.9-37.7"/ ), and most
(75%) at station depths <100m (median 50m, range
3.1-300m) (Fig. 3).
Discussion
Our data suggest that cobia eggs hatch in about 24 h
at 29°C. Ryder (1887) projected a 36h incubation time
at an unspecified temperature. Based on Pauly and
Pullin's (1988) predictive relationship to derive incuba-
tion time and a mean egg diameter of 1.24 mm from
this study, Ryder's cobia eggs were probably incubated
at about 25 °C. In cooler waters of the mid-Atlantic
Bight and northward during the spring/early summer
(i.e., ~20°C), projected incubation time increases to
56h. Cobia hatch at about 2.5mm based on collection
of vdld-caught early yolksac larvae (2.6-3.2 mm) with
unpigmented eyes and on the work of Hassler and
Rainville (1975).
674
Fishery Bulletin 90(4), 1992
501
IS <6.5 □ 6.5-10 ■tO.1-15 a 15.1-20
24 2S 2a 30 32
Temperature (C)
4C^
24 28 32
Salinity (ppt)
<25 25-50 51-100 >100
Depth Zone (m)
Figure 3
Distribution of cobia Rachycentron canadum larvae in the Gulf of Mexico with respect to hydrography. Larval stage/length class was
assigned as follows: preflexion, <6.5mmSL; flexion/early postflexion, 6.5-lO.OmmSL; late postflexion, 10.1-15.0mmSL; and transi-
tion larvae, 15.1-20.0 mm SL.
Our data on egg and oil globule diameter agree with
historical data (Ryder 1887, Joseph et al. 1964, Rich-
ards 1967, Hassler and Rainville 1975) except that our
mean oil globule diameter (0.45 mm) is greater than
that found for eggs from the Chesapeake Bay area
(0.37mm, Richards 1967; 0.38mm, Joseph et al. 1964).
Only two cobia eggs are previously illustrated, one in
early- and the other in midstage development (Ryder
1887). The diameter of the early-stage egg, however,
is considerably smaller than that of the midstage egg,
and the specific identification of the early egg is
unclear.
Cobia spawn during the day, since all embryos ex-
amined from the Gulf are at similar stages of develop-
ment (i.e., late stage after Ahlstrom and Moser 1980)
when collected during midmorning, except for one col-
lection of late-stage eggs near midnight. Furthermore,
daytime spawning cobia have been reported about 48
km southwest of Panama City, Florida (see Shaffer and
Nakamura 1989 for details) in waters we estimate at
82-165 m deep. Our data also show that cobia larvae
occur in both estuarine and shelf waters of the Gulf
(Figs. 1,3), primarily during May-September. The only
confirmed cobia eggs and yolksac larvae collected
together in the Gulf are from the Crystal River estuary
at station depths of 3-6 m. Early larvae (<6.8mm) are
also collected at stations within the 65-134 m isobath
range off Texas during September (Finucane et al.
1978b). The location of these collections suggests that
some spawning also occurs on the shelf 50-90 km from
the coast. Offshore waters beyond the edge of the con-
tinental shelf are relatively well sampled during May
(SEAMAP 1983-87) when histological analyses in-
dicated adult cobia are ripe (Thompson et al. 1991), but
no cobia larvae were identified. Seven cobia larvae (all
>9.5mm) were identified from beyond the 180m depth
contour during this study and all were collected off the
Mississippi River delta. Distribution of larvae centered
around the Mississippi River delta, however, may
reflect the intensity of neuston net sampling in this area
rather than actual distribution of spawning adults. Only
two larvae were collected off Florida during a com-
prehensive multiyear survey of eastern Gulf waters
>10m, both during August (Houde et al. 1979).
Seasonal occurrence and ecological data from along
the Atlantic coast of the United States support our find-
ings from the Gulf. Cobia eggs occur primarily between
May and August at surface water temperatures >20°C
(Joseph et al. 1964, Hassler and Rainville 1975,
Eldridge et al. 1977; W.F. Hettler and L. Settle, NMFS
Southeast Fish. Sci. Cent., Beaufort NC, pers. com-
mun.; P. Berrien, NMFS Northeast Fish. Sci. Cent.,
Sandy Hook NJ, pers. commun.; D. Hammond, S.C.
Dep. Wildl. Mar. Resour., Charleston SC, pers. com-
mun.). Eggs are collected in lower Chesapeake Bay
(Joseph et al. 1964), inlets to North Carolina estuaries
Ditty and Shaw Early life stages of Rachycentron canadum in northern Gulf of Mexico
675
(W.F. Hettler and L. Settle, pers. commun.), in coastal
waters 20-49 m deep (App. Table 1), and both near the
edge of the Florida Current and in the Gulf Stream
(Hassler and Rainville 1975, Eldridge et al. 1977). Off
North Carolina, cobia eggs are usually collected on
flood tides but few larvae are found in tidal inlets (W.F.
Hettler and L. Settle, pers. commun.). Cobia eggs and
larvae are usually collected in the upper meter of water
with surface-towed nets (Joseph et al. 1964, Hassler
and Rainville 1975 [implied], Eldridge et al. 1977; W.F.
Hettler, pers. commun.). Neither cobia study off the
Atlantic coast of the United States (Joseph et al. 1964,
Hassler and Rainville 1975) provides environmental
data, but eggs are successfully hatched at 19-35°/oo
(Hassler and Rainville 1975).
Similarities in larval morphology provide evidence of
a sister-group relationship between cobia and dolphin-
fishes (Coryphaenidae) rather than that previously
hypothesized between cobia and remoras (Echeneidi-
dae) (Johnson 1984). Larvae of both cobia and the
dolphinfishes share similar patterns of head spina-
tion: laterally swollen pterotics; a single, simple spine
on the supraorbital ridge of each frontal bone (except
in pompano dolphin C. equiselis, which may have
multiple spines along the ridge; JGD, pers. observ.);
a small posttemporal spine; and several spines along
the anterior and posterior preopercule, including an
enlarged spine on either side of the angle. However,
cobia have a small supracleithral spine (Dawson 1971,
this study) that dolphinfishes lack. Remoras complete-
ly lack head spines. Both cobia and the dolphinfishes
have epithelial spicules, a specialization unique to lar-
vae of these species (Johnson 1984). We found spicules
visible on the integument of both cobia and the dolphin-
fishes by 4mm (JGD, pers. observ.) and they cover the
entire body surface, except the pupil of the eye. Spicule
composition and function, however, are unknown
(Johnson 1984). Larval cobia are further separated
from superficially similar remoras by the presence of
large hook-like teeth on the dentary in remoras. Cobia
lack these teeth. Larval cobia differ from the dolphin-
fishes by the lack of dorsal and anal spines and a higher
vertebral count in the dolphinfishes (25 in cobia vs.
30-34 in dolphinfishes). Dolphinfishes also usually have
50-1- soft dorsal rays, whereas cobia have 27-33.
Acknowledgments
This study was supported by the Marine Fisheries Ini-
tiative (MARFIN) Program (contract nos. NA90AA-
H-MFlll and NA90AA-H-MF727). Thanks to those
who loaned us specimens or provided data: Karen
Burns, Mote Marine Lab, Sarasota FL; John Caruso,
Tulane University Museum, New Orleans LA; L. Alan
Collins and Churchill Grimes, NMFS Southeast Fish-
eries Science Center, Panama City FL; Wayne Forman
and Leroy Kennair, Freeport-McMoRan, New Orleans
LA; Bruce Comyns and Stuart Poss, Gulf Coast Re-
search Lab, Ocean Springs MS; Peter Berrien, NMFS
Northeast Fisheries Science Center, Sandy Hook NJ;
Bill Hettler and Larry Settle, NMFS" Southeast
Fisheries Science Center, Beaufort NC; and Karsten
Hartel, Museum of Comparative Zoology at Harvard.
Thanks also to the Southeast Area Monitoring and
Assessment Program (SEAMAP) and the Gulf States
Marine Fisheries Commission for providing specimens
and environmental data; to the Marine Resources,
Monitoring, Assessment and Prediction (MARMAP)
program for providing eggs collected during ichthyo-
plankton surveys; to Cathy Grouchy for illustrating
the egg and larvae; and to Joe Cope for computer
assistance.
Citations
Ahlstrom, E.H.. and H.G. Moser
1980 Characters useful in identification of pelagic marine fish
eggs. Calif. Coop. Oceanic Fish. Invest. Rep. 21:121-131.
Briggs, J.C.
1960 Fishes of worldwide (circumtropical) distribution. Copeia
1960(3);171-180.
Dawson, C.E.
1971 Occurrence and description of prejuvenile and early
juvenile Gulf of Mexico cobia, Rachycmitron canadum. Copeia
1971(1):65-71.
Eldridge, P.J.. F.H. Berry, and M.C. Miller III
1977 Test results of the Boothbay neuston net related to net
length, diurnal period, and other variables. S.C. Mar. Resour.
Cent. Tech. Rep. 18, 22 p.
Finucane, J.H., L.A. Collins, and L.E. Barger
1978a Determine the effects of discharges on seasonal abun-
dance, distribution, and composition of ichthyoplankton in the
oil field. In Jackson, W.B. (ed.), Environmental assessment
of an active oil field in the northwestern Gulf of Mexico,
1977-1978. NOAA Rep. to EPA {EPA-IGA-D5-E693-EO),
NMFS Southeast Fish. Sci. Cent., Galveston, 299 p.
Finucane. J.H., L.A. Collins, L.E. Barger, and J.D. McEachran
1978b Ichthyoplankton/mackerel eggs and larvae. Environ-
mental studies of the south Texas outer continental shelf,
1977. Final Rep. to Bur. Land Manage, by NMFS Southeast
Fish. Sci. Cent., Galveston, .504 p.
Gulf of Mexico & South Atlantic Fishery Management Councils
1985 Final Amendment 1, fishery management plan and en-
vironmental impact statement for coastal pelagic resources
(mackerels) in the Gulf of Mexico and south Atlantic region.
Gulf Mex. Fish. Manage. Counc, Tampa, FL.
Hardy, J.D. Jr.
1978 Development of fishes of the mid-Atlantic Bight, an atlas
of egg, larval, and juvenile stages. Vol. III. Aphredoderidae
through Rachycentridae. U.S. Fish. Wildl. Serv., Biol. Serv.
Prog. FWS/OBS-78/12. 394 p.
Hassler, W.W., and R.P. Rainville
1975 Techniques for hatching and rearing cobia, Rachycentron
canadum, through larval and juvenile stages. Univ. N.C. Sea
Grant Coll. Prog., UNC-SG-75-30, Raleigh, 26 p.
676
Fishery Bulletin 90(4), 1992
Houde, E.D., J.C. Leak, C.E. Dowd, S.A. Berkeley, and
W.J. Richards
1979 Ichthyoplankton abundance and diversity in the eastern
Gulf of Mexico. Univ. Miami Rep., BLM Contract AA550-
CT7-28, Miami, 546 p.
Hubbs, C.L., and K.F. Lagler
1958 The fishes of the Great Lakes region. Univ. Mich. Press,
Ann Arbor, 213 p.
Johnson, G.D.
1984 Percoidei: Development and relationships. In Moser,
H.G., et al. (eds.). Ontogeny and systematics of fishes, p.
464-498. Spec. Publ. 1, Am. Soc. Ichthyol. Herpetol., Allen
Press, Lawrence, KS.
Joseph, E.B., J.J. Norcross, and W.H. Massmann
1964 Spawning of the cobia, Rachycentron canadum. in the
Chesapeake Bay area, with observations of juvenile speci-
mens. Chesapeake Sci. 5:67-71.
Moser, H.G., and E.H. Ahlstrom
1985 Staging anchovy eggs. In Lasker, R. (ed.), An egg pro-
duction method for estimating spawning biomass of pelagic
fish: Application to the northern anchovy, Engraulis mordax,
p. 37-41. NOAA Tech. Rep. NMFS 36.
Pauly, D., and R.S.V. Pullin
1988 Hatching time in spherical, pelagic, marine fish eggs in
response to temperature and egg size. Environ. Biol. Fish.
22(4):261-271.
Richards, C.E.
1967 Age, growth, and fecundity of the cobia, Rachycentron
canadum, from Chesapeake Bay and adjacent mid-Atlantic
waters. Trans. Am. Fish. Soc. 96(3):343-350.
Richardson, S.L., and W.A. Laroche
1979 Development and occurrence of larvae and juveniles of
the rockfishes Sebastes crameri. Sebastes pinniger, and Sebastes
helvcnnafulatus (Family Scorpaenidae) off Oregon. Fish. Bull.,
U.S. 77:1-46.
Ryder, J. A.
1887 On the development of osseous fishes, including marine
and freshwater forms. Rep. U.S. Fish. Comm., Pt. 13. 1885
(1887):489-604.
SEAMAP (Southeast Area Monitoring and Assessment Program)
1983-87 Plankton data for 1982-1986. Fisheries-independent
survey data/National Marine Fisheries Service, Southeast
Fisheries Center. Produced by Gulf States Mar. Fish. Comm.,
Ocean Springs, MS (ASCII file).
Shaffer, R.V., and E.L. Nakamura
1989 Synopsis of biological data on the cobia Rachycentron
canadum (Pisces: Rachycentridae). NOAA Tech. Rep. NMFS
82 [FAO Fish. Synop."l53], 21 p.
Springer, V.G., and J. Pirson
1958 Fluctuations in the relative abundance of sport fishes as
indicated by the catch at Port Aransas. Texas 1952-1956.
Publ. Inst. Mar. Sci., Univ. Tex. 5:169-185.
Thompson. B.A., C.A. Wilson, J.H. Render, and M. Beasley
1991 Age. growth, and reproductive biology of greater amber-
jack and cobia from Louisiana waters. Year 1. Rep. to L^S.
Dep. Commer., NOAA, NMFS, Coop. Agreement NA90AA-
H-MF089, under Mar. Fish. Initiative (MARFIN) Prog.,
Coastal Fish. Inst., Louisiana St. Univ., Baton Rouge, 55 p.
Appendix
Table 1
Station location
collection, and environmental data for cobia Rachycentron cam
idum eggs
and larvae.
All specimens from the Gulf 1
of Mexico, except those loaned by the MCZ at Harvard and eggs from the NMFS Southeast Fisheries Science Center, Beaufort NC. |
Water temperature and salinity
values were at the surface. NA =
data not available.
Stn. depth
Temp.
Salinity
Size
Station
Date
Lat.
Long.
(m)
(°C)
7«,
Gear
N
(mm SL) Eggs
4353'
6-05-67
30°13
88°47
11
25.9
28.9
7
2
16.6-18.2
4357
6-05-67
30°02
88° 40
18
26.4
29.4
7
1
17.0
4355
6-10-68
29°24
88°17
55
32.0
36.6
7
2
12.6-15.3
4356
6-12-67
29° 19
88°14
73
30.3
27.8
7
3
13.6-14.2
4354
6-18-68
29°42
88°27
37
29.4
37.7
7
1
12.9
00807'
6-23-71
29°40
88°28
20
NA
NA
8
1
17.0
01613
May/Jun 67
27° 40
96° 59
20
NA
NA
8
1
7.8
01614
May/Jun 76
28°12
96°27
22
NA
NA
8
1
16.7
01687
May/Jun 76
27° 30
96°45
45
NA
NA
8
3
9.8-12.5
EPA IV-A^
7-13-77
28°51
94° 42
17
24.8
33.0
9
2
6.0-12.4
EPA V-A
7-13-77
28° 52
94°41
17
24.8
33.5
9
1
16.6
EPA V-B
7-13-77
28°51
94°42
17
24.7
33.0
9
1
16.6
EPA V-D
7-13-77
28°51
94°42
17
24.5
32.8
9
1
13.0
EPA II-B
7-14-77
28°53
94°41
17
25.5
34.0
9
1
19.5
BLM 11-3^
7-06-77
27°18
96°23
131
>25
36.0
10
1
3.8
BLM III
9-07-75
26°57
96°32
106
<25
36.0
11
1
5.9
BLM IV-3
9-07-77
26°10
96°24
91
>25
36.0
10
1
12.4
BLM III-2
9-08-77
26°, 58
96°48
65
>25
36.0
10
1
6.8
BLM 1-3
9-10-77
27°37
96°06
134
24.2
36.2
10
2
5.0-5.1
Ditry and Shaw: Early life stages of Rachycentron canadum in northern Gulf of Mexico 677
Appendix Table I (continued)
Station
Date
Lat.
Long.
Stn. depth
(m)
Temp. Salinity
(°C) "A.,
Gear
9114 A-23
7-18-84
28°56
82° 35
4.6
9121 C-1
7-18-84
28°56
82° 35
4.0
9179 H-3
7-18-84
28° 56
82°35
5.8
9008 H-4
7-18-84
28°56
82° 35
6.1
9266 I-l
7-18-84
28°56
82° 35
6.0
9545 1-2
7-31-84
28°56
82°35
3.7
9086 1-2
7-18-84
28°56
82° 35
4.1
9584 K-2
7-31-84
28° 56
82° 35
3.1
3088"
6-14-84
28° 58
90°33
13
5370
6-18-86
28°54
90°53
7
426
6-18-82
29°11
92°43
19
432
6-23-82
29°14
93°56
15
573
6-16-82
28°30
90° 00
100
1644
6-07-83
30°00
88° 02
26
1647
6-08-83
29°47
88°17
33
3166
6-24-84
29°00
95°00
15
3220
7-13-84
28°21
93°00
53
4484
9-17-85
29° 00
89°36
27
94374^
6-01-85
34°54
75°40
33
94505
8-21-68
38°07
70° 03
NA
AL8507-142«
7-28-85
36°50
75°27
21
AL8507-143
7-28-85
37°06
75° 11
36
DL8604-83
6-21-86
38°38
74°48
20
DL8604-152
6-25-86
36°47
75°14
26
EK8006-1
7-16-80
35°41
74° 58
49
EK8006-2
7-16-80
35°16
75° 14
24
P-1^'
9-12-86
28°48
89° 52
82
P-5
9-13-86
28°40
89°39
96
P-17
9-25-86
28°53
89°16
63
P-23
9-25-86
28°50
89° 05
195
P-31
9-26-86
28°56
88° 48
300
e
5-27-88
29°14
88°47
63
31
5-29-88
28°48
89°34
78
12
8-26-88
28°40
89°37
101
24
8-26-88
28°49
89° 43
70
39
8-27-88
28°50
89°21
72
44
8-27-88
28° 52
89°07
111
45
8-28-88
28°50
89°09
203
60
8-27-88
29°03
88°46
157
Stations:
' Gulf Coast Research Lab, Ocean Springs MS
-NMFS. Panama City Lab
■'Mote Marine Lab, Sarasota FL
"SEAMAP 1982-1986
^Museum of Comparative Zoology, Harvard
•^NMFS. Sandy Hook Lab, NJ
29.6
31.5
12
36.7
25.2
12
29.7
32.6
12
29.7
32.6
12
29.4
31.3
12
28.1
30.5
12
29.6
34.1
12
28.2
32.2
12
28.0
18.9
13
29.7
27.9
10
29.5
27.6
13
NA
30.2
13
29.4
33.3
10
25.0
NA
13
25.0
24.0
13
30.5
27.2
13
30.3
29.3
13
27.3
27.3
13
NA
NA
NA
NA
NA
NA
25.0
31.9
14
25.0
32.7
14
20.2
31.9
14
22.4
32.4
14
25.2
36.0
14
26.6
35.7
14
29.8
27.0
13
28.4
33.0
13
28.6
33.0
13
29.4
34.0
13
28.6
34.0
13
25.1
31.7
13
25.2
30.3
13
29.0
NA
13
28.8
26.8
13
30.1
29.5
13
29.6
28.0
13
29.6
30.5
13
29.5
29.5
13
Gear:
' Im, surface tow
*Neuston net (size unknown)
^0.5 X 1.0m neuston net, 0.505mm mesh
'°60cm bongo, 0.333mm mesh, oblique tow
"Im, 0.250mm mesh, oblique tow
'"Im, stepped-oblique tow, 0.505 mm mesh
'^lx2m neuston net, 0.948mm mesh
"60 cm bongo, 0.505mm mesh, oblique tow
N
Size
(mm SL)
6.8
3.2
7.5
3.1
2.6
2.6-4.0
4.5
6.7-7.8
10.5
16.0-19.5
22.3
5.0
12.5
12.0
16.0-21.0
25.0
10.3
6.0
10.0-14.5
10.1
11.8
11.0
10.2-18.5
10.7-12.3
9.5
12.2
10.0-14.0
14.0-19.0
13.0
10.0-24.5
9.5
15.0
Eggs
14
16
6
20
4
Abstract. - The basis for the
curious association between yellow-
fin tuna Thunnus albacares and
spotted dolphin Stenella attenuata in
the eastern tropical Pacific Ocean
has never been explained. Considera-
tion of the bioenergetics of the
associated tuna and dolphins sug-
gests that the association may be
based on the combined effects of a
shallow thermocline, overlapping
size (length) ranges of associated
yellowfin and young dolphins, con-
gruent diets, hydrodynamic con-
straints on swimming speeds of
dolphin schools, and social (care-
giving) behavior of dolphins. Insights
developed during construction and
exercise of comparative bioenerget-
ics models for the tuna and dolphin
suggest that tunas are more likely to
follow dolphins than dolphins to
follow tunas, and that the strength
of the association in a given area may
be related to oceanographic condi-
tions affecting prey distribution and
abundance.
Energetics of associated tunas
and dolphins in tlie eastern tropical
Pacific Ocean: A basis for the bond
Elizabeth F. Edwards
Southwest Fisheries Science Center, National Marine Fisheries Service, NOAA
P,0 Box 271, La Joila, California 92038
Manuscript accepted 29 July 1992.
Fishery Bulletin, U.S. 90:678-690 (1992).
In the eastern tropical Pacific Ocean,
yellowfin tuna Thunnus albacares
and spotted dolphin Stenella attenu-
ata form an association strong
enough that the fish can be captured
by capturing the associated dolphins
(e.g., Orbach 1977). The dolphins,
easier to locate than the tuna, form
the sighting cue for locating tuna
schools. Despite chases lasting on
average about half an hour (and occa-
sionally as long as 2-3 hours) the fish
tend to remain with the dolphins
throughout. Eventually the dolphins
tire and can be encircled, along with
the associated tunas, with a purse-
seine net.
Although the subject of substantial
conjecture (e.g., Perrin 1969, Orbach
1977, Au and Pitman 1986, Au 1991),
no definitive explanation exists for
the association, perhaps in part
because conjectures to date have
been qualitative rather than explicitly
quantitative. Quantifying the advan-
tages or disadvantages of the associa-
tion in terms of the energetics of its
component groups holds promise for
helping understand the bond, be-
cause such quantification can more
readily expose conceptual errors,
lead to unexpected insights, and form
the basis for testable hypotheses. Ex-
pressing relationships in terms of
energy flow (e.g., cost of finding
food, cost of reproduction, feeding re-
quirements, etc.) has often proved a
useful format for developing under-
standing of biological phenomena.
Following this precedent, I present
here bioenergetics models for both
tunas and dolphins in a "typical"
association in the eastern tropical
Pacific Ocean. I use these models to
estimate feeding rates of tuna and
dolphins, and discuss implications
concerning the ecological advantage
to tuna (or dolphins) when associated
with dolphins (or tuna).
Estimates of forage requirements
predict that tuna and dolphins should
experience severe competition under
some circumstances of prey distribu-
tion and abundance, but perhaps not
under others. Observations of over-
laps in sizes between associated tuna
and dolphins and of morphological
similarities between the animals have
implications for the importance of
swimming energetics to the associa-
tion.
These estimated forage require-
ments and considerations about
swimming energetics are discussed in
terms of their implications for deter-
mining which component (tuna or
dolphins) controls the association,
how the competition might be miti-
gated, when the association might be
more likely to occur, and how these
factors might be used to locate large
yellowfin tuna unassociated with
dolphins. The last point is important
in relation to current interest in
eliminating the practice of "dolphin-
fishing" in the eastern tropical
Pacific Ocean. "Dolphin-fishing" in-
volves location and capture of tuna
schools by locating and capturing
associated dolphin schools; as air-
breathers, the dolphins are more
easily sighted than the tuna due to
the dolphin's surface activity. Other
explanations for the bond, and poten-
678
Edwards; Associated tunas and dolphins in eastern tropical Pacific
679
tial conflicting evidence, are discussed briefly as they
relate to the energetics models and results presented
here.
Methods: Model development
and description
The tuna-dolphin association
The tuna-dolphin association occurs in the eastern
tropical Pacific Ocean (ETP) in a triangular region
roughly the size of the continental United States (~10
million km^), extending along the western coast of the
Americas from the tip of Baja California (~20°N) south
to Peru (~20°S) and seaward to ~140° W (Fig. 1). Total
productivity in this area tends to be low relative to all
other oceans, but high relative to other tropical oceans.
Ocean currents and winds generate a typical pelagic
environment in which areas of high productivity are
distributed in dynamic, nonrandom, complex patterns
(Fiedler et al. 1990, Fiedler 1992).
The ETP is characterized by an exceptionally shallow
surface mixed layer. In contrast to other areas of the
equatorial Pacific where the thermocline is generally
150-200 m deep (Kessler 1990), the depth of the ther-
mocline layer throughout much of the ETP extends
only 50-lOOm below the surface (Fig. 1). Water tem-
peratures in this wind-mixed layer are quite warm
(25-30 °C) and oxygen concentrations are high (Wyrtki
1966 and 1967, Fiedler et al. 1990, Fiedler 1992). Below
this layer, water temperatures fall relatively rapidly
(from ~27 to ~15°C) through the thermocline (usual-
ly 5-25 m vertical extent), stabilizing again below the
thermocline (Fiedler et al. 1990). Oxygen concentra-
tions also decrease relatively rapidly through the
thermocline, increasing again in cold water at greater
depths.
Strong dependence on warm water and on high con-
centrations of oxygen apparently force both tuna and
dolphins into this unusually shallow mixed layer. Tuna
must swim more or less constantly both to provide an
adequate flow of sufficiently-oxygenated water over
their gills and to locate adequate food supplies (e.g.,
Magnuson 1978, Olson and Boggs 1986). Yellowfin
tuna would likely have difficulty maintaining an ade-
quate energy balance swimming in the colder waters
below the mixed layer, nor can they afford being caught
for long in the oxygen minima characteristic of the
thermocline.
Dolphins are constrained to reside near the ocean sur-
face in order to breathe. Only temporary excursions
below the mixed layer are tolerable, both because of
this requirement for gaseous oxygen and because the
blubber layer of the tropical dolphins involved in the
tuna-dolphin association is too thin to maintain thermo-
100"E 120"' 140*
20'^ 100° SO^W
Figure 1
Depth of mixed layer in the area of the eastern tropical Pacific
Ocean characterized by associations between yellowfin tuna
Thunnus albacares and spotted dolphins Stenella attenuata.
Tuna-dolphin fishery occurs roughly in area delimited by the
300 m isocline.
nuetrality in waters much colder than that in the mixed
layer (unpubl. estimates). This is not necessarily a
disadvantage, as the major prey for associated tuna and
dolphins (small fish and squid; Perrin et al. 1973) also
tend to concentrate in this upper mixed layer, at least
periodically throughout a 24-hour day.
Although any individual tuna-dolphin association is
doubtless dynamic in the detaOs of its spatial configura-
tions and component individuals, the association in
general can be envisioned as a loose aggregation of
animals characterized by dolphins swimming relative-
ly near the ocean surface, separated vertically from the
tuna swimming below by only a few meters (Fig. 2).
Although several species of dolphins and two species
of tuna have been found to associate in the ETP, one
species of dolphin (spotted dolphin Stenella attenuata)
and one species of tuna (yellowfin Thunnus albacares)
comprise the majority (> 80%) of the associations (e.g.,
Orbach 1977, lATTC 1989). The remainder of this
paper assumes the "tuna-dolphin association" includes
only these two groups.
Energetics models
Both models followed the same format, using the stan-
dard bioenergetics approach of balancing food require-
ments against estimated energy costs for metabolism
and energy savings as growth in biomass (University
of Wisconsin Sea Grant 1989). The Wisconsin bioener-
getics model derives estimates of consumption by
680
Fishery Bulletin 90(4). 1992
Figure 2
Idealized representation of a typical association of yellowfin tuna Thunnus albacares and spotted dolphins
Stenella attenuata. Sizes and size-frequencies of tuna and dolphins are representative. Overlap in sizes be-
tween age-Ill yellowfin and neonate-lst yr dolphins (85-125cmTL) is emphasized.
iteratively fitting an energetics equation for growth in
body weight over time, to observed growth-rate curves
derived from field samples of the organism in question.
When the model growth curve simulates well the
observed growth curve, the other fluxes estimated by
the model are presumed to be reasonably accurate.
Specific rates (calories of flux ■ calories of animal " '
• day" ' ) of energy flux were estimated based on data
derived from various sources for individual tunas and
dolphins as a function of size. Rates of energy flux for
schools of dolphins and tuna were estimated as the sum
of weight-specific estimates for individuals in each
group.
Costs of reproduction were ignored for both yellowfin
and spotted dolphins; in the yellowfin model because
the model focuses on the sizes of yellowfin associated
with dolphins, which tend to be relatively immature
fish. Spawning activity in yellowfin does not occur in
fish much smaller than 80cm, and increases slowly to
the maximum activity in fish larger than ~ 150 cm
(Joseph 1963). Energy costs of reproduction for spotted
dolphins were omitted because the fraction of preg-
nant, lactating, or pregnant and lactating females in
a typical school at any time is relatively small (~25%;
see School composition).
Some of the energetics parameters reported here for
spotted dolphins are based on morphological measure-
ments from 4 dolphin specimens from the ETP; 3
spotted dolphins measuring 81-189cm total length
(TL), plus 1 spinner dolphin Stenella longirostris 114
cm in length. The 81 cm individual was a very late-term
fetus carried by the 189cm animal. Although this sam-
ple is very small, all morphological measurements from
these 4 animals fall well within the bounds of size-
related regressions of morphological characteristics
derived subsequently for a sample of 34 spotted
dolphins measuring 74-215cmTL (tip of rostrum to
fluke notch) (unpubl. data).
School composition The yellowfin model addresses
only those sizes of yellowfin found associated with
dolphins (relatively large age-II and age-Ill fish, 55-125
cmTL; Fig. 3). Based on catch records from the fisheiy,
an "average" association was assumed to include 500
yellowfin with an age composition of 65% age-II and
35% age-Ill fish per school (Ashley Mullin, lATTC, c/o
Scripps Inst. Oceanogr., La Jolla; unpubl. data from
commercial fishery).
Dolphin school composition was assumed to reflect
the apparent age distribution of the spotted dolphin
population, which in turn was assumed to appear as
the length (age) distribution of dolphins collected dur-
ing purse-seining operations in the ETP (Smith 1979,
Barlow and Hohn 1984; A. Hohn, NMFS Southwest
Edwards Associated tunas and dolphins in eastern tropical Pacific
681
40 50 60 70 80 90 100 110 120 130 140 150 160
FORK LENGTH (cm)
Figure 3
Sizes and ages of yellowfin tuna Thunnus albacares caught
with and without dolphins in the eastern tropical Pacific
Ocean, and length-interval during the first year of life by
spotted dolphins Sterwlla attenuata. Data include all years
1975-84, all areas fished, all fleets (U.S. plus non-U. S.). (Un-
publ. data from Ashley Mullin, lATTC, La Jolla).
meters, based on weight-length measurements from a
sample of 50 spotted dolphins ranging in size from 82
to 210cm TL.
Equations Each model included equations for specific
rates of consumption (Cgp), respiration (Rgp; including
both swimming activity ACT^p , and standard metab-
olism STDsp), heat of digestion (specific dynamic ac-
tion, SDAgp), and waste losses (excretion plus eges-
tion; WLgp). Specific rate of growth is estimated sim-
ply as the difference between consumption and the sum
of energy expenditures.
The form of the equation for each specific rate was
the same for both models, with the exception of Rgp ,
which was estimated for yellowfin using Boggs' (1984)
experimental results. Rgp was estimated for dolphins
following Magnuson's (1978) procedure for estimating
cost of swimming by carangiforms.
No effect of water temperature on consumption or
respiration rates appears in either model. Ambient
water temperature was assumed to be constant at
27°C, as most of the tuna-dolphin habitat occurs in
waters of this temperature.
Fish. Sci. Cent., La Jolla, unpubl. data). Proportions
of nursing calves (ages 0-2 yr), adolescents (ages 3-14
yr), sexually adult males (ages 15 and up), and sexual-
ly adult females (ages 1 1 and up) in an average school
were 0.05, 0.40, 0.25, and 0.30, respectively. Propor-
tions of adult females not pregnant or lactating, lac-
tating, pregnant, and pregnant and lactating animals
were 0.05, 0.15, 0.08, and 0.02, respectively.
Weight-length conversions The tuna model used,
as the calibration growth curve, the Gompertz fit de-
rived by Wild (1986) for yellowfin tuna from the ETP.
When necessary, body fork lengths in centimeters
(cmFL) were converted to wet weights in grams
(WWg) using the length-weight relationship (Alex
Wild, lATTC, La Jolla, unpubl. data for yellowfin tuna
from the ETP)
The calibration growth curve for expected size-at-age
in spotted dolphins was derived from equations and
figures in Hohn and Hammond (1985) and unpublished
data (A. Hohn, Southwest Fish. Sci. Cent., La Jolla).
Weight-length conversions assumed the relationship
WWkg = 1.4* 10-5 *TL2-95,
where WW^g is wet weight in kilograms, and TL is
total length (tip of rostrum to fluke notch) in centi-
Consumption Specific rate of consumption (Cgp;
calories food consumed • calories of animal "^ • day-^)
was estimated as
C,. = CONS,,,/CAL..
-■sp
CALan is total caloric content of an individual
yellowfin or spotted dolphin, estimated as a function
of wet weight in grams.
CAL,
CD* WW,
g'
where CD is caloric density (cal/g wet wt) of yellow-
fin tuna^ or spotted dolphins^.
CONSea] is total calories consumed per individual
per day, estimated as
CONSeal = CONSind*CDf,
where CDf is caloric density of food (cal/g wet wt) for
' 1440 cal/g wet wt (Boggs 1984).
-CDj = 1860 cal/g wet wt; average caloric density of four dolphins
measuring 81-189cmTL. Caloric density of each animal was deter-
mined as the sum of calories contained in blubbler, muscle, viscera,
and bone divided by total animal wet weight in grams. Average
caloric density of individual dolphins ranged from 1985 cal/g wet
wt in the 81 cm animal, to 1760 cal/g wet wt in the large adult female
(189cmTL). Assuming constant energy density for spotted dolphins
is acceptable, as spotted dolphins do not appear to exhibit any signifi-
cant seasonal, and little age-related, changes in thickness of their
blubber layer.
682
Fishery Bulletin 90(4|. 1992
yellowfin tuna^ or spotted dolphins^, and CONSjnd is
wet weight in grams of food consumed, estimated as
CONSind = C™ax*Pval*WWg,
where C^ax is maximum possible consumption (ex-
pressed as a fraction of wet weight) for the largest
yellowfin or dolphin, estimated as
r ^ r *WW cti
where ^Ca = 1.2 and ^Cb = -0.22 for yellowfin, or ^Ca
= 3.98 and 8Cb = -0.29 for spotted dolphins.
Pvai is an iteratively fitted unitless value in the
range 0-1 that, when "correct," results in the simu-
lated growth curve matching the observed growth
curve (University of Wisconsin Sea Grant 1989), and
WWg is body wet weight in grams.
Respiration (yellowfin tuna) Specific rate of respir-
ation (Rspi calories respired • calories of animal "^ •
day~^) for yellowfin tuna was estimated as
Rsp = (STD„-(-ACT„)*(20650/CD),
with energy costs of standard (STD^^.) and active
(ACTw) metabolism expressed in watts. The factor
^Energy density of yellowfin food was based on an assumed diet of
70% fish, 20% squid, and 10% inverteljrates (Olson and Boggs 1986).
with undigestible fractions of 0.124, 0.066, and 0.025, and caloric
densities of 1440, 1260, and 1000 cal/g wet wt, respectively. Average
ingested energy density (including the undigestible fraction) is 1380
cal/g wet wt.
'Energy density of spotted dolphin food changes with age (size).
Spotted dolphins nurse throughout their first year (Perrin et al.
1976). They do not begin to ingest solid food until their second year,
and they do not stop nursing entirely until their third year when
they are ~145cm in length. In this simulation, dolphins up to 1 yr
of age were assumed to consume only milk (2855 cal/g wet wt)
(Pilson and Walker 1970). Diet during the second year was assumed
to be the same as that for yellowfin tunas, with an ingested energy
density (CD,) of 1380 cal/g wet wt.
''Based on the assumption that maximum specific feeding rate for
very large yellowfin tunas (95000 g wet wt) would not exceed 10%/
day, then solving for the intercept C^ (i.e., C,, = 0.10/(95000-"")
yields 0^= 1.2. In practice, the exact value chosen for C„,„ is flex-
ible, as higher values simply reduce the fitted value of P^j, , and
vice versa.
''By analogy to walleye Stizosledion vitreum (Kitchell et al. 1977).
'Assuming maximum possible ration for adult spotted dolphins
(~75kg) would not exceed 15% of body weight/day (Sergeant 1969),
and with C„„ = 0.15, WW =85000g, and C„= -0.29, then C,,=
3.98.
"As a compromise between the unresolved arguments of Kleiber
(1961; Cb= -0.25) and Heusner (1982; 0^= -0.33) for scaling of
metabolic rate with size in mammals. This compromise was chosen
because consumption is not strictly a metabolic rate. While
Huesner's argument for metabolic rate is supported by data for
metabolic rate changes with size (see formulation for dolphin respira-
tion), no such data exist for consumption rates.
20650 converts watts to cal/day. Dividing by caloric
density of the animal (CD) produces the specific rate.
Weight-specific energy cost of standard metabolism
for yellowfin was assumed constant for all sizes of
yellowfin (Boggs 1984) as
STDw = 0.464 watts/g wet weight.
Energy cost of active metabolism (watts/g wet wt)
was estimated using Boggs' (1984) equations and data
for energy costs of activity in yellowfin,
ACT^. = F*VLG*FLH,
where VL is velocity in cm/sec and F( = 1.59 £"■*),
G( = 1.64), and H(= -1.28) are fitted parameters de-
rived from Boggs' (1984) laboratory studies on
yellowfin energetics.
Yellowfin were assumed to swim at length-specific
optimum-sustained cruising speeds, with velocity scal-
ing to fish length as
VL = VLa*FLVLb,
where 9VLa = 20.6, and i"VLh = 0.4.
Respiration (spotted dolphins) Specific rate of re-
spiration (Rsp; calories respired ■ calories of animal "^
■ day"^) for spotted dolphins was estimated as
R.
sp
(ACT,p + STDsp + HL,3p),
where ACTgp is specific rate of swimming activity,
STDgp is specific rate of standard (basal) metabolism,
and HLrsp is specific rate of residual heat loss.
Specific rates of swimming activity and standard
metabolism are estimated as
and
ACT.p = ACT,ai/CAL
STDsp = STDeai/CAL
Caloric cost of standard metabolism was estimated
as
STDeal = Sa*WW Sb,
"Intercept estimate based on lOOcniFL yellowfin swimming on
average 130cm/sec in situ (Holland et al. 1990).
'"Slope estimate based on theoretical and empirical studies by Weihs
(1973, 1981).
Edwards: Associated tunas and dolphins in eastern tropical Pacific
683
where "8^ = 1380, and 128^ = 0.67.
Caloric cost of activity!^ was estimated as
ACT,^ = PWR* 20650,
where 20650 converts watts to calories/day. Power re-
quired to swim (PWR) was estimated as
PWR = MP/(ME*PE),
where MP is mechanical power required to overcome
drag, ME is mechanical efficiency, ^^ and PE is "pro-
peller efficiency" (efficiency of propulsion by flukes)^'^.
MP (in watts) was estimated as a function of total drag
(Dti in dynes) and velocity (VL; in c/sec) as
MP = (Dt*L)/10^
where the factor 10^ converts the product Dt*L to
watts.
" Sa was assumed constant for all sizes of spotted dolphins. Given
an observed rate of 0.45mg 0, • g wet wt"' ■ hr ' for a spinner
dolphin Stenella longirostris weighing 68000 WW^ (Hampton and
Whittow 1976) and assuming 3.25 cal/mg 0„ (Elliot and Davidson
1975), then 2,386,800 (0.45*3.25*68000) calories are expended
daily in standard metabolism, and 83=1380 (2,386,800/68.000"").
The observed resting rate of oxygen consumption is consistent with
the range of resting rates (0.3-0.6 mg 0, ■ g wet wt ') reported
for bottlenose dolphins under various conditions (Hampton et al.
1971. Karandeeva et al. 1973, Hampton and Whittow 1976).
'- Heusner (1982) presents convincing statistical arguments that intra-
specific relationships between basal (standard) metabolism and body
weight in adult mammals are better described by the 2/3 power
than the 3/4 power proposed by Kleiber (1961). Heusner's argu-
ment is based on observed differences between adults of similar
species (e.g., breeds of dog); but Huesner's curve is also more
realistic because it predicts a relatively higher weight-specific rate
in smaller (younger) animals of a given species. This is more con-
sistent with Kleiber's (1961) observation that younger animals tend
to have elevated weight-specific metabolic rates compared not only
with adults of the same species, but with small adults of similar
species. In young marine mammals, weight-specific standard
metabolic rate is often at least twice the standard rate of adults
(Ashwell-Erickson and Eisner 1981, Lavigne et al. 1982). The
parameterization above results in weight-specific estimates of S
that are 2.3-1.3 times higher in dolphins measuring 80-140 cm TL
than in adult dolphins (~190cmTL). This differs by 0-11% (increas-
ing with increasing size) from basal metabolic rates of juvenile
through adult seals of similar weight (Ashwell-Erickson and Eisner
1981).
" Dolphins were assumed to swim steadily far enough below the sur-
face to eliminate the effects of surface drag (e.g., Hertel 1969).
This formulation ignores the costs of surfacing to breathe, and the
attendant increase in total distance swum to follow a sinusoidal
rather than a straight path through the water. Preliminary
estimates of these additional costs for individual dolphins of several
sizes, for reasonably realistic depths of dive and distance between
surfacings, ranged from 10 to 25% of steady swimming costs. As
this cost is relatively low, the dolphin model was not reformulated
to include these added costs of surfacing.
"ME = 0.20. by analogy to observed muscle efficiencies of terres-
trial mammals.
'■''PE = 0.85 by analogy to tunas (Magnuson 1978).
Total drag was estimated as a function of drag due
to body, fins, and movements by flukes as
Dt = (0.5*N*VL2*8„*Ct)/(1.0-FID),
where N is density of seawater (1.025g/cm3), S^ is
wetted surface area of the body, Cj is coefficient of
total drag, and FID is (fin -1- induced) drag. FID^^ is ex-
pressed here as the fractional increase in estimated
total drag due to adding the effects of fins and moving
flukes.
S„ is wetted surface area of the body, excluding flip-
pers, dorsal fin, and flukes, estimated as^^
S„ = 0.1636 *TL2i4.
Surface areas of fins are excluded from this calcula-
tion because fin drag is incorporated into the equation
for total drag as an increase of 21% over drag esti-
mated from body dimensions alone.
Ct was estimated from the formula for drag of sub-
merged streamlined bodies moving with constant
velocity
Ct = Cf*[l + (1.5*(Da/TL)3'2) + (7*(D,/TL)3)]
(Hoerner 1965, Webb 1975). Cf is the coefficient of
friction drag, and D^ is the maximum body diameter
(cm; derived from girth at axilla (Gg)) where
Ga = Gaa
■WWkgGab_
with Gaa = 25 and G^h = 0.28, based on measurements
of 50 spotted dolphins measuring 82-210cmTL.
Cf was estimated from the equation for streamlined
bodies moving submerged at constant velocity in tur-
bulent flow as
Cf = 0.072 Rl-1/5,
where Rl is Reynolds number, estimated here as
Rl = (TL*VL)/v,
where v is kinematic viscosity ( = 0.01 8tokes) assum-
ing turbulent flow at the boundary layer (Webb 1975),
and VL is velocity (cm/sec), estimated as
VL = VLa*TLVL\
■^FID was assumed = 0.21, based on the fraction of estimated total
(body -I- fin -I- induced) drag accounted for by (fin + induced) drag in
the 4-dolphin sample.
"Based on measurements of wetted surface area in the 34-dolphin
sample.
684
Fishery Bulletin 90(4). 1992
170
160
150
^ 140
« 130
i 120
- 110
o
> 100
90
80
70
DOLPHIN OPTIMUM VELOCITY
(Vop,=aL0''3)
80 100 120 140
LENGTH (cm)
Figure 4
Estimated optimum sustained swimming speed (curved line)
of yellowfin tuna Thunnus albacares and spotted dolphins
Ste-nella attenuata from the eastern tropical Pacific Ocean.
Lengths: fork length for tuna; rostrum to fluke notch ("total
length") for dolphins. Vertical bars indicate range of optimum
speeds predicted for sizes of tuna and dolphins occurring in
mixed associations. Arrows indicate observed average swim-
ming speeds of a radio-tagged 96 cm yellowfin tuna and of
tagged individual spotted dolphins swimming in situ. Size-
ranges for yellowfin tuna ages I-IV, and for spotted dolphins
from birth, are indicated above abscissa.
The term (1.0 -ME) in conjunction with ACT^p ex-
presses the fraction of total active metabolism that is
dissipated as heat, rather than converted to mechanical
energy. The term Hrsp was taken to be zero when the
estimate of Hrsp yielded a negative result. In this case,
all passive losses were more than offset by heat
generated by metabolism.
Specific rate of unavoidable passive heat loss (HLuspi
calories lost passively as heat ■ calories of animal"' •
day ') was estimated following Brodie's (1975) pro-
cedure for passive losses in large whales,
HL
usp
((21.18/BD J * (37.0 - Ta) * S^/IOOOO.O) * 24
WW„*(CDd/1000.0)
where '^BDa is average blubber depth, CD^ is caloric
density of spotted dolphins, 37.0 (°C) is the assumed
core temperature for spotted dolphins (Hampton and
Whittow 1976), T^ is ambient temperature (assumed
constant at 27°C), 21.18 is the conductivity factor for
whale blubber (Brodie 1975), and -"8^ is metabolic
surface area, estimated as
S,, = 0.84 *S».
where VLa = 20.6 and VLb = 0.43, assuming swimming
velocity scales with length in the same manner for both
spotted dolphins and yellowfin tuna (Fig. 4). Using the
same formula and parameters to predict velocity as a
function of length in both the tuna and dolphin models
maintains comparability between results from the two
models. As geometrically similar swimmers, hydro-
dynamic constraints should be approximately the same
for both tuna and dolphins.
Specific rate of residual heat loss (HL^sp; calories
heat lost in excess of that generated by active and stan-
dard metabolism, and specific dynamic action • calories
of animal"' • day')'* was estimated as
HLrsp = HL,3p-(ACT,p*(1.0-ME) + STD3p + SDAsp),
where HLrsp>0, otherwise HLrsp = 0.
"Because spotted dolphins are warm-blooded relative to their en-
vironment and because their blubber layer is not a perfect insulator,
they will constantly lose heat to surrounding water. If the sum of
estimated heat production generated by muscle activity, standard
metabolism, and specific heat of digestion equals or exceeds this
unavoidable passive loss, the term has no effect. Otherwise, the
additional heat loss was added to the animal's energy cost. In prac-
tice, the influence of the term was negligible, as differences be-
tween Hj and the sum of STD, ACT, and SDA were < 10%.
Unavoidable heat loss from fins and head is assumed
negligible, as blood flow to these areas can be adjusted
to minimize or maximize heat loss, as needed.
Specific dynamic action Specific rate of specific
dynamic action (SDA^p; calories lost as heat of diges-
tion • calories of animal' • day') was estimated as
SDAsp =
SDA*C
spi
where SDA (the fraction of consumption converted to
heat energy during digestion) = 0.15 for both yellowfin
tuna-' and spotted dolphins--.
Waste losses Specific rate of waste losses (WL^p ;
calories lost as feces or urine • calories of animal"' •
day"') were estimated as the sum of fractional losses
to egestion (F^) and excretion (Ua)
"BD, = 0.65cm, based on measurements of blubber depth at max-
imum girth for a sample of 72 spotted dolphins measuring 80-190
cmTL.
'"S,„ is the surface area of the body beneath the blubber layer. S„,
averaged 84% of S„ in the 4-dolphin sample.
^' Reflecting the relative high-protein low-carbohydrate diet ingested
by yellowfin tunas (Olson and Boggs 1986).
'" SDA is primarily a function of the protein content of ingested food,
and is ~15% for a variety of carnivores, including sea otters eating
clams and squid (10-13%, Costa and Kooyman 1984), harp seals
eating fish (17%, Gallivan and Ronald 1981). and various terrestrial
mammals fed a mixed diet (Kleiber 1961).
Edwards. Associated tunas and dolphins in eastern tropical Pacific
685
WL,p = (F, + Ua)*Csp,
where ^sp^ = 0.20 and 24Ua = 0.07 for yellowfin tuna;
Fa = 0.125 and Ua = 0.07 for spotted dolphins^s.
Growth Specific rate of growth (calories available for
growth • calories of animal" ^ • day " ^ ) was estimated as
Discussion: IVIodel impiications
The strict "result" of exercising the models is estima-
tion of food consumption by yellowfin tuna and spotted
dolphins of various sizes. This information alone is not
particularly helpful in furthering our understanding of
the tuna-dolphin bond. However, the process of model
Gsp = C,p-(R,p + SDA,p+WL3p).
Total calories available for growth (Gcai) is
Goal = Gsp*CALan.
Total grams wet- weight biomass available for growth
(Gwwg) is then
Gwwg = Gca|/CD.
The formulas and parameter values presented above
produce reasonable model estimates of the various
energy fluxes for both yellowfin tuna and spotted
dolphins (Edwards 1992).
Resuits: Estimated consumption
Despite the apparent similarity between yellowfin tuna
and spotted dolphins in food composition (prey type
and size-''), estimated food requirements for tuna and
dolphins differ considerably. Estimated food require-
ments for individual tuna and dolphins imply that each
dolphin requires 5-10 times more food per day than
each yellowfin tuna, depending on the sizes of the tuna
and dolphin being compared (Fig. 5). In a "tj^pical"
association of 200 dolphins and 500 tuna, total dolphin
requirements are still 2-5 times higher than total tuna
requirements per time-period (Fig. 6), despite the
greater number of tuna than dolphins.
^^ Based on the relative assumed nondigestible portions of tuna diet
items by analogy to similar items (Cummins and Wuycheck 1971).
-''Based on measurement of non-fecal excretion by carnivorous fish
(Brett and Groves 1979).
-^Together these processes probably account for 15-20% of ingested
food energy in spotted dolphins, as found for other small marine
mammals eating fish (Shapunov 1973, Ronald et al. 1984, Ashwell-
Erickson and Eisner 1981, Lavigne et al. 1982, and references
therein.)
-"Diet is undoubtedly an important factor in the tuna-dolphin associa-
tion, as associated yellowfin tuna and spotted dolphins apparent-
ly have nearly identical feeding preferences (Perrin et al. 1973).
Stomach contents of co-occurring tuna and spotted dolphins con-
sisted primarily of small pelagic schooling fish (e.g., mackerel Auxis
thazard) and squid of similar types and sizes.
-
Mature males
spotted
Dolphin ,-'''
Juveniles and
^^^S^^-'"^
mature females
^^^^^
^/"^^
ANNUAL RATION
•—Age III
Yellowfin
»--Age II
1 1 1 . 1 .
Tuna
4 6 8 10 12 14
AGE (year since birth)
16
Figure 5
Estimated annual ration for individual spotted dolphins
Stenella attenuata ages birth-18 yr, and yellowfin tuna Thun-
nus albacares ages II-III, occurring in mixed associations in
the eastern tropical Pacific Ocean.
CONSUMPTION vs.
SCHOOL SIZE
400
NUMBER IN SCHOOL
800
Figure 6
Estimated annual ration for schools of yellowfin tuna Thun-
nns albacares and spotted dolphins Stenella attenuata occur-
ring in mixed associations in the eastern tropical Pacific Ocean
(ETP). Solid circles indicate number of individual dolphins and
individual tuna in a typical mixed association. Ration estimates
for schools were based on average observed size-frequency
distributions of tuna and dolphin in the ETP.
686
Fishery Bulletin 90(4), 1992
development and comparisons of similar energy
fluxes in the completed models generated several in-
teresting observations with potentially significant
implications.
ticipate. The similarity in feeding preferences and
probable similarity in feeding behaviors provides one
explanation and suggests that tuna are more likely to
follow dolphins than the reverse.
Hydrodynamics and body length
Length frequencies of the tuna and dolphins in a typ-
ical association show a surprisingly strong overlap
between age-III yellowfin and neonate- 1st yr dol-
phins. Both animals begin their respective years at
~85cmTL, and complete the year at ~125cmTL
(Fig. 4). This is significant for two reasons. First, this
size range comprises the majority of the yellowfin
tuna found associated with dolphins (Fig. 3F. Second,
both animals have relatively stiff torpedo-shaped bodies
with stiff fins and carangiform swimming behavior.
Because theory predicts that optimum swimming
speeds (the speed at which the least energy is consumed
for a given distance covered) of geometrically-similar
swimmers will be comparable (Weihs 1973, Webb
1975), the similar body forms and swimming behaviors
of the tuna and the dolphins imply that optimum swim-
ming speeds will also be similar for either animal of
a given length.
Swimming speeds of sonic-tagged yellowfin tuna
measured in situ show that individual undisturbed
yellowfin, of the size most often found associated with
dolphins, choose in their natural environment to swim
on average at their predicted optimum cruising speed
(e.g., yellowfin 90-lOOcmFL swim at 100-130cm/sec;
Holland et al. 1990). Because yellowfin tuna tend to
associate in schools of like-sized individuals, the
expected speed of the tuna group is similar to the
expected speed of the individuals involved.
In contrast, tracking studies (Perrin et al. 1979) of
spotted dolphins in the ETP indicate that dolphin
schools swim on average not the speed most efficient
for the majority of the individuals in the school (i.e.,
~160-170cm/sec for large adults) but the speed most
efficient for the neonate-lst yr animals (~120 cm/sec;
Fig. 4).
These observations imply that yellowfin associating
with dolphin schools may do so at little or no added
hydrodynamic cost. The associated fish, unlike larger
or smaller sizes of yellowfin, need swim neither faster
nor slower than their apparently preferred optimum
in order to maintain an association with dolphins.
The observation that associating with dolphins may
cost tuna little does not explain why the tuna par-
' Figure 3 includes fish from all areas of the fishery, not just the
offshore areas where most dolphin fishing occurred during the years
these data were collected, causing dolphin-fish distribution to be
skewed to left.
Who follows whom
The higher forage requirements of dolphins both in-
dividually and as an association imply that dolphins
following tuna, particularly single dolphin schools
following single tuna schools, would fall far short of
meeting their daily energy requirements. Dolphin
schools might avoid this energy deficit by switching
from one tuna school to another, but they would have
to switch consistently from recently-successful to
soon-to-be-successful schools of foraging tuna. This
frequent switching could be difficult because it would
likely involve periods of searching at speeds greater
than sustainable by the young dolphins, in order to find
new tuna schools (and new patches of forage) faster
than the patches could be found by the current tuna
school.
Measurements of muscle mass and estimates of
power-time curves for various sizes of spotted dolphins
imply that the relatively small muscle mass of neonate-
lst yr dolphins probably cannot sustain speeds much
faster than their predicted optimum for any extended
length of time (unpubl. data). If searching for new
schools of tuna requires sustained accelerated swim-
ming, the young dolphins could have trouble keeping
up with the rest of the school. Because it is unlikely
that dolphins, as nursing mammals and highly social
animals, would simply leave their young behind, switch-
ing frequently from one tuna school to another may not
be a practical option.
The disparity in feeding requirements implies that,
while dolphins would probably be disadvantaged by
having to rely upon tuna to locate sufficient prey, the
tuna could recognize an advantage by following dol-
phins. The fish would then be associating with another
predator that is searching for the same prey, but which
must encounter that prey either more often or in con-
siderably larger patches than required by the tuna, per
time period.
However, the greater need of the dolphins for food
implies concomitantly that competition for resources,
if those resources are limited, could be fierce. The
schooling characteristics of the predators and prey,
coupled with feeding behaviors and differing sizes of
the predators, provide one possible explanation for the
ability of the smaller yellowfin tuna under some cir-
cumstances to persist in this potentially competitive
association despite the dolphin's greater size, and need
for food.
Edwards Associated tunas and dolphins in eastern tropical Pacific
687
WHEN SHOULD TUNA ASSOCIATE WITH DOLPHIN?
PREY
ABUNDANCE:
PATCH TYPE;
RARE
LARGE
LOW
FREQUENT
SMALL
RARE
LARGE
HIGH
FREQUENT
SMALL
ABILITY TO
LOCATE PREY:
D>T
YES
NO
YES
NO
D = T
NO
NO
NO
NO
D<T
NO
NO
NO
NO
Figure 7
Decision table predicting conditions under which yellowfin
tuna Thunnus albacares and spotted dolphins Stenella attenu-
ata should (or should not) associate.
Avoiding competition for food
As is characteristic of pelagic ocean systems, both
predators and prey in the ETP occur in clumped
distributions. Individuals occur in schools or aggrega-
tions separated by (often vast) distances devoid of other
individuals. The prey, like the yellowfin tuna, will tend
to occur in schools of like-sized individuals with similar
swimming speeds. Aggregations of tuna and dolphins
will typically consist of dolphins of assorted sizes ac-
companied by tuna of approximately one size. The
feeding strategy of the predators will involve searching
for a clump of prey, simultaneous (or nearly so) arrival
at the prey patch by both tuna and dolphins, and
repeated incursions by individuals of both predator
groups into the clump of prey wherein prey are seized
and swallowed whole individually.
Associated yellowfin tuna may be able to mitigate
this direct competition with dolphins for food on the
basis of the difference in size between the fish and the
feeding adult dolphins (~100cm vs. 200cmTL). Be-
cause the tuna are smaller, they have smaller maximum
stomach capacity (~400g wet wt for age-2 yellowfin,
~1100g wet wt for age-3 yellowfin; Olson and Boggs
1986) compared with spotted dolphins (~2000g wet wt
in adults; Bernard and Hohn 1989). Even if the smaller,
presumably more-agile tuna could seize individual prey
only as fast as the dolphins and no faster, they would
satiate more quickly than the dolphins.
As both groups would begin feeding at the same time,
when the prey concentration was maximum, the tuna
at any time would be relatively closer than the dolphins
to satiation, given the observed (average) relative pro-
portions of tunas and dolphins in a typical association.
The tuna would be filling their stomachs while the prey
were still relatively dense. Depending on the size of the
prey patch, dolphins might never succeed in satiating,
even though the tuna had their fill. Even if the prey
patch was sufficiently limited that neither group
achieved satiation, the tuna would always be relative-
ly more full at any given time. Thus, although the
dolphins require more prey overall, the tuna could suc-
ceed competitively by satiating sooner (being relative-
ly more successful) during any given prey encounter.
However, it may not always be to the tuna's ad-
vantage to associate with dolphins, even given this
scenario. The benefit (or not) can be assessed by
evaluating the relative advantages of associating or
not, given the range of possibilities for prey spatial
distribution and abundance.
WInen should the association occur?
The possibilities can be summarized in a simple deci-
sion table (Figure 7). At the extremes, prey abundance
may be either low or high and any given abundance
may be either homogeneously distributed (frequent) or
clumped (rare). The possibilities for locating prey are
that (1) dolphins are more adept than tuna, (2) both are
equally adept, or (3) dolphins are less adept than tuna.
The advantages for tuna to associate with dolphins can
be assessed for each cell in the table.
Consideration of each cell in the table suggests that
tuna may benefit from associating with dolphins only
when (1) prey are distributed in rare patches and (2)
dolphins are more adept than the tuna in finding these
patches. This would be true regardless of prey concen-
tration within the patches, because whenever tuna and
dolphin associate they will compete for food. If tuna
are more adept than dolphins at finding food, then
there will be no foraging-related advantage for the tima
to associate with their competitors. The tuna would be
able to find food more easily on their own than by
following dolphins, and would not have to risk sharing
these resources once located. If the tuna and dolphins
are equally adept, there is still no advantage, for the
same reason.
If the prey are distributed in relatively small but
numerous patches, there is still no advantage for tuna
to associate with dolphins, again because the spatial
frequency of schools would produce a relatively high
probability of tuna encountering the food on their own
without risk of sharing with their competitor. In addi-
tion, when patches are small, the tuna would be espe-
cially disadvantaged by having to compete with
dolphins because the presence of dolphins could pre-
vent the tuna from satiating, despite the fact that the
tuna would still be relatively more full than the dolphins
when the patch had been exhausted.
688
Fishery Bulletin 90(4), 1992
But when the prey are distributed in rare patches
and the dolphins are more adept than the tuna at
locating these patches, then tuna could benefit from
associating with dolphins because the fish could en-
counter food more often than if they were not asso-
ciated. This will be true regardless of the density of
the prey patch.
It is never the case that dolphins benefit energetically
from depending entirely on tuna for finding prey,
because dolphin forage requirements are so much
higher than tuna requirements.
These conclusions lead to the hypothesis that tuna-
dolphin associations should be more prevalent in areas
where oceanic conditions encourage strong clumping
of prey, and less prevalent when conditions encourage
a more homogeneous distribution of prey. I am current-
ly exploring, with a simulation model of tuna, the
movements of dolphin and prey in response to envi-
ronmental characteristics of the ETP (work in pro-
gress). Further studies correlating oceanic environmen-
tal characteristics with catches of various size-classes
of tuna are planned but not yet underway. If the sug-
gestions described above are borne out, it may be pos-
sible to identify areas of the ETP where large yellowfin
tuna could be captured without having to rely on
dolphin-associated fishing.
Caveats
This study assumes that average size of dolphin schools
remains constant at about 200 animals. This is the
average school size for spotted dolphins observed dur-
ing dolphin survey research cruises in the ETP. In fact,
neither school size nor school composition are constant.
Observers on both research and commercial vessels
report school sizes ranging from a few animals to many
hundreds. Scott (1991) reports diel changes in sizes of
schools sighted by tuna fishermen in the ETP.
However, these inconsistencies may not significant-
ly affect the implications of the energetics estimates
presented here. Average sizes of dolphin schools cap-
tured with tuna in the ETP are considerably larger
(400-600 animals) than the average school size ob-
served during research surveys because the fishermen
preferentially search and capture large schools of
dolphins, which tend to carry more tuna. Estimates
concerning the relative importance of tuna and dolphins
to energetics of the association are probably reasonably
similar for both large and small associations, because
in both cases the proportions of tuna and dolphins tend
to be similar (i.e., as the number of dolphins increases,
in general the number of associated tuna increases).
The study of diel differences (Scott 1991) shows that
school sizes of dolphins sighted in association with tuna
vary from a morning low to a late-afternoon high, but
the change is relatively small, from ~450 to ~600
animals on average.
Other exptanations for the bond
Other hypotheses have been proposed to explain the
tuna-dolphin association. The two most-often suggested
are the possibility that tuna perceive dolphin schools
as FADs (fish aggregating devices) or as protection
from sharks. Both of these factors may well contribute
to the strength of the bond; neither precludes the
energetics results discussed above.
The propensity for fish to collect around floating ob-
jects is well known, although the reasons are not yet
understood. Presumably, floating objects provide a
reference point for the aggregating tuna and in some
way increase foraging success, perhaps by concen-
trating prey items or by tracking convergence areas
where prey densities may be higher than elsewhere.
The FAD hypothesis has merit for the sizes of tuna
actually found with dolphins in the ETP, for two
reasons in particular. First, associating with dolphins
may increase foraging success for the associated tuna
because both tuna and dolphins are apparently seek-
ing the same prey and dolphins may be more adept at
finding it. Thus, tuna are associating with a FAD that
does not simply attract appropriate prey passively, but
actively searches and finds it. Second, tuna are required
to swim constantly in order to ventilate their gills. It
appears convenient that the average observed speed
of dolphin schools is also the optimum speed of the sizes
of tuna usually found associated with these schools.
Rather than circling a stationary FAD, tuna associated
with dolphin schools will cover a much larger area while
moving at their most efficient cruising speed, and will
cover that area in the presence of a sentient foraging
FAD.
The shark protection hypothesis derives from a com-
mon perception that dolphins actively protect their
young by driving sharks from their vicinity. If this is
so, tuna associating with dolphins may be associating
with the best of all possible FADs; a floating object that
moves at the tuna's optimal speed, moves in search of
the same prey the tuna would like to find, is probably
at least as adept as the tuna at finding that preferred
prey, and which provides protection against, rather
than increased risk of, predation (FADs of course con-
centrate not only fish, but also their predators).
Both the FAD and shark hypotheses assume that
tuna follow dolphins. Not all hypotheses assume that
tuna are the followers. Au and Pitman (1986) and Au
(1991) suggest, for example, that dolphins follow tuna
in order to take advantage of tuna foraging in conjunc-
tion with bird flocks. This would be an advantage for
dolphins during the actual feeding event. However, it
Edwards: Associated tunas and dolphins in eastern tropical Pacific
689
does not solve the problem that dolphins apparently
must locate not only the same type of prey as large
yellowfin tuna, but quite a bit more of it during any
given time-period. Following tuna does not appear ade-
quate to fulfill dolphin schools' energy requirements.
This fundamental difference in food energy require-
ments may be the single most important biological
factor underlying the association. Oceanographic
conditions (the shallow mixed layer) set the stage;
energetics requirements (hydrodynamics and foraging
patterns) appear to constrain the roles. Although the
definitive answer has yet to be demonstrated quan-
titatively, the energetics-based hypotheses presented
here are at least consistent with currently available
data. The tuna-dolphin association may be a conse-
quence of a combination of oceanography, hydro-
dynamics, foraging energetics, and life-history
characteristics, i.e, a consequence of the ecology of the
association's components.
Acknowledgments
This study could not have been completed without
generous sharing of data, time, and constructive ad-
vice by helpful individuals from, but not limited to, the
National Marine Fisheries Service, Southwest Fish-
eries Science Center, and the Inter- American Tropical
Tuna Commission. This work was completed while the
author was a National Research Council PostDoctoral
Research Associate at the Southwest Fisheries Science
Center, La Jolla, CA.
Citations
Ashwell-Erickson, S., and R. Eisner
1981 The energy cost of free existence for Bering Sea liarbor
and spotted seals. In Hood, D.W., and J. A. Calder (eds.), The
eastern Bering Shelf: Oceanography and resources, vol. 2, p.
869-899. Univ. Wash. Press, Seattle.
Au. D.W.
1991 Polyspecific nature of tuna schools: Shark, dolphin, and
seabird associates. Fish. Bull., U.S. 89:343-354.
Au, D.W.K., and R.L Pitman
1986 Seabird interactions with dolphins and tuna in the eastern
tropical Pacific. Condor 88(3):304-317.
Barlow, J., and A. Hohn
1984 Interpreting spotted dolphin age distribution. NOAA-
TM-NMFS-SWFC-48, Southwest Fish. Sci. Cent., La Jolla,
21 p.
Bernard, H.J., and A. A. Hohn
1989 Difference in feeding habits between pregnant and
lactating spotted dolphins {Stenella attenuata). J. Mammal.
70(1):211-215.
Boggs, C.H.
1984 Tuna bioenergetics and hydrodynamics. Diss. Abstr. Int.
45/02-B:482.
Brett, J.R., and D.D. Groves
1979 Physiological energetics. In Hoar, W.S., et al. (eds.),
Fish physiology, vol. 8, p. 280-354. Academic Press, NY.
Brodie, P.P.
1975 Cetacean energetics, an overview of intraspecific size
variations. Ecology 56:152-161.
Costa, D.P., and G.L. Kooyman
1984 Contribution of specific dynamic action to heat balance
and thermoregulation in the sea otter, Enhydra lutris.
Physiol. Zool. 57(2):199-203.
Cummins, K.W., and J.W. Wuycheck
1971 Caloric equivalents for investigations of ecological ener-
getics. Int. Assoc. Theor. Appl. Limnol. Proc. 18:1-158.
Edwards. E.F.
1992 Evaluation of energetics models for yellowfin tuna (Thun-
nus albacares) and spotted dolphin {Stenella atteyiuata) in the
eastern tropical Pacific Ocean. Admin. Rep. LJ-92-01, NMFS
Southwest Fish. Sci. Cent., La Jolla. 43 p.
Elliot, J.M., and W. Davidson
1975 Energy equivalents of oxygen consumption in animal
energetics. Oecologia (Heidelb.) 19:195-120.
Fiedler, P.C.
1990 Seasonal climatologies and variability of eastern tropical
Pacific surface waters. NOAA Tech. Rep. NMFS 109, 65 p.
Fiedler. P.C, L.J. Lierheimer. S.B. Reilly, S.N. Se.vton,
R.S. Holt, and D.P. DeMaster
1990 Atlas of eastern tropical Pacific oceanographic variability
and cetacean sightings, 1986-1989. NOAA-TM-NMFS-
SWFSC-144. Southwest Fish. Sci. Cent., La Jolla, 142 p.
Gallivan G.J.. and K. Ronald
1981 Apparent specific dynamic action in the harp seal (Phoca
groenlandica). Comp. Biochem. Physiol. 69A:579-581.
Hampton. I.F.G.. and G.C. Whittow
1976 Body temperature and heat exchange in the Hawaiian
spinner dolphin Stenella longirostris. Comp. Biochem.
Physiol. 55A:195-197.
Hampton. LF.G.. G.C. Whittow. J. Szekerczes, and S. Rutherford
1971 Heat transfer and body temperature in the Atlantic
bottlenose dolphin, Tursiops truncatus. Int. J. Biometrics
15(2-4):247-253.
Hertel, H.
1969 Hydrodynamics of swimming and wave-riding dolphins.
In Anderson, H.T. (ed.), Biology of marine mammals, p.
31-63. Academic Press, NY,
Heusner, A. A.
1982 Energy metabolism and body size. Part I. Is the 0.75
mass exponent of KJeiber's equation a statistical artifact?
Respir. Physiol. 48:1-12.
Hoerner. S.F.
1965 Fluid dynamic drag: Practical information on aero-
dynamic drag and hydrodynamic resistance. Published by
author, Midland Park, NJ, 468 p.
Hohn, A. A., and P.S. Hammond
1985 Early postnatal growth of the spotted dolphin in the off-
shore eastern tropical Pacific. Fish. Bull., U.S. 83:533-556.
Holland, K.M., R.W. Brill, and R.K.C. Chang
1990 Horizontal and vertical movements of yellowfin and
bigeye tuna associated with fish aggregating devices. Fish.
Bull., U.S. 88:493-507.
lATTC (Inter-American Tropical Tuna Commission)
1989 Annual report. 1988. Inter-Am. Trop. Tuna Comm.,
Scripps Inst. Oceanogr., La Jolla, 288 p.
Joseph, J.
1963 Fecundity of yellowfin tuna (Thunnus albacares) and skip-
jack (Katsuwonus pelamis) from the eastern Pacific
Ocean. Int.-Am. Trop. Tuna Comm. 7(4):257-292.
690
Fishery Bulletin 90(4). 1992
Karandeeva, O.G., S.K. Matisheva, and V.M. Shapunov
1973 Features of external respiration in the Delphinidae. In
Chapskii, K.K., and V.E. Salokov (eds.), Morphology and
ecology of marine mammals, p. 196—206. John Wiley, NY.
Kessler, W.S.
1990 Observations of Long Rossby Waves in the northern
tropical Pacific. J. Geophys. Res. 95(C4):5183-5217.
Kitchell, J.F., D.J. Stewart, and D. Weininger
1977 Application of a bioenergetics model to yellow perch
(Perca flavescens) and walleye {Stizostedion vitreum tntreum).
J. Fish. Res. Board. Can. 34:1922-1935.
Kleiber, M.
1961 The fire of life. John Wiley, NY, 453 p.
Lavigrne, D.M., W. Barcyard, S. Innes, and N.A. Oritsland
1982 Pinniped bioenergetics. In Mammals in the sea, vol. 4,
p. 192-235. FAG Fish. Ser. 5.
Magnuson, J.J.
1978 Locomotion by scombrid fishes: Hydrodynamics, mor-
phology and behavior. In Hoar, W.S., et al. (eds.). Fish phys-
iology, vol. 7, p. 239-313. Academic Press, NY.
Olson, R.J.. and C.H. Boggs
1986 Apex predation by yellowfin tuna (Thunnus albacares):
Independent estimates from gastric evacuation and stomach
contents, bioenergetics, and cesium concentrations. Can. J.
Fish. Aquat. Sci. 43(9):1760-1775.
Orbach, M.
1977 Hunters, seamen and entrepreneurs. The tuna seinermen
of San Diego. Univ. Calif. Press, Berkeley, 304 p.
Perrin, W.F.
1969 Using porpoises to catch tuna. World Fishing 18(6), 4 p.
Perrin, W.F., R.W. Warner, C.L. Fiseus, and D.B. Holts
1973 Stomach contents of porpoise, Stenetla spp., and yellowfin
tuna, Thunnus albafares, in mixed-species aggregations. Fish.
Bull., U.S. 71:1077-1092.
Perrin, W.F., J.M. Coe, and J.R. Zweifel
1976 Growth and reproduction of the spotted porpoise, SteneUa
attenuata, in the offshore eastern tropical Pacific. Fish. Bull.,
U.S. 74:229-269.
Perrin. W.F., W.E. Evans, and D.B. Holts
1979 Movement of pelagic dolphins (SteneUa spp.) in the
eastern tropical Pacific as indicated by results of tagging, with
summary of tagging operations, 1969-1976. NOAA Tech.
Rep. NMFS-SSRF-737, 14 p.
Pilson, M.E.Q., and D.W. Walker
1970 Composition of milk from spotted and spinner dolphins.
J. Mammal. 51(l):74-79.
Ronald, K., K.M. Keiver, F.W.H. Beamish, and R. Frank
1984 Energy requirements for maintenance and faecal and
urinary losses of the grey seal (Halichoerus grypus). Can. J.
Zool. 62:1101-1105.
Scott. M.
1991 The size and structure of pelagic dolphin herds. Ph.D.
thesis, Univ. Calif, Los Angeles, 165 p.
Sergeant. D.E.
1969 Feeding rates in cetacea. Fiskeridir. Skr. Ser.
Havunders. 15:246-2.58.
Shapunov, V.M.
1973 Food requirements and energj' balance in the black sea
bottlenose dolphin (Tursiops truncatus ponticiis bai-abasch). In
Chapskii, K.K., and V.E. Solokov (eds.). Morphology and
ecology of marine mammals, p. 207-212. John Wiley, NY.
Smith, T.D. (editor)
1979 Report of the status of porpoise stocks workshop, Aug.
27-31, La Jolla, CA. Admin. Rep. LJ-79-41, NMFS Southwest
Fish. Sci. Cent.. La Jolla. 138 p.
University of Wisconsin Sea Grant
1989 A generalized bioenergetics model of fish growth for
microcomputers. WIS-SG-87-245, Univ. Wise. Madison, 47 p.
Webb, P.W.
1975 Hydrodynamics and energetics of fish propulsion. Fish.
Res. Board Can. Bull. 190, 159 p.
Weihs, D.
1973 Optimal fish cruising speed. Nature (Lond.) 245:48-50.
1981 Voluntary swimming of speeds of two species of large
carcharhinid sharks. Copeia 1:219-222.
Wild, A.
1986 Growth of yellowfin tuna, Thunnus albacares. in the
eastern Pacific Ocean based on otolith increments. Int. -Am.
Trop. Tuna Comm. Bull. 18(6), 56 p.
Wyrtki, K.
1966 Oceanography of the eastern equatorial Pacific Ocean.
Oceanogr. Mar. Biol. Annu. Rev. 4:33-68.
1967 Circulation and water masses in the eastern equatorial
Pacific Ocean. Int. J. Oceanol. Limnol. 1(2):117-147.
Abstract. — A principal mechan-
ism underlying a production hypoth-
esis that artifical reefs increase en-
vironmental carrying capacity and
eventually the biomass of reef-asso-
ciated organisms is that these struc-
tures reduce predation on reef res-
idents. We tested this predation
mechanism with a series of field ex-
periments at two sites (inner-bay
sand-seagrass flat, and outer-bay
seagrass bed adjacent to coral reefs)
in Bahia de la Ascension, Mexico. We
examined survival of two size-classes
of juvenile Caribbean spiny lobster
Panulirus argus tethered in sea-
grass beds with and without access
to artificial lobster shelters, and at
different distances from the shelters.
The artificial shelters were concrete
structures (casitas) that simulate
lobster dens. Large juvenile lobsters
(56-65 mm CL) attained a relative
size refuge when tethered 60 m away
from casitas compared with smaller
(46-55 mm CL) lobsters. Conversely,
the small lobsters survived better
beneath casitas than did large lob-
sters. Small juveniles also survived
better at casitas or 30 m away from
casitas than at 15 m or 70 m away.
Observations indicated that the day-
time predator guild, composed pri-
marily of snappers (family Lutjani-
dae), seldom foraged more than 60 m
from casitas and were typically with-
in 15 m of casitas. There was also a
significant positive correlation be-
tween predation-induced lobster
mortality and numbers of snapper
associated with casitas at the inner-
bay site. Thus, tethering lobsters 70
m away from casitas appeared ade-
quate to examine survival of lobsters
in an environment uninfluenced by
daytime predators aggregating to
casitas. These results indicate that
(1) the relative importance of a lob-
ster-size refuge from predators
varies according to shelter availabil-
ity, and (2) that there is a nonlinear
relationship between predation risk
and distance from an artifical
shelter. Our results demonstrate that
casitas increase survival of small
juvenile lobsters but reduce survival
of larger juveniles. Small casitas
scaled according to body size may
enhance survival of large juvenile
lobsters in nursery habitats where
large conspecifics are removed from
large casitas.
Manuscript accepted 13 July 1992.
Fishery Bulletin, U.S. 90:691-702 (1992).
Artificial shelters and survival
of juvenile Caribbean spiny lobster
Panulirus argus: Spatial, habitat,
and lobster size effects*
David B. Eggleston
The College of U/illiam and Mary, Virginia Institute of Marine Science
Gloucester Point, Virginia 23062
Caribbean Marine Researcfn Center, Lee Stocking Island. Exuma Cays, Bahamas
Present address: College of Ocean and Fishery Sciences WH-10
University of Washington, Seattle, Washington 98195
Romuald N. Lipcius
The College of William and Mary, Virginia Institute of Marine Science
Gloucester Point, Virginia 23062
Caribbean Marine Research Center, Lee Stocking Island, Exuma Cays, Bahamas
David L. Miller
Department of Geography, State University of New York, Cortland. New York 13045
Artificial reefs are in use worldwide
as a means of increasing local abun-
dance of finfish and invertebrates
(see reviews by Bohnsack and Suth-
erland 1985, Grove and Sonu 1985,
Mottet 1985, Bohnsack 1989). The
use of artificial reefs to increase fish-
eries production remains controver-
sial because it is unknown whether
these structures (1) provide critical
resources that increase the environ-
mental carrying capacity and even-
tually the biomass of reef-associated
organisms (production hypothesis),
or (2) merely attract and aggregate
organisms from surrounding areas
without increasing total biomass (at-
traction hypothesis) (Bohnsack 1989).
The attraction hypothesis is an im-
portant consideration for artificial-
reef-based fisheries that may be vul-
nerable to overexpioitation. Thus,
there is a need for ecological inves-
tigations capable of assessing the im-
pact of artificial reefs upon species
distribution, abundance, and survival
* Contribution 1725 of the Virginia Institute
of Marine Science.
patterns, and the processes underly-
ing these patterns.
Artificial reef technology has tradi-
tionally been based on the assump-
tion that obligate reef dwellers (e.g.,
reef fishes and lobsters) are limited
locally or regionally by the availabil-
ity of shelter (Bohnsack 1989, Hixon
and Beets 1989, Eggleston et al.
1990 and references therein). Con-
versely, artificial reefs also concen-
trate numerous potential predators
(Hixon and Beets 1989, Eggleston et
al. 1990); increased predation pres-
sure at or near these structures could
outweigh the benefits from increases
in production. For instance, fishes
and lobsters normally dispersed over
a wide area could be concentrated
and consumed by predators more
rapidly in a smaller area. Thus, arti-
ficial shelters may either enhance or
reduce the survival of their inhabi-
tants, depending upon predator
responses. In this paper, we present
the results of a series of field ex-
periments comparing survival rates
of two size-classes of juvenile Carib-
bean spiny lobster Panulirus argus
691
692
Fishery Bulletin 90(4), 1992
Latreille, with and without access to artificial shelters
at different spatial scales in seagrass beds. We then
discuss these mortality patterns in terms of the relative
importance of lobster size, shelter availability, and
distance of lobsters from the artifical shelter. More-
over, we use daytime abundance and foraging ranges
of shelter-associated predators to speculate on the
mechanisms underlying these mortality patterns.
Juvenile P. argus inhabit shallow bays throughout the
tropical and subtropical western Atlantic where they
frequently aggregate during the day in crevices of coral
and rocky reefs (Berrill 1975, Herrnkind et al. 1975).
Gregarious behavior within dens probably enhances in-
dividual survivorship because spiny lobsters collectively
use their spinose antennae to fend off diurnally active
predators (Berrill 1975, Cobb 1981, Zimmer-Faust and
Spanier 1987, Eggleston and Lipcius 1992). However,
intra- and interspecific competition for suitable dens
can force smaller juvenile P. argus out of these dens
(Berrill 1975). Predation represents a major source of
mortality for juvenile spiny lobsters (Munro 1974,
Herrnkind and Butler 1986, Howard 1988, Smith and
Herrnkind 1992), and when individuals are displaced
or forced to shelter in an inadequate den they may be
subject to increased predation rates (Herrnkind and
Butler 1986, Eggleston et al. 1990).
Large juvenile and adult spiny lobsters are the focus
of intense commercial and recreational fisheries in
south Florida and the Caribbean, with the possibility
of regional overexploitation of spiny lobster fisheries
(U.S. Agency for International Development 1987).
Several Caribbean nations have met increased market
demand with the large-scale use of artificial shelters
to concentrate lobsters and facilitate harvest (e.g.,
Mexico-Miller 1989, Lozano-Alvarez et al. 1991;
Cuba— Cruz and Brito 1986; Bahamas— R.W. Thomp-
son, Dep. Fish., Nassau, Bahamas, pers. commun., May
1991). These artificial shelters, commonly referred to
as "casitas Cubanas" (see Fig. 1), attract and concen-
trate a broad size-spectrum of juvenile P. argus,
particularly in nursery areas (Eggleston et al. 1990,
Lozano-Alvarez et al. 1991).
Predation intensity in and around artificial shelters
is affected by numerous factors including the sizes of
predator, prey, and shelter (Hixon and Beets 1989,
Eggleston et al. 1990), and distance from the reef
(Shulman 1985). Moreover, since most crustaceans
have indeterminate growth (Hartnoll 1982), they must
continually search for larger shelters as they grow, a
process that involves predation risk that is inversely
related to body size (e.g., Scully 1983, Reaka 1987,
Vermeij 1987). Hence, we hypothesized that (1) the
relative importance of a lobster size refuge would vary
according to shelter availability, and (2) that the im-
pact of artificial shelters upon predation-induced mor-
tality of juvenile lobsters would vary according to the
distance of unprotected lobsters from these shelters.
We tested these hypotheses experimentally in the field
by quantifying the survival of tethered spiny lobster
juveniles in seagrass beds of Bahia de la Ascension,
Mexico. This bay is a productive nursery for juvenile
Panulirus argus and supports a commercial fishery for
large juveniles and adults (Miller 1989, Lozano-Alvarez
et al. 1991). Experimental factors included (1) presence
or absence of artificial shelter, i.e., casitas Cubanas,
(2) lobster size, (3) site, and (4) distance between
tethered, unprotected lobsters and artificial shelters.
Methods and materials
Study site
Tethering experiments were conducted in Bahia de la
Ascension, a large bay (~740km") within the Sian
Ka'an Biosphere Reserve, Mexico (19°45'N, 87°29'W)
(Fig. 2). Two experimental sites with contrasting
Figure 1
A large "casita Cubana"
constructed with a frame
of PVC-pipe and roof of
cement (177cm length x
118cm width x 6cm
height of opening).
Eggleston et al : Artificial shelters and survival of juvenile Panulirus argus
693
19°50'
habitats were chosen to compare
relative rates of predation: an
inner-bay sand-seagrass (Thalas-
sia tesUidinum) flat located at
the northwestern portion of the
bay, and an outer-bay seagrass
meadow adjacent to a coral reef
(Fig. 2). Seagrass and algal habi-
tats likely provide the only natu-
ral daytime refuge for juvenile P.
argus in this system because of
an apparent lack of crevices
(formed by rocky outcrops, patch
coral reefs, sponges, solution
holes, or undercut seagrass
banks). Anecdotal information
from lobster fishermen present
in Bahia de la Ascension prior to
the introduction of casitas
(around 1974) indicated that ju-
venile lobsters commonly resided
solitarily under dense stands of
Thalassia or complex red algae
(e.g., Laurencia), or aggregated
around existing structures such
as sponges or cobble. Moreover,
previous tethering experiments
with juvenile P. argus in this
system demonstrated that sea-
grass and algae provide some
protection for spiny lobster
juveniles from predators (R.N.
Lipcius et al., unpubl. data).
Differences in seagrass density between and within
sites were determined prior to experiments by measur-
ing dry-weight biomass (g) of Thalassia removed from
0.25 m^ plots. The inner-bay site was composed of
sparse seagrass patches (x Thalassia biomass 62.4
g/m-. A'' 6, SD 10.7) interspersed among coarse cal-
careous sand and coral rubble. The coral rubble was
covered mostly by green and red algae (Dasycladus
spp. and Laurencia spp., respectively), but also sup-
ported larger sponges. The outer-bay site was located
shoreward of a fringing coral reef and composed of
sand patches and patch corals interspersed among
moderate to dense seagrass beds (i Thalassia bio-
mass 111.6g/m2, N 6, SD 13.4, and 210.0g/m2, N 6,
SD 12.6, respectively). Further details of the study site
are described in Eggleston et al. (1990).
Artificial shelters
Our design of artificial lobster shelters was based on
"casitas Cubanas"— sunken wood and concrete struc-
tures that simulate lobster dens (Miller 1989) (Fig. 1).
87°30'
-" KReef
GULF OF '\_J
MEXICO
CUBA
€>
Figure 2
Study sites at Bahia de la Ascension, Mexico.
The large casitas used in this study (177x118x6cm)
were constructed with a reinforced concrete roof bolted
to a supporting PVC-pipe frame. Several physical prop-
erties of the casita appear to make it an optimal lobster
den: (1) shaded cover provided by the wide concrete
roof, (2) a low ceiling that excludes large piscine
predators, and (3) multiple den openings which are
smaller than the inner roof height of the casita (Fig.
1) (Eggleston et al. 1990). Hence, the use of casitas per-
mitted us to standardize den size and availability in dif-
ferent habitats.
Tethering experiments
and predator observations
Spiny, lobsters were collected from existing casitas and
held in traps for 1-2 days prior to initiation of each ex-
periment. Only intermolt lobsters exhibiting strong
"tail flipping" responses were used in tethering experi-
ments. Tethers were constructed by locking a plastic
cable-tie around the cephalothorax of a lobster, be-
tween the second and third walking legs, and securing
the cable-tie with cyanoacrylate cement. The cyano-
694
Fishery Bulletin 90(4), 1992
Inner-bay site
ra
c
o ;
CO :
ra
@
@
@
e
<M> <|. <M> <|> <«> <t> <g>
lOOm
Outer-bay site
60m
••.■wv ■
60m
<l> <^
60m
e
H
X
X X
/ X X X
X X
XXX
X X
X X
O X
e:x X
XXX
X X
XXX
X X
X X
60m
m
&
@
Figure 3
Schematic of casita layout at the (upper) inner-bay and (lower) outer-bay sites for the January 1989 experiment. S
casita. M = medium casita, L = large casita, and NC = no-casita station.
small
Eggleston et al Artificial shelters and survival of juvenile Panulirus argus
695
acrylate cement ensured that a piece of carapace re-
mained on the line as evidence of predator-induced mor-
tality. An empty cable tie without a piece of carapace
attached to it was scored as an escape. Each cable-tie
was connected with 301b test monofilament line either
to another cable-tie and attached to a shelter, or at-
tached to a square wire-metal frame that was posi-
tioned outside of the triangular casita station (Fig. 3)
on the seagrass bed with lead weights. The wire-metal
frame had the same length-width dimensions as the
large casita but did not provide shelter. The metal
frame was chosen over stainless-steel stakes because
stakes could not penetrate the underlying carbonate
platform at the inner-bay site. The metal frames were
visually inconspicuous because they were covered by
a thin layer of sediment. Tether lengths of 0.7m pro-
vided a foraging area of about l.Sm^ and prevented
tangling between adjacent lobsters. Although tether-
ing does not necessarily measure absolute rates of
predation, it does measure relative rates of predation
(Heck and Thoman 1981), which can serve to compare
mortality rates as a function of different experimen-
tal treatments.
We used a stationary visual census technique (Bohn-
sack and Bannerot 1986) to quantify the community
structure of potential predators associated with casita
and no-casita stations during the experimental period
(January and August 1989). Visual censuses were per-
formed between 10:00 and 14:00 hours with three
replicate samples taken during the experimental peri-
od. By performing the visual censuses during midday,
we maximized the visibility available for species iden-
tification. Nighttime observations were not performed
because our previous study (Eggleston et al. 1990) in-
dicated that the predator guild normally associated
with the casitas dispersed widely over the seagrass bed
at night. However, predator movements were observed
during one dawn and dusk crepuscular period.
We examined the daytime foraging ranges of casita-
associated predators by swimming along a transect
perpendicular to each casita. When potential predators
were observed, a float was set to mark the location,
whereupon a scuba diver then followed the predators
to assure that they were associated with the casita. Our
initial observations indicated that piscine predators
associated with casitas seldom moved more than
30-40 m away from a casita.
Experimental design
Before initiating the tethering experiments in 1989, we
deployed casitas at the inner-bay and outer-bay sites
during 1988. During July 1988 at the inner-bay site,
we positioned a row of six large casitas 25 m apart from
one another (Fig. 3). Each large casita had one medium
and one small casita placed 10m away, yielding six
stations with one small, medium, and large casita
arranged in a triangle (Fig. 3). At the outer-bay site
during August 1988, we positioned six small, medium
and large casitas equidistant between the shore and
reef line and arranged these in two rows, each contain-
ing three triangular stations (Fig. 3b). See Eggleston
et al. (1990) and Eggleston and Lipcius (1992) for a
complete description of the small and medium casitas
and their use in other field experiments. Two separate
tethering experiments were then performed during
January and August 1989.
The first experiment was performed during January
1989. In this study we examined the survival of two
sizes of juvenile lobsters with and without access to
shelter at both the inner-bay and outer-bay sites. Six
metal-frame, no-casita stations were placed 60-70 m
away and perpendicular to the casitas in sparse-to-
moderate-density Thalassia at both sites (Fig. 3).
Juvenile lobsters were divided into two size-classes:
small, 46-55 mm carapace length (CL) as measured dor-
sally from the base of the supraorbital spines to the
posterior border of the cephalothorax; and large,
56-65 mm CL. Lobsters were tethered for 7 days. Each
casita and no-casita station at both sites had six
tethered lobsters from either of two size-classes for a
total of 144 tethered lobsters (2 sites x 6 lobsters x
2 sizes X 2 treatments (casita vs. no-casita) x 3
replicate stations).
Based on our initial observations of predator forag-
ing ranges (see above), we assumed that our choice of
60-70 m for the no-casita station was well beyond the
foraging range of diurnally active predators, thereby
providing unbiased estimates of lobster survival in the
absence of artificial shelters (i.e., mortality estimates
were not biased towards finding significantly higher
predation rates on lobsters tethered within the forag-
ing range of casita-associated predators). However, our
observations during the January 1989 experiment in-
dicated that some predators moved nearly 60 m from
the casitas (see Results). Thus, although the 70 m
distance from the large casitas was probably beyond
the foraging range of casita-associated predators, the
60 m distance from the small and medium casitas was
probably not.
Before initiating the second tethering experiment,
we positioned a row of three large casitas equidistant
between the shore and reef line in July 1989 at the
outer-bay site (Fig. 4). In August 1989 we examined
how lobster survival varied with distance from the
casitas. Three metal-frame no-casita stations were
placed 15, 30, and 70 m away and perpendicular to the
large casitas (Fig. 4). Based on the foraging ranges of
predators during the January experiment (see above),
we assumed that 70 m was an adequate distance to
696
Fishery Bulletin 90(4). 1992
X X X X Xj^ j^
X X X x'^x x'*x''x''x''x X X X X X
X X X X X X „_^_ XXX XXX
xxxxxxxxx REEF xxxxxxxxxxx
xxxxxxxxxx xxxxxxxxxxx
X X XXXxX XXXXXXX x X X X X XXX
15 m
15 m
NC
in^'^'-r 401
NC
15 m
NC
15 m
NC
40 m
iNC
SHORELINE
Figure 4
Schematic of casita layout at the outer-bay site for the August
1989 experiment. S = small casita, M = medium casita, L
= large casita, and NC = no-casita station.
assess predation in an environment uninfluenced by the
casitas. This assumption held true for the diurnal pred-
ator guild during August 1989 (see Results). Only juve-
nile lobsters approximating the small size-class {x
53.2mmCL, range 45.2-59.0mmCL, N 72, SD 4.1)
were tethered for 7 days. We chose only small lobsters
in the second experiment because logistical considera-
tions limited us to one size-class, and we wanted to
verify that survival of small lobsters was enhanced
when residing beneath casitas (See Results for first ex-
periment below). Each casita and no-casita (i.e., metal
frame) station contained 6 tethered lobsters for a total
of 72 tethered lobsters (6 lobsters x 4 distances (0, 15,
30, and 70 m) x 3 replicate stations).
Table 1
(a) Three-way ANOVA table (model I) describing the effects
of site (inner-bay sand-seagrass flat, and outer-bay seagrass
bed adjacent to coral reefs), lobster Pamdirus ar^us size (small
46-55 mm CL; large 56-65 mm CL) and shelter availability
(casita vs. no-casita station 60 m away) on proportional mor-
tality rates (arc-sine square-root transformed) of tethered
lobsters during January 1989. •P<0.05, **P<0.01. ns
P>0.05.
Source of variation
SS
df
MS
F
Site
0.002
0.002
0.402 ns
Lobster size
0.040
0.040
7.174"
Shelter availability
0.001
0.001
0.006 ns
Site X lobster size
0.001
0.001
0.112ns
Site X shelter availability
0.001
0.001
0.235ns
Lobster size x shelter
0.023
0.023
4.179*
availability
Site X lobster size x
0.001
1
0.001
0.187ns
shelter availability
Error
0.089
16
0.006
(b) Ryan's Q tests of mean proportional mortality rates (arc-
sine square-root transformed) of tethered lobsters for the in-
teraction effect of lobster size x shelter availability. Treat-
ment levels not significantly different at the 0.05 level share
an underline. Treatment levels are arranged in increasing
order of proportional mortality.
Interaction
Shelter availability
Casita
No Casita
Lobster size
Small
Large
Lobster size
large
small
large small
Shelter availability
Casita
No Casita
No Casita
Casita
Lobsters were checked and predation losses scored
every 1-2 days during experiments. Fewer than 4%
of tethered lobsters escaped, and these were not used
in subsequent statistical analyses. Lobsters that were
eaten or missing were not replaced. Cumulative losses
were converted to proportional mortality/day/casita (or
station). Proportions were analyzed as a function of
shelter availability (casita vs. no casita), distance from
the casita (0, 15, 30, and 70 m), lobster size (small vs.
large), and site (inner-bay vs. outer-bay) with two- and
three-way, fixed-factor analyses of variance (ANOVA)
models (after procedures in Underwood 1981). Propor-
tional mortality was arc-sine square-root transformed
to meet assumptions of normality and homogeneity of
variance (Underwood 1981). In all cases, the variances
were homogeneous as determined by Cochran's C-test.
Differences among means were revealed by use of
Eggleston et al : Artificial shelters and survival of juvenile Panuhrus argus
697
JANUARY-1989
BAY SITE
O
S
_i
<
z
o
.10
.08
.06
.04
.02
O.OO-
LOBSTER SIZE
C3 SMALL
ENLARGE
,! I
m^
SHELTER
NO SHELTER
JANUARY-1989
REEF SITE
I
.10
.08
.06
.04
.02
0.00
SHELTER NO SHELTER
SHELTER AVAILABILITY
Figure 5
Results of field tethering of Panulins argus at the
inner-bay and outer-bay sites during January 1989,
describing mortality as a function of lobster size (small
46-55mmCL; large 56-65mmCL) and shelter
availability (casita vs. no casita). Values are mean pro-
portional mortality ■ casita"' • d"' resulting from a
total of 18 lobsters tested. Vertical bars are ISE.
H
O
cr
O
Q.
O
cr
AUGUST-1989
REEF SITE
.10
.08
.06
.04
.02-
0.00
PTW
DISTANCE FROM SHELTER (m)
Ryan's Q multiple comparison test (Einot and Gabriel 1975)
as recommended by Day and Quinn (1989).
Results
Tethering experiments
During January 1989, mortality of juvenile lobsters varied
significantly as a function of lobster size but not according
to site or shelter availability (i.e., tethered to casitas or
60-70 m away in seagrass) (Table la. Fig. 5). However, the
interaction effect of lobster size by shelter availability was
significant; this interaction effect was due to the significantly
higher mortality of small vs. large lobsters tethered in
seagrass, and by the significantly higher mortality of large
lobsters in casitas compared with those tethered in seagrass
(Table lb).
At the outer-bay site in August 1989, mortality rates of
small juvenile lobsters varied significantly according to
distance from the casita (i.e., 0, 15, 30, and 70m away from
the casita) (Fig. 6; one-way ANOVA; F 5.89, df 3, P<0.02).
Lobsters suffered significantly higher mortality rates when
tethered 15 and 70 m away from casitas than when tethered
to casitas or 30m away from casitas (Fig. 6; Q Ryan's test,
experiment-wise error rate 0.05).
Predator observations
The visual census of potential lobster predators at the inner-
bay site during January 1989 indicated two predatory crab
species (stone crab Menippe mercenaria, and a portunid
Portunus spinimamcs) and two piscine predators (gray snap-
per Lutjanus griseus, and schoolmaster snapper L. apodus)
associated with the casitas (Table 2). No potential pred-
ators were observed in the vicinity of the no-casita stations.
Mixed schools of gray snapper and schoolmaster snapper
were typically found within 10 m of large casitas. Schools
associated with small and medium casitas were usually
located within 5 m of the casitas. Observed movements of
snappers were seldom more than 15-20m from the shelters.
Similarly, two snapper species predominated at the outer-
bay site during January 1989: mutton snapper L. analis and
yellowtail snapper Ocyurus chrysurus (Table 2). Casitas at
the outer-bay site also attracted octopus (Octopus spp.),
green moray eel Gymnothorax funebris, the stone crab
Figure 6
Results of field tethering of Panulirus argus at the
outer-bay site during August 1989, describing mortality
of small juvenile lobsters (46-5.5 mm CL) as a function
of distance from the casita (i.e., 0, 15, 30, and 70 m away
from the casita). Values are mean proportional mor-
tality ■ casita"' • d"' resulting from a total of 18
lobsters tested. Vertical bars are ISE.
698
Fishery Bulletin 90(4). 1992
Table 2
Summary of results from visual census of potential lobster Panulirus argus
predators associated with 18 casitas of three sizes
(small,
medium, large) at two sites (inner-bay. outer-bay) during
10-16 January 1989 at Bahia de la Ascension.
Mexico.
Results below are
pooled from censusing 18 casitas on three
different sampl
ng dates. Fish size is fork length (cm) and crab
size is carapace width (cm).
Mean
Frequency/
Size (cm)
Total
1 TiHnnH 1 1 nl c/
casita size
(N 18)
Percent
frequency
Species
abundance
1 llUl V ILlUdlor
sample/casita
Mean
Min.
Max.
Inner-bay site
Large Casita
Lutjanus griseus (gray snapper)
213
11.8
18
100.0
23.4
9
37
Lutjanus apodus (schoolmaster snapper)
27
1.50
6
33.3
10.0
8
11
Menippe mercenaria (stone crab)
3
0.17
3
16.7
11.0
11
U
Medium Casita
Lutjanus griseus
12
0.66
12
66.7
7.5
6
10
Menippe mercenaria
2
0.13
2
11.1
7.0
4
10
Portunis spinimanus (portunid crab)
2
0.11
2
11.1
8.0
8
8
Small Casita
Lutjanus griseus
30
1.67
12
66.7
7.5
6
10
Lutjanus apodiis
2
0.11
1
5.6
9.0
9
9
Menippe mercenaria
1
0.06
1
5.6
4.0
—
—
Portunis spinimanus
4
0.22
Outer-bay site
2
0.1
8.5
7
12
Large Casita
Lutjanus analis (mutton snapper)
15
0.83
6
33.3
25.3
20
40
Ocyurus chrysurus (yellowtail snapper)
24
1.80
6
33.3
22.0
19
25
Menippe mercenaria
3
0.17
3
16.7
10.0
10
10
Octopus
3
0.16
3
16.7
-
-
-
Medium Casita
Lutjanus analis
4
0.22
4
22.2
12.8
10
15
Portunis spinimanus
2
0.11
2
11.1
9.5
6
13
Small Casita
Lutjanus analis
3
0.16
2
11.1
8.0
8
8
Gymnothorax funebris (green moray eel)
3
0.16
3
16.7
50.0
50
50
Portunis spinimanus
3
0.16
3
16.7
7.5
6
11
M. mercenaria, and the portunid crab P. spinimanus
(Table 2). As above, no potential predators were ob-
served in the vicinity of the no-casita stations, and
mixed schools of snapper seldom strayed more than
15-20 m from casitas. However, several large snapper
of both species (L. griseus at the inner-bay site and
L. analis at the outer-bay site) were observed ~60 m
from the casitas. We also witnessed a stone crab feed-
ing on a lobster tethered beneath a casita, and on two
separate occasions observed octopus feeding on
tethered lobsters beneath a casita.
During January 1989 at the inner-bay site, there was
a significant positive correlation between mean lobster
proportional mortality per day at a particular casita
station and the mean number of potential predators
occupying the same casita station (r 0.92, n 6,
P<0.01; Fig. 7). Conversely, there was no significant
correlation between lobster proportional mortality and
numbers of predators inhabiting casitas at the outer-
bay site {r 0.11, n 6, NS), nor between proportional
mortality and the sizes of piscine predators (mm total
length; TL) at both sites (inner-bay: r 0.64, n 6, NS;
outer-bay: r 0.59, n 6, NS).
Predator observations at the outer-bay site in August
demonstrated a more diverse predator guild than that
observed during January (compare Tables 2 and 3).
Although mutton snapper and yellowtail snapper were
abundant at large casitas, they were joined by larger
predators, including Nassau grouper Epinephelus
striatus and a great barracuda Sphyraena barracuda.
One barracuda was identified by particular scars near
the mouth and a broken tooth. This barracuda roamed
the entire experimental area. We also observed one
Nassau gi'ouper that moved between the 70 m no-casita
stations and the reef (see Fig. 4 for geography).
Another slightly smaller grouper moved back and forth
between the casitas, the 15 m no-casita stations, and
the reef.
Eggleston et al : Artificial shelters and survival of juvenile Panulirus argus
699
Table 3
Summary of results from visual census of potential lobster Pc
nulirus argv.s predators associated with three large casitas at the outer- |
bay site during 3-10 August
1989 at Bahia de la Ascension,
Mexico. Results below
are pooled from
censusmg
three casitas
on three
different sampling dates.
Fish size is fork
length (cm).
Mean
Frequency/
Size (cm)
Total
abundance
individuals/
sample/casita
casita size
(N 18)
Percent
frequency
Species
Mean
Min.
Max.
Lutjanus analis
12
1.33
9
100.0
20.3
15
30
(mutton snapper)
Ocyurus chrysurus
15
1.67
9
100.0
22.0
19
25
(yellowtail snapper)
Sphyraena barracuda
2
0.22
2
22.2
100.0
100
100
(great barracuda)
Epinephelus striatus
2
0.22
2
22.2
45.0
40
50
(Nassau grouper)
Dasyatis avmricana
1
0.11
1
11.1
60.0*
—
—
(southern stingray)
to
wingtip (cm).
* Measured from wingtip
JANUARY-1989
m
a
BAY SITE
ortality /
p
•
S 0.06 J
•^^"^
"to
c
0
^'^'^^ •
- 0.04 -
o
Q.
O
a.
• ^-^
8 12 16 20 24 28
Number of Predators
Figure 7
Mean proportional mortality of juvenile lobsters Panulirus
argus ■ large casita 'd'' compared against the mean
number of potential predators ■ casita station'' ■ d"' at the
inner-bay site during January 1989 (y = -0.0051 -i-0.0029x;
r- 0.85, n 6, P<0.009).
Discussion
The impact of artificial shelters upon juvenile spiny
lobster survival varied both by lobster size and the
distance of unprotected lobsters from shelter. During
our January 1989 experiment, which emphasized the
effects of lobster size and shelter availability, large
lobsters (56-65 mm CL) survived better than small
lobsters (46-55 mm CL) in sparse-to-moderate-density
seagrass (Thalassia) 60m from casitas. Conversely,
small lobsters survived better than large lobsters when
tethered beneath casitas. During our August 1989 ex-
periment, small lobsters survived better at casitas or
30 m away from casitas than 15 m or 70 m away. We
interpret these patterns in terms of the relative impor-
tance of shelter availability and body size upon lobster
survival, and then speculate on the influence of
artificial-shelter-associated predators and seagrass den-
sity relative to these patterns in lobster survival.
We reemphasize that predation estimates based on
tethering are likely biased by the technique and may
not reflect natural predation rates. For example,
lobster dens which are normally abandoned at night
may become "traps" for tethered lobsters because they
cannot effectively flee and conspecifics are not available
to help detect and repel predators. However, preda-
tion rates on early juvenile Panulirus argus tethered
in open sand, seagrass, and algal habitats in Florida
Bay were similar both day and night (Herrnkind and
Butler 1986, Smith and Herrnkind 1992). Moreover,
most casita-associated predators are widely dispersed
among the seagrass flats at night in Bahia de la Ascen-
sion, Mexico (Eggleston et al. 1990). Thus, we feel that
the tethering technique is not only useful for compar-
ing relative rates of predation between different size-
classes of juvenile spiny lobster, but also for compar-
ing predation rates between representative benthic
habitats (e.g., crevices, algal clumps, seagrass).
Results from our January 1989 experiment support
the hypothesis that large juvenile lobsters (56-65 mm
CL) attain a relative-size refuge from predation com-
pared with small juvenile lobsters (45-55 mm CL), and
that the relative importance of this size refuge varies
according to shelter availability. Increased predation
700
Fishery Bulletin 90(4). 1992
on small juvenile lobsters tethered in seagrass suggests
that sparse-to-moderate-density Thalassia does not
provide adequate protection from predators, and that
the addition of shelter greatly enhances survival for
these smaller juvenile lobsters. Thus, the use of arti-
ficial lobster shelters in sparse-to-moderate-density
Thalassia beds may effectively reduce predation-
induced mortality rates of small juvenile lobsters and
thereby enhance production of this size-class. However,
given the general relationship of increasing survival
with habitat complexity for many decapod crustaceans
(Heck and Thoman 1981, Wilson et al. 1987, Heck and
Crowder 1991 and references therein), the relative im-
portance of shelter availability upon survival of small
juvenile lobsters may be reduced in habitats with dense
Thalassia. Thus, further studies are required to under-
stand the relationship between shelter availability and
increasing habitat complexity upon survival of small
juvenile lobsters.
The reduced survival of large juvenile lobsters near
casitas compared with seagrass 60-70 m away during
the January 1989 experiment is consistent with our
previous results for this lobster size-class. For ex-
ample, survival of small lobsters (46-55 mm CL) in large
casitas was significantly higher than survival of large
lobsters (56-65 mmCL) (Eggleston et al. 1990). More-
over, large lobsters survived better in medium than in
large casitas (Eggleston et al. 1990). Eggleston et al.
(1990) suggested that medium casitas excluded pred-
ators that were able to prey on large lobsters, and
postulated that larger predators associated with large
casitas may selectively prey upon larger lobsters, due
to better visual perception with increasing predator
and prey size (Kao et al. 1985, Ryer 1988). The signifi-
cant positive correlation between the numbers of
predators (primarily gray snapper L. griseus) occupy-
ing specific casita stations and predation rates at these
same stations suggests that gray snapper may be the
principle predator of juvenile lobsters inhabiting casitas
at the inner-bay nursery site. Gray snapper (15cmTL)
have successfully attacked small early-juvenile lob-
sters tethered in Florida Bay (Herrnkind and Butler
1986).
The combined results from this study and previous
work in Bahia de la Ascension, Mexico (Eggleston et
al. 1990), suggest that juvenile lobsters would survive
better by leaving large shelters to take up residence
in smaller shelters or nearby seagrass habitats when
they reach a body size of ~56-65mmCL. This idea of
enhancing survival through size-specific emigration
from large shelters was partially supported during our
recent observations of habitat-specific and size-specific
patterns of shelter use by juvenile P. argus in Bahia
de la Ascension, Mexico. Our recent field observations
(Eggleston and Lipcius 1992) indicated that shelter-
seeking behavior of P. argus is highly flexible to local
social conditions (i.e., presence of conspecifics) and
shelter scaling. For example, in a habitat containing
very few conspecifics (e.g., outer-bay site), large juve-
nile lobsters chose smaller, safer medium casitas over
large casitas as predicted by our tethering results (this
study; Eggleston et al. 1990). However, in a habitat
containing large numbers of conspecifics (e.g., inner-
bay site), large juvenile lobsters occupied large casitas
with large conspecifics (Eggleston and Lipcius 1992).
The tethering technique in this study did not address
the potential benefits of gregarious residency to lobster
survival. Gregarious occupancy by more than the six
tethered lobsters appeared to be inhibited because of
the tethering technique, i.e., lobsters did not colonize
casitas containing tethered individuals (pers. observ.).
Since gregarious sheltering has been implicated as a
mechanism for reducing predator-induced mortality
(Berrill 1975, Herrnkind et al. 1975, Eggleston and Lip-
cius 1992), final conclusions regarding the impact of
casitas upon predation-induced mortality rates of large
juvenile lobsters must not only consider the size-specific
relationship between shelter-associated predators and
lobsters, but also the potential benefits of gregarious
sheltering.
Results from our August 1989 experiment support
the hypothesis that the impact of artificial shelters upon
predation-induced mortality of juvenile lobsters varies
according to the distance of unprotected lobsters from
these shelters. During the August experiment at the
outer-bay site, small lobsters survived equally well
whether they were tethered beneath casitas or 30 m
away. These tethering results, combined with obser-
vations on predator movements, suggest that 30 m is
beyond the daytime foraging range of most casita-
associated predators. However, the lack of a signifi-
cant correlation between the numbers of potential
predators at a specific casita station and predation
rates on lobsters at these same stations at the outer-
bay site during the January 1989 experiment suggests
that transient predators such as jacks (Caranx spp.),
groupers (Epinephelus spp.), sharks (Ginglymostoma
cirratum. and Sphyrma spp.), and stingrays (Dasyatis
spp.) may be moving from the nearby barrier reef (see
Figs. 3 and 4 for geography) and preying on tethered
lobsters. Gut contents of stingrays (Dasyatis spp.) and
bonnethead sharks S;)/i,i/rwa tiburo, captured at night
in nearshore Florida Bay waters, contained a high pro-
portion of early-juvenile spiny lobsters (Smith and
Herrnkind 1992). Nurse sharks Ginglymostoma cir-
ratum are also known predators of juvenile P. argus
(Cruz and Brito 1986). Thus, our observations on the
daytime abundance and movements of casita-associated
predators (i.e., primarily mutton and yellowtail snap-
per, Lutjanus analis and 0. chrysurus) at the outer-
Eggleston et al : Artificial shelters and survival of juvenile Panulirus srgus
701
bay site may not reflect potential predation intensity
as previously suggested for the inner-bay site.
Predation risk on artifical reefs usually decreases
with distance from a natural, larger reef. For example,
mortality of tethered juvenile grunts (family Poma-
dasyidae) in St. Croix, U.S. Virgin Islands, was 40%
higher at the reef edge than 20 m away (Shulman 1985).
Our results are somewhat consistent with those of
Shulman (1985) in that predation of lobsters decreased
from 15 to 30 m from the casitas. However, increased
predation rates from 0 to 15 m and from 30 to 70 m in-
dicate that predation risk does not simply decrease
linearly with increasing distance from the artificial reef
(casita). We hypothesize that the predator guild orig-
inating from the nearby barrier reef at the outer-bay
site (see Figs. 3 and 4 for geography) forages within
the adjacent seagrass habitat and is attracted to the
casitas, thereby leaving a relative "gap" in predator
abimdance between 15 and 60 m from the casitas. Thus,
predator encounter rates with lobsters tethered only
15 m from casitas were probably high relative to
lobsters tethered 30 m away. The patterns of survival
of small P. argus within close proximity to casitas (i.e.,
15 m) in this study are consistent with our previous
work in seagrass habitats of Bahia de la Ascension,
Mexico. For example, survival of small lobsters (46-55
mmCL) was significantly higher at medium and large
casitas than in seagrass 15m away (Eggleston et al.
1990). Predation rates also increased from 30 to 70m,
and predators not associated with the casitas, such as
Nassau grouper E. striatus, were observed moving
from nearby natural reefs to the 70 m no-casita stations
rather than from the casitas.
Resident piscivores set the upper limit of the number
and sizes of prey species that can occupy a given reef
(Hixon and Beets 1989, Eggleston et al. 1990). For ex-
ample, Hixon and Beets (1989) found an inverse rela-
tionship between the number of piscivorous fishes on
a reef and the maximum number of co-occurring poten-
tial prey fishes. The results from our study indicate that
large casitas are more effective at reducing mortality
on small juvenile lobsters than seagrass habitats, even
though seagrass and algal beds provide some refuge
for juvenile spiny lobsters (Herrnkind and Butler 1986;
R.N. Lipcius et al., unpubl. data). Hence, for small
lobsters, our results from both the January and August
experiments strongly suggest that artificial lobster
shelters such as casitas increase lobster production by
enhancing survival in nursery areas. However, our
results for the outer-bay site during January indicated
that survival of large juvenile lobsters was significantly
lower when tethered beneath large casitas compared
with nearby seagrass habitats. These results are con-
sistent with the notion of building artificial lobster
shelters that are scaled according to body size to en-
hance survival of larger juveniles in nursery habitats,
particularly in areas where large conspecifics are
removed from large casitas by the fishery (Eggleston
et al. 1990, Eggleston and Lipcius 1992). However, fur-
ther research on the impact of casitas upon lobster sur-
vival, growth rates, local and regional population struc-
ture, and benthic community structure will be required
to assess the efficacy of this technology as a fisheries
enhancement tool.
Acknowledgments
We thank L. Coba-Cetina, T. Camarena-Luhrs, and E.
Sosa-Cordero with the Centro de Investigaciones de
Quintana Roo, and J. Cohen, J. Eggleston, K. Kennedy,
and numerous Earth watch volunteers for their able
field assistance. Special thanks to Armando Lopez and
Sonja Lillvick for the fine accommodations and
logistical support, and R. Wicklund and G. Wenz of the
Caribbean Marine Research Center for their ad-
ministrative guidance. We thank J. Bohnsack, L. Jones,
M. Luckenbach, J. van Montfrans, and two anonymous
referees for critical comments. This work was funded
by Sigma-Xi, Earthwatch-The Center for Field Re-
search, the National Undersea Research Program of
the National Oceanic and Atmospheric Administration,
the Caribbean Marine Research Center, the Common-
wealth of Virginia, L.L. Glucksman, and the National
Science Foundation (INT-8617945 to D.M. and R.L.
and OCE 87-00414 to R.L. and A. Hines).
Citations
Berrill, M.
1975 Gregarious behavior of juveniles of the spiny lobster,
Panulirus argus (Crustacea: Decapoda). Bull. Mar. Sci. 25:
515-522.
Bohnsack, J. A.
1989 Are high densities of fishes at artificial reefs the result
of habitat limitation or behavioral preference? Bull. Mar. Sci.
44:631-645.
Bohnsack, J. A., and S.P. Bannerol
1986 A stationary visual census technique for quantitatively
assessing community structure of coral reef fishes. NOAA
Tech. Rep. NMFS 41. 15 p.
Bohnsack, J. A., and D.L. Sutherland
1985 Artificial reef research: A review with recommendations
for future priorities. Bull. Mar. Sci. 37:11-39.
Cobb, J.S.
1981 Behaviour of the Western Australian spiny lobster.
Panulirus cygnus George, in the field and laboratory. Aust.
J. Mar. Freshwater Res. 31:399-409.
Cruz, R., and R. Brito
1986 Ecologia de la langosta (Panulirus argus) al SE de Isla
de la .Juventud. I. Colonizacion de arrecifesartificiales. Rev.
Invest. Mar. 7:3-17.
702
Fishery Bulletin 90(4), 1992
Day, R.W., and G.P. Quinn
1989 Comparisons of treatments after an analysis of variance
in ecology. Ecol. Monogr. 59:433-463.
Eggleston, D.B., and R.N. Lipcius
1992 Shelter selection by spiny lobster under variable preda-
tion risk, social conditions and shelter size. Ecology 73:
992-1011.
Eggleston, D.B., R.N. Lipcius, D.L. Miller, and L. Coba-Cetina
1990 Shelter scaling regulates survival of juvenile Caribbean
spiny lobster Pamdirus argus. Mar. Ecol. Prog. Ser. 62:
70-88.
Einot, I., and K.R. Gabriel
1975 A study of the powers of several methods of multiple com-
parisons. J. Am. Stat. Assoc. 70:574-583.
Grove, R.S., and C.J. Sonu
1985 Fishing reef planning in Japan. /« D'ltri, F. (ed.), Ar-
tificial reefs: Marine and freshwater applications, p. 187-251.
Lewis Publ., Chelsea, MI.
Hartnoll, R.G.
1982 Growth. In Abele, L.G. (ed.) The biology of Crustacea,
vol. 2, p. 111-196. Academic Press, NY.
Heck, K.L. Jr., and L.B. Crowder
1991 Habitat structure and predator-prey interactions in
vegetated aquatic systems. In Bell, S.S., E.D. McCoy, and
H.R. Mushinsky (eds.). Habitat structure: The physical ar-
rangement of objects in space, p. 281-299. Chapman & Hall,
NY.
Heck, K.L. Jr., and T.A. Thoman
1981 Experiments on predator-prey interactions in vegetated
aquatic habitats. J. Exp. Mar. Biol. Ecol. 53:125-134.
Herrnkind, W.F., and M.J. Butler
1986 Factors regulating postlarval settlement and juvenile
microhabitat use by spiny lobsters Panulirus argus. Mar.
Ecol. Prog. Ser. 34:23-30.
Herrnkind, W.F., J. Vanderwalker, and L. Barr
1975 Population dynamics, ecologj' and behavior of spiny
lobsters, Panulirus argus, of St. John, U.S. Virgin Islands:
Habitation and pattern of movements. Sci. Bull., Nat. Hist.
Mus., Los Angel. Cty. 20:31-34.
Hixon, M.A., and J. P. Beets
1989 Shelter characteristics and Caribbean fish assemblages:
Experiments with artificial reefs. Bull. Mar. Sci. 44:666-680.
Howard, R.K.
1988 Fish predators of the western rock lobster {Panulirus
cygnus George) in a nearshore nursery habitat. Aust. J. Mar.
Freshwater Res. 39:307-316.
Kao, T., T.K. Wetterer, and N.G. Hairston Jr.
1985 Fish size, visual resolution and prey selectivity. Ecology
66:1729-1735.
Lozano-Alvarez, E., P. Briones-Fourzan, and B.F. Phillips
1991 Fishery characteristics, growth, and movements of the
spiny \ohsX.er Panulirus argus in Bahia de la Ascension, Mexico.
Fish. Bull., U.S. 89:79-89.
Miller, D.L.
1989 Technology, territoriality and ecology: The evolution of
Mexico's Caribbean spiny lobster fishery. In Barkes, F. (ed.).
Common property resources: Ecology and community-based
sustainable development, p. 185-198. Valhaven Press,
London.
Mottet, M.G.
1985 Enhancement of the marine environment for fisheries
and aquaculture in Japan. In D'ltri, F. (ed.). Artificial reefs:
Marine and freshwater applications, p. 13-112. Lewis Publ.,
Chelsea. MI.
Munro, T.
1974 The biology, ecology, and bionomics of Caribbean reef
fishes. Part VI. Crustaceans (spiny lobsters and crabs). Univ.
West Indies Zool. Dep., Res. Rep. 3:1-57.
Reaka, M.L.
1987 Adult-juvenile interactions in benthic reef crustaceans.
Bull. Mar. Sci. 41:108-134.
Ryer, C.H.
1988 Pipefish foraging: effects of fish size, prey size and altered
habitat complexity. J. Exp. Mar. Biol. Ecol. 48:37-45.
Scully, E.P.
1983 The behavioral ecologj' of competition and resource
utilization among hermit crabs. In Rebach, S., and D.W.
Dunham (eds.). Studies in adaptation: The behavior of higher
Crustacea, p. 23-56. John Wiley, NY.
Shulman, M.J.
1985 Recruitment and coral reef fishes: Effects of distribu-
tion of predators and shelter. Ecology 66:1056-1066.
Smith, K.N., and W.F. Herrnkind
1992 Predation on early juvenile spiny lobsters Panulirus
ar(/Mf;(Latreille): Influence of size and shelter. J. E.xp. Mar.
Biol. Ecol. 157:3-18.
Underwood, A.
1981 Techniques of analysis of variance in experimental marine
biology and ecology. Oceanogr. Mar. Biol. Annu. Rev. 19:
513-605.
U.S. Agency for International Development
1987 Caribbean marine resources: Oppurtunities for economic
development and management. U.S. Agy. Int. Develop.,
Wash. DC, 53 p.
Vermeij, G.
1987 Evolution and escalation: An ecological history of life.
Princeton LIniv. Press, Princeton, 527 p.
Wilson, K.A., K.L. Heck Jr., and K.W. Able
1987 Juvenile blue crab, Callinectes sapidus. survival: An
evaluation of eelgrass, Zostera marina, as refuge. Fish. Bull.,
U.S. 85:53-58.
Zimmer-Faust, R.K., and E. Spanier
1987 Gregariousness and sociality in spiny lobsters: Implica-
tions for den habitation. J. Exp. Mar. Biol. Ecol. 105:57-71.
Abstract.— Restriction-fragment
length polymorphism analysis of mito-
chondrial DNA (mtDNA) was used to
investigate the genetic basis of stock
structiire of the bluefish Pomatomus
saltatrix along the U.S. mid-Atlantic
coast, and to determine the degree
of genetic differentiation between
mid- Atlantic bluefish and Australian
conspecifics. A total of 472 young-of-
the-year (YOY) and yearling bluefish
collected in New Jersey, Virginia,
and North Carolina over a period of
3 years, and 19 YOY bluefish col-
lected in New South Wales, Australia
were analyzed with 9 informative re-
striction endonucleases. Despite con-
siderable mtDNA variation within
samples of U.S. mid-Atlantic blue-
fish, no significant genetic differen-
tiation was detected among spring-
spawned and simimer-spawned (YOY)
bluefish, YOY and yearling bluefish
from different geographic locations
along the mid- Atlantic coast, or year-
ling bluefish collected at the same
location in different years. Mid-
Atlantic bluefish differed from their
Australian conspecifics by three or
more restriction site differences, or
a mean nucleotide sequence diver-
gence of 1.96%. In addition, Austra-
lian bluefish demonstrated greatly
reduced levels of mtDNA variation
relative to the mid- Atlantic samples.
The results of this study suggest that
bluefish along the mid- Atlantic coast
comprise a single genetic stock and
that significant differentiation oc-
curs among geographically disjunct
populations of this widely distributed
marine fish.
Stock structure of the bluefish
Pomatomus saltatrix along
the mid- Atlantic coast*
John E. Graves
Jan R. McDowell
Ana M. Beardsley
Daniel R. Scoles
Virginia Institute of Marine Science, School of Marine Science
College of William and Mary, Gloucester Point, Virginia 23062
The bluefish Pomatomus saltatrix is
broadly distributed in temperate and
warm-temperate coastal waters of
the world's oceans (Briggs 1960),
although it is absent from the eastern
Pacific (Smith 1949). In the United
States, bluefish occur along the At-
lantic and Gulf coasts, supporting
large recreational and commercial
fisheries.
The movements and biology of the
bluefish, like many fishes along the
Atlantic coast, are closely tied to
large seasonal fluctuations in water
temperature (reviewed in Wilk 1977).
Spawning appears to be concentrated
in two spatially and temporally dis-
tinct events: a spring spawn at the
inside edge of the Gulf Stream in the
south Atlantic bight, and a summer
spawn in the shelf waters of the mid-
Atlantic bight (Kendall and Walford
1979). However, the presence of eggs
and larvae indicates that some spawn-
ing occurs throughout the year, espe-
cially in the southern portion of the
south Atlantic bight (Kendall and
Walford 1979, Collins and Stender
1988). Presumably, eggs and larvae
are transported by cross-shelf cur-
rents to estuaries along the Atlantic
coast which serve as nursery grounds
for the young bluefish.
The discrete temporal nature of the
two spawning events is evidenced by
a bimodal size distribution of juvenile
Manuscript accepted 29 July 1992.
Fishery Bulletin, U.S. 90:703-710 (1992).
•Contribution 1750 of the Virginia Institute
of Marine Science.
bluefish within the estuaries during
the middle and late summer (Nyman
and Conover 1988, McBride 1989), a
difference that is still evident in year-
ling fish and may persist until fish
reach 4 years of age (Lassiter 1962).
The extent to which each of the
major spawning events contributes
juveniles to specific areas appears to
vary annually (Chiarella and Conover
1990).
A general mixing of bluefish from
different coastal areas may occur at
the end of the first summer. Tagging
studies indicate that as water tem-
peratures cool, young bluefish move
out of the estuaries in a southerly
direction and probably overwinter in
the south Atlantic bight (Lund and
Maltezos 1970, Wilk 1977), while
adults move further offshore (Wilk
1977). As temperatures along the
mid-Atlantic coast warm in the
spring, there is a general movement
of bluefish up the Atlantic coast, with
larger bluefish making more exten-
sive migrations into northern waters
(Wilk 1977).
Although the seasonal movements
of bluefish may be conducive to a
mixing offish from different coastal
areas, mark and recapture studies
suggest that a large fraction of blue-
fish are recaptured in the same
general area in which they were
tagged (Lund and Maltezos 1970,
Wilk 1977). The degree to which this
fidelity affects stock structure is not
known.
703
704
Fishery Bulletin 90(4). 1992
Table 1
Sample size
, date,
location,
and age of bluefish Pomatomus saltatrix col- |
lected and
analyzed in this
study. YRL = yearling;
YOY = young-of- |
the-year.
Sample
n
Date
Location
Age
VA88
100
7/88
York River VA
YRL
VA89
102
7/89
York River VA
YRL
VA90
39
7/90
York River VA
YRL
NC88
83
7/88
Hatteras NC
YRL
NC89
57
7/89
Hatteras NC
YRL
NC90
40
7/90
Hatteras NC
YOY
NJ90-Sp
26
8/90
southern NJ
YOY
NJ90-SU
25
8/90
southern NJ
YOY
AU91
19
2/91
Port Stephens, N.S.W.,
Australia
YOY
The genetic basis of population structure of the bluefish is poorly
understood. Based on studies of morphological and scale char-
acteristics, Wilk (1977) suggested that two populations exist along
the mid-Atlantic coast. These populations correspond to the fish
which spawn off North Carolina in the spring, and those that
spawn in the northern mid- Atlantic during the summer. Lund and
Maltezos (1970) also concluded on the basis of mark and recap-
ture analysis that several populations are present along the mid-
Atlantic coast. Chiarella and Conover (1990) used scales from
summer-spawning fish in the New York Bight to back-calculate
length at age-1 and found that most summer-spawning fish had
lengths corresponding to a spring birthdate, a result not consis-
tent with spring- and summer-spawning stocks. They concluded
that the morphological and life-history differences found between
spring- and summer-spawned bluefish are probably ecophenotypic
in nature, and suggested that a direct genetic analysis of stock
structure was warranted.
In this paper, we present the results of a restriction-fragment
length polymorphism (RFLP) analysis of bluefish mitochondrial
DNA (mtDNA) among bluefish collected along the mid-Atlantic
coast over a period of 3 years. We employed RFLP analysis of
mtDNA to evaluate genetic differentiation between spring- and
summer-spawned bluefish collected at a single location at the same
time, among similarly-sized bluefish collected at the same loca-
tion over several years, and among bluefish collected during the
same year from the north and south mid-Atlantic coast, as well
as from a disjunct population in Australia.
Materials and methods
Experimental design and collections
Bluefish were collected along the mid-Atlantic coast during
1988-90, and in Australia during 1991 (Table 1). To test the
hypothesis that spring- and summer-spawned bluefish represent
genetically distinct stocks, young-of-the-year bluefish were col-
lected by trawl on New Jersey state survey cruises during August
1990 (NJ90-Sp, NJ90-SU, Table 1). Fish were classified as spring-
10
8-
NEW JERSEY
SPRING-SPAWNED
i
SUMMER-SPAWNED
-1 — t^ — I — i-
NORTH CAROLINA
AUSTRALIA
r ~l ~
"50 100 150 200
STANDARD LENGTH (mm)
250
Figure 1
Frequency distribution of standard lengths
among YOY bluefish Pomatomus saltatrix col-
lected in New Jersey. North Carolina, and Port
Stephens. N.S.W-, Australia. The New Jersey
fish were separated into spring- and summer-
spawned groups liased upon their standard
length on the date of capture relative to a stan-
dard length of 125 mm (Nyman and Conover
1988. McBride 1989).
or summer-spawned based on the date of
capture using a standard length of 125 mm
used as the cut-off between the two groups
August (Nyman and Conover 1988,
in
McBride 1989). The distribution of lengths
is presented in Figure 1.
Graves et al.: Stock structure of Pomatomus saltatnx along the mid-Atlantic coast
705
To obtain an estimate of the
degree of temporal genetic varia-
tion between bluefish year-
classes at a single collection loca-
tion, 1 -year-old (yearling) blue-
fish were purchased from com-
mercial fishermen on the York
River, Virginia during July 1988
(VA88), 1989 (VA89), and 1990
(VA90), and in Hatteras, North
Carolina during 1988 (NC88) and
1989 (NC89). The distribution of
lengths of the Virginia and North
Carolina samples is presented in
Figure 2.
An analysis of geographic pop-
ulation structure of highly vagile
fishes, like the bluefish, is prob-
lematic. The presence of an adult
bluefish in one geographic loca-
tion is not very meaningful, as
the fish could easily travel to
another location several hundred
kilometers away within a few
weeks. If discrete geographic
stocks of bluefish exist, such
stocks might be expected to sep-
arate at the time of spawning.
However, collection of adults at
this critical time is difficult since
bluefish spawn at the edge of the
continental shelf during the
spring and in the middle of the
shelf during the summer (Kendall
and Walford 1979). Thus we de-
cided to focus our study on their
products, YOY bluefish. Although
some mixing probably occurs
during cross-shelf transport, the
genetic composition of YOY blue-
fish should reflect the composi-
tion of the offshore spawning
population.
To determine genetic differentiation among bluefish
along the mid-Atlantic coast, samples of YOY individ-
uals were collected during summer 1990 in New Jersey
(described above) and purchased from commercial
fishermen in Hatteras, North Carolina (NC90). In ad-
dition, to obtain an estimate of the degree of mtDNA
differentiation between isolated bluefish populations,
a sample of 19 YOY bluefish was collected by hook-
and-line in Port Stephens, N.S.W., Australia during
February 1991 (AU91). The size composition of all YOY
collections is presented in Figure 1.
VIRGINIA
NORTH CAROLINA
20-
|-]
"1 1989
15-
10^
r-
_,
5-
-
0-
^^
-|~]
STANDARD LENGTH (mm)
STANDARD LENGTH (mm)
Figure 2
Frequency distribution of standard lengths among yearling bluefish Pomatomus saltatrix
collected in (left) the York River. VA during summer 1988, 1989, and 1990, and (right)
Hatteras, NC during summer 1988 and 1989.
mtDIMA analysis
Depending on size and quality of the bluefish, three
different procedures were used to analyze bluefish
mtDNA. The rapid isolation procedure of Chapman and
Powers (1984) was used to obtain mtDNA from
samples of lateral red muscle from the yearling bluefish
collected in 1988 and 1989. After digestion, restriction
fragments were separated electrophoretically on
0.8-1.5% agarose gels run at 2 volts/cm overnight and
visualized directly with ethidium bromide staining. For
those samples in which there was not sufficient mtDNA
706
Fishery Bulletin 90(4). 1992
for direct visualization, restriction digestions were
endlabeled before electrophoresis with a mixture of all
four 3^S nucleotide triphosphates using the Klenow
fragment (Maniatis et al. 1982). After electrophoresis,
gels were treated with a scintillation enhancer, dried,
and autoradiographs exposed at -70°C for 5 days.
Mitochondrial DNA was purified from YOY and
yearling bluefish collected in 1990 and 1991 following
the protocols of Lansman et al. (1981) and ^^S-end-
labeled restriction fragments were visualized auto-
radiographically after electrophoresis. Due to the ther-
mal history of many of these specimens, yields of
supercoiled mtDNA were low. In those instances, the
nuclear band containing both nuclear DNA and relaxed
mtDNA was collected and dialyzed as described for
mtDNA bands in Lansman et al. (1981), or mtDNA was
reisolated following the Chapman and Powers (1984)
protocol. For these samples, the Southern transfer and
Table 2
Distribution of mtDNA genotypes among
bluefish Pomatomus saltatrix
samples. Each letter represents the fragment pattern
for a
particular restriction endonuclease: from left to right
Aval,
HindUl, P
;mII, Oral,
EcoRV,Sstl.Pstl
, SstU, and Neil. A description |
of all fragment patterns and
sizes is available from the authors upon r
equest.
Composite
genotype
AAAAAAAAA
VA88
44
VA89
VA90
NC88
NC89
NC90
NJ90-Sp
NJ90-SU
AU91
Total
45
24
50
33
17
20
18
0
250
AAAAAAAAB
0
0
0
1
1
1
0
0
0
3
AAAAAAAAC
1
2
0
1
1
0
0
2
0
7
AAAAAAAAD
6
1
0
0
0
0
0
0
0
7
AAAAAAAAG
0
1
1
0
1
0
0
0
0
3
AAAAAAAAH
0
0
1
0
0
0
0
0
0
1
AAAAABAAA
0
0
1
0
0
0
0
1
0
2
AAAABAAAA
11
11
1
5
3
5
1
1
0
38
AAAABAAAB
0
0
0
0
1
0
0
0
0
1
AAAABAABA
0
1
0
1
0
0
0
1
0
3
AAAACAAAA
6
6
3
2
3
2
2
0
0
24
AAAACAAAC
0
0
0
0
0
1
0
0
0
1
AAAACAABA
0
0
0
0
0
1
0
0
0
1
AAAADAAAA
7
13
3
4
2
4
1
1
0
35
AAABAAAAA
0
0
1
0
1
0
0
0
0
2
AAACAAAAA
1
1
0
1
0
0
0
0
0
3
AAAEEAAAD
0
0
0
0
0
0
0
0
18
18
AAAEFAAAD
0
0
0
0
0
0
0
0
1
1
AABAAAAAA
0
2
0
1
1
0
0
0
0
4
AABABAAAA
3
4
0
2
1
3
1
0
0
14
AABABAAAB
1
0
0
0
0
0
0
0
0
1
AABABAAAC
0
0
0
0
1
0
0
0
0
1
AABABAAAE
0
2
0
2
0
1
0
0
0
5
AACAAAAAA
1
0
0
2
1
0
0
0
0
4
AACACAAAA
2
2
0
0
0
0
0
0
0
4
BAAAAAAAA
6
2
0
4
3
2
0
1
0
18
BAAAAAAAC
0
1
0
0
0
0
0
0
0
1
BAAACAAAA
5
1
0
2
0
0
0
0
0
8
BAAACAAAD
1
0
0
0
0
0
0
0
0
1
BAAACBAAA
0
1
0
0
0
0
0
0
0
1
BAAADAAAA
0
0
0
1
1
0
0
0
0
>>
BADAAAAAA
1
0
0
0
0
0
0
0
0
1
BADACAAAA
0
0
0
1
0
0
0
0
0
1
CAAAAAAAA
1
1
0
0
1
0
0
0
0
3
CAAAAAAAC
3
1
1
0
0
2
0
0
0
8
CAAABAAAC
0
1
0
0
0
0
0
0
0
1
DAAAAAAAA
0
2
0
1
1
1
0
0
0
5
DAAACAAAA
0
1
0
0
0
0
0
0
0
1
DACAAAAAA
0
0
0
0
1
0
0
0
0
1
EAAAAAAAF
0
0
0
1
0
0
0
0
0
1
FAAAAAAAA
0
1
0
1
0
0
0
0
0
2
Totals
100
102
36
83
57
40
26
25
19
469
Graves et al: Stock structure of Pomatomus ssltatnx along the mid-Atlantic coast
707
hybridization protocols of Maniatis et al. (1982) were
followed after digestion and electrophoresis. Highly
purified bluefish mtDNA, nick translated with biotin-
7-dATP, was used as a probe for mtDNA fragments.
Hybridization filters were visualized after strigency
washes using the BRL BlueGene Nonradioadtive
Nucleic Acid Detection System.
All mtDNA samples were digested with the follow-
ing nine restriction endonucleases used according to
the manufacturers' instructions: Aval, Dral, EcoRV,
HindUl, Neil, Pstl, Pvull, Sstl, and Sstll. The dif-
ferent restriction-fragment patterns produced by each
restriction endonuclease were assigned a letter, and a
composite mtDNA genotype, consisting of nine letters
representing the fragment patterns generated by each
of the restriction endonucleases, was constructed for
each individual. The nucleon diversity (Nei 1987) was
calculated for each sample and for the pooled samples.
The nucleotide sequence divergence among mtDNA
genotypes was estimated by the site approach of Nei
and Li (1979). The mean nucleotide sequence diversity
within samples and mean nucleotide sequence diver-
gence between samples were calculated following the
method of Nei (1987), with the latter value being cor-
rected for within-group diversity (Nei 1987). The dis-
tribution of genotypes was evaluated for homogeneity
among collections using the G-test (Sokal and Rohlf
1981); however, as several of the genotypes were
represented by one individual, we employed the Roff
and Bentzen (1989) Monte Carlo approach to estimate
the significance of heterogeneity
X" values determined from the
raw data.
Results
The analysis of 472 mid- Atlantic
bluefish with 9 restriction endo-
nucleases revealed 40 mtDNA
genotypes, and 2 mtDNA geno-
types were encountered among
19 Australian bluefish. A total of
77 restriction fragments was vis-
ualized, and the average individ-
ual was scored for 34 fragments,
accounting for approximately
1.4% of the mtDNA genome.
The restriction endonucleases,
HindlU and Pstl, revealed no
variant fragment patterns, while
the remaining seven enzymes
revealed from two (Sstl and
Sstll) to eight (Neil) different
fragment patterns. Restriction-
site gains or losses were inferred
from completely additive changes in fragment patterns.
Considerable RFLP variation was detected within
Atlantic bluefish samples (Table 2). The most common
mtDNA genotype, AAAAAAAAA, ranged in frequen-
cy from 0.43 (NC 1990 YOY) to 0.75 (NJ 1990 YOY).
The large number of variant genotypes resulted in
nucleon diversities ranging from 0.416 to 0.798 (Table
3). Because many of the variant genotypes differed
from the common genotype by several site changes, the
within-sample mean nucleotide sequence diversities
were also relatively high, varying from 0.63% to 1.49%.
In contrast to the mid-Atlantic bluefish, the Australian
sample was quite depauperate of variation. Of the 19
fish in the sample, 18 shared a common mtDNA geno-
type (AAAEEAAAD), and one fish had a genotype dif-
fering from the common type by a single site change
(Table 2). The lack of variation in the Australian sample
was reflected in a low nucleon diversity (0.105) and a
within-sample mean nucleotide sequence diversity of
0.07%.
Significant genetic differentiation was not found
between the samples of spring- and summer-spawned
YOY bluefish collected in New Jersey during the sum-
mer of 1990. The corrected mean nucleotide sequence
divergence between the two samples was extremely
small (0.02%), indicating that average sequence diver-
gence between two individuals randomly drawn from
either the spring- or summer-spawned sample was the
same as the divergence between two individuals ran-
domly drawn from each group.
Table 3
Genetic variation within bluefish Pomatom us saltatrix samples expressed as nucleon diver- |
sity and mean nucleotide
sequence diversity. The spring- and summer-spawned NJ YOY |
bluefish collections were
pooled {NJ90 combined) for comparison
with the NC90 YOY
sample, and all NJ, VA, and NC bluefish collections
were pooled (m
d- Atlantic combined)
for comparison with the AU91 YOY sample. YRL =
= yearling; YOY
= young-of-the-year.
Nucleon
Mean nucleotide
Sample
Age
n
diversity
sequence diversity
VA88
YRL
100
0.781
1.34%
VA89
YRL
102
0.777
1.41%
VA90
YRL
36
0.565
0.89%
NC88
YRL
83
0.632
1.15%
NC89
YRL
57
0.663
1.20%
NJ90-Sp
YOY
26
0.416
0.72%
NJ90-SU
YOY
25
0.467
0.63%
NJ90 combined
YOY
51
0.438
0.67%
NC90
YOY
40
0.798
1.49%
mid-Atlantic combined
372
0.696
1.23%
AU91
YOY
19
0.105
0.07%
708
Fishery Bulletin 90(4), 1992
Considerable genetic differen-
tiation was not detected among
samples of yearling bluefish col-
lected at the same site in differ-
ent years. The mean nucleotide
sequence divergences (Table 4)
among the VA88, VA89, and
VA90 collections, and between
the NC88 and NC90 samples,
were of the same magnitude as
the within-sample mean nucleo-
tide sequence diversities (Table
3). Consequently, when adjusted
for within-sample diversity (Nei
1987), the corrected mean nu-
cleotide sequence divergences
among samples were nearly zero
(Table 4).
Analysis of YOY bluefish from
the northern and southern mid-Atlantic bight revealed
little mtDNA genetic differentiation. The corrected
mean nucleotide sequence divergence between the com-
bined NJ90 YOY sample and the NC90 YOY collection
was 0.11%, suggesting little population structuring
along the mid- Atlantic coast. This inference was fur-
ther supported by an analysis of heterogeneity which
demonstrated no significant differences in the distribu-
tion of six major mtDNA genotypes (those occurring
in 10 or more of the 472 fish) and the pooled rare
genotypes among the seven mid-Atlantic collections
(Gh=39.5, 0.25<P<0.50). Heterogeneity x" analysis
of the distribution of all genotypes, including those
represented by a single individual, was performed using
the Monte Carlo simulation of Roff and Bentzen (1989).
A total of 320 of the 1000 randomizations produced x"
values greater than the original data set, indicating no
significant heterogeneity.
The low levels of mtDNA differentiation among mid-
Atlantic bluefish collections contrasted with the sub-
stantial difference encountered between the combined
mid-Atlantic bluefish and the Australian sample. The
average mid-Atlantic bluefish could be distinguished
from its Australian conspecific by three or more restric-
tion-site changes. Two of the site changes were unique
to the Australian sample, and the third (A^dl pattern
D) occurred at a low frequency (0.01) in the combined
mid-Atlantic sample. The corrected mean nucleotide se-
quence divergence between the Australian sample and
the combined mid-Atlantic bluefish samples was 1.95%.
Significant heterogeneity was noted among the pooled
samples when the Australian sample was included with
the mid-Atlantic bluefish (Gh = 177, p<0.001).
A sample of 10 yearling bluefish was analyzed from
the northeast Gulf of Mexico (Panama City, FL). Unlike
the Australian bluefish, all of the mtDNA genotypes
Table 4
Mean nucleotide sequence divergences (%) among selected bluefish Pomatom-us salta-
trix collections. Values are presented with and without correction for within-sample
variation.
Collections
Uncorrected
Corrected
Among collections at a single location over 2 or
more years
VA88 vs. VA89
1.39
0.11
VA88 vs. VA90
1.20
0.18
VA89 vs. VA90
1.20
0.05
NC88 vs. NC89
1.18
0.01
Between spring- and summer-spawned bluefish
NJ90-Sp vs. NJ90-SU
0.69
0.02
Between mid-Atlantic YOY fish
NJ90-combined vs. NC90
1.19
0.11
Between mid-Atlantic and Australian bluefish
mid-Atlantic combined vs. AU91
2.60
1.96
found in the Gulf of Mexico mtDNA individuals were
also present in the mid-Atlantic samples, and 7 of the
10 Gulf of Mexico bluefish had the common mid-
Atlantic mtDNA genotype. Because of the small size
of the Gulf of Mexico sample, it was not appropriate
to test for frequency differences between bluefish from
the mid- Atlantic coast and the Gulf of Mexico.
Discussion
Mid-Atlantic bluefish demonstrated considerable mtDNA
genotypic variation. It is difficult to directly compare
the nucleon diversities calculated in this study with
those from other studies because the value is sensitive
to the number of restriction sites surveyed, and
analyses employing larger numbers of restriction endo-
nucleases typically have higher nucleon diversities. The
value of 0.696 for the pooled mid-Atlantic bluefish
samples is higher than those reported for many marine
fishes surveyed with a larger number of enzymes (A vise
et al. 1989, Gold and Richardson 1991), and indicates
a relatively high degree of genetic variation within the
bluefish. This trend becomes more apparent when
mean nucleotide sequence diversities, a measure of
intrasample diversity that is much less sensitive to the
number of restriction sites surveyed, are compared.
The value calculated in this study for the pooled mid-
Atlantic samples, 1.23%, is higher than values reported
for many other marine fishes (Ovenden 1990).
The Australian bluefish demonstrated much less
variation than their mid-Atlantic conspecifics. The
sample of 19 Australian bluefish had a nucleon diver-
sity five times lower than the combined Atlantic
samples, and a mean nucleotide sequence diversity that
was an order of magnitude lower (Table 3). A similar
difference in the level of mtDNA variation between
Graves et al,: Stock structure of Pomatomus saltatnx along the mid-Atlantic coast
709
conspecific populations has been noted between Atlan-
tic and Pacific blue marlin (Graves and McDowell,
unpubl. data). The striking lack of variation within the
Australian sample could be the result of a smaller
effective population size of females resulting from
population bottlenecks, or may simply reflect a period
of isolation sufficient for the sorting of gene trees (Nei
1987, Avise et al. 1988, Chapman 1990, Bowen and
Avise 1990).
We found little evidence to support the hypothesis
that genetically distinct stocks of bluefish exist along
the mid-Atlantic coast. Although appreciable mean
nucleotide sequence divergences were found between
sampling locations (Table 4), when corrected for within-
group variation the values became extremely small, in-
dicating that most of the observed differentiation could
be accounted for by variation within the samples. The
lack of population structuring was also supported by
the homogeneous distribution of all genotypes and the
fact that the level of genetic divergence among sam-
pling locations was not appreciably greater than the
level of divergence among samples taken at any one
location in different years.
The extent of gene flow among populations can also
be inferred from the frequency distribution of rare
alleles (Slatkin 1989). An inspection of Table 2 indicates
that almost all mtDNA genotypes that occurred more
than once were found in different collections, sug-
gesting significant gene flow among sampling loca-
tions. For example, the genotype AAAABAABA,
which was present in three individuals, occurred in the
VA89, NC88, and NJ90-Su collections. An exception
to this pattern was presented by the genotype
AAAAAAAAD, which occurred seven times: in six
individuals of the VA88 sample and one individual of
the VA89 sample. However, an examination of bluefish
mtDNA genotypes not included in this analysis— be-
cause the individuals were greater than one year old,
or because they came from a sample that was too small
for inclusion in this analysis— suggests that the ob-
served distribution of the AAAAAAAAD genotype
may be an artifact of sampling error. The genotype was
present in two bluefish collected in 1988 (one in New
York and one in Connecticut) and in six bluefish col-
lected in 1989 (two in New York, two in Virginia, and
two in North Carolina).
In contrast to the genetic similarity among mid-
Atlantic samples, a large, consistent genotypic dif-
ference was noted between the mid-Atlantic bluefish
and a conspecific population in Australia. The corrected
mean nucleotide sequence divergence of almost 2% is
more than an order of magnitude larger than the values
detected among mid-Atlantic samples, and is similar
to values reported between northwest Atlantic and
Barents Sea capelin populations (Dodson et al. 1991)
or among populations of freshwater fishes of different
river systems (Bermingham and Avise 1986).
While significant genetic differentiation was found
between mid- Atlantic and Australian bluefish, no
major differences were detected between mid- Atlantic
bluefish and a small sample from the Gulf of Mexico.
Consistent restriction-site differences have been
reported between Gulf of Mexico and mid-Atlantic
populations of a number of marine organisms, including
horshoe crabs Limulus polyphemus (Saunders et al.
1986), oysters Crassostrea virginica (Reeb and Avise
1990), and black sea bass Centropristis striata (Bowen
and Avise 1990). These preliminary results suggest that
bluefish from the Gulf of Mexico and the mid- Atlantic
are not as genetically isolated as many other coastal
marine species, although much larger samples will have
to be surveyed to determine if significant mtDNA
genotypic frequency differences exist between the two
areas. Considering the high vagility of bluefish and
their continuous distribution around Florida, this result
is not unexpected.
The lack of significant genetic differentiation be-
tween spring- and summer-spawned bluefish is consis-
tent with the results of Chiarella and Conover (1990),
who found no correlation between the season in which
an adult bluefish spawned and the hatch-date of an in-
dividual. These data suggest that the bimodal distribu-
tion of YOY bluefish in mid-Atlantic estuaries results
from two major spawning events of the same popula-
tion of bluefish, rather than the participation of dif-
ferent stocks. The morphological differences found
between spring- and summer-spawned bluefish are
probably ecophenotypic, resulting from early-life-
history development in appreciably different environ-
ments. Similar morphological plasticity has been dem-
onstrated in many other marine fishes (Barlow 1961).
The high degree of genetic homogeneity detected
within mid- Atlantic bluefish is also consistent wath the
results of tag and recapture studies. While many
bluefish return to the same site for several years (Lund
and Maltezos 1970), migratory habits appear to change
with age (Wilk 1977). Thus, the potential exists for con-
siderable interchange, and it is important to note that
even small levels of exchange can prevent the accumu-
lation of genetic differentiation (Hartl 1988).
The results of this study cannot disprove the null
hypothesis that bluefish along the mid- Atlantic coast
share a common gene pool. There appears to be suffi-
cient gene flow to prevent the accumulation of even
slight genetic differences. Determining the magnitude
of exchange between geographic regions would require
an extensive tag and recapture program. Until such
data are available, the resource should be managed as
assumed in the Fishery Management Plan for the
Bluefish— as a single, genetically homogeneous stock.
710
Fishery Bulletin 90(4). 1992
Acknowledgments
Bluefish were kindly provided by Hunt Howell, Alice
Webber, Raoul Castaneda, Bill Andrews, Katy West,
Debbie Fabel, Steve Battaglane, and R. Bill Talbot.
This project resulted from a study initiated by Herb
Austin and Brian Meehan of the Virginia Institue of
Marine Science. Robert Chapman provided helpful ad-
vice with the statistical analysis. Critical reviews of the
manuscript were provided by John Olney and John
Musick. Funding for this research was provided by the
U.S. Fish and Wildlife Service (F-60-R) and the Com-
monwealth of Virginia.
Citations
Avise. J.C, R.M. Ball, and J. Arnold
1988 Current versus historical population sizes in vertebrate
species with high gene flow: A comparison based on mitochon-
drial DNA lineages and inbreeding theory for neutral muta-
tions. Mol. Biol. Evol. 5:331-344.
Avise, J.C. B.W. Bowen, and T. Lamb
1989 DNA fingerprints from hypervariable mitochondrial
genotypes. Mol. Biol. Evol. 6:258-269.
Barlow, G.W.
1961 Causes and significance of morphological variation m
fishes. Syst. Zool. 10:105-117.
Bermingham, E., and J.C. Avise
1986 Molecular zoogeography of freshwater fishes in the
southeastern United States. Genetics 113:939-965.
Bovk-en, B.W., and J.C. Avise
1990 Genetic structure of Atlantic and Gulf of Mexico popula-
tions of sea bass, menhaden, and sturgeon: Influence of zoo-
geographic factors and life-history patterns. Mar. Biol. (Berl.)
107:371-381.
Briggs, J.C.
1960 Fishes of world-wide (circumtropical) distribution.
Copeia 1960:171-180.
Chapman, R.W.
1990 Mitochondrial DN A analysis of striped bass populations
in Chesapeake Bay. Copeia 1990:355-366.
Chapman, R.W., and D.A. Powers
1984 A method for the rapid isolation of mitochondrial DNA
from fishes. Tech. Rep. UM-SG-TS-84-01, Md. Sea Grant
Prog., Univ. Md., College Park, 11 p.
Chiarela, L.A., and D.O. Conover
1990 Spawning season and first-year growth of adult bluefish
from the New York bight. Trans. Am. Fish. Soc. 119:45.5-462.
Collins, M.R., and B.W. Stender
1988 Larval king mackerel (Scomberomonis caralla), Spanish
mackerel (S. maculatus). and bluefish (Pomatomus saltatrix)
off the southeast coast of the United States, 1973-1980. Bull.
Mar. Sci. 41:822-834.
Dodson, J.J., J.E. Carscadden, L. Bernatchez, and F. Colombani
1991 Relationship between spawning mode and phylogeo-
graphic structure in mitochondrial DNA of north Atlantic
capelin Mnllotus mllimi.'f. Mar. Ecol. Prog. Ser. 76:1 103-1 13.
Gold, J.R., and L.R. Richardson
1991 Genetic studies in marine fishes. IV. An analysis of
population structure in the red drum (Sciaemrps ocellatus) using
mitochondrial DNA. Fish. Res. 12:213-241.
Hartl, D.L.
1988 A primer of population genetics. Sinauer Assoc, Sund-
erland. MA, 305 p.
Kendall. A.W. Jr., and L.A. Walford
1979 Sources and distribution of bluefish, Pomatcmius saltatrix,
larvae and juveniles off the east coast of the United States.
Fish. Bull.. U.S. 77:213-227.
Lansman. R.A., R.O. Shade. C.F. Shapira. and J.C. Avise
1981 The use of restriction endonucleases to measure mito-
chondrial DNA sequence relatedness in natural populations.
III. Techniques and potential applications. J. Mol. Evol.
17:214-226.
Lassiter, R.R.
1962 Life history aspects of the bluefish fish, Pomatomus
saltatm Linnaeus, from the coast of North Carolina. Master's
thesis, N.C. State College. Raleigh, 103 p.
Lund, W.A. Jr., and G.C. Maltezos
1970 Movements and migrations of the bluefish, Pomatomus
saltatrix. tagged in waters of New York and southern New
England. Trans. Am. Fish. Soc. 99:719-725.
Maniatis, T.. LF. Fritsch, and J. Sambrook
1982 Molecular cloning: A laboratory manual. Cold Spring
Harbor Lab., Cold Spring Harbor, NY, 545 p.
McBride, R.S.
1989 Comparative growth and abundance of spring versus
summer-spawned juvenile bluefish, Pomatomus saltatrix,
recruiting to New Youk bight. Master's thesis. State Univ.
New York, Stony Brook, 61 p.
Nei, M.
1987 Molecular evolutionary genetics. Columbia Univ. Press,
NY, 512 p.
Nei. M.. and W-H. Li
1979 Mathematical model for studying genetic variation in
terms of restriction endonucleases. Proc. Natl. Acad. Sci.
76:5269-5273.
Nyman, R.M., and D.O. Conover
1988 The relation between spawning season and the recruit-
ment of young-ofthe-year bluefish, Pomatomus saltatrix, to
New York. Fish. Bull., U.S. 86:237-250.
Ovenden, J.R.
1990 Mitochondrial DNA and marine stock assessment: A
review. Aust. J. Mar. Freshwater Res. 41:835-53.
Reeb, C.A., and J.C. Avise
1990 A genetic discontinuity in a continuously distributed
species: Mitochondrial DNA in the American oyster,
Crassostrea mrginica. Genetics 124:397-406.
Roff, D.A., and P. Bentzen
1989 The statistical analysis of mitochondrial DNA polymor-
phisms: x' and the problem of small samples. Mol. Biol.
Evol. 6:539-545.
Saunders. N.C. L.G. Kessler. and J.C. Avise
1986 Genetic variation and geographic differentiation in
mitochondrial DNA of the horshoe crab. Limulus polyphemus.
Genetics 112:613-627.
Slatkin. M.
1989 Gene flow and the geographic structure of natural popula-
tions. Science (Wash. DC) 236:787-792.
Smith. J.L.B.
1949 The sea fishes of southern Africa. Central News Agency,
Cape Town, 550 p.
Sokal, R.R., and F.J. Rohlf
1981 Biometry. W.H. Freeman. NY.
Wilk, S.J.
1977 Biological and fisheries data on bluefish, Pomatomus
saltatrix (Linnaeus). Tech. Ser. Rep. 11, Sandy Hook Lab.,
NMFS Northeast Fish. Sci. Cent., Highlands, NJ, 56 p.
Abstract. - Age, growth, and
mortality of larval Atlantic bumper
Chhroscombrus chrysurus were com-
pared between cruise samples col-
lected during August-September
1986 and September 1987 off the
Louisiana-Mississippi barrier islands.
Calcein-marked Atlantic bumper
otoliths (sagitta) were used for age
validation. The first growth incre-
ment formed on the sagitta approx-
imately 2 days after spawning, and
daily increments formed thereafter.
Length at hatching was estimated at
0.7-0.9 mm SL. Growth rates were
determined from sagitta and length-
frequency data. Highest growth
rates occurred in August 1986 (0.40
mm/day) and were associated with
highest mean temperature and zoo-
plankton standing stock estimates.
The length exponent for Atlantic
bumpers' dry weight-length relation-
ship was 3.25. Instantaneous daily
mortalities (M) ranged from 0.62 in
August 1986 to 0.17 in late Septem-
ber 1987.
Age validation, growtli,
and mortality of larval
Atlantic bumper (Carangidae:
Chloroscombrus chrysurus] in the
northern Gulf of Mexico
Deborah L. Leffler
Florida Marine Research Institute, Florida Department of Natural Resources
3 Jackson Street Fort Walton Beach. Florida 32548
Richard F. Shaw
Coastal Fisheries Institute. Center for Wetland Resources
Louisiana State University, Baton Rouge, Louisiana 70803-7503
Manuscript accepted 1 July 1992.
Fishery Bulletin, U.S. 90:711-719 (1992).
Atlantic bumper CMoroscombnis chry-
surus, a carangid, is an abundant
coastal pelagic fish that is widely
distributed in the western Atlantic
and Gulf of Mexico (Leak 1977). Ex-
ploratory fishing surveys indicate
that Atlantic bumper may be abun-
dant enough in the northern Gulf of
Mexico to harvest commercially (Juhl
1966, Bullis and Carpenter 1968,
Bullis and Thompson 1970, Klima
1971). Presently, Atlantic bumper
is mainly a commercial bycatch,
marketed primarily for petfood, with
little potential as a food fish in the
United States (Klima 1971, Leak
1977). It, however, may be an impor-
tant food source for many predatory
fish (Reintjes 1979).
Atlantic bumper spawn primarily
in nearshore coastal waters, espe-
cially off Louisiana and Mississippi
(Boschung 1957, Perret et al. 1971,
Christmas and Waller 1973, Ditty
1986, Shaw and Drullinger 1990),
and the larvae of this species were
most abundant during surveys off the
Louisiana-Mississippi (LA-MS) bar-
rier islands (Stuck and Perry 1982,
Leffler 1989). Larvae have been col-
lected from June to October, with
abundance peaks usually occurring in
July or August (Sabins 1973, Stuck
and Perry 1982, Williams 1983, Ditty
1986).
Very little early-life-history work
has been conducted on Atlantic
bumper (Shaw and Drullinger 1990).
Early-life-history information is
a critical component in estimating
future year-class strength (Gushing
1975, Leak and Houde 1987). For ex-
ample, slow larval growth rates in-
fluence mortality by extending the
duration of vulnerable larval stages
(Bannister et al. 1974, Houde 1987),
while a fast growth rate can possibly
increase interaction with predators
(Pepin 1991), thereby influencing re-
cruitment. Early-life-history data are
needed for Atlantic bumper to deter-
mine their ecological role and to
assist in the prudent development of
any directed fishery.
The abundance of Atlantic bumper
over a wide geographic range, their
perceived potential as a commercial
resource, and their probable eco-
logical importance as a forage fish,
provided the impetus for conducting
this larval age-and-growth study. The
goals of this ichthyoplankton study
were to (1) validate the periodicity of
growth Increments on larval and
juvenile Atlantic bumper otoliths, (2)
estimate Atlantic bumper length at
hatching, (3) estimate the age struc-
ture of the sample population, (4)
describe larval growth and mortality
rates, and (5) relate larval growth
71 I
712
Fishery Bulletin 90(4|, 1992
and mortality rates to environ-
mental parameters and food
availability.
Materials and methods
Sampling procedure
Atlantic bumper larvae were col-
lected during five cruises off the
Louisiana-Mississippi barrier
islands in the Gulf of Mexico
(Chandeleur Is., Ship I., and
Horn I.; 29°50'-30°15'N and
88°40'-89°00'W; Fig. 1). Three
cruises were completed in 1986
(5-7 Aug., 8-9 Sept., and 22-24
Sept.) and two in 1987 (8-10 and
24-26 Sept.). Adverse weather
conditions canceled the sched-
uled August 1987 cruise.
The sampling design consisted
of a 4x4 grid of stations (N 16)
randomly sampled on two con-
secutive nights, and a 3 x 3 grid
of stations (A'^ 9) randomly sam-
pled during daylight, starting 12
hours after the initiation of the
first nocturnal sampling. The
sampling grid had a fi.xed com-
pass orientation with respect to
three windowshade, subsurface
current drogues (five drogues were used in 1987) which
were released at the beginning of each cruise (Shaw
et al. 1988). The change to five drogues in 1987 allowed
for a more defined sampling grid. Surface-water
temperature and salinity, as well as water depth, were
recorded for each ichthyoplankton tow.
Three-minute surface tows were taken at ~1.0m/s
using a 60cm "bongo-type" plankton sampler fitted
with a flowmeter (General Oceanics model 2030). In
1986, samples were collected using a 202 jjm mesh net,
while in 1987 a 333^m mesh net was used. During the
two cruises in September 1987, the bongo sampler was
fitted with one 202 ^.(m mesh net and one 333 /urn mesh
net for comparisons of daytime collections. Atlantic
bumper collected using the two mesh sizes were placed
into 1 mm size-classes and tested for differences using
a Median test (a 0.05; SAS Inst. 1985). Ichthyoplank-
ton samples used for age determination were preserved
with 95% ethanol, stored in ice water, and later trans-
ferred to 70% ethanol in the lab.
Live larval and juvenile Atlantic bumper were col-
lected for an age-validation experiment and length-
weight measurement analysis by dipnetting the jelly-
V. .""'-"S^. ;i<. I.
Do,,pV!fiJjB^<Si
Dou^JJ^^
STUDY AREA
2^0-
Figure I
Location and dates of ichthyoplankton cruises during 1986-87 in which Atlantic bumper
Chloroscombrus chrysurus were collected for age determination. Shaded areas repre-
sent trajectory of the water mass followed by current drogues during 1986 cruises, and
diagonally-lined areas represent 1987 cruises.
fish Aurelia aurita with which the fish are often
associated (Reid 1954, Franks 1970). Fish were then
transferred to a cooler containing lOOppm calcein
(2,4-bis-[N,N'-di(carbomethyl)aminomethyl]fluorescein)
in 13 L of aerated ambient seawater to create a fluores-
cent mark in their otoliths using the method described
by Wilson et al. (1987). Fish were held between 6 and
12 h in the seawater-calcein solution and then trans-
ferred into a 127 L aquaria. Fish were held under a
12h/12h photoperiod in 23°C and 25ppt water and fed
ad lihidium on brine shrimp. Fish were sacrificed 2,
7, and 10 days after marking.
Lab analysis
Ichthyoplankton samples from the bongo net collections
were split once with a Folsom plankton splitter (Van
Guelpin et al. 1982). Chloroscombrus chrysurus larvae
were sorted, counted, and measured to the nearest
0.1mm standard length (SL). Preflexion larvae were
measured to the end of their notochord, otherwise
larvae were measured to the posterior tip of the
hypural plate. When more than 52 fish were present,
Leffler and Shaw Age, growth, and mortality of larval Chloroscombrus chrysurus
713
a random subsample of 50 fish were measured, as well
as the shortest and longest. Ethanol-related shrinkage
was assumed to be uniform for each fish collected and
preserved (3-min tow, alcohol preservation; see Radtke
1989).
Validation, age, and growth
Sagittal otoliths were removed from each Atlantic
bumper larvae using a dissection microscope equipped
with polarized light. The sagitta from nine postlarval
and juvenile Atlantic bumper (8. 3-25. 0mm SL) that
were immersed in the calcein-seawater solution were
prepared and viewed using the method described by
Wilson et al. (1987). Growth increments, following the
fluorescent mark, were counted at 400 x and verified
at 1000 X . The number of growth increments counted
from the calcein mark to the otolith edge were com-
pared with the number of days fish were held in cap-
tivity after marking.
Age estimation of larval Atlantic bumper was per-
formed using sagitta that were air-dried and mounted
in S/P Accu-mount 60 on a glass microscope slide. Most
larval otoliths were thin enough that only viewing
under a compound microscope was necessary to make
total increment counts and otolith radius measure-
ments. A few larger otoliths were ground with 600
WetorDry grit sandpaper and polished using 0.3/:/
Alumina 2 Alpha Micropolish until growth rings were
countable. The counting and measurement procedure
was enhanced by using a digital imaging system which
produced images on a video monitor at 400 x or 1000 x .
Independent increment counts were made twice by the
same person without knowledge of fish length or
previous otolith count. Only otoliths for which replicate
counts were identical were used in the analysis. Eleven
of the 170 otoliths prepared were discarded.
Separate linear growth equations of standard length
on increment number were developed for fishes col-
lected on the five cruises. These five equations were
compared using analysis of covariance (ANCOVA,
a 0.05; SAS Inst. 1985). Exponential and other non-
linear models (e.g., Laird-Gompertz) used to describe
larval growth were also tested (Campana and Neilson
1985). A General Linear Model ANOVA, followed by
a multiple comparisons test (Duncan, a 0.05; SAS Inst.
1985), were used to detect differences in surface-water
temperature between years, months, and cruises.
Zooplanl<ton biomass
Zooplankton displacement volumes (mL/m''^) were
determined (Yentsch and Hebard 1957) for each net
tow. A mean zooplankton standing stock value was
then calculated for each cruise and net mesh type. A
simple regression of zooplankton standing-stock values
(202 vs. 333^im mesh nets) was developed for both
September 1987 cruises. ANCOVA (a 0.05) was used
to test for differences between the two cruises. The
data from the two cruises were combined into one zoo-
plankton standing-stock regression to standardize the
values from the two mesh sizes.
Dry weight-length relationship
Larval and juvenile Atlantic bumper (N 120, 8.0-32.0
mmSL) collected by dipnetting for jellyfish, were
measured to the nearest O.lmmSL, oven dried for
6h at 62°C, and then weighed to the nearest 1.0 mg.
A log-log dry weight-length relationship was estab-
lished and described by the equation W = aL'^, where
W = logio dry weight (mg), and L = logio standard
length (mm). A 95% confidence interval placed around
the estimated slope (b) was used to test for differences
in the estimated length power term (b) and the classical
b estimate of 3.0 for adult fish (LeCren 1951) and 4.0
for larval fish (Power 1989).
Mortality
Atlantic bumper densities for each ImmSL category
were converted into mortality estimates following the
length-frequency method described by Essig and Cole
(1986). Sampling with respect to the windowshade
drogues allowed us to monitor larval densities from the
same mass of water for an entire collecting period. Only
nighttime collected larvae >2.0mm or <5.0mm were
utilized in our mortality estimates to minimize biases
from net avoidance by larger larvae or extrusion
through the mesh openings by the smallest larvae. The
descending limb of each age-frequency distribution cor-
responding to a length range of 2.0-5.0 mm SL was
described by the equation Dt=Doexp'"'^'>, where M =
the instantaneous daily mortality coefficient, Dt =
larval fish density at time t, Do = larval fish density in
the first fully recruited group (i.e., time = 0), and t =
time in days (Peebles and Tolley 1988). Mortality
estimates were tested for statistical differences be-
tween cruises and years using ANCOVA (a 0.05).
Results
Validation, age, and growth
Daily increment formation on Atlantic bumper sagitta
was validated using calcein. Each otolith from the nine
fish treated had distinct growth increments between
the green fluorescent calcein mark and the edge of the
otolith (Fig. 2). On each sagitta examined, the number
of increments counted after the calcein mark was
714
Fishery Bulletin 90(4). 1992
Figure 2
Photomicrograph (400 x ) of the transverse-sectioned
sagittal otolith of a 23.5 mm SL juvenile Atlantic
bumper Chlorosrombrus chrysui~us observed under
ultraviolet light. The lower light band displays the up-
take of calcein during the immersion process.
equivalent to the number of days the fish was held in
captivity. The slope (1.02) of a least-squares linear
regression (Fig. 3) was not significantly different from
1.0 (i-test, p>0.05), confirming daily increment forma-
tion in otoliths of larval and juvenile Atlantic bumper.
Larval Atlantic bumper have circular sagitta, with
a central core. Yolksac larvae (O.SmmSL, preserved
length) lacked increments. However, all other aged fish
between 1.0 and 5.0 mm (preserved length; N 158) had
countable increments (i.e., 1-11 increments or 3-13
days old; Fig. 4). Growth models were based only on
2-13 day-old fish.
Larval Atlantic bumper growth rates during the first
two weeks of life were best described using a linear
model. A separate growth curve was estimated for the
5-7 August 1986 data (Table 1). Growth curve com-
•o
10
o
o
y
*-
^
8
o
CO
7
(/)
6
♦^
c
5
-)
4
«
>.
T
CO
Q
2
1
Y = 0.25 + 1.02X, R =0.99
N = 9
01 23456789 10 11
Number of Increments
Figure 3
Regression of the number of otolith growth increments subse-
quent to the fluorescent calcein mark on the number of days
each fish was held in captivity before sacrificing. Numbers
associated with points represent overlapping values.
0)
?4
ni
>
T?
^
m
?0
_i
1-
0)
16
n
E
14
3
12
z
10
(Q
H
O
fi
1-
4
2
^ 1986 (N = 90)
[ZD 1987 (N = 70)
^
^
-S^ K^
10 11 12 13
Age (days)
Figure 4
Age distribution of Atlantic bumper Chloroscombrus chfysurux
larvae captured off the Louisiana-Mississippi barrier islands.
1986-87.
parisons for the two cruises in September 1986 (days
8-9 and 22-24) showed no significant differences
within month (intercept, p 0.44; slope, p 0.48). Similar-
ly, no significant difference was found between the two
September cruises in 1987 (days 8-10 and 24-26; in-
tercept, p 0.07; slope, p 0.42). Therefore, the paired
September data sets were combined into a single
regression for each year (1986 and 1987; Table 1).
Atlantic bumper length-frequencies displayed no sig-
nificant differences (p 0.93) between the two different
mesh sizes (202 vs. 333^m) during the 1987 daytime
Leffler and Shaw: Age, growth, and mortality of larval Chloroscombrus chrysurus
715
Table 1
Estimates of three linear growth equations used to describe the growth rate (mm/day) of larval Chloroscombrus ehrysui-us (0.8-4.8
mm) collected off the Louisiana-Mississippi barrier islands during 1986 and 1987, and the associated mean surface-water tempera-
tures (°C) including ranges. R- is the coefficient of determination for the respective models; L = standard length (mm); X = age
(days).
Sampling date
Number
fish Size
aged (range)
Equations R-
Growth
rate
(mm/day)
Mean surface-
temperature (°C)
(range)
5-7 August 1986
8-9, 22-24 Sept. 1986
8-10, 24-26 Sept. 1987
9 0.8-3.7
81 1.2-4.8
69 1.3-4.5
L = 0.40X-0.13 0.94
L = 0.26X-^0.70 0.61
L = 0.31X-i-0.71 0.72
0.40
0.26
0.31
29.6 (29.0-30.8)
28.4 (28.0-29.0)
27.8 (26.5-30.0)
C Y = 1 20 -f 0 05X, R =0 77
E ' N = 160
c
c
10 20 30 40 50 60 70
Otolith Radius (microns)
Figure 5
Regression of larval standard length (mm) on sagittal
otolith radius (^i) for Atlantic bumper Chloi-oscombrus
chi-ysiu-us lai-vae collected off the Louisiana-Mississippi
barrier islands, 1986-87.
collections. Comparisons of the growth curves for
September 1986 and 1987 and August 1986 indicated
a significant difference in both the August intercept
(p<0.04) and slope (p<0.03) of the regressions. Even
though the sample size (A'^ 9) was small, the observed
growth rate for August (0.40 mm/day) was significantly
higher than for September (0.26 mm/day in 1986 and
0.31 mm/day in 1987). The higher August growth rate
occurred at a higher mean surface-water temperature,
29.6°C (Table 1). In September of 1987, the water
temperature range (26.5-30.0°C) was wider than the
other sampling periods due to a cold front passing
through before the late-September cruise. There was,
however, no significant difference (/:» 0.11) in temper-
ature between months because of the low number of
cruises.
Atlantic bumper standard lengths were regressed on
the otolith radius (measured in microns; Fig. 5). The
coefficient of determination (r^) for the relationship
Table 2
Zooplankton standing-stock estimates (mL/nr'±SE) with
1987 values converted to equivalent 202 fim mesh net values,
based on the conversion study done during both September
cruises in 1987. The number of samples taken during each
cioiise is indicated in parentheses. The following equation was
used in the conversion: Y = 0.785X - 0.054 {R- 0.86).
1986
1987
Month
(202 ^m)
Converted
(333 Mm)
August
early Sept.
late Sept,
0.83±0.17
(34)
0.61±0.13
(25)
0.32 ±0.05
(41)
0.57 ±0.04
(40)
0.39 ±0.03
(40)
0.39 ±0.04
(40)
0.25 ±0.03
(40)
was 0.77 and the equation is L = 120 -i- 0.05R, where
L = standard larval length (mm) at the otolith radius,
R (Fig. 5). The relationship between the age and otolith
radius explained less variability (r^ 0.68) and fit the
following equation A = 2.96 -i- 0.13R, where A = age
in days at the otolith radius, R. Otolith radius was
observed to increase with both larval length and age.
Zooplankton blomass
Zooplankton mean biomass values for 1986 were sim-
ilar to the converted 1987 values. The 202 vs. 333/um
mesh regression equation Y = -0.054 -i- 0.785X
{r~ 0.86), where Y = the 333 f^m zooplankton stand-
ing-stock value, and X = the 202 ^^m zooplankton value,
was used to establish a correction factor to convert the
1987 zooplankton values into estimates comparable to
the 1986 values. The highest mean zooplankton bio-
mass estimate (0.83 mL/m^) was found in August 1986
(Table 2). The mean standing-stock estimates declined
throughout the September cruises within each year
(Table 2).
716
Fishery Bulletin 90(4). 1992
Dry weight-length relationship
The dry weight-length relationship for postlarval and
juvenile Atlantic bumper (Fig. 6) is described by the
exponential model Weight = 0.0016 Length^'-^s (r^
0.94), where weight = dry weight of the fish (mg) and
length = standard length (mm). The dry weight-length
power term for larval and juvenile Atlantic bumper,
3.25, is significantly different from the classical stan-
dard length-weight power term of 3.0 for adult fish
(LeCren 1951) and 4.0 for larval fish (Power 1989) at
the 95% confidence level (p>0.05).
Mortality
Instantaneous daily mortality (M) for larval Atlantic
bumper was significantly higher during August 1986
(F 13.8, p 0.03) than in either September 1986 or
130-
120-
Weight = 0.001 6Length ^'^^
110
.— ^ 100-
O) 90-
E, ao-
r2=0.94 /°
N = 125 /
B y4
$ 60
^ 50-
o 9^
Q 40
j/"" Q
30-
20
0 bJ^^°'
10-
.^rr'^r-l^^
0 5 ID IS 20 25 30 35
Standard Length (mm)
Figure 6
Relationship between dry weight and standard length of At-
lantic bumper Chloroscombrus chrysurus collected off the
Louisiana-Mississippi barrier islands, 1986-87.
1987, with September values decreasing during suc-
cessive cruises each year (Table 3). As a whole, how-
ever, the M values for all cruises in 1986 and 1987 were
similar (F 0.74, p 0.45).
Discussion
The age of larval and juvenile Atlantic bumper was
estimated from counts of growth increments on sagittal
otoliths. One growth increment formed daily on each
sagitta of Atlantic bumper between 8 and 25mmSL.
We, like others (Pritcher 1988, Fowler 1989, Parsons
and Peters 1989), assumed that this relationship held
true for smaller larvae (1-8 mm). We validated the
periodicity of otolith growth increments and estab-
lished an otolith age-and-growth analysis for larval
Atlantic bumper in the northern Gulf of Mexico.
Growth increments were not visible in the otoliths
of yolksac Atlantic bumper larvae (0.8mmNL), but at
least one increment was visible in l.OmmSL larvae.
The length at hatching appears to be between 0.7 and
0.9mmSL (after preservation) based on the larval
length measurements. Atlantic bumper larvae, there-
fore, appear to begin otolith increment deposition after
yolksac absorption, approximately 2 days after spawn-
ing (allowing 1 day each for egg incubation and yolksac
absorption). Pelagic species, such as Atlantic bumper,
often begin growth increment formation on their
sagitta at the time of yolksac absorption (Radtke 1984).
An isometric or linear relationship between the size
of otolith radius and standard length was revealed for
Atlantic bumper larvae. The variation observed in our
otolith radius-fish size relationship could be influenced
by grow^th- and age-related factors. For example, under
unfavorable environmental conditions the fish may not
continue to experience an increase in otolith radius or
fish size, while daily increment formation may continue
(Lyczkowski-Shultz etal. 1988, Secor and Dean 1989).
Table 3
Estimates of instantaneous daily mortality of larval Chloroscombrus chrysu
ras (2.0-5.0 mm SL) off the Louisiana-Mississippi barrier 1
islands were calculated using the length-frequency method. Total larval Atlantic bumper densities and total larval fish densities were |
included for each cruise in 1986 and 1987.
R- is the coefficient of determination for the respective models.
Instantaneous daily
Atlantic bumper
Total larval
Number
mortality estimates
total densities
fish density
Sampling dates of fish
(M)
R^ (#fish/100m-*)
(# fish/ 100 m-')
5-7 August 1986 1912
0.62
0.82 608.9
1838.1
8-9 Sept. 1986 576
0.35
0.98 121.7
799.8
22-24 Sept. 1986 291
0.18
0.86 227.9
599.9
8-10 Sept. 1987 573
0.30
0.90 62.2
262.4
24-26 Sept. 1987 122
0.17
0.92 42.4
298.2
Leffler and Shaw: Age. growth, and mortality of larval Chloroscombrus chrysurus
717
Temperature (Laurence 1978, Laurence et al. 1981,
Houde 1989) and food availability (Methot and Kramer
1979, Laurence et al. 1981, Lyczkowski-Shultz et al.
1988, Warlen 1988) play important roles in larval
growth and survival. Atlantic bumper growth rates
were highest in August 1986, when mean surface-water
temperatures and zooplankton biomass estimates were
greatest.
The Atlantic bumper growth rate calculated over the
two cruises in September 1987 may have been higher
than the September 1986 growth rate because of the
increase in zooplankton availability (Tables 1 and 2).
Zooplankton displacement volume values calculated
from the samples taken in 1986 declined from August
to September. Relative zooplankton biomass values
have peaked, however, as late as October off the
Chandeleur Is. within Chandeleur Sound (102,000
animals/100 m^; Gillespie 1971; Fig. 1). Our zooplank-
ton standing-stock estimates were high compared with
values obtained from Mississippi River plume fronts
during July 1987 (0.04-0.43 mL/m^; R.F. Shaw, un-
publ. data).
Atlantic bumper larvae had a dry weight-length ex-
ponent value of 3.25 which is similar to that of 3.32
determined for larval northern anchovy Engraulis
mordax (Lasker et al. 1970). This power term, however,
is lower than values determined for seven laboratory-
reared, cold-water marine larval species (3.76-4.77;
Laurence 1979), or the hypothesized standard value for
developing larval fish (4.0; Power 1989).
The highest Atlantic bumper instantaneous daily
mortality estimate (M 0.62), observed during August,
was similar to that reported for estuarine larval spotted
seatrout Cynoscion nehulosus (0.64; Peebles and Tolley
1988) and, to some extent, another carangid, jack
mackerel Trachurus symmetricus (0.80; Hewitt et al.
1985). Mortality estimates, which declined throughout
September 1986 (0.35-0.18) and 1987 (0.30-0.17), were
similar and were within the reported range for several
larval marine species (Essig and Cole 1986, Houde 1987
and 1989, Pepin 1991). The highest daOy mortality rate
was associated with highest temperatures, highest
macrozooplankton displacement volumes, and highest
larval Atlantic bumper densities (Tables 1-3). In late
September 1986, however, there was a low mortality
rate during a time of relatively high Atlantic bumper
densities, lower zooplankton biomass estimates, and
lower temperatures. Two factors— larval size and lower
water temperatures— may have influenced this lower
mortality rate (Weinstein and Walters 1981). Mean
larval Atlantic bumper standard lengths (1.2mm) were
similar for all the cruises. Lower surface-water tem-
peratures, therefore, may have enhanced survival.
reducing the Atlantic bumper mortality estimate.
Larval growth (i.e., daily development) and mortality
rates have been reported to increase wath temperature
(Houde 1989, Pepin 1991). The high growth-rate and
mortality estimate observed in August 1986 is consis-
tent with these findings.
The high natural mortality observed in August is
probably related to predation, based on two existing
theories. Larval Atlantic bumper are usually aggre-
gated in patches (Leffler 1989) and, therefore, may
offer exceptional feeding opportunities to any pred-
ator that encounters them (McGurk 1987). Pepin (1991)
suggested that increased mortality rates were asso-
ciated vdth increasing growth rates, resulting from in-
creased encounters with predators. These higher
growth rates require a higher intake of food, causing
increased activity which leads to increased predator
encounters.
Another possible cause for the high August mor-
talities may be associated with competition for limited
food resources, i.e., density-dependent mortality (Gush-
ing 1974). Food availability as indexed by the zoo-
plankton biomass estimate was highest during August,
but the high total larval fish density may have rapidly
depleted the food source, causing elevated mortalities.
Larval Atlantic bumper density was high during the
August cruise (608.9 larvae/100 m^) as was the total
larval fish density (1838.1 larvae/100 m^; Leffler
1989).
This study provides preliminary information on the
early life history of larval Atlantic bumper. Further
studies need to be conducted on larval Atlantic bumper
to determine the relationship between these early-life-
history parameters and fluctuating temperatures and
food availability.
Acknowledgments
The authors would like to thank J. Ditty, D. Drullinger,
R. Raynie, and K. Edds for assistance in the field and
laboratory. Also we would like to acknowledge
L. Rouse, E. Turner, G. Wilson, R. McMichael, and two
anonymous reviewers for critical reviews of the manu-
script, and M. Mitchell and B. McLaughlin for their
assistance in preparing the graphics. Special thanks to
the captains and staff at Gulf Goast Research Lab-
oratory, Ocean Springs, MS, for use of their boats and
aquarium facilities.
Financial support was provided through the Loui-
siana Sea Grant Gollege Program, a part of National
Sea Grant Gollege Program maintained by NOAA U.S.
Department of Gommerce.
718
Fishery Bulletin 90|4), 1992
Citations
Bannister, C.A.. D. Harding, and S.J. Lockwood
1974 Larval mortality and subsequent year-class strength in
the plaice (Pleuroiiectes platessa L.). In Blaxter, J.H.S. (ed.).
The early life history of fish, p. 21-37. Springer- Verlag, NY.
Boschung, H.T.
1957 The fishes of Mobile Bay and the Gulf Coast of Ala-
bama. Ph.D. diss., Univ. Alabama, Tuscaloosa, 626 p.
Bullis. H.R. Jr., and J.S. Carpenter
1968 Latent fisheries resources of the central West Atlantic
region. In de Witt, G. (ed.). The future of the fishing industry
of the United States, p. 61-64. Univ. Wash. Pubi. Fish., New
Ser. 4. Seattle.
Bullis, H.R. Jr., and J.R. Thompson
1970 Bureau of commercial fisheries exploratory fishing gear
research base, Pascagoula, Mississippi- July 1, 1967 to June
30, 1969. U.S. Fish Wildl. Serv., Circ. 351, 29 p.
Campana, S.E.. and J.D. Neilson
1985 Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci.
42:1014-1032.
Christmas. J.Y., and R.S. Waller
1973 Estuarine vertebrates, Mississippi. In Christmas, J.Y.
(ed.). Cooperative Gulf of Mexico estuarine inventory and
study-Mississippi, p. 320-439. Gulf Coast Res. Lab., Ocean
Springs.
Gushing, D.H.
1974 The possible density-dependence of larval mortality and
adult mortality in fishes. In Blaxter, J.H. (ed.). The early life
history of fish, p. 103-112. Springer-Verlag, NY.
1975 Marine ecology and fisheries. Cambridge Univ. Press,
Cambridge, 278 p.
Ditty, J.G.
1986 Ichthyoplankton in nertic waters of the northern Gulf of
Mexico off Louisiana: Composition, relative abundance and
seasonality. Fish. Bull., U.S. 84:935-946.
Essig, R.J., and C.F. Cole
1986 Methods of estimating larval fish mortality from daily
increments in otoliths. Trans. Am. Fish. Soc. 115:34-40.
Fowler, A.J.
1989 Description, interpretation and use of the microstructure
of otoliths from juvenile butterfly fish (family Chaetodon-
tidae). Mar. Biol. (Berl.) 102:167-181.
Franks, J.S.
1970 An investigation of the fish population within the inland
waters of Horn Island, Mississippi, a barrier island in the north-
ern Gulf of Mexico. Gulf Res. Rep. 3:3-104.
Gillespie, M.C.
1971 Analysis and treatment of zooplankton of estuarine
waters of Louisiana. In Cooperative Gulf of Mexico estuarine
inventory and study, Louisiana Phase IV, Biology, p. 108-175.
La. Wildl. Fish. Comm.. New Orleans.
Hewitt. R.P.. G.H, Theilacker, and N.C.H. Lo
1985 Causesof mortality in young jack mackerel. Mar. Ecol.
Prog. Ser. 26:1-10.
Houde. E.D.
1987 Fish early life dynamics and recruitment variability. In
Hoyt, R.D. (ed.), 10th annual larval fish conference, p.
17-29. Am. Fish. Soc. Symp. 2, Bethesda.
1989 Comparative growth, mortality, and energetics of marine
fish larvae: Temperature and implied latitudinal effects. Fish.
Bull.. U.S. 87:471-495.
Juhl, R.
1966 Experimental trawling survey along the Florida west
coast. Commer. Fish. Rev. 28(6); 1-5.
Klima, E.F.
1971 Distribution of some coastal pelagic fishes in the Western
Atlantic. Commer. Fish. Rev. 33(6);21-34.
Lasker, R., H.M. Feder, G.H. Theilacker, and R.C. May
1970 Feeding, growth, and survival of EngrauHs mordax
larvae reared in the laboratory. Mar. Biol. (Beri.) 5:345-353.
Laurence, G,C.
1978 Comparative growth, respiration, and delayed feeding
abilities of larval cod (Gadus morhua) and haddock (Melano-
grammus aeglefinus) as influenced by temperature during
laboratory studies. Mar. Biol. (Beri.) 50:1-7.
1979 Larval length-weight relations for seven species of North-
west Atlantic fishes reared in the laboratory. Fish. Bull., U.S.
76:890-895.
Laurence, G.C.. A.S. Smigielski. T.A. Halavik, and B.R. Burns
1981 Implications of direct competition between larval cod
(Gadus morhua] and haddock (Melanogrammus aeglefinus) in
laboratory growth and survival studies at different food den-
sities. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 178:304-311.
Leak, J.C.
1977 Distribution and abundance of carangidae (Pisces, perci-
formes) larvae in the eastern Gulf of Mexico, 1971-1974.
M.S. thesis, Univ. Miami, Coral Gables, 83 p.
Leak, J.C, and E.D. Houde
1987 Cohort growth and survival of bay anchovy, Anchoa
mitchilli. larvae in Biscayne Bay, Florida. Mar. Ecol. Prog.
Ser. 37:109-122.
LeCren, E.D.
1951 The length-weight relationship and seasonal cycle in
gonad weight and condition in the perch (Percaflavescens). J.
Anim. Ecol. 20(2):201-219.
Leffler, D.L.
1989 Composition, abundance, and small-scale distribution of
ichthyoplankton off the Louisiana-Mississippi barrier islands,
with special emphasis on the age, growth, and mortality of
Chloroscombrus chrysuriis. M.S. thesis, Louisiana State
Univ., Baton Rouge, 151 p.
Lyczkowski-Shultz, J., J. P. Steen Jr., and B.H. Comyns
1988 Early life history of red drum (Seiaenops oceIlatu,'i) in the
nort:h central Gulf of Mexico. MASGP-88-013, Gulf Coast Res.
Lab.. Ocean Springs, MS, 147 p.
McGurk. M.D.
1987 Natural mortality and spatial patchiness: Reply to
Gulland. Mar. Ecol. Prog. Ser. 39:201-206.
Methot, R.D. Jr., and D, Kramer
1979 Growth of northern anchovy, Engraulis mordax, larvae
in the sea. Fish Bull., U.S. 77:413-423.
Parsons, G.R., and K.M. Peters
1989 Age determination in larval and juvenile sheepshead,
Archoifargus prnhatocephalus. Fish. Bull., U.S. 87:985-988.
Peebles, E.B., and S.G. Tolley
1988 Distribution, growth, and mortality of larval spotted
seatrout, Cyriosckm n(bulos>t.'> : A comparison between two ad-
jacent estuarine areas of southwest Florida. Bull. Mar. Sci.
42(3):397-410.
Pepin, P.
1991 Effect of temperature and size on development, mortality,
and survival rates of the pelagic early life history stages of
marine fish. Can. J. Fish. Aquat. Sci. 48:503-518.
Perret, W.S., W.R. Latapie, J.F. Polland, W.R. Mock,
G.B. Adkins, W,J. Gaidry, and C.J. White
1971 Fishes and invertebrates collected in trawl and seine
samples in Louisiana estuaries. Section 1. In Cooperative Gulf
of Mexico estuarine inventory and study, Louisiana, Phase IV,
Biology, p. 39-105. La. Wildl. Fish. Comm., New Orieans.
Leffler and Shaw: Age, growth, and mortality of larval Chloroscombrus chrysurus
719
Power. J.H.
1989 Sink or swim: Growth dynamics and zooplankton hydro-
mechanics. Am. Nat. 133(5):706-721.
Pritcher. C.R.
1988 Validation of a technique for reconstructing daily pat-
terns in the recruitment of coral reef damselfish. Coral Reefs
7:105-111.
Radtke, R.L.
1984 Formation and structural composition of larval striped
mullet otoliths. Trans. Am. Fish. Soc. 113:186-191.
1989 Larval fish age. growth, and body shrinkage: Informa-
tion available from otoliths. Can. J. Fish. Aquat. Sci. 46:
1884-1894.
Reid, G.K.
1954 An ecological study of the Gulf of Mexico fishes in the
vicinity of Cedar Key. Florida. Bull. Mar. Sci. Gulf Caribb.
4:1-94.
Reintjes, J.W.
1979 Coastal herrings and associated species: A profile of
species or groups of species, their biology, ecology, and cur-
rent exploitation with economic and social informa-
tion. Prepared for Gulf of Mexico Fish. Manage. Coune. by
NMFS Southeast Fish. Sci. Cent., Beaufort NC, 170 p.
Sabins, D.S.
1973 Diel studies of larval and juvenile fishes of the Caminada
Pass area, Louisiana. M.S. thesis, Louisiana State Univ.,
Baton Rouge, 163 p.
SAS Institute Inc.
1985 SAS User's Guide: Statistics, 1982 Edition. SAS Inst.
Inc., Gary NC, .584 p.
Secor, D.H., and J.M. Dean
1989 Somatic growth effects on the otolith— fish size relation-
ship in young pond-reared striped bass, Morone saxatilis. Can.
J. Fish. Aquat. Sci. 46:113-121.
Shaw. R.F., and D.L. Drullinger
1990 Early-life-history profiles, seasonal abundance, and
distribution of four species of carangid larvae off Louisiana
during 1982 and 1983, NOAA Tech. Rep. NMFS 89, 37 p.
Shaw. R.F.. D.L. Drullinger, K.A. Edds, and D.L. Leffler
1988 Fine-scale spatial distribution of red drum, Sciaeriops
oretlatiLS. larvae. Contrib. Mar. Sci. 30(suppl.):109-116.
Stuck. K.C.. and H.M. Perry
1982 Ichthyoplankton community structure in Mississippi
coastal waters. In Fishery monitoring and assessment com-
pletion report, 1 Jan. 1977 to 31 Dec. 1981, p. VI-I-1 thru
VI-I-53. Proj. 2-296-R, Gulf Coast Res. Lab., Ocean Springs,
MS.
Van Guelpin, L., D.F. Markle. and D.J. Duggan
1982 An evaluation of accuracy, precision, and speed of several
zooplankton sampling techniques. J. Cons. Perm. Int. Explor.
Mer 40:226-236.
Warlen. S.M.
1988 Age and growth of larval gulf menhaden, Brevoortia
patronus, in the northern Gulf of Mexico. Fish. Bull., U.S.
86:77-90.
Weinstein, M.P., and M.P. Walters
1981 Growth, survival and production in young-of-year popula-
tions of Leiostomiis xanthurus Lacepede residing in tidal
creeks. Estuaries 4(3):185-197.
Williams. L.W.
1983 Larval fish assemblages of lower Mobile Bay. M.S.
thesis, Univ. South. Alabama, Mobile, 55 p.
Wilson, C.A., D.W. Beckman. and J.M. Dean
1987 Calcein as a fluorescent marker of otoliths of larval and
juvenile fish. Trans. Am. Fish. Soc. 116:668-670.
Yentsch, C.S., and J.F. Hebard
1957 A gauge for determining plankton volume by the mer-
cury immersion method. J. Cons. Cons. Int. Explor. Mer
22(2):184-190.
Abstract.- Mortality due to re-
tention of lobsters in derelict traps
was evaluated over a 2-year period
using two approaches. First, a string
of eight empty, single-chamber, plas-
tic traps was deployed at 40 m depth
off Oahu, Hawaii, and monitored
periodically by scuba during a 6-
month period in 1990. The traps
were stable and remained intact
despite adverse oceanic conditions.
Numerous entries and exits of lob-
sters were recorded. For the second
test, the ability of lobsters to exit
traps was tested in a series of field
and laboratory trials of trap strings
stocked with Hawaiian spiny lobster
Panulirus marginatus and slipper
lobster Scyllarides squammosus.
The number of lobsters that died in
stocked traps was less than 4% of the
test population and differed signifi-
cantly from zero only in the labora-
tory evaluation (x" 5.42, P 0.02).
The two species exited similarly;
however, the pattern of exits in lab-
oratory and field tests differed sig-
nificantly (x- 23.889, P0.03). All
lobsters exited within 56 days in a
pattern generally approximating an
exponential decline. Our evidence
suggests that little direct mortality
of lobsters is due to the inability to
exit traps, and consequently ghost
fishing by these traps is not con-
sidered a problem for spiny and slip-
per lobsters.
Evaluation of ghost fishing
in the Hawaiian lobster fishery
Frank A. Parrish
Thomas K. Kazama
Honolulu Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service. NOAA
2570 Dole Street, Honolulu Hawaii 96822-2396
Manuscript accepted 14 August 1992.
Fishery Bulletin, U.S. 90:720-725 (1992).
Continued fishing by lost traps has
become the focus of increasing con-
cern by both fishery managers and
scientists. The recent trend for trap
fisheries to replace their degradable
traps with designs made from more
persistent synthetic materials has
heightened the seriousness of pos-
sible ghost fishing by such unrecov-
ered traps. Ghost fishing has been
defined as the continued fishing of
irretrievable gear (Smolowitz 1978).
Such a definition fails to distinguish
between permanent entrapment and
temporary occupation of a trap (e.g.,
for feeding or shelter). Mortality oc-
curring in abandoned traps should be
measured to assess the impact of
ghost fishing on a particular fishery.
The phenomenon of ghost fishing
has been observed in a wide range of
trap fisheries with diverse trap
designs (e.g.. High 1976, Pecci et al.
1978, Smolowitz 1978, Paul 1983).
Despite this attention, few studies
have effectively assessed the ghost
fishing problem for any species. The
more rigorous evaluations rely on
continued underwater field observa-
tions of simulated lost traps (Breen
1990). With this method, mortalities
have been clearly demonstrated in
temperate fisheries in which animals
were unable to exit traps fitted with
nonreturn entrance devices. Features
such as spring-loaded doors and slick
plastic entrance chutes effectively
reduce the ability of some animals to
exit actual and simulated lost traps,
resulting in reported mortalities of
26-55% (High 1976, Miller 1977,
Breen 1987). Conventional wooden,
two-chamber or "parlor-type" traps
designed for the American lobster
Homarus americanus have produced
mortalities of 12-25% (Sheldon and
Dow 1975, Pecci et al. 1978, Smolo-
witz 1978).
Ghost fishing poses a potential risk
to at least some trap fisheries, and
such a risk requires assessment for
each species and trap configuration.
Tropical lobsters have been largely
neglected in the controlled evaluation
of mortality by ghost traps. Isolated
anecdotal reports of tropical lobsters
found in lost traps (Sutherland et al.
1983) and tank studies made to date
(Paul 1983) do little to predict realis-
tic, long-term effects of lobsters in-
teracting with modern traps in the
field.
Hawaii's lobster fishery targets
two species, Hawaiian spiny lobster
Panulirus marginatus and slipper
lobster Scyllarides squaiyirnosus. A
laboratory study by Paul (1983) used
California parlor-type traps made of
wire to determine the effectiveness
of escape vents in releasing under-
sized Hawaiian spiny lobster. Paul
(1983) suggested that ghost fishing
might occur in these traps and rec-
ommended that the Hawaiian fishery
consider incorporating degradable
escape panels to facilitate the escape
of adult lobsters. By 1985 the fishery
had adopted a more cost-effective
molded-plastic trap design as the
standard gear. Degradable panels
have not been included.
In 1989 the Hawaiian lobster indus-
try reported that more than 1 million
traps were set in the Northwestern
720
Parrish and Kazama: Ghost fishing in the Hawaiian lobster fishery
721
Hawaiian Islands (NWHI); an es-
timated 2000 of these traps were
unrecovered (Landgraf et al.
1989). The annual accumulation
of lost plastic traps on the banks
where commercial trapping oc-
curs must be considered a poten-
tial hazard to the lobster stocks.
No field studies have been done
on the interactions between lost
traps and the adults of the two
target lobster species. The objec-
tives of this study were to (1)
evaluate the persistence of lost
commercial traps under field con-
ditions, (2) estimate retention of
target species in plastic traps
with bait depleted, and (3) assess
mortality of lobsters unable to
exit traps.
Methods
:^^^i£s*S
Figure 1
Plastic trap used in the Hawaiian commercial lobster fishery.
Study sites
The prohibitive cost of prolonged, ship-supported
diving operations in the NWHI dictated that all field
experiments be conducted at Oahu. Sites close to the
windward shore of Oahu provided appropriate depths
(30-40 m) and habitat consistent with the NWHI com-
mercial banks. The area is known to harbor exploitable
numbers of the lobster fishery's target species. Its
heavy seas, strong bottom surge, and swift currents
(Bathen 1978) might mimic NWHI conditions and thus
could test the stability of lost traps. Traps placed in
such rough conditions, without surface markers, were
unlikely to be disturbed by fishermen and other recrea-
tional users.
Trap stability and faunal interactions
A string of eight empty traps was deployed in a linear
orientation from February to July 1990. The selected
area afforded two types of adjacent habitat— high-relief
rugose bottom, and hard relatively-flat bottom— allow-
ing comparisons of trap stability and use by lobsters
in the different environments. Individual traps were
set 10m apart, four on high-relief bottom and four on
adjacent even bottom. The molded-plastic traps (Fig.
1) used in the study were a standard commercial model
employed throughout the commercial fishery in Hawaii
(Fathoms Plus Marine Implementation, P.O. Box 6307,
San Diego, CA 92106). Each trap consisted of a single
chamber with two side entrances and was weighted
with about 10 kg of lead, as is conventional in the
fishery. Traps remained in place over a 6-month period.
They were observed monthly by scuba divers during
three dives conducted at 48-hour intervals. Physical
condition, movement, and contents of the traps were
noted on each dive, along with general observations of
the surrounding area. The monthly censuses recorded
the initial presence of lobsters from the surrounding
study site and any exits or entries over the following
4 days. The area in and under the traps was examined
for exoskeleton remnants that might indicate molting
or mortality. One additional trap with its hinge pins
removed was deployed on flat bottom near the trap
string to mimic a trap with corrodible hinge pins that
had deteriorated.
Trap stocl<ing experiment
In the summers of 1990 and 1991, traps of the same
type (Fig. 1) were deployed in the field and laboratory,
and stocked with spiny and slipper lobsters from the
NWHI to evaluate their ability to exit and the extent
of mortality. Prior to the traps being stocked, lobsters
had spent 3-8 days in transit in continuously-flushing
shaded bait wells where they were fed daily. Mean
carapace lengths were 87.6 mm (A^ 96, range 67.4-
121.7mm) for spiny lobster, and 83.3mm (A'' 96, range
50. 1-99.7 mm) for slipper lobster. Antennae were tagg-
ed with color-coded, plastic, self-locking, electrical ties
to permit visual identification of individuals without
their being handled during the experiment. Molt stage
722
Fishery Bulletin 90(4), 1992
of lobsters prior to their deployment in traps was determined
using Lyle's (1982) adaptation of Drach's (1939) staging
technique.
In the field test, 128 tagged lobsters were placed by divers
in 4 strings of 8 unbaited traps each (2 spiny and 2 slipper
lobsters per trap). Two strings were placed on and around
high-relief substrate, one in summer 1990 and the other in
summer 1991; two strings were placed on relatively-flat
hardbottom, at least 300 m from any relief that could pro-
vide lobsters shelter, in summer 1991. The contents of the
traps were checked every 48 hours until all tagged lobsters
had exited or died.
In the laboratory trials, 64 additional tagged lobsters
were placed in 16 traps (2 spiny and 2 slipper lobsters per
trap) in a large, shaded, outdoor concrete tank. Through-
out the tank, food and other suitable shelter were provided
outside the traps to encourage exiting. Contents of the traps
were monitored daily, and any lobsters found outside the
trap in which they were originally stocked were removed
from the tank. Lobster death totals in the laboratory (where
predation could not occur) and in the field were compared
in an attempt to separate losses by predation from other
mortality (e.g., starvation, conspecific aggression) in the
field.
This study employed a modified experimental cohort de-
sign to examine the effects of multiple categories (replicate,
habitat type, species) on exiting by lobsters. The design
permits cross-classified categorical analysis to be applied
(Fienberg 1987). Using chi-square tests, comparisons were
made between replicates, habitat types, and species in a
systematic order. Categories were collapsed or pooled when
justified by the lack of significant differences (Siegel and
Castellan 1988).
Results
Trap stability and faunal interactions
Estimated seas of 4-6 ft and currents of 1-2 knots were com-
mon at the study site. They produced no observable effect
on the physical integrity of the plastic traps. Movement of
traps across the substrate was not detected, despite frequent
observations of the interconnecting groundline actively mov-
ing in the bottom surge. The two halves of the trap without
hinge pins shifted 2cm apart. Over the 6-month period, the
traps became encrusted with sessile organisms, including
bryozoans, corals, and fish eggs. Occasionally adult fish
larger than the opening of the escape vent were found in
the traps; however, most of these departed through the trap
entrance as a diver approached.
Adults of both spiny and slipper lobsters local to the sur-
rounding study site entered the traps. Of the 12 such occur-
rences of lobsters recorded during the 6-month survey, 7
lobsters left before the last inspection of the monthly obser-
vation period, indicating that they did not occupy the traps
for more than 30 consecutive days. Three lobsters
were observed entering and exiting within the
same 48-hour observation period. Nine of the 12
lobsters were found in traps on even bottom.
One dead spiny lobster comprised the only mor-
tality observed within the 6-month field evalua-
tion. Postmortem examination and the presence
of lobster debris in the area around the trap sug-
gested death by predation. Sightings of known
predators such as octopus, eels, jacks, and sharks
were routine. Large eels often occupied the traps,
occasionally with lobsters.
Trap stocl<ing experiment
Molt-stage evaluation indicated that 27% of the
spiny lobsters and 1% of the slipper lobsters were
in the premolt stage at deployment. Mortality was
limited to seven spiny lobsters, five in the labor-
atory and two in the field. All of these lobsters
were in premolt stage at deployment and died
during or shortly after molting. Despite the
limited mortality, the number of deaths in the
laboratory trials differed significantly from zero
c
_o
C
O
o
ID
O
0
-1
2
3 -
■4
■5 --
6
N = 126 a
H
1 1 1 1 1 1 1 i' I
I ' I ' 1 1
u -
••
N = 59
b
1 -
•
'"-V
2 -
- ^
• •"^^
J -
^-^
• •«
»••••
5
■ ' 1 1 i 1 i i
1 ' ' 1 1
i 1 1 1 I' ' 1 1 1 1 ' '
f-i-H
Days
Figure 2
Persistence of occupancy of plastic traps stocked with
spiny Panulirus marginatus, and slipper Syllarides
squamosiis, lobsters (connbined) in the (a) field and (b)
laboratory.
Parrish and Kazama Ghost fishing in the Hawaiian lobster fishery
723
Oc^ 5.42, P 0.02). This was not true for the field trials
(x^ 2.06, P 0.15), and mortalities in the laboratory and
field were significantly different (x" 4.74, P 0.03).
Consequently, the two test situations were evaluated
separately, with all animals that died excluded from the
trap-occupation analysis.
Contingency tables were used to test for differences
in the exit distributions for various groups of the data
(Fienberg 1987). Within each species, the distributions
of exits were first compared between replicate trap
strings within the same habitat type and were found
not to be significantly different (spiny lobster— high
relief, x" 3.22, P 0.50; even substrate, x" 10.00, P
0.19; slipper lobster-high relief, x" 3.33, P 0.50; even
substrate x" 9.52, P 0.22). Therefore, the distributions
of exits for the two replicates were pooled within each
habitat type. Within each species, exits were then com-
pared for the effect of the two habitat types and were
found not significantly different (spiny lobster— x"
10.81, P0.21; slipper lobster-x" 4.53, P 0.72). The
distributions of two habitat types were then pooled
within each species, and exits of the two species were
not significantly different (x^ 16.93, P 0.08). Conse-
quently, the distributions of the two species in all field
trials were combined (A'^ 126 lobsters after 2 early mor-
talities) for further comparisons.
Within each species, exits observed in the tank were
compared with the field data pooled by replicate and
habitat type and were not significantly different (spiny
lobster-x^ 14.42, P0.21; slipper lobster-x^ 13.63,
P 0.09). Exits of spiny and slipper lobsters in the tank
were not significantly different (x" 11.55, P 0.32), and
the data were subsequently pooled (A^ 59 lobsters after
5 early mortalities). A comparison of exits of all lobsters
in the tank (pooled) and all lobsters in the field (pooled)
showed a significant difference (x" 23.889, P 0.03).
Half of the 126 spiny and slipper lobsters stocked in
the field and 33% of the lobsters stocked in the labor-
atory exited within 48 hours after being placed in the
traps. Ninety percent or more of the exits in both tank
and field trials occurred within the first 16 days. All
field animals had left by day 30, and all laboratory
animals by day 56. The overall exit pattern of the
lobsters suggested an exponential model. The data
were fitted to the log-transformed exponential function
In, (Nt/N,) = bt,
where Nt is the number of lobsters remaining after
time t from Nq lobsters initially stocked. The param-
eter b was estimated with a log-linear regression pro-
cedure for the field traps (b= -0.16/day; r^ 0.992,
SE 0.284, P<0.001) and for the laboratory traps (b =
-0.094/day, r^ 0.961, SE 0.632, P<0.001; Fig. 2).
Stocked spiny and slipper lobsters exited and re-
entered field traps in at least 13 instances; 6 lobsters
returned to the same trap. One spiny lobster was
observed exiting three traps within 6 days.
Discussion
Trap stability
The lack of structural damage and appreciable move-
ment of the plastic traps in the field contrasts with the
popular opinion of experienced fishermen that lost
traps break up and roll off the banks into waters
beyond the depth range of lobsters. Fishermen routine-
ly report movement of trap strings as a result of power-
ful swells moving across the commercial banks. Despite
the frequently observed movement of the groundline
by swells at the Oahu study site, it is likely that the
study site does not fully duplicate the power of the long-
term swells common in the NWHI. Lost traps may not
shift on the bottom as much as actively fished traps.
Buoys and interconnecting polypropylene line provide
additional lift and resistance to water motion; there-
fore, fully rigged strings of traps are more likely to
move than isolated traps severed from the groundline.
A 1990 systematic diving survey of 33 sites around 2
of the prominent NWHI commercial lobster fishing
banks revealed only 2 mangled derelict traps (F. Par-
rish, unpubl. data). The failure to locate large amounts
of lost gear may be partly explained by this survey be-
ing incidental to other work. Total gear losses in 1989
averaged about 1 trap/nm- over the total estimated
area of the lobster fishing grounds (Landgraf et al.
1989). Lost gear could be heavily concentrated in a few
of the more intensively fished areas that the survey
may have missed.
A trap manufacturer (Fathoms Plus Marine Imple-
mentation) has made available a corrodible pin which
is intended to allow the halves of plastic traps to even-
tually fall apart once the pin deteriorates. The fact that
our pinless trap remained relatively intact for 6 months
in the field suggests that the synthetic plastic clips on
the trap roof will continue to hold the trap together,
especially for fisheries conducted in calmer seas.
Mortality and movement of lobsters
Seven deaths among the 192 spiny and slipper lobsters
within the 56-day study represent low mortality when
compared with the natural mortality estimates from
the fishery population modeling by Haight and Polovina
(1992). Extrapolation of the experiment's percentage
of mortality from 2 months to 1 year (22%) is close to
half the fishery's annual estimated natural mortality
(40%). The fact that only animals that began the trials
724
Fishery Bulletin 90(4), 1992
in premolt condition died suggests that lobsters at this
stage are less fit. The probability that this mortality
would occur only with premolt individuals by chance
alone is < 0.001 (Agresti 1990). Increased vulnerabil-
ity to a poor physical environment, conspecific aggres-
sion, or predation have been associated with molting
(Conan 1985). It seems likely that the higher percent-
age of spiny lobster in premolt stage accounts for some
of the difference in mortality between spiny and slip-
per lobsters.
The significantly higher mortality observed in the
laboratory compared with the field does not support
the idea that undetected predation substantially af-
fected the field results. The relatively low absolute level
of total mortality suggests that such predation is prob-
ably minimal at the field site. However, with mortal-
ity being higher in the laboratory than in the field, no
estimate of predation is possible.
More than twice as many deaths were observed in
the laboratory as in the field, even though the field
trials involved twice as many lobsters and three times
as many were in the premolt stage. This suggests a less
healthy or fit laboratory population, which is consis-
tent with the significantly slower exit of pooled species
from traps in the laboratory versus in the field. Aspects
of the laboratory environment (e.g., water quality,
lighting, diet) may have degraded the physical condi-
tion of the lobsters or affected their behavior, inhibiting
their exit, or providing less inducement to leave the
traps than that encountered in the field. It seems like-
ly that our field assessment provides a better estimate
of natural exit patterns.
With a study design similar to ours, Munro (1974)
examined the rate of fish exiting unbaited traps. His
theoretical model suggests that the number exiting per
day may be a fixed fraction of the current trap occu-
pancy, and that catch eventually reaches an asymptote
when trap entrances are balanced by exits. Our number
of stocked lobsters declined approximately exponential-
ly, approaching zero asymptotically. However, the total
occupancy of traps declined daily until it reached a low
and varying level at which exits were roughly matched
by entrances of lobsters. This final, low level of trap
occupancy at the end of the stocking experiment seems
consistent with native occupancy observed during the
monthly field monitoring.
In our stocking tests, some individuals likely left a
trap and reentered it undetected between censuses,
particularly in the field test where the observation in-
terval was 48 hours. Based on independent probabilities
of exit and entry estimated from our field data, the
theoretical joint probability of such reentry was as high
as 0.06, and probably about 12 individuals left and
reentered the same trap undetected during the full
26-day field stocking experiment.
Conclusion
Our results indicate that spiny and slipper lobsters are
not restrained by lost molded-plastic traps for periods
long enough to cause serious harm. There is no evi-
dence that such lost traps result in increased mortal-
ity. The absence of any apparent trap-induced mortality
and the low incidence of identifiable in-trap mortality
due to predation suggest that ghost fishing by these
traps contributes little to the total mortality of the
population. Such traps, when unbaited and intact, may
best be considered short-term artificial shelters that
lobsters enter and exit occasionally, more or less at will.
Acknowledgments
Thanks are due to Steve Kaiser for advice on selection
of study sites and to Bill and Joanne Goebert for pro-
viding ready access to those sites. Ray Boland, Karl
Bromwell, Theresa Martinelli, and Leslie Timme
assisted in the rigorous program of field monitoring.
Greatly appreciated are the statistical and substantive
comments of Deborah Goebert, Robert Moffitt, James
Parrish, and Jeffrey Polovina.
Citations
Agresti, A.
1990 Categorical data analysis. John Wiley, NY, 558 p.
Bathen, K.H.
1978 Circulation atlas for Oahu, Hawaii. Misc. Rep. UNIHI-
SEAGRANT-MR-78-05, Univ. Hawaii Sea Grant Coll. Prog.,
Honolulu, 22 p.
Breen, P.A.
1987 Mortality of Dungeness crabs caused by lost traps in the
Fraser River estuary, British Columbia. N. Am. J. Fish.
Manage. 7:429-435.
1990 A review of ghostfishing by trap and gill nets. In Sho-
mura, R.S., and M.L. Godfrey (eds.). Proceedings, Second
international conference on marine debris, 2-7 April 1989,
Honolulu, p. 571-599. NOAA Tech. Memo. NMFS-SWFSC-
154, NMFS Southwest Fish. Sci. Cent., Honolulu.
Conan, G.Y.
1985 The periodicity and phasing of molting. In Wenner,
A.M. (ed.), Crustacean issues, factors in adult growth, vol. 3,
p. 73-99. A. A. Balkema, Boston.
Drach, P.
1939 Mue et d'intermue ehez les Crustaces Decapodes. Ann.
Inst. Oceanogr. 19:103-392.
Fienberg S.E.
1987 The analysis of cross-classified categorical data, 2d
ed. MIT Press, Boston, 198 p.
Haight W.R., and J.J. Polovina
1992 Status of lobster stocks in the Northwestern Hawaiian
Islands, 1991. Admin. Rep. H-92-02, NMFS Southwest Fish.
Sci. Cent., Honolulu. 19 p.
High. W.L.
1976 Escape of Dungeness crabs from pots. Mar. Fish. Rev.
38(4): 19-23.
Parrish and Kazama Ghost fishing in the Hawaiian lobster fishery
725
Landgraf, K.C., S.G. Pooley, and R.P. Clarke
1989 Annual report of the 1989 western Pacific lobster fishery.
Admin. Rep. H-90-06, NMFS Southwest Fish. Sci. Cent., 30 p.
Lyle, W.G.
1982 Molt stage determination in the Hawaiian spiny lobster
Panulirus marginatus. M.S. thesis, Univ. Hawaii, Honolulu,
29 p.
Miller, R.J.
1977 Resource underutilization in a spider crab industry.
Fisheries (Bethesda) 2(3):9-13.
Munro, J.L.
1974 The mode of operation of Antillean fish traps and the rela-
tionships between ingress, escapement, catch and soak. J.
Cons. Cons. Int. Explor. Mer 35:337-350.
Paul, L.M.B.
1983 Investigations into escape vent effectiveness and ghost
fishing in captive populations of the spiny lobster Panulirus
marginatus. In Grigg, R.W., and K.Y. Tanoue (eds.). Pro-
ceedings. Second symposium on resource investigations in the
Northwestern Hawaiian Islands, vol. 2, p. 283-295. Misc. Rep.
UNIHI-SEAGRANT-MR-84-01, Univ. Hawaii Sea Grant Coll.
Prog., Honolulu.
Pecci, K.J.. R.A. Cooper. CD. Newell, R.A. Clifford, and
R.J. Smolowitz
1978 Ghost fishing of vented and unvented lobster, Homanis
americamis, traps. Mar. Fish. Rev. 40(5-6):9-24.
Sheldon, W.W.. and R.L. Dow
1975 Trap contributions to losses in the American lobster
fishery. Fish. Bull., U.S. 73:449-451.
Siegel, S., and N.J. Castellan
1988 Nonparametric statistics, 2d ed. McGraw-Hill, NY,
399 p.
Smolowitz. R.J.
1978 Trap design and ghost fishing: An overview. Mar. Fish.
Rev. 40(5-6):2-8.
Sutherland, D.L., G.L. Beardsley, and R.S. Jones
1983 Results of a survey of the south Florida fish-trap fishing
grounds using a manned submersible. Northeast Gulf Sci.
6(2):279-183.
Abstract. - Movements of 25
yellowtail rockfish Sebastes flavidus
on Heceta Bank, off Oregon, were
studied by acoustical tagging and
tracking during the summers of
1988-90. Some fish were tracked dis-
continuously up to 1 month after
transmitters were inserted into their
stomachs. In each year, some fish re-
mained at the capture site after
release or returned after displace-
ment to a different release site. In
1990, the year of most intensive tag-
ging, 11 of 12 fish were detected
near the capture location 13 days
after release in August 1990, in-
cluding 3 of 4 fish displaced 0.5 nmi
(0.9km), all 4 fish displaced 2.0 nmi
(3.7 km), and all 4 of the fish released
at the capture site. One fish homed
overnight from the release site 0.5
nmi away. In September 1990, 1
month after release, eight of these
fish had dispersed up to 0.1-0.7nmi
(0.2-1. 3km) to the south of their cap-
ture location, suggesting a change
in site fidelity. Pressure-sensitive
transmitters showed that tagged
yellowtail rockfish usually remained
at midwater depths of 25-35 m, well
above the sea floor depth of ~75m.
Rapid descents to nearbottom depths
were common, but no obvious diel
vertical or horizontal migrations
were detected.
Movements of acoustically-tagged
yellowtail rockfish Sebastes flavidus
on Heceta Bank, Oregon
William G. Pearcy
College of Oceanography. Oregon State University. Corvallis, Oregon 97331-5503
Manuscript accepted 13 July 1992.
Fishery Bulletin, U.S. 90:726-735(1992).
The yellowtail rockfish Sebastes Jlavi-
dus is a common rockfish along the
west coast of North America. It is
caught by both commercial and rec-
reational fishermen and was one of
the most abundant rockfish species in
commercial groundfish landings from
the U.S. west coast from 1982 to
1990 (Pacific Fisheries Management
Council 1991).
Schools of yellowtail rockfish may
persist at the same location for many
years. Carlson (1986) reported that
a school of adult yellowtail rockfish
in southeastern Alaska consisted of
individuals from one or two year-
classes and had negligible recruit-
ment over an 11-year period. Be-
cause their aggregrations may be
site-specific with limited interchange
of adults, and because rockfish are
long-lived, late-maturing, and of low
fecundity (Gunderson et al 1980,
Love et al. 1990, Eldridge et al.
1991), overfishing or disturbances,
such as habitat modifications from
offshore mining or petroleum devel-
opment, may have long-lasting ef-
fects in a local area. On the other
hand, a rockfish species whose in-
dividuals move freely from reef to
reef may be less vulnerable to local-
ized disturbances (Love 1979). The
stability and areal range of rockfish
aggregations have important implica-
tions for assessment, availability, and
management of rockfish species.
The yellowtail rockfish is the most
abundant, large-sized schooling fish
seen from submersibles over the
shallow, rocky areas on the top of
Heceta Bank, a deep reef located
~55km off the central Oregon coast
(Pearcy et al. 1989; Figs. 1 and 2).
Large pelagic schools, sometimes of
a thousand or more individuals, were
observed over shallow portions of the
bank (<150m) during the summer.
Based on both observations from
submersible dives and the occurrence
of large echo-groups recorded by the
ships' echosounders, these schools
were often associated with pinnacles
or high-relief topography (Pearcy et
al. 1989).
During one dive, a school of yellow-
tail rockfish followed the submersible
along the bottom for over an hour
before abruptly turning and swim-
ming back toward the location where
the school was initially encountered
(Pearcy et al. 1989). This observation
and those of Carlson and Haight
(1972), who found that individual
rockfish returned to a home site in
southeast Alaska after being dis-
placed as far as 22.5km, suggest that
schools of yellowtail rockfish may
have home ranges centered around a
specific site on the bank.
Pelagic rockfishes, such as the
yellowtail rockfish, may range over
wider areas than benthic rockfishes.
However, little is known about the
vertical distribution or diel vertical
migrations of yellowtail rockfish, or
the relationships between vertical
and horizontal movements.
This study used acoustical tracking
to determine the horizontal and ver-
tical movements and site-specificity
of yellowtail rockfish on Heceta
Bank. In this paper, I define site-
specificity as the tendency of fish to
inhabit a specific localized area as
opposed to free-ranging or vagrant
726
Pearcy: Movements of acoustically-tagged Sebastes flavidus
727
125"30'
124'"30'
-f-
13"
30'
IS"
4-
-f-
h--
125"30-
125"00'
124''30
Figure 1
Location of Heceta Bank off Oregon. Hatched area of the bank encompasses the area of this study as shown in Figure 2 (depth
contours in meters).
behavior. Homing is defined as returning to a site
formerly occupied instead of to equally probable loca-
tions. Homing does not imply a direct, straight-line
course back to the home site.
Yellowtail rockfish are ideal for acoustical tracking
because they are common, large in size, and do not suf-
fer from the lethal embolisms of other rockfishes when
brought to the surface, but instead expel swimbladder
gases during decompression*.
* Bubbles of gas were observed emanating from the region of the
opercle as yellowtail rockfish were reeled from about 2-3 m depth
to the surface. By immersing fish in tanks aboard ship, these
bubbles were seen forming and being expelled from under the thin
skin between the last gill and the cleithrum anterior to the base
of the pectoral fin. Samples of the gas were collected in syringes
and analyzed with a microgasometer using the methods of Scho-
lander et al. (1955). The gas was comprised of about 75% oxygen,
indicating that gases from the swimbladder were released without
causing lethal embolisms when yellowtail rockfish were rapidly
decompressed.
728
Fishery Bulletin 90(4). 1992
44°05 '_
44°00
124° 55
124° 50
Figure 2
Topography of the southern portion of Heceta Bank and the
areas encompassed in Figures 3-6. Depths are in meters.
Methods
Yellowtail rockfish were captured with hook-and-line
during the daytime and placed in deck tanks with
circulating seawater. Acoustical transmitters were
inserted into the stomachs of large (42-54 cm fork
length), active, uninjured, unanesthetized fish with a
1cm diameter rod. Most fish were males. Fish were
released within 15min at the capture site or within
30min at displaced locations. Displaced fish were
released over bottom topography and depths known to
be inhabited by yellowtail rockfish (except in 1988), and
within 2nmi of the capture location to facilitate survey
of several release sites during a cruise. They were
tracked using a directional hydrophone and acoustical
receiver. Four fish tracked during 1989 were also
tagged with external Floy tags.
Preliminary tagging
Before this study began, a few tests were conducted
in circulating seawater tanks aboard ship or in the
laboratory to evaluate methods of tagging pelagic
rockfishes on five yellowtail and five black rockfishes
(Table 1). Dummy tags of the same dimensions and
weight as those used in the study were inserted into
the stomach or attached externally to the side of fish
under the dorsal fin. The external tag attachment,
Table 1
Information on tag retention in 5 yellowtail Sebastesflavidus,
and 5 black S. melanops, rockfishes aboard ship (S) and in the
laboratory ashore (L). ARM = tags with hooks (anti-regurgi-
tation mechanisms).
Days to regurgitate
Date S/L Tag (or die)
S. flavidus
Sep 1988
Aug 1990
S. melanops
Apr 1990
Jul 1990
External
Stomach
Stomach-ARM
Stomach
Stomach-ARM
Stomach
Stomach-ARM
(3)
.2,(6)
11
1
9
0.5
2,89
similar to the one used by Matthews et al. (1990) for
benthic rockfish, caused one fish to list to one side and
interfered with its swimming. It died after 3 days in
the ship's tank. Three fish with tags inserted in their
stomachs had normal orientation in deck tanks and
were more active. One of these fish died after 6 days.
Stomach insertion of tags was used in this study. This
method is quick and minimizes handling and trauma
to the fish (Stasko and Pinock 1977). The major disad-
vantage was possible regurgitation of tags (Table 1 and
Results), although Stasko and Pincock (1977) reported
that transmitters inserted into the stomachs of many
other species of fishes were not disgorged. During the
last year, to increase the retention of tags in the
stomachs, one or two small (no. 18 steel dry fly) hooks
were attached to the ends of tags with epoxy as anti-
regurgitation mechanisms (ARM's). Hooks protruded
2 mm from the tag. In experiments in large aquaria or
tanks in the laboratory, tags with ARM's stayed inside
either black rockfish S. melanops or yellowtail rockfish
a total of 2, 9, 11, and 89 days compared with 0.5, 1,
1, and 2 days (and one that died after 6 days) for con-
trols without ARM'S (Table 1).
Equipment
Acoustical tracking equipment (VEMCO Ltd.) was used
in this study, including a VR-60 receiver with preset
channel frequencies, a telemetry decoder and display
unit, directional hydrophone, and transmitters or tags
with five different crystal-controlled frequencies.
Transmitters were 16mm in diameter, 48-65mm in
length, with batteries of 4.5, 9, 21, and 60 days rated
life-span. In 1990, transmitters with the same frequen-
cies had different pulse widths and pulse periods which
Pearcy: Movements of acoustically-tagged Sebastes flavidus
729
were decoded and displayed by the receiver. The transmitters had
a rated range of 500-1500m.
Pressure-telemetering transmitters with battery durations of
4.5 days were used in 1989 and 1990. The pulse rates of these
transmitters were linearly proportional to pressure, and individual
calibrations were incorporated into the receiver program. The
manufacturer claimed accuracy was 5% of the full range, or 5 psi
(~3m depth), similar to my test of two tags lowered vertically
on a metered line at sea. Data on depths of fish and time of day
from these transmitters were printed at regular intervals aboard
ship and stored in the receiver. During 1989, depths and times
were recorded manually every 5min or less. During 1990, data
on time and depth were stored automatically by the receiver every
0.5 sec. Median depths were calculated for every 25-sec period and
plotted by computer.
Field procedures
Research was conducted using either the RV William A. McGaw,
a 32 m ship used to support submersible research, or the FV Cor-
sair, an 18m trawler. Echosounders were used to scout concen-
trations of fish over the shallow (60-90 m) portions of Heceta
Bank. When dense midwater schools of fish were detected,
weighted fishing lines with jigs were lowered to catch fish. Only
yellowtail rockfish were caught from these midwater schools,
which were usually at depths of 20-40 m. Often the schools were
so compact that our fishing weights bounced off fish at these
depths. If yellowtail rockfish were readily caught, our position was
recorded and an anchored surface float was released from the
Corsair to provide a fixed reference to prevent drifting off-station
and assist tracking of fish.
Transmitter signals were detected with a directional hydrophone
attached to the end of a 4 m rotatable pole mounted to the side
of'the vessels. The hydrophone pole was rotated through 360°
until the signal strength of a transmitter was maximal. Then the
vessel headed directly toward the transmitter. Signal strength
increased as the range closed. When signal strength was equally
high in all directions or when the direction of the signal decreased
rapidly, we assumed that the fish was in the vessel's immediate
vicinity and our location was then determined by LORAN C. Re-
peated positions for stationary transmitters on the bottom were
within O.lnmi (~180m) from one another. Repeatable accuracy
of Loran C for one vessel is about 100 m (Dugan and Panshin
1979).
Results
Horizontal movements
1 988 (Fig. 3) Four yellowtail rockfish were caught near the bot-
tom, tagged, and released during September 1988. Three fish were
released where they were caught, and the fourth was displaced
about 1 nmi seaward of its catch location. Fish 1 was caught and
released over a shallow (71m), high-relief rocky area of Heceta
Bank on 15 September (Fig. 3). Three locations were determined
44"02
44°00'
43 58' —
124° 55'
124° 51'
Figure 3
Locations and tracks of yellowtail rockfish
Sebastes flavidus tagged and released in 1988.
Fish 1, released 15 September at 1919 h ( • ), was
located 24 h later (■), and was found within the
open circle (O) during seven intervening fixes.
Fishes 2, 3, and 4 were released 13 September
( • ) and followed indicated paths. Fish 3 moved
less than 200 m.
immediately after release, three after 12 h,
and two after ~24h. All locations were
within 0.5 nmi of one another, and the last
was O.lnmi from the capture site. Fish 2 and
3 were caught, released, and detected once
over the southernmost shallow portion of
Heceta Bank at a depth of 80 m on 13 Sep-
tember. One of these fish was located ~0.75
nmi (1400 m) east of its capture location
after 7h. The other fish was found within
200m of the release site 17h after release.
To determine if a stationary transmitter
location was the result of a regurgitated tag,
the submersible Delta, with a separate
hydrophone and receiver, dove on Fish 3,
which remained close to the release site. The
ship maintained position over this transmit-
ter as the submersible was launched. Al-
though a strong signal was recorded from
the transmitter, its bearing changed fre-
quently, indicating that the tag was moving
and had not been regurgitated. This was
confirmed when the bearing of the transmit-
ter changed 180° as a school of several hun-
dred fish swam under the submersible. The
fish transmitting the signal had two exter-
nal Floy tags but was not seen.
730
Fishery Bulletin 90(4), 1992
The fourth fish was captured at the southern high spot of
Heceta Bank (80 m depth, same date and capture site as
Fishes 2 and 3) and released 27min after capture l.Snmi off-
shore in a habitat where yellowtail rockfish were rarely
seen— where the bottom was 150 m, flat, and comprised of
fine sediments. Between 1727h on 7 September and 0730h
the next day, this fish was tracked continuously (Fig. 3). It
moved to the northeast until 0400 h, turned south, but then
resumed its northeasterly course, ending up near the 75 m
depth contour just west of a shallow region of the bank, about
2nmi from its capture location.
1 989 (Fig. 4) Two experiments were conducted in 1989 to
further investigate horizontal movements: one involved
three fish caught and released with pressure-telemetering
transmitters at a station on 21 August (1 fish) and 24 Aug-
ust 24 (2 fish). The other experiment included six fish, three
of which were released at the capture site and three displaced
1.1 miles away, on 25 August. All fish were caught in mid-
water at ~79m.
Fish 2 in the first experiment was released at site A and
was tracked continuously for 11 h after release. During this
time it stayed within ~0.2nmi of the release site, which was
marked by a surface buoy. We returned to this location 36 h
later and found this fish 0.5nmi to the east. After 1.5h it
returned to the release site and was located several times
in this vicinity during the next 56 h (Fig. 4).
Two other fish were caught, tagged, and released 3 days
later, at site B, and tracked for about 24 h. Acoustical signals
of these two fish stayed within 0.2 nmi of the release loca-
tion during this period.
In the second experiment, six fish were caught at site B
on 25 August. Three fish (10, 11, 12) were released at the
capture location and three (7, 8, 9) were displaced 1.1 nmi
to the northeast of site C and released in 77 m of water.
When we returned to these locations 9 days later, two (7
and 9) of the three tags from fish displaced to site C were
detected and remained there over the next 36 h. Distinctive
double pings from these two tags were heard on the receiver,
indicating that the tags had been regurgitated and were on
the bottom.
Fish 10 and 11, which were released at capture site B, were
detected ~0.1nmi south of site B 9 days after capture.
Signals from the third fish (12) were not detected. The
transmitter from displaced Fish 8 was detected ~0.4nmi to
the east of the capture site. Within the next 36 h, this fish
moved to within 0.2 nmi of the capture location, and its last
position was 0.3 nmi from the capture site.
The submersible Delta was used to dive on one of the tags
that was stationary at the displacment location (site C). This
transmitter was found lying on top of a large rock.
1 990 (Figs. 5 and 6) During 1990, transmitters with ARM's
were inserted into 12 yellowtail rockfish. All fish were
caught during early evening (1900 h) on 15 August in mid-
^; .. 0 0 '
124 55'
12'i
Figure 4
Locations and tracks of yellowtail rockfish Sebastett
Jlat'idus released in 1989. See text for details. Dashed
line is the assumed path of Fish 8.
.;4° 03' -
;4° 02' ~
4 4° 01'
124° 510'
Figure 5
Tracks of 11 yellowtail rockfish Sehast.es Jlamdus cap-
tured at site A and released at sites A (dashed lines),
B (dotted lines), and C (solid lines) on 15 August 1990.
Symbols designate dates and times that positions were
obtained (see legend).
Pearcy: Movements of acoustically-tagged Setosfes flavidus
731
44 01,5 —
44 01 0
144 52 0
124 51 0'
Figure 6
Tracks of eight of the yellowtail rockfish Sehastes flamdus released on 15 August, one
month later during the period 16-18 September 1990. Symbols designate positions of
each fish.
water above a 68m rocky bottom. Four of these fish
were released at capture site A, four at 0.5 nmi to the
north (site B, bottom depth 70m), and four at 2.0 nmi
to the north (site C, bottom depth 87m) (Fig. 5). The
two release sites to the north of the capture site had
high-relief bottom topography, similar to the capture
site. In addition, schools of rockfish, similar in appear-
ance acoustically to those comprised of yellowtail rock-
fish, were observed in the vicinity of the two displace-
ment sites. Since yellowtail rockfish are numerous over
shallow (<100m) rocky ridge, boulder, and cobble
habitats of Heceta Bank, and schools of yellowtail
rockfish were seen from submersibles near release sites
B and C (Pearcy et al. 1989, Hixon et al. 1991), I as-
sumed that the transplant release sites were habitable
by yellowtail rockfish.
The morning after the releases, all four of the trans-
mitters in fish released at capture site A were detected
within 0.1 nmi of site A. One of the fish (23) released
0.5 nmi to the north returned to the capture site over-
night, after 17h. No transmitters were detected at the
other two release sites on the following day when the
ship passed over these locations and departed the bank.
Eleven of the twelve fish were located 13 days after
release when we returned to Heceta Bank, including
all four released at site C (2.0 nmi to the north) (Fig.
5). The missing transmitter was from site B. All 11 fish
were found at least once within 0.1 5 nmi of the capture
site (Fig. 5). These results are evidence for a strong
homing tendency.
Two fish that were caught and released at the orig-
inal capture location (site A) showed the most ex-
tensive short-term movements (Fig. 5). Fish 26 was
located 0.8 nmi and Fish 31 was
found 0.5 nmi north of site A dur-
ing the night and early morning
of 28-29 August, 14 days after
release. Both returned to site A
about 11 h later. Fish 27, dis-
placed to site C, was found 0.15
nmi east of capture site A and
then moved 0.24 nmi to the north
during a 2-hour period on the
evening of 29 August.
One month after releases, we
returned to the capture location
to study longer-term movements.
No transmitters were detected in
the immediate vicinity of the orig-
inal capture location. Using an ex-
panding rectangular search pat-
tern, 8 of the 12 transmitters were
discovered, all south and a distance
of ~0.1-0.7nmi from the capture
location. Locations of these fish
were determined over the next 2.5 days during three
periods between submersible operations. The fish were
scattered along a 1.1 nmi east-west axis (Fig. 6). Most
fish demonstrated short-term movements of over 0.1
nmi (our nominal error of navigation) during these 2.5
days. Two fish (23 and 28) moved ~0.5nmi. Only one
fish (22) ended up near the location where it was ini-
tially found on this cruise. None of these eight fish was
found closer than 0.1 nmi to the original capture site,
and most were 0.4 nmi away. There was no evidence
that these fish stayed in a common school or within a
small home range, as found earlier in the summer.
Vertical movements
Pressure-telemetering (depth sensor) transmitters
were used during 1989 and 1990, but due to problems
with the receiver, limited data were obtained. Figure
7 shows the maximum and minimum depths for 10-min
intervals for three fish monitored almost continuously
during 21-22 and 24-25 August 1989. Fish were usual-
ly in midwater, inhabiting depths of 25-50 m where the
bottom was ~75m. Short-duration vertical movements
were seen for all fish, usually rapid descent/ascent
("bounce") dives to or close to the bottom, followed by
rapid vertical ascents back to depths of 25-35 m. Fish
2 made nine of these "bounce dives" to the bottom dur-
ing the early morning of 24 August over about a 4h
period. Other than this series of dives, there was little
evidence for any diel patterns in the frequency of ver-
tical migrations of fish that were tagged. Fish 3 either
regurgitated its transmitter or rested on the bottom
after 0700 h on 25 August (Fig. 7).
732
Fishery Bulletin 90(4), 1992
X
I—
Q_
UJ
Q
0
25
50
75
0
25
50 [
75
0
25
50
75
0
25
50 ■
75
100
#2
#3
#5
AUG 21
AUG 22
AUG 24
AUG 25
AUG 25
AUG 25
^r-\m^ 1 f
12 16 20 0
HOUR
8 12
Figure 7
Depths of yellowtail rockfish Sebastes flavidus nos. 2. 3. and
5 measured with pressure-telemetering transmitters during
21-25 August 1989. Bottom depths 75-99 m during this period.
0
25
50
0
g 50
Q- 0
LJ
^ 25
50
0
25
50
75
1
Depths ol
and 6 mea
ing 14 Au
AUG 14 #2
AUG 14 #3
AUG 14 #5
V 1
AUG 14 #6
\ I-
1 12 13 14 15 15
HOUR
Figure 8
yellowtail rockfish Sebastes flavidus nos. 2, 3, 5,
sured with pressure-telemetering transmitters dur-
gust 1990. Bottom depths 75-80 m.
Vertical excursions of fish during late morning and after-
noon of August 1990 showed a similar pattern, with fish
occupying midwater depths and occasionally diving to deep
water (Fig. 8). The records for Fishes 2 and 3 show that these
fish descended toward the bottom immediately after release
and then rose to progressively shallower depths during the
next several hours. Synchronous vertical movements of
several fish were not common (Figs. 7, 8) but some did occur
among the three fish tracked during 25 August 1990 (Fig.
8). Sometimes fish dove as the vessel approached, perhaps
a response to ship noise (Ona and Godoe 1990), but at other
times fish descended when the vessel was not underway.
Maximum rates of descent for fish shown in Figure 8 were
0. 16-0.40 m/sec; maximum ascents were 0.1 5-0.31 m/sec.
Figure 9 shows the dive of Fish 5 after release, with the most
rapid descent during the first minute, and slower rates in
the next 3 minutes. Rapid vertical movements were also
observed during 1989, with maximum rates of descent of
0.15-0.45 m/sec, and rates of ascent of 0.15m/sec.
Discussion
Homing and horizontal movements
During all three years of the study, yellowtail rockfish on
Heceta Bank demonstrated site fidelity and homing. Dis-
placed fish returned from as far as 2nmi from their capture
site, and those released over rocky habitat and at similar
G
10
-\
-. 20
E
- \
I 30
\
IX
S 40
\^
50
^^'~~^^--,^
60
0 12 3 4
ELAPSED TIME (min)
Figure 9
Initial descent of yellowtail rockfish Sehaxtes Jlafidus
no. 5, starting at 1 154h, 14 August 1990, after release
at the surface (see Fig. 8).
depths returned to the location of capture in 1990.
Eleven of twelve fish tagged in 1990 returned to
or remained close to the original capture site 13
days after release. One fish that was displaced 0.5
nm returned overnight to the location where it
was captured.
Pearcy: Movements of acoustically-tagged Sebastes flavidus
733
Carlson and Haight (1972) also found that adult
yellowtail rockfish returned to their home site, some
from as far as 22.5km, some after displacement to
other yellowtail schools, and some after 3 months in
captivity. In both studies, yellowtail rockfish homed
even if released at sites where the habitat was similar
to that at the capture site and near other schools of
yellowtail rockfish. This demonstrates fidelity to a
home site.
Not all yellowtail rockfish demonstrate site fidelity,
however. Eight of ten recoveries of 153 yellowtail
rockfish tagged in Puget Sound were from the open
Washington coast, 58-2214 days after release, indicat-
ing an offshore migration probably related to the onset
of maturation of these fish (Mathews and Barker 1983).
In another study, Stanley (1988) tagged 4622 yellowtail
rockfish in Queen Charlotte Sound, British Columbia
during 1980 and 1981. As of 1987, the five that were
recovered moved from <10km to >300km. Of 9417
yellowtail rockfish tagged southwest of Vancouver
Island, 24 were recovered. Twelve moved < 10km while
others were recovered 23 to > 500 km from the tagging
location (Stanley 1988).
The degree of site fidelity and movement of yellowtail
rockfish may be related to the bottom topography of
the tagging location. This appears to be the case for
black rockfish, another offshore pelagic species. Culver
(1987; B.N. Culver, Wash. Dep. Fish., Montesano, pers.
commun.) found that black rockfish tagged over a
5-year period from rocky habitats of northern Wash-
ington exhibited "no significant movement," whereas
fish tagged in areas that had sandy sediments or small
pinnacles off the central Washington and northern
Oregon coast displayed appreciable movements. Per-
haps yellowtail rockfish on Heceta Bank, and other
rocky banks, are less mobile than those inhabiting areas
with level seafloors.
Carlson and Haight (1972) found that fish displaced
to sites across open water with depths >100m returned
to the site of capture with much less frequency than
did fish released along the adjacent coast in shallower
depths. One fish in my study, released in relatively deep
water off Heceta Bank and tracked continuously for
14 hours (Fish 4, Fig. 3), was not oriented toward its
home site. These observations suggest that homing is
most effective over relatively shallow water (<100m),
even though yellowtail rockfish are basically midwater
fish. Homing may also be influenced by topography.
Matthews et al. (1987) reported that displaced copper
and quillback rockfishes S. caurinus and S. maliger
returned to high-relief, but not to low-relief, reefs.
The sensory mechanisms and environmental cues
used for homing and home-site recognition by yellowtail
rockfish are not known. Possibly the fish on Heceta
Bank recognized familiar topography and prominent
"landmarks." Movements of up to 0.75nmi by yellow-
tail rockfish (Figs. 5 and 6) indicate that they do not
always have as small a home range and may range over
a large portion of the bank. Perhaps they learn visual
"landmarks" over much of the bank in this way.
One fish returned to its home site from 925 m after
only 11 hours, mainly during the night when recogni-
tion of visual landmarks would have been more diffi-
cult. This fish returned home more rapidly than sub-
strate-associated copper and quillback rockfishes that
took 8-25 days to return home after displacement of
only 500 m (see Matthews 1990 for this and summary
of homing by other rockfishes). This suggests oriented
or directed movement.
Eight of the twelve fish tagged in August 1990 were
relocated 1 month after release but were all south of
the capture location and scattered in an east-west direc-
tion. None was found within 0.1 nmi of the capture site.
This dispersal from the capture site suggests reduced
site fidelity and perhaps seasonal dissociation of indi-
viduals from the large schools observed earlier during
the summer. This dispersal may be associated with
seasonal changes, perhaps related to mating behavior
and the fact that most of the fish tagged were large
males. Carlson and Barr (1977) reported that the
spatial distribution and activity of yellowtail and dusky
{S. ciliatus) rockfishes differed markedly between
May-October, when they were seen in the water col-
umn and apparently actively feeding, and November-
April, when they withdrew into crevices between
boulders. Although no distinct seasonal changes are
known in the bathymetric distribution of yellowtail
rockfish (J. Tagart, Wash. Dep. Fish., Olympia, pers.
commun., Aug. 1991), the spatial distributions of other
species of rockfishes are known to change seasonally
(Miller and Geibel 1973, Patten 1973, Matthews et al.
1987). Several species of juvenile rockfishes are known
to move to deeper reefs with the onset of fall and
winter storms (Love et al. 1991). It would be interesting
to learn if the yellowtail rockfish of Heceta Bank
disperse and become more benthic during the late fall
and winter, and then if they eventually regroup at the
original capture location next spring after spawning,
or instead acquire new home sites on the bank. Studies
are obviously needed on seasonal and long-term move-
ments of yellowtail rockfish.
Diel vertical movements
Most yellowtail rockfish were seen swimming above the
bottom during submersible dives on Heceta Bank.
However, a few were observed resting on the sea floor.
More fish were observed inactive on the bottom dur-
ing night than day dives. The tagged yellowtail rockfish
of Heceta Bank were pelagic, swimming far above the
734
Fishery Bulletin 90(4), 1992
bottom most of the time. Data from pressure-tele-
metering tags show that fish dove toward the bottom
but remained there only briefly. Only one fish with a
pressure transmitter either rested on the bottom for an
extended period or disgorged its transmitter (Fig. 7).
Little is known about the diel vertical distribution of
rockfishes. Schools of S. entomelas and S. proriger are
known to rise off the bottom during the night and
become more diffuse than dense schools on the bottom
during the day (Leaman et al. 1990). Rockfish may
intercept vertically-migrating pelagic organisms that
constitute their primary prey, feeding closer to the sur-
face at night or during crepuscular periods and de-
scending with their prey during the day. Sometimes
vertically migrating prey, such as euphausiids, are
advected onto banks and seamounts and trapped near
the bottom during the day where they are devoured
by rockfishes (Isaacs and Schwartzlose 1965, Chess et
al. 1988, Genin et al. 1988, Hobson 1989). Euphausiids
are often the primary prey of adult yellowtail rockfish
(Lorz et al. 1983). About 50% of the diet by weight of
yellowtail rockfish from Heceta Bank was comprised
of euphausiids (Brodeur and Pearcy 1984). However,
vertically-migrating mesopelagic fishes and shrimp
were the primary food items of yellowtail rockfish col-
lected in deeper water (137m bottom depth) along the
southern edge of Astoria Canyon (Pereyra et al. 1969).
Yellowtail rockfish from Heceta Bank did not
demonstrate obvious diel changes in their behavior by
either rising closer to the surface at night or swimming
over deeper water to intercept more oceanic organisms.
Such behavior has been observed for other species of
rockfishes (Chess et al. 1988, Leaman et al. 1990), and
predatory shore fishes are known to migrate offshore
at night to feed in midwater (Hobson 1968). One
yellowtail rockfish on Heceta Bank with a pressure-
telemetering transmitter made more dives to the
bottom during night than day.
The reasons for dives to the bottom are unclear. One
possible explanation is that these dives assist the fish
in localizing their position on the bank and preventing
drift of the school away from their home station. Sur-
face currents often set the ship away from tagged fish
that appeared to be geostationary. Yellowtail rockfish
must be able to orient to a specific site and swim
against prevailing currents to maintain their position.
Tagging-tracking techniques
Sonic tags inserted into the stomachs of yellowtail
rockfish without retention hooks were useful for track-
ing fish for several days. Most fish showed detectable
movements up to 2 days after release. Horizontal
movements greater than the accuracy of fixes were
found in one fish 10 days later, but this was an excep-
tion. Depth sensor tags provided reliable information
on the retention of tags since fish were almost always
in midwater. Fish with depth transmitters remained
in midwater up to 5 days, the rated duration of the
batteries. One pressure-sensitive tag (Fish 3. Fig. 7)
was apparently regurgitated after 22 h and fell to the
bottom. If arm's were employed on tags, fish move-
ments were measurable for 1 month after release. One
ARM tag dropped to the bottom immediately after the
fish was released, demonstrating that restraining hooks
are not a guarantee that tags will stay in the stomach.
Eight of the twelve fish with non-pressure telemeter-
ing ARM tags that were relocated moved significant
distances 30 days after the release of fish, indicating
long-term retention of transmitters.
Effects of the transmitter on behavior of the fish are
not known. However, one fish apparently schooled soon
after release. Although fish dove toward the bottom
immediately after release, they rose to typical mid-
water depths after less than an hour. These observa-
tions suggest that the trauma of being caught, tagged,
and released, and the added weight of the transmit-
ter, did not have prolonged effects and tagged fish
behaved normally.
Acknowledgments
This research was supported by the Department of In-
terior's U.S. Minerals Management Service (Cooper-
ative Agreement 14-12-0001-30445). NOAA's National
Undersea Research Program provided supplemental
shiptime and submersible support. I thank J. P. Fisher,
B.N. Tissot, R. Albright, and the Captain of the FV
Corsair, W. Dixon, and his crew for conducting this
research at sea, J. P. Fisher for his help with analyses
of the data, and J. P. Fisher, A. Schoener, M.A. Hix-
on, and two anonymous reviewers for helpful com-
ments on the manuscript. A. Ebeling and M.A. Hixon
provided the microgasometer and advice to determine
the composition of gas escaping from decompressed
yellowtail rockfish.
Citations
Brodeur, R.D., and W.G. Pearcy
1984 Food habits and dietary overlap of some shelf rockfishes
(genus Sebasles) from the northeastern Pacific Ocean. Fish.
Bull., U.S. 82:269-293.
Carlson, H.R.
1986 Restricted year-class structure and recruitment lag within
a discrete school of yellowtail rockfish. Trans. Am. Fish. Soc.
115:490-492.
Pearcy: Movements of acoustically-tagged Sebsstes flavidus
735
Carlson, H.R., and L. Barr
1977 Seasonal changes in spatial distribution and activity of
two species of Pacific rockfishes, Sebastes flavidus and S.
ciliatus, in Lynn Canal, southeastern Alaska. Mar. Fish. Rev.
39(3):23-24.
Carlson, H.R., and R.E. Haight
1972 Evidence for a home site and homing of adult yellowtail
rockfish, Sebastes flavidus. J. Fish. Res. Board Can. 29:
1011-1014.
Chess. J.R. , S.E. Smith, and P.C. Fischer
1988 Trophic relationships of the shortbelly rockfish, Sebastes
jordani, off central California. Calif. Coop. Oceanic Fish.
Invest. Rep. 29:129-136,
Culver, B.N.
1987 Results from tagging black rockfish {Sebastes melanops)
off the Washington and northern Oregon coast. In Proc, Int,
rockfish syinp,, p. 231-239. Univ. Alaska Sea Grant Rep. 87-2,
Fairbanks.
Dugan, R.F., and D.A. Dashin
1979 How to get the most out of LORAN-C. Oregon State
Univ. Ext. Mar. Advis. Prog. SG-54, Corvallis, 12 p.
Eldridge, M.B., J. A. Whipple, M.J. Bowers, B.M. Jarvis and
J. Gold
1991 Reproductive performance of yellowtail rockfish, Sebastes
flavidus. Environ. Biol. Fish. 30:91-102.
Genin, A., L. Haury, and P. Greenblatt
1988 Interactions of migrating zooplankton with shallow
topography: Predation by rockfishes and intensification of
patchiness. Deep-Sea Res. 35:1,51-175.
Gunderson, D.R., P. Callahan, and B. Goiney
1980 Maturation and fecundity of four species of Sebastes.
Mar. Fish. Rev. 42(3):74-79.
Hixon, M.A., B.N. Tissot, and W.G. Pearcy
1991 Fish assemblages of rocky banks of the Pacific North-
west. Final Rep. to U.S. Dep. Int., Minerals Manage. Serv.,
MMS 91-0052, Oregon State Univ.. Corvallis, 409 p.
Hobson. E.S.
1968 Predatory behavior of some shore fishes in the Gulf of
California. U.S. Fish. Wildl. Serv. Res. Rep. 73, 92 p.
1989 Predation on ocean krill. Science (Wash. DC) 243:237.
Isaacs, J.D.. and R.A. Schwartzlose
1965 Migrant sound scatterers: Interaction with sea floor.
Science (Wash. DC) 150:1810-1813.
Leaman, B.M., R. Kieser, P. Withler and R.D. Stanley
1990 W.E. Richer hydroacoustic cruise to study rockfish
behavior off northern Vancouver Island. March 14-23, 1990.
Can. Manuscr. Rep. Fish. Aquat. Sci. 2091, 63 p.
Lorz, H.V., W.G. Pearcy, and M. Fraidenburg
1983 Notes on the feeding habits of the yellowtail rockfish,
Sebastes flavid'us, off Washington and in Queen Charlotte
Sound. Calif. Fish Game 69:33-38.
Love, M.S.
1979 Isolation of olive rockfish, Sebastes serranoides, popula-
tions off southern California. Fish. Bull., U.S. 77:975-983.
Love, M.S., P. Morris, M. McCrae, and R. Collins
1990 Life history aspects of 19 rockfish species (Scorpaenidae:
Sebastes) from the southern California Bight. NOAA Tech.
Rep. NMFS 87, 38 p.
Love, S.M., M.H. Carr. and L.J. Haldorson
1991 The ecology of substrate-associated juveniles of the genus
Sebastes. Environ. Biol. Fish. 30:225-243.
Mathews, S.B., and M.W. Barker
1983 Movements of rockfish {Sebastes) tagged in northern
Puget Sound, Washington. Fish. Bull., U.S. 82:916-922.
Matthews, K.R.
1990 A telemetric study of the home ranges and homing routes
of copper and quillback rockfishes on shallow rocky reefs. Can.
J. Zool. 68:2243-2250.
Matthews, K.R.. B.S. Miller, and T.P. Quinn
1986 Movement studies of nearshore demersal rockfishes in
Puget Sound, Washington. In Proc, Int. rockfish symp.. p.
63-72. Univ. Alaska Sea Grant Rep. 87-2. Fairbanks.
Matthews. K.R.. T.P. Quinn. and B.S. Miller
1990 Use of ultrasonic transmitters to track demersal rockfish
movements on shallow rocky reefs in Puget Sound, Wash-
ington. In Parker, N.. et al. (eds.). Fish marking techniques,
p. 375-379. Am. Fish Soc. Symp. 7, Bethesda.
Miller, D.J., and J.J. Geibel
1973 Summary of blue rockfish and lingcod life histories; a reef
ecology study; and giant kelp, Macrocystis pyrifera, experi-
ments in Monterey Bay. California. Calif. Fish Game Fish
Bull. 158, 137 p.
Ona. E., and O.R. Godoe
1990 Fish reaction to trawling noise: The significance to trawl
sampling. In Karp, W.A. (ed.) Developments in fisheries
acoustics: A symposium held in Seattle. 22-26 June 1987.
Rapp. P.-V. Reun. Cons Int. Explor. Mer 189:159-166.
Pacific Fisheries Management Council
1991 Status of the Pacific coast groundfish fishery through
1991 and recommended acceptable biological catches for
1992. Pac. Fish. Manage. Counc. Portland, 69 p.
Patten, B.G.
1973 Biological information on copper rockfish in Puget Sound,
Washington. Trans. Am. Fish. Soc. 102:412-416.
Pearcy, W.G.. D.L. Stein, M.A. Hixon, E.K. Pikitch, W.H. Barss,
and R.M. Starr
1989 Submersible observations of deep-reef fishes of Heceta
Bank, Oregon. Fish. Bull., U.S. 87:955-965.
Pereyra. W.T. , W.G. Pearcy, and F.E. Carvey
1969 Sebastes flavidus, a shelf rockfish reeding on mesopelagic
fauna, with consideration of ecological implications. J. Fish.
Res. Board Can. 26:2211-2215.
Scholander. P.P., L. Van Dam. C.L. Claff, and J.K. Kanwisher
1955 Micro gasometric determination of dissolved oxygen and
nitrogen. Biol. Bull. (Woods Hole) 109:328-334.
Stanley, R.D.
1988 Shelf rockfish (silvergray, yellowtail, and canary rock-
fish. In Fargo, J., et al. (eds.), Groundfish stock assessments
for the west coat of Canada in 1987 and recommended yield
options for 1988, p. 227-272. Can. Tech. Rep Fish. Aquat.
Sci. 1617.
Stasko, A.B., and D.G. Pincock
1977 Review of underwater biotelemetry, with emphasis on
ultrasonic techniques. J. Fish. Res. Board Can. 34:1261-1285.
Abstract.- Monte Carlo simula-
tion is used to quantify the uncer-
tainty in the results of sequential
population analysis and related sta-
tistics. Probability density functions
describe the measured or perceived
uncertainty in the inputs to the
assessment model. Monte Carlo sim-
ulation is then used to examine the
variability in the resulting parameter
estimates (stock sizes and fishing
mortalities), derived statistics (e.g.,
Fq i), and in the management regu-
lations necessary to achieve various
management objectives. We show
how relative frequency histograms of
the simulation results can be used to
describe the risk of not meeting a
given management goal as a function
of the catch quota selected. We also
show how to compute the expected
cost, in terms of potential yield fore-
gone, associated with picking a con-
servative quota. This enables one to
balance risks and costs or to allow
risk to vary within proscribed limits
while keeping the catch quota stable.
We illustrate the use of the Monte
Carlo approach with examples from
two fisheries: North Atlantic sword-
fish and northern cod.
A simple simulation approach to risk
and cost analysis, with applications
to swordfish and cod fisheries
Victor R. Restrepo
University of Miami, Rosenstiel School of Marine and Atmospheric Science
4600 Rickenbacker Causeway. Miami, Florida 33149
John M. Hoenig
Department of Fisheries and Oceans. Science Branch
P,0 Box 5667, St John's, Newfoundland AlC 5X1. Canada
Joseph E. Powers
Southeast Fisheries Science Center, National Marine Fisheries Service. NOAA
75 Virginia Beach Drive. Miami, Florida 33149
James W. Baird
Department of Fisheries and Oceans, Science Branch
PO Box 5667, St John's, Newfoundland AlC 5X1, Canada
Stephen C. Turner
Southeast Fisheries Science Center, National Marine Fisheries Service, NOAA
75 Virginia Beach Drive, Miami, Florida 33149
Manuscript accepted 29 July 1992.
Fishery Bulletin, U.S. 90:736-748 (1992).
Fishery managers recognize the
dangers of accepting parameter
estimates without consideration of
the variability inherent in the esti-
mates of fish stock status and related
parameters. Early strategies for
dealing with this were quite simple,
such as replacing the estimate of the
fishing mortality giving the max-
imum yield (Fn,ax) by a more conser-
vative value. Sensitivity analyses, in
which the effects of various pertur-
bations of the inputs are observed,
have commonly been employed to ob-
tain impressions of the probable
bounds on the errors (e.g.. Pope
1972, Pope and Garrod 1975). Re-
cently, various authors have used the
delta method (see Kendall and Stuart
1977: 246-248) to obtain analytical
expressions or numerical solutions
for the variances and covariances of
outputs from simple sequential
population analyses (SPAs) (Saila et
al. 1985, Sampson 1987, Prager and
MacCall 1988, Kimura 1989). These
solutions tend to be complex, only
asymptotically valid, and highly
model-specific. They have only been
worked out for the simplest SPA
models and some simple quota-
setting procedures (e.g.. Pope 1983).
It is possible to measure how well
the population estimates correspond
to trends in indices of abundance
when calibration procedures are ap-
plied to sequential population anal-
yses. The variance-covariance matrix
of the stock sizes estimated directly
in the optimization procedure is cal-
culated from the inverse Hessian or
its approximation (Seber and Wild
1989); the variance of any function of
these parameters is then approx-
imated by the delta method. But
estimates of standard errors of
population sizes obtained in this way
do not reflect all the variability in the
inputs and the uncertainty in the
model, because they are conditioned
on a number of assumptions such
as natural mortality being exactly
736
Restrepo et al. Monte Carlo simulation applied to Xiphias gladius and Cadus morhua
737
known. For example, similar trends in population abun-
dance may be obtained when two very different values
of natural mortality rate are assumed as inputs; despite
the simOar trends (and hence correlation with the abun-
dance indices), there may be large differences in the
absolute estimates of abundance.
In this paper we use the term "uncertainty" to
describe any variability or error that arises during the
stock assessment process. Uncertainty can enter into
an assessment in various ways. There may be uncer-
tainties in the values of the inputs, e.g., the total catch
may be estimated with error. Also, the formulation of
the assessment model may be subject to uncertainty,
and the analyst may make data-dependent decisions
during the analysis which are subject to error. The
degree to which these sources of error are incorporated
into the analyses will determine the perceived uncer-
tainty in the overall assessment results. If all sources
of error are not appropriately accounted for, then
estimates of the uncertainty in the assessment results
may be too small.
Monte Carlo simulation is a convenient tool for study-
ing a model's outputs given different types and levels
of error in the model's inputs (e.g., Restrepo and Fox
1988). In a sensitivity analysis framework. Pope and
Gray (1983) and Rivard (1983) used a Monte Carlo
approach to study the relative contribution of various
inputs to the overall uncertainty in total allowable
catch (TAC) estimates obtained from calibrated SPAs.
Francis (1991) used Monte Carlo simulation analysis
to construct risk curves describing the chances of not
meeting management objectives as a function of the
catch quota. In this paper, we present a general
method, also based on Monte Carlo simulation, to ac-
count for uncertainty in assessment results, including
the parameters directly estimated from the SPAs as
well as derived statistics used to set management
targets and allowable catches. We also show how the
simulation results can be used to quantify the risk (of
not meeting a management goal) associated with the
selection of a given TAC, and we describe a measure
of the cost of picking a conservative catch quota.
We apply the simulation method to swordfish
Xiphias gladius in the North Atlantic Ocean and cod
Gadus morhua off eastern Newfoundland and
southeastern Labrador. These fisheries are quite dif-
ferent in nature. Swordfish are highly migratory,
managed internationally with fishing mortality con-
trols, and the data set allows for the estimation of
only a few parameters in models with many con-
straints. Northern cod, on the other hand, are demer-
sal, managed by quota, and the availability of age-
specific survey indices allows for the estimation
of many parameters with a minimum number of
assumptions.
Quantifying uncertainty
by simulation
Suppose the only uncertainty in the inputs to an assess-
ment model concerns the value of the instantaneous
natural mortality rate, M, and that M could be
anywhere in the interval 0.15-0.25/yr with equal
likelihood. One could compute the assessment model
results for a large number of uniformly spaced values
of M in this interval (e.g., 100) and make histograms
of the results. This would represent the perceived in-
formation about the relative likelihood of the estimated
output taking on various values. If not all values of M
were believed to be equally likely, one could weight the
100 outputs by the probability associated with the cor-
responding inputs.
The above procedure becomes awkward when there
are a number of inputs subject to uncertainty, because
the number of combinations of input parameter values
becomes very large. An alternative is to use a Monte
Carlo approach in which values of the inputs are drawn
randomly from probability distributions. A sufficient-
ly large number of plausible input data sets are thus
generated and used to compute the assessment model
results such that the distributions of the estimated out-
puts are clearly defined. This may involve hundreds or
thousands of runs, depending on the types of data and
models used (in our work we found that 500-1000 data
sets were necessary to obtain stable results).
A typical assessment of a fish stock using SPA in-
volves three levels of analysis. First, data are prepared
for the SPA. This usually involves estimating and age-
ing the annual catch, and computing indices of abun-
dance for calibration. Second, the SPA itself is carried
out (it is also frequently termed "VPA," for Virtual
Population Analysis). In many cases, several SPAs are
carried out to examine the goodness-of-fits of the in-
put data to alternative model formulations or simply
to examine the sensitivity of the results to the alter-
native formulations. Third, derived statistics are com-
puted. These are commonly biological reference points
(Fmax, Fq i; Gulland and Boerema 1973), and forward
projections of stock status and catches under alter-
native management actions.
It is easy to see how the Monte Carlo approach
can be used to characterize the uncertainty in the
entire analysis process, starting with the raw data
collected for the first step in the above procedure.
For instance, the total annual catches and their pro-
portions at age can be obtained by resampling the
original data that led to the catch estimates, through
a non-parametric bootstrap (Efron 1982). These boot-
strapped catches would then be used in the SPAs,
whose results, in turn, would affect the values of
projected future catches.
738
Fishery Bulletin 90|4). 1992
0 18-
0 16-
0.14-
£■ 012-
!q
(0 0.1-
n
o
a. 0 08-
006-
O04-
0 02-
\
'\ n
..1
nr^
° ' 160 170 180 190 200 210 220 230 240 250 260
total catch (1000 mt)
Figure 1
Probability mass distribution (relative frequency) for estimates
of total catch of cod Gadus morhua necessary to have 1991
fishing mortality equal to that of 1990. Estimates were ob-
tained from 1000 simulated data sets analyzed by the ADAPT
approach. If a total catch of 210,000mt is selected (arrow),
the probability that fishing mortality will exceed the status
quo is estimated by the sum of histogram bar heights to the
left of 210,000 (i.e., the shaded portion).
In practice, however, the time and computer re-
sources required to carry out such a large-scale simula-
tion make it more practical to derive the input uncer-
tainty distributions from parametric statistical analyses
of data. This would involve assuming a distribution type
for the inputs and estimating their mean and variance.
For example, by virtue of the central limit theorem,
an estimated mean has an approximately normal
distribution if the sample size is sufficiently large.
Often, the distributions determined for some of the
inputs will not be based on a rigorous statistical treat-
ment of the data, but rather will represent educated
guesses about the likelihood of the inputs taking on par-
ticular values (this is probably most true for the natural
mortality rate, M, which is usually assumed and not
estimated). The outputs would then represent the ana-
lyst's personal uncertainty in the assessment results.
The above approach can be generalized to allow for
uncertainty in the formulation of the SPA model as
well. Suppose one believes that there is a 70% chance
that the fishing mortality rate in the last year does not
decline with age after a fully recruited age (this is often
known as a "flat-topped" partial recruitment curve),
and a 30% chance that it does decline ("dome-shaped"
partial recruitment). Then one could conduct 70% of
the simulations with an SPA that assumes the flat-
topped curve and 30% with the dome-shaped curve.
The resulting combination of outputs would reflect the
intuitive estimate of uncertainty about the SPA model
formulation. Similarly, the approach can also account
for uncertainty concerning data-dependent decision
making. For example, if several abundance indices are
available, one might subject each index to a preliminary
test to decide whether the index is acceptable for
calibrating the SPA, e.g., via analysis of residuals. One
can repeat this decision-making process for each of the
simulated data sets and thus account for the uncertain-
ty associated with screening indices.
In summary, the Monte Carlo approach to quantify-
ing uncertainty consists of generating a large number
of pseudo-data sets, drawn at random from specified
distributions, and carrying out the entire assessment
procedure for each data set. The distributions of the
assessment outputs and derived statistics are then simi-
marized, e.g., as histograms. The simulation is thus
viewed as a means for translating input and model
uncertainties (measured and/or perceived) into output
uncertainties.
Analyzing the consequences
of management options
Estimating risl<s
The simulation results can also be used to quantify the
uncertainty associated with a future management ac-
tion. An example is the determination of uncertainty
in the catch of the current year that would maintain
the fishing mortality at the level of the previous year
(Fstatus quo)- A point estimate might be computed as
follows. An SPA of some sort is used to estimate the
population size at the end of the previous year and the
fishing mortality during that year. Assuming that
recruitment in the current year is equal to the long-
term average, the estimated population size in the cur-
rent year can be computed. Finally, the harvest which
causes the population to experience the same fishing
mortality rate as that estimated for the previous year
can be computed. To quantify the uncertainty in this
result, the whole procedure can be repeated 1000 times,
each time perturbing each input to the SPA by a ran-
dom amount (as per the specified uncertainty distribu-
tions). This results in 1000 sets of estimates of popula-
tion size, fishing mortality, and natural mortality rate
which, together with a set of randomly-drawn values
for recruitment, can be used to generate 1000 esti-
mates of the total allowable catch which will cause the
fishing mortality to remain unchanged. These values
can be organized into a relative-frequency histogram
such as in Figure 1.
From the histogram, it appears that the most likely
(modal) value for achieving the management goal
Restrepo et al.: Monte Carlo simulation applied to Xiphias gladius and Gadus morhua
739
(Fgtatus quo) 's a quota of 220,000 mt, but the actual
value might be anywhere from ~170,000 to 260,000
mt. If the TAG is set at 220,000 mt, then these results
suggest there is roughly a 50% chance of the fishing
mortality increasing and a 50% chance of it decreas-
ing. Suppose one is risk averse and chooses a TAG of
210,000mt instead. What would be the perceived risk,
or probability of exceeding the target fishing mortal-
ity, under this quota?
The risk of exceeding the target fishing mortality
(Fstatusquo) is givcn by the proportion of the area under
the histogram to the left of the TAG chosen (Fig. 1).
Thus,
I
Prob(Fachieved>Ftarget) = Z.
i-1
P(i)
where p(i) is the probability mass (relative frequency
of outcomes) associated with the ith bar of the histo-
gram, and I is the number of bars to the left of the
chosen TAG. This probability can be computed for any
value of the TAG. In practice, the risk would be com-
puted by sorting in ascending order the 1000 catch
values obtained from the simulation, and then plotting
the cumulative count of outcomes less than any value
of the TAG versus that value of the TAG (Fig. 2). One
can also derive a family of risk curves. For example,
separate curves could be generated for the risk of ex-
ceeding Fgtatus quo by each of several amounts (in ab-
solute or relative numbers). For each of the 1000
simulation runs, one computes the value of Fgtatus quo
and the catch that causes current F to exceed the status
quo by the specified amount. The resulting histogram
of catches is summed to obtain the risk curve.
Estimating cost as yield foregone
If we choose a conservative value for the TAG in order
to ensure that risk of exceeding the target fishing mor-
tality will be small, then we are probably passing up
some of the yield we could have had in the short term
while still meeting our objective (e.g., see Bergh and Butterworth 1987). It is possible to describe this cost in
economic or biological terms. Here, we express the cost as the expected value of the potential yield foregone,
which we define as follows. For any TAG, x, let
160
140
120
100-
80
60
40
20
0
100 120 140 160 180 200 220 240
total catch (1000 mt)
160 180 200
total catch (1000 mt)
Figure 2
(a) Probability of exceeding the current (1990) cod Gadus
morhua fishing mortality and expected value of the potential
yield foregone (with 95% confidence hand determined as the
2.5th and 97.5th percentiles of the distribution resulting from
1000 simulations) as functions of total catch selected for
1991. (b) Probability of the 1991 fishing mortality exceeding
the 50% rule fishing mortality, and expected value of the yield
foregone (with 95% confidence band), as functions of the total
catch selected for 1991.
d(i)
0, yield associated with ith interval of histogram < x
1 , yield associated with ith interval of histogram > x.
Then,
E (potential yield foregone) = ^ p(i) d(i) (y(i) - x)
1 = 1
where E(ll) denotes the expectation operator, the summation is over all intervals of the histogram (Fig. 1), and
740
Fishery Bulletin 90|4). 1992
y(i) is the yield associated with the ith interval of the
histogram. The expected potential yield foregone can
be plotted against the corresponding TAG (Fig. 2).
Here, y(i)-x is a possible value of the yield foregone
provided it is non-negative; negative values are elim-
inated by the indicator function 6{\); p(i) is the prob-
ability that the yield foregone is equal to ci(i) (y(i)-x).
In practice, the expected yield foregone would be com-
puted by setting all simulated catches which are less
than the TAG equal to zero and then computing the
mean of the 1000 values minus the TAG. The results
can then be plotted versus the TAG for various choices
of TAG (Fig. 2).
It should be noted that this cost relates to the up-
coming year only. One can also calculate the fate of the
biomass left in the water at the end of the upcoming
year. That is, one can ask whether this biomass left in
the water will increase or decrease over the year. In
general, for a quantity of biomass left in the water, the
relative change in its biomass over the year is given by
relative change in unfished biomass =
IPaWa
Here, Pa is the proportion of the stock that is age a;
Wa, the average weight of animals at age a; M, the
(constant) instantaneous natural mortality rate; and the
summations are over all age groups of interest.
Trade-offs in decision malting
The manager can now choose how to trade off poten-
tial yield and risk. For example, consider the option
of a TAG of 210,000 mt as a means of maintaining the
fishing mortality at a constant level. From Figure 2a,
the perceived risk of the fishing mortality exceeding
the target mortality is ~14%. The expected value of
the potential yield foregone for this TAG is ~14,000
mt. If, instead, a TAG of 215,000mt is selected, the
risk of exceeding the target fishing mortality becomes
26% and the expected value of the potential yield
foregone becomes 10,000 mt. Thus, an increase in the
TAG of 5,000mt would almost double the risk of ex-
ceeding Fstatus quo while reducing the expected poten-
tial yield foregone by 30%.
Another way to present the results of the SPA sim-
ulations is to plot percentiles of output distributions
versus the TAG selected. For example, for each SPA
run on simulated data, one can take the estimated
population size and iteratively seek the fishing mor-
talities that will result in each of several TAGs. Then,
120 140 160 180 200 220 240 260
total catch (1000 mt)
Figure 3
Percentiles of the distribution of relative change (%) in biomass
of cod Gadus mo7-hua age 3 and above, as a function of the
total catch selected for 1991. Dotted lines: (top) 97.5th percen-
tile; (middle) 50th percentile; (bottom) 2.5th percentile.
for any value of TAG one can compute the median and
2.5th and 97.5th percentiles of the distribution of
fishing mortalities. Since instantaneous fishing mortal-
ity may not be meaningful to some interested parties
(such as fishing industry groups), one may wish to look
at the distribution of changes in population biomass
associated with particular choices of the TAG (Fig. 3).
Thus, we have two approaches which we can sum-
marize as follows. The first approach is to select a goal
or objective (such as Fgtatusquo or Fo i) and then quan-
tify the chances of achieving that goal as a function of
the TAG or effort restriction selected. The second is
to quantify the consequences of choosing different
quotas or effort restrictions. Both approaches are
useful to managers. A manager might first ask how a
specific management objective like Fo i can be met. A
graph similar to Figure 2a makes it clear that the trade-
off between risks and costs must be balanced. The
manager might also want to know the consequences
of picking particular quotas or effort restrictions. For
example, for economic or political reasons, it may be
difficult to stick with a management policy if a large
quota reduction is called for. In this case, the conse-
quences to the stock of maintaining the status quo or
reducing the quota by various intermediate amounts
may be of interest. A graph similar to Figure 3 may
be helpful for this.
Managers and industry have a strong interest in
maintaining stability in a fishery. Conflicts can easily
arise when annual assessments provide only point
estimates of the quota required to achieve a specified
goal. This is because random error in the estimates
Restrepo et al : Monte Carlo simulation applied to Xiphias gladius and Cadus morhua
741
implies that annual adjustments in the quota will be
proscribed even when no changes are in fact necessary.
Instead of letting the quota "float" from year to year,
one can stabilize the quota and let the risks float from
year to year. Thus, as long as the risks remain within
certain limits, there is no need to adjust the quota.
(Here, the risks can include potential stock collapse as
well as foregone potential yield.)
The sequential population
analysis model: ADAPT
The examples presented below use data from two very
different fisheries that are assessed with the same SPA
approach, known as ADAPT (Gavaris 1988). ADAPT
is widely used in the eastern United States and Canada.
Here we describe the basic method briefly and direct
the interested reader to more details in Parrack (1986),
Gavaris (1988), Conser and Powers (1990), and Powers
and Restrepo (1992).
The objective in ADAPT is to minimize deviations
between observed (age-specific) indices of abundance
and those predicted by what is commonly referred to
as virtual population analysis (VPA). Let the subscripts
t, a, and i denote time, age, and abundance-index
sequence number, respectively. The basic equations
governing the model are
Nat = Na,i,t,, ez«, (1)
Cat = Fat Na^i.t.i (eZa,-i)/Z,t, and (2)
I.t = q, Na, (l-e-Z3,)/Zat, (3)
where N = stock size in numbers of fish, C = catch
in numbers, I = index of relative stock abundance (each
index is associated with one or more ages which must
be specified), F = instantaneous fishing mortality rate,
Z = total instantaneous mortality rate (Z = F -i- M), and
q = coefficient of proportionality between relative
abundance and absolute abundance. Inputs to the
model are the catch, natural mortality, and relative
abundance indices. Given that there are T years of data,
A ages, and Y indices, a search algorithm, e.g., Mar-
quardt-Levenberg (Seber and Wild 1989), is used to
estimate the parameters q^ (i = 1 . . . Y) and Na, t + 1 (a =
2 ... A) that minimize the weighted residual sum of
squares:
RSS = min2:.I.A,(I,-i;)2, (4)
where the weights, Aj, may be input or estimated via
iteratively-reweighted least squares.
ADAPT, like other VPA calibration procedures, re-
quires model constraints in order to reduce the number
of parameters. Hence, the stock sizes for the last age
each year are not normally estimated but are instead
derived from a specified relationship between Fai and
Fa-i, f Additional constraints may be required when
the amount of relative-abundance data does not sup-
port the estimation of a large number of parameters.
Often, as in the swordfish example below, this involves
estimating the relative selectivities of the various age-
groups in year T in some fashion external to the calibra-
tion process. This leads us to add to our explanation
the notion that ADAPT is generally thought of as a
framework rather than a rigid model. Thus the reader
is likely to encounter applications that deviate from the
model in equations (1) through (3). For example, for
swordfish, A is a "plus" group consisting of ages A and
older. For cod in Atlantic Canada, the objective func-
tion (4) is modified to allow for lognormal errors. A
detailed presentation of some of the most commonly
used options in ADAPT can be found in Powers and
Restrepo (1992).
Assessment uncertainty: Application
to North Atlantic swordfish
Swordfish in the North Atlantic Ocean are assessed by
the International Commission for the Conservation of
Atlantic Tunas (ICCAT). Interest is centered on the
level of fishing mortality relative to reference values
(e.g., Fmax)i and on trends in mortality and stock
abundance. Potential management options involve
restrictions aimed at controlling fishing mortality. The
assessment procedure is continually changing as ex-
perience is gained. The procedure below was used for
the 1989 assessment (ICCAT 1990).
Assessment procedure
Nine age-groups were recognized in the commercial
catch, ages 1 to 9-1- . There were 11 years of catch-at-
age data from 1978 to 1988. Fleets from the United
States, Japan, and Spain accounted for most of the
catch. Eleven abundance indices were available based
on fleet-specific catch rates from the longline fisheries
(ICCAT 1990).
Details of this assessment of the stock are presented
in ICCAT (1990). Briefly, the procedure used was as
follows. (1) A separable virtual population analysis,
SVPA (Pope and Shepherd 1982), was computed in
order to obtain estimates of the age-effects or partial
recruitment in the last year for which data were avail-
able. Data from 1983 to 1988 were used for this under
742
Fishery Bulletin 90(4). 1992
the assumption that the selectivity pattern remained
stable during that period. For that analysis the terminal
fishing mortality was set to 0.2/yr and selectivity for
the oldest age group was 3.0. (2) The ADAPT ap-
proach to sequential population analysis was then
used for calibration, with each abundance index used
separately. A weighting factor for each index was ob-
tained by setting the weight for the ith index equal to
the reciprocal of the mean squared error after
calibrating with the index. In performing the calibra-
tions, ages 5 and above were assumed to be fully
recruited (8^=1.0 for a = 5,6,. . .,9 -i-) and the partial
recruitment for the other ages was as determined from
the SVPA (i.e., from step 1). (3) The set of weights
computed for the abundance indices were then rescaled
so that they summed to unity. (4) The weights were
then used in recalibrating the ADAPT SPA using all
of the abundance indices at once. In doing so, the
following constraints were used: Si was taken from
the separable virtual population analysis, S2 through
S5 were directly estimated through calibration, and Se
through 89+ were set equal to the estimated 85. The
objective function used in the calibration was to mini-
mize the weighted sum of the squared deviations from
the predicted abundance indices as in equation (4).
After the fishing mortalities and population sizes
were computed by the sequential population analysis,
the values of F„^s.\ ^•nd Fqi were calculated from yield-
per-recruit computations. Data from the terminal year
(i.e., the most recent year available) were used to pro-
ject the catch in the current year and then project the
catch for the next year. For this, recruitment in the
current year and the following year were assumed by
ICCAT to be equal to the long-term mean recruitment
obtained in the sequential population analysis. The pro-
jections were made for a variety of fishing mortalities,
specifically Fqi, F^^ax- and Fstatusquo-
Specification of uncertainty in the inputs
One thousand simulated data sets were analyzed using
a version of ADAPT written in FORTRAN 77 (avail-
able from the authors). The formulation of the problem
was made to mimic the 1989 ICCAT assessment for
North Atlantic swordfish. However, we emphasize that
the uncertainties in the inputs specified below are our
ad hoc choices and, although realistic, are intended
mainly for illustrative purposes.
Natural mortality Uncertainty in the natural mor-
tality rate (M) was specified as a uniformly-distributed
random variable in the interval 0.1-0.3/yr. The value
of 0.2 used by ICCAT (1990) is at the center of this
range, and the choice of a uniform distribution places
equal confidence in all values in the interval.
^
■o
"^e ^ ,
L^
Figure 4
Coefficients of variation of outputs from the sequential popula-
tion analyses of simulated swordfish Xiphias gladius data
sets, (a) Age- and year-specific estimates of population num-
bers; (b) age- and year-specific estimates of instantaneous
fishing mortality.
Catch-at-age Total annual catches were represented
by lognormally-distributed random variables with coef-
ficients of variation of 10% and expected values equal
to those in the assessment. A coefficient of variation
of 10% indicates that the catches are known with high
precision. The proportions of the total catch in any year
that make up each age component were assumed to
follow a multinomial distribution with expected values
Restrepo et al : Monte Carlo simulation applied to Xiphias gladius and Cadus morhua
743
82 83 84
Year
Figure 5
Distribution of swordfish Xiphias gladius recruitment esti-
mates, by year, from the sequential population analyses. Outer
lines are from the Monte Carlo simulations and show 95% con-
fidence bands (determined as the 2.5th and 97.5th percentiles
of the distribution resulting from 1000 simulations). Inner pair
of lines shows confidence bands obtained from the informa-
tion matrix after a single run of the ADAPT program using
actual data. Line with symbols gives median estimate for each
year from the simulations.
equal to the observed proportions and sample size equal
to 1% of the annual catch. This model for the uncer-
tainty was purely heuristic rather than based on mea-
sured variances.
Abundance (CPUE) indices The 11 available indices
from the longline fisheries were also assumed to be
lognormally distributed with a coefficient of variation
of 10%. We chose a value of 10% as a rough approx-
imation for all indices in all years. However, there is
no reason why each index could not have a different
coefficient of variation for each year depending on the
amount of data available.
Results of swordfish simulations
The simulations gave rise to 1000 sets of age- and year-
specific fishing mortality rates and population sizes. We
computed the coefficient of variation of these sets of
estimates for each age-year combination (Figs. 4a, b).
As expected, the coefficients of variation were highest
in the most recent year, 1988. Also, the age groups
which form the bulk of the catch (ages 3-5) were the
best determined. It is interesting to note that the coef-
ficients of variation of fishing mortality rates for ages
8 and 9 were consistently lower than those for pre-
ceding ages. This is due to the manner in which the
estimates for ages 8 and 9 were determined: it was
assumed that Fgt = Fgj (subscripts refer to age and
year, respectively), and these were computed as a
weighted average of fishing mortalities for ages 5-7.
Thus, the uncertainty in the estimates of fishing mor-
tality for the last two age-groups is solely a function
of the uncertainties in the estimates for ages 5-7. This
underscores the fact that the simulation results are
conditional not only on the input-uncertainty distribu-
tions but on the formulation of the model being fitted
as well.
The median recruitment (age 1) from the simulations
increased over time (Fig. 5). However, the 95% con-
fidence bands, defined by the 2.5th and 97.5th percen-
tiles of the 1000 estimates, are quite wide. The con-
fidence bands provided by the delta method for a single
run with the actual data are much narrower than the
ones obtained by the Monte Carlo approach. The
former confidence bands indicate there is no uncertain-
ty in the results for the converged part of the SPA in
contrast to the simulation results. This is because the
delta method results, based on the information matrix
of a single run, are conditional on the natural mortal-
ity rate, catch at age, etc., being known exactly where-
as the simulation accounts for uncertainty in these
inputs. For this reason, we believe the simulation
results are more reasonable.
Note that there appears to be very little interannual
recruitment variability in the time-series (Fig. 5). This
is probably due to the fact that fish ages were estimated
from lengths deterministically by inverting the
Gompertz growth equation, and this tends to blur the
age-groups.
The population of fish age 5 and above appears to
have declined rather steadily over time while the
weighted fishing mortality rate appears to have in-
creased (medians. Figs. 6a, b). Here, weighted fishing
mortality is defined as the mean of the fishing mortality
estimates for ages 5 through 9 + , computed with
weights proportional to the estimated population size
at age. Again, the confidence bands are very wide.
It should be noted that for each run the estimates
of fishing mortality. Fat, and population size, Nat, are
highly correlated not only with each other but also with
the value of natural mortality, M, used in the simula-
tion run. For this reason, it is appropriate to examine
trends in an estimated quantity one run at a time. We
computed the ratio of the weighted fishing mortality
in a given year t to the weighted F in the base year
(taken to be 1978 in this example) for each simulation
run (Fig. 7). The distribution of the fishing mortality
ratio in 1979 was centered around 1.0; the ratio in 1986,
1987, and 1988 was >1.0 in 100% of the runs, thus
clearly indicating that fishing mortality has increased.
This result is not obvious from examination of Figure
744
Fishery Bulletin 90(4). 1992
82 83 84
Year
Figure 6
Medians, 2.5th percentiles, and 97.5th percentiles of the out-
put distributions from the Monte Carlo simulations, (a)
Distribution of estimates of the population size of swordfish
Xiphius gladius aged 5 and above; (b) distribution of estimates
of fishing mortality for swordfish aged 5 and above.
6b and illustrates how very easily the Monte Carlo
approach lends itself to hypothesis testing.
Of course, the goals of an assessment are not re-
stricted to estimating population sizes and mortality
rates. Interest is often centered on catch projections
and quotas, effort regulations, and risk analyses. For
swordfish assessments, it is useful to contrast the
estimated current level of fishing mortality against
reference points such as Fq.i and Fmax- The uncertain-
ty in such comparisons (e.g., the ratio of current F to
Fo i) can easily be quantified using the Monte Carlo
procedure. For each simulation run, we computed the
multiplier that would be necessary to bring the esti-
mated vector of age-specific fishing mortalities in the
terminal year to the Fq.i and F^^^x levels (Fig. 8). For
the computations, we used the run-specific natural
3.5-
3-
2.5
1
■i 1.5-
LL
1-
0.5-
78 79 80 81 82 83 84 85 86 87 88
Year
Figure 7
Distribution of the ratio of swordfish Xiphias gladius fishing
mortality in year y to that in 1978 as a function of the year.
Vertical bars indicate 95% confidence intervals based on
percentiles; horizontal bars represent the median ratio.
400-
200
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5
Relative Change in Current F to Reach:
I Fmax j
I F0.1
Figure 8
Multipliers necessary to bring the vector of age-specific fishing
mortalities in the terminal year to the F^ ^ and F„„ levels,
for 1000 simulated data sets for swordfish Xiphias gladius.
mortality rate and the weight-at-age relationships used
by ICCAT in the 1989 assessment. No uncertainty was
specified for weight relationships although this could
easily be added if appropriate information were avail-
able. From Figure 8 it is evident that, to achieve the
Fo.i goal, fishing mortality must be cut to ~25% of its
current value. With respect to F^^^, it appears that
fishing mortality must be cut by ~50% (Fig. 8). Note,
however, that this conclusion is considerably less
Restrepo et al.: Monte Carlo simulation applied to Xiphias gladius and Gadus morhua
745
350
300
250-
& 200
u
2 150-
100
50-
0
r
U
p n n n
13 17 21 25 29
Projected Yield (1000 MT)
33
cnia
I 1990
Figure 9
Distribution of 1989 estimated swordfish Xiphias gladius
catches when fishing mortality is kept the same as in 1988
(open bars), and distribution of 1990 estimated catches when
fishing mortahty is equal to the midpoint between the 1988
fishing mortality and Fj ,, assuming 1989 fishing mortality
was the same as in 1988 (cross-hatched bars).
certain than that for Fq.i as evidenced by the fact that
the distribution of multipliers is broader for F^ax than
for Foi- But, as an anonymous reviewer pointed out,
it is interesting to note that the mode of both distribu-
tions is about one-third of the status quo F.
We also computed 1000 projected catches in weight
for 1989 with fishing mortality equal to that in 1988.
We then projected the catch for 1990 with fishing mor-
tality set at the midpoint between the fishing mortal-
ity in 1988 and F„.i (Fig. 9). This method gradually
reduces fishing mortality to minimize the short-term
impact of decreased landings on fishermen (see
Pelletier and Laurec 1990, for a discussion). Recruit-
ments for 1989 and 1990 were drawn randomly from
the empirical distribution of recruitments estimated
from 1978-87 on each iteration. If the fishing mortal-
ity does not change in 1989 from the level in 1988,
catches are likely to be somewhere around the 1988
yield of ~ 18,000 mt. The 1990 yields are likely to be
~ll,000-13,000mt.
Using the Monte Carlo results, it is equally simple
to obtain distributions of catches for fishing at other
exploitation levels or to obtain distributions of fishing
mortalities for fixed catch quotas. Similarly, the
distribution of other projected variables, such as the
spawning-potential ratio that results from various
catch and fishing mortality options, can be computed.
In doing so, it is important to have the values of the
inputs used in calibrating the SPAs (e.g., natural mor-
tality) stored in each iteration, so that the projection
computations use the same values.
Risks and costs: Application
to northern cod
We studied the cod fishery in NAFO Divisions 2 J + 3KL
and based our simulations on the data and methods
described in Baird et al. (1990). Additional data,
described below, were obtained from the files at the
Northwest Atlantic Fisheries Centre, St. John's, New-
foundland. The simulations reflect our owti perceptions
and experience about the sources and nature of the
uncertainties in the assessment. As with the swordfish
example, the selection of management objectives for
simulation was made for illustrative purposes.
This cod fishery is managed by quota. The assess-
ment uses trawl-survey data and commercial catch-rate
data to calibrate the SPA.
Assessment and simulation procedures
Only a brief description of the assessment procedure
is given here since the details are not important for
understanding the use of the simulation method. The
catch-at-age data for ages 3-13 for each year from 1978
to 1989 were taken from Table 7 of Baird et al. (1990).
Coefficients of variation of these catch estimates were
computed using the method of Gavaris and Gavaris
(1983); these coefficients were available in the files. The
coefficients of variation ranged from 2 to 17%. Age-
and year-specific catch rates from research-vessel
surveys for the period 1978-89 and associated coeffi-
cients of variation (Baird et al. 1990, table 23) were
used to tune the sequential population analysis. The
coefficients of variation were <30% in 87% of the
cases. Age- and year-specific catch rates from the off-
shore commercial trawl fishery for ages 5-8 for the
period 1983-89 were standardized by the method of
Gavaris (1980) for use as an index of abundance for
tuning the SPA (Baird et al. 1990, table 39). We
developed estimates of the coefficients of variation for
the commercial catch-rate indices. In all cases, these
were close to 10%. Natural mortality for this stock is
believed to be around 0.2/yr.
In the simulations, the point estimates of the inputs
were replaced by random variables with the same ex-
pected values and coefficients of variation as specified
above. Catch at age values were generated as normal
random variables, while the research-vessel and the
commercial catch rates were generated as lognormal
random variables. The value of the natural mortality
rate was generated as a uniform random number be-
tween 0.15 and 0.25/yr.
746
Fishery Bulletin 90(4), 1992
The specific formulation of the ADAPT model was
as follows. The research-vessel indices were obtained
in the fall and were assumed to represent population
size at the end of November. The commercial catch-
rate indices were assumed to represent population size
at the beginning of the year. The fishing mortality F
for the oldest age-group (13) was calculated as 50% of
the mean F for ages 7-9 weighted by population
number at age. The objective function to be minimized
differed from equation (4) in that lognormal errors were
assumed and the weights, Aj, were fixed to be 1.0.
Projections for 1990 and 1991 were made using the
same procedures used in the most recent annual assess-
ment (Baird et al. 1990). Population and fishing mor-
tality projections for 1990 were made by randomly
selecting a value for recruitment from the historical
set of estimated recruitments and assuming that (1)
the total catch in 1990 is 225,000 mt (the fixed Cana-
dian quota in place when the assessment was done in
1990, plus an additional 25,000 mt in expected foreign
catch), and (2) the partial recruitment (selectivity)
vector for 1990 is equal to that estimated for 1989 in
each simulated SPA.
Catch projections for 1991 were made in two ways.
In one, we set the fishing mortality for 1991 equal to
that for 1990 and solved for the catch. In the other,
we set the fishing mortality for 1991 equal to
min{(Fo.i+Fi99o)/2, 2 Fd}.
This is the 50% rule formulated by the Canadian
Atlantic Fisheries Scientific Advisory Committee
(Canada Department of Fisheries and Oceans 1991) for
a gradual movement towards Fq.i . We also computed
the fate of yield foregone and the distribution of popula-
tion changes for various choices of the total catch.
Results of cod simulations
We generated risk curves for two fishing mortality
objectives for 1991 (Figs. 2a, b). These curves can
be put in perspective by noting that the Canadian total
allowable catch for 1990 was 199,262 mt while the
total catch (Canadian plus international) may have been
as high as 235,000 mt. To have a 50% risk of increas-
ing the fishing mortality in 1991 over the 1990 level,
one would set the total catch at 225,500 mt; to have
a 50% chance of exceeding the fishing mortality asso-
ciated with the 50% rule would entail setting the total
catch at 1 63,000 mt. It appears that a cut in the TAC
would be necessary to have a reasonable chance of
preventing the fishing mortality from exceeding the
1990 value. Substantial cuts in the harvest would be
required to ensure a high probability of meeting the
50% rule.
For values of the TAC for which the risk is less than
25%, the expected value of the yield foregone is ap-
proximately a linear function of the TAC (Figs. 2a, b).
That is, for every change in the TAC of 1000 mt, the
expected yield foregone changes by ~1000mt. The fate
of biomass left in the water is to increase by ~13% in
a year (mean of 1000 simulations = median = 12.9%;
95% confidence band based on 2.5th and 97.5th percen-
tiles is 7.2% andl8.4%). The relative change in biomass
of fish aged 3 and above is also a linear function of the
TAC (Fig. 3). Note, however, that the relative change
in biomass cannot be determined very precisely as
evidenced by the wide confidence bands.
We presented results of catch projections for two
scenarios. Often, one might like to examine a larger
number of options. For example, if current fishing mor-
tality exceeds F^^, then one could explore various
ways to reduce fishing mortality in gradual steps as
well as exploring the consequences of various types of
"status quo" options. The simulation approach is ver-
satile enough to handle fixed catch, fishing mortality,
and biomass objectives, as well as objectives involving
relative change. Thus, one could have any of the follow-
ing objectives for fishing mortality: achieve F = 0.40/
yr, achieve F = Fo.i, reduce F by 40%, or adjust F so
that biomass changes a given fixed or relative amount.
In some fisheries, catch and population projections
may be highly dependent on the assumptions made
about recruitment. When this is the case, it may be
helpful to quantify the uncertainty separately for
various segments of the population. For example, we
computed the distribution of relative change in age 3 -t-
biomass of cod (from 1989 to 1991) for various choices
of the TAC. The wide confidence bands (Fig. 3) reflect
the large uncertainty in future recruitment. We could
have quantified the relative change in the biomass of
age 5 -I- fish. From the ADAPT run based on 1989 data,
we already have an estimate of age-3 biomass in 1989.
This biomass can be projected forward to age 5 in 1991;
hence, we do not need to generate a random value for
recruitment. The uncertainty in the biomass of age 5 -t-
fish should thus be smaller than the uncertainty in age
3 + biomass. Unfortunately, the latter quantity may be
of greater interest.
Conclusions
Monte Carlo simulation has long been regarded as a
very useful quantitative tool, especially for sensitivity
analysis (e.g., Pope and Gray 1983, Rivard 1983). It
is also quite useful for studying the properties of
specific assessment procedures (e.g., Mohn 1983,
Kimura 1989). Here, we follow Francis (1991) and use
it to quantify the risks of not meeting the objectives
Restrepo et al : Monte Carlo simulation applied to Xiphias gladius and Gadus morhua 747
for the fishery as a function of the management mea-
sures imposed. The simulation approach we present can
be used with assessment models other than ADAPT.
For example, one could use Monte Carlo simulation to
quantify the effects of uncertainty in input data, as-
sumptions, and model formulation on the outputs from
the CAGEAN (Deriso et al. 1985) or stock synthesis
(Methot 1990) methods. We believe that this simula-
tion framework is not only a versatile and intuitive
method to estimate uncertainty, risks, and costs, but
in many cases it may also be the only practical way to
incorporate some types of input uncertainty which are
not estimated statistically. Because the estimated
uncertainties in the model outputs are conditional on
what is known and what is assumed about the inputs,
failure to acknowledge possible sources of uncertain-
ty in a realistic manner may lead to overly optimistic
views of the uncertainties in the model outputs. The
Monte Carlo approach forces one to examine the nature
and magnitudes of the uncertainties in the inputs and
in the model formulation, and it allows one to study
how uncertainties are propagated through the assess-
ment and into the projections ultimately used for
management recommendations.
It appears feasible to quantify risks and costs for a
wide variety of management options when the assess-
ments are accomplished by any of a variety of analytical
models. It remains to determine what risks (and costs)
should be quantified, how much risk is acceptable, and
for how long. For example, we do not know how to
quantify the risk of stock collapse due to recruitment
failure, but we might wish to quantify the risk of the
spawning biomass falling below 20% of the virgin level
in three years out of five. If we assume that this
represents a dangerous situation (see Beddington and
Cooke 1983, Brown 1990, and Goodyear 1990, for
thoughtful discussions), then the risk should be kept
low. On the other hand, if we consider the risk of ex-
ceeding the economically-optimal fishing mortality
(however defined), then we might like the risk to be
close to 50%, i.e., as likely to be above the optimum
as below it. (Of course, we should consider the relative
costs of over- and undershooting the target mortality).
If F is not close to the economic optimum fishing mor-
tality, then one must also devise a way to determine
what is the best trajectory to take for arriving at the
long-term goal. It is beyond the scope of this paper to
address what are appropriate goals, biological refer-
ence points, and trajectories.
Finally, for any stock assessment, the results of a
Monte Carlo simulation study are necessarily condi-
tional on what is assumed about the sources of uncer-
tainty, including the model chosen for the assessment.
Since decisions about some of the sources of uncertain-
ty are subjective, the results are personal views of
uncertainty, risk, cost, etc. If three scientists assess
a given stock, they can generate three separate sets
of simulation outputs. The combination of their simula-
tions provides a picture of their collective uncertainty
about the assessment results. Alternatively, they can
agree that a minimal estimate of the uncertainty is
provided by the one set of results that are the least
uncertain.
A more detailed study of the relative sensitivities of
the assessment outputs and risk curves to the choice
of input distributions can be carried out via sensitivity
analysis (Miller 1974). In this Monte Carlo framework,
sensitivity analysis would consist of introducing plann-
ed perturbations to the input-uncertainty distributions
and then measuring the overall effect on the model's
outputs. This should aid in the identification of key in-
puts so that more effort could be placed on improving
their estimates. This is more difficult than it may seem.
A given input that is perturbed during the sensitivity
analysis (say, catch at age) will cause different degrees
of change in the various output distributions: stock
sizes, fishing mortalities, Fqi, projected catches, etc.
Furthermore, this impact may change over time. For
instance, assumptions about recruitment become very
dominant as the projections are made further ahead
in time. Nonetheless, sensitivity analysis can be very
useful in identifying trade-offs between the benefits of
precision and the cost of obtaining that precision.
Acknowledgments
Partial support for this study was provided through the
Cooperative Institute for Marine and Atmospheric
Studies by National Oceanic and Atmospheric Admin-
istration (NOAA) Cooperative Agreement NA90-RAH-
0075 and by the Canadian Government's Atlantic
Fisheries Adjustment Program (Northern Cod Science
Program). We thank Nicholas Payton for programm-
ing assistance and Peter Shelton, Al Pinhorn, Donald
Parsons, and two anonymous reviewers for helpful
comments.
Citations
Baird. J.W., C.A. Bishop, and W.B. Brodie
1990 The assessment of the cod stock in NAFO Divisions 2J,
3K and 3L. CAFSAC (Can. Atl. Fish. Sci. Adv. Comm.) Res.
Doc. 90/18.
Beddin^on, J.R., and J.G. Cooke
1983 The potential yield of fish stocks. FAO Fish. Tech. Pap.
242, 50 p.
Bergh, M.O., and D.S. Butterworth
1987 Towards rational harvesting of the South African anchovy
considering survey imprecision and recruitment variability. S.
Afr. J. Mar. Sci. 5:937-951.
748
Fishery Bulletin 90(4). 1992
Brown, B.
1990 Use of spawning stock size considerations in providing
fishery management advice in the North Atlantic— a brief
review. ICCAT (Int. Comm. Conserv. At). Tunas) Collect. Vol.
Sci. Pap. 32(2):498-506.
Canada Department of Fisheries and Oceans
1991 1991 Atlantic groundfish management plan 1990.
Ottawa, 114 p.
Conser. R.J., and J.E. Powers
1990 E.xtensions of the ADAPT tuning method designed to
facilitate assessment work on tuna and swordfish stocks.
ICCAT Collect. Vol. Sci. Pap. 32: 461-467.
Deriso, R.B.. T.J. Quinn II, and P.R. Neal
1985 Catch-age analysis with auxiliary information. Can. J.
Fish. Aquat. Sci. 42:815-824.
Efron, B.
1982 The jackknife, the bootstrap and other resampling plans.
Soc. Ind. Appl. Math, Philadelphia.
Francis, R.I.C.C.
1991 Risk analysis in fishery management. Northwest Atl.
Fish. Organ. Sci. Counc. Stud. 16:143-148.
Gavaris, S.
1980 Use of a multiplicative model to estimate catch rate and
effort from commercial data. Can. J. Fish. Aquat. Sci. 37:
2272-2275.
1988 An adaptive framework for the estimation of population
size. CAFSAC (Can. Atl. Fish. Sci. Adv. Comm.) Res. Doc.
88/29.
Gavaris, S., and C.A. Gavaris
1983 Estimation of catch at age and its variance for ground-
fish stocks in the Newfoundland region. Can. Spec. Publ. Fish.
Aquat. Sci. 66:178-182.
Goodyear, C.P.
1990 Spawning stock biomass per recruit: The biological basis
for a fisheries management tool. ICCAT Collect. Vol. Sci.
Pap. 32:487-497.
Gulland, J. A., and L.K. Boerema
1973 Scientific advice on catch levels. Fish. Bull.. U.S. 71:
325-335.
ICCAT (International Commission for the Conservation of
Atlantic Tunas)
1990 Appendix 10 to Annex 8. In Report for biennial period,
1988-89, Part II (1989). Madrid, Spain [Engl, version].
Kendall, M., and A. Stuart
1977 The advanced theory of statistics, vol. 1, 4th ed. Mac-
millan, NY.
Kimura, D.K.
1989 Variability, tuning, and simulation for the Doubleday-
Deriso catch-at-age model. Can. J. Fish. Aquat. Sci. 46:
941-949.
Methot, R.D.
1990 Synthesis model: An adaptable framework for analysis
of diverse stock assessment data. Int. N. Pac. Fish. Comm.
Bull. 50:259-277.
Miller, D.R.
1974 Sensitivity analysis and validation of simulation models.
J. Theor. Biol. 48:345-360.
Mohn, R.K.
1983 Effects of error in catch and effort data on tuning cohort
analysis, with a postscript on logistic production models. Can.
Spec. Publ. Fish. Aquat. Sci. 66:141-150.
Parrack, M.L.
1986 A method of analyzing catches and abundance indices
from a fishery. ICCAT (Int. Comm. Conserv. Atl. Tunas) Col-
lect. Vol. Sci. Pap. 24:209-221.
Pelletier, D., and A. Laurec
1990 Toward more efficient adaptive TAC policies with error-
prone data. Northwest Atl. Fish. Organ. SCR Doc. 90/100
(ser. N1835).
Pope, J.G.
1972 An investigation of the accuracy of virtual population
analysis using cohort analysis. Int. Comm. Northwest Atl.
Fish. Res. Bull. 9:65-74.
1983 Analogies to the status quo TACs: Their nature and
variance. Can. Spec. Publ. Fish. Aquat. Sci. 66:99-113.
Pope, J.G., and D.J. Garrod
1975 Sources of error in catch and effort quota regulations with
particular reference to variations in the catchability coeffi-
cient. Int. Comm. Northwest Atl. Fish. Res. Bull. 11:17-30.
Pope, J.G, and D. Gray
1983 An investigation of the relationship between the preci-
sion of assessment data and the precision of total allowable
catches. Can. Spec. Publ. Fish. Aquat. Sci. 66:151-157.
Pope, J.G., and J.G. Shepherd
1982 A simple method for the consistent interpretation of
catch-at-age data. J. Cons. Cons. Int. Explor. Mer 40:176-184.
Powers, J.E., and V.R. Restrepo
1992 Additional options for age-sequenced analysis. ICCAT
(Int. Comm. Conserv. Atl. Tunas) Collect. Vol. Sci. Pap. 39:
540-5.53.
Prager, M.H., and A. MacCall
1988 Sensitivities and variances of virtual population analysis
as applied to the mackerel, Scomber japonicus. Can. J. Fish.
Aquat. Sci. 45:539-547.
Restrepo, V.R., and W.W. Fox Jr.
1988 Parameter uncertainty and simple yield-per-recruit anal-
ysis. Trans. Am. Fish. Soc. 117:282-289.
Rivard, D.
1983 Effects of systematic, analytical, and sampling errors on
catch estimates: A sensitivity analysis. Can. Spec. Publ. Fish.
Aquat. Sci. 66:114-129.
Saila, S.B., E. Lorda, and H.A. Walker
1985 The analysis of parameter error propagation in simple
fishery models. Mar. Res. Econ. l(3):235-24(i.
Sampson, D.B.
1987 Variance estimators for virtual population analysis. J.
Cons. Cons. Int. Explor. Mer 43:149-158.
Seber, G.A.F.. and C.J. Wild
1989 Nonlinear regression. John Wiley, NY.
Abstract.— Sei whales Balenop-
tera horealis are noted for major
fluctuations in distribution, often in
response to local availability of prey.
An influx of sei whales occurred in
the southern Gulf of Maine during
summer 1986. Forty-seven individ-
uals (including four mothers with
calves) were photographically iden-
tified using natural markings, in-
cluding dorsal-fin notches, placement
of small circular scars on the ani-
mal's flank, and natural variation in
dorsal-fin shape and pigment swaths
along the dorsal surface behind
the blowholes. Seventeen of these
whales (36.1%) were photographed
on more than one day, and the period
between first and last sighting of
individuals ranged from one to 66
days. Only six animals were sighted
in more than one region in the south-
ern Gulf of Maine. Observed behav-
ior included traveling, nearsurface
skim feeding, lunge feeding, and
(rarely) "milling" or breaching. Group
sizes were small and variable. Two
individuals were matched to photo-
graphs taken in other regions in or
near offshore Gulf of Maine waters.
We hypothesize that the southern
Gulf of Maine represents a short-
term feeding site. The occurrence of
individuals without sufficient marks
for individual recognition suggests
that photoidentification is of limited
value in the study of this species.
Behavior of individually- identified
sei whales Balaenoptera borealis
dun'ng an episodic influx into the
southern Gulf of Maine in 1986
Mark R. Schilling
Cetacean Research Unit, P.O. Box 159, Gloucester, Massachusetts 01930
Irene Seipt
Cetacean Research Program, Center for Coastal Studies
Provincetown, Massachusetts 02657
Mason T. Weinrlch*
Cetacean Research Unit, P.O. Box 159. Gloucester, Massachusetts 01930
Steven E. Frohock
Atlantic Cetacean Research Center, P.O Box 1413, Gloucester, Massachusetts 01930
Anne E. Kuhlberg
Cetacean Research Unit, PO Box 159, Gloucester, Massachusetts 01930
Phillip J. Clapham
Cetacean Research Program, Center for Coastal Studies
Provincetown, Massachusetts 02657
Department of Zoology, University of Aberdeen, Aberdeen AB9 2TN, Scotland
Most species of baleen whales under-
take seasonal migrations between
high-latitude feeding grounds and
warmer breeding areas (Kellogg 1929,
Slijper 1962, Mackintosh 1965). Pop-
ulations often show annual variations
in local spatial distribution within
these areas (Wursig et al. 1985).
While the factors causing these varia-
tions are not well defined for breed-
ing grounds, it has been suggested
that they are explained on feeding
grounds by differences in prey distri-
bution (Whitehead and Carscadden
1988, Payne et al. 1990). Because of
the energetic demands upon large
whales (Lockyer 1981) and the unpre-
dictable distribution of their prey,
such areal variation would be ex-
pected if the animals were seeking to
Manuscript accepted 5 July 1992.
Fishery Bulletin, U.S. 90:749-755 (1992).
* Author to whom reprint requests should be
sent.
maximize their feeding efficiency.
Sei whales Balaenoptera borealis
have been reported to have greater
variation in distribution on their feed-
ing grounds than most baleen whale
species (Horwood 1987). Sei whales
have been reported in considerable
numbers for brief periods outside of
their regular range in Norway (1885,
1898, and 1919), Finland (1885), and
Scotland (1906) (Tomilin 1957, Jons-
gard and Darling 1977). Ingebrigtsen
(1929) reported large annual changes
in distribution off the Faroe Islands.
These changes are hypothesized to be
related to local increases in plank-
tonic productivity (International
Whaling Commission 1977, Horwood
1987).
Sei whales off the northeastern
United States and southeastern Can-
ada have been little studied. Mitchell
and Chapman (1977) hypothesized
749
750
Fishery Bulletin 90(4)^ 1992
the existence of a "stock" of sei whales centered
around Nova Scotia. During the spring these animals
are thought to occur on the southern edge of George's
Bank (Mitchell and Chapman 1977, CETAP 1982). Dur-
ing June and July, they move north to the southern
Scotian Shelf, then onto Brown's, Bacarro, and Rose-
way Bank from August to October (Sutcliffe and Brodie
1977). The lack of sightings in these areas, plus late-
winter/early-spring strandings in South Carolina, Loui-
siana, and Mississippi suggest a southward movement
after October (Mead 1977). The inshore waters of the
southern Gulf of Maine are rarely used by sei whales
(CETAP 1982, Payne et al. 1990).
In this paper, we report on the photoidentification,
occupancy patterns, surface behavior, and social
behavior of individual sei whales found in the Gulf of
Maine during an unexpected summer influx of this
species in 1986, documented through daily shipboard
surveys. We also report results of photographic iden-
tification of individual sei whales, and evaluate the
feasibility of such techniques for investigations of this
species.
Methods
Non-systematic surveys of the southern Gulf of Maine
were conducted daily (weather permitting) from mid-
April through October on commercial whale-watching
and research vessels operating out of Provincetown
and Gloucester, Massachusetts during 1980-91. In
1986, the only year of sei whale abundance, the number
of vessels collecting data varied. There were usually
six to nine 4-hr cruises daily from each port. Vessels
were 18-30 m long diesel-powered whale-watching
vessels and 6.7-14 m long research vessels powered by
sail, diesel, or outboard engines. Whale-watch cruises
were typically 4-5 hr in duration, while research-vessel
cruises often lasted from dawn until dusk.
Search effort by whale-watching vessels was concen-
trated on the southern and northern edges of Stell-
wagen Bank and the southern edge of Jeffrey's Ledge
because of the concentrations of whales there (Fig. 1).
Stellwagen Bank is a shallow glacial deposit with a sand
substrate, at 20-40m. Southern Jeffrey's Ledge has
a mean depth of 48 m, and is a mixture of sand, gravel,
and rocks. Depths surrounding both areas extend to
182 m. In both areas there is upwelling, caused by steep
topography which enriches the biological productivity
of the area, providing food for whales (Kenney and
Winn 1986).
For each sighting data included location, direction
and speed of animal movement (based on LORAN-C
readings taken every 5-10 min), environmental condi-
tions, behavioral information including respiration in-
CLOUCESTER
JEFFREY'S LEDGE
STELLWAGEN BANK
-7r, 05w
Figure 1
Study area in the Gulf of Maine.
tervals (to the nearest second) of individuals recog-
nizable by natural marks, notable non-respiratory sur-
face behavior, and associations among whales. Two or
more whales were considered associated if they were
in close proximity and consistently coordinated in the
timing and direction of their surfacings.
Cow/calf pairs were treated as single animals for an
analysis of occupancy periods (the time between first
and last sighting), assuming that the calf's movements
are determined by those of its mother. Calves were
designated based on the animal being considerably
smaller than any other animals, and on the continuous
association between two individuals for at least 30 min
or on more than one day. Personnel were experienced
in observing mother-calf pairs of humpback Megaptera
novaeangliae and fin Balaen.optera physalus whales. To
determine mean associated group size, mother and calf
pairs were counted as two individuals.
Fecal samples were scooped from the surface in a
5-gallon bucket on two days in August 1986 near north-
ern Stellwagen Bank. The material was frozen until
Schilling et al.: Behavior of Bataenoptera borealis during episodic influx
751
examination. Prey remains were identified to species
by staff of Allied Whale at the College of the Atlantic,
Bar Harbor, Maine.
High-speed (ISO 400) black-and-white film was used
in 35 mm single-lens reflex cameras equipped with
telephoto lenses (range 200-400 mm) to photograph
whales. When possible, three regions were photo-
graphed on each side of a whale: between the tip of the
snout and the blowhole, the flank between the blow-
hole and dorsal fin, and the dorsal fin.
Photos were classified as matchable or unmatchable
based on the same criteria defined by Seipt et al. (1990)
for fin whale photoidentification. Unmatchable photos
had poor focus, were not perpendicular to the whale,
or were too far from the whale to distinguish marks
clearly (generally photos taken at distances >100m
from the whale). For photos judged matchable, a whale
was classified as a unique individual based on the
presence of one or more of the following characters:
recognizable scars, a distinctive dorsal-fin shape (in-
cluding dorsal-fin notches), or detectable pigmentation
on the flank between the blowhole and dorsal fin. If
at least one of these features was not present in a
photograph, it was discounted.
Matches of individual whales were made within in-
dependent photographic collections of the Atlantic
Cetacean Research Center (Gloucester MA), the Center
for Coastal Studies (Provincetown MA), and the Ceta-
cean Research Unit (Gloucester MA) by personnel ex-
perienced in photo-identifying individual humpback or
fin whales. The independent catalogs of identified
whales from each organization were compared, with
matches being confirmed by both other groups,
resulting in a single collective catalog of identified
whales. Only matches agreed upon by all parties were
accepted.
In order to examine long-range movements of sei
whales, photographs were solicited from researchers
working in the Gulf of Maine for comparison with the
unified catalog described above. Each set of photo-
graphs obtained in different geographic areas or years
were treated separately.
Data for this study were stored on PC-based micro-
computers and statistical analyses were performed
using SPSS (1989) statistical software package for
PC's, including calculation of mean values and standard
deviations. A two-tailed <-test was used to compare
mean values for associated group sizes between dif-
ferent parts of the study area, and a x" test was used
to test potential differences in group sizes when a
mother-calf pair was present in the group (Zar 1984).
Results
Photoidentification
A total of 240 sei whale sightings took place between
29 June and 20 September 1986. Photographs were
taken on 182 sightings (75.8%). In 51 photographed
sightings (28.0% of all photographed sightings) the
animal could not be reliably identified because of poor
photographic quality.
Photoidentification of some individual sei whales was
possible using variation in dorsal-fin shape, placement
of small circular scars on either flank, and light pig-
ment swaths behind the dorsal fin (Fig. 2). A total of
47 identifiable non-calf sei whales and 4 calves were
photographed; of these, 19 were identifiable based on
notches in their dorsal fin alone, 4 based on the loca-
tions of small circular scars on the flanks, and 10 based
on both distinctive dorsal fins and circular scars. No
attempt was made to photoidentify calves. Twelve
whales were identified based on dorsal-fin shape and
pigment swaths behind the blowhole. One sei whale was
missing its dorsal fin, and one had a large white scar
visible on the lower portion of its right caudal pedun-
cle. In 12 cases, animals did not have distinctive marks
by which they could be identified reliably.
Occurrence and occupancy
The 47 individual sei whales were sighted on as few
as 1 and as many as 15 separate days (x 2.4, SD 3.0)
(Fig. 3). Seventeen individuals (36.2%) were observed
on more than one day; the mean period for resighted
animals between first and last sighting was 26.8 days
(SD 24.1, median 20.0). No individual sighting record
spanned more than 66 days from first to last sighting.
Twenty-six individuals (55.3%) were initially photo-
graphed on southern Stellwagen Bank, 15 (31.9%) on
northern Stellwagen Bank, 4 (8.5%) on Jeffrey's
Ledge, 1 (2.1%) in Massachusetts Bay (west of Stell-
wagen Bank), and 1 (2.1%) in the Great South Chan-
nel. Only 6 animals out of the 17 resighted were
photographed in more than one of these areas. Of
these, four were first seen on southern Stellwagen and
subsequently moved north on the Bank; one animal
showed the reverse pattern. The remaining one was
first photographed on southern Jeffrey's Ledge and
later resighted on northern Stellwagen Bank. One of
these whales moved between northern and southern
Stellwagen Bank at least five times; this individual also
had the greatest number of sightings (15) and the
longest period between first and last sighting.
752
Fishery Bulletin 90(4). 1992
Behavior and
movement patterns
A total of 752 respiration inter-
vals were recorded during 53 sei
whale sightings (from at least 15
different individuals). Inter-respir-
ation intervals showed a range of
2-928 sec, with a mean of 60.8
sec (SD 78.0 sec). The most com-
mon respiratory pattern was a
single breath followed by a short
dive of 45-90 sec. On only four
sightings did the whales sub-
merge for prolonged dives of
6-11 min, and they did so re-
peatedly within the observation.
The regular breath intervals
and lack of prolonged dives often
appeared associated with near-
surface feeding. In the most com-
mon feeding behavior, sei whales
would take a breath and then roll
45-90° around their longitudinal
axis while ~3m below the sur-
face. The mouth was often slight-
ly open as the animal swam for-
ward. There were four observa-
tions of lunge feeding, when the
whale rapidly surfaced with its
mouth opened. During lunges, no
rolling was observed. On one of
four occasions of lunging, the
same individual alternated lunges
with the more common nearsur-
face feeding behavior.
During feeding and swimming,
the whales often remained in an
area of ~0.5km2 for over an
hour. Whales would either change
swimming direction with each
breath, or travel in a straight line
for 10 min or less before revers-
ing, resulting in minimal net
movement at the surface.
Defecations were observed nine
times. Feces were bright red and contained chunks of
particulate matter. Only mandibles from the copepod
Calanus finmarchicus were subsequently identified in
fecal material.
While feeding and traveling comprised most of the
behavioral events (both time and number), milling (at
least three socializing with one another while moving
in apparently random directions, rolling, and remain-
ing on the surface continually for over 10 min) was seen
Figure 2
Photographic match of sei whale Balaenoptera borealis #22. utilizing both circular scars
and dorsal-fin notches as aids in identification. Note the pattern of circular scars below
the dorsal fin, particularly the large circular scar below and immediately posterior to
the dorsal fin, and the slightly blunted tip of the fin itself. Top photograph was taken
10 August 1986, bottom photograph on 22 August 1986.
four times. This was always associated with one whale
leaving the group either during or immediately after
the milling period. Breaching was seen once, when a
single animal breached twice in rapid succession.
Social behavior
Sei whales were seen in groups of 1—6 individuals.
Mean group size was 1.8, and was the same on both
Schilling et al,: Behavior of Balaenoptera boreahs during episodic influx
753
II I
II I M I, Ml I
I .11
I I I I I I
II II
July
August
September
DATE
Figure 3
Occurrence of individual sei whales Balaenoptera horealis in the southern Gulf of Maine
during 1986. Individual whales are ordered (top to bottom) by date of first sighting; each
row of marks represents the dates on which one individual was seen during the year.
Vertical marks represent single whales.
vidual associations were trans-
ient, generally lasting less than
24 hr.
Of 13 cow/calf pairs, 3 (23.0%)
were associated with 1 or more
other whales, while 48 (51.6%) of
the 93 non-calf groups involved
2 or more whales. Frequency of
association with another whale
was not statistically significant
between groups with and with-
out mother-calf pairs (x^ 3.7, p
0.06, 1 df). Cow/calf pairs were
never seen with more than one
associate.
Comparisons with
other data sets
S 60:64
89
75
10 20 30 10 50 60 70 80 90 100 130 160 200 300 500
Seconds Between Respirations
Figure 4
Histogram showing the distribution of breath intervals (see)
from sei whales Balaenoptera borealis recorded during 52
sightings in the 1986 season on northern Stellwagen Bank.
Note the scale change for the longer breath intervals due to
their infrequent occurrence.
northern and southern Stellwagen Banks, where most
observations took place (two-tailed i-test, p 0.93, 1 df).
Resightings of individual animals on successive days
were recorded 14 times. In cases where all members
of an associated group were photoidentified, only one
pair of animals was seen together on two days; one
whale from this pair was later sighted with a different
associate two days later. These data indicate that indi-
Sei whales identified in the south-
ern Gulf of Maine in 1986 were
compared with one 1984 photo-
graph from Georges Bank (At-
lantic Cetacean Research Center),
one from Stellwagen Bank in
1987 (Plymouth Marine Mammal
Center), one from Jeffrey's Ledge in 1988 (Cetacean
Research Unit), three from the Scotian Shelf in 1988
(Atlantic Cetacean Research Center and Nancy Miller),
and four from the Scotian Shelf in 1989 (New England
Aquarium). There were two matches. Sei whale 33 was
photographed on 11 June 1984 between Wilker's and
Oceanographer's Canyon on the southern edge of
Georges Bank (40°05'N, 68°19'W), 176nmi from the
1986 sighting; sei whale 19 was photographed on 28
August 1989 on the Nova Scotian Shelf (42°57'N,
65°09'W), 211nmi from its 1986 sighting.
Discussion
Photoidentification of individual animals has been ac-
complished for numerous baleen and toothed whales
(Katona et al. 1980, Dorsey 1983, Agler et al. 1990,
Seipt et al. 1990). Our results indicate that some sei
whales can be identified using variations in natural
markings. Dorsal-fin shape, natural pigment patterns,
and scarring were all useful features. The presence of
circular scars along the flank— hypothesized to be
caused by small sharks (Shevchenko 1977), lampreys,
or pathogenic microorganisms (Tomilin 1957, Rice
1977)— and dorsal-fin notches (unknown origin) both
facilitated identification of the individuals. While these
techniques work within a single season, the possibility
of acquiring new dorsal-fin notches, new scars, or
754
Fishery Bulletin 90(4), 1992
having scars change with age (Shevchenko 1977) might
make identification over a prolonged period difficult.
The lack of distinctive markings on some individuals
indicates that while photoidentification is useful in
studies of sei whales, it is not likely to allow identifica-
tion of all members of the population. Because we were
unable to determine the sex and/or age of the animals
involved, it is impossible to indicate whether distinc-
tive markings were related to age or sex.
The photographic matches of sei whales in the
southern Gulf of Maine to both Georges Bank and the
Scotian Shelf lend some support to the idea that whales
in the 1986 influx were from the closest-known geo-
graphic stock. However, given the dearth of knowledge
concerning the biogeography of this species, and
Brown's (1977) record of a sei whale moving 4000km
in 10 days, many or all of the animals reported here
may come from other locations in the North Atlantic.
During the year of the southern Gulf of Maine influx,
abnormally high levels of Calanus finmarchicus were
present on Stellwagen Bank (Payne et al. 1990). Our
findings therefore lend support to the hypothesis that
sei whales show annual areal fluctuations to take
maximal advantage of changes in local productivity
throughout their range. Whether each individual in-
dependently found the increased copepod productivity,
or whether social factors were involved in the influx,
remains an open question which our data do not
address.
Most of the sei whales in 1986 were seen during late
July and early August, with a secondary peak in early
September (Fig. 3). This suggests that the local pro-
ductivity provided a brief stop-over point during the
summer feeding season. Repeated resightings of a few
individuals during the study suggest that a small
number of animals found prey levels adequate to allow
a prolonged occupancy period.
Observed behavior of sei whales was similar to that
previously described (Tomilin 1957, International
Whaling Commission 1977, Horwood 1987). The rela-
tively small variation around the mean breath interval
shows a departure from the standard balaenopterid
pattern of hyperventilation, e.g., several breaths taken
in rapid succession followed by a longer diving period
(Gunther 1949, Leatherwood et al. 1976). The rolling
we observed during apparent feeding behavior differs
from that described by Tomilin (1957) who did not
observe this species roll during feeding.
Social groups observed in this study are similar to
those reported by studies of other balaenopterids, with
individuals being sighted either alone or in small groups
(Nemoto 1964, Dorsey 1983, Whitehead and Carlson
1988). Lockyer (1977) reported a mean associated
group size of 2.4, slightly higher than the 1.8 reported
here. Since she did not define an associated group,
however, direct comparisons are difficult. Further, if
there is a correlation between the associated group-size
and prey patch-size (such as that described for hump-
back whales by Whitehead 1983), it is possible that the
larger group size observed could be explained by the
more productive Antarctic waters. Our limited data
also suggest that, as in other baleen whales, cow/calf
pairs were more solitary than other animals, both in
frequency of association with other individuals and in
overall group size. This has previously been docu-
mented in humpback whales (Clapham and Mayo 1987),
gray whales Eschrichtius robustus (Swartz 1986), and
right whales Eubalaena australis (Payne 1986).
Acknowledgments
Many people helped gather field data including Cindy
Belt, Carole Carlson, Peggy Christian, Lisa Frohock,
David Mattila, Sharon Pittman, and many interns.
Polly Hamlin (CCS), Maribel Marcy (CRU), and Lisa
Frohock (ACRC) helped considerably in compiling raw
data. Scott Kraus (New England Aquarium), Nancy
Miller, Fred Wenzel, and Dave Wiley all provided
photographs of sei whales for comparison with our 1986
sightings. Dr. Steven Katona and his colleagues at the
College of the Atlantic generously took their time to
examine the fecal material. We thank the owners and
crews of the Dolphin Fleet, Cape Ann Whale Watch,
Cap't Bill and Sons Whale Watch, and Gloucester
Whale Watch for their logistical help and support.
Funding for this study came from the National Marine
Fisheries Service, the American Cetacean Society/Los
Angeles Chapter, the Essex County Ecologj' Center,
and Gloucester Whale Watch: we are indebted to
them all.
Citations
Agler. B.A., J. A. Beard, R.S. Bowman, H.D. Corbett,
S.E. Frohock, M.P. Hawvermale. S.E. Katona. S.S. Sadove,
and I.E. Seipt
1990 Finback whale. Balaenoptera physalus, photographic
identification: Methodology and preliminary results from the
western North Atlantic. Rep. Int. Whaling Comm. (Spec.
Issue 12):349-356.
Brown, S.G.
1977 Some results of sei whale marking in the Southern
Hemisphere. Rep. Int. WTialing Comm. (Spec. Issue l):39-43.
CETAP (Cetacean and Turtle Assessment Program)
1982 A characterization of marine mammals and turtles in the
mid- and north Atlantic areas of the U.S. outer continental
shelf. Final Report of the Cetacean and Turtle Assessment
Program, University of Rhode Island, to the Bureau of Land
Management, Washington D.C. Kingston, 450 p.
Schilling et al : Behavior of Balaenoptera borealis during episodic influx
755
Clapham, P.J., and C.A. Mayo
1987 Reproduction and recruitment of individually identified
humpback whales, Megaptera novaeangliae. observed in Massa-
chusetts Bay, 1979-1985. Can. J. Zool. 65:2853-2863.
Dorsey, E.M.
1983 Exclusive adjoining ranges in individually identified
minke whales in Washington state. Can. J. Zool. 61:174-181.
Gunther. E.R.
1949 Habits of the fin whale. Discovery Rep. 25:113-142.
Hammond, P.S., S.A. Mizroch, and G.P. Donovan (editors)
1990 Individual recognition of cetaceans: Use of photo-
identification and other techniques to estimate population
parameters. Rep. Int. Whaling Comm. (Spec. Issue 12), 440 p.
Horwood, J.W.
1987 The sei whale: Population biology, ecology, and manage-
ment. Croom Helm Ltd., Kent, England, 375 p.
Ingebrigtsen, A.
1929 Whales caught in the North Atlantic and other seas. Int.
Counc. Explor. Sea, Rapp. P.V. Reun. 56:1-26.
International Whaling Commission
1977 Report of the special meeting of the scientific commit-
tee on sei and Bryde's whales. Rep. Int. Whaling Comm.
(Spec. Issue l):l-9.
Jonsgard, A., and K. Darling
1977 On the biology of the eastern North Atlantic sei whale.
Rep. Int. Whaling Comm. (Spec. Issue 1):124-129.
Katona, S.K., P. Harcourt, J.S. Perkins, and S.D. Kraus
1980 Humpback whales: A catalog of individuals identified by
fluke photographs. College of the Atlantic, Bar Harbor, ME,
169 p.
Kellogg, R.
1929 What is known of the migration of some of the whalebone
whales. Smithson. Inst. Annu. Rep., 1928, Publ. 2981, p.
467-494.
Kenny, R.D., and H.E. Winn
1986 Cetacean high-use habitats of the northeast United States
continental shelf. Fish. Bull., U.S. 84:345-357.
Leatherwood, S., D.C. Caldwell, and H.E. Winn
1976 Whales, dolphins and porpoises of the western North
Atlantic: A guide to their identification. NOAA Tech. Rep.
NMFS CIRC-396, 176 p.
Lockyer, C.
1977 Some possible factors affecting age distribution of the
catch of sei whales in the Antarctic. Rep. Int. Whaling Comm.
(Spec. Issue l):63-70.
1981 Growth and energy budgets of large baleen whales from
the southern hemisphere. In FAO Advisory Committee on
Marine Resources Research, Working Party on Marine Mam-
mals (ed.). Mammals in the seas: Vol. 3. Large cetaceans, p.
379-487. FAO, Rome, 504 p.
Mackintosh, N.A.
1965 The stocks of whales. Fishing News (Books), Ltd., Lon-
don, 232 p.
Mead, J.
1977 Records of sei and Bryde's whales from the Atlantic coast
of the United States, the Gulf of Mexico, and the Caribbean.
Rep. Int. WhaMng Comm. (Spec. Issue 1): 113-1 16.
Mitchell, E.A., and D.G. Chapman
1977 Preliminary assessment of stocks of the Northwest Atlan-
tic sei whales. Rep. Int. Whaling Comm. (Spec. Issue 1):
117-120.
Nemoto, T.
1964 Schools of baleen whales in the feeding areas. Sei. Rep.
Whales Res. Inst. 18:89-110.
Payne, P.N., D. Wiley, S. Young, S. Pittman, P.J. Clapham, and
J. Jossi
1990 Recent fluctuations in the abundance of baleen whales
in the southern Gulf of Maine in relation to changes in prey
abundance. Fish. Bull., U.S. 88:687-696.
Payne, R.
1986 Long term behavioral studies of the southern right whale
(Euhalaena australis). Rep. Int. Whaling Comm. (Spec. Issue
10):161-167.
Rice, D.W.
1977 Synopsis of biological data on the sei whale and Bryde's
whale in the Eastern North Pacific. Rep. Int. Whaling Comm.
(Spec. Issue l):92-97.
Seipt, I.E., P.J. Clapham, C.A. Mayo, and M.P. Hawvermale
1990 Population characteristics of individually identified fin
whales Balaenoptera physalns in Massachusetts Bay. Fish.
Bull., U.S. 88:271-277.
Shevchenko, V.I.
1977 Application of white scars to the study of the location
and migrations of sei whale populations in Area III of the Ant-
arctic. Rep. Int. Whaling Comm. (Spec. Issue 1):130-134.
Slijper, E.J.
1962 Whales. Basic Books, NY, 475 p.
SPSS
1989 Statistical package for the social sciences (PC -h 3.1).
SPSS, Inc., Chicago.
Sutcliffe, W.H. Jr., and P.P. Brodie
1977 Whale distribution in Nova Scotia waters. Fish. Mar.
Serv. Tech. Rep. 22, Fisheries and Environment Canada, 22 p.
Swartz, S.L.
1986 Gray whale migratory, social, and breeding behavior.
Rep. Int. Whaling Comm. (Spec. Issue 8):207-229.
Tomilin, A.G.
1957 Mammals of the U.S.S.R. and adjacent countries. Vol.
9, Cetacea. Akad. Nauk SSSR, Moscow, 756 p. [in Russ.].
(Trans, by the Israel Prog. Sei. Transl., Jerusalem, 1967, 717
p.; avail. NTIS, Springfield VA, as TT-65-50086.)
Whitehead, H.
1983 Structure and stability of humpback whale groups off
Newfoundland. Can. J. Zool. 61:1391-1397.
Whitehead, H., and C. Carlson
1988 Social behavior of feeding finback whales off Newfound-
land: Comparisons with the sympatric humpback whale. Can.
J. Zool. 66:217-221.
Whitehead, H., and J.E. Carscadden
1985 Predicting inshore whale abundance— Whales and capelin
off the Newfoundland coast. Can. J. Fish. Aquat. Sei. 42:
976-981.
Wursig, B., E.M. Dorsey, M.A. Fraker, R.S. Payne, and
W.J. Richardson
1985 Behavior of bowhead whales Balaena mystice.tus summer-
ing in the Beaufort Sea: A description. Fish. Bull., U.S. 83;
357-377.
Zar, J.H.
1984 Biostatistical analysis. Prentice-Hall Inc., Englewood
Cliffs, 718 p.
Abstract.— Pelagic armorhead
Pseudopentaceros wheeleri are the
target of a directed trawl fishery on
many of the southern Emperor-
northern Hawaiian Ridge seamounts.
The population dynamics of armor-
head for the period 1970-90 were
reconstructed for Southeast Han-
cock seamount, the southernmost of
the seamounts commercially fished,
by using commercial catch-and-effort
statistics, various biological measure-
ments, and research stock-survey
data. The population declined almost
continuously from a 1972 high of
5500 metric tons (t) to a 1989 low of
25 1. In addition to the intense fish-
ery, this decline was due partly to the
sporadic pattern of armorhead re-
cruitment. Natural mortality rate
was estimated as 0.54/year; how-
ever, females had a higher mortality
rate than males.
Population dynamics of pelagic
armorhead Pseudopentaceros wheeleri
on Southeast Hancock Seamount
David A. Somerton
Honolulu Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA
2570 Dole Street, Honolulu, Hawaii 96822-2396
Present address: Alaska Fisheries Science Center, National Marine Fisheries Service, NOAA
7600 Sand Point Way NE, Seattle, Washington 981 15-0070
Bert S. Kikkawa
Honolulu Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA
2570 Dole Street, Honolulu, Hawaii 96822-2396
Manuscript accepted 5 July 1992.
Fishery Bulletin, U.S. 90:756-769 (1992).
Pelagic armorhead Pseudopentaceros
wheeleri have an unusual life history
that includes two distinct postlarval
phases: a pelagic juvenile phase, and
a demersal adult phase. During the
first 1.5-2.5 years of their lives (Uchi-
yama and Sampaga 1990), juvenile
armorhead inhabit the epipelagic
zone over a broad area of the north-
east Pacific, where they acquire large
fat reserves before migrating west-
ward to the southern Emperor-north-
ern Hawaiian Ridge (SE-NHR) sea-
mounts (Boehlert and Sasaki 1988,
Humphreys et al. 1989). After arriv-
ing at the seamounts, armorhead
mature and assume demersal habits.
Because of the rigors of spawning or
the inability to obtain sufficient prey
(Seki and Somerton, In prep.), adult
armorhead subsequently lose weight
to such an extent that they eventual-
ly become emaciated and moribund
(Humphreys et al. 1989).
Armorhead form dense nighttime
aggregations over the relatively flat
summits of the SE-NHR seamounts.
Soon after these aggregations were
discovered in 1967, they were sub-
jected to intense fishing effort first
by Soviet trawlers (Komrakov 1970)
and 2 years later by Japanese trawl-
ers (Sasaki 1986). The combined
annual catch of armorhead rapidly
increased and reached a high of
164,000 metric tons (t) in 1973
(Borets 1975, Takahashi and Sasaki
1977) before plummeting to 875 1 in
1978. This decline in catch was evi-
dently due to a decline in armorhead
abundance, because the Japanese
catch-per-unit-effort (CPUE) showed
a corresponding drop from a high of
54.0t/hr in 1972 to 0.4t/hr in 1978
(Sasaki 1986).
Although never a participant in
this fishery, the United States be-
came involved in 1976 when imple-
mentation of the Magnuson Fishery
Conservation and Management Act
extended its exclusive economic zone
(EEZ) to include the Hancock Sea-
mounts, the southernmost of the
SE-NHR seamounts supporting the
armorhead fishery. Although Soviet
trawlers ceased operations on the
Hancock seamounts after the juris-
dictional change, Japanese trawlers
continued to fish but were subject to
an annual harvest quota and were re-
quired to carry U.S. observers who
monitored the catch. Regardless of
these management efforts, catch
rates continued to decline and the
Japanese discontinued fishing on the
Hancock Seamounts in 1984. In re-
sponse to the apparent stressed con-
dition of the armorhead population,
the National Marine Fisheries Ser-
vice (NMFS) in 1985 initiated a stock
756
Somerton and Kikkawa Population dynamics of Pseudopentaceros wheelen
757
35°N-
30°-
25°
20°
%
\
■■■■\
c
SE Hancock Seamount
Colahan Smt
Kure Atoll
^ ^ Midway Island
^T-?e,„^,
atv,
9iia,
'^h
^la,
PACIFIC OCEAN
'''as
Main Hawaiian Islands
175°E
180°
175°W 170°
165"
160°
155°
Figure 1
Location and bathymetry of Southeast Hancock Seamount. Depth contours are in meters.
assessment pro-am to monitor armorhead abundance,
and in 1986 enacted a 6-yr moratorium prohibiting
trawl fishing on the Hancock Seamounts.
The population dynamics of pelagic armorhead have
been previously examined (Borets 1975, Wetherall and
Yong 1986); however, the results of these studies are
questionable because they were based on either Soviet
or Japanese catch-and-effort statistics but not both.
The present paper attempts to rectify the problem of
incomplete data by focusing solely on the stock of
armorhead inhabiting the Southeast (SE) Hancock Sea-
mount, where catch-and-effort statistics were supple-
mented with various biological and vessel performance
data after the initiation of the U.S. observer program
in 1977 and were completely replaced by research
stock-survey data when the commercial fishery ended.
With these additional data, it is possible to obtain ab-
solute estimates of armorhead abundance, rather than
relative estimates, and to extend the time-series of such
estimates beyond the termination of the commercial
fishery. In addition to developing a continuous record
of armorhead abundance since the initiation of the
fishery, this paper also includes estimates of the natural
mortality rate and annual recruitment of pelagic juve-
niles to the seamounts.
Materials and methods
SE Hancock Seamount
The Hancock seamounts consist of two peaks, North-
west (NW) and SE Hancock Seamount, separated by
61km and situated on the NHR ~293km northwest of
Kure Atoll, the northernmost of the Hawaiian Islands,
and ~287km southeast of Colahan Seamount, the
closest seamount supporting an armorhead fishery
(Fig. 1). The SE Hancock Seamount is shaped some-
what like a truncated cone with a relatively flat, smooth
summit and steep, rugged flanks (Fig. 1). This topog-
raphy, combined with the tendency of armorhead to
nocturnally migrate from the flanks to the summit
(Humphreys and Tagami 1986), constrained the com-
mercial trawl fishery to operate primarily on the sum-
mit (<300m) at night (Sasaki 1986).
758
Fishery Bulletin 90|4). 1992
Types of data and preliminary analysis
Since the types of data available for describing the
population dynamics of armorhead have changed with
time, it is convenient to separate the entire 1970-90
interval into three periods: (1) 1985-90, when NMFS
stock surveys were conducted but no commercial fish-
ing occurred; (2) 1978-84, when regulated Japanese
fishing occurred vdth U.S. observers aboard the
vessels; and (3) 1970-77, when unregulated Japanese
and Soviet fishing occurred.
Period 1 In the period 1985-90, NMFS conducted 10
armorhead stock-survey cruises to SE Hancock Sea-
mount. Although bottom trawls were occasionally used,
the primary sampling gear was a bottom longline.
Unlike trawls, longlines could be used on the steep
flanks of the seamount and allow sampling of the en-
tire population. Longlines consisted of 30 rigid poles
(droppers), each with 5 equally-spaced hooks on short
leaders, attached at 18 m intervals along a 600 m
groundline (Shiota 1987). On all cruises, longlines were
set perpendicular to the depth contours to maximize
the depth range sampled and were fished with the same
bait (squid), hook size (no. 20 circle), soak time (1 hr),
and fishing period (0800-1830 hr) to maintain constant
catchability. Starting in 1986, however, catchability
changed slightly when hook timers (small timing
devices that are activated when a fish strikes the hook;
Somerton et al. 1989) were installed on the leaders. To
estimate the effect of timers on armorhead catchabil-
ity, a comparison experiment was conducted in 1990
in which droppers were alternated with and without
timers along the longline. A correction coefficient ac-
counting for the effect of timers on catchability was
then estimated as the ratio of the armorhead catches
for droppers with timers to those without timers (this
ratio was 0.77).
Since preliminary information indicated that armor-
head density varied with depth on the seamount, stock
surveys were based on a depth-stratified sampling
design. Fishing depths were estimated by recording a
depth profile of the bottom as the longline was set, then
partitioning the measured distance between the term-
inal anchors into 30 equal intervals (the number of
droppers). To help correct for possible differences
between fathometer depths and actual fishing depths
due to horizontal drift while the longline sank, max-
imum depth recorders were placed on both anchors and
at the midpoint of all longline sets. Recorded maximum
depths were used instead of fathometer depths to
determine where the anchors and midpoint lay along
each depth profile.
When longlines were retrieved, the species identity
of each captured fish was recorded along with the
number of the hook on which it was caught. All fish
from each 5-dropper segment of the longline were then
placed together into a basket for later collection of the
following biological attributes: sex, fork length (FL,
mm), and body depth (BD, mm) which is the shortest
distance between the bases of the first anal spine and
the dorsal fin. In 1985, body weight (W, g) was also
measured on some specimens in addition to body depth.
Equations predicting BD from W and FL, and predict-
ing W from BD and FL, were calculated from these
data by using multiple regression. These equations are:
Females (n 436)
BD 86.69 - 0.19FL + O.IOW (R- 0.91)
W -936.25 + 9.06BD + 2.49FL (i?^ 0.90)
Males (n 476)
BD 75.55 - 0.18FL + O.IOW (R'~ 0.85)
W -934.02 + 7.52BD + 2.82FL (R- 0.87)
Although the depth distribution of armorhead on the
longline could be determined unambiguously with the
sampling procedure used, this was not true for the
depth distribution of any of the measured or derived
biological attributes, because the catch from each
5-dropper segment was aggregated before the attri-
butes were measured. As a means of approximating
such depth distributions, the biological attributes of in-
dividual fish within each segment group were randomly
assigned to the capture depths within the segment.
The relative abundance of armorhead during each
stock-assessment cruise was expressed as the mean
catch in numbers per hook (U) estimated as a weighted
average over four depth strata (<265, 265-300, 301-
400, 401-500 m). Algebraically, (U) is
U =
ZUiAi
i = l
4
i=l
(1)
where Ui is the catch per hook, and Aj is the bottom
area in depth stratum i. Values of Uj were corrected
for the influence of hook timers, and values of Aj were
estimated as planar areas between the strata depth
boundaries measured on a bathymetric map of the SE
Hancock Seamount (Fig. 1).
Since armorhead begin to lose weight after arriving
at the seamounts, we examined an index of relative
fatness (FI), defined as body depth divided by fork
length, as an index of post-recruitment age. Frequency
distributions of FI were calculated as weighted aver-
ages, where the weighting factors were proportional
to the estimated abundance of armorhead in each depth
stratum. Algebraically, this is expressed as
Somerton and Kikkawa: Population dynamics of Pseudopentaceros wheelen
759
1 Nij Ai U:
N^^ =
i=l
Z AiUi
i=l
(2)
where Njj is the number of fish in FI category j and
depth stratum i.
The frequency distributions of FI usually display two
or three distinct modes which are similar in appearance
to the modes often present in length-frequency distribu-
tions of temperate fishes. Since armorhead recruitment
to the seamounts is seasonal (Boehlert and Sasaki
1988), these modes were assumed to represent annual
cohorts of fish. To estimate the proportion of the pop-
ulation contributed by each cohort (P^ ) and the mean
and variance of its FI distribution, the FI distributions
were separated into their component distributions by
fitting a distribution mixture model using a procedure
developed for length-frequency data (Macdonald and
Pitcher 1979).
Several of the year-class modes in the FI distribu-
tions were so distinct that they could be followed
through the time-series in an orderly progression from
when they were fat (high FI) to when they were lean
(low FI). This feature of the FI distributions was used
in two ways. First, the rate at which armorhead
decrease in fatness was estimated by following the
particularly strong year-class recruiting to the sea-
mount in 1986. Mean FI of this cohort (jji^) on each
cruise was regressed against the time (in months), and
the time squared, since the cohort was considered fully
recruited to the seamount. Linearity of the relation-
ship was determined by the significance of the coeffi-
cient of the squared term.
Second, the instantaneous natural mortality rate (M)
of armorhead was estimated by following two cohorts,
one composed of armorhead recruiting to the seamount
in 1986 and the other composed of all armorhead pres-
ent on the seamount during the first cruise in 1985.
Relative abundance of each cohort at each time (t) was
first estimated as the product of the catch-per-hook and
the proportion of the population witWn the appropriate
cohort on each cruise; that is, P^ t Uf Instantaneous
natural mortality rate was then estimated by regress-
ing the natural logarithm of relative abundance against
the time (in months) since the cohort was considered
fully recruited. Analysis of covariance (ANCOVA) was
used to test whether the slopes of the regression lines
(i.e., the estimated values of M) differed between
cohorts. A best estimate of M was computed as the
average of the estimates for the two cohorts weighted
by the inverses of their variances. Additionally, M was
estimated for each sex separately, considering only the
1986 cohort. ANCOVA was again used to test whether
the estimated values of M differed between sexes.
These and all subsequent applications of ANCOVA
herein will first test a model with one slope and two
intercepts against a model wnth two slopes and two in-
tercepts. If such a test is not significant, then a model
with one slope and one intercept is tested against a
model with one slope and two intercepts.
Period 2 In the period 1978-84, Japanese trawlers
conducted 10 fishing trips to SE Hancock Seamount.
For all trips, U.S. observers recorded the weight of
armorhead caught and the duration of each trawl-haul.
In addition, fork length and body weight were recorded
for a random sample of armorhead drawn from each
haul.
The biomass of armorhead at the start of each fishing
trip was estimated by using the Leslie method (Leslie
and Davis 1939) in which the change in CPUE over
time is related to the cumulative catch removed.
Algebraically, this is expressed as
Ud = Bo q-q Kj
(3)
where Ud is the daily average catch (in kg) of armor-
head per hour of fishing on day d, K<j is the cumulative
catch of armorhead up to the beginning of day d, q is
the catchability coefficient, and B,, is the initial bio-
mass. Two parameters (q. Bo) were estimated by re-
gressing Ud on Kd ; variance of Bq was estimated using
the equation from Polovina (1986). Mean catchability,
qj , and its variance were calculated from the per-trip
estimates.
Frequency distributions of FI were computed the
same as for Period 1, except that body depths were not
measured but were estimated from body weights and
fork lengths using the regression equation previously
described. Since the distributions showed apparent
cohort modes similar to those observed during Period
1, the mean and variance of the FI distribution for each
cohort and the proportional contribution of the cohort
to the population were again estimated with a distri-
bution mixture model. The rate at which FI de-
creased with time was estimated by regressing the
mean FI of the cohort recruiting in 1980 against the
time (in months), since the cohort was considered fully
recruited.
Period 3 In the period 1970-77, Japanese trawlers
fished Hancock Seamounts for at least 1 month in every
year; however, data from NW and SE Hancock Sea-
mounts could not be separated. Soviet trawlers also
likely fished the Hancock Seamounts over this period,
but the available Soviet data (Borets 1975) were aggre-
gated over all seamounts and were not useful for deter-
mining the stock dynamics on the Hancock Seamounts.
760
Fishery Bulletin 90(4|. 1992
Relative abundance of armorhead was therefore based
solely on Japanese data and was calculated as the
reported monthly catch divided by the fishing effort
(in hr). Annual mean catch-per-hour (Ut) and its vari-
ance were calculated from the unweighted monthly
means. FI could not be calculated during this period
due to insufficient data.
Biomass estimation
Armorhead abundance could be estimated in absolute
terms as biomass only in Period 2. In the other periods,
abundance was estimable in relative terms as CPUE.
To allow estimation of biomass from CPUE in Periods
1 and 3 and to allow the merging of all three periods
into one continuous time-series, several parameters
were required that could be estimated only with data
from Period 2. For this reason, we will start by describ-
ing the biomass estimation procedures for Period 2.
Period 2 The initial biomass estimates obtained for
Period 2 (i.e., Leslie estimates of initial biomass in each
year, Bq t) may not include the total biomass of armor-
head on the SE Hancock Seamount. Instead, the ini-
tial estimates may include only the biomass of the
fishable stock or that portion of the stock occurring on
the summit at night and therefore vulnerable to trawls.
The question of whether Bot includes the total popula-
tion was addressed by testing the equality of two dif-
ferent estimators of annual recruitment to the sea-
mount. The first (Rj ) was calculated as the difference
between the estimated biomass in 1 year minus the
expected biomass surviving from the previous year.
Assuming that the catch was taken in a brief interval
at the start of the year, this relationship can be ex-
pressed as
Ri,t+i = Bo,t+i - (Bo,t - Ct) e
(4)
where B,, t and Bqi+i are the biomass estimates in
years t and t-H 1, Ct is the catch in year t, and e "^ is
the annual survival rate. The second (R2), is calculated
as the proportion of the biomass composed of recently
recruited fish:
R
2,t+l
= Bo t + 1 P
r.t + l I
(5)
where Pr,t+i is the proportion of the population com-
posed of the cohort recruiting in year t -1- 1 . If Bq, t
estimates include the total biomass, then they will be
appropriately scaled to Ct, and Ri will equal R2. But
if the Bo, t estimates are less than the total biomass,
then Ri will be greater than R9. Equality was tested
using the statistic
Z =
Ri - R9
VVar(Ri-R2)'
(6)
where Z was assumed to be distributed as a normal ran-
dom variable. Estimates of Var(Ri - R2) were com-
puted as described in the Appendix.
The fishable proportion of the stock (Pf) was esti-
mated in two stages. First, total biomass of the 1980
cohort was estimated for each year in 1980-84, when
the 1980 cohort represented more than 90% of the
total population, by using an age-structured analysis
(Megrey 1989) applied to a single cohort. Starting with
a known or assumed value of biomass at the beginning
of 1985, this analysis sequentially predicts biomass in
each preceding year by accounting for catch and
natural mortality. If the catch occurs over a short
period at the start of the year, total biomass of this
cohort in each year (B*t) can be expressed as
B*to^i = B*toe^' + Ctci
B*to-2 = B*to-i e^i + Cto-2
= B*to e^ + Cto-i eM + Cto-2 (7)
B'
tO-n
B
to
,nM
1 Cto-
i(n-i)M
i=l
where B* 0 is an estimate of total biomass at the start
of the last year (to) in the time-series (terminal bio-
mass), and Cto-i is the catch in year to_i. Second, the
proportion fished in each year (P,- , ) was then esti-
mated as Bo,t/B* , and mean Pf was then estimated as
the average of the five annual estimates. This estimate
of Pf, however, was not unique because it depended
on B*to, and B*to was chosen arbitrarily because no in-
dependent estimate was available. Therefore, the term-
inal biomass B*85 (the terminal fishing year was de-
fined as 1985 as a later convenience) was estimated
along with Pf. Assuming Pf is a constant, the two
parameters were estimated by minimizing the weighted
sum of squares of the Pf , with weights equal to the
inverse of the variance of each Pf t .
Once the estimate of mean Pf had been obtained,
corrected estimates of the initial biomass in each year
(i.e., corrected Leslie estimates) were estimated as
B
o.t -
Bo.t
Pf'
(8)
Somerton and Kikkawa: Population dynamics of Pseudopentaceros wheelen
761
and the mean annual biomasses were then estimated as
B*t =
qtPf
(9)
where Ut and q, are the mean annual CPUE and
catchability. Variance of B* was estimated with
methods described in the Appendix.
Period 3 Biomass during Period 3 was estimated
from the mean catch-per-hour of Japanese trawlers
(Ut)as
Ut
B*t = — ^, (10)
qjPf
where Qj is the mean catchability of Japanese trawlers
estimated for Period 2. Variance of B* was estimated
with methods described in the Appendix.
Period 1 Biomass during Period 1 was estimated
from longline catch-per-hook (Ut ) as
B*t
qi
(11)
where Wt is the mean individual weight of armorhead
caught during sampling Period t, and qi is the catch-
ability of the longlines. Estimation of qi required an
independent estimate of B*t for at least one of the
sampling periods, and the estimate chosen was the
terminal biomass of the 1980 year-class in 1985. Catch-
ability was thus estimated as
qi =
U85 P80 W:
85
B*.
(12)
80,85
where Ugs is the catch-per-hook, Wgs is the mean body
weight in 1985, B*8o,85 is the terminal biomass of the
survivors of the 1980 year-class at the start of 1985,
and Pgo is the proportion of the 1985 population com-
posed of the 1980 year-class survivors. Variances of
B*t and qi were estimated with the methods described
in the Appendix.
Spawning and recruitment biomasses
Spawning and recruitment biomasses were estimated
for Periods 1 and 2 in which FI information was avail-
able. Spawning biomass in each year (St ) was esti-
mated as:
St+i
B*t -
Ct
2
M
Pr.t),
(13)
where Prj is the proportion of B* comprised of newly-
recruited fish, and all other terms are as previously
defined. This formulation assumes that B* was always
estimated on 1 July, the midpoint of both the fishing
season and the stock-assessment cruise. Natural mor-
tality between 1 July and 31 December, the assumed
peak of spawning (Bilim et al. 1978), was accounted for
by the term e"'^'-, where M is the annual instantan-
eous natural mortality rate. P^t was estimated as
the proportion of the population (P^) within the modal
group with the largest mean FI, or as zero if no model
group had a mean FI>0.25. P^t was included in the
estimate of spawning biomass because, based on
samples collected on the August 1988 stock-assessment
cruise (R. Humphreys, NMFS Honolulu Lab., unpubl.
data), female armorhead appear to be nonreproductive
during the first spawning season after they recruit to
the seamounts. Recruitment biomass was estimated as
R, = |B*t + -^|Pr,f
(14)
This formulation assumes that recruitment occurs from
March to May (Boehlert and Sasaki 1988) and is com-
plete by the time B* is estimated. Since young armor-
head recruit to the seamounts at approximately 24-30
months of age (Uchiyama and Sampaga 1990), recruit-
ment follows spawning by 3 calendar years. Spawner-
recruit relationships were therefore examined using a
3-year lag between spawning and recruitment.
Results and discussion
The armorhead population on SE Hancock Seamount
fluctuated tremendously between 1970 and 1990 (Fig.
2) and declined steadily after the population high in
1972, except for small increases occurring in 1980 and
1986. Before the forces producing these changes (i.e.,
natural mortality, fishing mortality, and recruitment)
are examined, the potential biases and the precision of
the biomass time-series will be considered.
Biomass estimates
Construction of the time-series of biomass estimates
required (1) merging two time-series of CPUE data
that were non-overlapping in time and were from
distinctly different gear types, and (2) the conversion
of a relative measure of abundance (CPUE) into the
absolute measure of biomass. Since Japanese trawls
and research longlines were never used simultaneous-
ly, the time-series could not be merged by simply stan-
dardizing the catchability of one gear relative to the
other. Fortunately, however, the armorhead popula-
762
Fishery Bulletin 90(4). 1992
70 71 72 73 74 75 76 77 11
YEAR
z
o
00
<
o
QD
78 80 82 84 86 88 90
YEAR
Figure 2
Pelagic armorhead Pseudopentaceros wheeleri
biomass on Southeast Hancock Seamount dur-
ing periods of (A) high abundance (1970-78) and
(B) low abundance (1979-90). Prior to 1985,
when routine stock-assessment surveys were ini-
tiated, the biomass estimates are annual means
and are shown at the beginning of the year.
Starting in 1985, the biomass estimates are for
each stock-assessment survey and are shown for
the appropriate month.
tion on SE Hancock seamount was suffi-
ciently small and the fishing effort was suf-
ficiently large to allow use of the Leslie
method to estimate both the mean catch-
ability of the trawlers (cy) and the biomass
at the initiation of each fishing season
(Bq, i) during the period just prior to re-
placement of commercial trawling by
research longlining. The time-series was
merged by using the estimate of qj and Pf
to compute biomass from Japanese trawl
CPUE and by using the estimates of Bq t to
estimate cy and thereby compute biomass
from longline CPUE. Thus, the Leslie
method provided the means to merge the
two time-series and to express the resulting
time-series in terms of biomass.
Because success of this procedure rests on
the successful application of the Leslie
Table 1
Leslie
estimates
of initial
biomass (B,,) and catchability (q)
of pelagic
armorhead Pseudopentaceros wheeleri for each of the Japanese fishing trips |
during Period 2 (1978-84)
on SE Hancock Seamount.
Vessel
Days
Catch
B„
Year
Month
ID no.
fished
(t)
(t)
q
P(q = 0)
1978
May
1
11
204
198
0.00197
< 0.001
1979
Jun
2
18
68
80
0.00059
<0.005
1980
Aug
3
29
453
551
0.00047
<0.001
1981
Jun
2
5
161
297
0.00101
>0.10
Aug
3
20
44
55
0.00067
>0.10
1982
May
4
10
180
269
0.00064
0.10
Jul
2
12
8
17
0.00071
>0.10
1983
Jul
3
19
39
38
0.00066
<0.001
method, any biases in the biomass estimates are likely the result
of violations of the underlying assumptions. One of the most im-
portant assumptions of the Leslie method is that the change in
size of the population is solely due to removals by the fishery. In
practice, this requires that the population is closed to immigra-
tion and the fishery is sufficiently short and intense so that the
effect of natural mortality is negligible. In nearly all cases exam-
ined, these requirements were met; that is, fishing usually oc-
curred well after the spring peak in recruitment of pelagic juve-
niles (Table 1; Boehlert and Sasaki 1988) and the catch was usually
obtained in 1 or 2 months and represented a large fraction of the
estimated biomass (Table 1). One further indicator of the suc-
cessful application of the Leslie method is the significance of the
slope of the regression, or q. Although the estimates of q obtained
using the Leslie method were not always significantly greater than
zero, they were remarkably similar among years for each vessel
that had fished repeatedly (Table 1). Such similarity was used
as justification for using the non-significant estimates in later
calculations.
Bias could also result from violation of another assumption of
the Leslie method, that the entire population is equally \ailnerable
to the sampling gear. Such bias was considered likely when ini-
tial biomass estimates from Period 2 seemed too small to be con-
sistent with the observed catches. This apparent inconsistency was
examined statistically by testing the equality of two estimators
of recruitment, one that included catch (Rj ; Eq. 4), and one that
did not (R2; Eq. 5). The bias was confirmed since in 4 of the 5
years examined, R] was significantly (P<0.05) greater than R2,
a condition that could occur only if the biomass estimates were
too small.
The most likely explanation for the underestimation of armor-
head biomass is that the Bot estimates do not include the entire
population and instead include only the fishable population or the
part actually exposed to trawls. This result was surprising because
we believed that the population would be sufficiently mixed by
the nocturnal vertical migration so that all armorhead would be
equally vulnerable even though the trawls were topographically
restricted to only a part of the armorhead depth range. Our
Somerton and Kikkawa Population dynamics of Pseudopentaceros wheelen
763
finding, however, indicates either that mixing is minimal
or that the rate of mixing is relatively low compared with
the 2- to 4-week duration of a typical Japanese fishing trip.
Although the bias in Bot was corrected by estimating
the proportion of the stock vulnerable to trawling (Pf =
0.27), adequacy of this correction rests on the assump-
tion that Pf does not vary with time. However, Pf may
vary with time because the depth distribution of armor-
head may vary. For example, the proportion of the popula-
tion occurring in the shallowest depth stratum (<250m)
averaged 15% over the 10 research cruises, but ranged
from 0 to 40%. It is unclear if such variation in the
daytime distribution is reflected in the nighttime distribu-
tion, because armorhead do not feed at night and there-
fore cannot be sampled effectively with longlines (M.P.
Seki and D.A. Somerton, NMFS Honolulu Lab., unpubl.
data). Bias in the biomass estimates could additionally
occur if Pf depends on the degree of mixing of deep and
shallow fish, because Pf would likely be larger when
fishing periods were longer and less intense. Since fishing
periods tended to be longer during Period 3 than in Period
2, this would lead to an overestimate of biomass during
Period 3.
Precision in the estimates of biomass, which is ex-
pressed as the coefficient of variation (CV) to compensate
for the large range in biomass, was smallest in Period 1
(Fig. 3), because longline CPUE estimates were more
precise than trawl CPUE estimates. Expressed different-
ly, based on the mean CV over the period 1970-84 (ex-
cluding 1971 when data were not sufficient to estimate
the variance of Ut ), the 95% confidence intervals for B*x
was ± 1.70 B*x, which in all years includes zero. Over the
period 1985-90, however, the 95% confidence interval
was ±0.29 B*x, and never included zero.
Post-recruitment ageing
The estimates of natural mortality rate and annual
recruitment required estimates of the age distribution.
Although the ages of armorhead can be determined using
either daily or annual growth increments on their otoliths
(Uchiyama and Sampaga 1990), they are easily obtainable
only for individuals in the pelagic phase of their life
history, because somatic growth ceases once armorhead
recruit to the seamount (Humphreys et al. 1989) and
growth increments become so closely spaced that they are
exceedingly difficult to count (R. Humphreys, NMFS
Honolulu Lab., pers. commun.). Thus, we expressed age
on a scale relative to the presumed time of recruitment.
Such post-recruitment ages were based on the decrease
in FI over time.
Frequency histograms of FI display modes which can
be tracked sequentially from one histogram to the next
over time as they move from the right (high FI or fat)
to the left (low FI or lean). Two examples of this are the
Z 2.5
O
< 2.0
a:
<
>
^ 1-5
O
I .
i '".i UyVS -
Li- 0.5
Li_
LlJ
O
O 0.0
7
•• — •§►••-• •
0 75 80 85 90
YEAR
Figure 3
Coefficient of variation of the biomass estimates of pelagic
armorhead Pseudopentaceros wheeleri as a function of time.
large mode that appeared in 1980 and could be
followed until 1984 (Fig. 4A) and the large mode that
appeared in 1986 and could be followed until 1990
(Fig. 4B). Since the first appearance of these modes
was always associated with an increase in CPUE
(Fig. 2B), they were interpreted to represent cohorts
of fish that had recruited to the seamount.
To further substantiate our interpretation of the
modes, the rate of decrease in FI was examined for
consistency both over time and between the two
presumed year-classes. Plots of FI versus time ap-
peared to have slight curvature (Fig. 5), but for both
the 1980 and 1986 year-classes the curvature was
not significant (P>0.05). Changes in FI, therefore,
are proportional to changes in post-recruitment age.
The rates of decrease in FI of the 1986 (0.00169/mo)
and the 1980 (0.00157/mo) year-classes did not dif-
fer significantly (ANCOVA, P>0.05). In addition,
sexual equality in the rate of decrease in FI was
tested for the 1986 year-class alone, and the male
rate was not significantly different (ANCOVA, P>
0.05) from the female rate. Taken together, these
findings indicate that all armorhead decrease in FI
at approximately the same rate and that once estab-
lished by the recruitment of a strong year-class, the
coherency of an FI mode should be preserved over
time.
Natural mortality
Natural mortality was estimated from the change
in the relative abundance of two cohorts over time
during a period when no commercial fishing oc-
curred. The first of these cohorts, which consisted
764
Fishery Bulletin 90(4), 1992
fi-CO-O OOO9&OOQOQ
0.00
0.10
JANUARY 1985
JUNE 1985
^^^'y^'^^^rXiXiaooo—
O
P "■
cr 0.
£0,
o 0.
3.10-
3.05-
3.00 0000000
AUGUST 1986
OCTOBER 1986
10-
J^
"^^^ 1982
05-
00.
.^
, !^°^H^0090, . . _
O 0 O fi) O Q Q-» ■a O o » o o o
AUGUST 1987
;onWpnnoe&.
0.25
FATNESS INDEX
, ? 9"g-^t-Q.o oQoooooqooo
JANUARY 1988
AUGUST 1990
0.20 0.25 O.iO
FATNESS INDEX
APRIL
1987
0.05-
oja
«^
0.00 J
xjoQoqgrrti —
0
^=Q»W - .-
Figure 4
Frequency histograms of fatness index are shown for Japanese commercial catch of pelagic armorhead Pseudopentaceros wheeleri in
(A) each year during 1978-84 and (B) for each of the research cruises during 1985-90.
c
1
)
0
0 1986 COHORT
• 1980 COHORT
FATNESS INDE
0 0
0 en
0 8
0 •
•
0
0
0.15
C
)
10 20 30 40 5
0
TIME (MONTHS)
of the extant armorhead population in January 1985, was
identifiable for seven consecutive samplings and had an
instantaneous natural mortality (M) of 0.054/mo. The
second, which consisted of the year-class recruiting in
1986, was identifiable for eight consecutive samplings and
had an M of 0.044/mo (Fig. 6). Although the two estimates
were significantly different (ANCOVA, P<0.05), the
weighted average (0.045/mo or 0.54/yr) was chosen as
Figure 5
Decrease in fatness index of pelagic armorhead Pseudopen-
tnceroa wheeleri with time for the 1980 and 1986 year-classes
on Southeast Hancock Seamount.
Somerton and Kikkawa: Population dynamics of Pseudopentaceros wheelen
765
the most representative value. For both cohorts, log-relative
abundance was clearly a linear function of time, and M is
therefore invariant with age (Fig. 6A).
When M was estimated separately for each sex, considering
the 1986 cohort alone, the value for males (0.037/mo) was
significantly different (ANCOVA, P<0.001) from the value for
females (0.045/mo). Furthermore, the intercept of the regres-
sion, i.e., the log-relative abundance at the time of recruitment)
appeared to be smaller for males (4.27) than for females (4.94),
but the difference could not be tested because of the strong
difference in slopes (Fig. 6B). Taken together, these results in-
dicate that, at least for the 1986 cohort, females recruited to
the seamounts in greater abundance than males but subsequent-
ly died at a greater rate. It is also possible that the higher mor-
tality rate of females was primarily restricted to the first year
of residence on the seamount (Fig. 6B).
Since a sexual difference in mortality seemed inexplicable to
us, we examined the possibility that longlines preferentially
selected fat females. This was done by examining whether the
ratio of females to males, expressed as proportion female,
changed with FI similarly for trawls as for longlines. Since the
FI values of females and males decrease identically with time,
any change in female proportion with FI would indicate either
selective sampling or differential mortality, depending on
whether one or both gear types showed the change. To test
for such changes, female proportion was regressed on FI for
various samples. When these regressions were performed on
the longline samples, all 10 had a significant (P<0.05) positive
slope. When the regressions were performed on the research
trawl samples, four of five had a significant positive slope. Thus
a sampling bias is unlikely, unless both gears produced a
similar bias.
The estimate of natural mortality rate (0.54/yr) is more than
twice that reported in Borets (0.25/yr; 1975). His estimate,
however, was based on age data that were likely biased for two
reasons. First, the age range [i.e., 7 yr, ages 5-12], reported
in Borets (1975) appears to be excessive when compared with
the range estimated from modal progression through the
research FI histograms (4-5 yr). Second, the mean age of the
catch between 1968 and 1974 reported in Borets (1975) did not
decrease as would be expected in a developing fishery. On the
other hand, our estimate of natural mortality rate was less than
the rate implied in the studies of Uchida and Tagami (1984),
Humphreys et. al. (1989), and Uchiyama and Sampaga (1990),
which all suggested that armorhead were semelparous and, like
Pacific salmon Oncorhynchus spp., died soon after spawning.
Fishing mortality
Fishing mortality rate (F) can be estimated only between 1978
and 1983 when the total catch and effort on SE Hancock Sea-
mount are known with reasonable certainty, based on the U.S.
observer program. Over this period, F, which was calculated
as the estimated value of Qj x total annual effort, averaged
1.03/yr or roughly twice the natural mortality rate. The mean
p,
u
0 1985 COHORT A
f^ • 1986 COHORT
5
p^^^
LJ ^
^ —
Q-
^^^:::S^o
O 4
0\^^^^
o
^\^ •^^^
o
^ 3
■
2
0 12 24 36 48 60
R
U
• MALE B
0 FEMALE
^50
ZD ^
<5
Q_
^^^"""Sl
0
^"^"""""■■-^
4'
^•^^<° G-^
0
^^""""--^ ^^^"-^
0
^^"^^^^
^ 3
•^^^"^-#
2
til.
0 12 24 36 48 60
MONTHS
Figure 6
Change in log CPUE of pelagic armorhead Pseudo-
pentaceros u'heeleri versus time for the (A) cohort
composed of the extant popiUation in 1985 and the
cohort recruiting in 1986. and (B) for males and
females of the 1986 recruiting cohort.
exploitation rate (which was approximated as
(F/Z) (1-e-Z) where Z = F-i-M), was ~0.50.
Therefore, provided that no recruitment oc-
curred and that fishing effort was continuous
throughout the year, then an average of
roughly 50% of the population at SE Hancock
Seamount was removed annually by the fishery
over this period.
Recruitment
Annual recruitment to SE Hancock Seamount
was extremely intermittent between 1978 and
1990. When expressed in metric tons, the 1980
recruitment clearly dominated the entire
record (Fig. 7A). However, the total biomass
changed considerably over this period, and,
when expressed as a percentage of the total
biomass, the 1986 recruitment and— to a lesser
extent— the 1988 and 1990 recruitments were
also relatively important (Fig. 7B). For some
766
Fishery Bulletin 90(4). 1992
ROD
^ 500
- t
o
\
1- 400
'
y 300
-
ce
h 200
■ \
; \
^ 100
■ \
n *-! ' *-■■••-' ^•••-^
1980 1984 1988
YEAR
-CD
80
" T f -
1—
A A
5 60
-
O
/
cn 40
- /
LlJ
/ ^
Q_
/ /
20
- J • /-
o'
dX. y , i/,v
1980 1984 1988
YEAR
Figure 7
Annual recruitment to Southeast Hancock
Seamount expressed in (A) metric tons and
(B) percent of the population biomass of
pelagic armorhead Pseudopentaceros
wheeleri.
unknown reason, recruitment in the even years tended
to be larger than it was in the odd years (Mann- Whitney
test, P = 0.06).
Prior to 1978, it is difficult to estimate annual recruit-
ment because the data required to calculate FI were
not routinely collected. However, an examination of the
large increase in abundance in 1972 (Fig. 2A) cannot
be avoided. Interpretation of the 1972 increase is
troublesome because the evidence for an unusually
large recruitment is equivocal. A large recruitment, for
example, should have resulted in a large increase in the
proportion of the population comprised of fat ar-
morhead as it did in 1980 and 1986 (Fig. 4A,B), yet
the proportion of the population categorized as fat dur-
ing the period of maximum recruitment was smaller
in 1972 than it was in 1973 when recruitment did not
appear to be exceptionally large (Boehlert and Sasaki
1988). Armorhead measured in 1972, however, were
markedly smaller than in any other year (Takahashi
and Sasaki 1977, Borets 1975). One proposed explana-
tion for the small size is that the tremendous abundance
of armorhead in 1972 resulted in a density-dependent
suppression of growth in the pelagic phase (Borets
1977). This might also have reduced the FI of recruits
P 10^
2
•
Ul
tn
< 2
2 lO'^
-
o
•
QD
1—
Z
• •
^ 10^
•
1—
•
3
•
o
• •
^ in°
-
m 10 - -
10' 10^ 10^
SPAWNING BIOMASS (MT)
Figure 8
Annual recruitment versus spawning biomass of pelagic
armorhead Pseudopentaceros wheeleri 2 years earlier.
in
Ul
<
o
CO
0.0 0.2 0.4 0.6 0.8 1.0
JAPANESE CPUE (MT/HR)
Figure 9
Biomass of pelagic armorhead Pseudopentaceros wheeleri on
Southeast Hancock Seamount and Japanese trawl catch/hr
on all other southern Emperor-northern Hawaiian Ridge
seamounts during (•) Period 1 (1985-90) and (O) Period
2 (1978-84) when the Southeast Hancock Seamount popula-
tion experienced fishing.
and thereby masked the FI signature of a large
recruitment.
An alternate explanation for the 1972 increase in
abundance is that it is an illusion due to a rapid increase
in trawl catchability as the newly-developed fishery
progressed from an exploratory phase to a production
phase (Takahashi and Sasaki 1977, Uchida and Tagami
1984). If, however, the 1972 increase was due simply
Somerton and Kikkawa Population dynamics of Pseudopentaceros wheelen
Ibl
to changing catchability, then such an increase should
not be evident in the Soviet catch-and-effort data,
because the Soviet fishery had developed earlier and
was likely beyond its "fishing-up" phase. Soviet data
do display an increase from 1971 to 1972 (73-104
million fish/vessel day; Borets 1975), but this is con-
siderably less than that experienced by the Japanese
fishery. Thus, it is not entirely clear whether the ap-
parent increase in 1972 was real and due to recruit-
ment or an artifact due to changing catchability.
The dependance of recruitment on spawning biomass
was examined by Wetherall and Yong (1986) and found
to be essentially nonexistent, at least at the high levels
of spawning biomass extant during 1969-77. Recruit-
ment, however, must ultimately be limited by spawn-
ing biomass at low population levels; therefore, we
reexamined the relationship over the period 1980-90,
when the spawning biomass was considerably lower.
This was done by plotting, on a log-log scale, the esti-
mated spawning biomass on SE Hancock Seamount
against the estimated recruitment 2 years later (Fig.
8). Since a clear relationship is not evident, it is pos-
sible that recruitment and spawning biomass are only
weakly related even at the low population levels ex-
amined. There are, however, at least two other pos-
sible reasons why no relationship was found. First,
since no apparent genetic difference exists among the
armorhead collected at the various seamounts (Borets
1979), recruits to SE Hancock Seamount are likely the
progeny of the entire North Pacific population. If the
SE Hancock Seamount population does not vary con-
cordantly with the entire North Pacific population, any
relationship between recruitment and spawning
biomass would be obscured. However, plots of the
estimated biomass on SE Hancock Seamount against
Japanese CPUE on all SE-NHR seamounts show a
strong concordance (Fig. 9). Second, if spawning
biomass does exert an influence on recruitment, it may
do so only by limiting the maximum level attained.
Thus, at higher levels of spawning biomass, higher
levels of recruitment are possible— but not assured—
because of environmental variability. One interpreta-
tion of Figure 8 could therefore be that recruitment
did increase with spawning biomass, but at the higher
levels of biomass there were several environmentally-
poor recruitment years.
Management implications
Since armorhead do not grow after they recruit to the
fishery and therefore cannot be growth-overfished,
management strategies could be designed solely to
achieve some optimum level of spawning stock biomass
(SSB). One approach is to define this optimum SSB by
using a spawner-recruit relationship as is done for some
species of Pacific salmon (Ricker 1975). Another ap-
proach is to define it in terms of a fixed percentage
of the equilibrium biomass in the absence of a fishery
(Beddington and Cooke 1983). But in either case, the
spawning population must include the entire SE-NHR
population rather than the small component examined
here. In addition, some form of international agree-
ment controlling the armorhead catch will be required
before any management measures are effective.
Acknowledgments
We thank George Boehlert, Bob Humphreys, Bill
Lenarz, and Jeff Polovina for reviewing the manuscript
and offering helpful suggestions.
Citations
Beddington. J.R.. and J.G. Cooke
1983 The potential yield of fish stocks. FAO Fish. Tech. Pap.
242, 47 p.
Bilim, L.A., L.A. Borets. and M.K. Platoshina
1978 Characteristics of ovogenesis and spawning of the boar-
fish on the region of the Hawaiian Islands. In Fisheries
oceanography, hydrobiology, biology of fishes and other
denizens of the Pacific Ocean. Izv. Tikhookean. Nauchno-
Issled. Inst. Rybn. Khoz. Okeanogr. (TINRO) 102:51-57 [in
Russ.]. (Engl, transl. 106 by W.G. Van Campen, 1986, 9 p.;
avail. NMFS Honolulu Lab.)
Boehlert. G.W., and T. Sasaki
1988 Pelagic biogeography of the armorhead, Pseudopetitaceros
wheeleri, and recruitment to isolated seamounts in the north
Pacific Ocean. Fish. Bull., U.S. 86:453-466.
Borets, L.A.
1 9 75 Some results of studies on the biology of the boarfish (Pen-
taceros richardsoni). Invest, of the biology of fishes and
fishery oceanography. Izv. Tikhookean. Nauchno-Issled. Inst.
Rybn. Khoz. Okeanogr. (TINRO, Vladivostok) 6:82-90. (Engl,
transl. 97 by W.G. Van Campen. 1984, 9 p.; avail. NMFS
Honolulu Lab.)
1977 Dynamics of size-age composition of the boerfish {Pen-
taceros richardsoni). Invest, of the biology of fishes and
fishery oceanography. Izv. Tikhookean. Nauchno-Issled. Inst.
Rybn. Khoz. Okeanogr. (TINRO, Vladivostok) 8:65-69 [in
Russ.] (Engl, transl. 115 by W.B. Van Campen. 1987, 6 p; avail.
NMFS Honolulu Lab.)
1979 The population structure of the boarfish, Pentaceros
richardsoni, from the Emperor Seamounts and the Hawaiian
Ridge. J. Ichthyol. 19:15-20.
Efron, B., and G. Gong
1983 A leisurely look at the bootstrap, the jackknife and cross-
validation. Am. Stat. 37:36-48.
Humphreys, R.L. Jr., and D.T. Tagami
1986 Review and current status of research on the biology and
ecology of the genus Pseudopentaceros. In Uchida, R.N., et
al. (eds.). Environment and resources of seamounts in the North
Pacific, p. 55-62. NOAA Tech. Rep. NMFS 43.
768
Fishery Bulletin 90(4). 1992
Humphreys, R.L. Jr., G.A. Winans, and D.T. Tagami
1989 Synonymy and life history of the North Pacific pelagic
armorhead, Pseudopentaceros wheeleri Hardy. Copeia 1989:
142-153.
Komrakov, O.E.
1970 Distribution and fishery of the boarfish (Pentaceros
richardsoni) in the Hawaiian region. Avail, as Transl. 117
from NMFS Honolulu Lab.
Leslie, P.H., and D.H.S. Davis
1939 An attempt to determine the absolute number of rats on
a given area. J. Anim. Ecol. 8:94-113.
Macdonald, P.D.M., and T.J. Pitcher
1979 Age groups from size-frequency data: A versatile and
efficient method of analyzing distributions mixtures. J. Fish.
Res. Board Can. 36:987-1001.
Megrey, B.A.
1989 Review and comparison of age-structured stock assess-
ment models from theoretical and applied points of view. In
Edwards, E.F., and B.A. Megrey (eds.), Mathematical analysis
of fish stock dynamics, p. 8-48. Am. Fish. Soc. Symp. 6.,
Bethesda.
Polovina, J.J.
1986 A variable catchability version of the Leslie model with
application to an intensive fishery experiment on a multispecies
stock. Fish. Bull., U.S. 84:423-428.
Ricker, W.E.
1975 Computation and interpretation of biological statistics
of fish populations. Bull. Fish. Res. Board Can. 191, 382 p.
Sasaki, T.
1986 Development and present status of Japanese trawl fish-
eries in the vicinity of seamounts. In Uchida, R.N. , et al. (eds.).
Environment and resources of seamounts in the North Pacific,
p. 21-30. NOAA Tech. Rep. NMFS 43.
Seber, G.A.F.
1973 The estimation of animal abundance. Hafner Press, NY,
506 p.
Seki, M.P., and D.A. Somerton
In prep. Feeding ecology and daily ration of the pelagic armor-
head, Pseudopentaceros wheeleri, at Southeast Hancock Sea-
mount. NMFS Honolulu Lab., 30 p.
Shiota, P.M.
1987 A comparison of bottom longline and deep-sea handline
for sampling bottom fishes in the Hawaiian Archipelago.
Admin. Rep. H-87-5, NMFS Honolulu Lab., 18 p.
Somerton, D.A.. B.S. Kikkawa, and CD. Wilson
1989 Hook timers to measure the capture time of individual
fish. Mar. Fish. Rev. 50(2): 1-5.
Takahashi, Y., and T. Sasaki
1977 Trawl fishery in the central North Pacific seamounts [Kita
Taiheiyo chubu kaizan ni okeru tororu gyogryo. Hokuyo soko-
uo gyogyo— Shiryo (3)] Northern waters groundfish fishery-
Data (3). Div. North. Waters Groundfish Resour., Far Seas
Fish. Res. Lab., 45 p. [In Jpn.; Engl, transl. 22 by T. Otsu,
1977, 49 p.; avail.NMFS Honolulu Lab.]
Uchida, R.N., and D. Tagami
1984 Groundfish fisheries and research in the vicinity of sea-
mounts in the North Pacific Ocean. Mar. Fish. Rev.
46(2):l-:7.
Uchiyama, J., and J. Sampaga
1990 Age and growth of the pelagic armorhead Pseudopen-
taceros wheeleri from the Hancock Seamounts. Fish. Bull.,
U.S. 88:217-222.
Wetherall, J.A., and M.Y. Yong
1986 Problems in assessing the pelagic armorhead stock on
the central North Pacific seamounts. In Llchida, R.N. et al.
(eds.). Environment and resources of the seamounts in the
North Pacific, p. 73-86. NOAA Tech. Rep. NMFS 43.
Somerton and Kikkawa: Population dynamics of Pseudopentaceros wheelen
769
Appendix
Variances of several of the estimators described in Materials and Methods were approximated by using the Delta
method (Seber 1973) and assuming all covariance terms were negligible. Variance of (Rj - R-,) was estimated as
Var(Ri-Ro) = (P,,t,i-1)2 Var(Bo,t.i) + (B
o.t.i)2 Var(Pr,t,i) + e-2M [Var(Bo,t) + Var(Ct)+(Bo,t- Ct)^ Var(M)],
(15)
where all variables are defined in text Equations 5 and 6. Variances of Bq t and Bo,t+i were computed by using
the method in Polovina (1986). Variance of Ct was assumed to be negligible because catch was measured by U.S.
observers. Variance of M was estimated as the variance of the slope of the regression of log-relative abundance
on postrecruitment age (in years). Variance of Pr,t+i was estimated with a bootstrap method (Efron and Gong
1983). Bootstrap estimates were obtained from trawl samples of armorhead biological data by iteratively repeating
the following steps: (1) A subsample of n fish from each sample was randomly chosen with replacement, where
n is equal to the size of the original sample; (2) an FI frequency distribution was constructed from the subsample;
(3) Pr,t+i was estimated by fitting the distribution mixture model to the FI frequency distributions. In all cases,
variance of Pr,t+i was calculated as the variance among 100 bootstrap estimates.
Variance of B* during Period 2 was as
Var(B*t) =
qtPf
Var(Ut) +
UtPf
(qtPf)'
Var(qt) +
UtQt
(qtPf)'
Var(Pf)
(16)
where all variables are defined in Equation (9). Variance of Ut was estimated as the variance among the daily
U within each year. Variance of qt was estimated as the variance of the slope of the Leslie model. When more
than one vessel fished in each year, however, variance of qt was the average of the individual variance estimates
weighted by catch.
Variance of Pf was estimated by using a Monte Carlo model. Each iteration of the model consisted of generating
a random value of Bot for each year and a value of M, assuming all were normally distributed with means and
variances equal to the original estimated values. With these generated values, B* was estimated for each year
with Equation (9) and Pf t was estimated as Boj/B* . Mean estimates of Pf and B*go,85 (B*o in Eq. 7), were ob-
tained by using the iterative procedure to minimize the weighted sum of squares of the Pf t estimates. In all cases
variances of Pf and B*go,85 were estimated from 100 iterations of the Monte Carlo model.
Variance of B* during Period 3 was estimated using Equation (16) but with qt replaced by qj. Variance of
Uf was estimated as the variance among the monthly means. Variance of qj was estimated as the variance among
the q estimates in Table 1. Variance of Pf is the same as for Period 2.
Variance of B* during Period 1 was estimated as
Var(B*t) =
Wt
qi
Var(Ut) +
qi
Var(Wt) +
UtWt
q^
Var(q,),
(17)
where all variables are defined in Equation (11). Variance of Wt was estimated from the biological samples from
each research cruise. Variance of Ut was estimated as the variance of U among the four depth strata.
Variance of qi was estimated as
Var(q,) =
UssWss
B
80.85
Var(P8o) +
U85P80
B*80,85
Var(Wg5) +
P80W85
B*
80,85
VaKUgg) +
U85P80W85
B*,
80.85
Var(B
80,85.1
(18)
where all variables are defined in Equation (12). Variances of Wgj and Ugs were estimated as described above
for Wt and Uf. Variance of Pgo was estimated as described for Pr,t+i- Variance of B*8o,85 was estimated with the
previously described Monte (]arlo model.
Abstract.— An anomalous inabil-
ity to distinguish certain geograph-
ically-separated Chinook salmon On-
corhynchus tshaurytscha populations
of the Snake River and the Klamath
River from a survey of 18 polymor-
phic loci led to a prediction that
distinction would ultimately be found
through sampling of additional poly-
morphic loci. Recently published
studies involving pertinent groups
within each of these rivers included
data from an additional 15 polymor-
phic loci, and therefore allow a re-
examination of the relationships be-
tween these groups. Comparison of
results for the new studies shows the
formerly indistinguishable groups
from two areas to be as distinct from
one another as from other major
groupings of the species with a mean
genetic distance between popula-
tions of each river (0.014) that is
double that of the maximum within-
group genetic distance. Two newly-
resolved gene loci (niMDH-2* and
sMEP-1*) are particularly good at
distinguishing populations from the
two rivers. In addition to resolving
the anomalous similarity between
populations inhabiting geograph-
ically separated areas, the new re-
sults illustrate the care that must be
used in drawing inferences from
negative data.
Genetic isolation of previously
indistinguishable Chinook salmon
populations of the Snake and Klamath
Rivers: Limitations of negative data
Fred M. Utter
Robin S. Waples
David J. Teel
Coastal Zone and Estuarine Studies Division
Northwest Fisheries Science Center, National Marine Fisheries Service, NOAA
2725 Montlake Boulevard East, Seattle, Washington 981 12
Manuscript accepted 5 July 1992.
Fishery Bulletin, U.S. 90:770-777(1992).
A variety of characteristics can be
useful in distinguishing particular
groups of organisms from other
related groups. In humans, for in-
stance, major ancestral groups can be
identified by heritable morphological
traits, as well as by characteristic fre-
quencies of alleles detected by molec-
ular or immunological procedures.
Conversely, although two groups
lacking any distinguishing character-
istics may, in fact, be closely related,
the possibility of undetected differ-
ences often prevents a conclusive
determination of the degree of re-
latedness. For example, two cryptic
species of bonefishes in Hawaii were
considered members of a common
gene pool until biochemical genetic
analysis revealed that the two forms
diverged perhaps 20 million years
ago (Shaklee and Tamaru 1981).
Other examples of genetic distinc-
tions between and within species of
fishes previously considered to be
homogeneous are listed in Allendorf
et al. (1987).
The motivation behind our present
study was a puzzling instance of ap-
parent genetic similarity between
two geographically separated groups
of Chinook salmon Oncorhynchus
tfihawytscha. Indigenous chinook
salmon from the Klamath River and
spring- and summer-run chinook
salmon from the Snake River are
well differentiated from nearby
populations at several protein-coding
gene loci (Utter et al. 1989, Hartley
and Gall 1990, Waples et al. 1991,
Hartley et al. 1992). However, a com-
parison of the two river groups by
Utter et al. (1989) failed to distin-
guish them despite their substantial
geographic separation. The mouths
of the Snake and Klamath Rivers are
separated by a distance of almost 600
river-ocean miles, and a number of
ancestrally distinct groups of popula-
tions (Utter et al. 1989) are found in
intervening areas.
This apparent genetic similarity
was even more puzzling because of
substantial life-history differences
between chinook salmon from the
two rivers. The populations that were
not well differentiated in the Utter
et al. (1989) study included four
spring-run and two summer-run
populations from the Snake River
and two fall- and one spring-run
population from the Klamath River.
Utter et al. (1989) also sampled
fall-run fish from the Snake River,
but this population is genetically
quite different both from Snake
River spring- and summer-run fish
and chinook salmon from the Klam-
ath River. Whereas the fall-run fish
migrate to sea as subyearlings, the
other populations produce juve-
niles that spend an additional winter
in freshwater and outmigrate as
yearlings.
770
Utter et al.: Genetic isolation of Oncorhynchus tshawytscha of Snake and Klamath Rivers
771
Utter et al. (1989) speculated that the anomalously
high degree of genetic similarity between Klamath and
Snake River populations was due to coincidentally high
frequencies of the same common alleles (possibly a
reflection of restricted gene flow among populations
and reduced population sizes over an extended time in-
terval) rather than to a recent common ancestral origin.
Of the 25 polymorphic loci examined, only 18 were
variable in either the Snake or Klamath River groups,
and populations from these two areas had the lowest
average heterozygosities (0.027-0.045; Utter et al.
1989, App. A) of any populations included in the study.
Utter et al. (1989) predicted that additional genetic
surveys would ultimately reveal divergent frequencies
of alleles in the two areas. If such differences were not
found in more extensive studies, alternate explanations
for this apparent similarity would be required.
This paper retests and rejects the null hypothesis of
no genetic difference between these two groups based
on two recently published studies, which sample several
new populations and an additional 15 polymorphic loci.
Comparison of results for the new studies shows the
formerly-indistinguishable chinook salmon populations
of the Klamath and Snake River to be quite distinct,
with a mean genetic distance between populations of
each river (0.014) that is double that of the maximum
within-group genetic distance. In addition to resolving
the anomalous apparent similarity between these
chinook salmon populations of these geographically
separated areas, the new results illustrate the care that
must be used in drawing inferences from negative data.
Table 1
Collection data for samples of chinook salmon Oncorhynchus \
tshawytscha from the Klamath (Kl-
KIO; Bartleyetal. 1992)
and Snake (Sl-Sll; Waples et al. 1991) Rivers. Samples from |
hatchery stocks are marked by a dagger (T); other
samples
were from naturally spawning populations. Locations
included
in the study of Utter et al. (1989) are
' indicated by an
asterisk
(*). Run timing indicates the season of entry of adults into |
freshwater.
Map
Run
Sample
code Location
timing
size
Klamath River
Kl Omagar Creek
Fall
100
K2 Blue Creek
Fall
100
K3 Camp Creek
Fall
106
K4 Horse Linto Creek
Fall
100
K5 S. Fork Trinity River
Fall
100
*K6 Trinity RiverT
Fall
120
K7 Upper Salmon River
Fall
98
K8 Shasta River
Fall
100
K9 Bogus Creek
Fall
128
KIO Iron Gate Hatcheryt
Fall
99
Snake River
•SI Valley Creek
Spring
99
•S2 Sawtooth Hatcheryt
Spring
100
S3 Salmon River
Spring
99
S4 Marsh Creek
Spring
100
'S5 Johnson Creek
Summer
97
*S6 McCall Hatcheryt
Summer
100
S7 Secesh River
Summer
92
•S8 Rapid River Hatcheryt
Spring
100
89 Imnaha River
Summer
100
SIO Imnaha Hatcheryt
Summer
100
Sll Lostine River
Spring
100
Materials and methods
Our analyses used the data from Hartley et al. (1992)
for Klamath River populations and Waples et al. (1991)
for Snake River populations; comparisons also were
made with earlier data from Utter et al. (1989). Sam-
pling locations included 10 areas from the Klamath
River and 11 from the Snake River drainages (Table
1, Fig. 1). Samples of juvenile fish from hatcheries and
naturally-spawning populations were collected between
1986 and 1989 for the Klamath River, and 1989 and
1990 for the Snake River. Starch gel electrophoresis
for all three studies followed procedures described by
Aebersold et al. (1987). The data used in these analyses
were part of a larger baseline dataset used by manage-
ment agencies to help determine natal origins of
chinook salmon harvested in mixed-stock fisheries
(Shaklee and Phelps 1990).
Genetic nomenclature and abbreviations followed a
system suggested by Shaklee et al. (1989). Data were
collected from 21 enzyme systems and 30 presumptive
gene loci that were polymorphic in at least one of the
populations (Tables 2,3). The observed polymorphisms
were attributed to 26 disomic loci and 2 isolocus pairs
(sAAT-1,2* and sMDH-Bl,2* ; see Allendorf and Thor-
gaard 1984). A single, average allele frequency was
computed for each isolocus pair for purposes of com-
paring populations.
Genetic data were analyzed using the BIOSYS pro-
gram of Swofford and Selander (1981). Analyses in-
cluded calculation of unbiased pairwise genetic dis-
tances between populations (Nei 1978), unweighted
pair group method (UPGM) projection of a matrix of
these distances (Sneath and Sokal 1973), average
heterozygosities, and the number of alleles per locus.
Results and discussion
Our analyses focused on a comparison of genetic char-
acteristics between chinook salmon from the Klamath
and Snake Rivers. Discussion of population structure
within these two areas appears elsewhere, as do more
772
Fishery Bulletin 90(4). 1992
WASHINGTON
OREGON
OREGON
/
KiOrfiTI^
k sn^^^
'
^r
VK8 o
y~vK2LX
'y
I "
^
W'
0'
\
' MK7 ^^
l/
) "'
\^orse Linio Cr )
b
/•K6
Vk5 ^-w^
t/l\
-T\ \
'^\
CALIFORNIA
Figure 1
Sampling locations of chinook salmon Oncorhynch-us tshawyUcha in the Klamath and Snake River drainages. See Table 1 for names
of locations.
Utter et a\ : Genetic isolation of Oncorhynchus tshawytscha of Snake and Klamath Rivers
773
Table 2
Enzymes and loci examined (enzyme nos.
in parentheses) of chinook salmon Oncorhynchus tshawytscha.
Enzyme
Locus
Enzyme
Locus
Aspartate aminotransferase (2.6.1.1)
sAAT-1,2'
Tripeptide aminopeptidase (3.4.11.4)
PEPB-1*
sAAT-S*
Leucine-tjrrosine dipeptidase (3.4.-.-)
PEP-LT*
sAAT-i*
Malate dehydrogenase (1.1.1.37)
sMDH-BU-Z*
Adenosine deaminase (3.5.4.4)
ADA-1*
mMDH-1*
Alcohol dehydrogenase (1.1.1.1)
ADH*h
mMDH-2*
Aconitate hydratase (4.2.1.3)
sAH-1*
Malic enzyme (1.1.1.40)
sMEP-1*
mAH-h*
Mannose-6-phosphate isomerase (5.3.1.8)
MPI*
Glyceraldehyde-3-phosphate dehydrogenase (1.2.1.12)
GAPDH-3*
Phosphogluconate dehydrogenase (1.1.1.44)
PGDH*
Dipeptidase (3.4.13.11)
PEP A*
Phosphoglycerate kinase (2.7.2.3)
PGK-2*
Glutathione reductase (16.4.2)
GR*
Phosphoglucomutase (2.7.5.1)
PGM-2*
Hydroxyacylglutathione hydrolase (3.1.2.6)
HAGH*
L-Iditol dehydrogenase (1.1.1.14)
IDDH-r
Isocitrate dehydrogenase (1.1.1.42)
sIDHP-1'
Superoxide dismutase (1.15.1.1)
sSOD-l"
sIDHP-2'
Triose-phosphate isomerase (5.3.1.1)
TPH'
Lactate dehydrogenase (1.1.1.27)
LDH-B-r
LDH-C'
Table 3
Range of common allele frequencies in
samples of chinook sail
Tion Oncorhynchus tshau'ytscha from
the Snake and Klamath Rivers
reported in three investigations. Parenthetical entries summarize
data from studies (2) and (3), respectively, for those populations studied
in (1). Subset (A)
are loci common to all
studies; subset (B) are
isolocus pairs unique to study (1); subset (C) are loci newly resolved
in studies (2) and
(3).
(1)
(2)
(3)
Locus
Utter et al.
1989
Waples et al. 1991
Bartley et al. 1992
Snake
Klamath
Snake
Klamath
(A) sAAT-1.2*
0.981-1.000
0.995-1.000
0.957-1.000 (0.957-1.000)
1.000 (1.000)
sAAT-S*
0.994-1.000
0.995-1.000
0.965-1.000 (0,980-1.000)
0.985-1.000 (0.985-1.000)
ADA-1'
0.953-0.969
1.000
0.846-1.000 (0.894-1.000)
0.995-1.000 (1.000)
sAH-r
0.994-1.000
0.995-1.000
0.985-1.000 (0.990-1.000)
0.940-1.000 (0.940-1.000)
PEP A*
0.994-1.000
0.990-1.000
0.995-1.000 (0.995-1.000)
0.770-1.000 (0.930-1.000)
GR*
1.000
0.995-1.000
0.995-1.000 (0.985-1.000)
0.995-1.000 (1.000)
LDH-B2*
0.972-1.000
1.000
0.970-1.000 (0.970-0.995)
1.000 (1.000)
LDH-C*
0.976-1.000
1.000
0.920-1.000 (0.920-1.000)
0.890-1.000 (0.980-1.000)
PEPB-1'
0.944-0.976
0.949-0.990
0.904-0.985 (0.904-0.985)
0.860-1.000 (0.980-1.000)
sMDHB-1.2'
0.995-0.998
0.997-1.000
0.942-0.997 (0.944-0.990)
0.993-1.000 (0.997-1.000)
MPI'
0.910-0.953
0.975-0.990
0.770-0.990 (0.884-0.990)
0.860-1.000 (0.970-0.992)
PGK-2'
0.062-0.139
0.146-0.350
0.065-0.187 (0.065-0.187)
0.148-0.400 (0.148-0.320)
sSOD-1'
0.944-0.976
0.895-0.990
0.885-0.980 (0.939-0.980)
0.755-0.992 (0.845-0.992)
(B) sIDHP-1.2*
0.913-0.937
1.000
—
—
PGM-1.2*
1.000
0.942-0.990
-
-
(C) TPLJ,'
—
—
0.825-0.955
0.995-1.000
sAAT-J,*
—
—
0.919-1.000
0.985-1.000
ADH'
—
—
0.985-1.000
1.000
niAH-J,'
—
—
0.985-1.000
0.775-1.000
GAPDH-3*
—
—
1.000
0.871-l.OOOt
HAGH'
—
—
0.902-1.000
1.000
sIDHP-1*
—
—
0.783-0.950
0.992-1.000
sIDHP-2'
—
—
0.945-1.000
0.900-1.000
PEP-LT'
—
—
0.870-0.985
0.985-1.000
mMDH-1'
—
—
0.995-1.000
0.795-1.000
m.MDH-2'
—
—
0.490-0.800
0.905-1.000
sMEP-1'
—
—
0.010-0.079
0.150-0.465
PGDH*
—
—
1.000
0.910-1.000
PGM-2*
—
—
1.000
0.860-1.000
IDDH-1*
;t al.
1989
0.897-1.000
0.990-1.000
t Data from Gall
774
Fishery Bulletin 90(4). 1992
complete details of the individual studies. (Bartley et
al. 1992, Waples et al. 1991).
Variability within populations
The levels of genetic variation within populations were
evaluated using only the loci found to be polymorphic.
Because this restriction does not represent a random
sample of gene loci, values reported here are applicable
only for comparisons among populations included in
this study or with other studies using the same set of
loci. Indices of genetic variability were consistently
slightly higher in the Snake River samples; the average
number of alleles per locus was 1.63 vs. 1.51 for the
Klamath River, and the average heterozygosity was
0.079 vs. 0.065 (0.05>p>0.01 in both instances, based
on Mann- Whitney tests). Heterozygosities ranged from
0.058 to 0.090 in the Snake River populations and were
less uniform in the Klamath River groups, where both
the lowest (0.039 in Shasta River) and the highest
(0.126 in Omagar Creek) values were found. Neither
of these latter two populations were represented in the
initial study of the Klamath River by Utter et al. (1989).
The actual heterozygosity values reported here are
higher than those reported by Utter et al. (1989),
primarily because a number of new, very polymorphic
systems are included in the more recent analyses.
Nevertheless, Utter et al. (1989) also found a slightly
higher average heterozygosity in Snake River spring-
run and summer-nm chinook salmon (0.035-0.045) than
in those from the Klamath River (0.027-0.032). Based
on the new data, Waples et al. (1991) concluded that,
in comparison with other Columbia River populations.
Snake River spring-run and summer-run chinook
salmon have somewhat reduced levels of genetic vari-
ability, but that the difference is apparently not as large
as suggested by earlier studies (Utter et al. 1989,
Winans 1989).
Vanability between regions
Allele frequency distributions differed substantially
between the two regions at a number of gene loci.
Although three or more alleles were found at some of
these loci, most of the important differences were
reflected in differing frequencies of the common allele
(Table 3). Particularly large differences were found at
mMDH-2* and sMEP-1* (Fig. 2); for these loci, the
range of allele frequencies was nonoverlapping be-
tween regions, with substantially higher frequencies
of the common (i.e., iOO*) allele found in the Klamath
River samples at both loci.
Genetic differences between the two regions based
on data for all 30 loci are summarized in a phenogram
resulting from clustering of pairwise genetic distances
K6 K8 K10 K2 K1
1.0
° 7k7 7"k3 °
^^ K5 K4
0.9
Q
0.8
C\J
^88
rfoSII
D 0.7
E
dSIO
S9°.S4
0.6
hS5
nS1
bS2
0.5
" 0 33
• . 1 . 1
^0.0 0.1 0.2 0.3 0.4 0.5
sMEP-r
Figure 2
Plot of frequencies of common alleles of Klamath (K) and
Snake (S) River populations of chinook salmon Oiicorhynckus
tshawytscha at the mMDH-2' and sMEP-1' loci.
1 Blue Creek K2
-j j Camp Creek K3
1 South Fork Tnnity R K5
J Bogus Creek K9
I Shasta River K8
1 Iron Gate Hatchery K10
— Trinity River K6
1 — Marsh Creek S4
Hr Valley Creek SI
I Sawtooth Hatchery S2
|- Imnaha River S9
L Imnaha Hatchery S10
r- Secesh River S7
"L McCall Hatchery S6
Rapid River S8
1 — Johnson Creek S5
1 — Upper Salmon River S3
1 noma R.uor R11
1 1 1
002 0015 001
1 1
0 005 0
Genetic distance
Figure 3
UPGM projection of Nei
s genetic distances between Klamath
and Snake River populations of chinook salmon Oncorhynfhus |
tshawytscha.
(Fig. 3). The Snake and Klamath River populations are
separated by a mean genetic distance of 0.014, whereas
the within-river separations average 0.004 and 0.007,
Utter et al.: Genetic isolation of Oncorhynchus tshawytscha of Snake and Klamath Rivers
775
respectively. The present data, then, clearly identify
two genetically-distinct groups on the basis of the 30
polymorphic loci that were examined.
This genetic distinction clearly rejects a hypothesis
of a recent common ancestry for populations of these
regions. The topography of the clustering within
Klamath and Snake River groups and the relative
genetic distance between them are very similar to those
distinguishing Klamath River populations from other
genetically-distinct population groups of California and
the Oregon Coast based on a similar set of polymor-
phic loci (Hartley et al. 1992).
Comparison with previous information
Because the clear separation of Snake and Klamath
River populations reported here contrasts sharply with
the minimal differences detected between these groups
by Utter et al. (1989), an examination of results from
that earlier study is warranted. A direct comparison
of the original study with the two more recent studies
is complicated by (1) the addition of a number of new
gene loci in the more recent studies, (2) the greater
discriminatory capabilities for some loci used in the
newer studies, and (3) the more extensive sampling of
populations in the newer studies. A comparison of the
15 loci common to both the original and more recent
studies was made for the five Snake River sampling
sites (81, S2, S5, S6, S8) and two Klamath River sites
(K6, KIO) that were sampled in both investigations. In
general, very similar allele frequencies were found at
most loci in the two sets of samples (Table 3). None of
the allele frequency differences between the original
and the more recent studies exceeded 0.06 (atPEPA*
in the Klamath River comparisons). Thus, the more re-
cent samples confirm the minimal differences between
the two regions reported by Utter et al. (1989) based
on the loci and populations originally examined.
The improved resolution in the more recent studies,
therefore, can be attributed to an increase in the num-
ber and type of usable genetic characters. Particular-
ly important was the addition of 15 gene loci not in-
cluded in the earlier study (Table 3). Although regional
differences are strongest at mMDH-2* and sMEP-1*,
clear contrasts between the regions are also seen
at five other loci {mAH-Jt* . GAPDH-3*, HAGH*,
PEP-LT*, and TPI4*). In addition, the more recent
studies resolve individual loci that had previously been
considered isolocus pairs, which further enhanced the
discriminating power of two genetic systems. This ef-
fect was most apparent for the enzyme IDH. Utter et
al. (1989), as have other previous studies (e.g.. Utter
et al. 1987), reported variation for the isolocus pair
sIDHP-1,2* ; subsequently, Shaklee et al. (1990) showed
that it is possible to resolve the two loci individually.
Whereas the most extreme frequency difference
between the two regions at sIDHP-1,2* was 0.087
(1.0-0.913; Table 3) in the original study, the maximum
difference at sIDHP-1* in the newer studies was 0.217
(1.0-0.783). Similarly, the protocol of Gall et al. (1989)
for partitioning variation at the PGM-1,2* isolocus
increased the discriminatory power of this genetic
system.
General implications of the results
During the 1960s, the newly found capability to resolve
numerous genetic systems exhibiting Mendelian in-
heritance led to a flood of studies that continues to this
day (see Lewontin 1991). Protein electrophoresis has
been used extensively in fishery research and manage-
ment (Utter 1991); such data have proven particularly
useful in modifying previously held assumptions about
the genetic structure of fish species (Allendorf et al.
1987). The results discussed here are instructive with
regard to both the power and the limitations of such
information.
The power of Mendelian data lies in the identifica-
tion of genetic differences among individuals, popula-
tions and species. The regional differences among
populations of North American chinook salmon orig-
inally described by Utter et al. (1989) have also been
apparent in subsequent studies (Hartley and Gall 1990,
Waples et al. 1991, Bartley et al. 1992). These differ-
ences have generally been interpreted to reflect more
recent ancestries of populations within a particular
genetically-defined region than between populations of
different regions.
However, in spite of the power of electrophoretic
data to detect genetic differences when present, there
are limits to the conclusions that one can draw from
the failure to detect such differences. That is, although
a finding of a statistically-significant allele frequency
difference may provide evidence that gene flow is
restricted (or that some other evolutionary force is
operating), the inability to identify such differences
does not prove that genetic differences do not exist.
The present example, in which genetically divergent
groups were not well distinguished in a previous study,
emphasizes the potential significance of this limitation.
Although Utter et al. (1989) hypothesized that the ap-
parent similarity between Klamath and Snake River
chinook salmon was a coincidence that did not reflect
a common ancestral origin, the distinctness of the two
groups could not be demonstrated until new data
became available. The situation is analogous to a
classical genetic comparison between populations of
Drosophila pseudoobscura from Berkeley, California
and Bogata, Colombia, in which an initial apparent
genetic similarity was puzzling in view of the exten-
776
Fishery Bulletin 90(4|, 1992
sive geographic separation of the two regions (Lewon-
tin and Hubby 1966). A subsequent study that found
previously-unknown genetic variants (Singh et al. 1976)
demonstrated clear genetic differences between popu-
lations of each region.
The important message here is to beware of the
danger of drawing positive conclusions from negative
data. It should also be emphasized that problems of this
nature are not confined to genetic data; rather, the
limitations of nondiscriminatory information (i.e., the
power to reject the null hypothesis) should be con-
sidered in evaluating any kind of comparative data for
two or more samples.
Similar allele frequencies among samples, then, sup-
port but do not confirm hypotheses that the samples
are drawn from a common breeding group. This well-
established principle requires restatement from time
to time (e.g.. Utter 1981, Waples 1991). Such aware-
ness serves to safeguard against a premature conclu-
sion of identity for groups that are distinct and thus
may be subject to different management criteria.
In these instances it is important to recognize the
power of Mendelian data involving multiple polymor-
phic loci to detect differences between populations
when they do exist. For example, assuming that most
allozyme variation is neutral, it will take populations
that are divided into large units a considerable amount
of time before significant divergence will occur. Thus,
Atlantic herring Clupea harengus populations of the
eastern and western Atlantic Ocean that have likely
been isolated for thousands of years could not be distin-
guished because of similar allele frequencies at a num-
ber of polymorphic loci (Grant 1984). The observed
value for Wright's (1943) fixation index (F^i) of 0.0042
approximates an Fgt value of 0.003 expected for
neutral markers among populations of effective size of
1 million individuals separated over 3000 generations
(Nei and Chakravarti 1977). Such dynamics preclude
genetic distinction of these herring populations through
neutral genetic markers (and thus rejection of the null
hypothesis) even with very large samples of loci and
individuals. Under such circumstances, other criteria
(e.g., tagging data) are needed to determine whether
one or more populations is being sampled.
Finally, we note the complementary nature of rela-
tionships among populations indicated by many pheno-
typic traits on one hand and by most molecular genetic
markers on the other hand. A strong selective compo-
nent appears to be involved in the maintenance of
phenotypic traits such as timings of spawning and
migration (e.g., Ricker 1972, Helle 1981); consequent-
ly, relationships inferred from such traits tend to
reflect relative similarities in adaptations among pop-
ulations. Conversely, the apparent absence of strong
selection at most electrophoretically detectable loci
permits the estimation of relative degrees of gene flow
within and among regions (e.g., Chakraborty et al.
1978, Allendorf and Phelps 1981), and such estimations
provide useful insights about ancestral relationships.
In view of the complementary nature of these different
categories of genetic information, adequate sets of both
molecular markers (for clarifying ancestral relation-
ships) and phenotypic traits (for identifying adaptive
differences within lineages) should be included in
genetic surveys of a particular species whenever pos-
sible. Such adaptive differences have been noted within
a number of apparent ancestral groupings of chinook
salmon, including both spring- and fall-spawning migra-
tions within the Klamath River populations of the
species (Utter et al. 1989).
Acknowledgments
Research funded in part through contract DE-AI79-
89BP0091 with Bonneville Power Administration.
Citations
Aebersold, P., G. Winans, D. Teel, G. Milner, and F. Utter
1987 Manual for starch gel electrophoresis: A method for the
detection of genetic variation. NOAA Tech. Rep. NMFS 61,
19 p.
Allendorf, F., and S. Phelps
1981 Use of allelic frequencies to describe population struc-
ture. Can. J. Fish. Aquat. Sci. 38:1507-1514.
Allendorf, F., and G. Thorgaard
1984 Tetraploidy and the evolution of salmonid fishes. In
Turner, B (ed.). Evolutionary genetics of fishes, p. 1-53.
Plenum Press, NY.
Allendorf, F., N. Ryman, and F. Utter
1987 Genetics and fishery management: Past, present, and
future. 7« Ryman, N., and F. Utter (eds.). Population genetics
and fishery management, p. 1-20. Wash. Sea Grant Prog.,
Univ. Wash. Press. Seattle.
Bartley, D., and G. Gall
1990 Genetic structure and gene flow in chinook salmon pop-
ulations of California. Trans. Am. Fish. Soc. 119:55-71.
Bartley, D., B. Bentley. J. Brodziak, R. Gomulkiewicz.
M. Mangel, and G.A.E. Gall
1992 Geographic variation in population genetic structure of
chinook salmon from California and Oregon. Fish. Bull.. U.S.
90:77-100 (authorship amended per errata. Fish. Bull. 90(3):iii).
Chakraborty, R.. P. Fuerst, and M. Nei
1978 Statistical studies on protein polymorphism in natural
populations: 11. Gene differentiation between populations.
Genetics 88:367-390.
Gall. G.A.E. . B. Bentley, C. Panattoni, E. Childs, C. Qi. S. Fox,
M. Mangel, J. Brodziak, and R. Gomulkiewicz
1989 Chinook mi.xed fishery project 1986-1989. Report of the
Department of Animal Science and the Department of Applied
Mathematics, Univ. Calif., Davis, under Interagency Agree-
ment C-1335 with Calif. Dep. Fish & Game, 192 p.
Grant, W.
1984 Biochemical population genetics of Atlantic herr-
ing. Copeia 1984:357-364.
Utter et al,: Genetic isolation of Oncorhynchus tshawytscha of Snake and Klamath Rivers
777
Helle, J.
1981 Significance of the stock concept in artificial propaga-
tion of salmonids in Alaska. Can. J. Fish. Aquat. Sci. 38:
1665-1671.
Lewontin, R.
1991 Electrophoresis in the development of evolutionary
genetics: Milestone or millstone? Genetics 128:657-662.
Lewontin, R., and J. Hubby
1966 A molecular approach to the study of genetic hetero-
zygosity in natural populations. II. Amount of variation and
degree of heterozygosity in natural populations oi Drosophila
pseudoobscura. Genetics 54:595-609.
Nei. M.
1978 Estimation of average heterozygosity and genetic dis-
tance from a small number of individuals. Genetics 89:
583-590.
Nei, M., and A. Chakravarti
1977 Drift variances of F^, and G,, statistics obtained from a
finite number of isolated populations. Theor. Popul. Biol.
11:307-25.
Ricker, W.
1972 Hereditary and environmental factors affecting certain
salmonid populations. In Simon, R., and P. Larkin (eds.), The
stock problem in Pacific salmon, p. 19-161. H.R. MacMillan
Lectures in Fisheries, Univ. Brit. Col., Vancouver.
Shaklee, J., and S. Phelps
1990 Operation of a large-scale, multiagency program for
genetic stock identificiation. In Parker, N.C., et al. (eds.), Fish
marking techniques, p. 817-830. Am. Fish. See. Symp. 7,
Bethesda.
Shaklee, J., and C. Tamaru
1981 Biochemical and morphological evolution of Hawaiian
bonefishes {Albula). Syst. Zool. 30:125-146.
Shaklee, J., F. Allendorf, D. Morizot. and G. Whitt
1989 Genetic nomenclature for protein-coding loci in fish: Pro-
posed guidelines. Trans. Am. Fish. See. 118:218-227.
Shaklee, J., C. Busack, A. Marshall, M. Miller, and S. Phelps
1990 The electrophoretic analysis of mixed-stock fisheries of
Pacific salmon. In Ogita, Z.I., and C.L. Markert (eds.). Iso-
zymes: Structure, function, and use in biology and medicine,
p. 235-265. Wiley-Liss, NY.
Singh, R., R. Lewontin, and A. Felton
1976 Genetic heterogeneity within electromorphic alleles of
xanthine dehydrogenase in Drosophila pseudoobscura. Gen-
etics 84:609-629.
Sneath, P., and R. Sokal
1973 Numerical taxonomy. W.H. Freeman, San Francisco,
573 p.
Swofford, D., and R. Selander
1981 BIOSYS-1-A FORTRAN program for the comprehen-
sive analysis of electrophoretic data in population genetics and
systematics. J. Hered. 72:281-283.
Utter, F.
1981 Biological criteria for definition of species and distinct
intraspecific populations of salmonids under the U.S. En-
dangered Species Act of 1973. Can. J. Fish. Aquat. Sci. 38:
1626-1635.
1991 Biochemical genetics and fishery management: An
historical perspective. J. Fish. Biol. 39(Suppl. A):l-18.
Utter, F., D. Teel, G. Milner, and D. Mclsaac
1987 Genetic estimates of stock compositions of 1983 chinook
salmon, Oncarhynfkus tshawytscha, harvest off the Washington
coast and the Columbia River. Fish. Bull., U.S. 85:13-23.
Utter. F., G. Milner, G. StShl, and D. Teel
1989 Genetic population structure of chinook salmon in the
Pacific Northwest. Fish. Bull., U.S. 87:238-264.
Waples, R.
1991 Pacific salmon and the definition of "species" under the
Endangered Species Act. Mar. Fish. Rev. 53(3): 11-22.
Waples, R., D. Teel, and P. Aebersold
1991 A genetic monitoring and evaluation program for supple-
mented populations of salmon and steelhead in the Snake River
Basin. Annu. Rep. Res., Contract DE-A179-89BP00911, U.S.
Dep. Energy, 50 p. (Avail. Bonneville Power Admin., P.O. Box
3621, Portland OR 97208.)
Winans, G.
1989 Genetic variability in chinook salmon stocks from the
Columbia River Basin. N. Am. J. Fish. Manage. 9:47-52.
Wright, S.
1943 Isolation by distance. Genetics 28:114-138.
Differentiating Paralithodes iarvae
using teison spines: A taii of two species*
Gregory C. Jensen
Helle B. Andersen
David A. Armstrong
School of Fisheries WH-IO, University of Washington, Seattle, Washington 98195
Identification of larvae belonging to
closely related species of decapod
Crustacea is frequently dependent
upon few (often single) morpholog-
ical characters. Since larvae used in
descriptions are often the offspring
from a single captive, very little is
known about how intraspecific vari-
ability may affect the ability to dif-
ferentiate species. Two congeneric
decapods whose zoeae are separ-
ated on the basis of a single mor-
phological character are the red
king crab (RKC) Paralithodes cam-
tschaticus, and blue king crab
(BKC)P. platypus. Larvae of these
' Contribution 867 of the School of Fisheries,
University of Washington.
commercially-important species are
distinguished from each other by
the number of spines or processes
on the teison, RKC having 7 pairs
of spines and BKC 8 pairs, ex-
cluding a minute seta (Sato 1958,
Haynes 1984). However, when
viewed together there are other ap-
parent differences: BKC zoeae have
proportionately shorter rostrums
and carapace spines, larger bodies
at each stage, and larger eyes.
During the course of extensive
plankton sampling for king crab lar-
vae in Herendeen Bay within Port
Moller, Alaska, considerable vari-
ability in teison morphology was
noted. A large proportion of zoeae
resembling BKC were captured
BERING SEA
<?
PrtiMol IHanas
- 56'N
PACIFIC OCEAN
160 W
-r
Figure 1
Locations of Paralithodes spp. zoeae collections.
having an asymmetrical pattern of
8-1-7 teison spines, while others,
also appearing to be BKC in other
respects, had only 7 pairs; such in-
terspecific character overlap had
not been noted with these two
species near the Pribilof Islands,
Alaska (Armstrong et al. 1985). To
confirm the identity of these zoeae
and provide additional characters to
separate the two species, several
measurements were taken on spe-
cimens from Herendeen Bay and
the Pribilof Islands to establish a
stronger, more quantitative basis to
distinguish these larvae.
Materials and methods
Paralithodes zoeae were collected
near the Pribilof Islands, Alaska, in
May 1983 and April 1984, and from
Herendeen Bay, Alaska (Fig. 1) in
May and June of 1990, using either
a 505 /jm mesh Tucker trawl or a
60 cm bongo net with a mesh of 333
or 505f/m. Samples were preserved
in 5% buffered formalin in seawater
and later sorted for target species,
which were transferred to a solution
of 70% ethanol and 5% glycerol.
The number of teison spines (ex-
cluding a minute seta) was recorded
for each specimen, and three mea-
surements were taken: tip of ros-
trum to the anteriormost edge of
the eye ("rostrum length"), anterior
edge of eye to the tip of the pos-
teriolateral carapace spines ("cara-
pace length"), and the longest
dimension of the eye ("eye length,"
Fig. 2). Damaged or distorted spe-
cimens were not used for measure-
ments, but spine counts were re-
corded. A total of 608 larvae were
measured; 371 from Herendeen
Bay and the remainder from around
the Pribilof Islands. The ratios of
rostrum length to carapace length
were plotted against carapace
length for "normal" (i.e., 7-1-7
teison spine RKC and 8 -h 8 BKC)
Manuscript accepted 1 July 1992.
Fishery Bulletin, U.S. 90:778-783 (1992).
778
NOTE Jensen et al.. Differenuating Paralithodes larvae using telson spines
779
Figure 2
Measurements taken from Paralithodes spp. zoeae. (A)
rostrum length, (B) eye length, (C) carapace length.
0.7 n
C
• 7*7 Red King Crab
-■ 0.6-
• n 8*8 Blue King Crab
o
o
CO
Q.
2 0.5-
ni
O
f 0.4-
0)
_i
E
••• •
= 0.3-
Crt
o
a:
n o -
1
1 -■ 1 • 1
7 1.9 2.1 2.3 2.5 2.7
Carapace Length (mm)
Figure 3
Ratios of rostrum length/carapace length for first-stage
Paralithodes spp. collected near the Pribilof Islands, Alaska.
zoeae of both species in the two areas. Those with an
anomalous number of spines (8-i-7 and "BKC" with
7-1-7) were then similarly plotted for comparison.
Unless otherwise indicated, data are presented as
mean ± 1 standard deviation. ANOVAs were used to
test for significant differences (a < 0.05) between the
means of different groups, and the means compared
using Tukey's T method. A two-way ANOVA was used
to examine intraspecific variation in zoeae I between
the two areas.
Results
Plotted ratios of rostrum length/carapace length for
Pribilof Island specimens fell into two distinct clouds
representing those thought to be RKC (longer rostrum)
and BKC (Fig. 3). These were not so clearly separated
in the samples from Herendeen Bay; nevertheless, all
those with an asymmetric telson spine pattern or a
short rostrum, large body, and spine count of 7 -i- 7 fell
within the cloud of 8 -i- 8 BKC larvae (Fig. 4). These lar-
vae were designated as BKC and divided into groups
based on number of telson spines. The ratios for all of
these groups were significantly different (p<0.05) at
each zoeal stage from the means for RKC zoeae; Figure
5 shows combined results for both species at the two
locations. Both the carapace and eyes of BKC larvae
were significantly longer than those of RKC at all
c
Q.
to
O
o
cc
0.7
0.6-
0.5
0.4-
= 0.3-
0.2
•••
• 7+7 Red King Crab
O 7*7 Blue King Crab
A 7+8 Blue King Crab
D 8+8 Blue King Crab
A n
1.7 1.9 2.1 2.3 2.5
Carapace Length (mm)
2.7
Figure 4
Ratios of rostrum length/carapace length for first-stage
Paralithodes spp. zoeae collected from Herendeen Bay,
Alaska.
stages, but these differed considerably between the two
areas. Two-way ANOVA revealed no significant dif-
ference in rostrum/carapace length ratio for zoeae-I
BKC from the two areas, but significantly smaller
780
Fishery Bulletin 90(4). 1992
0.7
• Herendeen Red King Crab
■ Pribilof Red King Crab
0 Herendeen Blue King Crab
C
] Pribilof Blue King Crab
0.6
-
£1
Ol
c
0)
_I
• 47 V'
48
Length/Carapace
Ul
-
1
,55
49
10
E
o
I
(J
D13
0.4
-
85 (
60
1 14
57
'
41 (
)27
0.3
1
1
III IV
Stage
0.7
0.6
0.5
0.4
# Herendeen Red King Crab
■ Pnbilof Red King Crab
O Herendeen Blue King Crab
D Pribilof Blue King Crab
A Herendeen Blue King Crab 7 + 7
D 55
Ol5
A 14
0 46
162
22
• 56
18
Stage
Figure 5
Paralithodes spp. rostrum length/cara-
pace length by stage. Values are mean
ratios ( ± 1 SD); numbers are the number
of zoeae measured for each value.
eyes for Herendeen Bay BKC, suggesting that eye
length is not rehable for distinguishing the two species.
No intraspecific differences between areas were de-
tected for either rostrum/carapace length or eye length
for RKC.
The mean eye length of RKC larvae was significant-
ly less (p<0.05) at each stage than the means of all
groupings of BKC. In addition, the mean of the 7 + 7
zoeae-I BKC was significantly less than those with 8 + 8
(Fig. 6). BKC larvae from Herendeen Bay showed
great variation in number of telson spines. Only 45.9%
of all BKC larvae had a spine count of 8 + 8, the major-
ity having some other combination of 7, 8, or 9 spines
on each side of the telson. In contrast, the majority of
BKC zoeae from the Pribilof Islands (84%) had a spine
count of 8 + 8 (Fig. 7). RKC larvae showed little varia-
tion in telson spines, with only 0.9% of the zoeae from
Herendeen Bay and 3.2% from the Pribilof Islands
deviating from the count of 7 + 7.
Discussion
BKC zoeae are generally distinguished from RKC
zoeae by the presence of an additional pair of inner
spines on the telson (Fig. 8A,B), and for king crab
zoeae collected near the Pribilof Islands this was a
reliable character for separating the two species. Only
3.2% of the RKC zoeae collected from this area
deviated from the 7 + 7 pattern. Even though 16% of
the BKC zoeae differed from 8 + 8, this difference was
almost invariably in the form of an extra 1 or 2 inner
spines, making confusion with RKC unlikely. BKC
zoeae were visibly much larger as confirmed by their
greater carapace length, and had shorter rostrums and
larger eyes.
A substantially different pattern was apparent in
samples from Herendeen Bay, where 42% of zoeae-I
BKC were missing one spine from the telson (Fig. 8C)
and an additional 18% missing two spines (Fig. 8D).
Because the missing spines are the innermost of the
two pairs, those remaining tend to be considerably
longer than the single pair in RKC; yet as Figure 8D
Figure 6
Mean eye length ( ± 1 SD), and number of
Paralithodes spp. zoeae measured.
NOTE Jensen et al Differentiating Parahthodes larvae using telson spines
781
100
80
a 60-
a 40
Ol
20-
Prlbilof Island
L
1 1
III
100
80
60
40
20
IV I
Stage
Herendeen Bay
Figure 7
Number of telson spines of
Paralithodes platypus zoeae
collected from the Pribilof
Islands and Herendeen Bay.
I ''m
0 25mm
Figure 8
Telson spines of Paralithodes spp. zoeae. (A) P. platypus, "typical" 8 + 8 pattern; (B) P. camtschaticus
7 + 7; (C) P. platypus 8 + 7; (D) P. platypus 7 + 7 (identification based on eye and ratio of rostrum
length/carapace length).
782
Fishery Bulletin 90(4), 1992
0.6
■ Red King Crab (Sato 1956)
n Blue King Crab (Sato 1958)
• Rod King Crab (Marukawa 1 933)
r Blue King Ctab (Marukawa 1933)
▲ Red King Crab (Kutala 1960, 1964)
iL Blue King Ctab (Kurala 1964)
sK Blue King Crab (Hoffman 1 968)
■ ■
■
•
■
0,5
-
▲
Rostrum Length/Carapace Length
o
•
•
*
o
D
*
•
*
D
0,3
A
1 II III
IV
Stage
Figure 9
Rostrum/carapace length ratios of Paralithodes
taken from published illustrations.
spp. zoeae
shows, this difference is sometimes negUgible. How-
ever, BKC were readily distinguished from RKC by
their proportionately shorter rostrums, larger eyes,
and larger bodies, but the differences were not as great
as those seen in the Pribilof Islands. Zoeae in both loca-
tions appeared to gain additional telson spines with lar-
val stage, a pattern noted previously for other lithodid
larvae by Kurata (1964). While rostrum length alone
was not reliable for separating the two species (due to
intraspecific variation between the two areas), the pro-
portion of rostrum length to carapace length remained
constant and appears to be a useful method of differen-
tiating the two. Whether this is reliable in other areas
is not known, but it is consistent with illustrations in
published descriptions of the two species (Fig. 9).
Of particular interest is the intraspecific variability
exhibited both between the two areas and within the
population in Herendeen Bay. While there were no
significant differences between populations of RKC, all
linear measurements of BKC zoeae from Herendeen
Bay averaged 10-12% smaller than those of conspe-
cifics from the Pribilof Islands. The cause of this varia-
tion is not known, but environmental factors such as
temperature can affect both the number of decapod
larval stages (Knowlton 1974) and their morphology
(Shirley et al. 1987). Temperatures differed con-
siderably between the two areas at the time of larval
collection. At the Pribilof Islands in May 1983 the water
was 2-4°C, and -1-1.5°C in April 1984 (Armstrong
et al. 1985). In Herendeen Bay larvae stayed above a
40m thermocHne in water 2.5-8.5°C, and develop-
mental times were exceptionally fast (Wainwright et
al. 1991).
BKC have an extremely disjunct distribution (Somer-
ton 1985), and it is also possible that size differences
could be related to their reproductive isolation. But
although local environmental features or isolation may
explain differences between the two populations, they
are unlikely to account for the variation seen within
the relatively small scale of Herendeen Bay. The eyes
of BKC zoeae having a 7 -i- 7 spine pattern were inter-
mediate in length between the 8 -i- 8 or 8 -i- 7 BKC larvae
and RKC from Herendeen Bay (Fig. 7). The cause of
such differences cannot be known without appropriate
experiments and genetic studies, but the recent report
of an adult RKC-BKC hybrid from the Sea of Okhotsk
(Nizyayev 1991) raises the possibility that some inter-
breeding may occur within the confines of Herendeen
Bay.
Because all larvae were collected from the plankton
rather than hatched in captivity, we cannot state un-
equivocally that these differences in spine count are
due to intraspecific variation of BKC rather than a mix-
ture of other lithodid species. However, megalopae,
juveniles, and adults of RKC and BKC were collected
within Herendeen Bay during the course of this study;
despite extensive trawling, pot fishing, dredging, and
intertidal surveys, the only other lithodid found was
Hapalogaster grehnitzkii. Larvae matching the descrip-
tion for H. grebnitzkii were also the only other lithodid
zoeae occurring in the plankton samples. The large size,
shape, and position of the posterolateral carapace
spines, coupled with a lack of carapace sculpturing,
readily distinguish RKC and BKC zoeae from other
described species of Bering Sea lithodids. We believe
it is extremely unlikely that the variation is due to a
fourth, undescribed species.
No single character for reliably separating BKC and
RKC zoeae was observed, but since the number of
telson spines is useful for differentiating the two
species in some areas (e.g., Pribilof Islands), we sug-
gest using this count along with the rostrum/carapace
NOTE Jensen et a\ : Differentiating Paralithodes larvae using telson spines
783
length ratio until the extent of character overlap is
known. A ratio of >0.45 (RKC) or <0.45 (BKC) usual-
ly distinguished the species in our samples from both
areas. Eye measurements, like spine counts, vary with
area but can be useful when the rostrum or carapace
is damaged or distorted. In our samples the overlap in
telson spine counts was greatest in zoeael, but for-
tunately this was the stage when the two species could
be most reliably distinguished by the proportion of
rostrum to carapace length.
Acknowledgments
We thank Mike McGurk and Dave Warburton of
Triton, Ltd., for their help in the field, and Trent
McDonald, Pam Wardrup, and Tom Wainwright for
suggestions and advice. This study was funded by
Minerals Management Service of the Department of
the Interior, through an interagency agreement with
the National Oceanic and Atmospheric Administration,
Department of Commerce, as part of the Alaska Outer
Continental Shelf Environmental Assessment Program
(OCSEAP).
Citations
Armstrong, D.A., J.L. Armstrong, R. Palacios, G. Williams,
G.C. Jensen, and W. Pearson
1985 Early life history of juvenile blue king crab, Paralithodes
platypus, around the Pribilof Islands. In Melteff, B.R. (ed.),
Proceedings of the international king crab symposium,
p. 211-229. Rep. 85-12, Univ. Alaska Sea Grant Prog.,
Fairbanks.
Haynes, E.B.
1984 Early zoeal stages of Placetron wosnessenskii and Rhino-
lithodes wos-nfsseyiskii (Decapoda. Anomura, Lithodidae) and
revievi' of lithodid larvae of the northern North Pacific
Ocean. Fish. Bull., U.S. 82:315-324.
Hoffman, E.G.
1968 Description of laboratory-reared larvae of Paralithodes
platypus (Decapoda, Anomura, Lithodidae). J. Fish. Res.
Board Can. 25:439-455.
Knowlton, R.E.
1974 Larval developmental processes and controlling factors
in decapod Crustacea, with emphasis on Caridea. Thalassia
Jugosl. 10:138-158.
Kurata, H.
1960 Last stage zoea of Paralithodes with intermediate form
between normal last stage zoea and glaucothoe. Bull. Hok-
kaido Reg. Fish. Res. Lab. 22:49-56 [in Jpn., Engl, synop.].
1964 Larvae of decapod Crustacea of Hokkaido. 6. Lithodidae
(Anomura). Bull. Hokkaido Reg. Fish. Res. Lab. 29:49-65 [in
Jpn.. Engl. summ.].
Marukawa. H.
1933 Biological and fishery research on Japanese king-crab
Paralithodes camtschatica (Tilesius). J. Imp. Fish. E.xp. Stn.
4, 152 p. [in Jpn., Engl, abstr.].
Nizyayev, S.A.
1991 Finding of a hybrid crab specimen with the characters
of Paralithodes camtschatica and P. platypits in the Sea of
Okhotsk. Zool. Zh. 9:128-131 [in Russ.. Engl. summ.].
Sato, S.
1958 Studies on larval development and fishery biology of king
crab. Paralithodes camtschatica (Tilesius). Bull. Hokkaido
Reg. Fish. Res. Lab. 17:1-102 + plates [in Jpn., Engl. summ.].
Shirley, S.M., T.C. Shirley, and S.D. Rice
1987 Latitudinal variation in the Dungeness crab, Cancer
magister: Zoeal morphology explained by incubation
temperature. Mar. Biol. (Berl.) 95:371-376.
Somerton, D.A.
1985 The disjunct distribution of blue king crab, Paralithodes
platypus, in Alaska: Some hypotheses. In Melteff, B.R. (ed.).
Proceedings of the international king crab symposium, p.
13-21. Rep. 85-12, Univ. Alaska Sea Grant Prog., Fairbanks.
Wainwright, T.C, D.A. Armstrong, H.B. Andersen,
P. A. DinneL D.W. Herren. G.C. Jensen. J.M. Orensanz,
and J. A. Shaffer
1991 Port Moller king crab studies: Annual report. Fish. Res.
Inst. Rep. FRI-UW-9203, Univ. Wash., Seattle, 38 p.
A telemetric study of the home
ranges and homing routes of lingcod
Ophiodon elongatus on shallow rocky
reefs off Vancouver Island,
British Columbia
Kathleen R. Matthews
Pacific Southwest Research Station. U.S Forest Service
Box 245, Berkeley. California 94701
Lingcod Ophiodon elongatus are an
important commercial and recrea-
tional fish in the northeast Pacific
(Miller and Geibel 1973, Cass et al.
1990). In some areas lingcod show
serious population declines. In
Washington, after signs of deple-
tion, the fishery was closed in cen-
tral Puget Sound for 5 years (1978-
83) to allow rebuilding (Buckley et
al. 1984). Currently in central Puget
Sound, the now tightly-restricted
fishery allows only recreational use,
a 6-week opening, and a daily limit
of 1 fish. Similarly in British Colum-
bia, landings and average size of the
catch have declined and tighter reg-
ulations imposed (Richards and
Hand 1991). Because of lingcod de-
clines, it is crucial to understand
their life-history characteristics in
order to determine possible causes
of decline and to help recovery ef-
forts. One area of uncertainty re-
garding lingcod life history is their
movement behavior.
Despite a number of studies of
lingcod movements (Miller and Gei-
bel 1973, Mathews and La Riviere
1987, Jagielo 1990, Smith et al.
1990), many questions remain about
their movement patterns. Most
studies describe lingcod as seden-
tary (Miller and Geibel 1973,
Mathews and La Riviere 1987,
Smith et al. 1990); yet lingcod do
make migrations, and movement up
to 385 km has been documented
(Mathews and La Riviere 1987,
784
Jagielo 1990). Although not verified
through tagging, it is thought that
most lingcod movements, including
possible homing behavior, are re-
lated to spawning. Female lingcod
may seasonally leave deeper reefs
and move inshore to lay demersal
eggs that the shallower living males
guard. There is indirect evidence
for the inshore movement of fe-
males; an increase of larger gravid
females in the inshore catch occurs
during fall months just prior to
spawning (Miller and Geibel 1973).
Furthermore, some studies have
indicated the homing behavior of
lingcod, similar to many rocky reef
fishes (Hart 1943, Williams 1957,
Carlson and Haight 1972, Matthews
1990); i.e., when fish move away
for any reason (including spawning,
experimental displacement, etc.)
they will return to areas previous-
ly occupied (Gerking 1959). In an
early study in Canada, 4 of 14 dis-
placed lingcod returned 9.7km to
original capture sites (Hart 1943).
Additional evidence of lingcod
homing behavior came from an
attempt to enhance overfished
areas by transplanting lingcod
(Buckley et al. 1984). None of the
transplanted lingcod were resighted
at the release area (i.e., the en-
hancement was unsuccessful),
whereas nine of the transplanted
lingcod were caught close to the
original capture site (190 km from
release site).
Lingcod movement behavior has
implications for enhancement ef-
forts and habitat management. At-
tempts to rebuild populations in
overfished areas by transplanting
lingcod from areas of higher abun-
dance would be unsuccessful if ling-
cod simply returned to their original
home sites. Furthermore, move-
ment information is valuable be-
cause if lingcod preferentially home
to certain reefs, then those reefs
could be designated as management
reserves. Thus, any new knowledge
of lingcod homing will lead to a
better understanding of their move-
ment behavior and the effect of
rehabilitation efforts. The objec-
tives of this pilot study were to use
ultrasonic tagging to (1) describe
home ranges and movements of
lingcod on rocky reefs, and (2)
determine the homing routes of
displaced lingcod.
Methods
Study sites
The study was conducted during
April 1990 off eastern Gabriola
Island on the eastern side of Van-
couver Island, British Columbia
(Fig. 1). The area is characterized
by extensive shallow rocky reefs
and pinnacles. Depths encountered
during ultrasonic tracking were in
the range 3-35 m. Two reef areas
were chosen for the tracking work:
Gabriola reefs and Valdes reefs
(each is actually a series of small,
separate reefs) (Fig. 1). Both reefs
are approximately 15-30 m deep,
although shallower areas were
sometimes encountered. Although
bullkelp Nereocystis leutkeana is
present on these reefs during the
summer and fall, no surface kelp
was present during this April study.
Ultrasonic tagging
The transmitters (48 x 15 mm, 18 g
in air, 8.3g in seawater) were ex-
Manuscript accepted 29 July 1992.
Fishery Bulletin, U.S. 90:784-790 (1992).
NOTE Matthews: Movement behavior of Ophiodon elongatus off Vancouver Island
785
ternally attached using methods used for tuna (Holland
et al. 1985). The nylon loop on the transmitter anchored
one inelastic pull tie, with another tie wrapped around
the tag's opposite end. The two pull ties were inserted
through the dorsal musculature and cinched down to
prevent the transmitter from dangling.
Figure 1
Map of study area and general displacement directions for
Ophiodon elongatus fish nos. 7-11. Fish were captured and
displaced in the direction shown by arrow.
Lingcod were captured on hook-and-line, placed in
a seawater-filled cooler, and anesthetized (methomidate
hydrochloride). After tags were attached, lingcod were
allowed to recover in fresh (without anesthetic) sea-
water prior to release. Fish appeared completely recov-
ered from the anesthetic and tagging procedure ~5-10
min after tagging.
Eleven transmitters with replaceable batteries (Vem-
co V3-1H-R pingers, Vemco Ltd., Nova Scotia, Canada
B3L 4J4) were operated at five crystal-controlled fre-
quencies: 50.0 (2 tags), 60.0 (2 tags), 65.54 (3 tags),
69.0 (2 tags), and 76.8 (2 tags) kHz, corresponding to
pre-set channels on the Vemco VR-60 receiver. Tags
assigned to the same channel were easily differentiated
by their unique pulse period, which was automatically
decoded and displayed by the receiver. To locate the
transmitters, a Vemco V-10 directional hydrophone
was employed from a small boat. Once a tag was
located, the boat's position was determined, using
LORAN-C readings, depth, and visual compass bear-
ings of four charted features (buoys, lights, etc.), in the
study area on an almost daily basis (one of 21 tracking
days was missed due to boat breakdown) for the life
of the transmitter (21 d battery life). One reason this
study area was chosen was the presence of several
flashing lights and buoys which made navigation and
location determination easier, especially at night.
During 5-27 April 1990, 11 Hngcod (57.0-80.6cm
total length, TL) were tagged and monitored (Table 1).
Table 1
Summary of total lengths and duration of tracking
of lingcod Ophiodon elongatus equipped with ultrasonic transmitters in the Strait
of Georgia, British Columbia,
and monitored 5-27
April 1990 to
measure home ranges and homing routes. Nocturnal movement
occurred 24:00-06:00.
Tag
Length
Date tracking
Duration
no. Sex
(TLcm)
began
(d)
Nondisplaced lingcod home ranges
Controls
1 Female
67.5
4/5/90
20
Captured and released at Valdes reefs
2 Male
62.1
4/5/90
16
Captured and released at Valdes reefs
3 Male
59.5
4/9/90
19
Captured and released at Valdes reefs
4 Male
69.8
4/8/90
15
Captured and released at Gabriola reefs
5 Female
80.6
4/9/90
12
Captured and released at Gabriola reefs
6 Female
68.4
4/9/90
15
Captured and released at Gabriola reefs
Displaced lingcod homing routes
Distance Total time Movement Nocturnal
moved to return rate movement rate
(km) (h) (m/h) (m/h)
Experimentals
7 Male
66.2
4/12/90
14
1.0 40 25 83.3
8 Male
57.0
4/16/90
12
2.2 Did not return
9 Male
64.0
4/19/90
9
2.8 60 48 233.3
10 Male
64.5
4/23/90
5
2.8 35 80 233.3
11 Male
69.5
4/25/90
3
2.8 33 85 233.3
786
Fishery Bulletin 90(4), 1992
Up to seven tagged fish were deployed within the study
area at one time. One procedure was conducted on two
rocky reefs to determine the home ranges of Hngcod
during the day, night, and periods of strong current
which were predicted up to 14.81<;m/h (8.01<n) (Cana-
dian tide and current tables, Canadian Hydrographic
Service 1990). Six lingcod were captured, tagged,
released, and monitored at the original point of cap-
ture, either Gabriola reefs area or Valdes reefs. Their
geographic locations were determined on an almost
daily basis (the only exception being 18 April) for the
duration of the tag battery life.
To determine the homing routes of displaced lingcod,
five males were captured, tagged, and displaced up to
2.8km in two opposite directions (Fig. 1). Four of the
five displaced fish were moved south, while one fish
was moved north. Because there was no information
on the time required for lingcod to home, the first fish
was moved a short distance (1 km) to allow enough time
to track its homeward movement. Of the remaining
four lingcod, one was displaced 2.2km, and three were
displaced 2.8km. The movements were then monitored
almost continuously, with occasional rest breaks, until
the transmitter batteries expired. Displaced fish were
individually tracked because they required continuous
monitoring. When the displaced fish were first re-
leased, an attempt was made to stay with the fish for
several hours to detect any homeward movement.
The field schedule started with two teams each work-
ing 12 h on the boat for a full 24 h coverage. After the
first few days of field work, we changed the schedule
to devote our efforts to covering nighttime lingcod
movement (16:00-08:00). The entire study covered 21
tracking days during 5-27 April (only 18 April was
canceled due to boat problems) for a total of 336 track-
ing hours. When the nondisplaced fish were first re-
leased, an entire day was spent tracking those fish.
Subsequently, each fish was periodically (about 15-20
times/tracking day) checked for its position, which
allowed several tagged fish to be concurrently moni-
tored. The five displaced fish were followed one at
a time to ensure that their homing route could be
detected.
On two separate dates, I conducted scuba observa-
tions to search for tagged lingcod and to collect infor-
mation not available through telemetry. When a signal
cannot be located, it was impossible to determine
whether the tag battery has died or the fish has left
the area; hence, I searched underwater for two tagged
fish in their last recorded location after the signal could
no longer be detected. Also, to determine the accuracy
of telemetric locations, I used scuba observations to
compare underwater positions of tagged fish with those
provided by telemetry. Once the directional hydro-
phone positioned the boat directly over the tag's signal
I then anchored the boat, descended, and searched for
the tagged fish.
Results
Nondisplaced fish (controls)
The six lingcod (three females and three males) tagged
and released at Gabriola and Valdes reefs were general-
ly found close to release sites during the day, night,
and periods of swift current. When relocated, the six
lingcod were essentially in the same position (latitude-
longitude differences were within 0.01-0.02 nmi, which
is the normal resolution of the LORAN unit). Thus,
there was no detectable difference in their home range
size. Fish were monitored for 12-20d (x 16.2-1- 2. 9d)
until the tag batteries expired.
These telemetry findings were verified by visual
(scuba) resightings of two tagged lingcod. After deter-
mining the position of an ultrasonic tag, I later (within
a few minutes) observed the tagged fish sitting on the
bottom. These visual sightings also verified that the
tags were still attached to the fish. Furthermore, after
the signal had apparently died on two tagged fish, I
searched underwater and saw the two tagged lingcod
in their last recorded telemetry position.
The six individual lingcod were also monitored on six
separate nights when the current ranged from slack
to 10.4 km/h (5.6kn) for a total of 50h of nighttime
observation. Home ranges were similar to those ob-
served in the day. However, signals were louder sug-
gesting that the fish were out in the open, i.e., not
under a rock (Matthews et al. 1990).
Displaced fish (experimentals)
The five displaced lingcod remained close to release
sites and did not move for several hours following
release (Figs. 2-6). These first few hours (both during
the day and night) following release may be a recovery
period in response to capture, handling, and tagging.
Subsequently, four of the five displaced fish moved
back to the capture site. Each fish had returned to the
capture site by the end of the second night following
release. Return trips were confined to the immediate
vicinity of the Gabriola and Valdes Islands study area.
The four homing lingcod (nos. 7, 9-11) remained near
the release site for 4-6 h and returned to home sites
in 33-60h (Figs. 2-5). These four fish started their
homeward movements at night (20:30-06:00), and
movement terminated once it became light at ~06:00.
No clear pattern was detected in the homeward
movement, as lingcod did not appear to follow obvious
features such as depth contours or currents. Homing
lingcod traversed depths of 5-35 m. Occasionally the
NOTE Matthews: Movement behavior of Ophiodon elongatus off Vahcouver Island
787
1k
m
displacement
O
\
■ C home 07 00 until 4/25
o
o
. S-08 15-16 00
■ R S-16 40-20 00
D
J
1
N
1
V
Maximum 40 hours to return
]ldes Island ^^^^^
Figure 2
Positions of Ophiodon elongatus fish no. 7 (66.2 em male) cap-
tured from site C, moved 1km south to release site R, and
monitored 12-25 April. S = stationary.
2.8km displacement
04 15
C H-17 15 until 4/27
V.
S-06 00-0715 & S-18 30-01 00
r^^ Maximum 60 hours to return
'■ R S-22 30-04 00
Valdes Island
Figure 3
Positions of Ophiodon ekmgatus fish no. 9 (64.0cm male) cap-
tured from site C, moved 2.8km south to release site R, and
monitored 19-27 April. S = stationary, L = temporarily lost.
signals became quite strong, suggesting homing lingcod
were traversing areas without much rock relief. Ling-
cod encountered the deepest water (35 m) when cross-
ing open areas. The four lingcod returned to their
original capture sites where they remained until the
transmitter batteries died or the tracking project was
completed.
The first displaced fish (no. 7, a 66.2cm male) was
caught on 12 April in 10 m of water, moved 1km south,
and released in 18 m of water (Fig. 2). Because this was
the first release and I did not know what movement
to expect, I stayed directly over the fish from release
time (16:40) until 20:30. However, no movement was
detected. When I returned the following morning
(07:30), the fish had moved about halfway (500m, Fig.
2) back to the capture site. I stayed with the fish from
07:30 to 16:30, but no additional movement was
detected. When tracking resumed the following mor-
ning (07:00), the fish was back at the original capture
site where it remained for 12 d (until 25 April) when
the battery apparently died. This fish moved at least
1km, the displacement distance, in ~40h. Because all
homeward movement of the fish apparently occurred
at night, the remaining displaced fish were tracked at
night.
Fish no. 9 (64.0cm male) was captured on 19 April
in 12m of water, tagged and moved 2.8km south, and
released into water 18 m deep (Fig. 3). It was stationary
from release at 22:30 until 04:00 when it moved in a
northerly direction for 2h and stopped. Tracking was
terminated at 07:15; when it reconvened at 18:30, the
fish was in the same location. At 01:00 it moved to the
northwest and the northeast until 04:30 when its signal
was lost. Tracking was terminated at 08:30, and when
2 8km displacement
04 47 05 30
L ^-^ C
05 45 until 4/27
S-06 00-07 00 & S-16 30-03 10
v--^^
Maximum 35 hours to return
^ R S-22 30-01 00
Valdes Island ^^^
Figure 4
Positions of Ophiodon elongatus fish no. 10 (64.5 cm male) cap-
tured from site C, moved 2.8 km south to release site R, and
monitored 23-27 April. S = stationary, L = temporarily lost
(04:47-05:30).
it resumed at 17:15 the fish was back at the capture
site. I assumed that the fish homed between 04:30
(when the signal was lost) and daybreak because all
other lingcod movement occurred at night. It remained
at the capture site for 7 days until tracking ended on
27 April. The 2.8km return trip was completed in less
than 60 h.
Fish no. 10 (64.5cm male) was captured in 10 m of
water, tagged and displaced 2.8km south, and released
in water 18 m deep (Fig. 4). The fish was stationary for
2.5 h after release (22:30-01:00) and moved sporadically
from 01:00 to 06:00. The fish was stationary from 06:00
to 07:00 when tracking ended. When tracking
788
Fishery Bulletin 90(4). 1992
2 8 km displacement
R S-20 30-24 00
S-05 00-0630
& 19 00-03 00
Maximum 33 hours to return
C 05 45 until 4/27
Voldes Island
Figure 5
Positions of Ophiodon elongatus fish no. 1 1 (69.5 cm male) cap-
tiored from site C, moved 2.8km north to release site R. and
monitored 25-27 April. S = stationary.
2 2 km displacement
— \
■ C
-9
■ R
^1
N
1
Never returned
"^ Voldes Island ^^^-s.^
Figure 6
Positions of Ophiodon elongatus fish no. 8 (57.0 cm male) cap-
tured from site C, moved 2.2 km south to release site R, and
monitored 16-27 April.
reconvened at 16:30, the fish was in the same position
where it remained until 03:15. It then moved from
03:15 to 04:47 when the signal was temporarily lost.
When relocated at 05:30, it was followed back to the
capture site (05:30-05:45). It was at this site 4 days
later that tracking stopped. The 2.8km return trip took
about 35 h.
Fish no. 11, a 69.5cm male, was captured in 15m of
water, tagged and displaced 2.8km to the north, and
released in 12 m of water (Fig. 5). The fish remained
stationary from release (20:30) until 24:00, when it
moved to the west and south until 05:00. It remained
at this position until tracking ended at 06:30. Later that
day, the fish was relocated in the same location where
it remained until 03:00. At this time the fish moved
southeast and southwest and reached its original cap-
ture site at 05:45. It remained there until 27 April when
tracking ended. The 2.8km return trip took ~33h.
One fish did not return from displacement. Fish no.
8, a 57.0cm male was caught, tagged, and transplanted
2.2km (Fig. 6) to the release site. Tracking continued
for 12 d, during which time the fish apparently re-
mained at the release site and no movement was
detected. Because I was unable to make scuba obser-
vations at this site, it is possible that the tag was shed,
which would also result in a stationary signal.
Lingcod took 33-60h to return from their 1.0-2.8
km displacements for an average homing speed of
59.5 m/h (Table 1). Actually, their movement rate was
faster since they moved only at night. If averaged over
the total period when movement was documented
(6h, 24:00-06:00) for two consecutive nights (total of
12h) then the rates are 83.3-233m/h (x 195.8m/h or
1 175.0 m/d).
Discussion
Similar to intertidal fishes (Williams 1957) and several
species of rockfishes (Carlson and Haight 1972,
Matthews 1990), lingcod are another rocky reef fish
capable of homing. Ultrasonic tracking is limited by low
sample sizes due to tag cost and labor-intensive track-
ing. This study represents the first attempt to use ultra-
sonic telemetry to research lingcod movement
behavior, but this was limited to displaced males soon
after their nesting season. Additional work is necessary
to determine whether males behave differently (e.g.,
do not home) at other times of the year. Telemetry
would also be valuable to determine whether females
make inshore-offshore movements to relocate previous-
ly used areas.
Lingcod movement occurred at night (24:00-06:00,
Figs. 3-5) sometimes under dark, moonless skies. Little
work has been done on fish vision in cold-temperate
water systems (see review in Loew and McFarland
1990). Nevertheless, water at night is darker and has
lower visibility than during the day, and as Ebeling and
Bray (1976) point out, "...the relatively turbid,
temperate waters are often a dark and gloomy place
at night." Moreover, during our April tracking study,
most nights were overcast and rainy, further reduc-
ing the water's visibility. The low visibility at night
presumably precludes lingcod from using visual land-
marks which usually requires precise recognition of
specific features such as coral heads or rocks (Hasler
1966, Reese 1989). Still, an important question re-
mains: Why should lingcod move at night when visibil-
ity is better during the day? Perhaps their nocturnal
movement is to avoid predation since lingcod some-
NOTE Matthews: Movement behavior of Ophiodon elongatus off Vancouver Island
789
times crossed flat, open areas that had no hiding places.
Locally, harbor seals feed on lingcod, and recently their
numbers have dramatically increased (Olesiuk et al.
1990). On the other hand, perhaps lingcod are simply
more active at night, as my nighttime home-range
observations indicated, which would explain their
nighttime movement.
Lingcod homing was fairly directional and confined
to the immediate area of Gabriola Island. In contrast,
when displaced shorter distances (500 m), copper and
quillback rockfishes, which co-occur with lingcod,
moved along a bimodal northwest-southeast axis and
sometimes retraced that path before finally moving
in a westerly direction that led to their home site
(Matthews 1990). After displacement, initial movement
of lingcod was in the homeward direction only, i.e., no
back-and-forth movement between the release site and
home site was observed. The more direct and noctur-
nal homing in lingcod suggests they are navigating
rather than orienting along one compass course or rely-
ing on olfactory cues. Orientation can be ruled out
(Baker 1978, Able 1980) because lingcod successfully
homed from north and south displacements. It is cur-
rently unknown whether rocky-reef fish, including
lingcod, recognize olfactory cues.
Lingcod homing was fast in comparison with copper
and quillback rockfishes, which took 8-25 d to return
from 500 m displacements (Matthews 1990). From an
analysis of a large-scale tagging program during
1982-87, Smith et al. (1990) estimated that mean
dispersal rates for male and female lingcod were
500 m/d and 1040 m/d, respectively, similar to those
observed in the present study (1 173.7 m/24h). Pre-
sumably, lingcod movement rates vary depending upon
seasonal requirements (e.g., feeding, spawning, etc.).
Several hypotheses could explain why the smallest
male (no. 8) did not return from displacement. Perhaps
lingcod do not develop a resident or homing response
until they are older and larger. The length-maturity
relationship is determined by their geographic area
(Richards et al. 1990), and the 50% maturity level for
male lingcod at a similar latitude off the west coast of
Vancouver Island is 57.1cm. Thus, if fish no. 8 (57.0 cm)
was not sexually mature, it may have lacked the abil-
ity to home. Buckley et al. (1984) also noted a lack of
homing in small male lingcod. In that study, after
4. Syr, the smallest transplanted lingcod (a 57cm male)
remained close to the release site after most trans-
planted lingcod had apparently homed. Alternatively,
the lack of homeward movement could be due to tag
shedding, which would also produce a stationary signal.
This study revealed new information on lingcod hom-
ing behavior. After displacement up to 2.8km, lingcod
moved at night back to home sites within 60 h and
followed a fairly direct route. Because this was a pilot
study and I displaced only male lingcod, more track-
ing studies are needed to increase sample sizes, include
females, and attempt longer-distance transplants.
Whether they home and reuse spawning areas will be
important to document, as this information is crucial
if lingcod preserves are established. It does appear that
transplant attempts to rebuild lingcod stocks may be
ineffective with larger, older males but may be suc-
cessful if the lingcod are moved before they reach a
certain size or age.
Acknowledgments
Claudia Hand, John Candy, and Bronwyn Lewis ably
caught the lingcod and assisted in all phases of the field
work including the tortuous all-night trips. This
research was conducted while I held a Natural Sciences
and Engineering Research Council Visiting Scientist
Fellowship at the Canadian Department of Fisheries
and Oceans Pacific Biological Station under the spon-
sorship of Dr. Laura Richards. The Institute of Ocean
Sciences in Sidney kindly loaned us the use of the
research vessel Orca. Dr. Richards and two anonjmous
reviewers provided helpful comments which greatly im-
proved the manuscript.
Citations
Able, K.P.
1980 Mechanisms of orientation, navigation, and homing. In
Gauthreaux, S.A. Jr. (ed.), Animal migration, orientation, and
navigation, p. 283-373. Academic Press, NY.
Baker. R.R.
1978 Evolutionary ecoiogj' of animal migration. Holmes and
Meier Publ., Inc., NY. 1012 p.
Buckley, R., G. Hueckel, B. Benson, S. Quinnel, and M. Canfield
1984 Enhancement research on lingcod {Oph lodon elongatus)
in Puget Sound. Wash. Dep. Fish. Prog. Rep. 216. 93 p.
Canadian Hydrographic Service
1990 Canadian tide and current tables. Pacific Coast, vol. 5.
Carlson, H.R., and R.E. Haight
1972 Evidence for a home site and homing of adult yellowtail
rockfish, Sebastes flaiddiis. J. Fish. Res. Board Can. 29:
1011-1014.
Cass, A.J., R.J. Beamish, and G.A. McFarlane
1990 Lingcod (Ophiodon elongatm). Can. Spec. Publ. Fish.
Aquat. Sci. 109, 40 p.
Ebeling, A.W., and R.N. Bray
1976 Day versus night activity of reef fishes in a kelp forest
off Santa Barbara, California. Fish. Bull., U.S. 74:703-717.
Gerking, S.D.
1959 The restricted movement of fish populations. Biol. Rev.
Camb. Philos. Soc. 34:221-242.
Hart, J.L.
1943 Migration of lingcod. Fish. Res. Board Can., Prog. Rep.
Pac. Coast Stn. .57:3-7.
790
Fishery Bulletin 90(4), 1992
Hasler, A.D.
1966 Underwater guideposts. Univ. Wise. Press, Madison,
155 p.
Holland, K., R. Brill, S. Ferguson, R. Chang, and R. Yost
1985 A small vessel technique for tracking pelagic fish. Mar.
Fish. Rev. 47(4):26-32.
Jagielo, T.H.
1990 Movement of tagged lingcod Ophiodon ekmgatus at Neah
Bay, Washington. Fish. Bull.. U.S. 88:815-820.
Loew, E.R., and W.N. McFarland
1990 The underwater visual environment, /n Douglas, R., and
M. Djamgoz (eds.). The visual system offish, p. 1-43. Chap-
man and Hall, London.
Mathews, S,B., and M. LaRiviere
1987 Movement of tagged lingcod, Ophiodon elongatus, in the
Pacific Northwest. Fish. Bull., U.S. 85:153-159.
Matthews. K.R.
1990 An e.xperimental study of movement patterns and habitat
preferences of copper, quillback. and brown rockfishes on three
habitat tyises. Environ. Biol. Fishes 29:161-178.
Matthews, K.R., T.P. Quinn, and B.S. Miller
1990 Use of ultrasonic transmitters to track demersal rockfish
movements on shallow rocky reefs, hi Parker, N.C.. et al.
(eds.). Fish marking techniques, p. 375-379. Am. Fish. Soc.
Symp. 7. Bethesda.
Miller, D.J.. and J.J. Geibel
1973 Summary of blue rockfish and lingcod life history
histories: a reef ecology study; and giant kelp Macrocystis
pyrifera, experiments in Monterey Bay, California. Calif. Dep.
Fisli Game, Fish. Bull. 158:1-137.
Olesiuk, P.F., M.A. Bigg, G.M. Ellis, S.J. Crockford, and
R.J. Wigen
1990 An assessment of the feeding habits of harbour seal
(Phoca vitulina) in the Strait of Georgia, British Columbia,
based on scat analysis. Can. Tech. Rep. Fish. Aquat. Sci. 1730,
135 p.
Reese. E.S.
1989 Orientation behavior of butterflyfishes (family Chaetodon-
tidae) on coral reefs: Spatial learning of route specific land-
marks and cognitive maps. Environ. Biol. Fishes 25:79-86.
Richards, L.J., and CM. Hand
1991 Lingcod. /w Fargo, J., and B.M. Leaman (eds.), Ground-
fish stock assessments for the west coast of Canada in 1990
and recommended yield options for 1991, p. 19-42. Can. Tech.
Rep. Fish. Aquat. Sci. 1778.
Richards, L.J., J.T. Schnute, and CM. Hand
1990 A multivariate maturity m.odel with comparative analysis
of three lingcod (Ophiodon elongatus) stocks. Can. J. Fish.
Aquat. Sci. 47:948-959.
Smith, B.D., G.A. McFarlane, and A.J. Cass
1990 Movements and mortality of tagged male and female
lingcod in the Strait of Georgia, British Columbia. Trans. Am.
Fish. Soc. 119:813-824.
Williams, G.
1957 Homing behavior of California rocky shore fishes. Univ.
Calif. Publ. Zool. 59:249-284.
An investigation of bottienose dolpinin
Tursiops truncatus deatiis in East
Matagorda Bay, Texas, January 1990
W. George Miller
Naval Ocean Systems Center. Code 514, San Diego, California 92152
There are reports of massive mor-
talities of bottienose dolphins Tur-
siops truncatus over periods of
months in areas as large as the U.S.
Atlantic coast and the Persian Gulf.
From early June 1987 until March
1988, over 740 bottienose dolphins
(estimated at about 50% of the
coastal migratory stock) stranded
along the U.S. Atlantic coast from
New Jersey to Florida (Scott et al.
1988, Geraci 1989). Geraci con-
cluded the dolphins were poisoned
by brevetoxin, a neurotoxin pro-
duced by the red tide organism Pty-
chodiscus hrevis.
During 23 August to 30 October
1986, 527 dead dolphins were found
on the eastern and western shores
of the Persian Gulf. Several dead
turtles, dugongs, and one 6.1 m un-
identified whale were also found,
along with many fish that washed
ashore. Of the 78% of mammal car-
casses identified to species, 64%
were bottienose dolphins, 34% were
humpback dolphins Sousa chinen-
sis, 1.7% were common dolphins
Delphinus delphis, and 0.3% were
finless porpoises Neophocoena pho-
coenoides. The dead dolphins in-
cluded adults, neonates, calves, and
juveniles. Cause(s) of the deaths
could not be determined, since only
four animals were necropsied
(Anonymous 1986).
The subject of this study is an
unusual stranding of 26 T. trun-
catus that occurred in January 1990
around East Matagorda Bay (EMB),
Texas. There are no previous re-
ports of this number of strandings
in a relatively small area in a single
day. On 20 January 1989, a helicop-
ter pilot reported the stranded dol-
phins to the U.S. Coast Guard, who
notified the Texas Marine Mammal
Stranding Network at Texas A&M
University and Texas Parks & Wild-
life Department (TPW). These or-
ganizations collected 26 carcasses,
23 from within the Bay and 3 from
the Gulf side of East Matagorda
Peninsula. I performed necropsies
on the dolphins on 24-25 January to
determine cause of death.
Methods
Examination of dolphins
Each T. truncatus dolphin was as-
signed an identification number and
its stranding location noted (Fig. 1).
State of decomposition was noted:
freshly dead with no bloating (1
animal), detectable bloating, or
severe decomposition. Animals
were sexed and weighed, and length
was measured via a straight-line
from the notch in the tail flukes to
the most rostral aspect of the man-
dible. Measurements of blubber
thickness were taken at six loca-
tions along the animal's left side
using the standard protocol of the
Naval Ocean Systems Center (NOSC)
(Fig. 2). Skin condition and abnor-
mal marks or deteriorated areas
were recorded. Condition and posi-
tion of thoracic and abdominal
organs were noted before removal
and collection of tissue samples.
Site inspection and
background information
On 26 January 1990, 1 conducted an
aerial survey of the stranding site,
comparing the actual configuration
of East Matagorda Bay with an ex-
isting map (NOAA nautical chart
#11319) to determine exit routes for
dolphins from the Bay to the deeper
waters of the outer coast. The main
exit is a narrow cut connecting the
Bay and the Gulf (Fig. 1).
The Texas Parks and Wildlife
Department monitored water-tem-
perature changes in the Bay almost
daily during 15-29 December 1989
(Fig. 3). The Bay was completely
frozen over for 2.5 days with par-
tial ice remaining for 4 days. On 22
December, a helicopter pilot flew
close to the Bay to observe about 12
dolphins swimming and breaking
ice (~5cm thick) in a 4-7 km area in
the east-central region of the Bay.
Rapidly-moving weather systems
from the north with strong norther-
ly winds can significantly lower
tidal levels in the Bay (Steve Mar-
witz, Texas Parks & Wildlife, Rock-
port, TX 78382, pers. commun.).
Within the period 15-22 December
1989, when two cold-weather sys-
tems moved through the area, an
estimated range for the mean low
tide level was 30-60 cm below nor-
mal (Mark Mazot, Tex. Parks Wildl.
Dep., pers. commun., Feb. 1990);
however, there were no official
measurements. Thus it is possible
that lowered water depths around
the periphery of the Bay could have
impeded dolphin movement be-
tween the Bay and the Gulf of Mex-
ico via the Caney Creek Gulf Cut or
the intercoastal canal.
Results
Of the 26 Tursiops truncatus ex-
amined, 23 dolphins were from
within East Matagorda Bay and
consisted of 6 mature males (MM),
5 immature males (IM), 7 mature
Manuscript accepted 9 September 1992.
Fishery Bulletin, U.S. 90:791-797 (1992).
791
792
Fishery Bulletin 90(4), 1992
• = DOLPHIN CARCASS
«« = SWING BRIDGE
ICC = INTERCOASTAL CANAL
M = ACCESS CHANNEL
• PALACIOS
15 mi
Figure 1
Map of East Matagorda Bay.
Texas, showing location of
recovered bottlenose dolphin
carcasses.
Total Length
Site 1) Dorsal midline 10cm to 15cm caudal to the blow hole.
Site 2) Dorsal midline 10cm to 15 cm cranial to the insertion of
the dorsal fin.
Site 31 Lateral midline at the midpoint of the dorsal fin.
Site 4) Ventral midline 10cm cranial to umbilicus.
Site 5) Dorsolateral aspect of tail stock at level of the anus.
Site 6) Lateral midline at level of the anus.
Figure 2
Blubber measurement sites on bottlenose dolphins collected
in East Matagorda Bay, based on Naval Ocean System Center
collection procedures.
females (MF) (4 with fetus), and 5 immature females
(IF); 2 females and 1 male were from outside the Bay
on East Matagorda Peninsula. I could not make an ac-
curate determination of time of death because decom-
position varies markedly depending on environmental
conditions. The condition of the carcasses at necropsy
suggested that death occurred ~5-10 days prior to
siting, with the exception of one freshly dead animal
collected outside the Bay on East Matagorda Peninsula.
Table 1 shows blubber thickness measured at NOSC
(site 2), length, sex, and weight data from 24 East
Matagorda Bay dolphins for which there were complete
data, and from a comparison group of 16 Texas coast
dolphins which stranded over the period 1981-89.
Average blubber thickness for 21 stranded dolphins
recovered in the Bay was 12.7mm, while the average
thickness for the comparison group of 16 dolphins
recovered on the Texas coast during the winter months
of November-March 1981-89 was 18.6mm at the same
measurement site. This difference was significant (Stu-
dent's t test, P< 0.001). In addition, the subcutaneous
fat layer that is prominent between the blubber and
skeletal muscles in healthy robust dolphins (Ridgway
and Fenner 1982) was greatly reduced or absent in all
of the dolphins taken from inside the Bay.
A linear regression was used to (1) show the relation-
ship of blubber thickness (mm) to weight (kg) for the
comparison group of dolphins stranded along the Texas
coast during winter months and (2) test the similarity
of slopes between the Texas-coast and East Matagorda
Bay groups (Fig. 4). In Texas coast strandings, there
was a significant positive correlation between blubber
thickness and body weight (r 0.9); while in the Bay
group, blubber thickness decreased with increased
weight (r -0.38).
NOTE Miller: Deaths of Tursiops truncatus in East Matagorda Bay
793
15
• WATER TEMPERATURE
D AIR TEMPERATURE
20 22 24 26
DECEMBER 1989
30
22 DEC 89:
Dolphins seen breaking ice
inEMB
25 DEC 89:
Ice cover begins to thaw
on EMB
26 DEC 89:
Collection of fish kill data
in EMB begins
3 JAN 90:
At least three dolphins seen
following trawling boat in EMB
20 JAN 90:
26 dead dolphins observed
24/25 JAN 90:
Necropsies performed
Table 1
Summary data from 24 bottlenose dolphins Tursiops truncatus taken
in East Matagorda Bay and Peninsula, January 1990, and from Texas
coast bottlenose dolphins recovered in November-March 1981
-89 (data
from Texas Marine Mammal Stranding Network).
East Matagorda
Texas coast
Blubber
Blubber
thickness Length Weight
thickness Length
Weight
(mm) (cm) Sex (kg)
(mm) (cm) Sex
0<g)
Peninsula
12 100 F
14
9 198 M 68
12 115 M
17
15 240 F 123
15 205 F
91
28 235 F NA
16 212 M
104
Bay
17 219 M
109
4 216 M 114
18 177 F
59
5 269 M 182
18 206 F
107
5 282 F 170
18 221 F
102
7 262 F 139
18 235 F
139
8 288 M 189
18 240 F
141
9 262 M 159
18 235 F
143
10 275 M 180
20 227 M
136
11 176 F 43
20 179 M
120
11 216 F 91
24 280 M
150
11 270 M 170
25 237 M
193
13 191 M 55
28 260 M
216
13 258 M 141
14 207 F 82
15 245 F 114
15 261 F 136
15 255 F 150
15 256 F 166
16 256 M 150
22 220 F 82
22 254 F 145
26 218 M 77
Figure 3
Water and air temperatures at East Mata-
gorda Bay (EMB) during December 1989,
and chronology of significant events related
to January 1990 strandings of bottlenose
dolphins. Air temperatures recorded at Pala-
cios, Texas (the closest recording weather
station, 15 mi away), (from Naval Oceanogr.
Command Detachment. Asheville, NC).
Water-temperature data from Texas Parks
& Wildl. Dep., Fish. Div., Coastal Branch,
Palacios, TX. Water temperatures in East
Matagorda Bay for the dates 1, 15. and 29
January 1992 were 13.5, 16, and 14.5°C,
respectively.
Two mature females (MF) and
a mature male (MM) had thick-
ened hepatic capsules. One MF
had marked lobulation and in-
creased fibrous tissue throughout
the liver. Five mature animals (2 MM, 3 MF)
had an unusual and unidentified thin, smooth
creamy-white layer on the endothelial surface
of the hepatic portal vessels.
Three animals had abnormalities associated
with the gastrointestinal (GI) system. An im-
mature female (IF) had a section of small
□ EMB DOLPHINS • TEXAS COAST DOLPHINS
30
III INI I ll I
I I I I I I I I I M I I 1 I I I I t I I I I ! I II I I ( I I I
50
100 150
WEIGHT (kg)
200 250
Figure 4
Linear regression of blubber thickness and weight for
bottlenose dolphins recovered from East Matagorda
Bay, in January 1990, and for comparison group of
bottlenose dolphins from the Texas coast during
November-March 1981-89.
794
Fishery Bulletin 90(4|, 1992
intestine ~1.5m long containing extremely-hard dehy-
drated feces, and its stomach contained partially-
digested fish and bones. Two IF had peritoneal adhe-
sions throughout the GI tract. One of these had a
serofibrinous exudate on the serosal surfaces of the en-
tire small intestine, and the gastric compartments were
empty. Four animals (2 MF, 2 IM) had nematodes in
the forestomach and fundic chamber. Clear, crystallized
deposits adhered to the parietal and visceral surfaces
of the thoracic and abdominal cavities of all animals.
These deposits were < 1mm in size, felt "gritty," and
imparted a "sandpaper-like" texture to the surface, a
condition not uncommon in decomposed dolphins in the
region (Raymond J. Tarpley, Texas A&M Univ., Col-
lege Station, pers. commun., March 1990).
Every mature animal in the Bay group had hard,
white, spherical deposits in the pancreatic interstitial
tissue. These deposits were <2mm in size and were
scattered throughout the central pancreas. When
crushed with a knife, the deposits were the same white
color and consistency throughout.
Stomach contents were noted in 19 of the 23 Bay
animals. Stomachs of 6 animals (3 MM, 2 IF, 1 MF)
were void of food. Ten animals had unidentified fish,
bones, and scales in the stomach. Three animals (2 MF,
1 IM) had undigested and partially-digested fish in the
forestomach; in two of these animals, there was a 30cm
undigested fish in the esophagus.
No other gross abnormalities were noted in the
respiratory, cardiovascular, renal, musculoskeletal, or
reproductive systems of the Bay dolphins. Eyes were
too decomposed for examination. Data concerning in-
fectious agents (viral, bacteriological, fungal, etc.) could
not be obtained because of advanced decomposition of
the carcasses.
Discussion
Several factors might have contributed to the East
Matagorda Bay dolphin mortality. First, an abnormally
rapid drop in water temperature which resulted in the
Bay freezing over; second, abnormally low tidal levels,
possibly preventing exit from the Bay; and third,
striped mullet, an important food source for the
dolphins, may have been significantly depleted by the
freeze. The poor condition of Bay dolphins was in-
dicated by the ~89% of males and 80% of females in
states of emaciation or near-emaciation, based on
minimum weight-length guidelines established by
Ridgway and Fenner (1982) (Fig. 5). In addition,
average blubber thickness of the Bay dolphins was a
third less than that of the Texas-coast dolphins during
winter, based on records over the previous 9-year
period.
240
220
200
1
- 160
-n ^ , 1 < I < \ < I T
Minimum Weight (Males)
200 220 240 2li0 280 300
LENGTH (cm)
220
200
180
S160
I 140
O
5 120
100
_ ' 1 ' 1 ' 1
1 '
1
1
Minimum Weight (Females)
—
—
-
• ^^
•
-
L y
X
••
—
-
"
— ^^
—
> —
-
y^
•
^^ •
-
— ^^ • •
—
^^ 1,1,1
1 1
1
1
220 240 260
LENGTH (cm)
Figure 5
Weight-length data for male and female bottlenose dolphins
from East Matagorda Bay, plotted on Ridgway and Fenner's
(1982) minimum weight-length graph.
Gunter (1941) and Gunter and Hildebrand (1951)
reported on the death of fishes and other organisms
during severe cold periods along the Texas coast. In
1940, water temperature fell from 18.3°C to -3.9°C
in 4 hours (Gunter and Hildebrand 1951). Concerning
dolphins, Gunter (1941) write, "It is probably worth
recording that two porpoises, T. truncatus, were
stranded in St. Charles Bay by the low tide and were
forced to remain there, only partially submerged, dur-
ing the coldest days of the freeze. They did not die and
it was reported that they escaped when the tide rose."
There are other reports of bottlenose dolphins in frozen
seas; for example, Manton (1986) reports that T. trun-
catus have been seen breaking ice in the northern part
of the Adriatic Sea. There are no records of dolphin
deaths associated with other recent freezes in East
Matagorda Bay, i.e., in 1983-84 or February 1989;
however, local fishermen stated that the only previous
sightings of dead dolphins (reported as 4 or 5) in the
Bay followed the 1983-84 storm. No data on water
NOTE Miller: Deaths of Tursiops truncatus in East Matagorda Bay
795
temperatures or duration of ice on the Bay was avail-
able for the 1983-84 freeze.
There may have been no possible escape route for
the dolphins because of the very low water level and
the ice formation on the surface of the Bay. Smith et
al. (1983) state that ice may impede the movement of
dolphins in an area, and Shane (1980) studied the
distribution of bottlenose dolphins in southern Texas
and found that some animals had a home range that
was limited to shallow bays. In our study, local fish-
ermen stated that they repeatedly saw the same
animals, which they could recognize by marks on the
dorsal fin and flukes, and that the approximate number
of dolphins in the Bay usually was "in the20's." If the
dolphins in East Matagorda Bay were resident, then
many of the older animals stranded in January 1990
likely had experienced and survived the severe weather
conditions in 1983 when the Bay froze over.
There are no precise data available to accurately
determine the food biomass available to the EMB
dolphins during and after the December 1989 freeze,
but it is possible that an essential food source was not
available. Fish mortality is greatest during a rapid
decrease in water temperature
(Springer and Woodburn 1960).
Data from Dailey et al. (1991a)
show that the relative abundance
(gillnet entrapment technique,
n/hour) of subadult and adult
striped mullet Mugil cephalus along
the Texas coast in spring 1989 was
double that of previous years, while
the relative abundance (bag-seine
entrapment technique, n/ha) of
juveniles in 1989 was only 60% of
the value for the two previous
years. Following the December
1989 freeze, the relative abundance
of subadult and adult striped mullet
in the spring of 1990 was far below
that of spring 1989, while the
relative abundance of juveniles (for
recruitment to the population) was
380% higher in 1990 than it was in
1989. The large increase in relative
abundance of young for spring
recruitment to the population
following the December 1989 freeze
has been attributed to a lack of
adult predator fish (Lawrence
McEachron, Texas Parks Wildl.,
Rockport, TX 78382, pers. com-
mun., Oct. 1991).
Table 2 shows estimated freeze
kills for a variety of marine fish
species in East Matagorda Bay for periods in 1983-84,
February 1989, and December 1989 (McEachron et al.
1991). Although freezes are common on the Texas
coast, fish kills of the magnitude of the December 1989
freeze in the Bay had never before been recorded, with
Mugil cephalus mortality estimated to be over 2.5
million fish.
And thus, a compounding problem for the dolphins
in the December 1989 freeze is the unprecedented kill
of striped mullet. It is probable that a food source
essential to the Bay dolphins was severely depleted at
a critical time when the dolphins needed calories. Bar-
ros (1992), Barros and Odell (1990), and Cockcroft and
Ross (1990) show that bottlenose dolphins utilize a
variety of food resources, composed primarily of fish
(4 to 6 major prey species common to their respective
areas) and cephalopods (primarily 1 species common
to their respective areas), and occasionally crustaceans.
Pryor et al. (1990) suggest that mullet has been a staple
dolphin food for centuries. Gunter (1942) reported on
prey in freshly-killed, presumably-healthy T. truncatus:
1 from deeper waters and 33 from the shallows of
Aransas and St. Charles Bays, Aransas County, Texas.
Table 2
Estimated marine fish-freeze kill
in East Matagorda
Bay for the periods 1983-84, 1
February 1989, and December 1989 (McEachron et al. 1991). ND =
= no data.
February
December
Species
1983-84
1989 "
1989
Micropogonias undulatus
Atlantic croaker
700
100
<100
Pogonias cromis
Black drum
3300
89,900
5600
Brevoortia patronus
Gulf menhaden
7100
100
2400
Arius felis
Hardhead catfish
ND
200
ND
Lagodon rhomboides
Pinfish
300
200
100
Sciaenops ocellatus
Red drum
500
23,500
400
Cynoscion arenarius
Sand seatrout
ND
ND
ND
Archosargus probatocephalus
Sheepshead
4300
11,600
200
Leiostomus xanthurus
Spot
ND
ND
ND
Paralichthys lethostigma
Southern flounder
ND
ND
ND
Cynoscion nebulosus
Spotted seatrout
900
170,400
600
Mugil cephalus
Striped mullet
178.400
21,200
2,684,100
Other fish
19,400
67,400
6,400
Total fish
214,900
384,600
2.699.800
796
Fishery Bulletin 90(4). 1992
Although Gunter found 12 species offish and 1 shrimp,
83% of fish consumed were Mugil cephalus.
A significant difference between stranded (presum-
ably ill) and net-caught/capture-kiiled (presumably
healthy) dolphins is that stranded dolphins (Barros
1992, Barros and Odell 1990) have a high percentage
of empty stomachs (empty or < Ig, 32-54%) while net-
caught or captured dolphins (Cockcroft and Ross 1990,
Gunter 1942) have a very low percentage of empty
stomachs (<3%). The reason for this discrepancy is not
documented, but ill dolphins often have a decreased ap-
petite or may not be able to catch food. Another reason
for a high percentage of empty stomachs in the Bay
dolphins may be lack of food availability. Of 19 Bay
dolphins examined, 32% had empty stomachs and 37%
had only unidentifiable bones and scales (no flesh).
Gunter (1942) observed 34 killed specimens of T. trun-
catus; none of the stomachs were void of food. In addi-
tion, Gunter (1942) showed that the average number
of recognizable fish/stomach was 18, whereas the 15%
of EMB dolphins that had eaten recently had no more
than 2 recognizable fish/stomach. These data, along
with the Texas Parks & Wildlife fish freeze-kill and
biomass data, indicate that food was in short supply
for the Bay dolphins. I suspect that many of them might
have survived if they had sufficient nutrition.
Ridgway and Fenner (1982) state that the blubber
may thin as weight loss progresses to emaciation, and
reduced blubber thickness at necropsy is one sign of
emaciation. Studies on healthy, well-fed dolphins at the
Naval Ocean Systems Center in San Diego show that
T. truncatus have thicker blubber as body weight in-
creases, and that T. truncatus may respond within 2
weeks to water-temperature changes by increasing or
decreasing blubber thickness for cooler or warmer
temperatures, respectively (William A. Friedl, NOSC,
Kaneohe, HI, pers. commun., Nov. 1990). Level of star-
vation may not be the only reason for differences in
blubber thickness between EMB and Texas coast
dolphins: the EMB dolphins might originally have had
thinner-than-normal blubber resulting from living in a
shallow bay with higher-than-average water tempera-
tures (29, 25, and 19°C monthly average water tem-
peratures in EMB for September, October, and
November 1989, respectively); or the normal prey field
in EMB might be limited compared with other areas.
Further work on blubber constituents and factors af-
fecting blubber thickness is needed to determine if blub-
ber thickness is an indicator of starvation as a cause
of death.
The December 1989 EMB freeze, in which temper-
atures stayed near freezing for about 4 days, resulted
in devastation of the dolphins' most-likely major food
source, the striped mullet. The dolphins' emaciated
condition, the substantial reduction in their blubber
thickness, lack of food in their stomachs, the assess-
ment that dolphins lived for 2 weeks following the
freeze, and the EMB fish freeze-kill and biomass data
suggest, in addition to any direct effects to the dolphins
of the extreme cold, that decimation of the food
resource contributed to this acute dolphin mortality
event.
Acknowledgments
I thank the following organizations and individuals for
their contributions that made this report possible:
Texas Marine Mammal Stranding Network (Dr. Ray-
mond Tarpley, founder of the TMMSN; Gina Barron,
for her untiring efforts and organizational skills; Elsa
Haubold and all the other members of the network who
have graciously devoted their time); Texas Parks and
Wildlife Department (Steve Marwitz and Lawrence
McEachron, Rockport, Texas); U.S. Coast Guard;
Houston Helicopter (Mike Boyaki and Dave Beard); Dr.
James Calvin, Department of Statistics, Texas A&M
University, College Station; and Naval Oceanographic
Command Detachment, Asheville, North Carolina.
Special thanks to Dr. Sam H. Ridgway (Naval Ocean
Systems Center) for his critique and support for this
work, and to F.G. Wood for his suggestions. Dr. Ray-
mond J. Tarpley (TMMSN and Texas A&M University),
provided NOSC with information on blubber measure-
ments taken from dolphins stranded on the Texas coast
from 1981 to June 1989.
Citations
Anonymous
1986 Regional organization for the protection of the marine
environment, Report of the first meeting of e.xperts on mor-
tality of marine animals. Kuwait. Nov. 22-23. 1986.
Barros, N.B.
1992 Food habits. In Hansen, L.J. (coordinator), Report on
investigation of 1990 GuK of Mexico bottlenose dolphin strand-
ings, p. 41-46. Contrib. MIA-9293, NMFS Southeast Fish.
Sci. Cent., Miami.
Barros, N.B., and D.K. Odell
1990 Food habits of bottlenose dolphins in the southeastern
United States. In Leatherwood, S., and R.R. Reeves (eds.).
The bottlenose dolphin, p. 309-328. Academic Press, San
Diego.
Cockcroft, V.B., and G.J.B. Ross
1990 Food and feeding of the Indian Ocean bottlenose dolphin
off southern Natal, South Africa. In Leatherwood, S., and
R.R. Reeves (eds.), The bottlenose dolphin, p. 295-308. Aca-
demic Press, San Diego.
Dailey, J. A., J.C. Kana, and L.W. McEachron
1991 Trends in relative abundance in size of selected finfish
and shellfish along the Texas coast: November 1975-December
1989. Manage. Data Ser. 53. Texas Parks Wildl. Dep., Fish.
Wildl. Div., Coastal Fish. Br., Austin, 241 p.
\
NOTE Miller: Deaths of Tursiops truncatus in East Matagorda Bay
797
Geraci, J.R.
1989 Clinical investigation of the 1987-1988 mass mortality
of bottlenose dolphins along the U.S. central and south Atlantic
coast. Report to the National Marine Fisheries Service, U.S.
Navj', Office of Naval Research, and the Marine Mammal Com-
mission, April 1989.
Gunter, G.
1941 Death of fishes due to cold on the Te.xas coast, January,
1940. Ecology 22:203-208.
1942 Contributions to the natural history of the bottlenose
dolphin, Tursiops truncatus (Montague), on the Texas coast,
with particular reference to food habits. J. Mammal. 23:
267-276.
Gunter. G., and H.H. Hildebrand
1951 Destruction of fishes and other organisms on the south
Texas coast by the cold wave of January 28-February 3,
1951. Ecology 32:731-736.
Manton, V.J. A.
1986 Water management. In Bryden, M.M., and R. Harrison
(eds.). Research on dolphins, p. 189-208. Clarendon Press,
Oxford.
McEachron, L.W.. G.C. Matlock, C.E. Bryan. P. Unger, T.J. Cody,
and J.H. Martin
1991 Winter mass mortality of animals in Texas bays. Texas
Parks Wildl. Dep., Fish. Wildl. Div., Coastal Fish. Br., Austin,
47 p.
Pryor, K.. J. Lindbergh, S. Lindbergh, and R. Milano
1990 A dolphin-human fishing cooperative in Brazil. Mar.
Mammal Sci. 6(l):77-82.
Ridgway. S.H.. and C.A. Fanner
1982 Weight-length relationships of wild-caught and captive
Atlantic bottlenose dolphins. J. Am. Vet. Med. Assoc.
181(11):1310-1315.
Scott. G.P.. D.M. Burn, and L.J. Hansen
1988 The dolphin die-off: Long-term effects and recovery of
the population. Proc, Oceans '88, p. 819-823. IEEE, NY.
Shane, S.H.
1980 Occurrence, movements, and distribution of bottlenose
dolphins, T. truncatus, in Southern Texas. Fish. Bull., U.S.
78:593-601.
Smith, T.G., J.R. Geraci, and D.J. St. Aubin
1983 Reaction of bottlenose dolphins, Tursiops truncatus, to
a controlled oil spill. Can. J. Fish. Aquat. Sci. 40:1522-1525.
Springer, V.G., and K.D. Woodburn
1960 An ecological study of the fishes of the Tampa Bay
area. Fla. State Board Conserv. Prof. Pap. Ser. 1:1-104.
/
Application of otolith microchemistry
analysis to investigate anadromy
In Chesapeake Bay striped bass
Morone saxatilis*
David H. Secor
The University of Maryland System. Center for Environmental and Estuarine Studies
Chesapeake Biological Laboratory, Solomons, Maryland 20688-0038
Management of Chesapeake Bay and
coastal striped bass Morone saxa-
tilis fisheries is affected by migra-
tion of large Chesapeake adults in-
to coastal waters. Tagging studies
during the 1930s and 1950s indi-
cated that a small percentage of
Chesapeake striped bass contribute
to the coastal fishery (AHadykov and
Wallace 1952, Mansueti 1961, Mass-
man and Pacheco 1961). However,
work on age- and sex-specific migra-
tion patterns (Chapoton and Sykes
1961, Kohlenstein 1981) suggested
that about half of the females aged
3 + migrate out of the Bay. The cur-
rent consensus appears to be that
young striped bass remain in or
near the tributary in which they
were spawned for 2 or 3 years;
thereafter most males remain in
the Bay, while a substantial number
of females migrate out of the Bay
and remain in coastal waters until
sexually mature (Chapman 1987,
Setzler-Hamilton and Hall 1991).
Although facultative anadromy is
suggested by tagging studies, age-
and sex-specific rates of anadromy
remain largely unknown (ASMFC
1990).
Wave-length dispersive electron
microprobe analysis of strontium/
calcium ratio (Sr/Ca) in otoliths has
recently been employed as a method
for distinguishing between fresh-
water and marine life-history phases
• Contribution 2368, Center for Environmen-
tal and Estuarine Studies, The University
of Maryland System.
of individual fishes (Casselman 1982,
Radtke et al. 1988, Kalish 1990). Sr
is substituted for Ca into the lattice
of aragonitic calcium carbonate
(Kinsman and Holland 1969), and in
otoliths the rate of substitution is in
proportion to its abundance in the
endolymph (Kalish 1989). Sr con-
centration in seawater is more than
one order of magnitude greater
than in freshwater (Bagenal et al.
1973, Radtke et al. 1988, Kalish
1990, Ingram and Sloan 1992).
Therefore, Sr levels in otoliths of
fish exposed to seawater should be
substantially higher than those ex-
posed to freshwater.
Sr/Ca ratio in otoliths of anad-
romous striped bass was analyzed
to determine its usefulness in chart-
ing individual migratory histories.
In a prospectus, Coutant (1990) sug-
gested a similar application to in-
vestigate patterns of estuarine use
by Chesapeake Bay and Roanoke
River striped bass. Here, I looked
for a seasonal pattern in otolith
Sr/Ca ratios that was consistent
with anadromous behavior. An an-
nual cycle of low Sr/Ca ratios dur-
ing spring (exposure to Sr-poor
freshwater) and high ratios during
fall and winter (exposure to Sr-rich
saltwater) was expected in large
adults. If such a pattern existed,
then further research and applica-
tion would be justified. Analysis of
Sr/Ca composition could be applied
to problems of migratory behavior,
spawning, hatchery contribution
to coastal stocks, definition of life-
history traits, environmental degra-
dation (Coutant 1990), and conse-
quences of anadromy to recruit-
ment (e.g., KaHsh 1990).
In this investigation, I related
Sr/Ca ratios to annuli which are
assumed to form in spring (see Dis-
cussion). I used a less traditional
definition for annulus, " . . .a ridge
or a groove in or on the [hard] struc-
ture. . ." (Wilson et al. 1987), be-
cause opaque and translucent zones
did not adequately describe the
microstructure observed under
scanning electron microscopy or
light microscopy.
Methods
Sr/Ca ratios were examined for five
large adults from the Chesapeake
Bay and South Carolina (Table 1).
Adults from the Chesapeake were
presumed to be anadromous based
on their size (Setzler-Hamilton and
Hall 1991); the South Carolina
population is a freshwater popula-
tion, resident to the Santee-Cooper
watershed (Secor et al. 1992).
Chesapeake Bay fish (n 3) were col-
lected by charterboat fisherman
from Solomons, Maryland during
the May 1991 "Maryland Trophy
Season", presumably caught in up-
per Bay waters. South Carolina fish
were collected at a 1989 fishing
tournament. Otoliths were removed,
cleaned in 10% sodium hypochlorite
solution (bleach), and rinsed with
deionized water. They were em-
bedded in Spurr epoxy, sectioned in
a transverse plane with a Buehler
Isomet saw, and mounted on a glass
slide. Otoliths were polished (see
Secor et al. 1991) until all annuli
were visible with transmitted light
on a compound microscope. Otolith
sections were further polished with
3 Jim alumina to limit any surface
structure that could cause artifacts
Manuscript accepted 13 July 1992.
Fishery Bulletin, U.S. 90:798-806 (1992).
798
NOTE Secor: Otolith microchemistry analysis of Morone saxatilis anadromy
799
Table 1
Striped bass Morone saxatilis from Chesapeake Bay (MD and Juv) and |
Santee-Cooper (SC) populations used in electron microprobe
analyses.
TL
Weight
ID
Population Sex Age
(cm)
(kg)
MD-1
Chesapeake Female 21
119
15.5
MD-2
Chesapeake ? 8
94
7.3
MD-3
Chesapeake ? 9
93
8.1
SC-1
Santee-Cooper Female 6
80
4.8
SC-2
Santee-Cooper Female 5
81
5.4
Juv-1
Patuxent River Juvenile 0
—
—
in microprobe analysis (Kalish 1991). Otolith sections were carbon-
coated in a high-vacuum evaporator.
A sagitta from a juvenile striped bass sampled from the Patux-
ent River (Chesapeake Bay tributary) was similarly prepared and
polished so that the core and all increments were sectioned (Secor
et al. 1991). The juvenile's parentage, a 20kg female that was
assumed to be migratory based on its size (Kohlenstein 1981), was
known because the juvenile was a marked hatchery fish released
as a 9-day-old larva.
X-ray intensities for Sr and Ca elements were quantified using
a JXA-840A JEOL wave-length dispersive electron microprobe
(Central Facility for Microanalysis, Univ. Maryland, College Park
MD 20742), with Calcite (CaCOg) and Strontianite (SrCOg) as
standards. Striped bass otoliths were resilient to high-beam power
densities compared with previous work on salmonid otoliths
(Kalish 1990) and showed no diffusion of elements over a 32-sec
counting period (Table 2). This permitted analysis of small
Table 3
Summary statistics for Sr/Ca ratios of Chesapeake Bay and Santee-Cooper
samples. All ratio statistics have been multiplied by 1000 for presentation
purposes. Age is given in parentheses below each sample. Step = distance
between sampled points along transect.
Sample
Transect
Step
(fim)
N
X
SE
Mode
Median
MD-1
1
20
99
2.753
0.094
3.4
2.8
(21)
2
20
99
2.645
0.095
2.3
2.6
3
13
60
3.713
0.124
4.7
4.0
(age>7)
MD-2
_
13
99
2.974
0.155
0
3.0
(8)
MD-3
1
20
100
2.385
0.097
2.8
2.4
(9)
2
13
70
2.323
0.086
1.7
2.3
SC-1
—
20
130
0.937
0.061
0
0.9
(6)
SC-2
—
13
99
0.241
0.054
0
0
(5)
Table 2
Effect of counting time on
strontium and cal- 1
cium counts.
Accelerating voltage
= 25 kV,
probe curreni
= 20 nA, sample size
= 5(im-.
Note that Sr and Ca show no
decline with count- |
ing time w
lich would
indicate
sample
destruction.
Counts/sec
Seconds
counted
Sr
Ca
0
585
16763
4
586
16678
8
576
16725
12
589
16869
16
586
16748
20
578
16847
24
570
16828
28
592
16850
32
603
16813
sample points (5 x 5fim) at high accelerating
voltage (25 kV) and probe current (20 n A).
Background and peak counting times were
each 20 sec for Sr, and 5 sec for Ca. Back-
ground counting times were equally divided
below and above the peak position. The
detection limit for Sr was 580 ppm. Preci-
sion was calculated at <1% for Ca counts
and 8.2% for Sr counts (at Sr/Ca= 0.003)
(1.96o; Goldstein et al. 1981). The electron
beam caused a physical disruption (a pit) at
the section's surface which limited the prox-
imity of adjacent points that could be accur-
ately sampled. Initial analyses of Chesa-
peake sample otoliths at "step" distances of
8/.im resulted in no Sr X-ray counts. This
was probably due to physical disruptions
among adjacent points because surface
structure can cause artifacts in microprobe
analysis (Kalish 1991). Analysis was there-
fore conducted at 13 and 20j:.(m step sizes
where positive counts occurred (Table 3).
Transects (700-2600 ^.im in length) across
annuli in the otolith sections were selected.
The electron microprobe sampled 60-130
points along these transects. Each point re-
quired ~70sec of microprobe time. X-ray in-
tensities were calculated using the ZAP pro-
cedure (Reed 1975), normalized to stan-
dards, and converted to elemental (atomic
weight) ratios.
Due to their close proximity, individual
points were not always visible in probed
otolith sections. To relate Sr/Ca ratios to the
800
Fishery Bulletin 90(4). 1992
opaque zones of annuli checks,
probed sections were viewed
under a compound microscope
and transect distances between
annuli measured with an ocular
scale. Because the distance be-
tween each microprobe mea-
surement was known, distances
between measurements can be
converted to distances between
annuli. Distances between annuli
(annular increments) became
narrow with increasing age (<50
i^m) (Fig. 1), and points did not
always sample directly on annuli.
Therefore, it was necessary to
assign an annulus to the closest
sampled point. Points between
annuli were assumed to sample
age in linear proportion. For in-
stance, if 10 points were sampled
between annuli 5 and 6, then
points would be assigned ages
5.0, 5.1, 5.2,. . .6.0. A replicate
scan was performed on two of
the otolith samples. In the
juvenile's otolith, not all daily in-
crements were visible along the
transect with scanning electron
microscopy or light microscopy.
Therefore, Sr/Ca ratios were
related to standard length using
an otolith/fish-length relation documented for the
Potomac River population (Houde and Rutherford
1992).
Results
Mean Sr/Ca ratios in Chesapeake striped bass were
three to four times greater than Santee-Cooper striped
bass (Table 3; Figs. 2, 3). This trend is consistent with
a salinity influence on the ratio, because Santee-Cooper
striped bass are confined to freshwater and both
Santee-Cooper females were sexually mature. Al-
though substantial variation occurred in Sr/Ca between
South Carolina samples, both samples were near the
electron microprobe's detection limit of Sr/Ca (Sr/Ca
= 0.0008). Instrumental precision decreases markedly
as the detection limit is approached, which may pro-
duce spurious variation. Peaks and nadirs in the Sr/Ca
ratios were apparent in Chesapeake striped bass, and
in fish >age-4, these patterns generally showed an an-
nual cycle (Fig. 2). This is most apparent for sample
MD-1. Because low Sr/Ca ratios can be associated with
s.
^ — —
ftu.n^
(
•
m^ , MD-i
•
' A 2
1
t
•
— . ■- ft ■ ^^^^— 1
0010 30KU
X30 1mm WD20
Figure 1
Back-scatter electron micrograph of otolith from striped bass Morone saxatilis Sample
MD-1. Twenty-one annuli are clearly visible along the sulcus (s) and sulcal ridge. Transects
1 and 2 were performed at 20fim step size; Transect 3 began at the 7th annular check
and 13fjm step size. The probed transect previous to Transect 3 was performed at S^m
step size and resulted in no positive Sr counts. Note that individual probed points are
visible in a series of physical disruptions (pits) for Transects 1 and 2.
freshwater excursions, results indicate yearly migra-
tion for this large female.
Lack of agreement among replicate transects (Fig.
2) probably was due to the manner in which ages were
assigned, spatial resolution, and within-sample vari-
ability. Probed points of replicate transects could not
be directly "lined up" with respect to annuli. This
offset occasionally resulted in the interpretation that
an annuli was associated with a peak in one transect
and a nadir in the other transect (e.g., annuli 15, 18,
and 19 in Transect 1 vs. these annuli in Transect 2
for MD-1; Fig. 2). Transects 1 and 2 for Sample MD-1
(20;jm step size) sampled few points between succes-
sive annuli at older ages, and the accuracy with which
points could be assigned to annuli was less (Fig. 4).
Transect 3 for MD-1 (13fim step size) clearly shows
increased resolution of the ratio across annular in-
crements. Similarly, Transect 2 (13jjm) for MD-3
revealed several more peaks and nadirs after the
5th annulus than did Transect 1 (20^m). The overall
Sr/Ca ratio was significantly different between
Transects 1 and 2 for MD-3 (Table 4, t 3.06, p<
0.01). Replicate transects in MD-1, where step size
NOTE Secor: Otolith microchemistry analysis of Morone saxatilis anadromy
801
Sample: MD-1 , Transect 1
Sample: MD-2
5 :
CO
!i! 4-
12 3 4 5 6 7 8 9 10 1112 13 14 15 16 17 18 19 2021
ST 8
2 3 4 5 6 7 8
3
i ▼ i)HTTt ^%\
m 7
& 6
T T T T T
1 3^
15
tk) u 4
^
Mean Ra
1-
0 -
Mean Ratio
1=
3 2
I ^
W 0
MeanRatio Mean Ratio
23456789
age
1 2 3 4 5 6 7 8 9 101112 1314151617 18 19 20 21
age
Sample: MD-1 , Transect 2
Sample: MD-3, Transect 1
5 p
1 2 3 4 5 6 7 8 9 10 111213141516171819 20 21
5
,0 1 23456789 10
^ 4-
E o
Ki ^'^ ililililT
^^4
Mean Ratio
■
§ 3 -
o
Mean Ratio .2 '^
9 2
E
y 2
3
1 ^:
CO f
Or
3
55
0
7
12 3 4 5 6 7 8 9 10 1112 13 14 15 16 17 18 19 20 21
age
01 23456789 10
age
Sample: MD-1 , Transect 3
Sample: MD-3, Transect 2
6 :
UJ 5 -
IS 3
O
2 1 r
w
0 -
1 2 3 4 5 6 7 8 9 10 1112 131415 16 1718 19 20 2
ititttTTfTtTT
1
1
5 2 3 4 5 6 7 8 9 10
Mean Ratio
.§3:
■S Me
O 2t
an Ratio
E , — , MeanRatio
c .1 ^
O 1 ■ Q
^ [ 1^
OK —
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2
age
23456789 10
age
Transects of Sr/Ca ratios across annuli increments of Mary
and ages > 5 for each sample. See Table 3 for step sizes. A
nadirs in ratios which were consistent with an annual cycle
Figure 2
land striped
rrows indicat
but did not
)ass Morone saxatilis. Mean ratios are plotted for ages <5
e presumed freshwater excursions. Circled arrows indicate
coincide with annuli.
was 20^m for both transects, did not significantly
differ.
Despite the differences among transects, the over-
all trend in Chesapeake fish was a nadir in Sr/Ca ratio
at or shortly after anntili that coincided with spring
spawning runs (Merriman 1941, Robinson 1960).
This trend occurred only in fish >4 years old. There
was a significant increase in the overall Sr/Ca ratio
in fish >5 years for two of three Chesapeake fish
(Table 4).
802
Fishery Bulletin 90(4). 1992
Sample; SC-1
5
-
4
-
E
3
u
ra
:
1 ,
?
i-
II
3
-
1
1)
1
n
00
n
[4
1
i
III
LM^U
" " Mean Ratio
Sample: SC-2
CO
lii 4
><
i 3
o
9 1
in
\±±
1 li i ""
- - Mean Ratio
12 3 4 5 6
age
Figure 3
Transects of Sr/Ca ratios across annuli of adult female Santee-
Cooper (SC) striped bass Morone saxatilis. Mean ratios are
plotted across all ages.
24
20 -
O
0.
Q 16
LU
_l
Q.
s
<
CO 12
u_
O
>
o
g 8
O
m
J Sample MD-1 , Trial 2 (Step=20/jm)
>
Strontium was not detected in the Patuxent River
juvenile striped bass otolith until it reached ~8mmSL
(Fig. 5). Because larvae less than this size tend to utilize
freshwater nurseries (Houde and Rutherford 1992),
this further verified that freshwater residence results
in low levels of otolith Sr.
Discussion
Annulus formation
Rate and season of annulus formation in striped bass
otoliths are critical to interpretation of the results on
annual and seasonal changes in otolith microchemistry.
Heidinger and Clodfelter (1987) validated the hypoth-
esis of yearly annulus formation for young (<5 years
old) striped bass. However, no directed research has
documented the time of annulus formation in striped
bass otoliths or scales despite their widespread use in
fisheries (e.g., Beamish and McFarlane 1983). Several
investigators have made the observation that annuli are
not observed until spawning season in scales (Merriman
1941, Robinson 1960) and otoliths (M. White, S.C.
Wildl. Mar. Resour., Bonneau SC 29431, pers. com-
mun.). Based on these limited observations and the
general trend of spring annulus formation in other
North American temperate fishes, it was assumed that
annular check formation occurred during or just prior
to the spawning season (February-April).
Salinity effect on otolith
microchemistry
Sample MD-1, Trial 3 (Step=13/jnn)
SOum Contour. Slep=13(jn<- — ^------a.
StV^rn Contour, Step=2Qunv
2 3 4 5 6 7
Figure 4
Number of sampled points between annuli checks at two step sizes. Contours for a
hypothetical SOpim annular increment are plotted at 13 and 20jim step sizes.
Because Chesapeake samples
had substantially higher Sr/Ca
ratios than South Carolina sam-
ples, there appears to be a salin-
ity effect on the ratio. This con-
clusion is further substantiated
by the juvenile otolith that was
examined and showed nondetect-
able Sr/Ca ratio during the early-
larval period, a time when Chesa-
peake tributary larvae generally
occur above the salt-wedge
(Uphoff 1989, Houde and Ruth-
erford 1992). Patterns in otolith
Sr/Ca in adults were consistent
with expected seasonal changes
in ambient salinity. The range of
ratios found for Chesapeake
striped bass was similar to those
found by Kalish (1989) for 12
marine species (Sr/Ca 0.0018
NOTE Secor: Otolith microchemistry analysis of Morone saxatilis anadromy
803
Table 4
Comparison of ratios between probed points
<5 or
> 5 years
. Significant |
differences (p< 0.001)
are shown by <
in asterisk.
Sample
Transect
Age<5
AgeS>5
t
-
SD
N
-
SD
N
MD-1
1
2.17
0.89
44
3.22
0.68
55
6.62*
2
2.16
0.89
44
3.03
0.64
55
5.09'
MD-2
-
2.96
1.50
65
3.01
1.66
34
0.15
MD-3
1
2.15
0.93
78
3.20
0.67
22
4.95*
2
2.01
0.62
35
2.61
0.70
35
3.78*
><
E 3
O
E2h
CO 1
Sample: PAX Juvenile-1
4 6
Estimated SL (mm)
Figure 5
Transect of Sr/Ca ratios for the early-larval period from a
juvenile striped bassMorcm« saxatilis sampled from the Patnx-
ent River, 1991. Transect distance was converted to standard
length using regression of standard length on otolith length
for Potomac River striped bass larvae (Houde et al. 1992).
Transect points were converted to larval lengths assuming
a linear growth rate and constant otolith length:fish length
relationship.
-0.0062) and 1 freshwater species (Sr/Ca 0.0005-
0.0010).
Radtke (1984) and others (Townsend et al. 1989,
Radtke et al. 1990) have shown an inverse relationship
between temperature and otolith Sr/Ca ratio. If there
were such an inverse relationship in adult striped bass
otoliths, then ratios would increase during fall and
winter and decrease during spring and summer, a pat-
tern which would to some degree parallel the pattern
seen for anadromous striped bass.
Kahsh (1989, 1991) in recent directed research found
no temperature relationship for otolith Sr/Ca ratio, and
suggested that seasonal changes in fish physiology can
cause incidental correlation between temperature and
Sr/Ca ratios. Based on evidence for seasonal, growth.
and age effects on Sr/Ca ratios, he pos-
tulated that during certain periods of active
metabolism, Ca-binding proteins are more
abundant in the serum which results in a
higher relative fraction of free Sr available
for deposition onto the otolith. If Sr/Ca
ratios in the otolith are controlled by phys-
iological processes alone, then a different
pattern of Sr/Ca ratios would be expected
compared with those observed for striped
bass, i.e., Sr/Ca ratios would tend to rise in
late-winter and early-spring during vitello-
genesis but might also be high during peri-
ods of active growth. However, physiolog-
ical effects such as sexual maturation and stress could
explain both the increase in Sr/Ca ratio after the 5th
annulus in Samples MD-1 and MD-3, and seasonal
(subannular) patterns which varied among samples
(e.g., the major peak which proceeded the 6th annulus
in Sample MD-2; Fig. 2).
Although results exist for few species, the magnitude
of the salinity effect found in this and other studies
(Casselman 1982, KaJish 1989 and 1990) may be greater
than differences expected due to physiological condi-
tion (Kalish 1989, 1991) and temperature alone (Radtke
1984, Townsend et al. 1989, Radtke et al. 1990). Similar
to my findings, Kalish (1989, 1991) reported a three-
to four-fold difference in Sr/Ca ratio between groups
of young rainbow trout exposed to either freshwater
or saltwater. Casselman (1982) reported a three-fold
difference in Sr/Ca ratio between the marine and fresh-
water life-history phases of American eel. In labora-
tory-rearing studies on larval herring Clupea harengus,
temperature effects resulted in no more than a two-
fold difference in Sr/Ca ratios (Townsend et al. 1989,
Radtke et al. 1990), although a complementary field
study conducted by Townsend et al. (1989) showed that
temperatures of 1-12°C had a four-fold effect on Sr/Ca
ratio. Physiological condition has been associated with
an approximate two-fold effect on Sr/Ca ratio in juve-
nile Australian salmon Arripis trutta (Kalish 1989).
A three-fold difference in Sr/Ca ratio in otoliths is
consistent with the probable influence of ambient con-
centrations of Sr and Ca, since the Sr/Ca ratio is at
least four times greater in saltwater than in freshwater
(Casselman 1982, Radtke et al. 1988, Kalish 1989 and
1990). Further, Berg (1968) has shown substantially
less physiological discrimination against Sr than Ca in
scale formation, and Kalish (1989) shows excellent cor-
respondence between otolith microchemistry and the
chemical composition of endolymph that bathes the
otolith. Therefore, ambient levels of Sr could be
reflected in the otolith's microchemistry (Mugiya and
Takahashi 1985, Kalish 1989) dependent upon the
degree of physiological discrimination against Sr.
804
Fishery Bulletin 90(4). 1992
Otolith microchemistry and
migratory history
The otolith microchemistry method offers great poten-
tial to address questions related to time of maturation
and frequency of spawning. A distinct positive shift in
Sr/Ca ratio at 5 years in samples MD-1 and MD-3 could
be indicative of maturation or onset of coastal migra-
tion. Current estimates of age-at-maturation for
Chesapeake population females indicate that <30% of
females are mature by age 5 years (Maryland DNR
1991). Kohlenstein (1981) showed through a tagging
study that the majority of female striped bass migrate
by 5 years. Lack of a shift in MD-2 might indicate that
this individual was a male or had a different migration
history.
All Maryland striped bass samples showed annual
peaks and valleys in Sr/Ca ratios. Based on a salinity
effect, it can be inferred that valleys represent excur-
sions into strontium-poor freshwater habitats. Assum-
ing that large, mature adults venture into freshwater
or low-salinity habitats to spawn, then spawning fre-
quency can be estimated.
Precision error and spatial resolution of the electron
microprobe analysis is critical in the proposed applica-
tion of charting individual migratory histories. Preci-
sion errors were indicated by differences in Sr/Ca pat-
terns and overall level between transects of the same
sample (e.g., MD-3). Changing spatial resolution be-
tween measurements of 20 and 13f^m permitted
greater resolution of seasonal (subannular) patterns.
A more complete series of measures along a transect
is taken at lower step sizes because gaps between
measured points become narrower. This effect could
explain variation in mean Sr/Ca levels among transects
for the same sample. Alternatively, lack of agreement
between transects could indicate machine precision
limits in detecting Sr/Ca levels.
A decline in otolith growth rate wdth age also caused
a decrease in spatial resolution (Fig. 4). At a 13fim step
size, four or five measurements were taken per annular
increment in fish >7 years. Therefore, each measure-
ment can correspond to several months of the fish's
life. This would explain why nadirs in Sr/Ca ratio rarely
approached zero after the 5th annulus. Tagging studies
indicate that adult striped bass can migrate between
freshwater and coastal habitats within a month (Man-
sueti 1961, Chapoton and Sykes 1961, Waldman et al.
1990). Peaks and nadirs observed in otolith Sr/Ca ratio
may therefore represent temporally-averaged values.
Laboratory verification studies are planned to gauge
the spatial sensitivity of otolith microchemistry to
resolve changes in ambient salinity.
Other life-history applications
An ingenious application of the Sr/Ca method to early-
life-history consequences of anadromy was made by
Kalish (1990). He was able to detect Sr in the core of
salmonid otoliths (the earliest deposited material).
Under the rationale that maternally-derived protein in-
fluenced offspring otolith microchemistry, it was pos-
sible to segregate offspring on the basis of whether
they originated from eggs spawned by anadromous
(high Sr/Ca) or nonanadromous (low Sr/Ca) females. In
my study, a single striped bass juvenile of knowTi anad-
romous parentage had no detectable Sr in the otolith
core. In contrast to salmonid embryos and larvae,
striped bass obtain relatively small amounts of mater-
nal protein and lipids, and the period of endogenous
feeding is considerably shorter. Also, the chorion of
striped bass eggs is highly permeable; therefore, am-
bient concentrations of Sr could have a greater in-
fluence on otolith microchemistry.
Substantial variation in Sr/Ca ratio occurred for
young adult (<5 years old) Chesapeake fish. In all
samples, values ranged below detection limits. Pre-
sumably these values represent excursions unrelated
to spawning by young fish into freshwater systems.
Freshwater and slightly saline environments in the
upper reaches of Chesapeake Bay tributaries may serve
as foraging grounds. Future research could analyze the
duration and seasonality of freshwater habitation by
fishes that reside in the Chesapeake Bay.
Further verification studies are needed to establish
whether estuarine and marine phases can be distin-
guished using Sr/Ca ratios. Ratios tended to continu-
ously increase following nadirs, and this pattern could
indicate exposure to waters of increasing salinity. A
verification study could be carried out by probing the
last deposited otolith material (the edge) for Sr/Ca and
relating that ratio to the salinity in which the striped
bass was sampled in the field or through laboratory
rearing studies (Kalish 1989, Townsend et al. 1989,
Radtke et al. 1990). A key comparison will be that be-
tween samples from estuarine habitats (salinity 5-20
ppt) and marine habitats ( > 32 ppt). Should differences
be detectable between these groups, then it will be
possible to infer detailed patterns of anadromy and
related life-history traits.
Acknowledgments
Dr. Philip Picoli at the Central Facility for Microanal-
ysis, University of Maryland, generously provided
expertise and assistance with the electron microprobe.
Drs. Ed Houde, John Kalish, and David Townisend gave
useful criticisms on earlier versions of this manuscript.
NOTE Secor: Otolith microchemistry analysis of Morone saxatilis anadromy
805
Dr. John Dean made available samples of otoliths from
South Carolina striped bass. Bunky's Charter Boat Ser-
vice provided Chesapeake Bay samples. This research
was supported by the U.S. Fish and Wildlife Service
Emergency Striped Bass Study.
Citations
ASMFC (Atlantic States Marine Fisheries Commission)
1990 Amendment 4 to the Atlantic States Marine Fisheries
Commission Interstate Striped Bass Management Report.
Fish. Manage. Rep. 16. Versar Inc., Columbia, MD, 200 p.
Bagenal, T.B., F.J.H. Mackereth, and J. Heron
1973 The distinction between brown trout and sea trout by the
strontium content of their scales. J. Fish Biol. 5:555-557.
Beamish, R.J., and G.A. McFarlane
1983 The forgotten requirement for age validation in fisheries
biology. Trans. Am. Fish. Soc. 112:735-743.
Berg, A.
1968 Studies on the metabolism of calcium and strontium in
freshwater fish. I. Relative contribution of direct and intestinal
absorption. Mem. 1st. Ital. Idrobiol. 23:161-196.
Casselman, J.M.
1982 Chemical analysis of the optically different zones in eel
otoliths. In Loftus, K.H. (ed.), Proc, 1980 North American
eel conference, p. 74-82. Ont. Fish. Tech. Rep. Ser. 4, Ont.
Minist. Nat. Resour.
Chapman, R.W.
1987 Changes in the population structure of male striped bass,
Morone saxatilis, spawning in three areas of the Chesapeake
Bay from 1984 to 1986. Fish. Bull., U.S. 85:167-170.
Chapoton, R.B., and J.E. Sykes
1961 Atlantic coast migration of large striped bass as evidence
by fisheries and tagging. Trans. Am. Fish. Soc. 90:13-20.
Coutant, C.C.
1990 Microchemical analysis of fish hard parts for reconstruct-
ing habitat use: Practice and promise. In Parker, N.C., et
al. (eds.), Fish marking techniques, p. 574-580. Am. Fish. Soc.
Symp. 7, Bethesda.
Goldstein, J. I., D.E, Newberry, P. Echlin, D.C. Joy. C. Fiori, and
E. Lifshin
1981 Scanning electron microscopy and x-ray microanalysis.
Plenum Press, NY, 675 p.
Heidinger, R.C., and K. Clodfelter
1987 Validity of the otolith for determining age and growth
of walleye, striped bass, and smallmouth bass in power cool-
ing ponds. In Summerfelt, R.C. (ed.). Age and growth offish,
p. 241-251. Iowa State Univ. Press, Ames.
Houde E.D., and E.S. Rutherford
1992 Egg production, spawning biomass and factors influ-
encing recruitment of striped bass in the Potomac River and
Upper Chesapeake Bay. Final Rep. to Maryland Dep. Nat.
Resour., Contract CB89-001-003. Univ. Maryland. Cent. En-
viron. Estuarine Stud., Ref. [UMCEES]-CBL 92-017, 313 p.
Ingram, B.L., and D. Sloan
1992 Strontium isotopic composition of estuarine sediments
as paleosalinity-paleoclimate indicator. Science (Wash. DC)
255:68-72.
Kalish, J.M.
1989 Otolith microchemistry: Validation of the effects of
physiology, age and environment on otolith composition. J.
Exp. Mar. Biol. Ecol. 132:151-178.
1990 Use of otolith microchemistry to distinguish the progeny
of sympatric anadromous and non-anadromous salmonids.
Fish. Bull,, U.S. 88:657-666.
1991 Determination of otolith microchemistn,-: Seasonal varia-
tion in the composition of blood plasma, endolymph and otoliths
of bearded rockcod Pseudophyds barbatus. Mar. Ecol. Prog.
Ser. 74:137-159.
Kinsman, D.J.J. . and H.D. Holland
1969 The co-precipitation of cations with CaCOj— IV. The co-
precipitation of Sr^* with aragonite between 16° and 96°C.
Geochim. Cosmo. Acta. 33:1-17.
Kohlenstein, L.C.
1981 On the proportion of the Chesapeake Bay stock of striped
bass that migrates into the coastal fishery. Trans. Am. Fish.
Soc. 110:168-179.
Mansueti, R.J.
1961 Age, growth and movements of the striped bass. Roccus
saxatilis. taken in size selective fishing gear in Maryland.
Chesapeake Sci. 2:9-36.
Maryland DNR
1991 Investigation of striped bass in Chesapeake Bay.
USFWS Federal Aid Project F-42-R-3, 1989-1990. Md. Dep.
Nat. Resour., Tidewater Admin.. 193 p.
Massman, W.H., and A.L. Pacheco
1961 Movements of striped bass in Virginia waters of the
Chesapeake Bay. Chesapeake Sci. 2:37-44.
Merriman, D.
1941 Studies on the striped bass, Roccus saxatilis, of the Atlan-
tic Coast. U.S. Fish Wildl, Serv. Fish. Bull. 50:1-77.
Mugiya, Y., and K. Takahashi
1985 Chemical properties of the saccular endolymph in the rain-
bow trout, Salmo gairdneri. Bull. Fac. Fish. Hokkaido Univ.
36:57-63.
Radtke, R.L.
1984 Formation and structural composition of larval striped
mullet otoliths. Trans. Am. Fish. Soc. 113:186-191.
Radtke, R.L., R.A. Kinzie III, and S.D. Folsom
1988 Age at recruitment of Hawaiian freshwater gobies. En-
viron. Biol. Fishes 23:205-213.
Radtke, R.L., D.W. Townsend. S.D. Folsom, and M.A. Morrison
1990 Strontium: Calcium ratios in larval herring otoliths as
indicators of environmental histories. Environ. Biol. Fishes
27:51-61.
Reed, S.J.B.
1975 Electron microprobe analysis. Cambridge Univ. Press,
Cambridge, 400 p.
Robinson, J.B.
1960 The age and growth of striped bass (Roccus saxatilis) in
California. Calif. Fish Game 46:279-290.
Secor, D.H., J.M. Dean, and E.H. Laban
1991 Manual for otolith removal and preparation for micro-
structure examination. Baruch Press, Univ. South Carolina,
Columbia, 85 p.
Secor, D.H., J.M. Dean, T.A. Curtis, and F.W. Sessions
1992 Effect of female size and propagation methods on larval
production at a South Carolina striped bass [Morone saxatilis)
hatchery. Can. J. Fish Aquat. Sci. 49:1778-1787.
Setzler-Hamilton, E.M., and L.W. Hall Jr.
1991 Striped bass Morone saxatilis. In Funderburk, S.L.. et
al. (eds.). Habitat requirements for Chesapeake Bay living
resources, 2d ed., p. 13-1-13-31, plus maps. Living Resources
Subcommittee, Chesapeake Bay Program, Annapolis.
806
Fishery Bulletin 90(4). 1992
Townsend, D.W., R.L. Radtke, M.A. Morrison, and S.D. Folsom
1989 Recruitment implications of larval herring overwinter-
ing distributions in the Gulf of Maine, inferred using a new
otolith technique. Mar. Ecol. Prog. Ser. 55:1-13.
Uphoff, J.H.
1989 Environmental effects on survival of eggs, larvae and
juveniles in the Choptank River, Maryland. Trans. Am. Fish.
Soc. 118:251-263.
Vladykov, V.D., and D.H. Wallace
1952 Studies of striped hass Roccus saxatilis (Walbaum) with
special reference to the Chesapeake Bay region during
1936-1938. Bull. Bingham Oceanogr. Collect., Yale Univ.
14(1):132-177.
Waldman, J.R., D.J. Dunning, Q.E. Ross, and M.T. Mattson
1990 Range dynamics of Hudson River striped bass along the
Atlantic coast. Trans. Am. Fish. Soc. 119:910-919.
Wilson, C.A. et al.
1987 Glossary. /« Smnmerfelt, R.C., and G.E. Hall (eds.). Age
and growth of fish, p. 527-530. Iowa State Univ. Press, Ames.
Leatherback turtle captured by ingestion
of squid bait on swordfish longiine
Robert A. Skillman
George H. Balazs
Honolulu Laboratory, Southwest Fisheries Science Center
National Marine Fisheries Service, NOAA
2570 Dole Street, Honolulu, Hawaii 96822-2396
The leatherback turtle Dermochelys
coriacea is the only species of the
family Dermochyidae. The other six
extant marine turtles are hard-
shelled members of the family Che-
loniidae. The leatherback inhabits
the pelagic marine environment, ap-
parently only leaving to breed in
coastal waters. With recorded dives
to 475m, it is among the world's
deepest-diving vertebrates (Eckert
et al. 1986). With weights up to 916
kg, it is the world's largest turtle
(Eckert and Luginbuhl 1988). The
leatherback is listed as endangered
under the U.S. Endangered Species
Act, the International Union for
Conservation of Nature, and the
Convention on International Trade
in Endangered Species. Conse-
quently, fishery interactions involv-
ing the leatherback are of concern.
This paper reports an interaction
with longiine gear while fishing for
swordfish.
Leatherbacks ingest and become
entangled in marine debris (Balazs
1985), and they are taken by oper-
ative fishing gear (Nishemura and
Nakahigashi 1990). Entanglement
has been reported with lobster pot
lines (Lazell 1976), drift nets (Balazs
1982, Wetherall et al. In Press);
pelagic longiine (Witzell 1984,
Tobias 1991); gillnets (Margaritoulis
1986); and swordfish Xiphias
gladius tangle nets (Frazier and
Brito Montero 1990). Interactions
with tuna and swordfish longiine
fishing have involved entanglement
and foul-hooking, particularly with
the leatherback's long flippers
(Honolulu Star-Bulletin 1935,
Witzell 1984, Dollar 1991, Tobias
1991, USFWS 1969). In the Hawaii
swordfish fishery, sightings of
leatherbacks and reported interac-
tions are not rare, particularly in
the area of the seamounts above the
Northwestern Hawaiian Islands
(Robert Dollar, NMFS Honolulu
Lab., pers. commun.). It is not un-
common for leatherbacks to become
entangled in driftnets set north
of Hawaii between 30° and 45°N.
However, virtually nothing is
known about their overall distribu-
tion, abundance, and life history in-
cluding stock structure (Wetherall
et al., In press). The nearest col-
onies of nesting leatherbacks occur
in the eastern Pacific along the
coast of Mexico and Costa Rica and
in the western Pacific in peninsular
Malaysia. To our knowledge, inges-
tion of baited hooks has not been
reported in the literature.
Leatherbacks are known to feed
on gelatinous, pelagic animals.
These include the medusa of sycho-
zoan coelenterates (true jellyfish)
(Bleakney 1965, Brongersma 1969)
and hydrozoan coelenterates (Por-
tuguese man-of-war Physalia are-
thusa) (Bacon 1970). Davenport
(1988) and Davenport and Balazs
(1991) have suggested the potential
importance of bioluminescence in
the predation of free-swimming
colonial tunicates (pyrosomas) by
leatherbacks during the night or on
deep dives. Neither fish (tuna bait)
nor squid (swordfish bait) have been
cited in the literature as prey of
leatherbacks. Accordingly, Witzell
(1984) stated that leatherbacks are
not likely to be taken on a baited
hook.
The present paper presents docu-
mentation of a leatherback captured
after ingesting squid bait on sword-
fish longiine gear. The chemical
light sticks used to attract sword-
fish may have attracted the leather-
back to the gear.
On 24-25 January 1991, while ex-
perimental longiine fishing opera-
tions were being conducted for
swordfish from the NOAA research
ship Toumsend Cromwell, a leather-
back turtle was hooked and released
alive at lat. 26°58.3'N, long. 168°
53.5'W. The turtle swam vigorous-
ly while being hauled next to the
research vessel and after being
released. The hook line could be
seen coming from the turtle's
mouth, but the exact location of the
hook was not apparent. No blood or
external injuries were apparent. A
tree branch lopper on the end of an
extendable fiberglass pole was used
to cut the hook line a few centi-
meters from the turtle's mouth. The
estimated carapace length of the
turtle was 170 cm. The turtle was
too large to haul on board, and the
prevalence of sharks, including blue
shark Prionace glauca, made it im-
possible to enter the water for ac-
curate measurement or tagging.
Other site specifics included the
following: 2400 m bottom depth,
21.4°C sea surface-water tempera-
ture, 18.9°C air temperature,
150-180cm sea swells, northeaster-
ly trade winds at 15 kn, and approx-
imate depth of the upper mixed
layer at 85 m.
Details of the set and gear are as
follows. The longiine gear consisted
of ~16km of 4.0 or 3.2mm mono-
filament main line suspended with
floats every 3 hook lines. The gear,
with 206 hooks, was set on 24
Manuscript accepted 6 August 1992.
Fishery Bulletin, U.S. 90:807-808 (1992).
807
808
Fishery Bulletin 90(4). 1992
January 1991 starting at lat. 27°01.720'N, long.
168°56.153'W, in the vicinity of an unnamed seamount
some 940 km east of Midway Is. and 150 km north-
northeast of Raita Bank, Northwestern Hawaiian Is.
The gear was hauled on 25 January beginning at lat.
26°59.094'N, long. 168°57.810'W. The turtle was
taken on the first hook of a 3-hook basket located about
mid-set. This hook was set at 1818h and hauled at
0907 h the next day, for a soak time of 14 h 49 min.
Because the hook timer (a 7.5x3 cm cylinder of clear
plastic resin with an embedded clock chip; Somerton
et al. 1988) on that line was lost, an estimate of the
time of hooking is not available. The hook timer on the
second hook-line, with full bait remaining, was set off
at 051 Ih; the hook timer on the first hook of the
previous basket, with the bait missing, was set off at
0159 h. The float lines, made of polypropylene rope,
were 9m long. The hook droppers, made of 2.1mm
monofilament, were 13 m long with a 60g weighted
swivel 4 m from the hook. Thus, the depth of hook 1
was nominally ~22m, unless altered by currents, since
the first hook-line of each basket was attached within
3 m of the float. Green, 12 h chemical light sticks were
placed ~2m above each of the 206 hooks (the light
sticks still glowed weakly at the time of hauling). Each
hook was baited with a whole, previously-frozen Argen-
tinean squid {Illex sp.), weighing ~0.34kg.
While entanglement of leatherbacks in pelagic long-
line and other gears has been described, ours is ap-
parently the first report of a hook and bait being eaten.
Chemical light sticks used on swordfish longline may
impose an added hazard for leatherbacks by simulating
natural prey. The magnitude of the take, the level of
mortality or serious injury, and the impact on the
leatherback stock are unknown. Additional data on the
take by pelagic fisheries as well as information on
leatherback feeding habits, stock structure, and popula-
tion dynamics would be needed to evaluate the impact
of the take.
Citations
Bacon, P.R.
1970 Studies on the leatherback turtle Dermochelys coriacea
(L.), in Trinidad, West Indies. Biol. Conserv. 2:213-217.
Balazs, G.H.
1982 Driftnets catch leatherback turtles. Oryx 16:428-430.
1985 Impact of ocean debris on marine turtles: Entanglement
and ingestion. In Shomura. R.S., and H.O. Yoshida (eds.),
Proceedings, Workshop on the fate and impact of marine
debris, 27-29 November 1984, Honolulu, p. 387-429. NCAA
Tech. Memo. NMFS-SWFC-54, NMFS Honolulu Lab.
Bleakney, J.S.
1965 Reports of marine turtles from New England and eastern
Canada. Can. Field-Nat. 19:120-128.
Brongersma, L.D.
1969 Miscellaneous notes on turtles, IIB. Proceedings, K.
Ned. Akad. Wet. Ser. C. Biol. Med. Sci. 72:90-102.
Davenport, J.
1988 Do diving leatherbacks pursue glowing jelly? Br.
Herpetol. Soc. Bull. 24:20-21.
Davenport, J., and G.H. Balazs
1991 'Fiery bodies'— Are pyrosomas an important component
of the diet of leatherback turtles? Br. Herpetol. Soc. Bull.
37:33-38.
Dollar, Robert A.
1991 Summary of swordfish longline observations in Hawaii,
July 1990-March 1991. Admin. Rep. H-91-09, NMFS Hono-
lulu Lab.. 13 p.
Eckert, K.L.. and C. Luginbuhl
1988 Death of a giant. Mar. Turtle Newsl. 43:2-3.
Eckert, S.A., D.W. Nellis, K.L. Eckert, and G.L. Kooyman
1986 Diving patterns of two leatherback sea turtles {Dermo-
chelys coriacea) during internesting intervals at Sandy Point,
St. Croix, U.S. Virgin Islands. Herpetologica 42(3):3'81-388.
Frazier, J.G., and J.L. Brito Montero
1990 Incidental capture of marine turtles by the swordfish
fishery at San Antonio, Chile. Mar. Turtle Newsl. 49:8-13.
Honolulu Star-Bulletin
1935 [Photograph with a caption indicating a 200 kg leather-
back turtle was found entangled in the line and hooks of a sam-
pan, a Japanese-style fishing boat probably using longline
gear.] Honolulu Star-Bulletin, 8 April 1935, p. 3.
Lazell, J.D. Jr.
1976 This broken archipelago. Cape Cod and the islands, am-
phibians and reptiles. Demeter Press Book, Grafelfing, Ger-
many, p. 191.
Margaritoulis, D.N.
1986 Captures and strandings of the leatherback sea turtle,
Dermochelys coriacea, in Greece (1982-1984). J. Herpetol.
20(3):471-474.
Nishemura, W.. and S. Nakahigashi
1990 Incidental capture of sea turtles by Japanese research
and training vessels: Results of a questionnaire. Mar. Turtle
Newsl. 51:1-4.
Somerton, D.A., B.S. Kikkawa. and CD. Wilson
1988 Hook timers to measure the capture of individual
fish. Mar. Fish. Rev. 50(2): 1-5.
Tobias, W.
1991 Incidental catch a continuing problem in the Mediterra-
nean. Mar. Turtle Newsl. 51:10-12.
USFWS (U.S. Fish & Wildlife Service)
1969 Cruise report, USFWS ship Townsend Cromwell, cruise
44. USFWS Hawaii Area Biol. Lab., 4 p. [Avail. NMFS
Honolulu Lab.]
Wetherall, J. A.. G.H. Balazs. R.A. Tokunaga. and M.Y.Y.Yong
In press Bycatch of marine turtles in North Pacific high-seas
driftnet fisheries and impacts on the stocks. In Proc, Int.
North Pac. Fish. Comm., Nov. 4-6, 1991, Tokyo.
Witzell. W.N.
1984 The incidental capture of sea turtles in the Atlantic U.S.
Fi.shery Conservation Zone by the Japanese tuna longline fleet,
1978-81. Mar. Fish. Rev. 46(3):56-58.
Reproductive biology of tlie swordfisii
Xiphias gladius in tlie Straits of
Florida and adjacent waters
Ronald G. Taylor
Michael D. Murphy
Florida Marine Research Institute, Department of Natural Resources
100 Eighth Avenue SE, St. Petersburg. Florida 33701-5095
The swordfish Xiphias gladius Lin-
naeus inhabits all tropical, subtrop-
ical, and temperate oceans of the
world, including the Mediterranean
Sea and the Gulf of Mexico. In the
western Atlantic, it is found from
Newfoundland to Argentina (Palko
et al. 1981, Nakamura 1985). Sword-
fish occur in the Florida Straits at
all times of the year. Prior to 1970,
swordfish were pursued primarily
by recreational fishermen. During
the 1970s, the fishery in Florida at-
tracted displaced Cuban-Americans
and New England longline fisher-
men, and by 1980 commercial land-
ings from the east coast of Florida
had reached nearly ISOOmt (Berke-
ley and Irby 1982).
Little is known about the repro-
ductive biology of swordfish in the
western Adantic. Ovchinnikov (1970)
and Berkeley and Houde (1980)
reported contradictory findings on
male and female sizes-at-maturity.
Wilson (1984) reported that males
mature at younger ages than do
females in the U.S. south Atlantic.
Descriptions of swordfish spawning
season and spawning grounds have
been based on the temporal and
areal distribution of infrequently-
collected larvae and juveniles (Ara-
ta 1954, Tibbo and Lauzier 1969,
Markle 1974, Grail et al. 1983). Our
research objectives were to deter-
mine the size- and age-at-maturity,
spawning season, and approximate
spawning grounds of swordfish in
the Straits of Florida and adjacent
waters. Data were collected as part
of a joint Florida Marine Research
Institute and University of Miami
investigation of the fishery and biol-
ogy of the swordfish. Samples gath-
ered during this research have been
used to develop a method to deter-
mine the ages of swordfish and to
describe their growth (Berkeley and
Houde 1984).
Methods and materials
Swordfish were sampled from rec-
reational and commercial catches
made off southeast Florida (Fig. 1)
from June 1977 through November
1980. Each year most of the collec-
tions were made April through Sep-
tember. Samples were taken at
least once each month over the 2.5
yr sampling period, except in De-
cember when no samples were
taken either year. Because of the
varied conditions of landed sword-
fish, a variety of length measure-
ments (to the nearest cm) were
taken: total length (TL), distance
from the tip of the bill to the mid-
point of the line connecting the
distal edges of the caudal-fin lobes;
fork length (FL), from the tip of the
bill to the distal end of the central
ray of the caudal fin; lower jaw to
fork length (LJFL), from the tip of
the lower jaw to the distal end of the
central ray of the caudal fin; eye to
fork length (EFL), from the poste-
rior margin of the eye's bony orbit
to the distal end of the central ray
of the caudal fin; and trunk length
(TRNKL), from the posterior mar-
gin of the gill cavity to the point of
least circumference of the caudal
peduncle. Lower jaw to fork length
is used throughout this paper unless
otherwise noted. For fish measured
only for TL, FL, EFL, or TRNKL,
LJFL was estimated using the
appropriate regression equation
(Table 1). Whole weight (W) was
determined to the nearest pound
and converted to kilograms for our
analyses. Portions of ovaries and
testes were collected and preserved
in Davidson's fixative (Humason
1972). Whole gonads, macroscop-
ically judged ripe or mature based
on the presence of transparent
eggs, were preserved and then
weighed to the nearest gram.
Swordfish maturity was described
using histological features to define
gonadal development. Subsamples
of preserved gonads were embedded
in paraffin, sectioned at Gj^m, stained
with Mayer's haematoxylin and
eosin, and mounted for microscopic
examination. Swordfish were as-
signed to one of eight developmen-
tal classes following Murphy and
Taylor (1990) and based on the ap-
pearance of histological features
described by Grier (1981) for males
and Wallace and Selman (1981)
for females. These developmental
classes and the mean observed
oocyte diameters are (1) Immature,
<20^im- (2) Developing, llt^m; (3)
Maturing, IGO^m; (4) Mature, 434
Mm; (5) Gravid, 723 ^m; (6) Spawn-
ing/Partially Spent, 823 Mm; and (7)
Spent, 181 Mm. The relationship
between swordfish maturity and
length was described for each sex
using maturity data for fish grouped
into 10 cm size-classes. A logistic
distribution function was fit to the
percentages of mature fish ( > Glass
4) and the midpoints of their size-
classes (Saila et al. 1988) in order
to predict a maturity schedule. A
similar distribution function was
generated for maturity against age.
Manuscript accepted 2 July 1992.
Fishery Bulletin, U.S. 90:809-816 (1992).
809
810
Fishery Bulletin 90(4). 1992
r
20 mi
Figure I
Areal extent of sampling locations (shaded area) off southeast Florida during June
1977-November 1980. Triangles indicate where female swordfish Xiphias glndius were
found with histological evidence for recent (postovulatory follicles) or imminent (hydrated
oocytes) spawning (see "Methods and materials").
Table 1
Linear regressions of lower jaw to fork length (LJFL) on total length (TL). fork length |
(FL), eye to fork length (EFL), or
trunk length (TRNKL); and nonlinear
regressions
of whole weight (W) on lower jaw to fork length anc
LJFL on W, for swordfish Xiphias \
gladius off southeast Florida (regression analysis,
SAS 1982).
Equation
N
Range
r^
LJFL = -6.03 + 0.662 (TL)
401
32-432 em TL
0.982
LJFL = -5.51 + 0.714(FL)
100
30-396 cm FL
0.983
LJFL = 8.89 + 1.076(EFL)
316
68-249cmEFL
0.995
LJFL = 15.71 + 1.402(TRNKL)
324
18-189cmTRNKL
0.987
W = 1.050 X lO-'i LJFL'"»
127
27-281 cm LJFL
0.973
LJFL = 48.58^^°^
127
0.090-168. OkgW
0.980
Ages for the swordfish that com-
prise this data set were deter-
mined by Berkeley and Houde
(1984), who counted unvalidated
age marks found on thin-sections
of the second anal-fin spine.
Temporal differences in mean
oocyte diameters were used to
define spawning season. To de-
termine mean oocyte diameter
for each individual, 100 oocytes
in a common lamella were mea-
sured with an ocular micrometer.
Mean oocyte diameters were cal-
culated for all collections in a
given month and plotted to ex-
amine monthly changes. The dis-
tribution of oocyte diameters was
also examined within individuals
to determine whether swordfish
undergo multiple spawns or a
single spawn each year.
We used the distribution of
swordfish captured in near-term
spawning condition to delimit
their spawning grounds off south-
east Florida. Histological fea-
tures indicative of recent or im-
minent spawning included post-
ovulatory follicles and hydrated
oocytes (DeMartini and Fountain
1981, Hunter and Macewicz 1985).
For a variety of fishes, it has
been found that oocytes hydrate
during the late-afternoon or eve-
ning just prior to spawning {Ser-
rifus politics, DeMartini and Foun-
tain 1981; Engraulis mordax,
Hunter and Macewicz 1985; Cyno
scion nebulosus, Brown-Peterson
et al. 1988; Sciaenops ocellatus,
Fitzhugh et al. 1988). Postovula-
tory follicles are identifiable only
for a short time. Following spawn-
ing, they are rapidly absorbed
(within 6h for Callionymus enne-
actis, Takita et al. 1983) and
quickly become indistinguishable
from other atretic structures
(within 2d in Engraulis mordax,
Hunter and Macewicz 1985).
Batch fecundity was estimated
gravimetrically from counts of
ova >750Mm diameter in a 2-3 g
portion from the midsection of
each preserved ovary (n 7).
NOTE Taylor and Murphy Reproductive biology of Xiphias gladius in Straits of Florida
Results
Sex and length data were collected from 554 swordfish: 211
females 72-281 cm, and 343 males 82-235cm. Gonads were avail-
Table 2
Observed and predicted percentages of mature (
> Class 4, see
"Methods
and materials") swordfish Xip/ims
gladius in 10 cm LJFL length
intervals.
Predicted percentages were calculated from logistic distribution functions |
fit to observed maturity data (see
'Results"). Numbers in parentheses are |
numbers of fish examined.
Lower jaw
Male
Female
to fork length
interval midpoint
Observed
Predicted
Observed
Predicted
(cm)
%(N)
%
%(N)
%
80
0(1)
0
90
0(3)
11
0(2)
0
100
33(3)
24
0(7)
0
110
46(11)
45
0(17)
0
120
64(11)
69
0(19)
1
130
90(10)
85
0(14)
3
140
96(24)
94
0(8)
5
150
93(14)
98
0(2)
10
160
100(11)
99
0(3)
18
170
100(14)
100
50(4)
30
180
100(11)
100
60(10)
47
190
100(8)
100
56(16)
64
200
100(6)
100
71(14)
78
210
100(5)
100
75(16)
87
220
100(1)
100
100(10)
93
230
100(1)
100
100(6)
97
240
100(3)
98
250
100(5)
99
260
100(2)
100
270
280
100(3)
100
Table 3
Observed and predicted percentages of mature (» Class 4, see "Methods |
and materials") swordfish Xi-pt
ias gladiiis by assigned age group (ages from
Berkeley and Houde 1984). Predicted percentages were calculated from
logistic distribution functions fit to observed maturity data (see "Results").
Numbers in
parentheses are
numbers of fish examined.
Age
Male
Female
Observed
Predicted
Observed Predicted
(years)
%(N)
%
%{N) %
0
0(1)
4
1
30(10)
29
0(13) 0
2
80(10)
79
0(18) 1
3
93(14)
97
0(6) 5
4
92(12)
100
18(11) 14
5
100(12)
100
38(8) 37
6
100(8)
100
60(10) 66
7
100(3) 87
8
100(2)
100
83(6) 96
9
100(4) 99
10
100(1) 100
able for histological processing from 295 fish
(133 males and 162 females), of which 149
were ascribed ages.
Male swordfish mature at a smaller size
and younger age than do females. Males
begin to mature at ~ 100 cm at age 1 (Tables
2, 3). The proportion of mature males in our
samples increased rapidly thereafter, and all
males were mature by 160 cm or age 5. In
contrast, the smallest mature females were
~170cm or age 4, and all females were
mature by 220 cm or age 9. The predicted
length at 50% maturity, based on the fit of
logistic distribution functions to the percent-
age mature within size-classes, was signif-
icantly less (approximate ^test, Sokal and
Rohlf 1981; t' 36.5, df 33, P<0.001) for
males (112cm) than for females (182cm).
Likewise, age at 50% maturity was signif-
icantly younger for males (1.4yr) than for
females (5.5yr; t ' 17.7, df 16, P<0.001). The
logistic distribution functions for maturity
by size-class and age fit observed data well
(Tables 2, 3) and are as follows:
Males:
% Mature = i/(i + e(-oo976(LJFL-n2)))
(n 15, r- 0.980)
% Mature = i/(i + e(-2-223(AGE-i.40)))
(n 8, r2 0.990)
Females:
% Mature = i/(i + e(-o.«690(LJFL-i82)))
(n 20, r2 0.966)
% Mature = i/(i + e(-i-234(AGE-5.45)))
{n 10, r2 0.976).
Swordfish from southeast Florida waters
demonstrate group-synchronous oocyte
maturation {sensu Wallace and Selman
1981). This pattern of oocyte development
is characterized by the presence of at least
two distinct groups of dissimilar-sized
oocytes during the spawning season. Ova-
ries from all swordfish in our samples con-
tained a dominant group of oocytes <200f.im
diameter (Fig. 2). All oocytes within Matur-
ing ovaries (Class 3) were <200/^m, although
lipid deposition suggests preparation for ac-
tive vitellogenesis. Mature ovaries (Class 4)
contained an additional distinct group of
vitellogenic oocytes at 200-600 ^^m. A third
group of oocytes, 600-llOO^im, were pres-
ent only in Gravid and Spawning/Partially
Spent fish (Classes 5 and 6). This largest
Fishery Bulletin 90(4), 1992
20
10
l^
5 -
CLASS 3
n = 975
40
20
> 15
S ^
lU 0
c
u.
%
I I I I I I I I I I I I M I M I I M I I ] I I I I M I 1 I I I I M I I I I I
200 400 600 800 1000 1200
CLASS 4
n = 950
I I I I I I I I I I ' " I I I I fi "I I I ' I ' I I I ' ' ' I ' I I I ' ' ' I " I I
0 200 400 600 800 1000 1200
I I I i| II i|
1000 1200
'I" 'I
200 400 600 800 1000 1200
OOCYTE DIAMETER {urn)
Figure 2
Percent frequency of oocyte diameters for swordfish Xiphias
gladius collected during the spawning season off southeast
Florida with ovaries in Classes: (3) Maturing, (4) Mature. (5)
Gravid, and (6) Spawning/Partially Spent (see "Methods and
materials").
group of oocytes proceeds through final maturation and
represents the clutch (Wallace and Selman 1981) to be
shed during the next spawning event.
The reproductive season for swordfish is protracted.
Mature or actively spawning females were found dur-
ing each month sampled, except January. Gravid or ac-
tively spawning males were found during all months
sampled (Fig. 3). The greater numbers of spawning fish
taken from late-spring to midsummer suggests in-
creased spawning activity then. In addition, mature
swordfish show a sharp increase in their maximum
oocyte diameters to >800^m beginning in April and
extending through July, indicating peak spawning
activity then (Fig. 4).
Histological features (postovulatory follicles and
hydrated oocytes) provide evidence for active sword-
fish spawning off the Atlantic coast of Florida from
about 24°40'N in the Straits of Florida southwest of
Duck Key to about 28°25'N just west of Cape Canaver-
al (Fig. 1). The easternmost location where fish in
spawning condition were collected was 16 km east of
Grand Bahama Island at 480 m depth. About 25% of
the fish collected at the westward extent of our sam-
pling area (about the 200 m contour) exhibited evidence
of recent or imminent spawning.
There was a preponderance of smaller males in our
spring and summer samples. Overall, males significant-
ly dominated the catch (345 M:216 F; x'- 29.7, df 1,
P<0.001); although at sizes >200cm, females signifi-
cantly outnumbered males (22 M:67 F; x" 22.8, df 1,
P<0.001). There was no histological indication of
hermaphroditism throughout the 72-281 cm length
range. During spring and summer, males were domi-
nant (266 M:152 F; r 31.1, df 1, P<0.001); but dur-
ing the fall and winter there was no significant differ-
ence in abundances of sexes (79 M:64 F; x" 1.6, df 1,
P>0.05).
The batch fecundity of the seven swordfish sampled
was 1.4-4.2 million eggs for swordfish 177-281 cm
and 69-268 kg (Table 4). There was no obvious rela-
tionship between our estimates of batch fecundity
and length or weight. The correlation coefficients
between batch fecundity and length (r 0.21) and be-
tween batch fecundity and weight (r 0.64) were not
significantly different from zero (table of critical values
for correlation coefficients, df 5, P>0.05, Rohlf and
Sokal 1981).
Discussion
Sizes-at-maturity have been reported for swordfish in
the Atlantic and Pacific Oceans and Mediterranean
Sea (Ovchinnikov 1970, Berkeley and Houde 1984,
DeMetrio et al. 1989). However, in these reports
maturity was determined by visual inspection of
gonads, and no explanations were given as to whether
the sizes reported were for first maturity or for 50%
maturity. Berkeley and Houde (1984) reported that
western Atlantic male and female swordfish mature at
~21kg and 74 kg, respectively. This corresponds
closely to the lengths-at-50%-maturity we present for
swordfish off southeast Florida. Using equations in
Table 1, we determined 50% maturity of males at 18 kg
(112 cm LJFL converted to whole weight) and 50%
maturity of females at 77kg (182cm). Ovchinnikov
(1970) reported that male swordfish "reach maturity"
in the Atlantic at about 100 cm (type of measure
unknown). However, he reported a smaller size-at-
maturity for females (70cm). DeMetrio et al. (1989)
reported that male swordfish in the eastern Mediter-
ranean Sea first begin to mature at 82- 105 cm LJFL
and that nearly all are mature when > 135cm. Females
NOTE Taylor and Murphy: Reproductive biology of Xiphias gladius in Straits of Florida
50
i 30
UJ
3
o
UJ 20
a. MALE
n = 117
El CLASS 4
n CLASS S
■ CLASS 6
D CLASS 7
n CLASS 8
H B ff 1 1
NS
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
COLLECTION MONTH
>
U
z
UJ
o
UJ
20-
b. FEMALE
n = 71
M CLASS 4
a CLASS 5
■ CLASS 6
D CLASS 7
n CLASS 8
^^m f" ■"! ^^B
I
inj
Fl.H.^.H
NS
1 I T
JUL AUG SEP OCT NOV DEC
JAN FEB MAR APR MAY JUN
COLLECTION MONTH
Figure 3
Monthly frequencies of gonad development classes for mature (> Class 4,
see "Methods and materials") swordfish Xiphias gladius collected off
southeast Florida during the period June 1977-November 1980. NS indicates
no samples taken.
1200^
3
1
2
1000
1
»
?
J 800-
UJ
H
UJ
S 600-
5
1 .
\
)
>
1
1
>
UJ
^ 400-
O
O
O
200
^
\
\
>
■
NS
Jf
kN FEB MAR APR
MAY JUN JUL AUG
COLLECTION MONTH
SEP
OCT
NOV
DEC
in the Mediterranean first reach matur-
ity at 106-135cm LJFL, and about 50%
of the females > 135 cm are mature. This
is in contrast to our findings that the
smallest mature females off Florida were
~170cm LJFL and that 50% of females
reached maturity when ~180cm.
Female swordfish from the Pacific
Ocean mature at sizes comparable to or
somewhat smaller than those in the west-
ern Atlantic. From 1949 to 1958, Yabe
et al. (1959) examined 372 female sword-
fish taken by a Japanese longline fleet in
the North Pacific. They found only five
mature females and suggested that
females in the North Pacific begin to
mature at 150-170cm EFL (170-192
LJFL). Kume and Joseph (1969) con-
cluded that female swordfish off Califor-
nia begin to mature at slightly smaller
sizes, 139cm EFL (158cm LJFL), and
are regularly found in ripe condition at
>170cm (192cm LJFL).
The ages-at-maturity we estimated for
male (1.4yr) and female (5.5yr) swordfish
off southeast Florida are somewhat
younger and older, respectively, than
those determined for male and female
swordfish collected farther north in the
U.S. South Atlantic. From a relatively
small sample collected off North Carolina
and South Carolina, Wilson (1984) deter-
mined that males mature between ages
2 and 3 (15 mature males of 24 exam-
ined). Females mature between ages 4
and 5 (2 mature females of 18 examined).
Although Wilson (1984) relied on a dif-
ferent technique (thin-sectioned otoliths)
than did our study (thin-sectioned spines)
to determine swordfish ages, he com-
pared the two aging techniques and found
they provided similar swordfish ages
through at least age 5.
Female swordfish mature at a younger
age in the eastern Mediterranean Sea
than in the western Atlantic or Pacific
Oceans. In the eastern Mediterranean,
Figure 4
Monthly mean and range (vertical bars) of individual
mean oocyte diameters for all mature (Class >4) sword-
fish Xiphias gladius collected off southeast Florida dur-
ing the period June 1977-November 1980. Numbers
indicate sample size; NS indicates no samples taken.
814
Fishery Bulletin 90(4). 1992
Table 4
Estimates of batch fecundity for seven swordfish Xiphias gladitis sampled off southeast
Florida, including lower jaw to fork length, age (Berkeley and Houde 1984), whole weight,
gonad development class, and oocyte diameters.
Batch
fecundity
(millions)
Lower jaw
to fork length
(mm)
Age
(yr)
Whole
weight
(kg)
Gonad
class
Hydrated oocyte size
N X range
1.398
256
6
107
6
1754
0.98
0.83-1.16
2.814
252
9
223
6
1288
1.04
0.75-1.14
2.836
207
8
116
6
1656
1.56
1.30-1.77
3.071
177
—
69
6
2036
1.55
1.33-1.73
3.125
233
—
170
6
2121
1.35
1.17-1.57
4.220
281
—
268
6
2912
1.54
1.34-1.68
4.220
256
9
210
6
1232
1.11
0.96-1.23
DeMetrio et al. (1989) found mature female swordfish
as young as age 2, with most mature by age 3. How-
ever, Yabe et al. (1959) suggested that female sword-
fish in the North Pacific mature at age 5 or 6. This is
in agreement with our age-at-50%-maturity for females
(5.5 years). Although we found no information on
maturation of male swordfish from the Pacific, most
male swordfish in the eastern Mediterranean reach
maturity by age 2 (DeMetrio et al. 1989), again similar
to our findings off southeast Florida.
The observed protracted spawning of swordfish off
southeast Florida, with peak activity during April
through July, agrees with reported spawning seasons
determined from temporal changes in the abundance
of larvae and juveniles. Taning (1955) reported that
spawning off Florida and elsewhere in the North Atlan-
tic occurs throughout the year, with peak larval abun-
dances during February through April. The temporal
distribution of larval abundance in the western North
Atlantic suggests that swordfish spavra during Decem-
ber through September, with a peak in April (Arata
1954, Markle 1974, Grail et al. 1983). Off the coast of
southern California, gonads of female swordfish are
inactive from late-August through mid-November
(Weber and Goldberg 1986). This period of inactivity
coincides with low mean oocyte diameters of sword-
fish in the western Atlantic (Fig. 4), suggesting that
swordfish in the eastern Pacific and western Atlantic
may have similar annual spawning seasons.
Two histological features present in ovarian tissue—
hydrated oocytes and postovulatory follicles— provided
evidence that female swordfish spawn off the coast of
southeast Florida between the Florida Keys and Cape
Canaveral. The presence of hydrated oocytes indicates
an imminent spawn, certainly within 12h. Young post-
ovulatory follicles found in many female swordfish pro-
vide evidence that spawning occurred within 24 h of
capture. Further, short-term stud-
ies of the movement of swordfish,
using acoustic tags, have shown
that fish that were likely mature
(70- 140 kg) remained in the same
general area (<90km from tag-
ging) for up to 5d during the
peak of the spawning season
(April) off Baja California (Carey
and Robison 1981). More exten-
sive movement occurred for a
70kg swordfish tracked for 2.5d
in November in the vicinity of
Cape Hatteras, North Carolina.
This fish traveled 240km in 67 h,
heading from cold continental
shelf waters off Cape Hatteras to
warmer Sargasso Sea waters.
The combination of this information on net random
movement of swordfish with our observed histological
evidence for recent or imminent spawning suggests
that spawning occurs in our sampling area. Addition-
ally, Grail et al. (1983) found major concentrations of
swordfish larvae and juveniles in the western Atlantic
in the waters near the Lesser Antilles, in the Yucatan
Straits, and in the Florida Straits, implying the pres-
ence of a large spawning population in these areas.
Changes in sex ratios as swordfish increase in size
apparently result from differences in grow^th between
the sexes and possibly from seasonal differences in
their distribution. Swordfish larger than ~230cm are
female because males have shorter life spans and
slower growth rates (Berkeley and Houde 1984, Wilson
1984). The dominance of males in our summer collec-
tions may be explained by seasonal differences in the
distribution of the sexes. Guitart-Manday (1964) found
a similar preponderance of males (72%) in samples from
the mainly summertime commercial fishery off Cuba.
Beckett (1974) reported that few males were taken in
the northern swordfish fisheries in waters <18°C,
whereas in more tropical latitudes, males account for
67-100% of the catch. Off southeast Florida, surface-
water temperatures remain >18°C throughout the
year, ranging from ~22°C in February to 29°C in
August (Atkinson et al. 1983). Our observed seasonal
changes in sex ratios of swordfish collected off
southeast Florida imply that whereas some males and
females are year-round inhabitants of these waters,
more females than males move north during the sum-
mer, which results in a dichotomous distribution that
becomes more acute the farther north the fish migrate.
At the northern extent of the range in the western
North Atlantic, off New England and the Canadian
maritimes, most, if not all, fish captured are females
(Lee 1942, Tibbo et al. 1961).
NOTE Taylor and Murphy: Reproductive biology of Xiphias gisdius in Straits of Florida
815
Our estimates of batch fecundity (1.4-4.2 million ova)
are comparable to the estimate of Yabe et al. (1959)
and to estimates for smaller swordfish by Uchiyama
and Shomura (1974). Our estimates are somewhat less
than those made for larger fish collected by the latter
authors. Yabe et al. (1959) estimated the fecundity of
a swordfish 186 cm TRNKL (276 cm LJFL from Table
1) to be 3-4 million ova (1.2-1. 6mm in diameter).
Uchiyama and Shomura (1974) estimated the fecundity
of an 80kg swordfish (185cm LJFL from Table 1) to
be 3.0 million ova, which closely agrees with our deter-
minations. However, they estimated that a 200 kg
swordfish (244 cm LJFL from Table 1) would have a
fecundity of ~6 million ova, which is more than our
estimates for swordfish over 200 kg.
Our estimates of fecundity and maturity schedules
can be used in analyzing the effect of fishing on the
spawning-stock biomass or spawning potential of
swordfish in the U.S. South Atlantic (e.g., spawning-
stock biomass per recruit analysis, Gabriel et al. 1989).
Recent assessments of the status of swordfish in the
Atlantic have utilized several techniques, including
dynamic pool models ("yield per recruit"), to determine
the effects that fishing and age at entry to the fishery
have on yield (ICCAT 1991). Our quantitative estimates
of the female maturation process and fecundity can be
used in further analyzing the effect of fishing on the
abundance of mature swordfish (and by Implication the
production of new recruits).
Acknowledgments
Credit for collecting the samples and data for this
research is given to E. Houde, S. Berkeley, E. Irby,
J. Jolley, and D. Nickerson. Data collected for this
study were funded in part by a grant/cooperative
agreement from the National Oceanic and Atmospheric
Administration through the National Sea Grant Col-
lege, Grant NA80AA-D-00038. The views expressed
herein are those of the authors and do not necessarily
reflect the views of NOAA or any of its subagencies.
We thank L. French, J. Leiby, M. Myers, and anony-
mous reviewers for their editorial assistance.
Citations
Arata, G.F. Jr.
1954 A contribution to the life history of the swordfish, Xiphias
gladius Linnaeus, from the South Atlantic coast of the United
States and the Gulf of Mexico. Bull. Mar. Sci. Gulf Caribb.
4:183-243.
Atkinson, L.P., T.N. Lee, J.O. Blanton, and W.S. Chandler
1983 Climatology of the southeastern United States Continen-
tal Shelf waters. J. Geophys. Res. 88:4705-4718.
Beckett, J.S.
1974 Biology of swordfish. Xiphias gladius L., in the North-
west Atlantic Ocean. In Shomura, R.S., and F. Williams
(eds.), Proc, Int. billfish symp.; Part 2, Review and contributed
papers, p. 105-106. NOAA Tech. Rep. NMFS SSRF 675.
Berkeley, S.A., and E.D. Houde
1980 Swordfish dynamics in the Straits of Florida. Int. Counc.
Explor. Sea CM 1980/H:59 Pelagic Fish Committee, 11 p.
1984 Age determination of broadbill swordfish, Xiphias
gladius. from the Straits of Florida, using anal fin spine sec-
tions. In Prince, E.D., and L.M. Pulos (eds.). Proc. Interna-
tional workshop on age determination of oceanic pelagic
fishes: Tunas, billfishes, and sharks, p. 137-143. NOAA Tech
Rep. NMFS 8.
Berkeley. S.A., and E.W. Irby
1982 Analysis of catch, effort and landing data in the sword-
fish longline fishery of the Florida Straits. Int. Counc. Explor.
Sea CM 1982/H:53, 12 p.
Brown-Peterson, N., P. Thomas, and C.R. Arnold
1988 Reproductive biology of the spotted seatrout, Cynoscion
nebulosus, in South Texas. Fish. Bull., U.S. 86:373-388.
Carey, F.G.. and B.H. Robison
1981 Daily patterns in the activities of swordfish, Xiphias
gladius, observed by acoustic telemetry. Fish. Bull., U.S.
79:277-292.
DeMartini, E.E., and R.K. Fountain
1981 Ovarian cycling frequency and batch fecundity in the
queenfish, Seriphus politus : Attributes representative of serial
spawning fishes. Fish. Bull., U.S. 79:547-555.
DeMetrio, G., P. Megalofonou, S. Tselas, and N. Tsimedides
1989 Fishery and biology of the swordfish Xiphias gladius L.,
1758 in Greek waters. In Savini, M., and J.F. Caddy (eds.),
Report of the second technical consultation on stock assess-
ment in the eastern Mediterranean, p. 135-145. FAO Fish.
Rep. 412.
Fitzhugh, G.R.. T.G. Snider III, and B.A. Thompson
1988 Measurement of ovarian development in red drum {Sciae-
nops ocellatus) from offshore stocks. Contrib. Mar. Sci.
(Suppl.) 30:79-83.
Gabriel, W.L., M.P. Sizenmore, and W.J. Overholtz
1989 Analysis of spawning stock biomass per recruit: An ex-
ample for Georges Bank haddock. N. Am. J. Fish. Manage.
9:383-391.
Grail, C, D.P. DeSylva, and E.D. Houde
1983 Distribution, relative abundance, and seasonality of
swordfish larvae. Trans. Am. Fish. Soc. 112:235-246.
Grier, H.J.
1981 Cellular organization of the testis and spermatogenesis
in fishes. Am. Zool. 21:345-357.
Guitart-Manday, D.
1964 Biologla pesquera del emperador o pez de espada, Xiphias
gladius Linnaeus (Teleostomi: Xiphiidae) en las aguas de
Cuba. Poeyana Inst. Biol. (Habana), Ser. B. 1, 37 p. [transl.
by Foreign Language Div., Canada].
Humason, G.L.
1972 Animal tissue techniques. W.H. Freeman, San Fran-
cisco, 641 p.
Hunter, J.R., and B.J. Macewicz
1985 Measurement of spawning frequency in multiple spawn-
ing fishes. In Lasker, R. (ed.). An egg production method for
estimating spawning biomass of pelagic fish: Application to
the northern anchovy, p. 79-94. NOAA Tech. Rep. NMFS 36.
Fishery Bulletin 90(4), 1992
ICCAT (International Commission for the Conservation of
Atlantic Tunas)
1991 Reference paper on 1991 swordfish stock assessments
by SCRS swordfish assessment group. SCRS/91/16, Int.
Comm. Conserv. Atl. Tuna, Madrid, 19 p.
Kume, S., and J. Joseph
1969 Size composition and sexual maturity of billfish caught
by the Japanese longHne fishery in the Pacific Ocean east of
130°W. Bull. Far Seas Fish. Res. Lab. (Shimizu) 2:115-162
[in Engl., Japanese summ.].
Lee, R.E.
1942 The occurrence of female swordfish in southern New
England waters, with a description of their reproductive con-
dition. Copeia 2:117-119.
Markle, G.E.
1974 Distribution of larval swordfish in the Northwest Atlan-
tic Ocean. In Shomura, R.S., and F. Williams (eds.), Proc,
Int. billfish symp.; Part 2. Review and contributed papers, p.
252-260. NOAA Tech. Rep. NMFS SSRF 675.
Murphy, M.D., and R.G. Taylor
1990 Reproduction, growth, and mortality of red drum Sciae-
rwps ocellatus in Florida waters. Fish. Bull., U.S. 88:531-542.
Nakamura, I.
1985 Billfishes of the world. FAG Fish. Sj-nop. 125(5), 58 p.
Ovchinnikov, V.V.
1970 Mech-ryba i parusnikovye (Atlanticheskii okean. Ekologia
i funktsional naya morfologiya). (Swordfish and billfishes of
the Atlantic Ocean. Ecology and functional mor-
phology.) Nauch-Issled. Inst. Ryb. Knoz. Okeanogr., Kalin-
ingrad, 106 p. [Transl. by Israel Prog. Sci. Transl., 77 p.; avail.
NTIS, Springfield, VA, TT71-50011.]
Palko, B.J., G.L. Beardsley, and W.J. Richards
1981 Synopsis of the biology of the swordfish, Xiphias gladius
Linnaeus. NOAA Tech. Rep. NMFS Circ. 441, 21 p.
Rohlf, F.J., and R.R. Sokal
1981 Statistical tables, 2d ed. W.H. Freeman, NY. 219 p.
Saila. S.B., C.W. Reckseik. and M.H. Prager
1988 Basic fishery science programs, a compendium of micro-
computer programs and manual of operation. Elsevier, NY,
230 p.
SAS
1982 SAS user's guide: Basics, 1982 ed. SAS Inst., Inc., Gary
NC, 923 p.
Sokal, R.R., and F.J. Rohlf
1981 Biometry, 2d ed. W.H. Freeman, NY, 859 p.
Takita, T., T. Iwamoto, S. Kai, and I. Sogabe
1983 Maturation and spawning of the dragonet, Callionymus
ennfactis. in an aquarium. Jpn. J. Ichthyol. 30(3):221-226.
Taning, A.V.
1955 On the breeding areas of the swordfish (Xip/iios). Pap.
Mar. Biol. Oceanogr. Deep Sea Res. Suppl. to Vol. 3:3248-3251.
Tibbo, S.N., L.R. Day, and W.F. Doucet
1961 The swordfish {Xiphias gladius L.), its life-history and
economic importance in the northwest Atlantic. Fish. Res.
Board Can. Bull. 130:1-47.
Tibbo, S.N., and L.M. Lauzier
1969 Larval swordfish (Xiphias gladius) from three local-
ities in the western Atlantic. J. Fish. Res. Board Can. 26:
3248-3251,
Uchiyama, J.H., and R.S. Shomura
1974 Maturation and fecundity of swordfish from Hawaiian
waters. In Shomura, R.S., and F. Williams (eds.), Proc, Int.
billfish symp.; Part 2, Review and contributed papers, p.
142-148. NOAA Tech. Rep. NMFS SSRF 675.
Wallace, R.A., and K. Selman
1981 Cellular and dynamic aspects of oocyte growth in tele-
osts. Am, Zool, 21:325-343.
Weber, E.C., and S.R. Goldberg
1986 The sex ratio and gonad indices of swordfish, Xiphias
gladius, caught off the coast of southern California in 1978.
Fish. Bull., U.S. 84:185-186.
Wilson, C.A. Ill
1984 Age and growth aspects of the life history of billfishes.
Ph.D. diss,, Univ, South Carolina, Columbia, 180 p,
Yabe, H.S., S. Ueyanagi, S. Kikawa, and H. Watanabe
1959 Studies of the life history of the swordfish, Xiphias
gladius Linnaeus. Rep. Nankai Reg. Fish. Res. Lab. 10:
107-150 [in Jpn., Engl, summ.; transl. by CEP Multilingual
Serv. Div., Dep. Sec. of State, Canada].
Fishery Bulletin Index
Volume 90 (1-4), 1992
List of Titles
90(1)
1 Estimating trends in abundance of dolphins associated with
tuna in the eastern tropical Pacific Ocean, using sightings data
collected on commercial tuna vessels, by Stephen T. Buckland,
Karen L. Cattanach, and Alejandro A. Anganuzzi
13 Morphology, systematics, and biology of the double-lined
mackerels (Grammatorcynus, Scombridae), by Bruce B.
Collette and Gary B. Gillis
54 Geographic variation in cranial morphology of spinner dolphins
Stenella longirostris in the eastern tropical Pacific Ocean, by
Michael E. Douglas, Gary D. Schnell, Daniel J. Hough, and
William F. Perrin
197 Association between the sessile barnacle Xenobalanus globi-
cipitis (Coronulidae) and the bottlenose dolphin Tursiops trun-
catus (Delphinidae) from the Bay of Bengal, India, with a sum-
mary of previous records from cetaceans, by Arjuna Rajaguru
and Gopalsamy Shantha
203 Lack of biochemical genetic and morphometric evidence for
discrete stocks of Northwest Atlantic herring Clupea harengus
harengus, by Susan E. Safford and Henry Booke
211 Variability of monthly catches of anchovy Engraulis encrasi-
colus in the Aegean Sea, by Konstantinos I. Stergiou
90(2)
217 Age, growth, and reproduction of the goosefish Lophius
amencanus (Pisces: Lophiiformes), by Michael P. Armstrong,
John A. Musick, and James A. Colvocoresses
231 Annual reproductive cycle of oocytes and embryos of yellowtail
rockfish Sebastes flavidus (Family Scorpaenidae), by Michael
J. Bowers
77 Geographic variation in population genetic structure of chinook
salmon from California and Oregon, by Devin Bartley, Boyd
Bentley, Jon Brodziak, Richard Gomulkiewicz, Marc Mangel,
and Graham A.E. Gall (authorship amended per errata, Fish.
Bull. 90(3):iii)
101 Fecundity, spawning, and maturity of female dover sole Micro-
stnmus pacificus. with an evaluation of assumptions and preci-
sion, by J. Roe Hunter, Beverly J. Macewicz, N. Chyan-huei
Lo, and Carol A. Kimbrell
129 Comparisons of early-life-history characteristics of walleye
pollock Theragra chalcogravima in Shelikof Strait, Gulf of
Alaska, and Funka Bay, Hokkaido, Japan, by Arthur W.
Kendall Jr. and Toshikuni Nakatani
139 Exploitation models and catch statistics on the Victorian fishery
for abalone Haliotis rubra, by Paul E. McShane
147 ITQs in New Zealand: The era of fixed quota in perpetuity,
by Michael P. Sissenwine and Pamela M. Mace
161 Seasonality in reproductive activity and larval abundance of
queen conch Strombiis gigas, by Allan W. Stoner, Veronique
J. Sandt, and Isabelle F. Boidron-Metairon
171 Predicting effects of dredging on a crab population: An
equivalent adult loss approach, by Thomas C. Wainwright,
David A. Armstrong, Paul A. Dinnel, Jos6 M. Orensanz, and
Katherine A. McGraw
183 Comparison of feeding and growth of larval round herring
Etrumeus teres and gulf menhaden Brevoortia patronus, by
Weihzong Chen, John J. Govoni, and Stanley M. Warlen
190 Analytical correction for oversampled Atlantic mackerel
Scomber scombrus eggs collected with oblique plankton tows,
by Denis D'Amours and Francois Gregoire
243 Age, growth, and reproduction of jewfish Epinephehis itajara
in the eastern Gulf of Mexico, by Lewis H. Bullock, Michael
D. Murphy, Mark F. Godcharles, and Michael E. Mitchell
250 Genetic patchiness among populations of queen conch Strom-
bus gigas in the Florida Keys and Bimini, by Donald E. Camp-
ton, Carl J. Berg Jr., Lynn M. Robison, and Robert A. Glazer
260 Detecting environmental covariates of Pacific whiting Merluc-
cius productiis growth using a growth-increment regression
model, by Martin W. Dorn
276 Influence of sectioning otoHths on marginal increment trends
and age and growth estimates for the flathead Platycephalus
speculator, by Glenn A. HjTides, Neil R. Loneragan, and Ian
C. Potter
285 Metamorphosis and an overview of early-life-history stages in
Dover sole Microstomus pacificus, by Douglas F. Markle, Phillip
M. Harris, and Christopher L. Toole
302 Estimating stock abundance from size data, by Michael L.
Parrack
328 Biology of two co-occurring tonguefishes, Cynoghssus arel and
C. lida (Pleuronectiformes:Cynoglossidae), by Arjuna Rajaguru
368 Inverse method for mortality and growth estimation: A new
method for larval fishes, by David A. Somerton and Donald
R. Kobayashi
376 Seasonal distribution of river herring Alosa pseiidaharengus
and A. aestivalis off the Atlantic coast of Nova Scotia, by Heath
H. Stone and Brian M. Jessop
390 Long-term coded wire tag retention in juvenile Seiaenops
ocellatus, by Britt W. Bumguardner, Robert L. Colura, and
Gary C. Matlock
817
818
INDEX: TITLES Fishery Bulletin 90(1-4), 1992
395 Growth and mortality of Lutjanus vittus (Quoy and Gaimard)
from the North West Shelf of Australia, by Tim L.O. Davis
and Grant J. West
405 Correlation of winter temperature and landings of pink shrimp
Penaevs duorarum in North Carolina, by William F. Hettler
407 Growth of five fishes in Texas Bays in the 1960s, by Gary C.
Matlock
552 Management advice from a simple dynamic pool model, by
Grant G. Thompson
561 A Bayesian approach to management advice when stock-
recruitment parameters are uncertain, by Grant G. Thompson
574 Eastern Pacific species of the genus Umbrina (Pisces: Sciae-
nidae) with a description of a new species, by H.J. Walker Jr.
and Keith W. Radford
412 A mortality model for a population in which harvested indi-
viduals do not necessarily die: The stone crab, by Victor R.
Restrepo
417 Optimal course by dolphins for detection avoidance, by Carlos
A.M. Salvad6, Pierre Kleiber, and Andrew E. Dizon
421 Effects of microprobe precision on hypotheses related to otolith
Sr:Ca ratios, by Christopher L. Toole and Roger L. Nielsen
90(3)
429 Population characteristics of individually identified humpback
whales in southeastern Alaska: Summer and fall 1986, by
C. Scott Baker. Janice M. Straley, and Anjanette Perry
439 Precision of recruitment predictions from early life stages of
marine fishes, by Michael J. Bradford
454 Post-yolksac larval development of two southern California
sculpins, Clinocottus analis and Orthonopias triads (Pisces:
Cottidae), by Richard F. Feeney
469 A genetic analysis of weakfish Cynoscion regalis stock struc-
ture along the mid-Atlantic coast, by John E. Graves, Jan R.
McDowell, and M. Lisa Jones
476 An assessment of the exploitable biomass of Heterocarpus
laevigatus in the main Hawaiian Islands. Part 2: Observations
from a submersible, by Robert B. Moffitt and Frank A. Parrish
483 Variability in spiny lobster Panulirus margirmtus recruitment
and sea level in the Northwestern Hawaiian Islands, by Jeffrey
J. Polovina and Gary T. Mitchum
494 An assessment of the exploitable biomass of Heterocarpus
laevigatus in the main Hawaiian Islands. Part 1; Trapping
surveys, depletion experiment, and length structure, by
Stephen Ralston and Darryl T. Tagami
505 Interannual variation and overlap in the diets of pelagic juvenile
rockfi.sh (Genus: Sebastes) off central California, by Carol A.
Reilly, Tina Wyllie Echeverria, and Stephen Ralston
516 Age and growth of red hind Epinephelus guttatus in Puerto
Rico and St. Thomas, by Yvonne Sadovy, Miguel Figuerola,
and Ana Roman
529 Early life history of the tautog Tautoga onitis in the Mid-
Atlantic Bight, by Susan M. Sogard, Kenneth W. Able, and
Michael P. Fahay
540 Fish-habitat associations on a deep reef at the edge of the
Oregon continental shelf, by David L. Stein, Brian N. Tissot,
Mark A. Hixon, and William Barss
599
Behavioral reactions of humpback whales Megaptera novae-
angliae to biopsy procedures, by Mason T. Weinrich, Richard
H. Lambertson, Cynthia R. Belt, Mark R. SchilHng, Heidi J.
Iken, and Stephen E. Syrjala
Spatial and temporal distribution of juvenile Atlantic cod Gadv^
morhua in the Georges Bank-Southern New England region,
by Susan E. Wigley and Fredric M. Serchuk
607 Larval development of two sympatric flounders, Paralichthys
adspersus (Steindachner, 1867) and Paralichthys microps
(Gunther, 1881), from the Bay of Coquimbo, Chile, by Humberto
N. Zuniga and Enzo S. Acufia
621 An estimate of the tag- reporting rate of commercial shrimpers
in two Texas bays, by R. Page Campbell, Terry J. Cody. C.E.
Bryan, Gary C. Matlock, Maury F. Osborn, and Albert W.
Green
625 Power to detect linear trends in dolphin abundance: Estimates
from tuna-vessel observer data, 1975-89, by Elizabeth F.
Edwards and Peter C. Perkins
90(4)
633
642
659
668
678
Abundance, distribution, and settlement of young-of-the-year
white seabass Atractosion nobilis in the Southern California
Bight, 1988-89, by Larry G. Allen and Michael P. Franklin
Depth, capture time, and hooked longevity of longline-caught
pelagic fish: Timing bites offish with chips, by Christofer H.
Reproduction in American lobsters Homarus americamis
transplanted northward to St. Michael's Bay, Labrador, by
Frank A. Boothroyd and Gerald P. Ennis
Larval development, distribution, and ecology of cobia /;ac%-
centron canadum (Family: Rachycentridae) in the northern Gulf
of Mexico, by James G. Ditty and Richard F. Shaw
Energetics of associated tunas and dolphins in the eastern
tropical Pacific Ocean: A basis for the bond, by Elizabeth F.
Edwards
691 Artificial shelters and survival of juvenile Caribbean spiny
lobster Panulirus argus: Spatial, habitat, and lobster size
effects, by David B. Eggleston, Romuald N. Lipcius, and David
L. Miller
703 Stock structure of the bluefish Pc/matomus saltatrix along the
mid-Atlantic coast, by John E. Graves. Jan R. McDowell, Ana
M. Beardsley, and Daniel R. Scoles
INDEX: TITLES Fishery Bulletin 90(1 -4). 1992
il9
711 Age validation, growth, and mortality of larval Atlantic bumper
(Carangidae: Chloroscombru^ cht-ysunt^) in the northern Gulf
of Mexico, by Deborah L. Leffler and Richard F. Shaw
720 Evaluation of ghost fishing in the Hawaiian lobster fishery,
by Frank A. Parrish and Thomas K. Kazama
726 Movements of acoustically-tagged yellowtail rockfish Sebastes
flavidus on Heceta Bank, Oregon, by William G. Pearcy
736 A simple simulation approach to risk and cost analysis, with
applications to swordfish and cod fisheries, by Victor R.
Restrepo, John M. Hoenig, Joseph E. Powers, James W. Baird,
and Stephen C. Turner
749 Behavior of individually-identified sei whales Balaenoptera
horealis during an episodic influx into the southern Gulf of
Maine in 1986, by Mark R. Schilling, Irene Seipt, Mason T.
Weinrich, Steven E. Frohock, Anne E. Kuhlberg, and Phillip
J. Clapham
756 Population dynamics of pelagic armorhead Psetidopentaceros
wheeleri on Southeast Hancock Seamount, by David A. Somer-
ton and Bert S. Kikkawa
778 DifferentiatingParaKiAodes larvae using telson spines: A tail
of two species, by Gregory C. Jensen, Helle B. Andersen, and
David A. Armstrong
784 A telemetric study of the home ranges and homing routes of
lingcod Ophiodon elongatus on shallow rocky reefs off Van-
couver Island, British Columbia, by Kathleen R. Matthews
791 An investigation of bottlenose dolphin Tiirsiops truncatiis
deaths in East Matagorda Bay, Texas, January 1990, by
W. George Miller
798 Application of otolith microchemistry analysis to investigate
anadromy in Chesapeake Bay striped bass Morone saxatilis,
by David H. Secor
807 Leatherback turtle captured by ingestion of squid bait on
swordfish longline, by Robert A. Skillman and George H. Balazs
809 Reproductive biology of the swordfish Xiphias gladiiis in the
Straits of Florida and adjacent waters, by Ronald G. Taylor
and Michael D. Murphy
770 Genetic isolation of previously indistinguishable chinook salmon
populations of the Snake and Klamath Rivers: Limitations of
negative data, by Fred M. Utter, Robin S. Waples, and David
J. Teel
Fishery Bulletin Index
Volume 90 (1-4), 1992
List of Authors
Able. Kenneth W. 529
Acuna, Enzo S. 607
Allen, Larry G. 633
Andersen, Helle B. 778
Anganuzzi, Alejandro A. 1
Armstrong, David A. 171, 778
Armstrong, Michael P. 217
Baird, James W. 736
Baker, C. Scott 429
Balazs, George H. 807
Barss, William 540
Bartley, Devin 77
Beardsley, Ana M. 703
Belt, Cynthia R. 588
Bentley, Boyd 77
Berg, Carl J. Jr. 250
Boggs, Christofer H. 642
Boidron-Metairon, Isabelle F. 161
Booke, Henry 203
Boothroyd, Frank A. 659
Bowers, Michael J. 231
Bradford, Michael J. 439
Brodziak, Jon 77
Bryan, C.E. 621
Buckland, Stephen T, 1
Bullock, Lewis H, 243
Bumguardner, Britt W. 390
Campbell, R. Page 621
Campton, Donald E. 250
Cattanach, Karen L. 1
Chen, Weihzong 183
Chyan-huei Lo, N. 101
Clapham, Phillip J. 749
Cody, Terry J. 621
Collette, Bruce B. 13
Colura, Robert L. 390
Colvocoresses, James A. 217
D'Amours, Denis 190
Davis, Tim L.O. 395
Dinnel, Paul A. 171
Ditty, James G. 668
Dizon, Andrew E. 417
Dorn, Martin W. 260
Douglas, Michael E, 54
Edwards, Elizabeth F. 625, 678
Eggleston, David B. 691
Ennis, Gerald P. 659
Fahay, Michael P. 529
Feeney, Richard F. 454
Figuerola, Miguel 516
Franklin, Michael P. 633
Frohock. Steven E. 749
Gall, Graham A.E. 77
Gillis, Gary B. 13
Glazer, Robert A. 250
Godcharles, Mark F. 243
Gomulkiewicz, Richard 77
Govoni, John J. 183
Graves, John E. 469, 703
Green, Albert W. 621
Gregoire, Francois 190
Harris, Phillip M. 285
Hettler, William F. 405
Hixon, Mark A. 540
Hoenig, John M. 736
Hough, Daniel J. 54
Hunter, J. Roe 101
Hyndes, Glenn A. 276
Iken, Heidi J. 588
Jensen, Gregory C. 778
Jessop, Brian M. 376
Jones, M. Lisa 469
Kazama, Thomas K. 720
Kendall, Arthur W. Jr. 129
Kikkawa, Bert S. 756
Kimbrell, Carol A. 101
Kleiber, Pierre 417
Kobayashi, Donald R. 368
Kuhlberg, Anne E. 749
Lambertson, Richard H. 588
Leffler, Deborah L. 711
Lipcius, Romuald N. 691
Loneragan, Neil R. 276
Mace. Pamela M. 147
Macewicz. Beverly J. 101
Mangel. Mark 77
Markle. Douglas F. 285
Matlock, Gary C. 390, 407, 621
Matthews, Kathleen R. 784
McDowell, Jan R. 469, 703
McGraw, Katherine A. 171
McShane. Paul E. 139
Miller, David L. 691
Miller. W. George 791
Mitchell, Michael E. 243
Mitchum, Gary T. 483
Moffitt, Robert B. 476
Murphy. Michael D. 243, 809
Musick, John A. 217
Nakatani, Toshikuni 129
Nielsen, Roger L. 421
Orensanz, Jose M. 171
Osborn. Maury F. 621
Parrack. Michael L. 302
Parrish, Frank A. 476, 720
Pearcy, William G. 726
Perkins, Peter C. 625
Perrin, William F. 54
Perry, Anjanette 429
Polovina, Jeffrey J. 483
Potter. Ian C. 276
Powers, Joseph E. 736
Radford, Keith W. 574
Rajaguru, Arjuna 97, 328
Ralston, Stephen 494, 505
Reilly, Carol A. 505
Restrepo, Victor R. 412, 736
Robison, Lynn M. 250
Roman, Ana 516
Sadovy, Yvonne 516
Safford, Susan E. 203
Salvado. Carlos A.M. 417
Sandt. Veronique J. 161
Schilling. Mark R. 588, 749
Schnell, Gary D. 54
Scoles, Daniel R. 703
Secor, David H. 798
Seipt, Irene 749
Serchuk, Fredric M. 599
Shantha, Gopalsamy 197
Shaw, Richard F. 668, 711
Sissenwine, Michael P. 147
Skillman, Robert A. 807
Sogard. Susan M. 529
Somerton. David A. 368, 756
Stein, David L. 540
Stergiou, Konstantinos L 211
Stone, Heath H. 376
Stoner, Allan W. 161
Straley. Janice M. 429
Syrjala, Stephen E. 588
Tagami, Darryl T. 494
Taylor, Ronald G. 809
Teel. David J. 770
Thompson, Grant G. 552, 561
Tissot, Brian N. 540
Toole, Christopher L. 285, 421
Turner. Stephen C. 736
Utter, Fred M.
(70
171
Wainwright. Thomas C.
Walker. H.J. Jr. 574
Waples. Robin S. 770
Warlen, Stanley M. 183
Weinrich. Mason T. 588, 749
West, Grant J. 395
Wigley, Susan E. 599
Wyllie Echeverria, Tina 505
Zufiiga, Humberto N. 607
820
Fishery Bulletin Index
Volume 90 (1-4), 1992
List of Subjects
Abalone 139
Abundance
dolphin, eastern tropical Pacific 625
eggs
mackerel, Atlantic 190
fish, Oregon continental shelf 540
herring, river 376
larvae
conch, queen 161
seabass, white, young-of-the-year 633
whale, humpback 429
Abundance estimates— see also Population
studies
dolphins, eastern tropical Pacific 1
fisheries stocks 302
shrimp, deepwater 476, 494
Aegean Sea fishery
anchovy 211
Age determination
bumper, Atlantic 711
otoliths
flathead 276
hind, red 516
jewfish 243
tautog, juvenile 529
urohyal bones
Lutjanus vittus 395
vertebrae, goosefish 217
Age-size estimation
jewfish 243
Lutjanus vittus 395
tautog, juvenile 529
tonguefish 328
whiting. Pacific 260
Age validation
marginal increment analysis
hind, red 516
oxytetracycline
hind, red 516
Alaska, southeastern
whale, humpback 429
Alewife— see Herring, river
Alosa
aestivalis— see Herring, river
pseudoharengus—see Herring, river
Anchovy 211, 439
Archosargus probatocephalus—see
Sheepshead
Armorhead, pelagic 756
Artificial reefs
lobster, spiny 691
Artificial shelters
lobster, spiny 691
Atlantic Ocean
bluefish 703
Atlantic Ocean, northwest
goosefish 217
Atractoscion nobilis—see Seabass. white
Australian fisheries
abalone 139
Lutjanus vittus 395
Balaenoptera borealis—see Whale, sei
Barnacle
associations with cetaceans 197
Bass, striped 798
Bayesian statistics 561
Behavior
shrimp, deepwater 476
tuna-dolphin associations 678
whale, humpback 588
whale, sei 749
Bimini
conch, queen 250
Bluefish 703
Brevoortia patronus—see Menhaden, gulf
California Bight, southern
seabass. white 633
California, central
rockfish, juvenile 505
California, southern
sculpin larvae 454
Canada, Labrador
lobster, American, transplant 659
Cancer magister—see Crab, Dungeness
Caribbean
conch, queen 161
hind, red 516
grouper 516
Catch estimation— see also Population
studies
abalone 139
anchovy 211
herring, river 276
temperature and diel effects
herring, river 376
water temperature effects
shrimp, pink 405
Catch-per-unit-effort
longline, Hawaii 642
Catch rates
shrimp, deepwater 494
Cetaceans
barnacle associations 197
Chile
flounder 607
Paralichthys
adspersus 607
microps 607
Chloroscombrus chrysurus—see Bumper,
Atlantic
Classification— see Taxonomy
Clinocottus anaiis 454
Clupea harengus—see Herring
Clupea harengus harengus—see Herring,
northwest Atlantic
Cobia, early life history 668
Cod 439
Cod, Atlantic 599, 736
Conch, queen 161, 250
Crab
blue king 778
Dungeness 171
red king 778
stone 412
Croaker 574
Current, subtropical counter
lobster, spiny 483
Cynoglossus
arel—see Tonguefish
Hda— see Tonguefish
Cynoscion
nebulosus—see Seatrout, spotted
regalis—see Weakfish
Delphinus delphis—see Dolphin, common
Density
shrimp, deepwater 476
Depth
distribution, Atlantic cod 599
Dermochelys coriacea—see Turtle,
leatherback
Dietary overlap
rockfish. juveniles 505
Distribution
cobia, larvae and juveniles 668
cod, Atlantic 599
herring, river 376
seabass, white, young-of-the-year 633
tautog larvae 529
Dolphin 417
bottlenose 197, 791
common 1, 625
eastern tropical Pacific spp. 1, 625
spinner 1,54
Central American 54
eastern 54, 625
pantropical 54
whitebelly 625
spotted 1, 54, 625, 678
Dredging effects
crab, Dungeness 171
Drum
black 407
red 390, 407
Early-life-history studies
bumper, Atlantic 711
Eelpouts 540
Egg studies
mackerel, Atlantic 190
pollock, walleye 129
Eggs
conch, queen 161
821
822
INDEX: SUBJECTS Fishery Bulletin 90(1-4). 1992
Embryos— see also Larval studies
rockfish, yellowtail 231
Energetics
dolphin, spotted 678
tuna, yellowfin 678
Engraulis
encrasicolus—see Anchovy
mordax—see Anchovy
Epinephelus
guttatussee Hind, red
itajara—see Jewfish
Estuarine fishes, Texas 407
Etrumeus te7-es— see Herring, round
Fecundity
goosefish 217
pollock, walleye 129
sole, Dover 101
Feeding— see Food habits
Fish, marine 439
Fishery
abalone 139
crab, stone 412
jewfish 243
lobster, spiny 483
longline 642
shrimp 621
shrimp, deepwater 476, 494
shrimp, pink 405
trap, Hawaiian lobsters 720
Fishery interactions
tuna-dolphin 1, 625
Fishery management
abalone 139
models 552, 561
quotas, New Zealand 147
Flatfish 328
Florida Keys
conch, queen 250
Flounder 607
Flounder, southern 407
Food habits
dolphin, bottlenose 791
dolphin, spotted 678
interannual variation
rockfish, juveniles 505
juveniles, rockfish 505
bocaccio 505
chilipepper 505
shortbelly 505
widow 505
yellowtail 505
larvae
herring, round 183
menhaden, gulf 183
tonguefish 328
tuna, yellowfin 678
whale, humpback 429
Funka Bay
pollock, walleye 129
Gadus morhua—see Cod
—see Cod, Atlantic
Genetic studies
conch, queen 250
herring, northwest Atlantic 203
Genetic studies (continued)
salmon, chinook 77, 770
stock identification, weakfish 469
Geographic studies
salmon, chinook 77
Geographic variation
dolphin, spinner 54
Ghost fishing 720
Goosefish 217
Gravimatorcymis
bicarinatus—see Mackerel, shark
—see Mackerel, double-lined
biiitwatis—see Mackerel, double-lined
Grays Harbor
crab, Dungeness 171
Grouper 516
Growth rates— see also Age-size estimation
estuarine fishes, Texas 407
flathead 276
goosefish 217
hind, red 516
jewfish 243
Lutjanus vittus 395
shrimp, deepwater 494
tautog, juvenile 529
tonguefish 328
whiting. Pacific 260
Growth studies
bumper, Atlantic 711
larvae
herring, round 183
menhaden, gulf 183
model, larval fish 368
whiting. Pacific 260
Gulf of Alaska
pollock, walleye 129
Gulf of Maine
whale, sei 749
Gulf of Mexico
bumper, Atlantic 711
cobia, early life history 668
dolphin, bottlenose 791
Gulf of Mexico fishery
crab, stone 412
jewfish 243
Habitat
conch, queen 161
fish, Oregon continental shelf 540
Habitat studies
shrimp, deepwater 476
Hake, Pacific— see Whiting, Pacific
Haliotis rubra— see Abalone
Hawaiian Islands
dolphin 54
pelagic fishes 642
shrimp, deepwater 476, 494
Hawaiian Islands fishery
lobster, spiny 483
Hermaphroditism, protogynous
hind, red 516
Herring 439
blueback— see Herring, river
northwest Atlantic 203
river 376
round 183
Heterocarpus
laex'igatus 476
—also see Shrimp, deepwater
ensifer 476
Hind,' red 516
Homarus americanus—see Lobster,
American
Hydrodynamics
dolphin, spotted 678
tuna, yellowfin 678
Ichthyoplankton 190
Impact assessments
fishery, Dungeness crab 171
India fishery
tonguefish 328
Jewfish 243
Juvenile studies
cobia 668
cod, Atlantic 599
pollock, walleye 129
rockfish 505
bocaccio 505
chilipepper 505
shortbelly 505
widow 505
yellowtail 505
seabass, white 633
sole, Dover 285
Larvae
crab
blue king 778
red king 778
Larval studies
bumper, Atlantic 711
Clinocottus analis 454
cobia 668
conch, queen 161
herring, round 183
menhaden, gulf 183
model
growth 368
mortality 368
Orthonopias triads 454
pollock, walleye 129
rockfish, yellowtail 231
sculpins, marine 454
seabass, white 633
sole, Dover 285
Length frequency
shrimp, deepwater 494
Length-weight relationship
dolphin, bottlenose 791
tonguefish 328
Length studies— see Age-size estimation
Life history
sole, Dover 285
Life history, early
fishes, marine 439
Lingcod 784
Lobster
American 659
Hawaiian spiny 720
slipper 720
INDEX: SUBJECTS Fishery Bulletin 90( 1-4), 1992
823
Lobster (continued)
spiny 483, 691
Longline fishing
capture time 642
depth 642
hooked longevity 642
turtle capture 807
Lophius americanus—see Goosefish
Lutjanus vittus 395
Lycodes pacificus—see Eelpouts
Mackerel
Atlantic 190
double-lined 13
shark 13
Management— see also Fishery management
Mathematical methods
analytical correction
egg sampling bias 190
Bayesian statistics 561
simulations
Monte Carlo 736
spectral analysis, anchovy catch 211
Maximum sustainable yield
sole, rock 552, 561
Megaptera novaeangliae—see Whale,
humpback
Menhaden, gulf 183
Merliiccius productus—see Whiting, Pacific
Metamorphosis
sole, Dover 285
Methods
electron microprobe
sole, Dover 421
line transect, dolphin 1, 417
oblique plankton tows 190
photographic identification
whale, humpback 429
Microchemistry
sole, Dover 421
Microstomus pacificus—see Sole, Dover
Mid-Atlantic Bight
tautog 529
Migration— see Movements
Mitochondrial DNA
bluefish 703
Models
abundance
fisheries stocks 302
detection avoidance
dolphin 417
dynamic pool, rock sole 552, 561
egg-per-recruit, abalone 139
energetics, tuna-dolphin
associations 678
entrainment
crab. Dungeness 171
exploitation, abalone 139
growth rate
whiting. Pacific 260
larval fish
growth rate 368
Monte Carlo simulation 368
mortality rate 368
mortality rate
crab, stone 412
Models (continued)
recruitment
fishes, marine 439
yield-per-recruit, abalone 139
Morone saxatilis—see Bass, striped
Morphology
cobia larvae 668
cranial variations
dolphin, spinner 54
herring, northwest Atlantic 203
sole, Dover 285
Mortality
armorhead, pelagic 756
bumper, Atlantic 711
dolphin, bottlenose 791
larvae, marine fish 439
Mortality rates
crab, stone 412
fishes, marine 439
Lutjanus vittus 395
shrimp, deepwater 494
Movements
bass, striped 798
herring, river 376
larvae
conch, queen 161
lingcod 784
rockfish, yellovrtail 726
Mugil cephalus—see Mullet, striped
Mullet, striped 791
New Zealand fisheries, quotas 147
North Carolina fishery
shrimp, pink 405
Nova Scotia coast fishery
herring, river 376
Oncorhynchus tshawytscha—see Salmon,
Chinook
Ophiodon elongatussee Lingcod
Oregon continental shelf
fish-habitat associations 540
Orthonopias triads 454
Otoliths
ageing, flathead 276
bass, striped 798
growth
flathead 276
herring, round 183
jewfish 243
menhaden, gulf 183
SR:CA ratio
bass, striped 798
sole, Dover 421
Pacific Ocean
armorhead, pelagic 756
dolphin, spotted 678
tuna, yellowfin 678
Pacific Ocean, eastern tropical
dolphin 625
dolphin abundance 1
dolphin, spinner 54
Panulims
argus—see Lobster, spiny
marginatus— see Lobster, Hawaiian spiny
Panulims (continued)
marginatus— see also Lobster, spiny
Paralichthys
adspersus 607
microps 607
lethostigma—see Flounder, southern
Paralithodes
camtschatiais—see Crab, red king
platypus— see Crab, blue king
Penaeus duorarum—see Shrimp, pink
Photo identification
whale, sei 749
Photoperiod effects
conch, queen 161
Plaice 439
Platycephalus speculator— see Flathead
Pleuronectes
bilineatus—see Sole, rock
platessa—see Plaice
Pogonias cromis—see Drum, black
Pollock, walleye 129
Pomatomus saltatrix—see Bluefish
Population dynamics
sole, rock 552. 561
Population studies
armorhead. pelagic 756
conch, queen 250
genetic structure 770
lobster
American 659
spiny 483
Lutjanus vittus 395
pollock, walleye 129
salmon, chinook 77
shrimp, deepwater 476
whale, humpback 429
Predation— see also Mortality rates
lobster, spiny 691
Productivity
abalone fishery 139
Pseudopentaceros wheeleri—see
Armorhead, pelagic
Rachycentnnn canadum—see Cobia
Recruitment
fishes, marine 439
lobster, spiny 483
Reproductive behavior
conch, queen 161
Reproductive biology
conch, queen 161
goosefish 217
jewfish 243
lobster, American 659
pollock, walleye 129
rockfish, yellowtail 231
sole. Dover 101
swordfish 809
tautog 529
tonguefish 328
whale, humpback 429
Reproductive maturity
goosefish 217
hind, red 516
jewfish 243
sole, Dover 101
824
INDEX: SUBJECTS Fishery Bulletin 90(1-4), 1992
Reproductive maturity (continued)
swordfish 809
Risk and cost analysis in fisheries 736
Rockfish 505
bocaccio 505
chilipepper 505
pygmy 540
rosethorn 540
sharpchin 540
shortbelly 505
widow 505
yellowtail 231, 505, 540, 726
Salmon
Chinook 77, 770
Sciaenops ocellatus—see Drum, red
Scomber scombrus—see Mackerel, Atlantic
Sculpins, marine 454
Scyllarides squammosus—see Lobster,
slipper
Seabass, white 633
Sea level-recruitment interaction
lobster, spiny 483
Seasonal studies
herring, river 376
Seatrout, spotted 407
Sebastes—see Rockfish
entomelas—see Rockfish, widow
JlavidiLs—see Rockfish, yellowtail
goodei—see Rockfish, chilipepper
helvomaculatus—see Rockfish, rosethorn
jordani—see Rockfish, shortbelly
paucispinis—see Rockfish, bocaccio
wilsoni—see Rockfish, pygmy
zacentrus—see Rockfish, sharpchin
Sexual dimorphism
dolphin, spinner 54
Sexual maturity— see Reproductive
maturity
Sheepshead 407
Shrimp
deepwater 476, 494
pink 405
Size estimation— see Age-size estimation
Snapper 395
Sole
Dover 101, 285 ,421, 540
rock 552, 561
Spawning— see also Reproductive Biology
tautog 529
Starvation
dolphin, bottlenose
blubber thickness as indicator 791
Stenella
attenuata—see Dolphin, spotted
longirostris—see Dolphin, spinner
centroaTnericana—see Dolphin,
Central American spinner
longirostris—see Dolphin, pantropical
spinner or Dolphin, whitebelly
spinner
orientalis—see Dolphin, eastern
spinner or Dolphin, whitebelly
spinner
Stock assessment
model 302
Stock identification
bluefish 703
genetics, weakfish 469
herring, northwest Atlantic 203
Strandings, bottlenose dolphin 791
Strovibus gigas—see Conch, queen
Submersible observations
fish habitat 540
Submersible surveys
shrimp, deepwater 476
Surveys, trapping
shrimp, deepwater 494
Survival— see Mortality rates
Swordfish 736, 809
Tagging
acoustic
lingcod 784
yellowtail rockfish 726
Tagging studies
coded-wire tag retention
drum, red 390
reporting rate 621
shrimp fishery 621
Tautog 529
Tautoga onitis—see Tautog
Taxonomy
Ctinocottus analis 454
croaker 574
flounder 607
mackerel
double-lined 13
shark 13
Orthonopias triads JtSJt
Paraiichthys
adspersus 607
microps 607
sculpins, marine 454
Umbrina
analis 574
Taxonomy (continued)
Umbrina (continued)
busingi 574
dorsalis 574
galapagorum 574
reedi 574
roncador 574
wintersteeni 574
xanti 574
Temperature
distribution, Atlantic cod 599
Temperature effects
conch, queen 161
dolphin, bottlenose 791
Texas
dolphins, bottlenose 791
East Matagorda Bay 791
shrimp fishery 621
Theragra chalcogramma—see Pollock,
walleye
Thunnus albacares—see Tuna, yellowfin
Tonguefish 328
Transplant
lobster, American 659
Tuna
yellowfin 678
Tursiops tru)icatus—see Dolphin,
bottlenose
Turtle, leatherback 807
Umbrina
analis 574
busingi574
dorsalis 574
galapagorum 574
reedi 574
roncador 574
wintersteeni 574
xanti 574
Vertebrae
goosefish 217
Weakfish 469
Whale
humpback 429, 588
sei 749
Whiting, Pacific 260
Xenobalanus globicipitis—see Barnacle
Xiphias gladius—see Swordfish
Yield, potential
shrimp, deepwater 494
Statement of Ownership,
Management and
Circulation
t Required by 39 U S.C 36851
1 A Title c( Piililicalion
ie PUBIICAIION NO
2 Dele of Filing
Fishery Bulletin
3
6
6
-
3
7
0
9-16-92
3 Ffpqoency of Issue
Quarterly
3A. No of issues Published
Annuelly
4
3B Annuel Subsciiption Ptice
24.00 domestic
30.00 foreign
4 Complete Ma-lmg Address of Known Olftce ol Publication tSirtet. Clly. County, Stale and ZiP*4 Code) fStH priMrrx)
HMFS Scientific Publications Office
7600 Sand Point Way NE, Seattle, WA 98115-0070
5 Conifilete M.niing Address ol the HesdQuerted ot Genersl Business OMices o( lh# Pubtlsha' (Noi pnnirr)
National Marine Fisheries Service , NOAA/DOC
1335 East-West Highway, Silver Spring, MP 20910
6 Full Names and Complete Mailing Address of Publisher, Editor, and Managing Editor (Tkii turn MUST MJT be Nank)
Publisher l\,'nu- ..-1,1 Cf^'iflfii Mu,I.Hg_ A.Urrtfl
Sher (\,i'iif :.-ut C. '■>,,■! f I, Ktuil.Hg *.Uri-tfl
Willis Hobart, Scientific Publications Office
7600 Sand Point Way NE, Seattle, WA 98115-0070
FdltO' A. I.". .I».l (.mifUl, Sl.ilhnti,t,frful , , , .
Hancy Peacock, Scientific Publloatlong Office
7600 Sand Point Way NE, Seattle, WA 98115-0070
Managtig Editor iN.inir and Coinplrir Mailing Addretsi
Same as above
7 Owiiei (II fi-iti.t h\ u .r'7'ii'iiK'i'i, 1/1 iitiPif iind uJJrut mutt br Muled and -iha Immedtalfly ihfriwidfr ihe miiiift iinil lUliUtsitt fif fii;lhnlilffi uiwiinjt iii holihnu
I luiuni •'• Ki-it (i/ I, Hill ami'uni n/ wo. t //nof nxitfd by a corporaiion. ihr nantrs and addrntrt •>/ ibr inditidual imntri wtui he kmhi 1/ cKiinl h\ a piinnmhip
I" iiihir wu'wrjioriiird firm, in name and addre^i. ai n-ell at ihui of rath IndnirtunI must he gl**n If the pubHcain't it publlihed h\ a mmprofii i)'ganl:<uinn. in
ii.iHtr ami luUlms muu
be Milled 1 flirni musi br compleied >
Full Nam*
Complvt* Mallng Addrtaa
U, S, Dept.
of Commerce
14th St.
, NW, Washington, DC 20230
8 Knrvvn Bonrtliolder
Secuimes 'li ihrrr n
. Moitgagees and Other Security Holders Owrtlng of Holding
Percent of Mor* ol Total Amounl ol Bonds, Mortgages or Other
Full Nam*
NnfJF
9 Fo' Compleifon bv Nonprofit Oigannanons Authorlted To Mall at Special Rates CDAfM Srmon 424 12 onhl
Tlip purpose (unction, and nonprodi status ol this oiganiiation and the oiempt Status (or Federel ncome tax purposes iChrck nnri
\ J Has Noi Char>gori During
I ifl Pti-.C.I.iin '? Months
□ Mas Changnd Dutmg
Precei
icedmg \2 Munths
l/f •hanged. puNnhrr imoi <i,h>fr ryl»'<„fl,m -/
• li.inBr I'llh rAiI tnilri'irtu }
Emenl ond Nature of ClrrulsHo
A Total No. Copies l\ri Prrtt Kun)
B Paid and'or Requested Circulation handled bv U.S. Supt.
1 Sales thiough riealsrs and earners, sireal vendoi^ and counter tales
2 Mail Suhscription
lP.„.l ..».l ,>r re4>ie\n
C Total Paid and/or Requested Ciiculatlon
,Sum ■■/ /('«; ,md 1082)
D Free Disinbutiort by Mail. Carrier or Other Means
Samples, Complimentary, and Other Fre« Coplts
E Total Distribution iSum of C and D)
F Copies Nol Distributed
1 Oilier use. led over, unaccounted, spoiled after printing
Averege No Copies Each Issue During
Preceding 1 2 Months
2074
2080
of Documents, Washingtork, DC 20240
1600
1455
2055
19
Acliinl No Cop.os of SiM||le ts'.uo
Published Noaiosi h< rilmg Ontc
600
1457
^57
23
2 RBturn (rom News Agents
G TOTAL iSum of E, Fl and 2— should equal net press nin jhriHTi in A\
I certify that the statements made by
fne above are correct and complete
PS Form 3526. January 1991
Superintendent of Documents Subscriptions Order Form
I — I YES, enter my subscription as follows:
Order Processing Code:
*5178
Charge your order.
It's Easyl
To fax your orders (202) 512-2233
subscriptions to FISHERY BULLETIN (FB) for $24.00 per year ($30.00 foreign).
The total cost of my order is $ . Price includes regular domestic postage and handling and is subject to change.
Please Choose Method of Payment:
I I Check Payable to the Superintendent of Documents
n GPO Deposit Account | | | | | | | l-fl
(Company or Personal Name)
(Please type or print)
(Additional address/attention line)
(Street address)
I I VISA or MasterCard Account
rm
(City, State, ZIP Code)
(Credit card expiration date)
(Daytime phone including area code)
Thank you for
your order!
(Purchase Order No.)
YES NO
May we make your name/address available to other maflers? I I I I
(Authorizing Signature) 11
Mail To: New Orders, Superintendent of Documents
PO. Box 371954, Pittsburgh, PA 15250-7954
The National Marine Fisheries Service (NMFS) does not approve, recommend or en-
dorse any proprietary product or proprietary material mentioned in this publication.
No reference shall be made to NMFS, or to this publication furnished by NMFS, in
any advertising or sales promotion which would indicate or imply that NMFS ap-
proves, recommends or endorses any proprietary product or proprietary material
mentioned herein, or which has as its purpose an intent to cause directly or indirect-
ly the advertised product to be us"d or purchased because of this NMFS publication.
5330 8.-.
Fishery Bulletin
Guidelines for Contributors
Form of manuscript
The document should be in the following
sequence: Title Page, Abstract (not re-
c[uired for Note), Tfext, Acknowledgments,
Citations, Text footnotes. Appendices,
Tables, Figure legends, and Figures.
Title page should include authors' full
names and mailing addresses and the
senior author's telephone and FAX
numbers.
Abstract Not to exceed one double-
spaced typed page. Should include a
sentence or two explaining to the gen-
eral reader why the research was under-
taken and why the results should be
viewed as important. Abstract should
convey the main point of the paper and
outline the results or conclusions. No
footnotes or references.
Text A brief introduction should por-
tray the broad significance of the paper.
The entire text should be intelligible to
readers from different disciplines. All
technical terms should be defined, as
well as all abbreviations, acronyms, and
symbols in text, equations, or formulae.
Abbreviate units of measure only when
used with numerals or in tables and
figures to conserve space. Measure-
ments should be expressed in metric
units, with other equivalent units given
in parentheses. Follow the U.S. Govern-
ment Printing Office Style Manual,
1984 ed., and the CBE Style Manual,
5th ed. Fishery and invertebrate no-
menclature should follow the American
Fisheries Society Special Publication 12
(for fishes), 16 (for mollusks), and 1 7 (for
decapod crustaceans).
Text footnotes should be numbered
in Arabic numerals and typed on a sep-
arate sheet from the text. Footnotes are
not used for reference material or per-
sonal communications, but rather to
explain or define terms in the text and
for contribution numbers on the title
page.
Informal sources Personal communi-
cations, unpublished data, and untitled
manuscripts in preparation are noted
parenthetically in the text (full name.
affiliation, brief address including zip
code, and month and year when appro-
priate).
Acknowledgments Gather all ac-
knowledgments into a brief statement
at the end of the text. Give credit only
for exceptional contributions and not to
those whose contributions are part of
their normal duties.
Citations All titled sources should be
listed in the Citations section, including
unpublished and processed material. In
text, cite as Smith and Jones (1977) or
(Smith and Jones 1977); if more than
one citation, list chronologically (Smith
1936, Jones 1975, Doe In press). All
sources cited in the text should be listed
alphabetically by the senior authors'
surnames under the heading CITA-
TIONS. Abbreviations of periodicals
and serials should conform to Serial
Sources for the BIOSIS Data Base'".
Indicate whether sources are in a langu-
age other than English. For informal
literature, include address of author or
publisher. Authors are responsible for
the accuracy of all citations.
Tables should supplement, not dupli-
cate, the text. Each table should be
numbered and cited consecutively, with
headings short but amply descriptive so
that the reader need not refer to the
text. For values less than one, zeros
should precede all decimal points. In-
dicate units of measure in column head-
ings; do not deviate from the unit of
measure within a column. Table foot-
notes should be noted consecutively in
Roman letters across the page from left
to right and then down. Since all tables
are typeset, they need not be submitted
camera-ready.
Figures Photographs and line draw-
ings should be of professional quality-
clear and concise— and reducible to 42
picas for full-page width or to 20 picas
for a single-column width, and to a max-
imum 55 picas high. All graphic ele-
ments in illustrations must be propor-
tioned to insure legibility when reduced
to fit the page format. Line weight and
lettering should be sharp and even. Let-
tering should be upper and lower case.
and vertical lettering should be avoided
whenever possible (except for vertical,
y, axis). Zeros should precede all deci-
mal points for values less than one. Re-
productions of line artwork are accepted
in the form of high-quality photographic
prints from negatives or photomechan-
ical transfer (PMT). Halftones should be
sharply focused with good contrast.
Micron rules should be inserted on elec-
tron micrographs, even when magnifi-
cation is included in the figure legend.
There should be clear distinction be-
tween identifying letters (press-on or
overlay) and background of photograph.
Label each figure in pencil on the back.
Send only xerox copies of figures to the
Scientific Editor; originals or photo-
graphic prints will be requested later
when the manuscript is accepted for
publication.
Copyright Government publications
are in the public domain, i.e., they are
not protected by copyright.
Submission of manuscript
Disks Authors are encouraged to re-
tain manuscripts on word-processing
storage media (diskettes, floppy disks)
and submit a double-spaced hardcopy
run from the storage media. Submit
disks as MS-DOS "print" or "non-
document" files (often called "ASCII
files"). If a disk cannot be converted to
an ASCII file, the author should indicate
on the disk the source computer and
software language along with the file
name. Either 5V4-inch or SVz-inch disks
from IBM-compatible or Apple/Macin-
tosh systems (non-graphics only) can be
submitted, double-sided/double-density
or high-density, limiting each file to 300
kilobytes. All 8-inch word-processing
disks (e.g., Wang or NBI) must be con-
verted onto 5V4- or 3V2-inch MS-DOS
print disks.
Send original hardcopy and two dupli-
cated copies to:
Dr. Ronald W. Hardy, Scientific Editor
Northwest Fisheries Science Center
National Marine Fisheries Service,
NOAA
2725 Montlake Boulevard East
Seattle, WA 98112-2097
Copies of published articles and notes
The senior author and his/her organiza-
tion each receive 50 separates free-of-
charge. Additional copies may be pur-
chased in lots of 100.
MBI. WHOI LIBRAHV
iiiH nup
WWMIIIMJIJIJI
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11