Fishery Bulletin
'ATES O^ ^
Vol. 86, No. 1
Marine Biological Laboratory '
LIBRARY
JUL 6 1000
Woods Hole, m^IS^^'^^^^QS
oplank-
1
25
LOEB, VALERIE J., and OMAR ROJAS. Interannual variation of ichthyopl
ton composition and abundance relations off northern Chile, 1964-85
VETTER, E. F. Estimation of natural mortality in fish stocks: A review ...
NOTARBARTOLO-DI-SCIARA, GUISEPPE. Natural history of the rays on the
genus Mobula in the Gulf of California 45
WILLIAMS, AUSTIN B. Notes on decapod and euphausiid crustaceans, continen-
tal margin, western Atlantic, Georges Bank to western Florida, USA 67
WARLEN, STANLEY M. Age and growth of larval gulf menhaden, Brevoortia
patronus , in the northern Gulf of Mexico 77
O'BRIEN, LORETTA, and RALPH K. MAYO. Sources of variation in catch per
unit effort of yellowtail flounder, Limanda ferruginea (Storer), harvested off the
coast of New England 91
McEACHRON, LAWRENCE W., JEFF F. DOERZBACHER, GARY C. MATLOCK,
ALBERT W. GREEN, and GARY E. SAUL. Reducing the bycatch in a commer-
cial trotline fishery 109
BRUCE, B. D. Larval development of blue grenadier, Macruronus novaezelandiae
(Hector), in Tasmanian waters 119
COWAN, JAMES H., JR., and RICHARD F. SHAW. The distribution abundance,
and transport of larval sciaenids collected during winter and early spring from the
continental shelf waters off west Louisiana 129
HAMNER, WILLIAM M., GREGORY S. STONE, and BRYAN S. OBST. Behavior
of southern right whales, Eubalaena australis, feeding on the Antarctic krill,
Euphausia superba 143
Notes
PENSON, JOHN B., JR., ERNEST 0. TETTY, and WADE L. GRIFFIN. An econo-
metric analysis of net investment in Gulf shrimp fishing vessels 151
SHIRLEY, SUSAN M., and THOMAS C. SHIRLEY. Appendage injury in Dunge-
ness crabs. Cancer magister, in southeastern Alaska 156
(Continued on back cover)
Seattle, Washington
U.S. DEPARTMENT OF COMMERCE
C. William Verity, Jr., Secretary
NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION
William E. Evans, Under Secretary for Oceans and Atmosphere
NATIONAL MARINE FISHERIES SERVICE
Fishery Bulletin
The Fishery Bulletin carries original research reports and technical notes on investigations 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. Beginning 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 instead of being issued individually. 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.
SCIENTIFIC EDITOR, Fishery Bulletin
Dr. Andrew E. Dizon
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 271
La Jolla, CA 92038
Editorial Committee
Dr. Jay Barlow
National Marine Fisheries Service
Dr. William H. Bayliff
Inter-American Tropical Tuna Commission
Dr. George W. Boehlert
National Marine Fisheries Service
Dr. Robert C. Francis
University of Washington
Dr. James R. Kitchell
University of Wisconsin
Dr. William J. Richards
National Marine Fisheries Service
Dr. Bruce B. CoUette
National Marine Fisheries Service
Dr. Tim D. Smith
National Marine Fisheries Service
Mary S. Fukuyama, Managing Editor
The Fishery Bulletin (ISSN 0090-0656) is published quarterly by the Scientific Publications GfTice, National Marine Fisheries Service,
NOAA, 7600 Sand Point Way NE, BIN C15700, Seattle, WA 98115. Second class postage is paid in Seattle, Wash., and additional
offices. POSTMASTER send address changes for subscriptions to Fishery Bulletin, Superintendent of Documents, U.S. Government
Printing Office, Washington, DC 20402.
Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated.
The Secretary of Commerce has determined that the publication of this periodical is necessary in the transaction of the public business
required by law of this Department. Use of funds for printing this periodical has been approved by the Director of the Office of
Management and Budget.
Fishery Bulletin
■ Marine Biolomlil^Z^
CONTENTS / LIBRARY
JUL 6 ma
Vol. 86, No. 1 I January 1988
LOEB, VALERIE J., and OMAR ROJAS. Interannual variation oPrc'Sfhj^^k^k-Mass.
ton composition and abundance relations off northern Chile, 1964-85
VETTER, E. F. Estimation of natural mortality in fish stocks: A review 25
NOTARBARTOLO-DI-SCIARA, GUISEPPE. Natural history of the rays on the
genus Mobula in the Gulf of California 45
WILLIAMS, AUSTIN B. Notes on decapod and euphausiid crustaceans, continen-
tal margin, western Atlantic, Georges Bank to western Florida, USA 67
WARLEN, STANLEY M. Age and growth of larval gulf menhaden, Brevoortia
patronus , in the northern Gulf of Mexico 77
O'BRIEN, LORETTA, and RALPH K. MAYO. Sources of variation in catch per
unit effort of yellowtail flounder, Limanda ferruginea (Storer), harvested off the
coast of New England 91
McEACHRON, LAWRENCE W., JEFF F. DOERZBACHER, GARY C. MATLOCK,
ALBERT W. GREEN, and GARY E. SAUL. Reducing the bycatch in a commer-
cial trotline fishery 109
BRUCE, B. D. Larval development of blue grenadier, Macruronus nouaezelandiae
(Hector), in Tasmanian waters 119
COWAN, JAMES H., JR., and RICHARD F. SHAW. The distribution abundance,
and transport of larval sciaenids collected during winter and early spring from the
continental shelf waters off west Louisiana 129
HAMNER, WILLIAM M., GREGORY S. STONE, and BRYAN S. OBST. Behavior
of southern right whales, Eubalaena australis, feeding on the Antarctic krill,
Euphausia superba 143
Notes
PENSON, JOHN B., JR., ERNEST O. TETTY, and WADE L. GRIFFIN. An econo-
metric analysis of net investment in Gulf shrimp fishing vessels 151
SHIRLEY, SUSAN M., and THOMAS C. SHIRLEY. Appendage injury in Dunge-
ness crabs, Cancer magister, in southeastern Alaska 156
iContinued on next page)
Seattle, Washington
1988
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washing-
ton DC 20402— Subscription price per year: $16.00 domestic and $20.00 foreign. Cost per
single issue: $9.00 domestic and $11.25 foreign.
Contents — Continued
CURRNES, KENNETH P., CARL B. SCHRECK, and HIRAM W. LI. Reexamina-
tion of the use of otolith nuclear dimensions to identify juvenile anadromous and
nonanadromous rainbow trout, Salmo gairdneri 160
BUTLER, JOHN L., and DARLENE PICKETT. Age-specific vulnerability of
Pacific sardine, Sardinops sagax, larvae to predation by northern anchovy, En-
graulis mordax 163
EPIFANIO, CHARLES E., DAVID GOSHORN, and TIMOTHY E. TARGETT.
Induction of spawning in the weakfish, Cynoscion regalis 168
The National Marine Fisheries Service (NMFS) does not approve, recommend or
endorse any proprietary product or proprietary material mentioned in this publi-
cation. 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 approves, recommends or endorses any proprietary product or proprietary
material mentioned herein, or which has as its purpose an intent to cause directly
or indirectly the advertised product to be used or purchased because of this NMFS
publication.
AWARDS
The Publications Advisory Committee of the National Marine
Fisheries Service has announced the best bublications authored
by the NMFS scientists and published in the Fishery Bulletin for
1986 and the Marine Fisheries Review for 1985. only effective
and interpretive articles which significantly contribute to the
understanding and knowledge of NMFS mission-related studies
are eligible, and the following papers were judged as the best
In meeting this requirement.
Fishery Bulletin 1986 — Starvation-induced mortalitv of young
sea-caught jack mackerel, Trachurussymmetrlcus, determined
with histological and morphological methods, by Gail M.
Theilacker. Fish. Bull., U.S. 84: 1-17. Gail M. Theilacker is with the
Southwest Fisheries Center La Jolla Laboratory National Marine
Fisheries Service, NOAA, La Jolla, CA.
Marine Fisheries Review 1985 — Biology of the red sea urchin,
Strongylocentrotus franclscanus, and its fishery in California, by
Susumu Kato and Stephen c. schroeter. Mar. Fish. Rev 47(3):i-20.
Susumu Kato is with the Southwest Fisheries center Tiburon
Laboratory National Marine Fisheries Service, NOAA, Tiburon, CA,
and Stephen C. schroeter is with the Department of Biological
Sciences, university of California, university Park, CA.
INTERANNUAL VARIATION OF ICHTHYOPLANKTON COMPOSITION
AND ABUNDANCE RELATIONS OFF NORTHERN CHILE, 1964-83
VaLKRIK J LOEB' AND OMAR ROJAS^
ABSTRACT
Larval fishes were collected off of northern Chile during winter (July-September) ichthyoplankton
surveys undertaken in 1964-70, 1972-73, and 1983. The 19-year timespan included a wide variety
of hydrographic conditions in the Humboldt Current area (cold years. El Nino events, and intervening
transition years); it also included the decline and collapse of the anchoveta fisheries and increases of
sardine, mackerel, and jack mackerel stocks off northern Chile and Peru. The ichthyoplankton data
are examined in relation to ambient hydrographic conditions as well as to possible chronological
changes in environmental conditions which led to the increased Chilean sardine stocks and anchoveta
fishery collapse.
More coherent patterns come from considerations of larval fish species composition in 1964-69 and
1970-73 data sets than from years of "similar" hydrographic conditions. A marked shift in relative
abundances of nonfished mesopelagic species in 1969-70 is associated with changes within long-term
physical data bases from Chile and Peru suggesting a large-scale environmental change. Sardine
stock growth began with successful larval survival of 1968-69 and later year classes. Anchoveta stock
decline began in 1972 probably due to poor larval survival. Affiliation of anchoveta and coastal
species larval abundance implies that they are similarly influenced by coastal processes. An atmos-
pherically driven oceanic circulation change beginning in the late 1960's and possibly involving
onshore presence of subtropical and or oceanic waters and altered coastal processes may have been
responsible for the changes in the northern Chilean fish assemblages.
The Humboldt Current region, like the other
major eastern boundary current systems (Califor-
nia, Benguela, and Canary Currents) is domi-
nated by pelagic schooling fish stocks including
anchoveta (Engraulis), sardine iSardinops),
hake iMerluccius), mackerel (Scomber), jack
mackerel (Trachurus), and bonita iSarda) (Par-
rish et al. 1983). These fish stocks, like those in
the other eastern boundary current areas, exhibit
extreme population fiuctuations. Most notable in
the past 30 years are the collapses of Peruvian
and Chilean anchoveta stocks in the mid-1970's
and their succession by sardine and, to a lesser
extent, mackerel and jack mackerel stocks (San-
tander and Flores 1983; Serra 1983).
Hydrographic complexity and variability are
characteristic of eastern boundary current sys-
tems. Included in the Humboldt Current region
are equatorial, subequatorial, subantarctic, and
antarctic oceanic water masses; northward flow-
ing currents and opposing countercurrents; and
wind driven, seasonally variable coastal up-
welling (Wyrtki 1967). Additionally, the region is
iMoss Landing Marine Laboratories, P.O. Box 450, Moss
Landing, CA 95039.
2Instituto de Fomento Pesquero, Avenida Pedro de Valdivia
2633, Casilla 1287, Santiago, Chile.
subject to 1) large seasonal and longer period
fluctuations in advection of water masses of
markedly different properties and 2) large in-
terannual differences in the timing and intensity
of seasonal upwelling processes (Bakun 1987;
Bernal et al. 1983; Parrish et al. 1983; Robles et
al. 1976). The clearest and generally considered
most important of the nonseasonal processes in-
fluencing the biology of the current system is the
El Nino phenomenon (Bernal et al. 1983; Guillen
1983). El Nino events off Peru and Chile are
marked by large-scale atmospherically driven
southward and coastward advection of warm,
high-salinity equatorial and subequatorial sur-
face waters, weakening of coastal upwelling (or
upwelling of warm nutrient-poor waters), and
weakening of subsequent phytoplankton blooms.
These El Nino or warm-water periods are vari-
able in their intensity and duration (Guillen
1983; Santander and Flores 1983). In contrast to
these periods are more "normal" cold-water
events resulting from atmospherically driven in-
tensification of northward flowing cold, low-
salinity subantarctic waters and seasonal up-
welling of cold, nutrient-rich water. Major El
Nino events occurred in 1891, 1925-26, 1940-41,
1957-58, 1965, 1972-73, 1976, and 1982-83;
Manu.script accepted September 1987.
FISHERY BULLETIN: VOL. 86, NO. 1, 1988.
FISHERY HULLKTIN: VOL 8(i, NO 1
major cold events over the past 20 years occurred
in 1964, 1967-68, 1970-71, and 1974-75 (Guillen
1983).
The decline and ultimate collapse of the an-
choveta fisheries of Peru began in 1970 and was
finalized by the intense 1972-73 El Nino; the
northern Chilean stock decline started in 1972
and was finalized by 1977. Factors facilitating
these declines are generally believed to include
overfishing and the devastating effects of the El
Nino on anchoveta spawning behavior and in-
tensity as well as on subsequent recruitment.
Competition and/or predation pressure result-
ing from increasing abundances and distributions
of sardine and mackerel have also been hy-
pothesized (Santander and Flores 1983; Serra
1983).
Because of the great socioeconomic value of the
dominant pelagic fish species of the Peru-Chile
ecosystem, their population fluctuations have re-
ceived a great deal of attention over the past 20
years. However, coincidental changes in the com-
position, abundance, or spawning intensities of
other commercially less important and non-
harvested species have not been examined. Infor-
mation on the changes of these unfished species
in relation to hydrographic conditions and fluctu-
ations of the dominant pelagic fish stocks provide
additional insight into the ecology of the Hum-
boldt Current and may elucidate possible causes
for the dramatic changes which occurred during
the 1970's.
In the present work we examine the abundance
and composition of total ichthyoplankton assem-
blages collected off of northern Chile (lat. 18°-
24°S) during 1964-73 and 1983 in relation to am-
bient hydrographic conditions. "Normal" cold
water as well as warm-water and El Nino events
occurred during the 19-yr sampling span. We also
examine our results with respect to possible
chronological change in environmental condi-
tions which led to the 1977 anchoveta fishery col-
lapse off northern Chile. Our results may be ap-
plicable for interpreting coincidental changes
in the Peruvian ecosystem and may also be
broadly applicable for studies of similar changes
in the other eastern boundary current ecosys-
tems.
METHODS
Samples were collected during 1964-73 and
1983 ichthyoplankton surveys conducted by the
Instituto de Fomento Pesquero. The area most
intensively surveyed was a narrow coastal strip
extending between Arica and Antofagasta (lat.
18°-24°S, long. 70°-72°W; Fig. 1). This area in-
cludes one of two major anchoveta (Engraulis rin-
gens ) spawning grounds off Chile and the pri-
mary sardine iSardinops sagax) spawning area
off Chile prior to 1973 (Fig. 2A, B). All samples
used for interannual comparisons were collected
during late July-September following peak win-
ter anchoveta and sardine spawning periods. Be-
tween 21 and 87 samples from the 18°-24°S area
were analyzed for each of 11 cruises (Table 1). In
one case data from two cruises (August and Sep-
tember 1968) were pooled to provide adequate
coverage. Sampling was done annually from 1964
to 1970 and in 1972 and 1973. There was a 10-yr
hiatus before regular sampling was resumed in
1983.
The 1964-73 samples were collected with
Hensen nets (0.28 m^ mouth opening; 300 ixm
mesh). Prior to 1973 the vertical net hauls were
50-0 m; in 1973 haul depth was increased to 100
m. The 1983 100-0 m vertical hauls were made
with WP2 nets (0.25 m^ mouth opening;
UNESCO 1968) of 300 ^JLm mesh. Samples were
preserved using buffered 5% formalin solution.
Sea surface temperature and salinity data were
collected at most sampling stations for all but two
winter cruises; these data are lacking for 1970
and salinity data are minimal for 1967.
All fish eggs and larvae were removed from
samples, and invertebrate zooplankton biomass
was measured. Wet weight displacement volume
was measured for 1964-73 samples; in 1983 the
Yashnov (1959) technique modified by Robertson
(1970) was used. A calculated correction factor
of 1.44 (±3.34) was applied to the 1983 biomass
values to permit comparison with the earlier
data.
All fish larvae were identified to lowest taxon
possible and counted. We herein treat the larvae
of six commercially important species (anchoveta
[Engraulis ringens], Pacific sardine [Sardinops
sagax], jack mackerel [Trachurus murphyi; also
known as T. symmetricus in U.S.A.], chub mack-
erel [Scomber japonicus]. South Pacific men-
haden [Ethmidium maculatum], and hake [Mer-
luccius gayi]) separately from the other 35
identified taxa. These six species are referred to
as the "PL" (larvae of pelagic schooling species).
The other larval taxa considered together are the
"OL". The PL and OL categories are treated sepa-
rately because abundances of the PL (especially of
anchoveta and sardine) mask abundance rela-
LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE
90° 75° 60°
45<
z
<
UJ
u
o
<
a.
15
- 30'
45'
60'
Figure 1. — Ichthyoplankton study area between Arica and Antofagasta, Chile (18°-24°S), 1964-83.
tions of the OL. The OL are further separated into
myctophid, "other mesopelagic" and "coastal"
fractions.
Ichthyoplankton abundances are expressed as
numbers per 10 m^ sea surface. Flow meters were
not used with the vertical Hensen net hauls;
numbers of larvae caught in each tow are multi-
plied by 30 to provide numbers per 10 m^ esti-
mates. A flow meter used with WP2 nets provided
water volume measurements and more accurate
abundance estimates. Based on these measure-
ments the conversion factor applied to Hensen net
catches appears to be reasonable: water volumes
filtered by 100-0 m vertical WP2 net hauls aver-
76*
74°
"I T"
COQUIMBO
VALPARAISO
FALC AHUANO
VALDIVIA
PTO MONTT
_ai aJl_
18
20
22
24'
26 °
28 °
30°
32
34'
36'
38'
- 40'
42<
44'
FISHKKY lUlI.I.KTIN: VOl. Hfi. NO 1
72° 70°
I 8 °
S
- 20 °
22 o
-24
26
-28
- 30
- 32
- 34
- 36
-38
- 40'
- 42
44
Figure 2. — Spawning areas of (A) anchoveta iEngraiilis ringens) and (Bi sardine tSardinops sagax) off Chile based on egg
abundances during July-September ichthyoplankton surveys, 1964-73.
LOEB and ROJAS: ICHTHYOPLANKTON COiMPOSITION AND ABUNDANCE
Table 1. — Cruises yielding samples used for examination of ichthyoplankton abundance and composition variations
oft northern Chile Only samples from 1 8 -24 S are used for interannual comparisons. Data from two 1 968 cruises are
combined; data from cruise 71(4)69CD are used for analysis of sampling depth-related catch differ-
ences. N = number of samples used in ichthyoplankton analyses Tow types: H = Hensen net; WP2 = WP2 net;
V = vertical.
Year
Cruise
Dates
Location
N
Tow
Type
Depth
(m)
1964
06(3)64GE
16 08-23 09
18 20',
, 23 38'S
70 ir.
, 71 50'W
68
H
V
0-50
1965
13(3)65CD
15 08-09 09
18 20'
, 23 50'S
70 00',
, 72 08'W
76
H
V
0-50
1966
25{3)66CD
21 08-31 09
18 28'
, 23 52'S
7016'
, 72 16'W
72
H
V
0-50
1967
37(3)67CD
17 08-1009
18 25',
, 23 45'S
70 05'
, 71 38'W
59
H
V
0-50
1968
47(3)68NO
25 08
18 28'
, 23 OO'S
70 05'
, 70 58'W
37
H
V
0-50
49(3)68NO
29 09
18=26'
, 23 01'S
70 11'
, 70 59'W
41
H
V
0-50
1969
70(3)69NO
23 08-25 08
18^29'
, 23 01'S
70 09'
, 71 10'W
35
H
V
0-50
71(4)69CD
71(4)69CD
03 12-17 12
03 12-17 12
28 29'
28 29'
, 38 OO'S
, 38 OO'S
71 22'
71 22'
, 73 55'W
, 73 55'W
43
39
H
H
V
V
0-100
0-50
1970
86(3)70NO
25 09-26 09
19°27'
, 21 58'S
70 14'
, 71 03'W
21
H
V
0-50
1972
109(3)72NO
04 09-15 09
18 '29'
, 22^58 'S
70 '10'
, 71 25'W
87
H
V
0-50
1973
130(3)73CP
28 07-08 08
18 17'
, 23 OS'S
7005'
, 72 20'W
42
H
V
0-100
1983
277(3)83CP
07 08-15 09
18°33'
. 23 48'S
70 09'
, 71 38'W
38
WP2
V
0-100
aged 28.3 m'^ yielding a raw count to numbers per
10 m'- conversion factor of 35; this 179f increase in
conversion factor is associated with a 12*7^ de-
crease in mouth opening of WP2 vs. Hensen nets.
Larval fish diversity is expressed as total num-
bers of taxa per sampling period and mean num-
bers of taxa per tow.
Various parametric and nonparametric tests
were used for statistical analyses. Differences in
mean abundances are tested with 2-tailed Z tests
and Mann Whitney U tests (Dixon and Massey
1969; Conover 19711. Similarity of abundance
ranks within data sets are tested with Kendall's
concordance iW) test (Tate and Clelland 1969)
and Spearman's rho (p) correlation test (Conover
1971). Significant values resulting from these
tests are indicated, but due to multiple testing
these values should be used only as indicators of
the relative strengths of relationships. Percent
similarity indices (PSI; Whittaker 1975) are used
for comparisons of species percentage composi-
tion. Because PSIs are strongly influenced by
abundant species, we apply these tests to the OL
fraction as well as to total larvae. We define as
"high" all PSI values >80, as "moderate" PSI's
>65 and <80, and as "low" values <65.
SAMPLING CONSIDERATIONS
Sampling Depth Differences
The 100 m sampling depths in 1973 and 1983
potentially effect direct comparisons of abun-
dance estimates and species composition in these
vs. earlier data sets owing to individual species'
depth distributions. Evaluation of depth-related
sampling differences is possible through a com-
parison of data obtained from coincidental 50-0
m and 100-0 m Hensen net samples taken during
December 1969 at 28°-38"S (Table 1). These data
indicate that abundance estimates and percent-
age contribution of the PL to total larvae were
higher in 50 m tows while those of the OL were
higher in 100 m tows (Table 2). Four of five PL
species caught were more abundant and frequent
in the 50 m samples. The greatest differences
were for Trachurus murphyi, which was 5x more
abundant and 3.6 x more frequent, and Merluc-
cius gayi , which was present only in 50 m sam-
ples. Abundance estimates were higher for 11 of
16 OL taxa in 100 m tows although, with a few
exceptions, catch frequencies were similar. The
greatest difference was for Triphoturus mexi-
FISHERY BULLETIN: VOL 86, NO. 1
Table 2. — Catch comparisons of 50-0 m and 100-0 m Hensen net tows taken at lat. 28 -38 S, long. 71 -74'W during
3-17 December 1969. Abundances expressed as mean and standard deviations of numbers per 10 m2, F = percent
frequency of occurrence in samples Depth-related catch differences are shown as ratios of 50:100 m abundance
estimates, species percentage contribution to total identified larvae, PL and OL percentage contribution to total larvae, and
taxonomic diversity (mean numbers per tow and total numbers of taxa). N = number of samples. PL = larvae of pelagic
species; OL = other larval taxa.
50-
-0 m
100-
■0 m
(N
= 43)
(/V =
39)
Taxon
X
(S)
F
%
X
(S)
F
%
Ratio
Engraulis ringens
34.2
(122.8)
23.2
13.42
22.3
69.4)
15.4
7.23
1.53
Clupea bentincki
6.3
( 28 2)
9.3
2.47
1.5
6.7)
5.1
0.49
4.20
Ethmidium maculatum
0.7
( 4.6)
2.3
0.27
0.8
4.8)
2.6
0.26
0.88
Merluccius gayi
1.4
( 6.4)
46
0.55
Trachurus murphyi
4.2
( 14.0)
9.3
1.65
0.8
4.8)
2.6
0.26
5.25
Total PL
46.8
(131.1)
34.9
1836
25.4
71.0)
23 1
8.24
1.84
Bathylagus nlgngenys
0.7
( 4.6)
2.3
0.27
0.8
4.8)
2.6
0.26
0.88
Vinciguerria lucetia
10.5
( 33.9)
14.0
4.12
16.9
75.6)
7.7
5.48
0.62
Diogenlchthys spp.
31.4
(107.0)
20.9
12.32
37.7
79.9)
28.2
12.22
0.83
Hygophum bruuni
121.4
(238.2)
58.1
47.64
166.2
291.0)
53.8
53.86
0.73
Protomyctophum sp.
3.1
9.2)
10.3
1.00
Diaphus sp.
77
( 18.6)
18.6
3.02
9.2
24.0)
17.9
2.98
0.84
Lampanyctus parvicauda
7.7
( 18.6)
163
3.02
6.9
21.2)
15.4
2.24
1.12
Lampanyctus sp.
1.4
{ 6.4)
4.6
0.55
1.5
6.7)
5.1
0.49
0.93
Triphoturus mexicanus
7.7
( 21.8)
14.0
3.02
18.5
57.4)
20.5
5.99
0.42
Scopelosaurus sp.
0.8
4.8)
2.6
0.26
Normanichthys crockeri
15.3
( 38.9)
18.6
6.00
20.0
66.5)
10.3
6.48
0.76
Sebastes sp.
1.4
( 9.1)
2.3
0.55
0.8
4.8)
2.6
0.26
1.75
Blennild A
0.7
( 4.6)
2.3
0.27
Blenniid D
0.7
( 4.6)
2.3
0.27
0.8
4.8)
2.6
0.26
0.88
Bothid
0.7
( 4.6)
2.3
0.27
Unid. 2
0.7
( 4.6)
2.3
0.27
Small damaged myctophids
4.8
9.2
Other unidentified
13.2
7.0
Total OL
226.0
(316.0)
86.0
81.59
299.4
405.9)
82.0
91.78
0.84
Total larvae
272.8
(340.7)
324.8
400.0)
0.84
No. taxa/tow
2.5
( 2.0)
2.2
2.0)
1.13
Total no. taxa
18
17
0.94
canus which was 2.4 x more abundant and 1.5 x
more frequent in 100 m tows. Because of the large
catch variability, none of the species abundance
differences nor the abundance differences of the
PL, OL, and total larvae are significant (Z tests,
P's all >0.10). Additionally, species abundances
within all positive tows from the two sampling
depths are not significantly different (Mann
Whitney U tests, P > 0.10 in all cases).
The overall species composition of 50 and 100 m
tows was similar. Despite greater proportions of
PL in 50 m tows, the PSI value from comparisons
of total species lists was high (87.7). Species per-
centage contribution within the OL fraction of the
two tow types was also quite similar (94.5). Spe-
cies abundance rankings within the two total lar-
val data sets are significantly correlated
(p = +0.80; P < 0.01 ). Species diversity estimates
(total numbers of taxa and mean numbers of taxa
per tow) are also similar.
From these comparisons it is apparent that the
PL predominantly occur within the upper 50 m.
Similar shallow (e.g., <50 m ) distributions have
been described for dominant PL species off of Peru
(anchoveta, sardine, and hake; Sameoto 1982).
The generally lower 100 m abundance estimates
of these species is puzzling, but suggests possi-
bly shorter sampling time and/or less efficient
sampling within the upper 50 m of these tows.
Higher catch frequency and abundance of T.
mexicanus in 100 m tows suggest that large pro-
portions (e.g., 30-60%) of these larvae are at 50-
100 m.
As a result of these catch differences we suggest
caution in making direct numerical comparisons
between the 1973 and 1983 vs. earlier data sets.
Although the overall compositions and abun-
dance relations should not be markedly altered,
some accommodation should be allowed for the
percentage contributions and across-year abun-
dance ranks of PL species (especially Trachurus)
and Triphoturus mexicanus.
LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE
Net Type Differences
There are no data available for a direct evalua-
tion of catch differences between the vertical
Hensen and WP2 net tows. However, comparisons
of data from the 1983 WP2 net hauls (Table 3) do
not indicate that this net is more or less efficient
than the Hensen net. Mean abundances and spe-
cies diversity (total numbers of taxa and mean
numbers of taxa per tow) are within the ranges of
values from Hensen net tows.
Day-Night Catchi Considerations
Day and night sample data have been combined
for each cruise. Day samples (0600 -1800 hours)
outnumbered (generally 55-64%) night samples
during all cruises. In nine cases there were non-
significant day-night larval catch differences (Z
tests, P >0.05) and overall similarity in night:
day catch ratios (0.74-1.8:1; mode = 1.2:1). One
cruise had a significantly higher night vs. day
catch (2.5 X; P < 0.01); this cruise and one with
Table 3. — Abundance estimates and diversity of larval fishes and abundance estimates of zooplankton collected in July-September
samples off northern Chile (18 -24-S), 1964-83^ Larval fish abundances as mean numbers per 10 m^. PL = larvae of commercially
important pelagic species; OL = other larval taxa (myctophid, other mesopelagic, and other categories). Number of larvae is total raw count
of identified larvae for each sampling period. Larval fish diversity expressed as total number of identified taxa and mean number of taxa
per tow. Zooplankton abundance is mean displacement volume (cc/10 m2). N.A. = data not available.
Mean abundance for
Taxon
1964
1965
1966
1967
1968
1969
1970
1972
1973
1983
Engraulis ringens
302.0
36.7
3,478.3
224.7
72.3
181.7
620.0
1,634.1
1,816.4
—
Sardinops sagax
5.3
5.1
12.5
4.6
21.1
24.9
1.4
327.2
52.1
594.9
Ethmidium maculatum
—
—
—
—
—
2.6
—
0.7
—
—
Trachurus murphyi
2.6
7.1
10.8
3.1
0.4
6.8
5.7
100.3
2.1
—
Scomber japonicus
—
0.4
—
—
—
—
—
25.9
—
—
Merlucaus gayi
—
—
—
—
—
—
—
—
0.7
2.6
Total PL
309.9
49.3
3,501.6
232.4
93.8
216.0
627.1
2,088.2
1,871.3
597.5
Bathylagus nigrigenys
4.0
5.1
6.7
9.7
6.2
0.9
7.2
56.6
35.0
31.5
Vinciguema lucetia
3.1
0.4
2.9
6.1
3.1
2.6
21.4
54.8
18.6
3.7
Stemoptyx diaphana
—
—
—
—
—
—
—
0.7
0.7
1.0
Chaultodus sp.
—
—
—
—
—
—
—
0.3
—
—
Stomias spp.
—
—
—
—
—
—
—
—
—
1.9
Lestidiops pacificum
—
—
—
—
—
—
—
0.3
0.7
—
Melamphaes sp.
—
—
—
0.5
0.4
—
—
0.3
2.1
—
Beryciform
—
—
—
—
—
—
1.4
—
—
—
Other mesopelagics
7.1
5.5
9.6
16.3
9.7
3.5
30.0
113.0
57.1
38.1
Diogenichthys spp.
6.2
18.9
14.6
30.5
7.3
14.6
21.4
59.3
39.3
114.2
Hygophum bruuni
—
2.8
1.2
3.6
—
—
—
—
—
0.9
Hygophum atratum
—
—
—
—
—
—
—
—
—
6.7
Metelectrona ventralis
—
—
—
—
—
—
—
—
■ —
0.9
Myctophum nitidulum
0.4
—
0.4
—
0.4
0.9
—
4.8
4.3
4.6
Diaphus sp.
—
0.4
0.4
—
—
—
1.4
0.3
—
4.5
Lampanyctus parvicauda
41.0
32.0
65.8
63.0
10.8
38.6
5.7
29.7
12.1
15.5
Lampanyctus spp.
—
—
0.8
—
1.1
—
—
—
0.7
—
Triphoturus mexicanus
31.8
15.8
20.8
29.0
4.6
31.7
11.4
19.7
23.6
33.6
fvlyctophids
79.4
69.9
104.0
126.1
24.2
85.8
39.9
113.8
80.0
180.9
Normanichtys crockeri
17.2
1.5
281.2
5.6
2.3
26.6
10.0
2.4
31.4
—
Sebastes sp.
0.4
—
2.5
1.0
0.8
—
—
—
0.7
—
Gadiform D
—
—
0.8
—
—
—
—
—
—
—
Macrourid A
—
—
—
0.5
—
—
—
—
0.7
—
Macrourid C
—
—
—
0.5
—
—
—
—
—
—
Blenniid A
0.4
—
0.8
—
1.2
—
—
—
1.4
—
Blenniid B
—
—
0.4
—
—
—
—
—
—
—
Blenniid C
—
—
—
1.5
—
—
—
0.3
—
—
Blenniid D
5.3
0.4
16.7
7.1
4.2
2.6
—
2.8
8.6
1.0
Gobiesocid A
0.4
—
0.8
—
—
—
—
1.0
0.7
—
Gobiesocid B
—
—
—
—
0.4
—
—
—
—
—
Unid. 1
—
—
—
—
0.8
^
—
—
5.7
8.6
Unid. 2
—
—
0.4
—
—
—
—
0.3
12.1
—
Unid. 3
—
—
—
—
0.4
1.7
—
0.3
0.7
—
Unid 4
—
—
—
—
—
—
—
0.3
0.7
—
Ophidiid
—
—
—
1.0
0.8
—
—
—
—
—
Hippoglossina sp.
—
—
—
—
—
—
—
0.3
—
—
Other larvae
23.7
1.9
303.6
17.2
10.9
30.9
10.0
7.7
62.7
9.6
Total OL
110.2
77.3
417.2
159.6
44.8
120.2
79.9
234.5
199.8
228.4
FISHERY BULLETIN: VOL. 86, NO 1
Table 3. — Continued.
Mean abundance for
Taxon
1964
1965
1966
1967
1968
1969
1970
1972
1973
1983
Larval fish abundance and di
iversity
Total ID PL
309.9
493
3,501 6
2324
938
216.0
627 1
2,0882
1,871 3
597.5
Total ID OL
110.2
77,3
417.2
159 6
448
120.2
799
234 5
1998
2286
Total ID larvae
420.1
126.6
3,9188
3920
138.6
336.2
707.0
2,322.7
2,071.1
826 1
Unid missing OL
10.2
4.4
170
220
3.4
11.9
7.2
104.5
57.2
76.6
Total larvae
430.3
131.0
3.9358
4140
142.0
348.1
714.2
2,427.2
2,1283
9027
Number of larvae
953
321
9,406
771
360
650
495
6,738
2,900
893
Number of taxa
15
14
20
19
20
13
12
26
25
19
Number of taxa tow
22
16
32
25
1.4
1.9
23
5.3
4 1
3.7
Number of samples
68
76
72
59
78
35
21
87
42
38
Zooplankton abundance
333.9
2422
406.5
370.8
279.2
141.0
N.A
301 7
100.6
168.7
Number of samples
85
124
72
59
78
35
110
42
30
a 1.8:1 night:day catch ratio were represented
by fairly equal day {55-5T/( ) and night sam-
ples.
TAXONOMIC PROBLEMS
The 576 samples used for interannual compari-
sons yielded a total of 41 taxa including 19 spe-
cies, 7 genera, and 11 higher taxa (Table 4). The
PL and most mesopelagic forms were identified to
species. During several cruises there were large
proportions of small Diogenichthys spp. (Myc-
tophidae) larvae which could not be identified to
species. As a consequence, data on the two spe-
cies, D. atlanticus and D. laternatiis, were lumped
to permit reasonable between-year taxonomic
composition comparisons. In all but one cruise, D.
laternatiis dominated (77-1007f ) the identifiable
Diogenichthys larvae. Total within-year Dio-
genichthys spp. abundances were multiplied by
proportions of identified D. laternatus and D. at-
lanticus larvae to provide between-year abun-
dance rankings for each species.
Species identifications of coastal forms are lim-
ited by inadequate taxonomic information and by
the presence of generally early larval develop-
mental stages in samples. These larvae are pri-
marily classified at familial and ordinal levels.
Because the classifications include few multispe-
cies groupings and those were numerically rare
the taxonomic limitations offer no severe analyti-
cal problems.
Largest taxonomic problems occurred in
cruises when large numbers of small unidentifi-
able larvae were caught (e.g., 1972 and 1983;
Table 3). Additionally, most cruises had "miss-
ing" larvae (e.g., "other larvae" enumerated when
the samples were first processed but not ac-
counted for during later species identification
work). With the exception of 1983 the uniden-
tified and missing larvae made up <5% of the
total larval abundance for each sampling pe-
riod.
RESULTS
Overall Ichthyoplankton Composition
The 576 July-September samples used for in-
terannual comparisons yielded a total of 23,487
identified larvae. These larvae were dominated
(85.17r) by PL species (Table 4). Overall domi-
nants were anchoveta (Engraulis ringens; 74.37c)
and sardine (Sardinops sagax; 9.3%). The other
PL species were relatively rare: Trachurus mur-
phyi contributed 1.2% and Scomber japonicus,
Merluccius gayi , and Ethmidium maculatum to-
gether formed 0.3% of the total. The larval abun-
dances of these species off of Chile are strongly
influenced by sampling time and location. Mer-
luccius gayi occurs primarily to the south of the
study area (24°-43°S) and Scomber japonicus and
Trachurus murphyi have later summer (Novem-
ber-February) spawning peaks.
The OL were dominated by mesopelagic fishes
(18 taxa, 10.6% of total larvae). Myctophids were
most abundant (8.0%) primarily because of the
large numbers of Diogenichthys spp., Lampanyc-
tus parvicouda , and Triphoturus mexicanus,
which together made up 7.7% of the total. One
bathylagid (Bathylagus nigrigenys) and one
gonostomatid (Vinciguerria lucetia ) were also rel-
atively abundant (together 2.5%). Coastal fish
larvae (14 taxa) made up 4.2% of the total; a
scorpaeniform (Normanichthys crockeri; 3.4%)
and blenniid (Blenniid D; 0.4%) dominated this
group.
Eighteen taxa were relatively frequent (e.g., in
8
LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE
>50% of cruises) and/or abundant across the 10
sampling periods (Table 4). These taxa (four PL
species, seven myctophid taxa, two other meso-
pelagic species, and five coastal forms) made up
99.69; of the identified larvae; they also con-
tributed 97-1009r of the identified larvae (91.2-
99.6% of total larvae) and included the top 9-12
ranked taxa within each sampling period (Table
5).
Table 4— Ichthyoplankton species collected in July-September
samples oft northiern Chile (18 -24 S, 70 -72 W), 1964-83. Rela-
tive abundance (ROA) and percentage composition based on
summed cruise mean abundances (no. 10 m2) of all identified
forms. Frequency is number of total 10 sampling periods when
taxon was caught. Categories are PL (commercially important
pelagic species) and other taxonomic components (M = myc-
tophids: OM = other mesopelagic taxa; C = coastal forms).
Fre-
Cate-
Taxon
ROA
%
quency
gory
Engraulis ringens
1
74.29
9
PL
Sardinops sagax
2
9.33
10
PL
Normanichthys crockerl
3
3.36
9
C
Diogenichthys latema-
tus + D. atlanticus
4
2.90
10
M
Lampanyctus parvicauda
5
2.79
10
M
Tnphoturus mexicanus
6
1.97
10
M
Bathylagus nigngenys
7
1.44
10
OM
Trachurus murphyi
8
1.23
9
PL
Vinciguerna lucetia
9
1.04
10
OM
Blenniid D
10
0.43
9
C
Scomber japonicus
11
023
2
PL
Myctophum nttidulum
12
0 14
7
M
Unid. 1
13
0.13
3
C
Unid. 2
14
0.11
3
C
Hygophum bruuni
15
0.075
4
M
Diaphus sp.
16
0.062
5
M
Hygophum atratum
17
0.060
1
M
Sebastes sp.
18
0050
5
C
Blenniid A
19
0.034
4
C
Ethmidium macu latum
20.5
0030
1
PL
Merluccius gayi
20.5
0.030
2
PL
Unid. 3
22
0028
4
C
Melamphaes sp
235
0.026
3
OM
Gobiesocid A
23.5
0026
4
C
Lampanyctus spp.
25
0024
3
M
Sternoptyx diaphana
26
0,020
3
OM
Stomias spp.
27
0.017
1
OM
Blenniid C
285
0016
2
C
Ophidiid
285
0016
2
C
Beryciform
30
0.012
1
OM
Macrourid A
31
0011
2
Lestidiops pacificum
32.5
0.009
2
OM
Unid. 4
32.5
0.009
2
C
Metelectrona ventralis
34
0.008
1
M
Gadiform D
35
0.007
1
Macround C
36
0,004
1
Blenniid B
37.5
0.0036
1
C
Gobiesocid B
37.5
0.0036
1
C
Chaultodus sp.
39.5
0.0030
1
OM
Htppoglossina sp.
39.5
0.0030
1
C
Total number of identified larvae:
23,487
Total number of samples:
576
Total number of taxa:
41
INTERANNUAL VARIATIONS IN
ABUNDANCE AND COMPOSITION
The ichthyoplankton demonstrated extreme in-
terannual variations in abundance and composi-
tion (Tables 3, 5; Fig. 3A, B). Most obvious are the
1) total larval and PL abundance peaks of 1966,
1972, and 1973 and 2) shift from anchoveta to
sardine dominance between 1973 and 1983. The
maximum total larval abundance values in 1966,
1972, and 1973 were 2.4-30 x higher than those
of other years; the PL had 3-71 x higher abun-
dances during these vs. other years. Interannual
abundance variations during the seven years of
lower abundance were also large for total larvae
(to 6.9 X ) and the PL (to 12.7 x ). Anchoveta domi-
nated the ichthyoplankton through 1973 (29.0-
88.8% ) and was directly responsible for the ex-
treme abundance variations; anchoveta were
absent from the 1983 samples, and sardine larvae
(72.0%) contributed to the moderately high total
larval and PL abundances (Table 5).
The OL fraction had less extreme abundance
variations than the PL: maximum 1966, 1972,
1973, and 1983 mean abundance values ranged
from 1.4 to 9x those of the other years; mean
values within the six years of lower abundance
varied to 3.8 x. Unlike the PL, OL abundance
fluctuations were not attributed to any one taxo-
nomic component (Fig. 3B): the 1966 peak was
largely due to coastal taxa (73% of OL); myc-
tophids and other mesopelagic taxa equally domi-
nated the 1972 peak (48%); myctophids domi-
nated (79%) in 1983; and all three components
were relatively abundant (29-40%) in 1973. Myc-
tophids were generally the dominant component
(50-90% ) during the six years of lower OL abun-
dance. Abundance fluctuations (as range of mean
abundance values) across the 10 years were less
extreme for myctophids (8.1 x ) than for the other
mesopelagic taxa (32.3 x) and coastal forms
(160x).
In contrast to the ichthyoplankton, July-Sep-
tember zooplankton biomass values were rela-
tively constant between years and exhibited only
a 4x range in values (Table 3).
Despite large between-year variability in rela-
tive proportions of the PL and OL, there is a sig-
nificant agreement of their ranked mean abun-
dances across the 10 years (p = +0.81, P < 0.01).
There is also a general agreement of the ranked
mean abundances of PL and the three OL compo-
nents across the 10 years (W = 0.44; P = 0.05).
These categories were generally more abundant
9
FISHERY BULLETIN: VOL. 86, NO 1
Table 5. — Comparisons of relative abundances of dominant larval fish taxa collected off of northern Chile (18-
24°S, 70°-72 W) during July-September sampling periods, 1964-83. Relative abundances within each year are
presented as (A) percentage contribution to total identified larvae and (B) ranked abundance. Taxa are listed in
order of total summed 10-yr mean abundances
Taxon
1964
1965
1966
1967
1968
1969
1970
1972
1973
1983
A. Percentage contribution to total identified
larvae
Engraulis nngens
71.89
28.99
88.76
57.32
52 20
54.05
87.69
70 35
87.70
—
Sardinops sagax
1.26
4.03
0.32
1.17
15 23
7.41
0.20
14.09
2.52
72.01
Normanichthys crocken
409
1.18
718
1.43
1.66
7.91
1.41
0.10
1.52
—
Diogenichthys spp.
1.48
14.93
0.37
7.78
5.27
4.34
303
2.55
1.90
13.82
Lampanyctus pan/icauda
9.76
25.28
1.68
16.07
7.80
11.48
0.81
1.28
058
1 88
Tnphoturus mexicanus
7.57
12.48
0.53
7.40
332
9.43
1 61
085
1.14
4.07
Bathylagus nigrigenys
095
4.03
0.17
2.47
4.48
027
1.02
244
1.69
381
Trachurus murphyi
0.62
5.61
0.28
0.79
0.29
2.02
0.81
4.32
0.10
—
Vinciguerna lucetia
0.74
0.32
0.07
1.56
2.24
0.77
3.03
2.36
0.90
0.45
Blenniid D
1.26
0.32
0.43
1.81
3.03
0.77
—
0.12
0.42
012
Scomber japonicus
—
0.32
—
—
—
—
—
1.12
—
—
Myctophum nitidulum
0.10
—
0.01
—
0.29
0.27
—
0.21
0.21
0.56
Unid. 1
—
—
—
—
0.56
—
—
—
0.27
0.96
Unid. 2
—
—
0.01
—
• —
—
—
0.01
0.57
—
Hygophum bruuni
—
2.21
0.03
0.92
—
—
—
—
—
0.11
Diaphus sp.
—
0.32
0.01
—
—
—
0.20
0.01
—
0.54
Hygophum atratum
—
—
—
—
—
—
—
—
—
0.81
Sebastes sp.
0.10
—
0.06
0.26
0.58
—
—
—
0.03
—
Other taxa
0.19
—
0.09
1.02
3.03
1.28
0.20
0.19
0.43
0.78
B. Ranked within-year abundance
Engraulis nngens
1
1
1
1
1
1
1
1
1
—
Sardinops sagax
6.5
6.5
7
9
2
5
10
2
2
1
Normanichthys crocken
4
9
2
8
9
4
5
12
5
—
Diogenichthys spp.
5
3
6
3
4
6
2.5
4
3
2
Lampanyctus parvicauda
2
2
3
2
3
2
7.5
7
8.5
5
Triphoturus mexicanus
3
4
4
4
6
3
4
9
6
3
Bathylagus nigrigenys
8
6.5
9
5
5
12.5
6
5
4
4
Trachurus murphyi
10
5
8
11
17
7
7.5
3
13.5
—
Vinciguerria lucetia
9
11.5
10
7
8
9
2.5
6
7
10
Blenniid D
6.5
11.5
5
6
7
9
—
11
10
13.5
Scomber japonicus
—
11.5
—
—
—
—
—
8
—
—
Myctophum nitidulum
12.5
—
18.5
—
17
12.5
—
10
12
8
Unid. 1
—
—
—
^
12.5
—
—
—
11
6
Unid. 2
—
—
18.5
—
■ —
—
—
20.5
8.5
—
Hygophum bruuni
—
8
12
10
—
—
—
—
—
15.5
Diaphus sp.
—
11.5
18.5
—
—
—
10
20.5
—
9
Hygophum atratum
—
—
—
—
—
—
—
—
—
7
Sebastes sp.
12.5
—
11
13.5
14
—
—
—
21
—
in 1966, 1972, and 1973 and relatively rare in
1965 and 1968. The ranked abundance patterns of
each of the components differ from one another
(e.g., all pairwise correlation coefficients
[p = —0.21 to +0.61] are nonsignificant). Larval
diversity is strongly correlated with total larval
abundance (p = +0.88, P < 0.01).
There are no significant correlations between
abundance ranks of invertebrate zooplankton
biomass and total larvae (p = +0.27) or any of the
larval categories (p = -0.03 to +0.22; P > 0.05 in
all cases).
Species Abundance Variations
and Relations
The top 10 ranking larval fish taxa were caught
during at least 9 of the 10 sampling periods (Table
4). All of these taxa exhibited large interannual
abundance fluctuations (Table 3). Most marked
were the abundance changes of anchoveta, sar-
dine, and coastal species Normanichthys crockeri.
This latter species (rank 3 in overall abundance)
was frequently abundant prior to 1983; like an-
choveta it was absent from 1983 samples. Among
the 10 taxa only Triphoturus mexicanus had
<10x changes in mean abundance values; <20x
changes occurred for Lampanyctus parvicauda
(11.5X) and Diogenichthys spp. (18.4x); all other
taxa had >20x mean abundance changes over
the 10 years. The abundance fluctuations of these
10 taxa are primarily responsible for the interan-
nual abundance and composition variations
(Table 5; Fig. 3A, B).
10
LOEB and ROJAS: ICHTHYOPLANKTON COMI'OSITION AND AHUNDANCE
A. TOTAL ICHTHYOPLANKTON
E
O
a.
t :1 Anchoveta
Other PL
I I Sordine
■ OL
B.
OL CATEGORIES
II Other mesopeiogic toxo
M yctophids
Coastal toxa
*j>\ij
83
Figure 3. — Mean abundance (numbers per 10 m2) of (A) total ichthyoplankton and major PL components and
(Bl total OL and major OL components collected off northern Chile (18'-24°S) during July-September sampling
periods, 1964-83.
Between-year comparisons of the species per-
centage compositions of total larvae give a wide
range of PSI values (3.3-95.0; Table 6A) which
primarily reflect similarity in percentages of an-
choveta; 51% of these values are moderate to high
(e.g., >65). Highest values (91.5-95.0) come from
comparisons between 1966, 1970, and 1973 when
anchoveta contributed >87% of the larvae. High
values (80.7-82.5) also result from comparisons
between 1964, 1967, and 1969 and result from
moderate anchoveta abundance (54.0-71.9%) and
relatively similar proportions of other taxa. Low-
est values (3.3-31.0) result from comparisons of
1983 vs. all other years and reflect the absence of
anchoveta larvae in 1983 samples. The 1973 and
1983 PSI values are little affected (e.g., <2.3) by
11
FISHERY BULLETIN: VOL. 86, NO 1
Table 6. — Between-year percent similarity index (PS!) values from compari-
sons of (A) total larvae and (B) \he OL (ottier larval taxa) fraction collected
during July-September sampling periods off northern Cfiile (18-24 S), 1964-
83.
PS! values for
1965
1966
1967
1968
1969
1970
1972
1973
1983
A. Total larvae
1964
52.44
79.94
8222
71.25
80.70
79.70
77.89
80.30
10.30
1965
X
33.95
67.41
55.54
62.38
38.16
45.32
38.74
28.47
1966
X
62.69
57.83
65.08
91.55
73.65
91 88
3.32
1967
X
77.96
8247
67.75
68.21
66.82
18.05
1968
X
79.39
62.81
76.39
63.68
30.95
1969
X
62.95
69.67
63.50
18.80
1970
X
79.06
94.95
7.31
1972
X
79 42
22.02
1973
X
8.95
B. OL fraction
1964
69.88
47.01
75.76
57.56
85.16
45.99
36.08
51.29
33.15
1965
X
29 22
88.07
60 36
68.42
55.43
54.30
47.13
54.39
1966
X
34.96
3678
50.09
30 53
26 15
37 50
18.48
1967
X
70.21
71.01
53.92
52.55
55.68
49.12
1968
X
58.46
54.85
61.57
67 92
51.92
1969
X
48.98
39.22
51 92
37.18
1970
X
74.35
68.40
60.21
1972
X
66 23
58.74
1973
X
58.61
accommodations for possible sampling depth-
catch differences of Trachurus gayi and Triphotu-
rus mexicanus.
When the PL are excluded, comparisons be-
tween the OL taxa yield generally lower PSI val-
ues than those of the total larvae; only 2¥7( of the
18.5-88.1 values are moderate to high (Table 6B).
Moderate to high values (69.9-88.1) come from
comparisons between 1964, 1965, 1967, and 1969,
and in part result from similar proportions of
Lampanyctus parvicauda (32.1-41.4%) and Tri-
photurus mexicanus (18.2—28.99^) during those
years. Moderate values (66.2-74.4) also come
from comparisons of 1967 vs. 1968 and 1969 (sim-
ilar proportions of L. parvicauda, T. mexicanus,
and Diogenichthys [12.2-39.5%!); 1970 vs. 1972
(similar proportions of Vinciguerria lucetia and
Diogenichthys [23.4-26.8%!); and 1972 vs. 1973
(similar proportions of Bathylagus nigrigenys, V.
lucetia, and Diogenichthys [9.3-25.3%!). Lowest
PSI values (<30) result from comparisons of 1966
vs. 1965, 1970, 1972, and 1983, and are due
largely to extreme dominance by Normanichthys
crockeri (67.4% of OL) in 1966. Recalculations to
accommodate for possible depth-related increased
catches of T. mexicanus in most cases decrease
1973 and 1983 PSI values (e.g., by 2.6-7.2) and in
two cases (1968 and 1970 vs. 1973) change the
value characterization from moderate (67.9 and
68.4) to low (63.8 and 64.5). With one exception
(1972 vs. 1973, PSI = 67.0)
justed values are low.
all of the other ad-
Species Across-Year Ranked
Abundance Patterns
Individual species across-year abundance rank-
ings demonstrate a variety of patterns. Three pat-
terns are shared by nine of the more frequently
occurring taxa (Table 7). These involve 1) a
group formed by anchoveta and three coastal
forms; 2) a group formed by one myctophid and
two other mesopelagic species; and 3) a species
pair consisting of sardine and a myctophid. An-
other species pair (two myctophids) can be formed
if the 1973 and 1983 abundances of T. mexicanus
are adjusted.
Group I includes anchoveta, Normanichthys
crockeri, Blenniid D, and Sebastes sp. (Table 7).
There is a significant concordance among these
species as to years of highest (1966 and 1973) and
lowest (1965 and 1983) abundance (W = 0.69,
P < 0.01). The abundance rankings of anchoveta
and A'^. crockeri (p = +0.79) and of Blenniid D and
Sebastes (p = +0.88) are significantly correlated
(P < 0.01). None of the correlations between spe-
cies of the two pairs are significant due to differ-
ences in 1967-68 vs. 1970-72 relative abun-
dances.
The three Group II species, Diogenichthys later-
12
LOEB and RO.I.XS ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE
notus. Bathylagus nigrigenya , and Vinciguerna
lucetia, have a concordance of higher abundances
in 1972, 1973, and 1983 vs. other years (W = 0.81,
P<0.01). The abundance rankings of B. ni-
grigenys are strongly correlated with those of D.
laternatus and V. lucetia (p = +0.82, P < 0.01 in
both cases); the correlation between D. laternatus
and V . lucetia is not significant.
The abundance rankings of sardine and Mycto-
phum nitidulum (Pair I species) are significantly
correlated (p = +0.96, P<0.01); both species
were rare or absent in 1970 and most abundant in
1972, 1973, and 1983.
With adjustment to 1973 and 1983 abundances
of T. mexicanus, its abundance rankings are
strongly correlated with those of Lampanyctus
parvicauda (Pair II species; p = +0.84, P < 0.01).
Highest relative abundances of both species were
in 1964, 1966, 1967, and 1969.
Three relatively frequent species iTrachurus
murphyi, Diogenichthys atlanticus , and Diaphus
sp.) do not conform to any of the above patterns.
If the 1973 Trachurus abundance is adjusted to
accommodate for possible undersampling, its
abundance pattern is similar to that of anchoveta
(p = +0.68, P = 0.05) but not to any of the other
Group I species (p = +0.03 to +0.46).
VARIATIONS IN ABUNDANCE AND
COMPOSITION RELATIVE TO
HYDROGRAPHIC CONDITIONS
Ichthyoplankton abundances in the 10 years
sampled show no consistent patterns relative to
warm water-cold water events (Table 8). High PL
and OL abundances occurred during strong El
Nino events (1972 and 1983) and during cold or
transition years immediately following El Nihos
(1966 and 1973). Lowest abundances of both frac-
tions were associated with the 1965 El Nino and
warm 1968. Neither the PL nor OL have signifi-
cant correlations with ranked (high to low) July-
Table 7. — Larval fish taxa grouped according to similar across-year (1964-83) ranked abundance patterns.
Significant agreement of group rankings indicated by Kendall's concordance {W) values. Correlations between
species pair rankings indicated by Spearman's rhio (p) values. Significant values at P £ 0.05 are indicated, but
note use of multiple testing. " = abundance ranks adjusted to accommodate for possible sampling deptti
related catch differences (Table 2).
Abundance rank foi
r
Species
1964
196£
. 1966
1967
1968
1969
1970
1972
1973
1983
Group 1
Engraulis nngens
Normanichthys crocken
Blenniid D
5
4
4
9
9
9
1
1
1
6
6
3
8
8
5
7
3
7
4
5
10
3
7
6
2
2
2
10
10
8
Sebastes sp.
5
8
1
2
3
8
8
8
4
8
E nngens-N- crocken
Blenniid D-Sebastes sp,
E nngens- Blenniid D
E- nngens-Sebastes sp.
N. croc/(en -Blenniid D
W =
P =
P =
P =
P =
P =
0.69 (P
+ 0.79 (P
+ 0.88 (P
+0.59
+0.44
+ 0.61
o o o
V II V
N. crocken-Sebastes sp.
P =
+ 0.48
Group II
Diogenichthys laternatus
Bathylagus nigngenys
Vinciguerna lucetia
10
9
6.5
5
8
10
7.5
6
8
4
4
4
9
7
6.5
7.5
10
9
6
5
2
2
1
1
3
2
3
1
3
5
B nigngenys-D. laternatus
B nigrigenys-V. lucetia
D. laternatus-V lucetia
W =
P =
P =
P =
0.81 (P
+ 0.82 (P
+ 0.82 (P
+ 0.52
<0.01)
<0.01)
< 0.01)
Pair 1
Sardinops sagax
Myctophum nitidulum
7
6
8
9
6
6
9
9
5
6
4
4
10
9
2
1
3
3
1
2
S sagax-M. nitidulum
P =
+ 0.96 (P<0.01)
Pair II
Lanpanyctus parvicauda
Tnphoturus mexicanus ' '
3
1
5
6
1
4
2
3
9
10
4
2
10
8
6
5
8
9
7
7
L. parvicauda-T. mexicanus"
P =
+ 0.84 (P
<0.01)
13
FISHERY BULLETIN VOL H6, NO 1
Table 8. — Range, mean, standard deviation and ranked (high to low) values of sea
ichthyoplankton sampling periods, 1964-83. N = number of observations. N.A. ^ data not
(1976), Bernal et al. (1982), and Kelly and Blanco (1983).
1964
1965
1966
1967
1968
Temperature (°C)
N
85
128
72
55
81
Range
13.5-17.9
13.6-18.1
13.7-17.1
13.3-16.0
14.3-18.6
X
15.7
16.5
15.3
14.8
16.5
(S)
(1.0)
(1.0)
(0.8)
(0.8)
(0.9)
Rank
7
3.5
8
9
3.5
Salinity (%o)
N
84
124
72
4
81
Range
34.51-34.99
34.80-35.40
34.53-35.03
34.74-34.82
34.69-35.23
X
34.74
35.06
34.83
34.78
34.92
(S)
(0.12)
(0.12)
(0.11)
(0.04)
(0.12)
Rank
9
2
6
8
4
Hydrographic
condition:
Cold
El Nino
Transition
Cold
Warm
September mean temperature and salinity val-
ues; larval diversity (mean number of taxa/tow)
also shows no correlation with these values
(Table 9). Within the PL, anchoveta were most
abundant during years immediately following El
Nihos (cold 1966, transition 1973), the 1972 El
Nirio, and cold 1970; lowest abundances were dur-
ing the 1965 and 1983 El Nihos (Table 8). There
are no significant correlations between ranked
anchoveta abundances and ranked values of tem-
perature or salinity (Table 10). Sardine larvae
were most abundant during and after the 1972 El
Niiio; prior to this moderate abundances and rela-
tively large percentage contributions to the
ichthyoplankton occurred only during the warm
1968-69 period (Tables 5, 7). Lowest sardine
abundances were during cold years 1964, 1967,
and 1970 and the 1965 El Nino. Despite low abun-
dances during the 1965 El Nino, there is a signif-
icant positive correlation between ranked sardine
abundance and temperature (p = +0.69, P
< 0.05). Ranked larval anchoveta and sardine
abundances are not correlated (p = -0.07).
The only apparent warm-cold year abundance
pattern among the OL categories is that of the
coastal taxa; this group had lowest abundances
during the 1965, 1972, and 1983 El Nihos and
highest abundances in subsequent 1966 and 1973
transition years. The ranked abundance pattern
of this category has negative correlations
(P < 0.05) with ranked temperature (p = -0.69)
and salinity (p = -0.68) values (Table 9). Both
the myctophid and other mesopelagic categories
appear to have abundance patterns unrelated to
warm-cold hydrographic conditions (Table 9).
Table 9. — Correlations of across-year abundance ranks of
zooplankton and ichthyoplankton categories with ranked (high to
low) mean temperature and salinity values from nine July-Septem-
ber sampling periods off northern Chile, 1964-83. Correlations
based on Spearman's rho tests. Significant values at P s 0.05 are
indicated, but note use of multiple testing. PL = larvae of pelagic
species; OL = other larval taxa.
Temperature
Salinity
(^■C)
(%o)
Zooplankton
-0.48
-0.37
PL
-0.10
-0.17
OL
-0.02
+ 0.05
Larval diversity:
(mean no. taxa/tow)
+ 0.12
0.00
Myctophids
+ 0.02
+0.12
Other mesopelagic
taxa
+ 0.28
+ 0.12
Coastal taxa
-0.69 P<0.05
-0.68 P = 0.05
Zooplankton biomass values show negative but
nonsignificant correlations with temperature and
salinity (Table 9). This is in agreement with the
time-series analysis results of Bernal et al. (1983)
which demonstrated no consistent relations of
zooplankton biomass with cold- or warm-water
events.
Species Groups and
Hydrographic Conditions
The species groups formed by similarity of
between-year abundance ranks demonstrate both
positive and negative correlations with cold- and
warm-year conditions (Table 10). Group I and
Pair II and their member species have negative
correlations with ranked temperature and
salinity values indicating a tendency for higher
14
LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE
surface temperature (°C) and salinity (%o) values during July-September
available Hydrographic "condition" descriptions for these penods are from Robles et al.
1969
1970
1972
1973
1983
Temperature ("C)
N
35
N A
110
43
24
Range
14.6-17.5
15.5-18.5
14.3-17.4
15.1-19.0
X
16.0
17.2
15.8
17.2
(S)
(1.0)
(0.7)
(0.9)
(0.9)
Rank
5
1.5
6
1.5
Salinity (%o)
N
35
N.A.
108
43
20
Range
34.60-35.19
34.75-35.96
34.60-35.13
34 81-35.39
X
34.87
35.05
34.82
35.09
(S)
(0.16)
(0.18)
(0.13)
(0.19)
Rank
5
3
7
1
Hydrographic
condition:
Warm
Cold
El Nirio
Transition
El Nino
Table 10. — Correlations of across-year abundance ranks of larval
fish species groups and member species with ranked (high to low)
mean temperature and salinity values from nine July-September
sampling periods, 1964-83. Correlations based on Spearman's
rho tests. Group correlations based on ranks of summed within-
year ranks of member species. Significant values at P < 0.05 are
indicated but note use of multiple testing. " = abundance ranks
adjusted to accommodate for apparent large sampling depth-
related catch differences.
Temperature
Salinity
Species
(C)
(%o)
Group 1
-0.78
P < 0.05
-0.78 P<0.05
Engraulis nngens
-0.52
-0.62
Normanichthys crocken
-0.69
P < 0.05
-0.72 P<0.05
Blenniid D
-0.74
P < 0.05
-0.77 P<0.05
Sebastes sp.
-0.67
-0.55
Group II
+ 0.27
+ 0.20
Diogenichthys laternatus
+ 0.45
+ 0.50
Bathylagus nigngenys
+ 0.27
+ 0.20
Vinciguerna lucetia
+ 0.12
-0.14
Pair 1
+ 0.70
P < 0.05
+ 0,49
Sardinops sagax
+ 0.69
P < 0.05
+ 0.50
Myctophum nitidulum
+ 0.61
+ 0.35
Pair II
-0.62
-0.57
Lampanyctus parvicauda
-0.68
P = 0.05
-0.48
Tnphoturus mexicanus"
-0.50
-0.53
abundances during colder, lower salinity periods.
The rankings of Group I (based on ranks of
summed within-year member species ranks) are
significantly correlated (P < 0.05) with both
temperature and salinity (p = -0.78 in both
cases). Within this group the rankings of
Normanichthys crockeri and Blenniid D are cor-
related (P < 0.05) with temperature and salinity
(p = -0.69 to -0.77); the correlation of Sebastes
with temperature is also relatively strong
(p = -0.67). Within Species Pair II, Lampanyctus
parvicauda abundance has a relatively strong
negative correlation with temperature (p =
-0.68).
Group II and Pair I and their member species
have positive correlations of ranked abundance
with temperature and (with one exception) salin-
ity values (Table 10) suggesting a tendency for
higher abundances during warmer, higher salin-
ity conditions. These correlations are all non-
significant and generally weak for Group II and
its member species (Bathylagus nigrigenys , Vin-
ciguerria lucetia, and Diogenichthys laternatus).
Pair I has a positive correlation (P < 0.05) with
temperature (p = +0.70) primarily due to sardine
abundance ranks.
Species Percentage Composition
Relative to Hydrographic Conditions
Ichthyoplankton percentage composition shows
no striking warm year vs. cold year related
patterns. Total ichthyoplankton composition
comparisons between years of "similar" hydro-
graphic conditions do not give overall higher PSI
values than do comparisons between years of dif-
ferent conditions (Table llA). PSI values (ranges,
means, and proportions of high and moderate val-
ues) from comparisons of cold, transition, and
warm years are similar. However, highest values
(91.6-94.9) come from cold vs. transition year
(1970 vs. 1966 and 1973) and between-transition
year (1966 vs. 1973) comparisons. Additionally,
intercomparisons of the transition, warm, and
El Niilo years give relatively lower values than
do cold-year comparisons. Comparisons between
El Nino years give generally PSI low values.
15
FISHERY BULLETIN: VOL. 86, NO. 1
Table 11. — Within- and between-hydrographic period ichthyoplankton composition
comparisons presented as range, mean, and standard errors of percent similarity index
(PSI) values and numbers (N) out of total comparisons hiaving moderate to high (e.g.,
-65) values. A. Total larvae. B. OL (other larval taxa) fraction.
PSI values for
Cold
Transition
Warm
El Nino
A. Total larvae
Cold years
Range
67 8-822
62.7-94.9
62.8-82.5
7.3-79.1
(1964, 1967, 1970)
X
76.6
79.4
73.0
46.5
(SE)
( 4.5)
( 5.2)
( 3.6)
( 9 6)
N
33
5/6
4 6
49
Transition years
Range
57.8-65.1
3.3-79.4
(1966, 1973)
X
91.9
625
39.7
(SE)
( 1-6)
(12.9)
N
1/1
1/4
2/6
Warm years
Range
188-76.4
(1968, 1969)
X
(SE)
79.4
52.3
( 9.2)
N
1/1
2/6
El Nmo years
Range
220-45.3
(1965, 1972, 1983)
X
(SE)
N
31.9
( 7.0)
03
B. OL
Cold years
Range
46 0-758
30 5-68 3
47.0-85,2
332-88 1
(1964, 1967, 1970)
X
58.6
48.0
64.6
57.6
(SE)
( 8,9)
( 5.7)
( 5.4)
( 5.9)
N
1,3
1/6
3/6
3/9
Transition years
Range
36.8-67.9
18 5-66.2
(1966, 1973)
X
37.5
51.7
41.0
(SE)
( 6.4)
( 7.8)
N
0/1
1 4
1 6
Warm years
Range
37,2-68,4
(1968, 1969)
X
(SE)
58.5
53,1
( 5.2)
N
0/1
1/6
El Nino years
Range
54.3-58.7
(1965, 1972. 1983)
X
(SE)
N
55.8
( 1.5)
03
The OL percentage composition similarly does
not demonstrate clear hydrographically related
patterns (Table IIB). As with the total larvae,
comparisons of cold vs. other years yield most of
the moderate to high PSI values. Highest values
(88.1 and 85.2) come from comparisons of cold vs.
El Nino (1965 vs. 1967) and cold vs. warm (1964
vs. 1969) years. Comparisons within and between
transition, warm and El Nino years give primar-
ily low values. Recalculation of PSIs to accommo-
date for Triphoturus mexicanus lowers mean val-
ues for comparisons with 1973 and 1983 by only
0.4-2.3 and does not affect the overall results.
Chronological Considerations of
Species Composition
When the total larval and OL PSI data are con-
sidered in terms of chronological rather than hy-
drographic periods, various patterns become ap-
parent (Table 12). For the total ichthyoplankton,
comparisons within the 1964-69 data set and be-
tween this and the 1970-73 data set give similar
means, ranges, and proportions of moderate to
high values. In contrast, comparisons within the
1970-73 data set provide more similar values and
a significantly higher mean value than results
from comparisons within the 1964-69 set (Z test,
P < 0.01). This suggests that, despite the varied
hydrographic conditions represented during the
1970-73 period, conditions were favorable for a
repeated fairly similar anchoveta-dominated
ichthyoplankton assemblage during July-
September months.
Chronologically grouped comparisons of the OL
fraction provide somewhat different patterns
from those of the total ichthyoplankton (Table
12B) and indicate a marked change in species
16
LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE
proportions between 1964-69 and later years.
Comparisons within the 1964-69 OL data set
yield all of the high and most of the moderate PSI
values. The mean PSI value from these compari-
sons is significantly higher (P < 0.05) than that
from 1964-69 vs. 1970-73 comparisons. As with
the total larvae, the 1970-73 PSI values are sim-
ilar and moderately high and the 1983 compari-
son values are relatively low compared with other
Table 12. — Within- and between-time period ichthyoplank-
ton composition comparisons presented as ranges,
means, and standard errors of percent similarity index
(PSI) values and numbers (A/) out of total comparisons
having moderate to high (e.g.. -65) values. A. Total larvae.
B. OL {other larval taxa) fraction.
PSI values for
1964-69
1970-73
1983
A. Total larvae
1964-69
Range
33 9-825
38 2-91.9
3 3-31.0
X
67.4
67.7
18.2
(SE)
{ 3.6)
( 3.6)
( 4.3)
N
9 15
11 18
06
1970-73
Range
79.1-95.0
7.3-22.0
X
84.4
12.8
(SE)
( 5.2)
( 4.7)
N
3/3
03
B. OL
1964-69
Range
29.2-88.1
26.2-67.9
18.5-54.4
X
60.2
48.4
40.7
(SE)
( 4.6)
( 2.5)
( 5.6)
N
715
1 18
06
1970-73
Range
66.2-74.4
58.6-60.2
X
69.7
592
(SE)
( 2.4)
( 0.5)
N
33
0/3
years. Accommodation for T. mexicanus abun-
dance reduces slightly (1.1-3.1) the mean values
from comparisons with 1970-73 and 1983 data
sets and strengthens the significance of difference
(P<0.01) between 1964-69 and 1964-69 vs.
1970-73 mean values.
The difference in OL PSI values between 1964-
69 and 1970-83 is, to a great extent, due to abun-
dance shifts of Group II and Pair II species. The
abundance ranks of all three Group II species
(Table 7) indicate significantly higher abun-
dances during 1970-83 than in earlier years
(Man Whitney U tests, all P's < 0.05). Addition-
ally, the averaged abundance estimates from
these four years are significantly higher than
from earlier years (Z tests; P < 0.01 for Bathyla-
gus nigrigenys and Vinciguerria lucetia,P < 0.05
for Diogenichthys spp.). Together B. nigrigenys ,
V. lucetia, and Diogenichthys spp. (primarily D.
laternatus ) contributed 46.5-72.8^7^ of the OL col-
lected during 1970-83 compared with 5.8-37.1%
during 1964-69 (Table 13).
In contrast. Pair II species Lampanyctus parvi-
cauda and Triphoturus mexicanus were rela-
tively less abundant during 1970-83 than in pre-
vious years. These two species contributed
21-66% of the OL during 1964-69 compared with
<22% during 1970-83 (Table 13). With adjust-
ments to T. mexicanus abundance these species
proportions in 1973 and 1983 decrease to 11.8%
and 13.2%, and those of the Group II species in-
crease to 49.9% and 71.4%, respectively. The rela-
tive abundance decrease of Pair II was primarily
Table 13.— Percentage contribution by dominant OL (other larval taxa) species collected
sampling penods, 1 964-83. Species arranged according to group affiliations based on across
patterns.
during July-September
■year ranked abundance
Percentage contribution for
1964
1965
1966
1967
1968
1969
1970
1972
1973
1983
Group 1
Normanichthys crocken
Sebasles sp.
Blenniid D
1561
0.36
4.81
1.94
0.52
67.40
060
4.00
3.51
0.63
4.45
5.15
1.79
940
22.13
2.16
12.52
1.02
1.19
15.72
0.35
4.31
0.44
Group total
20.78
2.46
72.00
8.59
16.34
24.29
12.52
2.21
20.38
0.44
Group II
Bathylagus nigrigenys
Vinciguerria lucetia
Diogenichthys spp.
3.63
2.81
5.63
6.60
0.52
24.45
1.61
0.70
350
608
3.82
19.11
13.87
6.94
16.33
0.75
2.16
12.15
9.01
26.78
26.78
24.14
23.37
25.29
17.53
9.31
19.68
13.78
1.62
49.96
Group total
12.07
31.57
5.81
2901
37.14
15.06
62.57
72.80
46.52
65.36
Pair 1
Myctophum nitidulum
0.36
—
0.10
—
0.89
0.75
—
2.05
2.15
2.01
Pair II
Lampanyctus parvicauda
Triphoturus mexicanus
37.21
28.86
41.40
20.44
15.77
4.99
3947
18 17
24.16
10.29
32.11
26.37
7.13
14.27
12.67
8.40
6.06
11.82
6.78
14.70
Pair total
66.07
61.84
20,76
57.64
34.45
58 48
21.40
21.07
17.88
21.48
17
FISHERY BULLETIN: VOL. 86. NO 1
due to the increased numbers of Group II species
and decreased numbers of L. parvicauda . The
1970-83 averaged abundance of L. parvicauda is
significantly lower (Z test; P < 0.01) than that of
1964-69. Triphoturus mexicanus averaged abun-
dance (both adjusted and unadjusted values)
is similar (P > 0.05) between the two time peri-
ods.
DISCUSSION
The northern Chilean ichthyoplankton data set
is obviously weakened by lack of information
from the 1974-82 period; this missing informa-
tion is critical for an appreciation of the temporal
extent and relative constancy of the apparent
ichthyoplankton composition change in 1970-73
vs. earlier years. This data set also suffers from
limited seasonal coverage which prohibits exami-
nation of between-year variations in spawning
time and intensity as a cause of interannual
abundance fluctuations and apparent composi-
tion change. However, the existing data set does
provide coherent coverage over varied hydro-
graphic conditions between 1964 and 1973 and is
sufficient to test for correlations with short-term
(e.g., year to year) fluctuations in hydrographic
conditions.
The large interannual changes in abundance
and composition of the northern Chilean ichthy-
oplankton can to a certain extent be related to
interannual changes of hydrographic conditions
in the Humboldt Current. This has been demon-
strated through correlations of ranked tempera-
ture and salinity values and abundances of
coastal species, sardine, and Lampanyctus parvi-
cauda (Tables 9, 10). The temperature and salin-
ity values used in these correlation tests
represent ambient conditions during the July-
September spawning period and therefore
possibly reflect only conditions affecting egg and
early larval (e.g., to stages capable of substantial
net avoidance) survival. These values do not nec-
essarily reflect longer term conditions affecting
abundance, distributions, and fecundities of adult
populations or later larval survival and recruit-
ment. However, there is a generally good corre-
spondence between these values and reported
longer term hydrographic conditions in the Hum-
boldt Current over the 19-yr timespan (e.g.,
Table 8; Robles et al. 1976; Bernal et al. 1983;
Guillen 1983; Bakun 1987).
Despite significant correlations between abun-
dances of some ichthyoplankton components and
temperature and salinity values, there is no ap-
parent consistency of total larval or OL species
percentage compositions during years of
"similar" hydrographic conditions (Table 11).
More coherent patterns emerge from consider-
ations of the 1964-69 and 1970-73 data sets
(Table 12). This chronological separation is also
supported by the ranked abundance patterns of
the various species groups and pairs {Table 7).
Among the least confusing across-year abun-
dance patterns demonstrated by the ichthy-
oplankton are 1) generally greater abundance
of Group II species after 1969, 2) greatest abun-
dance of Pair I species after 1970, and 3) pre-
dominantly higher abundances of one of the Pair
II species prior to 1970 (Table 7). Associated with
the Group II and Pair II abundance patterns are
large shifts in their relative proportions
(Table 13).
The shift from relatively large percentage con-
tributions by Lampanyctus parvicauda and
Triphoturus mexicanus to larger proportions of
Diogenichthys spp., Bathylagus nigrigenys , and
Vinciguerria lucetia after 1969 is notable. The
abundances of these mesopelagic species, unlike
those of anchoveta and sardine, are not directly
influenced by man's fishing activities and so may
be interpreted as indicators of environmental
change. Furthermore, the timing of these species
absolute and relative abundance changes pre-
ceded by several years the dramatic changes in
anchoveta and sardine stocks off of northern
Chile (Fig. 4) and so cannot be directly related to
biological consequences of change in the domi-
nant pelagic schooling fish stocks.
Although fragmentary, there is evidence for a
change in zooplankton biomass values off north-
ern Chile (18°-24°S) occurring in 1969 (Fig. 5)
which, like OL percentage composition, suggests
a possible environmental change. Time series
analysis of quarterly zooplankton biomass values
during 1964-73 indicate generally lower biomass
during 1969-73 relative to the 1964-68 period.
As with total larval abundance (Table 9), these
zooplankton biomass variations do not appear to
be related to warm year-cold year events (Bernal
et al. 1983).
The changes in OL composition and zooplank-
ton biomass suggest that there was subtle but
large-scale (low-frequency) environmental tran-
sition occurring in the 1969—70 period. Various
indications of environmental change occurring
about this time are present in long-term physical
data bases from Chile and Peru. Predominantly
18
LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE
3,0
1955
83 84
Figure 4.— Total catch and catch of dominant species taken in the northern Chilean pelagic fishery, 1955-84.
negative sea surface atmospheric pressure
anomalies occurred off of Arica from 1960 to 1972
with a strong negative anomaly occurring in
1969; predominantly positive anomalies occurred
there after 1972 (Fig. 6A). A similar but less ex-
treme change from negative or neutral anomalies
to predominantly positive anomalies occurred off
Iquique (20"S) in 1970 (Fig. 6B; Kelly and Blanco
1983).
Off Peru (5°-15°S) the wind driven turbulent
mixing index of surface waters shows a general
increase during and after the 1972 El Nino event
(Fig. 7A; Bakun 1987). A probable result of this
increased turbulence is an increase in standard
deviations associated with monthly temperature
values; standard deviations above the 30-yr mean
generally persisted throughout the year from
1972 to 1984 (Fig. 7B) and suggest increased
physical variability and heterogeneity in this
later period. Comparable data sets from northern
Chile are not available to determine if these lat-
ter two features were also characteristic of the
Chilean area.
How these observed atmospherically related
changes could be related to changes in the marine
environment off northern Chile is uncertain. It is
possible that the observed changes in atmos-
pheric pressure off Arica and Iquique have associ-
ated changes in advection of water mass and fau-
nal sources. Bernal et al. (1983) discussed El Nino
related changes in water mass distribution off
Chile in 1972 and 1973 relative to cold-year 1967.
These changes involved southerly extensions of
oceanic subtropical and equatorial subsurface
waters, strengthening of the spring-summer ther-
mocline, and cessation of coastal upwelling.
These authors did not examine water mass distri-
butions in the 1968-70 period. However, lowered
zooplankton biomass starting in 1969 and the OL
composition change starting around 1970 suggest
that the hydrographic conditions attributed to the
1972 El Nino may have been an intensification of
19
w
ZOOPLANKTON BIOMASS
W
FISHERY BULLETIN: VOl- 86, NO 1
W
2 0
1 .0
o
■o
c
o
to
I 0 -
- 2 0
—
— 1
1
1 1
1
-
19
64
9
S5
1966
1967
1966
969
1970
1971
972
1973
Year
Figure 5. — Time series estimates of quarterly zooplankton biomass values from northern Chile
(18°-24°S), 1964-73, standardized accordmg to the long term standard deviation. W = warm
years; C = cold years. From Bernal et al. 1983.
SEA SURFACE
ATMOSPHERIC PRESSURE ANOMALIES
3 -
3 -
3 -
2 -
B
-
1 -
/^ 1
V^
\/
L-A
/V
r^
A
r\
f
V
A
\
\
\
r\
\ ''
yt
\'
V
1 -
0
J ^
V
H
V^
/•M
-J
V \H
'V
1
^
yl
V
\
I960 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82
Year
Figure 6. — Sea surface atmospheric pressure anomalies off of (A) Arica (18'Sl and (B) Iquique (20''Sl, 1960-82. From
Kelly and Blanco 1983.
20
LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE
TURBULENT MIXING INDEX
350 -
300 -
250 -
200 -
55
65 70
Year
STANDARD DEVIATIONS
4 0-
3 5
3 0
2 5
2.0
I 5
1.0
B
Std. dev
12- mo- running meon of std. dev.
.— Long - Itrtn meon of tid. dev.
85
Yeor
Figure 7.— (A) Surface layer turbulent mixing index values and (B) standard deviations and 12-mo running mean of standard
deviations associated with mean monthly sea surface temperature values off of Peru (5°-15°S), 1953-84. Dotted lines represent
long-term mean values of each index. From Bakun 1987.
conditions initiated during the warm 1968-69 pe-
riod; these could possibly have persisted and in-
tensified again during the 1976 and 1983 El Nino
events.
It is also likely that the 1970 change in OL
composition is related to onshore advection of
northern or oceanic water masses and associated
faunal assemblages, but this cannot be confirmed.
All involved species are relatively abundant in
coastal Peruvian and south eastern tropical
Pacific waters (Ahlstrom 1971, 1972; de Castillo
1979; Santander and de Castillo 1979) and in
more southern coastal Chilean waters (Table 2),
but their wintertime relative abundances in these
areas have not been documented. It is also possi-
ble that the changes in species composition are
related to locally lowered zooplankton abundance
(e.g., that the Group II species are relatively more
successful than Lampanyctus parvicauda and
Triphoturus mexicanus during periods of lowered
secondary productivity levels). Alternatively, the
observed change could be due to altered seasonal
spawning activity which is not treated in the
present study. However, whatever the cause,
there is evidence for an environmental change
in the study area, and this may be also implicated
in changes occurring within the PL ichthyoplank-
ton fraction.
Increased abundances of sardine and Mycto-
phum nitidulum (Pair I species) during and after
1972 may be further evidence for a changed
marine environment off of northern Chile. Addi-
tionally, the significant correlation between lar-
val sardine abundance and temperature
(Table 10) suggests that elevated temperatures
may have been important for increased spawning
activity and/or increased success of hatching and
early larval survival. Given this observation one
may speculate that the increased sardine catches
after 1973 (Fig. 4) are related to increased fre-
quency of warm-water events in the 1964-84 pe-
riod relative to earlier years. Increased sardine
21
FISHERY BULLETIN: VOL. 86, NO. 1
catches in 1973 resulted from apparently ex-
tremely good survival and recruitment of individ-
uals spawned during the warm 1968-69 period
(Serra"^). Similarly, good survival of the large
1972 El Nino spawn could explain the huge catch
increases in 1976 and later years (Fig. 4). How-
ever, mean biomass estimates of age groups con-
tributing to the Chilean fisheries catch from 1974
to 1981 (Serra 1983) indicate increasing contribu-
tions after the 1967 year class with marked in-
creases beginning with the 1970 year class; this
suggests that factors other than temperature
(e.g., environmental change starting in 1969-70)
may also be responsible for increased larval sur-
vival and recruitment. A possible cause is in-
creased nearshore influence of equatorial and
subtropical waters (Santander and Flores 1983).
Because the sardine abundance increase was ini-
tiated prior to the 1972 anchoveta decline off
northern Chile (Fig. 4), it is difficult to implicate
reduced anchoveta competition as the cause of the
early sardine population growth in this area.
The grouping of anchoveta with three coastal
species (Group I), and significant correlation of
anchoveta and Normanichthys crockeri larval
abundances are extremely interesting and imply
that the spawning intensity and/or early stage
survival of these four species are influenced in
similar ways by interannual changes off northern
Chile. Unfortunately, little is known about the
natural histories or population abundances of the
coastal species. Because of the group composition,
it is logical to suspect that coastal processes are
important factors influencing their larval abun-
dance. The significant negative correlation of the
group as a whole, and of two of the coastal species,
with ranked temperature and salinity values
(Table 10), suggests that coastal upwelling and/or
increased coastal influence by subantarctic
waters, and theoretically enhanced food supplies,
are important factors.
Given the present data set and information
from recent publications, a case can be made for a
low-frequency environmental change influencing
the abundances of anchoveta and sardine larvae
as well as the larvae of coastal and mesopelagic
species during the 1964-84 period. The Chilean
OL composition suggests an environmental
change (e.g., an atmospherically related oceanic
circulation change) starting with the 1968-69
warm-water event. This coincided with apparent
successful survival of sardine larvae and
markedly increased recruitment by 1968 and
later year classes despite varied warm water-cold
water events between 1968 and 1973. Physiologi-
cal anomalies of Peruvian anchoveta stocks in
1971 suggest that these fishes may have experi-
enced environmental change at that time. Unusu-
ally low proportions (e.g., 40% vs. typically 90%)
of potential spawning-sized fish were sexually
mature during the 1971 spawning season and fat
content of the 1971-72 catch was anomolously
high, indicating unusually low transfer of body
fat to gonadal products (Sharp 1980). Starting
with the 1972 El Nino was 1) an obvious in-
creased incidence of penetration of subtropical
surface waters toward the Peruvian coast, 2) co-
incidental onshore and southward expansion of
sardine spawning activity off both Peru and
Chile, 3) southward expansion of Peruvian an-
choveta spawning activity into new spawning
areas between 14°S and 18°S (e.g., to northern
Chile), and 4) a succession of years of poor an-
choveta larval survival off Peru and Chile (San-
tander and Flores 1983; Serra 1983). Environ-
mental conditions favorable for growth of sardine
populations, as well as of mackerel and jack
mackerel populations, off both Chile and Peru
have persisted since the early to mid-1970's (San-
tander and Flores 1983; Serra 1983).
The lack of Chilean ichthyoplankton data from
the 1974-82 period precludes evaluation of the
constancy of altered species composition during
that time. However, there are indications that
change is once more occurring off northern Chile.
Preliminary analysis of ichthyoplankton samples
collected between Arica and Antofagasta during
4-14 August 1985 indicates a clear dominance by
anchoveta larvae at a markedly higher mean
abundance level than encountered in the 1964-
83 samples; sardine and Trachurus larval abun-
dances are comparable to those in the 1973 sam-
ples (Table 14). The other species have not yet
been analyzed, but Normanichthys crockeri is
Table 14. — Mean abundance estimates and standard errors (num-
bers per 10 m2) and percent frequency of occurrence (F) of PL
taxa, OL and total larvae collected in 81 1 00-0 m WP2 net samples
off northiern Chile (18°-24°S) during 4-24 August 1985.
Species
(SE)
(F)
3Serra, R. Unpubl. manuscr. Subsecretaria de Pesca,
Teatinos 120, Piso 11, Of. 44, Santiago, Chile.
Engraulls ringens 5,535.3 (1,844.1) (90.4)
Sardinops sagax 63.8 ( 24.9) (26.6)
Trachurus murphyi 1.2 ( 0.9) (2.1)
Otfier species 233.7 ( 28.4) (92.5)
Total 5,834.0 (1,838.2)
22
LOEB and ROJAS: ICHTHYOPLANKTON COMPOSITION AND ABUNDANCE
once again noted to be among the abundant OL
taxa. It will be of great interest to see the recruit-
ment resulting from this 1985 anchoveta spawn.
If the 1985 anchoveta year class is relatively suc-
cessful during this period of continued large sar-
dine stocks, it will lend support to the idea that
the anchoveta-sardine population fluctuations
have been primarily regulated by low-frequency
hydrographic events. In a long-term context such
events may not be unusual to the Humboldt Cur-
rent area. Fish scales present in sediment records
from coastal Peruvian waters indicate that sar-
dine replacements of typically dominant an-
choveta stocks have occurred at infrequent inter-
vals over the past 11,000 years (De Vries and
Pearcy 1982). In light of this, the ecological
events of the past 20 years may be naturally oc-
curring, physically mediated, but probably fish-
eries enhanced, fluctuations in this hydrographi-
cally complex region. Because of our limited time
reference, we have not previously acknowledged
such fluctuations as being "normal".
ACKNOWLEDGMENTS
We extend our appreciation to Dick Parrish and
Andy Bakun, Pacific Fisheries Environmental
Group (SWFC/NMFS/NOAA), whose interest in
eastern boundary current comparison studies has
made this work possible. We also thank Rodolfo
Serra and Gary Sharp for their valuable com-
ments and discussions of the manuscript. Appre-
ciation is extended to numerous other people in-
cluding ship captains and crew and laboratory
assistants, especially Hernan Miles, who have
helped in sample collection and processing over
the 20-yr period.
This work was funded by NMFS/NOAA Solici-
tation WASC-84-00075.
LITERATURE CITED
Ahlstrom, E. H.
1971. Kinds and abundance of fish larvae in the eastern
tropical Pacific, based on collections made on
EASTROPAC I. Fish. Bull., U.S. 69:3-77.
1972. Kinds and abundance of fish larvae in the eastern
tropical Pacific on the second multivessel EASTROPAC
survey, and observations on the annual cycle of larval
abundance. Fish. Bull., U.S. 70:1153-1242.
Bakun, A.
1987. Monthly variability in the ocean habitat off Peru as
deduced from maritime observations, 1953-84. In D.
Pauly and I. Tsukayama (editors), the anchoveta and its
ecosystem, p. 46-74. International Center for Living
Aquatic Resources Management (ICLARM), Manila.
BERNAL, p. a., F. L. RoBLES, AND O. RoJAS.
1983. Variabilidad fisica y biologica en la region merid-
ional del sistema de corrientes Chile-Peru. FAO Fish.
Rep. 291:683-711.
CONOVER, W. J.
1971. Practical nonparametric statistics. John Wiley
and Sons, N.Y., 462 p.
DE Castillo, O. S.
1979. Distribucion y variacion estacional de larvas de
peces en la costa Peruana. Inf. Inst. Mar Peru 63:1-
32.
De Vries. T J . and W G. Pearcy.
1982. Fish debris in sediments of the upwelling zone off
central Peru: a late quarternary record. Deep-Sea Res.
28(1A):87-109.
DiXON. W J., AND F. M. Massey, Jr.
1969. Introduction to statistical analysis. McGraw-Hill,
N.Y., 638 p.
Guillen, O
1983. Condiciones oceanograficas y sus fluctuaciones el el
Pacifico sur oriental. FAO Fish. Rep. 291:607-658.
Kelly, R , and J. L. Blanco.
1983. Fluctuaciones ambientales y su relacion con la
abundancia de recursos pelagicos en la zona norte-centro
de Chile. Inst. Fom. Pesq. 830040, 22 p.
Parrish. R H , A Bakun. D M Husby, and C S Nelson.
1983. Comparative climatology of selected environmental
processes in relation to eastern boundary current pelagic
fish reproduction. FAO Fish. Rep. 291:731-777.
Robertson. A
1970. An improved apparatus for determining plankton
volume. Fish. Bull., S. Afr. 6:23-26.
ROBLES, F L , E ALARCON. AND A ULLOA.
1980. Water masses in the northern Chilean zone and
their variations in the cold period (1967) and warm
periods ( 1969, 1971-73). Proceedings of the workshop on
the phenomenon known as "En Nino", p. 83-174.
UNESCO.
Sameoto. D
1980. Distribution and abundance of six species of fish
larvae in Peruvian waters and their relationship with the
physical and biological environment. Bol. Inst. Mar
Peru Callao 5:164-170.
Santander, H , and O S de Castillo
1979. El ictioplancton de la costa Peruana. Bol. Inst.
Mar Peru Callao 4:69-112.
Santander. H . and R Flores
1983. Los desoves y distribucion larval de cuatro especies
pelagicas y sus relaciones con las variaciones del ambi-
ente marino frente al Peru. FAO Fish. Rep. 291:835-
867.
Serra. J R
1983. Changes in the abundance of pelagic resources
along the Chilean coast. FAO Fish. Rep. 291:255-284.
Sharp. G D
1980. Report of the workshop on effects of environmental
variation on survival of larval pelagic fishes. In G. D.
Sharp (editor). Workshop on the effects of environmental
variation on the survival of larval pelagic fishes, Lima,
Peru, April-May 1980. Workshop Report No. 28, p. 15-
59. Intergovernmental Oceanographic Commission,
UNESCO, Paris.
Tate. M W , and R C Clelland
1957. Nonparametric and shortcut statistics in the social,
biological and medical sciences. Interstate Printers and
Publishers, Danville, IL, 171 p.
23
FISIIKRY BUI.LKTIN VOL HH. NO 1
UNESCO Wyktki, K
1968. Zooplankton sampling. Monographs on oceano- 1967. Circulation and water masses in the eastern equa-
graphic methodology 2. Imprimeries Fopulaires, torial Pacific Ocean. Int. J. Oceanol. Limnol 1:1 17-147.
Geneva. 174 p. Yashnov, V A
WmiTAKKK. K H 1959. A new model of a volume meter for rapid and pre-
1975. Communities and ecosystems MacMillan Pub- cise plankton evaluation undt-r field conditions. Zool.
lishing Co., N.Y., 385 p. Zh. (Moscowl 38:1741-1744.
24
ESTIMATION OF NATURAL MORTALITY IN FISH STOCKS:
A REVIEW
E. F VetterI
ABSTRACT
The instantaneous rate of natural mortality (M) is an important but poorly quantified parameter m
most mathematical models of fish stock dynamics. This report reviews methods used commonly to
estimate M for fish stocks, sensitivity of some common fishery models to values chosen for M, and
evidence refuting the common assumption that a constant value can be an adequate approximation
of A/ within single stocks.
With the exception of simple surplus production
models (e.g., Schaefer 1954; Pella and Tomlinson
1969) all mathematical models offish stock dy-
namics include as a parameter the instantaneous
rate of natural mortality (M). The models do not
require explicitly any particular form for M; it
can be constant or can vary in any imaginable
form. But because natural mortality has proved
extremely difficult to measure directly, M is as-
sumed almost universally to be some constant
specific to whatever stock is being modeled. This
is particularly true for analyses of commercial
fish stocks, which often require estimates of M
only for the postrecruit ages. Decreases in natural
mortality with increasing age during egg and
postlarval stages are so dramatic compared to ap-
parent changes during postrecruitment ages (e.g.,
Gushing 1975) or compared to differences be-
tween different sexes, collection sites, seasons,
years, cohorts, or stocks within species, that vari-
ations in M during these later (postrecruitment)
ages are often assumed negligible.
Whether this assumption is in fact acceptable is
the subject of this report. The answer is no, it is
probably not acceptable in most cases. That an-
swer follows from the information presented in
Sections II through V, with the following conclu-
sions:
Section II: Current methods for estimating nat-
ural mortality: a review of methods used cur-
rently to estimate M in fish populations. All of
these methods have strong limitations or disad-
vantages.
^Southwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038.
Manuscript accepted October 1987.
FISHERY BULLETIN: VOL. 86, NO. 1. 1988.
Section III: Sensitivity of fishery models to
choices for M: a review of the sensitivity of some
standard fishery models to different choices for
input value(s) of M. Existing studies show that
the models are sensitive and that sensitivity is
affected not only by the values chosen for M, but
also by interactions between M and the values
chosen for other parameters in the models.
Section IV: Evidence for nonconstant M; factors
influencing death rate: a review of factors as-
sumed or shown to affect M in fish stocks. Avail-
able information implies that many such factors
exist, acting alone or in concert.
Section V: Evidence for nonconstant M; vari-
ability within and between groups: a review of
existing quantitative evidence for the extent of
variability in M between but especially within
stocks. Because almost all fishery models focus on
single stocks, variability within stocks (as op-
posed to between stocks) is the most important
question. Some studies show strong differences
between mortality rates of various groups offish;
some do not. Those which do not have tended to
assume there would be none, and have often used
catch curve regression analysis to derive a single
estimate from data combined over many groups
(usually years) of data. The few studies from
which it is possible to determine ranges of esti-
mates show differences of at least 50 to 100% be-
tween minimum and maximum estimates for sin-
gle groups (e.g., stocks) offish.
The report's major conlcusions are that natural
mortality is far from constant for many fish
stocks, and that this variability is extensive
enough that it should not be ignored. Analyses of
25
FISHERY BULLETIN: VOL. 86, NO. 1
fish stock dynamics need much more rigorous es-
timates of within-stock variability (both trends
and variance) in M for exploited fish stocks.
II. CURRENT METHODS FOR
ESTIMATING NATURAL MORTALITY
Three methods are used currently or have been
proposed to estimate M in fish populations:
1) analysis of catch data, usually from commer-
cial fisheries but also from sampling programs
specifically conducted for stock assessment (this
includes mark-recapture studies), 2) correlations
of M with other life history parameters, and
3) estimation of deaths due to predation. I de-
scribe below each method in turn, listing both
advantages and disadvantages of each.
Catch-Analysis Methods
Methods for deriving estimates of natural mor-
tality from catch data are based on measuring
decreases in abundance, either relative or abso-
lute, in groups of fish during two or more succes-
sive periods of time. Groups may be distinguished
on the basis of any identifiable characteristic,
e.g., size (length or weight), age, sex, location and
time of capture, or some identifiable tag or mark.
The most common grouping is by age, for two
reasons. First, age has been considered histori-
cally the most important factor potentially affect-
ing estimates of mortality rate and subsequent
results from the most commonly used fishery
models (e.g., Heincke 1913; Baranov 1918). This
is probably because the methods were developed
for temperate water fisheries which tend to have
obvious annual reproductive cycles, so that indi-
vidual year classes are often relatively easy to
distinguish. Second, the earliest method of esti-
mating M (catch curve analysis, discussed below)
depends on determining the rate during succes-
sive ages.
Regardless of the grouping criterion, methods
for estimating M use generally one of two types of
data. The first type is simply subsamples of un-
marked catch. These subsamples contain fish se-
lected randomly and classified into groups on the
basis of size (length or weight). The second type is
mark-recaptures, in which previously marked in-
dividual fish can be identified and classified after
recapture into groups on the basis of this positive
identification. Estimates of mortality are usually
derived from samples of unmarked fish by analy-
sis of resulting "catch curves" (Ricker 1975). Be-
cause it has been used so frequently, catch curve
analysis is discussed below in some detail.
With marking it is possible to follow the history
of individual fish, so many different types of esti-
mation procedures exist for deriving estimates of
mortality from mark-recapture data (e.g., Ricker
1975; Jones 1979; Brownie et al. 1985). Because
so many variations are possible, marking experi-
ments are discussed only generally, stressing
the basic advantages and disadvantages of mark-
ing data relative to data from unmarked samples,
in deriving estimates of mortality from these
data.
Size-frequency distributions from unmarked
subsamples of catch (the first type of data) are
converted usually to age-frequency distributions,
on the basis of previously determined relation-
ships between age and length or age and weight.
Subsequent analyses concentrate on analyzing
this resulting curve of age-composition (e.g.,
Ricker 1975). Abundance usually decreases expo-
nentially with size (or age) in this type of sample.
Converting the abundances to their logarithmic
values often results in a relatively linear decrease
during most exploited ages (or sizes), after some
initial increase in vulnerability. Graphs of these
logged-frequency distributions are usually called
"catch curves", and their analysis, "catch curve
analysis". "Catch curve analysis" generally con-
sists of determining the best-fit straight line
through the decreasing portion of the logged-
frequency distribution, because if the decrease in
abundance is truly exponential, the slope of this
line through the log-transformed data is the in-
stantaneous rate of decrease in abundance (e.g.,
Ricker 1975).
There are two basic t3T)es of catch curves, dis-
tinguished on the basis of when the data were
collected and how many groups are represented in
the curves. The first, horizontal catch curves, in-
cludes data from several groups (e.g., size or age
classes) collected at a single point in time (or com-
bined from two or more points in time). Thus,
horizontal catch curves reflect "ancient history".
The individuals contributing to the frequency dis-
tribution were not originally all members of the
same group. To use this type of catch curve, one
must assume that for each successive age, risk of
mortality has been historically the same for all
individuals achieving that age. If this has not
been the case, the catch curves may show various
types of curvature in the descending leg, but ab-
sence of curvature is no guarantee that the rates
have in fact been constant.
26
VETTER: NATURAL MORTALITY IN FISH STOCKS
Thus, horizontal curves are subject to the ex-
tremely restrictive assumption that the groups
from which the data were collected must be in
steady state relative to each other, i.e., their rela-
tive abundances must be constant through time.
If this is true, then a graph of data collected at a
single point in time, which may include, for ex-
ample, individuals from 5 consecutive year
classes displayed as frequencies at 5 consecutive
ages, will look the same as the 5 graphs of data
that will result from collecting samples during 5
consecutive years (ages) from each of the 5 year
classes. If these conditions are not met, simple
linear fitting to determine a single estimate for
mortality will be inappropriate.
The second type of catch curve, longitudinal,
includes data collected from a single identifiable
group over a protracted period of time. Most often,
this will be a single cohort of fish such as single
year class, sampled during successive years. Lon-
gitudinal curves are not subject to the assumption
of steady state, but do share with horizontal catch
curves several other severe disadvantages. These
include 1) groups must be adequately identifi-
able; 2) groups must be closed to migration, so
that changes in abundance are due only to fishing
or natural mortality, or if migration does occur, it
must occur in proportion to the age distributions
in the local groups; 3) samples must represent
adequately the true composition of the groups in
nature; 4) rate(s) of mortality must be relatively
constant between groups over time, so that the
log-transformed frequency distributions are truly
linear (e.g., Jensen 1984); 5) compensatory rela-
tionships between stock levels and natural mor-
tality, or fishing mortality and natural mortality,
must not be present.
Methods for estimating M, which assume to
greater or lesser degrees that the conditions listed
above are met, have been described repeatedly.
The methods tend to fall into two categories.
Methods in the first category estimate M from
catch records of unexploited or lightly exploited
groups of fish. In these groups, F equals or ap-
proximates zero. Therefore, the observed rate of
decrease (Z ) equals or approximates M, because Z
equals the sum of F and M (e.g., Heincke 1913;
Baranov 1918; Ricker 1947; Beverton and Holt
1957; Robson and Chapman 1961; Pauly 1982;
Munro 1982; and among others).
Methods in the second category estimate M by
determining Z at various levels of fishing effort,
then using the observed relationship between Z
and effort to predict, via regression analysis or
manipulation of various ratios, the value of Z
at zero effort (e.g., Silliman 1943; Beverton and
Holt 1957; Paloheimo 1961; Lander 1962; Chap-
man and Murphy 1965; Paulik and Robson 1969;
Gulland 1983; Butler and MacDonald 1979;
Fournier and Archibald 1982; Caddy 1984; and
others).
These methods are most appropriate for analyz-
ing catches of unmarked fish. Accurate results
depend strongly on accurate measures of catch
per unit effort (CPUE) and constant catchability
(q) because if these conditions (in addition to
those listed above) are not met, observed relation-
ships between abundances in different sample
groups may not reflect true differences between
groups in situ.
Marked fish present fewer problems. Advan-
tages include 1) concentration on measuring rela-
tive rather than absolute differences between
abundances of different groups, 2) immigration
need not be considered, as entire original groups
are known to carry marks, and 3) with suffi-
ciently large samples, it becomes possible to test
for differences in mortality rate between different
groups (e.g., between ages, between sexes, or be-
tween sampling sites), rather than having to as-
sume that such effects are negligible.
Reviews and descriptions of various mark-
recapture methods appear in papers by Seber
(1973), Ricker (1975), Jones (1979), and Brownie
et al. (1985). Some of the newer types of marking
analyses can solve many of the most vexing prob-
lems associated with traditional catch curve anal-
ysis (e.g.. Reed and Davies 1980; Hochbaum and
Walters 1984; Burnham and Andersen 1984;
Burnham et al. 1984; Brownie et al. 1985).
Several disadvantages unique to marking oper-
ations counteract these advantages, however,
even with the newer methods. These disadvan-
tages include various types of mark-induced
effects on mortality rates, behavior, and vulnera-
bility to capture, as well as mark loss, unrepre-
sentative mixing of marked fish with their origi-
nal groups prior to recapture (e.g., Ricker 1975),
and especially in commercial fisheries, under-
reporting or incorrect reporting of recaptures.
Both analysis of catch curves from unmarked
fish and analysis of mark-recapture data have the
advantage of requiring only catch (and usually
effort) data, and these data can generally be col-
lected by sampling catches from commercial fish-
eries. However, in addition to problems specific to
each method, they have in common one or more
other major disadvantages: 1) inability to distin-
27
FISHERY BULLETIN; VOL 86, NO 1
guish between losses (or gains) from migration or
recruitment versus losses due to fishing or natu-
ral causes, 2) imprecision in the estimates of M
obtained (e.g., Beverton and Holt 1957; Taylor
1958; Bishop 1959; Paloheimo and Dickie 1966
Ricker 1975, 1977; Doubleday 1976; Pauly 1980
Larkin and Gazey 1982; Paloheimo 1980, 1982
Myers and Doyle 1983; Roff 1984), 3) sensitivity
to size-specific mortality affecting the estimated
age-structure of the group (Ricker 1969), 4) errors
in estimates of age, such that abundances-at-age
derived from age-length conversions are unrepre-
sentative, 5) where analyses are conducted on
data combined over two or more cohorts, the un-
likely condition that mortality rates were in fact
similar for all cohorts, and 6) problems inherent
in the analyses themselves (e.g.. Barlow 1984).
Disadvantages 1, 4, and 5 may not apply to
marked fish. Disadvantage 5 does not apply to
single cohorts. But collections from marked
groups and single cohorts are still vulnerable to
the other problems.
Further, although in principle it would be pos-
sible to estimate M for different ages, times, or
places, most commonly in practice a single,
fishery-wide constant M is estimated by pooling
data from throughout the fishery. By implication,
the analyst is assuming that the exploited stock
was more or less in steady-state over all times and
areas of catch so that M was relatively constant
while that data set was collected and while (his-
torically) the observed age-distributions were
being created. In fact, substantial evidence exists
that M is not constant, either within a single
stock over time (age) or between stocks of a given
species in different areas (Sections IV and V).
A final disadvantage is that catch-curve analy-
ses are fundamentally unmechanistic, generated
simply by charting changes in abundance. Catch-
curve analyses cannot predict the effect of
changes in factors that control M; thus there is
little hope of predicting M in the future should
conditions change.
Life History Methods
A second approach to estimating the instanta-
neous rate of natural mortality in fish stocks is
based on the observation that M often correlates
strongly with life history parameters, such as
growth rate, age at sexual maturity, costs of re-
production, and maximum age (Table 1).
Typically in such studies, analytical formulas
are derived from theoretical relationships be-
tween the various parameters (e.g., Beverton
1964), or empirical formulas are derived from re-
gression of M against one or more of the parame-
ters (e.g., Hoenig 1983). These models have two
significant advantages: 1) they require minimal
amounts of data, and 2) they are useful in demon-
strating broad trends across species and in devel-
oping ecological theory. But because they produce
only a single and often very imprecise estimate of
M for any given group offish, they are not partic-
ularly effective for generating precise estimates
of natural mortality or for determining the exis-
tence or extent of trends and variability in M for
given stocks. They will also be no better than the
methods used to estimate the values of M used in
the regressions.
Table 1. — Studies relating instantaneous rate of natural mortality to life history traits in fish.
Traits
Species
Source
^^max. ^'f' ~''-inf' nietabolic rate,
Various
Beverton and Holt 1959
reproduction
^max. !<• '-inf. "'-asm. f'Shing
clupeids, engraulids
Beverton 1963
^^V,n,
general
Ursin 1967
'max' 'maxbiomass, '*•
general
Alverson and Carney 1975
growth rate
young fish
Ware 1975
'-asm. gonad size, condition factor
gadoids
Jones and Johnston 1977
'max
general
Blinov 1977
gonad body weight index.
Ms/w, r^3„ /.,„
general
Gunderson 1980
tV|n(, L|n), k. water temperature
175 stocks
Pauly 1980
energy cost of reproduction
general
Myers and Doyle 1983
max
various
Hoenig 1983
weight
various
Peterson and Wroblewski 1983
'<. '-inf. '-asm
various
Roff 1986
^Maximum age,
2Von Bertalanffy growth parameter.
3Maximum length.
''Length at age of sexual maturity.
^Maximum weight.
6Age at occurrence of cohort's max biomass.
''Age of sexual maturity.
28
VETTER: NATURAL MORTALITY IN FISH STOCKS
Predation Methods
A third class of estimators extends single spe-
cies cohort analysis to a multispecies assemblage
incorporating the major predators and alterna-
tive prey of the stock in question. Single species
cohort analysis is used to estimate population
abundances and annual values for the instanta-
neous rate of fishing mortality (F) for single
groups, usually year classes, of fish (e.g., Pope
1972; Ricker 1975; Gulland 1983). The multispe-
cies extension simply combines cohort analyses
for several species (e.g., Anderson and Ursin
1977). The methods all generate estimates of M as
the sum of some constant rate of nonpredatory,
nonfishing mortality plus the total estimated flux
of prey (stock) to each of the major predators. This
feeding flux to predators is estimated by first
using cohort analyses to reconstruct population
sizes of the various groups of predator and prey,
then combining these population sizes with ob-
served growth rates for the predators and with
estimated preferences for various prey. Thus it
becomes possible to estimate the predatory com-
ponent of M.
Versions of the method have been described by
Anderson and Ursin (1977), Majkowski (1981),
and Pope and Knights (1982). Applications in a
marine system (North Sea) have been described
by Anderson and Ursin (1977), in an ecosystem
context, by Laevastu et al. (1982) and in lake
systems by Forney (1977) and Stein et al. (1981).
The predation method has been developed pri-
marily from analyses of marine systems, espe-
cially the North Sea, and much of the literature
exists only as "mimeos" or notes associated with
ICES (International Council for the Exploration
of the Seas) activities. The most readily available
discussion of this approach appeared in Mercer
(1982), which includes a critical review and dis-
cussion by Ursin (1982) of the various methods.
Several other discussions appear in Pauly and
Murphy's (1982) volume of collected papers from
a symposium on theory and management of trop-
ical fisheries. Most of these papers specifically
address tropical multispecies systems, but the
concepts are broadly applicable. References to
other, often less accessible, works can be found in
these two general references.
The predation method is elegant in concept but
often difficult to apply. Studies by Forney (1977)
and Stein (1981) had the distinct advantages of
limited species numbers in a small system, and
direct quantification of stomach contents. Yet
even in lake systems, the sampling problems of
estimating Z, population abundances, and so
forth, remain often as intractable as in large
marine systems. The two greatest problems are 1)
the difficulty in defining vulnerability and prefer-
ence functions for the various prey stocks (e.g.,
Ursin 1982) and 2) the need to include cohort
analyses of all the major interacting species, some
or many of which may not be available commer-
cially (and for which therefore data will be
scarce).
Despite these problems the approach can cer-
tainly generate, for stocks that suffer heavy
predatory mortality from other fished stocks,
more realistic estimates of M than approaches
that simply generate a globally fixed and invari-
ant M. More importantly (and in contrast to the
age-frequency or life history methods) the preda-
tion method has the advantage of being mecha-
nistic. Predation-related causes and conse-
quences of age, size, site, stock, geographic, or
time trends in M can be investigated via pertur-
bation and sensitivity analysis in computer simu-
lation studies or, alternatively, investigated
through analysis of existing catch data. It be-
comes possible (not necessarily feasible) to inves-
tigate the implications of varying age or abun-
dance structures of interacting fishery resources.
Thus the predation approach has considerable
conceptual appeal for fairly simple systems in
which 1 ) predation is the major force controlling
prey abundance, 2) predators have few alterna-
tive prey, 3) the possibility can be ignored that
predators prefer moribund prey which were about
to die anyway, and 4) all major species of predator
and prey are sought commercially so that data on
abundances and feeding preferences are or can be
made available.
Unfortunately, the number of systems satisfy-
ing these requirements appears to be fairly small,
and of course where predation is a relatively
small fraction of M, the multispecies predation
method will be particularly ineffective.
III. SENSITIVITY OF FISHERY MODELS
TO CHOICES FOR M
Although catch-analysis, life history, and pre-
dation methods all exist currently for estimating
M in fish stocks, in practice the only method used
extensively is the first — direct estimation of M
from analysis of catch structure. Thus the discus-
sion below of model sensitivity to M is based on
this type of estimate. The conclusions reached are
29
FISHERY BULLETIN: VOL. 86, NO. 1
not specific to this one method. Model sensitivity
to a given derived value of M will be the same,
regardless of the method used to derive the value.
General Patterns
Sensitivity analyses of M in fishery models
have evolved through two phases. Earlier studies
noted the influence of M on estimates of maxi-
mum yield (Fmax) or maximum yield per recruit
{(Y/R)jnax\ and on F^ax (the fishing pressure re-
quired to produce maximum yield) in Beverton-
Holt yield models (Beverton and Holt 1957; Hen-
nemuth 1961; Francis 1974; Parks 1977; Bartoo
and Coan 1979; Bulgakova and Efimov 1982).
More recently, as cohort analyses have become
more popular, more attention has been directed
toward assessing the influence of M on age-
specific estimates of stock sizes (A^, ) and fishing
mortalities (F, ) produced by these models (Mur-
phy 1965; Pope 1971; Ricker 1971; Agger et al.
1973; Doubleday 1976; Ulltang 1977; Doubleday
and Beacham 1982; Pope and Shepard 1982; Sims
1982a, 1982b, 1984). A few other studies have
investigated the effect of M on estimates of maxi-
mum sustainable yield (MSY) or total biomass
(Francis 1976; Deriso 1982; Beddington and
Cooke 1983; Tyler et al. 1985).
Most of these studies have used a single, invari-
ant value for M . Model sensitivity is then as-
sessed by comparing model results using some
"best" estimate of M, to results using one (or
rarely, more) pair(s) of M values some arbitrary
percentage above and below the best estimate.
Only a few studies exist of the effects of noncon-
stant M, where M varies in different groups of
fish within a given stock. These include Beverton
and Holt's (1957) example of density -dependent
M in plaice, and several investigations of age-
specific M (Parks 1977; Ulltang 1977; Bartoo and
Coan 1979; Sandland 1982; Bulgakova and Efi-
mov 1982; Caddy 1984; Tyler et al. 1985).
No study to date has specifically addressed the
problems of estimating values of M for a full fish-
ery analysis, leading from cohort analyses (using
M to estimate F, , Ni , and recruitment R ) to esti-
mates of yield or yield-per- recruit using the same
M(s) and R subsequently in the Beverton-Holt
formulas. Also, no study to date has addressed the
possibility and consequences of differing patterns
of variability in M , although it has been sug-
gested in one case (Ulltang 1977) that random
variations will be unimportant if the rate is con-
stant (on average) over the fished ages.
In general, the earlier analyses with yield mod-
els assuming a constant M show that higher esti-
mates of M lead to 1) lower estimates of y^ax oi"
iY/R)jnax (because fewer survive to be caught),
2) higher estimates of Fj^ax 'yo^J must fish a bit
harder to catch a given amount of those left), and
3) lower estimates of age at first capture it^.; be-
cause it pays to catch them before they die, rather
than waiting for them to grow bigger but less
abundant).
Including density-dependence tends to exag-
gerate these trends, at least for plaice in the
North Sea (Beverton and Holt 1957). Including
age-structured M in yield models also affects the
estimates, but not necessarily in a straightfor-
ward manner. As described below in the section
on numeric results, change in model output for a
given change in M depends not just on the values
chosen for M, but also on those chosen for the
other parameters. M is not an independent
parameter in these models.
Analyses with cohort or virtual population
models which assume a constant value for M
show that in general the effect of increasing M is
to increase estimates of N, (because the higher M
is, the more fish died in addition to those being
caught) and to decrease estimates of F,. The data
show only Z , which is the sum of M and F, . As-
suming Z has been constant, a decrease in F,
requires an increase in M. If Z has been variable,
the lower F, may be explained on the basis of
higher A^,, a smaller proportion of which (F,)
would account for the observed catch.
The actual effect, particularly on estimates of
A^,, is not necessarily that simple. As with yield
models, a given change in M does not always
produce the same change in model output. The
result depends also on values chosen for other
parameters; M is not an independent param-
eter.
In cohort analysis the results (estimates of A'^,
and F, ) are particularly sensitive to the relative
sizes of F and M (i.e., to the exploitation ratio
E = F/(F + M)). The effect of assuming an incor-
rect value (or series of values) of M tends to build
up as the analysis proceeds backward in time.
This is because with every time step backward
the catch (C) is inflated by the factor M in order
to estimate at that time the size of the entire
stock, not just the size of the catch. That is
A^, =A^, + i + C,(F, +M)/F,
where F, satisfies the catch equation
(1)
30
VETTER: NATURAL MORTALITY IN FISH STOCKS
C, =N,,i (FJF, + A/)(e'
(2)
If M is large relative to F (i.e., the exploitation
ratio is low), then errors in A^, can increase pro-
gressively and become quite large at the younger
ages (e.g., Agger et al. 1973; Murphy 1965; Ull-
tang 1977; Sims 1982a, 1982b, 1984).
Numeric Results
Although general responses of various models
can be determined simply by inspection of the
analytic models themselves, the quantitative
change to expect in the result (output) for a quan-
tified change in M (input) is not always immedi-
ately obvious. This is because M tends to occur
more than once in various formulas. For example,
M appears in both the numerator and denomina-
tor in the solution to the Beverton-Holt yield
equation (Ricker 1975).
Y = FA^oe'-^'-WxdAM + F)
-3e'-*'"V(M +F + k)
+ 3e(-2*^V(M +F + 2k)
e'-s^'-'/CM +F ^3k)).
(3)
So, rather than derive analytical expressions
(e.g., Sims 1984), I resort below to a simpler ap-
proach. Sensitivity of fishery models to changes of
given magnitude in M is assessed by comparing
percent change reported in model response (out-
put) to percent change in M (input). In cases for
vector (age or density-dependent) M, I have
merely described the shape of the M -vector. For
these different vectors, I report the percent
change in the result due to switching from a vec-
tor of one shape to a vector of another shape.
Yield Models
At least four studies (Beverton and Holt 1957;
Hennemuth 1961; Francis 1974; Bartoo and Coan
1979) have shown that errors in estimates of
M propagate into roughly equal errors in esti-
mates of (y//? )maxj but with sign reversed (Table
2). For example, a 10% overestimate in M will
lead to approximately 10% underestimate of (Y/
R 'max- An equally important result is that the
actual magnitude of the effect induced depends
strongly not just on the error in M, but on the
values chosen for the other parameters in the
model.
In another study, Beddington and Cooke (1983)
used the Beverton-Hol formulation to investi-
gate the influence of M (constant; 0.1 to 0.8
year"M, t^ (0 to 4 years), and K (the von Berta-
lanffy growth parameter; 0.1 to 0.5 year"M on
MSY (maximum sustainable yield), expressing
the result as "MSY as a % of Bq," where Bq is the
initial or recruited biomass. Higher percentages
indicate that more of the original biomass is
being taken at MSY. Increasing M by a factor of
8 (0.1 to 0.8 year-i) increased MSY/Bq by a factor
of about 4 to 8, depending on the particular values
of tf. and K. Again, errors in M produced roughly
the same relative error in the result; and again
the actual effect of any given change in M de-
Table 2— Sensitivity of estimated maximum yield per recruit ((V/Rj^ax) 'o changes in instantaneous
rate of natural mortality (M) and other input conditions. Sensitivity of (>^'W)max ^^d of changes in Ware
expressed as percentage difference from nominal responses at nominal (best-guess) M. Symbols
are: t^ = age-at-first-capture, F = instantaneous rate of fishing mortality, M = nominal value for
M. Frances (1974) used an age-structured simulation model. All other citations used standard yield-
per-recruit analyses.
Input conditions
% change % change in
in M (^/'^Jmax Species
Source
fc = constant (3.72)
F = variable
/W = 0.10
-1-50
-50
-20
-^30
fp = variable
F = constant (0.73)
M = 0.10
-1-50
-50
-60
-^50
F = constant (0.95)
M = 0.8
+ 20
-20
-21
+ 32
/W = 0.8
-HO
-10
-14
-M6
M = 0.60
+ 2b
-20
plaice
plaice
Beverton and Holt 1957
Beverton and Holt 1957
yellowfin tuna Hennemuth 1961
yellowfin tuna Francis 1974
yellov\rfin tuna Bartoo and Coan 1979
31
FISHERY BULLETIN: VOL. 86, NO. 1
pended on the values chosen for the other
parameters.
Pope and Garrod (1973) present another exam-
ple of sensitivity in MSY to values chosen for M.
They describe briefly the consequences of using
an incorrect constant for M of cod stocks when
estimating the F required to generate MSY
(Fmsy*- Underestimating M by 507r (assumed
M = 0.1 year^^; true M = 0.2 year"M leads to a
choice of Fmsy that is 67% too high. Overestimat-
ing M by 50% (assumed M = 0.3 year ^ true
M - 0.2 year"M underestimated Fmsy by 50%.
The simulations described above tested the ef-
fects of choosing alternative constant values for
M. Choosing a vector alternative can also have
significant effects; again, the magnitude of the
effect depends on the values chosen for other
parameters. Beverton and Holt (1957) showed
that incorporating density-dependence in M for
plaice decreased (y/i?)max by 12%, when holding
tc constant at 3.72 years and letting F vary. Con-
versely, holding F constant and letting t^, vary
decreased {Y/R )max by about 37%.
Age-dependent values for M were compared
with age-constant values by Bartoo and Coan
(1979), Bulgakova and Efimov (1982), and Tyler
et al. (1985). In their analysis of Atlantic yel-
lowfin tuna stocks, Bartoo and Coan found that
replacing an assumed constant M of 0.8 year"^
with an age-structured M increasing from 0.1
year~^ at age 0 to 1.2 year~^ at age 7, increased
(Y/R )max by 17% (from 6 to 7 kg).
Estimating total yield (7,) rather than {Y/R )
and estimating R as a function of constant versus
age-specific M in analysis of catch curves for rela-
tively unexploited stocks of Pacific ocean perch
and Oregon hake, Bulgakova and Efimov (1982)
found that replacing a constant (age-averaged) M
with age-variable M tended to increase estimated
Yi when fish recruited fairly late to the fishery,
but decreased Yf if the fish recruited early. This is
because of the interaction between the values as-
sumed for M (constant or age-variable) and the
value calculated for R from each type of mortality
curve.
Starting with a given value for recruitment at
age 6 years (from Efimov 1976), they calculated R
twice for ages 4 and 8 years — once with age-
averaged M and once with age-specific M. Be-
cause in this set of data the age-averaged M was
generally higher than the age-specific M at the
tested ages of recruitment (ages 4, 6, or 8 years),
back-calculations with age-averaged (i.e., con-
stant) M predicted fewer recruits than back-
calculations with age-specific M. With fewer re-
cruits and generally higher M , potential yield at
later ages obviously must drop. Differences in
predicted potential yield ranged from about
-30% at ^4 (age-specific estimate lower than age-
averaged estimate, when fish were assumed to
recruit to the fishery at age 4 years) to +15% at
^6 (age-specific estimate higher) and to +60% at
Tyler et al. (1985) tested (among other things)
the effects of ignoring "true" age-structure in M
and using instead a constant value in estimating
stock biomass using Deriso's (1980) delay-
difference model. They did the tests on catch data
generated by Walter's (1969) age-structured sim-
ulation model of cod, using three different (input)
age structures for M in Walter's model. After gen-
erating "catch data" from Walter's model, they
analyzed the simulated data set using Deriso's
model with constant M (= 0.5 year~^). The age
structures tested were 1) mortality increasing
and then decreasing with age (Walter's original
mortality vector spanning ages 3 to 12 years; age-
averaged M = 0.55 year~\ range = 0.33 to 0.70
year~^), 2) mortality increasing with age (ages 7
to 12 years; average M = 0.5 year~^, range 0.3 to
0.7 year"^) and 3) mortality decreasing with age
(ages 7 to 12 years; average M = 0.5 year"\
range 0.7 to 0.3 year M. In all three cases Deriso's
model with constant M misestimated the "true"
biomass generated by Walter's model (with age-
structured values for M). The differences were
relatively small, however: —13% for the increas-
ing and then decreasing series, +19% for the de-
creasing series, and +4% for the increasing
series. These differences were due to the differ-
ences in M, and not the differences in model
structure; generating and analyzing biomass
with the same constant M in both models led to a
discrepancy of only 0.5%.
By analogy to life history patterns in other
adult animals, M (after recruitment into most
fished stocks) is more likely to increase with age
than to cycle or decrease. By implication, the sim-
ulation results from the increasing series are
probably most realistic. If so, the effects of ignor-
ing age-structure in favor of using a constant M
may be relatively small (5 to 20%), at least for the
cod stock simulated in this study. But the results
obviously depend again not just on correctly
choosing the values for M, but on the values cho-
sen for the other parameters. In this case, Tyler et
al.'s (1985) results imply that age-structure in M
can be relatively unimportant, at least when the
32
VETTER: NATURAL MORTALITY IN FISH STOCKS
assumed constant is evenly bracketed by the
"true" age-structure in M .
Further simulations by Tyler et al. (1985) using
a wide range of constant values for M (0.4 to 1.4
year M and the growth rate parameter rho
(mean Ford growth coefficient for the fishable
stock; 0.46 to 1.6) showed that incorrect guesses of
M (and rho) could produce errors up to 1,000^^ in
estimated biomass. More realistic ranges for the
two parameters (0.4 to 0.8 year" ^ forM, 0.6 to 1:2
for rho), extending about 509r above and below
the "true" values for these parameters, induced
much lower error in biomass estimates (about the
same order of magnitude, 50 to 100% below and
above the "true" biomass). As before, changes
(errors) of a given amount in M (expressed as
fraction or percentage of the original value) ap-
pear to produce about the same amount of change
(expressed as percent of original value) in simple
estimates of yield, depending on the conditions of
other parameters in the model.
Chatwin (1958) compared estimates of Yj^ax
from lingcod populations. Rather than compare
constant and age-variable values for M, he as-
sumed several different values for an average
(constant) M in adults, but assumed that M in-
creased from the assumed average for adults to
higher values in both juveniles and senescent
fish. He reports no quantitative results but states,
as found above, that increasing the average M,
for a given F, considerably decreased l^max' that
decreasing M increased Yrna-x' snd that size at
first capture changed relatively little with those
changes in M.
These comparisons between age-structured
versus constant M, or between different constants
have demonstrated that effects on results can be
large for some combinations of parameters yet
small for others. Alternative choices drawn from
apparently realistic parameter values lead to rel-
atively small differences in estimates of M.
Specific amounts of change depend strongly not
only on the values chosen for M, but also on the
value of M relative to values chosen for the other
interacting parameters in the yield models. For
most choices of parameter values, sensitivity of
output is roughly equal to perturbation of input.
Cohort Analyses
Effects of interactions between changes in M
and values chosen for other parameters is even
more obvious in stock reconstruction analyses
(e.g., cohort analysis and virtual population anal-
ysis (VPA)). These analyses are used to "recon-
struct" estimates of stock abundance during pre-
vious years, based on catch data and assumptions
about the value(s) of M during those previous
years. Studies of sensitivity to M in Beverton-
Holt types of yield or biomass assessments were
usually empirical, based on analyses of catch data
from specific fisheries. Studies of sensitivity to M
in VPA and cohort analysis include both theoret-
ical and empirical studies; i.e., simulations using
totally contrived data sets (e.g.. Agger et al.
1973), analyses of specific data sets (e.g.. Pope
1971; Doubleday and Beacham 1982) and combi-
nations of analytical evaluations and analysis of
specific data sets (e.g., Doubleday 1976; Ulltang
1977; Sims 1982a, 1982b, 1984).
Simple analyses of sensitivity to M, in which M
is varied but all else is held constant, include
1) Pope's (1972) analysis of Atlantic yellowfin
tuna, in which he found that replacing constant
M with age-structured M (higher Ms for older
fish) produced lower estimates for fishing mortal-
ity {F, ) in the later ages, but had little effect on
estimates for the younger ages, and 2) Doubleday
and Beacham's (1982) statement that I07c error
in constant M translated into 9 to 149f error in
estimates of i? (at age 3) for cod in the Gulf of St.
Lawrence.
Somewhat more complicated analyses are pre-
sented by Ulltang ( 1977 ) and Sims ( 1982a, 1982b,
1984). Ulltang evaluated the effects on model pre-
dictions of F, and A'^, , of several types of variation
in M. These included no variation (uniformly con-
stant M), M constant within years but varying
randomly between years, M varying with age,
and M varying with season. Sims evaluated the
effects of choosing various constants for M on esti-
mates of A'^,, and derived an analytical expression
relating variance in M to expected variance in
estimates of abundance.
In Ulltang's simulations, increasing (decreas-
ing) a constant M by 50*^ (from 0.2) decreased
(increased) F by about 207c ("true" F's ranging
from 0.4 to 0.8). Creating a data set with M vary-
ing randomly from one year to the next, then an-
alyzing those data with an assumed constant M,
Ulltang (1977) found that the Z calculated from
the constant-M model was on average the same as
the "true" Z from the random-M model. He con-
cluded that random fluctuations in M will cancel
out during analysis and so can be ignored. Ull-
tang assessed the influence of age-dependent M
compared with constant M by generating a catch
curve with age-variable M (decreasing curvilin-
33
FISHERY BULLETIN: VOL 86, NO 1
early from 0.3 at age 1 to 0.1 at age 10, average
about 0.2) and F equal to 0.2, then analyzing the
catch with F equal to 0.2 or 0.6, and M equal
either to 0.1 or 0.2. Choice of M made little differ-
ence in estimates of stock size for the case of high
F (0.6), because most of the deaths were due to
(observed) fishing. When F was low (0.2), stock-
size estimates were much more sensitive to incor-
rect choices for M , because most of the deaths
were due in this case to M, which was unmea-
sured and therefore unobserved.
Ulltang (1977) simulated seasonal changes in
M by concentrating all deaths in either the first or
last quarter of a year. Estimated stock sizes (N, )
changed relatively little; with F = 1.2 and
M - 0.4, A'^, was a maximum of 10% higher if all
deaths occurred first quarter, 10% lower if all oc-
curred in the last quarter.
A serious problem with the conclusions reached
by Ulltang (1977) is also common to all the other
studies discussed above; they are based on rela-
tively few combinations of values for the various
parameters, and relatively few simulations. For
example, the conclusion that random errors in M
will tend to even out is intuitively attractive, pro-
vided the time scale of variation is short relative
to the generation time of the fish. In fact random
variation in M did even out in the two sets of
simulations he conducted. But the examples he
chose included only one set of ages (2 to 10 years),
with relatively high values of F (0.5 to 0.8 year"^)
compared to the values tested for M (0.1, 0.3
year"M. The gravity of consequences from choos-
ing an incorrect M depends very heavily on the
size of M relative to the size of F, i.e., on E. Had
he chosen different values for his simulations, he
might have reached very different conclusions.
This is probably the basis for the discrepancy be-
tween Ulltang's conclusion that seasonal effects
are minor, versus Sims' (1984) conclusion that
seasonal effects can be quite large, if M is high.
Sims (1984) attempted to overcome this prob-
lem (trying to draw general conclusions from the
results of simulations based on particular, or rel-
atively few, sets of parameters) by analytically
deriving formulas for relative error in stock-size
estimates, and then testing the formulas with
data from actual fisheries. He used this approach
twice: once to assess the effects of seasonality
(Sims 1982a) and once to consider in general the
effects of different choices (errors) for constant M
(Sims 1984). But his results (and equations for
error) show clearly that error in estimated stock
sizes depends on several parameters and that the
effects of one can be strongly dependent on the
values chosen for the others. Choosing a high M
(0.6 year~M and concentrating catch during the
first quarter of the year overestimated R by 20%;
concentrating catch during the last quarter
underestimated R by 23% (compared with the
10% error found by Ulltang).
Within the same analysis, reducing M by half
(to 0.3 year^ ^) reduced the error in R by half, but
the same reduction of error inR was also achieved
by leaving M high and reducing F . In assessing
specifically the effects of error in M on error in R ,
Sims (1984) showed very different effects on esti-
mates of/? in heavily fished versus lightly fished
cohorts of Atlantic bluefin tuna. Changing M by
50% led to changes in estimated R of 60 to 260%
in the lightly fished cohort, but only to relatively
smaller changes of 35 to 70% in the heavily fished
cohort. Again, the magnitude of the error in
model predictions depended not just on the mag-
nitude of M, but on its relationship to the other
parameters in the catch equation, particularly F.
Errors (expressed as percentage change in out-
put for a given change in input) in model output
in the simulations described above, all of which
tried to use apparently realistic values for model
parameters, rarely exceeded 50%, and were often
less than the error introduced into values chosen
for M . By implication, the effects of incorrectly
guessing M may be relatively unimportant if M is
relatively small (e.g., in this situation not more
than about 0.5 year"M and relatively invariant,
although the actual magnitude of effect due to
any given percentage change in M depends on the
values chosen for other parameters.
So, inaccurate estimates of M might be impor-
tant or they might not. It all depends on the mag-
nitude and variability of M within a given stock
(or group). Although untested, it seems likely
that estimates of M for groups in which M varies
little and is relatively low, are more likely to be
reasonably accurate than estimates of M from
groups in which M is large and variable. The fol-
lowing section reviews evidence that M does in
fact vary both within and between groups of fish,
and the succeeding section reviews evidence for
the magnitude of that variability in ostensibly
similar groups.
IV. FACTORS INFLUENCING DEATH
RATE
Despite the fact that in most fishery models, M
is assumed to be constant for all exploited ages in
34
VETTER; NATURAL MORTALITY IN FISH STOCKS
any given stock, abundant evidence exists to the
contrary. Natural mortality has been shown to
vary with age, density, disease, parasites, food
supply, predator abundance, water temperature,
fishing pressure, sex, and size. Evidence for rela-
tionships between these factors and M, and se-
lected references for each, are presented below.
Changes in mortality rate with age, within sin-
gle groups of fish, have been demonstrated and
discussed more frequently than changes with any
other factor. References include, among others,
Baranov (1918, plaice), Sette (1943, Atlantic
mackerel), Ricker (1945, 1947, lake fish; 1969,
1975, various species), Beverton and Holt (1959,
many species of marine fish), Beverton (1963, en-
graulids and clupeids), Boiko (1964, sturgeon).
Gushing (1975, plaice), Blinov (1977, fish in gen-
eral), Bulgakova and Efimov (1982, Oregon hake
and sea perch), Sandland (1982, fish in general).
Smith (1985, clupeoids), Roff (1986, fish in gen-
eral). Evidence for changes with senescence for
fish in general has been discussed or documented
by, among others, Woodhead (1979) and Craig
(1984).
Although specific patterns vary with species
(e.g., Woodhead 1979), in general M is extremely
high during egg and larval stages (e.g., 2 to 10%
per day in plaice and clupeoids (Cushing 1975;
Smith 1985)), falls precipitously during the juve-
nile period, becomes relatively stable during in-
termediate adult ages and increases again with
senescence. But even during these relatively sta-
ble mid-adult ages, changes in M with age can be
substantial, particularly in short-lived fish (e.g.,
Ricker 1947, stunted versus "normal" whitefish).
Changes in natural mortality rate with size
(rather than age) within single groups of fish
(usually stocks), have been discussed by Baranov
(1918, plaice), Ricker (1969, size-selective mortal-
ity in general). Ware (1975, larval fish), and
Peterson and Wroblewski (1984, many species).
Differences in natural mortality rate between
populations of the same species in different envi-
ronments, or even in different areas of a single
environment (e.g., a single lake) are documented
by Ricker (1947), Kennedy (1954), and Schupp
(1978). Year-to-year differences in natural mor-
tality rates of single stocks from a given area are
shown by Pope and Knights (1982, plaice) and by
Henderson et al. (1983, whitefish). Density-
dependent changes in M are discussed by Bever-
ton and Holt (1957), Cushing (1967), Tyler and
Gallucci (1980), Backiel and LeCren (1978), Jones
(1982), and others. Differences in M between
sexes have been documented by Beverton and
Holt (1957, plaice), Ricker (1947, rock bass), and
others. Changes in natural mortality rate related
to the cost of reproduction have been discussed
by Jones and Johnston (1977), Roff (1984), and
others.
Other factors that affect M either alone or in
combination with other factors include disease
and parasitism (reviewed by Lester 1984), starva-
tion (Hewitt et al. 1985; Theilacker 1986: larval
anchovy), physiological state (Smith 1985), and
fishing pressure (Ursin 1982; Munro 1982). Addi-
tional examples are cited by Beverton and Holt
(1957), Anderson and Ursin (1977), Sissenwine
(1984), and Hunter (1984).
Most of the factors listed above (e.g., age, size,
sex) are indirect influences on M . The most im-
portant factor directly affecting natural mortality
rate is probably predation; this is implied by a
large body of literature describing changes in
prey community composition and abundance fol-
lowing changes in composition and abundance of
predators (e.g., Carpenter et al. 1985).
Direct evidence that predators account for most
natural mortality in fish stocks is difficult to
gather (Section II). To quantify the fraction of M
due to predation, one must know, not only rela-
tive changes in abundance, but absolute popula-
tion density of all predators and prey together
with consumption rates and prey preferences of
all the predators. Although this is rarely possible,
at least two studies from freshwater systems do
present quantified estimates of predatory mortal-
ity in relation to available prey. Forney (1977)
quantified predation mortality in a relatively
simple, unmanipulated lake system where there
were few species of predator and prey. Combining
stomach-content estimates of prey consumed with
trawl-sample estimates of predator and prey
abundance, he concluded that 30 to 100% of yel-
low perch production was consumed by walleye,
their principal predator. In a manipulated sys-
tem. Stein et al. (1981) assessed predatory mor-
tality of young tiger muskellunge after they were
stocked in a small pond and lake. During the time
of the study, a single predator (largemouth bass)
accounted for 25 to 45% of losses to natural mor-
tality.
In marine systems evidence for the relative im-
portance of predation can be gleaned from com-
paring total natural mortality with estimated
predatory mortality based on abundance of preda-
tors and feeding preferences. For example, multi-
species cohort analyses reported by Pope and
35
FISHERY BULLETIN: VOL. 86, NO. I
Knights ( 1982) show predatory mortality as 80 to
909c of M for age-0 cod, whiting, and haddock in
the North Sea (the fraction of M due to predatory
mortahty cannot be assessed accurately in the
older ages because predators appropriate to these
sizes were not included in the analysis). In an-
other example, estimating M from energy flow
models, Sissenwine (1984) demonstrated that
predation in the Georges Bank ecosystem can ac-
count for all production by prey fish; nonpreda-
tory mortality was negligible.
Thus a multitude of factors, acting alone or in
concert, can be expected to produce variations in
M between individuals within single groups of
fish, as well as between groups. Differences can
be expected between species, between stocks
within species, and from place-to-place and time-
to-time within given stocks. In the following sec-
tion, I will review more completely existing evi-
dence for, and the extent of, this expected
variability in M .
V. VARIABILITY WITHIN AND
BETWEEN GROUPS
As discussed above (Section III), simulation
studies generally show that effects of choosing a
particular value or set of values for M can range
from insignificant to considerable, depending in
part on the model used, in part on the values
chosen for other parameters, and in part on the
form chosen for the estimate(s) of M. Authors
suggest that in the future, simulations should be
conducted with a range of values for M, to bracket
probable values (e.g., Beverton and Holt 1957;
Tyler et al. 1985).
The problem with this advice is identifying the
appropriate range and distribution of M for any
given group offish. Obviously, wide ranges for M
will lead to great discrepancies between model
predictions based on one end of the range or the
other. It has been shown above, however, that
model output can be relatively insensitive to
small changes in M . This is particularly true if F
is much larger than M (i.e., if the stock is highly
exploited so that losses to fishing far exceed losses
to natural mortality). The problem is determining
whether, for a given stock in situ, changes in M
are in fact large or small. Compensatory changes
in M, in response to changes in F, will further
confound the problem, because variations in M
will then be a function of the value(s) of F, in
addition to the suite of other factors that may be
affecting estimates of M .
M does appear to vary considerably between
groups offish. Estimates of M compiled by Pauly
(1980) (Fig. 1) for 175 stocks and species offish
worldwide differ greatly between groups, ranging
from a minimum of about 0.1 year"^ to several
unusual values as high as 7.0 year ^ Even
within a group as ostensibly homogeneous as the
tunas, the range of estimated mortality constants
spans the majority of the common values (0.2 to
2.0 year"i. Murphy and Sakagawa 1977).
Estimates of variability in M within groups of
fish are much less common, but are actually more
important than the obvious differences between
groups with obviously different characteristics
such as differing lifespans. Most fishery analyses
are directed toward understanding or predicting
dynamics of single stocks (single groups offish).
The most important considerations for natural
mortality parameter values in these single-
species analyses are whether and if so over what
values M varies for the group of fish in question.
But measuring trends or variability in natural
mortality rates withingiven groups (e.g., stocks)
is difficult and, with the exception of trends with
age, rarely attempted. This is primarily because
the only extant methods for estimating M depend
either directly or indirectly on analysis of catch
data (Section II), and catch data are prone to
many well known (but largely unsolved) prob-
lems.
Problems with analysis of catch data fall into
two general categories: 1) problems with sam-
pling procedure, such that fish are caught or
counted out of proportion to their true abundance
and 2) problems with fish appearing or disappear-
ing from the "unit stock" due to causes other than
birth or natural mortality (i.e., migration, fishing
mortality, or tagging mortality), again resulting
in catch data that do not represent the true struc-
ture of the stock. If sampling biases can be over-
come, the problem reduces to partitioning total
disappearance offish into fractions owing to fish-
ing, tag mortality, and migration. The first parti-
tion can be eliminated by studying unfished popu-
lations, the second by quantifying tag mortality,
and the third by studying only closed or tagged
populations.
Unfortunately, very few sampled populations
satisfy completely even one of these criteria. Re-
gardless, we still need at least some crude esti-
mates of M in order to determine whether M truly
varies enough to invalidate the standard assump-
tion in fisheries models that M is effectively con-
stant during exploited ages. The question here
36
VETTER NATURAL MORTALITY IN FISH STOCKS
>-
o
z
UJ
3
O
UJ
QC
U.
28
26-
24-
22-
20-
18-
16-
14-
12
10-
8-
6-
4-
2-
0
NATURAL MORTALITY ESTIMATES: FROM PAULY (1980)
(FISH STOCKS)
T
nni
Ji
[,i i"i'"('|"i'i'i I I I I'i'i I I I I I I I'l'i'i'i'n I I I iTi I I I I I I I I I I I I I I I I I I iMi I I I'l'i I
0.5 1.0 1.6 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
ESTIMATED NATURAL MORTALITY (M)
Figure 1. — Frequency of estimated instantaneous annual rates of natural mortality (M) in 175 different fish stocks, populations, or
species. Estimates include both freshwater and marine species. Data from Pauly 1980.
concerns variability of M within groupings that
would be used commonly to estimate M , such as
stocks of single species, rather than general pat-
terns across species. If M truly varies relatively
little during these ages (so that the log of catch-
at-age decreases linearly with age), if the age
structure has been relatively constant histori-
cally (so that catch curves are actually linear,
rather than curvilinear as seen in stocks with
inconstant age structure, e.g., chapter 2, Ricker
1975), and if catch curves actually reflect rela-
tively accurately this constancy and low variabil-
ity, then most estimates of M derived from analy-
sis of appropriately processed catch curves cannot
help but be relatively close to the true rate. Model
predictions, although in theory sensitive, would
in practice be fairly robust to any particular value
chosen from the true range of values for M .
Despite the potential problems with accuracy
or precision of existing estimates from single
groups of fish, I list in Table 3 most of the esti-
mates available for unexploited populations, and
some of the few existing estimates from exploited
populations. My purpose is to identify the appar-
ent range of variability in M within single stocks.
The estimates are drawn from references cited by
Pauly (1980) and other sources. Only references
that reported multiple estimates for M are in-
cluded, thus excluding most of the references re-
viewed. Because these estimates are derived from
catch data, the stated ranges are "apparent",
rather than demonstrably the "true" values.
Estimated rates of natural mortality are not
particularly constant for either unexploited or ex-
ploited groups, and are only slightly less variable
within stocks than they are within species. Al-
though the range of rates within groups may ap-
pear relatively small compared to the total range
of rates reported for all fish species (e.g., 0.36 to
0.56 for sauger from Lake Nipigon [Table 3] vs.
approximately 0.1 to 3.0 for most species listed
by Pauly 1980), the maximum and minimum
rates reported for single groups differed by at
least 50% in 20 of the 22 comparison listed in
37
FISHERY BULLETIN: VOL. 86, NO. 1
Table 3. In at least one case maximum and mini-
mum estimates differ by as much as a factor of 7
(i.e., young vs. old whitefish in Shakespeare Is-
land Lake, Ricker 1947, Table 3).
The range of reported estimates of M for species
(rather than single groups or stocks within a spe-
cies as compared above) is even greater. Even the
least variable estimates differed by a factor of
1.75 (75%, male vs. female plaice, Beverton 1964).
In whitefish, the species for which the most esti-
mates exist, maximum estimates are 20 times
greater than minimum estimates (Table 3).
Table 3. — Ranges in estimates of instantaneous rate of natural mortality in unexploited and exploited fisfi populations. 'Wr^ax'''^min is
expressed as ffie ratio between tfie maximum (Mmax) and minimum (/Wmm) values reported for M for that species, Values in parenttieses
are total range of estimates and ratios for those species where multiple reports exist.
Common
Age
Species name
name
Body of water
Sex
(years)
M range
'Wmax^'^mir
, Source
1) Unexploited populations:
Amboplites rupestris
rock bass
Nebish Lake
m
10-12
1.47-2.1
1.49
Ricker 1947
Nebish Lake
f
10-14
1.1-1.6
1.45
Ricker 1947
Nebish Lake
both
10-14
1.08-1.56
(1.08-2.1)
1.44
(2.01)
Ricker 1947
Stizostedion canadensis
sauger
Lake Nipigon
8-14
0.36-0.56
1.56
Ricker 1947
Coregonus clupeaformls
whitefish
Lake Opeongo
6-13
0.53
Ricker 1947
Shakespeare Island Lake
11-27
0.08-0.60
7.51
Ricker 1947
Great Slave Lake
17-22 .
0.71-0.99
1.39
Kennedy 1953
Lake Nueltin
13-15
0.84
Kennedy 1963
Lake IVIcDonald
11-14
9-10
1.34
1.66
(0.08-1.66)
(20.75)
Kennedy 1963
Kennedy 1963
Leuachthys sardinella
Ikroavik Lake
6-10
0.2-1.4
7.00
Wohlschlag 1954
Cnstovomer namayacush
Great Slave Lake
1-26
0.31-1.61
5.19
'Kennedy 1954
Great Slave Lake
15-23
0.49-0.92
1.88
'Kennedy 1954
Great Slave Lake
15-23
0.52-0.75
(0.31-1.61)
1.44
(5.19)
2Kennedy 1954
Perca fluviatilus
perch
River Thames
m
3-8
0.56-0.98
1.75
Williams 1967
River Thames
f
3-7
0.2-0.64
3.20
3Williams 1967
River Thames
juv
3-5
0.53-1.69
(0.2-1.69)
3.19
(8.45)
3Williams 1967
Leuciscus leuciscus
dace
River Thames
5-11
0.36-1.31
3.64
"Williams 1967
Alburnus alburnus
bleak
River Thames
3-8
0.6-2.4
4.00
■•Williams 1967
Rutilus rutilus
roach
River Thames
River Stour
2-11
3-12
0.22-1.38
0.44
(0.22-1.38)
6.27
(6.27)
"Williams 1967
Mann 1973
Cheilodactylus macropterus
tarakihi
Chatham Islands
Chatham Islands
New Zealand
5-35
5-22
0.03
0.08
0.15
(0.03-0.15)
(5.00)
Vooren 1977
Vooren 1977
Vooren 1977
II) Exploited Populations:
Pleuronectes platessa
plaice
North Sea
f
5-13
0.08
Beverton 1964
North Sea
m
5-13
0.14
(0.08-0.14)
(1.75)
Beverton 1964
Brevooiiia petronus
gulf menhaden
Gulf of fvlexico,
1-3
0.7-1.6
2.29
5Ahrenholz 1981
Central (1969, 1971)
Gulf of fvlexico,
1-3
0.88-0.98
1.11
SAhrenholz 1981
Eastern (1969,
1971)
Gulf of IVIexico,
1-3
1.17-1.23
1.05
SAhrenholz 1981
Western (1969,
1971)
Gulf of IVIexico,
1-3
0.95-1.2
1.26
SAhrenholz 1981
all areas (1969,
1971)
(0.7-1.6)
(2.29)
Gadus morhua
cod
North Sea
0.5-1
0.59-1.46
2.47
7Pope and
Knights 1982
Coregonus clupeaformls
whitefish
Lake Huron
3.8
0.34-1.67
4.91
SHenderson et al.
1983
'Increasing with age.
2Year to year variation (1946-52); ages 15-23 combined.
3Not consistent with age.
"Generally increasing with age.
SAssuming 20% tag loss rate,
spor tag loss rates from 10 to 30%.
78 different years (1967-75).
823 different year classes (1947-75).
38
VETTER: NATURAL MORTALITY IN FISH STOCKS
As discussed previously, these different esti-
mates can lead to at least as great a difference in
results produced by fishery analyses such as yield
models or stock reconstruction analyses (Sec-
tion II).
Reported differences in estimates of M for
whitefish stocks living in Shakespeare Lake com-
pared with other relatively small lakes (e.g., Lake
McDonald, Table 3) are particularly significant.
Because both stocks are of the same species and
living in more or less similar environments
(small lakes), one might easily (and incorrectly)
assume that both have the same rate of natural
mortality; but they did not. The lower rates oc-
curred in the stock occupying a small lake with no
predators. This is a clear example of the effect
that environment, particularly the predator envi-
ronment, can have on the realized rate of natural
mortality in a fish stock. Obviously, choosing a
single appropriate constant for this species would
be difficult. Choosing an appropriate species-
specific constant for some of the other species
with multiple estimates might be difficult as well
(e.g., rock bass, lake trout, perch, roach, tarakihi,
or menhaden. Table 3).
None of these studies from either unexploited
or exploited stocks support the assumption that M
is constant for any given stock or species, nor are
these within-stock ranges particularly narrow. In
addition, treatment of the original catch data
may have in some cases obscured the "true" vari-
ability. Ricker (1947) and Kennedy (1953, 1954,
1963), for example, use a 3-yr smoothing tech-
nique to reduce the effects of unequal recruit-
ment; this also serves to reduce variability that
may actually be due to differences in natural mor-
tality. Also, single estimates from data collected
during only one or two years of sampling (e.g.,
Wohlschlag 1954; Williams 1967; Mann 1973;
Vooren 1977) can be seriously biased by annual
changes in either recruitment or mortality rates.
If the estimates reported above are even approxi-
mately accurate, it is apparent that the range of
possible values for M is wide, and that variability
can be considerable even within single stocks.
A solution to this problem of choosing a reason-
able value for M, at least for long-lived fish, is
suggested by the possibility that variation in M
(not just the mean value) may be related to max-
imum lifespan. Fish that live for many years
must naturally have lower mortality rates than
more short-lived fish. These lower rates may also
be less variable in the longer lived stocks, if as in
many other biological processes, variability is
proportional to the mean. This could account for
the ubiquity and apparent effectiveness of the
constant 0.2 year"\ used almost universally for
the long-lived (20 to 30 years) and well-studied
fish stocks from northern European seas (e.g.,
Beverton 1964). If so, assuming a constant M
might be valid for these longer lived stocks.
Unfortunately, the few studies cited above do
not support this attractive idea. Although in gen-
eral, mortality rates decease as lifespan in-
creases, the variability in estimates does not ap-
pear to follow the same trend. This may be due
partially to the relatively similar lifespans (10 to
20 years) for most of the species for which esti-
mates exist. But the apparent range in rates for
the shortest lived species cited above (Ahrenholz
1981, Brevoortia patronus, ages 1 to 3 years, M
range 0.7 to 1.6 year"^) is certainly not greater
than ranges reported for the longer lived white-
fish (Henderson et al. 1983, Coregonus clu-
peaformis, ages 10+ years, M range 0.34 to 1.67
year"^).
VI. SUMMARY AND
RECOMMENDATIONS
Thus it appears that rates of M, or at least rates
of M derived by existing estimation methods, do
in fact vary widely within many fish stocks. Be-
cause the variations appear to be considerable
and because the results from fishery models can
be sensitive to large variations in M , one must
conclude that assuming constancy without proof
can have serious consequences for fishery man-
agement.
A better approach may be to discard the notion
that a single "best" estimate of M can be found,
and instead try to tailor estimates of M to local
groups, based on some combinations of the meth-
ods discussed in Section III. Obviously, practical
considerations of time and resources will limit the
accuracy and precision with which M can be esti-
mated. Also, the estimates in the studies re-
viewed here are prone to all the artifacts men-
tioned in the previous sections. True rates of
natural mortality, and their variability, are still
very poorly known for even the great stocks of
commercial fish in temperature regions that have
been subject to continuous exploitation for
decades. Careful, repeated tagging experiments
probably hold the most promise for determining
with any reasonable degree of accuracy, rates of
natural mortality in fish stocks. But even these
have inherent problems that are not easily
39
FISHERY BULLETIN: VOL 86, NO. 1
solved. There remains a great need both for new
methods, and refinements of the old.
ACKNOWLEDGMENTS
This review was initiated in response to a re-
quest for Deus ex machina in mortality estima-
tion for fish stocks, elicited from participants at
the 35th Annual Tuna Conference held at Lake
Arrowhead, CA, 20-22 May 1984, during a discus-
sion organized in an attempt to identify the single
most important but least well-estimated parame-
ter in fishery models. The author apologizes for
the lack of divine inspiration herein, but extends
grateful thanks to Chris Boggs, Andy Dizon, John
Hoenig, Pierre Kleiber, Robert Olsen, and two
anonymous reviewers for their insightful discus-
sion and helpful comments during preparation of
this review.
LITERATURE CITED
Agger, P , I Boetius, and H Lassen
1973. Errors in the virtual population analysis: The
efiects of uncertainties in the natural mortality
coefficients. J. Cons. Int. Explor. Mer 35:98.
Ahrenholz, D W
1981. Recruitment and exploitation of gulf menhaden,
Brevoortia patronus. Fish. Bull., U.S. 79:325-335.
Alverson, D L , AND M J Carney
1975. A graphic review of the growth and decay of
population cohorts. J. Cons. Int. Explor. Mer 36:133-
143.
Anderson, K P , and E Ursin
1977. A multispecies extension to the Beverton and Holt
theory of fishing, with accounts of phosphorus circulation
and primary production. Medd. Dan. Fisk. Havunders,
7:319-435.
BaCKIEL, T , AND E D. LECREN.
1978. Some density relationships for fish population
parameters. In S. Gerking (editor), Ecology of
freshwater fish production, p. 279-302. Blackwell
Scientific, Oxf.
Barlow. J
1984. Mortality estimation: biased results from unbiased
ages. Can. J. Fish. Aquat. Sci. 41:1843-1847.
Baranov. F I
1918. On the question of the biological basis of
fisheries. Reports of the Department of Fisheries and
Scientific and Industrial Investigations, Vol I, Part 1.
(Engl, translat.; Library, Southwest Fisheries Center,
P.O. Box 271, La Jolla, CA.)
BARTOO, N W , AND A L COAN
1979. Changes in yield recruit of yellowfin tuna Thunnus
albacares under the ICCAT minimum size
regulation. ICCAT Coll. Vol. Sci. Pap. VIII:120-138,
BEDDINGTON, J R . AND J G COOKE.
1983. The potential yield offish stocks. FAO Fish. Tech.
Rep. 242, 47 p.
Beverton, R J H
1963. Maturation, growth and mortality of clupeid and
engraulid stocks in relation to fishing. Rapp. P. -v. Cons,
int. Explor. Mer 154-44-6J.
1964. Differential catchability of male and female plaice
in the North Sea and its effect on estimates of
abundance. Cons. Perm. Int. Explor. Mer Rapp. et
P.-v. 155, p. 103-112.
Beverton, R J H , and S J Holt
1957. On the dynamics of exploited fish popu-
lations. Fish. Invest. Minist. Agric, Fish. Food (G. B.),
Ser. II, 19:21-42.
1959. A review of the lifespan and mortality rates of fish
in nature and their relation to growth and other
physiological characteristics. In CIBA Foundation.
Colloquium on aging V. The lifespan of animals, p.
142-180. Churchill, Lond.
Bishop, Y.
1959. Errors in estimates of mortality obtained from
virtual populations. J. Fish. Res. Board Can. 16:73-90.
Blinov, V V
1977. The connection between natural mortality and re-
production of food fishes and modeling of the processes
involved. J. Ichthyol. 16:1033-1036.
BoiKO, E G
1964. Evaluating natural mortality in the Azov Zander
iStizostedion lucioperca ). Fish. Res. Board Can. Transl.
Ser. No. 541, 15 p.
Brownie. C , D. R Anderson, K Burnham, and D. S. Robson.
1985. Statistical inference from band recovery data - a
handbook. 2d ed. U.S. Fish Wildl. Serv. Resour. Publ.
# 156.
BULGAKOVA. T I. AND Y N EFIMOV.
1982. A method for forecasting catch by consideration of
the dependence of natural mortality on age. J. Ichthyol.
22:24-32.
Burnham, K., and D. R Andersen.
1984. Tests of compensatory vs. additive hypotheses of
mortality in mallards. Ecology 65:105-112.
BuRHAM, K , G C White, and D R Anderson
1984. Estimating the effect of hunting on annual survival
rate of adult mallards. J. Wildl. Manage. 48:350-361.
Butler, S A , and L. L MacDonald.
1979. A simulation study of a catch-effort model for esti-
mating mortality rates. Trans. Am. Fish. Soc. 108:353-
357.
Caddy, J F
1984. Method of solving for natural mortality rate from
stocks with different schedules of growth and mortal-
ity. Can. J. Fish. Aquat. Sci. 41:1226-1230.
Carpenter, S R , J F. Kitchell. and J R Hodgson
1985. Cascading trophic interactions and lake productiv-
ity. Bioscience 35:634-639.
Chapman, D G , and G I Murphy
1965. Estimates of mortality and population from survey-
removal records. Biometrics. 21:921-935.
Chatwin, B M
1958. Mortality rates and estimates of theoretical yield in
relation to minimum commercial size of ling cod (.Ophid-
ian elongatus ) from the Strait of Georgia, British Colum-
bia. J. Fish. Res. Board Can. 15:831-849.
Craig. J F
1984. Aging in fish. Can. J. Zool. 63:1-8.
Gushing, D. H
1967. The possible density-dependence of larval mortality
and adult mortality in fishes. In J. H. S. Blaxter (edi-
tor), The early life history of fishes, p. 103-
111. Springer- Verlag, N.Y,
40
VETTER: NATURAL MORTALITY IN FISH STOCKS
1975. The natural mortality of the plaice. J. Cons. Int.
Explor. Mer 36:150-157.
Deriso. R B
1980. Harvesting strategies and parameter estimation for
an age-structured model. Can. J. Fish. Aquat. Sci.
37:268-282.
1982. Relationship of fishing mortality to natural mortal-
ity and growth at the level of maximum sustainable
yield. Can. J. Fish Aquat. Sci. 39:1054-1058.
DOUBLEDAY. W G
1976. A least squares approach to analyzing catch at age
data. ICNAF Res. Bull. No. 12, p. 69-81.
DOUBLEDAY. W G , AND T D BEACHAM
1982. Southern Gulf of St. Lawrence cod: a review of
multispecies models and management advice. In M. C.
Mercer (editor), Multispecies approaches to fisheries
management advice, p. 133-140. Can. Spec. Publ. Fish.
Aquat. Sci. No. 59,
Efimov, Y N
1976. Assessment of the magnitude of the reserves of
Pacific hake by the virtual population method. [In
Russ.) In Abstracts of All-Union Conference of Young
Scientists, Fisheries Exploitation in the Resources of the
Worlds Oceans.
Forney. J L
1977. Reconstruction of yellow perch iPerca flavescens)
coherts from examination of walleye iStizostedwn vit-
reum vitreum ) stomachs. J. Fish Res. Board Can.
34:925-932.
FOURNIER, D . AND C P ARCHIBALD
1982. A general theory for analyzing catch at age
data. Can. J. Fish. Aquat. Sci. 39:1195-1207.
Francis. R C
1974. TUNPOP. a computer simulation model of the yel-
lowfin tuna population and the surface tuna fishery of the
eastern Pacific Ocean. Inter-Am. Trop. Tuna Comm.,
Bull. 16:234-258.
1976. Relationship of fishing mortality to natural mortal-
ity at the level of maximum sustainable yield under the
logistic stock production model. J. Fish. Res. Board
Can. 31:1539-1542.
GULLAND. J A
1983. Fish stock assessment. A manual of basic meth-
ods. FAOAViley series on food agriculture. Vol 1. John
Wiley and Sons, N.Y., 223 p.
Gunderson, D R
1980. Using R-K selection theory to predict natural mor-
tality. Can. J. Fish. Aquat. Sci. 37:2266-2271.
Heincke, F
1913. Investigations on the Plaice. General report 1. The
plaice fishery and protective regulations. Rapp. Cons.
Explor. Mer 17A, p. 1-153.
Henderson. B A . J Collins, and J. A Reckahn
1983. Dynamics of an exploited population of whitefish
(Coregonus dupeaformis I in Lake Huron. Can. J. Fish.
Aquat. Sci. 40:1556-1567.
Hennemuth.R. C.
1961. Year class abundance, mortality and yield-per-
recruit of yellowfin tuna in the eastern Pacific Ocean,
1954-1959. Inter-Am. Trop. Tuna Comm., Bull. 6:1-51.
Hewitt. R . G. H Theilacker. and N C H Lo
1985. Causes of mortality in young jack mackerel. Mar.
Ecol. Prog. Ser. 26:1-10.
Hochbaum. G S . AND C J Walters.
1984. Components of hunting mortality in ducks: a man-
agement analysis. Can. Wildl. Serv, Occas. Pap. # 52.
Hoenig, J
1983. Empirical use of longevity data to estimate mortal-
ity rates. Fish. Bull., U.S. 82:898-903.
Hunter, J R
1984. Inferences regarding predators in the early life
stages of cod and other fishes. In E. Dahl, D. S.
Danielsen, E. Moksness, and P. Solemdal (editors), The
propagation of cod Gadus morhua . L. Flodevigen Rapp.
1:533-562.
Jensen, A L
1984. Non-linear catch curves resulting from variation in
mortality among subpopulations. J. Cons. Int. Explor.
Mer 41:121-124.
Jones. R
1979. Materials and methods used in marking experi-
ments in fishery research. FAO Fish. Tech. Pap. No.
190, 134 p.
1982. Ecosystems, food chains, and fish yield. In D.
Pauly and G. Murphy (editors). Theory and management
of tropical fisheries, p. 195-240. ICLARM Conf Proc. 9.
Jones. R . and C Johnston
1977. Growth, reproduction, and mortality in gadoid fish
species. In J. H. Steele (editor). Fisheries mathemat-
ics, p. 37-62. Acad. Press, N.Y.
Kennedy, W A
1953. Growth, maturity, fecundity and instability in un-
exploited whitefish iCoregonus dupeaformis) of Great
Slave Lake. J. Fish. Res. Board Can. 10:413-441.
1954. Growth, maturity, and mortality in the relatively
unexploited lake trout, (Cristivomer namaychush),
of Great Slave Lake. J. Fish. Res. Board Can. 11:827-
852.
1963. Growth and mortality of whitefish in three unex-
ploited lakes in northern Canada. J. Fish. Res. Board
Can. 20:265-272.
Laevastu. T , F Favorite. H A Larkin.
1982. Resource assessment and evaluation of the dynam-
ics of the fisheries resources in the northeastern Pacific
in numerical ecosystem models. In M. C. Mercer (edi-
tor), Multispecies approaches to fisheries management
advice, p. 70-81. Can. Spec. Publ. Fish. Aquat. Sci. 59.
Lander. R H
1962. A method of estimating mortality rates from
changes in composition. J. Fish. Res. Board Can.
19:159-168.
Larkin, P. A , and W Gazey.
1982. Applications of ecological simulation models to
management of tropical multispecies fisheries. In D.
Pauly and G. Murphy (editors). Theory and management
of tropical multispecies fisheries, p. 123-140. ICLARM
Conf Proc. 9.
Lester. R J G
1984. A review of methods for estimating mortality due to
parasites in wild fish populations. Helgol. Meersunters.
37:53-64.
MA.JK0WSKI, J.
1981. Application of a multispecies approach for assessing
the population abundance and age-structure of fish
stocks. Can. J. Fish. Aquat. Sci. 38:421-431.
Mann. R H K
1973. Observations on the age, growth, reproduction, and
food of the roach, Rutilus rutilus (L.), in two rivers in
England. J. Fish. Biol. 5:707-736.
Mercer, M C
1982. Multispecies approaches to fisheries management
advice. Can. Spec. Publ. Fish. Aquat. Sci. 59, 169 p.
41
FISHERY BULLETIN; VOL. 86, NO. 1
MUNRO, J L
1982. Estimation of biological and fishery parameters in
coral reef fisheries. In D. Pauly and G. I. Murphy (edi-
tors), Theory and management of tropical multispecies
fisheries, p. 71-83. ICLARM Conf Proc. 9.
Murphy, G. I.
1965. A solution to the catch curve. J. Fish. Res. Board
Can. 22:1965.
Murphy. T. C, and G T. Sakagawa
1977. A review and evolution of estimates of natural mor-
tality rates of tunas. In ICCAT Coll. Vol. Sci. Pap.
VI(SCRS-1976):117-123.
Myers, R A., and Doyle, R W.
1983. Predicting natural mortality rates and
reproduction-mortality trade offs from life-history
data. Can. J. Fish. Aquat. Sci. 40:612-620.
Paloheimo, J E
1961. Studies on estimation of mortalities I. Comparison
of a method described by Beverton and Holt and new
linear formula. J. Fish. Res. Board Can. 18:645-662.
1980. Estimation of mortality rates in fish popula-
tions. Trans. Am. Fish. Soc. 109:378-386.
1982. Estimation offish mortality. Trans Am. Fish. Soc.
111:535-537.
Paloheimo, J E , and L M Dickie.
1966. Food and growth of fishes III. Relations among food,
body size, and growth efficiency. J. Fish. Res. Board.
Can. 23:1209-1247.
Parks, W W
1977. Cohort and equilibrium yield-per-recruit analyses
for the Atlantic bluefin tuna fisheries system, accounting
for two system configurations and two natural mortality
models. ICCAT Coll. Vol. Sci. Pap. VI:385-401.
PAULIK, J , AND D S ROBSON.
1969. Statistical calculations for change-in-ratio estima-
tors of population parameters. J. Wildl. Manage. 33:1-
27.
Pauly, D
1980. On the interrelationships between natural mortal-
ity, growth parameters, and mean environmental tem-
perature in 175 fish stocks. J. Cons. Int. Explor. Mer
39(2):175-192.
1982. Studying single-species dynamics in a tropical mul-
tispecies context. In D. Pauly and G. Murphy (editors).
Theory and management of tropical fisheries, p. 33-
70. ICLARM Conf Proc. 9.
Pauly, D., and G. Murphy (editors).
1982. Theory and management of tropical fish-
eries. ICLARM Conf Proc. 9, 360 p. International
Center for Living Aquatic Resources Management,
Manilla, Phillipines.
PELLA, J J . AND P K TOMLINSON.
1969. A generalized stock-production model. Inter-Am.
Trop. Tuna. Comm., Bull. 13:416-496.
Peterson. J , and J S Wroblewski
1984. Mortality rate of fishes in the pelagic ecosys-
tem. Can. J. Fish. Aquat. Sci. 41:1117-1120.
Pope. J G
1971. Note on cohort analysis and age-specific fishing
mortality. ICCAT Tuna Meeting Working Paper, 2 p.
1972. An investigation of the accuracy of virtual popula-
tion analysis using cohort analysis. ICNAF Res. Bull. 9,
p. 65-74.
Pope, J G , and D J Garrod.
1973. A contribution to the discussion of the effect of error
on the action of catch quotas and effort quotas. Ann.
Meet. Inst. Comm. Northwest Atl. C. Fish. 1973, Res.
Doc. 119, Serial #3074.
Pope, J. G , and B J Knights.
1982. Simple models of predation in multispecies fisheries
for considering the estimation of fishery mortality and its
effects. In M. C. Mercer (editor), Multispecies ap-
proaches to fisheries management advice, p. 64-
69. Can. Spec. Publ. Fish. Aquat. Sci. 59.
Pope, J. G., and J, G Shepard
1982. A single method for the consistent interpretation of
catch-at-age data. J. Cons. Int. Explor. Mer 40:176-
184.
Reed, R. B., and W D Davies
1980. Estimating rate of exploitation from tag returns
and fishing effort. Proc. Conf Southeast. Assoc. Fish.
Wildl Agencies 34:151-157.
RICKER, W. E
1945. Natural mortality among Indiana bluegill sun-
fish. Ecology 26:111-122.
1947. Mortality rates in some little-exploited populations
of fresh-water fishes. Trans. Am. Fish. Soc. 71:114-128.
1969. Effects of size-selective mortality and sampling bias
on estimates of growth, mortality, production, and
yield. J. Fish. Res. Board Can. 26:479-541.
1971. Derzhavin's biostatistical method of population
analysis. J. Fish. Res. Board Can. 28:1666-1672.
1975. Computation and interpretation of biological statis-
tics offish populations. Bull. Fish Res. Board Can. 191,
382 p.
1977. The historical development, /n J. A. Gulland (edi-
tor). Fish population dynamics, p. 1-26. Wiley-
Interscience, Lond.
RoBSON. D S., AND D G Chapman
1961. Catch curves and mortality rates. Trans. Am.
Fish. Soc. 90:181-189.
ROFF, D A
1984. The evolution of life history parameters in teleosts.
Can. J. Fish. Aquat. Sci. 41:989-1000.
1986. Predicting body size with life history model.
Bioscience 6:316-323.
Sandland, R. L.
1982. Estimation, inference, and data analysis for log-
linear models in tagging studies. J. Cons. Cien. 40:290-
303.
Schaefer, M B
1954. Some aspects of the dynamics of populations impor-
tant to the management of commercial marine fish-
eries. Inter-Am. Trop. Tuna. Comm., Bull. l(2):27-56.
Schupp. D
1978. Walleye abundance, growth, movement and yield in
disparate environments in a Minnesota lake. Am. Fish.
Soc. Spec. Pub. 11:58-65.
Seber, G a F.
1973. The estimation of animal abundance and related
parameters. Griffin, Lond., 506 p.
Sette, O E
1943. Biology of the Atlantic mackerel (Scomber scom-
brus) of North America. U.S. Fish. Wildl. Serv., Fish.
Bull. 50:149-237.
Silliman, R p.
1943. Studies on the Pacific pilchard, or sardine
iSardinops caerula ) 5. A method of computing mortali-
ties and replacements. U.S. Fish Wildl. Serv., Spec. Sci.
Rep. 24, 10 p.
Sims. S. E
1982a. The effect of unevenly distributed catches on stock
42
VETTER: NATURAL MORTALITY IN FISH STOCKS
size estimates using virtual population analysis (cohort
analysis). J. Cons. Int. Explor. Mer 40:47-52.
1982b. Algorithms for solving the catch equation forward
and backward in time. Can. J. Fish. Aquat. Sci. 39:197-
202.
1984. An analysis of the effect of errors in the natural
mortality rate on stock-size estimates using virtual popu-
lation analysis (cohort analysis). J. Cons. Int. Explor.
Mer 41:149-153.
SISSENWINE. M P
1984. Why do fish populations vary? In R. M. May (edi-
tor). Exploitation of marme communities, p. 59-94.
Springer-Verlag, N.Y.
Smith. P E
1985. Year-class strength and survival of 0-group clu-
peoids. Can, J. Fish. Aquat. Sci. 42:69-82.
Stein. R A , R F Carline. and R S Hayward.
1981. Largemouth bass (Micropterus salmoides) preda-
tion on stocked tiger muskellunge. Trans. Am. Fish.
Soc. 110:604-612.
Taylor, C C
1958. Natural mortality rate of Georges Bank had-
dock. U.S. Fish Wildl. Serv., Fish Bull, 58:1-7,
Theilacker. G H
1986. Starvation-induced mortality of young sea-caught
jack mackerel, Trachurus synimetricus, determined with
histological and morphological methods. Fish, Bull,,
U,S, 84:1-17,
Tyler, A V . and V Gallucci
1980. Dynamics of fished stocks. In R. T. Lackey and
L. A. Nielsen (editors). Fisheries management, p. 111-
149, John Wiley & Sons, NY,
Tyler, A V , L L Sebring, M C Murphy, and L F. Murphy.
1985. A sensitivity analysis of Deriso's delay-difference
equation using simulation. Can. J. Fish. Aquat. Sci.
42:836-841.
Ulltang. O
1977. Sources of errors in and limitations of virtual popu-
lation analysis (cohort analysis). J. Cons. Int. Explor.
Mer 37:249-260.
Ursin, E
1967, A mathematical model of some aspects of fish
growth, respiration and mortality, J, Fish, Res, Board
Can, 24:2355-2452,
1982, Multispecies fish stock and yield assessment in
ICES, In M, C, Mercer (editor), Multispecies approaches
to fisheries management advice, p, 39-47, Can, Spec.
Publ, Fish, Aquat, Sci, 59,
VOOREN.C M
1977, Growth and mortality of tarahiki (Pisces: Cheilo-
dactylidae) in lightly exploited populations. N.Z. J.
Mar. Freshwater Res. ll(l):l-22.
Walters, C J
1969, A generalized computer model for fish population
studies. Trans, Am, Fish, Soc, 98:505-512,
Ware, D M.
1975. Relation between egg size, growth and natural mor-
tality of larval fish. J. Fish. Res. Board Can. 32:2503-
2512,
Williams, W P
1967, Growth and mortality of 4 species of fish in the
Rivers Thames and Reading. J, Anim, Ecol, 266:695-
720,
Wohlschlag, D E
1954, Mortality rates of whitefish in an Arctic
lake. Ecology 35:388-396,
Woodhead, a D
1979, Senescence in fish. In P, J, Miller (editor), Fish
phenology: Anabolic adaptations in teleosts, p, 179-
205, Symp, Zool, Soc, Lond, 44,
43
NATURAL HISTORY OF THE RAYS OF THE GENUS MOBULA
IN THE GULF OF CALIFORNIA
Giuseppe Notarbartolo-di-Sciara1
ABSTRACT
Mobulid rays, which abound during summer in the southern Gulf of California, southern Baja
California, Mexico, were monitored for a period of four years during a study of their biology. A total
of 262 specimens belonging to four species of Mobula were examined. Mobula thurstoni was the most
abundant l58'/f of the catch), followed by M. japanica (30%), M. munkiana (9'5f ), and M. tarapacana
(3'^). The study area served as a nursery ground for M. thurstoni, a summer feeding and mating
ground for M. thurstoni and M. japanica, and a wintering ground for M. munkiana and young
M. thurstoni ; M. tarapacana was rare. Data on size, weight, sex ratio, life history, seasonality, feeding
habits, behavior, habitat, and symbionts are presented for each species. Size segregation was a
common feature of M. thurstoni , M. japanica , and M. munkiana ; sex segregation was not evident. An
extreme degree of feeding specialization was noted: summer prey were almost exclusively the eu-
phausiid Nyctiphanes simplex: the mysid Mysidium sp. dominated in the winter. A key to the genus
Mobula in the Gulf of California is presented as an aid for species identification.
This paper reports on natural history aspects of
rays of the genus Mobula (Mobulidae), a poorly
known group of elasmobranchs commonly called
manta rays or devil rays, frequent in the Gulf of
California. A good early overview of the family
Mobulidae was given by Gill (1908). Cadenat
(1960) described the natural history of the mobu-
lids of tropical west African waters, based on
specimens which were occasionally captured by
the local fishermen. However, with the exception
of observations carried out with some regularity
by Coles (1910, 1913, 1915, 1916a, 1916b) of Mo6-
ula olfersi (= M. hypostoma ) and Manta birostris
off North Carolina, most of the available litera-
ture is purely anecdotal and deals with occasion-
ally encountered or harpooned specimens. Long-
term field investigations of devil ray ecology and
behavior are wanting. As a result, mobulids are
among the least known of the batoid taxa. This
was recognized by Bigelow and Schroeder (1953)
in their comprehensive review of the knowledge
of this family. No major contribution to the
understanding of any aspect of mobulid biology
has since been published.
Regular fisheries for mobulids were not known
to exist, because mobulid meat is generally con-
sidered of little market value. However, in 1981,
during a reconnaissance trip to the southern Gulf
iScripps Institution of Oceanogi'aphy, University of Califor-
nia, San Diego, La Jolla, CA 92093; present address: Museo
Civico di Storia Naturale, corso Venezia 55, 20121 Milano,
Italy.
Manu.script accepted October 1987.
FISHERY BULLETIN: VOL. 86, NO. 1. 1988.
of California (Mexico), in the vicinity of La Paz,
Baja California Sur, a regular, seasonally impor-
tant fishery was discovered. This activity afforded
the opportunity to study several aspects of the
natural history and the ecology of these batoids.
Preliminary oral interviews revealed that local
fishermen in the Gulf of California knew of, and
routinely captured, four species of devil rays, in
addition to the well-known giant manta ray,
Manta birostris. This information contrasted
with the scientific literature, where only two
mobulid species, Manta birostris and Mobula lu-
casana, were reported for the area (Beebe and
Tee-Van 1941; Fowler 1944; Castro-Aguirre
1965). The confusing state of mobulid taxonomy
demanded a revisionary work of the genus Mob-
ula (Notarbartolo-di-Sciara 1987), and a discus-
sion of the systematics of Manta in the eastern
Pacific (Notarbartolo-di-Sciara in press). Such ef-
fort permitted designation of names for all species
of Mobula found in the Gulf of California: M.
thurstoni (Lloyd 1908), of which M. lucasana
Beebe and Tee Van (1938) is a junior synonym;
M. japanica (Miiller and Henle 1841); M. tara-
pacana (Philippi 1892); and M. munkiana
Notarbartolo-di-Sciara (1987), which had not
been described before. Many of the reports of
M. lucasana { = M. thurstoni ) from Central and
South America (Beebe and Tee- Van 1941; Fowler
1944; Nichols and Murphy 1944; Barton 1948;
Castro-Aguirre 1965; Chirichigno 1974; Pequeno
1983) undoubtedly refer to other species of Mob-
ula . A key to the genus Mobula in the Gulf of
45
F1SHP:RY BULLETIN: VOL 86. NO 1
0)
3
a:
D
O
46
NOTARBARTOLO-DI-SCIARA: NATURAL HISTORY OF MOBULA
California is presented as an aid to future studies
of mobulids from this region.
Working relationships were established with
the local fishing communities, and their activities
were intermittently monitored between 1981 and
1984. Captured rays were examined and mea-
sured before their pectoral fins were filleted;
stomach contents and reproductive organs were
examined later. Information was gathered on
size, weight, sex ratios, life history, seasonality,
feeding habits, habitat, behavior, and symbionts
of four species of rays belonging to the genus Mob-
ula (M. thurstoni, M. japanica, M. munkiana,
and M. tarapacana ). Detailed descriptions and
morphometries of those species are given by
Notarbartolo-di-Sciara (1987). The manta ray,
Mania birostris, was also occasionally captured
(Notarbartolo-di-Sciara in press), but is not
treated in the present study.
METHODS
Although mobulids are locally said to be abun-
dant on both sides of the southern Gulf of Califor-
nia, for logistic convenience collecting trips were
made only to the peninsular coast (Fig. 1). The
fishing cooperative based at Punta Arena de la
Ventana was selected as the prime collecting site,
because mobulids were caught there more consis-
tently than at other localities. Fishing camps on
Isla El Pardito, at Cueva de Leon, Ensenada de
los Muertos, and Bahia de los Frailes were also
sources of study material. Other fishing commu-
nities, such as Juncalito, San Evaristo, El Sar-
gento. La Ventana, and San Jose del Cabo were
occasionally visited, but yielded no data because
mobulids were not specifically sought by the fish-
ermen. Seven field trips were made. Six were
short-term (24 January-8 February 1981, 25
November 1981, 16-21 December 1981, 20-23
December 1982, 19-26 January 1984, and 28 Oc-
tober-1 November 1984); one lasted almost six
months (26 January-15 July 1983).
Mobulids of all available species and sizes are
caught with nets and harpoon; their meat is fil-
leted out of the pectoral fins for human consump-
tion and used as shark bait. Gill nets are either
strung just under the surface or are set on the
bottom perpendicular to shore, usually at depths
between 10 and 200 m. Fishing vessels were 5-7
m fiberglass launches, locally called "pangas",
powered by an outboard engine. Fishing occurred
within a radius of about 15 km from a base camp.
Nets are checked once a day, early in the morn-
ing. Rays weighing up to approximately 100 kg
were hauled on board, larger specimens were
towed ashore. Rays that were dead in the nets,
after several hours (i.e., three unsexed specimens
of Mobula thurstoni), were often partially de-
stroyed by gammarid amphipods (locally called
"plaga"), and were unmarketable.
Specimens were weighed and measured before
being processed by the fishermen. Weights (WT)
were taken with calibrated spring-scales. Rays
lighter than 20 kg were weighed to the nearest
pound with a 50-lb scale; weights were subse-
quently converted to kg. Heavier rays were
weighed to the nearest kg with a 150 kg scale.
Specimens which exceeded 150 kg (all postnatal
M. tarapacana ) were cut in four pieces and
weighed separately. Ten percent was then added
to the total weight to compensate for body fluid
loss. All the specimens could not be weighed, as
occasionally a large number of rays were beached
simultaneously, and because of the intense heat
the fishermen could not delay their processing.
A set of 29 measurements was taken for mor-
phometric analysis and systematic purposes.
Methods and results are presented in
Notarbartolo-di-Sciara (1987). Measurements
relevant to the present paper were disc width
(DW), greatest dimension between outermost tips
of pectoral fins, pelvic fin length, from anterior
margin of vent to tip of pelvic, and clasper length,
from anterior margin of vent to tip of clasper.
Most specimens were discarded after measuring
and sampling. All preserved specimens were
deposited in the Marine Vertebrate Collection of
the Scripps Institution of Oceanography. Raw
data listing all specimens examined and pre-
served can be found in Notarbartolo-di-Sciara
(1985).
The size and shape of the testes were inspected
in male specimens, and the ducti deferentes were
cut slightly above the genital papilla. Presence or
absence of seminal fluid was determined by
running a finger in the caudal direction over the
ducts anterior to the cut. Clasper length in
thousandths of disc width (DW) was plotted
against DW to determine size at maturity, and
the presence or absence of seminal fluid was
noted. Relative size and contents of uteri and
nidamental glands were examined in female spec-
imens, and right and left ovaries were compared.
The diameter of the largest ovum was plotted
against DW to determine size at maturity of fe-
male Mobula. Eggs were extracted from the
germinative epithelium and their greatest di-
47
KISHKRY [BULLETIN: VOL 86, NO 1
ameter was measured to the nearest 0.1 mm with
a steel dial caliper.
Stomach contents, if not larger than approxi-
mately 200 cc, were sampled whole; otherwise,
the bolus was made homogeneous by stirring, and
about 200 cc were preserved. Stomach content
samples were fixed and preserved in 10% buffered
formalin. The stomach content of each sample
was thoroughly agitated, separated with a plank-
ton strainer (mesh size 0.5 mm), rinsed of forma-
lin in deionized water, and blotted for 30 seconds
on blotting paper. The lump was then molded into
a cylindrical shape, and a portion of one end was
separated to make up 1 g of wet weight, measured
to the nearest 0.1 g. The subsample was then
placed with water in a gridded tray, and exam-
ined under a dissecting microscope. Contents of
the spiral intestine were discarded, because the
small crustacean prey was rapidly digested.
Feeding habits were analyzed quantitatively
by computing the Index of Relative Importance
(%IRI) (Pinkas et al. 1971; Hyslop 1980) for each
prey species. The IRI combines percentage by
number iN), mass (M), and frequency of occur-
rence (F) in the formula:
mi = (%N + %M) X %F
Prey items were identified, when possible, to
lowest taxa or species, then the %N of all prey
species within each subsample was calculated.
When more than one species was present, all
items were individually counted. To obtain the
%M term of the equation, mean mass was calcu-
lated for each species by measuring the length of
each item contained in five randomly selected
squares on a tray, calculating the average length
of each prey species, and obtaining mass values
from Miller's (1966) Plankton Conversion Tables,
where mass is related to length for all main
planktonic taxa. The %N from all subsamples
were summed, and the percent from the new sum
was calculated, to calculate %IRI for each prey
species. The same procedure was applied to %M
and 7cF. The total 7(N was then added to the total
7cM, and that sum was multiplied by the total
%F, to obtain total IRI for each prey species, from
which the %IRI was calculated. When few items
(e.g., copepods) were found among a large amount
of partially digested euphausiid or mysid
shrimps, the possibility of reconstructing the
shrimp number within the subsample by count-
ing the digestion-resistant eye pairs was dis-
carded to avoid bias in favor of the shrimp frac-
tion. The following method was adopted instead:
all odd prey items were counted and measured,
and their total mass was obtained from Miller's
tables; this was subtracted from the total weight
(1 g) of the subsample. The remaining weight was
divided by the mean weight of each individual
item, calculated by averaging the lengths of all
available intact specimens, and obtaining the cor-
responding weight in Miller's tables. A potential
biasing factor existed, when only a few prey re-
mains were found (e.g., when a relatively uncom-
mon item occurred alone in a stomach, therefore
contributing a value of 100 7(N and 7<M to the
total IRI). This factor was avoided by considering,
for quantitative treatment, only those stomachs
which contained more than 1 g (wet weight) of
recognizable food. All stomachs that had <1 g of
contents were in fact virtually empty, and the few
items found in them were treated only qualita-
tively.
Remoras were fixed in 10% formalin, preserved
in 50% isopropanol, and deposited in the Marine
Vertebrate Collection, Scripps Institution of
Oceanography. Parasitic copepods were fixed and
preserved in 50% isopropanol and sent to the
Long Beach State University for taxonomic iden-
tification and study.
RESULTS AND DISCUSSION
A total of 262 mobulid rays, belonging to four
species iMobula thurstoni, M. japanica, M.
miinkiana, and M. tarapocana) were examined
between 1981 and 1984. Of these, M. thurstom
was the most abundant species, constituting 58%
of the total catch, followed by M. japanica (30%),
M. munkiana (9%), and M. tarapacana (3%).
KEY TO THE SPECIES OF MORULA IN
THE GULF OF CALIFORNIA
Mobula can be distinguished from Manta by the
mouth on the lower surface of the head rather
than being subterminal, and by the presence of
toothbands in both jaws. Moreover, Manta grows
to a greater size, and the size of its head, relative
to the body, is much greater than in Mobula.
la. Branchial filter plates fused
M. tarapacana
(Spiracle in an elongated longitudinal
slit, dorsal to plane of pectoral fins. Teeth
tessellated, surface of crown pitted, buc-
cal edge comblike. Thick cover of acumi-
48
NOTARBARTOLO-DI-SCIARA: NATURAL HISTORY OF MOHri.A
nate denticles. Large adult size, often
exceeding 3 m in width. Dorsal side
greenish brown. Ventral side anteriorly
white, posteriorly gray).
lb. Branchial filter plates separate 2
2a. Spine on base of tail M. japanica
(Tail very long with a line of white tuber-
cles on both sides. Spiracle a short
transversal slit, dorsal to plane of pec-
toral fins. Teeth not in contact with each
other, at least twice as high as the crown
is wide; crown subtriangular in apical
view. Medium-large adult size, approach-
ing 2.5 m in width. Dark blue to black on
dorsal side, apex of dorsal fin white; ven-
tral side white. Skin rough to the touch).
2b. No spine on base of tail 3
(Spiracle small, subcircular, ventral to
plane of pectoral fins. Skin smooth to the
touch).
3a. Base of tail dorsally depressed
M. thurstoni
(Double curvature of anterior margin of
pectoral fin. Surface of tooth crown
rugose. Medium adult size, exceeding 1.8
m of width. Dark blue to black on dorsal
side, apex of dorsal fin white; ventral side
white, with a dark greenish patch near
the posterior margin of each pectoral fin,
and with a pattern of dark and shiny sil-
very pigmentation on distal half of pec-
torals).
3b. Base of tail laterally compressed
M. munkiana
(Anterior margin of pectoral fin straight
to weakly convex. Surface of tooth crown
smooth. Small adult size, barely exceed-
ing 1 m of width. Dorsum mauve gray;
ventral side white, tips of pectoral fins
gray).
Mobula thurstoni (Lloyd 1908)
Local name: cubana de lomo azul
Eighty one males (disc width range 630-1,770
mm), 69 females (210-1,801 mm), and three spec-
imens of undetermined sex (941-1,494 mm), were
caught at four stations (Punta Arena de la Ven-
tana. Cueva de Leon, Ensenada de los Muertos,
and Isla El Pardito) and their adjacent waters,
between 7 February 1981 and 30 October 1984.
Overall and seasonal size-frequency distributions
for M. thurstoni are given in Figure 2. Mean size
varied with season, smaller rays being dominant
in winter catches, medium sizes prevailing in the
summer. The difference between mean disc width
(DW) in winter and overall mean DW is highly
significant (T-value = 5.189, df=169, P
<< 0.001). There were no significant differences
between male and female DWs (T-value
= 0.3767, df = 145, P > 0.5), with the exception of
the November to February period, when females
were larger ( T-value = 2.331, df = 12, P < 0.05).
A total of 105 specimens (210-1,770 mm DW)
were weighed. The WT/DW relationship is best
described by the equation:
WT - 4.817 X 10-8 ^Y)^^)2.^8
r = 0.99
WT is given in kg, DW in mm. The largest speci-
men in the sample was a female; DW was not
measured because the fishermen had already
started filleting the pectoral fins. Calculated DW,
regressed from disc length, cranial width, and
upper toothband length, was 1,801 mm (multiple
correlation coefficient = 0.99). The second largest
specimen was also a female, 1,799 mm DW. The
largest male had a DW of 1,770 mm and weighed
53 kg. The smallest freshly caught specimen was
876 mm DW and weighed 6.4 kg. The smallest
postnatal specimen was a male, the carcass of
which was found drying on the beach in Ensenada
de los Muertos. Its calculated DW, regressed from
toothbands length, was 864 mm (multiple correla-
tion coefficient = 0.99).
Overall ratio of males to females caught was
1.18 {N = 148). Catch sex ratios varied with sea-
son. Females appeared to be dominant in winter
(ratio of males to females 0.27; A'^ = 14). The re-
verse was true in March, in favor of males. A
significant difference from a 1:1 ratio (chi square
testP > 0.05) was not noted. Geographical segre-
gation, either of sex or size, was not apparent for
M. thurstoni during the warmer months when a
wide array of size classes and both sexes were
found in the same fishing area. Males and fe-
males were occasionally harpooned from the same
group basking at the surface. This fact argues
against behavioral sex segregation. Winter data,
however, were suggestive of size segregation at
that time of year. It was common knowledge
among local fishermen that during the winter
months all M. thurstoni caught are small. The
bimodal size-frequency distribution for early
49
FISHERY BULLETIN: VOL. 86, NO. 1
OVERALL
Xtot= 13456 ±4022 (N=I48)
V,f
13449+5513 {N-79)
30 -] H% - I356.3±59.27(N = 66)
20-
10-
z
UJ
o
UJ
q:
u.
lij
o
V)
CD
<
30nx,<„:
20-
10-
'•%%
MARCH -APRIL
1211 1±I52 30 (N = I5)
II28.81I6I 34(N = I0)
I346.0±370.37(N = 4)
£22.
rr-irT^rrr-iEin. i .h^.
30
20-
10-
JUNE
Xtot= 1401 22+39.45 (N=76)
X^j.:l38l42±55 22(N^43)
Xjj:l42703±55 I0(N = 33)
NOVEMBER -FEBRUARY
Xtot = 95969±42 75(N-I6)
Xj.jt = 86733±48l(N-3)
X^j =987 46±52 26(N = II)
D TOTAL
E3 ??
□73. 1
MAY
Xtot= 1435 4 + 69 87 (N = 37)
Xj.^ = 1436.61 107.98 (N: 20)
Xj^ = l433.9±87.27(N=l7)
r^m. , .mm
i
lA.
JULY
X,0T = 15I275±341.72(N = 4)
X^^ = 1409 67 ± 385.40 ( N = 3)
K^
:|80I(N=I)
950 1150 1350 1550 1750 950 1150 1350 1550 1750
DISC WIDTH (mm)
Figure 2. — Size-frequency distributions of Mobula Ihurstoiii (means ±2 SE).
spring (Fig. 2) suggested that larger rays began to
move into the area in the spring from their un-
known wintering grounds.
Mature testes are large, elongated structures
attached by the mesorchia to the upper anterior
wall of the pleuroperitoneal cavity, on either side
of the vertebral column. A large epigonal organ is
associated with each testis. Both testes appeared
to be functional. They were usually about the
same size, although occasionally the left testis
was nearly 25''f larger. The paired ducti defer-
entes remain separated from each other through-
out their length and open into the cloaca at the tip
of the urogenital papilla through two distinct
pores, rather than merging in the urogenital
sinus, as in most elasmobranchs (Daniel 1934),
including Manta ehrcnhcrgi (Gohar and Bayoumi
1959). Clasper length was plotted against DW for
43 M. thurstoni (Fig. 3), to determine the onset of
male maturation. Rapid increase in relative size
of the claspers, beginning at a DW of about 1,500
mm, was concomitant with the incipient presence
of abundant seminal fluid in the lower portion of
each ductus deferens. The pelvic fin area, and
50
NOTARBARTOLO-ni-SCIARA NATl'RAl. HISTORY OV MORULA
especially the tissue at the bases of claspers, of
the larger males appeared swollen and congested
in May and June, and some of the skin had as-
sumed a pink coloration. Similar observations in
carcharhinid and odontaspid sharks have been
linked with mating activities (Springer 1960;
Gilmore et al. 1983).
The ovaries in M. thursfoni are paired, elon-
gated organs located inside the pleuroperitoneal
cavity, analogously to the testes, and are con-
nected to a large epigonal organ. Eggs are pro-
duced within the germinative epithelium. The
largest eggs were found at the anterior end of the
ovary. Only the left ovary develops and is func-
tional, whereas the size and aspect of the right
ovary remains comparable with those of the im-
mature stage. Asymmetry is also present in the
oviducts, the left uterus being usually the largest
in mature females. It consists of a voluminous,
thick-walled expansion of the lower tract of the
oviduct; its lumen is lined with a highly devel-
oped epithelium consisting of elongated, flattened
villi (trophonemata), a well-known mobulid (Gill
1908; Setna and Sarangdhar 1950; Wourms 1977)
and rhinopterid (Schwartz 1966; Smith and Mer-
riner 1986) feature. In several instances both
uteri were found to contain a viscous, whitish or
greenish substance. Oviducts open separately
into the cloaca. A progression of sexual maturity
in female mobulids was evident from the exam-
ined ovary's developmental condition. In the im-
mature female the germinative epithelium is a
narrow, leaf-shaped band, tapering at both ends,
located opposite to the mesovarium (facing the
center of the cavity) along the ovary's longitudi-
nal axis. In mature females the germinative ep-
ithelium takes over most of the ovary's ventral
side, making room for the mature ova. To deter-
mine the size at maturity of female M. thurstoni,
the diameter of the largest ovarian egg was plot-
ted against DW (19 specimens. Fig. 4). An egg
growth plateau was not evident, because data on
Figure 3. — Relationship between clasper size and body size
in Mobula thurstoni.
50
I 40
g^30
_i -^
o
UJ "
Q. £ 20
OT ^
< s
i 10
• ••mtn«l fluid pr»««nl
O ••minal fluid aba«nl
y
o ca>
to
500 1000 1500
DISC WIDTH (mm)
15
E
>
o
(-
CO
LU
o
<
•
••
-
•• •
•
••
•
•
-
•
•
•
•
••
•
•
1
1
]
Figure 4. — Relationship between size of largest ovum and
body size in Mobula thurstoni.
1000 1500
DISC WIDTH (mm)
2000
51
FISHKKY BULI.KTIN VOL 86. NO 1
the larger sizes were insufficient, and therefore
the maximum egg diameter was not l^^nown. It
appears from the scanty available data that fe-
male M. thurstoni began to mature at a DW of
about 1,500 mm (the point at which the slope of
the curve becomes steeper).
Mobulo thurstoni, like all mobulids, is a vivi-
parous matrotroph ( Wourms 1981 ), the near-term
embryo being three orders of magnitude larger
than the mature egg. Uteri and nidamental
glands of 68 females were inspected. No adujt
female exammed from March through June iN
= 55) was pregnant. Embryos were found in July
and October. Two embryos from females caught
in July were near-term. Four embryos found in
October were in an early stage of development.
All females (A^ = 4) inspected in October were
pregnant, and all embryos were in the same de-
velopmental stage, suggesting coordinated breed-
ing activity. The largest female (DW 1,801 mm)
had a single embryo, which appeared to be in the
final stage of fetal development, fully pigmented,
the yolk sac completely absorbed, and the umbili-
cus a mere scar (Fig. 5A). The embryo's DW was
630 mm; its WT, 3.4 kg. It occupied the left uterus
with the rostrum pointed forward. Its pectoral
fins were folded dorsally, the right pectoral on
top. The cephalic fins were almost totally un-
rolled and extended ventrally towards the mid-
line of the body. The uterus, with the embryo
inside, occupied roughly one third of the female's
pleuroperitoneal cavity. Lack of space inside the
cavity and the distended skin on the abdomen
made it apparent that no other embryo had been
recently expelled or aborted. Uniparity appears to
be a common pattern within the genus Mobiila
(Hill 1862; Gill 1908; Coles 1913, 1916b; Barnard
1925; Setna and Sarangdhar 1950; Cadenat 1960;
Wallace 1967; Capape and Zaouali 1976;
Figueiredo 1977). Only Risso (1826) asserted that
Cephaloptera giorna (= M. mohular) may have
one or two young, but his statement was not doc-
umented. Since the smallest free-swimming spec-
imen noted had a DW of 864 mm, the average size
at birth for M. thurstoni is probably between 650
and 850 mm DW, and a WT of approximately 4.5
kg. The second largest female (DW 1,799 mm)
had also only one embryo in her left uterus. The
embryo was unpigmented, with disc 210 mm
wide, and weighed 173 g. Like the term-embryo,
its rostrum was pointing forward; unlike it, how-
ever, its pectoral fins were folded ventrally.
Mating, parturition, and early mobulid life his-
tory take place in the shallower portion of a popu-
lation's range, not an uncommon elasmobranch
feature. McLaughlin and O'Gower (1971) dis-
cussed inshore movements in the mating Port
Jackson shark Heterodontus portusjacksoni , as
did Springer (1960) for the sandbar shark Eu-
lamia milberti (= Carcharhinus plumbeus). One-
year-old gray reef sharks, Carcharhinus am-
bhrhynchos, were observed in French Polynesia
in shallower waters than adults by Nelson and
Johnson (1980). A similar result was reported for
the hammerhead shark, Sphyrna lewini , by
Clarke (1971) in Hawaii and by Klimley ( 1983) in
the Gulf of California. Bullis (1967) hypothesized
an upward movement to shallower depths for
newborn marbled cat sharks, Galeus arae. There
is likely an advantage for juveniles to remain in
relatively protected areas during the earlier
stages of their life, when they are most vulnera-
ble to predation (Springer 1967).
Examination of 139 stomachs indicated that
M. thurstoni was extremely specialized in its
feeding habits. Eighty one (58.3%) stomachs were
empty, or had only traces of food (<1 g wet
weight). The remaining 58 stomachs (41.7%) had
quantifiable contents. All recognizable prey
items were planktonic crustaceans (with the ex-
ception of a few fish eggs, one nematod, and a
small coleopteran, probably ingested accidentally
when it was floating). They were listed, ranked by
decreasing ^IRI, in Table 1. Mobula thurstoni
Table 1 —Prey species found m 57 stomachs of Mobula thurstoni, ranked by decreasing Index
of Relative Importance (IRI). N = percentage of prey species by number; M = percentage of
prey species by mass; F = percent frequency of occurrence of prey species.
Prey species
N
%N
M
°/oM
%F
IRI ,
%IRI
Nyctiphanes simplex
4,940.8
86 70
4.982.0
87 40
877
15,268 1
97,90
Mysidium sp.
6352
11 10
631.2
11.10
12.3
273 '
1.75
Copepoda
99.9
1 80
74.5
1.31
15.8
49 ;
Megalopa larvae
12.0
0.21
4.1
0.07
14.0
3.95 1
Hyperiid amphipods
6.8
0.12
5.8
0.10
5.3
1.18 '
[ 0,35
Fisfi eggs
1,4
0.03
1,8
0.03
5.3
0.302 \
Nematoscelis difi.
0.5
001
1.1
002
1.8
0.050 1
Stomatopod larvae
0.7
0.01
0.6
0.01
1.8
0.041 'J
52
NOTARBARTOLO-DI-SCIARA: NATURAL HISTORY OF MOBULA
B i
Figure 5. — A: term-embryo of Mobula thurstoni . The scale in the photograph is in centimeters. B; embryo of M.
tarapacana . The scale bar equals 5 cm.
53
KISHEKY BULLETIN: VOL 86, NO. 1
fed mostly on adult and juvenile euphausiids,
Nyctiphanes .s///?p/c.v. The niysid Mysidium sp.
(underscribed, Thomas E. Bowman^) was second
in order of importance. Mysids and euphausiids
were never found together in the same stomach.
The overall importance of the two food items dif-
fered by two orders of magnitude. All other prey
species found in the stomachs were rare and prob-
ably fortuitously ingested. These included one
zoea larva and the following copepod species:
Undinula vulgaris, Eucalanus subcrassus, E.
subtenuis, Ternora discaudata, Scolecithrix
danae, Nannocalanus minor, Euchaeta remana,
Euchaeta sp., and Labidocera diandra. Diet
varied with season (Fig. 6A), with mysids being
dominant from December through March, and eu-
phausiids during the warmer months. Diet varied
with predator size (Table 2): smaller individuals
fed both on euphausiids and mysids; the larger
rays fed only on euphausiids. This result probably
reflects the predominance of smaller rays during
winter, when fewer euphausiids are available,
rather than an ontogenetic change in food prefer-
ences.
Table 2. — Size differences in the diet of Mobula thurstoni.
%IRI
Prey species
DW < 1 ,300
(n = 21)
mm DW
> 1 ,300 mm
(n = 36)
Nyctiphanes simplex
Mysidium sp.
Other
79.04
18.13
2.83
99.91
0
0.09
Two remoras (Echeneididae) were occasionally
found on large M. thurstoni: Remora remora (3
specimens; range: 98-200 mm SL), and R.
albescens (3 specimens; 93-100 mm SL). Crus-
tacean parasites were encountered: Pupulina
minor (Copepoda: Caligidae), Ecthrogaleus den-
ticulatus (Copepoda: Pandaridae) sparsely on the
skin, and Ecthrogaleus disciarai (Benz and Deets
1987) in large patches on the dorsal surface, En-
tepherus laminipes (Copepoda: Cecropidae) from
the branchial filter plates, Eudactylina oliveri
(Copepoda: Eudactylinae) from the gill lamellae,
and Kroeyerina sp. (Copepoda: Kroyeriidae) from
the olfactory lamellae.
Mobula thurstoni was usually observed at the
surface in coastal waters of Bahia de la Ventana,
Cueva de Leon, and Bahia de los Muertos, often
within a few hundred meters of land and occa-
2Thoma.s E. Bowman, Smithsonian Institution, Washington,
D.C., pers. commun. 1984.
sionally as far as 6 km. When sighted offshore, it
was sometimes found over considerable depths
O500 m), although it appeared to be more abun-
dant in shallower, neritic waters. Mobula
thurstoni was always caught in the shallower
part of the nets, usually at a depth of < 100 m. The
greatest part of the catch, however, was surface-
dwelling rays. Beginning in mid-April, numerous
M. thurstoni were consistently seen in the early
morning hours cruising slowly at the surface.
They would frequently pause, conspicuous on
calmer days, with the tips of their pectoral fins
protruding out of the water. This behavior is well
known in mobulids (Norman and Eraser 1937); it
has been observed also in connection with mating
activities in M. olfersi ( = M. hypostoma ) by Coles
(1910). During such occasions, fishermen could
easily approach the rays and harpoon them, be-
fore startling them and causing them to dive. Re-
peated captures within the same aggregation re-
vealed that rays of various sizes and both sexes
could be found together. While at the surface,
M. thurstoni was usually solitary or in small,
nonpolarized groups (2-6), rather than in larger
aggregations or schools. The species was fre-
quently seen jumping out of the water in spectac-
ular, often reiterated somersaults; it was recog-
nized by the distinctive ventral markings.
It is not known to what extent mobulids make
use of the sea bottom. Beebe and Hollister (1935)
observed a group of 12 small devilfish (most likely
Mobula ) lying on the sandy substrate off Frigate
Islet, in the British West Indies. Bigelow and
Schroeder (1953) speculated that Manta spends
much of its time resting quietly on the seafloor.
During an experiment organized in conjunction
with Sea World of San Diego, aimed at establish-
ing whether M. thurstoni could survive in a con-
fined environment, five young specimens were
captured with gill nets and kept in a large pen (6
m in diameter) anchored in 2.5 m of water in
Ensenada de los Muertos. None of the rays sur-
vived 24 hours of captivity; the reasons for their
deaths were not clear, although the particularly
stressful capturing method appeared as a likely
cause. During that experiment the negatively
buoyant rays (sinking tail-first as soon as they
stopped swimming) spent a great deal of time
resting on the bottom, and were able to circulate
water through their gills while resting, by a syn-
chronized maneuvering of the oral valve and of
the gill covers (as judged from the flow made vis-
ible by the numerous particles suspended in the
water). A frequent method of turning around
54
NOTARBARTOLO-Dl-SCIARA: NATURAL HISTORY OF MOBULA
XII -II
IV
VI VII VIM
IX
XI
Figure 6. — A. seasonal variation of the relative importance ofNyctiphanes simplex and Mysidium sp.
in the diet of Mobula thurstoni. B. seasonal variation of the abundance of adult and juvenile A^.
simplex in the coastal areas of the southwestern Gulf of California (modified from Brinton and
Townsend 19801. C. captures of M. tarapacana . D. captures of M. munkiana . E. mean number of
daily captures of M. japanica; bars represent 2 SE on either side of the mean. F. mean number of
daily captures of M. thurstoni: error bars as in E. (*): although specimens were also captured
between December and February, data are not comparable with spring and summer captures because
catch effort was minor and inconsistent in the colder months.
55
FISHKKY BULI.KTIN: VOL 86, NO 1
(e.g., when swimming towards the wall of the
pen) was to dive vertically in a tight circle until
swimming in the opposite direction in an inverted
position, and then spinning around the body axis
to brmg the dorsal side up, rather than turning by
banking to the right or to the left.
An indication of the seasonal abundance of
M. thiirstoni in the surface waters of the study
area was obtained by the mean number of rays
caught daily from March to July 1983 (Fig. 6;
Table 3). Mean daily catch should be taken as a
rough indication of the relative abundance of M.
thurstom rather than as a precise index because
the fishing effort was difficult to quantify. Mean
effort, however, was roughly constant from March
through July because the mean monthly number
of working boats (about 20) and the number and
size of the nets set then was constant. Further-
more, the fishermen would harpoon a ray every
time they had the opportunity to do so. Peak of
abundance was in June, a result which appears to
be consistent with the fishermen's past experi-
ence, despite the 1983 abnormally high water
temperatures (Cane 1983). In July the number of
M. thurstom caught had dropped drastically, and
most of the catch consisted of M. japanica . No
information was obtainable for the August-Sep-
tember period. Eighteen specimens were cap-
tured during six fall and winter field trips (24
January-8 February 1981; 25 November 1981;
16-21 December 1981; 20-23 December 1982;
19-26 January 1984; 28 October-1 November
1984), but that figure was not comparable with
other data because part of the fishing cooperative
migrated south to Los Frailes during the cooler
months. It is common knowledge, however,
among the local fishermen, that M. thurstom in
the colder season is present, but in fewer numbers
than during the summer.
The study area constitutes a feeding, mating
and nursery ground for M. thurstoni. The eu-
phausiid Nyctiphanes simplex , the main diet item
and the only food of the adults when in the area,
is the most abundant and widespread euphausiid
in the Gulf of California, and has been observed in
dense swarms (Brinton and Townsend 1980). Al-
though it is found in the study area year-round,
its juvenile and adult stages are most abundant
between February and August, peaking in June
on the west side of the Gulf of California (Brinton
and Townsend 1980). The seasonal abundance of
M. thurstoni in the southern Gulf thus seems to be
closely related to the seasonal abundance of its
main prey. It is impossible to describe the general
Table 3— Mean number of daily captures of Mobula thurstom
a - total monthly number of captures; b monthily number of days
of monitonng; X = mean number of daily captures; SE = standard
error of tfie mean; so = standard deviation.
Month
a
6
b
X
2SE
SD
range
March
16
038
054
1.1
0-4
April
9
13
0.69
0.69
1.3
0-4
May
35
7
5.00
1.95
2.6
1-9
June
77
15
5.10
332
6.4
0-23
July
5
9
0.56
0.34
1.0
0-3
movement and life history pattern of M. thurstom
in the Gulf of California from the fragmentary
information available. The scanty data, however,
suggest the following: 1 ) adult male and nonpreg-
nant adult female M. thurstoni enter the area in
spring to feed and to mate, 2) pregnant females
segregate from the rest of the population in
spring (as is also suggested by the slight predom-
inance of males in spring and early summer),
3) gestation period is one year and females give
birth to one young every two or more years, 4) the
young are born in the study area or near it in
midsummer and remain there throughout their
early life, and 5) in late summer, when the num-
bers of adult and juvenile Nyctiphanes simplex
decline due to intense heating of the water (Brin-
ton and Townsend 1980), adult M. thurstom leave
the area, whereas the young switch their diet
from euphausiids to mysids. Further investiga-
tions are needed for additional corroboration of
these hypotheses.
Mobula japanica
(MuUer and Henle 1841)
Local name: cubana de lomo bianco
A total of 78 specimens, 34 males (DW range
1,316-2,386 mm) and 44 females (1,470-2,302
mm), were caught at three stations (Punta Arena
de la Ventana, Cueva de Leon, and Ensenada de
los Muertos) and adjacent waters, between 16 De-
cember 1981 and 13 July 1983. Overall and sea-
sonal size-frequency distributions for M. japanica
are shown in Figure 7. With the exception of
April, when females were larger (T -value
= 4.697, df = 3, P < 0.02), there are no significant
size differences between the sexes (T -value
- 0.535, df= 76, P > 0.5). Most of the rays in the
sample were large; only three were < 1,900 mm
DW.
Twenty-seven specimens (size range: 1,316-
2,285 mm DW; 18.6-115 kg) were weighed. The
WT/DW relationship is described by the equation:
56
NOTARBARTOLO-DI-SCIARA: NATURAL HISTORY OF MOBULA
WT = 4.29 X 10-10 (DW)'^^
r = 0.98
where WT is given in kg, DW in mm. The overall
male to female ratio was 0.89 (A'^ = 781. Females
dominated the June through December period;
males were predominant in April and May. Sex
ratios, however, never significantly differed from
1 (x" test P > 0.05). Sex segregation, behavioral
or geographical, was never observed in M. japan -
ica (both sexes were caught together in nets and
by harpoon); sampling bias (N = 3) may explain
why only males were caught in May. Geographi-
cal size segregation, by contrast, was an evident
feature of sexually mature specimens in the Gulf
of California (Fig. 7).
No pregnant females were found, although in
some specimens the left uterus had a flabby and
dilated appearance, suggesting recent delivery.
Tissues at the base of the claspers of most of the
larger males were swollen and reddened in June
and July when the tips of the claspers were flex-
ible and the rhipidion could be easily spread, and
in doing so a white, viscous fluid would ooze from
the hypophyle; all this suggested mating activity.
The clasper length-DW relationship for M Japan -
ica (Fig. 8) did not exhibit a clear pattern as in M.
30 -I
20
10 -
OVERALL
Xtot = 2I25 I±3674 (H-7Q)
^^^ = 21016167 12 (N-34)
Xjj -2143.3139 28 (N = 44)
frrn r-[7^ r-TH , i
.^^
I.
i
DECEMBER
Xtot = I86I4± 385 74 {H-5)
X^^ = I3I6(N = I)
Xj^:l997.8± 352 22(N = 4)
rmrTTi. i ., i .. i ., i .. i ,. i .1 ^. i ,. i
>
u
z
LU
O
UJ
o
en
m
<
30 n
20 -
10 -
APRIL
X,o,^ 2156 0 166 58 (N^5)
X^^ = 2l050i28.84(N-3)
H%
= 22325 + 53.00(N = 2)
mm,
MAY
Xtot = X^^= 2057 33 1122 56 (N^ 3)
.■ I ■■ I ■■ I
■■ I .Fxl— I,
30
20-
10 -
JUNE
Xtot=2I45 5131.32 (N^26)
X^<?^2I40 4 150 78(N=I0)
5(jo =2I48.814I04(N-I6)
D TOTAL
E3 ^^
0 %%
I ■■ I ■■ I
i.
n]za_
JULY
XT0T = 2l466i42.94 (N = 39)
7^^ = 2132.2 + 88.06 (N= 17)
X^j =2157713598 (N = 22)
I .n-i.
I ■■ I .r-iT^
ii
hi
4.
i
1450 1650 1850 2050 2250 1450 1650 1850 2050 2250
DISC WIDTH (mm)
Figure 7. — Size-frequency distribution.s of Mobula japanica (means ±2 SE).
57
FISUKKY BULLETIN: VOL H6. NO 1
Figure 8. — Relationship between clasper size and body size
in Mobula japan ica .
^150
O o
Z.<2
liJ^125
UJ "
CO "O
^«
O §100
# s«mlnal fluid preteni
O seminal fluid absent
A no data
.\.
1500 2000
DISC WIDTH (mm)
2500
thurstoni, since most of the specimens of M.
japanica were of larger size classes. It was appar-
ent, however, that at a DW between 2,050 and
2,150 mm there is a high degree of variability in
the relative size of claspers, which tended to sta-
bilize at higher values (DWs >2,150 mm), indi-
cating that male sexual maturity in M. Japanica
begins at a DW of 2,100 (±50) mm. Lack of infor-
mation on smaller specimens prevented a clear
understanding of the onset of female sexual ma-
turity (Fig. 9). Large eggs were found in speci-
mens as small as 2,070 mm DW, possibly indicat-
ing that female M. japanica began to mature at
that size.
Only 19 (24%) of 78 specimens had quantifiable
stomach contents (>1 g wet WT). The remaining
59 stomachs (76%) were empty or had only traces
of food. All M. japanica fed largely on the eu-
phausiid Nyctiphanes simplex (Table 4); no
mysids were found. Other species occurring in the
stomachs, including copepods, megalopa larvae,
stomatopod larvae, hyperiid amphipods, caridean
decapods (Crangon sp., Pasiphaea sp., and one
alpheid decapod), and one cumacean, had an over-
all relative diet importance of only 0.38%>. Mobula
japanica were therefore very similar in feeding
habits to large M. thurstoni. Information on the
Figure 9. — Relationship between size of largest ovum and
body size in Mobula japanica .
40
•
•
^^
E
^ 30
•
\.
5
•
<
• •
Q
2
•
3 20
•-
•
>
• •
O
•
•
1-
V)
•
0
•
LU
Sio
-
• •
<
_l
'
1
1
2000 2200
DISC WIDTH (mm)
2400
Table 4. — Prey species found in 19 stomachs of Mobula japanica, ranked by decreas-
ing Index of Relative Importance (symbols as in Table 2).
Prey species
N
%N
M
%/W
%F
IRI
%IRI
Nyctiphanes
simplex
Copepoda
Other
1,890.64
4.61
4.74
99.51
0.24
0.25
1,869.85
1.15
29.00
9841
0.06
1.53
100.00
31 58
36.84
19,792
9.47
65.58
99 62
0.05
0.33
58
NOTARBAKTOI.O-DI SCIARA: NATl'HAI. HISToHY OF MOBVLA
winter diet of M.Japanica was lacking, as four of
the five specimens collected then had empty stom-
aches; the fifth, a large female, contained a small
fragment of a partially digested fish carcass.
Since quantifiable stomach contents were found
only in large rays between April and July, no size
or seasonal differences in the diet of M.japanica
could be detected.
Mohiila Japanica was often found carrying
Remora remora , usually seen clinging to the out-
side of body, but found once inside a spiracle. Six
specimens of R. remora (range 109-217 mm SL)
were collected from M. japanica . Only one speci-
men of/?, albescens (97 mm SL) was found, in the
mouth cavity of a M.japanica. A pilot fish, Nau-
crates ductor (Carangidae) also associated with
M. japanica, swam alongside a harpooned ray
that was being towed inshore and remained for
some time at the water's edge, where the ray was
beached. Mobula japanica was parasitized by the
following crustaceans: Nerocila acuminata
(Isopoda: Cymothoidae), Pupulina brevicauda
and P . minor (Copepoda: Caligidae) on the skin;
Eudactylina oliveri (Copepoda: Eudactylinae) in
the gills; and Kroeyerina sp. (Copepoda: Kroyeri-
idae) among the olfactory lamellae. Unidentified
trypanorhynch cestodes were occasionally found
within the pleuroperitoneal cavity.
Habitat preference of M . japanica did not ap-
pear to differ from that of M. thurstoni . However,
the use of the habitat differed seasonally: in April
and May, when M. thurstoni was abundant at the
surface, M. japanica was never seen, and few
specimens were bottom gillnetted during those
months. Conversely, M.japanica, in June and
July, was seen in the late morning hours at the
surface in groups of several individuals swim-
ming parallel to the shore. Occasionally speci-
mens were seen in water <1 m deep. Mobula
japanica is not known to school, and I never ob-
served schooling.
Coles (1910) reported that M . olfersi (= M. hy-
postoma ) utters a "musical, bell-like bark" when
dying. A similar account was given by Risso
(1810) of Cephalopterus massena {= M . mobular).
This information led subsequent authors (Nor-
man and Fraser 1937; Bigelow and Schroeder
1953) to wonder whether mobulid rays are capa-
ble of producing sounds while in the water. Sound
production is a fairly widespread phenomenon
among bony fishes (Fish and Mowbray 1970;
Tavolga 1971); however, elasmobranchs lack the
traditional structures used by teleosts to generate
sound, i.e., the swim bladder and bony skeletal
parts (Marshall 1962), and recognizable sounds
have not been recorded from these animals
(Backus 1963). Sound production among elasmo-
branchs has been reported only for the Atlantic
cownose ray, Rhinoptera bonasus (Myrberg
1981); in that case clicks and scraping sounds
were presumably produced with the dental plates,
elicited when strongly prodding three rays which
were confined in a tank (Fish and Mowbray
1970). Mobula japanica, when beached alive,
often emitted a distinctive noise which could have
been the equivalent of Coles' "bark". This noise,
however, was apparently caused by the periodic,
spasmodic contractions of the mandibular, pha-
ryngeal, and hypobranchial musculature of the
asphyxiating ray, which forced air from the
mouth cavity out of the gill openings through the
meshlike branchial filter plates. Although under-
water sonic recordings have never been made, it
seems unlikely that under normal circumstances
any audible sound could be produced in this fash-
ion by submerged mobulids.
This area served as a spring and summer feed-
ing and mating ground for adult M. japanica,
rather than as a pupping or nursery ground, as
indicated by the lack of small-sized specimens.
Seasonal abundance of M .japanica in the surface
waters was indicated by the catch data (Table 5)
and is comparable to the seasonal abundance of
M . thurstoni (Fig. 6). No M . japanica were ob-
served in March; in April and May they occurred
occasionally. By mid-June large numbers ap-
peared in the nearshore surface waters near
Punta Arena de la Ventana, and were easily har-
pooned. Most of the July mobulid catch consisted
of M .japanica, when the numbers of M. thurstoni
had declined. Data are lacking for the August-
October period, therefore it was impossible to tell
whether the peak of abundance occurred in July
or later. Fishermen's reports were not clear, al-
though there was agreement on an overall decline
of mobulid abundance in late summer. Mobula
japanica fed exclusively on the euphausiid A^yc-
tiphanes simplex, and its numbers apparently de-
clined concomitant with the late summer decline
Table 5. — Mean number of daily captures of Mob-
ula japanica (symbols as in Table 1).
Month
a
b
X
2SE
SD
range
March
0
16
0
0
0
0-0
April
5
13
0.4
0.36
0.7
0-2
May
3
7
0.4
0.40
0.5
0-1
June
26
15
1.7
1.02
2.0
0-6
July
39
9
4.3
2.89
4.3
0-14
59
FISHERY BULLETIN: VOL 86, NO. 1
of abundance oC their prey (Brinton and
Townsend 1980). Some M . japanica , however, in-
cluding individuals both large and small, were
found in this region throughout December. Like
M. tharstoni , winter catch data are not compara-
ble because of differences in fishing effort, when
M . Japanica is apparently caught less frequently.
Mobula miinkiatia
NotJirbartolo-di-Sciara 1987
Local name: tortilla
Twenty-four specimens, 10 males (DW range
686-900 mm) and 14 females (719-1,097 mm),
were caught at four stations (Punta Arena de la
Ventana, Ensenada de los Muertos, Bahia de los
Frailes, and Isla El Pardito) and adjacent waters,
between December 1982 and October 1984. Size-
frequency distributions for M. munkiana (Fig. 10)
revealed that female mean size was greater than
male, although not significantly (T-value
= 1.724, df = 22,P > 0.1). Seasonal differences in
size-frequency distribution could not be examined
because M . munkiana were only collected during
the fall and winter. All 10 freshly captured speci-
mens were weighed (size range: 686-1,097 mm
DW; 4.1-11.8 kg). Their WT/DW relationship is
described by the following equation:
WT= 1.041 X 10-«(DW)2 34
r = 0.95.
WT is given in kg, DW in mm.
The largest specimen in the sample, a female,
was one of the largest "tortillas" ever seen. There
are no data on size at birth, as no embryos were
found. Lack of knowledge of the size of the young
of the year also prevented insight on size segrega-
>-
o
z
LlJ
o
UJ
q:
u.
LlI
O
CE
<
10
5 -
p?rn r-p^
Xtot = 886.5 ±33.4 mm (M = 24)
X^^ = 853.8 ± 39.5mm (N = IO)
X^^ =909.9 ±47. 1 mm (N:|4)
D TOTAL
m is
0 ??
mm
650 750 850 950 1050
DISC WIDTH (mm)
Figure 10. — Size-frequency distribution.s oi' Mubula munki-
ana (means ±2 SE).
tion. Term-embryos in M . rochebrunei , a closely
related, similar sized species from west Africa,
were 340-350 mm wide (Cadenat 1960). Since
Mobula at birth has a DW of about 1/3 of the
adult, size at birth would be about 350 mm DW.
This information argues in favor of size segrega-
tion in M. munkiana. Male to female ratio was
0.71 (N = 24), insignificantly different from 1 (x"
test P > 0.05). Both sexes were caught in the
same net sets, indicating that males and females
school together, and that there was no sex segre-
gation, either geographic or behavioral.
A dried, twisted male carcass, for which mor-
phometries could not be obtained, with a calcu-
lated DW of 895 mm, had long, well-developed
claspers, markedly protruding beyond the pelvic
fins. Based on other mobulid species, this condi-
tion indicates sexual maturity. A second speci-
men, with a disc 686 mm wide, had small and
pliable claspers, and the ratio between clasper
length and pelvic fin length, both measured from
tip to anterior margin of vent, was 0.88. At this
ratio, both M. thurstoni and M . japanica are im-
mature. Two M . munkiana , 871 and 872 mm DW,
differed greatly in the relative size of their
claspers: one possessed slightly longer claspers
than the pelvic fins (ratio = 1.10), and an incipi-
ent hardening of the cartilage was apparent; in
the other specimen the claspers were much
shorter than the pelvics (ratio = 0.84), and still
soft. This information suggested that male sexual
maturity in M. munkiana began at about 870 mm
DW.
The largest female specimen was sexually ma-
ture, as it had a large, flaccid left uterus, and the
enlarged left ovary consisted mainly of about 30
macroscopic ova (size range 2-15.2 mm).
Ten stomachs were examined: four were empty,
three contained unidentifiable whitish matter,
and three were full of planktonic crustaceans. A
list of prey species ranked by decreasing 9rIRI is
given in Table 6. Mysidium sp. appeared to be the
main staple of M. munkiana's diet. Mobula
munkiana is thus similar to the wintering young
of M. thurstoni. One of the stomachs contained
coarse coral fragments and small gastropod
shells, perhaps ingested by the ray while foraging
on mysids near a sandy substrate. The west
African species M . rochebrunei appeared to have
similar feeding habits, as Cadenat (1960) found
mysids and a few larger postlarval stomatopods
in the stomachs of several specimens.
I have no record of remoras associating with
M. munkiana. The only parasitic crustacean
60
NOTARBARTOLO-Dl-SCIARA: NATURAL HISTORY OF MOBULA
found was Pupulina cf. minor (Copepoda: Caligi-
dae) from the skin.
Among M. miinkiana's distinguishing features
are its neritic preferences combined with its so-
cial habits. This is the only mobulid species in the
Gulf of California that was consistently seen in
schools. It is not known whether this is a seasonal
behavior, or a permanent ethological feature of
the species. Schools appear as a conspicuous dark
patch, sometimes a few tens of meters in diame-
ter, as they slowly cruise along the coastline in
shallow water. The presence of the school is often
also highlighted by the frequent, simultaneously
leaping individuals, which betray its position
from a long distance. Similar behavioral traits
(schooling and leaping) have been reported for
two closely related species from the Atlantic,
M. hypostoma (Bancroft 1829; Coles 1910, 1916a)
and M. rochebrunei (Cadenat 1960). During
leaps, M. munkiana occasionally reached a
height of about two DWs. Two types of leaps were
observed: rising vertically head first and landing
flat with the belly on the sea surface with a loud
clap (breach), and spinning one to three times
around the main transverse body axis (somer-
sault).
A salient feature of M. munkiana's ecology in
this area is its winter occurrence when all other
mobulids are absent or at their lowest numbers.
Mobula munkiana apparently subsists then
chiefly on the mysid shrimp, Mysidium sp., which
is also the main food for young wintering M.
thurstoni . However, M . munkiana frequents the
area occasionally in summer: two specimens were
caught by surface gill net in Bahia de la Ventana
in July 1983. Even during the season in which it
is most abundant, M. munkiana is seen in
"pulses", as its occurrence at any particular loca-
tion is spotty. It may occur in large numbers at
one location for a few days, and then be absent for
1 or 2 weeks. This observation suggests the possi-
bility that M . munkiana lives in large concentra-
tions, perhaps composed by several schools, which
travel along the coast. A similar phenomenon was
observed off the Senegal coast by Cadenat (1960)
in M . rochebrunei, a species which is closely re-
lated to M. munkiana both morphologically and
ecologically.
It is conceivable that mobulids in the northern
half of the eastern tropical Pacific mate and give
birth in summer, based on the few term and near-
term embryos found in summer in M . thurstoni
and M . japanica, and from anatomical evidence
of mating activity in adult males M . thurstoni
and M .japanica . That such a hypothesis can also
be extended to M . munkiana is supported by lack
of reproductive activity in any of the specimens
collected during the fall or winter, and that new-
born and young-of-the-year are missing from the
sample. This evidence corroborates the hypothe-
sis that the local waters are a wintering ground
for M. munkiana, which then migrates into an
unknown area (perhaps the northern Gulf of Cali-
fornia) during the warmer season for mating and
pupping.
The possible causes of this ecological difference
between M. munkiana and the other mobulids
are many, and open to speculation. Mysid abun-
dance may be declining in summer in the south-
ern Gulf of California, and M . munkiana perhaps
migrates to areas where this crustacean or re-
lated species abound during the warmer season.
Alternatively, M. munkiana could be excluded
from this region in spring by competition with the
incoming, larger M. japanica and adult M.
thurstoni. Finally, M. munkiana may be moving
during the summer into an area which is more
suitable for its reproductive needs. Unfortu-
nately, this recently discovered species is very
little known, and it has been reported only from
the Gulf of California and Ecuador, although its
distribution probably extends to other coastal
areas of the tropical east Pacific (Notarbartolo-di-
Sciara 1987).
Table 6^ — Prey species found in three stomachs of Mobula munkiana ranked by decreasing
Index or Relative Importance (symbols as m Table 2). Identifiable copepod species included
Undinula vulgans, Rhincalanus nasutus, and Scolecithnx danae. The stomatopods found
were "erhithrus" larvae. One unidentified food item was a fragment of a larger crustacean,
probably an euphausnd.
Prey species
N
°/oN
M
°/oM
%F
IRI
%IRI
Mysidium sp. 293.40 97.80 287.40 95.80 100.00 19.360
Stomatopod larvae 4.65 1.55 9.06 3.02 66.67 304
Copepoda 1.59 0.53 1.44 0.48 66.67 67.3
Other 0.37 0.12 2.10 0.70 66.67 54.7
97.84
1.54
0.34
0.28
61
FISHERY BULLETIN: VOL. 86, NO. 1
Mob III (I tarapacatm (Philippi 1892)
Local name: vaquetilla
Mobula tarapacann is not a common species in
the study area. Seven specimens were collected,
one of which, a premature male embryo, was ex-
pelled by a large female while she was being
landed. Of the postnatal individuals, two were
male (DW range 2,476-2,494 mm), and four were
female (2,704-3,052 mm). All were caught in
Bahia de la Ventana between 9 June and 30 Octo-
ber 1983. All but two of the specimens were
weighed. The following equation describes the
WT/DW relationship for M. tarapacana (where
WT is given in kg, DW in mm):
WT = 2.378 X 10'*^(DW)2-92
r = 0.998.
Although all sampled postnatal M . tarapacana
were large, smaller individuals are known from
the area, as can be seen in photographs taken at
Punta Arena de la Ventana in summer 1981
(courtesy Felipe Galvan Magaha, CICIMAR, La
Paz, Mexico; also Greg B. Deets"^). This informa-
tion argues against geographical size segregation
of M. tarapacana . Data on the embryo provide no
indication of size at birth, since it was still far
from term. Pale pigmentation was apparent only
around the head and pelvic regions, and the ex-
ternal yolk sac was present (Fig. 5B). The em-
bryo, expelled tail first, was alive at birth. Judg-
ing by its size it had filled the left uterus
completely and must have been the sole develop-
ing embryo.
The small size of the sample does not permit
any clear inference on size at sexual maturity for
M . tarapacana . Some indication, however, can be
obtained by comparison with similar species. Of
the two postnatal males, the specimen with a disc
2,476 mm wide appeared to be immature: no sem-
inal fluid was found in the ducti deferentes, the
testes were small and apparently little developed,
and the ratio between clasper length and pelvic
fin length was 0.94. Conversely, the second post-
natal male, with a disc of approximately the same
width (2,494 mm), possessed claspers longer than
pelvics (ratio = 1.14), and the testes were well de-
veloped. Thus, sexual maturity in male M . tara-
pacana begins around a DW of 2,400-2,500 mm.
The specimen with a DW of 2,704 mm, one of two
nonpregnant females, had a bulky left ovary, con-
taining numerous large eggs; the largest, 32 mm
in diameter, weighed 12 g. Similar features ap-
peared to be associated with sexual maturity in
female M. thurstoni and M. Japanica. The left
ovary of another specimen, DW 2,831 mm, was
smaller, and the diameter of the largest ovum
was 18.6 mm, indicating that a DW of 2,700-
2,800 mm denotes a transitional stage for female
M. tarapacana , in which both mature and non-
mature individuals can occur.
Twelve echeneidids were recovered from
M . tarapacana . Three were Remora remora (size
range: 108-229 mm SL), and nine were R.
albescens (74-159 mm SL). The following crusta-
cean parasites were also found: one cymothoid
isopod (still in an unidentifiable aegathoid stage)
and Pupulina [lores (Copepoda: Caligidae) on the
skin, Entepherus laminipes (Copepoda: Cecropi-
dae) on the branchial filter plates, and Eu-
dactylina sp. (Copepoda: Eudactylinae) in the
gills.
Mobula tarapacana is strictly a summer and
fall visitor to this region (Fig. 6C). This species is
often found farther from the coast than
M . thurstoni and M. japanica, and may have
more pelagic habits. Four of the five stomachs
examined (all from specimens caught in summer)
were almost empty. Only traces of food were
found among the folds of the stomach epithelium.
Prey included four species of copepods (Acartia
sp. , Pontella sp., Tetnora discaudata , and Undin-
ula vulgaris ), hiperiid amphipods, one brachi-
uran (family Calappidae), one euphausiid, two
caridean decapods (one of which belonged to the
family Alpheidae), megalopa and stomatopod lar-
vae, and a fish egg. The fifth stomach, from a late
October capture, contained the remains of 27
fishes (probably carangids 15-30 cm long, and a
smaller anchovy-like species). Small tetraodon-
tids had been found before in the stomach of a
M . tarapacana caught in Bahia de la Ventana
(Felipe Galvan Magaha'*). On this basis it is im-
possible to determine whether M. tarapacana is a
specialized ichthyophagous ray, with the few
crustacean items accidentally ingested while
swimming, or a generalized feeder. The mesh size
of this species' branchial filter plates is indeed
greater than in other Mobula species
(Notarbartolo-di-Sciara 1987). However, filter-
•*Greg B. Deets, Long Beach State University, CA, pens, com-
mun. 1984.
^Felipe Galvan Magana, CICIMAR, La Paz, Mexico, pers.
commun. 1983.
62
NOTARBARTOLO-DI-SCIARA NATURAL HISTORY OK MOBrLA
feeding on planktonic Crustacea still appears
to be a feasible foraging technique for M. tara-
pacana, judging from the size of its branchial
sieve as it compares with the average-sized crus-
tacean prey.
SUMMARY AND CONCLUSIONS
Four species ofMobula were found in the south-
ern Gulf of California. The most abundant spe-
cies, M. thurstoni , was present year-round, but
only the smaller individuals were seen during the
winter. The bulk of the population, including the
adults, appeared in early spring. Numbers began
declining in July. Mobula japanica , the second
most abundant species, was comprised of only
large individuals; numbers progressively in-
creased from March throughout July. Large M.
japanica were rare in winter, but were occasion-
ally caught then. Mobula tarapacana is the rarest
mobulid in the area, yet its presence as a summer
and fall visitor is well known and predictable; it
is believed by the local fishermen to be more
abundant farther offshore. All three species share
a similar pattern of peak summer seasonal abun-
dance. The reverse is true for Mobula munkiana ,
it being most abundant in winter, and almost to-
tally absent during the rest of the year. It is not
known where any species goes when not seen in
the area. Seasonal migrations within the
epipelagic habitat to different areas of the
Panamic region are likely, but unverifiable be-
cause of the present lack of knowledge of the oc-
currence of identified Mobula species south of the
Gulf of California. Alternatively, devil rays may
spend part of the year in midwater, or near the
sea bottom, therefore disappearing from sight and
reach.
There is a striking similarity between the
array of mobulid species found in the Gulf of Cali-
fornia (and probably along the Pacific coast of
tropical America) and the mobulid fauna from the
tropical waters off west Africa. The family is rep-
resented in both areas by Manta birostris and by
four species oi Mobula : M . thurstoni; M Japanica
(reported from west Africa as M. rancureli by
Cadenat 1959); M . tarapacana (reported as M.
coilloti for African waters by Cadenat and Ran-
curel 1960 and Stehmann 1981); and a small gi-e-
garious form, represented in the Gulf of Califor-
nia by M . munkiana and off west Africa by the
closely related M . rochebrunei (Notarbartolo-di-
Sciara 1987). Tropical coastal areas off west
America and west Africa are known to be among
the most productive tropical waters in the world,
because of comparable large-scale atmospheric
and oceanographic circulation patterns (Sverdrup
et al. 1942). It is conceivable that the ecological
similarity between these two regions is reflected
in similar faunal associations, especially as far as
low levels of the tropic chain (e.g., plankton-
feeding vertebrates) are concerned.
The Gulf of California presents a unique envi-
ronment in the eastern Pacific Ocean, with ex-
treme annual water temperature ranges, wind-
induced mixing and upwellings, and subsequent
great productivity (Roden 1964; Brusca 1980).
Upwelling is caused along the peninsular coast by
the southerly winds prevailing during the
warmer months. This environment apparently
creates optimal conditions for the existence of the
euphausiid shrimp A^yc^/p/?a/?es simplex , which is
found in great abundance in the neritic habitat
between spring and midsummer, before the in-
tense August heat causes a decline in its numbers
(Brinton and Townsend 1980). The following data
are combined in Figure 6 to provide an overview
of the possible relationship between the seasonal-
ity of predator and prey in the study area: a) the
relative importance of Nyctiphanes simplex and
Mysidium sp. in the diet of M. thurstoni; b) the
relative abundance of A'^. simplex; the occurrence
of M. tarapacana (c) and M . munkiana (d) in the
catch; and the relative abundances of M .japanica
(e) and M . thurstoni (f) (no data on the biology of
Mysidium sp. are available).
Young M . thurstoni and all M . munkiana ex-
amined in winter appeared to subsist largely on
Mysidium sp., whereas adult M. thurstoni and
M .japanica caught during the warmer months
fed exclusively on N . simplex . An extreme degree
of feeding specialization was evident in all mobu-
lid species in which quantitative analyses of the
stomach contents was possible; most prey forms,
other than N . simplex and Mysidium sp., were so
rare that they were probably ingested acciden-
tally. Stenophagy was linked to feeding special-
ization in another myliobatiform species, the
mollusk-feeder Rhinoptera bonasus (Schwartz
1966; Smith and Merriner 1985). These results
suggest that devil rays are highly efficient in lo-
cating and selecting their preferred food. They
may be aided during this behavior by their prey's
habit of swarming. Competitive interaction is to
be expected between sympatric species-pairs
which are closely related both taxonomically and
ecologically. Food-resource partitioning is known
to occur in sympatric species-pairs of skates
63
FISHERY BULLETIN: VOL. 86, NO. 1
(McEachran et al. 1976). This condition, however,
is not necessarily true when the sought-after re-
sources are not in short supply (Zaret and Rand
1971). This may be the case of M. thurstoni and
M . japanica feeding together on A^. simplex when
the abundance of euphausiids is at its peak. Com-
petition should occur, however, in late summer,
when prey numbers decline. It would be interest-
ing to determine whether the slight morphologi-
cal and behavioral differences between poten-
tially competing species pairs (M. thurstoni/
M. munkiana in winter, M. thurstoni IM . japan-
ica in spring and summer) influence or reflect
partitioning of their habitat when food resources
become limiting, as was described for both fresh-
water (Werner and Hall 1977) and marine
teleosts (Hixon 1980; Larson 1980).
This overview of the ecology and natural his-
tory of mobulids in the Gulf of California is based
on field investigations made chiefly in 1983, a
year in which the El Nino perturbation was par-
ticularly severe (Cane 1983). Although in terms
of fishermen's experience the year 1983 was not
unduly different, as far as mobulid relative abun-
dance and seasonality are concerned, the abnor-
mally high water temperatures resulting from El
Nino may have affected the devil rays studied in
subtle ways; therefore this investigation should
be repeated in a normal year.
According to the fishermen, the abundance of
sharks (mostly carcharhinids and sphyrnids) on
which their activity is based is declining. This
decline will probably result in an increase of mob-
ulid fishing effort. It is of concern that 12% of the
specimens of M. thurstoni caught were immature
(DW <1,500 mm).
ACKNOWLEDGMENTS
I owe deep gratitude to the many persons who
assisted me in this investigation: Richard H.
Rosenblatt, Theodore H. Bullock, Paul K. Dayton,
William E. Evans, and Walter H. Munk, mem-
bers of my doctoral committee; Edward Brinton,
Robert Cowen, Abraham Fleminger, Nicholas
Holland, Margaret Knight, Spencer Luke,
William Newman, Mark Grygier, Jeff
Schweitzer, George Shor, George Snyder, and
Fred White of the Scripps Institution of Oceanog-
raphy; Thomas Bowman (United States National
Museum, Washington, D.C.); Daniel Brooks (Uni-
versity of British Columbia, Vancouver); Greg
Deets (Long Beach State University); Dennis
Bedford and Robert Lea (California Department
of Fish and Game); Fay Wolfson (Hubbs Marine
Research Institute); Alexis Fossi (Institut Na-
tional des Techniques de la Mer, Cherbourg,
France); Felipe Galvan Magaiia (Centro Inter-
diciplinario de Ciencias Marinas, La Paz, Mex-
ico); Lalo Cuevas, Marcelo Geraldo, Juan Lucero,
and their colleagues of the Cooperativa Pesquera
de Punta Arena de la Ventana; Steven Kramer
(National Marine Fisheries Service, Honolulu,
Hawaii); Carl A. Jantsch and Steven D. Kamol-
nich (Sea World, Inc., San Diego). The Hubbs
Marine Research Institute (San Diego) loaned a
sailing vessel, the "Fling", for the field study;
Rodney Black helped in the outfitting of the ves-
sel and in sailing it to the Gulf of California. This
investigation was supported in part by a grant
from the Foundation for Ocean Research (San
Diego).
LITERATURE CITED
Backus. R H
1963. Hearing in elasmobranchs. In P. W. Gilbert
(editor), Sharks and survival, p. 243-254. D. C. Heath &
Co., Lexington, MA.
Bancroft, E. N
1829. On the fish known in Jamaica as the
sea-devil. Zool. J. 4:444-457.
Barnard. K H
1925. A monograph of the marine fishes of South Africa.
Part 1. Ann. S. Afr. Mus. 21:1-415.
Barton, O
1948. Color notes on Pacific manta, including a new
form. Copeia 1948:146-147.
BEEBE, W , AND M A HOLLISTER
1935. The fishes of Union Island, Grenadines, British
West Indies, with the description of a new species of
stargazer. Zoologica, N.Y. 19:209-224.
Beebe. W., and J Tee- Van
1938. Eastern Pacific Expeditions of the New York Zo-
ological Society. XV. Seven new marine fishes from
Lower California. Zoologica, N.Y. 23(15):299-301.
1941. Eastern Pacific expeditions of the New York Zoolog-
ical Society. XXVIII. Fishes from the Tropical Eastern
Pacific. (From Cedros Island, Lower California, south to
the Galapagos Islands and northern Peru). Part 3. Rays,
mantas and chimaeras. Zoologica, N.Y. 26(26):245-
280.
Benz. G W . AND G B Debts
1987. Echthrogaleus disciarai sp. nov. (Siphonostoma-
toida: Pandaridae), a parasitic copepod of the devil ray
Mobula lucasana Beebe and Tee- Van, 1938 from the Sea
ofCortez. Can. J. Zool. 65:685-690.
BiGELOW. H B , AND W C SCHROEDER
1953. Sawfishes, guitarfishes, skates and rays. In J.
Tee- Van, C. M. Breder, A. E. Parr, W. C. Schroeder, and
L. P. Schultz (editors), Fishes of the western north At-
lantic, Memoir 1, part 2, p. 1-514. Sears Foundation for
Marine Research, New Haven.
BrINTON, E . AND A W TOWNSEND
1980. Euphausiids in the Gulf of California - the 1957
64
NOTAKliAIMOI.O 1)1 SflARA NATl'KAl. HISTORY OF .\f()IUl.A
cruises. Calif. Coop Oifaiiic Fi.sh Inve.st. Rt'p. 21:211-
•J.'i.'').
Bki-.sca. R C
1980. Common intertidal invcrti-brate.-i of thf (iulf of
California Univ. Arizona Pres.s, Tust<in. 513 p
BLl.l.lS. H H ,Jk
1967. Depth segregations and distribution of sex-
maturity groups in the marbled catshark. Galena
arav. In P. \V. Cilbert, R. F. Mathewson, and I). P. Rail
(editors). Sharks, skates, and rays. p. 141-148. Johns
Hopkins Press. Baltimore.
Cadknat. J
1959. Notes d'lchtyologie ouest-africaine. XXV. Descrip-
tion dune Mohiila de grande taille. a aiguillon caudal, de
Cote dlvoire: Mobula rancurvli. sp. nov. Bull. Inst. F"r.
Afr. Noire Ser. A 21 :i;326-1331.
19(i(). Notes d'ichtyoJogie ouest-africaine. XXIX. Les
Mobulidae de la cote occidentale d'Afrique. Bull. Inst.
Fr. Afr. Noire Ser. A 22:1053-1084.
CaHENAT. J . A.M) P Ra.MLRKI,
1960. Notes dichtyologie ouest-afncaine. XXVI. Descrip-
tion dune nouvelle espece de Mobulidae de la Cote
divoire: Mobula cotllnti - Bull. Inst. Fr. Afr. Noire Ser.
A 22:283-288.
Cank. M a
1983 Oceanographic events during El Nino. Science
222:1189-1195.
CAPAPE. C . AND J ZaoL'AL!
1976. Notes on the presence of the devilfish Mobula rnob-
ular (Bonnaterre. 1788) (Selachii. Rajiformes) in
Tunisian waters. Doriana 5i 223): 1-8.
CastroAci'IRRE. J L
1965. Feces sierra, rayas, mantas y especies afines de
Mexico. An. Inst. Nac. Invest. Biol.-Pesq. 1:171-256.
Chirkhicno. N.
1974. Clave para identificar los peces marinos del Peru.
Informe n. 44. Callao: Institute del Mar del Peru, 387 p.
Clarke, T A
1971. The ecology of the scalloped hammerhead shark,
Sphyrna Icuini. in Hawaii. Pac. Sci. 25i2):133-144.
Coi.ES R J
1910. Observations on the habits and distribution of cer-
tain fishes taken on the coast of North Carolina. Bull.
Am. Mus. Nat. Hist. 28:337-348.
1913. Notes on the embryos of several species of rays, with
remarks on the northward summer migration of certain
tropical forms observed on the coast of North Caro-
lina. Bull. Am. Mus. Nat. Hist. 32:29-35.
1915. Notes on the sharks and rays of the Cape Lookout,
N.C. Proc. Biol. Soc. Wash. 28:89-94.
1916a. Natural history notes on the devil fish, Manta
btrostna (Walbaumi and Molbula olfersi iMuller). Bull.
Am. Mus. Nat. Hist. 35:649-657.
1916b. Notes on Radcliffe's sharks and rays of Beau-
fort. Copeia 32:45-47.
Daniel J F
1934. The elasmobranch fishes. Univ. California Press,
Berkeley. 332 p.
FlC.l'EIREDO, L
1977 Manual de peixes marinhos do sudeste do Brasil.
Vol. 1. Cacoes, rayas e quimeras. Universidade de Sao
Paulo. Museu de Zoologia. Sao Paulo, 105 p.
Fish. M P . a.nu W H Mowbray
1970. Sounds of western north Atlantic fishes: a reference
of biological underwater sounds. Johns Hopkins Press,
Baltimore, 207 p.
Fowler. H W
1944. Results of the fifth George Vanderbilt South Pacific
Expedition il941i. The fishes. Monogr. Acad. Nat. Sci.
Phila. 6:57-583.
GlI.L.T
1908. The story of the devil-fish. Smithson. Misc. Col-
lect. 52(1816):155-180.
GlLMORE.R G J W DoDRILL. ANDP A LiNLEY
1983. Reproduction and embryonic development of the
sand tiger shark, Odontaspis taurus iRafinesque). Fish.
Bull., U.S. 81:201-225.
GoHAR. H a F , AND A R Bayou.mi
1959. On the anatomy of Manta eherenbergi with notes on
Mobula kuhlii. Publ. Mar. Biol. Stn. AI Ghardaqa
10:191-238.
Hill, R
1862. The devil-fish in Jamaica. Intell. Obs. 2:167-176.
Hixon. M a
1980. Competitive interactions between California reef
fishes of the genus Embiotoca . Ecology 61:918-931.
Hyslop. E J
1980. Stomach contents analysis - a review of methods
and their application. J. Fish. Biol. 17:411-429.
Klimley. a P
1983. Social organization of schools of the scalloped ham-
merhead shark, Sphyrna lewini (Griffith and Smith), in
the Gulf of California. Ph.D Thesis, University of Cali-
fornia, San Diego, Scripps Institution of Oceanography.
Larson. R J
1980. Competition, habitat selection, and the bathymetric
segregation of two rockfishes iSebastes) species. Ecol.
Monogr. 50i2):221-239.
Lloyu. R E
1908. On two species of eagle-rays with notes on the skull
of the genus Ceratoptera . Rec. Indian Mus. (Calcutta)
2(2):175-180.
McEachran. J D., D F BoESCH, and J A MUSICK
1976. Food division within two sympatric species-pairs of
skates (Pisces: Rajidaei. Mar. Biol. (Berl.) 35:301-317.
McLaughlin. R H . and A K OGower.
1971. Life history and underwater studies of a heterodont
shark. Ecol. Monogr. 41(4):271-289.
Marshall. N B
1962. The biology of sound-producing fishes. Symp.
Zool. Soc. Lond. 7:45-60.
Miller, J K
1966. Biomass determination of selected zooplankters
found in the California Cooperative Oceanic Fisheries
Investigations. Scripps Inst. Ocean. Ref Ser. 66-15. 16
P-
MULLER. J , AND J HENLE.
1841. Systematische Beschreibung der Plagiostomen.
Verlag von Veit & Co., Berlin, 204 p.
Myrberg, A A . JR
1981. Sound communication and interception in
fishes. In W. N. Tavolga, A. N. Popper, and R. N. Fay
(editors). Hearing and sound communication in fishes, p.
397-425. Springer- Verlag, New York. Heidelberg, and
Berlin.
Nelson, D R., and R. H Johnson.
1980. Behavior of the reef sharks of Rangiroa, French
Polynesia. Natl. Geogr. Res. Rep. 12:479-499.
NICH0L.S. J T . AND R C Ml rphy
1944. A collection of fishes from the Panama Bight,
Pacific Ocean. Bull. Am. Mus. Nat. Hist. 83(4):217-
260.
65
FISHKKY BULLKTIN: VOL. 86, NO. 1
NOKMAN, J K , AND F C FKASKR
1937. Giant fishes, whales, and dolphins W. W Norton,
N.Y., .361 p.
NOTARBARTOLO-DISC'IAKA, G
1985. A revisionary study of the genus Mohula
fiafinesque 1810, with the description of a new species,
and natural history notes on east Pacific mobu-
lids. Ph.D Thesis, Univ. California, San Diego.
University Microfilms International No. 8510912, Ann
Arbor, Michigan, 346 p.
1987. A revisionary study of the genus Mohula
Rafinesque, 1810 (Chondnchthyes: Mobulidaei, with the
description of a new species. Zool. J. Linn. Soc. 91:1-91.
In press. Myliobatiform rays fished in the southern Gulf
of California (Baja California Sur, Mexico) (Chon-
drichthyes: Myliobatiformesl. Proc. Fifth Symposium of
Marine Biology, La Paz, B.C.S., Mexico. 24-26 Oct. 1984.
Pequeno. G
1983. La condrictiofauna de las regiones de Chile y
California-Oregon: comparacion preliminar. In P. M.
Arana (editor). Proceedings of the International Confer-
ence on Marine Resources, p. 253-267. Pacific, Vina del
Mar, Chile.
Phii.ippi. R A
1892. Algunos peces de Chile. An. Mus. Nac. Chile, Sec.
1, Zool. 3:1-17.
PiNKAS. L.MS Oliphant. and I L K Iverson
1971. Food habits of albacore, bluefin tuna, and bonito in
Californian waters. Calif Dep. Fish Game Fish Bull.
1.52, p. 1-82.
RiSSO. A
1810. Ichtyologie de Nice, ou histoire naturelle des pois-
sons du departement des Alpes Maritimes. Paris F.
Schoell, 388 p.
1826. Histoire naturelle des principales productions de
1 Europe meridionale et particulierement de celles des
environs de Nice et des Alpes Maritimes. Vol. 3. Paris:
F. G. Levrault, 480 p.
RODEN, G I
1964. Oceanographic aspects of Gulf of California. Am.
Assoc. Pet. Geol., Symp. Mar. Geol. Gulf Calif 3:30-58.
Schwartz. F J
1966. Embryology and feeding behavior of the Atlantic
cownose ray, Rhmoptera honasus. Assoc. Islands Mar.
Lab. Caribb. 7th Meeting, Barbados, 24-26 August 1966,
p. 15.
SETNA, S. B , AND P, N, Sarangdhar
1950. Breeding habits of Bombay elasmobranchs. Rec.
Indian Mus. (Calcutta I, 47:107-124.
Smith. J W . and J. V. Merriner,
1985. Food habits and feeding behavior of the cownose
ray, Rhmoptera honasus, in lower Chesapeake
Bay. Estuaries 8:305-310.
1986. Observations on the reproductive biology of the
cownose ray, Rhmoptera honasus, in Chesapeake
Bay. Fish. Bull., U.S. 84:871-877.
Springer, S
1960. Natural history of the sandbar shark Eulamia mil-
berti. U.S. Fish Wildl. Serv., Fish. Bull. 61:1-38.
1967. Social organization of shark populations. In P. W.
Gilbert, R. F. Mathewson, and D. P. Randall (editors).
Sharks, skates and rays, p. 149-174. Johns Hopkins
Press, Baltimore.
Stehmann, M
1981. Batoid fishes. /« W. Fischer, G. Bianchi, and W. B.
Scott (editors), FAO species identification sheets for fish-
ery purposes. Eastern central Atlantic, fishing areas 34,
37 (in part), Vol. 5. Rome: FAO.
Sverdrup, H. U , M W Johnson, and R H Fleming
1942. The oceans. Prentice-Hall, N.Y., 1087 p.
Tavolga. W. N.
1971. Sound production and detection. In W. S. Hoar
and D. J. Randall (editors). Fish physiology. Vol. 5, p.
135-205. Academic Press, N.Y.
Wallace, J H
1967. The batoid fishes of the east coast of southern
Africa. Part 2. Manta, eagle, duckbill, cownose, butterfiy
and stingrays. S. Afr. Assoc. Mar. Biol. Res., Ocean.
Res. Inst. Invest. Rep. 16:1-55.
Werner, E E , and D J Hall
1977. Competition and habitat shift in two sunfishes
(Centrarchidae). Ecology 58:869-876.
Wourms. J P
1977. Reproduction and development of chondrichthyan
fishes. Am. Zool. 17:379-410.
1981. Viviparity: the maternal-fetal relationship in
fishes. Am. Zool. 21:473-515.
Zaret, T M . AND A S Rand
1971. Competition in tropical stream fishes: support for
the competitive exclusion principle. Ecology 52:336—
342.
66
NOTES ON DECAPOD AND EUPHAUSIID CRUSTACEANS,
CONTINENTAL MARGIN, WESTERN ATLANTIC, GEORGES BANK
TO \XTSTERN FLORIDA, USA
Austin B Williams'
ABSTRACT
Twenty-six species of decapod crustaceans in 16 families and 1 species of euphausiid are reported from
the outer continental shelf, submarine canyons, and nearby slope of the eastern and southeastern
United States. Station data are given for all collections made with the aid of submersible and surface
vessels. Bathymetric and geographic distributions are summarized for six species (Lithodes maja,
Munida forceps. M. longipes, Chacellus filiformis, Dissodactylus juvenilis . Euchirograpsus ameri-
canus ) whose ranges are extended. Comparative descriptive notes are given for other forms that have
uncertain identities and need further study (Alpheus cf. amblyonyx, sp. near Ligur. Mumdopsis cf.
transtridens ).
Records of decapod crustaceans from the outer
continental shelf, submarine canyons, and
nearby slope off the eastern United States have
accumulated in my files to the point that it seems
appropriate to publish them with notes on the
samples taken in that habitat. Specimens were
collected with the aid of deep and shallow water
submersibles and surface vessels operated cooper-
atively by the Lamont-Doherty Geological Labo-
ratory, Columbia University, Palisades, NY. with
sponsors listed hereinafter, and between Rutgers
University, Center for Coastal and Environmen-
tal Studies, and Department of Horticulture and
Forestry, New Brunswick. NJ; the Northeast
Fisheries Center. National Marine Fisheries Ser-
vice (NMFS). NOAA, Woods Hole. MA; and the
Southeast Fisheries Center Laboratories, NMFS.
NOAA, at Panama City, FL, and Pascagoula, MS
(Able et al. 1982, 1987; Grimes et al. 1980a. 1986;
and Lamont-Doherty records). Vessels (see Table
1 1 and their sponsoring institutions were DSRV
Alvm, Woods Hole Oceanographic Institution,
Woods Hole, MA; RV JSL I and JSL II, Harbor
Branch Oceanographic Institution, Inc., Fort
Pierce, FL; RV Cape Henlopen , University of Del-
aware, Lewes, DE; RV Eastward , Duke Univer-
sity, Beaufort. NC; RV Endeavor, University of
Rhode Island, Kingston, RI; RV Gyre, Texas
A&M University, Galveston, TX. The specimens
have been deposited in the crustacean collection
'Systematics Laboratory, National Marine Fisheries Service,
NOAA. U.S. National Museum of Natural History, Washing-
ton. DC 20560
Manuscript accepted August 1987.
FISHERY BULLETIN: VOL. 86. NO. 1, 1988.
of the United States National Museum of Natural
History (USNM).
Bathymetric and geographic distributions of
many decapod crustacean species from the North
American continental shelf in the western At-
lantic were reviewed by Squires (1965), Williams
and Wigley (1977), Wenner (1982), and Williams
(1984). Wenner and Boesch (1979) and Wenner
and Windsor (1979) included a deeper dwelling
component in their treatments of epibenthic deca-
pods collected from the continental shelf and
slope. Collections by workers from the contribut-
ing institutions listed above include 26 species of
decapod crustaceans in 16 families, some uniden-
tified fragments, and 1 species of euphausiid
listed in Table 1. Named localities are listed from
north to south (see Figures 1 and 2); successive
visits are arranged chronologically, and species
present in each collection are inventoried alpha-
betically.
Species whose reported bathymetric or geo-
graphic ranges are extended by presence in these
collections are discussed below. Some of the forms
have uncertain identities that may be clarified
after more thorough study of samples from the
outer shelf-upper slope environment.
SUPERFAMILY ALPHEOIDEA
Family Alpheidae
Alpheus cf. amblyonyx Chace 1972. Three lots of
specimens key out to A. amblyonyx (see Chace
1972), but they differ in several respects from it.
67
FISHERY BULLETIN: VOL. 85, NO 1
Table 1 . — Records of decapod crustaceans from submarine canyons, outer
listed from norlti to south, successive visits are arranged chronologically, and
in footnote. ' = Bathymetric or geographic range extension.
Locality
Depth Cruise/
(m) station
Date
W
■w
39.97'W
Lydonia Canyon
40^21 'N, 67 41
40°22'N, 67' 41
40°21.69N, 67
700-790 Gyre
462 Alvin 1269
906 Alvin 1270
8 June 1982
18 Sept. 1982
19 Sept. 1982
Oceanographer Canyon
40"23.7'N, 68'07.8'W
415-680
Eastv/ard
35982
15 IVIay 1979
40^6 3'N, 68'07.2'W
660-1,424
Eastward
35991
19 IVIay 1979
Slope Area III between Hydrographer and Veatch Canyons
39°50'N, 69°25'W
500-2,000
Endeavor
21 Oct. 1981
Slope Area II betv^^een Toms and Meys Canyons
39°13.05'N, 72°30.93'W
174
JSL 1 1081
1 Aug. 1981
39°12.64'N, 72"30.87'W
200
JSL 1 1081
1 Aug. 1981
39 11.06' N, 72°33.66'W
219
JSL 1 1082
1 Aug. 1981
Hendnckson Canyon
39°03.52'N, 72°28.45W
1,420-1,425
Alvin 1118
19 July 1981
Baltimore Canyon
38 09.6'N, 73 49.2'W
155-160
Eastward
35940
2 May 1979
38°07.9'N, 73°48.8'W
185-190
Eastward
35938
2 IVIay 1979
38°08.5'N, 73°49.9'W
280-570
Eastward
35944
2 IVIay 1979
38°09.6'N, 73°49.2'W
155-160
Eastward
35940
2 IVIay 1979
38°08.7N, 73°49.3'W
148-174
Eastward
35939
2 IVIay 1979
38°11.7'N, 73°52.8'W
1 50-245
Eastward
35942
2 IVlay 1979
38°03.6'N, 73°46.2'W
910-925
Eastward
35946
3 IVIay 1979
38 07.6'N, 73°46.8'W
152-163
Eastward
35947
3 May 1979
38''06'N, 73=50'W
200-800
Eastward
35979
7 May 1979
38°04.93'N, 73°47.79'W
1,024
Alvin 1109
11 July 1979
38°08'N, 73°50.7'W
600-700
B10-1
28 May 1981
34°04'N, 73°46'W
1,000-1,200
Cape Henlopen
CM02
13 June 1981
38°09'N, 73°51'W
171-381
JSL 1 1083
2 Aug. 1981
38°09.99'N, 73'51.86'W
381
JSL 1 1083
2 Aug. 1981
38°11'N, 73°51'W
165
JSL 1 1088
4 Aug. 1981
38°11'N, 73°51'W
165-244
JSL 1 1088
4 Aug. 1981
38°10.1'N, 73°52.2'W
225
JSL 1 1084
4 Aug. 1981
68
WILLIAMS: DECAPOD AND EUPHAUSIID CRUSTACEANS
continental shelf and upper slope off the eastern United States. Named localities are
species inventoried in each collection are listed alphabetically. Family abbreviations
Species
N. Sex
Notes, familyi
■ Lamont-
Doherty --
Munida valida Smith
Pagurus politus (Smith)
'Munidopsis cf. transthdens
Pequegnat and Pequegnat
Pandalus propinquus
G 0. Sars
Id
19
Dredge. Ga
Pag
Ga
Pan
Panadalus propinquus
Id, 39
Day dredge.
Metacrangon lacqueti agassizii
Id
Dredge, Cr
(Smith)
Nematocarcinus ensifer (Smith) 1 spec.
Camera. N
Euprognatha rastellifera
1 9 ovig.
M
Stimpson
'Euchirograpsus americanus
19
Gr
Eumunida picta Smith
19
Ch
Munida ins ins
1d
Ga
A. Milne Edwards
(parasitized)
'Munidopsis cf. transtndens
Id
Ga
Cancer borealis Stimpson
1 juv.
Can
'Euchirograpsus americanus
1 d , 3 9
Gr
A. Milne Edwards
Euprognatha rastellifera
Id
Day dredge. M
Cancer borealis
19
Day dredge, Ca
Euprognatha rastellifera
Id
Day dredge, M
'Munida longipes
1d, 29 ovig.
Day dredge, Ga
A. Milne Edwards
Pagurus politus (Smith)
19
Day dredge, Pa
'Chacellus filiformis Guinot
Id
Day dredge. Go
Euprognatha rastellifera
Id
Day dredge, M
Munida ins ins
3d, 39
Day dredge, Ga
Plesionika edwardsii (Brandt)
1 9 ovig.
Day dredge. Pan
Cancer borealis
2 juv.
Day dredge, Ca
Cancer irroratus Say
Id
Day dredge, Ca
Euprognatha rastellifera
Id
Day dredge, M
Cancer borealis
Id
Day dredge, Ca
Sergestes arcticus
19
Day dredge, S
Collodes robustus Smith
1 9 ovig.
Day dredge, M
Eumunida picta
4 juv.. Id, 59
Camera sled, Ch
Lebbeus polaris (Sabine)
19
Camera sled, H
'Munidopsis cf. transtridens
1d
Airplane wreck, Ga
Sergestes arcticus Kroyer,
99
Bongo net, S
euphausiid and copepods
Sergestes arcticus
19
Bongo net, S
Bathynectes longispina
Id
Po
Stimpson
'Lithodes maja (Linnaeus)
19
L
Cancer irroratus
Id
Ca
Munida ins ins
Id
Ga
Munida ins ins
Id
Ga
Pagurus politus
Id
Pag
69
FISHERY BULLETIN: VOL. 86, NO. 1
Table 1. — Continued.
Depth
Cruise/
Locality
(m)
station
Date
38' 09.51 'N, 73°51.25'W
338
JSL 1 1089
5 Aug. 1981
38'11'N, 73 SrW
402
JSL 1 1090
5 Aug. 1981
38 02N, 73 45'W
200-1,400
Gyre
26 May 1982
Norfolk Canyon
37°03'N, 74°37'W
215
Rutgers
JSL 1 1093
7 Aug. 1981
Lydonia Canyon
40'27.25'N, 67°42.30'W
160-178
JSL 1 1070
23 July 1981
Veatch Canyon
40 03'20"N, 69 45'23"W
124-126
JSL 1 1074
25 July 1981
40TO00"N, 69°45'25"W
123-128
JSL 1 1075
26 July 1981
39 59'51"N, 69°35'00"W
213-244
JSL 1 1076
26 July 1981
40"00.91'N,
70°50.79'W
243-304
JSL II 901
29 July
1984
40 0 1.29 'N,
7050.63'W
4003.24N,
71 05.01 'W
193-209
JSL II 902
29 July
1984
40"02.96N,
70 20.84'W
181-195
JSL II 909
2 Aug.
1984
40'12.73'N,
70'19.77'W
103-104
JSL II 910
2 Aug.
1984
40 12.51 'N,
70'20.22'W
103-104
JSL II 910
2 Aug.
1984
40"02.79'N,
70 11.95'W
183-337
JSL II 911
3 Aug.
1984
40°03.23'N,
70''12.0rW
213-327
JSL II 912
3 Aug.
1984
Off eastern Florida
28=43'N, 80 02'W 137 JSL I 1565 1 Oct. 1984
28°42.5'N, 80°02.8'W 98-114 JSL I 1566 2 Oct. 1984
Off western Florida
27=51 .2N, 84°53.7'W 250-260 JSL I 1673 24 Sept. 1985
27°56.2'N, 84°43.9'W 167-175 JSL I 1676 26 Sept. 1985
'A, Alpheidae; Cal, Calappidae, Can. Cancndae; Ch, Cliirostylidae; Cr, Crangonidae; Ga,
M, Majidae. N, Nematocarcinidae: Pag, Paguridae; Pan, Pandalidae; Pi, Pinnotheridae; Po,
70
WILLIAMS: DECAPOD AND EUPHAUSIID CRUSTACEANS
Species
N. Sex
Notes, family'
Rochinia crassa
A Milne Edwards
Pandalus propmquus
Sergestes arcticus
Id
29
36
M
Pan
Camera, S
Bathynectes longispina
1 juv.
Rutgers
Po
'Munida forceps
A. Milne Edwards
IcJ
Bot t 1 1 2°C, Ga
Munida ins ins
Pagurus politus
Acanthocarpus alexandn
Stimpson
'Chacellus filiformis
'Munida forceps
Munida ins ins
'Alphieus cf. amblyonyx Chace
'Alpfieus cf. amblyonyx
Munida ins ins
'Munida forceps
'Goneplacid'' crab
'Chacellus filiformis
'Munida forceps
'Munida forceps
Meganyctipfianes norvegica
(M. Sars)
'Munida forceps
'Chtacellus filiformis'^
Meganyctiptianes novegica
Meganyctiphanes novegica
Meganyctiptianes novegica
Alpfieus cf. amblyonyx
Munida forceps
Crab (xanthid or goneplacid
frags.)
'Alpfieus cf. amblyonx
Munida forceps
Id, 42 ovig.
Bot. t.
1 1 .4'C, Ga
1d
Bot. t.
11.4 C, Pag
29
Bot. t.
11.5 C, Ca
36 (1 JUV.)
Bot. t.
11.2 C, Go
49 (1 juv.,
3 ovig.)
Bot. t.
12.2'C, Ga
1 (5 , 1 2 ovig.
Bot. t.
12.2'C, Ga
Id
1d
H
1 JUV,
H
Id
Ga
1d
Ga
Go
Id, 19
Go
2d
Ga
1d
Ga
1
E
1 juv., 1 I ovig
Ga
Frag, of chela
Go
45
E
2
E
8
E
2d, 19, 1 JUV.
Tilefisfi burrow, A
29
Tilefisl
T burrow, Ga
1
Tilefish burrow, Ga
19
Tilefish burrow, Ga
12.1
C, A
1
Tilefishi burrow, Ga
*Near Ligur Sarato
'Munida forceps
' Dissodactylus juvenilis
Bouvier
2d. 1 9 Tilefish burrow. Hi
Id, 25 ovig. Tilefish burrow, Ga
1 juv.
1 V ovig. With Clypeaster
ravenellii A.
Agassiz, 13.8'C, Pi
Galatheidae. Go. Goneplacidae; Gr, Grapsidae: E. Euphausiidae; H, Hippolytidae; L. Lithodidae;
Ponunidae; S, Sergestidae.
71
FISHERY BULLETIN: VOL. 86, NO. 1
69
68 45
67 44 66 65 43 64 42
\
V J--*
.\ :.;;/ ^^ -^-^ . X ..• M \ /
70
\
^ i;i
,-'-\_-J ,'' NORTHEAST CHANNEL
41
\
64
45
/
r ■■■'
//'"' ,,^•''■''^^
71
\
./■
/^■'
;"' '.•■■ vVCORSAIR
65
Y
J ;'••■ * '^ /
40
44
/
/
\\)k
' 1 •' <n ■ '
72
\
1 ^
■: S ■■ :j
i^v^S. •■•.... . ,-J^LYDONIA
/A .•■^V'Tgilbert
-^/ ; ,'TOCEANOGRAPHER /
66
39
43
73
\
-V
VL ^ ..'JrWELKER
,^° Y
:• ~3jrHYDR0GRAPHER \
67
Vv
... •■ ,-_/VEATCH
42
/
A(f
,'_/^TLANTIS y
38
74
\
1^
f
.■/ /block ^
68
41
/
&
f"
V-f^ HUDSON y
37
75
\
iC./ —
\
. //hendrickson
69
/v
\
M ATLANTIC
/^
;
p
.■ '
40
76
/
>
"H
';^ WILMINGTON OCEAN <^
f:i BALTIMORE
36
70
39
%
^
* !)
.it WASHINGTON
35
77
V
g
^y\ NORFOLK
71
38
/
■\ \
78
\
^
^^X VI
34
72
/
\
x'S
/ -^Xj p. CAPE HATTERAS \ y \
37
78
77 36
76 35 75 74 34 73
Figure l. — East coast of United States showing continental shelf, slope and submarine canyon
areas from which species listed in Table 1 were collected. Base map adapted from Uchupi ( 1965) and
Veatch and Smith (1939). Contours m m: dotted = 100, dashed = 200, solid = 1,000.
72
WILLIAMS: DECAPOD AND EUPHAUSIID CRUSTACEANS
83 31
30 81
80 29
79
28 78
V
\ ^ A
/
\
■\K
78
27
79
y 26
25
80
85
27 84
83 26
82
25 81
Figure 2. — Florida peninsula including continental platform showing slope localities from which species listed
in Table 1 were collected. Base map adapted from Uchupi (1965). Contours in m; dotted = 100, dashed = 200,
solid = 1.000.
Some of the more obvious differences for the
Veatch Canyon material are: Rostrum relatively
shorter in relation to basal antennal article. Dor-
sal spines of telson more distally positioned, ante-
rior pair at about midlength of telson but subject
to some variation. Antennal scale with distal
spine exceeding antennular peduncle. Major
chela of mature male missing, but juvenile with
fingers bent mesad; dactyl moderately arched in
profile and greatly overreaching fixed finger,
somewhat twisted, compressed proximally and
dorsally producing thin dorsal margin, external
surface somewhat concave, occlusive surface
lacking plunger but broadened and strongly calci-
fied distally, fitted to obliquely flattened occlu-
sive surface of fixed finger lacking socket but
provided with small stout tooth on mesial sur-
face; palm with obsolescent dorsal and ventral
notches.
In the Florida material, the differences are:
Chela relatively stout, fingers stout and thick,
dactyl opening and closing in oblique plane, tip
rounded, bearing short plunger fitting into shal-
low socket on occlusive surface of fixed finger,
latter with 2 short spines on mesial surface; palm
with shallow notch on dorsal margin and slight
offset on ventral margin, outer surface smooth
but base of dactyl flanked by distodorsal groove
and longer mesial groove. Second pleopod of male
with appendix masculina exceeding appendix in-
terna. Uropodal exopod with lateral margin end-
ing in single sharp tooth and rather long, uncol-
ored movable spine; endopod lacking distal spines
but bearing subterminal tuft of setae on dorsal
surface.
Alpheus amblyonyx is distributed from Quin-
tana Roo (type locality, Bahia de la Ascen-
sion), Yucatan Peninsula, Mexico, to Puerto Rico,
Saint Thomas and Dominica; sublittoral (Chace
1972).
Family Hippolytidae
Three small shrimps, two males and one female
from JSL 1673 in the Gulf of Mexico off Florida,
represent an undescribed species resembling
members of the genus Ligur Sarato, 1885 from
the western Mediterranean and Indo-Pacific re-
gion (see Holthuis 1947, 1955). The specimens
were associated with burrow systems of the blue-
line tilefish, Caulolatilus microps Goode and
Bean.
73
FISHERY BULLETIN: VOL. 86, NO. 1
SUPERFAMILY PAGUROIDEA
Family Lithodidae
Lithodes maja (Linnaeus 1758). Southern limit
extended from Sandy Hook, NJ (see Williams
1984) to Baltimore Canyon.
SUPERFAMILY GALATHEOIDEA
Family Galatheidae
Munida forceps A. Milne Edwards 1880.
Geographic range extended from south of Norfolk
Canyon, 36°43.2'N, 74°38.0'W, 252 m (Wenner
1952), to Veatch and Lydonia Canyons off south-
ern New England, 103-337 m.
The distinctive color pattern of this species was
described on 14 October 1981 from specimens pre-
served in formalin 25 July 1981. Carapace (Fig.
3), salmon color with lavender submesial spots on
gastric region and interrupted U-shaped bands of
same color in nested series on mesogastric region,
posterior to cephalic groove, and arching across
posterior and posterolateral parts. Oblique red
lines on lateral wall of carapace below suture,
most prominent band along anterior edge, contin-
ued dorsally anterior to antennal peduncle and
ending on lateral side of supraocular spine. Me-
dian band of same intensity on epistome and
labium. Paler oblique lateral band on basal an-
tennular article. Some flecks of red on merus of
chelipeds and cross banding on fingers of some
individuals.
Munida longipes A. Milne Edwards 1880.
Northern limit extended from off Cape Lookout,
NC (Williams 1984) to Baltimore Canyon.
Munidopsis cf. transtridens Pequegnat and Pe-
quegnat 1971. Munidopsis transtridens is
known only from the holotype female taken in the
southeastern Gulf of Mexico at 1,280 m. The spec-
imens reported here from Baltimore, Hendrick-
son, and Lydonia Canyons off New Jersey and
southern New England, 906-1,425 m, are all
males. They resemble M. transtridens but differ
from it in rostral characters (both longer and
shorter, variably narrower or broader, in degree
of lateral convexity) and in having chelae strik-
ingly larger than the slender ones of the holotype.
Although these differences may be attributable to
sexual dimorphism, provisional identification
seems best until more material is available for
study.
Figure 3. — Munida forceps, male. Dorsal view of carapace, dia-
grammatic representation of lavender bands on salmon ground
color, carapace length 17.7 mm to base of supraocular spine.
SUPERFAMILY XANTHOIDEA
Family Goneplacidae
Chacellus filiformis Guinot 1969. Geographic
range extended from the northern Gulf of Mexico
east of the Mississippi River Delta and off the east
coast of Florida, 328-400 m (Guinot 1969), to Bal-
timore and Lydonia Canyons, 160-244 m.
Goneplacid crabs were driven out of secondary
burrows in walls of larger burrows constructed by
tilefish, Lopholatilus chamaeleonticeps Goode and
Bean, with rotenone. The poison did not kill the
crabs but caused them to emerge from the burrow
systems enough that they could be collected by
74
WILLIAMS: DECAPOD AND EUPHAUSIID CRUSTACEANS
"slurp gun". These burrow systems in Pleistocene
clay, referred to by Warme et al. (1978) and
Cooper and Uzmann (1980) as "Pueblo Villages",
shelter a number of invertebrate and vertebrate
species (Able et al. 1982; Bowman 1986; Grimes
et al. 1980a, 1980b, 1986). Goneplacids are rare in
collections made from surface vessels probably
because trawls or grabs cannot efficiently sample
the burrow systems in which these crabs have
been observed. Galatheids from shallower bur-
rows are more open to capture by conventional
means (Churchill B. Grimes^).
To the brief color description quoted from
Chace by Guinot (1969), the following can be
added from notes made 14 October 1981 on ma-
ture males, females, and juveniles that were pre-
served in formalin 25 July 1981, and personal
communication from Churchill Grimes (fn. 2).
Carapace dorsally spotted with red on off-white
background. Same type of spots on pterygosto-
mian, subocular, epistomial, and subbranchial
areas, on external maxillipeds, and on merus, car-
pus, and propodus of chelipeds (dorsally, later-
ally, and mesially). Spots tending to coalesce
along front of carapace and on chelae. Red color
more diffuse on dorsal or exposed surfaces of
walking legs, becoming more distinct and intense
with increasing size. Fingers of chelae black.
Dactyls of walking legs white except for darkened
tips, but setae pinkish. There is some variation in
pattern on individual crabs.
There is variation also in the length of the male
first pleopod, both in the USNM series of speci-
mens studied by Guinot (1969) and in the new
material reported here. In some specimens of the
latter, this appendage exceeds or at least reaches
the distal edge of the telson, whereas it is shorter
in specimens previously reported from localities
further south. In the latter, the third abdominal
segment is more angled laterally than in speci-
mens from the north. Thus, there seem to be some
differences between the northern and southern
populations.
SUPERFAMILY GRAPSIDOIDEA
Family Grapsidae
Euchirograpsus americanus A. Milne Edwards
1880. Geographic range extended north from off
^Churchill B. Grimes, Southeast Fisheries Center Panama
City Laboratory, National Marine Fisheries Service, NOAA,
.3500 Delwood Beach Road, Panama City, FL .32407-7499, pers.
commun. February 1982.
Oregon Inlet, NC (Williams 1984) to Oceanogra-
pher Canyon at the edge of Georges Bank and
nearby continental slope at 155-200 m.
SUPERFAMILY PINNOTHEROIDEA
Family Pinnotheridae
Dissodactylus juvenilis Bouvier 1917. The
ovigerous female from the Gulf of Mexico off west-
ern Florida, though similar in general features to
D. juvenilis, is very large for that species. In a
recent review of the genus Dissodactylus , Griffith
(1987) reported D. juvenilis from north of Yu-
catan and the Mississippi Delta. Members of the
genus are found in association with clypeastroid
echinoids (Schmitt et al. 1973), as was this speci-
men in a sample that included Clypeaster ravenel-
lii A. Agassizi.
ACKNOWLEDGMENTS
The following persons brought these records to
my attention through requests for identifications:
Barbara Hecker and Dennis T. Logan, Lamont-
Doherty Geological Laboratory, Columbia Uni-
versity, Palisades, NY; Kenneth W. Able, Center
for Coastal and Environmental Studies, and
Churchill B. Grimes, Department of Horticul-
ture and Forestry, Rutgers University, New
Brunswick, NJ (CBG now with Southeast Fish-
eries Center Panama City Laboratory, National
Marine Fisheries Service, NOAA, Panama City,
FL). Maureen E. Downey identified the echinoid.
Ruth E. Gibbons drafted the maps and Keiko Hi-
ratsuka Moore figured the galatheid. The
manuscript was critically reviewed by B. B. Col-
lette and D. L. Felder.
LITERATURE CITED
Able, K W , C B Grimes, R A Cooper, and J. R Uzmann
1982. Burrow construction and behavior of tilefish,
Lopholatilus chamaeleonticeps, in Hudson Submarine
Canyon. Environ. Biol. Fishes 7(3):199-205.
Able. K. W., D. C. Twitchell. C. B. Grimes, and R. S. Jones.
1987. Tilefishes of the genus Caulolatilus construct
burrows in the sea floor. Bull. Mar. Sci. 40:1-10.
Bowman. T H.
1986. Tridentella recava , a new isopod from tilefish
burrows in the New York Bight (Flabellifera: Tri-
dentellidael. Proc. Biol. Soc. Wash. 99(2):269-273.
Chace. F A, Jr
1972. The shrimps of the Smithsonian-Bredin Caribbean
Expeditions with a summary of the West Indian
shallow-water species (Crustacea: Decapoda:
Natantiaj. Smithson. Contrib. Zool. 98:1-179.
75
FISHERY BULLETIN: VOL. 86, NO 1
Cooper. R A , and J R Uzmann.
1980. Ecology of juvenile and adult Homarus. In J. S.
Cobb and B. F. Phillips (editors), The biology and man-
agement of lobsters. Vol. II, Ecology and management,
ch. 3, p. 97-142. Acad. Press, NY.
Griffith. H
1987. Taxonomic revision of the genus Dissodactylus
(Crustacea: Brachyura: Pinnotheridae). Bull. Mar. Sci.
40:397-422.
Grimes, C B , K W Able, and R S Jones.
1986. Tilefish, {Lopholatilus chamaeleonticeps), habitat,
behavior and community structure in mid-Atlantic and
southern New England waters. Environ. Biol. Fishes
15(41:273-292.
Grimes, C B . K W Able, and S C Turner
1980a. A preliminary analysis of the tilefish, Lopholatilus
chamaeleonticeps fishery in the Mid-Atlantic
Bight. Mar. Fish. Rev. 42(11):13-18.
Grimes, C G , K W Able, S C. Turner, and S J Katz
1980b. Tilefish: Its continental shelf habitat. Under-
water Nat. 12(41:34-38.
Guinot, D
1969. Recherches preliminaires sur les groupements na-
turels chez les Crustaces Decapodes des Brachyoures.
VII. Les Goneplacidae (suite et fin). Bull. Mus. Natl. Hist.
Nat. Ser. 2, 41(3):688-724.
HOLTHUIS, L B
1947. The Decapoda of the Siboga-Expedition. Part IX.
The Hippolytidae and Rhynchocinetidae collected by the
Siboga and Snellius Expeditions with remarks on other
species. Siboga-Exped. Monogr. 39aS, 100 p.
1955. The recent genera of caridean and stenopodidean
shrimps (Class Crustacea, Order Decapoda, Supersection
Natantia) with keys for their determination. Zool.
Verb. (Leiden) 26:1-157.
PEQUEGNAT, W E , AND L H Pequegnat
1971. New species and new records o{ Munidopsis (Deca-
poda: Galatheidae) from the Gulf of Mexico and
Caribbean Sea. Texas A & M Univ. Oceanogr. Stud., 1
(suppl.):3-24.
Schmitt. W L , J C McCain, and E S Davidson
1973. Decapoda I, Brachyura I, Fam. Pinnotheridae. In
H.-E. Gruner and L. B. Holthuis (editors), Crustaceorum
Catalogus 3, 160 p. Dr. W. Junk B. V.-Den Haag.
Squires, H J
1965. Decapod crustaceans of Newfoundland, Labrador
and the Canadian eastern Arctic. Fish. Res. Board
Can., Manuscr. Rep. Ser. (Biol.) 810, 212 p.
UCHUPI, E.
1965. Map showing relation of land and submarine topog-
raphy. Nova Scotia to Florida. U.S. Geol. Surv. Misc.
Geol. Invest., Map 1-451 (Sheets 1-3).
Veatch, a C, and P a Smith
1939. Atlantic submarine valleys of the United States
and the Congo Submarine Valley. Geol. Soc. Am. Spec.
Pap. 7, 101 p.
Warme. J E , R A Slater, and R A Cooper
1978. Bioerosion in submarine canyons. In D. J. Stanley
and G. Kelling (editors). Sedimentation in submarine
canyons, fans, and trenches, ch. 6, p. 65-70. Dowden,
Hutchinson & Ross, Inc., Stroudsburg, PA.
Wenner, E L
1982. Notes on the distribution and biology of Galathei-
dae and Chirostylidae (Decapoda: Anomura) from the
Middle Atlantic Bight. J. Crustacean Biol. 2(3):360-
377.
Wenner, E L , and D F Boesch
1979. Distribution patterns of epibenthic decapod Crus-
tacea along the shelf-slope coenocline. Middle Atlantic
Bight, USA. Bull. Biol. Soc. Wash. 3:106-133.
Wenner, E L , and N. T Windsor
1979. Parasitism of galatheid crustaceans from the Nor-
folk Canyon and Middle Atlantic Bight by bopyrid
isopods. Crustaceana 37(3):293-303.
Williams, A B.
1984. Shrimps, lobsters, and crabs of the Atlantic coast of
the eastern United States, Maine to Florida. Smithson.
Inst. Press, Wash., DC, 550 p.
Williams, A B , and R L Wigley
1977, Distribution of decapod Crustacea off northeastern
United States based on specimens at the Northeast Fish-
eries Center, Woods Hole, Massachusetts. U.S. Dep.
Commer., NOAA Tech. Rep. NMFS Circ. 407, 44 p.
76
AGE AND GROWTH OF LARVAL GULF MENHADEN,
BREVOORTIA PATRONUS, IN THE NORTHERN GULF OF MEXICO
Stanley M. Warlen'
ABSTRACT
Experiments on laboratory-spawned and -reared larval gulf menhaden, Brevoortia patronus , showed
that they formed one otolith growth increment per day and that the increments could be used to
estimate their age. Wild larvae from collections in the northern Gulf of Mexico along three transects
I Cape San Bias, Florida; Southwest Pass, Louisiana; and Galveston, Texas) were aged. Gompertz
growth equations were used to describe the relationship between age and standard length for larvae
collected at various locations, and in different seasons and years. MANOVA tests and subsequent
pairwise tests were used to test for differences among these growth curves. For the most extensive
data set (Southwest Pass, Louisiana), there were significant differences in growth between early
season (December) and late season (February) larvae. Early season larvae grew faster than late
season larvae. Growth of larvae also differed among December collections and among February
collections. The growth model for the pooled data for all wild larvae predicted that they grew from 2.4
mm SL at hatching to 20.4 mm SL at 62 days.
Gulf menhaden, Brevoortia patronus , is the most
abundant commercial finfish in the Gulf of Mex-
ico and, with 883,500 metric tons (t) landed in
1985 (U.S. National Marine Fisheries Service
1986); it constitutes the largest fishery in the
United States. Some aspects of the oceanic early
life history of this clupeid are known and are re-
viewed by Turner (1969), Christmas and Waller
(1975), Lewis and Roithmayr (1981), Govoni et al.
(1983), and Shaw et al. (1985a). However, virtu-
ally nothing is known about the age and growth
of the larvae, much less how these parameters
vary spatially and temporally. Daily growth in-
crements on otoliths of larval fishes can be used
as an indicator of their age, and once the use of
this technique, first described by Pannella ( 1971 ),
is validated for the larvae of an individual spe-
cies, their ages can be estimated with confidence
and growth rates can be determined. Intraspecific
growth may be compared for larvae from different
areas and seasons (Lough et al. 1982), and from
this it may be possible to ascertain how biotic and
abiotic environmental variables affect larval
growth and survival. The objectives of this study
are to 1) validate the periodicity of increment for-
mation in otoliths of larval gulf menhaden, 2)
estimate larval growth rates, 3) compare growth
rates of larvae from different locations and times,
4) estimate spawning times, and 5) examine pos-
•Southeast Fisheries Center Beaufort Laboratorv, National
Marine Fisheries Service, NOAA, Beaufort, NC 28516-9722.
Manuscript accepted September 1987
FISHERY BULLETIN. VOL. 86, NO. 1, 1988.
sible relationships between larval growth and
surface water temperature. This work was part of
a larger project designed to investigate the early
life history of several economically important
fishes and the marine planktonic food webs that
support their growth and survival in the northern
Gulf of Mexico.
METHODS
Spawning and Larval Rearing
Adult gulf menhaden were collected near Gulf
Breeze, FL, and transported to the Beaufort Lab-
oratory, Beaufort, NC (Hettler 1983). After a pe-
riod of acclimation, adults were induced to spawn
in the laboratory. The resultant larvae were used
in experiments to validate the periodicity of in-
crement formation on their otoliths and the age at
first increment formation.
Beginning February 1983, several thousand
newly spawned gulf menhaden eggs were trans-
ferred to a tank containing 90 L of filtered sea-
water. The static water in this tank, kept at
20.5° ± 0.5°C throughout the experiment, was
continuously aerated and the salinity maintained
at 31 ± 17cc. Photoperiod was 12 hours light:12
hours dark. A food concentration of 25 rotifers
(Brachionus plicatilis) mL"^ was maintained. A
green alga, Nanochloris sp., was added periodi-
cally as food for the rotifers and to aid in remov-
ing toxic metabolites. The otoliths of larvae sam-
77
KISHERY BULLETIN; VOL. 86, NO. 1
pled at 10-, 17-, 24-, and 31-d posthatch were ex-
amined.
In January 1984 additional larvae were reared
to compliment results of the earlier experiment.
Smaller tanks with 60 larvae in 10 L of filtered
water were used. Experimental conditions were
the same as for the first experiment. The otoliths
of larvae sampled at 7-, 14-, and 20-d posthatch
were examined.
Larval Collections
Larval gulf menhaden were collected in the
northern Gulf of Mexico during six cruises of the
RV Oregon II. Sampling stations (Fig. 1) along
transects LA (off Louisiana) and FL (off Florida)
were occupied during 11-19 December 1979, 5-
15 February 1980, and 2-12 December 1980 and
along transects LA, FL, and TX (off Texas) during
9-24 February 1981, 2-13 December 1981, and
4-16 February 1982. Transect LA is near the
Mississippi River outflow off Southwest Pass, LA;
transect FL is southwest of Cape San Bias, FL;
and transect TX is located off Galveston Bay, TX.
Sampling stations were in water depths of 18, 91,
and 183 m except off Texas where only the 18 and
91 m depths were sampled.
A multiple opening-closing net and environ-
mental sensing system (MOCNESS) as described
by Wiebe et al. ( 1976) were the primary sampling
gear used to capture larvae. Additional samples
were taken in oblique tows with a 60 cm bongo
frame also fitted with 505 fjim mesh nets. Samples
were collected day and night and were preserved
in 95Vf ethanol (final concentration ^757^ ) within
5 minutes of collection. The ethanol was changed
in all samples at least once after initial preserva-
tion to prevent dissolution of otoliths in fish from
any samples that may have been inadequately
preserved. Data from larvae collected at all sta-
tions within a transect were combined for that
transect.
Estimating Age and Growth
All gulf menhaden larvae were measured to the
nearest 0.1 mm standard length (SL). The largest
otolith pair (sagittae) was teased from the sur-
rounding tissue, cleaned in distilled water, and
then placed on a glass microslide under a thin
layer of Flo-Texx-^ mounting medium.
Otoliths were viewed with a compound micro-
scope fitted with a television camera. Growth in-
crements were counted from otolith images on a
video monitor at magnifications of at least 400 x.
An increment appeared as a light, wide incremen-
tal band and a dark, narrow, discontinuous band
(Tanaka et al. 1981). Increments were generally
clearly discernable and easily counted (Fig. 2).
Estimated age was the number of increments
counted plus an empirically derived value for the
2Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
96 00' 94° 00' 92° 00' 90° 00' 88° 00' 86° 00' 84° 00'
30° 00'
28° 00'
26° 00'
Figure l. — Location of sampling sites from which larval gulf menhaden were collected during crui-ses of the RV Oregon II in
December 1979-81 and February 1980-82.
78
WARLEN; AGE AND GROWTH OF LARVAL GULF MENHADEN
Figure 2. — Photomicrograph of a saggital otohth with 22 increments from a 17.4 mm SL
field collected larval gulf menhaden. Scale bar represents 10 pim. Growth increments
appear as pairs of wide incremental and narrow discontinuous bands.
number of days from spawning to first increment
formation. Results of the laboratory experiments
established the periodicity of otolith increment
formation.
A spawning date was assigned each ageable
larva by using the estimated age of the fish in
days to back-calculate from the date of capture. It
was assumed that there were no differences in
either the age at initial increment deposition or
the otolith increment deposition rate between lo-
cations and seasons and that the rate was not a
function of temperature, food, or photoperiod.
Average growth of larvae was described by the
Laird version (Laird et al. 1965) of the Gompertz
growth equation (Zweifel and Lasker 1976) fitted
to estimated age and size at time of capture for
fish from all cruises and transects. To stabilize
the variance of length over the observed age in-
terval, length data were log-transformed and
model parameters were estimated from the log-
transformed version of the growth equation. The
model was fit to data for each transect within
each cruise and for pooled data from all cruises.
Potential differences in the overall growth
curves among years and between seasons for lar-
vae caught off Louisiana and between years ( 1981
and 1982) for larvae caught off Louisiana and
Texas were examined by treating the parameters
of the Gompertz equation as dependent variables
in two-way multivariate analysis of variance
(MANOVA) designs. A one-way MANOVA de-
sign was used to test for differences
among transects (LA, FL, TX) within one season
(February 1982). Following significant
MANOVA results, prespecified pairwise
Hotelling's T'^ test comparisons (Bernard 1981, as
modified by Hoenig and Hanumara 1983) were
made using the Bonferroni procedure (Harris
1975) to provide conservative tests of statistical
significance. Bonferroni critical values for these
individual tests were equal to the overall error
rate (significance level = 0.05) divided by the
number of possible comparisons in the particular
MANOVA design. The emphasis in the compari-
sons was to look for overall differences in the
growth of larvae using these statistics as a guide
and not to look for differences in individual
parameters of the growth models.
Hotelling's T^ test and MANOVA both require
that the data fit a multivariate normal distribu-
tion and that the variance-covariance matrices of
the populations are not different (Harris 1975).
These assumptions are difficult to test and are
almost certainly not valid for real data sets (par-
79
FISHERY BULLETIN: VOL 86. NO 1
ticularly field data), but they may be nearly valid
for many sets of data (Harris 1975). No direct test
of normality in a trivariate, joint probability dis-
tribution is available (Bernard 1981), but bias
arising from nonnormal, multivariate, joint dis-
tributions is minimized with large sample sizes
(Bernard 19811. While methods are available to
test the hypothesis of equal variance-covariance
matrices (e.g.. Box's modification of Bartlett's
test), these methods are very sensitive and even
minor differences between group dispersions will
likely be discovered (Pimentel 1979). In any
event, the use of MANOVA in this paper relies on
variance-covariance matrices estimated from
nonlinear regressions, and these are not
amenable to testing. However, both MANOVA
and Hotel ling's tests are extremely robust even
under violation of the assumptions of homo-
scedasticity and multivariate normality (Harris
1975).
RESULTS
Increment Formation
The age of gulf menhaden at formation of the
first otolith growth increment was estimated
from laboratory-reared larvae. The intercept (2.6
days) of the regression of the number of growth
increments on known posthatch age of 36 larval
gulf menhaden (Fig. 3) was used to estimate
posthatch age at formation of the first increment.
This value was added to the time from spawning
to hatching which at 20°C is 2 days (Powell'M.
This sum (4.6 days) is the estimated time from
spawning to formation of the first increment.
Hence, it was necessary to add 5 days to each
increment count to estimate the age of larval gulf
menhaden from spawning.
The periodicity of increment formation was as-
certained from the regression of the number of
growth increments on the known age (Fig. 3). The
slope did not differ significantly (/-test, P < 0.05)
from 1.0, and thus, on the average, one otolith
growth increment was formed per day in
laboratory-reared larvae up to 31 days after
hatching. Results of a second experiment (Table
1 ) confirmed this periodicity. The age of gulf men-
haden larvae estimated from otolith increment
counts ( + 5) closely approximated the known ages
of 51 laboratory-reared larvae. Mean estimated
age of larvae differed by < 1 day from the known
3A. B. Powell, Southeast Fisheries Center Beaufort Labora-
tory, National Marine Fisheries Service, NOAA, Beaufort, NC
28516-9722, pers. commun. February 1986.
CO
H
Z
111
liJ
DC
O
z
O
o
o
tr
LU
m
Z
35
30
25
20
15
10
No. increments=-2.617+ 0.921(known age)
r= 0.928
n= 36
5-
10 15 20 25
KNOWN AGE (days)
30
35
Figure 3. — Regression of the number of
growth increments on the known
posthatch age of 36 laboratory-reared
gulf menhaden. Standard error of the
slope IS 0.108.
80
WARLEN; AGE AND GROWTH OF LARVAL GULF MENHADEN
Table 1. — Standard length (mm) and estimated age (num-
ber of otolith growth increments +5) of laboratory-reared
larval gulf menhaden. Values in parentheses are 95% inter-
val estimates.
Known age Number of Mean estimated
(days) fish age
Mean SL
7
14
20
13
16
22
7.8 (±0.38)
13.2 (±1.14)
19.3 (±0.77)
4.7 (±0.28)
6.4 (±0.38)
8.0 (±0.42)
ages and the 957c confidence intervals included
the known age in each of the three groups. Some
of the variation in the number of gi'owth incre-
ments observed in known age larvae (Fig. 3,
Table 1) may have been due to 1) poor growing
conditions during rearing that could have re-
sulted in reduced growth in underfed larvae, 2)
variation in the inception of increment formation
as has been observed in other species (Laroche et
al. 1983; Fives et al. 1986), and 3) faintness of
growth increments in some larvae. In contrast,
increments on otoliths of field collected larvae
(Fig. 2) were usually very regular and distinct
and were more easily observed than those on
otoliths of laboratory-reared larvae. I assumed
that the gi'owth increment deposition rate was
also daily in wild larvae examined in this study.
Age and Growth of Larvae
Average growth of larval gulf menhaden dur-
ing their first two months of life was described by
the Gompertz growth model for pooled length at
age data for 2,003 fish representing collections
from all six RV Oregon II cruises (Table 2, Fig. 4).
Larvae ranged in age from 5 to 62 days (x=24A
days) and in SL from 3.4 to 28.0 mm (x = 12.6
mm). In the log-transformed model, age ac-
counted for 82% of the variation in length. Gulf
menhaden were predicted to have grown from 2.4
mm SL at hatching to 20.4 mm at age 62 days.
The size at hatching estimated from the Gom-
pertz equation was only slightly less than the
hatching size, 2.6-3.0 mm SL, observed in the
laboratory (Hettler 1984). Age-specific growth
rates declined from — 79f/day at age 10 days to
<0.4%/day at age 60 days. Maximum absolute
growth rate occurred when gulf menhaden larvae
were 7.9 mm SL and 13 days old.
The asymptotic length of larvae (21.5 mm SL),
determined from the variables of the growth
equation, is approximately the size when larvae
begin to transform into juveniles. This transfor-
mation, described by Lewis et al. (1972) for the
closely related Atlantic menhaden, B. tyrannus,
apparently ends when the fish reach 28-30 mm
SL (Suttkus 1956).
In all instances except one transect (TX Decem-
ber 1981, where there was no convergence in the
parameter values in the computer fitting proce-
dure and the model would not fit the data), the
Gompertz growth model could be used to describe
the growth of gulf menhaden larvae from each
cruise and transect (Figs. 5-7, Table 2). The
growth model for the FL December 1980 larvae
approximates an exponential form because of the
exceptionally low value for a. This may be due to
the preponderance of small, young larvae.
GROWTH COMPARISONS
Louisiana - Seasons and Years
There were statistically significant differences
(MANOVA, P < 0.001) in the growth curves for
larvae caught off Louisiana for two seasons (De-
cember, February) and three years (1979-80,
1980-81, 1981-82). To determine if differences ex-
isted between seasons in each year and among
any two years within each season, I selected 9 of
the possible 15 pairwise comparisons for testing.
The Bonferroni critical value in these tests was
0.0033 (0.05/15). The inability to fit a Gompertz
growth model to the TX December 1981 data pre-
cluded a comparison with the larvae collected off
Texas in February 1982.
Pairwise comparisons for within years data for
Louisiana larvae showed significant differences
(P < 0.003) in growth curves between early sea-
son (December) and late season (February) for
each year. Faster growth of early season larvae is
evident if the respective curves (Figs. 5; 6a, c; 7a,
c) are compared. For any age, the predicted size is
greater for early season than for late season lar-
vae. Only for the third year did the length at age
40+ days of February-caught larvae exceed that
for December-caught larvae.
In similar comparisons for larvae caught off
Louisiana in December of all three years, there
were significant differences (P< 0.003) in the
growth curves (Figs. 5a, 6a, 7a) for any two years.
As judged by the predicted size at any age, larvae
appeared to grow faster in 1979 than in either
1980 or 1981. While the curves for the 1980 and
1981 larvae overlapped, larvae from 1980 were
larger at 30+ days than were the 1981 larvae.
Significant differences were also found among the
curves (Figs. 5b, 6c, 7c) for larvae caught in
81
F1SHF:KV BUI.l.KTlN VOL Hli, NU 1
Table 2.— Estimates of Gompertz growth model parameters and mean age (days) and mean SL (mm) for
larval gulf menhaden collected m the northern Gulf of Mexico during the winters 1979-80, 1980-81, and
1981-82. fl2 is the coefficient of determination for the respective models.
Number
fish
Growth model parameters^
fvlean
estimated
tVlean
Date
Transect'
aged
R2
'-(O)
A(0)
a
age (d)
SL (mm)
Winter 1979-80
Dec. 1979
LA
42
0863
2 768
(1.270)
0 1701
(0,0711)
00809
(00190)
28 1
174
Feb. 1980
LA
324
0954
2.131
(0 138)
0 1592
(00105)
00710
(0 0031)
302
12.7
Winter 1980-81
Dec. 1980
LA
191
0931
2888
(0.188)
0 1125
(0,0097)
00496
(0.0044)
220
118
Dec. 1980
FL
80
0701
3418
(0.994)
00561
(0 0370)
0.0001
(0,0407)
14 9
79
Feb. 1981
LA
338
0849
2702
(0.231)
0 1159
(0 0120)
00577
(0,0049)
246
11.7
Feb. 1981
TX
223
0526
5839
(0.579)
00305
(00088)
0,0125
(0,0110)
21,5
10.3
Winter 1981-82
Dec. 1981
LA
370
0921
0384
(0.076)
04780
(0,0433)
0 1240
(0,0054)
21 7
13.3
Dec. 1981
TX
114
—
.1
3
,1
25,4
11.9
Feb. 1982
LA
191
0,736
0807
(0,337)
02729
(0,6093)
0,0851
(0,0090)
31 2
15.5
Feb. 1982
FL
88
0.624
2278
(0,824)
0 1067
(0,0454)
0,0394
(0,0205)
17,6
89
Feb 1982
TX
42
0946
1 798
(0,392)
0,1276
(0,0340)
00500
(0,0142)
225
100
All years
All data pooled
2,003
0822
2,355
(0.098)
0,1345
(0,0059)
0,0608
(0,0020)
24.4
12.6
iLA=IVIississippi River Delta (Southwest Pass Louisiana); FL = Cape San Bias, FL; TX = Galveston, TX.
2L|0| = length at hatching, A(0) = specific growth rate at hatching, « = exponential decline in A,o)- Values in
parentheses are estimated standard errors from the nonlinear regressions.
3Gompertz growth model did not fit the data.
E
E
X
»-
CD
Z
lU
_l
Q
<
Figure 4.— Growth of gulf menhaden
larvae collectccl in the winters of 1979-
80, 1980-81, and 1981-82 m the north-
ern Gulf of Mexico. The Gompertz growth
model was used to describe the pooled
data. Two through nine coincident data
points are labelled with their numeral.
Coincident points of 10 and above are la-
belled A, B, etc.
CO
30
25
20
15
10
1
1
1 1
1
3
1 1 1
1 1
1 1
2 2231 1 1111 1 11 1121
1 112 11 13313221331
21 4 222635231 8 3687ia<T 2 1
224235B253223445«7B32 2
8562 22 82 5 I 2J-9r536831 221
122 B8C49221AS3251 623352 11
1 1 780DB425*35253342 1214 1
4 5238 1 CBie58 31314 23
327586F9NBC65 6272331 12
34EB9A56Q7H561311 11 1
1 274E6E>C6K5H28 111 1
2194EaC6546383331 1
55E5BaC665A2A2 1
" ::: _ 31 1 i
L59422
-0.0608t
n= 2,003
10
20
30
40
50
60
70
ESTIMATED AGE (cJays)
82
WARLEN; AGE AND GROWTH OF LARVAL GULF MENHADEN
E
E
I
(-
o
z
LU
_l
Q
cr
<
Q
W
a
?
1
20
1 1 >
' 12^
''^1 1 1
1
1
15
"
A 1 1
1
10
-
/
t
' \
5
-
1
1
1
1
Louisiana
Dec 1979
1 1
10
20
30
40
50 60
20
15
10-
11 11 12
1 2 2232221 1 123 11
114"
" ' 1
322 1 1
833122
1 32
212 1
223
121
10
20
30
40
Louisiana
Feb 1980
50
60
ESTIMATED AGE (days)
Figure 5. — Growth of larval gulf menhaden collected in the winter
1979-80 in the northern Gulf of Mexico. The Gompertz growth model
was used to describe the data. Coincident data points are labelled as in
Figure 4.
February in three years. Larvae caught in 1980
grew faster than larvae caught in 1981 and
larvae caught in 1982 up to 25 days, thereafter
1980 and 1982 had very similar size at age esti-
mates.
Louisiana vs. Florida vs.
Texas - February 1982
There were no significant differences
(MANOVA, P = 0.212) among the growth curves
for larvae caught in February 1982 off Louisiana,
Texas, and Florida (Figs. 7c, d, e), and hence no
pairwise comparisons were necessary.
Louisiana vs. Texas - February
1981-82
Statistically significant differences (MAN-
OVA, P < 0.002) in larval growth were observed
in the LA and TX transects from February 1981-
82. Pairwise comparisons indicated significant
differences (P < 0.008) in the growth of larvae
collected off Texas in 1981 (Fig. 6d) and 1982 (Fig.
7e) and in growth between LA 1981 (Fig. 6c) and
TX 1981 collections. The earlier pairwise com-
parisons had already shown a significant differ-
ence in growth of larvae from LA February 1981
and February 1982 collections (Fig. 7c), but none
for growth of larvae caught in the LA February
1982 and TX February 1982 collections. Two
other potential tests, between transects of differ-
ent areas and different years, were not considered
to be meaningful.
Larvae caught off Louisiana in February 1981
were larger at age 18+ days than were larvae
caught off Texas in February 1981, and might be
considered to be faster growing fish. There was a
83
FISHERY BULLETIN: VOL 86, NO 1
E
E
I
H
o
z
LU
_J
Q
cr
<
Q
Z
<
I-
25
20
15
10
5-
25
20
15
10
5-
Louisiana
Dec 1980
j_
Louisiana
Feb 1981
Florida
Dec 1980
-L, I
1 1
1 1
2 2
31J3512 1
1 4 244
2 126263M?22
3122ttaft?7311
511 1
__J3323 23 1
2 12131
11112
10
20
30
40
50
60
10
20
30
40
Texas
Feb 1981
50
60
ESTIMATED AGE (days)
Figure 6.— Growth of larval gulf menhaden collected in the winter 1980-81 in the northern Gulf of Mexico. The Gompertz growth
model was used to describe the data. Coincident data points are labelled as in Figure 4.
statistical difference in growth of larvae caught
off Texas in February 1981 and 1982, and it ap-
pears that the 1982 larvae grew at a faster rate.
Conclusions from these statistical differences in-
volving TX February 1981 larvae collections
should be viewed with caution because of the rel-
atively poor fit (r^ = 0.526) of the model. Inade-
quacies, such as the lack of larvae <13 or >31
days old, in that data set probably resulted in the
relatively poor parameter estimates (Table 2).
Additional sampling would be necessary to fur-
ther test the hypotheses of differences in growth
between geographic areas in the northern Gulf of
Mexico and between years for Texas larvae.
84
Estimated Spawning Times
Gulf menhaden larvae observed in this study
were estimated to have been spawned from mid-
October to mid-February (Fig. 8). The limited ex-
tent of seasonal sampling precluded estimation of
the probable total range of the spawning season.
Most larvae captured in December and February
had been spawned in November and January re-
spectively (Fig. 8). The considerable overlap in
spawning times of larvae caught the same month
in different years is a reflection of the similarity
of sampling dates. The relatively narrow distribu-
tion of spawning dates for larvae caught off Flor-
WARLEN; AGE AND GROWTH OF LARVAL GULF MENHADEN
20
15
10
E
3
20
I
(-
(0
z
lb
LU
_l
D
CC
10
<
O
z
<
t-
b
co
20
15-
10-
Louisiana
Dec 1981
I
b
-
2
41 1
173 721
1 1252925
1 21435131
1 13262
112
122
611
3
1
1
Texas
Dec 1981
1 1
1
1 1 1
1 1 1512 21;
21246212 "
2 1611
12 51 i^21 1 t
r4121 1 1
1212 ^\ 14 31
nil 222
21 1
2 1 1
Louisiana
Feb 1982
_L
Texas
Feb 1982
J_
J-
10 20 30 40
ESTIMATED AGE (days)
50
60
Florida
Feb 1982
I
10 20 30 40 50
ESTIMATED AGE (days)
60
Figure 7. — Growth of larval gulf menhaden collected in
the winter 1981-82 in the northern Gulf of Mexico. The
Gompertz model was used to describe the data from all
transects except Texas December 1981 where it could
not be made to fit the data. Coincident data points are
labelled as in Figure 4.
ida in both December 1980 (Fig. 6b) and in Febru-
ary 1982 (Fig. 7d), off Louisiana in February 1982
(Fig. 7c), and off Texas in February 1981 (Fig. 6d)
represent larvae from fewer cohorts.
DISCUSSION
Laboratory observations indicate that larval
gulf menhaden on the average form one growth
increment per day on their otoliths and that
counts of these increments can be used to esti-
mate age. Otoliths of larval gulf menhaden are
thin and round, and the increments are generally
easily counted and consequently are ideally
suited for ageing. The most closely spaced incre-
ments, those occurring near the focus, were at
least 1.5 |xm wide and were above the 0.2 jxm
resolution of the light microscope (Campana and
Neilson 1985). First increment formation occurs
about 5 days after spawning and probably coin-
cides with first exogenous feeding. This is sup-
ported by Hettler ( 1984) who found that gulf men-
haden eggs hatched at about day 1.7 at 19°-20°C.
Four days after hatching larvae had functional
85
FISHERY BULLETIN: VOL. 86, NO. 1
(42)
1979
D ILA
(324)
I — C
1980
LU
<
Q
LU
CC
Q.
<
O
(191)
(80
)l— {JH
FL
rn — HLA
1980
(338)h
(223)1 —
1981
LA
c±i ^L/
-Cp ITX
(88)1 IJHFL
1982
DEC I Jan I FEB '
SPAWNING DATE
Figure 8. — Schematic plots of the spawning times of larval gulf menhaden collected in the northern
Gulf of Mexico during 6 cruises of the RV Oregon II from December 1979 to February 1982. In each
distribution the vertical line is the median value and 50% of the data points fall within the block. Lines
beyond the boxes represent the range of data points. The value in parentheses to the left of each
distribution is the number of fish.
mouths and were 4.5 mm SL. However, develop-
mental rates are probably temperature depen-
dent (Powell and Phonlor 1986), and hence larvae
at lower temperatures would be older at first feed-
ing.
The Gompertz growth model appears to ade-
quately describe the growth of larval gulf men-
haden in most cases. Except where data are some-
what limited (Figs. 6b, d; 7b, d) the fit of the
model is relatively good and the r'^ is >0.73 for
each transect (Table 2). Gompertz gi'owth models
have been used (Zweifel and Lasker 1976; Methot
and Kramer 1979; Laroche et al. 1982; Warlen
and Chester 1985) to describe growth of larval
fishes where the length-age plots are nonlinear
and upper asymptotes were apparent.
Average growth rate of larval gulf menhaden to
day 60 was 0.30 mm/day throughout its oceanic
existence. This rate was very similar to that, 0.28
mm/day (estimated from figure 2 of Hettler 1984),
for larvae reared in the laboratory at 20° ± 2°C
for 60 days. However, wild-caught larvae were
from wider extremes in water temperature, with
mean early season (December) temperatures
from 17.4° to 21.2°C and late season (February)
12.9° to 16.4°C. The growth rate of larval Atlantic
herring, Clupea harengus , up to 50 days old was
similar and varied between 0.23 and 0.30 mm/day
(Lough et al. 1982). However, gulf menhaden lar-
vae grew slower than the fast growing but rela-
tively short-lived engraulids — bay anchovy, An-
choa mitchilli (Fives et al. 1986) and northern
anchovy, Engraulis mordax (Methot and Kramer
1979).
Only a small number of larvae from all the
collections were 2^50 days old. Larvae of this age
86
WARLEN; ACE AND GROWTH OK LARVAL C.rLF MENHADEN
were either not in the sampling area or were inac-
cessible to the fishing gear used. Although the
latter cannot be fully discounted, the former pos-
sibility is most likely, since larvae as they g:-ow
are known to be transported (Shaw et al. 1985b)
toward estuaries. Larvae are about 15-25 mm SL
(estimated from Suttkus 1956) when they enter
estuaries in Louisiana, and the smallest immi-
grating larvae are estimated from the growth
model <Fig. 4) to have been at least 30 days old.
Larvae then are probably 30-70 days old when
they enter Gulf of Mexico estuaries. This range is
very similar to that for Atlantic menhaden, B.
tyranniis, entering North Carolina estuaries (S.
M. Warlen, unpubl. data). The so-called "larval
drift" period for gulf menhaden is probably closer
to 4-10 weeks than the 3-5 weeks surmised by
Reintjes (1970).
Growth of larval gulf menhaden in the north-
ern Gulf of Mexico varied both spatially and tem-
porally. For three consecutive years there were
significant differences in gi'owth for early season
(December) and late season (February) larvae
caught off the Mississippi River Delta. The in-
crease in length for early season larvae was
greater than for larvae hatched in late season.
Environmental conditions in this area differed
between early season and late season. Mean
water temperature measured during the Decem-
ber 1979, 1980, and 1981 cruises were 17.4°, 19.4°,
and 21.2°C, respectively, while in February 1980,
1981, and 1982 the temperatures were 13.8°,
15.7°, and 14.7°C, respectively. Although not
shown experimentally for gulf menhaden larvae,
there is evidence that larvae of some marine
fishes grow faster at higher temperatures (Lau-
rence 1978; Laurence et al. 1981). Jones (1985)
also associated higher water temperatures with
higher growth of larvae and found that increase
in length of Atlantic herring larvae hatched early
in the season was greater than for larvae hatched
late in the season.
Growth rate of larvae caught in the same sea-
son but in different years was inversely related to
mean water temperature. Larvae caught in De-
cember off Louisiana showed a trend of higher
growth (1981 < 1980 < 1979) at lower respective
mean water temperatures (21.2°, 19.4°, 17.4°C);
similarly, the growth rates for larvae caught in
February (1981 < 1982 < 1980) was higher at
lower respective mean water temperatures ( 15.7°,
14.7°, 13.8°C). Other environmental factors in ad-
dition to temperature may also affect the growth
rate of larval menhaden. Food availability that
can be an important growth-limiting factor for
larval fishes, may be determining the relative
growth rates at the lower temperatures in Febru-
ary. On the basis of limited data on the zooplank-
ton (pelecypod larvae, copepod nauplii, cope-
podites, and adult copepods) that could serve as
food for gulf menhaden larvae (Govoni et al.
1983), food availability (number/100 m'^) is
highest in February 1980, lower in 1982, and low-
est in 1981. Analogous food abundance data is not
available for the December cruises, but levels
would probably need to be higher on a per fish
basis to support the higher metabolism concomi-
tant with the higher December mean water tem-
peratures (17.4°-21.2°C). The main and interac-
tion effects of growth-limiting (food abundance)
and growth-regulating (temperature) factors on
larval gulf menhaden growth still must be deter-
mined experimentally, preferably using labora-
tory-spawned and -reared larvae.
The apparent growth advantage enjoyed by
menhaden larvae spawned early in the season
(November) is only typical for a small part of the
population. The largest segment of the popula-
tion, those spawned in the peak months of Janu-
ary and February (Christmas and Waller 1975)
and that immigrated to estuaries in February-
April, typically had slower growth. Although
slower growing, larvae spawned later in the sea-
son may be more successful in reaching estuaries.
Guillory et al. (1983) found a negative relation-
ship between temperature and gulf menhaden re-
cruitment into Louisiana estuaries. The larger
estuarine recruitment later in the season may be
related to winter-early spring (January-April)
predominant west-northwest longshore flow of
coastal water within and just outside the coastal
boundary front producing longshore advective
transport and of lesser importance by episodic,
short-term cross-shelf advection associated with
cold fronts (Shaw et al. 1985b). They hypothesized
that longshore currents facilitated the movement
of larvae toward shore. Only for December did
they note a reverse flow (eastward) that would
not allow larvae to be transported toward estuar-
ies west of the Mississippi delta.
The between transect comparisons of growth
rate of larvae caught off Texas and Louisiana in
February showed a difference in 1981 but not in
1982. Again higher growth rates were associated
with higher mean water temperatures. Larvae
from the LA February 1981 sample (x water tem-
perature = 15.7°C) grew faster than larvae from
the TX February 1981 sample U water tempera-
87
FISHERY HULl.KTIN: VOL. 86, NO 1
ture = 12.0°C). Where no significant differences
in growth were found, i.e., between LA and TX
February 1982 samples, the respective mean
water temperatures were 14.7° and 14.4°C. Nei-
ther of those growth curves were significantly dif-
ferent from the curve for larvae from the FL
February 1982 sample where the mean water
temperature was 16.4°C. However, the paucity of
larvae >23 days old caught off Florida in Febru-
ary 1982 (Fig. 7d) suggests that comparisons of
that data set with the data sets for larvae caught
off Louisiana and Texas in February 1982 would
be of little value.
The estimated spawning period for gulf men-
haden extended from mid-October to mid-
February (Fig. 8). These results agree with Fore
(1970) and Christmas and Waller (1975) who,
using the occurrence of eggs and larvae, esti-
mated that gulf menhaden spawned from mid-
October through March. Gonad weight-body
weight ratios of adults (Lewis and Roithmayr
1981) and morphological and physiological fea-
tures of ovarian tissue (Combs 1969) also indicate
that spawning extends from October to early
March. Based on the movement of late larvae into
Lake Pontchartrain, Suttkus (1956) presumed
that gulf menhaden spawning began in October
and ceased in February. He suggested that the
beginning and end of the spawning period fluctu-
ates from year to year, and that there is no spawn-
ing activity during the spring and summer
months as Higham and Nicholson (1964) have
reported for the closely related Atlantic men-
haden.
Most larvae caught in December were spawned
in November (Fig. 8) regardless of the year. Lar-
vae caught in February were spawned mostly in
January but estimated spawning dates extended
from mid-December to mid-February. For any
given cruise, larvae from off Texas and Louisiana
were spawned at about the same time. There was
also considerable overlap in the spawning dates
in any cruise off Florida and the other areas. The
distribution of the central 507^ of spawning dates
from the Louisiana sample in February 1980 ex-
tended over a 29-d period and was wider than for
any other data set. This unusually wide distribu-
tion may have been due to the presence of two
distinct cohorts, one spawned in late December
and one in late January, collected on the Febru-
ary 1980 cruise. Combs (1969) found that this
species had intermittent total spawning. Lewis
and Roithmayr (1981) inferred that gulf men-
haden were intermittent, or fractional spawners.
Christmas and Waller (1975) noted a modal tem-
poral distribution of eggs in the region from the
Mississippi delta to east of Cape San Bias. Bal-
dauf sampled young menhaden in the lower
Neches River, TX, from November through April
and found two incoming populations from which
he suggested that there may have been two
spawning peaks. Only in the larval collections of
December 1981 did spawning date distribution
appear to be bimodal; 7 and 20 November for Lou-
isiana and 8 and 19 November for Texas. Future
sampling throughout the spawning season will be
necessary to determine the seasonal periodicity
and peaks of gulf menhaden spawning. Relative
numbers of larvae in cohorts within the spawning
season could then be compared with measure-
ments of environmental conditions as a test of the
match-mismatch hypothesis (Cushing 1975) and
to further test, as Methot (1983) has done,
whether larvae spawned during favorable envi-
ronmental periods constitute the greatest per-
centage of the year class.
ACKNOWLEDGMENTS
I thank the following persons of the Beaufort
Laboratory: M. Boyd who extracted, mounted,
and aged otoliths, W. Hettler who spawned gulf
menhaden and furnished eggs for the laboratory
experiments, A. Chester who advised on statisti-
cal procedures and problems, D. Ahrenholz and
A. Powell who reviewed early drafts of the
manuscript, and the crew and scientists on the
RV Oregon II cruises. P. Ortner of the Atlantic
Oceanographic and Meteorological Laboratories,
NOAA, Miami, FL, provided raw data from which
zooplankton counts were summarized. This re-
search was supported by a contract from the
Ocean Assessments Division, National Ocean
Service, NOAA.
LITERATURE CITED
Bernard. D R
1981. Multivariate analysis as a means of comparing
growth in fish. Can. J. Fish. Aquat. Sci. 38:233-236.
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.
1975. Location and time of menhaden spawning in the
4Baldauf R. J. 1954. Survey and study of surface and sub-
surface conditions in and around Beaumont, Texas. Biological
survey of the Neches River in the region of Beaumont,
Texas. Texas A&M Res. Found., Mimeo. Rep., 184 p.
88
WARLEN; AGE AND GROWTH OF LARVAL GULF MENHADEN
Gulf of Mexico. Gulf Coast Res. Lab.. Ocean Springs,
MS, 20 p.
Combs. R M
1969. Embryogenesis, histology, and organology of the
ovary of Breioortia patronus . Gulf Res. Rep. 2:333-
434.
CUSHINO. D H
1975. Marine ecology and fisheries. Cambridge Univ.
Press. Cambridge, Eng., 278 p.
Fives. J M , S M Warlen. and D E Hoss
1986. Aging and growth of larval bay anchovy, Anchoa
mitchilli, from the Newport River estuary, North Caro-
lina. Estuaries 9(4B):362-367.
Fore. P L
1970. Eggs and larvae of the Gulf menhaden. In Report
of the Bureau of Commercial Fisheries Biological Labo-
ratory, Beaufort. N.C., p. 11-13. U.S. Fish Wildl. Serv.
Circ. 341.
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,
Leiostomus xanthurus, and Micropogonias undula-
tus. Mar. Ecol. Prog. Ser. 13:189-199.
GUILLORY. V . J GEAGHAN. AND J. ROUSSEL
1983. Influence of environmental factors on gulf men-
haden recruitment. La. Dep. Wildl. Fish. Tech. Bull. 37,
32 p.
Harris. R J
1975. A primer of multivariate statistics. Acad. Press,
N.Y., 332 p.
Hettler, W F
1983. Transporting adult and larval gulf menhaden and
techniques for spawning in the laboratory. Prog. Fish-
Cult. 45:45-48.
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.
HiGHAM, J R . AND W R NICHOLSON
1964. Sexual maturation and spawning of Atlantic men-
haden. Fish. Bull., U.S. 63:255-271.
Hoenig. N a . AND R G Hanumara
1983. Statistical considerations in fitting seasonal growth
models for fishes. Cons. Int. Explor. Mer, CM. 1983/
D:25, 25 p.
JONES. C
1985. Within-season differences in growth of larval At-
lantic herring, Clupea harengus harengus . Fish. Bull.,
U.S. 83:289-298.
Laird. A K . S A Tyler, and A D Barton
1965. Dynamics of normal growth. Growth 29:233-
248.
Laroche. J L . S L Richardson, and A A Rosenberg
1982. Age and growth of a pleuronectid, Parophrys ve-
tulus, during the pelagic larval period in Oregon coastal
waters. Fish. Bull., U.S. 80:93-104.
Laurence. G C
1978. Comparative growth, respiration and delayed feed-
ing abilities of larval cod (Gadus morhua ) and haddock
{Melanogrammus aeglefinus) as influenced by tempera-
ture during laboratory studies. Mar. Biol. (Berl.) 50:1-
7.
Laurence, G C . A S Smigielski. T A Halavik. and B R
Burns
1981. Implications of direct competition between larval
cod {Gadus morhua I and haddock i Melanogrammus ae-
glefinus ) in laboratory growth and survival studies at
different food densities. Rapp. P. -v. Reun. Cons. int. Ex-
plor. Mer 178:304-311.
Lewis. R M . and C M Roith.mayr
1981. Spawning and sexual maturity of gulf menhaden,
Brevoortia patronus . Fish. Bull., U.S. 78:947-951.
Lewis. R M . E P H Wilkins, and H R Gordy
1972. A description of young Atlantic menhaden, Bre-
voortia tyrannus, in the White Oak River estuary, North
Carolina. Fish. Bull., U.S. 70:115-118.
Lough. R G . M Pennington. G R. Bolz. and A A Rosenberg
1982. Age and growth of larval Atlantic herring, Clupea
harengus L., in the Gulf of Maine-Georges Bank region
based on otolith growth increments. Fish. Bull., U.S.
80:187-199.
Methot. R D. JR
1983. Seasonal variation in survival of larval northern
anchovy, Engraulis mordax. estimated from the age dis-
tribution of juveniles. Fish. Bull., U.S. 81:741-750.
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.
Pannella. G
1971. Fish otoliths: Daily growth layers and periodical
patterns. Science (Wash., D.C.) 173: 1124-1127.
Pimentel. R a
1979. Morphometries, the multivariate analysis of biolog-
ical data. Kendell/Hunt Publ. Co., Dubuque. lA, 276
P-
Powell. A B , and G Phonlor
1986. Early life history of Atlantic menhaden, Brevoortia
tyrannus , and gulf menhaden, B . patronus . Fish. Bull.,
U.S. 84:991-995.
Reintjes. J W
1970. The gulf menhaden and our changing estuar-
ies. Proc. Gulf Caribb. Fish. Inst. 22:87-90.
Shaw. R F . J H Cowan. Jr , and T L Tillman
1985a. Distribution and density of Brevoortia patronus
(gulf menhaden) eggs and larvae in the continental shelf
waters of western Louisiana. Bull. Mar. Sci. 36:96-
103.
Shaw. R F . W J Wiseman, Jr . R E Turner. L J Rouse. Jr .
R E Condrey. and F J Kelly. Jr
1985b. Transport of larval gulf menhaden Brevoortia pa-
tronus in continental shelf waters of western Louisiana:
A hypothesis. Trans. Am. Fish. Soc. 114:452-460.
Suttkus, R D
1956. Early life history of the gulf menhaden. Brevoortia
patronus. in Louisiana. Trans. N. Am. Wildl. Conf
21:390-407.
Tanaka. R. Y Mugiya. and J Yamada
1981. Effects of photoperiod and feeding on daily growth
patterns in otoliths of juvenile Tilapia nilotica. Fish.
Bull., U.S. 79:459-466.
Turner. W R
1969. Life history of menhadens in the eastern Gulf of
Mexico. Trans. Am. Fish. Soc. 98:216-224.
US National Marine Fisheries Service
1986. Fisheries of the United States, 1985. U.S. Natl.
Mar. Fish. Serv. Curr. Fish. Stat. 8380, 121 p.
Warlen, S. M.. and A. J Chester.
1985. Age, growth, and distribution of larval spot, Leios-
tomus xanthurus, off North Carolina. Fish. Bull., U.S.
83:587-599.
89
FISHERY BULLETIN: VOL. 86, NO. 1
WiEBK. P H . K H Burt. S H Boyd, and A W Morton Zweifei,. J R . and R Lasker
1976. A multiple opening/closing net and environmental 1976. Prehatch and posthatch growth of fishes — a general
sensing system for sampling zooplankton. J. Mar. Res. model. Fish. Bull., U.S. 74:609-621.
34:313-326.
90
SOURCES OF VARIATION IN CATCH PER UNIT EFFORT
OF YELLOW! AIL FLOUNDER, LIMANDA FERRUGINEA (STORER),
HARVESTED OFF THE COAST OF NEW ENGLAND
LORETTA O BrIEN AND RALPH K. MaYO^
ABSTRACT
Factors affecting variability in commercial catch per unit effort (CPUE) of yellowtail flounder were
examined in order to establish a basis for standardizing fishing effort. Analysis of variance ( ANOVA)
procedures were employed to test for differences in CPUE among vessel tonnage class, fishing area,
and depth zone and the interactions between tonnage class and area, and tonnage class and depth.
Vessel tonnage class and fishing area accounted for highly significant IP < 0.01) sources of variation
in CPUE whereas depth was not significant (P > 0.05) in most cases. Interactions between tonnage
class and stock area were also highly significant in all cases. A series of annual fishing power
coefficients was computed for each tonnage class relative to a standard for each stock based on
parameter estimates obtained by fitting the CPUE observations to a linear model with tonnage class
as the independent variable. Deviations of annual fishing power coefficients from the 20-year mean
were found to exhibit significant first order autocorrelations. Consequently, annual coefficients were
computed over the entire 1964-83 period by incorporating tonnage class, annual and seasonal effects
as independent variables in a three-way linear model. Although the standardized CPUE estimates
obtained from this procedure are similar to those obtained by previous methods, the revised proce-
dures described in this paper insure adequate representation of all vessel classes engaged in the
yellovfcrtail fishery in the CPUE calculations.
Fishing effort and resulting catch per unit effort
(CPUE) indices are routinely used in assessing
the impact of commercial fishing operations on
stock abundance. The traditional concept that ag-
gregate CPUE indices may be used to measure
annual changes in relative stock abundance is
based on the principal assumption that the catch-
ability coefficient (q) either remains constant
over all fleet components, or that nominal effort is
adjusted to account for differences in relative effi-
ciencies (Pope and Parrish 1964; Kimura 1981).
Variation in q may be due to persistent differ-
ences in fishing power of various types of gear or
to technological innovations which may be gradu-
ally introduced over time (Gulland 1964; Sis-
senwine 1978). Biological interactions such as
changes in availability of a species due to sea-
sonal distribution patterns or to annual changes
in abundance may also affect the overall catcha-
bility of demersal species (Garrod 1964; Pope and
Garrod 1975). Variability in catchability coeffi-
cients may be taken into account by relating nom-
inal fishing effort of each fleet component to some
chosen standard category.
'Northeast Fisheries Center Woods Hole Laboratory, Na-
tional Marine Fisheries Service. NOAA, Woods Hole, MA
02543.
Manuscript accepted October 1987.
FISHERY BULLETIN; VOL. 86. NO. 1, 1988.
Numerous authors have described the basic
procedures for calculating relative fishing power
of various fleet components. Beverton and Holt
(1957) provided evidence to suggest that the dis-
tribution of logarithms of fishing power factor/
vessel tonnage ratios could be described by a nor-
mal curve while Gulland (1956) employed an
analysis of variance (ANOVA) model of log
CPUE. The properties of the ANOVA model were
further examined by Robson (1966) who extended
the techniques developed by Gulland (1956) and
formally specified the analysis of Beverton and
Holt (1957) as a two-way multiplicative ANOVA
model. Stern and Hennemuth (1975) employed
the method of Robson (1966) in their analysis of
fishing effort in the U.S. Georges Bank haddock
fishery using depth fished and vessel tonnage as
classification variables. In a previous study,
Rounsefell (1957) computed standardized log
CPUE indices to determine relative abundance of
several co-occurring species on Georges Bank ac-
cording to depth. More recently, Gavaris (1980)
and Kimura (1981) have developed modifications
of the ANOVA model to estimate annual stan-
dardized CPUE indices from time series of catch
and effort data by incorporating a year effect in
the model.
Standardized annual CPUE indices based on
91
I-ISIIKKY lUU.l.KTIN: VOL, 86, NO. 1
criteria established by Lux (1964) have been rou-
tinely used to monitor relative abundance of
three stocks of yellowtail flounder, Limanda fer-
ruginea (Storer), in the commercial fishery off the
New England coast (Fig. 1 ). Lux calculated CPUE
indices for otter trawlers ranging from <26 gross
registered tons (GRT) to 100 GRT based on trips
in which yellowtail flounder accounted for bQ^/< or
more of the total landed weight between 1942 and
1961. A fishing power coefficient was then com-
puted for each of several GRT categories as the
ratio of CPUE to a standard GRT category CPUE
for the entire timespan. A separate set of fishing
power coefficients was computed for each of the
three stocks. Lux's (1964) work improved upon an
earlier analysis of yellowtail flounder CPUE by
Royce et al. ( 1959) which was based only on rela-
tively small vessels ranging in size from 5 to 50
GRT that dominated the fishery during the
1940's.
Since 1964, numerous technological innova-
tions have drastically changed the character of
the New England fishing fleet as traditional side
trawlers have gradually given way to larger,
more efficient stern trawlers equipped with so-
phisticated electronic navigation and hydroa-
coustic devices. This gradual alteration in the
fleet characteristics over time suggests that pre-
viously documented relationships among vessel
categories may no longer be applicable to the cur-
rent fishery, and that use of nominal effort in
CPUE calculations will tend to overestimate rela-
tive abundance in the more recent years
(Westrheim and Foucher 1985). Long-term de-
clines in yellowtail flounder abundance on each of
the principal fishing grounds (Clark et al. 1984)
also indicate that current catchability coefficients
may differ from previous values. Accordingly, up-
dated fishing power coefficients are required to
adequately assess changes in effective fishing ef-
fort and CPUE which have occurred during the
past two decades. Further, to obtain annual effort
and CPUE estimates over such a broad period of
years, techniques for computing relative fishing
power should incorporate a time element in the
analysis.
In this paper we examine variation in CPUE
with respect to fishing area, depth, vessel tonnage
class, season, and year for three stocks of yellow-
tail flounder on Georges Bank, Southern New
England, and Cape Cod grounds between 1964
and 1983. Before evaluating differences in rela-
tive fishing power among vessel classes, we inves-
tigate potential interactions between tonnage
class and area and tonnage class and depth
within each year, and partition the data to mini-
mize tonnage class-area interactions. For each
stock, fishing power coefficients are examined for
68 67 66
Figure 1.— Yellowtail flounder stocks off the coast of New England (After Lux 1963).
92
O'BRIEN AND MAYO: CPUE OF YELLOWTAIL FLOUNDER
annual and seasonal interactions. A three-way
linear model, incorporating annual and seasonal
components, is employed to compute relative
fishing power coefficients over the entire 20-yr
period and estimate annual standardized CPUE
indices.
DISTRIBUTION OF CATCH AND
EFFORT
Commercial exploitation of yellowtail flounder
began in the late 1930's following the decline of
the winter flounder fishery (Royce et al. 1959).
Nominal catches^ for the three grounds combined
rapidly increased to 31,500 metric tons (t) in 1942
but subsequently declined to 5,500 t in 1954.
Landings by U.S. vessels gradually increased to a
record high of 36,900 t in 1963, but declined again
to 10,500 t in 1978 (Fig. 2). Distant water fleet
(DWF) catches were also substantial during this
2Nominal catch defined as live weight equivalent of landings,
excluding discards.
period, peaking at 20,700 t in 1969. Overall
catches from the three fishing grounds have re-
cently increased to 30,800 t in 1983, although
1984 landings declined to 15,500 t (Clark et al.
1984). The decline in catch during the 1940's was
not due to overfishing (Royce et al. 1959) but may
have been related to a warming trend in the re-
gion which affected recruitment (Sissenwine
1974). The more recent decline between 1969 and
1978, however, has been attributed to increased
fishing effort by both domestic and distant water
fleets (Brown et al. 1980).
In the early 1940's the size of vessels fishing for
yellowtail flounder varied from 5 to 75 GRT. The
predominant vessels on Southern New England
and Cape Cod grounds ranged from 26 to 50 GRT;
on Georges Bank, the dominant vessels were in
the 51-75 GRT range. By the mid-1960's larger
vessels had begun to enter the fishery, increasing
the maximum size to 215 GRT. During this period
the size range of the dominant vessels on South-
ern New England grounds and on Georges Bank
had increased to 51-72 and 73-104 GRT, respec-
55 n
O
CO
O
O
o
I
O)
c
C
<0
50-
45
'cJj
S 40
30-
Total
DWF
nVja-yy
1940
1945
1950
1955
1960 1965
Year
Figure 2. — Yellowtail flounder landings (metric tons) by United States and distant water fleet (DWF) vessels
from the combined Georges Bank, Southern New England, and Cape Cod grounds, 1940-84.
93
FISHKKY MUI.LKTIN: VOL. 86, NO 1
tively. Vessels fishing the relatively nearshore
Cape Cod grounds remained in the 34-50 GRT
range. Larger vessels continued to enter the fish-
ery during the 1970's and, by 1983, several ves-
sels were in the 311-400 GRT range. Since 1964,
vessel CRT's have been categorized by tonnage
class (TO as given in Table la.
A review of Lux's (1964) data from 1942 to 1961
and the distribution of more recent yellowtail
flounder landings from 1964 to 1983 (Figs. 3-5)
reveal that vessels of similar size have continu-
ally fished the same general areas over the past
40 years. The TC 21-24 vessels fish primarily on
Southern New England and Cape Cod grounds
(Fig. 3 1, although TC 24 vessels occasionally
enter the southwest part of Georges Bank. The
TC 25-33 vessels fish on both Georges Bank and
Southern New England grounds (Fig. 4), while
TC 41-43 vessels concentrate on Georges Bank
(Fig. 5). Although TC 41 vessels operate at times
on the eastern part of the Southern New England
grounds, the TC 42 and 43 vessels fish exclusively
on Georges Bank.
The distribution plots (Figs. 3-5) reveal a grad-
ual phaseout of smaller (TC 21-24) vessels on the
inshore Southern New England and Cape Cod
grounds and a concurrent increase in the activity
of large (TC 41-43) vessels on Georges Bank,
Southern New England, and, to a lesser extent,
on the Cape Cod grounds. In evaluating trends in
CPUE we must ask whether these changes in the
yellowtail fishery over the past 20 years (i.e., the
shift in the predominant vessel size on two of the
Table 1. — Definition of vessel tonnage classes (a) and depth
ranges and corresponding zones (b) included in analysis of vari-
ance of yellowtail flounder CPUE.
a Gross registered
tonnage
Vessel tonnage
(range)
class
5-10
21
11-15
22
16-22
23
23-33
24
34-50
25
51-72
31
73-104
32
105-150
33
151-215
41
216-310
42
311-400
43
b Depth range
Depth
(m)
zone
1-55
1
56-110
2
111-183
3
three grounds and the addition of larger vessels to
the fieet on all three grounds) affect CPUE as
calculated by the traditional method (Lux 1964).
If the same size range of vessels (5-100 GRT) had
fished for yellowtail flounder throughout the
years, a shift in the dominant vessel class would
not affect CPUE estimates since effort would be
standardized against the same class and is, there-
fore, relative. However, the maximum vessel size
has increased and the predominant TC now repre-
sents vessels larger than 100 GRT. Since landings
and effort data contributed by these larger vessels
were not incorporated into previous CPUE calcu-
lations, CPUE estimates will not necessarily rep-
resent overall fleet performance in recent years.
The following procedures, therefore, were devel-
oped to calculate new fishing power coefficients
that encompass the entire size range of vessels
currently in the fishery.
METHODS OF ANALYSIS
Catch and effort data recorded by trip were ob-
tained from Northeast Fisheries Center (NEFC)
detailed commercial landings files. Fishing effort
or days fished (df) is defined on a 24-h basis as
number of hours of actual fishing time divided by
24. Only trips landing 50% or more of yellowtail
flounder were analyzed; trips included within the
qualification level generally accounted for 70-
90% of the total yellowtail landings over the en-
tire period, except on Cape Cod grounds where
qualified trips accounted for 40-60% of the total.
Catch per day fished (CPUE) was computed for
each trip and transformed to In CPUE since pre-
liminary analysis indicated a positive correlation
of the arithmetic mean CPUE with the variance.
Use of the log transformation, however, stabilized
the variance and created a lognormal distribution
as noted by Gulland (1956) and Steel and Torrie
(1980).
Trips landing between 1964 and 1983 are clas-
sified in the data base by vessel tonnage class,
statistical area, and depth zone fished. Vessels
ranging in size from 5 to 310 GRT (Table la) and
statistical areas corresponding to the three major
stocks were selected for analysis as follows:
Georges Bank (areas 522-525), Southern New
England (area 526 and 537-539), and Cape Cod
(areas 514 and 521) (Fig. 6). Because of their spo-
radic representation, TC 21-23 vessels have been
excluded from the Georges Bank analyses and
have been combined as one category on the South-
ern New England grounds. Depth zones 1, 2, and
94
O'BRIEN AND MAYO CPUE OF YELLOWTAIL FLOUNDER
O — — — ■
■ "X^
to o
IT) O
o o o o
f\j m =r u
— LJ
o —
z a:
a: «-
O —
. . . .
^
o \r>
o o o
— nj m
• •»!*
ST^'^
CM
I
1— (
bo
aj
c
c
o
<u
CO
en
>
cS
CO
00
05
T3
c
(8
05
O
05
in
CO
05
bo
C
■3
c
(8
c
3
CO
c
o
3
95
FISHERY BULLETIN: VOL 86, NO. 1
CO
CO
lO
(M
w
(V
CO
1
«
U-)
ca
OJ
CD
(D
LlJ
be
tx
CO
C
—i
c
(_) rn
o
CO
■w
Lu cn
^M
O —
a;
cr -■
«
z cc
z cr
O LiJ
t— >-
u
a
CO
00
OS
1—1
-a
a
CD
to
o
05
tH
O
t~
05
r-l
U5
CO
o>
Z
o
oj o in o in o
z
— f\j i/l r- IT)
_ (_,
o —
Z (T
cr "-
—1 UJ
O l/i o o o
oo O Ul O
— Aj in
_,
O UJ
03 ts^
z a
O UJ
bo
S
1
OJ
-a
c
3
o
c
_o
"-^
3
'u
in
(5
D
O
96
O'BRIEN AND MAYO: CPUE OF YELLOWTAIL FLOUNDER
CO
I
CO
be
<a
c
c
<u
>
;2
CO
00
Oi
c
o
09
bo
C
s
,2
>H
a>
-a
e
3
o
t
3
Q
P
2
97
FISHERY BULLETIN: VOL. 86. NO. 1
Figure 6. — Yellowtail flounder fishing grounds defined by U.S. statistical area are as follows: Southern New
England. 526-539; Georges Bank, 522-525; Cape Cod, 514 and 521.
3 (Table lb) were also selected from six possible
zones based on the bathymetric distribution of
yellowtail flounder.
Trip data were aggi'egated at different levels of
spatial resolution to examine variability in
CPUE over the entire region and within each of
the three established stocks. Two-way ANOVAs
with interaction were performed on annual data
sets using the BMDP statistical software progi'am
P4V (Dixon 1981). Given the large number of ob-
servations in each analysis, the more rigorous
99% significance level was chosen to test the null
hypothesis (no significant differences) since
relatively small differences in mean CPUE can
produce statistically significant results. The
ANOVA was performed initially to test for differ-
ences in CPUE among all tonnage classes and
statistical areas and to determine the overall ex-
tent of tonnage class-area interactions. Second-
ary analyses were performed to examine the ef-
fect of tonnage class-area and tonnage class-depth
interactions among and within each of the stocks.
All subsequent tests for significance of tonnage
class, area, and depth main effects were per-
formed with the interaction effects absorbed in
the error sum of squares. Estimates of annual
geometric mean CPUE were obtained by combi-
nations of tonnage class, stock area, and depth
from the row and column means provided by the
P4V software (Tables 2, 3).
A standard vessel class was selected for each of
the three stocks for use in calculation of fishing
power coefficients based on the prevalence of the
vessel class in the fishery and its relative contri-
bution to the landings over the 20 years. The TC
32 category was chosen as the standard for both
Georges Bank and Southern New England stocks,
and the TC 25 class was chosen as the standard
for the Cape Cod stock. Within each stock annual
fishing power coefficients were derived for each
tonnage class relative to the standard by fitting
In CPUE to a one-way linear model using the
GLM procedure of the Statistical Analysis Sys-
tem (SAS Institute 1982) as follows:
U = CC + ^ [^jXj] + ^.
Annual deviations of the coefficients from the
20-yr mean were tested for first order autocorre-
lation using the Durbin-Watson test statistic
(Neter and Wasserman 1974). A time component
was subsequently incorporated in the linear
model to account for annual trends in the data;
seasonal effects were also included by classifying
the data according to calendar quarter. The ini-
tial year (1964) and the fourth quarter were se-
lected as reference categories. The general model
is specified as:
98
O'BRIEN AND MAYO: CPUE OF YELLOWTAIL FLOl'NDER
Table 2. — Geometric mean CPUE^ (landings per day fished, metric tons) for Georges Bank, Southern New England, and Cape Cod
yellowtail flounder trips by vessel tonnage class, 1964-83.
Vessel tonnage
class
Year
Vessel tonnage class
Year
21
22
23
24
25
31
32
33
41
42
21
22
23
24
25
31
32
33
41 42
Georges Bank
1964
—
—
—
3.65
4.17
4.88
5.04
4.58
4.52
2.14
1974
—
1.08
1 62
231
2.14
2.07
2.07
2.64
1.30 —
1965
—
—
—
2.41
3.06
3.62
3.55
3.39
4.97
—
1975
—
0.57
1.57
1.13
1.29
1.33
1.55
1.55
1.72 —
1966
—
—
1.18
1.25
2.27
276
2.78
2.76
2.71
—
1976
—
0.72
0.90
0.53
0.86
1.13
1 32
1.69
2.06 —
1967
—
—
—
—
2.20
2.74
2.96
278
265
—
1977
—
0.93
1.73
1.44
1.77
1.61
1.82
1.98
1.81 —
1968
—
—
—
3.88
2.99
3.47
3.55
384
3.53
2.66
1978
—
—
2.72
1.15
1.58
1.52
2.04
2.26
282 —
1969
—
—
—
3.28
2.82
3.06
3.30
3.02
2.81
3.88
1979
0.79
1.74
2.28
1.59
2.17
2.32
2.93
2.88
3.64 —
1970
—
—
—
3.74
2.40
2.88
308
2.81
2.70
—
1980
—
1.02
246
1.91
2.54
2.68
2.81
3.07
3.00 —
1971
—
—
—
1.66
2.07
2.70
252
2.24
2.35
—
1981
—
1.88
2.50
1.43
2.58
2.76
2.61
3.28
2.23 —
1972
—
—
—
—
1.96
224
239
2.29
1.96
—
1982
249
1.96
2.42
2.29
3 10
3.20
3.72
3.98
3.97 —
1973
—
—
—
1.38
2.64
2.61
2.81
3.05
2.76
3.20
1983
8.81
1.74
2.51
3.35
3.43
3.27
3.65
3.57
3.79 —
1974
—
2.52
—
—
2.79
2.13
2.23
2.19
1.30
3.42
pane Cnd
1975
—
—
1.93
1.91
1.64
1.93
2.04
1.72
3.19
*»
^i^l^f u »
1976
1.35
1.69
1.77
1.84
2.16
2.06
1964
—
1.65
1.38
1.94
2.64
2.85
2.76
—
— —
1977
1.17
1.66
1.77
226
1.81
1965
—
1.45
1.20
1.59
2.16
3.56
2.12
2.07
— —
1978
1.67
1.50
1.72
2.03
282
1966
1.13
1.07
1.23
1.48
2.55
3.64
1.76
—
— —
1979
2.69
1.31
1.80
2.41
2.37
2.82
3.64
1967
—
0.78
0.90
1.56
2.46
3.21
1.95
2.62
— —
1980
0.30
1.09
0.90
2.88
2.62
3.27
3.00
1968
—
1.16
0.99
1.89
2.65
4.11
2.42
—
1.13 —
1981
_
1.12
1.24
1.98
2.48
2.61
223
1969
—
1.66
1.33
202
2.77
3.64
2.47
—
— —
1982
1.88
0.48
2.02
1.71
2.10
2.53
397
2.07
1970
—
1.03
1.05
2.34
2.33
3.64
26.96
234.18
2.91 —
1983
—
—
—
—
0.79
2.46
1.97
2.21
3.79
—
1971
1972
1.71
1.17
1.63
1.89
1.16
2.43
2.02
2.04
1.85
2.24
2.02
1.90
1.46
2.45
1.79
3.36 —
1.93 —
Southern New England
1973
1.02
1.32
1.06
1.87
1.99
1.90
1.91
1 63
284 —
1964
234
1.94
6.15
4.88
4.66
4.44
5.04
4.59
4.09
—
1974
0.83
1.15
0.90
1.59
2.08
1.74
1.82
1.87
1.37 —
1965
1.77
1.55
3.87
3.62
3.12
3.56
3.81
4.05
1.50
—
1975
0.77
1.13
1.23
1.42
1.92
1.45
1.64
1.09
1.19 —
1966
1.21
306
3.19
2.93
2.52
261
2.56
2.35
3.03
—
1976
0.34
1.31
1.38
1.69
1.93
1.42
1.54
1.88
2.15 —
1967
1.80
1.14
294
3.38
323
2.67
2.66
2.31
1.80
—
1977
0.28
0.86
1.25
1.50
2.03
1.45
1.42
2.19
2.57 —
1968
2.24
4.15
4.69
4.42
4.02
3.41
3.62
3.63
3.53
—
1978
—
0.84
1.59
1.90
2.11
1.96
1.53
1.92
5.99 —
1969
265
372
5.17
3.81
4 10
3.39
3.43
4.01
2.81
—
1979
0.98
0.91
1.42
1.83
2.10
2.24
1.74
2.23
3.12 —
1970
7.74
3.28
4.76
3.77
3.96
3.08
3.64
2.87
2.70
—
1980
0.26
0.58
1.30
1.73
2.18
2.24
2.07
2.19
3.34 —
1971
11.72
1.60
4.06
3.14
3.32
2.85
2.87
3.16
2.35
—
1981
0.70
0.76
1.17
1.69
2.24
2.40
1.22
1.85
2.99 —
1972
—
1.96
3.19
2.62
3.30
3.01
2.96
3.04
1.96
—
1982
0.54
0.71
1.22
1.79
2.02
2.41
1.57
1.94
2.37 —
1973
—
1.01
1.05
2.26
2.66
2.29
2.22
2.12
2.76
—
1983
0.69
1.51
1.23
1.96
1.91
1.29
1.53
1.46
0.83 —
^Calculated as exp
1 nX In
' landings I
effort I
20nly one trip by tonnage classes 32 and 33 in 1970 on Cape Cod grounds.
u
a + 2! lP#(/] + e
where U = \n CPUE,
a = intercept estimate,
P,Y = model parameter estimates in loga-
rithmic units for category 7 for ton-
nage class, season, and year,
X,j = dummy variable for tonnage class,
season, and year ( = 1 when category
j occurs; = 0 otherwise), and
e — error term.
All tests for significance of main effects were
based on the above model without interaction.
Separate ANOVAs were also performed to exam-
ine first order interactions. Parameter estimates
obtained from the model without interaction were
retransformed following methods described by
Bradu and Mundlak (1970) to derive unbiased
fishing power, seasonal, and annual coefficients.
Annual coefficients corresponding to the 1965-83
period were multiplied by the reference year
CPUE to compute annual standardized CPUE es-
timates.
RESULTS
Smaller vessels (TC 21-24) generally exhibited
the lowest CPUE indices in all three areas, al-
though TC 21-23 vessels were not represented on
Georges Bank (Table 2). Catch rates of medium
vessels (TC 25; 31-33) were similar to each other,
and were generally greater than those for TC 21-
24 vessels. Mean CPUE indices for the largest
vessels (TC 41 and 42) were more variable, but
generally were equal to or greater than those
99
FISHERY BULLETIN: VOL. 86, NO. 1
Table 3. — Geometric nriean CPUE^ (landings per day fished, metnc tons) for Georges Bank,
Southern New England, and Cape Cod yellowtail flounder tnps by depth zone, 1964-83.
Georges Bank
So
ut
herr
New England
pth zone
Cape Cod
Depth zone
De
Depth zone
Year
1
2
3
Combined 1
2
3
Combined
1
2
3 Combined
1964
3.92
4.91
2.74
4.80
5.05
4.96
4.71
503
2.13
2.36
—
2.25
1965
3.76
3.48
—
3.55
3.61
3.88
2.28
366
1.61
1.80
2.36
1.71
1966
3.02
2.62
1.29
2.71
2.71
2.66
0.71
2.71
1.99
2.13
1.70
2.07
1967
3.19
2.69
2.81
2.79
2.97
2.77
1.14
2.91
1.52
2.16
—
1.80
1968
3.38
3.51
—
3.49
3.70
3.97
3.24
3.76
2.15
2.31
—
2.24
1969
3.36
3.11
—
3.16
3.62
3.93
1.58
3.66
2.16
2.66
—
2.27
1970
2.81
2.99
1.00
2.93
3.44
3.68
1.62
3.49
1.10
2.24
234.18
1.40
1971
2.20
2.30
2.75
2.28
3.26
2.96
—
3.14
1.03
1.94
—
1.46
1972
2.28
2.37
1.34
2.35
3.28
3.25
2.74
3.27
1.56
1.88
—
1.82
1973
2.30
3.06
2.85
2.91
2.41
2.55
4.40
2.47
1.75
1.83
2.27
1.81
1974
2.10
2.32
2.90
2.30
2.31
2.02
—
2.19
1.86
1.78
—
1.82
1975
1.85
1.99
1.29
1.98
1.59
1.63
—
1.60
1.53
1.38
1.32
1.46
1976
1.75
2.06
—
2.03
1.34
1.56
—
1.43
1.54
1.88
2.93
1.70
1977
1.92
2.15
1.67
2.14
1.92
1.97
—
1.94
1,70
1.89
1.27
1.77
1978
1.78
2.09
1.98
2.07
2.45
2.17
—
2.32
1.90
1.71
3.49
1.86
1979
2.47
2.79
2.04
2.74
2.88
2.39
—
2.76
2.03
2.03
2.47
2.03
1980
2.58
3.50
1.53
3.37
2.71
3.51
1.13
3.04
2.06
2.31
2.38
2.16
1981
2.26
2.88
1.54
2.76
2.46
2.78
—
2.58
1.74
1.96
5.67
1.79
1982
2.34
2.66
2.53
2.62
3.20
3.54
4.53
3.31
1.81
1.44
0.91
1.67
1983
1.98
2.28
3.04
2.26
3.08
3.13
5.91
3.10
1.70
1.41
—
1.62
1 Calculated as exp
1
n-:i In (
andings
effort
)]
20nly one tnp in depth zone 3 in 1970 on Cape Cod grounds.
corresponding to medium and small vessels, par-
ticularly in the later years (Table 2). The initial
ANOVAs performed over all statistical areas re-
vealed highly significant iP < 0.01) differences in
CPUE for tonnage class and area main effects in
each of the 20 years (Table 4). The interaction of
tonnage class and area was also highly signifi-
cant in all years, suggesting that relative fishing
power of the individual vessel classes varies ac-
cording to area. ANOVA results for the compari-
son of CPUE among stocks were highly signifi-
cant for area main effects in 19 out of 20 years,
and the tonnage class-stock area interaction term
was highly significant in all years (Table 4).
Grouping the data according to stock tended to
reduce the amount of significant tonnage class-
area interaction within each stock, although dif-
ferences among tonnage classes remained highly
significant.
On Georges Bank the differences in CPUE were
highly significant for statistical area and tonnage
class main effects in 80 and lOO'/r of the years,
respectively, while the tonnage class-area inter-
action was highly significant in only 40% of the
years. Differences due to area on Southern New
England grounds were highly significant in all
years except 1978, and differences due to tonnage
class were highly significant in all years. The in-
teraction term was highly significant in 70% of
the years. Differences due to area on the Cape Cod
grounds were highly significant in all years ex-
cept 1975, and differences due to tonnage class
were highly significant for all years. The interac-
tion was highly significant in only 35% of the
years (Table 4).
Differences in CPUE by depth zone were gener-
ally not significant. Depth main effects yielded
Table 4. — Frequency with which highly significant (P < 0.01) re-
sults were obtained from analysis of vanance (ANOVA) tests of
yellowtail flounder annual CPUE data. (Total number of years
tested = 20.) N/A = Not applicable (tests not performed).
Main effects
Area
Tonnage class
Depth
All areas
2020
20/20
N/A
Among stocks
19 20
N/A
N'A
Within stocks
Georges Bank
16 20
20/20
10 20
So. New England
1 9, 20
20/20
7/20
Cape Cod
19/20
20/20
3/20
Interactions
Tonnage class Tonnage class
■ area
■ depth
All areas
20/20
N/A
Among stocks
20/20
N/A
Within stocks
Georges Bank
8/20
4/20
So. New England
14/20
2/20
Cape Cod
7/20
1/20
100
O'BRIEN AND MAYO: CPUE OF YELLOWTAIL FLOUNDER
highly significant differences in only 50, 35, and
IdVc of the years for Georges Bank, Southern New
England, and Cape Cod grounds, respectively
(Table 4), while the tonnage class-depth interac-
tion was highly significant in no more than 20%
of the years on each of the three grounds.
Interaction between tonnage class and statis-
tical area throughout the region was highly sig-
nificant in all cases. Further analyses yielded
highly significant differences in CPUE among the
three stocks and highly significant tonnage class-
stock interactions which suggests that relative
fishing power among vessel classes is not consis-
tent from stock to stock, implying a need for com-
puting a separate set of fishing power coefficients
for each stock. Within each stock, differences in
CPUE among statistical areas and tonnage
classes were also highly significant in most cases,
although the tonnage class-area interaction was
not.
Standardized CPUE
Annual fishing power coefficients, obtained by
retransforming linear model parameter esti-
mates for each tonnage class, are presented in
Table 5 by stock. Cursory examination of the coef-
ficients reveals distinct trends throughout the 20-
yr series. On Georges Bank and Southern New
England grounds, fishing power coefficients for
the smaller vessels (TC 23-25) relative to the
standard declined over time, whereas coefficients
for the larger vessels increased over time. On
Cape Cod grounds the coefficients increased for
the smaller vessels (TC 23 and 24) although
trends were less pronounced. These trends are
illustrated graphically by plotting annual devia-
tions from the 20-yr average in Figures 7-9. A
Durbin-Watson test for first order autocorrela-
tion of the annual deviations (Neter and Wasser-
man 1974) was significant for most tonnage
classes within each of the stocks, suggesting the
presence of a substantial tonnage class-time in-
teraction.
The three-way linear model, modified to in-
clude interaction terms, also revealed highly sig-
nificant tonnage class-year as well as tonnage
class-season and year-season interactions within
each of the three stocks (Table 6). When interac-
tions are significant, they can be examined in
detail or absorbed in the error term when testing
for main effects. Since tonnage class effects have
already been examined on an annual basis, the
interaction terms were excluded from the three-
way model used to obtain parameter estimates for
tonnage class, season, and year. The model is
specified as follows:
f/ = a + X [Pi, Xy + ^2j X2j + Paj Xsj] +
where ^y, P2/, P3; = model parameter estimates
in logarithmic units for
category j for tonnage
class, season, and year,
Xij , X2J, X2,j
dummy variables for ton-
nage class, season, and
year ( = 1 when category 7
occurs; = 0 otherwise).
ANOVA results obtained without interaction
are presented in Table 6 for each of the three
stocks. For Georges Bank and Southern New
England stocks, year accounts for the greatest
reduction in error sums of squares; on Cape Cod
grounds tonnage class accounts for the greatest
overall reduction.
Coefficients for tonnage class, year, and season,
derived from model parameter estimates for the
combined 1964-83 period are presented in
Table 7. Tonnage class coefficients for Georges
Bank and Southern New England are relatively
homogeneous, as compared with those obtained
for Cape Cod grounds, owing to the narrower
range of vessel tonnage classes which have con-
sistently exploited these fisheries. Seasonal coef-
ficients exhibit the same pattern on Georges
Bank and Southern New England with the
highest catch rates occurring during the third
quarter; on Cape Cod grounds the highest catch
rates occur during the second quarter. Trends in
annual coefficients are similar on all three
grounds. Standardized CPUE indices based on
the annual coefficients are illustrated in Figure
10, and traditional indices based on the methods
of Lux (1964), as given by Clark et al. (1984), are
also presented for comparative purposes.
Although each series indicates similar trends,
CPUE indices obtained from the linear model for
Georges Bank and Southern New England have
remained slightly higher than the traditional in-
dices since the early 1970's. Prior to this, the tra-
ditional CPUE indices were greater than the re-
vised indices. On Cape Cod grounds, differences
between the two series are considerably greater,
particularly in the early years.
101
Georges Bank
1-l
TC24
0.5-
•
■
-0.5-
•
c ■^^
o
65 67 89 71 73 75
77 79 81 83
4-»
- ^1
TC25
> 0.5-
•
.
£J 0-
*
, '
— -0.5-
(0
Zj ■' 1 I i I I I 1
J- 65 87 89 71 73 75
1 1 I 1
77 79 81 83
C
<
1-
TC31
0.5-
.
-0.5-
* '
1 1 1 1 1 I
65 67 89 71 73 75
1 1 1 1
77 79 81 83
Table 5.
KIS51EKY BULLETIN: VOL 86, NO 1
-Annual fishing power coeflicients calculated by vessel tonnage
Georges Bank, Southern New England,
Vessel tonnage class^
Year 21 22
23 24
25
31 32
33
41
42
Georges Bank
1964
073
083
0.97
0.91
0,90
043
1965
068
0.86
1.02
095
1,40
—
1966
0.45
0.81
0.99
099
097
—
1967
—
0.74
0.93
094
0,89
—
1968
1.09
084
098
1 08
088
0,75
1969
0.99
0.85
0.93
0.92
0,95
1.17
1970
1.21
0.78
0.94
0.91
094
0.92
1971
0.66
0.82
1.07
089
1,12
—
1972
—
082
0.94
096
0,93
—
1973
not
0.49
0.94
0.93 1.00
1 09
1,09
1.14
1974
calculated
—
1 25
096
098
097
1.53
1975
1.00
0.99
0.85
1.06
1 04
1 66
1976
0.73
0.92
0.96
1.17
1,01
—
1977
—
0.66
0.94
1.28
1,67
—
1978
—
0.97
0.87
1,18
1,54
—
1979
0.55
0.76
1.02
1.19
1.35
—
1980
0.42
0.35
1.10
1.25
1.43
—
1981
0.45
0.50
0.80
1.05
1.16
—
1982
0.23
0.96
0.82
1.21
1.41
099
1983
—
0.40
1.25
1 12
1.26
—
Southern New England
1964
- 20.95
— 0.97
0.92
0.88
0.91
0.81
1965
— 0.86
— 0.95
0.82
0.93
1 06
0.39
1966
— 1.12
— 1.14
0.98
1.02
092
1.18
1967
— 0.94
— 1.27
1.22
1.00
0.87
068
1968
— 1.21
— 1.23
1.11
0.94
1,01
098
1969
— 1.32
— 1.11
1.19
0.99
1.17
0.82
1970
— 1.23
— 1.04
1.09
085
0.79
0.74
1971
— 1.06
— 1.10
1 16
0.99
1.10
082
1972
— 0.99
— 088
1.12
1.01
1,03
0.66
1973
— 0.46
— 1.01
1.20
1 .03 1 .00
0,95
1.24
'Standard vessel class on Georges Bank and Southern New England
rounds = 32. Standard vessel class on Cape Cod grounds = 25.
1-
TC33
0.5-
.
0.5-
-1-
[ I 1 1 1 1 1 1 I 1
0.5
0
66 67 69 71 73 75 77 79 81 83
TO 41
~~i I I I I I — I — I — I — I
65 67 69 71 73 75 77 79 81 83
1
0.5
TC42
~i I I I — I — I — I — I — I — I
66 67 69 71 73 75 77 79 81 83
Figure 7. — Deviations in annual fishing power from the
1964-83 20-yr mean for major vessel tonnage classes fish-
ing on Georges Bank.
Year
102
O'BRIEN AND MAYO: CPUE OF YELLOWTAIL FLOUNDER
class relative to a standard class vessel fishing for yellowtail flounder on
and Cape Cod grounds, 1964-83.
Vessel
tonnage class^
Year
21
22
23
24
25
31
32
33
41 42
Southern New England
1974
—
0.67
—
1.11
1 03
1 00
1.27
0.62
1975
—
072
—
0.73
0.83
085
1.00
1.11
1976
—
0.66
—
0.40
0.66
086
1.28
1.56
1977
—
087
—
0.79
0.97
088
1.09
0.99
1978
—
1 33
—
0.56
0.78
0.75
1.11
1.38
1979
—
069
—
0.54
0.74
079
098
1.24
1980
—
082
—
0.68
0.90
0.95
1.09
1.07
1981
—
0.92
—
0.55
099
1.06
1.26
0.86
1982
—
0.64
—
0.62
0.84
0.86
1.07
1.07
1983
—
0.68
—
092
0.94
0.89
0.98
1.04
Cape Cod
1964
—
062
0.52
0.73
1.08
1.04
—
—
1965
—
067
0.55
0.73
1.65
0.98
0.96
—
1966
044
042
048
0.58
1.42
0.69
—
—
1967
—
032
0.36
0.64
1.30
0.79
1.07
—
1968
—
044
0.37
0.71
1.55
0.91
0.99
0.43
1969
—
060
048
0.73
1.31
0.89
—
—
1970
—
0.44
0.45
1.01
1.56
(3)
(3)
1.25
1971
—
0.57
0.93
1.19
1.10
0.93
1.20
1.65
1972
0.93
088
063
1 09
1.09
0.79
0.97
1.05
1973
0.51
066
053
0.94
1.00
0.95
0.96
0.82
1.43
1974
0.40
0.55
0.43
0.76
083
0.87
0.90
0.66
1975
0.40
059
0.64
0.74
0.75
085
0.57
0.62
1976
0 18
068
072
088
0.73
0.80
0.97
1.12
1977
0.14
042
061
0.74
0.71
0.70
1.08
1.26
1978
—
040
075
0.90
0.93
0.73
0.91
2.84
1979
0.47
043
0.68
0.87
1.06
0.83
1.06
1.48
1980
0 12
027
0.60
0.80
1.03
095
1 01
1.53
1981
0.31
0.34
052
0.75
1.07
0.54
0.83
1.34
1982
0.27
035
060
088
1.19
0.77
0.96
1.17
1983
0.36
0.79
0.64
1.02
0.68
0.80
0.77
0.43
2Vessel classes 21, 22, and 23 combined on Southern New England
grounds
3|nsuf1icient data for these categories.
Southern New England
c
o
1-
05-
TC 21-22-23
0
0.5-
-1-
1 1 ; 1 \ 1 1 1 ^ 1
0.5-
0
-0.5
65 67 69 71 73 75 77 79 81 83
TC31
— I 1 1 1 1 1 \ 1 1 I
65 67 69 71 73 75 77 79 81 83
TC33
<0
D
C
C
<
Figure 8. — Deviations in annual fishing power from the
1964-83 20-yr mean for major vessel tonnage classes fish-
ing on Southern New England grounds.
1
0.5 H
0
-0.5-
-I — I — I — I — I — I — I — I r
65 67 69 71 73 75 77 79 81 83
TC25
— I 1 1 1 1 1 1 I I I
65 67 69 71 73 75 77 79 81 83
—I — 1 1 1 1 1 1 1 1 I
66 07 69 71 73 76 77 79 81 83
1-
0.5-
TC41
0.5-
-1-
— 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1
65 67 69 71 73 75 77 79 81 83
Year
103
FISHERY BULLETIN VOL 86. NO 1
Cape Cod
1-
0 5
0
-0 5-
TC23
— I — I — I — I — 1 — I — I — I — I — I
65 67 69 71 73 75 77 79 81 83
1
0.5
0
-0.5
-1
TC32
—1 — I — I — I — I — I — I — I — I — I
65 67 69 71 73 75 77 79 81 83
Figure 9.— Deviations in annual fishing power from the
1964-83 SO-yr mean for major vessel tonnage classes fish-
ing on Cape Cod grounds.
>
Q
<o
C
C
<
1
0.5-
0
-0 5
1-1
0.5
0
-0.5-
-1
TO 24 '
0.5-
—\ 1 1 1 1 1 1 1 1 1
65 67 69 71 73 75 77 79 81 83
0
-0.5
-1
TO 31
—\ — I — I — I — I — I — I — I — I — I
65 67 69 71 73 75 77 79 81 83
1
0.5
0
-0.5
-1
Year
TC33
— I — I — I — I — I — I — I — I — I — I
65 67 69 71 73 75 77 79 81 83
•TC41
— I \ 1 1 1 1 1 1 1 I
65 67 69 71 73 75 77 79 81 83
Georges Bank
I ' I ' I ' I ...
64 66 68 70 72 74 76 78 80 82 84
Year
>•
CD
Q
0}
a
w
c
,o
Southern New England
n ' I •—] I |— ' I I I — ■— I — I — I 1 I I — I
64 66 68 70 72 74 76 78 80 82 84
Year
4-1
■D
0)
JO.
(f)
IT
>~
CD
Q
a3
Q.
CO
c
,o
Cape Cod
I ' I ' I ' I ' I ' I
64 66 68 70 72 74 76 78 80 82 84
Year
Traditional CPUE index
Linear Model CPUE Index
Figure 10. — Trends in annual yellowtail flounder CPUE (metric tons per day fished) calculated with
traditional CPUE based on Lux (1964) and annual coefficients (linear model CPUE).
104
O'BRIEN AND MAYO; CPUE OF YELLOWTAIL FLOUNDER
Table 6. — ANOVA results obtained from a three-way linear model
incorporating year, quarter, and tonnage class for Georges Bank,
Southern New England, and Cape Cod stocks of yellowtail floun-
der.
Table 7. — Tonnage class, quarter, and year coefficients derived
from a three-way linear model for the 1964-83 period for yellowtail
flounder stocks on Georges Bank, Southern New England, and
Cape Cod grounds.
Sum of
Mean
F
Georges
Southern
Source
df
squares
Georges
square
Bank
value
P
Tonnage
Bank
New England
Cape Cod
With interaction
class
Model
211
1.432 92
6.79
29.53
<0.01
21
0.92
0.32
Year
19
908.42
47.81
207.88
<0.01
22
0.92
0.50
Qtr
3
107.36
35.79
155.59
<0.01
23
0.92
0.57
TC
9
84.24
9.36
40.70
<0.01
24
057
0.93
0.80
Yr'Tc
103
116.81
1.13
4.93
<0.01
25
0.80
0.96
1.00
Yr*Qtr
57
202.29
3.55
15.43
<0.01
31
0 96
0.92
1.03
Qtr'Tc
20
13.80
0.69
3.00
<0.01
32
1.00
1.00
0.84
Error
18,437
4.171.68
023
33
1.08
1.09
1.00
Total
18,648
5.604.60
030
41
1.13
1.05
1.24
Without inte
iraction
31
1.100.02
35.48
147.83
<0.01
42
1.19
—
—
Model
Quarter
Year
19
908.42
47.81
199.21
<'0.01
Qtr
3
107.36
35.79
149.11
<0.01
1
0.96
1.00
1.07
TC
9
84.24
936
39.00
<0.01
2
0.89
1.06
1.15
Error
18,617
4.504.58
0.24
3
4
1.09
1.00
1.18
1.00
0.90
1.00
Total
18,648
5 604 60
n .^n
Southern New England
Year
With interaction
1964
1.00
1.00
1.00
Model
217
2,85995
13.18
33.79
<0.01
1965
0.73
0.72
0.84
Year
19
2,085.95
109.79
281.50
<0.01
1966
0.55
0.54
0.84
Qtr
3
111.50
37.17
95.30
<0.01
1967
0.54
0.60
0.81
TC
6
59.80
9.97
25.56
<0.01
1968
0.68
0.79
0.93
Yr*Tc
114
257.12
2.26
5.78
<0.01
1969
0.61
0.77
1.06
Yr-Qtr
57
304.89
535
13.72
<0.01
1970
0.58
0.75
1.00
Qtr-Tc
18
40.69
2.26
5.80
<0.01
1971
0.49
0.63
0.95
Error
26,879
10.612.21
0.39
1972
0.45
0.65
0.88
Total
27,096
13,472.16
0.50
1973
1974
0.55
0.42
0.49
0.45
0.84
0.77
Without interaction
1975
0.36
0.30
0.72
Model
28
2.257.25
80.62
196.62
<0.01
1976
0.37
0.25
0.76
Year
19
2,085.95
109.79
267.77
<0.01
1977
0.37
0.36
0.74
Qtr
3
111.50
37.17
90.66
<0.01
1978
0.34
0.39
0.83
TC
6
59.80
9.97
24.31
<0.01
1979
0.48
0.52
0.85
Error
27,068
11.214.91
0.41
1980
0.55
0.56
0.78
Total
27,096
13.472,16
0.50
1981
0.46
0.54
0.80
Cape Cod
1982
1983
0.44
0.40
0.71
0.70
0.77
0.78
With interaction
Model
247
2.438.73
987
2468
<:0.01
Year
19
166.49
8.76
21.91
<0.01
Qtr
TC
3
9
174.44
1,331.92
58.15
147.99
145.37
369.98
<0.01
<0.01
DISCUSSION
Yr'Tc
135
526.18
3.90
9.74
<0.01
Yr'Qtr
57
167.62
2.94
7.35
<0.01
The analytical approach adopted
in this paper
Qtr'Tc
24
72.09
3.00
7.51
<0.01
is based on _ _
the hypothesis that CPUE of yellow-
Error
19,097
7,731.24
0.40
tail flounder differs among the various tonnage
Total
19.344
10.169.97
0.53
classes of vessels and ge _ _
sographic regions associ-
Without interaction
Model 31
1,672.84
53.96
122 64
<0.01
ated with the fishery. In all of the analyses, the
Year
19
166.49
876
19.92
<0.01
null hypothesis (i.e., no
1 significant differences)
Qtr
TC
3
9
174.44
1,331.92
58.15
147.99
132.15
336.34
<0.01
<0.01
was rejected only when the probabil
iity of obtain-
Error
19.313
8.497.13
0.44
ing a greater F statistic
was <0.01.
Even at this
Total
19.344
10.169.97
0.53
probability
level, statistically signi
ficant results
were often obtained when difi'erences among vari-
able levels appeared to be minimal due primarily
to the large number of observations included in
most analyses.
The initial series of ANOVAs, based on pooled
CPUE data from all statistical areas encompass-
105
FISHEKY BULLETIN; VOL. 86, NO. 1
ing the three stocks, provided sufficient evidence
to reject the null hypothesis. In each of the 20
years analyzed, the main effects of tonnage class
and statistical area represented highly signifi-
cant (P < 0.01) sources of variation. The tonnage
class-area interaction term was also highly sig-
nificant in all cases, implying that vessels of var-
ious tonnage classes exhibit different CPUE
trends relative to each other in different areas.
The initial results established the basis for fur-
ther investigations. In subsequent analyses, the
data were grouped to test the null hypothesis that
no significant differences in CPUE existed among
the three traditionally accepted stock definitions
(Georges Bank, Southern New England, and
Cape Cod). The highly significant differences ob-
tained in 19 out of 20 years indicate that catch
rates differ among the three stocks. The resulting
highly significant tonnage class-stock area inter-
action term obtained from the ANOVAs in all
years suggests that standardization of CPUE
among tonnage classes should be performed sepa-
rately for each stock.
Analysis of variance within each stock provided
the final basis for performing the standardized
CPUE calculations. In these tests, the rejection of
the null hypothesis for tonnage class main effects
in all years for each stock suggests that separate
fishing power coefficients must be calculated for
each tonnage class even though the coefficients
are similar in many cases. The ANOVA results
also indicated that differences in CPUE among
statistical areas within each stock were highly
significant in SO^f or more of the years implying
that, within each stock region, yellowtail abun-
dance is not homogeneous. This is not surprising
since yellowtail flounder are prevalent only on
certain grounds within each geographic region.
Further analyses of the data by depth indicated
no overall significant differences in CPUE be-
tween the two primary depth zones (1-55 m and
56-110 m) where yellowtail flounder are consis-
tently caught.
The infrequent number of significant interac-
tions on Georges Bank and Cape Cod grounds
relative to Southern New England (Table 4) sug-
gests a greater independence of the tonnage class
and area main effects with respect to CPUE, i.e.
both large and small vessels exhibited relatively
similar changes in mean CPUE among statistical
areas within each stock. In choosing data sets for
computing fishing power coefficients we sought to
minimize the amount of interaction among the
vessel tonnage classes and geographic areas in-
volved. This criterion was met to a greater extent
for the Cape Cod and Georges Bank stocks than
for the Southern New England stocks. It appears
that yellowtail flounder inhabiting this region
are subject to a more complex set of interactions
perhaps due to temperature and bottom type. We
decided, however, to accept the results for each of
the three stocks and proceed with the calculations
of fishing power coefficients.
Annual fishing power coefficients computed for
each vessel tonnage class fishing on Georges
Bank, Southern New England, and Cape Cod
grounds provided a basis for examining the con-
sistency in relative fishing power of individual
tonnage classes over time. Annual deviations for
Georges Bank and Southern New England
grounds indicated a gradual change in relative
fishing power of most tonnage classes between
1964 and 1983, and tests for autocorrelation of
residuals indicated significant time effects. On
these grounds, larger vessels exhibited higher
catch rates relative to the standard in the later
years as compared with the earlier years. Since
many of the larger vessels have been replaced in
recent years by newer vessels which are, pre-
sumably, equipped with more sophisticated elec-
tronics, any attempt to relate CPUE to stock
abundance must account for such technological
advances.
Similarly, changes in seasonal availability are
often great enough to mask interannual variation
in stock abundance. Thus, the presence of signifi-
cant tonnage class-season interactions may be ex-
plained by the ability of certain vessel classes to
effectively target seasonal concentrations. Since
peak spawning of yellowtail flounder occurs dur-
ing late spring (Lux 1964), the presence of high
seasonal coefficients during the second and third
quarters is not surprising.
By specifying the model to include tonnage
class, annual, and seasonal components, we have
attempted to account for technological and sea-
sonal availability factors which interact with
temporal changes in abundance. Although other
factors could be incorporated in the model to ac-
count for a larger portion of the variation in
CPUE, analyses of historical commercial fishing
operations of this type are often limited to those
attributes which can be directly linked to land-
ings records (Kimura 1981; Westrheim and
Foucher 1985). An alternate approach adopted by
Stern and Hennemuth (1975) involved the use of
a study fleet of selected vessels whose characteris-
tics and fishing practices were closely monitored.
106
O'BRIEN AND MAYO CPUE OF YELLOWTAIL FLOUNDER
In our study, factors were selected for inclusion
in the model based on prior knowledge of fleet
characteristics and seasonal and spatial distribu-
tion patterns of the species. Despite this, the
three attributes incorporated in the final model
accounted for 15-25% of the total variation in
CPUE, depending on the stock. Undoubtedly,
other factors such as experience of the captain,
net design and rigging, and variation in local fish
abundance contribute substantially to overall
variation in catch rates.
Differences between the annual CPUE esti-
mates based on Lux's original fishing power coef-
ficients and the recalculated indices occur in
many cases because of shifts in the vessel compo-
sition of the fleet over the past 20 years. The in-
clusion of larger vessels in the more recent years,
particularly on Georges Bank and Southern New
England grounds, may account for the consis-
tently higher CPUE estimates obtained for these
areas since the mid-1970's. On Cape Cod grounds,
CPUE estimates differ substantially prior to this
time. Lux (1964) has stated that a relatively low
proportion of the landings from this area were
used in his CPUE computations and, conse-
quently, the indices were not considered to be as
valid a measure of relative abundance as those
obtained for Georges Bank and Southern New
England. Our analyses for Cape Cod grounds,
based on data for the period since 1964, are sub-
ject to the same concerns since a large proportion
of the yellowtail flounder landings continues to be
taken incidentally.
Although the revised standardized CPUE esti-
mates presented in this paper are based on a dif-
ferent standardization technique, trends are gen-
erally similar to those obtained previously. The
revised procedure, however, accounts for seasonal
and technological influences and insures com-
plete representation of all vessel classes engaged
in the yellowtail fishery.
ACKNOWLEDGMENTS
We wish to sincerely thank Stephen H. Clark
for his advice throughout this study, and for crit-
ically reviewing the manuscript. Michael J. Fo-
garty reviewed the final draft and advised on
statistical procedures. We are also grateful for the
suggestions provided by two anonymous referees.
LITERATURE CITED
Beverton, R J H . AND S J Holt
1957. On the dynamics of exploited fish popula-
tions. Fish. Invest., Lond., Ser. 2, 19:1-533.
Bradu, D. andY Mundlak
1970. Estimation in lognormal linear models. J. Am.
Stat. Assoc. 65(329):198-211.
Brown. B E . M P Slssenwine. and M M McBride.
1980. Implication of yellowtail flounder stock assessment
information for management strategies. U.S. Dep.
Commer., NMFS, NEFC, Woods Hole Lab. Ref Doc. No.
80-21, 12 p.
Clark, S. H., M. M. McBride, and B Wells.
1984. Yellowtail flounder assessment update - 1984.
U.S. Dep. Commer., NMFS, NEFC, Woods Hole Lab. Ref.
Doc. No. 84-39, 30 p,
Dixon, W J (editor),
1981. BMDP statistical software, Univ. Calif. Press,
Berkeley, p. 388-412.
Garrod. D J
1964. Effective fishing effort and the catchability coeffi-
cient q. Rapp. Cons. int. Explor. Mer 155, No. 14, p. 66-
70.
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.
Gulland. J A
1956. On the fishing effort in English demersal fish-
eries. Fish. Invest., Lond., Ser. 2, 20(51:1-41.
1964. Catch per unit effort as a measure of abundance.
Rapp. Cons. int. Explor. Mer 155, No. 1, p. 8-14.
KiMURA. D K
1981. Standardized measures of relative abundance based
on modelling log (c. p. u.e.), and their application to Pacific
ocean perch (Sebastes alutus). J. Cons. int. Explor. Mer
39:211-218.
Lux. F E
1963. Identification of New England yellowtail flounder
groups. Fish. Bull, U.S. 63:1-10.
1964. Landings, fishing effort, and apparent abundance
in the yellowtail flounder fishery. Int. Comm. North-
west Atl. Fish. Res. Bull. No. 1, p. 5-21.
NETER. J . AND W WaSSERMAN
1974. Applied Linear Statistical Models. Richard W.
Irwin, Inc., Homewood, IL, 843 p.
Pope. J A . and B B Parrish
1964. The importance of fishing power studies in abun-
dance estimation. Rapp. Cons. int. Explor. Mer 155, No.
17, p. 81-89.
Pope, J G . and D J Garrod
1975. Sources of error in catch and effort regulations with
particular reference to variations in the catchability coef-
ficient. Int. Comm. Northwest Atl. Fish. Res. Bull. 11,
p. 17-30.
Robson, D S.
1966. Estimation of the relative fishing power of individ-
ual ships. Int. Comm. Northwest Atl. Fish. Res. Bull. 3,
p. 5-14.
Rounsefell. G a
1957. A method of estimating abundance of groundfish on
Georges Bank. Fish. Bull., U.S. 113:264-278.
RoYCE, W. F., R J Buller, E. D. Premetz
1959. Decline of the yellowtail flounder (Limanda fer-
ruginea) off New England. Fish. Bull., U.S. 59:169-
267.
SAS Institute
1982. SAS User's Guide: Statistics. 1982 ed. SAS Insti-
tute Inc., Gary, NC.
107
FISHERY BULLETIN: VOL. 86, NO. 1
SissENWiNE. M P Stern. H , ,Jk , and R C Hknnemuth
1974. Variability in recruitment and equilibrium catch of 1975. A two-way model for estimating standardized fish-
the Southern New England yellowtail flounder fish- ing crfort applied to the U.S. haddock fleet. Rapp. Cons.
ery. J. Cons. int. Explor. Mer 36:15-26. int. Explor. Mer 168:44-49.
1978. Is MSY an adequate foundation for optimum Westrheim, S J , AND R. P. FOUCHER.
yield? Fisheries 3(6):22-42. 1985. Relative fishing power for Canadian trawlers land-
Steel. R G D, AND J H TORRIE. ing Pacific cod iCiadus macrocephalus) and important
1980. Principles and procedures of statistics. McGraw- shelf cohabitants from major offshore areas of western
Hill, N.Y., 633 p. Canada, 1960-1981. Can. J. Fish. Aquat. Sci. 42:1614-
1626.
108
REDUCING THE BYCATCH IN A COMMERCIAL TROTLINE FISHERY
Lawrence W McEachron,' Jeff F Doerzbacher,- Gary C Matlock,^
Albert W. Green,^ and Gary E, Saul-
ABSTRACT
Reducing the bycatch of red drum, Sciaenops ocellatus, and spotted seatrout, Cynoscion nebulosus, in
the Texas commercial trotline fishery is desirable. Hook placement within the water column was
examined as a means of accomplishing this objective. The commercial trotline fishery was simulated
in the Laguna Madre during February 1985 through January 1986. Requiring placement of trotline
hooks on bottom will reduce bycatch of red drum, spotted seatrout, and other nonmarketable fishes
and improve operational efficiency of commercial fishermen without significantly reducing catch of
black drum, Pogonias cromis, a target commercial species. Other than crab and shrimp being more
effective baits than oleander leaves, no other generalization could be made concerning baits and
seasons.
Longlines catch species unwanted or legally non-
retainable by fishermen and have been regulated
to reduce the bycatch of nontargeted species
(South Atlantic Fishery Management Council
1985). Trotlines (Figs. 1, 2) are a specialized long-
line used in shallow (<4 m) Texas estuaries to
catch fish (Simmons and Breuer 1962; Breuer
1973, 1974, 1975; Matlock 1980). Red drum,
Sciaenops ocellatus, and spotted seatrout,
Cynoscion nebulosus, were the primary targets
until 1981 when their sale was prohibited be-
cause of overfishing (Matlock et al. 1979; Anony-
mous 1979, 1981, 1983). The effort has since been
redirected toward black drum, Pogonias cromis.
Regulations requiring the use of circle hooks and
placement of the mainline under water were en-
acted to reduce the bycatch of red drum and spot-
ted seatrout. However, a bycatch still occurs. This
study was conducted to determine if the bycatch
could be further reduced by additional regulation
of where in the water column hooks are fished and
bait types.
MATERIALS AND METHODS
The catch on trotlines with hooks placed on the
bottom or in the top of the water column was
compared by simulating commercial fishing tech-
niques in the Laguna Madre, TX (Fig. 3). Bottom
trotlines were set with the mainline on the bot-
tom. Top trotlines had the mainline floated with
iTexas Parks and Wildlife Department, P.O. Box 1717, Rock-
port, TX 78382.
2Texas Parks and Wildlife Department, 4200 Smith School
Road, Austin, TX 78744.
the hooks suspended in water >0.6 m deep to
insure hooks fished in the water column. Texas
Parks and Wildlife Department (TPWD) trotlines
were set in the same area as commercial trotlines.
Commercial fishermen were contacted by tele-
phone within 24 hours prior to TPWD sets to de-
termine areas of commercial activity. All TPWD
trotlines were at least 15 m apart.
Trotlines with 100 hooks each were built ac-
cording to commercial fishermen specifications
(McEachron et al. 1985). The mainline (182.9 m
long) consisted of #36 nylon twine, knotted twice
every 1.8 m for swivel (1/0 black brass barrel)
placement (Figs. 1, 2). Hooks (#8 Mustad
39960ST) were attached by a 610-686 mm long
staging (56.7 kg test monofiliament) to the swivel
at 1.8 m intervals. Stakes (51 cm x 76 mm) and/or
anchors were placed on each end to stretch the
mainline. Floats (3.8 L) were attached to the
mainline every 15 hooks for navigation identifi-
cation.
Eighteen trotlines were set overnight each
month in both the upper and lower Laguna Madre
during 1 February 1985 through 31 January
1986. Six (3 top; 3 bottom) were set during each of
two monthly sampling periods (first and last 15
days of the month). Another six sets were made in
either the first or last half of each month; the
period was randomly selected each month. Each
trotline was baited completely with one of three
bait types — cut portions of blue crab, Callinectes
sapidus; dead shrimp, Penaeus sp.; or oleander,
Nerium sp. leaves — so that all bait types were
used on both top and bottom trotlines during
every period. These baits represented the most
Manuscript accepted October 1987.
FISHERY BULLETIN; VOL. 86. NO 1, 1987.
109
<5-*
5.1 cm X 7.6 cm
WOODEN STAKES
Figure 1.— Top trotline.
1 cm X 7.6 cm
WOODEN STAKES
Figure 2. — Bottom trotline.
commonly used baits by commercial trotline fish-
ermen (McEachron et al. 1980, 1986).
Fishes caught were identified (Hoese and
Moore 1977; Robins et al. 1980), counted (Table
1), and total length (TL) was measured to the
nearest 1 mm. Data were pooled into fall (Septem-
ber-November), winter (December-February),
spring (March-May), and summer (June-Au-
gust) to examine seasonal variation.
A catch rate (No. /line • h) for black drum; red
drum; spotted seatrout; hardhead catfish Arius
felis ; and total fishes was computed for each trot-
line set by dividing the number caught by the
number of hours fished. Catch rates were trans-
formed to log (catch rate + 1) and analyzed using
a four-factor fixed-effects model analysis of vari-
ance (AOV). The four factors were 1) hook place-
ment, at two levels — top and bottom; 2) bait, at
three levels — crab, shimp, and leaves; 3) bay, at
two levels — upper Laguna Madre and lower La-
guna Madre; 4) season, at four levels — fall, win-
ter, spring, and summer.
Diff'erences in main effect means were evalu-
ated with Ducan's multiple range test. However,
when significant first-order interactions were
found, comparisons were made within levels of
the interacting factors using the mean square
error (MSE) from the AOV.
Total lengths of each species were analyzed in
a nested AOV to investigate differences among
the four factors. However, because fish were not
caught in all factor level combinations, factors
and/or factor levels for each species were elimi-
nated from analyses. Spotted seatrout lengths
110
McEACHRON ET. AL: BYCATCH IN TROTLINE FISHERY
Figure 3.— Texas coast.
Table 1 —Number of fishes caught on trotlines in the upper and
lower Laguna Madre during February 1985-January 1986.
Upper
Lower
Laguna
Laguna
Species
Madre
Madre
Total
Arlus felis
977
1,652
2,629
Sciaenops ocellatus
352
658
1,010
Pogonias cromis
67
265
332
Cynoscion nebulosus
29
103
132
Micropogonias undulatus
36
51
87
Opsanus beta
34
1
35
Archosargus probatocephalus
1
24
25
Dasyatis americana
0
17
17
Dasyatis sabina
6
7
13
Elops saurus
4
4
8
Orthopristis chrysoptera
1
7
8
Bagre marinus
1
4
5
Lagodon rhomboides
0
5
5
Paralichthys lethostigma
1
3
4
Rhinoptera bonasus
0
4
4
Chilomycterus schoepfi
0
3
3
Ophichthus gomesi
3
0
3
Cynoscion arenanus
0
2
2
Negapnon brevirostris
0
1
1
Trachinotus carolinus
0
1
1
All species
1.512
2,812
4,324
were pooled for both bay systems because an in-
sufficient number of spotted seatrout were caught
for individual bay analyses. Factor levels elimi-
nated from length analyses were leaves and win-
ter from hardhead catfish, leaves and crab from
spotted seatrout, and fall, spring, and summer
from black drum. Spring and winter red drum
lengths were pooled. Lower Laguna Madre data
only were used for red drum, black drum, and
hardhead catfish length analyses. Each measured
fish length was an observational unit of a trotline
set. Sets were a random factor nested within fixed
main effect combinations. The nested set effect
mean square was used for testing other effects
when the set effect was significant. However, the
AOV yields approximate F values because un-
equal numbers of fish were caught among sets.
SAS procedures (SAS Institute, Inc. 1980, 1982)
were used for all analyses. The significance level
for each AOV test was a = 0.01 because the AOV
used to examine catch rates of each species had 15
potential F tests. This alpha value assured that
111
KKSHEKY BULLETIN: VOL. 86, NO. 1
the family level of significance would not exceed
0.15. All other tests were made with a signifi-
cance level of u = 0.05. Mean catch rates and con-
fidence intervals computed from transformed
data were back-transformed for tabular and
graphic presentation (Elliott 1979).
RESULTS
Fishing trotlines on the bottom reduces bycatch
without affecting catches of black drum, the
target species. A significant difference could not
be detected between black drum catch rates on
top and bottom trotlines regardless of bait, sea-
son, or bay (Tables 2, 3). Catch rates for hardhead
catfish, red drum, spotted seatrout and total
fishes were significantly lower on bottom trot-
lines than on top trotlines (Tables 2, 3). Differ-
ences in catch rates between top and bottom trot-
lines for hardhead catfish, red drum, and total
fishes did not vary significantly among seasons
and baits but did vary between bays based on the
first-order interactions (Tables 4, 5). A significant
second-order interaction of position x bay x bait
for spotted seatrout revealed that differences be-
tween top and bottom trotlines were affected by
both bait and bay but not by season (Fig. 4).
No significant difference was found in red drum
catch rates among baits nor in spotted seatrout
catch rates among seasons (Tables 2, 3). All other
main effects were significant for catch rates of all
species and total fishes. Of the possible first-order
interactions involving bait, season, and bay, only
season x bait for hardhead catfish, red drum, and
total fishes, bait x bay for black drum and total
fishes, and season x bay for black drum, hard-
head catfish, and total fishes were significant (Ta-
bles 4, 5). The second-order interaction of
bait X bay x season for total fishes was signifi-
cant (Fig. 5).
No significant differences were found in mean
lengths of black drum, hardhead catfish, red
drum, and spotted seatrout between top and bot-
tom trotlines (Tables 6-8). Significant differences
in mean length of hardhead catfish were detected
for main effects of bait and season (Table 7).
DISCUSSION
Management objectives could be better met by
requiring placement of trotline hooks on bottom
than by allowing hooks to be fished from the sur-
face. Red drum and spotted seatrout mortality
would be reduced without significantly affecting
TS
r
ro
CL
o
c-
o
4.^
(0
m
£
-)
QJ
(0
o
C
r
OJ
If)
o
i^
c
Q)
nj
•o
r
O
O
re
^o
o^
LO
_)
05
()
CO
<0
0)
I/I
O
■n
OJ
(/)
c
(U
(5
3
Q.
Ifl
(0
>
E
55
2
•o
Q.
CL
3
n
(1)
"O
c
.
(0
r
<fi
0)
(0
o
5
p
"n
0)
m
.c
QJ
j_
Q)
-n
(0
m
n
(D
E
7^
jC
T)
C
Ji:
9
o
(0
«
a.
J3
c
O
to
(0
03
r*
n
Q)
V-
r
D
r'
01
o
(D
ta
(1)
CD
■D
r
>.
o
(0
3
,->"
c
(0
—i
■D
in
(»
«
^—
o
o
£:-
m
(TJ
(/I
3
(0
JO
u
01
t_
LL
m
Ol
.c
C
11
3
c
■D
—
C
o
2
o
CO
CO
~ —
0)
(11
en
(fl
■D
*-
c
r
(D
o
to
(J
(0
J2
o
01
*-'
1-
(0
JD
o
>.
</)
n
t
m
</)
Q)
*^
c
V
o
n
■^
c
F
(0
o
Q)
t;
>
o
i
JD
(11
_j
(11
<
1-
0)
E
E
3
C/5
Q.
C/)
(0
LL
0)
i_ (0
Q) C
S CD (0
S (0-5
01
^ (tl
Q. 3 T3
0.01(0
^(0 3
(/)
(0
6
E
o
o
CD
en
o
(D
Q.
CD
(JD
t~-
CM
CO
O)
0
CO
in
0°
(£)
CM
CO
•^
■^
'-
0
0
0^
T- 0
0
CO
cp
CM
cp
CD 0
o^
d
0
t^
0
CO
-* 1^
0 0
d
in
0
in
d
0
0 CD
0
CM
0
0
CO
0
d
0
0
d
rg
~
?
in
—
in
in
00
, —
T—
t^ °
0
CO
to
CM
0
CD
0
r^®
CO 0
in
cp
CD
0
0
cp
r^ 0
0 ^
0
CO
0
,_
0
tT
'^co
d c\j
d
■t
d
0
0 in
0
5
0
0
in
0
d
d
d
d
r?
in
CO
CM
£"
LO
CJ)
CO
■*
t^
CO ""
CO
OM
cr>
CO
CO
0
t-"
r- <D
cp
O)
cp
C\J
cp
i- 0
■^ CO
CM
to
CM
in
0
4
r~- ^
0 r-
0
CO
0
d
d r-
0
cnI
0
in
0
d
d
d
d
£0
—
CD
cm"
0
00
s
(J)
CM
^
0)9
CM
CO
C\J
0
■^
0
CNJ^
■.- 0
(3
r--
cp
CD
in 0
P6
CO
(D
0
00
0
N-
^6
0 ••-
d
•^
d
•^
d
0
0 r~~-
0
CM
0
0
CO
0
d
d
d
d
CM
CO
^
£■
CO
01
O)
0
CO
•^
^ 0
■^
■^
CD
CM
00
0
in®
r- (D
CM
(D
CD
cp
CM
0
tt 0
q .;.
■<T
in
6>
0
(D
-"-cfi
d in
d
in
d
CM
d
0 TJ-
0
CO
0
CD
0
d
d
d
d
?
<B
5^
CM
CM
CM
r-.
1^ 0
in
CO
(J)
"-
CO
0
00^
■.- 0
CD
CD
CO
cp
0
cp
0 0
P6
CvJ
CO
0
h-
0
cb
^CD
0 T-
d
1 —
d
CD
d
0
0 tt
0
CM
0
0
CO
0
d
d
d
d
P^
CO
0
(D
in
CvJ
CM
M-
O)
t-°
0
■^
CO
'-
0
0
^^
■-- 0
O)
<o
0
cp
cp
CO 0
p ^.
0
in
4
0
4
^co
0 0
d
in
d
CO
d
0
0 r-.
0
0
0
0
0
d
d
d
d
—
in
ai
—
pr
5
TT
(j>
en
CD
^°
00
CO
in
""
CO
0
0
coP
■<a- 0
CO
C)
in
CD
CD
r^ 0
P r^
CM
CO
CM
CO
"^co
0 Cvl
d
CO
d
d
CM
0 O)
0
CM
T—
0
■*
0
d
d
d
d
05"
CO
cm"
—
^
(D
O)
CD
T—
CO
CM°
Tt
1^
CO
r^
0
inP
t^ 0
r^
CD
CM
CD
0
0
CM *-
°(0
(D
CO
t^
0
CO
C^liS
0 •*
d
CO
d
00
d
0
0 <y>
0
in
0
0
r-~
0
d
d
d
d
CO
ST
CM
G^
CM
in
■*
CD
0
r~-
0)0
1^
CM
r-
0
in
0
a)P
CO 0
0
cp
'^
d
0
d
•.- 0
°in
CM
in
0
.,^
0
yL
Pco
0 eg
d
CO
d
CO
d
0
0 CD
0
0
0
CM
d
d
d
d
d
<5"
CD
CO
—
in
(D
CO
in
■^
■^
0)°
(D
in
O)
CM
.^
0
OlP
'^ 0
CO
0
0
CD
CO
cp
CO 0
°co
'«■
CM
r^
0
CM
^O)
0 CO
d
, —
d
co
d
CM
0 CO
0
■<t
0
l-~
d
(0
d
d
3
0
d
d
JZ
E
■5
01
U)
3
"O
(0
0)
E
3
"D
TO
0
(0
■5
0)
0
1-
m
0)
a.
CD
I
cr
CO
cn
en
en
0)
<D
(D
0
0
0
(11
ni
(11
n
Ci
Q.
en
en
en
r
r
r
0
0
0
(0
(0
CO
01
0
0)
^_
^_
^_
C3
C)
Cl
CD
•<a-
00
-t
0
CM
II
II
II
^^^
(NJ
0
112
McEACHRON ET AL: BYCATCH IN TROTLINE FISHERY
oo
CD
£>
m
C
(0
-p
00
CD
k_
J3
0)
c
3
"D
(0
TO
c
3
O)
TO
_l
(/5
CO
E
o
o
n
c
TO
a.
o
c
o
QJ
TO
sz
o
TO
O
"CJ
0)
E
c
TO
c
TO
£
>
O
<
TO
^_
3
O
3
CD
.<
O q;
O 0)
ii
.,
en
o
Sf
E
TO
3
3
C/5
a:
o
c
a>
o
o
TO
3
O
TO
C/5
>
^i;r;i:'5'~ir'~'~'~oowcD'-r--
ooooininoooocDoooo-^
OOCDOOOCDOcioCDOOOcb
»— OCDTj-cooOCXJCOOCDCMCDlfl'l-CO
r--coa)inior-.cococDLnoT}--t-coo>
C\J
lO CO o o
r^'^Lrii-ocbcNJcNJo
CO -^
00 OCVJinCMCDCD'— COO'J-CDO'3''-T-r—
CO cocO'-cDTrojcocDininocnocococo
o ocDLnr-'.-coc\iooa)ooocoa5cot-^r-^
CO h^'-OJCOOOini-Tj-^OOOCMOCO
oj "-cvJocDdooocbcbcbczicbocbcri
■^'-'— '— lOt-i-t-i— ,-^C\JCDCOCO
oooocjTj-ocoooair-^-'-Ttoo
cbc3cbc)C)cbocboc3ooodcD
V V V V V V V
cococD''-a5<DcotDcooO'-ir)cor^
CMcoT-LncocDC\j'<j-Oin'.-tor^eD'3;
T-iri-.-rO'-^dtbcb'^'
rr CD T- •■- •^
CO o o T- o o
c\ia)ocDocoa5LniOLOcoa50T-r}-T-
LO'-cD'— oocDC\icoLnr~-c\jcDC\icDco^
cDcDmr^cococ\ji-cocDO'^r-r-.co(£)
u7r--^i-ooc\jococ\ioooooc\j
CDi-dcbdcbdocbdcDdcbdcbiri
■.- ,- O T- CO
o o ■>- o o
■'-maioo-^'j-oo-^Ln
cDcbcbcDcJcbcbcbcodcDcbcridd
COOO'-S-CDt^OOCDCMCOOfOitCD
OOCDi-CpOOr^CO-^COOOCOCOOOCOCD
Lric\icMihc\ic>cb->-T-cbuSc3CDdc)
CM -^ •>- CM
CD tncOOOOCDOOOCMO-r-OOO-^CDCO-r-
co ■^■^CMr^cM-^'^'-in-^LnO'-ocM'-
00 ■^'-oooo-'-oooooooor--
CM ooooooooooooooo---
d cDdcbcDdcDdcDcbcbcbcbcDdcbd
'-oo-^T-cDCM-.-co-'-r^ococDLnco
p-r-oo'^oooooococD-'j-or-
CDOcDcricJdcocricDCDebcbcDCDcb
CM''--^cocMa)t~^rgcDcoooocMO
i-r--.T3-cpt^T}-r~.-<ircMCMCMC\jcoi-cD
co-^iricDCDcocbcoLrid-^ddcMd
>- cDi^CM''--'-'<a-o-^Tj-mcMcococor^oo
LO coi-rO'-cj3CDoocO'-Ti-Lnt^LnocM(0
CD t^CMOJCOOCO-^'^OO-^Oi-COCMCO
in CMOCMOOOOOOdOOOOOO-^
CO cbcDCDcbdcbcDcbdocDdcbdcbcNJ
[i^z:i;°^^f^'~'^'^'^cr5oococD
coooocqo5LnO'-oooa)<DOCD
cbddddcbdcocbdcbcDdcDd
V V V V V
CMO-'-T-COunCOh-COTtOOCDCO'-CM
COCD'<J;00'----COCMCOh-->-'.-0'>^-.-
dcd"^dcbcb<3r^'r^cricDC3C3c\id
^ CM T-
cnCMtDOLnCOCOinOTOOTCOCO^COCM
'-inoocoooocot^ooo'— ocD'-'-
OCMt~^C0OOOC\|T-LnOOOCMOO
OOOOOOOOOOOOOOOh^
cbdcDocbcDdcbdcricDcbddcbd
'-CMCO'-CMCO-'-CMCOCOCMCOCOCOCO'V
CO
oo
c
o
(/)
(1)
</>
o §
~ 00 "' -5
CO tr TO >. in
O TO Q> TO O
Q- CD en m Q.
c
o
uo
</5
cn
•*^
TO
>-
TO
Q)
TO
E
n
U)
^
X
X
X
00
c
c
c
o
o
o
oo
E
Q)
00
>,
O)
>.
TO
XI
c
o
TO
0)
en
c
o
U)
TO
0)
E .
CD
TO
If)
E
to
>.
en
>.
TO
E
B
oo
>,
TO
<^ B B 5 ^ ^ ,_
TO TO TO O
CD m CD Q.
O TO
CL CD
O ^
g A
JO TO
c
'c
O)
C3)
(/>
CO
c
.^^
o
>.
c
TO
Q)
^
TO
X
O)
•^^
c
TO
J3
T3
TO
•o
OJ
C
^
TO
(0
>,
Q)
TO
CJ
n
0)
X
Q.
05
c
TO
o
'^
OJ
CO
■D
O
C
Q.
3
O
g
O
B
TO
TO
u.
C
^
O
CD
15
C
o
Q)
■o
(U
O
E
E
TO
o
03
In
0)
c
£
TO
>.
^
C
TO
■D
a>
a>
2
g
o
1
^
^
03
LLI
C
—i
CO
<
TO
0)
03
< CD
a>
>
CO
TO
o
CO
0)
o
o
_l
d
d
Q.
< CD
E
CO
CM
c
CO
in
n
o
o
en
o o
< <
CM CO
■•- CO
O -r-
CD CD
■* O
•<t in
o o
< <
in CO
oo CO
■■- CM
d d
< CD
CD r^
■■- CD
C\J ■.-
d d
< <
LO in
•>- CD
00 CD
(U QJ
■O JJ
TO TO
E
22
03
TO TO
03
C C
>-
3 3
0)
O) 03
TO
TO TO
_l _l
CD
k_ k_
03 03
1=1- g
Q. o
D -J
03
C
TO
0)
2
03
CL
Q.
3
II
03
c3 =>
TO
03 -.^
? i
g^
c o
TO •
O
^ CL
C ^-'
03 >,
.£ TO
>..y
TO ^
X -^
^ 03
C
o
03
S 2?
05 TO
T3
C
TO
03
C
T3
TO 03
03
03
O
03
<u <n 03
05 TO -o
;^ ™
c
o
03
TO
O
03
TO
O
TO
O
03
t 03 O
— TO — i
C 03 -I
TO 03 iu
03 ^ il
5 — "O
I >- TO
I -Q 2
^"S
TO
- I §
m o o3
2 P "1
x:
03
2
^
_l
TO
_J
O
"O
TO
03
x:
"O
2
_l
TO
I
D
c
o
03
TO
03
cn
< CD m <
CO Tj- r^ CD
03 tT ■.- tT
OJ -"t CD 00
d d d d
< < < <
CD ■>- CO in
CM O •* O
CO O -^ CO
c:i d d> d
< CD < <
o in in CD
00 O CO ■•-
o c\j o o
cs d d d
< < < <
CO oo CO o
O CM CO O
o o o o
d d d d
< CD < <
h- 00 r-- CD
■^ 00 t^ o
o ■.- o 1-
d d d d
< CD < <
oo ■^ in TT
CD CD CO •*
O TJ- o o
d d d d
< < < <
O TJ- CD O
o CM r~. o
■•- C\J o ••-
d d d d
O < O O
CM oo CO in
•,-■,- r-- ■*
CM T- T- o
d d d d
CD < CD CD
O O ■* -r-
"- o ■* o
CO CM ■^ CM
d d d d
< < < <
■* CD CD r^
•■- CM o in
CD CO O 1^
d d y-^ d
03
03
w g'E
= -E i E
TO •> CL 3
LL S CO Cn
113
0.12
0.10-
O.OB
E
3 0.06
< 0.04
a.
I
u
< 0.02
o
0.00-
— I —
ULM
FISHERY BULLETIN; VOL. 86, NO. 1
SHRIMP
TOP
CRAB
LEAVES
TOP
TOP
BOTTOM
BOTTOM
— I
LLM
— I
ULM
— I
LLM
— I
ULM
— I
LLM
BAY SYSTEM
Figure 4.— Significant second-order interaction among positions, baits, and bay system for spotted
seatrout catch rates. LLM = lower Laguna Madre, ULM = upper Laguna Madre.
Figure 5. — Significant second-order interac-
tion among bait types, bay system, and sea-
sons for total fishes catch rates. LLM = lower
Laguna Madre, ULM = upper Laguna
Madre.
- 1.5
\
• 1.2
E
I 0.9
<
K
I
O
I-
<
u
uj 0.6 •
0.3
0.0 ,.
l.B
1.5 ■
1.2
0.9
0.6
0.3
0.0-
FALL
CRAB
SHRIMP
LLM
ULM
LEAVES
LLM
ULM
I I
WINTER SPRINS
SEASON
SUMMER
114
McEACHRON ET. AL: BYCATCH IN TROTLINE FISHERY
Table 6. — Mean length (mm t 1SE) of black drum, hardhead catfish, red drum, and spotted
seatrout caught on top and bottom trotlines by bait, bay and season during February 1985-
January 1986 Number in parentheses is number of fish measured.
Position
Bait
Species
Top
Bottom
Crab
Shrimp
Leaves
Black drum
539 ± 10
550 ± 13
566 ± 1 1
493 ± 12
588 i 21
(177)
(146)
(172)
(108)
(43)
Hardhead catfish
338 ± 1
343 ± 2
346 ± 1
328 ± 2
337 ± 3
(1.363)
(675)
(1,201)
(651)
(186)
Red drum
521 ± 4
504 ±9
540 ± 6
481 ±6
542 ± 7
(657)
(173)
(260)
(336)
(234)
Spotted seatrout
439 ± 9
420 i 24
471 ± 23
426 ± 10
445 ± 19
(111)
(18)
(17)
(86)
(26)
Bay
Season
Upper
Lower
Laguna
Laguna
Species
fvladre
(vladre
Fall
Winter
Spnng
Summer
Black drum
484 ± 19
560 ±8
464 ± 20
538 ±8
630 ± 21
416 ± 19
(67)
(256)
(35)
(201)
(71)
(16)
Hardhead catfish
327 ± 1
348 ± 1
329 ±2
368 ±2
333 ±2
345 ±2
(824)
(1,214)
(538)
(277)
(724)
(499)
Red drum
506 ± 7
523 ± 5
493 ± 1 1
506 ± 5
546 ±9
551 ± 10
(308)
(522)
(133)
(427)
(114)
(146)
Spotted seatrout
411 ± 15
443 ± 10
410 ±21
404 ± 1 2
482 ± 25
478 ± 13
(29)
(100)
(25)
(51)
(16)
(37)
Table 7. — Summary of results of the AOV s of mean length for black drum (winter season only), hardhead catfish (excludes leaves and
winter season), and red drum (winter and spring seasons combined) on top and bottom trotlines in lower Laguna Madre during February
1985-January 1986. NA = not analyzed.
Black drum
Hardhead catfish
Red drum
Source of
Sum of
Sum of
Sum of
variation
df
squares
F
PR > F
df
squares
F
PR>F
df
squares
F
PR>F
Total
176
1,941,401
1,050
2,310,898
521
6,484,793
Position
1
5,978
0.15
0.70
1
80
0.02
0.90
1
6,122
0.26
0.61
Bait
2
15,082
0.19
083
1
33,650
6.60
0.01
2
89,151
1.90
0.15
Season
NA
3
89,202
5.83
<0.01
2
133,190
2.85
0.06
Position X bait
2
49,063
0.69
0.55
1
8,242
1.62
0.21
2
10,913
0.23
0.79
Season ^ position
NA
3
1,517
0.10
0.96
2
789
0.02
0.98
Season x bait
NA
3
6,132
0.40
0.75
4
68,535
0.73
0.57
Season '; position x bait
NA
3
10,269
0.67
0.57
4
63,388
0.68
0.61
Set (position x bait)
27
10,906,616
8.26
-O.OI
NA
NA
Set (season '- position
■ bait)
NA
89
453,750
3.22
<0.01
91
2,129,491
2.61
<0.01
Error
144
704,611
946
1,499,474
413
3,706,235
Table 8. — Summary of results of the AOV of mean length for spot-
ted seatrout (shnmp bait only) on top and bottom trotlines in upper
and lower Laguna Madre combined dunng February 1985-
January 1986.
Source of
variation
df
Sum of
squares
PR>F
Total
85
750,362
Position
1
601
0.04
0.84
Season
3
58,796
1.32
0.29
Position X season
3
58,462
1.31
0.29
Set (position ^ season)
28
416,855
3.42
<0.01
Error
50
217,744
115
FISHERY BULLETIN: VOL 86, NO 1
black drum catches. Operational efficiency of
commercial fishermen should improve with less
handling of nontarget species. Mortality of non-
target fishes would decrease because they would
not be caught and subsequently handled. For red
drum and spotted seatrout that are caught, sur-
vival would be high for those released back into
the water. Survival of released red drum caught
on trotlines in winter and summer and of spotted
seatrout in winter was 100^^ (Martin et al. 1987).
About 509^ of the spotted seatrout died in summer
cage studies; but few commercial trotlines are
fished during this period (TPWD unpubl. data).
Thus, the goal of reducing the catch of nontarget
species and reducing mortality due to trotlines
can be achieved with minimal impact on the com-
mercial fishermen.
Interactions between bay system and the other
three factors for some species probably reflect dif-
ferences in relative abundance. Fewer fish were
available to be caught in upper Laguna Madre
than in lower Laguna Madre (Crowe et al. 1986).
The effects of bait and season on trotline catches
cannot be determined in bay systems where the
fish abundance approaches zero.
No spotted seatrout were caught on crab bait on
bottom in either bay; but they were caught on all
baits on top in the lower Laguna Madre leading to
the significant second-order interaction of posi-
tion X bay X bait. This condition was not unex-
pected because spotted seatrout are predomi-
nately sight feeders (Vetter 1977), and might not
take baits on bottom as readily as baits suspended
in the water column.
Crab and shrimp were more effective baits than
oleander leaves for all four species. No other gen-
eralizations could be made concerning baits and
seasons. Selection of crab or shrimp as the bait of
choice for reducing bycatch while maximizing
black drum catch is unclear because catch rates
for black drum and red drum were greater on crab
than shrimp, especially in winter.
ACKNOWLEDGMENTS
We would like to thank all Laguna Madre field
personnel who diligently collected the samples.
Tom Heffernan, Ed Hegen, Lynn Benefield,
Maury Ferguson, and Tony Maciorowski re-
viewed the manuscript.
LITERATURE CITED
Anonymous
1979. Saltwater finfish research and management in
Texas. A report to the Governor and the 66th
Legislature. Tex. Parks Wildl. Dep., Coastal Fish.
Branch, PWD Rep. No. .3000-59, 21 p.
1981. Saltwater finfish research and management in
Texas. A report to the Governor and the 67th
LegLslature. Tex. Parks Wildl. Dep., Coastal Fi.sh.
Branch, PWD Rep. No. ;J000-10K, ,31 p.
1983. Saltwater finfish research and management in
Texas. A report to the Governor and the 68th
Legislature. Tex. Parks Wildl. Dep., Coastal Fish.
Branch, PWD Rep. No. 3000-154, 48 p.
Breuer, J P
1973. A survey of the juvenile and adult food and game
fish of the Laguna Madre. Tex. Parks Wildl. Dep.,
Coastal Fish. Branch, Proj. Rep. 173-202.
1974. Juvenile and adult food and game fish of the
Laguna Madre. Tex. Parks Wildl. Dep., Coastal Fish.
Branch, Proj. Rep. 109-130.
1975. Biological studies in the lower Laguna Madre of
Texas, 1975. Tex. Parks Wildl. Dep., Coastal Fish.
Branch, Proj. Rep. 158-196.
Crowe, A L , L W McEachron, and P C Hammerschmidt.
1986. Trends in relative abundance and size of selected
finfish in Texas bays: November 1975-December
1985. Tex. Parks Wildl. Dep., Coastal Fish. Branch,
Manage. Data Ser. No. 114, 259 p.
Elliott. J. M
1979. Some methods for the statistical analysis of
samples of benthic invertebrates. Freshwater Biol.
Assoc, Sci. Publ. No. 25, 160 p.
HoESE. H D , AND R. H Moore
1977. Fishes of the Gulf of Mexico, Texas, Louisiana, and
adjacent waters. Texas A&M Univ. Press, College
Station, 327 p.
Martin. J H , K W Rice, and L W McEachron
1987. Survival of three fishes caught on trotlines. Tex.
Parks Wildl. Dep., Coastal Fish. Branch, Manage. Data
Ser. No. Ill, 21 p.
Matlock, G. C
1980. History and management of the red drum fishery.
In Proceedings Colloquium on red drum and sea-
trout, p. 37-54. Gulf States Mar. Fish. Comm. No.
5.
Matlock, G C , P L Johansen, and J P Breuer
1979. Management of red drum in a Texas estuary - a
case study. Proc. Annu. Conf Southeastern Assoc.
Fish Wildl. Agencies 33:442-450.
McEachron, L W , A W Green, G. C Matlock, and G E. Saul.
1985. A comparison of trotline catches on two hook types
in the Laguna Madre. Tex. Parks Wildl. Dep., Coastal
Fish. Branch, Manage. Data Ser. No. 86, 44 p.
1986. Evaluation of the commercial trotline fishery in
the Laguna Madre during fall 1984. Tex. Parks Wildl.
Dep., Coastal Fish. Branch, Manage. Data Ser. No. 93,
25 p.
McEachron. L W . G C Matlock. A R Martinez, and J P.
Breuer
1980. Evaluation of natural, leaf, vegetable, worm and
cork baits used on trotlines in upper and lower Laguna
Madre, Texas (September 1977-October 1978). Tex.
Parks Wildl. Dep., Coastal Fish. Branch, Manage. Data
Ser. No. 8, 68 p.
Robins, C R , R M Bailey, C E Bond. J R Brooker, E A
Lachner, R N Lea, and W. B Scott.
1980. A list of common and scientific names of fishes
from the United States and Canada. (4th ed. ) Am.
116
McEACHRON ET AL: BYC'AIVH IN TROTLINE FISHERY
Fish. Soc. Spec. Pub. No. 12, 174 p. SOUTH ATLANTIC FISHERY MANAGEMENT COUNCIL.
SAS Institute. iNC 1985. Source document for the swordfish fishery
1980. SAS supplemental library user's fjuide. SAS management plan. South Atl. Fish. Manage. Counc,
Institute Inc. Cary, NC, 202 p. Charleston, SC.
1982. SAS users guide: Statistics. SAS Institute Inc. Vetter, R D
Cary, NC, 584 p. 1977. Respiratory metabolism of, and niche separation
Simmons, E G . and J P Breuer between two co-occurring congeneric species, Cynoscion
1962. A study of redfish, Sciaenops ocellata Linnaeus nebulosus and Cynoscion arenarius in a south Texas
and black drum, Pogonias cromis Linnaeus. Publ. estuary. MA Thesis, Univ. Tex, 113 p.
Inst. Mar. Sci., Univ. Tex. 8:184-211.
117
LARVAL DEVELOPMENT OF BLUE GRENADIER, MACRURONUS
NOVAEZELANDIAE (HECTOR), IN TASMANIAN WATERS
B D BruceI
ABSTRACT
The development of Macruronus novaezelandiae is described and illustrated from both reared speci-
mens and larvae from Tasmanian waters. Eggs of M. novaezelandiae are pelagic, spherical (1.08-1.18
mm diameter), and have a single oil droplet (0.36-0.42 mm diameter). Eggs hatch after 55-60 hours
at 14°-19°C. Larvae are 2.2-2.3 mm at hatching. Characteristic pigmentation, a myomere count of
78-80, and the sequence of fin development separate M. novaezelandiae from other known gadiform
larvae. Development is direct, with no marked change in body morphology. Fin development proceeds
in the sequence: second dorsal, anal, first dorsal, pelvic, caudal, pectoral. However, adult fin comple-
ments are reached in the sequence: first dorsal, pelvic, anal, second dorsal, caudal, pectoral.
Caudal development is late in Macruronus. Flexion begins at 20 mm and is not complete until 28
mm. The caudal fin is based on two ural centra, four hypurals, two epurals, and a parhypural. X and
Y bones are present although they are not readily distinguishable from dorsal and anal pterygio-
phores.
The genus Macruronus comprises four nominal
species, which occur in southern temperate conti-
nental shelf and slope regions. Two species,
Macruronus novaezelandiae and M. magellanicus
support commercial fisheries. The blue grenadier,
M. novaezelandiae , forms the basis of fisheries in
New Zealand and Australia where total annual
catches range up to 97,750 and 1,100 t respec-
tively (Patchell 1982; Wilson 1981, 1982).
Macruronus magellanicus is fished commercially
off South America. The remaining species, M.
maderensis and M. capensis , are known only from
a limited number of specimens (Svetovidov 1948;
Cohen 1986). Despite their economic importance
and widespread distribution, very little is known
of the early life history of any member of the
genus. Patchell (1982) identified winter spawn-
ing grounds on the west coast of the South Island
for New Zealand populations of M. novaeze-
landiae and similarly Wilson (1981, 1982) has
suggested a winter spawning, on the west coast of
Tasmania, for Australian M. novaezelandiae.
This paper presents the first published informa-
tion on the larvae of Macruronus .
In 1984, the Division of Fisheries Research of
the Commonwealth Scientific and Industrial Re-
search Organization established a multidisci-
ICSIRO Division of Fisheries Research, GPO Box 1538, Ho-
bart, Tasmania 7001, Australia; present address: South Aus-
tralia Department of Fisheries, GPO Box 1625, Adelaide, South
Australia 5001, Australia.
Manu.scnpt accepted September 1987.
FISHERY BULLETIN: VOL. 86. NO. 1, 1988.
plinary program to investigate the biology and
ecology of blue grenadier in Tasmanian waters.
An integral part of this program was a study of
larval ecology. As such, it was first necessary to
establish criteria for the identification of blue
grenadier larvae. This paper describes the larval
development of M. novaezealandiae from Tasma-
nian waters.
MATERIALS AND METHODS
Specimens were obtained from samples col-
lected aboard the CSIRO Fisheries Research Ves-
sel Soela between April 1984 and September
1985. Details of sampling strategies, locations,
and procedures will be described in a subsequent
manuscript. Larvae were obtained by sampling
with a rectangular midwater trawl (RMT 1+8;
Baker et al. 1973), aim diameter ring net (500
[im mesh), and free-fall, vertical drop nets of 64
|jLm and 200 fxm mesh (Heron 1982). Juvenile
specimens were obtained with an Engels 352
pelagic trawl fitted with a 10 mm liner.
Newly hatched larvae were reared from eggs
stripped and fertilized at sea. Eggs and milt
stripped from ripe adults trawled from 500 m
were mixed in 1 L plastic jars filled with sea-
water. Despite the jars being located in a sea-
water bath, incubation temperatures varied con-
siderably (14°-19°C). On return to the
laboratories at Hobart, the eggs were transferred
to 2 L glass jars and placed in a constant temper-
119
FISHERY BULLETIN: VOL. 86, NO. 1
ature incubation chamber set at 14.0° ± 0.2°C. In-
cubation jars were not aerated, and no attempt
was made to feed the larvae.
All specimens used for description were fixed in
a IQ% formalin-seawater solution buffered with
sodium p-glycerophosphate and later transferred
to a 59^ solution.
This description is based on a series of 74 lar-
vae, 2.2-34.2 mm in length, although comments
on pigment and meristic variability stem from
routine examination of several hundred speci-
mens. A representative series of larvae is de-
posited with the South Australian Museum, Ade-
laide, South Australia.
Developmental terminology follows Ahlstrom
et al. (1976). Body measurements follow Matarese
et al. (1981). Length measurements are reported
as notochord length, NL (i.e., from the snout tip to
the end of the notochord) in preflexion and flexion
larvae, and standard length, SL (i.e., from the
snout tip to the posterior margin of the superior
hypural elements) in postflexion larvae and juve-
niles. Larvae were measured under a dissecting
microscope fitted with an ocular micrometer and
a camera lucida. Juveniles were measured with
vernier calipers.
Meristic counts and examination of ossification
sequences were made on specimens cleared and
stained using Alizarin Red S-KOH-glycerine
(Hollister 1934). Caudal osteology follows Inada
(1981), Marshall and Cohen (1973), and Monod
(1968).
Vertebral counts include the first vertebrae,
the neural spine of which is fused to the supraoc-
cipital crest (Marshall 1966), and both ural cen-
tra. Vertebral centra were counted as ossified
only when a complete band of stain connected
both neural and haemal spines.
RESULTS
Identification of M. novaezelandiae larvae was
based on their typical gadiform morphology
(large head, compact gut, tapering body form),
myomere count, and the development of confluent
dorsal-caudal-anal fins (see section on Distin-
guishing Features). Identification of field-
collected specimens was confirmed by comparison
with reared larvae.
Distinguishing Features
Prior to median fin development, myomere
counts are useful in separating M. novaezelandiae
larvae (78-80) from similarly pigmented morid
(41-72), macrourid (10-16 + 70 > 100), gadid
(39-64) and other known merlucciid larvae (48-
58) which they superficially resemble (Marshall
and Iwamoto 1973; Fahay and Markle 1984;
present study).
Both M. novaezelandiae and most morid larvae
show moderately pedunculate pectoral fins, a fea-
ture common in gadiform larvae with delayed
caudal development (Fahay and Markle 1984).
Macrourid larvae, in contrast, have very promi-
nently stalked pectorals and can further be sepa-
rated from M. novaezealandiae and most morids
by precocious development of the pelvic fin.
Size at caudal flexion and the sequence of fin
development are also useful in separating M.
novaezelandiae from all Merluccius species. In
Merluccius , notochord flexion generally begins at
about 9 mm and the caudal fin is the first to form
(Dunn and Matarese 1984; Fahay and Markle
1984). Macruronus novaezelandiae larvae do not
begin caudal flexion until approximately 20 mm,
and the caudal fin is the second last to form.
Macruronus novaezelandiae larvae have 1-3
prominent melanophores along the ventral mid-
line of the tail (although variable in appearance,
see section on Trunk and Tail Pigmentation)
and a double series of dorsal melanophores. When
expanded, melanophores in these two regions
coalesce to give the appearance of a broad
postanal band. Postanal banding patterns are
widespread in gadoid larvae (Fahay and Markle
1984); however, unlike many gadoid larvae, M.
novaezelandiae lacks pigment at the notochord
tip.
At larger sizes M. novaezelandiae larvae de-
velop long-based dorsal and anal fins confluent
with the caudal fin. Other gadoid larvae with this
configuration have markedly different pigmenta-
tion (see Fahay and Markle 1984 for details).
Ophidiiform larvae have confluent dorsal, caudal,
and anal fins but can be separated from M. no-
vaezelandiae by their lack of a separate first dor-
sal fin and general lack of body pigment (see Gor-
don et al. 1984).
Development
Embryonic development has not been treated
in detail here as it is the subject of a manuscript
in preparation by G. Patchell (Fisheries Research
Centre, Wellington, New Zealand).
The pelagic eggs of blue grenadier are spheri-
cal, with an unsegmented yolk and a smooth cho-
120
BRUCE: LARVAL DEVELOPMENT OF BLUE GRENADIER
rion. Late-stage eggs are 1.08-1.18 mm in diame-
ter with a single oil droplet of 0.36-0.42 mm di-
ameter (Fig. lA). Reared larvae hatch at 2.2-2.3
mm after 55-60 hours ( 14'-19"C). Newly hatched
larvae have a posteriorly positioned oil droplet
and adopt a head down position in rearing con-
tainers. Yolk absorption was incomplete in speci-
mens reared to 3.7 mm (6 days posthatch), al-
though the smallest field-collected larvae (3.6
mm) had already completed yolk absorption. The
anus opened laterally to the right in all reared
larvae and 95*^ of field-collected larvae. The anus
becomes symmetrical by 5.1 mm. A lateral anal
opening in M. novaezelandiae is consistent with
the developmental pattern reported for other gad-
iform species (Marak 1967; Matarese et al. 1981;
Fahay and Markle 1984; Dunn and Vinter 1984).
Field-collected larvae are moderately elongate
with the greatest body depth (16-229? body
length) occurring at or near the pectoral fin base
(Table 1). Head length as a proportion of body
length (BL) remains relatively constant at about
22% BL throughout the larval phase, decreasing
to about \1'7( BL in juveniles. Eye diameter de-
creases from 99f BL in preflexion larvae to A'7c BL
in juveniles. Depth at anus remains relatively
constant at about 139^ BL in larvae and juveniles.
Distances from the snout tip to the first dorsal fin
and from the snout tip to the anal fin decrease
slightly during development from about 279?^ BL
to 2l7c BL and 519r BL to 469? BL respectively.
in the number of melanophores and their degree
of expansion. Although Badcock and Merrett
(1976) suggested the appearance of melanophores
can change on a diurnal rhythm, in the blue
grenadier larvae examined, there was no conspic-
uous relationship between time caught and me-
lanophore expansion.
Head Pigmentation
Newly hatched larvae (2.2-2.3 mm) have
melanophores concentrated in front and behind
the eye (Fig. IB). Melanophores increase in num-
ber and extend over the sides of the head and
snout by 3.3-3.5 mm (reared larvae, Fig. IC).
Melanophores migrate dorsally to the top of the
head by 3.6 mm (Fig. ID). Eyes become pig-
mented at this size in reared larvae. By 4.5 mm,
the dorsal pigment on the head consists of a group
of 3-11 melanophores scattered over the hind-
brain and posteriorly to above the cleithrum. Pig-
ment gradually extends over the midbrain, with 1
or 2 melanophores usually present between the
eyes by 5.3 mm. Melanophores develop externally
over these initial mid- and hindbrain spots and
extend posteriorly as a double row to the dorsal
fin anlage by 7.2 mm. Dorsal pigment gradually
intensifies: melanophores increase in number
and form a cap over mid- and hindbrains by 16.0
mm. Melanophores extend down between the eyes
to the tip of the maxilla by 12.0 mm. Internal
Table 1 — Body proportions of larvae and juveniles of Macruronus novaezelandiae (expressed as percentage NL or SL): mean, standard
deviation, range.
Preflexion
Flexion
Postflexion
Juvenile
Body proportions
sample size
42
4
4
2
X
SD
range
X
SD
range
X
SD
range
X
SD
range
length (mm)
8.9
4.5
(3.6-19.0)
23.5
2.3
(20.6-26.1)
300
2.9
(27.6-34.2)
1890
1.4
(188.0-190.0)
head length
227
1.7
(18.3-24.7)
23.7
1.0
(223-24.7)
22.3
1.4
(20.5-23.6)
17.6
0.4
(17.3-17.9)
eye diameter
92
0.7
(8,1-10.3)
7.9
0.4
(7.3-8.3)
7.3
0.7
(6.5-8.0)
4.2
0.4
(4.4-5.2)
snout length
6.1
0.9
(4.6-7.7)
6.2
0.6
(5.7-7.0)
5.9
0.5
(5.4-6.5)
4.9
0.1
(4.8-5.0)
depth at pectoral
223
1.7
(21.0-24.4)
17.9
0.6
(17.2-18.5)
16.6
2.0
(13.7-18.5)
13.1
0.1
(13.0-13.2)
depth at anus
120
2.4
(8.2-15.2)
13.1
0.3
(12.6-13.4)
13.0
0.8
(12.3-13.8)
12.7
0.6
(12.3-13.1)
snout to first
dorsal fin
27.5
1.3
(25.2-29.3)
26.6
0.8
(26.0-27.7)
25.3
0.8
(24.2-26.0)
20.7
0.2
(20.6-20.9)
snout to anal fin
51.4
0.9
(50.0-52.6)
50.4
0.5
(49.6-50.6)
46.6
1.5
(45.0-48.4)
46.5
1.2
(45.7-47.4)
Pigmentation
Although pigmentation in M. novaezelandiae is
variable, certain features persist that, when com-
bined with meristic and morphometric informa-
tion, enable identification. Variation in the ap-
pearance of pigmentation is a result of differences
pigment expands over the forebrain in larvae
from 9.0 to 15.0 mm.
Ventral pigment on the head first develops in
4.2 mm larvae as 3-5 melanophores between the
dentaries. The number of melanophores increases
to 10-12 by 12.0 mm.
121
FISHERY BULLETIN: VOL, 86, NO. 1
Figure 1. — Development of Macruronus novaezelandiae: A)
Late stage egg 1.08 mm diameter, oil droplet 0.37 mm di-
ameter, B) 2.2 mm.; C) 3.5 mm.; D) 3.6 mm.; E) 5.3 mm.; F)
7.2 mm.; G) dor.sal view of above; H) 12.0 mm.; I) 24.2 mm.
postanal myomeres omitted. A-C reared specimens, D-I
field-collected.
BRUCE: LARVAL DEVELOPMENT OF BLUE GRENADIER
123
FISHERY BULLETIN; VOL. 86, NO. 1
The onset of dentary pigment is variable; no
pigment may be present on some larvae as large
as 7.0 mm. Most larvae develop 1 or 2 melano-
phores over the posterior section of the dentary by
5.3 mm and add melanophores anteriorly along
its length, with 5 or 6 usually present by 7.1 mm.
Two melanophores are often present around the
otic capsule by 7.0 mm, but they are obscured by
overlying tissue in 10.0 mm larvae. Scattered
melanophores develop over the pterotic region by
25.0 mm, but the operculum and preoperculum
remain largely unpigmented, even in the largest
specimen examined (34.2 mm).
Trunk and Tail Pigmentation
Newly hatched larvae have melanophores on
the body above the yolk sac and ventrally on the
tail. Some pigment is also present on the yolk sac
near the developing gut and scattered over the oil
droplet. Pigment forms a cap over the gas bladder
by 4.2 mm. Melanophores are gradually added to
the lateral surfaces of the gut throughout the lar-
val period until the entire gut (including the ven-
tral surface) becomes pigmented by 30.0 mm.
Dorsal pigment first appears on larvae 3.8-4.5
mm as scattered melanophores at approximately
607c NL. Melanophores rapidly increase in num-
ber and form a double row, extending from 51% to
67% NL in larvae of 5.0 mm. Lateral melano-
phores may also develop above the body midline
in this region. Concurrently, a similar double row
of melanophores appears and extends posteriorly
from the head (Fig. IG). The head and tail rows
join by 10.5 mm. Melanophores appear posteri-
orly in the caudal region by 29.0 mm forming a
twin series one either side of the developing dor-
sal fin. Pigment also appears internally on the
dorsal surface of the vertebrae in larvae of 9.5 mm
and extends anteriorly to approximately 50% SL
and posteriorly to the last vertebrae by 34.0 mm.
Single melanophores appear on the dorsal fin
ray bases by 14.0 mm and are present on all bases
by 29.0 mm.
Pigment along the ventral midline of the tail
appears in newly hatched larvae as a diffuse re-
gion that extends posteriorly from the yolk sac to
75-82% NL. This contracts to 1-3 melanophores
(most commonly 2) located 52-65% NL in larvae
of 3.8-4.0 mm. Additional melanophores (up to 6)
may appear later, but the initial 1-3 melano-
phores persist throughout the larval period. In
larvae larger than 7.0 mm, the initial 1-3
melanophores appear internally above the anal
fin ray bases and are gradually obscured by both
overlying musculature and external melano-
phores. These ventral melanophores on the tail
are a useful diagnostic character, although their
appearance varies, depending on their degree of
expansion. This variability in melanophore ap-
pearance is particularly evident in small larvae
where expanded ventral melanophores may ex-
tend over the lateral surfaces of the body to al-
most the dorsal area (Figs. ID, 2).
Lateral pigment gradually intensifies through-
out the larval period, excepting the area immedi-
ately above the gut, which remains largely devoid
of pigment even in the largest specimen (34.2
mm).
Morphological Variability
Macruronus novaezelandiae larvae showed
some size variation in development. In general,
specimens captured in ring net and RMT samples
appeared to develop features at slightly smaller
sizes than those taken from drop net samples.
This is likely a result of difi"erential shrinkage of
specimens caught by the different capture sys-
tems. Hay (1981) reported that considerably more
shrinkage occurred in Pacific herring when lar-
vae were killed prior to fixation and that shrink-
age increased with tow length. Ring net and RMT
tows varied in duration from 15 to 110 minutes,
with most larvae dead by the time the net was
retrieved and the catch fixed. Drop net sampling,
in contrast, lasted for, at most, 3 minutes dura-
tion, and many larvae were still alive on fixation.
Some variability in development can also be ex-
pected in field-collected larvae as a reflection of
past history (e.g., feeding success), although it is
unlikely such variations would account for the
observed differences between larvae caught by
different techniques.
Meristics and Osteology (Table 2)
Head and Axial Skeleton
In laboratory-reared larvae, jaw development
was first visible after 3.5 days (posthatch) with a
functional mouth present in larvae of 5.5 days
(3.7 mm). Pigmentation of the eyes also occurred
at this time suggesting that larvae were ready for
first feeding. The smallest larva stained was a
field-collected specimen 3.7 mm NL. The maxilla,
premaxilla, dentary, and cleithrum were all ossi-
fied in this specimen.
124
BRUCE: LARVAL DEVELOPMENT OF BLUE GRENADIER
Figure 2. — Variability of ventral pigment on the tail in 4.9 mm larvae of Macruronus novaezelandiae.
Ossification of branchiostegals begins in larvae
of 4.6 mm, with the full complement of 7 ossified
by 11.5 mm. Gill rakers are first discernible in
larvae of 9.4-9.9 mm with the full complement of
7 + 22-23 present by 28.9 mm.
Ossification of neural and haemal spines gener-
ally precedes that of the vertebral centra. Ossifi-
cation of centra, neural spines, and haemal spines
occurs sequentially from anterior to posterior pro-
ceeding slowly in larvae less than 9.0 mm in
length and then more rapidly until the full com-
plement is ossified by 23.2 mm. Elements associ-
ated with the caudal complex are the last to os-
sify.
Fins
Completion of fin development in M. novaeze-
landiae occurs in the sequence: first dorsal and
pelvic (almost simultaneously), anal, second dor-
sal, caudal, pectoral.
Pelvic fins first appear in larvae of 5.7— 5.8 mm
as slight swellings either side of the gut. They do
not form distinct buds until 6.9 mm. Ossification
125
FISHERY BULLETIN: VOL 86, NO 1
Table 2.— Meristic counts from cleared and stained larval and juvenile Macruronus novaezelandiae Specimens between dashed lines are
undergoing notochord flexion, a = specimen damaged; b juveniles not stained.
Length
(mm)
37
3.9
4.2
4.6
4.8
5.2
6.0
74
9.4
9.9
11.5
16.3
17.4
19.8
dorsal
4 + 28
0 + 19
9 + 74
12 + 84
12 + 86
13 + 87
Fin rays
anal pectoral pelvic
Branchi-
ostegal
rays
Gill rakers
upper lower total
Total
centra
Neural
spines
Haemal
spines
18
4
60
73
76
86
4
8
8
8
1
1
3
3
5
6
6
7
7
7
7
8
6
12
15
15
15
8
6
12
16
18
20
5
42
38
55
70
71
73
1
1
2
2
2
3
6
55
54
58
70
70
74
37
34
42
55
55
57
Caudal
elements
3
3
4
23.2
12+ 100
90
9
8
7
5
17
22
76
74
57
5
26.1
13 + 99
90
13
8
7
5
21
26
76
74
57
5
28.9
13 + 99
91
9
8
7
7
22
29
78
76
57
a
188
13 + 94
90
20
8
7
7
22
29
b
b
b
b
190
13 + 96
90
20
8
7
7
23
30
b
b
b
b
may start as early as 9.4 mm with the full comple-
ment (8 rays) present by 16.3 mm. Ossification
proceeds from the outer to the innermost rays.
The second dorsal fin anlage is visible in larvae
of 5.7 mm. Bases are first visible by 6.9 mm, with
ray ossification commencing by 7.3 mm. Al-
though the anal fin anlage does not form until 6.9
mm, complete ossification is reached before that
of the second dorsal. Distinct anal fin bases are
first visible in 7.2 mm larvae and ossification has
consistently begun by 9.9 mm. The full comple-
ment of anal rays is present by 21.0 mm and for
the second dorsal, by 23.2 mm.
The first dorsal starts development slightly
later than the second dorsal, although it is the
first fin to complete ossification. The full comple-
ment of 12 or 13 elements is present by 16.3 mm.
Pectoral buds were first observed in larvae 4.5
days posthatch (3.2 mm). However, the pectoral
fin is the last to complete development. Ossifica-
tion of pectoral rays starts by 16.3 mm; a 34.2 mm
specimen had only 15 ossified rays, still short of
the 20 rays of juveniles. Sequence of ossification
is from upper to lower.
The caudal fin anlage first appears on the ven-
tral surface of the notochord just anterior to the
tip in larvae of 10.4 mm. Flexion begins at 20 mm
and is usually complete by 28 mm. Ossification of
all caudal elements was incomplete in a 34.2 mm
specimen. Insufficient material of the appropriate
size was available to define the completion of cau-
dal ossification.
The caudal complex in M. novaezelandiae is
based on two ural centra, two epurals, a superior
hypural (HP3 + 4), inferior hypural (HPl + 2),
and a parhypural (Fig. 3). Eight to nine rays ar-
ticulate with these elements — one or two rays on
the second epural, three rays on the superior hy-
pural, two on the inferior hypural, and one ray
each on the first epural and the parhypural. Sin-
gle rays also articulate with the elongate neural
and haemal spines of the first preural centrum. X
and Y bones are present although they are not
readily distinguishable from dorsal and anal
pterygiophores. Total caudal fin ray counts are
low (12 or 13).
Additional caudal structures occurred in one of
the six specimens examined. This specimen had a
twin haemal spine on the first preural centrum
and greatly elongated haemal spines on preural
centra 3-8 (1.3-1.4 times the length of corre-
sponding neural spines, Fig. 3).
DISCUSSION
The general morphology and pigmentation of
M. novaezelandiae larvae show broad similarities
to Merluccius and to gadine gadids. Characteris-
tic differences between M. novaezelandiae and
Merluccius species occur in fin structure and the
sequence of fin development. In Merluccius, the
caudal fin is the first to form, followed by the
pelvic. In Macruronus, caudal development is
late with the caudal fin being the second last to
126
BRUCE: LARVAL DEVELOPMENT OF BLUE GRENADIER
Additional
Haemal Spine
Figure 3. — Caudal osteology of a juvenile Macruronus novaezelandiae (181 mm SL).
X = X bone, Y = Y bone, EP = epural, SH = superior hypural (hypurals 3 + 4), IH
= inferior hypural (hypurals 1 + 21, PH = parhypural, U = ural centra, PU = preural
centra.
form. The pectoral fin in Macruronus larvae is
more markedly stalked than in Merluccius.
Fahay and Markle ( 1984) suggested that this pec-
toral modification in larvae with delayed caudal
development may be a compensatory response as-
sociated with swimming.
Although the larvae of the remaining mer-
lucciid genera (Lyconus and Lyconodes) are cur-
rently unknown, fin structure and position should
be useful in separating these from Macruronus.
Based on adult features, pelvic insertion should
distinguish Macruronus (pelvics inserted behind
pectorals) from Lyconus (opposite) and Lyconodes
(abdominal). Additionally, Lyconus has only a
single dorsal fin and no caudal fin.
The caudal fin of M. novaezelandiae is similar
to Muraenolepis in its confluence with dorsal and
anal fins (Fahay and Markle 1984). This similar-
ity extends to the undifferentiated X and Y bones
and the total caudal fin ray count (12 or 13) re-
ported by these authors. However, unlike Mu-
raenolepis, M. novaezelandiae has radials fused to
the spines of the first preural centrum, which is
the more typical gadoid condition.
Variability in the structure and appearance of
bones associated with the caudal fin has been re-
ported for other Macruronus species. Marshall
( 1966) observed double neural arches and "super-
numary elements" in M. magellanicus. Indeed,
variability in gadiform caudal structure ap-
pears not to be unusual with examples in several
taxa (Markle 1982). Unfortunately, insufficient
specimens in the appropriate 35-150 mm size
range were available to assess developmental
characteristics of these variations in blue
grenadier.
ACKNOWLEDGMENTS
I thank R. Thresher, J. Gunn, J. Leis, and A.
Miskiewicz for their reviews of the manuscript. I
also thank D. Furlani for sorting the samples and
for her considerable patience in the laboratory.
This work was supported by a grant from the
Fisheries Industry Research Trust Account.
NOTE: Since the acceptance of this paper, the
embryological work by A. Patchell (see section on
Development) has been published in New
Zealand Journal of Marine and Freshwater Re-
search Vol. 21, No. 2. That paper includes a simi-
lar larval developmental sequence to that re-
ported here.
LITERATURE CITED
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 north-
west Atlantic. Bull. Mar. Sci. 26:285-402.
Badcock, J R . AND N R Merrett
1976. Midwaterfishes in the eastern North Atlantic. I.
Vertical distribution and associated biology in 30° N, 23°
W, with developmental notes on certain myctophids.
Frog. Oceanogr. 7:3-58.
Baker. A C, M R Clarke, and M. J Harrls
1973. The N.I.O. combination net (RMT 1 + 81 and further
developments of rectangular midwatertrawls. J. Mar.
Biol. Assn. U.K. 53:167-184.
127
FISHERY BULLETIN: VOL. 86, NO, 1
Cohen. D M
1986. Merlucciidae. /a; M. M. Smith and P. C. Heemstra
(editors). Smiths' sea fishes, p. 324-326. Springer-
Verlag, N.Y.
Dunn. J R . and A C Matarese
1984. Gadidae: Development and relationships. In H. G.
Moser, W. J. Richards, D. M. Cohen, M. P. Fahay, A. W.
Kendall, Jr., and S. L. Richardson (editors). Ontogeny
and systematics of fishes, p. 283-299. Am. Soc. Ichth.
Herpet. Spec. Publ. 1.
Dunn. J R . and B M Vinter.
1984. Development of larvae of the safiron cod, Eleginus
gracilis, with comments on the identification of gadid
larvae in Pacific and Arctic waters contiguous to Canada
and Alaska. Can. J. Fish. Aquatic Sci. 41:304-318.
Fahay. M P . and D F Markle
1984. Gadiformies: Development and relationships. In
H. G. Mo.ser. W. J. Richards, D. M. Cohen, M. P. Fahay,
A. W. Kendall, Jr., and S. L. Richardson (editors), On-
togeny and systematics of fishes, p. 265-283. Am. Soc.
Ichth. Herpet. Spec. Publ. 1.
Gordon. D J . D F Markle, and J E. Olney.
1984. Ophidiiformes: Development and relationships.
In H. G. Moser, W. J. Richards, D. M. Cohen, M. P. Fahay,
A. W. Kendall, Jr., and S. L. Richardson (editors). On-
togeny and systematics of fishes, p. 308-319. Am. Soc.
Ichth. Herpet. Spec. Publ. 1.
Hay, D E
1981, Effects of capture and fixation on gut contents and
body size of Pacific herring larvae. Rapp. P. -v. Reun.
Cons. int. Explor. Mer 178:395-400.
Heron. A C
1982. A free fall plankton net with no mouth obstruc-
tions. Limnol. Oceanogr. 27:380-383.
Hollister, G.
1934. Clearing and dyeing fish for bone study. Zoologica
12:89-101.
Inada. T
1981. Studies on merlucciid fishes. Bull. Far Seas Fish.
Res. Lab. 18:1-172.
Marak.R R
1967, Eggs and early larval stages of the offshore hake,
Merluccius albidus. Trans Am. Fish. Soc. 96:227-228.
Markle, D F
1982. Identification of larval and juvenile Canadian At-
lantic gadoids with comments on the systematics of gadid
subfamilies. Can. J. Zool. 60:3420-3438.
Marshall. N B
1966. The relationships of the anacanthine fishes,
Macruronus, Lyconus and Steindachneria. Copeia
1966:275-280.
Marshall, N B , and D M Cohen.
1973. Order Anacanthini (Gadiformes). Fishes of the
western North Atlantic. Mem. Sears Found. Mar. Res.
l(6):479-495.
Marshall, N, B,, and T Iwamoto,
1973, Family Macrouridae, In D. M. Cohen (editor).
Fishes of the western North Atlantic, Part 6, p. 496-
665. Mem. Sears Found. Mar. Res. Yale Univ. 1.
Matarese, A C . S L Richardson, and J R Dunn
1981. Larval development of Pacific tomcod, Microgadus
proximus, in the northeast Pacific Ocean with compara-
tive notes on larvae of walleye pollock, Theragra chalo-
gramnia, and Pacific cod, Gadus macrocephalus (Gadi-
dae). Fish. Bull., U.S. 78:923-940.
Monod, T
1968. Le complexe urophore des poissons teleosteens.
Mem. Inst. Fond. Afr. Noire. 81:1-705.
Patchell, G J
1982. The New Zealand hoki fisheries 1972-82. Fish.
Res. Div. Occas. Publ. 38, 23 p.
Svetovidov, a N
1948. Fauna of the USSR, Fishes, Gadiformies. Zool.
Inst. Acad. USSR 9(4):l-222.
Wilson, M A
1981. Blue grenadier spawning grounds. Fintas 4:9-10.
1982. Spawning blue grenadier caught off Cape Sorell.
Fintas 4:13.
128
THE DISTRIBUTION, ABUNDANCE, AND TRANSPORT OF
LARVAL SCIAENIDS COLLECTED DURING WINTER AND EARLY SPRING
FROM THE CONTINENTAL SHELF WATERS OFF WEST LOUISIANA^
James H Cowan, Jr.^ and Richard F. Shaw^
ABSTRACT
The larvae of six species of Sciaenidae were collected in continental shelf waters off west Louisiana
on five midmonthly ichthyoplankton cruises from December 1981 to April 1982. Ranked in order of
abundance these species were sand seatrout, Cynoscion arenarius; Atlantic croaker, Micropogonias
undulatus; spot, Leiostomus xanthurus; black drum, Pogonias cromis; southern kingfish, Menticir-
rhus amencanus; and banded drum. Larimus fasciatus. Total larva density was highest in April, and
the high densities were associated with the coastal boundary layer, a horizontal density front caused
by an intrusion of fresher water onto the inner shelf that probably issued from the Atchafalaya River
east of the study area. Spawning by sand seatrout began in January, two months earlier than
previously reported, and first occurred offshore of midshelf but moved shoreward as the season
progressed. Analysis of length-frequency data suggest that spot probably began to spawn in Novem-
ber, one month earlier than once thought. Both sand seatrout and Atlantic croaker larvae were
captured at higher rates at night than during the daytime. Sand seatrout larvae appear to be
somewhat surface oriented while spot may undergo vertical migration. Interpretation of the sciaenid
data support a previously developed transport hypothesis involving gulf menhaden larvae and west-
northwest alongshore advection within and just outside of a horizontally stratified coastal boundary
layer.
Members of the perciform family Sciaenidae are
an important sport and commercial fishery re-
source along the United States coast of the Gulf of
Mexico and are perhaps the most prominent
group of northern Gulf inshore fishes. Sciaenids
exceed all other families in numbers of species
(18) and in numbers of individuals or biomass;
they are among the top four families with Mugili-
dae, Engraulidae, and Clupeidae (Gunter 1938,
1945; Moore et al. 1970; Franks et al. 1972; Hoese
and Moore 1977). Of the six species of sciaenids
captured during this study, only the banded
drum, Larimus fasciatus, is not commonly sought
by both sport and commercial fishermen.
Many of Louisiana's sciaenids spawn in coastal
or offshore waters. They have pelagic eggs and
young which are then transported into estuaries
(Johnson 1978 for review). The seasonal impor-
tance of Louisiana's estuaries as nursery grounds
'Louisiana State University Contribution No. LSU-CFI-86-
08.
^Center for Wetland Resources, Louisiana State University,
Baton Rouge, LA 70803-7503; present address: Center for Envi-
ronmental and Estuarine Studies, University of Maryland,
Chesapeake Biological Laboratory, Box 38, Solomons, MD
20688-0038.
^Center for Wetland Resources, Louisiana State University,
Baton Rouge, LA 70803-7503.
for postlarval and juvenile sciaenids is well docu-
mented (Cowan 1985 for review), and several
summary works are available which contain tax-
onomic and biological information on adult
sciaenids (Pearson 1929; Suttkus 1955; Guest and
Gunter 1958; Hoese and Moore 1977; Johnson
1978; Powles and Stender 1978; Barger and John-
son 1980; Barger and Williams 1980; Mercer
1984a, b). In contrast, there is little information
about sciaenid ichthyoplankton assemblages in
Gulf continental shelf waters, their offshore and
coastal distribution, or the oceanic current sys-
tems which influence their estuarine recruit-
ment.
This study provides such early life history in-
formation by determining larva distribution,
abundance, and length frequency; by document-
ing spawning location (depth and distance from
shore) of winter and early spring-spawned
sciaenids off west Louisiana; and by analyzing
larval sciaenid distribution with respect to known
water circulation patterns and a larval gulf men-
haden, Brevoortia patronus, transport hypothesis
in the shelf waters of the northwestern Gulf of
Mexico (Shaw et al. 1985b). Recruitment implica-
tions of the observed distribution, larva age struc-
ture, and transport of sciaenids in Louisiana
waters are also discussed.
Manuscript accepted September 1987.
FISHERY BULLETIN: VOL 86. NO. 1, 1988.
129
FISHERY BULLETIN: VOL 86. NO 1
METHODS AND MATERIALS
Detailed sampling methodology has been pre-
sented elsewhere (Shaw et al. 1985 a, b). Briefly,
larval sciaenids were collected off west Louisiana
on a sampling grid consisting of 37 stations on 5
transects (Fig. 1) during 5 midmonthly cruises
from December 1981 to April 1982. Ichthyoplank-
ton samples were analyzed from the 335 |xm mesh
net side of an opening and closing, 60 cm, paired
"bongo type" plankton sampler fitted with Gen-
eral Oceanics' flowmeters (model no. 2030). Most
plankton collections (125 of 187 total) consisted of
10-min stepped oblique tows from near bottom to
surface. Nets were set closed and opened just
prior to the stepped ascent. Each tow had five
steps with a retrieval rate between steps of 20
m/minute; towing speed was about 1 m/second
(2 knots). The object of the 10-min tow was to
filter approximately 100 m*^ of water. This process
increased the water volume filtered per unit
depth at the shallow stations relative to deeper
stations. This discrepancy is acceptable since the
alternative would be to compare 17-s shallow-
station oblique tows with 9-min deep-station tows
at a uniform retrieval rate (Houde 1977). At se-
lected stations (A-3, 6, 9; B-1; C-6; D-1; E-3, 6, 9;
Fig. 1), only 10-min simultaneous surface and
near-bottom horizontal tows (31 surface and 31
near-bottom) were made to determine if sciaenid
larvae were vertically stratified. Larva total
length (TL) was measured to the nearest 0.1 mm.
Larva densities are reported as standardized
catch rates at a station (density = larvae/100 m"^).
A four-way analysis of variance (ANOVA) was
performed on logio transformed [(no. larvae/100
m^) + 1] data to determine the spatial (vertical
and horizontal), temporal, and diel patterns of
species density and distribution. The four main
effects tested were month (January-April); sta-
tion depth group (d.g.) (d.g. 1 < 10 m, 10 m < d.g.
2 < 14 m, 14 m < d.g. 3 < 24 m and d.g. 4 > 24 m);
day-night (2000 hours < night < 0500 hours);
and horizontal tow type (surface vs. near-bottom).
Data from the December cruise were not included
as only the A transect was completed due to ad-
verse weather conditions.
Two methods of current estimates were utilized
(following Shaw et al. 1985b): 1) instantaneous
current profiles taken at each station and
2) continuous surface and near-bottom current
■♦Reference to trade names does not imply endorsement by the
National Marine Fisheries Services, NOAA.
meter measurements at two sites (H and S; Fig.
1). The instantaneous number of larvae trans-
ported on each transect was calculated by using
the equation D x U x M = number of larvae per
meter per second where D = larva density
darvae/m') from either oblique tows or from the
mean of the horizontal tows (i.e., average of sur-
face and near-bottom catch rates), U = depth-
averaged water velocity (m/s) determined from
instantaneous current meter profiles at each sta-
tion, and M = water depth (m) at each station.
Distribution diagrams and length-frequency
histograms were generated for each cruise for the
three most abundant sciaenid species. Inspection
of these data along with current measurements
allowed a comparison with the previously men-
tioned transport hypothesis.
RESULTS AND DISCUSSION
Total Sciaenids
A total of 5,225 larval sciaenids accounted for
9.1% of the fish larvae collected. In December
through February, samples were dominated by
Atlantic croaker, Micropogonias undulatus, and
spot, Leiostomus xanthuriis. In March and April
samples contained mostly sand seatrout,
Cynoscion arenarius. In all, six species of sciaenid
larvae were collected: sand seatrout (N = 4,100);
Atlantic croaker {N = 567); spot (A^ = 264); black
drum, Pogonias cromis {N = 68); southern king-
fish, Menticirrhus americanus (N = 53); and
banded drum (A'^ = 13). Additional Menticirrhus,
not identifiable to species, accounted for 160 more
specimens (Table 1). A more detailed examina-
tion of the data on the three most abundant
sciaenid species follows.
Sand seatrout,
Cynoscion arenarius
A total of 4,100 sand seatrout larvae was col-
lected making it the most abundant sciaenid
taken during the study. Larval sand seatrout den-
sities were highest in April (Table 1) with a mean
of 46.1 larvae/100 m"^; mean density in February
and March was 0.3 and 2.9/100 m'^, respectively,
and 1 larva was collected in January. Larvae
were distributed mostly over the midshelf in
February but highest concentrations were later
found inshore and towards the east (Fig. 1). Over
the course of study, larvae were found in temper-
atures and salinities ranging from 14° to 21°C and
130
COWAN AND SHAW: LARVAL SCIAENIDS COLLECTED OFF WEST LOUISIANA
Figure l. — Density distribution of sand seatrout, Cynoscion arenarius, larvae by month,
February-April 1982. Densities are as follows: o = 0: • >0-10; • >10-50; • >50-99; •
>99-250; 0 250 100 m3 of water filtered from all plankton tow types. Also shown is the
station sampling grid with moored current meter sites (H and S) and selected iso-
baths. SP = Sabine Pass, CR = Calcasieu River, and MR = Mermentau River.
131
KISHKRY BULLETIN: VOL, 86, NO. 1
Table 1. — Total number, months of occurrence and monthly density of
sciaenid larvae collected in west Louisiana shelf waters from December 1981
to April 1982.
Months of
occurrence and
Total
number
density (No. /1 00
m3)
Taxa
Dec.
Jan.
Feb.
IVIar.
Apr.
Cynoscion arenarius
4,100
—
1 larva
0.3
2.9
46.1
Micropogonias undulatus
567
20
2.2
32
22
02
Leiostomus xanthurus
264
7.8
1.3
0.9
0.1
—
Menticirrhus sp.
160
—
—
—
—
19
Pogonias cromis
68
—
1 larva
0.5
0.2
0.2
Menticirrhus americanus
53
—
—
—
—
0.6
Larimus fasciatus
13
—
—
—
—
0.2
from 15 to 36 ppt (Table 2) and at station depths
ranging from 5 to 70 m, but most were collected
inside the 18 m isobath. Larval sand seatrout
density increased in April with many stations ex-
hibiting densities in excess of 250 larvae/100 m'^;
the larvae appeared to be associated with a
freshet of water on the shelf, probably issuing
from the Atchafalaya River east of the study area
(Shaw et al. 1985a). The presence of riverine
runoff on the shelf, which was most evident in
March and April 1982, caused the development of
an oceanic salinity front referred to as the coastal
boundary layer, 10-35 km from shore (Wiseman
et al. 1987).
Observed spawning seasonality and location
for sand seatrout are in part consistent with pre-
viously published information (Shlossman and
Chittenden 1981 for review). The presence of a 4
mm TL larva in January indicates some spawn-
ing had taken place at least 2 months earlier than
Table 2. — l^onthly data summaries at time of capture for three sciaenid larvae
(Cynoscion arenarius, Micropogonias undulatus. and Leiostomus xanthurus) col-
lected in west Louisiana shelf waters from December 1981 to April 1982.
Total
Species'month number
Length range
(mm TL)
Temperature
range ( C)
Salinity
range
(ppt)
Depth
range
(m)
Cynoscion arenarius
January 1
4
14
35
18
February 20
2.5-4.5
(mode = 2-3)
14-20
34-36
15-70
March 203
1.5-10.5
(mode = 2-3)
14-18
25-36
5-40
April 3,876
1.5-20.5
(mode = 2-3)
20-21
15-36
5-70
Micropogonias undulatus
December 28
2.5-10.5
(mode = 3-4)
12-20
30-36
10-65
January 158
2.5-10.5
(mode = 4-5)
10-18
30-36
5-70
February 221
2.5-17.5
(mode = 14-15)
11-17
27-36
5-40
March 144
2.5-19.5
(mode =14-15)
14-20
25-36
5-115
April 16
11.5-18.5
(mode= 17-18)
20.5
22
7
Leiostomus xanthurus
December 110
2.5-7.5
(mode = 3-4)
14-18
30-36
16-65
January 89
2.5-13.5
(mode = 4-5)
10-18
30-36
5-40
February 62
3.5-15.5
(mode= 12-13)
10-17
28-36
5-40
March 3
3.5-16.5
14-17.5
26-36
11-40
132
COWAN AND SHAW: LARVAL SCIAENIDS COLLECTED OFF WEST LOUISIANA
previously reported. Monthly length-frequency
data for sand seatrout show that larvae as large
as 11 mm TL were first present in March samples
(Fig. 2A). Based on the estimated growth rate
determined for sand seatrout (Cowan 1985; Shaw
et al., in press), an 11 mm larva could be as old as
65 days; this further supports January spawning.
Sand seatrout are reported to spawn from March
to August, during two discrete periods — one in
March-May, the other August-September, with
little spawning between the two peak periods
(Hoese 1965; Daniels 1977; Shlossman and Chit-
tenden 1981).
An examination of distribution and length-
frequency data (Figs. 1, 2A) suggests that most
spawning initially took place in midshelf to off-
shore waters at depths ranging from 15 to 80 m or
to about 175 km from shore. As the season pro-
gressed into March and April, spawning location,
as determined by the presence of larvae <3.0 mm
TL, was more inshore (5-18 m) with few small
larvae occurring at depths >25 m.
Other than the indication that spawning may
move from offshore to inshore waters as the sea-
son progresses, this spatial information agrees
with the limited life history data available on
sand seatrout. Most spawning has been shown to
occur in the shallow waters of the Gulf of Mexico,
primarily between 7 and 15 m in depth (Gunter
1945; Moffet et al. 1979; Shlossman and Chitten-
den 1981). Running ripe C. arenarius have been
captured in deepter waters (70-90 m) in February
and March, but no spawning was indicated
(Franks et al. 1972; Perry 1979).
In a four-way ANOVA employed to determine
patterns of larval sand seatrout density and dis-
tribution, month was a highly significant main
effect (P < 0.01; Table 3) reflecting spawning sea-
sonality and the magnitude of the density in-
crease in April. The test for interaction between
month and day-night was employed to determine
if daytime gear avoidance was evident as size and
mobility of larva increased. Most sand seatrout
collected, however, were small and no clear
monthly modal increase in larva size was evident
(Fig. 2A). The significant interaction (P <0.01)
was probably due to an increased catch in oblique
tows at night as the season progressed (0.0 in
January, 64.9/100 m"^ in April). The significant
interaction between month and depth group and
the highly significant depth group main effect
(P < 0.01) represents the shift in larva concentra-
tion from midshelf early in the study, to a more
coastal distribution in March and April (Fig. 1).
Mean larva density was greatest in depth group 1
(23.7/100 m'^) followed by depth groups 2, 3, and 4
(12.7, 9.2, and 0.3/100 m^, respectively). The third
main effect, day vs. night tows, was highly sig-
nificant (P < 0.01); many more sand seatrout
larvae were collected at night (averge catch rates
in all night (74) tows combined = 21.6/100 m"^
vs. day (113) tows = 7.4/100 m^). Highest night-
time catches occurred in oblique (49) tows (26.9/
100 m'^) while the day-oblique-catch rate aver-
aged 7.6/100 m'^ in 76 tows. Overall, average
catch rate was highest in oblique (125) tows
(14.6 larvae/100 m-^), followed by surface (31)
tows (9.2/100 m^), and then bottom (1.9/100 m^;
31 tows). Intrepretation of the data suggests that
Table 3. — Summary data from four-way analysis of variance done
on logio transformed [(larvae 100 m3) + 1] data from ichthiyoplank-
ton samples collected from January to April 1982. Tfie results are
for A. Cynoscion arenarius. B, Micropogonias undulatus, and C.
Leiostomus xanthurus. The four main effects tested were months
(Jan. -Apr.), station depth group (d.g. 1 < 10 m, 10 m < d.g. 2 < 14
m, 14 m < d.g. 3 < 24 m and d.g. > 24 m), day - night (2000
hours s night < 0500 hours) and horizontal tow type (surface vs.
near-bottom).
Source
df
PR
r2 = 0.75
A. Dependent variable:
Log 10 [(Cynoscion arenarius 1^00 m3) + 1]
Model 21 O.OOOr*
r^^onth 3 0.0001"
Depth group 3 0.0001"
Day-night 1 0.0026"
Horizontal tow type 1 0.2574 (NS)
IVIonth vs. Day-night 3 0.0001"
l\/lonth vs. Depth group 9 0.0001"
Day-night vs. Tow type 1 0.4180 (NS)
Error 177
Corrected Total 1 98
B. Dependent variable:
Logio [(Micropogonias undulatusHOO m3) + 1]
Model 21 0.0001" r2 = o.63
Month 3 0.0045"
Depth group 3 0.3551 (NS)
Day-night 1 0.0001"
Horizontal tow type 1 0.4448 (NS)
Month vs. Day-night 3 0.2168 (NS)
Month vs. Depth group 9 0.0001"
Day-night vs. Tow type 1 0.1288 (NS)
Error 177
Corrected Total 1 98
C. Dependent variable:
Logio [(Leiostomus xanthu^us/^00 m3) + 1]
Model 21 0.0001" /-s = 0.51
Month 3 0.0001"
Depth group 3 0.0033"
Day-night 1 0.1875 (NS)
Horizontal tow type 1 0.3216 (NS)
Month vs. Day-night 3 0.2138 (NS)
Month vs. Depth group 9 0.0001"
Day-night vs. Tow type 1 0.0324*
Error 177
Corrected Total 198
* = Statistically significant (P < 0.05).
" = Highly significant (P < 0.01).
(NS) = Not significant.
133
FISHERY BULLETIN: VOL. 86, NO. 1
00
CM
II
O
ill
Q
CD
O
!• I ■ I ■ I ' I ' I
8Q Q O p O o
ift ^ r^ 04 ^
00
U1
II
z
z'
<
-J
C
L
[
r°
i
E
CM
CM
II
z
CD
S 2
I ■ I ■ I ■
S § CO
8 2
8
II
z
m'
K
to
II
z
m
[
[
— I—
o
-T —
O
i:
[
<
z
o
Cvj
o
CM
II
z
if
<
S
r ■ I ■ 1
So Q O p O O
u^ ^ r-1 "N --
w'::.
r"
^Kis;.
( :....
r:::„
r
r— ' — 1 — •— I — •"
r"
e
II
z
of
<
Z
' E
(0
o
II
z
(0
C
2 I
UJ o
-I w
-I 9
I-
O
nVlOl dO iN30a3d
II
z
of
OL
<
o
•o
c
(0
3
c
3
(O
c
o
en
o
&
CD
oj
3
C
n:
c
F
QJ "O
fe iJ
F
^
S wi
Z
O
.O CD
U 0)
7
5 "O
UJ
-J
O^S
-1
< -c
<
O)
H
D
O
Cvi S
1-
CO "
^ o
^
^ E
Q- D
< c
j|_
(U
£ ,
E
0)
O
Q)
a
(0
■o
c:
a>
CO
O
(0
134
COWAN AND SHAW: LARVAL SCIAENIDS COLLECTED OPT WEST LOUISIANA
sand seatrout larvae v^'ere somewhat surface
oriented.
Atlantic croaker,
Micropogon ias undiilatus
The second most abundant sciaenid taken was
Atlantic croaker (A'^ = 567). Larval Atlantic
croaker density was greatest in February at 3.2/
100 m'^, but density was relatively constant from
December through March (Table 1). Mean densi-
ties for December, January, March, and April
were 2.0, 2.2, 2.2, and 0.2/100 m^, respectively.
Atlantic croaker was the only sciaenid collected
in all months of the study. Their overall distribu-
tion (all sizes combined) was generally uniform
over most of the shelf (Fig. 3) except in March and
April when they were more often found inshore.
Recently spawned larvae (<3.0 mm TL) were also
collected over much of the shelf at station depths
ranging from 15 to 115 m or from about 20 to 200
km from shore. However, most small larvae were
collected near midshelf about 65-125 km from
shore. In December and January the majority of
the larvae were small. By April, no recently
spawned individuals were collected (Fig. 2B). For
the study overall, larvae were found in salinities
and temperatures ranging from 22 to 36 ppt and
from 10 to 20.5°C (Table 2).
Spawning by Atlantic croaker in Gulf of Mexico
waters is reported to occur from September to
March, with a distinct peak in October (Hoese
1965; Sabins and Truesdale 1974; White and
Chittenden 1977; Benson 1982) and to occur pri-
marily offshore over a wide area (Pearson 1929;
Hildebrand and Cable 1930; Wallace 1940; Haven
1957; Bearden 1964; Hoese 1965; Nelson 1967).
Atlantic croaker larvae, however, have been
taken on the outer continental shelf off Texas
from September to May (Finucane et al. 1979).
As with sand seatrout, a four-way ANOVA was
used to determine patterns in larva density and
distribution (Table 3). Larval Atlantic croaker
density by month was a highly significant main
effect (P <0.01). Densities at the end of their
spawning period were low, increased only slightly
in February, and then dropped off by April (Table
1). The interaction between month and day-night
was not significant. The highly significant inter-
action between month and depth group was not
surprising. Larvae were in more offshore waters
early in the study while later becoming more
abundant inshore (Fig. 3). However, as a main
effect, depth gi'oup was not significant. Larval
Atlantic croaker mean densities for depth groups
1 through 4 were 3.9, 0.8, 0.4, and 0.7/100 m^,
respectively. Day-night, as a main effect, was
highly significant (P < 0.01). Larval Atlantic
croaker density was over 5 times higher at night
(all tow types combined) than during the day (3.7
vs. 0.7/100 m'^). However, the interaction between
day-night and horizontal tow type was not signif-
icant. The fourth main effect tested, horizontal
tow type, was not significant. Average catch rates
at the surface and near-bottom were similar (1.0
and 1.8/100 m'\ respectively).
Spot, Leiostomiis xcmthurus
The third most abundant sciaenid collected was
spot (A^ = 264). Density of spot larvae was highest
in December at 7.8 larvae/100 m'^ (Table 1). How-
ever, the high December value must be viewed
with a consideration of the abbreviated cruise
track for that month and the resultant reduction
in spatial coverage. Mean densities for January to
March were 1.3, 0.9, and 0.1/100 m"^, respectively.
No spot larvae were collected in April. In general,
larva density was low and their distribution was
uniform over the shelf out to the 40 m isobath,
about 130 km offshore (Fig. 4). Spot were col-
lected in temperatures and salinities ranging
from 10° to 18°C and from 26 to 36 ppt (Table 2),
and at stations with depths ranging from 5 to 65
m.
Larvae >7 mm TL in our mid-December (Fig.
2C) collections and small larvae (<3.0 mm TL) in
all but the last cruise indicate that spawning
probably began by at least November and contin-
ued through March. Spawning occurred from
near midshelf (about 65 km) out to 175 km from
the coast. Data presented here partly concur with
previously published information on spot spawn-
ing periodicity. In the northern Gulf, spawning
reportedly occurs from late December to March,
peaking in January, and takes place well offshore
in moderately deep water (Pearson 1929; Kilby
1955; Townsend 1956; Dawson 1958; Springer
and Woodburn 1960; Pacheco 1962; Nelson 1967;
Joseph 1972; Music 1974; Sabins and Truesdale
1974).
A four-way ANOVA indicated that month, as a
main effect for spot larvae, was highly significant
(P < 0.01 ), which probably reflects the decreasing
catch rates seen from January to March (Table 3).
The interaction between month and depth group
was also highly significant (P < 0.01) as was
depth group as a main effect. Larval spot
135
FISHERY BULLETIN VOL 86. N(J 1
DEC
-yT ,4^-
-^
^?^IV^
° • :
--^.^^^^
o
o
0 o
.
-
o
"
o
o
0
o
o
o o o
o
o
o
o
c
o
o
-
MAR
1
o
1
APR
Figure 3— Density distribution of Atlantic croaker, Micropogonias undulatus, larvae by month, December 1981— April 1982. Densities are
as follows: o = 0; • >0-10; • > 10-50; • >50;-99; • >99-250; # >250/100 m3 of water filtered from all plankton tow types.
136
COWAN AND SHAW: LARVAL SCIAENIDS COLLECTED OFF WEST LOUISIANA
i.Tr'if"
1
^u^
*^^\
-^ 0 o ■
o
0
o
° O O .
o
O 0 o
o o
o
o
0
o
FEB
o
1 1
o
o
■ o o 0
o o
o o
o
o
0
o
o
O 0
o o o
o
o
.
o
o
0
o
o
o
o
1
MAR
o
Figure 4 — Density distribution of spot, Leiostomus xanthurus. larvae by month, December
1981-Marcti 1982. Densities are as follows: o = 0; ■ >0-10; • > 10-50; • >50-99; • >99-
250; 9 >250/100 m3 of water filtered from all plankton tow types.
137
FISHERY BULLETIN: VOL. 86, NO. 1
densities were higher offshore in the early part of
the study and then greater inshore during Febru-
ary and March. Depth group 4 had the highest
mean density (1.3/100 m-'^) followed by depth
groups 1 and 3 (0.4/100 m'^ each) and depth
group 2 (0.1/100 m^). Day-night comparisons
proved nonsignificant as a main effect for spot
larvae. The average catch rates for all day and
night tows were identical (0.5/100 m^). However,
the interaction between day-night and horizontal
tow type was statistically significant (P < 0.05).
In this case, vertical migration and stratification
may be indicated. Average catch rate of spot lar-
vae during the day at the surface was 0.1/100 m^,
while near the bottom it averaged 1.6. Con-
versely, nighttime average catch rate at the sur-
face was 1.0/100 m-^ while the near-bottom rate
averaged 0.04. These are very low densities but
the vertical differences are an order of magnitude
and their reversing pattern suggests that spot
larvae were stratified and undergoing diel verti-
cal migration. Daytime bottom and nighttime
surface average catch rates were higher than for
oblique (O) tows (day, 0 = 0.45/100 m^; night,
O = 0.49/100 m^). Average catch rates for surface
and near-bottom tows, regardless of time of day,
were 0.5 and 0.9/100 m\ respectively.
As previously mentioned, other larval sciaenid
species (i.e., black drum, banded drum, southern
kingfish) were collected during these cruises.
However, relatively few individuals were cap-
tured (Table 1), making information on their dis-
tribution inconclusive.
TRANSPORT ANALYSIS
Alongshore advection within and just outside
the coastal boundary layer in the northwestern
Gulf of Mexico has been hypothesized as the
major mechanism transporting gulf menhaden
larvae to the estuaries in western Louisiana,
rather than across-shelf transport from directly
offshore. In contrast, such direct across-shelf
transport has been demonstrated for sciaenids
and other species along the U.S. mid-Atlantic
coast (Nelson et al. 1976; Norcross and Austin
1981; Miller et al. 1984). The data collected for
sciaenid larvae (all species combined) were exam-
ined in light of this Gulf hypothesis. Larval
sciaenid densities were less than those for gulf
menhaden but similarities in distribution were
evident. Both larval sciaenid and gulf menhaden
densities were highest at midshelf early in the
study. By March and April the highest densities
were found towards the east and inshore and were
associated with a horizontal density front (coastal
boundary layer) caused by an intrusion of fresher
water onto the shelf.
The along-transect length-frequency patterns
exhibited by larval sciaenids and gulf menhaden
were also similar. No apparent increase in size
was seen until gulf menhaden larvae were on the
inner shelf or sciaenids were on the mid- to inner
shelf. The expected pattern of a gradual increase
in larva size from offshore to inshore, which
would result if there were significant across-shelf
(south to north) transport, was not evident in ei-
ther data set. Off the North Carolina coast, War-
len (1981) and Miller et at. (1984) showed that
ages and lengths of both spot and Atlantic
croaker larvae increased systematically toward
shore in an area where winter water currents fa-
vored across-shelf (west to east) transport.
During the winter of 1981-82, moored current
meter data from sites H and S (Fig. 1) indicated
that flow was directed primarily alongshore in
the west-northwest direction. Several researchers
have reviewed the circulation in the northwest-
ern Gulf (Nowlin 1971; Kelly et al. 1982; Crout
1983). It was not until Cochrane and Kelly (1986)
developed their comprehensive circulation model
for the Louisiana-Texas continental shelf, how-
ever, that the ocean current patterns, which led to
the hypothesized larva transport model, were
fully documented. Flow in nearshore coastal
waters is westward all year except in summer
when it usually reverses, while farther offshore
flow is eastward all year (Cochrane and Kelly
1986).
To quantify transport, larval sciaenid densities
were combined with the vertically averaged, in-
stantaneous current measurements. The resul-
tant curves present the number of larvae trans-
ported per unit time at each station (Fig. 5). Early
in the winter, highest sciaenid larva transport
(mostly Atlantic croaker and spot) was located
midshelf. Later (March and April), transport val-
ues were higher inshore and reflected the in-
crease in larval sciaenid density (primarily sand
seatrout). Overall, larva transport was primarily
westward and ranged from about 0.05 to 4.0 lar-
vae/meter per second.
Although the oceanographic data collected
were insufficient to precisely quantify onshore
transport rates, an estimate was obtained by
using the mean current vectors from the near sur-
face meter at site H (Fig. 1) from 24 January to 12
May 1982 (14.33 cm/second alongshore westward
138
COWAN AND SHAW LARVAL SCIAENIDS COLLECTED OFF WEST LOUISIANA
U)
U1
\
■«
n
eg
k
o I—-—
MAR
(/)
r5Csi»-0'-fMr)'*
/
1-' I
»- o »- oj r)
09999
t-8,uj avAbvn
oc
a.
<
3 8 8 3 8 3
009990
-I — ' — »■ o
3
CO
CO
<D
E
c
ID
k_
k.
O
0>
3
O
a>
c
c
CO
c
5 uj
T3 -^
(1)
c
r
0
^
t
F
c
0
OJ
0
Q)
(/>
?.
c
11
0
^
CO O)
<D 01
8 g
CO c
(0 0)
~ a:
^^
O o
E c
o I
I"
o
0)
Q
•o
LU a>
- U5
O S
< 2
o
a.
2 <o
2 a)
<-> 2
<" to
II
5 o
JO (D
t: <2
o ^
Q- 9
<o £
^ CO
li
1 1
to a>
y CD
139
FISHERY BULLKTIN: VOL. 86, NO. 1
and 1.75 cm/second shoreward). Based on that av-
erage shoreward advection rate we calculated
that larvae could be passively transported 98 km
in the onshore direction in 65 days. Examination
of length frequency and age at capture data
(Cowan, in press) suggest that larval Atlantic
croaker arrive in nearshore coastal waters, on the
average, 60-90 days after hatching. Most small,
newly hatched Atlantic croaker larvae were col-
lected approximately 100 km offshore. Although
the onshore component of advective transport is
small in comparison with the average alongshore
component, the estimate of shoreward transport
rate is reasonable when age of larvae is consid-
ered.
CONCLUSIONS-
RECRUITMENT IMPLICATIONS
Across-shelf transport appears to be an order of
magnitude smaller than alongshore advective
transport in the northwestern Gulf shelf waters
during winter and spring. Sciaenid larvae col-
lected offshore in the study area, at midshelf and
beyond, would probably be lost to the estuaries in
western Louisiana. Those offshore larvae would
be transported towards north Texas estuaries, or
back to the east if they were far enough offshore,
since there is evidence for an easterly counter
current (Kelly et al. 1982; Cochrane and Kelly
1986).
Sand seatrout are common in west Louisiana
estuaries (Herke et al. 1984) and were the most
abundant sciaenid larvae collected in this study.
They spawn, in general, more inshore (Fig. 1)
than Atlantic croaker or spot. Conceivably, many
of the sand seatrout collected in the study area
inside the coastal boundary layer on the inner
shelf would have recruited to Louisiana estuaries.
Still, large numbers of postlarval sciaenids,
other than sand seatrout, enter the estuaries in
west Louisiana each year. Atlantic croaker and
spot were the 3rd and 21st most abundant fish, of
117 species collected, in the Calcasieu River
Basin, the largest estuary in west Louisiana
(Herke et al. 1984). However, the distribution and
transport analyses indicate that most spot and
Atlantic croaker larvae directly offshore at least
would not have recruited to the Calcasieu Basin.
Interpretation of these data suggests that the
source of the sciaenid postlarvae shown to season-
ally recruit to the Calcasieu estuaries must be
east of the study area. Interpretation of data sum-
marizing several years of northern Gulf shrimp-
trawl collections suggests that, during the spawn-
ing season, a sufficient concentration of adults
exists to the east of our study area (Darnell et al.
1983).
In the fall and winter, high concentrations of
Atlantic croaker, and to a lesser extent spot, have
been found between the 20 and 40 m depth con-
tours (65 and 125 km offshore) in an area east of
the sampling grid. The area and timing of high
concentration coincides with the reported spawn-
ing location and period for both Atlantic croaker
and spot. If indeed this concentration represents a
spawning distribution, it would help explain why
so few Atlantic croaker and spot larvae, relative
to the number of juveniles seen in estuaries, were
collected in this and previous Gulf of Mexico
ichthyoplankton studies. Unless collections were
made in or near the spawning area, single-station
densities would be low as eggs and larvae were
dispersed. Furthermore, this study demonstrates
the need for understanding both biological (verti-
cal distribution, age and growth, behavior, etc. of
larvae) and physical (ocean currents, estuarine-
shelf exchange, etc.) processes which may influ-
ence estuarine recruitment.
ACKNOWLEDGMENTS
We would like to thank Wm. Wiseman,
L. Rouse, Jr., and S. Dinnel for their discussion
and assistance in interpretation of the physical
oceanographic data. We gratefully acknowledge
E. Turner, J. Geaghan, M. Fitzimmons,
B. Thompson, and W. Herke for critically review-
ing this manuscript.
Funding was provided by a Louisiana Depart-
ment of Wildlife and Fisheries, U.S. Department
of Energy and LSU Center for Wetland Resources
cooperative agreement No. DE-FC96-81P010313.
Additional support was given by the Department
of Marine Sciences, Louisiana Sea Grant College
Program and the Coastal Fisheries Institute.
LITERATURE CITATIONS
Barger, L. E., and a. G. Johnson
1980. An evolution of marks on hardparts for age
determination of Atlantic croaker, spot, sand seatrout,
and silver seatrout. U.S. Dep. Commer., NOAA Tech.
Memo. NMFS-SEFC-22, 5 p.
Barger, L E , and M. L. Williams.
1980. A summarization and discussion of age and growth
of spot, Leiostomus xanthurus Lacepede, sand seatrout,
Cynoscion arenarius Ginsburg, and silver seatrout,
Cynoscion nothus (Holbrook), based on a literature
review. U.S. Dep. Commer., NOAA Tech. Memo.
140
COWAN AND SHAW: LARVAL SCIAENIDS COLLECTED OFF WEST LOUISIANA
NMFS-SEFC-14, 15 p.
Bearden, C. M.
1964. Distribution and abundance of Atlantic croaker,
Micropogon undiilatus, in South Carolina. Bears Bluff
Lab. Contrib. 38, 27 p.
Benson, N G (editor).
1982. Life history requirements of selected finfish and
shellfish in Mississippi Sound and adjacent areas. Biol.
Serv. Program, U.S. Dep. Inter., Fish Wildl. Serv.
FWS OBS-78/12, 314 p.
Cochrane, J D., and F. J Kelly, Jr
1986. Low-frequency circulation on the Texas-Louisiana
continental shelf. J. Geophys. Res. 91(C9): 10,645-
10.659.
CoWAN, J H. jR
1985. The distribution, transport and age structure of
drums (family Sciaenidae) spawned in the winter and
early spring in the continental shelf waters off west
Louisiana. Ph.D. Thesis, Louisiana State Univ., Baton
Rouge, 182 p.
In press. Age and growth of Atlantic croaker, Micro-
pogonias undulatus. larvae collected in the coastal
waters of the northern Gulf of Mexico as determined by
increments in saccular otoliths. Bull. Mar. Sci. 42(3).
Crout, R L
1983, Wind-driven, near-bottom currents over the west
Louisiana inner shelf. Ph.D. Thesis, Louisiana State
Univ., Baton Rouge, 117 p,
Daniels. K
1977. Description, comparison, and distribution of larvae
oi Cynoscion nebulosus and Cynoscion arenarius from the
northern Gulf of Mexico. M.S. Thesis, Louisiana State
Univ., Baton Rouge, 48 p.
Darnell. R N , R E Defenbaugh. and D Moore
1983. Northwestern Gulf shelf bio-atlas; a study of the
distribution of demersal fishes and panaeid shrimp on
soft bottoms on the continental shelf from the Rio Grande
to the Mississippi River Delta. Open File Rep. No.
82-04, Miner. Manage. Serv.. Gulf Mex. OCS Reg. Off.,
Metairie, LA, 438 p.
Dawson, C E
1958, A study of the biology and life history of the spot,
Leiostomus xanthurus Lacepede, with special references
to South Carolina. Bears Bluff Lab. Contrib. 28, 48 p.
Finucane. J H , L A Colllns. L E Barger. and J D
McEachran
1979. Environmental studies of the south Texas outer
continental shelf. Ichthyoplankton'mackeral eggs and
larvae. NOAA Final Rep. to Bur. Land Manage. Wash.,
D.C., 504 p.
Franks. J S.J Y Christmas, W L Siler, R Combs, R Waller,
and C Burns
1972. A study of nektonic and benthic faunas as related to
some physical, chemical and geological factors. Gulf
Res. Rep. 4:1-147.
Guest. W. C, and G. Gunter.
1958. The seatrout or weak fishes (genus Cynoscion) of
the Gulf of Mexico. Gulf States Mar. Fish. Comm. Tech.
Summ. No. 1, 40 p.
Gunter. G
1938. Seasonal variations in abundance of certain
estuarine and marine fishes in Louisiana, with
particular reference to life histories. Ecol. Monogr.
8:313-346.
1945. Studies on marine fishes of Texas. Publ. Inst. Mar,
Sci., Univ. Tex. 1:1-190.
Haven, D S
1957. Distribution, growth, and availability of juvenile
croaker, Micropogon undulatus, in Virginia. Ecology
38:88-97.
Herke. W H , B D Rogers, and J. A Grimes.
1984. The study of the seasonal presence, relative
abundance, movements, and use of habitat types by
estuarine-dependent fishes and economically important
decapod crustaceans on the Sabine National Wildlife
Refuge. La. Coop. Fish. Res. Unit Final Rep., two vols.,
603 p.
HiLDEBRAND. S F . AND L E CABLE
1930. Development and life history of fourteen teleostean
fishes at Beaufort, N. C. Bull. U.S. Bur. Fish. 46:383-
488.
HOESE. H D
1965. Spawning of marine fishes in the Port Aransas,
Texas area as determined by the distribution of young
and larvae. Ph.D. Thesis, Univ. Texas, Austin, 144 p,
HOESE, H. D.. AND R. H MOORE
1977. Fishes of the Gulf of Mexico, Texas, Louisiana and
adjacent waters. Texas A&M Univ. Press, College
Station, 327 p.
HOUDE.E D
1977. Abundance and potential yield of the round
herring, Etrumeus teres, and aspects of its early life
history in the eastern Gulf of Mexico. Fish. Bull., U.S.
75:61-89.
Johnson, G. D.
1978. Development of fishes of the Mid-Atlantic Bight.
An atlas of egg, larval and juvenile stages. Vol. IV,
Carangidae through Ephippidae. Biol. Serv. Program,
U.S. Dep. Inter., Fish Wildl. Serv. FWS/OBS-78/12,
314 p.
Joseph, E B
1972. The status of the sciaenid stocks of the middle
Atlantic coast. Chesapeake Sci. 13:87-100.
Kelly. F J . Jr . J E Schmitz, R E. Randall, and J. D.
Cochrane
1982. Physical oceanography. In R. W. Hann, Jr. and
R. E. Randall (editors). Evaluation of brine disposal from
the Bryan Mound site of the Strategic Petroleum Reserve
Program, p. 1-144. Final report of the 18 month
post-disposal studies. Texas A&M Univ. Res. Found.,
College Station.
KiLBY, J
1955. The fishes of two Gulf coastal marsh areas of
Florida. Tulane Stud. Zool. 2:175-247.
Mercer, L P
1984a. A biological and fisheries profile of spotted
seatrout, Cynoscion nebulosus. N.C. Dep. Nat. Resour.
Community Dev., Div. Mar. Fish., Spec. Sci. Rep. 40,
87 p.
1984b. A biological and fisheries profile of red drum,
Sciaenops ocellatus. N.C. Dep. Nat. Resour.
Community Dev., Div. Mar. Fish., Spec. Sci. Rep. 41,
89 p.
Miller, J M , J P Reed, and L J. Pietrafesa.
1984. Patterns, mechanisms and approaches to the study
of migrations of estuarine-dependent fish larvae and
juveniles. In J. D. McCleave, G. P. Arnold, J. J. Dodson,
and W. H. Neill (editors). Mechanisms of migrations in
fishes, p. 209-225. Plenum Publ. Corp., N.Y.
Moffet, a W., L W. McEachron, and J. G. Key.
1979. Observations on the biology of sand seatrout
(Cynoscion arenarius) in Galveston and Trinity Bays,
141
FISHERY BULLETIN VOL 86, NO. 1
Texas. Contrib. Mar. Sci. 22:163-172.
Moore. D . H A Brusher. and L Trent.
1970. Relative abundance, seasonal distribution, and
species composition of demersal fishes ofT Louisiana and
Texas, 1962-1964. Contrib. Mar. Sci. 15:45-70.
Music. J L.Jr
1974. Observations of the spot (Leiostomus xanthurus ) in
Georgia's estuarine and close inshore ocean waters. Ga.
Dep. Nat. Resour., Coastal Fish. OfT. Contrib. Ser. 28,
29 p.
Nelson. W R
1967. Studies on the croaker, Micropogon undulatus
Linneaus, and the spot, Leiostomus xanthurus Lacepede
in Mobile Bay, Alabama. M.S. Thesis, Univ. Alabama,
Tuscaloosa, 85 p.
Nelson, W R , N C Ingham, and W E Schaaf
1976. Larval transport and year-class strength of Atlantic
menhaden, Brevoortia tyrannus. Fish. Bull., U.S.
75:23-41.
NoRCRoss. B L AND H M Austin
1981. Climate scale environmental factors affecting year
class fluctuations of Chesapeake Bay croaker
Micropogonias undulatus. Va. Inst. Mar. Sci. Spec. Sci.
Rep. No. 110, 78 p.
NOWLIN, W D, Jr.
1971. Water masses and general circulation of the Gulf of
Mexico. Oceanol. Int. 6(2):28-33.
PACHECO, a. L
1962. Age and growth of spot in lower Chesapeake Bay,
with notes on distribution and abundance of juveniles
in the York River system. Chesapeake Sci. 3:18-
28.
Pearson, J C
1929. Natural history and conservation of redfish and
other commercial sciaenids on the Texas coast. Bull.
U.S. Bur. Fish. 44:129-214.
Perry, A
1979. Fish of Timbalier Bay and offshore Louisiana
environments collected by trawling. Rice Univ. Stud.
65:537-545.
POWLES, H , AND B W STENDER
1978. Taxonomic data on the early life history stages of
sciaenids of the South Atlantic Bight of the United
States. S.C. Mar. Resour. Comm. Tech. Rep. No. 31,
64 p.
Sabins, D S . AND F. M Truesdale.
1974. Diel and seasonal occurrence of immature fishes in
a Louisiana tidal pass. Proc. Annu. Conf. Southeast.
Assoc. Game Fish Comm. 28:161-171.
Shaw, R F , J H Cowan, Jr , and T L Tillman
1985a. Distribution and abundance of Brevoortia
patronus (gulf menhaden) eggs and larvae in the
continental shelf waters of western Louisiana. Bull.
Mar. Sci. 36:96-103.
Shaw, R F , B D Rogers, J H Cowan, Jr , and W H Herke
In press. Ocean-estuary coupling of ichthyoplankton and
nekton in the northern Gulf of Mexico. Am. Fish. See.
Symp. 3:77-89.
Shaw, R F , W J Wiseman, Jr , R E Turner, L J Rouse, and
R E CONDREY
1985b. Transport of larval gulf menhaden Brevoortia
patronus in continental shelf waters of western
Louisiana: a hypothesis. Trans. Am. Fish. Soc.
114:452-460.
Shlossman, P a , AND M E Chittenden, Jr.
1981. Reproduction, movements and population dynamics
of the sand seatrout, Cynoscion arenarius. Fish. Bull.,
U.S. 79:649-669,
Springer, V G , and K D Woodburn
1960. An ecological study of the fishes of the Tampa Bay
area. Fla. Board Conserv. Mar. Lab. Prof. Pap. Ser.
No. 1, 104 p.
Suttkus, R. D.
1955. Seasonal movements and growth of the Atlantic
croaker (Micropogon undulatus ) along the east
Louisiana coast. Proc. Gulf Caribb. Fish. Inst.
7:151-158.
TOWNSEND, B C , Jr
1956. A study of the spot, Leiostomus xanthurus
Lacepede, in Alligator Harbor, Florida. M.S. Thesis,
Florida State Univ., Tallahassee, 43 p.
Wallace, D H
1940. Sexual development of the croaker, Micropogon
undulatus, and distribution of the early stages in
Chesapeake Bay. Trans. Am. Fish. Soc. 70:475-482.
Warlen, S. M
1981. Age and growth of larvae and spawning time of
Atlantic croaker in North Carolina. Proc. Annu. Conf.
Southeast. Assoc. Game Fish. Comm. 34:204-214.
White, M. L., and M E Chittenden
1977. Age determination, reproduction and population
dynamics of the Atlantic croaker, Micropogon
undulatus. Fish. Bull., U.S. 75:109-123.
Wiseman, W J , Jr , R E Turner, F J Kelly, Jr, L J Rouse, Jr.,
AND R F Shaw.
1987. Analysis of biological and chemical associations
near a turbid coastal front during winter 1982. Contrib.
Mar. Sci. 29:141-151.
142
BEHAVIOR OF SOUTHERN RIGHT WHALES, EUBALAENA AUSTRALIS ,
FEEDING ON THE ANTARCTIC KRILL, EUPHAUSIA SUPERBA
William M Hamner,^ Gregory S. Stone,^ and Bryan S. Obst^
ABSTRACT
Southern right whales, Eubalaena australis, were observed in 3 successive years on the western side
of the Antarctic Peninsula. These whales do not appear to be from the well-documented Valdes,
Argentina population. The whales we observed were feeding on Antarctic krill, Euphausia superba.
When krill were at the surface right whales surface-skimmed at high speed, with upper jaw lifted
above the water surface. In heavy weather one right whale "tail-sailed" at slow speed, with head
submerged and apparently feeding. When krill were organized in subsurface schools, right whales
engaged in subsurface feeding, diving repeatedly in place, apparently working a given school. One
whale swam directionally to the only known large school of krill in the area and fed intensively,
rested on the surface, then began a second feeding bout. Whales hyperventilated, false fluked, and
fluked prior to feeding dives. These are the first detailed observations of feeding behavior of right
whales in Antarctic waters and suggest that coastal Antarctica may have been (and may become
again) a regular part of the summer feeding range of the species.
Right whales are among the rarest of the great
whales, having been hunted almost to extinction
a century ago. The southern right whale, Eubal-
aena australis, has been studied only recently
and only during the austral winter when the
whales aggregate inshore to bear calves and to
mate (Clarke 1965; Payne 1976, 1986; Best 1981;
Aguayo and Torres 1986). Because right whales
were commercially extinct by the mid-1850's,
very little has been learned about their ecology
from the 20th century whaling industry. Informa-
tion on feeding, migration, stock structure, and
reproductive biology was collected for most other
Antarctic mysticete whales during the heyday of
whaling in this century (e.g.. Mackintosh 1965;
International Whaling Commission reports 1964-
present). The small number of surviving right
whales (ca. 29c of historic levels in the Southern
Hemisphere, Breiwick and Braham 1984) has
made it difficult for researchers to study this spe-
cies. Our current understanding of its feeding and
calving ecology in the Southern Hemisphere
comes from observations made primarily off
Peninsula Valdes, Argentina (Payne 1986).
In Antarctic waters south of lat. 60°S, more
than 30 sightings of right whales have been re-
ported previously (Berzin and Vladimirov 1981;
Goodall and Galeazzi 1986; Ohsumi and Kasa-
iDepartment of Biology, University of California, Los Ange-
les, CA 90024.
2College of the Atlantic, Bar Harbor, ME 04609.
Manuscript accepted September 1987.
FISHERY BULLETIN: VOL. 86. NO. 1. 1988.
matsu 1986). Most were in the vicinity of the
South Orkney Islands, 9 were near the Antarctic
Peninsula, 6 were in the Pacific sector of the
Antarctic, and 2 were south of Africa. We have
observed southern right whales during 3 consecu-
tive austral summers near the western shore of
the Antarctic Peninsula, and the sightings re-
ported herein and in Stone and Hamner (in press)
are the most southerly as well as the most de-
tailed observations.
We sighted one individual during the 1983-
84 austral summer (also recorded by Captain
P. Lenie in the log of the RV Hero ; Goodall
and Galeazzi 1986), two individuals in 1984-85,
and eight in 1985-86, four of which we indi-
vidually identified. In 1986 a fortunate combina-
tion of fair weather and available ship time
permitted us to make the first extended uninter-
rupted observations on the behavior of right
whales feeding on the Antarctic krill, Euphausia
superba.
METHODS
Right whales are distinguished by the absence
of a dorsal fin and regions of cornified skin (cal-
losities) on the head, jaws, and chin. Individual
whales were identified by standard methods,
using video tapes and telephotographs of head
callosities and scarring patterns on the head and
back (Payne et al. 1983; Kraus et al. 1986). When
possible, we dropped large disks of plywood of
143
FISHERY BULLP:TIN: VOL 86, NO 1
known diameter next to the whale and included it
in the photograph to measure whale size.
Behavioral patterns were observed from the
ship's bridge and recorded on a portable computer
as they occurred, using a program which timed
entries of encoded behaviors or coments to the
nearest second. Krill schools were recorded on a
Simrad echosounder and the identity of the or-
ganisms causing the echograms was verified by
net samples taken with an Isaacs-Kidd midwater
trawl and by divers' visual confirmation of krill
schools near the surface.
RESULTS
7 January 1984: At 0500 hour, north of Cape
Murray near Two Hummock Island, one right
whale was feeding at the surface with its upper
jaw lifted above the water, swimming at high
speed (estimated at 8-9 knots by the ship's cap-
tain) in feeding runs of 15-20 seconds, which we
recorded on video. Three humpback whales
nearby were diving in one specific location. Th,e
right whale repeatedly changed direction be-
tween surface runs so that its horizontal direc-
tional feeding excursions did not take it far from
the vicinity of the vertically diving humpbacks.
During these powerful filter-feeding runs enor-
mous amounts of water were displaced, cascading
beside and behind the right whale and producing
a large wake.
15 January 1985: A cow and a calf were swim-
ming slowly at the surface some 500 m from the
eastern shore of Anvers Island. On approach by
the ship the whales swam slowly into shallow
water where we could not follow. It was near dusk
and we could not get photographs for future iden-
tification.
7 January 1986: We encountered one southern
right whale and six humpbacks at 1830 hour at
lat. 63°46'S, long. 61°13'W, between Trinity and
Hoseason Islands. We photographed the head and
body of the right whale for subsequent identifica-
tion. We followed the whale for approximately 2
hours, recording diving times, surface intervals,
and breathing rates. The whale frequently
changed directions underwater and consequently
we often failed to see the whale immediately
when it resurfaced, so breathing rate data for this
behavioral sequence are incomplete. The whale
appeared to have captured krill on at least one
dive because when the whale surfaced it repeat-
edly and briefly opened and closed its mouth, with
baleen visible, a behavior presumably associated
with separation of krill and water prior to swal-
lowing the prey (Watkins and Shevill 1976).
About 50 cape petrels, Daption capensis, alighted
on the water and fed at the surface around the
whale. When the whale's jaw movements ceased,
the birds soon stopped feeding, but they remained
on the water and did not follow the whale when it
swam away at the surface.
2 March 1986: We observed one right whale at
the northern end of the Neumayer Channel,
where a large iceberg was grounded on a 93-
fathom rise 2 miles east of Iceberg Point. The
wind was blowing from the north at 20-24 knots
and a strong surface current was flowing south,
producing a bow wave on the grounded iceberg.
The right whale repeatedly swam NE of the berg,
raised its tail high out of the water at 90° to the
wind, submerged its head, and "sailed" downwind
past the iceberg, a behavior previously noted for
right whales in Argentina (Payne 1976). Soon
after we first saw the whale, it stopped tail-
sailing and began diving, still along the N-S tran-
sect near the iceberg where it had been sailing.
The presence of the ship did not cause the whale
to alter its back-and-forth swimming rhythm or
direction. We waited until the whale began one of
its N-S transects past the iceberg and followed
about 100 m behind it with the ship. A large
school of krill was present on the east side of the
iceberg. We recorded no other schools in the vic-
nity. The whale was accompanied by three female
fur seals. The seals constantly darted about the
head of the whale when it surfaced after long
dives and appeared to annoy the whale, because
several times the whale repeatedly slashed its
head sideways when the seals swam too close.
3 March 1986: At 1030 hour, we spotted a single
southern right whale near the mouth of Andvord
Bay on the Antarctic Peninsula. The whale was
swimming SSW at about 3 knots, making short
dives that lasted about 19 seconds {N = 12,
SD = 9.0 seconds), with brief surface intervals
that averaged 6.1 seconds (A^ = 12, SD = 2.6 sec-
onds) (Fig. 1). The whale then stopped diving but
continued to swim SSW toward the NE tip of
Lemaire Island, swimming mostly at the surface
for approximately 90 minutes. During this period
the whale appeared unconcerned with the ship,
which remained 50-100 m behind it, but when
the whale neared an iceberg that was hard
aground near Lemaire Island, it turned suddenly
at a right angle to its prior course and swam be-
tween the iceberg and the rocks. The ship was
nonetheless able to follow the whale through the
144
HAMNER ET AL : FEEDING BEHAVIOR OF SOUTHERN RIGHT WHALES
0
30
60
90
120
150
180
1M1
35
65
95
125
155
185
10
15
20
25
1
40
45
50
55
70
75
80
85
U
100
105
no
115
130
135
140
145
160
165
170
175
1
V/
190 195 200
Time in minutes
205
30
60
90
r
120
150
180
210
Figure 1. — Dive record of the right whale constantly observed for 3.5 hours on 3 March 1986. 44 minutes
(arrowl: The ship approached to within 10 m of the whale for i.d. photographs and size measurement. 54
minutes: The whale hyperventilated and dove in an area without krill, then swam northward. 92 minutes:
The whale began diving on scattered small krill schools while still traveling N and NNW. 140 minutes: The
whale stopped and began diving on one large concentration of krill. 182 minutes: The whale rested on the
surface, moving its jaws. 194 minutes: The whale began a second feeding bout ca. 400 m further south.
narrow channel. Thereafter the whale ignored
the ship and altered course to NNW, still swim-
ming at the surface. The ship pulled ahead of the
whale to measure its length and this may have
caused the short dive noted at ^ = 43 minutes.
However, the whale calmly surfaced again within
10 m of the ship and watched us while we pho-
tographed it next to the wooden disk. The whale
was 11m long. The whale then continued to swim
at the surface to the NNW. During the period that
the whale swam at the surface without diving the
average time between breaths was 50.8 seconds
(n = 55; SD = 10.8). At ^ = 54 minutes the whale
began to hyperventilate, took three short dives,
lifted the flukes partially clear of the water (false
fluking I on the third dive, and then fully raised
the flukes on the fourth dive, which lasted 210
seconds. The whale then remained at the surface
for about 30 minutes while swimming northward.
This pattern of hyperventilation prior to a long
dive occurred before every long dive sequence
which was preceded by a surface interval of at
least 4 minutes (Fig. 2). We used this criterion to
restrict our dive selection for this plot because
there is some indication that there is also a brief
recovery period involving hyperventilation after
long dives. Of the last 19 breaths that were taken
during the 90 seconds before the 5 long dives plot-
ted in Figure 2, 18 were less than 30 seconds
apart, with a mean interval of 15 seconds
(A^ = 19, SD = 7.3). During the time preceding
this 90-s hyperventilation period, the whale aver-
aged 1 breath every 48 seconds iN - 17, SD =
12.4), not significantly different from the time of
50.8 seconds recorded between breaths during
long surface intervals without dives. During hy-
perventilation, therefore, breathing rate in-
creased by about 3 x .
At ^ = 92 minutes we began to record scattered
small krill schools on the echosounder and the
whale began to dive erratically, with some rea-
sonably long dives, but most quite short. Of 23
dives, 13 (56%) were shorter than 10 seconds; the
time averaged for all dives was 75.3 seconds
145
FISHERY BULLETIN: VOL 86, NO 1
70 r
3 2
Minutes prior to dive
0
Figure 2. — Breath sequences prior to 5 dives lasting from 93 to 279 seconds. During the 90 seconds
immediately preceding dives, the mean interbreath interval was 15 seconds. Prior to this 90-s period
of hyperventilation the mean interbreath interval was 48 seconds.
(A^ = 20, SD = 74.6). During this period of erratic
diving the whale continued to swim N, then
NNW. At about t = 140 minutes the whale
stopped traveling NNW and began a series of 10
dives, the second being the longest, followed by
successively shorter dives. These dives were fol-
lowed by rather uniform surface intervals and all
10 lasted significantly longer (x = 183.5 seconds,
N = 10, SD = 89.2) than the previous mean of 75
seconds. At the end of the dive sequence, at 182
minutes, the whale stopped swimming entirely
and floated motionless at the surface, occasion-
ally moving its jaws. This inactivity lasted about
12 minutes and then the whale began a second
series of dives. We called these dive sequences,
which consisted of regular length surface inter-
vals interspersed with a series of longer dives of
progressively decreasing duration, "feeding
bouts" because at ^ = 140 seconds the whale had
reached the only major aggregation of krill in the
vicinity, as verified by sonic records (Fig. 3). We
saw dense schools of krill and isolated krill at the
surface from the bow of the ship, and we captured
krill in three successive hauls with the 1 m
Isaacs-Kidd midwater trawl.
During the first feeding bout the whale slowed
from its steady 3-knot swimming speed and
moved slowly at about 1 knot, but no longer in
any specific direction, finally swimming some
400 m to the south before beginning the second
feeding bout. The whale had fed on krill earlier
also, because it defecated during the feeding se-
quence and the feces, as determined by later mi-
croscopic examination, were composed entirely of
well-digested euphausiids.
The display of a false fluke prior to the high
fluke initiating a long dive did not necessarily
indicate presence or absence of krill, but when
krill were present, it was a highly significant pre-
dictor of the length of the dive. In the presence of
prey, when the whale showed its flukes once (i.e.,
did not false-fluke preceding the dive), dive dura-
tion averaged 91.2 ± 13.0 seconds {N ^ 5),
whereas when both a false fluke and a high fluke
preceded the dive, the dive averaged 234.7 ± 64.4
seconds (N = 1).
In March and April 1986 we saw a total of eight
southern right whales. Of these we distinguished
four as individuals on the basis of video tapes and
photographs of callosities and body markings.
One of these whales was the same individual that
we had observed in January some 70 miles fur-
146
HAMNER ET AL : FEEDING BEHAVIOR OF SOUTHERN RIGHT WHALES
0( ■
I/)
(V
<D
E
c
Cl
O)
Q
20
40
S5:<s,<?
.J4%?*
^
<»
60
80
=:Sf-
kilometer
Figure 3. — Echogram of krill school on 3 March 1986 on which right whale stopped swimming and began
feeding bout.
ther north, as confirmed by comparing video
recordings of both sightings. Photographs of these
four individuals were subsequently compared
with the catalog of southern right whales that
overwinter near Peninsula Valdes, Argentina
(Payne and Rowntree 1984), but none of the
Antarctic individuals were in that compendium of
some 623 individuals.
DISCUSSION
Our observations of southern right whales
along the Antarctic Peninsula suggest that
coastal Antarctica may have been (and may well
become again) a regular part of the summer feed-
ing range of the species. These sightings indicate
a more extensive distribution in the Southern
Ocean than heretofore reported, an extension cor-
roborated by the comparison of recent and his-
toric summer distributional patterns by Ohsumi
and Kasamatsu (1986). Southern right whales
were hunted almost to extinction in the shallow
coastal embayments where they overwinter more
than 100 years before whaling began in the
Southern Ocean (Harmer 1928; Townsend 1935).
It is not surprising, therefore, that there are only
a few scattered records of southern rights in
Antarctic waters, fluke observations so to speak.
During the past 50 years southern rights have
been protected, and several domes of southern
right whales have increased (Best 1981; Ohsumi
and Kasamatsu 1986; Whitehead et al. 1986), and
a commensurate extension of the summer feeding
range into areas formerly occupied by these
whales may be occurring.
It is interesting that none of the 4 individual
whales for which we have photographic identifi-
cations appear in the catalog of 623 individual
southern rights for the Valdes, Argentina breed-
ing population (Payne and Rowntree 1984). We
believe that the right whales we encountered in
Antarctic waters probably winter elsewhere, per-
haps along the Chilean coast, which has been less
well surveyed than has the east coast of South
America (Cardenas et al. 1986).
Payne (1976) described tail-sailing among the
Valdes population of southern right whales and
suggested that this may be a form of play behav-
ior. In Antarctic waters this behavior may be a
method used to forage on krill. The right whale
that we observed tail-sailing had three fur seals
in attendance, constantly darting about its head,
apparently feeding on the same prey as the
whale. Furthermore, tail-sailing occurred next to
a grounded iceberg at a specific location that was
repeatedly transected by the whale, the only spot
in the vicinity where krill were detected on the
ship's sonar.
147
FISHERY BULLETIN: VOL. 86, NO I
Northern right whales feed by both surface-
skimming and subsurface feeding, apparently
holding the jaws agape below the surface much
like they do when at the surface (Watkins and
Schevill 1976, 1979). Surface-skimming by north-
ern right whales that feed on copepods and juve-
nile euphausiids occurs at speeds of about 3 knots
with relatively little turbulence (Watkins and
Schevill 1976, 1979). The one southern right
whale which we observed surface-skimming, and
which was also mentioned in Goodall and
Galeazzi (1986), swam at about 8 knots, with an
immense amount of turbulence. This whale un-
doubtedly was feeding on Euphausia superba.
The extremely high speed of this whale when
surface-skimming may be related to type of prey.
Euphausia superba is a large euphausiid that
swims up to 30 cm/second, and it is quite adept at
avoiding scuba divers and other dark objects
(Hamner et al. 1983; Hamner 1984). It may be
that right whales adjust the speed of their surface
feeding runs to overcome the swimming speed of
the particular prey that they pursue. Fin whales
also exhibit differences in breathing and diving
rhythms when feeding on euphausiids and on
schooled fish (Watkins et al. 1984).
The southern right for which we have a 3.5-h
record in 1986 fed exclusively below the surface.
It swam for about an hour at the surface almost
directly NNW toward the only patch of krill that
we had located previously via the echo sounder
during repeated transects across the mouth of
Andvord Bay. It is not understood how baleen
whales locate prey. They might echolocate, they
might listen for krill noises, or they might re-
member where they had prior success and return
to the general vicinity and then begin to hunt
randomly. Although our data do not allow us to
choose among these possibilities, it appears that
this particular southern right whale did navigate
directly to patches of prey, as suggested by
Watkins and Schevill (1979) for northern right
whales. However, some subsurface searching
probably occurs as well. The whale that we fol-
lowed for the longest period of time made a rea-
sonably long dive of 210 seconds in the middle of
its surface swim toward the krill patch. It appar-
ently did not find anything to eat (nor did the
sonar indicate the presence of krill), and it made
no more significant dives for about 30 minutes.
When the whale reached an area with small
scattered krill schools, its dives were generally
short, and the temporal spacing and duration of
dives and surface intervals were erratic. When
the whale reached a large school of krill it stopped
swimming between dives, dives lasted longer,
surface intervals between dives became more reg-
ular, and successive dives were progressively
shorter. After 10 of these dives the whale floated
without any swimming movements at the surface
for about 12 minutes while occasionally rattling
its baleen and slightly lifting its upper jaw.
The whale clearly anticipated its dives. It hy-
perventilated for about 90 seconds prior to all of
the long dives. Hyperventilation is a common pre-
cursor to diving in air-breathing vertebrates
(Kooyman et al. 1981), but this respiratory pat-
tern apparently has not been reported previously
for a baleen whale.
The data that we present in support of this feed-
ing pattern are admittedly modest, consisting of
only one continuous 3.5-h data sequence for one
individual. Nonetheless, this data set is truly
unique and it includes a series of statistically sig-
nificant changes of behavior at the surface and at
depth which are correlated with prey distribu-
tion. Since the changes of behavior are real, they
merit interpretation. One hypothesis to account
for the increase and then the decrease in dive
length after encountering prey is that the whale
dove to its full capacity in order to maximally
exploit an opportunity to feed. With repeated
dives its dive capacity may have diminished, pro-
ducing first long, then shorter and shorter dives
(Fig. 1). In contrast, this breathing pattern has
not been observed in grey whales, which exhibit a
sustained capacity for repetitive, nearly
monotonic diving while feeding (Obst et al. in
prep.). It is possible that right whales, being more
specialized for surface-skimming, become ex-
hausted more readily during sustained diving.
An alternative hypothesis to account for this
pattern of behavior is that the whale captured
krill with such ease on the first long dive of 348
seconds that a "digestive bottleneck" (Kenward
and Sibly 1977) developed; subsequent dives
would be progressively shorter because the stom-
ach (or some storage chamber) filled more rapidly
than the food could be processed. The long surface
interval of immobility followed by a second feed-
ing bout might then represent the time taken to
clear such a chamber (e.g., the first of the three
stomach pouches) in preparation for the next bout
of feeding. If this hypothesis is correct, it implies
that once a right whale has found a particularly
favorable school of krill, it fills up very rapidly.
This interpretation contrasts markedly with a
popular impression of great whales as continuous
148
HAMNER ET AL.: FEEDING BEHAVIOR OF SOUTHERN RIGHT WHALES
feeders that harvest prey while swimming lazily
through diffuse planktonic resources. Herwig et
al. (1984) presented preliminary evidence that
microbial fermentation of ingesta may occur in
the forestomach of baleen whales, a process which
may be expected to limit the throughput rate of
food in accordance with continuous-flow, stirred
tank reactor kinetics (Penry and Jumars 1987). It
is tempting to speculate that the feeding behavior
of right whales, among the largest endotherms on
earth, may be constrained by a digestive bottle-
neck analogous to that described for humming-
birds (Karasov et al. 1986), the smallest of en-
dotherms. These hypotheses warrant subsequent
testing through more extensive time-budget
analyses of right whales and other balaenopter-
ans on their feeding grounds, such as described
here and in Watkins et al. (1984).
ACKNOWLEDGMENTS
We thank P. Hamner, M. Lang, and the crews
of the RV Hero and RV Polar Duke for field assis-
tance. R. Jedas wrote the computer program for
recording behavior in real time. R. Payne and V.
Rowntree were of great assistance in comparing
photographs of individual Antarctic whales and
of right whales occurring off Peninsula Valdes,
Argentina. P. Hamner provided editorial assis-
tance. This research was supported by NSF
grants DPP83-02852 and DPP86-14821.
LITERATURE CITED
Aguayo. L a , AND N D Torres
1986. Records of the southern right whale, Eubalaena
australis (DesmouHns, 1822), from Chile between 1976
and 1982. Rep. int. Whaling Comm., Spec. Issue 10,
p. 159-160.
BERZIN, a a . AND V L Vladimirov
1981. Changes in the abundance of whalebone whales in
the Pacific and Antarctic since the cessation of their ex-
ploitation. Rep. int. Whaling Comm. 31:495-499.
Best, P
1981. The status of right whales (Eubalaena glacialis) off
South Africa, 1969-1979. S. Afr. Dep. Agric. Fish., In-
vest. Rep. 123, 44 p.
Breiwick, J M . AND H W Braham (editors).
1984. The status of endangered whales. Mar. Fish. Rev.
46(4):l-64.
Cardenas, J C , S S Stutzin. C Cabello, and J Oporto
1986. The first steps to cetacean conservation and man-
agement in Chile. Final Report to the World Wildlife
Fund, Project No. WH-445. World Wildl. Fund U.S.,
Wash., DC.
Clarke, R
1965. Southern right whales on the coast of Chile. Nor.
Hvalf.-Tid. 54:121-128.
Goodall, R N P . AND A R Galeazzi
1986. Recent sightings and strandings of southern right
whales off subantarctic South America and the Antarctic
Peninsula. Rep. int. Whaling Comm., Spec. Issue 10, p.
173-176.
Hamner, W M.
1984. Aspects of schooling in Euphausia superba. J.
Crust. Biol. 4 (Spec. No. l):67-74.
Hamner, W M , P P Hamner, S W Strand, and R W Gilmer.
1983, Behavior of Antarctic krill, Euphausia superba:
chemoreception, feeding, schooling, and molting.
Science 220:433-435.
Harmer, S. F.
1928. A history of Whaling. Proc. Linn. Soc. Lond.
140:51-95.
Herwig, R P , J T Staley, M K. Nerini, and H W Braham
1984. Baleen whales: preliminary evidence for forestom-
ach microbial fermentation. Appl. environ. Microbiol.
47:421-423.
International Whaling Commission
1964-present. Reports. Office of the Commission, Cam-
bridge.
Karasov, W H , D Phan, J M Diamond, and F L Carpenter
1986. Food passage and intestinal nutrient absorption in
hummingbirds. Auk 103:453-464.
Kenward. R E , and R M Sibly.
1977. A woodpigeon (Columba palumbus) feeding prefer-
ence explained by a digestive bottle-neck. J. Appl. Ecol.
14:815-826.
Kooyman, G L . M A Castellini, and R W Davis.
1981. Physiology of diving in marine mammals. Ann.
Rev. Physiol. 43:343-356.
Kraus, S D , K E MooRE, C. A Price, M J Crone, W A.
Watkins, H E Winn, and J. H. Prescott.
1986. The use of photographs to identify individual North
Atlantic right whales (Eubalaena glacialis ). Rep. int.
Whaling Comm., Spec. Issue 10, p. 145-151.
Mackintosh, N A.
1965. The stocks of whales. Fishing News (Books) Ltd.,
Lond.
Ohsumi, S . AND F Kasamatsu
1986. Recent off-shore distribution of the southern right
whale in summer. Rep. int. Whaling Comm., Spec.
Issue 10, p. 177-185.
Payne, R S
1976. At home with right whales. Natl. Geogr. Mag.
149:322-339.
1986. Long term behavioral studies of the southern right
whale (Eubalaena australis ). Rep. int. Whaling Comm.,
Spec. Issue 10, p. 161-167.
Payne, R. S., O Brazier, E M. Dorsey, J. Perkins, V. Rowntree,
AND A. Titus.
1983. External features in southern right whales (Euba-
laena australis) and their use in identifying individu-
als. In Payne, R. S. (editor), Communication and be-
havior of whales. Westview Press, Boulder, CO,
p. 371-445.
Payne, R. S., and V. J. Rowntree.
1984. Un catalogo fotografico de ballenas individuales
avistadas en las aguas alrededor de la Peninsula Valdes,
Argentina. 1st ed. World Wildlife Fund, Center for
Long Term Research, Lincoln, MA.
Penry, D L , and P A Jumars.
1987. Modeling animal guts as chemical reactors. Am.
Nat. 129:69-96.
149
FISHERY BULLETIN: VOL. 86, NO. 1
Stone. G S . and W M HamnKR baleen whales: Eubalaena glacialis. Balaenoptera bore-
In press. Humpback and right whales in the Gerlache alis, Megaptera novaeangliae , and Balaenoptera
Strait on the Antarctic Peninsula. Polar Rec. physalus. J Mammal. 60:155-163.
TowNSEND. C H Watkins, W A . K E Moore. J Sigur-jonsson, D Wartzok, and
1935. The distribution of certain whale species as shown G N DI S( IARA
by logbook records of American whaleships. Zoologica 1984. Fin whale (Ba/aenop<erap/!vsa/us 1 tracked by radio
(N.Y.) 19:1-50. in the Irminger Sea. Rit Fiskideildar 8:1-14.
Watkins. W A. AND W E Schevill Whitehead, H. R Payne, and M Payne
1976. Right whale feeding and baleen rattle. J. Mam- 1986. Population estimate for the right whales off Penin-
mal. 57:58-66. sula Valdes, Argentina, 1971-1976. Rep. int. Whaling
1979. Aerial o'bservation of feeding behavior in four Comm., Spec. Issue 10, p. 169-171.
150
NOTES
AN ECONOMETRIC ANALYSIS OF
IVfET INVESTMENT IN
GULF SHRIMP FISHING VESSELS'
The major capital inputs in the Gulf shrimp fish-
ery are vessels and gear. Early vessels, built
mostly of wood, employed drag seines, cast nets,
and fixed traps to catch shrimp. Today, much
larger and more powerful vessels, equipped with
sophisticated fishing gear and accessories, trawl
the Gulf of Mexico. The last 20 years has seen a
remarkable substitution of steel and fiberglass
vessels for wooden vessels. The factors underlying
this aggregate investment trend in the Gulf of
Mexico (hereafter referred as the Gulf) shrimp
fishery, however, have yet to be examined.
Economic analysis of the fishing industry has
increased in recent years because of the growing
importance of world fish stocks. However, these
studies have been focused predominantly at the
micro or firm level (Thompson et al. 1970^;
Wilson et al. 1970; Juhl 1974; Watson 1977;
Griffin et al. 1978; Prochaska and Cato 1981).
Moreover, accounting for a comprehensive mean
of the cost of capital has been ignored in past
research efforts. Yet, the cost of capital is likely a
major factor in fishing vessel investment deci-
sions in the Gulf shrimp fishery. The long run
profitability of the sector and its exposure to fi-
nancial risk depends, to a large extent, upon its
capital structure and fluctuations in the cost of
debt and equity capital. Futhermore, this deter-
minent of aggregate investment behavior repre-
sents a major channel through which macroeco-
nomic policy actions are transmitted to the Gulf
shrimp fishery.
The purpose of this study is to estimate an
econometric model of annual real net investment
in fishing vessels in the Gulf and to determine the
sensitivity of investment decisions in the indus-
try to fluctuations in the cost of equity and debt
capital. This study begins by examining the indi-
vidual factors that affect the expansion of the
iTechnical Article No. TA-20803 of the Texas Agricultural
Experiment Station, Texas A&M University System. This re-
search was supported in part by the Texas A&M University Sea
Grant College Program under Grant No. NA81AA-D-00092.
2Thompson, Russel G., R. W. Callen, and L. C.
Wolken. 1970. Optimal investment and financial decisions
for a model shrimp fishery firm. Unpubl. Rep. Texas A&M
Univ., TAMU-SG-70-205.
stock of steel, wood, and fiberglass vessels in the
Gulf fieet. The effects of alternative macroeco-
nomic policies on investment expenditure trends
in the Gulf shrimp fishery are then studied. The
final section of this paper presents some conclud-
ing remarks.
Determinants of Net Investment
The aggregate investment model used in this
study is based upon the neoclassical theory of ag-
gregate investment behavior. The determinants
of the desired capital stock of fishing vessels as
well as the relationship between this desired cap-
ital stock and net investment behavior of Gulf
shrimp fishermen are specified in this section.
Desired Stock of Fishing Vessels
In making investment decisions, competitive
firms add to their existing stock of capital as long
as the present value of the periodic net cash flows
generated by an additional unit of capital exceeds
its net purchase price. This condition for any par-
ticular type of fishing vessel (e.g., wood, steel,
fiberglass) can be stated algebraically as follows^:
2 pidX/dKj) - (dTi/dKj) - r{dDt/8Kj}
t=i
idPt/dKj) (1 + p)-' > Qjia - ic)
X U + ^ dRjt/dKj]a + p)-'] (1)
where variable p represents the real price fisher-
men expect to receive per unit of output in the
Gulf of Mexico shrimp industry, X represents the
quantity of shrimp expected to be harvested, Kj is
the real stock of thej'^ category of fishing vessels,
Tt and P/ represent the tax payment and principal
payment due in period t expressed in constant
3The variables without any subscripts are expected at the
time the investment is made to remain constant over time.
FISHERY BULLETIN: VOL. 86, NO. 1, 1988
151
dollars, D, represents debt outstanding in period
t, r is the real rate of interest on debt capital, p is
the real after-tax opportunity rate of return on
equity capital desired by fishermen, Qj is the real
price paid for the^'^ category of fishing vessels at
the retail level, a is the proportion of investment
financed with equity capital, i^. is the investment
tax credit rate, and Rjt represents the real level of
replacement investment in the j^^ category of
fishing vessels required in period t to offset losses
in productive capacity due to wearout.
The entire term on the left-hand side of the
inequality sign in Equation (1) represents the
present value of the additional net cash flows gen-
erated by a permanent addition to the^'^ category
of fishing vessels. It is assumed that both the in-
terest and the principal payments, ((tP/SK^) and
r(dD/BK, ), vary over time as further expenditures
are required to maintain the productive capacity
of this addition to the capital stock at its original
level. The right-hand side of Equation (1) repre-
sents the initial downpayment minus the invest-
ment tax credit plus the present value of all fu-
ture cash outlays required to maintain the stock
of the j^^ category of fishing vessels at its new
level.
To maximize the present value of their equity.
Gulf shrimp fishermen would continue to add to
the stock of the j^^ category of fishing vessels
until Equation (1) holds as an equality. Equiva-
lently, maximization of the present value of
owner equity requires thaf*
pidX/dKj) = j^^
a
ic + Z - A
(1 - iy)
(2)
where F, represents the present value of the
stream of capacity depreciation of they '^^ category
of vessels and iy is the average income tax rate.
The term Z represents the present value of the
stream of after-tax interest payments and princi-
pal payments on debt while A represents the
present value of the stream of tax depreciation
allowances that can be claimed for each dollar of
investment as the stock of vessels is maintained
at its new level. ^ The right-hand side of this ex-
4Equation (1) as well as the derivation of the implicit rental
price of vessels (c) assume that fishermen expect real prices (p)
and the marginal physical product of vessels dX/dK to remain at
current levels. These and other assumptions which allow us to
treat many components of Equation (1) as consoles are consis-
tent with those employed in Penson et al. (1981) and Coen
(1975).
pression thus represents the implicit rental price
of the J^^ category of fishing vessels.
The concept of the implicit rental price of capi-
tal has been widely employed in previous studies
of investment behavior as a determinant of the
capital stock which firms desire to hold (e.g., Coen
1968, 1975; Penson et al. 1981). Equation (2) sug-
gests that the implicit rental price of they*^^ cate-
gory of fishing vessels will increase if their pur-
chase price, the cost of debt, and equity capital, or
income tax rates increase. These effects, however,
will be offset to some extent by an increase in
the investment tax credit rate, the deductibil-
ity of tax depreciation allowances, and interest
expenses.
Let us assume that output in this industry is a
function, in part, of the stock of fishing vessels
and that this production relationship is of the
Cobb-Douglas form. Letting p^ represent the par-
tial production elasticity associated with the
stock of the 7'^ category of fishing vessels (Kj),
the marginal physical product for these vessels
can be expressed as follows:
idX/dKj) = i^jiX/Kj).
(3)
Substituting Equation (3) into Equation (2), the
desired stock of they'*^^ category of fishing vessels
at the end of year t can be expressed as follows:
K*jt^^j{pX/Cjh
(4)
where Cj represents the expected implicit rental
price of they '^'^ category of fishing vessels given by
the right-hand of Equation (2). Equation (4) im-
plies that the desired stock ofj^^ category of fish-
ing vessels is directly related to the expected real
gross income from Gulf shrimp fishing and is in-
^The nominal value of Z (the present value of the stream of
after-tax loan payments) in Equation (2) is equal to
n
il -iyW 2j d,(l + ^)''{1 + p)-'
i = 1
n
+ ^ ^ (e - d,)(l -I- (J))-'(l -h p)-'
( = 1
while the real value of A (the present value of the stream of
depreciation allowances is equal to
i^(hl(p + A> + pi> + ^))
where d^ is equal to the nominal interest payment on a loan of
one constant dollar, (J) is the inflation rate, e is the nominal
amortized loan payment on a loan of one current dollar, and 8 is
the tax depreciation rate given by 2ln where n is the service life
of the vessel.
152
versely related to the expected implicit rental
price of these vessels. Similar equations could be
developed for other inputs used in the shrimp
fishing effort.
Desired Net Investment in Fishing Vessels
New fishing vessels are acquired by Gulf
shrimp fishermen both to expand their productive
capacity and to replace losses in the productive
capacity of existing vessels. This partitioning of
observed gross investment into net investment
and replacement investment for the j^^ category
of fishing vessels can be expressed definitionally
as follows:
Njt = Kj, - K,,
h - ^jt
(5)
where 7,7 represents the level of real gross invest-
ment in the 7'^ category of fishing vessels in year
t while Rf is the real replacement investment
needed to offset annual capacity depreciation of
these vessels. The variables K; and K/- 1 represent
the productive capital stock of the 7'^ category of
fishing vessels the end and the beginning of the
year, respectively. Given Equations (4) and (5),
the following relationship between the desired
stock of the j^^ category of fishing vessels and
current real net investment in these durable
inputs can be defined:
A^
jt
QjiK*j,
Kj,-i)
(6)
where 0 < 0^ < 1 and where 0^ represents the par-
tial adjustment coefficient that describes the
speed of adjustment of actual stocks to desired
levels for thej'^ category of fishing vessels. Sub-
stituting Equation (4) into Equation (6) and as-
suming an adaptive expectations hypothesis for
(pXICj)t, the following compound geometric ex-
pression is obtained:
A^^, = 0^ \^jUpXICj)t + (1 - ^j)Njt-i
+ %{l - \j)Kjt-2
0A
Jt-l + \^Jt
(7)
where \j is the adaptive expectations coefficient
and [Xji represents the error term. Since Kjt-2 is
equal to Kjt^i - Nf-i, Equation (7) reduces to the
following estimating equation:
A^,, = bjo + bji (pXICj)t + bj2 Kjt-i
where 60 is the intercept, 61 = 8 px, 62 = -8X,
63 = (1 - \) (1 - 0) and \xt is once again the ran-
dom disturbance term. The estimates of the 61
and 63 coefficients are expected to be positive
while the value of 62 is expected to be negative.^
Equation (8) thus represents the general form of
the equations to be econometrically investigated
in this study.
Data
The time series data used in this study consist
of annual observations for each variable in Equa-
tion (8) over the 1965-77 period. This time period
represents the only period for which investment
expenditure information is available.
The productive capital stock of the^'^'^ category
of fishing vessels is comprised of a series of differ-
ent vintages of vessels or
Kj, = Ijt + (1 - hj^)Ijt.^ + (1 - A,i - hj2)\^_^ + . . .
+ (\-hj^-hj2- .. .- hjr,)Ijt-n (9)
where /i^, is the fraction of thej*^^ category of fish-
ing vessel's original productive capacity lost in
the i^^ year of its service life. The value of A, is
represented by (1 - <}>)'" ^, where <t> = 2/n and n is
the assumed service life.^ In a related matter,
the present value of the stream of capacity de-
preciation of a vessel {Fj) was computed as fol-
lows:
+ bj3Njt-i + iXjt
(8)
6The net investment model expressed in Equation (8) can be
seen as a part of a simultaneous equation system that includes
other investment equations as well as supply equations for all
inputs and the production function for the fishing industry. The
specification of the complete simultaneous system of equations
and measurement of time series data needed to simultaneously
estimate the 6, coefficients in Equation (8) are beyond the scope
of this study. Since the disturbance terms for this set of invest-
ment equations are likely correlated, the seemingly unrelated
regression equations estimator was employed. The disturbance
terms given by this estimator were also examined for autocorre-
lation. The estimated rho coefTicient in this small sample was
shown to be insignificant in all cases. Finally, the predicted
rather than actual value of A'^, _ j was used in estimating Equa-
tion (8) to address the pyossibility of correlation between the
lagged dependent variable and the disturbance term.
''While a geometric decline in productive capacity has been
assumed for fishing vessels, recent studies indicate that the
productive capacity of equipment and machinery deteriorates at
a lower rate in the early period than in latter years. Coen (1975)
suggested that equipment and machinery deteriorate as they
age, though not necessarily at a geometric rate. For farm trac-
tors, a concave decay pattern represents the best proxy for the
capacity depreciation pattern suggested by engineering consid-
erations (Penson et al. 1981). The true pattern which underlines
actual capital spending decisions in the fishing industry could
not be examined due to inadequate data.
153
Fj = ^hj,a + p)
(10)
(=1
Data from the National Marine Fisheries Ser-
vice were used as annual observations on the
nominal value of gross investment in Gulf fish-
ing vessels (U.S. Department of Commerce 1965-
77^). These values were deflated to real terms
using the industrial price index. The quality of
the time series for real net investment in fishing
vessels, N/, depends on how well the annual
values of/, reflect quality changes in vessels over
time.
The annual levels of the implicit rental price of
vessels (c) were computed using the definition
outlined on the right-hand side of Equation (2).
Coen (1975) assumed that the real after-tax rate
of return desired on equity capital, p, is constant
over the economic life of the investment. Follow-
ing the lead of Coen, a value for p of 5% was
employed in this study.
The real rate of interest on nonreal estate loans
at commercial banks, r, along with annual rate of
inflation equals the nominal rate of interest on
debt capital. Annual values for all these variables
were obtained from US. Department of Com-
merce publications. The annual values for the
fraction of investment expenditures that are debt
financed (v];) used in computing Z were found by
dividing the annual change in total debt in the
fishing industry by the annual level of gross
investment in durable inputs provided by the Na-
tional Marine Fisheries Service. The time series
for a, the fraction of capital expenditures fi-
nanced with internal equity capital, was equal to
one minus the percentage debt financed
(l-il/).
Investment tax credit rate, i^, was equal to 7%
during the 1965-68 period, 0% during the 1969-70
period, and I09c during the 1971-77 period. The
maximum corporate income tax was assumed to
represent i^ for the Gulf shrimp fishery. The
double-declining balance method was assumed in
determining the present value of the stream of
annual tax depreciation allowances in A .
The time series data on prices paid for vessels,
a component of the rental price, were measured
using cost data collected from shrimp vessel
builders. Griffin et al. (1978) have shown that
vessel length, material of construction, and year
of purchase were the most significant factors de-
termining the price of a vessel. The equation esti-
mated in that study was used to extend available
vessel price information over the entire time pe-
riod covered by this study.
Econometric Results
Statistical as well as economic criteria can be
employed to evaluate the estimated equations for
the various categories of fishing vessels. The eco-
nomic criteria include the reasonableness of the
elasticities for the economic variables and as well
the partial production elasticities implied for the
production function.
Empirical estimates of the annual real net in-
vestment model for steel, wooden, and fiberglass
vessels indicate statistically significant coeffi-
cients for all but one of the explanatory variables
at the 10% level or less (Table 1). The lone excep-
tion was the coefficient associated with the lagged
capital stock variable in wooden vessel model,
which was not significantly different from zero at
less than the 20% level. All the coefficients associ-
ated with the explanatory variables have the
signs hypothesized earlier in this paper. Finally,
the coefficients on the lagged dependent variable
satisfy the constraint of being both greater than
zero and less than one.
Table 1 . — Estimated coefficients for the annual net investment
model for fishing vessels. Gulf Shrimp Fleet, 1965-77.
Vessel
Constant
(pX/c),
K,-:
N,-^
type
(bo)
(bi)
{b2)
(b3)
Steel
-95.3895
4.3770
-9.1302
0.6318
M6.21)
(6.23)
(2.34)
Wooden
384.3848
0.1132
-3.1944
0.9089
(1.79)
(0.93)
(1.21)
Fiberglass
0.3819
0.2990
-0.9529
0.2765
(5.37)
(1.74)
(2.88)
8U.S. Department of Commerce. 1965-77. Vessel charac-
teristics data. National Marine Fisheries Service, NOAA,
Wash., D.C.
1 Numbers in parentheses indicate absolute values of f -statistic.
Economic criteria employed in evaluating the
reasonableness of the empirical results and in
comparing the investment behavior among vessel
types include the partial production elasticity of
fishing vessels (p). This elasticity is given by
-61/62 which is computed using the estimated
beta coefficients in Equation (7). It appears that
steel vessels, with a partial production elasticity
of 0.479, are highly productive and play an impor-
tant role in the supply of fish. The fishing sector,
unlike other sectors of the economy, depends pri-
154
marily on one major capital input — vessels.
Therefore, the high partial production elasticities
recorded for steel vessels and fiberglass vessels
(0.314) are no surprise. Even though wooden ves-
sels appear to incur more repair and maintenance
costs, attract a lower quality crew, and, for that
matter, are less efficient than the other vessel
types, the low partial production elasticity of
0.033 is surprising. This low partial production
elasticity may have been caused by the fact that
the instrument for A'^^ - i in the wooden vessels
equation used to address the issue of the relation-
ship between the lagged dependent variable and
the error term did a poor job of explaining Nf - i.
An examination of the elasticities associated
with the (pX/c )/ term, computed at the mean, re-
veals that real net investment in steel vessels is
the most sensitive to changes in prices, interest
rates, taxes, and the other factors captured in this
variable. An elasticity of 7.28 associated with this
economic variable was computed for steel vessels.
This means that a 1% change in the (pX/c)t vari-
able causes real net investment in steel vessels to
change by 1.289c. This high investment response
to changes in these economic relationships could
be attributed to the fact that steel vessels, by far
the most productive (as evidenced by the high
partial production elasticity reported earlier), are
the most durable and the most capital intensive.
The elasticity associated with the (pX/c)t term in
the fiberglass and wooden vessel equations were
5.35 and 3.11, respectively. This would suggest
that macroeconomic policy actions would have a
substantially greater effect on real net invest-
ment in steel vessels than, say, wooden vessels.
Impact of Changes in
Cost of Capital
The impact of high real interest rates on the
growth of selected sectors in the economy has
been of great concern in the 1980s. The sensitiv-
ity of annual real net investment in fishing ves-
sels to changes in the real rate of interest is exam-
ined in this section by simulating the estimated
equations under annual real rates of 5 and 109?^.
In the short run (3 years), an increase in the real
rate of interest on debt capital from 5 to 10%
would cause real net investment in fishing ves-
sels to decrease by 3.04%. Annual real net invest-
ment in these fishing vessels would decrease by
15.88% in the long run. As the real rate of interest
on debt capital increases, it becomes more diffi-
cult to justify the purchase of additional vessels
owing to their rising marginal factor cost. Given
the fact that 67% of the cost of new fishing vessels
is normally financed with debt capital, it is not
surprising that rising real interest rates on debt
captial have a significant negative effect on the
long run expansion of the Gulf fleet.
The real cost of equity capital, which reflects
the opportunity cost of the fisherman's own funds,
has a less dramatic effect on annual real net in-
vestment in fishing vessels. This can be at-
tributed to the fact that only 33% of the cost of
new fishing vessels are financed with equity cap-
ital. The short run impact of an increase in the
real cost of equity capital from 5 to 10% translates
into only a 1.76% decrease in annual real net
investment in fishing vessels in the short run.
This same change in the cost of equity capital
would result in a 12.32% decrease in annual real
net investment in the long run.
Summary and Conclusions
This study evaluated aggregate investment be-
havior by fishermen for steel, wooden, and fiber-
glass fishing vessels in the Gulf of Mexico shrimp
fishery and examined the implications of changes
in the cost of acquiring debt and equity capital on
the industry's investment response. This study
showed statistical justification for the theoretical
model of aggregate investment behavior for all
three vessel types.
It is quite evident that the cost of capital plays
an important role in influencing the investment
decisions in the Gulf shrimp fishing industry.
Macroeconomic policies that lead to high real in-
terest rates depress real net investment in this
fishery. Capital expenditures for steel vessels are
the most sensitive to changes in real interest
rates while wooden vessels are the least sensitive.
While low real interest rates are desirable for
stimulating investment activities in the general
economy, they would add to the overcapitaliza-
tion problem which currently exists in the Gulf
shrimp fishing industry. Finally, this study un-
derscores the need to reinitiate efforts to collect
data on gross investment expenditures for differ-
ent categories of fishing vessels in the Gulf fleet.
Literature Cited
COEN, R M
1968. Effects of tax policy on investment in manufactur-
ing. Am. Econ. Rev. 58:200-211.
1975. Investment behavior, the measurement of deprecia-
ble and tax policy. Am. Econ. Rev. 65:59-74.
155
Griffin, W L.J P Nichols, R G Anderson. J E Buckner.and
C M Adams
1978. Cost and returns data: Texas shrimp trawlers, Gulf
of Mexico, 1974-75. Dep, Agr, Econ. Tech. Rep,, TAMU-
SG-76-601, Texas A&M Univ.
JUHL. R
1974. Economics of the Gulf of Mexico industrial and food-
fish trawlers. Mar. Fish. Rev. 36(ll):39-42.
Kmenta. J
1971. Elements of econometrics. MacMillan Company,
N.Y.
Penson, J B . R G Romain, and D W Huches
1981. Net investment in farm tractors: An econometric
analysis. Am. J. Agr. Econ. 63:629-635.
Prochaska. F J. andJ C Cato
1981. Economic conditions in the Gulf of Mexico shrimp
industry: 1960-81. Food Resour. Econ. Staff
Rep., Univ. Florida, Gainesville, FL.
Watson, J W . Jr . and C McVea, Jr
1977. Development of a selective shrimp trawl for the
southeastern United States shrimp fisheries. Mar.
Fish. Rev, 39(101:18-24,
Wilson, R R , R G Thompson, and R W Callen,
1970, Optimal investment and financial strategies in
shrimp fishing, Dep. Agr. Econ. Tech. Rep, TAMU-SG-
70-218. Texas A&M Univ.
John B Penson, Jr,
Ernest O, Tetty
Wade L Griffin
Department of Agricultural Economics
Texas A&M University
College Station, TX 77843-2124
APPENDAGE II^JURY IN DUNGENESS
CRABS, CANCER M AGISTER, IN
SOUTHEASTERN ALASKA
The Dungeness crab, Cancer magister, is com-
mercially important along the western coast of
the United States. Like many decapod crus-
taceans, it can autotomize and regenerate ap-
j>endages to heal wounds and limit injury.
Studies of appendage injury may be useful in
assessing the physical condition of crustacean
populations and the impact of fishing on commer-
cially important species. Incidences of appendage
loss in the field have been reported for species of
crabs other than C. magister (McVean 1976;
McVean and Findlay 1979; Needham 1953). Ap-
pendage loss was studied in adult Dungeness
crabs in Washington (Cleaver 1949) and Oregon
(Waldron 1958) and for juvenile crabs in the Co-
lumbia River estuary (Durkin et al. 1984).
In this study we examined adult Dungeness
crabs in southeastern Alaska to determine the
incidence of missing, regenerating, and damaged
appendages. Temporal incidence of appendage in-
jury was compared to the molting and mating
periods of the crabs and to the commercial fishing
season for Dungeness crabs.
Materials and Methods
Adult Dungeness crabs were collected from Icy
Strait and the Excursion Inlet fjord near Glacier
Bay, AK (lat. 135°30'N, long. 58°25'W), from May
through November 1984-85. Data were obtained
by monthly surveys of commercially caught
crabs. Crab pots (Waldron 1958) were set at
depths of 7 to 20 m and remained in the water for
3 to 11 days. All crabs were held in live tanks
(<24 hours) before they were measured on board
ship. In southeastern Alaska, pots are equipped
with escape rings to permit release of crabs with
carapace widths <165 mm, but sublegal-sized
crabs are often found in the catch.
Carapace width (excluding the 10th anterolat-
eral spines), wet weight, and sex were recorded
for each crab. Carapace condition was graded as
soft-shell (recent molt), new-shell, worn-shell, or
skip-molt (Somerton and Macintosh 1983). The
number and identities of missing, damaged, or
regenerating chelipeds and walking legs were
recorded. An appendage with a cracked cuticle or
missing dactyl was considered damaged. Ap-
pendages smaller in length and diameter than
intact appendages were designated regenerating.
Combined missing, damaged, and regener-
ating appendages are referred to as injured
appendages.
Interrelationships between variables were de-
termined with Pearson correlations (SAS 1985).
Means were compared with Student's ^ -tests, and
chi-square analyses were used to determine if
multiple autotomies occurred by chance (Steel
and Torrie 1960). Data are presented as means
± 1 standard error of the mean.
Results
Males comprised 65% and females 35% of the
878 Dungeness crabs examined. Average cara-
pace widths were 169 ± 0.6 and 159 ± 0.7 mm for
males and females, respectively. Wet weights
were 1,102 ± 9 g for males and 884 ± 14 g for
females. The greatest number of female crabs was
caught in July, and the greatest number of males
in August.
156
FISHERY BULLETIN; VOL 86, NO, 1, 1988.
Seventy-five percent of the crabs were intact,
with no appendage injuries. Twenty-five percent
of all crabs had injured limbs; 18% had missing,
5% had regenerating, and 2% had damaged ap-
pendages. No relationship existed between cara-
pace width and appendage injury.
Most of the Dungeness crabs sampled were in
the worn-shell condition (67%). Twenty-eight per-
cent of all crabs were new-shelled. Only 1% were
soft-shelled and 4% were skip-molts. Correlations
between carapace condition and appendage in-
jury were not significant.
No significant differences existed in appendage
injury between male and female Dungeness
crabs. Injuries were bilaterally symmetrical ex-
cept for the 3d walking leg which was missing
more frequently on the left side (P < 0.05). Con-
sidering only those crabs with missing legs, a
total of 246 legs were missing with a mean of
1.5 ±0.1 missing legs/crab. Ninety-seven crabs
had legs missing on the right side and 98 had legs
missing on the left side. The maximum number of
missing legs per crab was 5. Sixty percent of the
crabs had 1 leg missing, 20% had 2 missing legs,
and 12% were missing 3 or more legs. Of the crabs
with missing legs, 63% were males and 37% were
females.
Forty Dungeness crabs had regenerating legs,
with a mean of 1.2 ± 0.1 regenerating legs per
crab. Sixty percent of those crabs had 1 regenerat-
ing leg, 10% had 2 regenerating legs, and 3% had
3 or more regenerating legs. The maximum num-
ber of regenerating legs per crab was 4. Of the
crabs with regenerating legs, 73% were males
and 27% were females.
Seventeen crabs had damaged appendages with
a mean of 1.1 ± 0.1 damaged appendages/crab. Of
the crabs with damaged appendages, 82% were
males and 18% were females.
The observed number of Dungeness crabs with
2 or more missing appendages was significantly
higher (P <0.01) than expected for both sexes,
indicating that appendage loss was not due only
to chance.
Appendage injury was significantly correlated
with date, with more injuries occurring later in
the year. The number of Dungeness crabs with
missing appendages was significantly correlated
with date for both males and females (P < 0.01).
The lowest percentage of crabs with injured ap-
pendages occurred in July (4.8%, both sexes com-
bined) and increased to a maximum of 34.3% in
November. The percentage of male crabs with re-
generating appendages did not vary significantly
over time and was about 6% for all months. How-
ever, the percentage of female crabs with regener-
ating legs increased from 0% in May to 10.5% in
November (P < 0.01). Male crabs with damaged
appendages increased from 0% to 8.5% from May
to October (P < 0.05) and then decreased to 1.7%
in November.
Chelipeds and 1st and 4th walking legs were
injured most frequently. The hierarchy for fre-
quency of injury for female crabs (chelipeds > 4th
walking legs > 1st, 2d, and 3d walking legs) dif-
fered slightly from the hierarchy for males (che-
lipeds > 1st walking legs > 4th walking legs > 2d
and 3d walking legs). Months in which high per-
centages of crabs had injured chelipeds also had
high percentages with injured 1st (males) and 4th
(females) walking legs (Fig. 1).
The temporal incidence of appendage injury in
Dungeness crabs was compared to life history
events and to the commercial crab fishing season
in southeastern Alaska (Fig. 2). The season
opened 15 June and closed 15 August, reopened
1 October and closed 28 February 1986. Ap-
pendage injuries were low in July and increased
157% from July to August, a period of simulta-
neous molting, mating, and fishing. An addi-
tional increase in appendage injury of 43% oc-
curred in October, even though the fishery was
closed from 16 August to 30 September.
Discussion
Pot samples are biased towards larger sized
Dungeness crabs because of the size of the mesh
on the pot and the presence of two escape rings
with diameters of 11 cm. However, 62% of the
crabs collected for this study were either male
crabs with carapace widths <165 mm or were
females. Very few soft-shell crabs were caught,
even though molting was occurring during part of
the sampling period. Our dependence on commer-
cial crabbers for data collection restricted us to
sampling mostly during the open fishing season
when most of the crabs were not in the soft-
shelled condition.
Twenty-five percent of the Dungeness crabs
sampled in southeastern Alaska had appendage
injuries. In other studies of Dungeness crabs in
Washington, Oregon, and the Columbia River es-
tuary, 18%, 32%, and 62%, respectively, of the
crabs were injured (Cleaver 1949; Waldron 1958;
Durkin et al. 1984). The crabs examined in our
study were held for up to 24 hours in crowded
tanks on board ship before being measured and
157
Appendage Injury in Male Crabs
30
25
in
01
oi
(O
c
01
a.
a.
<
B
A ppendage Injury in Fe m ale Crabs
m
01
C31
ID
T3
C
01
Q
30
Figure 1. — Monthly percentages of male (A) and female (B) Dungeness crabs with injured (missing + regenerating + damaged)
chelipeds and walking legs.
•o
9
60-
50-
4 0-
30-
20-
1 0-
FISHING
d" MOLTING
9 MOLTING
J
FISHIlie
1 15 31
May
—I 1 — T"^ — I n — I 1 1
15 30 15 31 15 31 15 30 15 31
Jun Jul Aug Sep Oct
s
s
s
S
s
s
s
N
s
s
s
\
\
\
\
15 30
Nov
Males
l\\l Females
Figure 2. — Temporal relationships of percentages of male and female Dun-
geness crabs with injured (missing + regenerating + damaged) append-
ages to the crabs' molting and mating periods and the commercial crab
fishing season in southeastern Alaska.
were sometimes observed grasping other crabs,
but very few autotomized limbs were found in the
tanks.
The estimate of appendage injury may be low if
Dungeness crabs with injured appendages were
less likely to enter pots than intact crabs. In an-
other study, the observed number of Carcinus
maenas missing 2 or more legs was higher than
expected if multiple autotomies occurred by
chance, which was interpreted to mean that in-
jured C. maenas enter pots as readily as intact
crabs (McVean 1976). Because there were more
Dungeness crabs with 2 or more missing legs in
our collections than would be expected if multiple
158
autotomies occurred by chance alone, our data
could be similarly interpreted to suggest that
Dungeness crabs with injured appendages
showed little decrease in pot-entering ability and
that our estimate of injury was accurate. For this
interpretation of the chi-square results to be
valid, one must assume that all injuries occurred
before the crabs entered pots and that injury did
not occur within the pots.
Appendage injury in Dungeness crabs was bi-
laterally symmetrical except for the 3d walking
leg. Interestingly, Easton (1972) demonstrated
that 3d walking legs of Hemigrapsus oregonensis
were the most easily autotomized. In other stud-
ies, both bilateral symmetry and asymmetry have
been reported for different species of crabs
(Durkin et al. 1984; Needham 1953). Asymmetri-
cal appendage loss has been associated with crabs
that move predominantly in one direction, while
symmetrical leg loss occurs in crabs that move
randomly (Needham 1953).
The chelipeds, followed by the 1st and 4th
walking legs, were most vulnerable to injury.
Limb loss has been correlated in other studies
with degree of exposure of the limb; the outermost
limbs, the longest limbs and limbs with postures
that afford little protection are most frequently
lost (Needham 1953). Anterior limbs are lost
more frequently than posterior limbs (Needham
1953). The chelipeds are the most anterior and
one of the most exposed appendages on Dunge-
ness crabs and are frequently used in aggressive
threat postures. After loss of chelipeds, the 1st
walking legs remain as the most anterior, ex-
posed limbs and therefore, the most vulnerable.
The 4th walking legs are the most posterior and
also very exposed limbs on an intact crab.
A significantly greater number of Dungeness
crabs with 2 or more missing legs was observed
than expected if multiple autotomies occurred by
chance, indicating an increased susceptibility to
subsequent appendage loss after initial injury
(Needham 1953; Easton 1972).
The correlations between appendage injury and
date were significant but may not be biologically
important. Although these correlations were sig-
nificant, the r^ values (square of the coefficient of
variation) were low. Increased appendage injury
later in the year may be related to other factors.
Soak times, the length of time pots were left in
the water, were longer later in the year. Dunge-
ness crabs may cannibalize other crabs while con-
fined in pots (Waldron 1958). There may also be
delays between time of injury and subsequent au-
totomy and regeneration. Regeneration of legs in
Dungeness crabs is usually completed after 2 or 3
molts (Cleaver 1949).
Over the sampling period, only b^c of all crabs
had regenerating appendages while 187^ had
missing appendages. The discrepancy may be due
to increased mortality of the crabs following in-
jury (McVean and Findlay 1979), or by the effi-
cient, yearly removal of legal-sized, injured crabs
by the commercial fishery.
When temporal incidence of appendage injury
was compared with the opening and closing of the
commerical Dungeness crab fishing season, con-
siderable appendage injury occurred when the
fishery was closed. Closure of the fishery tradi-
tionally occurs during the crabs' mating period,
when a high percentage of soft-shelled female
crabs are present in the population. There was,
however, some overlap in late July and early
August in fishing, molting, and mating. Exclud-
ing damage by humans, potential causes of ap-
pendage injury are aggression between males
competing for females, the cheliped-to-cheliped
mating embrace of Dungeness crabs that can last
up to a week, cannibalism, and the increased vul-
nerability of females which molt prior to mating
(Butler 1960; Durkin et al. 1984). Damage to
Dungeness crabs can also result from other fish-
ing gear such as trawls (Reilly 1983), but no other
commercial fisheries occurred in the study area
while the Dungeness crab fishery was closed.
The results of our studies indicate that Dunge-
ness crabs in southeastern Alaska are in com-
parable condition to adult populations of Dunge-
ness crabs examined in Washington and Oregon,
in terms of appendage injury. Further studies are
needed to investigate the effect of appendage in-
jury on survival of Dungeness crabs and the con-
tribution of handling injury and mortality of
crabs in the commercial fishery.
Acknowledgements
T. Meyers and D. Erickson provided technical
assistance for which we are very appreciative. We
would like to thank T. Olsen, C. Kondzela, and
D. Sterritt for assistance in data collection. This
research was funded by Alaska Sea Grant project
R/06-20.
Literature Cited
Butler, T. H.
1960. Maturity and breeding of the Pacific edible crab,
159
Cancer magister Dana. J. Fish. Res. Board Can. 17:641-
646.
Cleaver, F. C.
1949. Preliminary results of the coastal crab {Cancer
magister) investigation. Wash. Dep. Fish., Biol. Rep.
No. 49A, 82 p.
DuRKiN, J. T., K. D. Buchanan, and T. H. Blahm.
1984. Dungeness crab leg loss in the Columbia River estu-
ary. Mar. Fish. Rev. 46(l):22-24.
Easton, D. M.
1972. Autotomy of walking legs in the Pacific Shore crab
Hemigrapsus oregonensis. Mar. Behav. Physiol. 1:209-
217.
McVean, a.
1976. The incidence of autotomy in Carcinus maenas
(L.) J. Exp. Mar. Biol. Ecol. 24:177-187.
McVean, A., and I. Findlay.
1979. The incidence of autotomy in an estuarine popula-
tion of the crab Carcinus maenas. J. Mar. Biol. Assoc.
U.K. 59:341-354.
Needham, a. E.
1953. The incidence and adaptive value of autotomy and
of regeneration in Crustacea. Proc. Zool. Soc. Lond.
123:111-122.
Reilly, p. N.
1983. Effects of commercial trawling on Dungeness crab
survival. P. W. Wild and R. N. Tasto (editors), In Life
history, environment, and mariculture studies of the
Dungeness crab. Cancer magister, with emphasis on the
central California fishery resource, p. 165-174. Calif
Dep. Fish Game Fish Bull. 172.
SAS Institute, Inc.
1985. SAS user's guide: statistics. 5 ed. Cary, NC, 956
P-
Somerton, D. a., and R. A. MACINTOSH.
1983. Weight-size relationships for three populations in
Alaska of the blue king crab, Paralithodes platypus
(Brandt, 1850) (Decapoda, Lithodidae). Crustaceana
45:169-175.
Steel, R. G. D., and J. H. Torrie.
1960. Principles and procedures of statistics with special
reference to the biological sciences. McGraw-Hill Book
Co., N.Y. 481 p.
Waldron, K. D.
1958. The fishery and biology of the Dungeness crab
(Cancer magister Dana) in Oregon waters. Fish Comm.
Oreg., Contrib. 24, 43 p.
Susan M. Shirley
Thomas C. Shirley
Juneau Center for Fisheries and Ocean Sciences
University of Alaska- Fairbanks
11120 Glacier Highway
Juneau. AK 99801
REEXAMINATION OF THE USE OF OTOLITH
NUCLEAR DIMENSIONS TO IDENTIFY
JUVENILE ANADROMOUS AND
NONANADROMOUS RAINBOW TROUT,
SALMO GAJRDNERI^
Otoliths are a potential source of taxonomic char-
acteristics for identifying stocks offish (Ihssen et
al. 1981). Differences in dimensions of the otolith
nucleus have provided a basis for separating win-
ter from summer races of steelhead, anadromous
rainbow trout, Salmo gairdneri. In addition,
otoliths provided data from which to distinguish
steelhead from resident nonanadromous forms as
well (McKern et al. 1974; Rybock et al. 1975).
Neilson et al. (1985) studied the development of
sagittal otoliths in resident rainbow trout and
steelhead from south-central British Columbia,
and identified sources of variability in the size of
otolith nuclei. However, they were unable to find
morphometric differences between the two forms
of trout. They concluded that the usefulness of
dimensions of the otolith nucleus for separating
steelhead from resident rainbow trout was much
more limited than that suggested by Rybock et al.
(1975) for rainbow trout in the Deschutes River,
Oregon.
The difference in mean length of the otolith
nuclei between the rainbow trout studied by Ry-
bock et al. (1975) and those studied by Neilson et
al. (1985) suggested either population differences
or differences in defining the nuclear boundary.
These disparate results, which led to opposite con-
clusions, limit the usefulness of measurements of
otolith nuclei for the racial identification of juve-
nile rainbow trout until the source of these differ-
ences is better understood. Consequently, to de-
termine whether juveniles of the two forms could
be distinguished by differences in dimensions of
otolith nuclei, we measured the nuclei in sagittae
from steelhead and resident rainbow trout col-
lected from the same Deschutes River, OR, loca-
tions used by Rybock et al. (1975). We used the
definitions proposed by Rybock et al. and by Neil-
son et al. (1985), and compared our measure-
ments for the two forms with each other and with
published values.
Methods
Resident rainbow trout and steelhead were col-
iQregon State University Agricultural Experiment Station
Technical Paper No. 8279.
160
FISHERY BULLETIN: VOL. 86, NO. 1, 1988.
lected from three locations in the Deschutes
River, OR. Resident rainbow trout, which were
collected from the main stem near the mouth of
Nena Creek in March 1985, were mature and
smaller (280-450 mm FL) than the steelhead,
and, based on analyses of scales and otoliths
(McKern et al. 1974), had never entered salt-
water. Juvenile progeny of steelhead were col-
lected from Round Butte Hatchery on the
Deschutes River in 1984. Wild juvenile rainbow
trout (<200 mm FL) of unknown parental origin
were collected in 1984 and 1985 from Bakeoven
Creek, an important spawning tributary for steel-
head in the Deschutes River.
Sagittae removed from rainbow trout were
stored in 90"^^ ethanol for up to two months. Be-
fore they were viewed, one otolith from each pair
was mounted (concave face up) with epoxy on a
glass slide. The back of the slide was blackened
with indelible ink. The otolith was ground by
hand with 600 grit wet sandpaper and periodi-
cally inspected under a light microscope at 100 x
until the microstructure of the nucleus, as de-
scribed by Neilson et al. (1985), was visible. The
otolith was rinsed with 59f HCl for several sec-
onds to remove scratches and improve resolution.
To reduce bias, we coded each slide with a ran-
dom number and ordered the slides sequentially
for viewing. Otoliths were examined with a Zeiss^
dissecting microscope at 125 x. A camera lucida
attachment enabled us to use a computer digitizer
to measure three dimensions of the otolith. In
measuring length and width of the central nu-
cleus, we used the first growth increment encir-
cling all the central primordia, which was the
nuclear boundary defined by Neilson et al. (1985).
In addition, we measured the maximum length
along the longest axis through an area defined by
the first metamorphic check, a narrow hyaline
ring surrounding an opaque ring with a hyaline
center, to replicate the measurements of Rybock
et al. (1975).
We used analysis of variance (ANOVA) to test
for significant differences in each dimension of
the otolith nuclei among groups in our study.
Where adequate data were available, we tested
for significant differences between groups in our
study and similar groups described by Rybock et
al. (1975) and Neilson et al. (1985) for mean di-
mensions of otolith nuclei. Neilson et al. (1985)
showed that the mean length of otolith nuclei for
rainbow trout incubated at 6.5°C was signifi-
cantly less than those for trout incubated at 9.5°
or 15.0°C. Because of this discrepancy, we evalu-
ated the potentially confounding effects of incuba-
tion temperature on the comparisons of otolith
dimensions between our samples and those of Ry-
bock et al. (1975), by testing the hypothesis that
water temperatures during 1967-69 were higher
than those during 1982-83. We used a paired t-
test of average daily water temperatures recorded
by the U.S. Geological Survey on the 1st and 15th
day of each month from 1 January to 1 August
during 1967-69 and 1982-83 (U.S. Department
of the Interior Geological Survey 1967, 1968,
1969, 1982, 1983 1. These dates represent the incu-
bation periods for most of the resident rainbow
and steelhead trout sampled in our study and by
Rybock et al. (1975). Incubation temperature for
steelhead at Round Butte Hatchery is from hatch-
ery records. We estimated spawning and incuba-
tion periods for resident rainbow and steelhead
trout on the basis of reports of the Oregon Depart-
ment of Fish and Wildlife (Fessler 1972) and per-
sonal observations.
Results
For each dimension, we failed to reject the hy-
pothesis (a = 0.05) that rainbow trout collected
from different populations for our study had
otolith nuclei of the same size (Table 1). There-
fore, we concluded that these dimensions could
not be used to discriminate between the resident
and steelhead forms of rainbow trout sampled in
our study.
Water temperatures during 1967-69 were
slightly greater than those during 1982-83
(t = 2.03, df = 14, P = 0.03). Mean difference be-
tween the two periods was 0.8°C. Spawning dates
for resident rainbow trout and steelhead differ;
steelhead spawn from January to April and resi-
Table 1. — Means, standard errors (in parentheses), and sample
size for three otolith dimensions in resident rainbow trout and steel-
head from three Deschutes River populations.
2Reference to trade name does not imply endorsement by the
National Marine Fisheries Service, NOAA.
Dimensions of nuclei
Nucleus
Nucleus
Check
Populations
No of
length
width
length
compared
fish
(mm)
(mm)
(mm)
Resident
44
0.173
0.070
0.323
rainbow trout
(0.006)
(0.003)
(0.012)
Hatchery
30
0.190
0.070
0.349
steelhead
(0.006)
(0.002)
(0.009)
Suspected
32
0.178
0.069
0.312
wild steelhead
(0.006)
(0.002)
(0.007)
161
dent rainbow trout spawn from May to mid-July
(Fessler 1972). Mean water temperature during
the period of steelhead egg incubation was 8.4°C
for 1967-69 and 7.6°C for 1982-83. Mean water
temperature during the period when resident
rainbow trout eggs were incubating in the main
stem of the river was 12.6°C in 1967-69 and
11.9°C in 1982-83. Incubation temperature for
steelhead at Round Butte Hatchery was 10°C and
did not vary.
The dimensions of otolith nuclei from resident
rainbow trout and steelhead in our study were
indistinguishable from those in fish from British
Columbia. No significant difference (a = 0.05) in
mean length of otolith nuclei existed between the
British Columbia steelhead incubated at 9.5° or
15°C and suspected wild steelhead from Bakeoven
Creek or Round Butte Hatchery steelhead incu-
bated at 10°C. Among resident rainbow trout, the
mean length of otolith nuclei for fish from the
Deschutes River was also not significantly differ-
ent from that for fish from British Columbia incu-
bated at 9.5° or 15°C. Because Rybock et al. ( 1975)
did not provide variances, we were unable to test
the hypothesis that means from our study coin-
cided with theirs. However, mean length and
width of otolith nuclei in our study were 29 and
559c less, respectively, for resident rainbow trout
and 49-70% less, respectively, for steelhead than
those studied by Rybock et al. (1975).
Discussion
The similarity of our results to those of Neilson
et al. (1985), who used similar methods, might be
expected for different populations under similar
genetic and environmental control. The disparate
results of our study and that of Rybock et al.
(1975) for the same populations after little ge-
netic change (based on comparisons of unpub-
lished, biochemical genetic data for these popula-
tions from 1972 to 1974 and 1984 to 1986) and
little environmental change partly reflected the
use of different definitions for the nucleus. We
defined the nuclear boundary as the first growth
ring surrounding all the fused primordia,
whereas Rybock et al. (1975) defined the nucleus
as the hyaline area in the center of the otolith
that is bounded by a metamorphic check formed
at hatching; they resolved the check by rendering
the otolith with HCl. We also measured the
length of the check surrounding the nucleus, as-
defined by Rybock et al. (1975), which we found
either to correspond with the area enclosed by the
first check or to increase in density of growth
increments surrounding both the central and ros-
tral primordia. The close similarity between our
estimate for Round Butte Hatchery steelhead
(0.349 mm) and the mean calculated by Rybock et
al. (1975) for steelhead (0.354 mm) suggested sim-
ilar checks. It is unclear, however, why values for
resident rainbow trout for this dimension and the
results of tests to discriminate races differed be-
tween the two studies. Rybock (1973) noted that
the nuclear check could not be distinguished in
29% of the otoliths and that the use of HCl may
have caused the frequent confusion between the
metamorphic check and other groups of daily
growth rings. The grinding and polishing of
otoliths greatly reduce this source of error. Neil-
son et al. (1985) also discouraged the use of meta-
morphic checks as boundaries because the causal
links between checks and developmental events,
such as hatching, have not yet been established.
Neilson et al. (1985) demonstrated that nuclear
length increased significantly with increase in in-
cubation temperature from 6.5° to 9.5°C but not
from 9.5° to 15°C. Although average water tem-
peratures in the Deschutes River were 0.8°C
lower during 1982-83 than in 1967-69, it is un-
likely that such differences completely explain
the greater estimates of mean length and width of
otolith nuclei in the earlier study by Rybock et al.
(1975). Rybock et al. (1975) calculated mean nu-
clear lengths and widths of 0.354 and 0.230 mm
for steelhead and 0.243 and 0.154 mm for resident
rainbow trout in the Deschutes River. Our esti-
mates were 29-70% less than their estimates for
a 0.8°C difference; whereas under controlled con-
ditions in British Columbia, mean nuclear length
for resident rainbow trout at 6.5°C was 18% less
for resident rainbow trout and 21% less for steel-
head than the nuclear length for fish incubated at
9.5°C, a difference of 3°C (Neilson et al. 1985).
Comparisons of otolith nuclear dimensions be-
tween resident rainbow trout and steelhead incu-
bated at similar temperatures would establish
whether significant differences exist for these
measurements between the two races from the
Deschutes River. The use of a common definition
of nuclear boundaries would allow better com-
parisons between studies. However, given the dis-
parate results of our study, which were similar to
the results of Neilson et al. (1985), and the origi-
nal study for steelhead and resident rainbow
trout in the Deschutes River, as well as our fail-
ure to discriminate between races using both nu-
clear definitions proposed by Neilson et al. (1985)
162
and Rybock et al. (1975), we believe that popula-
tion differences do not explain the differences in
results between the studies of Rybock et al. (1975)
and Neilson et al. (1985). Furthermore, our study
provided strong evidence to support the conclu-
sion of Neilson et al. (1985) that the usefulness of
measurements of otolith nuclei to identify sym-
patric juvenile progeny of resident rainbow trout
and steelhead reared in the wild may be limited.
Acknowledgments
We thank Jeff Light for his advice on grinding
and polishing otoliths to resolve their nuclear di-
mensions and Eric Volk for his review of this
manuscript. This research was funded by
Bonneville Power Administration, U.S. Depart-
ment of Energy, Agreement No. DE-A179-
83BP13499.
Literature Cited
Fessler, J L.
1972. An ecological and fish cultural study of summer
steelhead in the Deschutes River, Oregon. Fed. Aid
Fish. Prog. Rep. Proj. No. F-88-R-1. Oreg. State Game
Comm., Portland, OR, 47 p.
IHSSEN. P E . H E BooKE. J M Casselman, J M McGlade, N. R.
Payne, and F M Utter
1981. Stock identification: materials and methods. Can.
J. Fish. Aquat. Sci. 38:1838-1855.
McKern. J. L., H. F HoRTON, and K V. Koski
1974. Development of steelhead trout (Salmo gairdneri)
otoliths and their use for age analysis and for separating
summer from winter races and wild from hatchery stocks.
J. Fish. Res. Board Can. 31:1420-1426.
Neilson. J D . G L Geen, and B Chan
1985. Variability in dimensions of salmonid otolith nu-
clei: implications for stock identification and microstruc-
ture interpretation. Fish. Bull., U.S. 83:81-89.
Rybock. J. T
1973. Use of otoliths to differentiate juvenile steelhead
trout from juvenile rainbow trout in the lower Deschutes
River, Oregon. M.S. Thesis, Oregon State University,
Corvallis, OR, 44 p.
Rybock. J T , H F Horton, and J L Fessler.
1975. Use of otoliths to separate juvenile steelhead trout
from juvenile rainbow trout. Fish. Bull., U.S. 73:654-
659.
United States Department of the Interior Geological
Survey
1967. Water resources data for Oregon, Pt. 2. U.S. Dep.
Inter., Geol. Surv., Water quality records, 163 p.
1968. Water resources data for Oregon, Pt. 2. U.S. Dep.
Inter., Geol. Surv., Water quality records, 145 p.
1969. Water resources data for Oregon, Pt. 2. U.S. Dep.
Water quality records, 137 p.
1982. Water resources data for Oregon water year 1982.
U.S. Dep. Inter., Geol. Surv., Eastern Oregon, Vol. 1,
206 p.
1983. Water resources data for Oregon water year 1983.
U.S. Dep. Inter., Geol. Surv., Eastern Oregon, Vol. 1,
202 p.
Kenneth P. Currens
Carl B Schreck
Hiram W. Li
Oregon Cooperative Fishery Research Unit
Oregon State University
Corvallis. OR 973313
3Cooperators are Oregon State University, Oregon Depart-
ment of Fish and Wildlife, and U.S. Fish and Wildlife Service.
AGE-SPECIFIC VULNERABILITY OF
PACIFIC SARDINE, SARDINOPS SAGAX, LARVAE
TO PREDATION BY
NORTHERN ANCHOVY, ENGRAULIS MORDAX
To a large degree interannual variability in re-
cruitment determines the size of pelagic fish pop-
ulations. Recruitment to the Pacific sardine,
Sardinops sagax, population off California varies
from year to year over several orders of magni-
tude and is unrelated to spawning stock size
(Murphy 1966; MacCall 1979). Variable mortal-
ity rates in the first year of life must determine
year-class strength, although the sources of this
variability are unknown. Mortality rates in the
earliest stages are size specific with highest rates
in the egg and yolk-sac stage (Ahlstrom 1954;
Butler 1987) and may contribute to variability in
year-class strength (Smith 1985).
The sources of mortality of sardine larvae have
yet to be investigated. In other pelagic larvae,
mortality is due to either starvation or predation,
and starvation is significant only during the brief
period after the onset of feeding (O'Connell 1980;
Hewitt et al. 1985; Theilacker 1986; Owen et al.
1987). In sardines, significant mortality occurs
during the egg and yolk-sac stages (Ahlstrom
1954) and this mortality can only be due to preda-
tion. Variable mortality in older larval and juve-
nile sardines may also contribute to variability in
recruitment, and this mortality, as in other
fishes, may also be due to predation (Hunter
1984).
The objective of this paper was to determine the
size-specific vulnerability of Pacfiic sardine lar-
vae to predation by adult northern anchovies,
Engraulis mordax. The vulnerability of cape
anchovy and northern anchovy larvae to
FISHERY BULLETIN: VOL. 86, NO. 1, 1988.
163
cannibalism has been investigated by Brownell
(1985) and Folkvord and Hunter ( 1986) and found
to be an important source of mortality. In this
paper the vulnerability of sardine larvae will be
compared with that of anchovy larvae and differ-
ences in the biology of sardines and anchovies
will be discussed.
Our approach was to observe the avoidance be-
havior of Pacific sardine larvae in response to
predatory attacks by northern anchovy adults.
Adult northern anchovy were chosen as a preda-
tor because the northern anchovy was the most
abundant pelagic fish in the California Current
region during the waning years of the sardine
fishery and because its planktivorous diet in-
cludes fish eggs and larvae (Loukashkin 1970;
Hunter and Kimbrell 1980).
Materials and Methods
Experimental Fishes
The Pacific sardine larvae used in the experi-
ments were reared from eggs spawned in the lab-
oratory. Adult Pacific sardines were collected off
San Diego and held in 175 m"^ aquarium for six
months. Males and females with developing go-
nads were isolated in spawning tanks and in-
jected with 250 mg human chorionic go-
nadotropin and on the following day injected with
200 units pregnant mare serum and 20 mg
salmon pituitary extract. On the third day fertil-
ized eggs were collected from the spawning tank.
Larval rearing procedures follow those described
by Hunter (1976). Temperature in the rearing
tanks was maintained at 21°C.
Apparatus and Procedures
Experimental apparatus and procedures were
the same as those described by Folkvord and
Hunter (1986) but will be briefly outlined here.
Experimental predators were two groups of 5
adult northern anchovy (range of standard
lengths 8.4-9.2 cm). Predators were maintained
in two rectangular fiberglass tanks (0.75 x 2.15
X 0.83 m = 1.35 m"^) and fed adult brine shrimp
except on days of experimental observation. Sea-
water was supplied continuously to the tanks ex-
cept during experiments. The, temperature in the
observation tank ranged from 16.2° to 22.8°C
(mean = 20.1°C). Two 100 W incandescent lamps
produced 2,000-3,000 mc at the surface of each
tank. A black plastic tent enclosing a window on
one side of the tank provided a darkened observa-
tion chamber.
Each trial consisted of the encounter of three
prey with the predators. Prey were introduced
into the observation tank with a clear glass
beaker. Initial feeding behavior of the predators
is quite variable but becomes less variable as the
predators become accustomed to prey in the tank.
For this reason, prior to each experiment adult
Artemia were introduced as prey for five consecu-
tive trials to standardize predator behavior. After
the preliminary trials with Artemia, three trials
with sardine larvae were alternated with one
trial with brine shrimp until 18 trials with sar-
dine larvae were completed. Each experiment
was concluded with a trial of brine shrimp to test
for satiation.
The number of observations for each larval size
class was the total number of predator-prey inter-
actions observed among larvae in that size class.
The mean standard length was determined from
20 larvae sampled randomly from the rearing
tank on the day of each behavior experiment. The
numbers of observations for each size class (mean
SL) were 8.0 mm, 41; 11.3 mm, 51; 12.1 mm, 114;
12.7 mm, 46; 14.1 mm, 81; 17.6 mm, 104; and 19.6
mm, 69. Experiments were not extended to larger
sizes due to insufficient numbers of larvae.
Classification of Behavior
Prey behavior was scored only when the preda-
tor attacked a prey. Four measures of predator-
prey interactions were calculated: predator at-
tack distance, the distance from which the
predator responded to the prey and initiated its
attack; frequency of avoidance response; fre-
quency of escapes; and predation rate (percentage
of larvae captured during the 5-min trials). An
avoidance response was a change in speed or tra-
jectory of a larvae before the predator had com-
pleted its attack by closing its mouth. An escape
was defined as a larval response in which the
predator failed to capture the larvae in a single
attack. Typically adult anchovy make a single
attack on a prey item and do not pursue a prey
that escapes (Folkvord and Hunter 1986) but
rather continue searching the tank. Thus attacks
on one prey item were recorded twice if the first
attack was unsuccessful. Although predator at-
tack distance was recorded, this measurement is
highly subjective and comparison with the meas-
urements of other observers is suspect. We did not
analyze predator attack distance for this reason.
164
To compare response and escape behaviors of
Pacific sardine with those reported by Folkvord
and Hunter (1986) for northern anchovy at simi-
lar stages of development, mean lengths were
converted to ages using field growth rates back-
calculated from otolith increment widths for each
species. Confidence limits of the percentage of lar-
vae responding or escaping attack were estimated
assuming the binomial distribution.
Results
Probability of Response to Attack
The youngest larval stages of both sardine and
anchovy were the most vulnerable to predation.
Only n7c of 8 mm sardine larvae (smallest size
tested) responded to attack by adult anchovy.
With increasing size more sardine larvae re-
sponded to attack. At 20 mm, the largest size
tested, 61% of the larvae responded to attack. The
response rate of Pacific sardine larvae was consis-
tently lower than that of northern anchovy larvae
of similar lengths (Fig. 1). Although this differ-
ence in responsiveness could be due to differences
in the observer, it may also be explained by the
difference in age of anchovy and sardine at the
same length. Sardine larvae are about 6.2 mm
when they begin feeding (age = 5 days from fertil-
ization at 17°C), whereas first-feeding anchovy
larvae are only about 4.3 mm (age = 5 days from
fertilization at 17°C) (Zweifel and Lasker 1976).
Sardine larvae also grow faster than anchovy lar-
100
o 80 1-
z
a.
O 60
(A
U
OC
40
20 -
ANCHOVY
V.--
SARDINE
8 12 16
LENGTH (mm)
24
Figure 1. — Increase by size of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, responding to attack by adult northern an-
chovy and 95% confidence intervals. Data on anchovy larvae
from Folkvord and Hunter (1986).
vae at the same temperature (Butler and Rojas de
Mendiola 1985). Thus, sardine larvae are younger
at a given size than anchovy larvae.
Since the latency of response to attack must be
related to the development of the central nervous
system (Webb 1981; Webb and Corolla 1981), it
may be more appropriate to compare sardine lar-
vae with anchovy larvae of the same age. For that
reason lengths of the larvae of both species were
converted to age using growth rates measured in
the field (Methot and Kramer 1979; Butler 1987).
Comparison of the percentage of larvae respond-
ing to attack at a given age (Fig. 2) reveals no
significant difference in the rate of development
of response to attack. Thus, the escape response
develops at the same rate in Pacific sardine and
northern anchovy, and the difference in propor-
tion of larvae responding at a given size (Fig. 1)
is due to the difference of size at hatching and
the difference in growth rates of the two
species.
100
o
z
a.
<
O
If)
LU
o
OC
UJ
Q.
80 -
60 -
40
20
-
_ h
ANCHOVY . -'-
- -|
r .••■■
..?'
r .■■•••
r J
^
L
, [
kj
•■••n
1 '
1
SARDINE
1
10
20
AGE (days)
30
40
Figure 2. — Increase by age of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, responding to attack by adult northern an-
chovy and 95% confidence intervals. Size categories of reared
larvae have been converted to ages using growth rates esti-
mated from the field. Data on anchovy larvae from Folkvord
and Hunter (1986).
Probability of Escaping Attack
The ability to successfully avoid attack in-
creased with size of Pacific sardine as well as
northern anchovy. Few small larvae of either spe-
cie escaped attack by adult northern anchovy.
Only 3% of 8 mm sardine larvae escaped attack
and the percentage of larvae escaping increased
to only 11% for 17 mm larvae and 13% for 20 mm
165
larvae (Fig. 3). The proportion of small anchovy
larvae escaping attack was also low {67c ) but in-
creased with size to 737r of 22 mm larvae
(Folkvord and Hunter 1986). The numbers of lar-
vae escaping attack were significantly different
between anchovy and sardines at sizes larger
than about 13 mm. Conversion of lengths to age
using field growth rates does not eliminate the
differences between sardine and anchovy (Fig. 4).
Sardine larvae older than 20 days were more vul-
nerable to predation than anchovy larvae of the
same age (Fig. 4).
100 I—
o
z
5
z
o
a
V)
z
lU
O
a.
UJ
Q.
80
60
40
20
ANCHOVY .f
V..-- ■
-SARDINE
8 12 16
LENGTH (mm)
20
24
Figures. — Increase by size of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, escaping attack by adult northern anchovy and
95'^ confidence intervals. Data on anchovy larvae from
Folkvord and Hunter (1986).
1 \J\J
_
I
i 80
-
-]
p
.,.•'1
o
z
o
Q. 60
V)
\u
a.
^ 40
o
-
.
1
ANC
i- <
HO
r>-'
1
UJ 20
~ <
■.C-^'
J-
SARDINE
0
, ^
-l_. 1
1 F t r
c
) 1
0
20 30
40
A
GE (
days)
Figure 4. — Increase by age of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, escaping attack by adult northern anchovy and
95% confidence intervals. Size categories of reared larvae have
been converted to ages using growth rates estimated from the
field. Data on anchovy larvae from Folkvord and Hunter (1986).
Discussion
The proportion of Pacific sardine larvae re-
sponding to attack and escaping attack increased
with size and with age. Our results differ from
those reported by Folkvord and Hunter (1986) for
anchovy larvae in the rate at which sardine lar-
vae respond and escape attacks at given sizes and
ages. It should be noted that, although the
methodology was the same, the observers were
different. This difference could affect rate of re-
sponse to attack. It also should be noted that the
size of adult anchovy used by Folkvord and
Hunter (1986) ranged from 83 to 89 mm SL,
whereas the size range was 84-95 mm SL in our
study and that the size of predator influences the
number of larvae escaping (Folkvord and Hunter
1986). The slightly larger size of predators used in
this study is not sufficient to explain difference in
escapement, nor is the difference in observer
likely to affect the rate of escapement since the
observer's task is to examine whether the larvae
are escaping or are being eaten.
The greater vulnerability to predation of sar-
dine larvae than anchovy larvae has interesting
implications. In general, larger larvae are less
vulnerable to predation than small larvae. Bailey
(1984) and Bailey and Batty (1983) compared the
vulnerability of cod, flounder, plaice, and herring
larvae to predation by invertebrate predators.
They found that herring larvae were the least
vulnerable larvae because herring were more re-
active and had the greatest escape speeds. Sar-
dine larvae are larger at hatching and at a given
age are larger than anchovy larvae. In our exper-
iment sardine larvae react to predatory attacks at
similar rates as anchovy larvae, but escape attack
at a much lower rate.
This difference in vulnerability to attack may
be due to differences in swimming behavior. An-
chovy larvae swim using beat and glide locomo-
tion (Hunter 1972). The escape behavior is usu-
ally a burst of swimming from a motionless
position (Folkvord and Hunter 1986). We ob-
served that sardine larvae, however, swim contin-
uously and they respond to attack by changing
direction and increasing speed. This difference in
swimming mode may affect escape behavior in
two ways. The escape behavior of sardine larvae
may be less flexible than that of anchovy larvae
because the direction the sardine larvae takes is
largely determined by its trajectory. Since sar-
dine larvae cruise, their scope for activity (escape
behavior) may be limited. Anchovy larvae accel-
166
erate from a standing start and have the possibil-
ity of moving in a number of directions. We spec-
ulate that the beat and glide behavior of anchovy
larvae may not only be hydrodynamically more
efficient (Weihs 1974) but also may reduce the
vulnerability to predation.
Acknowledgments
Roderick Leong maintained sardines in the lab-
oratory and spawned the adults after manipula-
tion with hormones. Sandor Kaupp reared sar-
dine larvae in the laboratory. Arild Folkvord and
Clelia Booman elaborated experimental proce-
dures. John R. Hunter commented on the
manuscript.
Literature Cited
Ahlstrom, E H
1954. Distribution and abundance of egg and larval popu-
lations of the Pacific sardine. Fish. Bull., U.S. 56:83-
140.
Bailey, K. M.
1984. Comparison of laboratory rates of predation on five
species of marine fish larvae by three planktonic inverte-
brates; effects of larval size on vulnerability. Mar. Biol.
(Berl.) 79:303-309.
Bailey. K M . and R S. Batty
1983. A laboratory study of predation on Aurelia aurita on
larval herring (Clupea harengus ): experimental observa-
tions compared with model predictions. Mar. Biol.
(Berl.) 72:295-301.
Brownell, C. L.
1985. Laboratory analysis of cannibalism by larvae of the
Cape anchovy Engraulis capensis. Trans Am. Fish. Soc.
114:512-518.
Butler, J. L.
1987. Comparisons of the early life history parameters of
Pacific sardine and northern anchovy and implications
for species interactions. PhD Thesis, University of
California at San Diego, San Diego, 242 p.
Butler, J L., and B. Rojas de Mendiola
1985. Growth of larval sardines off Peru. CalCOFI Rep.
XXVI:113-118.
Folkvord. A., and J R Hunter
1986. Size-specific vulnerability of northern anchovy, En-
graulis mordax, larvae to predation by fishes. Fish.
Bull., U.S. 84:859-869.
Hewitt, R P . G H Theilacker. and N C H Lo
1985. Causes of mortality in young jack mackerel. Mar.
Ecol. Prog. Ser. 26:1-10.
Hunter. J R
1972. Swimming and feeding behavior of larval an-
chovy, Engraulis mordax. Fish. Bull., U.S. 70:821-
838.
1976. Culture and growth of northern anchovy, Engraulis
mordax, larvae. Fish. Bull., U.S. 74:81-88.
1984. Inferences regarding predation on the early life
stages of cod and other fishes. Flodevigen rapportser.,
1., ISSN 0333-2594 The propagation of cod Gadus
morhua L, p. 533-562.
Hunter. J R . and C A. Kimbrell
1980. Egg cannibalism in the northern anchovy, En-
graulis mordax. Fish. Bull., U.S. 78:811-816.
Loukashkin, a S
1970. On the diet and feeding behavior of the northern
anchovy, Engraulis mordax (Girard). Proc. Calif Acad.
Sci. 37 (Ser. 4):419-458.
MacCalL, A. E.
1979. Population estimates for the waning years of the
Pacific sardine fishery. Calif Coop. Oceanic Fish. In-
vest. 10:72-82.
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.
Murphy, G I.
1966. Population biology of the Pacific sardine (Sardinops
caerulea). Proc. Calif Acad. Sci. 34:1-34.
OConnell, C, p.
1980. Percentage of starving northern anchovy, En-
graulis mordax, larvae in the sea as estimated by histo-
logical methods. Fish. Bull., U.S. 78:475-484.
Owen, R W , J L Butler, N C H Lo. A Alvarlno. G H Thei-
lacker, J R. Hunter, J. Hakanson, and Y Watanabe.
In press. Spawning and survival patterns of larval an-
chovy in contrasting environments — a site intensive ex-
periment. Cal. Coop. Fish. Invest. Rep.
Smith, P E.
1985. Year-class strength and survival of 0-group clu-
peoids. Can. J. Fish. Aquat. Sci. 42 (Suppl. 11:69-82.
Theilacker. G H
1986. Starvation-induced mortality of young sea-caught
jack mackerel, Trachurus symmetricus , determined with
histological and morphological methods. Fish. Bull.,
U.S. 84:1-15.
Webb, P. W
1981. Responses of northern anchovy, Engraulis mordax,
larvae to predation by a biting planktivor, Amphiprion
percula. Fish. Bull., U.S. 79:727-735.
Webb. P W . and R T Corolla
1981. Burst swimming performance of northern anchovy,
Engraulis mordax, larvae. Fish. Bull., U.S. 79:143-
150.
Webb. P W , and D Weihs.
1986. Functional locomotor morphology of early life his-
tory stages of fishes. Trans. Am. Fish. Soc. 115:127.
Weihs, D.
1974. Energetic advantages of burst swimming of
fish. J. Theoret. Biol. 48:215-229.
ZWEIFEL, J E , AND R. LaSKER
1976. Prehatch and posthatch growth of fishes — a general
model. Fish. Bull., U.S. 74:609-621.
John L. Butler
Darlene Pickett
Southwest Fisheries Center
National Marine Fisheries Service, NOAA
P.O. Box 271
LaJolla.CA 92117
167
INDUCTION OF SPAWNING IN
THE WEAKFISH, CYNOSC/ON REGALIS
Reproductive activity in the weakfish, Cynoscion
regalis, is associated with extensive, north-south
migrations that result in spawning in the estuar-
ies of the Middle Atlantic Bight during the late
spring and early summer. Spawning is appar-
ently related to increasing water temperature
and day length, but there have been no experi-
mental investigations of specific factors that con-
trol this process in the weakfish. In contrast, gen-
eral aspects of the reproductive biology of the
species are well known (reviewed by Mercer
1983). Both males and females become sexually
mature at 1 year of age, and remain sexually
active throughout their lifespan (10^ years).
Spawning involves external fertilization of eggs
by pairs or small aggregations of fish.
There has been limited study of larval develop-
ment in the laboratory, and descriptions of
growth and development of larval weakfish come
entirely from field investigations (Lippson and
Moran 1974). Weakfish larvae resulting from
gametes stripped from sexually mature adults
captured in the field have been reared for a few
days on natural zooplankton diets (Public Service
Electric and Gas Company 1984), but no informa-
tion is available on mortality or growth rates on
prescribed diets and rations. In contrast, Houde
and Taniguchi (1981) and Taniguchi (1981, 1982)
have conducted extensive investigations of the ef-
fects of diet, ration, and temperature on growth
and survival of larval spotted seatrout, Cynoscion
nebulosus, under laboratory conditions. Similar
studies have been conducted with other sciaenids
including red drum, Sciaenops ocellatus (Holt and
Arnold 1983; Holt et al. 1981), spot, Leiostomus
xanthurus (Powell and Gordy 1980), and
bairdiella, Bairdiella icistia (May 1974).
The present paper describes a technique for in-
duction of spawning in a laboratory population of
weakfish and provides preliminary information
on early development and growth of weakfish lar-
vae.
Methods
Sixteen adult weakfish, C. regalis, (approxi-
mately 30-45 cm and 0.5-1.5 kg) were captured
in September 1984 in Delaware Bay by hook-and-
line. Five of these fish were dissected and found to
have regressed gonads. The remaining 11 fish
were placed in two large tanks (2,000 L) con-
nected to a 20,000 L recirculating system that
delivered temperature-controlled seawater to
each tank at 10 L/minute. Water in the recircu-
lating system was replaced approximately
monthly. Ordinary white room light was provided
by two 1.25 m fluorescent lamps positioned 1 m
above the surface of the water in each tank. Ini-
tial conditions in the tank were similar to ambi-
ent conditions in Delaware Bay in September:
18°-19°C, 30%c, and 12 hours light:12 hours dark.
Temperature and salinity in the tanks were mea-
sured daily and pH approximately weekly. The
pH was maintained between 7.0 and 7.6 by
additions of new seawater to the system; this was
accomplished by replacing 40% of the water in the
system with new seawater approximately
monthly.
Fish began to feed 5-7 days after capture. Diet
consisted of an ad libitum ration of sliced squid
with weekly additions of penaeid shrimp or fresh
calf liver. After approximately one month in the
laboratory, the fish were subjected to a prescribed
regimen of temperature and photophase (Fig. 1).
Temperatures were lowered and light phase
shortened over a period of three weeks until con-
ditions reached 8 hours light, 13°-14°C; this ap-
proximated winter conditions on the continental
shelf off Cape Hatteras where adult fish are
known to overwinter (Merriner 1976). Fish were
held under these conditions for 11 weeks after
which temperature was gradually raised and
light phase increased until spring conditions of 14
hours light and 22°-23°C were reached. Fish were
held at these conditions throughout a period of
extended spawning activity. After spawning ac-
tivity ceased, six newly captured fish were placed
in the system to replace fish that had died during
the previous year, and the process of gradually
changing temperature and photoperiod to winter
conditions was repeated (Fig. 1).
Fertilized eggs were buoyant and exited the
tanks at the surface via stand pipes that emptied
into a sump tank. The drain was located at the
bottom of the sump tank allowing eggs to accu-
mulate at the water surface. The presence or ab-
sence of eggs in the sump was determined daily
with a fine-meshed dip net. After collection, eggs
from each spawning were allowed to hatch in 20 L
plastic aquaria filled with 5.0 ixm filtered sea-
water at 30%c and 23°C. Eggs exposed to gentle
aeration under these conditions hatched in 24-36
hours.
Larvae were cultured in 2 L beakers filled with
filtered seawater (25°C, 30%f, and 5 mg/liter
168
FISHERY BULLETIN: VOL 86, NO. 1, 1988.
MAY JUNE JULY AUG SEPT
o
LU
IT
a:
Ld
Q_
25
24
23
22
21
20 H
19
18
17
16
15-1
14
13
12
II
10
▼ T
16
C/)
3
O
1^
X
X
8
1-
(3
-z.
4
LU
_l
>-
0
<
O
OCT NOV
DEC
JAN
FEB MAR APR
MAY
Figure 1. — Temperature-photophase regime used to condition northern weakfish, Cynoscion regalis, to spawn
in the laboratory. Values are 7-d means. Standard deviations are plotted for temperature. Arrows on time axis
indicate dates of spawning. A = 1984-85. B = 1985-1986.
169
chloramphenicol) at a density of 25 larvae/liter.
Larvae were fed rotifers, Brachionis plicatilis, be-
ginning two days after hatching; on the seventh
day after hatching, brine shrimp Artemia sp.
were added to the diet.
Early larval development from spawning to
first feeding was determined for larvae from three
separate spawnings. Newly hatched larvae (6-15
hours old) were pipetted into an aquarium and
10-15 removed immediately and preserved in
70% ethanol. Additional larvae were sacrificed
daily for nine days. These samples were examined
to relate early development and age.
Results and Discussion
Three adult fish died of undermined causes dur-
ing the period of winter conditions in 1984-85.
Four additional fish died after jumping from the
tanks. The initial spawning by the remaining
population of four fish occurred five weeks after
spring conditions were achieved. This was fol-
lowed the next day by another spawning. Spawn-
ing continued for nine weeks at 10-14 d intervals.
Spawning episodes usually consisted of produc-
tion of fertilized eggs over 2 successive days.
While actual spawning was never observed, it al-
ways occurred between sunset and 08:00 the fol-
lowing day.
After cessation of spawning, the four fish were
removed from the tanks and their sex deter-
mined. Two of the fish drummed when handled
and were clearly males. At least one (and proba-
bly both) of the two remaining fish were females.
The fish were returned to the system after exam-
ination.
The temporal sequence of the spawning events
(two successive days at 10-14 d intervals) sug-
gests that each female may have spawned as
many as four times during the 9-wk period. This
is in contrast to reports for natural populations of
C. regalis (e.g., Merriner 1976), but has been
reported for laboratory populations of other
sciaenid fishes. For example, Arnold (1984) ob-
served 82 spawning events in a laboratory popu-
lation of 12 C. nebulosus over a 27-mo period and
52 spawning events in a similar population of six
S. ocellatus over a 3-mo period. It is not clear why
the C. regalis in our investigation ceased spawn-
ing after several months of long days and high
temperatures while the C. nebulosus in Arnold's
system continued to spawn over a much longer
period. Perhaps this is related to the smaller an-
nual variation in photophase and temperature
typical of C. nebulosus habitats. However, go-
nadal resorption has been reported for at least
one other sciaenid species iB. icistia) held in the
laboratory for extended periods of long day-length
(May 1974).
Spawning from the 1985-86 conditioning pe-
riod first occurred four weeks after spring
conditions were reached with a second spawning
nine days later. Further spawning in the 1985-
86 population did not occur because of an
unidentified infection resulting in the death of all
10 fish in the system over a period of a few weeks.
Autopsies revealed that all fish had highly
developed ovaries or testes at the time of
death.
Newly hatched larvae (6-15 hours old) had a
yolk sac, no mouth, and little development of the
eyes. By 24-36 hours after hatching, the yolk sac
had been virtually absorbed, the mouth was just
beginning to form, and the eyes were not yet pig-
mented. By 48-60 hours, larvae had a completely
formed mouth and digestive system, and the eyes
were pigmented. Larvae were capable of feeding
at this stage.
Chloramphenicol improved larval survival,
which was as high as 24% over 11 days. This
survival is comparable to that reported for other
sciaenid larvae (Holt et al. 1981; Houde and
Taniguchi 1981; Holt and Arnold 1983). However,
growth was less than the maximum seen in the
laboratory for C. nebulosus. After 11 days C. re-
galis larvae in the present experiments had
grown from a mean, posthatching size of 2.7 mm
(36 fxg dry weight) to 4.5 mm (235 \i.g dry weight).
In contrast Houde and Taniguchi (1981) found
that one group of C. nebulosus larvae reached a
size of 13.6 mm (7,082 fxg dry weight) in 12 days
when reared on a concentrated ration of natural
zooplankton at very low stocking density and
high temperature (32°C). However, when fed a
rotifer diet at comparable temperatures and
stocking densities, growth of C. nebulosus larvae
was somewhat less than that of C. regalis larvae
in the present experiments.
Our results show that the spawning cycle of
weakfish can be manipulated to produce repeated
spawnings without the aid of hormone injections.
While the fish appear to resorb their gonadal tis-
sue after several months of exposure to long day
length and high temperature, the differential ma-
nipulation of several groups of fish could allow
year-round production of fertilized eggs. Further-
more, survival and growth of larvae produced in
this manner appear comparable to survival and
170
growth of other sciaenids reared in the labora-
tory.
Acknowledgments
This study was supported by funds provided by
the University of Delaware Sea Grant College
Program and by the U.S. Department of Fish and
Wildlife through the Delaware Department of
Natural Resources and Environmental Control.
Paul Cosby, Big John Ellsworth, Anne Marie Ek-
lund, John Ewart, Paul Grecay, Anne Hastings,
Kathleen Little, Leslie Picoult, and Peter Rowe
were indispensible in the capture, care, and feed-
ing of the adult and larval weakfish used in the
investigation.
Literature Cited
Arnold, C R
1984. Maturation and spawning of marine finfish. In
Carl J. Sindermann (editor). Proceedings of the seventh
U.S. -Japan meeting on aquaculture, marine finfish cul-
ture held at Tokyo, Japan, October 3-4, 1978, p. 25-
27. U.S. Dep., NOAA Tech. Rep. NMFS 10.
Holt, G J , and C R. Arnold
1983. Effects of ammonia and nitrite on growth and sur-
vival of red drum eggs and larvae. Trans. Am. Fish. Soc.
112:314-318.
Holt, G J , R Godbout, and C R. Arnold
1981. Effects of temperature and salinity on egg hatching
and larval survival of red drum, Sciaenops ocel-
lata. Fish. Bull., U.S. 79:569-573.
Houde, E D.. and a K Taniguchi
1981. Marine fish larvae growth and survival. Effects of
density-dependent factors: spotted seatrout iCynoscion
nebulosus ) and lined sole iAchirus lineatus ). Report to
U.S. Environmental Protection Agency, Environmental
Research Laboratory, Narragansett, R.I. EPA 600/3-81-
052. Natl. Tech. Inf Serv. PB82-101395, 66 p.
LiPPSON, A J , and R L. Moran
1974. Manual for identification of early developmental
stages of fishes of the Potomac River estu-
ary. Environmental Technology Center, Martin Mari-
etta Corporation, Baltimore, MD, 282 p.
May,R C
1974. Effects of temperature and salinity on fertilization,
embryonic development, and hatching in Bairdiella icis-
tia (Pisces:Sciaenidae), and the effect of parental salinity
acclimation on embryonic and larval salinity toler-
ance. Fish. Bull., U.S. 73:1-22.
Mercer. L P.
1983. A biological and fisheries profile of weakfish,
Cynoscion regalis. N.C. Dep. Nat. Resour. Comm. Dev.
Div. Mar. Fish., Spec. Sci. Rep. No. 39, 107 p.
Merriner, J V.
1976. Aspects of reproductive biology of the weakfish,
Cynoscion regalis, in North Carolina. Fish. Bull., U.S.
76:18-26.
Powell, A B , and H R Gordy
1980. Egg and larval development of the sp)ot Leiosto-
mus xanthurus (Sciaenidae). Fish. Bull., U.S. 78:701-
714.
Public Service Electric and Gas Company.
1984. Salem Generating Station 316 (b) Demonstration.
Appendix XI. Weakfish (Cynoscion regalis): A synthesis
of information of natural history with reference to occur-
rence in the Delaware River and Estuary and involve-
ment with the Salem Generating Station. Public Ser-
vice Electric and Gas Company, Newark, NJ.
Taniguchi, A K.
1981. Survival and growth of larval spotted seatrout
(Cynoscion nebulosus ) in relation to temperature, prey
abundance and stocking densities. Rapp. P. -v. Reun.
Cons. int. Explor. Mer 178:507-508.
1982. Growth efficiency estimates for laboratory-reared
larval spotted seatrout fed microzooplankton or roti-
fers. In C. F. Bryan, J. V. Conner, F. M. Truesdale (ed-
itors), Proceedings of the Fifth Annual Larval Fish Con-
ference, p. 6-11. Louisiana State University,
Cooperative Fishery Research Unit, Baton Rouge, LA.
Charles E Epifanio
David Goshorn
Timothy E. Targett
College of Marine Studies
University of Delaware
Lewes, DE 19958
171
ERRATA
Fishery Bulletin: Vol. 85, NO. 4
Botton, Mark L., and John W. Ropes, "Populations of horseshoe crabs, Limulus
polyphemus , on the northwestern Atlantic continental shelf," p. 809.
Page 809, Table 3, footnote was omitted:
'Depth at the end of this tow was 439 m
Henwood, Tyrrell A., and Warren E. Stuntz, "Analysis of sea turtle captures and
mortalities during commercial shrimp trawling," p. 814.
Page 814, paragraph 1, line 3, the equation should read:
E = (nets * length/30.5 m) * (min/60)
INFORMATION FOR CONTRIBUTORS TO THE FISHERY BULLETIN
Manuscripts submitted to the Fishery Bulletin will reach print faster if they conform to the following instructions.
These are not absolute requirements, of course, but desiderata.
CONTENT OF MANUSCRIPT
The title page should give only the title of the paper,
the author's name, his/her affiliation, and mailing ad-
dress, including ZIP code.
The abstract should not exceed one double-spaced
page.
In the text, Fishery Bulletin style, for the most part,
follows that of the U.S. Government Printing Office
Style Manual. Fish names follow the style of the Amer-
ican Fisheries Society Special Publication No. 12, A List
of Common and Scientific Names of Fishes from the
United States and Canada , Fourth Edition, 1980.
Text footnotes should be typed separately from the
text.
the literature cited. The accuracy of the literature
cited is the responsibility of the author. Abbreviations
of names of periodicals and serials should conform to
Serial Sources for the BIOSIS Data Base. (Chemical Ab-
stracts also uses this system, which was developed by
the American Standards Association.)
Common abbreviations and symbols, such as mm,
m, g, L, mL, mg, °C (for Celsius), %, %o, and so forth,
should be used. Abbreviate units of measure only when
used with numerals. Periods are only rarely used with
abbreviations.
We prefer that measurements be given in metric
units; other equivalent units may be given in parenthe-
ses.
FORM OF THE MANUSCRIPT
Figures and tables, with their legends and headings,
should be self-explanatory, not requiring reference to
the text. Their placement should be indicated in the
right-hand margin of the manuscript.
Preferably figures should be reduced by photography
to 53/4 inches (for single-column figures, allowing for
50% reduction in printing), or to 12 inches (for double-
column figures). The maximum height, for either width,
is 14 inches. Photographs should be printed on high
quality or glossy paper.
Do not send original drawings to the Scientific Editor;
if they, rather than the photographic reductions, are
needed by the printer, the Scientific Publications Office
will request them.
Each table should start on a separate page. Consis-
tency in headings and format is desirable. Vertical rules
should be avoided, as they make the tables more expen-
sive to print. Footnotes in tables should be numbered
sequentially in arable numerals. To avoid confusion
with powers, they should be placed to the left of numer-
als.
Acknowledgments, if included, are placed at the
end of the text.
Literature is cited in the text as: Lynn and Reid
(1968) or (Lynn and Reid 1968). All papers referred to in
the text should be listed alphabetically by the senior
author's surname under the heading "Literature Cited."
Only the author's surname and initials are required in
The original of the manuscript should be typed,
double-spaced, on white bond paper. Please triple space
above headings. We would rather receive good dupli-
cated copies of manuscripts than carbon copies. The se-
quence of the material should be;
TITLE PAGE
ABSTRACT
TEXT
LITERATURE CITED
TEXT FOOTNOTES
APPENDIX
TABLES (Each table should be numbered with an
arable numeral and heading provided).
LIST OF FIGURES (Entire figure legends)
FIGURES (Each figure should be numbered with
an arable numeral; legends are desired)
ADDITIONAL INFORMATION
Send the ribbon copy and two duplicated or carbon
copies of the manuscript to:
Dr. Andrew E. Dizon, Scientific Editor
Fishery Bulletin
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 271
La Jolla, CA 92038
Fifty separates will be supplied to an author free of
charge and 50 supplied to his organization. No covers
will be supplied.
Contents — Continued
CURRNES, KENNETH P., CARL B. SCHRECK, and HIRAM W. LI. Reexamina-
tion of the use of otolith nuclear dimensions to identify juvenile anadromous and -^
nonanadromous rainbow trout, Salmo gairdneri 160
BUTLER, JOHN L., and DARLENE PICKETT. Age-specific vulnerability of
Pacific sardine, Sardinops sagax, larvae to predation by northern anchovy, En-
graulis mordax 163
EPIFANIO, CHARLES E., DAVID GOSHORN, and TIMOTHY E. TARGETT.
Induction of spawning in the weakfish, Cynoscion regalis 168
GPO 791-008
kt^'
Co.
Fishery Bulletin
•^'■^rEs o<
P^^^^^^^^
Vol. 86, No. 2
ocr^8
ms
April 1988
Migratibns- of jjoho' sa^lmon,
PEARCY, WILLIAM G., and JOSEPH P. FISHI
Oncorhynchus kisutch, during their first summer in the ocifta'n-?'*.^.^*^.;^^. T?: . . . 173
DUTIL, J.-D. and J.-M. COUTU. Early marine Hfe of Atlantic salmon/Sa?7m>
salar, postsmolts in the northern Gulf of St. Lawrence 197
MURPHY, MICHAEL L., JOHN F. THEDINGA, and K V. KOSKI. Size and diet
of juvenile Pacific salmon during seaward migration through a small estuary in
southeastern Alaska 213
BOLZ, GEORGE R., and R. GREGORY LOUGH. Growth through the first six
months of Atlantic cod, Gadus morhua , and haddock, Melanogrammus aeglefinus ,
based on daily otolith increments 223
NYMAN, ROBERT M., and DAVID O. CONOVER. The relation between spawn-
ing season and the recruitment of young-of-the-year bluefish, Pomatomus salta-
trix , to New York 237
JAHN, A. E., D. M. GADOMSKI, and M. L. SOWBY. On the role of food-seeking
in the suprabenthic habit of larval white croaker, Genyonemus lineatus (Pisces:
Sciaenidae) 251
WILLIAMS, AUSTIN B. New marine decapod crustaceans from waters influenced
by hydrothermal discharge, brine, and hydrocarbon seepage 263
MARTIN, JOEL W., FRANK M. TRUESDALE, and DARRYL L. FELDER. The
megalopa stage of the Gulf stone crab, Menippe adina Williams and Felder, 1986,
with comparison of megalopae in the genus Menippe 289
SHENKER, JONATHAN M. Oceanographic associations of neustonic larval and
juvenile fishes and Dungeness crab megalopae off Oregon 299
DAGG, M. J., P. B. ORTNER, and J. AL-YAMANI. Winter-time distribution and
abundance of copepod nauplii in the northern Gulf of Mexico 319
HERRNKIND, WILLIAM F., MARK J. BUTLER IV, and RICHARD A. TANKERS-
LEY. The effects of siltation on recruitment of spiny lobsters, Panulirus
argus 331
KIRKLEY, JAMES E., and DALE E. SQUIRES. A limited information approach
for determining capital stock and investment in a fishery 339
POLACHECK, TOM. Analyses of the relationship between the distribution of
searching effort, tuna catches, and dolphin sightings within individual purse seine
cruises 351
(Continued on back cover)
Seattle, Washington
U.S. DEPARTMENT OF COMMERCE
C. William Verity, Jr., Secretary
NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION
William E. Evans, Under Secretary for Oceans and Atmosphere
NATIONAL MARINE FISHERIES SERVICE
James W. Brennan, Assistant Administrator for Fisheries
Fishery Bulletin
The Fishery Bulletin carries original research reports and technical notes on investigations 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. Beginning 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 instead of being issued individually. 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.
SCIENTIFIC EDITOR, Fishery Bulletin
Dr. Andrew E. Dizon
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NCAA
P.O. Box 271
La Jolla, CA 92038
Editorial Committee
Dr. Jay Barlow
National Marine Fisheries Service
Dr. William H. Bayliff
Inter-American Tropical Tuna Commission
Dr. George W. Boehlert
National Marine Fisheries Service
Dr. Robert C. Francis
University of Washington
Dr. James R. Kitchell
University of Wisconsin
Dr. William J. Richards
National Marine Fisheries Service
Dr. Bruce B. CoUette
National Marine Fisheries Service
Dr. Tim D. Smith
National Marine Fisheries Service
Mary S. Fukuyama, Managing Editor
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. Second class postage is paid in Seattle, Wash., and additional
offices. POSTMASTER send address changes for subscriptions to Fishery Bulletin, Superintendent of Documents, U.S. Government
Printing Office, Washington, DC 20402.
Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated.
The Secretary of Commerce has determined that the publication of this periodical is necessary in the transaction of the public business
required by law of this Department. Use of funds for printing this periodical has been approved by the Director of the Office of
Management and Budget.
Fishery Bulletin
CONTENTS / LIBRARY
OCT 2 8 )98g '
Vol. 86, No. 2 I April1988
PEARCY, WILLIAM G., and JOSEPH P. FISHERi..,Migrfcns of coho salr^^n, :
Oncorhynchus kisutch , during their first summer in thr nrrTHl i i , J.73
DUTIL, J.-D. and J.-M. COUTU. Early marine life of Atlantic salmon, Salmo
salar, postsmolts in the northern Gulf of St. Lawrence 197
MURPHY, MICHAEL L., JOHN F. THEDINGA, and K V. KOSKI. Size and diet
of juvenile Pacific salmon during seaward migration through a small estuary in
southeastern Alaska 213
BOLZ, GEORGE R., and R. GREGORY LOUGH. Growth through the first six
months of Atlantic cod, Gadus morhua , and haddock, Melanogrammus aeglefinus ,
based on daily otolith increments 223
NYMAN, ROBERT M., and DAVID O. CONOVER. The relation between spawn-
ing season and the recruitment of young-of-the-year bluefish, Pomatomus salta-
trix , to New York 237
JAHN, A. E., D. M. GADOMSKI, and M. L. SOWBY. On the role of food-seeking
in the suprabenthic habit of larval white croaker, Genyonemus lineatus (Pisces:
Sciaenidae) 251
WILLIAMS, AUSTIN B. New marine decapod crustaceans from waters influenced
by hydrothermal discharge, brine, and hydrocarbon seepage 263
MARTIN, JOEL W., FRANK M. TRUESDALE, and DARRYL L. FELDER. The
megalopa stage of the Gulf stone crab, Menippe adina Williams and Felder, 1986,
with comparison of megalopae in the genus Menippe 289
SHENKER, JONATHAN M. Oceanographic associations of neustonic larval and
juvenile fishes and Dungeness crab megalopae off Oregon 299
DAGG, M. J., P. B. ORTNER, and J. AL-YAMANI. Winter-time distribution and
abundance of copepod nauplii in the northern Gulf of Mexico 319
HERRNKIND, WILLIAM F., MARK J. BUTLER IV, and RICHARD A. TANKERS-
LEY. The effects of siltation op recruitment of spiny lobsters, Panulirus
argus 331
KIRKLEY, JAMES E., and DALE E. SQUIRES. A limited information approach
for determining capital stock and investment in a fishery 339
POLACHECK, TOM. Analyses of the relationship between the distribution of
searching effort, tuna catches, and dolphin sightings within individual purse seine
cruises 351
(Continued on next page )
Seattle, Washington
1988
For sale by the Superintendent of Documents, U.S. Government Printing OfTice, Washing-
ton DC 20402 — Subscription price per year: $16.00 domestic and $20.00 foreign. Cost per
single issue: $9.00 domestic and $11.25 foreign.
Contents — Continued
WATSON, CHERYL, ROBERT E. BOURKE, and RICHARD W. BRILL. A compre-
hensive theory on the etiology of burnt tuna 367
BROWN-PETERSON, NANCY, PETER THOMAS, and CONNIE R. ARNOLD.
Reproductive biology of the spotted seatrout, Cynoscion nebulosus, in South
Texas 373
Notes
CHEN, CHE-TSUNG, TZYH-CHANG LEU, and SHOOU-JENG JOUNG. Notes
on reproduction in the scalloped hammerhead, Sphyrna lewini, in northeastern
Taiwan waters 389
COLLINS, MARK R., and CHARLES A. WENNER. Occurrence of young-of-the-
year king Scomberomorus cavalla, and Spanish, S. maculatus, materials in
commercial-type shrimp trawls along the Atlantic coast of the southeast United
States 394
DEW, C. BRAXTON. Stomach contents of commercially caught Hudson River
striped bass, Morone saxatilis , 1973-1975 397
VERNET, MARIA, JOHN R. HUNTER, and RUSSELL D. VETTER. Accumula-
tion of age pigments (lipofuscin) in two cold-water fishes 401
MULLIN, M. M., and E. R. BROOKS. Extractable lipofuscin in larval marine
fish 407
Notices: NOAA Technical Reports published during the last 6 months of 1987 . . . 416
The National Marine Fisheries Service (NMFS) does not approve, recommend or
endorse any proprietary product or proprietary material mentioned in this publi-
cation. 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 approves, recommends or endorses any proprietary product or proprietary
material mentioned herein, or which has as its purpose an intent to cause directly
or indirectly the advertised product to be used or purchased because of this NMFS
publication.
MIGRATIONS OF COHO SALMON, ONCORHYNCHUS KISUTCH ,
DURING THEIR FIRST SUMMER IN THE OCEAN
William G. Pearcy and Joseph P. Fisher'
ABSTRACT
Marked juvenile coho salmon caught in fine-meshed purse seines during the summers of 1981-84 off
Oregon and Washington generally demonstrated northward migrations from their rivers of ocean
entrance. Northward movements in summer were preceded by southerly movements during spring,
probably caused by southerly advection. Catch rates and sizes offish caught in different months and
regions of the coast also indicated northerly movements of both yearling and subyearling coho
salmon. Despite this movement, the average catch of juvenile coho salmon per purse seine set along
the coasts of Washington and Oregon in late summer, including marked fish from the Columbia
River, was still a substantial proportion of that in May and June soon after ocean entrance, suggest-
ing that many coho did not migrate great distances. Additionally, recoveries of marked juvenile coho
salmon by sports and commercial fishermen from Alaska to California and by scientists in Alaska
were generally in the region of release. These data indicate that migrations of juvenile coho are of
limited extent during their first summer in the ocean and are not strong support for an earlier
conclusion that juvenile coho salmon from the Columbia River, Oregon, and California may form a
large proportion of the stocks of this species that migrate northward along the coastal belt in Cana-
dian and Alaskan waters each summer.
Although there has been Httle research on juve-
nile salmon during their first summer at sea, this
phase of the life history may be critical to survival
and recruitment to fisheries (Hartt 1980). High-
est ocean mortality is thought to occur early in
marine life (Foerster 1968; Parker 1968; Ricker
1976). Production (catch and escapement) of adult
coho salmon, Oncorhynchus kisutch, in the Ore-
gon Production Index (OPI) Area (from Leadbet-
ter Point, WA, to Monterey Bay, CA) is usually
accurately predicted in one year by the number of
precocious males (jacks) returning to index
streams in the previous year (Gunsolus 1978; Or-
egon Department of Fish and Wildlife 1982;
Pacific Fishery Management Council 1986).
Hence survival from jacks to adults is fairly con-
stant from year to year. Because survival rates
from smolt to adult are variable (Nickelson 1986),
however, variable year-class survival must occur
before the time that jacks return, after only a few
months in the ocean. This relationship, and the
positive correlation between coastal upwelling
and survival of OPI coho salmon (Gunsolus 1978;
Scarnecchia 1981; Nickelson 1986), strongly sug-
gest that the first few months in the ocean consti-
tute the "critical period" in determining subse-
quent adult survival.
iCollege of Oceanography, Oregon State University, Corval-
lis, OR 97331.
Between 1976 and 1985 the production of coho
salmon in the OPI area drastically declined, de-
spite large increases in the number of public and
private smolt releases (Oregon Department of
Fish and Wildlife 1982; Nickelson 1986). Reduced
upwelling and ocean productivity, perhaps cou-
pled with density-dependent mortality, is one of
the hypothesized causes for this decrease in sur-
vival (Scarnecchia 1981; Peterman and Rout-
ledge 1981; McCarl and Rettig 1983; McGie 1984;
Nickelson 1986). To understand the mechanisms
affecting survival of juvenile salmonids at sea, we
must first know where salmon reside at the time
of their high and variable mortality. Are the
smolts highly migratory, immediately leaving
local coastal waters and migrating into waters of
the Gulf of Alaska (Hartt and Dell 1986), or are
they nonmigratory, spending their early ocean
life in local coastal waters?
This paper summarizes research on the move-
ments and migrations of coho salmon during their
first summer in the ocean in the northeastern
Pacific Ocean based on purse seine catches made
mainly in coastal waters off Oregon and Wash-
ington. A few records of migrations of tagged ju-
venile (age .0)^ coho salmon were given by God-
Manuscript accepted November 1987.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.
2The numeral preceding and following the decimal indicate
the number of winters spent in fresh water and in the ocean,
respectively.
173
FISHERY BULLETIN: VOL. 86, NO. 2
frey (1965) and French et al. (1975), but by far the
most comprehensive data were provided by Hartt
(1980) and Hartt and Dell (1986). All of these
studies, however, were based on recovery of ma-
ture or maturing coho in the year following tag-
ging and on tagging in northern waters from the
Strait of Juan de Fuca to the Alaska Peninsula.
The only other studies of juvenile salmon in the
ocean off Washington and Oregon have been re-
stricted to within 24 km of the Columbia River
(Dawley et al. 1981) or to coastal waters during
1980 (Miller et al. 1983). Our 1979-85 research,
covering large areas along the coast, provides ex-
tensive and unique data on the movements of ju-
venile coho salmon during their first summer in
the ocean.
METHODS
Purse seines, our primary sampling method,
were used to sample juvenile salmonids during
1979-85 (Table 1). Cruises were in coastal waters
off Oregon in 1979-80, off Oregon and southern
Washington in 1981, and off Oregon and the en-
tire Washington coast during the summers of
1982-85 (Fig. 1). During July 1984, sets were also
made from northern California (lat. 40°32'N) to
northern Vancouver Island (50°26'N). Except for
the exploratory cruises off Oregon in 1979 and
1980, purse seine sets were usually made along
east-west transect lines (Fig. 1). Sets started at
the 37 m (20-fathom) contour, and continued at
9.3 km (5-mi) intervals farther offshore, usually
until no salmonids were captured. Repeat sets
were sometimes made when fish with missing
adipose fins were common, indicating the pres-
ence of coded wire tagged (CWT) fish. In 1985,
special sets were made in the vicinity of the Co-
lumbia River plume. Detailed sampling data are
provided in Pearcy (1984) and our cruise reports
(Wakefield et al. 1981; Fisher et al. 1982, 1983;
Fisher and Pearcy 1984, 1985).
The mesh size of the seines were the same dur-
ing all years, 32 mm (stretch), with 32 mm or
smaller mesh in the bunts of the seines. The seine
was 495 m long except in 1981 (457 m). Depths
that seines fished, sometimes measured with a
depth gauge on the lead line, varied among years
from about 20 m to 65 m (Table 1).
Generally, purse seine sets were "round hauls",
where the seiner and the skiff made a circle with
the net. The seine was fully pursed after about
one-half its length was aboard (half-purse sets).
All sets were "blind". We attempted to use sonar
on some cruises to locate concentrations of sal-
monids but were unsuccessful. Radar was some-
times used to determine the distance between the
seiner and the skiff when a semicircle was made
with the net. Each round haul encompassed about
17,000 m2 (1981) or 19,000 m2 (1979-85). To de-
termine the direction of movement of fish, eight
"half-round" hauls, or "semicircular" sets, were
made in 1979, where the entire net formed an
open semicircle. Paired sets were made in close
succession, with sets open in a northern and a
southern direction, at four locations. The seine
was open for the same duration (15-45 minutes,
depending on location) in each paired set while
Table 1. — Summary of number of purse seine sets and latitudinal range of sampling, 1979-85.
Dates of
No. of
Purse
seine
Lengtfi
Depthi
Year
cruises
setsi
Latitudinal range of sampling
(m)
(m)
1979
18-29 June
56
Cape Disappointment to Cape Arago
46°20'
-43°18'
495
20
1980
20-28 June
36
Cape Disappointment to Alsea River
46°20'
-44°30'
495
20
1981
16-25 May
63
Willapa Bay to Alsea River
46°35'
-44°25'
495
20
9-18 June
67
Willapa Bay to Cut Creek
46°35'
-43°1 1 '
495
20
9-19 July
67
Willapa Bay to Alsea River
46°35'
-44°25'
457
49
8-19 Aug.
66
Willapa Bay to Cut Creek
46°36'
-44°1 1 '
457
49
1982
19 May-2 June
62
Waatch Point to Siuslaw River
48°2r
-44°00'
495
265
7-22 June
57
Quinault River to Yachiats
47°21 '
-44°20'
495
265
4-14 Sept.
42
Quinault River to Yachats
47°20'
-44°19'
495
265
1983
16-27 May
56
Waatcfi Point to Yachats
48°2r
-44°20'
495
249
9-27 June
58
Waatcfi Point to Four Mile Creek
48°20'
-43°00'
495
249
15-24 Sept.
53
Waatch) Point to Coos Bay
48°20'
-43°28'
495
249
1984
4-20 June
69
Waatcfi Point to Coos Bay
48''20'
-43°27'
495
249
9 July-3 Aug.
65
Winter Harbor, B.C. to False Cape. CA
50°26'
-40°32'
495
249
1-15 Sept.
63
Waatcfi Point to Siuslaw River
48°20'
-44°00'
495
249
1985
29 May-25 June
112
Sea Lion Rock to Coos Bay
48°00'
-43°27'
495
225
^Quantitative sets.
^Measured with depth guage.
174
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
Figure l. — Locations of purse seine transects
off the Oregon and Washington coasts.
WAATCH POINT
SEA LION ROCK
DESTRUCTION ISLAND
QUINAULT RIVER
GRAYS HARBOR
WILLAPA BAY
CAPE DISAPPOINTMENT
SEASIDE
NEHALEM BEACH
CAPE LOOKOUT
WECOMA BEACH
YAOUINA HEAD
YACHATS
SIUSLAW RIVER
COOS BAY
FOUR MILE CREEK
125
124"
the vessel and skiff towed the seine only fast
enough to maintain a constant net opening.
The purse seine catches were either dip-netted
from the bunt of the seine while it was alongside
the vessel, brailed aboard, or hauled aboard in the
bunt, depending on the composition and size of
the catch.
In 1979 and 1980, juvenile salmon preserved in
formalin were identified ashore. In 1981-85, ju-
venile salmon were identified to Species at sea,
fork length (FL) was measured to the nearest mil-
limeter and then they were individually wrapped
in labelled plastic bags and frozen. All salmonids
with marks or missing adipose fins were frozen.
When large numbers of juvenile salmonids were
caught in a set, most unmarked fish were re-
leased after they were measured.
In order to increase the numbers of marked fish
released into our study area we marked about 1.5
million coho smolts in 1981 and 835,000 in 1982
using fluorescent pigment propelled by com-
pressed air (see Phinney et al. 1967) prior to their
175
FISHERY BULLETIN: VOL. 86. NO. 2
transport from Oregon Aqua-Foods, Inc. (OAF)
hatchery to their ocean release facilities at
Yaquina Bay or Coos Bay, OR.
In the laboratory ashore, species identifications
were confirmed and individuals remeasured and
reexamined for both fluorescent marks (under ul-
traviolet light 1981-82) and missing adipose fins
or other marks (1979-85). Coded-wire tags from
the heads of salmonids with missing adipose fins
were decoded by personnel from the Oregon De-
partment of Fish and Wildlife, Clackamas Labo-
ratory.
Juvenile or age .0 (first year in the ocean) coho
salmon were distinguished from adult or age .1
(second year in the ocean) coho salmon by exami-
nation of size-frequency histograms and scales.
The division between age .0 and .1 coho pro-
gressed from approximately 300 to 420 mm FL
from May to September, in most years. Most coho
salmon migrated to the ocean a little over one
year after hatching (age 1.0), but OAF released
large numbers of subyearling (age 0.0) smolts
into Yaquina Bay and Coos Bay. These two age
groups of smolts were distinguished by the radial
distance to the 21st circulus on scales removed
from the preferred area (Clutter and Whitesel
1956) of the fish. The accuracy of this method for
distinguishing known age 0.0 and 1.0 fish was
approximately 85-90%. In the years 1981-85,
scales from 52% of the 4,222 juvenile coho sam-
pled were analyzed. The estimated numbers of
age 0.0 and 1.0 fish represented in different
geographic areas and cruises were then extrapo-
lated from their proportions in each 10 mm length
interval.
Distances traveled and movement rates were
estimated from actual distances between sites of
release and entry into the ocean, and from
straight-line distances between ocean entry and
recapture locations for CWT or fluorescent
marked juvenile coho salmon that were recovered
in the ocean within 10 days of release. These dis-
tances and swimming speeds are minimal esti-
mates.
In addition to purse seining, fine-meshed
monofilament gill nets were used off the Oregon
coast (ca. lat. 45°00'N, long. 124°21'W) during 24
and 25 July 1985, from the training vessel Oshoro
Maru , to determine depth and direction of swim-
ming of juvenile salmonids. Surface and subsur-
face nets were used. The surface gill nets were
2,050 m long, and fished from depths of 0-6 m
with 11 mesh sizes ranging from 29 to 121 mm
(stretch). The subsurface nets were 500 m long
and consisted of four mesh sizes ranging from 29
to 42 mm; they were suspended below large (300-
400 mm) mesh to fish at depths of 5-12 m. Four
sets were made in an east-west direction with
soak-times of about 4-9 hours. As the gill nets
were hauled, the direction that each juvenile
salmonid was heading when caught, and its depth
in the net (upper, middle, or lower section) were
noted. Each juvenile salmonid was given a con-
secutive number and frozen for later identifica-
tion. Comparisons of catch rates in the surface
and subsurface nets were based on equal lengths
of the four mesh sizes of the subsurface net, stan-
dardized to 10 hours of fishing time.
Information on the location of landings of
marked juvenile coho by commercial and sports
fishermen was provided by the Pacific Marine
Fisheries Commission (PMFC), (PMFC 1980,
1981, 1984a, b, c, 1985a, b), from lists of non-
standard recoveries (Johnson PMFC unpubl.
data), and from state agencies. The actual num-
bers of tagged fish, and the total numbers of
tagged fish estimated from the proportions of the
catch sampled are reported.
To determine if juvenile coho salmon were sex-
ually precocious "jacks", we examined testes from
542 juvenile males caught in July 1981 and 1984,
in August 1981, and in September 1982, 1983,
and 1984. All developed and some undeveloped
testes (ribbonlike, with no thickening), as deter-
mined by visual inspection, were weighed (123)
and gonadal-somatic indices (GSI = testes wt./
body wt. X 100) were determined.
RESULTS AND DISCUSSION
Swimming Direction
Of the 106 juvenile coho salmon captured dur-
ing June in paired, half-round purse seine sets,
all but two were in the sets open to the south
(Table 2). This suggests that juvenile coho salmon
Table 2. — Catches of coho salmon In semicircular
purse seine sets open to the south (S) and north (N) off
Oregon, June 1979.
Location
Km
offshore
Age
S
.0
N
Age .1
S N
Clatsop Spit
Clatsop Spit
Clatsop Spit
Newport
12.6
18.5
18.5
9.4
57
37
7
3
0
0
0
2
2 6
15 37
6 8
0 9
98%
2%
28% 72%
176
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
were swimming to the north during this sampHng
period. Maturing fish over 300 mm FL (age .1
echo) showed the opposite trend.
Miller et al. (1983) made several hundred
paired purse seine sets open to the south and
north during three cruises off the northern
Oregon-southern Washington coasts in 1980.
During their May-June cruise, they caught 76%
of the juvenile coho salmon, 80% of the chinook
salmon, and almost all the steelhead trout in sets
open to the south, indicating northward move-
ment. We note a positive relation between the
proportion of juvenile salmon caught in south-
facing sets in their three cruises and strength of
upwelling during these cruises (mean daily
Bakun indices of 52, 39, and 19 m^ s"! 100 m'^
coastline at 45°N, 125°W in May, July, and Au-
gust, respectively (Mason and Bakun 1986)), sug-
gesting that surface currents to the south result-
ing from Ekman transport may be cues for
orientation of salmon smolts.
Hartt (1980) and Hartt and Dell (1986) found
that 83% of the combined species of juvenile
salmonids caught in 19 paired purse sets along
the coast from Cape Flattery, WA to Yakutat, AK
were caught in sets held open to the southeast and
only 17% in sets open to the northwest and north.
They concluded that juvenile salmonids tended to
migrate in a northwest direction along the coast
during July-September.
Of the 100 juvenile coho salmon (135-315 mm
FL) caught in the gill nets set in an east-west
direction off the Oregon coast in July 1985, 90
coho were caught as they approached the south-
ern face of the gill net (heading north) and 10 in
the northern face (heading south). Jaenicke et al.
(1984) reported that 63% of the juvenile coho
caught in a surface gill net fished off southeastern
Alaska in July moved north at night, but only 6%
moved to the north during the day.
Available data indicate that most juvenile coho
salmon caught off Oregon and southern Washing-
ton, as well as juveniles farther to the north, are
predominantely swimming in a northerly direc-
tion during summer months.
Depth Distribution
One-half of the juvenile coho salmon caught in
gill nets set off the Oregon Coast in 1985 were in
the upper 2 m of the surface gill net (Table 3).
Catches in the surface net exceeded those in the
subsurface net, except for the last set that fished
during daylight hours, indicating that juvenile
coho salmon were most common in the upper 4 m
of the water column.
Other information on the vertical distribution
of maturing coho and other species of salmon
caught in gill nets or with longlines in oceanic
waters also indicates that they usually swim near
the surface, between 0 and 20 m (Manzer 1964;
Godfrey 1965; Godfrey et al. 1975). Machidori
(1966), for example, fished gill nets from the sur-
face to 50 m and caught 79% of the coho salmon in
the upper 10 m of the gill net. Although catches in
gill nets at different depths may be biased by ver-
tical differences in avoidance reactions to the net
or swimming speeds (Hartt 1975), acoustical
methods have also shown that salmon are usually
distributed near the surface (Susuki and Sonoda
1972; Lord et al. 1976). We conclude that most
juvenile coho salmon in coastal waters and ma-
turing coho in oceanic waters reside at depths
above 20 m, the minimum depth that our purse
seine fished. We recognize, however, that matur-
ing coho and other species of salmon may feed in
deeper water. Some salmon (including coho
salmon) caught in surface gill nets in the oceanic
waters of the Gulf of Alaska contained prey in
their stomachs characteristic of mesopelagic
depths (200-1,000 m), suggesting that some indi-
viduals may feed well below the thermocline
(Pearcy et al. in press).
Table 3. — Catches of juvenile coho salmon in four gill net sets
in 50 m lengths of 29, 33, 37, and 42 mm mesh at different
depths and times 24-25 July 1985, each set adjusted to 10-h
fishing duration.
Depth
in meters
Surface net
Subsurface net
Times of set
0-2
2-4
4-6
5-7
7-9
9-12
0913-1702
2001-0104
0248-0701
0830-1737
2.5
25.0
42.8
1.1
3.8
11.5
2.4
0
0
0
2.4
0
1.3
0
9.5
6.6
1.3
0
16.6
7.7
0
1.9
0
5.5
Total catch
Percent of
total catch
51
50.3
16
12.5
1
1.7
11
12.3
15
18.0
6
5.2
North-South Trends in Catch per Set
and Sizes of Juvenile Coho
Variations in the average catches and sizes of
juvenile coho salmon in purse seine sets in differ-
ent regions of the Oregon-Washington coast dur-
ing the summer provide indirect evidence for
north-south coastal movements. Histograms
showing average catches per set for 10 mm size
177
FISHERY BULLETIN: VOL. 86, NO. 2
groups of juvenile coho, classified as age 0.0 or 1.0
from scale analysis, are shown in Figure 2, for
1981-84 in three regions: (A) Cape Flattery, WA
to Grays Harbor, WA (called Washington), (B)
Willapa Bay, WA to Nehalem Bay, OR (Columbia
River region), and (C) Cape Lookout, OR to Coos
Bay, OR (Oregon) (Fig. 1). In May of 1981, 1982,
and 1983, average catch per set of yearling (age
1.0) coho generally decreased from the southern
to the northern regions. Catches v^ere highest off
Oregon (Area C) or the Columbia River (Area B)
and lowest off Washington (Area A) in May of
1982 and 1983. This trend was reversed later in
the summer. In June of 1981, 1983, and 1984,
lowest catches were found in the Oregon region.
By August or September 1981-84, highest
catches consistently occurred off the Columbia
River or Washington and few yearling fish were
caught off Oregon. These shifts in abundance sug-
gest a northerly movement of age 1.0 smolts dur-
ing the summer. Highest catch rates occurred in
May and June of 1981 and 1982 when an average
of over 10 juvenile coho salmon were caught in
most sets.
Subyearling or age 0.0 coho salmon released
from private facilities at Yaquina and Coos Bays
provide more direct evidence on movements. Sub-
yearling coho salmon clearly demonstrated
northward dispersal. They were most common in
our catches of July 1981 and September 1982,
1983, and 1984 (Fig. 2). They were apparently
more numerous than age 1.0 coho salmon in the
Oregon region during June-August 1981 and
September 1982, and in the Oregon and Columbia
River regions in September 1983. The catches and
proportions of age 0.0 coho salmon increased off
Oregon during the summer because they were
released from hatcheries later in the summer
than yearling coho salmon. They were found in
the most northern region sampled late in the
1981
6
-
MAY
5
-
4
-
3
-
2
[-
/SET
J
-TV.
X 6
O
^ 5
" 4
3
-
2
-
"
1
Jl
wW.ljb.nv ,..y
0
20 30
JUNE
JULY
....Mljlito^
i^i,i^»i
AUGUST
B
C
11*11(^1
30
10
20
30
40
FORK LENGTH (cm)
Figure 2. — Catch per purse seine set of age 1.0 (open) and age 0.0 (shaded) juvenile coho by 10 mm length groups during different
months, 1981-84, for three regions of the Oregon-Washington coast: (A) Cape Flattery to Grays Harbor, WA, (B) Willapa Bay, WA
to Nehalem Beach, OR, (C) Cape Lookout to California.
178
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
summer of all years. Their abundance, as that of
age 1.0 coho salmon, also increased during the
summer off Washington where they intermingled
with age 1.0 coho salmon. Because the Oregon
region included the release locations of all age 0.0
coho salmon, our figures provide no information
on southward movements of these fish.
The mean lengths of both age 0.0 and 1.0 coho
salmon increased from the southern to the north-
ern areas during most months. Larger age 0.0 and
1.0 coho salmon were caught off Washington than
Oregon during the late summer, 1981-84 (Fig. 2),
providing corroborative evidence for northward
migration of coho salmon. Larger, and pre-
sumably older, fish were found farther to the
north than smaller fish.
Despite northward movements, many yearling
coho salmon did not migrate out of the sampling
area, but remained in coastal waters off Oregon
and Washington during the entire summer. Mean
coastwide catch per set of yearling coho salmon in
August 1981 and September 1982, 1983, and 1984
1982
I-
LJ
if)
6
5
4
3
21-
6
5
4
3
21-
51- J
4
MAY
TTfllTTTTTTtlll l^[
JUNE
t^t^ri^iM^rT^f rr
SEPTEMBER
A
¥.«
^
¥^
iWiT'i I
^
T^T*T^ IllllTTTfftfl
^ .UJJlmnifii
B
^JMpMk
T^P»
<L¥
»'ri^i f I
C
IM^
10
.p..Q.,,,n,f:T;i, ,f77.
20 30 10 20 30 10 20 30 40
FORK LENGTH (cm)
Figure 2. — Continued.
179
FISHERY BULLETIN: VOL. 86, NO. 2
1983
6
5
4
3
21-
MAY
■ I nfXb. .yi iGv
JUNE
■ ■f;w{ttlkbp.
SEPTEMBER
A
fffllfflVIIVI
^T^f^^WP^^^'^^^''''' T y T
CO
D ■
5 .
4-
3 ■
2-
1- r
,,dJ,U.i.b|
B
i*t f »'f^^*yT^ I
6
■
-
-
5
-
■
•
4
-
-
•
3
-
•
■
2
-
.
■
1
JOCII.
-•
^
1
•
10 20
30 10 20 30 10
FORK LENGTH (cm)
Figure 2.— Continued.
c
1^1 I I I t [ I I I I [I
20 30 40
was 61%, 42%, 81%, and 77% respectively, of that
in June of the same years. In September 1983, a
strong El Nino year (Pearcy and Schoener 1987),
almost all yearling juvenile coho salmon were
caught at the extreme northern transect off Cape
Flattery, but in August or September of other
years they were more evenly distributed off the
Columbia River and the Washington coast.
In July 1984, we sampled both north (15 sets off
the west coast of Vancouver Island) and south (5
sets off northern California) of Washington and
Oregon. Catches per set of yearling and subyear-
ling juvenile coho salmon were higher off the
Columbia River (5.1), the area of greatest smolt
production, and off Washington (3.5) than off
Vancouver Island (1.8), Oregon (1.7), and Califor-
nia (1.2). This shows that as late as July juvenile
coho salmon occurred in coastal waters of all re-
gions and were not concentrated off Vancouver
Island or California.
180
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
1984
H
UJ
C/)
\.
X
5
4
3|-
2
I
6
5
4
3
2
I
6
5
4
3
2
JUNE
rftflMru,
JULY
1^1 1 iiTMpiiiriii'JiiiriVpi^i I 'i -rf^'^^
SEPTEMBER
A
Ml^l I p ■ I I
^
pir^Tr7i.»,r>|
B
.oM^
rp.iP, ■ ,
c
1 — f^
30 10
-r-
10 20 30 10 20 30 10 20 30
FORK LENGTH (cm)
Figure 2.— Continued.
40
The Question of "Jacks"
Are the juvenile coho salmon off Oregon and
Washington in late summer relatively nonmigra-
tory because they are sexually precocious? GSI
were almost always <0.1% for those fish visually
classified as "undeveloped". GSI's from fish with
"developed" testes ranged from 0.29?^ to 1.0% in
July 1981 and 1984 in fish >250 mm FL; from
0.3% to 5.6% in August 1981, mostly in fish >280
mm FL; and from 2.4% to 6.6% (except for one
value at 0.6) in September 1982, 1983, and 1984
in fish >300 mm FL (Fig. 3). In August 1981, and
clearly in September 1982, 1983, and 1984, two
distinct groups of fish were evident: "jacks", with
developing testes (GSI >0.3% August or GSI
>2.0% September), and "nonjacks", which
showed no development (GSI <0.1%).
The total numbers of jacks and nonjacks in each
50 mm length group were estimated for the catch
during August 1981 and September 1982, 1983,
and 1984 from the ratio of jacks and nonjacks in
the sample (Table 4). Only 8.4%, 4.8%, 5.2%, and
2.8% of all juvenile fish (male and female) were
"jacks" in August 1981 and September 1982,
1983, and 1984, respectively. However jacks com-
181
FISHERY BULLETIN: VOL, 86, NO. 2
X
LU
Q
<
-z.
o
3
-
-
2
D 1981 n= 12
A 1984 n=9
-
1
n
A A A
,D ^ * °
A
1 1 1 1 1
1 1 1
6
5
4
3
2
I
0
9
8
7
6
5
4
3
2
I
0
-
-
D 1981 n=55
D
D
D
D
D
°D
D
-
D
D
° a
-
.iiniLtniiM imiiiiiriijiiinjii .Viiiiiim ii. i
i 1
■ 1 t 1 1
1
1 1
0 1982 n=9
• 1983 n=27
A 1984 n=ll
•
-
-
A
•
A
^ A O
o
o
-
_
O
.
-
^ •a
• 0°
o
-
A
-
-
, M . , r . m^m
• m^}A
^AaftMB . an . . .
, 1 , • ,
20
25
30
35
40
45
FORK LENGTH (cm)
Figure 3. — Gonadal-somatic index (testis wt/total body wt) x 100 of juvenile coho salmon vs.
length of fish for July, August, and September 1981-84. Only data for those testes actually
weighed are shown.
prised a higher percentage of fish larger than 300
mm FL in August 1981 and larger than 350 mm
FL in September 1982, 1983, and 1984. These
results indicate that most juvenile coho salmon
caught off Oregon and Washington were not sexu-
ally precocious. Thus, the relatively large catches
of juvenile coho salmon in late summer are ex-
plained by lack of strong migrational tendencies
of juvenile coho salmon in this region and not by
a high proportion of precocious "jacks" that re-
sided in this region as a prelude to re-entry of
streams for spawning.
182
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
Table 4. — The percent of coho salmon jacks and males, by length groups, in the total catch,
Augjst 1981 and September 1982-84.
Fork
Number
Number
Est.
Est. %
length
of fish
%
of
Total
total
jacks
Date
(mm)
examined
males
jacks'
catch
jacks
of total
Aug. 1981
<200
71
60.6
0
111
0
0
201-250
63
60.0
1
115
2
1.7
251-300
55
47.3
4
104
8
7.7
301-350
16
81.3
4
22
6
27.3
351-420
10
80.0
8
19
15
78.9
Total
215
59.5
17
371
31
8.4
Sept. 1982
<200
56
55.4
0
125
0
0
201-250
21
76.2
0
54
0
0
251-300
22
63.6
0
109
0
0
301-350
33
69.7
3
97
9
9.3
351-420
13
53.8
5
28
11
39.3
Total
145
62.8
8
413
20
4.8
Sept. 1983
<200
16
62.5
0
18
0
0
201-250
23
47.8
0
25
0
0
251-300
23
56.5
0
71
0
0
301-350
39
61.5
4
77
8
10.4
351-420
2
0
1
3
2
66.7
Total
103
58.3
5
194
10
5.2
Sept. 1984
<200
6
33.3
0
15
0
0
201-250
38
47.4
0
69
0
0
251-300
90
50.0
2
128
3
2.3
301-350
27
55.6
12
31
2
6.5
351-420
6
33.3
2
7
2
28.6
Total
167
49.1
16
250
7
2.8
'Jack Is defined as a male whose testes wt. /total body wt. x 100 > 0.3% in August and >2.0% in Septem-
ber.
Movements of Marked Fish
Direct evidence of movements of juvenile coho
salmon was obtained from capture of marked fish
containing coded wire tags or marked with fluo-
rescent pigment. The generalized pattern of
movements that emerges for 1981-85 is an initial
movement of most juvenile coho salmon to the
south soon after ocean entry in May and June and
then a reversal of movement with most fish mi-
grating to the north by August and September
(Figs. 4-8). These trends are discussed for fish
originating from the Columbia River, Oregon
coastal, Washington coastal, and private hatch-
eries.
Columbia River
Juvenile coho salmon originating from hatch-
eries on the Columbia River were usually recov-
ered south of the Columbia River in May. This
trend was especially obvious in May 1982 when
all 22 marked fish which were recovered moved
south, some as far as 175 km (Fig. 5). In May
1981, all but one of 14 marked Columbia River
fish were caught to the south, three as far as 180
and 204 km (Fig. 4). In May 1983, all four fish
were taken south of the mouth of the Columbia
River (Fig. 6).
During June and July of all years, marked Co-
lumbia River coho salmon were recovered in
nearly equal proportions both north and south of
the river mouth, except in June 1982 when 15 of
17 fish were found to the south (Figs. 4—8). By
September, all marked Columbia River coho
salmon were captured north of the river, includ-
ing fish captured off the Quinault River in Sep-
tember 1982 and off Cape Flattery in September
1984. Fish were also caught close to the mouth of
the Columbia River in July, August, and Septem-
ber, indicating that some marked juvenile coho
salmon did not undertake extensive migrations at
sea.
In two sets on the Wecoma Beach Transect on 1
June 1982 we caught 17 marked juvenile coho
salmon released between 30 April and 6 May
from six hatcheries on the Columbia River. Based
on downstream migration rates for these groups
to Jones Beach (Dawley et al. 1985) and assuming
similar rates from Jones Beach to the ocean, these
fish had probably been in the ocean for <10 days
before recapture. This indicates that some juve-
183
FISHERY BULLETIN: VOL. 86, NO. 2
AUG JULY JUNE
1981
Figure 4. — North-south movements of marked juvenile echo salmon captured
in purse seines, May-August 1981. The width of the lines are approximately
proportional to the number offish. Numbers at end of arrows indicate number
offish captured. Arrows without numbers and thin lines represent single fish.
Inshore-offshore movements are not shown. Dashed lines indicate latitudinal
extent of sampling.
nile coho salmon released from hatcheries at
about the same time tended to stay together dur-
ing their downstream migration in the Columbia
River and during early residency in the ocean.
Oregon Public Coastal Hatcheries
We captured marked fish originating ft-om pub-
lic Oregon coastal hatcheries both north and
south of the latitude of ocean entrance in May. A
total of five fish were found to the south, while 11
fish were found to the north in May (Figs. 4-8).
With the exception of one coho salmon from the
Umpqua River in June 1983 and two from the
Rogue River in July 1984 (Figs. 6, 7), the other 25
fish taken after May were captured north of
where they entered the ocean. Northerly move-
ments into Washington waters occurred by June
1983 and 1985 (Figs. 6, 8).
The southward movements of two juvenile coho
salmon released from the Rogue River (south of
Cape Blanco) and captured off northern Califor-
184
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
I I I I Ku I I I I 1 IV q
-CAPE FLATTERY
SEPT JUNEV MAY
1982
-rCAPE FLATTERY
Figure 5. — North-south movements of marked juvenile coho
salmon captured in purse seines, May, June, and September
1982. The width of the lines are approximately proportional to
the number of fish. Numbers at end of arrows indicate number
offish captured. Arrows without numbers and thin lines repre-
sent single fish. Inshore-offshore movements are not shown.
Dashed lines indicate latitudinal extent of sampling.
JUNE
1983
Figure 6. — North-south movements of marked juvenile coho
salmon captured in purse seines, May, June, and September
1983. The width of the lines are approximately proportional to
the number of fish. Numbers at end of eirrows indicate number
offish captured. Arrows without numbers and thin lines repre-
sent single fish. Inshore-offshore movements are not shown.
Dashed lines indicate latitudinal extent of sampling.
nia during July 1984 are notable (Fig. 7). They
were captured in our only cruise into California
waters and represent the only recoveries of
marked juvenile coho salmon originating from
hatcheries south of Cape Blanco in all six years of
sampling. Although ocean sampling was limited
south of Coos Bay, if juvenile coho salmon from
southern Oregon and northern California hatch-
eries had migrated north of Coos Bay, we
would expect them to be represented in our
catches. The fact that they were not caught in this
northern region, but two were caught after swim-
ming to the south, suggests that juvenile coho
salmon originating in streams south of Cape
Blanco may migrate south, possibly occupying
the region of intense coastal upwelling off north-
ern California during their first summer in the
ocean. The catch of over 70% of the adult coho
185
FISHERY BULLETIN: VOL. 86, NO. 2
ll
SEPT
UMPQUA
RIVER
JUNE 10-25
.coos BAY
MAY 29 - JUNE 5
1985
±
CAPE BLANCO
I I I I I
— 43=
JULY
1984
Figure 8. — North-south movements of marked juvenile coho
salmon captured in purse seines, 29 May-5 June and 10-25
June. The width of the lines are approximately proportional to
the number of fish. Numbers at end of arrows indicate number
offish captured. Arrows without numbers and thin lines repre-
sent single fish. Inshore-offshore movements are not shown.
Dashed lines indicate latitudinal extent of sampling.
Figure 7. — North-south movements of marked juvenile
coho salmon captured in purse seines, June-September
1984. The width of the lines are approximately propor-
tional to the number of fish. Numbers at end of arrows
indicate number of fish captured. Arrows without num-
bers and thin lines represent single fish. Inshore-
offshore movements are not shown. Dashed lines indi-
cate latitudinal extent of sampling.
186
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
salmon from the Rogue River in the troll fishery
off California (R. Garrison'^) is further evidence
for a southern distribution of this stock. Marked
juvenile coho salmon from California hatcheries
were reported from the sports fishery off southern
Oregon, however, as will be shown later.
Washington Coastal Hatcheries
Juvenile coho salmon from Washington public
hatcheries demonstrated southerly movements,
sometimes into Oregon waters, during May 1981
and 1982 (Figs. 4, 5). During June 1982, 1984,
and 1985, Washington coastal fish were found
both north and south of ocean entry. Except for
one fish that moved north in September 1984, no
Washington coastal fish were taken in August or
September of other years, suggesting that most
Washington fish may have migrated out of our
sampling area by late summer.
Oregon Private Hatcheries
All marked juvenile coho salmon originating
from Yaquina and Coos Bays that we captured at
sea were from private hatcheries. Those from
Yaquina Bay were mainly age 0.0 smolts from
OAF, those from Coos Bay were either age 1.0
smolts from Anadromous, Inc. or age 0.0 smolts
from OAF. Forty-one recoveries of marked OAF
fish released from Yaquina Bay were caught to
the north while only 4 were to the south of
Yaquina Bay. In general, more juvenile coho
salmon from Yaquina Bay were captured in late
than early summer, and distances traveled to the
north were largest (up to 413 km) for fish caught
in later summer (Figs. 4-8). All recoveries of
marked Anadromous, Inc. and OAF fish released
into Coos Bay were to the north in all years.
Large northerly movements were demonstrated
by some of these fish (Figs. 4, 6, 8). Since our
sampling in the ocean usually did not extend
south of Coos Bay, recoveries of these fish are
biased to the north; however, strong northward
movements of these stocks were indicated.
Rates of Movement
Recoveries of marked juvenile coho salmon in
the ocean provided information on the minimum
rates of movement from hatchery release to cap-
3R. Garrison, Oregon Department of Fish and Wildlife, Cor-
vallis, OR 97330, pers. commun. December 1983.
ture in the ocean. Some fish moved rapidly
through estuaries into the ocean. We captured
some tag-groups in the ocean only a few days after
the median date of capture at Jones Beach (75 km
from the ocean) as reported by Dawley et al.
(1985): 5 fish after 2 days in 1981, 6 fish after
3-11 days in 1981, 8 fish after 1-14 days in 1982,
and 5 fish after 3-8 days in 1983. Dawley et al.
(1986) found average rates of movement of 14-23
km d"^ for marked groups of coho smolts from
areas of release on the Columbia River to river
km 75, and rates of movement that were 40%
faster from river km 75 to the lower Columbia
River estuary and to the ocean plume. These re-
coveries and those reported by Miller et al. (1983)
for yearling chinook salmon and steelhead trout
indicate rapid movements of juvenile salmonids
of over 20 km d^^ through the Columbia River
estuary.
Some juvenile coho salmon released from
Yaquina Bay and Coos Bay also demonstrated
rapid movements into and in the ocean, e.g., 17
Anadromous, Inc. fish were captured 11 km north
of Coos Bay only two days after release in June
1983 (Table 5). Myers (1980) described an expo-
nential decrease in the catches of juvenile coho
salmon released from the OAF facility into
Yaquina Bay; about one-half the fish from
marked groups remaining in the bay after 1.7-9.0
days. Juvenile coho salmon apparently emigrate
rapidly from estuaries into the ocean.
Some of the marked fish recovered within 10
days of release demonstrated rapid movements
down-rivers or in the ocean. Twenty-four fish
traversed 10 km d~^ or more largely in the ocean,
in both north and south directions (Table 5). Four
fish released in bays or in the ocean moved over
18.8 km d~^ Two of these swam to the north,
presumably against coastal currents. These
speeds are equivalent to 1.7 body lengths (BL) per
second or more and suggest that some fish must
be traveling nearly straight courses during
24-h days, since 1-3 BL s"^ are thought to be
optimal cruising speeds for small (<20 cm)
pelagic fishes (Weihs 1973; Ware 1978). These
maximum rates of movement for purse seine
caught juvenile coho salmon are similar to
those estimated by Hartt (1980) and Hartt and
Dell (1986) for tagged sockeye salmon during
their first summer in the ocean: 14-27 km d"^
for 11 Fraser River fish and 6-14 km d"^ for
10 Skeena River fish that were between about
8 and 23 cm in length during the migration
period.
187
FISHERY BULLETIN: VOL. 86. NO. 2
Table 5. — Release Information and mean travel speeds in kilometers per day and body length (BL)
per second for CWT and fluorescent-pigment marked age .0 coho salmon recovered in the ocean
within 10 days of release. CR ^ Columbia River: OAF = Oregon Aqua Foods, Inc., (OAF Yaquina
is 3.7 km from ocean; OAF Coos Is 14 km from the ocean), Anad. - Anadromous Inc. (7.4 km from
the ocean).
Median
Direction
Release
Date
days to
Mean
Mean
Mean
of
Year
No.
location
released
recovery
FL
km/d
BUS
movement
1981
1
Big Cr. (CR)
8 June
3
153
22.2
1.7
S
1
Tanner Cr. (CR)
6 July
6
138
41.0
3.4
N
1
OAF Yaquina
11 May
7
140
4.0
0.3
S
1
OAF Yaquina
12 June
2
124
14.5
1.4
N
2
Anad. Coos
8 June
9
179
2.3
0.2
N
2
OAF Yaquina
10-15 June
2
124
3.7
0.3
N
1
OAF Yaquina
10-15 June
2
123
10.2
1.0
N
1
OAF Yaquina
10-15 June
2
126
18.8
1.7
N
11
OAF Coos
5-9 June
10
122
1.9
0.2
N
1
OAF Yaquina
10-15 June
5
124
24.0
2.2
S
1
OAF Coos
5-9 June
7
136
20.1
1.7
N
1983
17
Anad. Coos
26 June
2
156
11.2
0.8
N
1984
1
OAF-Offshore
7 June
7
143
11.3
0.9
N
1985
6
Tongue Pt. (CR)
24 IVIay
6
151
9.5
0.7
S
1
Offshore-22 km
30 l^ay
0.9
134
22.0
1.9
S
1
Naselle River
20 l^ay
9
143
7.8
0.6
S
1
Cowlitz River
31 IVIay-6 June
2.5
144
88.0
7.1
N
Effects of Ocean Currents
The tendency for juvenile coho salmon to move
to the south early in the summer and to the north
later in the summer (Figs. 4—8) may be related to
advection of water and the size, orientation, and
swimming speeds of fish. Generally, surface cur-
rents are to the south off Oregon and Washington
in the early summer owing to prevailing north-
westerly winds (Hickey 1979; Huyer 1983).
Southward flow averaging 17-34 km d"^ (Huyer
et al. 1975, 1979) has been measured near the
surface. May and June are periods of peak outflow
of the Columbia River, so fish entering the ocean
at this time, especially in the Columbia River
plume, may be displaced to the south by advection
of surface waters. Southward flow is at a maxi-
mum in the coastal jet which is strongest (—22 km
d~^) during the spring about 15-20 km from
shore (Kundu and Allen 1976; Huyer et al. 1979).
Since currents can be equivalent to 1.7 BL s"^ for
a 15 cm smolt, advection alone could explain the
southward movement of most marked Columbia
River fish during May and June but not the rapid
northward movement of fish during this period
(see Figures 4-7).
Coastal Oregon fish were often found to the
north in May and June, but these fish were usu-
ally substantially larger and generally released
much earlier in the spring than Columbia River
fish, and were presumably better able to swim
against the current. Later in the summer when
Columbia River hatchery fish had grown larger,
movement was also predominately northward. In
August and September southward velocities of
surface coastal currents are diminished and the
mean may be near zero (Huyer et al. 1975).
Northward movements during the summer off
Oregon and Washington generally cannot be ex-
plained by passive drift and in most years must
entail active, oriented swimming.
The northern El Nino of 1982-83, which had
severe effects on the growth and survival of adult
and jack coho salmon (Pearcy et al. 1985; Johnson
1984; Pearcy and Schooner 1987; Fisher and
Pearcy in press), also appeared to affect the distri-
bution of juvenile coho salmon. During Septem-
ber of 1983 nearly all the seine-captured juvenile
coho were taken along our northernmost transect,
off Cape Flattery, WA (Fig. 2). In other years
juvenile coho salmon during late summer were
common and more equally distributed from the
Columbia River northward. In the summer of
1983 juvenile coho salmon may have moved far-
ther north, or more likely those to the south may
have experienced higher mortality, as a result of
northerly currents (Huyer and Smith 1985),
warm temperatures and low productivity (Pearcy
et al. 1985; Chung 1985) that prevailed off Ore-
gon.
188
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
Recoveries of CWT Juvenile Fish
in Ocean Fisheries
Data on ocean location of landings of juvenile
CWT coho salmon reported in sports and commer-
cial fisheries in the ocean along the west coast of
North America, 1977-83, provide valuable infor-
mation on ocean migrations of marked fish, al-
though these data are biased by differences in
legal minimum sizes, time and duration of open
season, and effort in the different regions. The
summary of all years shows that, except for Cali-
fornia, most of the recoveries of juvenile coho
salmon during their first summer in the ocean
were in the general region of their ocean entry
location (Table 6). Both the actual number offish
reported and the estimated total numbers (in
parentheses) support our earlier conclusion that
many juvenile coho salmon off Oregon and Wash-
ington are not highly migratory. All (20) of the
actual recoveries of marked juvenile coho salmon
that were released in southeastern Alaskan
waters were from southeastern Alaska. Ninety-
seven percent of the recoveries of marked fish
released in British Columbia waters were landed
in British Columbia; only two were landed in
Alaska. Most (86%) marked juvenile fish from
Puget Sound hatcheries were caught in the
Sound, and more were recovered in British Co-
lumbia fisheries (13%) than in coastal Washing-
ton fisheries (<1%), probably due to the smaller
size limits for coho in British Columbia as well as
migratory patterns. Half of the actual numbers of
recoveries of juvenile coho salmon liberated into
Washington coastal waters were landed in
Washington coastal ports, followed by British Co-
lumbia (29%) and the Columbia River region
(17%). Only one was landed in Alaska and two in
Oregon ports (Garibaldi and south). Juvenile coho
salmon originating from Columbia River hatch-
eries had a broader distribution of recoveries in
other regions. Only 40% of Columbia River fish
were caught in this region, 41% were taken in
northern regions, including two (1%) in Alaska.
The remaining 19% were captured off Oregon.
The majority (73%) of Oregon coastal fish were
recovered off Oregon, followed by the Columbia
River region, Washington coast, and British Co-
lumbia. None was reported from Alaska, but 10
(2%) were from California ports. All marked Cali-
fornia fish were recovered from the Columbia
River region and farther south. Most (87%) were
landed in Oregon. The few recoveries of Califor-
nia fish off California is undoubtedly influenced
by the larger size limits in this than other fish-
eries.
Table 6. — Recoveries of coded wire tagged juvenile
coho in the ocean fisheries 1 977-83. Estimated
total numbers are
in parentheses
Landing area
»«..
5-
&
^
i
O^
i
#«"^
Release area
CO'
/I"
/
S.E. Alaska
20
(39)
0
0
0
0 0 0
British
2
1,086
24
2
0 0 0
Columbia
(2)
(1,735)
(90)
(8)
Puget Sound
0
201
1,352
9
1 2 0
(729)
(5,262)
(40)
(2) (9)
Washington
1
71
7
125
42 2 0
Coast
(5)
(316)
(42)
(451)
(151) (77)
Columbia
2
24
4
39
67 31 0
River
(133)
(14)
(162)
(310) (164)
Oregon Coast
0
18
3
21
62 308 10
(107)
(13)
(83)
(213) (1,137) (45)
California
0
0
0
0
1 138 19
(4) (552) (200)
'Sports catches are not expanded. The estimated total number
CWTs recovered in the sports fisheries. Preliminary data.
expanded commercial catch + actual number
189
FISHERY BULLETIN: VOL. 86, NO. 2
In general, the legal size limits increased from
north to south, which could result in more recov-
eries of juvenile coho salmon in northern than
southern regions. Thus, these data do not provide
evidence that a large proportion of juvenile coho
salmon from British Columbia and waters to the
south made northward migrations into Alaskan
waters before or during the commercial and
sports salmon seasons. Movements of Washing-
ton, Columbia River, and Oregon fish into British
Columbian waters were common however.
Hunter (1985) expanded the catches of CWT
juvenile coho salmon caught along the west coast
of North America during 1978-80 to the total
landed plus estimated "drop-off' mortality (fish
that were hooked and died without being landed)
of both tagged and untagged hatchery groups.
Calculations of the percentage returns from dif-
ferent release and recapture areas are similar to
ours (Table 6). The highest percentage of returns
were from the areas of release for all areas except
for California, and a higher proportion of the
catches of Washington coastal, Puget Sound, Co-
lumbia River, and Oregon coastal stocks were re-
ported north than south of the area of release.
Are Juvenile Coho Highly Migratory?
Based on our observations on movements of
marked fish, north-south and seasonal trends in
abundance and size, and directional purse seine
sets during the summer, we conclude that many
juvenile coho salmon from Oregon and Washing-
ton coastal streams and the Columbia River are
transported by currents to the south in May and
June but then migrate north later in the summer.
The mean catches per set of yearling coho salmon
in August and September are a large fraction of
those in June, indicating that in the years studied
many juvenile coho salmon in coastal waters of
Oregon and Washington were not highly migra-
tory. Moreover, more marked juvenile hatchery
coho salmon were caught in ocean fisheries in the
region of release than in distant waters. Recover-
ies of juvenile coho salmon released from hatch-
eries south of Cape Flattery were rare in northern
waters off Alaska and relatively few were recov-
ered in British Columbia (Table 6). In addition,
the positive correlation between upwelling off Or-
egon and survival of hatchery coho salmon from
the Columbia River, Oregon, and California
(Nickelson 1986) also argues for a close coupling
of OPI coho salmon with a local, not a distant,
environmental event during the time that year-
class survival is determined. All of these trends
suggest that most juvenile coho salmon from this
area are not highly migratory and that many usu-
ally remain in coastal waters near their sites of
ocean entry during their first summer in the
ocean, and perhaps during their entire ocean life.
In years of unfavorable ocean conditions, how-
ever, movements may be more extensive or mor-
tality may be higher, as suggested by the very low
catches of juvenile coho salmon in purse seine sets
south of Cape Flattery during September 1983,
the year of the recent strong El Nino.
Although Pacific salmon are renown for their
long foraging migrations in the subarctic Pacific,
coho salmon demonstrate both nonmigratory and
highly migratory behavior. Milne (1950) found
immature coho salmon almost year-round in
Georgia Strait and concluded that two types of
coho salmon exist in British Columbia waters:
"ocean" and "inshore" types, the "ocean" type
spending most of its ocean life in coastal and off-
shore waters and the "inshore" type in inside
waters such as Georgia Strait. Healey (1978)
caught "inshore" juvenile coho salmon in purse
seines in Georgia Strait during summer, fall, and
winter months. Similarily, large numbers of coho
salmon originating from streams of Puget Sound
remain in the Sound throughout their marine life
(Haw et al. 1967). Young coho salmon have also
been found in the winter and spring, many
months after seawater entry in Yaquina Bay
(Myers 1980) and other Oregon estuaries
(J. Nicholas^). Hartt and Dell (1986), in their im-
pressive study of juvenile salmonids of the north-
eastern Pacific during 1956-70, recognized these
two migratory patterns of coho salmon. They
found juvenile coho salmon in waters off Vancou-
ver Island and in the Strait of Juan de Fuca
throughout the summer and fall, and concluded
that some coho salmon spend their entire marine
life in "inside" waters and make only limited
ocean migrations.
What Proportion of Juvenile Coho
from Oregon and
Washington Migrate North?
The tagging experiments reported by Hartt and
Dell (1986) and Godfrey (1965) provide convinc-
ing evidence for long-distance migrations of coho
■IJ. Nicholas, Oregon Department of Fish and Wildlife, Cor-
vallis, OR, 97331, pers. commun. May 1986.
190
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
salmon during their first summer in the ocean.
Based on recoveries of maturing coho salmon that
were tagged a year earlier at sea during April-
October 1956-70, Hartt and Dell (1986) con-
cluded that juvenile coho salmon from the Colum-
bia River, Oregon, and California may form a
large proportion of the coho stocks that migrate
north along the coast each summer. Of the 70
recoveries of tagged fish that were released be-
tween Kodiak Island and 56°N, 37% were recov-
ered the following year in the area of the Colum-
bia River and Oregon-California; of the 59
recoveries offish released between 56°N and Cape
Flattery, 47% were recovered in these southern
regions. In all, 25% of the recoveries were from
Oregon-California, 16% from the Columbia
River, 14% from Washington, 33% from British
Columbia, and 12% from Alaska.
Loeffel and Forster (1970) concluded that pat-
terns of radioactive ^^Zn in juvenile coho salmon
collected in the northeastern Pacific supported
the concept of a northerly migration from Oregon
and Washington into the Gulf of Alaska during
the summer. They found that juvenile coho
salmon captured off the west coast of Vancouver
Island in June and July 1967 contained ^^Zn, pre-
sumably originating from neutron activation of
Columbia River water used to cool the nuclear
reactors at Hanford, WA. ^^Zn levels decreased
in fish caught farther to the north (54°42'N-
58°24'N) in July-September of 1967. The authors
thought the low concentrations in northern sam-
ples represented background levels and that fish
with relatively high levels of ^^Zn had associated
with the Columbia River plume and subsequently
migrated north from the Oregon- Washington re-
gion. They found low ^^Zn levels in 1968, how-
ever, and no pronounced latitudinal gradients.
Furthermore, they reported none of the many fin-
marked juvenile coho salmon released from Ore-
gon and Washington hatcheries in 1967 and 1968
north of Juan de Fuca Strait. Hence their evi-
dence for northward movements of Columbia
River or Oregon-Washington coho salmon was
equivocal.
During June and July 1984, research was con-
ducted with the NMFS Auke Bay Laboratory in
waters from northern California to southwest
Alaska from the FV Pacific Warwind and Bering
Sea , both making round hauls with the same size
of purse seine, to sample juvenile coho in waters
north of Oregon and Washington: 37 sets were
made in coastal waters of British Columbia, and
39 were made in coastal waters and 29 in inland
waters (bays, inlets, and ^ords) of southeastern
Alaska. Of the 371 juvenile coho salmon captured
in these regions, 77% were caught in inland
waters of southeastern Alaska. The seven CWT
juvenile coho salmon captured were all from
Alaska inland waters and all originated from
Alaska hatcheries (Auke Bay Laboratory 1984a).
A later cruise in southeast Alaska by the Auke
Bay Laboratory in August 1984 caught eight
CWT coho salmon, also all from inland waters
and from Alaskan hatcheries (Auke Bay Labora-
tory 1984b).
Of the 14 CWT juvenile coho salmon collected
in other purse seines, gill nets, and special troll
gear in waters of southeastern Alaska during
1982, 1983, and 1985, 12 originated from Alaska
hatcheries and 2 originated from Washington
hatcheries (Auke Bay Laboratory 1983; Jaenicke
et al. 1984; Orsi et al. 1987). Table 6 shows that
only 5 of 25 CWT juvenile coho salmon caught in
Alaskan waters during 1977-83 were from hatch-
eries south of Alaska, indicating that most juve-
nile coho salmon caught in southeastern Alaska
during the summer originated from Alaska and
not from southern regions.
Hartt (1980) and Hartt and Dell (1986) recog-
nized that their data did not indicate the propor-
tion of southern stocks that made northerly mi-
grations, but they concluded that a large
proportion is probable, since juvenile coho salmon
were consistently caught in most seine sets
throughout the area sampled. They estimated
that the average density of juvenile salmonids in
coastal waters between 56°N and 60°N off south-
eastern Alaska during August and September
1964-68 was 1,500 km"^. The average density of
juvenile coho salmon in this area during these
two months was only 82 km~^ (Hartt and Dell
1986, app. A). During August and September
1981-84, the average density of juvenile coho
salmon in our round hauls between Cape Flat-
tery, WA and Cape Arago, OR to 37 km offshore
was 350 km"^, several times the estimates of
Hartt and Dell for the same months of the year.
This suggests that juvenile coho salmon may be
found in higher densities off Oregon and Wash-
ington than southeastern Alaska during late
summer, assuming that distributions and abun-
dances in the late 1960s and early 1980s were
similar. This trend for higher abundances of juve-
nile coho salmon off Oregon and Washington than
in coastal waters farther north was also found
during July 1984 (Table 7), although average
catches off Washington and Oregon were not as
191
FISHERY BULLETIN: VOL. 86, NO. 2
Table 7. — Average catches of juvenile coho salmon in purse
seines sets in coastal waters along the west coast of North America
during July 198412
Area
No. per set
No. per km2
Sitka-Juneau
Ketchikan-Sitka
Cape Scott-Dixon Entrance
Vancouver Is.
Washington
Oregon No. California
1.29
0.58
1.90
1.80
3.76
2.59
68
3
100
95
198
136
'Cruise Report. Drum Seiner FV Bering Sea, Coastwide NWAFC/OSU
Cooperative Study, Ecology of Juvenile Salmon in Coastal and Inside Waters
of Soutfieast Alaska, 28 June-26 July 1984, NWAFC Auke Bay Laboratory,
National Marine Fisfierles Service, NCAA, P.O. Box 115, Auke Bay, AK
99821 .
2Fisher and Pearcy (1984).
large as in some earlier years owing to low sur-
vival (Fig. 2; Pearcy 1984; Fisher and Pearcy in
press).
Comparisons of the estimates of total juvenile
yearling coho salmon abundances off Oregon and
Washington with the production of coho smolts in
the Oregon Production Area (Columbia River to
California) also suggests that many juvenile coho
resided off Oregon and Washington during the
summer. By expanding our catches per m^ to the
region sampled, we estimated that the numbers of
juvenile yearling coho salmon in areas surveyed
by our purse seine sampling during August or
September 1981-84 were 6.3%, 6.5% 5.1%, and
5.2%, respectively, of the numbers of hatchery
and wild smolts released in the Columbia River
and in Oregon (T. Lichatowich^). The areas in-
cluded in these estimates were roughly 83%, 62%-,
51%, and 68% of the total area from Cape Flattery
to Cape Arago out to 37 km offshore. Recognizing
that the entire area was not sampled, that year-
class strength of coho salmon in this region is
probably established soon after ocean entrance
(Fisher and Pearcy in press), and that early
marine mortality may be inversely related to size
(Parker 1968; Ricker 1976) so that much of the
ocean mortality has occurred by late summer,
these percentages probably represent a substan-
tial portion of the surviving OPI coho smolts. In
fact, they are several times higher than the
smolt-to-adult survival of 1.3-2.8% for OPI public
hatchery coho salmon (excluding Rogue River
and California hatcheries) during 1981-84 (R.
Kaiser^).
5T. Lichatowich, Oregon Department of Fish and Wildlife,
P.O. Box 59, Portland, OR 97207, pers. commun. September
1987.
6R. Kaiser, Oregon Department of Fish and Wildlife, Hatfield
Marine Science Center, Newport, OR 97265, pers. commun.
September 1987.
We conclude, therefore, that a major fraction of
the juvenile coho salmon from Oregon and Wash-
ington hatcheries did not undertake distant mi-
grations into the Gulf of Alaska in recent years.
This is not necessarily in conflict with Hartt and
Dell's (1986) data, since they established the pres-
ence of Oregon and Washington coho salmon in
northern waters but not the proportion of total
production that undertakes this migration. On
the other hand, neither the stocks of coho nor
oceanographic conditions have remained con-
stant over the period from 1956 to 1985 when
these two studies were conducted. Wild coho
smolts exceeded hatchery smolts in the Oregon
Production Area before 1961 (Nickelson 1986) but
comprised <12% of the smolts in 1980-85 (R.
Kaiser fn. 6). Perhaps wild smolts from the OPI
area had different migratory patterns than hatch-
ery smolts do today and migrated rapidly into
northern waters soon after ocean entrance. This
may explain Nickelson's (1986) finding that sur-
vival of hatcherv, but not wild coho smolts, was
significantly correlated with coastal upwelling off
Oregon.
Ocean conditions have also changed over this
period. The late 1960's were accompanied by
strong upwelling along the coast compared to
weak upwelling in the early 1980's (Nickelson
1986; Mason and Bakun 1986). McLain (1984),
Norton et al. (1985), and Royer (1985) illustrated
that sea surface temperatures and sea levels in-
creased in the northeastern Pacific between 1976
and 1984. These factors and associated changes in
ocean circulation could explain differences in mi-
gratory behavior of coho salmon between 1960's
and 1980's. If currents provide orientational cues
to migration, cues facilitating northerly move-
ments may be reduced during years of weak up-
welling and weak Ekman transport from the
north. Ocean conditions may have modified mi-
gratory patterns, as they possibly have for the
migration of Fraser River sockeye salmon around
Vancouver Island (Groot et al. 1984; Hamilton
1985).
ACKNOWLEDGMENTS
We thank D. Larden and his crew of the FV
Pacific Warwind and Bering Sea for their cooper-
ation and competence during purse seining opera-
tions at sea; A. Chung, R. Brodeur, J. Shenker,
W. Wakefield, D. Gushee, C. Banner, J. Long, K.
Krefft, and C. Wilson for their help on cruises;
J. Norton and the Oregon Department of Fish and
192
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
Wildlife Clackamas Laboratory for decoding
coded-wire tags; K. Johnson for providing data on
marked fish; W. McNeil and R. Severson for as-
sisting in fluorescent spray-marking Oregon
Aqua-Food's Inc. smolts; Northwest and Alaska
Fisheries Center for the loan of a purse seine;
Northwest and Alaska Fisheries Center Auke
Bay Laboratory for the loan of electronarcosis
equipment; and the Faculty of Fisheries, Hok-
kaido University, and the TV Oshoro-Maru for
conducting gill net research. A. Hartt,
H. Jaenicke, R. Brodeur, R. Gowan, D. Hankin,
and L. Botsford all provided helpful comments on
the manuscript.
This research was supported by Oregon State
University Sea Grant College Program (Grant
No. NA 81AA-D-00086, Project R/OPF-17), the
Northwest and Alaska Fisheries Center (Con-
tract 81-ABC-00192, 83-ABC-00102, 84-ABC-
0009, and NA-85-ABH0002J), the Oregon De-
partment of Fish and Wildlife, and Oregon
Aqua-Foods, Inc.
LITERATURE CITED
Auke Bay Laboratory.
1983. Cruise report. NOAA fishery research vessel John
N. Cobb, JC-83-03. Juvenile salmon purse seining proj-
ect in coastal and inside waters of southeastern
Alaska. NWAFC Auke Bay Lab., Natl. Mar. Fish.
Serv., NOAA, 30 p.
1984a. Cruise report drum seines. F/V Bering Sea.
Coastwide NWAFC/OSU Cooperative study. Ecology of
juvenile salmon in coastal and inside waters of south-
eastern Alaska, 28 June-26 July 1984. NWAFC Auke
Bay Lab., Natl. Mar. Fish. Serv., NOAA, 9 p.
1984b. Cruise report, NOAA RV John N. Cobb Cruise
84-03. Juvenile salmon purse seining project in coastal
and inside waters of southeastern Alaska, July 31-Au-
gust31, 1984. NWAFC Auke Bay Lab., Natl. Mar. Fish.
Serv., NOAA.
Chung, A
1985. Relationships between oceanographic factors and
the distribution of juvenile coho salmon (Oncorhynchus
kisutch) off Oregon and Washington, 1982-83. M.S.
Thesis, Oregon State Univ., Corvallis, 116 p.
Clutter. R I., and L E. Whitesel.
1956. Collection and interpretation of sockeye salmon
scales. Int. Pac. Salmon Comm. Bull. 9:1-159.
Dawley. E. M , R. D Ledgerwood, T H Blahm, C. W. Sims, J. T.
DuRKiN, R A Kirn, A E Rankis, G E Monan, and F J.
OSSIANDER
1986. Migrational characteristics, biological observa-
tions, and relative survival of juvenile salmonids enter-
ing the Columbia River estuary, 1966-1983.
Final Report, Bonneville Power Administration Project
No. 81-102. Northwest and Alaska Fish. Cent., Natl.
Mar. Fish. Serv., NOAA, Seattle, 256 p.
Dawley. E M . R. D. Ledgerwood, and A. Jensen
1985. Beach and purse seine sampling of juvenile
salmonids in the Columbia River estuary and ocean
plume, 1977-1983. Vol. 2: Data on marked fish recover-
ies. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/
NWC-75, 397 p.
Dawley, E M., C. W Sims. R. D. Ledgerwood, D. R. Miller, and
J G Williams
1981. A study to define the migrational characteristics of
chinook and coho salmon in the Columbia River estuary
and associated marine waters. Final Rep. Northwest
and Alaska Fish. Cent., Natl. Mar. Fish. Serv., NOAA,
68 p.
Fisher, J P , and W G Pearcy
1984. Studies of juvenile salmonids off northern Califor-
nia, Oregon, Washington and Vancouver Is-
land. 1984. Oreg. State Univ. Sea Grant Coll. Pro-
gram (ORESU-T-OGl), 24 p.
1985. Studies of juvenile salmonids off the Oregon and
Washington coast, 1985. Oreg. State Univ. Sea Grant
Coll. Program (ORESU-T-004), 31 p.
In press. Growth of juvenile coho salmon [Oncorhynchus
kisutch ) off Oregon and Washington, USA in years of
differing coastal upwelling. Can. J. Fish. Aquat. Sci.
Fisher, J P , W G Pearcy, and A. W Chung
1982. Studies of juvenile salmonids off the Oregon and
Washington coast, 1982. Oreg. State Univ. Sea Grant
Coll. Program (ORESU-T-83-003), 41 p.
1983. Studies of juvenile salmonids off the Oregon and
Washington coast, 1983. Oreg. State Univ. Sea Grant
Coll. Program (ORESU-T-001), 29 p.
FOESTER, R. E.
1968. The sockeye salmon Oncorhynchus nerka. Fish
Res. Board Can. Bull. 162, 422 p.
French, R. R., R G Bakkala. and D. F. Sutherland.
1975. Ocean distribution of stocks of Pacific salmon,
Oncorhynchus spp., and steelhead trout, Salmo gaird-
nerii, as shown by tagging experiments. U.S. Dep.
Commer., NOAA Tech. Rep. NMFS SSRF 689, 89 p.
Godfrey, H
1965. Salmon of the North Pacific Ocean. Part IX Coho,
chinook and masu salmon in offshore waters. 1. Coho
salmon in offshore waters. Int. North Pac. Fish. Comm.
Bull. No. 16, 39 p.
Godfrey, H., K A Henry, and S. Machidori.
1975. Distribution and abundance of coho salmon in off-
shore waters of the North Pacific Ocean. Int. North
Pacific Fish. Comm. Bull. No. 31, 80 p.
Groot, C , L Margolis, and R. Bailey.
1 984 . Does the seaward migration of Eraser River sockeye
salmon (Oncorhynchus nerka) smolts determine the
route of return of adults? In J. D. McCleave, G. P.
Arnold, J. J. Dodson, and W. H. Neill (editors). Mecha-
nisms of migration in fishes, p. 283-292. Plenum Press,
N.Y.
GUNSOLUS, R. T.
1978. The status of Oregon coho and recommendations for
managing the production, harvest, and escapement of
wild and hatchery reared stocks. Oreg. Dep. Fish.
Wildl. Fish Div. Proc. Rep., 59 p.
Hamilton, K.
1985. A study of the variability of the return migration
route of Eraser River sockeye salmon (Oncorhynchus
nerka). Can. J. Zool. 63:1930-1943.
Hartt, A C
1975. Problems in sampling Pacific salmon at sea. Int.
North Pac. Fish. Comm. Bull. 32:165-231.
193
FISHERY BULLETIN: VOL. 86, NO. 2
1980. Juvenile salmonids in the oceanic ecosystem — the
critical first summer. In W. J. McNeil and D. C.
Himsworth (editors), Salmonid ecosystems of the North
Pacific, p. 25-57. Oreg. State Univ. Press.
Hartt, a. C , AND M. B Dell.
1986. Early oceanic migrations and growth of juvenile
Pacific salmon and steelhead trout. Int. North Pac.
Fish. Comm. Bull. 46:1-105.
Haw. F . H. O. Wendler, and G Deschamps.
1967. Development of Washington State salmon sports
fishery through 1964. Wash. Dep. Fish. Res. Bull. 7,
191 p.
Healey, M. C.
1978. The distribution, abundance, and feeding habits of
juvenile Pacific salmon in Georgia Strait, British Colum-
bia. Dep. Fish. Environ. Can., Fish. Mar. Serv. Tech.
Rep. No. 788, 49 p.
HiCKEY, B M
1979. The California Current System - hypotheses and
facts. Prog. Oceanogr. 8:191-279.
Hunter. M. A.
1985. The 1976-1978 brood coho model. Wash. Dep.
Fish. Prog. Rep. No. 222, 146 p.
HUYER. A
1983. Coastal upwelling in the California Current Sys-
tem. Prog. Oceanogr. 12:259-284.
HUYER, A , R O PiLLSBURY, AND R L. SMITH.
1975. Seasonal variation of the alongshore velocity field
over the continental shelf off Oregon. Limnol.
Oceanogr. 20:90-95.
HuYER, A., AND R. L. Smith.
1985. The signature of El Niiio off Oregon, 1982-1983.
J. Geophy. Res. 90:7133-7142.
HUYER, A., E. J. C. SOBEY, AND R. L. SMITH.
1979. The spring transition in currents over the Oregon
continental shelf J. Geophy. Res. 84:6995-7011.
Jaenicke. H. W., R D Brodeur, and T Fujii.
1984. Exploratory gillnetting from the Oshoro-Maru for
juvenile salmonids off southeastern Alaska, 24-25 July
1982. Bull. Fac. Fish. Hokkaido Univ. 35:154-160.
Johnson, S L.
1984. The effects of the 1983 El Nirio on Oregon's coho and
Chinook salmon. Oreg. Dep. Fish Wild. Info. Rep. 84-8,
40 p.
KuNDU, P. K., AND J S Allen
1976. Some three-dimensional characteristics of low-
frequency fluctuations near the Oregon coast. J. Phys.
Oceanogr. 6:181-199.
LOEFFEL, R. E., AND W. O FORSTER
1970. Determination of movement and identity of stocks
of coho salmon in the ocean using the radionuclide zinc-
65. Res. Rep. Fish Comm. Oreg. 2:1-12.
Lord. G . W C Acker. A. C Hartt. and B J Rothschild.
1976. An acoustic method for high-seas assessment of mi-
grating salmon. Fish. Bull., U.S. 74:104-111.
Machidori, S
1966. Vertical distribution of salmon (Genus Oncor-
hynchus) in the North-western Pacific. Hokkaido Reg.
Fish. Res. Lab. 31:11-17.
Manzer, J I
1964. Preliminary observations on the vertical distribu-
tion of Pacific salmon (Genus Oncorhynchus ) in the Gulf
of Alaska. J. Fish. Res. Board Can. 25:1085-1089.
Mason, J. E , and A Bakun
1986. Upwelling index update, U.S. West Coast, 33N-
48N latitude. U.S. Dep. Commer., NOAA Tech. Memo.
NMFS No. 67, 81 p.
McLain, D R
1984. Coastal ocean warming in the Northeast Pacific,
1976-83. In W. G. Pearcy (editor), The infiuence of
ocean conditions on the production of salmonids in the
North Pacific, p. 61-86. Oreg. State Univ. Sea Grant
Coll. Program (ORESU-W-001).
McCarl, B. a., and R. B. Rettig
1983. Influence of hatchery smolt releases on adult
salmon production and its variability. Can. J. Fish.
Aquat. Sci. 40:1880-1886.
McGlE.A M
1984. Evidence for density dependence among coho
salmon stocks in the Oregon Production Index area. In
W. G. Pearcy (editor). The influence of ocean conditions
on the production of salmonids in the North Pacific, p.
37-49. Oreg. State Univ. Sea Grant Coll. Program
(ORESU-W-83-001).
Miller, D R., J G. Williams, and C. W. Sims.
1983. Distribution, abundance and growth of juvenile
salmonids off Oregon and Washington, summer 1980.
Fish Res. 2:1-17.
Milne, D J
1950. The difference in the growth of coho salmon on the
east and west coasts of Vancouver Island in 1950. Fish.
Res. Board Can. Prog. Rep. Pac. Coast Biol. Sta. 85:80-
82.
Myers, K, W. W.
1980. An investigation of the utilization of four study
areas in Yaquina Bay, Oregon, by hatchery and wild
juvenile salmonids. M.S. Thesis, Oregon State Univ.,
Corvallis, 234 p.
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.
Norton, J , D McLain. R Brain, and D Husby
1985. The 1982-83 El Nirio event off Baja and Alta Cali-
fornia and its effect on ocean climate context. In W. S.
Wooster and D. L. Fluharty (editors). El Niiio North, p.
44-72. Univ. Wash. Sea Grant Program.
Oregon Department of Fish and Wildlife.
1982. Comprehensive plan for production and manage-
ment of Oregon's anadromous salmon and trout. Part II.
Coho salmon plan. Oreg. Dep. Fish Wildl., var. p.
Orsi, J A., A G Celewycz, D. G Mortensen. and K A
Herndon
1987. Sampling juvenile chinook salmon (Oncorhynchus
tshawytscha) and coho salmon (O. kisutch) by small
trolling gear in the northern and central regions of south-
eastern Alaska, 1985. U.S. Dep. Commer., NOAA Tech.
Memo. NMFS F/NWC-115, 47 p.
Pacific Fisheries Management Council
1986. Preseason Report 1. Stock abundance analysis for
1986 ocean salmon fisheries. Pac. Fish. Manage.
Counc, var. p.
Pacific Marine Fisheries Commission.
1980. 1977 Pacific salmonid coded wire tag recoveries.
Regional Mark Processing Center for Pacific Coast
States, Portland, OR.
1981. 1978 Pacific salmonid coded wire tag recoveries.
Regional Mark Processing Center for Pacific Coast
States, Portland, OR.
1984a. 1979 Pacific salmonid coded wire tag recoveries.
Regional Mark Processing Center for Pacific Coast
194
PEARCY and FISHER: MIGRATIONS OF COHO SALMON
States, Portland, OR.
1984b. 1980 Pacific salmonid coded wire tag recoveries.
Regional Mark Processing Center for Pacific Coast
States, Portland, OR.
1984c. 1981 Pacific salmonid coded wire tag recoveries.
Regional Mark Processing Center for Pacific Coast
States, Portland, OR.
1985a. 1982 Pacific salmonid coded wire tag recoveries.
Regional Mark Processing Center for Pacific Coast
States, Portland, OR.
1985b. 1983 Pacific salmonid coded wire tag recoveries.
Regional Mark Processing Center for Pacific Coast
States, Portland, OR.
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.
Pearcy. W G
1984. Where do all the coho go? The biology of juvenile
echo salmon off the coasts of Oregon and Washington.
In W. G. Pearcy (editor). Influence of ocean conditions on
the production of salmonids in the North Pacific, p. 50-
60. Oreg. State Univ. Sea Grant Coll. Program
(ORESU-W-83-001).
Pearcy, W G , R D Brodeur, J M Shenker, W. W Smoker, and
Y Endo
In press. Food habits of Pacific salmon and steelhead
trout, midwater trawl catches, and oceanographic condi-
tions in the Gulf of Alaska, 1980-1985. Bull. Ocean
Res. Inst. Univ. Tokyo.
Pearcy, W . J Fisher. R Brodeur, and S Johnson
1985. Effects of the 1983 El Niiio on coastal nekton off
Oregon and Washington. In W. S. Wooster and D. L.
Fluharty (editors). El Nino North, p. 188-204. Univ.
Wash. Sea Grant Program.
Pearcy. W G , and A Schoener
1987. Changes in the marine biota coincident with the
1982-1983 El Nino in the northeastern subarctic Pacific
Ocean. J. Geophys. Res. 92, No. 013:14417-14428.
Peterman, R M . and R D Roittledge
1981. Experimental management of Oregon coho salmon
(Oncorhynchus kisutch ): Designing for yield of informa-
tion. Can. J. Aquat. Sci. 40:1212-1223.
Phinney, D E , D M Miller, and M L Dahlberg.
1967. Mass-marking young salmonids with fluorescent
pigment. Trans. Am. Fish. Soc. 96:157-162.
Richer. W. E
1976. Review of the growth rate and mortality of Pacific
salmon in salt water, and non-catch mortality caused by
fishing. J. Fish. Res. Board Can. 33:1483-1524.
Royer, T. C.
1985. Coastal temperature and salinity anomalies in the
northern Gulf of Alaska, 1970-84. In W. S. Wooster and
D. L. Fluharty (editors), El Nino North, p. 107-115.
Univ. Wash. Sea Grant Program.
SCARNECCHIA, D. L.
1981. Effects of streamflow and upwelling on yield of wild
coho salmon (Oncorhynchus kisutch ) in Oregon. Can. J.
Fish. Aquat. Sci. 38:471-475.
SUSUKI, T., AND H SONODA
1972. On a trial fish finder for salmon and its experimen-
tal results. Bull. Jpn. Soc. Sci. Fish. 38:463-469.
Wakefield, W. W., J. P. Fisher, and W G Pearcy.
1981. Studies of juvenile salmonids off the Oregon and
Washington coast, 1981. Oreg. State Univ. Sea Grant
College Program Ref 81-13, 51 p.
Ware, D M
1978. Bioenergetics of pelagic fish: Theoretical change in
swimming speed and ration with size. J. Fish. Res.
Board Can. 35:220-228.
Weihs, D
1973. Optimal fish cruising speed. Nature 245:48-50.
195
EARLY MARINE LIFE OF ATLANTIC SALMON, SALMO SALAR,
POSTSMOLTS IN THE NORTHERN GULF OF ST. LAWRENCE
J -D DUTIL AND J.-M. COUTU'
ABSTRACT
Postsmolts of Atlantic salmon, Salmo salar, having spent some 2-4 months in the marine environ-
ment were collected in the nearshore zone of the northern Gulf of St. Lawrence. From a back-
calculated smolt length of 135 mm they had increased in length to 265 mm (212 g) on 1 September,
and 306 mm (320 g) on 30 September. The rate of increase in length averaged 1.65 mm/day over more
than 2 months. Individuals reached 35 cm and 500 g in late autumn. Postsmolts moved in small shoals
near the surface and were possibly more active at dawn and dusk. In midsummer, stomach contents
changed from insects and gammarids to sand lance, 40-100 mm in length; vertically migrating
crustaceans also occurred in the stomachs in autumn. River origin of these postsmolts is not known.
The possibility of their belonging to a particular subgroup of some north shore stocks is examined in
relation to scale patterns and size of gonads. This occurrence of postsmolts near shore in late summer
and presumably their late movement out of the Gulf of St. Lawrence indicate the directional nature
of smolt migration to distant feeding areas should be reconsidered. Low sea temperature is hypothe-
sized to trigger the movement out of the Gulf of St. Lawrence. Continual presence of postsmolts in a
shallow layer at the surface could prove to be extremely valuable in forcasting movements and
production.
Many papers have been published concerning the
biology of Atlantic salmon, Salmo salar, but very
little has been said concerning the postsmolt
stage. This stage has been defined as ". . .the
juvenile salmon from the time that it leaves the
river as a smolt until the onset of wide annulus
formation on the scales at the end of the first
winter in the sea" (Allan and Ritter 1977). This
paper presents new data on the Atlantic salmon
postsmolts (hereafter referred as postsmolts) in
the northern Gulf of St. Lawrence, reviews our
current knowledge on the biology of postsmolts,
and points to biological and environmental fac-
tors potentially limiting the success of their early
life in marine environments.
Published data on postsmolts are mainly lim-
ited to stocks in the Baltic Sea. Routes of migra-
tion have been described based on the locations
and time of early recapture from smolt releases in
Sweden and Finland (Carlin 1959; Larsson and
Ateshkar 1979; Ikonen and Auvinen 1984, 1985;
Jutila and Alapassi 1985). Data on predators are
limited (Soikkeli 1973; Valle 1985), and most of
the material concerns predation on smolts in
streams and estuaries (Larsson 1985). Many
analyses of stomach contents have been pub-
'Ministere des Peches et des Oceans, Gouvernement du
Canada, 850, route de la Mer, C.P. 1000, Mont-Joli, Quebec,
Canada G5H 3Z4.
lished, particularly on smaller postsmolts (re-
viewed by Christensen and Larsson 1979; Jutila
and Toivonen 1985). Data on rate of growth
(Ikonen and Auvinen 1985) and rate of mortality
(Carlin 1959) are lacking. However, Baltic
salmon spend their entire sea life in the brackish
waters of the Baltic and nearly 80% of smolt pro-
duction originates from hatcheries (Anonymous
1984). Thus the information derived from salmon
in the Baltic should be extended to other stocks
only with caution.
Publications on postsmolts in the northern At-
lantic and Gulf of St. Lawrence mentioned small
salmon as bycatches of commercial fisheries and
described the distribution of recaptures from
smolt release programs. The earliest report on
postsmolts in the Gulf of St. Lawrence claimed
that small salmon, referred to as "ouananiche" by
local fishermen, were regularly taken near shore
in herring nets in autumn (Comeau 1909).
Kendall (1935) also reported such incidental
catches for the New England coast. Elson (1953)
recorded a bycatch of more than 1,000 marked
postsmolts from one locality in the Bay of Fundy
in the period 1951-53 and reported their mean
length. There are also limited records of
postsmolts taken off France (Vibert 1953) and in
the Gulf of St. Lawrence (Caron 1983) from
smolts tagged in streams. Recently, information
on movements has been derived from tag returns
Manuscript accepted December 1987.
FISHERY BULLETIN; VOL. 86, NO. 2, 1988.
197
FISHERY BULLETIN: VOL. 86, NO. 2
of smolts released in spring in New England and
caught in summer in Canada (Meister 1984). Ru-
mors of bycatch in herring nets along the coast of
the northern Gulf of St. Lawrence in autumn pro-
vided an occasion to acquire some knowledge con-
cerning the elusive postsmolt. Production of
salmon in the sea may well be limited by the
success of smolts in the marine environment.
Materials and Methods
Postsmolts were collected between Bale Trinite
and Riviere-au-Tonnerre in the northern Gulf of
St. Lawrence, in 1982, 1983, 1984, and 1985 (Fig.
1). Fishermen contacted in summer 1982 col-
lected postsmolts in late summer and autumn as
bycatch in herring gill nets. Four of them were
asked in 1983 to monitor the catch of postsmolts
in experimental gill nets in late-September in 4
locations (Bale Trinite, Riviere Pentecote, Port
Cartier, and Sept-Iles). We also monitored sta-
tions in Bale Trinite in 1983 (23 September-11
October), in Baie Trinite and Port Cartier in 1984
(21 August-20 October), and in Sept-Iles in 1985
(20 August-4 October). Finally, smolts and early
postsmolts were collected in seines in June and
July, during an eel marking program in the estu-
ary of Grande Trinite River at Baie Trinite (Fig.
1).
Fishermen used standard herring gill nets in
1982, but custom-made gill nets were used in
1983, 1984, and 1985. Custom-made gill nets had
5 sections of increasing mesh sizes (50.8, 57.2,
63.5, 69.9, and 76.2 mm stretched) covering the
range in mesh sizes of herring gill nets in the
northern Gulf of St. Lawrence. Stretched mesh
sizes were determined by measuring 10 meshes
per section. Sections were 6 m deep and 10 m long.
In 1983, 1984, and 1985, postsmolts were
recorded by section individually. In 1984 and
1985, their position in the nets was recorded more
precisely: floating lines had numbered buoys and
a string divided the nets into 2 halves horizon-
tally. Time of the catch was also recorded. The gill
nets were usually visited at 2-h intervals be-
tween 0600 and 1800, as sea conditions allowed.
They were left fishing overnight. Gill nets were
all set at the surface and near shore (<2 km).
Mean air temperatures for 1982 to 1985 were
drawn from Environment Canada meteorological
summaries for Sept-Iles airport. Temperature
recorders were also tied to nets in 1983, 1984, and
R.MOISIE
SEPT-ILES
PORT CARTIER.
R.PENTECdTE*/
NORTH SHORE
GULF
OF
ST. LAWRENCE
Figure 1. — Locations of the areas investigated in the northern Gulf of St.
Lawrence (shaded area).
198
DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON
1985 to get hourly records of temperature near
the surface, but portions of these records are miss-
ing.
Fish were preserved at -15°C for less than
4 months in 1982 and less than 2 months in 1983,
1984, and 1985. Fork length was measured to
nearest mm and weight to nearest g. Condition
index values were calculated as
C/ = sin 1
100 ■ W
0.5
CI = condition index; W = weight (g); L = length
(mm); s = slope of the length- weight relationship
for age groups combined. Scales, usually taken
below the dorsal fin close to the lateral line, were
cleaned and mounted on glass slides. Stomachs
were preserved in formalin (10%) or alcohol (70%)
to be analyzed later for their contents. The sex
was determined visually and checked histologi-
cally on a subsample of postsmolts collected in
1982. Gonads were weighed to nearest mg and
only in 1982. Gonadosomatic index (GSI) was cal-
culated as
GSI =
GW
TW-GW
100
where GW = gonad weight; TW - total weight.
Scales showing no focal regeneration were ex-
amined for age determination by 2-4 readers.
Readings were discussed and a consensus was
reached in most cases. Reported ages are smolt
ages plus 2-4 months. Fork length at smoltifica-
tion (mm) was back-calculated as
LS =
(LC - 34) ■ RS
RC
+ 34
where LS = back-calculated smolt length; LC
= body length of post smolt; RS = scale radius at
smolt check; RC = scale radius of postsmolt.
Length at scale formation is assumed to be 34
mm. Scales were also examined for any mark in
the postsmolt zone that could be of potential in-
terest. Hence the number of individuals showing
a transition zone between the riverine and
marine increments was determined in 1982, and
the number of individuals showing a summer
check was determined in 1982, 1983, 1984, and
1985. Scale radii were measured along the
postero-anterior axis on 4 scales per individual in
1982, and on 1 scale per individual in 1983, 1984,
and 1985. Projected scales were measured at con-
stant magnification on a digitizing pad connected
to a personal computer. Fork length at formation
of a summer check (mm) was back-calculated for
1982 and 1983 as
LF =
(LC - 34) • RM
RC
+ 34
where LF = back-calculated length; LC = body
length of postsmolt; RM = scale radius at summer
check; RC = scale radius of postsmolt.
Stomach contents were examined in detail in
1982. Results are expressed as percentage of oc-
currence (number of stomachs containing an item
divided by the number of stomachs examined)
and percentage by number (the count number of
an item divided by the count number of all items).
Food items that could not be identified are re-
ferred to as remains. In as far as their condition
allowed, prey fish were usually measured to the
nearest cm. Stomach contents examined in 1983,
1984, and 1985 were consistent with the 1982
conclusions, so the results are not presented.
Results
The Age and Sex Composition
Three-year-old female salmon dominated the
catch both in 1982 and 1983 (Table 1). Postsmolts
ranged from 2 to 5 years in 1982 and from 2 to 4
years in 1983. Five-year-old salmon made up less
than 4% of the catch in 1982. Three-year-old and,
to a lesser extent, 4-yr-old salmon dominated the
catch both in 1982 and 1983 (82% of the catch);
mean smolt age was 3.2 ± 0.7 years. Age composi-
tion in 1983 differed from 1982 because of an in-
crease in the percentage of 2-yr-old salmon (12%
to 21%) and a decrease in the percentage of 5-yr-
old salmon (x^ = 15.96, P < 0.01). The sex ratio
(male:female) for 1982 and 1983 combined was
Table 1 . — The age and sex composition of the catch of postsmolt
Atlantic salmon along the north shore of the Gulf of St. Lawrence,
1982 and 1983.
Sex
Smolt
age
Year
2
3
4
5
Total
1982
Female
26
114
79
7
226
Male
19
67
49
7
142
1983
Female
25
58
15
0
98
Male
8
26
23
0
57
Total
Female
51
172
94
7
324
Male
27
93
72
7
199
199
FISHERY BULLETIN; VOL. 86, NO. 2
0.62. This ratio did not change between years (x^
= 7.92, P < 0.01) and was significantly less than
1.0 (binomial test: z = -5.37, P < 0.001). There
was no trend between sex ratio and smolt age in
1982 (x^ = 1.21, P = 0.75), but the percentage of
males tended to increase with smolt age in 1983
(X^ - 12.67, P - 0.002).
Length at Smoltification
The mean length at smoltification was esti-
mated by back-calculation at 135 ± 15 mm for
postsmolts collected in 1982 and 1983. Scale ra-
dius was linearly related to body length both in
1982 (4 scales measured per fish, P < 0.0001) and
in 1983 (1 scale measured per fish, P < 0.0001);
probability values are those of F-tests from anal-
ysis of variance. Back-calculated smolt lengths
were normally distributed (P = 0.80): 75% of the
postsmolts ranged between 120 and 150 mm at
smoltification. There were also 7% individuals in
the 160-200 mm range. Differences in mean
smolt length were not significant between the
ages and the sexes both in 1982 (F = 1.84,
P = 0.08) and in 1983 (F = 1.05, P = 0.39), and,
pooling the ages and the sexes, between 1982 and
1983 (^ = 1.32, P >0.15).
Rate of Increase in Size
The rate of increase in size of postsmolts was
very rapid in summer both in 1982 and 1983.
Postsmolts collected in 1982 between mid-August
and mid-October in = 383; mean date is 1 Sep-
tember) averaged 265 ± 25 mm (range 195-328)
and 212 ± 58 g (range 92-389). In 1983,
postsmolts collected between mid-September and
mid-October (n = 155, mean date is 30 Septem-
ber) averaged 306 ± 17 mm (range 258-362) and
320 ± 57 g (range 192-565). There was no differ-
ence in mean size between males and females and
between age-classes in 1982 (P > 0.40) and 1983
(P > 0.20); probability values are those of F-tests
from 2-way analysis of variance. From a mean
smolt length of 135 mm and assuming smolts mi-
grated to estuaries 15 June, postsmolts grew at a
rate of 1.65 mm/day during a 2.5-mo (15 June —
1 September) and a 3.5-mo (15 June-30 Septem-
ber) period in 1982 and 1983, respectively.
This estimate is conservative because the rate
of increase in length tended to decrease late in the
sampling period. Postsmolts steadily increased in
length and weight in the period mid-August to
mid-September 1982 (Figs. 2, 3). From mid-
September, the rate tended to slow down. The
inclusion of data for 1983, collected later in the
autumn, corroborates this observation indicating
that conditions changed in late-September in
1982 and 1983.
Length-weight relationships were examined
for 1982 and 1983 separately. The analysis of co-
variance showed that males and females exhib-
ited the same length-weight relationship, both in
1982 (P - 0.53) and 1983 (P - 0.42). Similarly,
length-weight relationship did not change be-
tween age groups in 1982 (P - 0.06) and 1983
(P = 0.49). The covariance for 1982 was nearly
significant because the slope for 5-yr-old
postsmolts, based on 15 individuals, was larger
than for other age groups. However, mean condi-
tion index values by age revealed no significant
difference between age groups in 1982 (P = 0.11)
or 1983 (P = 0.28).
Since there was also no significant difference in
the length- weight relationship between 1982 and
1983 (P = 0.14), the data were pooled. Thus the
length-weight relationship for postsmolts col-
lected in this study can be described as a single
regression:
log W = (2.8280 • log L) - 4.5336
P < 0.001, n = 539
where W = weight (g); L = length (mm).
Maturation of Gonads
Postsmolts were all immature both in 1982 and
1983, but differences were observed between
males and females in 1982. Testes averaged
48 ± 6 mg (95% C.L., n = 124) for a mean male
gonadosomatic index value of 0.025% ± 0.003%
(95% C.L., n = 124). There was no significant dif-
ference in the mean value of either parameter
between age groups (P > 0.52). Testis weight in-
creased in time and as body length and body
weight increased, but again there was no differ-
ence between age groups (Table 2). However, the
gonadosomatic index did not change in time
(P = 0.10) or as postsmolts' size increased
(P = 0.16 for body length and P = 0.10 for body
weight), suggesting that changes in size of the
testes were not allometric in male postsmolts in
the autumn period. Regressions were tested and
compared following Snedecor and Cochran (1967)
and Sokal and Rohlf (1969).
There was more variability in the data for fe-
200
DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON
360-.
350
340-
330
320H
310
300-
290-
E
E
I-
O
z
UJ
280
270
260
250
240
230
220
210
200
190-
180-
• *♦ • .
* A *
01 06 11 16 21 26 31 05 10 15 20 25 30 05 10 15 20 25 30
AUGUST SEPTEMBER OCTOBER
DATE
Figure 2. — Seasonal changes in length of Atlantic salmon postsmolts collected in 1982 and 1983.
Table 2. — Functional regressions of gonad weight on date, body length, and
body weight of Atlantic salmon male postsmolts {n = 124). Lengths and
weights are log-transformed.
Date
Length (mm)
Weight (g)
Regression
Covariance
(age groups)
P< 0.00011
P = 0.31 n.s.
P< 0.00012
P = 0.56 n.s.
P< 0.00013
P = 0.47 n.s.
1/ = 0.0168 X- 5.41 66
2y = 7.8496 X - 20.3350
3V = 2.6798 X- 7.5298.
201
FISHERY BULLETIN: VOL. 86, NO. 2
O)
lU
460
440
420
400
380
360
340
320
300
280
260
240
220
200
180
160
140
120
100
(473) (565) (464)
(488)
01 06 11 16 21 26 31 05 10 15 20 25 30 05 10 15 20 25 30
AUGUST
SEPTEMBER
DATE
OCTOBER
Figure 3. — Seasonal changes in weight of Atlantic salmon postsmolts coUecteci in 1982 and 1983.
males. Though the size of ovaries did not differ
significantly between age groups (P = 0.07), go-
nadosomatic index values increased in older fe-
male postsmolts (P = 0.0006). Mean values
ranged from 133 mg (0.06%) for 2-yr-old post-
smolts to 205 mg (0.10%) for 5-yr-old postsmolts
(Table 3). The regressions of ovary weight on body
length, body weight, and time were all significant
except for 5-yr-old females (Table 4), and differed
between age groups particularly in elevation
(P < 0.05, n - 213). The regressions of gonadoso-
matic index values on the same variables were
Table 3. — Gonad weight (mg) and gonadosomatic index (GSI) (%)
of Atlantic salmon female postsmolts: mean values and confidence
limits (n = 213).
Gonad
Mean
weight
95% C.L.
interval
GSI
Age
group
Mean
95% C.L.
interval
2
3
4
5
133
152
175
205
124-144
142-163
161-189
194-217
0.064
0.077
0.092
0.095
0.060-0.067
0.073-0.082
0.087-0.098
0.091-0.099
202
DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON
not significant indicating that, in females as in
males, changes in size of the gonads were not
allometric (body length: P = 0.06; body weight:
P = 0.73; time: P - 0.68).
Scale Marks
Scales used for age determination were also ex-
amined for any mark that could be of use in stock
identification. Many individuals exhibited a tran-
sition zone on their scales. Circuli in this zone
were more wide-spaced than circuli laid down in
earlier summers, but they were more narrow-
spaced than circuli formed in summer in a marine
environment. This mark was present on 73% of
the scales in both males and females (72.6 and
73.87f) but tended to decrease as smolt age in-
creased: 84% (age 2 years), 75% (3 years), 68% (4
years), and 62% (5 years). However, the trend was
not significant (x^ = 5.74, P > 0.10).
Scales also exhibited summer checks in 1982,
1983, and 1984. The number of circuli between
the smolt mark and the summer check was
10.9 ± 2.6 (mean ± SD) (range 7-15) in 1982 and
9.7 ± 2.2 (range 5-15) in 1983. The ratio between
the radius to summer check and the radius to
smolt mark averaged 2.00 ± 0.30 (range 1.49-
2.45) in 1982, and 1.67 ± 0.19 (range 1.26-2.32)
in 1983. The overall percentage of occurrence was
low in 1982 (3.6%) and involved only specimens
collected in the area Bale Trinite-Pointe aux
Anglais where 11 out of 13 postsmolts examined
showed a summer check. The incidence of sum-
mer checks on scales increased markedly in 1983
(55%) and 1984 (75%), and summer checks were
no longer restricted in distribution. Examination
of data concerning postsmolts possessing a sum-
mer check showed no relationship between the
check and measured biological variables except
perhaps in 1982. The gonads of males having a
summer check on their scales in = 5) were heav-
ier than in males having no summer check (120
mg and 48 mg respectively). Their gonadosomatic
index was also higher (GSI = 0.042% and 0.025%
respectively). This was not so in females. Infor-
mation on gonads is not available for 1983, 1984,
and 1985.
Back-calculated lengths at summer check for-
mation were larger in 1982: 22 ± 2.5 cm in 1982
and 20 ± 2 cm in 1983. Postsmolts measured 265
mm on 1 September. With a mean rate of increase
in length of 1.65 mm/day, the check must have
been formed some 26 (1982) to 40 (1983) days
earlier, i.e., in late-July. This is confirmed by ex-
trapolation of the length-date plot (Fig. 2). No
salmon were noted possessing 2 summer checks
on their scales.
Food Items
Drastic changes took place in prey selection be-
tween postsmolts collected in summer and in au-
tumn. The stomach contents of 40 salmon ranging
up to 188 mm in length (70 g) collected in the
second half of July 1984 in the estuary of the
Grande Trinite River, near Bale Trinite, revealed
a low diversity in prey items, indicative of a tran-
sition period between the riverine and marine en-
vironment. Small fish remains were observed in
only 5 out of 39 stomachs containing food re-
mains. Invertebrates were observed in 38 stom-
achs, classes Insecta and Crustacea occurring in
87 and 92% of the stomachs respectively. Food
items consisted mainly of 2 families: Chironomi-
Table 4. - Functional regressions of gonad weight on date, body length, and
body weight of Atlantic salmon female postsmolts (n =213). Lengths and
weights are log-transformed.
Significance
Functional regression
Date
Length
(mm)
Weight (g)
2: P = 0.0001
3 : P = 0.0001
4 : P = 0.0058
5 : P = 0.26 n.s.
covariance : P
2
3
4
5
2
3
4
5
P < 0.0001
P < 0.0001
P < 0.0001
P = 0.34 n.s.
covariance
P < 0.0001
P < 0.0001
P < 0.0001
P = 0.28 n.s.
covariance
Y = 0.0075 X
Y = 0.0044 X
Y = 0.0052 X
2.7127
1.9052
2.0229
0.04
y = 3.8015 X- 10.1089
y = 2.9105 X - 7.8659
/ = 3.1893 X -8.4701
P = 0.02
Y= 1.31 10 X- 3.9428
Y = 1 .0385 X - 3.2227
Y= 1.1763 X- 3.4659
P = 0.04
203
FISHERY BULLETIN: VOL. 86, NO. 2
dae (95% of insects by number) and Gammaridae
(92% of crustaceans by number).
Later in summer and autumn, postsmolts con-
sumed mainly small fish. Stomach contents were
analyzed for 373 out of 385 postsmolts collected in
1982. There were 109 stomachs with no food re-
mains (29%). They were most prevalent in the
first half of August: 46%, 1-15 August; 25%, 16-
31 August; 26%, 1-15 September; 25%, 16 Sep-
tember-31 October. Fishes dominated the list of
prey items as they occurred in 238 out of 264
(90%) stomachs containing food remains, includ-
ing 200 postsmolts (84%) that fed exclusively on
small fish. Fishes could be identified in 157 stom-
achs. Diversity was low, capelin, Mallotus villo-
sus, occurring in 16 stomachs (10%) and sand
lance, Ammodytes americanus, in 145 stomachs
(92%). Ammodytes americanus dominated in
terms of percentage by number (94%). Postsmolts
consumed smaller A. americanus, in the 40-100
mm range (Fig. 4). Stomachs examined in 1983,
1984, and 1985 confirmed these observations.
Invertebrates were found in 69 out of 264 stom-
achs (26%) containing food remains. Eighteen
postsmolts had only invertebrates in their stom-
achs (26%). In contrast with smolts in the estua-
rine environment, postsmolts did not rely on in-
sects; insects occurred in only 8 stomachs (3%),
whereas crustaceans occurred in 61 stomachs
(23%) (respectively, 12 and 88% by number). Fur-
thermore, the class Amphipoda no longer domi-
nated the crustaceans (Table 5).
>-
O
z
u
o
ui
oc
u.
20 40 60 80 100 120 140 160
LENGTH - CLASS (mm)
Figure 4. — Length distribution of sand lance in the stomachs of
Atlantic salmon postsmolts collected in 1982.
Horizontal and Vertical Position
in the Nets
Postsmolts were not randomly distributed in
the nets in 1984. They occurred most frequently
(78% of the individuals) in the top half of the nets
(binomial test; z = 4.62, P < 0.001; Siegel 1956).
Furthermore, 25 out of 74 salmon occurred alone
in the nets, but many also occurred in clusters.
Positions of postsmolts are indicated by number
of nearest buoy on the head-line for those sets
having taken 2 salmon and more (stations visited
at 2-h intervals usually but also nets set
overnight) (Table 6). Distributions are likely not
random in sets 3, 6, 7, 9, 10, and 11 and most
particularly in sets 13 and 15. The catch was low
in midday: 3 salmon between 0900 and 1200 and
5 between 1200 and 1500. This increased to 12
between 1500 and 1800. The majority were
caught later than 1800 (36) and in the morning
between 0600 and 0900 (20).
Finally, positions in the nets were analyzed in
terms of selectivity. The gear used in 1982 could
not be controlled. Fishermen reported using
Table 5. — Crustacean organisms in the stomachs of Atlantic
salmon postsmolts collected in the penod August-October 1982,
based on 39 stomachs containing identifiable crustaceans.
Crustacean
order
Percentage
of
occurrence
Percent-
age by
number
Main organisms
Euphausia-
cea
Decapoda
Amphipoda
87
28
15
68
24
8
Meganyctiphanes norvegica
Thysanoessa inermis
Thysanoessa raschi
Chionoecetes opilio
(larvae)
Table 6. — Positions of Atlantic salmon postsmolts in nets by num-
ber of nearest buoy on head-line for catches of 2 salmon and more.
Set
No. of
no.
smolts
Number of nearest buoy to each smolt
1
2
11
32
2
2
12
23
3
2
14
15
4
2
15
23
5
2
19
27
6
2
22
24
7
2
23
23
8
2
23
39
9
2
25
27
10
2
23
27
27
11
4
16
21
21
43
12
5
16
24
33
42 44
13
5
23
23
23
23 23
14
6
12
17
18
18 22 42
15
9
11
12
12
12 13 13 13 14 14
204
DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON
62 mm stretched mesh nets (range 57-70 mm);
mesh size is regulated. The nets used in 1983,
1984, and 1985, had 5 sections of increasing
stretched mesh sizes: 50.8, 57.2, 63.5, 69.9, and
76.2 mm. Catches were recorded for individual
mesh sizes in 1983 combining the 4 stations.
Modal length increased only slightly as mesh size
increased. Furthermore, the distribution for the
69.9 mm mesh was skewed to the right, indicat-
ing no larger sized postsmolts were present.
There was also no catch in the 50.8 mm mesh, but
this may have been because of the small mesh
section being made of a coarser material in 1983.
Postsmolts were frequent in = 49) in the 57.2 mm
mesh, but the distribution for this mesh does not
suggest the existence of smaller postsmolts as the
size range is similar to that of the 63.5 mm mesh
and covers the size range for the 5 meshes com-
bined. Catches declined from a maximum in the
63.5 mm mesh (56) to 40 and 9 in the 69.9 and
76.2 mm mesh, respectively. Therefore it is un-
likely that there was any bias except perhaps
against smaller postsmolts.
Fall Movement out of Coastal
Reaches
Postsmolts gradually left the nearshore area in
late-September. Fishermen in 1982 started col-
lecting postsmolts in mid-August. Their bycatch
declined from mid-September and was low in Oc-
tober. The catch declined partly because most
commercial fishermen were asked to return no
more than 20 salmon each, and they reached this
limit early (Table 7). Fishing was initiated later
in 1983 and took place over a shorter period (23
September-11 October), but a 68^ decline in the
catch was observed between the period 29 Sep-
tember-5 October and the period 5-11 October.
Finally in 1984, postsmolts came near the coast
near the end of August and moved out in mid-
September (Table 8) so that no salmon were
caught in the period 20 September-20 October.
This movement out of the nearshore zone was
associated with decreasing near-surface tempera-
tures in autumn. Temperatures measured near
the surface closely followed the mean air temper-
atures recorded in Sept-Iles (Fig. 5). Since most
postsmolts travelled near the surface, mean air
temperatures were assumed to reflect prevailing
conditions for postsmolts. In 1982, postsmolts
were abundant until mean air temperatures de-
clined below 5°C, i.e., in early October. The situa-
tion was similar in 3 stations out of 4 in 1983,
particularly in the Bay of Matamek River near
Sept-Iles and in Riviere Pentecote, 2 stations
closer to our monitoring station. In 1984, this de-
cline in mean air temperature occurred earlier
(26 September), but surface temperatures were
also lower than air temperatures, and the catch
declined (Table 8) as soon as surface tempera-
tures fell below 5°C (mid-September.) Hence, low
temperatures were associated with a movement
of postsmolts out of the nearshore zone. However,
the reverse is not true: results for 1983 and 1984
suggest that postsmolts do not necessarily move
towards nearshore zones when prevailing tem-
perature conditions are favorable.
Variations in seasonal abundance are masked
by a general decline in salmon abundance near
shore from 1982 to 1985. Fishing effort could not
be assessed in 1982, but the incidence of
postsmolts in coastal herring nets in 1982 was
such that it can reasonably be concluded they
were more abundant than in 1983, 1984, or 1985.
Relative abundance can be assessed for 1983,
1984, and 1985. Bale Trinite and Port Cartier
stations were operated in 1983 and 1984 using
similar nets at the same locations each year and
showed that postsmolts were more abundant in
1983 than in 1984 (Table 9). Finally, a station
was monitored near Sept-Iles in 1985 in an area
where the best catches were made in 1982, using
Table 7. — Time distribution of the commercial
bycatch of Atlantic salmon postsmolts in 1982.
Period
Number
10-31 August
01-15 September
16-30 September
01-15 October
16-31 October
210
118
21
24
8
Table 8. — Catch of Atlantic salmon postsmolts by
period and locality in 1984.
Locality
Baie
Port
Period
Trinite
Cartier
21-24 August
1
4
24-27 August
0
0
27-30 August
0
1
30 August-02 September
0
14
02-05 September
4
8
05-08 September
6
9
08-11 September
0
2
11-14 September
3
16
14-17 September
0
8
17-20 September
0
0
20 September-20 October
0
0
205
20
15
10
5
0
-5
20
15
10
P 5
uj 0
cc
<
£ 20
Q.
15
10
5
0
-5
20
15
10
5
0
-5
FISHERY BULLETIN: VOL. 86, NO. 2
1982
UJ
Sept-lles
air temperature
^ 05 10 15 20 25 30 04 09 14 19 24 29 04 09 14 19 24 29
1983
Port Cartier
sea temperature
05 lb i5 20 25 30 04 09 14 19 24 29 04 09 14 19 24 29
1984
Bale Trinite
sea temperature
05 io i5 20 25 30 04 09 14 19 24 29 04 09 14 19 24 29
1985
Sept-lles
sea temperature
I r
I ~
05 10 15 20 25 30 04 09 14 19 24 29 04 09 14 19 24 29
AUGUST
SEPTEMBER
DATE
OCTOBER
Figure 5. — Mean daily air temperature in Sept-lles (— -), and mean daily sea temperature at 0.5 m ( — ) in Port Cartier
(1983), at 6 m (---) in Baie Trinite (1984), and at 0.5 m (---) and 6 m ( ) in Sept-lles (1985).
206
DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON
Table 9. — Relative abundance in terms of catch per unit of eHort
(CPUE) at 2 stations run in 1983 and 1984 using similar nets at tfie
same sites.
Baie Trinite
Port Cartier
Period
Catch
E
CPUE
Catch
E
CPUE
1983
23 Sept.- 11 Oct.
54
38.5
1.4
40
20.0
2.0
1984
21 Aug. -31 Aug.
1
20.5
0.1
5
32.5
0.2
30 Aug. -17 Sept.
15
54.5
0.3
57
73.0
0.8
17 Sept.-23 Sept.
0
22.0
—
0
17.0
—
23 Sept.- 11 Oct.
0
55.5
—
0
45.5
—
11 Oct.-18 Oct.
0
9.5
—
0
21.0
—
'CPUE: catch of 1 net in 24 hours.
similar nets to 1983 and 1984. The catch was
nearly nonexistent: 5 salmon for 120 unit effort
( 1 unit effort, is 1 net x 24 hours). Thus, based on
limited observations, numbers of salmon near
shore in summer and autumn seem to be highly
variable from year to year.
DISCUSSION
Postsmolts of Atlantic salmon stay much longer
near our coasts than is usually believed. Though
early months in the marine environment have
been shown to have a marked influence on
salmon runs 1 and 2 years later (e.g., Christensen
and Larsson 1979; Scarnecchia 1983, 1984),
postsmolt biology has been a neglected area of
investigation. North American smolts are as-
sumed to migrate rapidly out of the estuaries of
their home rivers, to feeding areas located far out
in the North Atlantic east of the Grand Bank
(Templeman 1968; Reddin 1985) and north to
Labrador and Greenland (Saunders 1966; Tem-
pleman 1967). They return 1 or 2 years later to
spawn in home rivers. This study, and early
records, indicate that some postsmolts remain in
coastal areas as late as autumn before moving
offshore. This was clearly suggested by Comeau
(1909) who stated that 0.5-1.5 lb postsmolts were
regularly taken in autumn along the north shore
of the Gulf of St. Lawrence. Fishermen inter-
viewed in 1982 on the north shore of the Gulf of
St. Lawrence, from Pointe-des-Monts to Blanc
Sablon, declared incidental catches of postsmolts,
mainly in the months of August and September.
The majority declared taking postsmolts each
year. Smolts were also shown to linger in estuar-
ies of the north shore of the Gulf of St. Lawrence
(this study; Power and Shooner 1966; Randall and
Power 1979). Huntsman (1939) mentioned their
occurrence in autumn in herring nets near the
mouth of Gaspe Bay. Fall catches also occurred in
New England (Kendall 1935). Recently, smolts
released in New England were caught as post-
smolts in coastal areas of Canada (Meister 1984).
There are stocks in the Bay of Fundy (as the
stocks in the Baltic) that do not go on extensive
migrations in the North Atlantic (Huntsman
1939; Jessop 1976): postsmolts of these stocks are
regularly taken in herring nets in Passa-
maquoddy Bay and off Grand Manan Island (El-
son 1953, 1964; Allen et al. 1972). Hence the pres-
ence of postsmolts near shore in autumn (or in
summer and autumn) is a characteristic of the
marine life of Atlantic salmon in North America.
Timing of migration has been described for
hatchery-reared smolts released in Sweden
(Larsson 1974) and Finland (Jutila and Alapassi
1985). Tags were returned mostly from a distance
of less than 10 km between days 0 and 10, 20-50
km between days 10 and 20, and 50-100 km 2
months past their release, in the brackish waters
of the Gulf of Bothnia (Jutila and Alapassi 1985).
Behavior of postsmolts is similar to that of 1-
and 2-sea-year salmon. Postsmolts in this study
occurred mainly near the surface as indicated by
their distribution in the nets. LaBar et al. (1978)
concluded that smolts migrated near the surface
in the Penobscot estuary. Templeman (1967,
1968) also found salmon to occur near the surface
in the Northwest Atlantic: 62% occurred in the
top 0.6 m and 90% in the top 1.5 m in July and
August 1965. Similarly in 1966, most salmon
were taken in the top 1.5 m, the number caught
decreasing sharply below 0.6 m. Similar observa-
tions were made on Baltic salmon (Carlin and
Lundin 1967; Christensen 1968). There is less in-
formation available on schooling. Postsmolts did
not regularly have a clustered distribution in
nets, but considering that a net does not retain all
the salmon striking it, there were still many in-
stances of salmon moving in schools. Thurow
(1968) came to the same conclusion for older
salmon in the Baltic. Templeman (1967) pre-
sented limited evidence for salmon in the North-
west Atlantic, but reached negative conclusions
later (Templeman 1968). Finally there are lim-
ited data in the literature concerning the rhythm
of activity of salmon in the marine environment.
Christensen and Lear (1980) showed that in West
Greenland best catches occurred early in the
morning (0600-0800), decreased sharply between
0800 and 1000, and were low between 1000 and
1400. The nets were not set between 1400 and
207
FISHERY BULLETIN: VOL. 86, NO. 2
0600. Catches in this study were also low in mid-
day. Either this is a reflection of 2 peaks of activ-
ity, dawn and dusk, as is common in salmonids in
freshwater, or this is possibly due to salmon
avoiding the nets in high light levels. Thus
postsmolts and 1- and 2-sea-year salmon appear
to have a similar behavior at sea. They move in
small schools close to the surface and are possibly
more active at dusk and dawn.
Food items in the stomachs of postsmolts
changed markedly in summer and indicated a low
diversity of prey. This is in contrast to findings
reported for salmon in the Northwest Atlantic.
Grande Trinite River postsmolts had fed mostly
on chironomids and gammarids in late-July.
Baltic salmon postsmolts fed mainly on aerial in-
sects though small fish and crustaceans also oc-
curred in the stomachs of postsmolts in the south-
ern Baltic (reviewed by Christensen and Larsson
1979). Jutila and Toivonen (1985) also found
aerial insects to be the dominant food items in the
stomachs of small postsmolts ( <20 cm) in the Gulf
of Bothnia (Baltic). They observed that post-
smolts were not selective and must have fed near
the surface (20 cm surface layer). Postsmolts col-
lected later in the present study relied mainly on
small sand lance. Insects and gammarids had
been replaced by vertically migrating crus-
taceans such as Meganyctiphanes noruegica
(Kulka et al. 1982). Thurow (1968) estimated 25
cm to be the length threshold for piscivorous feed-
ing by Baltic salmon. In the present study, this
size was likely reached in the first half of August
1982. This coincides with a major change in stom-
ach contents and a high percentage of stomachs
containing no food. The data on postsmolts in
July are too limited to suggest that sand lance
abundance could limit the early success of
postsmolts at sea, but potential relationships in
late summer should be tested as was done for
capelin by Reddin and Carscadden (1981). Data
on 1- and 2-sea-year salmon indicate they will
readily feed on a diversity of prey items, main
items including Arctic squid, Gonatus fabricii;
paralepids, Paralepis coregonoides ; and lantern
fishes (Lampenyctus sp., Notoscopelus sp.) (Tem-
pleman 1967, 1968; Lear 1980). Sand lance and
capelin are dominant items in West Greenland
and on the coast of Newfoundland (Lear 1972,
1980), and on the Grand Bank (Reddin 1985).
Reddin ( 1985) observed major changes in stomach
contents between salmon on the Grand Bank
(sand lance and capelin) and east of the Grand
Bank (Bathylagidae, Paralepis sp., and crus-
taceans), emphasizing that salmon are not selec-
tive predators.
The rate of increase in mean length averaged
1.65 mm/day in the Gulf of St. Lawrence over the
summer period in 1982 and 1983. This value is
based on the hypothesis that smolts migrated to
estuaries in mid-June. Smolt migration took
place in the first half of June in Restigouche River
in the southern Gulf of St. Lawrence (Peppar
1982) and in the second half of June in Grande
Trinite River in the northern Gulf of St.
Lawrence (Caron 1984). Downstream migration
of smolts peaked at various dates in June in West-
ern Arm Brook in western Newfoundland (Chad-
wick 1981). The calculated rate of increase is also
based on a mean back-calculated smolt length of
135 mm. Length at smoltification averaged 125-
130 mm in Grande Trinite River (mean and SD:
127.5 ± 12.3, n = 88, in 1981; 125.8 ± 10.9,
n = 92, in 1982; see also Caron 1984). Matamek
River and Moisie River smolts measured 125-150
mm (Schiefer 1972). They measured 150 mm in
Little Codroy River (Murray 1968) and 174 mm in
Western Arm Brook (Chadwick 1981). There are
no data in the literature concerning the rate of
increase in size of smolts and postsmolts in the
marine environment. Postsmolts in the Bay of
Fundy reached a mean length of 296 mm in mid-
August 1952 (Allen et al. 1972), some 3 cm more
than postsmolts in this study: 265 mm and
306 mm on 1 September and 30 September. How-
ever the high value of the power exponent of the
length-weight relationship as compared with
salmon in Newfoundland and Labrador (Lear
1973) indicates postsmolts were not in poor condi-
tion. There was possibly a decline in the rate of
increase in length in mid-summer as suggested
by the large proportion of scales showing a sum-
mer check (false-annulus) in 1983. This occurred
10 circuli from the smolt check in mid-summer in
postsmolts 20-22 cm in length, i.e., prior to this
study period, and may have been produced as a
response to a shortage of prey or to deteriorating
environmental conditions. Elson (1953) also no-
ticed the frequent occurrence of a slowing of
growth 6-10 circuli out from the last parr an-
nulus. The percentage of occurrence of the check
varied between locations (1982) and between
years (1982 < 1983 < 1984). Therefore it is not
likely to be a response to a change in postsmolt
biology such as a scheduled shift in prey selection.
However, the summer check can be thought of as
a potential tool for stock discrimination. Some
26-32 circuli are formed before the first sea an-
208
DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON
nulus is formed (Lear and Misra 1978) at a length
of 46-50 cm in the first half of April (Munro
1970).
Presence of postsmolts near shore in late sum-
mer in the northern Gulf of St. Lawrence, as re-
ported in this study, and presumably their late
movement out of the Gulf of St. Lawrence indi-
cate that the directional nature of the migration
should be reconsidered. There are some smolts
that do not head towards the high seas as soon as
they reach the estuaries. They seem to roam
nearby unless prevailing conditions are not favor-
able. Temperature can be hypothesized as trig-
gering the late movement of postsmolts out of the
northern Gulf of St. Lawrence. Saunders (1986)
reviewed the thermal biology of Atlantic salmon
and suggested the thermal range for salmon in
the sea is lower than for juvenile salmon in fresh-
water. Salmon occur mainly at temperatures
ranging from 4° to 8°C in the Northwest Atlantic
(Templeman 1968; May 1973; Reddin 1985). Post-
smolt movements out of the nearshore area took
place in a short period as temperature was de-
creasing, between mid-September and mid-
October. Postsmolts were more abundant in 1982
and 1983 as mean air temperature ranged
between 4° and 10°C in early autumn. In 1984,
sea temperature decreased rapidly from more
than 15°C in late-August down to 2°C in mid-
September; postsmolts vanished from the near-
shore area as temperature declined below 4°C.
However in 1985, they did not come near the coast
though sea temperature ranged between 8° and
12°C. Saunders et al. (1975) reported the lethal
temperature of salmon in seawater to be -0.7°C.
This precludes the possibility of salmon over-
wintering in the Gulf of St. Lawrence unless they
return to freshwater, as do some salmon in the
Koksoak River (Cote et al. 1984; Robitaille et al.
1984a, b), or move down to midwater, a behavior
described for salmon in the Baltic in response to
high temperatures (>12°C) at the surface (Aim
1958). Comeau (1909) reported postsmolts found
in the stomach of seals off Pointe-des-Monts in
January and February. Low sea temperature has
been hypothesized as limiting the passage of Kok-
soak River smolts (Ungava Bay) to West Green-
land in some years, thereby resulting in an estu-
arine population (Power 1969, 1981). This
situation might also be hypothesized to occur in
the Gulf of St. Lawrence. For instance in 1983,
mean air temperature, not to mention minimal
temperature, decreased from 4° to 0°C and less in
a short period near the end of October. Masses of
seawater carried by gyres east and west of Anti-
costi Island, and presenting momentarily favor-
able conditions, can get surrounded by masses of
seawater at lower temperature. Should salmon
rely on temperature as a cue for their movement
out of the Gulf of St. Lawrence, then late mi-
grants could not escape as conditions deteriorate.
The origin of postsmolts collected in this study
is not known. They may be a particular subgroup
of some north shore stocks. Postsmolts in this
study smoltified earlier and at a smaller size than
stocks in northern Newfoundland (Chadwick
1981). However, their origin cannot be deter-
mined based on smolt length or smolt age distri-
butions. For instance, there is a general tendency
for increasing smolt age with latitude, but there
is much variability in the data at latitudes below
52°. Data for salmon stocks in rivers near 50°
latitude range from 3 to 4 years (Power 1981).
Furthermore, postsmolts in this study had an age
distribution similar to that of salmon in the Port-
aux-Basques (Newfoundland) drift net fishery.
Port-aux-Basques salmon migrated to rivers all
around the Gulf of St. Lawrence (Belding and
Prefontaine 1938).
Postsmolts in this study may also be from a
particular subgroup of individuals, such as late-
migrant smolts. Power and Shooner (1966) and
Randall and Power (1979) observed remnants of
the smolt migration feeding in river estuaries on
the north shore of the Gulf of St. Lawrence in July
and August. Furthermore coho salmon released
in mid- to late-summer did not leave the general
area of release (Mahnken and Joyner 1973).
Since grilse are known not to migrate as far as
2- and 3-sea-year salmon in the Northwest At-
lantic (Ruggles and Ritter 1980), postsmolts in
this study can be thought of as potential grilse.
However, the only indication in that direction
that we have is the observation that some males
having a summer check on their scales in 1982
had a higher gonadosomatic index than males
having no summer check on their scales (0.042
and 0.025% respectively). However, this is based
on a small number of postsmolts as few salmon
exhibited a summer check in 1982 and, unfortu-
nately, no gonads were preserved in 1983-85.
There are some stocks maturing mainly as grilse
among the north shore stocks, but grilse are
nearly exclusively males in these stocks (Schiefer
1972; Caron 1984). Postsmolts in this study were
62% females. Female grilse are common in New-
foundland (Chadwick 1981; Power 1981). There
are no published data on Anticosti stocks. Poten-
209
FISHERY BULLETIN: VOL. 86, NO. 2
tial grilse have been reported from the Grand
Bank in offshore fisheries in the Northwest At-
lantic (Reddin 1985), and it has been suggested
that stocks maturing as grilse in the Bay of
Fundy may not leave the general area throughout
their entire marine life (Jessop 1976).
Knowledge on the marine biology of Atlantic
salmon postsmolts has been a neglected area of
research. Their continual presence in sea surface
waters could prove to be extremely valuable in
forecasting salmon movements and production
(Chadwick 1982; Scarnecchia 1984). Potential
studies include mortality rates in the period of
transition (July) and the relationship between
low temperatures (3°-4°C) and postsmolt migra-
tion out of the Gulf of St. Lawrence. Biological
data in general are needed to be included in mod-
els forecasting salmon runs in the North Atlantic.
ACKNOWLEDGMENTS
The authors wish to thank Fisheries and
Oceans personnel, including many summer stu-
dents, having assisted in field work on the North
Shore, particularly S. Cloutier, M. Fortin, M.
Laverdiere, Y. Lavergne, B. Legare, B. Mercille,
M. Michaud, M. Poirier, and others. J. Boulva
believed in the need to study postsmolt biology
and actively supported this study. G. Shooner,
G. Morin, and team assisted in scale reading and
in examining preserved postsmolts in 1982 and
1983. J. R. Robitaille assisted in scale data analy-
sis. Fishermen between Pointe-des-Monts and
Blanc Sablon kindly discussed the incidence of
postsmolts in their nets and participated in sam-
pling in 1982 and 1983. P. Bertrand of the Minis-
tere du Loisir de la chasse et de la Peche (Quebec)
shared his knowledge of the bycatch in herring
nets. Many thanks also to the population of Bale
Trinite. E. M. P. Chadwick, G. Power, and
R. L. Saunders kindly reviewed an earlier version
of the manuscript.
LITERATURE CITED
Allan. I R H . and J A Ritter.
1977. Salmonid terminology. J. Cons. int. Explor. Mer
37:293-299.
Allen. K. R.. R. L. Saunders, and P. F. Elson
1972. Marine growth of Atlantic salmon (Salmo salar) in
the Northwest Atlantic. J. Fish. Res. Board Can.
29:1373-1380.
Alm, G
1958. Seasonal fluctuations in the catches of salmon in
the Baltic. J. Cons. int. Explor. Mer 23:399-433.
Anonymous
1984. Report of the Baltic salmon and trout working
group. ICES CM. 1984/ Assess:ll.
Belding, D L., and G Prefontaine
1938. Studies on the Atlantic salmon. II-Report on the
salmon of the 1937 Port-aux-Basques (Newfoundland)
drift-net fishery. Contrib. Inst. Zool. Univ. Montreal 3,
58 p.
Carlin, B
1959. Results of salmon smolt tagging in the Baltic area.
Rapp. P.-v. R^un. Cons. int. Explor. Mer 147:89-96.
Carlin, B., and H.-E Lundin
1967. The size and position of the salmon caught in drift
nets in the Baltic. Swed. Salmon Res. Inst. Rep. 14, 6
P-
Caron. F.
1983. Migration vers I'Atlantique des post-saumoneaux
(Salmo salar) du Golfe du Saint-Laurent. Nat. can.
(Rev. Ecol. Syst.) 110:223-227.
1984. Rapport d'op6ration de la riviere Trinite, 1984.
Gk)uvemement du Qu6bec, Ministere du Loisir, de la
Chasse et de la Peche, Direction gen^rale de la faune, 26
P
Chadwick, E M. P
1981. Biological characteristics of Atlantic salmon smolts
in Western Arm Brook, Newfoundland. Can. Tech. Rep.
Fish. Aquat. Sci. 1024, iv + 45 p.
1982. Dynamics of an Atlantic salmon stock (Salmo
salar) in a small Newfoundland river. Ph.D Thesis,
Memorial Univ. Newfoundland, St. John's, 267 p.
Christensen, O.
1968. Report on an experimental fishery with salmon
drift-nets. ICES CM. 1968/B:11.
Christensen, O , and P.-O Larsson
1979. Review of Baltic salmon research. ICES Coop.
Res. Rep. 89, 124 p.
Christensen O., and W. H. Lear.
1980. Distribution and abundance of Atlantic salmon at
West Greenland. Rapp. P.-v. R6un. Cons. int. Explor.
Mer 176:22-35.
Comeau, N. a.
1909. Life and sport on the north shore of the lower St.
Lawrence and Gulf Telegraph Printing Company, Que-
bec.
COTfi, Y., I. Babos, and J A. Robitaille.
1984. Caract6ristiques scalim6triques des saumons du
Koksoak (Ungava, Quebec). Nat. can. (Rev. Ecol. Syst.)
111:401-409.
Elson, P F
1953. Growth and migration of PoUett River salmon. St.
Andrews Biol. Stn. Annu. Rep., p. 146-148.
1964. Post-smolt Atlantic salmon in the Bay of Fundy.
St. Andrews Biol. Stn. Annu. Rep., E20-E23.
Huntsman, A G.
1939. Conference on salmon problems. Publ. Am. Assoc.
Adv. Sci. 8:86-106.
Ikonen, E., and H Auvinen
1984. Migration of salmon in the Baltic Sea, based on
Finnish tagging experiments. ICES CM. 1984/M:4.
1985. Migration of salmon post-smolts (Salmo salar) in
the Baltic Sea. ICES CM. 1985/M:19.
Jessop. B M
1976. Distribution and timing of tag recoveries from native
and normative Atlantic salmon (Salmo salar) released
into Big Salmon River, New Brunswick. J. Fish. Res.
Board Can. 33:829-833.
210
DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON
JUTILA. E.. AND T ALAPASSI.
1985. Recaptures of salmon post-smolts (Salmo salar)
during the first summer after release in Finnish tagging
experiments. ICES CM. 1985/M:28.
JUTILA. E , AND J TOIVONEN
1985. Food composition of salmon post-smolts (Salmo
salar) in the northern part of the Gulf of Bothnia. ICES
CM. 1985/M:21.
Kendall. W C
1935. The fishes of New England, the salmon family.
Part 2. The salmons. Mem. Bos. Soc. Nat. Hist. 9:1-
166.
KULKA. D W , S Corey, and T D Iles
1982. Community structure and biomass of euphausiids
in the Bay of Fundy. Can. J. Fish. Aquat. Sci. 39:326-
334.
LaBar, G W , J D McCleave. and S M Fried.
1978. Seaward migration of hatchery-reared Atlantic
salmon iSalmo salar) smolts in the Penobscot River estu-
ary, Maine: open-water movements. J. Cons. int. Ex-
plor. Mer 38:257-269.
Larsson. P -O
1974. Migration of the Swedish west coast salmon stocks.
Swed. Salmon Res. Inst. Rep. 3, 11 p.
1985. Predation on migrating smolt as a regulating factor
in Baltic salmon, Salmo salar, populations. J. Fish Biol.
26:391-397.
Larsson, P -O., and S Ateshkar.
1979. Laxsmoltens vandring fraan Lulealven.
Fiskeritidskrift Finl. 1:8-9.
Lear, W H
1972. Food and feeding of Atlantic salmon in coastal areas
and over oceanic depths. ICNAF Res. Bull. 9:27-39.
1973. Size and age composition of the 1971
Newfoundland-Labrador commercial salmon catch.
Fish. Res. Board Can. Tech. Rep. 392, 43 p.
1980. Food of Atlantic salmon in the West Greenland-
Labrador Sea area. Rapp. P. -v. Reun. Cons. int. Explor.
Mer 176:55-59.
Lear, W H , and R K Misra
1978. Clinal variation in scale characters of Atlantic
salmon (Salmo salar) based on discriminant function
analysis. J. Fish. Res. Board Can. 35:43-47.
Mahnken, C, and T Joyner
1973. Salmon for New England fisheries. Ill: Developing
a coastal fishery for Pacific salmon. Mar. Fish. Rev.
35(10):9-13.
May, AW
1973. Distribution and migrations of salmon in the
Northwest Atlantic. Int. Atl. Salmon Found. Spec.
Publ. Ser. 4:373-382.
Meister. a L
1984. The marine migrations of tagged Atlantic salmon
(Salmo salar) of USA origin. ICES CM. 1984/M:27.
Munro. W R
1970. Notes on the salmon long-lining cruise by the R.V.
"Jens Chr. Svabo" off Faroe, April 1969. ICNAF Res.
Doc. 70/40, 8 p.
Murray, A R
1968. Smolt survival and adult utilisation of Little Cod-
roy River, Newfoundland, Atlantic salmon. J. Fish.
Res. Board Can. 25:2165-2218.
Peppar, J L
1982. Atlantic salmon smolt investigations, Restigouche
River system. New Brunswick. Can. MS Rep. Fish.
Aquat. Sci. 1641, vii + 15 p.
Power, G.
1969. Salmon of Ungava Bay. Arct. Inst. N. Am. Tech.
Pap. 22, 72 p.
1981. Stock characteristics and catches of Atlantic
salmon (Salmo salar) in Quebec, and Newfoundland, and
Labrador in relation to environmental variables. Can.
J. Fish. Aquat. Sci. 38:1601-1611.
Power, G., and G. Shooner.
1966. Juvenile salmon in the estuary and lower Nabisipi
River and some results of tagging. J. Fish. Res. Board
Can. 23:947-961.
Randall, R G , and G Power.
1979. Atlantic salmon (Salmo salar) of the Pigou and
Bouleau Rivers, Quebec. Environ. Biol. Fish. 4:179-
184.
Reddin, D. G.
1985. Atlantic salmon (Salmo salar) on and east of
the Grand Bank. J. Northwest Atl. Fish. Sci. 6:157-
164.
Reddin, D. G., and J. E Carscadden.
1981. Salmon-capelin interactions. Can. Atl. Fish. Sci.
Adv. Comm. Res. Doc. 81/2, 38 p.
RoBiTAiLLE, J A . Y. Cote, G. Hayeur, and G. Shooner.
1984a. Particularit^s de la reproduction du saumon atlan-
tique (Salmo salar) dans une partie du reseau Koksoak,
en Ungava. Rapp. tech. can. sci. halieut. aquat. 1313,
vii + 33 p.
1984b. Croisssance estuarienne du saumon atlantique
(Salmo salar) dans le fleuve Koksoak, en Ungava.
Rapp. tech. can. sci. halieut. aquat. 1314, vii + 23 p.
RUGGLES. C P., AND J. A RlTTER
1980. Review of North American smolt tagging to assess
the Atlantic salmon fishery off West Greenland. Rapp.
P.-v. Reun. Cons. int. Explor. Mer 176:82-92.
Saunders, R L.
1966. Some biological aspects of the Greenland salmon
fishery. Atl. Salmon J., Summer: 17-23.
1986. The thermal biology of Atlantic salmon: influence
of temperature on salmon culture with particular
reference to constraints imposed by low tempera-
ture. Inst. Freshwater Res. Drottningholm Rep. 63,
38 p.
Saunders. R L . B C Muise, and E. B. Henderson.
1975. Mortality of salmonids cultured at low temperature
in seawater. Aquaculture 5:243-252.
Scarnecchia. D L
1983. Age at sexual maturity in Icelandic stocks of At-
lantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci.
40:1456-1468.
1984. Climatic and oceanic variations affecting yield of
Icelandic stocks of Atlantic salmon (Salmo salar). Can.
J. Fish. Aquat. Sci. 41:917-935.
SCHIEFER, K
1972. Ecology of Atlantic salmon, with special reference
to occurrence and abundance of grilse, in north shore
Gulf of St. Lawrence rivers. Ph.D Thesis, Univ. Water-
loo, Ontario, 129 p.
SlEGEL, S
1956. Nonparametric statistics for the behavioral sci-
ences. McGraw-Hill Book Co., N.Y., 312 p.
Snedecor. G W . AND W G. Cochran
1967. Statistical methods. Iowa State Univ. Press,
Ames, 10, 593 p.
SOIKKELI, M.
1973. Tagged salmon smolts in the diet of the Caspian
tern. Laxforskningsinst. Medd. 3:1-5.
211
FISHERY BULLETIN: VOL. 86, NO. 2
SoKAL, R R , AND F J RoHLF. 5:62-85.
1969. Biometry. The principles and practice of statistics Thurow, F.
in biological research. Freeman and Co., San Francisco, 1968. On food, behavior and population mechanisms of
CA., 776 p. Baltic .salmon. Swed. Salmon Res. Inst. Rep. 4, 16 p.
Templeman, W. Valle, E
1967. Atlantic salmon from the Labrador Sea and off West 1985. Predation of birds on salmon and sea trout smolts
Greenland, taken during A. T. Cameron cruise, July- and post-smolts. ICES CM. 1985/M:22.
August 1965. ICNAF Res. Bull. 4:5-40. ViBERT, R
1968. Distribution and characteristics of Atlantic salmon 1953. Voyages maritimes des saumons et retour h la riv-
over oceanic depths and on the bank and shelf slope areas iere natale. Bull. Fr. Piscic. 170:5-23.
off Newfoundland, March-May 1966. ICNAF Res. Bull.
212
SIZE AND DIET OF JUVENILE PACIFIC SALMON
DURING SEAWARD MIGRATION THROUGH A SMALL ESTUARY
IN SOUTHEASTERN ALASKA
Michael L. Murphy, John F. Thedinga, and K V. Koski'
ABSTRACT
To assess competition and predation among juvenile Pacific salmon iOncorhynchus spp.) migrating
through the estuary of Porcupine Creek, a small stream in southeastern Alaska, their size and diet
were determined in 1979 and 1981. Mean fork length (FL) during May and June increased from 32
to 73 mm (1.5 mm/day) for pink salmon, O. gorbuscha; from 39 to 51 mm (0.4 mm/day) for chum
salmon, O. keta; and during June and July, from 99 to 165 mm (1.6 mm/day) for coho salmon, O.
kisutch. Prey, in order of importance, included larval fish (mostly Gadidae), larval molluscs (Mesogas-
trofKxla), and calanoid copepods for pink salmon; larval molluscs, larvaceans, and hyperiid amphipods
for chum salmon; and fish (Clupea harengus pallasi, Ammodytes hexapterus, and Gadidae), insects,
and larval decapods (Brachyrhyncha) for coho salmon. No pink or chum salmon were found in the coho
salmon stomachs. Prey size for pink and chum salmon was similar (median, 0.4 mm long for both
species), and much smaller than that of coho salmon (median, 2.3 mm). Diet overlap was greater
between pink and chum salmon than between either species and coho salmon. Pink salmon, however,
ate almost exclusively (95%) pelagic prey, whereas chum salmon ate both pelagic (74%) and epiben-
thic (26%) prey. Rapid ecirly growth and differences in diet probably help minimize predation and
competition among salmon during seaward migration.
The early marine life stage of juvenile Pacific
salmon iOncorhynchus spp.), during transition
from freshwater to seawater, is important in de-
termining brood-year survival and subsequent
adult returns (Manzer and Shepard 1962; Parker
1968); their survival rate is lowest during this
time (Parker 1968; Bax 1983). Salmon often
school in large concentrations in estuaries as they
migrate seaward, and are more likely to deplete
food supplies and compete for food than after they
disperse to the sea (Bailey et al. 1975; Feller and
Kaczynski 1975). Survival depends on size
(Parker 1971; Healey 1982), and competition for
food can depress early growth (Peterman 1984)
and prolong vulnerability to predators (Taylor
1977; Walters et al. 1978). Size and diet of juve-
nile salmon in an estuary, therefore, determine
the potential for predation and competition and
can greatly affect survival.
As salmon aquaculture expands and more juve-
nile salmon are released into estuaries, competi-
tion and predation among salmon may increase
(Johnson 1974). To optimize hatchery production
and avoid adversely affecting wild stocks, an
1 Northwest and Alaska Fisheries Center Auke Bay Labora-
tory, National Marine Fisheries Service, NOAA, P.O. Box
210155, Auke Bay, AK 99821.
Manuscript accepted November 1987.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.
understanding is needed of how different stocks of
salmon interact in estuaries. This paper compares
size and diet of juvenile pink, O. gorbuscha;
chum, O. keta; and coho, O. kisutch, salmon
to assess potential predation and competition
between the species during their seaward mi-
gration through the estuary of a small, pristine
stream.
STUDY AREA
This study was conducted in the estuary of Por-
cupine Creek, the only salmon stream flowing
into Steamer Bay in southeastern Alaska (Fig. 1).
The estuary (about 5.5 km long) consists of a 1.5
km stream reach that is periodically inundated
by tides, and a 4 km series of three estuarine
basins. At low tide, the inner and middle basins
are small (2 and 7 ha, respectively) and shallow
(14 and 16 m, respectively) compared with the
outer basin (120 ha and 42 m deep). The littoral
zone ranges from low-gradient mudflats to steep
cobble. Bottoms of the basins are level and com-
posed of shell, gravel, and mud.
During low tide, the inner and middle basins
are partially isolated from the outer basin and the
main part of Steamer Bay by tidal rapids 1-3 m
deep. Salinity is lower in the inner and middle
213
FISHERY BULLETIN: VOL. 86, NO. 2
Figure 1. — Aerial photo of study site in the inner part of Steamer Bay, southeastern Alaska,
showing the Porcupine Creek estuary at low tide and location of smolt traps used by Thedinga
(1985).
basins (24-29%c) than in the outer basin (28-
30%c), but temperature does not differ between
basins in spring and summer (11°-13°C from May
to September 1981). Heavy tidal flushing, partic-
ularly during spring tides, results in a diverse
community within the estuary; e.g., eel grass,
Zostera; Dungeness crabs, Cancer magister; bull
kelp, Nereocystis; and rock scallop, Hinnetes. A
detailed description of the study area is in Merrell
and Koski (1978) and Koski (1984).
Porcupine Creek, upstream of tidal influence, is
4.5 km long and has an average discharge of
214
MURPHY ET AL.: SIZE AND DIET OF JUVENILE PACIFIC SALMON
about 0.5 m"^/second. Its watershed is forested by
mature western hemlock, Tsuga heterophylla ,
and Sitka spruce, Picea sitchensis. Annually,
5,000-75,000 adult pink salmon and 200-4,000
chum salmon spawn in the creek from late July to
October, and 250-600 adult coho salmon spawn
from late September to November (Koski 1984).
Pink and chum salmon fry typically migrate from
Porcupine Creek from late March to mid-May
(Koski^). Coho salmon smolts migrate from late
April to early June, but over 90% usually migrate
in late May (Thedinga 1985).
METHODS
Six stations, one each on the east and west sides
of the three basins (Fig. 1) were sampled by a
beach seine 37 m long, with 1.6 cm stretch mesh
on the wings, and a central bag of 6 mm stretch
mesh. The seine tapered from 2 m deep at the
central bag to 1 m deep at each end. In 1979, only
one station in each basin was seined about every
4 days from 16 May to 12 June. In 1981, all six
stations were seined biweekly from 26 May to
7 July and monthly thereafter through 11
November. Seines were set parallel to and about
40 m from shore by a skiff, and retrieved from
shore. Setting and retrieval were accomplished
within 10 minutes.
All fish caught were identified and counted.
Fork lengths (FL) were measured to the nearest
millimeter from a random sample of <25 salmon
per species, station, and sampling period. Stom-
ach contents were collected only in 1981 from <10
salmon per species and station in May, June, and
July. Contents were collected from anesthetized
fish by flushing the stomach with water from a
syringe (Meehan and Miller 1978; Koski and
Kirchhofer 1984) and preserved in 5% formalde-
hyde. Prey were later identified, counted, mea-
sured, and weighed.
For diet analysis, the index of relative impor-
tance (IRI) was calculated, where
IRI = {% number + % weight) (% frequency of
occurrence)
(Pinkas et al. 1971). Diet overlap between salmon
species was calculated (McCabe et al. 1983):
2 2 X'Y,
C =
1 = 1
2xf . J^Yf
1=1
1 = 1
2K V. Koski, Northwest and Alaska Fisheries Center Auke
Bay Laboratory, National Marine Fisheries Service, NOAA,
P.O. Box 210155, Auke Bay, AK 99821, pers. commun. October
1985.
where C = overlap coefficient and X, and Y, are
proportions of the total diet of salmon species X
and Y, respectively, contributed by prey taxon i of
s prey taxa. Diet overlap was calculated sepa-
rately for proportions based on prey number and
weight. Prey were also classified as epibenthic or
pelagic to assess overlap in foraging mode (Feller
and Kaczynski 1975). Epibenthic prey were poly-
chaetes, gammarid amphipods, harpacticoid cope-
pods, barnacle cyprids, and cumaceans. Pelagic
prey were calanoid copepods, euphausiids, barna-
cle nauplii, cladocerans, larvaceans, larval deca-
pods (Brachyrhyncha), hyperiid amphipods, and
fish (eggs, larvae, and juveniles).
RESULTS
Size
In May 1979, pink salmon were the size of
newly emergent fry, about 32 mm FL (Fig. 2).
Average length increased 1.5 mm/day, to 73 mm
on 12 June 1979. In 1981, pink salmon averaged
73 mm FL in late May and early June. Changes
in average FL in 1981 could not be calculated
because most migration occurred before sampling
began.
Average FL of chum salmon increased slower
than that of pink salmon. Mean FL of chum
salmon increased 0.4 mm/day in both years, from
39 mm to 51 mm in 1979, and from 60 mm to 78
mm in 1981 (Fig. 2). Chum salmon averaged
about 10 mm FL longer in 1981 than in the same
period in 1979. Chum salmon were not found in
the estuary after early July, except for two fry
caught in the outer basin in November.
Average FL of coho salmon was nearly con-
stant, between 85 and 110 mm, throughout May
and early June in both 1979 and 1981 (Fig. 2).
Average FL of coho salmon in the estuary during
this period was influenced by an influx of Porcu-
pine Creek migrants, which averaged between 75
and 96 mm FL (Thedinga 1985). After the migra-
tion from Porcupine Creek in 1981 (=9 June),
average FL increased 1.6 mm/day to 165 mm by
20 July. Average FL then decreased to 85 mm in
215
FISHERY BULLETIN: VOL. 86, NO. 2
uu
80
i>
60
40
•
/
/
/
- /
(a) pink
20
4
-
'l 1 1
1 1 1
E 80
E
O 60
u
-» 50
Z
O 40
< 30
0)
ni
._^k, ^^ (^ CHUM
220
180
140
100
60
(c) COHO
3-^-tH^4
>-
4—.^
/
/
"n 1 1 1 1 r
15 MAY 1 JUN 15 JUN 1 JUL 15 JUL 1 AUG
Figure 2.— Length of salmon in 1979 and 1981. Data shown are
means and ranges for pooled samples from all stations on each
sampling date. Data for pink salmon in 1981 are omitted be-
cause of small sample sizes. (▲ = 1979, • = 1981.)
August and 106 mm in September, after most
smelts had left and a few new smelts entered the
estuary.
Diet
A wide variety ef prey was eaten by the three
salmon species, but usually only one or two prey
taxa dominated the diet (Table 1). Pink salmon
ate mostly larval molluscs (Mesogastropoda) and
larval fish (mostly Gadidae) in May, and calanoid
copepods in June. Chum salmon ate mostly larval
molluscs in May; larval molluscs, larvaceans, and
cladocerans in June; and hyperiid amphipods and
larval decapods in July. Coho salmon ate mostly
fish and insects in May and June, and fish and
larval decapods in July. The identifiable fish prey
of coho salmon consisted of 53% Pacific herring,
Clupea harengus pallasi; 45% cod (Gadidae); and
2% Pacific sand lance, Ammodytes hexapterus. No
identifiable pink or chum salmon were in the coho
stomachs. Catch of coho, but not that of the other
salmon, was significantly correlated (r = 0.46,
P < 0.001) with aggregate catch of herring, sand
lance, and cod, indicating that coho salmon con-
gregated near prey schools.
Diet overlap was higher between pink and
chum salmon than between either species and
coho salmon (Table 2). Diet overlap between pink
and chum salmon was especially high in May
when both species ate large numbers of larval
molluscs. If based on prey weight, diet overlap
between pink and coho salmon was negligible. If
based on prey number, however, overlap was
>50% in June when both pink and coho salmon
ate large numbers of calanoid copepods. Diet
overlap between chum and coho salmon was con-
sistently low, especially when based on prey
weight.
Of the 12 most important prey taxa in May and
June, when all 3 salmon species were present in
the estuary, only 4 differed significantly
iP < 0.05) in mean number per stomach between
pink and chum salmon, whereas 9-10 differed
significantly between the two species and coho
salmon (Table 3). Compared with pink salmon,
chum salmon ate more harpacticoid copepods,
cladocerans, and insects. Coho salmon ate fewer
small plankton and more fish than did the other
salmon species. Coho salmon averaged fewer than
20 total prey items, compared to more than 100 in
pink salmon and 200 in chum salmon.
Coho salmon ate larger prey than did the other
salmon (Fig. 3). Median prey length for coho
salmon was 2.3 mm, compared with 0.4 mm for
pink and chum salmon. Coho salmon generally
selected larger individuals of each prey taxon —
particularly larger calanoid copepods, gammarid
amphipods, euphausiids, and larval decapods —
than did pink and chum salmon (Table 4). Offish
prey, coho salmon ate mostly juveniles, whereas
pink and chum salmon ate mostly eggs and lar-
vae.
As they grew larger, all three salmon species
selected larger prey. Numbers of hyperiid am-
phipods, euphausiids, and fish larvae — all rela-
tively large prey — were positively correlated
with FL of pink or chum salmon, whereas num-
bers of cladocerans and larvaceans — both rela-
tively small prey — were negatively correlated
with chum salmon FL (Table 5). Numbers of
calanoid copepods and fish were positively corre-
lated with coho salmon FL, whereas the number
216
MURPHY ET AL.: SIZE AND DIET OF JUVENILE PACIFIC SALMON
Table 1 . — Stomach contents of juvenile salmon in Porcupine Creek estuary, 26 May-7 July 1981 . %A/ is percent by
number, %W is percent by wet weight, %F0 is percent frequency of occurrence of fish with prey item /, and %IRI is
percent of total sum of IRI for all prey taxa. IRI = (%/V + %W)%fO. Taxa are omitted if %IRI is s3 for all salmon
species.
Pink salmon
Chum salmon
Coho salmon
Prey taxon
%N
%W
%F0
%IRI
%/V
%W
%F0
%IRI
%/V
%W
%F0
%IRI
26-29 May 1981
Mollusc larvae
47
2
71
28
59
36
100
60
2
0
4
0
Calanoida
12
2
57
7
5
4
71
5
0
0
4
0
Harpacticoida
1
0
29
0
11
9
90
12
17
1
18
3
Cladocera
1
0
43
1
14
5
77
10
0
0
4
0
Cumacea
0
0
0
0
2
9
32
3
16
1
41
6
Euphausiacea
11
2
71
8
0
2
42
1
0
0
0
0
Decapod larvae
Shhmp
9
5
71
8
0
1
13
0
0
0
0
0
Crabs
8
2
60
4
3
3
60
2
2
0
12
0
Fish eggs
and larvae
8
86
57
44
0
21
13
2
1
6
4
0
Fish
0
0
0
0
0
0
0
0
22
62
48
38
Insects!
1
0
29
0
1
5
52
2
29
23
100
47
Total
98
99
100
95
95
97
89
93
94
(Number of stomachs)
(7)
(31)
(27)
9-10 June 1981
Mollusc larvae
9
5
100
8
24
16
73
23
0
0
7
0
Calanoida
47
52
100
58
5
5
65
5
22
0
43
9
Harpacticoida
3
5
25
1
4
4
65
4
0
0
7
0
Cladocera
13
7
75
9
19
8
81
17
1
0
23
0
Decapod larvae
Shhmp
1
1
75
1
0
1
20
0
0
0
0
0
Crabs
4
11
100
5
7
12
100
8
19
0
21
4
Larvacea,
Oikopleura
9
5
100
8
20
13
85
23
7
0
14
1
Fish eggs
and larvae
6
5
100
6
11
12
35
6
0
0
0
0
Fish juveniles
0
0
0
0
0
0
0
0
10
94
71
70
Insects^
0
1
50
0
3
10
75
8
28
2
50
14
Total
92
92
96
93
91
94
87
96
98
(Number of stomachs)
(4)
(26)
(14)
7 July 1981
Calanoida
—
—
—
—
4
1
43
3
16
0
24
5
Hypehidea
—
—
—
—
12
30
88
45
4
0
29
2
Decapod larvae
Shrimp
—
—
—
—
61
40
14
18
1
0
6
0
Crabs
—
—
—
—
3
10
86
9
55
5
53
25
Fish eggs
and larvae
—
—
—
—
0
3
57
2
0
0
0
0
Fish juveniles
—
—
—
—
0
0
0
0
9
94
50
65
Insects^
—
—
—
—
6
4
100
14
5
0
24
2
Total
—
—
—
—
86
88
91
90
99
99
(Number of stomachs)
(0)
(7)
(17)
1 Mostly adult Diptera.
of insects was negatively correlated. As a conse-
quence of the selection of larger prey as the
salmon grew, total prey weight increased with
salmon FL, whereas total prey number did not
(Table 5). Although pink and chum salmon ate
prey of similar size, they foraged differently (Fig.
4). Pink salmon consumed about 95% pelagic
prey; chum salmon, only 74%. Individual taxa
changed, but the importance of pelagic prey did
not change significantly between sampling peri-
ods, estuary basins, or salmon FL classes.
Diet of coho salmon, on the other hand, varied
widely depending on salmon FL, date, and loca-
tion (Fig. 4). Pelagic prey increased from 1% of
total prey for coho salmon <80 mm FL to 80% for
those >100 mm FL. Coho salmon ate fewer
pelagic prey in May, when most coho were in the
inner basin and feeding mainly on insects, than
in July when most were in the outer basin and
feeding mainly on fish (Table 1). Analysis of vari-
ance, however, showed that differences between
basins and sampling periods were not significant
217
FISHERY BULLETIN: VOL. 86, NO. 2
Table 2. — Diet overlap (McCabe et al. 1983) based on number (n)
and weight (W) of prey by sampling period for juvenile salmon in
Porcupine Creek estuary, 26 May-7 July 1981. The number of
stomach samples is in Table 1 .
Sampling
period
Pink vs
n
chum
W
Pink vs
n
coho
W
Chum vs. coho
n W
26-29 (vlay
9-10 June
7 July
0.87
0.47
0.40
0.38
0.07
0.54
0.09
0.00
0.13 0.08
0.29 0.00
0.10 0.01
Table 3. — Comparison of mean number of the 12 most important
prey and total of all prey per salmon stomach from Porcupine Creek
estuary, 26 N^ay- 1 0 June 1 98 1 . Means followed by the same letter
are not significantly different (Kruskal-Wallis analysis of variance,
P>0.05) compared within a row and across columns. The number
of stomach samples is in Table 1 .
Prey item
Pink salmon Chum salmon Coho salmon
Mollusc larvae
28 a
Barnacle larvae
4a
Calanoida
40 a
Harpacticoida
2a
Cladocera
10a
Cumacea
Oa
Euphausiacea
5a
Decapod larvae
11 a
Larvacea
6a
Fish eggs and
larvae
3a
Fish
0 a
Insects
Oa
86 a
5a
12 b
19 b
37 b
5a
1 a
12a
20 a
11 a
Oa
4b
Total
115a
216 a
0 b
4 b
6 b
19 b
Table 4. — Mean length (mm) of prey items in salmon stom-
achs from Porcupine Creek estuary, 26 May-7 July 1981.
Mean prey length within prey taxa was significantly greater
for coho salmon than for pink or chum salmon (sign test,
n = 12 means, P=0.03 and P =0.003, for coho salmon vs.
pink and chum salmon, respectively). The number of stom-
ach samples is in Table 1 .
Pink
Chum
Coho
Prey item
salmon
salmon
salmon
Mollusc larvae
0.5
0.5
0.4
Barnacle larvae
0.4
0.7
0.4
Calanoida
1.2
1.2
4.8
Harpacticoida
1.3
1.1
1.3
Cladocera
0.6
0.6
0.7
Cumacea
2.4
2.5
2.8
Hyperiidea
3.3
2.4
3.0
Gammaridea
1.5
1.8
5.2
Euphausiacea
3.5
3.0
18.0
Decapod larvae
2.8
2.0
3.5
Larvacea
0.8
0.9
1
Fish, all life
stages
2.0
1.2
23.9
Insects
2.8
2.2
3.5
T — I — I — I — I I I "1 — I — *T — I — T — I — I — r
0.5 2.5 4.5 6.5 8.5 10.5 12.5 >1 4
Prey Length (Interval Midpoint, mm)
Figure 3. — Relative frequencies of length of prey eaten by pink,
chum, and coho salmon in the Porcupine Creek estuary in 1981.
Total prey measured were 687 in 11 pink, 5,634 in 63 chum, and
1,179 in 53 coho salmon.
Table 5. — Spearman rank correlations between number of prey
items and fork length of juvenile salmon. Because of the large
number of correlations tested, significance levels were adjusted by
multiplying the probability P by the number of tests for each salmon
species. The number of stomach samples is in Table 1 .
Prey item
Pink salmon Chum salmon Coho salmon
Mollusc larvae
-0.32
Barnacle larvae
-0.36
Calanoida
-0.51
Harpacticoida
-0.30
Cladocera
-0.06
Cumacea
-0.20
Hyperiidea
-0.35
Euphausiacea
0.84
Decapod larvae
0.43
Larvacea
-0.36
Fish eggs
-0.33
Fish larvae
and juveniles
0.86
Insects
0.61
Total prey
number
-0.40
Total prey
weight
0.85
0.05
-0.02
0.10
-0.03
-0.34*
0.04
-0.37-
0.36*
-0.19
-0.50*
-0.18
-0.04
-0.11
-0.04
0.35*
0.18
0.22
0.33*
0.19
0.04
0.28
0.19
0.15
0.25
1
1
0.36*
0.71"
0.09
0.66**
'None present.
'None present in any stomachs.
*Adjusted probability P < 0.05.
**Adjusted probability P < 0.01.
after adjusting for differences in coho salmon FL
(Table 6). Thus, changes in diet were mainly re-
lated to increasing size of coho salmon as they
migrated through the estuary.
DISCUSSION
Both size and diet can affect predation and com-
petition among juvenile salmon in an estuary. A
218
MURPHY ET AL.: SIZE AND DIET OF JUVENILE PACIFIC SALMON
100
80
60-
40-
20
-A)
. Pink
' Chum
Coho
^100^
Q.
O
■5. 80
«
<D
Q.
^ 60
c
0)
(J
« 40-
Q.
20
26May
9June
7July
B
I
100-f
80
60
40
20
Inner Basin
Middle Basin
Outer Basin
O-l
/
/
/
/
^<
/
50-59 60-79 80-99
Salmon Length (mm)
> 100
Figure 4. — Number of pelagic prey as percent of total prey
eaten by individual salmon compared between sampling periods
(A), estuary basins IB), and salmon fork length classes (C) in the
Porcupine Creek estuary in 1981. Symbols are means; bars are
±2 SE of the means. Symbols in B and C are the same as in A .
Pelagic prey are defined in the text.
salmon's size mainly influences its vulnerability
to predators, whereas its diet determines poten-
tial competition for food. Size and diet, however,
are not independent. Salmon change their diet as
they grow, which helps relieve competition be-
tween salmon of different size, and a poor diet
slows their growth, which prolongs vulnerability
to predation.
Table 6. — Analysis of variance of percentage pelagic
prey of coho salmon, with sampling period and estuary
basin as factors and salmon fork length as covariate.
Factors, covariate, and interactions were adjusted
simultaneously before assessing significant (Kim and
Kohout 1975).
Source
of variation
df
Mean
square
F
P
Length
Sampling period
Basin
Residual
Total
1
2
2
47
52
8,458
1,580
654
900
1,519
9.4
1.8
0.7
0.004
0.184
0.489
Because of similar diets, pink and chum salmon
are potential competitors. Although diets of pink
and chum salmon in the Porcupine Creek estuary
were similar in prey size and some prey taxa,
however, pink salmon fed almost solely on pelagic
prey , whereas chum salmon foraged both pelagi-
cally and epibenthically; such differences may
help reduce competition. Competition probably
also was reduced because, as the salmon grew
larger, they switched to larger prey. Coho salmon
probably did not compete for food with the other
two species because the coho fed on larger, differ-
ent prey.
Rapid early growth of salmon is important in
reducing vulnerability to predators (Parker 1971;
Taylor 1977). For example, hatchery pink salmon
fry raised for 60 days (to 40 mm FL) before release
into an estuary in southeastern Alaska survived
68% better at sea than did fry released immedi-
ately after emergence (Martin et al. 1981).
Marine survival also is higher for year classes of
larger (9-11 cm FL) than for smaller (6-8 cm FL)
sockeye salmon, O. nerka , smolts (Foerster 1954).
Coho salmon smolts from Porcupine Creek in
1978 averaged 99 mm FL and their survival was
twice that of the 1979 smolts, which averaged
only 91 mm FL (Thedinga 1985). Smolt size and
migration timing, however, interact complexly to
influence marine survival of coho salmon (Bilton
1978).
Growth of juvenile salmon in estuaries usually
inferred from changes in mean size
IS
(LeBrasseur and Parker 1964; Healey 1978), but
these estimates are subject to bias. In this study,
changes in mean size of fish in the catches on
successive dates could underestimate real growth
for two reasons: 1) small individuals may have
migrated continuously into the estuary from
219
FISHERY BULLETIN: VOL. 86, NO. 2
freshwater and 2) larger individuals may have
migrated continuously from the estuary to the
sea. Conversely, growth could be overestimated if
predators of salmon selected small individuals
(Parker 1971). In addition, although the inner
and middle basins are semiisolated from adjacent
marine waters during low tide, juvenile salmon
from adjacent waters could easily enter the estu-
ary, especially the outer basin, during flood tide
and mix with salmon from Porcupine Creek.
Estimates of salmon growth in estuaries and
nearshore marine waters are variable, but gener-
ally range between 1 and 2 mm/day. LeBrasseur
and Parker (1964) estimated pink salmon growth
to be 0.9 mm/day during the first 30 days at sea,
and Healey (1978) estimated pink salmon growth
during summer to be 1.0 mm/day; our estimate
was 1.5 mm/day. Our estimate for chum salmon
at 0.4 mm/day was considerably less than that of
Healey (1978) at 1.5 mm/day, also based on
change in mean length; however, our estimate for
coho salmon of 1.6 mm/day was similar to that of
Healey (1978) at 1.2 mm/day. Summer growth
back-calculated from scales of salmon from the
Sea of Okhotsk was about 1.6 mm/day for pink
and chum salmon (Birman 1969).
Because of their initial small size, pink and
chum salmon particularly are vulnerable to
predators including juvenile coho salmon (Parker
1971). Several authors have suggested that a
major share of pink salmon mortality in the first
weeks at sea results from juvenile coho salmon
predation (Parker 1971; Kaczynski et al. 1973;
Hargreaves and LeBrasseur 1985), but such pre-
dation has not been found in field collections.
Parker (1971) demonstrated predation by juve-
nile coho salmon on pink salmon fry in the labora-
tory, and juvenile coho salmon are known preda-
tors of salmon fi:y in fi*eshwater (Hunter 1959;
Koski and Kirchhofer 1984). However, we have
not found any published data that show predation
by juvenile coho salmon on other salmon in estu-
aries or marine waters. Predation by juvenile
coho salmon on pink salmon fiy migrating ft-om
freshwater does occur in the tidal-influenced
reach of Porcupine Creek (Koski and Kirchhofer
1984), but such predation apparently does not ex-
tend into the estuarine basins. Many fishes have
been identified as predators of pink and chum
salmon in estuaries, including Pacific herring
(Thorsteinson 1960), sea-run cutthroat trout,
Salmo clarki; cod, Gadus macrocephalus ; and
sculpin, Leptocottus armatus, (Bax et al. 1977).
We speculate that predation by coho salmon on
salmon fiy may occur only under circumstances
in which the coho salmon are combined with
small fry as they migrate from freshwater.
The period of vulnerability of pink and chum
salmon fry to predation by juvenile coho salmon is
probably relatively short. Within the first 3
weeks after entering the estuary, pink salmon fiy
can grow larger than the prey fish of juvenile coho
salmon. In the laboratory, juvenile coho salmon
ate the smallest pink salmon available and did
not eat any larger than about 50 mm FL (Parker
1971), which coincides with the largest fish eaten
by coho salmon in our study. At a growth of 1
mm/day, pink salmon entering the estuary at 32
mm FL will outgrow predation by coho salmon
smolts in 18 days. In Porcupine Creek, most pink
and chum salmon migrated fi-om the stream sev-
eral weeks before coho salmon, which enables
them to grow large enough to avoid predation by
coho salmon in the estuary. Thus, early migration
and rapid growth of pink and chum salmon fiy
probably are important in reducing predation by
coho salmon.
In the Porcupine Creek estuary, competition
and predation probably were slight. Competition
for food was minimal, as evidenced by the rapid
salmon growth, because of differences in prey and
foraging mode and because regular tidal flushing
probably replenished food supplies, as in Traitors
Cove, AK (Bailey et al. 1975). Natural stocking
levels in the estuary also probably were below
thresholds where competition for food would de-
press survival. Predation by coho salmon on pink
and chum salmon was avoided because the pink
and chum salmon migrated earlier than coho
salmon and rapidly grew too large for the
coho to handle. Thus, in this natural system, com-
petition and predation probably were unimpor-
tant because of moderate stocking levels, rapid
growth, and differences in diet and timing of mi-
grations. In systems with hatchery inputs, how-
ever, stocking levels would probably be higher
and salmon size and timing of migrations differ-
ent than in natural systems, which could increase
competition and predation.
Stocking levels and timing of hatchery releases
of juvenile salmon in estuaries are important in
minimizing competition and predation (Myers
1980). Hatchery releases should avoid combining
large concentrations of pink and chum salmon fi-y
so as not to deplete food supplies. Conversely, re-
leases during low predator abundance and good
growing conditions — high food availability and
warm temperature — could increase grov^rth and
220
MURPHY ET AL.: SIZE AND DIET OF JUVENILE PACIFIC SALMON
survival. Early releases of coho salmon could in-
crease predation on pink and chum fry (Johnson
1974), especially fry <50 mm FL, if the releases
coincide with fry migrations through the estuary.
ACKNOWLEDGMENTS
Our thanks to R. Brodeur, G. Grette,
C. Hawkes, D. Kirchhofer, and R. Walter for help
in study design and field sampling. Thanks also
to J. Hard, J. Landingham, T. Merrell, Jr.,
D. Mortensen, and J. Pella for reviewing the
manuscript. The Fisheries Research Institute of
the University of Washington identified and
measured prey from stomach samples.
LITERATURE CITED
Baily. J. E., B. L. Wing, and C R. Mattson
1975. Zooplankton abundance and feeding habits of fry of
pink salmon, Oncorhynchus gorbuscha , and chum
salmon, Oncorhynchus keta, in Traitors Cove, Alaska,
with speculations on the carrying capacity of the
area. Fish. Bull., U.S. 73:846-861.
Bax, N J.
1983. Early marine mortality of marked juvenile chum
salmon (Oncorhynchus keta) released into Hood Canal,
Puget Sound, Washington, in 1980. Can. J. Fish.
Aquat. Sci. 40:426-435.
Bax, N J , E O. Salo, B. P Snyder, C. A Simenstad, and W. J.
Kinney
1977. Salmon outmigration studies in Hood Canal: a sum-
mary—1977. In W. J. McNeil and D. C. Himsworth (ed-
itors), Salmonid ecosystems of the North Pacific, p. 171-
201. Oregon State Univ. Press, Oregon State Univ. Sea
Grant Program, Corvallis.
BiLTON, H. T
1978. Returns of adult coho salmon in relation to mean
size and time at release of juveniles. Fish. Mar. Serv.
Tech. Rep. 832, 73 p. Dep. Fish. Environ., Pac. Biol.
Stn., Nanaimo, B.C.
BiRMAN, I B
1969. Distribution and growth of young Pacific salmon of
the genus Oncoryhynchus in the sea. Probl. Ichthyol.
9:651-666.
Feller, R. J., and V. W. Kaczynski.
1975. Size selective predation by juvenile chum salmon
(Oncorhynchus keta) on epibenthic prey in Puget
Sound. J. Fish. Res. Board Can. 32:1419-1429.
FOERSTER, R. E.
1954. On the relation of adult sockeye salmon
{Oncorhynchus nerka) returns to known smolt seawEird
migrations. J. Fish. Res. Board Can. 11:339-350.
Hargreaves, N B., and R J LeBasseur
1985. Species selective predation on juvenile pink
(Oncorhynchus gorbuscha ) and chum salmon (O. keta ) by
coho salmon (O. kisutch). Can. J. Fish. Aquat. Sci.
42:659-668.
Healey, M. C
1978. The distribution, abundance, and feeding habits of
juvenile Pacific salmon in Georgia Strait, British Colum-
bia. Fish. Mar. Serv. Tech. Rep. 788, 49 p. Dep. Fish.
Environ., Pac. Biol. Stn., Nanaimo, B.C.
1982. Timing and relative intensity of size-selective
mortality of juvenile chum salmon (Oncorhynchus keta)
during early sea life. Can. J. Fish. Aquat. Sci. 39:952-
957.
Hunter, J G.
1959. Survival and production of pink and chum salmon
in a coastal stream. J. Fish. Res. Board Can. 16:835-
886.
Johnson, R. C.
1974. Effects of hatchery coho on native Puget Sound
stocks of chum salmon fry. In D. R. Harding (editor).
Proceedings of the 1974 Northeast Pacific Pink and
Chum Salmon Workshop, p. 102-109. Dep. Environ.,
Fish. Vancouver, B.C.
PCaczynski, V W , R. J Feller, J. Clayton, and R. J. Gerke.
1973. Trophic analysis of juvenile pink and chum salmon
(Oncorhynchus gorbuscha and O. keta) in Puget
Sound. J. Fish. Res. Board Can. 30:1003-1008.
Kim, J O , AND F J. Kohout.
1975. Analysis of variance and covariance: subprogram
ANOVA. 2d ed. In N. H. Nie, C. H. Hull, J. G.
Jenkins, K. Steinbrenner, and D. H. Brent (editors).
Statistical package for the social sciences, p. 398-
433. McGraw-Hill, N.Y.
KOSKI, K V
1984. A stream ecosystem in an old-growth forest in
southeast Alaska: Part I. Description and characteristics
of Porcupine Creek, Etolin Island. In W. R. Meehan,
T. R. Merrell, Jr., and T. A. Hanley (editors), Fish and
wildlife relationships in old-growth forests. Proceedings,
p. 47-55. Am. Inst. Fish. Res. Biol., Juneau, AK.
Available from J. W. Reintjes, Rt. 4, Box 85, Morehead
City, NC 28557.
KosKi, K V , AND D. A. Kirchhofer.
1984. A stream ecosystem in an old-growth forest in
southeast Alaska. Part FV: Food of juvenile coho salmon,
Oncorhynchus kisutch in relation to abundance of drift
and benthos. In W. R. Meehan, T. R. Merrell, Jr., and
T. A. Hanley (editors). Fish and wildlife relationships
in old-growth forests, Proceedings, p. 81-87. Am.
Inst. Fish. Res. Biol., Juneau, AK. Available from
J. W. Reintjes, Rt. 4, Box 85, Morehead City, NC
28557.
LeBrasseur, R J , AND R R Parker
1964. Growth rate of central British Columbia pink
salmon (Oncorhynchus gorbuscha ). J. Fish. Res. Board
Can. 21:1101-1128.
Manzer, J I., AND M. P. Shepard
1962. Marine survival, distribution and migration of pink
salmon (Oncorhynchus gorbuscha) off the British Colum-
bia coast. H. R. MacMillan Lectures in Fisheries, p.
113-122. Symposium on Pink Salmon 1960, Univ.
British Columbia, Vancouver.
Martin, R. M , W. R. Heard, and A. C. Wertheimer.
1981. Short-term rearing of pink salmon (Oncorhynchus
gorbuscha ) fry: effect on survival and biomass of return-
ing adults. Can. J. Fish. Aquat. Sci. 38:554-558.
McCabe, G. T , JR , W D MUIR, R L. Emmett, and J. T. DURKIN.
1983. Interrelationships between juvenile salmonids and
nonsalmonid fish in the Columbia River estuary. Fish.
Bull., U.S. 81:815-826.
Meehan, W R., and R A Miller.
1978. Stomach flushing: effectiveness and influence on
survival £ind condition of juvenile salmonids. J. Fish.
Res. Board Can. 35:1359-1363.
221
FISHERY BULLETIN: VOL. 86, NO. 2
Merrell, T. R.. Jr., and K V. Koski.
1978. Habitat values of coastal wetlands for Pacific coast
salmonids. In P. E. Greeson, J. R. Clark, and J. E. Clark
(editors). Wetland functions and values: the state of our
understanding, p. 256-266. Proceedings of the Na-
tional Symposium on Wetlands, American Water Re-
source Association, Minneapolis, MN.
Myers, K W. W.
1980. An investigation of the utilization of four study
areas in Yaquina Bay, Oregon, by hatchery and wild
juvenile salmonids. M.S. Thesis, Oregon State Univer-
sity, Corvallis, 234 p.
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.
1971. Size selective predation among juvenile salmonid
fishes in a British Columbia inlet. J. Fish. Res. Board
Can. 28:1503-1510.
Peterman. R M.
1984. Density-dependent growth in early ocean life of
sockeye salmon iOncorhynchus nerka). Can. J. Fish.
Aquat. Sci. 41:1825-1829.
PiNKAS, L., M S OLIPHANT, and I. L. K. IVERSON.
1971. Food habits of albacore, bluefin tuna, and bonito in
California waters. Calif Dep. Fish Game, Bull. 12,
105 p.
Taylor, S. G.
1977. The effect of timing of downstream migration on
meirine survival of pink salmon iOncorhynchus gor-
buscha). M.S. Thesis, Univ. Alaska, Southeastern
Senior College, Juneau, 40 p.
Thedinga, J F.
1985. Smolt scale characteristics and yield of coho
salmon, Oncorhynchus kisutch, smolts and adults from
Porcupine Creek, southeastern Alaska. M.S. Thesis,
Univ. Alaska, Juneau, 93 p.
Thorsteinson, F. V
1960. Herring predation on pink salmon fry in a south-
eastern Alaska estuary. Trans. Am. Fish. Sec. 91:321-
323.
Walters, C. J., R. Hilborn, R. M. Peterman, and M J Staley.
1978. Model for examining early ocean limitation of
Pacific salmon production. J. Fish. Res. Board Can.
35:1303-1315.
222
GROWTH THROUGH THE FIRST SEX MONTHS OF ATLANTIC COD,
GADUS MORHUA , AND HADDOCK, MELANOGRAMMUS AEGLEFINUS,
BASED ON DAILY OTOLITH INCREMENTS^
George R. Bolz and R. Gregory Lough^
ABSTRACT
Daily growth increments of otoliths from larval and juvenile Atlantic cod and haddock were enumer-
ated, and growth curves were derived describing the first six months of life. Growth for both species
was best described by Gompertz-type curves. Inverse regressive methods were employed to construct
general models with confidence limits for predicting age (days) for given standard lengths (mm) from
hatching through the juvenile period. Microstructural analysis of the otoliths did not discern a
settling check at the time when the fish would be expected to leave the pelagic lifestyle for the
demersal one, which indicates that the transition is neither physiologically stressful nor abrupt.
Fluctuations in the year-class strength of fish
stocks are thought to be determined by the rate of
mortality during the first year of life (Moser 1981;
Lough et al. 1985; Neilson and Geen 1986; and
others). Calculation of reliable mortality rates,
assessment of the influences of size-selectivity,
and establishment of precise hatching dates and
times during a given year when loss to recruit-
ment is greatest are dependent upon accurate age
and abundance estimates. Recently, investigators
have suggested that mortality during the postlar-
val and juvenile periods may be as critical as that
occurring in the egg and larval life stages (Cohen
and Grosslein 1982; Sissenwine 1984). Investiga-
tion of this hypothesis by the Northeast Fisheries
Center (NEFC) has been ongoing since 1984.
Enumeration of daily growth increments de-
posited on fish otoliths provides the best method
for the age determination of larvae and juveniles
needed for generating growth curves and estimat-
ing mortality (Essig and Cole 1986). An excellent
review of past and current methodologies em-
ployed in the study and application of otolith mi-
crostructure may be found in Campana and Neil-
son (1985).
Atlantic cod and haddock are both spring
spawners on Georges Bank (Sherman et al. 1984)
and have pelagic eggs and larvae that undergo
similar development. Transformation to the juve-
nile life stage occurs around 20-30 mm SL, or 2-3
months from hatching (Fahay 1983). The transi-
tion from the pelagic to demersal habitat of the
adults takes place sometime after transforma-
tion, usually by 6-8 cm in midsummer, and re-
cent field observations by the NEFC indicates the
transition is a gradual process with considerable
variability.
In an earlier study by Bolz and Lough (1983),
growth curves were developed for larval Atlantic
cod and haddock based on otolith analysis that
defined growth from hatching (4-5 mm SL)
through the first two months of life (ca. 20 mm
SL). Juvenile Georges Bank Atlantic cod and had-
dock are not fully vulnerable to bottom-trawl
gear (Clark et al. 1982), and growth curves based
on groundfish surveys conducted by the NEFC in
the autumn and spring are inaccurate for fish
younger than about six months of age. The pri-
mary goal of the work reported here was to derive
age-at-length curves for field-caught Atlantic cod
and haddock describing their growth from hatch
until they are fully available to capture by
bottom-trawl survey gear. A secondary objective
was to determine if a check ring, a wide incremen-
tal band indicative of physiological or environ-
mental changes, was deposited during the juve-
nile's transition from the pelagic to the demersal
mode of life.
IMARMAP Contribution FED/NEFC 87-15, Northeast Fish-
eries Center Woods Hole Laboratory, National Marine Fish-
eries Service, NOAA, Woods Hole, MA 02543.
^Northeast Fisheries Center Woods Hole Laboratory, Na-
tional Marine Fisheries Service, NOAA, Woods Hole, MA
02543.
Manuscript accepted December 1987.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.
METHODS
Atlantic cod and haddock larvae and juveniles
were collected on six cruises conducted by the
NEFC's RV Albatross IV and RV Delaware II on
223
FISHERY BULLETIN: VOL. 86, NO. 2
Georges Bank during the springs of 1981, 1983,
and 1984 and the summers of 1984 and 1985.
Sample dates and station locations where larvae
and juveniles were collected for otolith analysis
are given in Table 1. The samples were collected
with either 1) a continuous double-oblique haul
using a 61 cm bongo net sampler (0.505 and 0.333
mm mesh) deployed to a maximum depth of 100 m
(Posgay and Marak 1980), 2) a 1 m MOCNESS^
fitted with nine 0.333 mm mesh nets which sam-
pled discrete vertical strata from the bottom of
the water column to the surface, 3) a 10 m MOC-
NESS (3 mm mesh) with five nets fished in the
same manner as the 1 m MOCNESS (Wiebe et al.
1976, 1985), or 4) a Yankee 36 otter trawl towed
for 30 minutes (Grosslein 1974). Stations with
high densities of Atlantic cod and haddock larvae
and juveniles in good condition were selected dur-
ing the cruises for otolith analysis. The fish were
removed immediately following the haul and pre-
served in 95% ethanol.
In the laboratory, larvae and juveniles repre-
sentative of the entire size-range collected were
selected for analysis. The standard length, as well
as several other morphometric measurements of
each larva or juvenile, was measured to the
nearest 0.1 mm prior to removal of their otoliths.
The 2 sagittae, 2 lapilli, and, when possible, 2
asterisci were dissected from the fish and, except-
ing juvenile sagittae, mounted whole on micro-
scope slides with Permounf*. The growth incre-
ments (Fig. lA) on most of these otoliths were
^Multiple Opening/Closing Net and Environmental Sensing
System.
''References to trade names do not imply endorsement by the
National Marine Fisheries Service, NOAA.
Table 1 . — Station information for Atlantic cod and haddock specimens collected for otolith analysis by 61
cm bongo net (0.505 mm mesh) oblique hauls (6B5), 1 m MOCNESS (0.333 mm mesh) discrete vertical
hauls (1M3), 10 m MOCNESS (3.0 mm mesh) discrete vertical hauls (10M), and Yankee 36 otter trawl
(Y36) duhng the 1981, 1983, and 1984 survey seasons.
Time GMT
Bottom
Number of
Lat.
Long.
W
(Night or
depth
M&ll
Station
N
Date
day)
Gear
(m)
Cod
Haddock
1981
Albatross IV
81-03
54
4ri0'
67=06'
24 April
1235(D)
6B5
62
19
—
55
4ri3'
67=02'
24 April
1330(D)
6B5
62
10
—
56
4ri8'
66=58'
24 April
1450(D)
685
66
16
—
57
41°22'
66=55'
24 April
1630(D)
6B5
66
13
—
58
41=26'
66=51 '
24 April
1840(D)
6B5
71
12
—
160
41°22'
67=00'
26 April
0645(N)
1M3
63
32
—
1981
Albatross IV
81-05
190
40°57'
67=19'
22 May
0300(N)
1M3
76
—
8
197
40°55'
67=13'
25 May
1200(D)
1M3
80
—
16
205
40°55'
67=09'
26 May
1130(D)
1M3
80
—
6
211
41-11'
67=35'
27 May
1200(D)
1M3
49
—
27
215
4ri2'
67=36'
27 May
2330(D)
1M3
40
—
19
1983
Albatross IV
83-03
415
40°54'
67=32'
13 May
1816(D)
1M3
74
7
2
418
40°56'
67=35'
14 May
0456(D)
1M3
71
11
16
421
40°5r
67=34'
14 May
1026(D)
1M3
68
2
13
432
40°47'
67=26'
15 May
1636(D)
1M3
89
1
—
434
40°46'
67=24'
15 May
2229(D)
1M3
93
2
—
438
41°05'
67=47'
16 May
1147(N)
1M3
54
—
15
440
4r09'
67=54'
16 May
1646(N)
1M3
52
3
3
442
4r08'
67=48'
16 May
2222(N)
1M3
40
10
—
444
4r09'
67=55'
17 May
0504(N)
1M3
52
13
—
1984
Albatross IV
84-05
519
4ri9'
67=19'
18 June
0319(N)
10M
47
—
30
1984
Delaware II
84-07
76
40°53'
66=22'
15 Aug.
1744(D)
Y36
66
1
1
85
4r50
66=26'
16 Aug.
1045(D)
Y36
78
—
1
88
41=49'
66=23'
16 Aug.
1430(D)
Y36
62
—
1
89
41°47°
66=18'
16 Aug.
1604(D)
Y36
68
—
4
90
4r47'
66=24'
16 Aug.
1715(D)
Y36
78
—
3
91
41=47'
66=30'
16 Aug.
1821(D)
Y36
72
—
1
93
41°45'
66=30'
16 Aug.
2024(D)
Y36
75
—
1
94
41=42'
66=25'
16 Aug.
2124(D)
Y36
75
—
1
98
41=47'
66=1 1 '
17 Aug.
0249(N)
Y36
70
4
—
1984
Albatross IV
84-09
18
41=49'
66=16'
12 Sept.
0845(N)
Y36
70
—
9
19
41=52'
66=21'
12 Sept.
0951 (N)
Y36
90
1
12
224
BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK
Figure 1. — A. Scanning electron micrograph for a portion of the sagitta from a 5-yr-old Atlantic cod, 79 cm SL, showing daily growth
increments. Bar of photograph represents 100 (xm. B. Sagitta from 47-d-old Atlantic cod larva, 13.3 mm SL (630x). Bar of
photograph represents 20 ji.m. nc = nuclear check, yc = yolk-sac check.
225
FISHERY BULLETIN: VOL. 86, NO. 2
discernible without any further preparation.
Sagittae from fish >25 mm SL were mounted in
epoxy resin and were ground, above and below,
with carborundum paper (600 grit). The resulting
thin section was secured to a microscope slide
with epoxy resin and etched with 6% EDTA (pH
7.0). Both the grinding and etching procedures
were monitored periodically by viewing the
sagitta under a dissecting microscope.
The sagittae were then viewed under a Zeiss
compound microscope with transmitted light. The
number of growth increments were counted from
the image projected by a drawing tube onto a
Zeiss MOP Digital Image Analyzer System.
Under transmitted light each growth increment
was composed of a light and dark ring (Fig. IB),
which corresponded to the heavily calcified incre-
mental zone and the organic-rich discontinuous
zone of Watabe et al. (1982). Depending on the
size of the otolith, magnifications used ranged
from 400 X to 1,000 x . Three counts were made on
one of the 2 sagittae from each larva or juvenile,
and those otoliths with a repeatable increment
count of >90% were used in the growi:h analysis.
The other sagitta was counted once for compari-
son, as were the 2 lapilli. The number of incre-
ments on the 2 asterisci also were enumerated. It
was found in the previous study (Bolz and Lough
1983) that the asterisci were not detectable at
hatching, in contrast to the sagittae and lapilli,
but appeared later in the larval period. This was
reflected in the asterisci having on average 27
fewer growth increments than the sagittae. In
those instances where the sagittae and lapilli
were particularly difficult to read, the number of
asteriscal increments plus 27 was consulted as an
additional check. Maximum and minimum di-
ameters and planar surface area of the entire
otolith were measured routinely on all sagittae,
lapilli, and asterisci.
The differential shrinkage of Atlantic cod and
haddock larvae and juveniles with respect to
standard length was corrected using Theilacker's
algorithm ( 1980), which is specified and discussed
in Bolz and Lough (1983). All lengths referred to
in the results and discussion portions of this
paper are reported as corrected lengths.
RESULTS
Haddock Larval and Juvenile Growth
From analysis of the 189 larval and juvenile
haddock, ranging from 3.5 to 123.4 mm SL, we
found that growth was best described by a
Gompertz-type curve. Previous uses of the Gom-
pertz growth curve and methodology for fitting
the curve are described in Pennington (1979),
Lough et al. (1982), and Messieh et al. (1987). The
variance was stabilized by using the natural log
form of the growth equation, and parameters
were derived by nonlinear estimation techniques
resulting in the relationship:
ln(L) = 1.1987 + 4.8438(1 -e
0.0088R\
(1)
where L = standard length in mm, and
R = number of days (increments) from
hatch.
A plot of the Gompertz curve fitted to the natural
log of standard length vs. age in days is shown in
Figure 2.
The predicted hatch-length fi-om the curve of
3.32 mm falls within the range of previous studies
(Colton and Marak 1969; Fahay 1983). An aver-
age growth rate of 0.24 mm/day (Table 2) for the
first 30 days is also reasonable (Laurence 1978;
Laurence et al. 1981) and agrees with the earlier
study of Bolz and Lough (1983). As a generalized
model the Gk)mpertz equation described haddock
growth through the first six months (175 days), at
which point it intersected (Fig. 3) the von Berta-
lanffy growth curve generated fi'om an analysis of
adult haddock by Clark et al. (1982):
Table 2. — Mean standard length at age, 95% confidence limits,
and growth rate (mm/day and %/day) of larval and juvenile had-
dock from hatch through 200 days estimated from the Gompertz
growth model fit.
Mean
95% confidence
limits
Growth
Growth
Age
(d)
length
(mm)
rate
rate
Lower
Upper
(mm/day)
(%/day)
0
3.32
3.22
3.41
0.14
4.22
10
4.99
4.88
5.11
0.20
4.01
20
7.27
7.13
7.41
0.26
3.58
30
10.25
10.07
10.42
0.34
3.32
40
14.03
13.80
14.26
0.42
2.99
50
18.70
18.39
19.02
0.51
2.78
60
24.34
23.89
24.80
0.61
2.55
70
30.98
30.34
31.63
0.71
2.32
80
38.63
37.74
39.53
0.82
2.12
90
47.27
46.08
48.51
0.91
1.95
100
56.88
55.30
58.50
1.01
1.77
110
67.37
65.35
69.45
1.09
1.63
120
78.67
76.14
81.28
1.17
1.49
130
90.66
87.56
93.86
1.23
1.35
140
103.23
99.52
107.08
1.28
1.23
150
116.26
111.88
120.81
1.32
1.14
160
129.63
124.54
134.92
1.35
1.04
170
143.21
137.38
149.28
1.36
0.95
180
156.88
150.29
163.76
1.37
0.87
190
170.54
163.17
178.25
1.36
0.80
200
184.09
175.92
192.63
1.35
0.73
226
BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK
JZ
-P
cn
c
X)
L.
O
"D
C
o
-p
cn
200.0
100.0 -■
50.0 -■
25. 0 --
12.0
-Q 0088R
ln(L) = 1. 1987 + 4.8438(1 - g " )
-I — I — ^-
-I — \ — U
4 — I — U
J I \ L
J I L
25
50
75
100
125
150
175
AgG in Days
Figure 2. — Gompertz growth curve and equation fitted to plot of In standard length and number of otolith increments (estimated age
in days) for 189 larval and juvenile haddock collected on Georges Bank.
L = 738.0(1 - e-0.3763[(fl+D)/365-0.1649])^ (2)
where D = Julian date of hatch.
Based on the 1981 season, an average hatch-date
of 15 April {D = 105) was employed in the present
model. An average length of 19.9 cm would have
been attained on 1 January, by fisheries science
convention the date at which an individual is con-
sidered to be 1-year-old.
The predicted hatch-length of 4.02 mm was
within known limits (Colton and Marak 1969).
The average growth rate of 0.21 mm/day (Table 3)
through the first month was slightly lower than
that of haddock, which is consistent with previous
findings (Bolz and Lough 1983). At approxi-
mately 192 days the larval and juvenile growth
curve intersected the von Bertalanffy curve calcu-
lated for adult Atlantic cod by Penttila and Gif-
ford (1976):
Atlantic Cod Larval and
Juvenile Growth
Although there were few larger individuals
amongst the 157 larval and juvenile Atlantic cod
examined, the apparent pattern was similar to
that seen in haddock. A Gompertz growth curve
also provided a good fit when the natural log of
standard length (range: 4.6-104 mm) was plotted
(Fig. 4) against age in days (range: 7-151):
ln(L) = 1.3915 + 6.2707(1 - e-o.oo53/?)_
(3)
L = 1481 0(1 - e-01200((i?+D)/365-0.6160)^ (4)
For purposes of the model a mean hatch-date of 15
March (D = 74) was assumed. An average At-
lantic cod would be expected to have achieved a
length of 26.1 cm by 1 January (Fig. 5).
Predictability
Since it is desirable, especially during field sur-
veys when direct analysis of otoliths is impossi-
ble, to be able to predict age from a given length,
227
FISHERY BULLETIN: VOL. 86, NO. 2
SL
-t->
cn
c
Qi
L.
O
X)
c
D
-P
tn
200.0
-
*
100.0
-
•
50.0
-
^^
-. 3783 <(R- 1055/365-
ln<L) = 6.6039 * Ind - e
Adult von Bertalanffy Growth Curve
Clark at al. (1982)
1849)
25.0
12.0
•5P^
jf^' '
6.0
•
>?
i^
-0. 0088R
ln(L) = 1. 1987 * 4.8438C1 - e )
3.0
n
/
•
1
Larval and Juvenile Gompertz Growth Curve
1 1 1 I
1
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dqc
Jan
Figure 3. — Haddock growth model with an assumed mean hatch-date of 15 April projected through mid-January. An average
length of 19.9 cm would have been attained on 01 January, by fisheries science convention the date at which an individual is
considered to be 1-year-old.
inverse regression (Draper and Smith 1966) was
performed on the Atlantic cod and haddock
grovii;h curves to estabhsh confidence intervals
for predicting age from a given standard length.
In its reduced form the equation obtained for had-
dock was
X^ _ ln(l - (Xq ± Q.023^(((Xo - 0.2990)2/7.7959) + (1 + l/n))^^^))
Xi -0.0088
(5)
where X^^ and Xi = upper and lower confidence
limits,
Xo=l-e-ooo88« and
n = sample size.
Figure 6 shows the fitted growth curve bracketed
by 95% confidence intervals.
Performing the same calculations on the At-
lantic cod growth curve yielded the relation-
ship:
X„, _ ln(l - (Xq ± 0.022t{{iXo - 0.1918)^/0.9294) + (1 + l/n))^'^))
X,
0.0053
(6)
where Zo = l-e-o 0053ft
228
BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK
E
JZ
4->
m
c
Q)
XI
O
X)
c
D
C/)
200.0 -
100.0 -
50. 0 -
25.0 •-
12.0
6.0 -
3.0 -
ln(L) = 1.3915 * 6.2707(1 - g-°-°°53Rj
J I \ L
J I L
J 1 L
J I L
25
50
75
100
125
150
175
AgG in Days
Figure 4. — Gompertz growth ciirve and equation fitted to plot of In standard length and number of otolith increments (estimated age
in days) for 157 larval and juvenile Atlantic cod collected on Georges Bank.
Table 3. — Mean standard length at age, 95% confidence limits,
and growth rate (mm/day and %/day) of larval and juvenile Atlantic
cod from hatch through 200 days estimated from the Gompertz
growth model fit.
Mean
95% confidence
limits
Growth
Growth
Age
(d)
length
(mm)
rate
rate
Lower
Upper
(mm/day)
(%/day)
0
4.02
3.79
4.27
0.13
3.37
10
5.56
5.31
5.82
0.18
3.24
20
7.55
7.30
7.82
0.23
3.05
30
10.11
9.85
10.37
0.29
2.87
40
13.32
13.03
13.62
0.36
2.70
50
17.31
16.88
17.75
0.44
2.60
60
22.19
21.48
22.92
0.54
2.43
70
28.08
26.95
29.26
0.64
2.31
80
35.11
33.40
36.91
0.76
2.19
90
43.39
40.91
46.03
0.90
2.10
100
53.05
49.58
56.77
1.04
1.98
110
64.18
59.48
69.26
1.19
1.87
120
76.90
70.69
83.65
1.35
1.77
130
91.27
83.26
100.05
1.52
1.68
140
107.38
97.24
118.57
1.70
1.58
150
125.27
112.65
139.30
1 88
1.51
160
144.99
129.52
162.30
2.06
1.43
170
166.55
147.85
187.61
2.25
1.35
180
189.95
167.61
215.26
2.43
1.28
190
215.17
188.79
245.24
2.61
1.22
200
242.18
211.34
277.52
2.79
1.16
Figure 7 shows the Atlantic cod growth curve
bracketed by 95% confidence intervals. Tables 4
and 5 provide predicted ages of Atlantic cod and
haddock for given standard lengths with 70% and
95% confidence limits.
Otolith Growth
In the earlier study of larval Atlantic cod and
haddock (Bolz and Lough 1983), it was found that
the sagittal rings (one incremental and one dis-
continuous zone) were segregated into distinct re-
gions separated by thicker, darker discontinuous
zones referred to as checks or check rings. Two
"heavy rings" were noted in the larvae: 1) a nu-
clear check surrounding a central, amorphous
core and 1 or 2 irregular rings, and 2) a yolk-sac
check 2-8 increments farther outward (Fig. IB).
The present study corroborated the existence of
these two checks. Although each otolith was care-
fully examined for the presence of a settling
check, no regularly occurring heavy ring could be
discerned beyond the yolk-sac check in either the
haddock or the Atlantic cod juveniles. It should,
229
FISHERY BULLETIN: VOL. 86, NO. 2
E
E
\^
JZ
c
Q)
L
O
"□
C
o
CO
200.0
100.0
50.0
25.0
12.0
6.0
3.0
-0.0053R
In(L) = 1.3915 * 6.2707(1 - e )
Larval and Juvonile Gompertz Growth Cunva
ln(L) = 7.6622 * Ind
-. 1200((R*74)/3e5*.eiBro.
Adult von Bortalanffy Growth Curve
Panttila and Glfford (1976)
Mar Apr
Mav
Jun
Jul
Aug
Sep
Oct
Nov
Dec Jan
Figure 5. — Atlantic cod growth model with assumed mean hatch-date of 15 March projected through mid-January. An average
length of 26.1 cm would have been attained by 01 January, by fisheries science convention the date at which an individual is
considered to be 1-year-old.
D
Q
CD
<
y
200
;
^
175
-
^
^
^
^-'^^.^
150
-
^
^
l^
^ - '^
125
-
^
•
lUU
-
/-
/
y
^
75
—
/
/ y
/ /
y
/
y
50
—
f
25
n
1
f
1
111)
1 1 . 1 < 1
1 1
1 1 1 1 1
I 1 1
1 1 I 1 1 1 1 1
25
50
75
100
125
150
175
230
Standard Length (mm)
BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK
200
175
D
Q
W
CD
<;
150
-
125
_^
^ .^--"''^
/ ^^^-^""^
/ ^^'''''^ ^ -^
.
/ '^''■^'^ --
100
-
■^ ^...^ . ^ '"
■
y /-^ • • -^
_
/ ^X ^ '^
75
-
"
/ .X /
50
-
i
3r>
25
n
- itl
/
J — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — I — 1_. 1 1 1 1 1
25
50
75
100
— I — I — 1 I I I I I I I
125 150 175
Standard Length (mm)
Figure 7. — Inverse regression of Atlantic cod growth curve with 95% confidence intervals for predicting age in days for a given
standard length (mm).
however, be noted that the size range available
for study limited the search for a settling check to
individuals >90 mm SL and does not preclude the
possibility that thinner, less discernible checks
may be found when greater numbers of juveniles
50-90 mm SL are analyzed.
In both haddock and Atlantic cod diametral
growth (fjim) of the sagittae, lapilli, and asterisci
was linearly related to standard length (mm)
throughout the larval and juvenile periods. The
high correlation (r > 0.98) of this relationship
and its good agreement with measurements made
by Bergstad (1984) would allow the sagittal di-
ameter to be used as a check on the predictability
model outlined in Equations (5) and (6) for esti-
mating age from standard length. Estimated ages
for haddock larvae and juveniles based on maxi-
mum otolith diameters may be obtained with the
following equation:
Figure 6. — Inverse regression of haddock growth curve with
95% confidence intervals for predicting age in days for a given
standard length (mm).
Y = 28.390 + 2.413Xi + 21.561X2 + 73.841X3
(7)
where Y = estimated age in days,
Xi = sagittal diameter in mm,
X2 = lapillus diameter in mm, and
X3 = astericus diameter in mm.
Table 6 provides a comparison of estimated ages
derived from otolith diameters with observed
ages derived from the number of daily incre-
ments. Although multiple regression analysis
using the three otolith diameters yielded a high
correlation coefficient (r = 0.9890) and nonsignif-
icant ^-values, the 95% confidence limits are
quite broad (±2 weeks) and should be used with
caution. Use of the sagittal diameter alone
(Y = 35.945 + 18.484Xi) provided a good fit
(r = 0.9861) for juveniles >90 mm SL but was a
poor age predictor for younger fish. If only the
sagitta is available for analysis, the relationship:
231
Table 4. — Predicted age in days with 70% and 95% confidence
limits of larval and juvenile haddock for a given standard length.
FISHERY BULLETIN: VOL. 86, NO 2
Table 5. — Predicted age in days with 70% and 95% confidence
limits of larval and juvenile Atlantic cod for a given standard length.
Observed
standard
length
(mm)
Predicted
age
(d)
70% confidence
limits
95% confidence
limits
Observed
standard
length
(mm)
Predicted
age
(d)
70% confidence
limits
95% confidence
limits
Lower
Upper
Lower
Upper
Lower
Upper
Lower
Upper
5
10.0
7.1
13.0
4.4
15.9
5
6.7
2.1
11.3
-2.1
15.7
10
29.3
25.8
32.8
22.7
36.2
10
29.6
24.5
34.8
19.9
39.8
15
42.3
38.4
46.3
34.9
50.1
15
44.5
39.0
50.1
34.0
55.5
20
52.5
48.2
56.9
44.4
61.1
20
55.8
49.9
61.8
44.7
67.6
25
61.1
56.5
65.8
52.4
70.4
25
65.0
58.9
71.3
53.4
77.5
30
68.6
638
73.7
59.4
78.7
30
72.9
66.5
79.6
60.8
86.0
35
75.4
70.3
80.8
65.7
86.1
35
79.9
73.2
86.8
67.3
93.5
40
81.7
76.2
87.4
71.4
93.0
40
86.1
79.2
93.3
73.1
100.3
45
87.5
81.7
93.5
76.7
99.5
45
91.8
84.7
99.2
78.3
106.5
50
92.9
86.9
99.3
81.6
105.6
50
97.0
89.7
104.7
83.2
112.2
55
98.1
91.8
104.8
86.3
111.4
55
101.9
94.3
109.8
87.6
117.6
60
103.1
96.5
110.0
90.7
117.0
60
106.4
98.7
114.6
91.8
122.6
65
107.8
101.0
115.1
95.0
122.4
65
110.7
102.7
119.1
95.7
127.3
70
112.4
105.3
120.0
99.0
127.6
70
114.7
106.6
123.3
99.4
131.8
75
116.8
109.4
124.8
103.0
132.7
75
118.6
110.3
127.4
102.9
136.1
80
121.1
113.5
129.4
106.8
137,7
80
122.3
113.8
131.3
106.3
140.2
85
125.3
117.4
133.9
110.4
142.6
85
125.8
117.1
135.0
109.5
144.1
90
129.5
121.2
138.4
114.0
147.4
90
129.2
120.3
138.6
112.5
147.9
95
133.5
124.9
142.8
117.5
152.2
95
132.4
123.4
142.0
115.5
151.6
100
137.5
128.6
147.1
121.0
156.9
100
135.6
126.4
145.4
118.3
155.1
105
141.4
132.2
151.4
124.3
161.6
105
138.6
129.2
148.6
121.0
158.6
110
145.2
135.8
155.6
127.6
166.2
110
141.5
132.0
151.7
123.7
161.9
115
149.0
139.3
159.8
130.9
170.8
115
144.4
134.7
154.8
126.3
165.1
120
152.8
142.7
163.9
134.1
175.4
120
147.2
137.3
157.7
128.7
168.3
125
156.6
146.1
168.1
137.2
180.0
125
149.9
139.9
160.6
131.1
171.3
130
160.3
149.5
172.2
140.4
184.6
130
152.5
142.3
163.4
133.5
174.3
135
164.0
152.9
176.3
143.4
189.2
135
155.1
144.8
166.1
135.8
177.3
140
167.6
156.2
180.4
146.5
193.9
140
157.6
147.1
168.8
138.0
180.2
145
171.3
159.5
184.6
149.5
198.5
145
160.0
149.4
171.4
140.2
183.0
150
175.0
162.8
188.7
152.5
203.2
150
162.4
151.6
174.0
142.3
185.7
155
178.6
166.0
192.8
155.5
207.9
155
164.8
153.8
176.5
144.4
188.5
160
182.3
169.3
197.0
158.5
212.7
160
167.1
156.0
179.0
146.4
191.1
165
185.9
172.6
201.2
161.4
217.5
165
169.3
158.1
181.4
148.4
193.7
170
189.6
175.8
205.4
164.3
222.4
170
171.5
160.2
183.8
150.4
196.3
175
193.3
179.0
209.6
167.3
227.4
175
173.7
162.2
186.2
152.3
198.8
Y = 11.875 + 112.654Xi, r = 0.9129
(8)
DISCUSSION
should be used for the larval and postlarval pe-
riod.
The equation for the estimated ages of larval
and juvenile Atlantic cod (Table 7) is as follows:
Y = 48.202 + 8.628Xi - 121.908^2 + 139.733Z3,
r = 0.9292. (9)
When using only the sagittal diameter, the fol-
lowing relationship should be applied to larvae
and postlarvae:
Y = 19.364 + 89.560Xi, r = 0.8659. (10)
Unlike the tedious laboratory methods needed for
the enumeration of otolith increments, gross
measurements on a limited number of juvenile
otoliths could be performed at sea.
Despite the tedious methodology required for
enumerating daily growth increments in larval
and juvenile otoliths, the present work suggests
that it is feasible to construct age-length keys for
Atlantic cod and haddock similar to those com-
monly applied in adult population studies (Clark
et al. 1982). The value of such growth data is
based on several assumptions, however. Since all
conclusions drawn depend upon it, reasonable as-
surance of the day-increment relationship in the
species being investigated is critical (Beamish
and McFarlane 1987; Geffen 1987). Confidence in
the growth models generated here for Atlantic
cod and haddock may be found in the following
inferences: 1) the predicted hatch lengths of 3.32
mm for haddock and 4.02 mm for Atlantic cod fall
within known limits; 2) a high correlation for the
length-at-age data with the rearing experiments
of Laurence et al. (1981); and 3) the smooth-
232
BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK
Table 6. — Estimated age in days based on otolith diame-
ters with 95% confidence limits for larval and juvenile had-
dock compared with observed age derived from number of
daily increments.
Observed
age
(d)
Estimated
age
(d)
95%
confidence
limits
Lower
Upper
0.0
1.5
-12.3
15.4
10.0
11.3
-2.5
25.1
20.0
21.1
7.4
34.9
30.0
30.9
17.2
44.6
40.0
40.7
27.0
54.4
50.0
50.5
36.8
64.2
60.0
60.3
46.7
73.9
70.0
70.1
56.5
83.7
80.0
79.9
66.2
93.6
90.0
89.7
76.0
103.3
100.0
99.5
85.8
113.1
110.0
109.3
95.6
122.9
120.0
119.0
105.3
132.8
130.0
128.8
115.1
142.6
140.0
138.6
124.8
152.5
150.0
148.4
134.5
162.3
160.0
158.2
144.3
172.2
170.0
168.0
154.0
182.0
180.0
177.8
163.7
191.9
Table 7. — Estimated age in days based on otolith diame-
ters with 95% confidence limits for larval and juvenile At-
lantic cod compared with observed age derived from num-
ber of daily increments.
Observed
age
(d)
Estimated
age
(d)
95%
confidence
limits
Lower
Upper
0.0
7.1
-6.9
21.1
10.0
15.8
2.1
29.5
20.0
24.5
11.0
38.0
30.0
33.2
19.9
46.5
40.0
41.9
28.7
55.1
50.0
50.6
37.5
63.8
60.0
59.3
46.2
72.5
70.0
68.0
54.8
81.2
80.0
76.7
63.4
90.1
90.0
85.4
71.9
98.9
100.0
94.1
80.4
107.9
110.0
102.8
88.8
116.9
120.0
111.6
97.2
125.9
130.0
120.3
105.5
135.0
140.0
129.0
113.8
144.1
150.0
137.7
122.1
153.3
160.0
146.4
130.3
162.5
170.0
155.1
138.5
171.7
180.0
163.8
146.6
180.9
ness with which the larval and juvenile curves
flow into those independently developed for the
adults (Clark et al. 1982; Penttila and Gifford
1976).
The predictive models for Atlantic cod and had-
dock have to be viewed as general in nature, and
the widening of the confidence intervals with in-
creasing length (Tables 4, 5) must be kept in
mind. Natural variability of length-at-age and
difficulty in the preparation and reading of
otoliths increases as the fish becomes older and
makes precise age determinations extremely dif-
ficult. For example, the ability to predict correctly
the age of an individual haddock at the 70% con-
fidence level decreases from ±3 days at 5 mm SL
to ±2 weeks at 175 mm. In spite of this problem,
otolith aging of field-caught larvae and juveniles
provides a degree of precision not possible with
indirect methods based on size-frequency analy-
ses (Ebert 1973). Refinement of the estimated
means and the reduction and stabilization of the
variance should result as a greater number of
otoliths are analyzed in the future.
Microstructural examination of larval Atlantic
cod and haddock otoliths clearly delineated check
rings related to hatching and yolk-sac absorption
(Bolz and Lough 1983). Both of these transitions
are abrupt, and the dark, thick discontinuous
zones readily observable on the otoliths are a re-
flection of metabolic disturbances undergone at
these times. Although additional check rings
were noted in 3 or 4 of the juvenile otoliths, there
was no regularity with respect to age of their oc-
currence. In these individuals the checks were
probably the result of physiological trauma in-
duced by disease or injury since calcium carbon-
ate secretion ceases not only with the metabolic
changes accompanying transitional phases but
during times of stress (Morales-Nin 1987). It was
suspected that a distinct check, similar to the set-
tling check found by Victor (1982) in the bluehead
wrasse, Thalassoma bifasciatum , would be found
demarcating the transition from the pelagic to
the demersal mode of life with its accompanying
changes in diet and activity levels. No check rings
were found in the transition period (50-100 days)
on the otoliths analyzed. This suggests that an
abrupt metabolic disturbance does not occur at
this phase of the fish's life and that settling near
the bottom takes place over an extended period of
time (1-2 months) even for individual fish. This
agrees with a preliminary finding for Scotian
Shelf gadoids by Campana and Neilson (1985).
However, in a recent study by Mahon and Neilson
(1987) on the gut contents of Scotian Shelf had-
dock, they concluded that the transition from
pelagic to demersal life occurred relatively sud-
denly, less that a month for the individual fish.
Apparently, change to the demersal life stage is
not stressful for Atlantic cod and haddock, at
least as a metabolic manifestation recorded in
their otoliths.
233
FISHERY BULLETIN: VOL. 86. NO. 2
When used in conjunction with length-
frequency data collected throughout the year, the
Atlantic cod and haddock growth curves pre-
sented in this report should allow accurate esti-
mates of the following: 1) peak hatching dates, 2)
the number of cohorts produced within a given
season, 3) intraseasonal changes in growth and
mortality rates of cohorts, and 4) which part of the
spawning curve the recruits originated from
(Methot 1983). In the future year-to-year com-
parison of deviations in these estimates could
lead to the construction of viable recruitment
models permitting the early prediction of year-
class strength.
ACKNOWLEDGMENTS
We gratefully acknowledge the technical help
and guidance provided by Susan Houghton with
the scanning electron microscope portions of this
paper and by Michael Pennington with the statis-
tical analyses.
LITERATURE CITED
Beamish, R J . and G. A McFarlane.
1987. Current trends in age determination methodology.
In R. C. Summerfelt and G. E. Hall (editors), Age and
growth offish, p. 15-42. Iowa State Univ. Press, Ames.
Bergstad, 0 A.
1984. A relationship between the number of growth incre-
ments on the otoliths and age of larval and juvenile cod,
Gadus morhua L. In E. Dahl, D. S. Danielssen, E. Mok-
ness, and P. Solemdal (editors). The propagation of cod
Gadus morhua L, p. 251—272. Flodevigen rapportser. 1,
1984.
BoLZ, G R., and R. G. Lough.
1983. Growth of larval Atlantic cod, Gadus morhua, and
haddock, Melanogrammus aeglefinus , on Gteorges Bank,
spring 1981. Fish. Bull., U.S. 81:827-836.
Campana. S. E , and J. D. Neilson.
1985. Microstructure of fish otoliths. Can. J. Fish.
Aquat. Sci. 42:1014-1032.
Clark, S. H . W J. Overholtz, and R. C. Hennemuth.
1982. Review and assessment of the Georges Bank and
Gulf of Maine haddock fishery. J. Northwest Atl. Fish.
Sci. 3:1-27.
Cohen. E. B., and M D Grosslein.
1982. Food consumption by silver hake (Merluccius bilin-
earis) on Georges Bank with implications for recruit-
ment. In G. M. Calliet and C. A. Simenstadt (editors),
Gutshop '81, fish food habits studies, p. 286-294. Proc.
Third Pac. Workshop, Washington Sea Grant, Univ.
Wash. Press, Seattle.
COLTON, J B , AND R R. MaRAK.
1969. Guide for identifying the common planktonic fish
eggs and larvae of continental shelf water. Cape Sable to
Block Island. U.S. Bur. Commer. Fish., Biol. Lab.,
Woods Hole, Mass., Lab. Ref. Doc. 69-9, 43 p.
Draper, N R , and H. Smith.
1966. Applied Regression Analysis. Wiley, N.Y.
Ebert, T a
1973. Estimating growth and mortality rates from size
data. Oecologia 11:281-298.
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.
Fahay, M P
1983. Guide to the early 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.
Geffen, a J
1987. Methods of validating daily increment deposition in
otoliths of larval fish. In R. C. Summerfelt and G. E.
Hall (editors). Age and growth offish, p. 223-240. Iowa
State Univ. Press, Ames.
Grosslein, M. D.
1974. Bottom trawl survey methods of the Northeast
Fisheries Center, Woods Hole, Massachusetts, U.S.
ICNAF Res. Doc. 74/86, Ser. No. 3332.
Laurence, G. C.
1978. Comparative growth, respiration and delayed feed-
ing abilities of larval cod (Gadus morhua) and haddock
{Melanogrammus aeglefinus) as influenced by tempera-
ture during laboratory studies. Mar. Biol. (Berl.) 50:1-
7.
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 ae-
glefinus ) in laboratory growth and survival studies at
different food densities. In R. Lasker and K. Sherman
(editors). The early life history of fish, p. 304-311.
Rapp. P.-v. Reun. Cons. int. Explor. Mer 178.
Lough, R. G., G R. Bolz, M. Pennington, and M D Grosslein
1985. Larval abundance and mortality of Atlantic herring
(Clupea harengus L.) spawTied in the Georges Bank and
Nantucket Shoals areas, 1971-78 seasons, in relation to
spawning stock size. J. Northwest Atl. Fish. Sci. 6:21-
35.
Lough, R. G., M. Pennington, G R. Bolz, and A A Rosenberg.
1982. Age and growth of larval Atlantic herring, Clupea
harengus L., in the Gulf of Maine-Georges Bank region
based on otolith growth increments. Fish. Bull., U.S.
80:187-199.
Mahon, R., and J D Neilson.
1987. Diet changes in Scotian Shelf haddock during
pelagic and demersal phases of the first year of life.
Mar. Ecol. Prog. Ser. 37:123-130.
Messieh, S. N., D S Moore, and P. Rubec.
1987. Estimation of age and growth of larval Atlantic her-
ring as inferred from examination of daily growth incre-
ments of otoliths. In R. C. Summerfelt and G. E. Hall
(editors), Age and growth of fish, p. 433-442. Iowa
State Univ. Press, Ames.
Methot, R D, Jr.
1983. Seasonal variation in survival of larval northern
anchovy, Engraulis mordax, estimated from the age dis-
tribution of juveniles. Fish. Bull., U.S. 81:741-750.
MoraleS-Nin, B.
1987. Ultrastructure of the organic and inorganic con-
stituents of the otoliths of the sea bass. In R. C. Sum-
merfelt and G. E. Hall (editors). Age and growth offish,
p. 331-343. Iowa State Univ. Press, Ames.
234
BOLZ AND LOUGH: GROWTH OF ATLANTIC COD AND HADDOCK
MOSER, H G.
1981. Morphological and functional aspects of marine fish
larvae. In R. Lasker (editor), Marine fish larvae, p. 90-
131. Univ. Wash. Press, Seattle and Lond.
NEILSON, J. D., AND G. H. Geen.
1986. First-year growth rate of Sixes River chinook
salmon as inferred from otoliths: effects on mortality and
age at maturity. Trans. Am. Fish. Soc. 115:28-33.
Pennington. M R
1979. Fitting a growth curve to field data. In J. KOrd.,
G. P. Patil, and C. Taillie (editors). Statistical distribu-
tions in ecological work, p. 419-428. Int. Coop. Publ.
House, Fairland, MD.
PENTTILA, J. A , AND V. M. GiFFORD.
1976. Growth and mortality rates for cod from the
Georges Bank and Gulf of Maine areas. Int. Comm.
Northwest Atl. Fish, Res. Bull. No. 12, p. 29-36.
POSGAY. J A , AND R R MARAK
1980. The MARMAP bongo zooplankton sampler. J.
Northwest Atl. Fish. Sci. 1:91-99.
Sherman. K , W Smith, W Morse. M Berman, J Green, and
L. Ejsymont.
1984. Spawning strategies of fishes in relation to circula-
tion, phjloplankton production, and pulses in zooplank-
ton off the northeastern United States. Mar. Ecol. Prog.
Ser. 18:1-19.
SiSSENWINE, M p.
1984. Why do fish populations vary? In R. May (editor),
Exploitation of marine communities, p. 59-94.
Springer-Verlag, N.Y.
Theilacker, G H.
1980. Changes in body measurements of larval northern
anchovy, Engraulis mordax, and other fishes due to han-
dling and preservation. Fish. Bull., U.S. 78:685-692.
Victor, B. C.
1982. Daily otolith increments and recruitment in two
coral-reef wrasses, Thalassoma bifasciatum and Halicho-
eres bivittatus. Mar. Biol. (Berl.) 71:203-208.
Watabe, N . K. Tanaka. J Yamada, and J. M Dean.
1982. Scanning electron microscope observations of the
organic matrix in the otolith of the teleost fish Fundulus
heteroclitus (Linnaeus) and Tilapia nilotica (Linnaeus).
J. Exp. Mar. Biol. Ecol. 58:127-134.
Wiebe, P H , K. H Burt, S H Boyd, and A W Morton.
1976. A multiple opening/closing net and environmental
sensing system for sampling zooplankton. J. Mar. Res.
34:313-326.
Wiebe, P. H., A. W. Morton. A M Bradley, R H Backus.
J. E. Craddock, V Barber, T. J. Cowles, and G R Flierl.
1985. New developments in the MOCNESS, an apparatus
for sampling zooplankton and micronekton. Mar. Biol.
(Berl.) 87:313-323.
235
THE RELATION BETWEEN SPAWNING SEASON AND
THE RECRUITMENT OF YOUNG-OF-THE-YEAR BLUEFISH,
POMATOMUS SALTATRIX, TO NEW YORK^
Robert M. Nyman^ and David O Conover^
ABSTRACT
The association between oceanic spawning season and the recruitment of young-of-the-year (YOY)
bluefish, Pomatomus saltatrix, to the inshore waters of New York was studied by estimating the
spawn dates of recruited fish collected in the shore zone from the number of growth increments in
their otoliths. Field collections on the south shore of Long Island showed that recruitment of 3-6 cm
fork length fish occurred as a distinct pulse during the last week of May in 1985 and the second week
of June in 1986. Length-frequency distributions were generally unimodal and most fish collected later
could be attributed to this one recruitment episode. The frequency of otolith ring deposition in YOY
bluefish was determined by marking the otoliths of field-caged fish with an injection of tetracycline,
and then periodically subsampling these over the ensuing 61-day period. Regression analysis indi-
cated a 1:1 relation between the number of days since marking and the number of rings beyond the
mark. Back-calculation to the time of first ring deposition revealed that field-collected YOY bluefish
from Long Island were spawned primarily in the March-April spawning season reported to occur
south of Cape Hatteras. Relatively few fish were collected from the summer spawning season that
reportedly occurs in the Middle Atlantic Bight. Almost all of these summer-spawned fish were
collected from the Hudson River.
The bluefish, Pomatomus saltatrix, supports a
major recreational fishery along the Atlantic
coast of the United States. In 1985, more bluefish
by weight were caught than any other marine
fish, accounting for over 24% of the total marine
recreational catch (U.S. Department of Com-
merce 1986). Despite the importance of bluefish
to the recreational fishery, very little is known of
its early life history.
Bluefish are found over different portions of the
continental shelf from Florida to Nova Scotia at
various times of the year (Bigelow and Schroeder
1953; Wilk 1977; Gilmore 1985). Based on de-
scriptions of the temporal and spatial abundance
of larvae, Kendall and Walford (1979) suggested
that there are primarily two distinct spawning
periods and regions: a spring spawning in the
South Atlantic Bight at the edge of the Florida
Current (see also Collins and Stender 1987), and
a summer spawning in the Middle Atlantic Bight
iContribution No. 588 of the Marine Sciences Research Cen-
ter, State University of New York, Stony Brook, NY.
^Marine Sciences Research Center, State University of New
York, Stony Brook, NY 11794-5000; present address:
Chesapeake Biological Laboratory, University of Maryland,
Solomons, MD 20688.
^Marine Sciences Research Center, State University of New
York, Stony Brook, NY 11794-5000.
Manuscript accepted November 1987.
FISHERY BULLETIN; VOL. 86, NO. 2, 1988.
midway over the continental shelf (see also Morse
et al. 1987). They further proposed that the
spring-spawned larvae are transported north-
ward in the slope waters and then move inshore,
spending their first summer in the bays and estu-
aries of the Middle Atlantic Bight. Summer-
spawned larvae, according to Kendall and Wal-
ford, spend their first summer at sea or enter the
estuaries of the Middle Atlantic Bight only
briefly before migrating southward with the
onset of winter. A minor spawning season extend-
ing from September to November off the coast of
Georgia and Florida (Collins and Stender 1987)
involves fish resident to the South Atlantic Bight
(Kendall and Walford 1979).
The purpose of this study is to evaluate Kendall
and Walford's hypothesis by back-calculating the
spawn dates of young-of-the-year (YOY) bluefish
that have recruited to inshore waters, from the
number of grov^i;h increments in their otoliths.
First, we describe the timing and pattern of re-
cruitment of YOY bluefish to one segment of the
mid-Atlantic coastline: Long Island, NY. If
spawning is episodic, and if YOY bluefish from
each spavining period enter New York waters,
then length-frequency distributions of field col-
lections should be multimodal. Next, we verify
that otolith increment deposition has a daily peri-
237
FISHERY BULLETIN: VOL. 86, NO. 2
odicity in P. saltatrix. Finally, the spawning sea-
son(s) of YOY bluefish recruiting to New York is
determined by ageing and back-calculating to the
date of first ring deposition.
METHODS
Seine Collections
The temporal abundance and length-frequency
distribution of YOY bluefish was estimated by
seining 2-4 times per month from April to Octo-
ber at several sites on Long Island and in the
Hudson River (Fig. 1). In 1985 and 1986, three
sites in Great South Bay on the south shore of
Long Island were sampled: Smith Point County
Park, Fireplace Neck, and the Carmans River.
Seining was conducted with a 0.6 cm mesh, 30 m
net set from shore, either on foot or from a small
boat. Water temperature was recorded at each
site and date. In 1986, a site on the north shore of
Long Island, Setauket Harbor, was also sampled.
A few samples were taken in the fall by angling
with rod and reel. All specimens were frozen for
later measurement of fork length (FL) and
weight, and extraction of otoliths.
Additional specimens captured in 1986 from
Jamaica Bay and the Hudson River were pro-
vided by the New York Department of Environ-
mental Conservation (NYDEC). Their sampling
was conducted with a 60 m seine (1.2 cm mesh) set
from a boat.
Otolith Preparation and Analysis
The sagittae were mounted concave side down
on a glass microscope slide with cyanoacrylate
(instant glue). Two layers of masking tape were
applied on either side of the otolith. The slide was
then turned upside-down and sanded on a strip of
wet 1200 grit wet-dry sandpaper. The masking
tape ensured that the otolith was sectioned on a
consistent plane and helped prevent grinding
past the nucleus. Once the nucleus was reached,
the otolith was polished on wet felt, using levi-
gated alumina polishing compound. Three repli-
cate counts of each otolith were made under a
Zeiss'* compound microscope with transmitted po-
larized light at 125-312X. If the three counts dif-
fered by more than 10% (which occurred in about
1 out of every 10 otoliths), an additional count
was made and the outlier discarded. The three
final counts were then averaged. The total length
of each otolith was measured (nearest 0.1 mm)
with a dissecting microscope using an ocular
micrometer.
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
7 4 "loo'
73° |00
i
N
SOUTH BAY
JAMAICA BAY
ATLANTIC OCEAN
73°]00'
72 °|00'
Figure 1. — Map of the study area with sampling locations cis indicated.
238
NYMAN and CONOVER: YOUNG-OF-THE-YEAR BLUEFISH
Frequency of Ring Deposition
The frequency of growth ring deposition was
determined by marking the otoliths of fish with
tetracycline and then subsampling the marked
fish at various periods of time thereafter (Cam-
pana and Neilson 1982). Sixty YOY bluefish (7-
10 cm FL) were captured by seine in Flax Pond,
Old Field, NY (Fig. 1) and were transported to the
Flax Pond Laboratory of SUNY Stony Brook. The
fish were anesthetized in a solution of MS-222 (30
mg/L) and given an intraperitoneal injection of
tetracycline (100 mg/kg offish). After injection,
all fish were placed in a 1.3 x 1.3 m cylindrical
floating cage constructed out of 5 mm plastic
mesh and anchored in Flax Pond. The fish were
fed chopped Menidia menidia twice a day, and
dead bluefish were removed daily. Samples of 5-
10 healthy fish were periodically taken from the
cage using a dip net and frozen until the otoliths
could be excised. The experiment was terminated
61 days after the injections.
After preparation as described above, the
tetracycline-treated otoliths were viewed on a
Zeiss compound microscope using reflected ultra-
violet (UV) light at 160-400 X. Tetracycline fluo-
resces upon exposure to UV light, thus enabling
the location of the marked ring to be determined.
The UV light was then turned off, and the num-
ber of rings from the mark to the edge of the
otolith was counted under transmitted white
light. Each otolith preparation was coded so that
the reader did not know the true age. Three repli-
cate counts were conducted on each otolith.
RESULTS
Temporal Abundance and Length
Frequency
Great South Bay
The appearance of YOY bluefish in the shore
zone was abrupt in both years of the study. In
1985, no YOY bluefish were caught in weekly
samples until 28 May when a catch per unit effort
(CPUE = no. fish per seine haul) of 14.0 was
recorded (Fig. 2a). Corresponding water tempera-
ture was about 20°C. CPUE declined steadily
thereafter through October with two exceptions:
the large collections on 10 and 28 July were each
due to an unusually large number offish in single
seine hauls in the Carmans River. In 1986, YOY
bluefish were first caught on 10 June when the
water temperature was 24°C. The maximum
CPUE (45.3) was obtained on 16 June and was
followed by a decrease in CPUE in subsequent
collections (Fig. 2b).
Length-frequency distributions in 1985 showed
the progression of a single mode through mid-
August (Fig. 3a). Newly recruited fish in late May
were 3-6 cm FL. Subsequent samples showed an
increase in the mean and range of fish lengths,
probably due to somatic growi;h of the initial re-
cruits. There was no evidence of new 3-6 cm re-
cruits entering the shore zone later in the year
(Fig. 3a). Although seining continued until
November, very few YOY bluefish were caught
after August. An additional sample {n = 8) taken
on 16 September by angling from a pier on Great
South Bay had a mean fork length of 17.8 cm and
a range of 14.6-19.5 cm. Length-frequency data
from 1986 (Fig. 3b) show a very similar pattern to
that in 1985: a single length mode appears in
June and these fish increase in size through the
summer. Few YOY bluefish were caught in Au-
gust, September, or October.
Size at recruitment to the shore zone was simi-
lar in both years of our study: mean length of the
1985 and 1986 year classes at first appearance in
the shore zone was 4.6 and 4.5 cm respectively
(Fig. 4). However, because the 1986 year class
first appeared in the shore zone two weeks later
than did the 1985 year class (Fig. 2), the mean
lengths of 1986 year class were less than those of
1985 on comparable dates in June and early July.
By mid-July, however, this difference in mean
length of the two year classes was no longer ap-
parent. Both year classes reached a size of about
13-14 cm by late August when they rarely ap-
peared in our seine collections.
Setauket Harbor
In 1986, the YOY bluefish did not appear on the
north shore of Long Island at Setauket Harbor
until 3-6 weeks after they first appeared in Great
South Bay. Collections at Setauket Harbor were
small at first with only one individual being
caught on 1 July and three on 8 July. It was not
until 22 July that catches similar in number to
those in Great South Bay were being obtained.
These fish had similar mean lengths (10.2 cm,
n = 87, on 22 July; 11.9 cm, n = 22, on 5 August;
13.9 cm, n = 17, on 20 August) to those on com-
parable dates from Great South Bay (of. Fig. 4).
Length-frequency distributions by date were uni-
modal.
239
a.
^ 20
UJ
'^ 16
X
^ 12
UJ
CL
a
FISHERY BULLETIN: VOL. 86, NO. 2
(12)
(2)
-(2)(2)(4)(3)
0 0 0 0
(3)
(13)
T
(7)
(9)
(5)
(7)
3)
(7)
(7)
(10)
(6) 0
MAY
1 \ 1
JUNE JULY AUG SEPT OCT
FlGlTRE 2.— Catch per unit effort (CPUE) of YOY bluefish from
Great South Bay, NY, plotted with the mean water temperature
Jamaica Bay and the Hudson River
The length-frequency distributions of YOY
bluefish from Jamaica Bay in June, July, and
August were similar to those from Great South
Bay (Fig. 5). Fish lengths in June were unimodal.
Subsequent collections contained progressively
larger fish that were also unimodal in length dis-
tribution.
Sampling in the Hudson River began on 16
July 1986 and continued through 8 October. The
size ranges of YOY bluefish in the July and Au-
gust samples were similar to those from Great
South Bay, although the length distribution on 30
July appears bimodal (Fig. 6). The length distri-
butions from the 10 and 23 September collections
were especially broad. In particular, the 23 Sep-
tember sample contained a group offish that were
much smaller (10-14 cm FL) than the mean size
at this time in Great South Bay (Fig. 4), together
with a second group of larger fish that correspond
more closely in size with those collected else-
where (18-24 cm).
Frequency of Ring Deposition
In tetracycline-injected YOY bluefish, the
number of rings beyond the tetracycline mark (Y)
and the number of days after injection (X) had a
1:1 correspondence (Fig. 7). The relationship was
described by the equation Y = 0.97 LY - 0.287
in = 27, r = 0.996). The slope did not differ signif-
icantly from 1.0 (^test, P > 0.1).
Growth rate of the caged fish was slightly
greater than that of field fish and survival was
high (80%) with mortalities occurring only in the
first few days of the 61-d experiment. The in-
crease in mean fork length was 1.7 mm/day
among the caged fish, as compared with about 1.3
mm/day for fish from field collections during the
same time period (Fig. 4). Hence, the caged fish
did not appeEir to be adversely affected by confine-
ment.
240
NYMAN and CONOVER: YOUNG-OF-THE-YEAR BLUEFISH
b. '°'-
^ 26
o
o
a. 22
liJ
18
< 14
10,
J-
^ 20|-
UJ
UJ
en 16
iZ 12
3
a
(3)
45.3
J
(9)
"(12) (8) (10(3)
0 OOP
— I r
(3)
(10)
(10)
(10)
(10)
(13)
(7)
(e)(4)
MAY
JUNE
1 r
JULY AUG SEPT OCT
Figure 2. — Continued — on each sampling date, (a) 1985;
(b) 1986. Number of seine hauls is in ().
Back-Calculated Date of
First Ring Deposition
A representative sample of 169 YOY bluefish
(n - 88 from 1985, n = 81 from 1986) captured in
Great South Bay were aged by counting the total
number of otolith rings. The date of first ring
deposition for each aged fish was then calculated
based on the date of capture. In both 1985 and
1986, the dates of first ring deposition for YOY
bluefish were predominantly in March and April
(Fig. 8a, b).
Four fish from each of the two apparent length
modes in the 30 July collection from the Hudson
River (Fig. 6) were aged to determine if these
represented a difference in spawning season. The
fish examined were 7.8—13.8 cm in size, and back-
calculated dates of first ring deposition extended
from 7 to 30 April. Hence, these fish could all be
attributed to the same spring spawning period as
those from the south shore of Long Island.
However, YOY bluefish from the smaller (10-
14 cm) size class caught on 23 September in the
Hudson River (Fig. 6) were also aged and their
back-calculated dates of first ring deposition were
found to be predominantly in June and July, and
to a lesser extent in May (Fig. 8c). These dates
differed greatly from those offish captured earlier
in the year in the Hudson River, and along the
south shore of Long Island.
The relationship of ring number and fork
length for each year was best described by the
following equations: Y = 132.308X - 29.890 in
1985 and Y = 95.532X + 1.186 in 1986, where X
is log fork length and Y is the number of rings
(Fig. 9). The slopes of these regressions differed
significantly (ANCOVA, P < 0.001). Total otolith
length and fork length were highly correlated
(r > 0.99) and increased isometrically. Total
otolith length and ring number also had a high
correlation (r = 0.91).
241
FISHERY BULLETIN: VOL. 86, NO. 2
a.
o
'Z.
UJ
Z)
o
LU
15;
10;
28 MAY
85
5 \
II 1
32 ■
22;
24 JUNE
1
85
12 ;
1
.
40
10 JULY
85
20
Jl.
15
10;
2 8 JULY
85
5 '■
II
. •■iklMI
1,
15 1
10;
2 AUG
85
5 \
■ Li
3 6 9 12 15 18 21 24
FORK LENGTH (cm)
>-
o
LU
Z)
o
UJ
(T
l_L_
5
10
10 JUNE 86
22 JULY 86
20 AUG 86
I illJUili , I
X
3 6 9 12 15 18 21 24
FORK LENGTH (cm)
Figure 3.— Length-frequency histograms for YOY bluefish (no.) from Great South Bay, NY. (a) 1985; (b) 1986.
242
NYMAN and CONOVER:
YOUNG-OF-THE-YEAR BLUEFISH
20
p
18
-
/
16
-
,/
E
1 /
/
o
14
—
/
X
>
-7^-—
1-
12
—
<>
/
z
V
X
/
llJ
./
>
_l
10
—
1
A
^
J
o
8
-
1^
^
0-
V i/
z
6
-
/♦^ — ¥
<
^
• = 1985
UJ
4
2
1
1
• = 1986
1 1 1
JUNE
JULY
AUG SEPT
DATE
OF
COLLECTION
Figure 4. — Mean fork length (cm) of YOY bluefish (no.) captured with a beach seine in
Great South Bay in 1985 and 1986. Vertical bars are 95% confidence intervals. The last
sample in 1986 was caught by angling.
60
5
01 AUG
3 6 9 12 15 18 21 24 3 6 9 12 15 18 21 24
FORK LENGTH (cm)
Figure 5. — Length-frequency distributions for YOY bluefish (no.) fi^m Jamaica Bay, NY, 1986.
243
FISHERY BULLETIN: VOL. 86, NO. 2
10
30 JULY
*T*- -T
13 AUG
• f "-y
0 SEPT
.uIL
ii iJ I
23 SEPT
3 6 9 12 15 18 21 24
FORK LENGTH (cm)
The analysis was then expanded to the remain-
ing samples of YOY bluefish from Great South
Bay that had not been aged. The above length-age
equations were used to estimate date of first ring
deposition from the dates of capture for all YOY
captured in each year of sampling from Great
South Bay. This exercise revealed that the vast
majority of YOY bluefish in our collections from
Long Island had dates of first ring deposition in
late March, April, and early May (Fig. 10). The
weighted mean date of first ring deposition was 8
April 1985 and 14 April 1986.
The age-length equations for YOY bluefish
from Great South Bay were not applied to collec-
tions from Jamaica Bay or the Hudson River. Pre-
liminary analyses suggested that the age-length
equation for fish from the Hudson River differs
substantially from those in Great South Bay,
probably owing to a difference in growth rate.
Geographic variation in the pattern of recruit-
ment and in the age-length relationships of YOY
bluefish are being further investigated.
DISCUSSION
Recruitment of YOY Bluefish
to New York
In both 1985 and 1986, the arrival of YOY blue-
fish on the south shore of Long Island was abrupt.
Within about a 1-wk period, CPUE went from 0.0
to 14-18 fish/seine haul. CPUE then remained
high for the next two months until declining in
August and September when the fish probably
became too large to be efficiently sampled by our
techniques. These data suggest that the YOY
bluefish recruit to the shore zone as a sudden
pulse. The timing of this recruitment event is ap-
parently variable, differing by about two weeks
among the two years of our study. The appear-
ance of fish 3-6 weeks earlier on the south shore
(Great South Bay) than on the north shore (Se-
tauket Harbor) of Long Island suggested that
these fish arrive from offshore waters to the
south.
Temperature probably influenced the time of
arrival of YOY bluefish in the shore zone. In both
years of our study, YOY bluefish appeared as
temperatures reached about 20°-24°C. In Octo-
ber, after temperatures dropped to the middle
Figure 6. — Length-frequency distributions for YOY bluefish
(no.) from the Hudson River, NY, 1986.
244
NYMAN and CONOVER: YOUNG-OF-THE-YEAR BLUEFISH
Figure 7. — Relation between number of days
since marking with tetracycline and the number
of growth increments beyond the mark in YOY
bluefish maintained in a field cage.
q:
<
2
O
>-
UJ
CD
(n
I-
z
UJ
70
60
50
40 -
30
S 20
u
z
10
Y= 0.97IX- 0.287
r = 0.996
1
1
1
1
1
10 20 30 40 50 60
DAYS SINCE INITIAL MARKING
70
O
UJ
o
UJ
on
u.
Figure 8. — Back-calculated date of first ring deposi-
tion for YOY bluefish (no.) as determined by count-
ing daily growth rings, (a) and (b) represent fish
from Great South Bay in 1985 and 1986, respec-
tively, (c) is for fish seined from the Hudson River
on 23 September 1986.
MAR APR MAY JUN JUL MAR APR MAY JUN JUL
DATE OF FIRST RING DEPOSITION
245
FISHERY BULLETIN: VOL. 86, NO. 2
180 -
160 -
■
140 -
■
Ul
CD
Z
1 — 1
cr
Ll
o
o
z
120 -
100 -
80 -
60 -
40 -
20 -
1986
X
■ Xx , X %^^f0^
1985
**^ ^ X
^x
■ 1985
X 1986
0 -
1
1 1 1 1 r 1
0.4 0.6 0.8 10 12
LOG^O FORK LENGTH (cm)
Figure 9. — Relation between log fork length (X) and number of otolith rings (7) for YOY
bluefish from Great South Bay, NY. The regression equations are 7 = 132.308X - 29.890
(n = 88) for 1985 and y = 95.532A: + 1.186 (n = 81) in 1986.
teens, we no longer captured YOY bluefish. Oben
(1957) noted that in the Black Sea, YOY bluefish
appeared in the shore zone at temperatures of
18°-24.5°C, and left the shore zone in October and
November when temperatures dropped to 13°-
15°C.
Length-frequency distributions of YOY blue-
fish from the south shore samples showed only a
single mode that attained progressively larger
size through the summer and fall, and corre-
sponded to the initial recruitment offish. If multi-
ple spawning and recruitment events contributed
YOY bluefish to Long Island, multimodal length-
frequency distributions should have been ob-
served. The unimodal distributions suggested
that only one spawning period contributed the
majority of YOY bluefish to Long Island.
Interannual variation in the length-age rela-
tionship of YOY bluefish was observed. Although
recruitment occurred two weeks earlier in 1985
than in 1986, the empirical mean lengths at re-
cruitment were similar (Fig. 4). Postrecruitment
growth, however, was slower in 1985 than 1986 so
that empirical mean lengths became similar by
mid-July. Correspondingly, the slope of the
length-age regressions differed significantly
among years: YOY bluefish at an age of about
50-70 days had greater fork lengths in 1985 than
in 1986 (Fig. 9), but the reverse was true among
older, larger fish. Apparently, the growth rate of
YOY bluefish prior to recruitment was higher in
1985 than 1986, but this pattern among the two
years was reversed during the period of postre-
cruitment growth.
Validation of Daily Otolith Rings
Our experimental results demonstrate that
otolith ring deposition is daily in YOY bluefish. A
regression slope of 0.971 indicates a one-to-one
correspondence between number of days after in-
jection and the number of rings beyond the tetra-
cycline mark. This outcome is not particularly
surprising because numerous studies have shown
that increment production is daily, particularly
in the early life history when somatic growth is
rapid (Brothers et al. 1976; Campana and Neilson
1985; Jones 1985). Cases where ring periodicity is
reportedly less-than-daily have involved subopti-
mal growing conditions (Geffen 1982, 1987; Rice
et al. 1985). In our study, the confinement of YOY
bluefish in a field cage apparently had little effect
on growth rate, or the production of daily growth
increments. The field cage allowed for natural
light, temperature, and salinity variations that
the fish would normally have experienced in na-
246
NYMAN and CONOVER: YOUNG-OF-THE-YEAR BLUEFISH
O
u
c:
CD
Z)
cr
QJ
o
u
d
CD
Z)
O"
QJ
30
20
10
0
60
50 -
40 -
30 -
20 -
10 -
0
150
180
Julian Date of First Ring Deposition
Figure 10. — Estimated date of first ring deposition for all YOY bluefish caught in Great
South Bay in 1985 (n = 561) and in 1986 {n = 868) using the respective age-length equa-
tions in Fig. 9.
ture. Feeding periodicity was probably the pri-
mary artifact of confinement that could have af-
fected the rate of ring production in caged fish.
However, Marshall and Parker (1982) showed
that feeding periodicity did not significantly af-
fect ring production in sockeye salmon,
Oncorhynchus nerka.
We were unable to determine directly the num-
ber of days between spawning and first ring depo-
sition because numerous attempts to capture
running-ripe females for initiating experiments
on eggs and larvae were unsuccessful. However,
most species of fish deposit the first daily growth
increment within a few days of hatching (Broth-
ers et al. 1976; Radtke and Dean 1982; McGurk
1984; Radtke 1984; Davis et al. 1985). Recent ev-
idence suggests this is also true in bluefish. Lar-
vae captured off Long Island in 1987 had about
seven otolith increments at a total body length of
4-5 mm (R. K. Cowen and D. O. Conover, unpubl.
data). Based on the rate of development at 20°C in
the laboratory observed by Deuel et al. (1966),
247
FISHERY BULLETIN: VOL. 86, NO. 2
larvae would reach this size in about 5-7 days
posthatch. If so, first ring deposition would
roughly coincide with hatching. Bluefish hatch in
48 hours at 20°C (Deuel et al. 1966), so the day of
first ring deposition probably follows the date of
spawning by about 2-4 days.
Spawning Seasons
Along the Atlantic Coast
Published studies of larval bluefish distribu-
tions along the Atlantic coast suggest the exis-
tence of three temporally and spatially distinct
spawning seasons: spring and fall spawning sea-
sons in the South Atlantic Bight and a midsum-
mer spawning in the Middle Atlantic Bight. In
the only synoptic study covering most of the U.S.
east coast, Kendall and Walford (1979) described
two periods of high larval abundance: One peak
occurred in March and April on the outer shelf of
the South Atlantic Bight, and the other peak was
in July and August midway over the continental
shelf of the Middle Atlantic Bight. Subsequently,
Powles (1981) and Collins and Stender (1987) also
found the highest abundance of bluefish larvae in
the South Atlantic Bight (Cape Canaveral to
Cape Fear) to be in April and May. Collins and
Stender, however, noted the existence of a lesser
peak in larval abundance during September-
November. This fall spawning season in the
South Atlantic Bight was further confirmed by
Finucane and Collins (in press) based on the
gonad condition of bluefish from Georgia and the
Carolinas. In the Middle Atlantic Bight off Vir-
ginia, Norcross et al. (1974) found that eggs and
larvae of bluefish first appeared in June, peaked
in abundance in July, and persisted into August.
Similar observations on the timing of the summer
spawning season in the Middle Atlantic Bight
were presented by Sherman et al. (1984) and
Morse et al. (1987).
Lassiter (1962) provided additional evidence
of the existence of relatively discrete spawning
seasons in bluefish. He showed that the dis-
tribution of back-calculated lengths at age one
has a distinctly bimodal pattern among adult
fish from North Carolina. Size at age 1 tended
to be either about 14 cm or 28 cm. Lassiter
showed that the bimodal pattern could not be
explained as a difference in growth rate, and
suggested that there must be two distinct
spawning seasons such that one group of fish had
a first growing season about twice as long as the
other.
Spawn Dates of YOY Bluefish
from New York
Back-calculation to the day of first ring deposi-
tion for YOY bluefish recruiting to Great South
Bay in 1985 and 1986 demonstrated that these
fish were spawned primarily in March and April
(Figs. 8, 10). Fish that were spawned in July-
August were rarely captured by us on Long Island
in 1985 or 1986, despite continued sampling into
October.
Recruitment to Jamaica Bay and the Hudson
River in July and August 1986 involved YOY
bluefish of about the same size as those fi-om
Great South Bay. Though the size range of fish
fi-om the Hudson was slightly greater than those
from Long Island, fish aged from each of the two
modes appearing in the July Hudson River sam-
ples (Fig. 6) were all spawned during April within
about three weeks of each other. The apparent
bimodality in July is probably a sampling arti-
fact. Hence, Jamaica Bay and Hudson River fish
collected in July and August can be attributed to
the same spawning season as those fi^om Great
South Bay.
Length-frequency distributions ft"om the Hud-
son River in September, however, contained a
group of unusually small bluefish, and back-
calculation showed that they were spawned pre-
dominately in June and July (Fig. 8c). These fish
probably resulted fi-om the summer spawning
season in the Middle Atlantic Bight. Examina-
tion of gonads from adult fish captured during
1986 suggested that the running-ripe males and
mature females were most abundant during late
June and July off Long Island (L. Chiarella and
D. O. Conover, unpubl. data). Hence, at least,
some summer-spawned YOY bluefish do recruit
to the shore zone of the Middle Atlantic Bight.
They were, however, much less abundant than
spring-spawned YOY bluefish in our 1985 or 1986
samples.
Spawning by bluefish in the spring is known to
occur only in the South Atlantic Bight (Kendall
and Walford 1979; Collins and Stender 1987).
Water temperatures over the shelf north of Cape
Hatteras are probably too low for bluefish to
spawn in March and April: average shelf water
temperatures in the Middle Atlantic Bight range
from 5° to 14°C in March and April (Ingham
1986). Virtually no eggs and larvae (Morse et al.
1987) and comparatively few adult bluefish
(Gilmore 1985) are captured in plankton or trawl
surveys north of Cape Hatteras in March and
248
NYMAN and CONOVER: YOUNG-OF-THE-YEAR BLUEFISH
April. Moreover, the time of arrival of YOY blue-
fish on Long Island actually precedes the summer
spawning season in the Middle Atlantic Bight.
We therefore conclude that YOY bluefish recruit-
ing to the Middle Atlantic Bight in late spring
come from spawnings in the South Atlantic
Bight.
Larval Transport
The physical mechanisms that account for the
transport of bluefish larvae fi"om the South At-
lantic Bight to New York are not clear. Spawning
in the South Atlantic Bight occurs primarily over
the outer half of the continental shelf (Powles
1981; Collins and Stender 1987), and some larvae
may be entrained by the Gulf Stream and carried
northward into the slope waters of the Middle
Atlantic Bight (Kendall and Walford 1979).
Neuston net collections in April have shown that
bluefish larvae are periodically abundant on both
sides of the Gulf Stream-shelf water interface off
Cape Hatteras (Kendall and Walford 1979).
Collins and Stender (1987) found a negative cor-
relation between larval size and latitude in the
South Atlantic, but their sampling may not have
extended far enough north (i.e., they did not sam-
ple above Cape Fear).
If the Gulf Stream is responsible for the north-
ward transport, a mechanism by which larvae
avoid being advected too far offshore would ap-
pear to be necessary. According to our results, the
interval between spawning and recruitment to
Long Island is about 45-60 days, whereas the
surface flow of the Gulf Stream at lat. 36°N is
about 104 km/day (Iselin 1936). Hence, larvae re-
maining in the Gulf Stream for an extended pe-
riod would be transported far off the shelf. Reten-
tion near the shelf could be achieved by entering
the slope waters at an appropriate time.
The abrupt appearance of YOY bluefish in the
shore zone suggests that the onshore migration is
a temporally distinct event, perhaps triggered by
vernal warming of the shelf. Because the circula-
tion of the slope and shelf waters of the Middle
Atlantic Bight is toward the southwest (Sherman
et al. 1984), the cross-shelf migration must to
some extent involve active swimming.
Very few summer-spawned YOY bluefish were
captured in our study. This may not be surpris-
ing, however, because the prevailing currents
over the midshelf off Long Island would carry lar-
vae to the southwest. If so, summer-spawned fish
would be found along the coast fi"om approxi-
mately New Jersey to Cape Hatteras. We caution,
however, against any general conclusion concern-
ing the lack of summer-spawned fish in New
York. There could, for example, be substantial
year-to-year variation in the recruitment level of
spring- and summer-spawned fish along any par-
ticular segment of the U.S. coast. These issues are
now being examined by extending our sampling
to southern latitudes.
ACKNOWLEDGMENTS
We thank Robert Cerrato, Robert K. Cowen,
and Peter Woodhead for reviewing the manu-
script and Stephen Heins, Melanie Meade and
Louis Chiarella for assistance in the field. Byron
Young and Kim McKown of the NYDEC gra-
ciously provided samples from the Hudson River
and Jamaica Bay. An earlier version of this paper
was submitted by R.M.N, to the Graduate School
of the State University of New York at Stony
Brook in partial fulfillment of the requirements
for an M.S. degree in Marine Environmental Sci-
ences. Initial funding was provided by grants
from the Sport Fishery Research Foundation
(D.O.C./R.M.N.) and the Montauk Marine Basin.
Later funding was provided by grants to D.O.C.
fi-om the NYDEC through the Dingell-Johnson
Federal Aid in Sport Fish Restoration Act and
by the New York Sea Grant Institute through
the NOAA Office of Sea Grant, U.S. Department
of Commerce, under Grant No. NA86AA-D-
SG045.
LITERATURE CITED
BiGELOW, H. B., AhfD W. C. SCHROEDER.
1953. Fishes of the Gulf of Maine. U.S. Fish and Wildlife
Service, Fish. Bull. 53:1-577.
Brothers. E. B, C P. Mathews, and R. Lasker.
1976. Daily growth increments in otoliths from larval and
adult fishes. Fish. Bull., U.S. 74:1-8.
Campana, S. E., and J. D. Neilson.
1982. Daily growth increments in otoliths of starry floun-
der (Platichthys stellatus ) and the influence of some envi-
ronmental variables in their production. Can. J. Fish.
Aquat. Sci. 39:937-942.
1985. Microstructure of fish otoliths. Can. J. Fish.
Aquat. Sci. 42:1014-1032.
Collins, M. R., and B. W. Stender.
1987. Larval king mackerel Scomberomorus cavalla,
Spanish mackerel (S. maculatus), and bluefish (Po-
matomus saltatrix ) off the southeast coast of the United
States, 1973-1980. Bull. Mar. Sci. 41:822-834.
Davis, R D., T. W. Storck. and S. J Miller.
1985. Daily growth increments in the otoliths of young-of-
the-year gizzard shad. Trans. Am. Fish. Soc. 114:304-
306.
249
FISHERY BULLETIN: VOL. 86, NO. 2
Deuel. D G , J R. Clark, and A J Mansueti
1966. Description of embryonic and early larval stages of
bluefish, Pomatomus saltatrix. Trans. Am. Fish. Soc.
95:264-271.
FiNUCANE, J. H., AND L A. COLLINS.
In press. Reproductive biology of bluefish, Pomatomus
saltatrix, from the southeastern United States. North-
east Gulf Sci.
Geffen, A J
1982. Otolith ring deposition in relation to growth rate in
herring (Clupea harengus) and turbot (Scophthalmus
maximus) larvae. Mar. Biol. (Berl.) 71:317-326.
1987. Methods of validating daily increment deposition in
otoliths of larval fish. In R. C. Summerfelt and G. E.
Hall (editors). Age and growth offish, p. 223-240. Iowa
State Univ. Press, Ames.
Gilmore, J.
1985. Oceanic distribution, abundance, and migration of
bluefish along the east coast of the United States. M.S.
Thesis, State University of New York, Stony Brook,
60 p.
Ingham, M C
1986. Sea surface temi>eratures in the northwestern At-
lantic in 1985. NAFO SCR Doc. 86/75.
ISELIN, C. O'D.
1936. A study of the circulation of the western North At-
lantic. Pap. Phys. Oceanogr. Meteorol. 4:1-101.
Jones, C. J.
1985. Determining age of larval fish with the otolith in-
crement technique. Fish. Bull., U.S. 84:91-103.
KENDALL, A. W , Jr. and L. a. Walford
1979. Sources and distribution of bluefish, Pomatomus
saltatrix, larvae and juveniles off the east coast of the
United States. Fish. Bull., U.S. 77:213-227.
Lassiter, R. R.
1962. Life history aspects of the bluefish, Pomatomus
saltatrix (Linnaeus), fi-om the coast of North Carolina.
M.S. Thesis, North Carolina State College, Raleigh, NC,
91 p.
Marshall. S L , and S. S. Parker.
1982. Pattern identification in the microstructure of sock-
eye salmon (Oncorhynchus nerka) otoliths. Can. J.
Fish. Aquat. Sci. 39:542-547.
McGURK, M. D.
1984. Ring deposition in the otoliths of larval Pacific her-
ring, Clupea harengus pallasi . Fish. Bull., U.S. 82:113-
120.
Morse, W W , M P Fahay, and W. G. Smith
1987. MARMAP surveys of the continental shelf from
Cape Hatteras, North Carolina, to Cape Sable, Nova Sco-
tia (1977-1984). Atlas No. 2. Annual distribution pat-
terns of fish larvae. U.S. Dep. Commer., NOAA Tech.
Memo. NMFS-F/NEC-47, 215 p.
NoRCROss, J J , S L Richardson, W H Massmann, and E. B.
Joseph
1974. Development of young bluefish (.Pomatomus salta-
trix ) and distribution of eggs and young in Virginian
coastal waters. Trans. Am. Fish. Soc. 103:477-497.
Oben, L C
1957. About the drifting approach of fingerling bluefish
(Pomatomus saltatrix) (Linnaeus) to the shores of the
Black Sea in the Karadaga region (1947-1954). Kara-
dagskaia Biol. Stn. Tr. (4):155-157.
POWLES, H.
1981. Distribution emd movements of neustonic young of
estuarine dependent iMugil spp., Pomatomus saltatrix)
and estuarine independent (Coryphaena spp.) fishes off
the southeastern United States. Rapp. P. -v. R^un.
Cons. int. Explor. Mer 178:207-209.
Radtke, R L
1984. Formation and structural composition of larval
striped mullet otoliths. Trans. Am. Fish. Soc. 113:186-
191.
Radtke, R L., and J M Dean.
1982. Increment formation in the otoliths of embryos, lar-
vae and juveniles of the mummichog, Fundulus heterocli-
tus. Fish. Bull., U.S. 80:201-215.
Rice, J A , L B Crowder. and F P Binkowski.
1985. Evaluating otolith analysis for bloater Coregonus
hoyi: do otoliths ring true? Trans. Am. Fish. Soc.
114:532-539.
Sherman, K , W. Smith, W Morse, M Berman, J Green, and
L EllSYMONT.
1984. Spawning strategies of fishes in relation to circula-
tion, phytoplankton production, and pulses in zooplank-
ton off the northeastern United States. Mar. Ecol. Prog.
Ser. 18:1-19.
US. Department of Commerce.
1986. Recreational fishery statistics for 1985. U.S. Dep.
Commer., NOAA, NMFS, 130 p.
WILK, S J
1977. Biological £md fisheries data on bluefish, Po-
matomus saltatrix (Linnaeus). NMFS, NEFC, Sandy
Hook Lab. Tech. Ser. Rep. No. 11, 56 p.
250
ON THE ROLE OF FOOD-SEEKING IN THE SUPRABENTHIC HABIT
OF LARVAL WHITE CROAKER, GENYONEMUS LINEATUS
(PISCES: SCIAENIDAE)
A. E. Jahn, D. M Gadomski, and M. L Sowby'
ABSTRACT
Fish larvae and their prey were sampled from discrete depths within the bottom meter and at
middepth near the 15 m depth contour off southern California. The smallest white croaker larvae
(<2.7 mm NL) occurred mostly at middepth. Mid-sized larvae (2.7 mm to the beginning of flexion)
were almost all collected at the two depths nearest the bottom. All preflexion-stage larvae ate small
(50-300 (im in length) prey, chiefly rotifers, copepod nauplii, tintinnids, and invertebrate eggs.
Although small and mid-size larvae ate these items in different proportions, this difference could not
be ascribed to vertical distribution. Diet of the largest larvae, flexion and postflexion (roughly 5-15
mm), consisted mainly of copepods and differed by >90% from diets of smaller larvae. Though largest
larvae were only captured 50 cm above the bottom, their prey, with one exception (amphipods), were
more abundant at or above 1 m. It was concluded that the observed suprabenthic concentration of
older white croaker larvae was probably not motivated by food-seeking.
Disparity between concentrations of food re-
quired for survival and growth of laboratory-
reared fish larvae and observations of average
concentrations of food organisms in the ocean has
led to the widely accepted idea that aggregations
offish larvae and their food must frequently over-
lap in nature (see reviews by Theilacker and
Dorsey [1980] and Hunter [1981]). Direct and in-
direct evidence for the importance of overlapping
concentrations of larvae and their prey (Lasker
1975, 1978; Govoni et al. 1985; Buckley and
Lough 1987) comes from sampling at fronts and
discontinuities in the pelagic environment. One
interface that attracts many zooplankters is the
seabed itself (Hamner and Carleton 1979; Wish-
ner 1980; Sainte-Marie and Brunei 1985). On the
southern California continental shelf, the seabed
serves as a surface of aggregation for larvae of
numerous fish species (Brewer et al. 1981; Schlot-
terbeck and Connally 1982; Barnett et al. 1984;
Jahn and Lavenberg 1986) and other zooplankton
(Clutter 1969; Barnett and Jahn 1987) and of
large-zooplankton biomass (Jahn and Lavenberg
1986). While it is tempting to suggest a trophic
advantage to the suprabenthic habit of the fish
larvae, near-bottom concentrations of organisms
actually eaten by larval fishes have yet to be
demonstrated along the open coast.
iNatural History Museum of Los Angeles County, 900 Expo-
sition Boulevard, Los Angeles, CA 90007.
In all cases reported, concentration in the near-
bottom zone was greater in older larvae and,
when observations permitted, greater during the
day than at night (Brewer and Kleppel 1986;
Jahn and Lavenberg 1986). The phenomenon is
therefore thought to be behavioral. Possible ad-
vantages of such behavior, including avoidance of
midwater predators, maintenance of position on
the shelf, and increased encounters with high
concentrations of food, have been discussed else-
where (Barnett et al. 1984; Brewer et al. 1984;
Brewer and Kleppel 1986; Jahn and Lavenberg
1986). In discussing the near-bottom schooling
behavior of a larval clupeoid in Japan, Leis (1986)
stated, "knowledge of the biology of epibenthic
fish larvae is too rudimentary to allow a clear
assessment of the advantages and disadvan-
tages. . . ." Whatever the advantages, a seemingly
more answerable question about the near-bottom
habit is what causes the larvae to behave as they
do? In another study from Japan, Tanaka (1985)
showed that juvenile red sea bream, Pagrus
major, exploited suprabenthic copepod popula-
tions, and he speculated that the distribution of
prey was a template for the descent of the fish
from midwaters and its subsequent migration
into estuaries. The question addressed in the
present study was whether the fine-scale layering
of larval fishes was a direct response to that of
their prey field.
Because of the immediate behavioral aspect of
Manuscript accepted Februar>' 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.
251
FISHERY BULLETIN: VOL 86, NO 2
the question posed, a 1-d study was thought
appropriate. Though environmental conditions
on this day might differ from "average", fish lar-
vae were assumed to be capable of a constant
array of behaviors. In other words, response (if
any) of the larvae to the vertical distribution of
their prey was assumed to be a deterministic
rather than a statistical phenomenon. If their
vertical distribution resembled that of their prey,
then food-seeking would remain a plausible ex-
planation for the near-bottom habit; if not, then
other stimuli must be considered important in
shaping these near-bottom concentrations of fish
larvae.
The sampling was planned for daylight hours,
when most feeding by larvae was expected to
occur (Hunter 1981; Govoni et al. 1983). Late win-
ter was chosen because in this season peak larval
abundances of several species of interest to us
(northern anchovy, Engraulis mordax; white
croaker, Genyonemus lineatus; California hal-
ibut, Paralichthys californicus; and sometimes
queenfish, Seriphus politus, often overlap
(Lavenberg et al. 1986). A survey cruise in late
February found moderate-to-high abundance of
the first three species plus California sardine,
Sardinops sa^ox, (all >0.2 m"'^, Lavenberg un-
publ. data), and so this study was scheduled for 19
March 1985 off Seal Beach, CA (lat. 33°41'N,
long. 118°05'W; for a map, see Jahn and Laven-
berg 1986).
As it happened, we chanced to encounter condi-
tions that were less typical than those found on
the February cruise. Only one fish species, white
croaker, was abundant enough to merit analysis,
and an uncommonly reported prey item, rotifera,
was important for small larvae. The diet of
various-sized larvae with respect to the abund-
ance of prey organisms at an array of heights
above the seabed was nevertheless useful in ques-
tioning whether food-seeking shaped the ob-
served larval distribution.
METHODS
Field
At the hour of 0750 PST, an array of Interocean
model S4^ electromagnetic current meters was
set out over the 15 m isobath, with current meters
1, 4, and 8 m above the seabed. These meters were
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
set to record average current vectors and tem-
perature at 5-min intervals. The vessel (RV West-
wind) was then anchored some 200 m seaward of
the current meter array. A Nielson model NCH
fish pump, rated at 227 m'^ h^ at a 2 m head, was
used to sample fish larvae and zooplankton. The
end of the hose was tethered between a 200 kg flat
steel weight and several subsurface floats, with a
pulley arrangement such that divers could adjust
the distance between hose mouth and seabed. A
similar setup was previously found to give re-
peatable, fine-scale resolution at vertical sep-
arations of 25 cm (Jahn and Lavenberg 1986).
Sampling heights above the bottom were 50 cm,
1 m, and 6.7 m. The 15.2 cm diameter hose was
nearly horizontal at the tether point, so that
nominal sampling strata were z ± 7.6 cm. Ves-
sel surge, transmitted through the stiff hose,
caused occasional downward excursions of some
10 cm.
Accompanying each pump sample was a cast of
water bottles for phytoplankton and microplank-
ton analysis. Rigid arrays of horizontally held 4 L
Niskin bottles (of. Owen 1981) were used to take
water samples simultaneously from 25, 50, and
100 cm above the bottom. The bottle array was
designed to be tripped by messenger, but poor
performance led to diver-implemented use after
the second cast. A midwater sample, 7.5 m below
the surface, was obtained via a single Niskin bot-
tle for each sample set.
The sampling plan thus consisted of duplicate
pump samples from each of three strata, each
pump sample to be accompanied by a set of bottle
samples from four standard heights, three within
1 m of the seabed and one at midwater column.
One-liter samples from the bottles were fixed in
Lugol's solution for later identification of phyto-
plankton and microplankton. Pump samples of
15-min duration (approximately 35 m^) were
mainly directed into an overboard, 330 |xm mesh
plankton net for retention of large zooplankton
and ichthyoplankton. Unexpected problems in
reading an inline flowmeter required that vol-
umes be estimated as 2.4 m"^ min"^, based on pre-
vious experience with the pump under similar
conditions aboard the same vessel. To collect
smaller zooplankton, a 5 cm diameter hose led
from the intake side of the fish pump to a 100 jjtm
mesh plankton net. This small-meshed net was
suspended over a watertight box, which was
marked such that exactly 0.5 m'^ could be subsam-
pled for animals too small to be quantitatively
retained by the large net. This subsample, which
252
JAHN ET AL : FOOD-SEEKING LARVAL WHITE CROAKER
was first seived through 330 p.m mesh, took about
10 minutes to obtain; the portion retained on the
330 |jim mesh was added to the contents of the
large plankton net. All pump samples were pre-
served in 5% formalin.
Laboratory
All fish larvae and eggs were sorted from the
large zooplankton samples and identified. All
specimens of white croaker, the only species
abundant in all six collections, were measured
with an eyepiece micrometer in units of 0.024,
0.062, or 0.159 mm, depending on magnification.
Length was measured from tip of snout to end of
straight (NL) or fiexed (FL) notochord or to the
end of the hypural plate when this margin was
vertical (SL). A further designation of de-
velopmental stage indicated the amount of yolk
present: "free embryos" (Balon 1975) had a
relatively massive yolk sac and may or may not
have had functional eyes and mouths; more
advanced individuals with a much-reduced or
totally resorbed yolk sac, fully pigmented eyes,
and an apparently functional mouth were
designated "feeding-stage" larvae, or simply
"larvae".
All larvae, plus a maximum of 20 free embryos
with apparently functional mouths from each col-
lection, were dissected for gut contents analysis
by methods described in Arthur (1976) and
Gadomski and Boehlert (1984). Length, rather
than width, of prey items was measured, because
it was considered a more conservative property of
often crushed specimens and because our concern
was not so much with what the larvae could eat
(Hunter 1981) as with what they did eat. Lengths
of prey items (of copepods, cephalothorax length)
were recorded in 50 fxm classes up to 200 iJim, by
100 ji-rn classes from 200 \x.m to 1 mm, and by 0.5
mm classes at larger sizes. In a few cases, these
size categories were inconvenient, and more in-
clusive ranges were used.
Water bottle samples of phytoplankton and mi-
crozooplankton were prepared following proce-
dures in Utermohl (1931). From a thoroughly
agitated sample, a 50 mL subsample for net
phytoplankton was taken and placed in a settling
chamber overnight (about 14—18 hours). Cells
were identified and counted in 10 ocular fields,
and mean density (cells per liter) calculated as
the number counted scaled by the proportion of
the area of the 10 fields (20.6 mm^ total) to the
area of the slide (510.7 mm^).
Microzooplankton was filtered from a 500 mL
subsample onto a 35 fxm mesh screen, washed
from the screen into a 50 mL settling tube and
allowed to settle overnight. All organisms >50
fxm were counted and identified to taxon and size
category, using the same system as for larval fish
gut contents. Densities were scaled to number per
liter.
The 100 ^jLm zooplankton samples were concen-
trated to 200 mL, then subsampled twice using a
10 mL Stempel pipette. Organisms were identi-
fied and classified to size categories as described
above for larval fish prey. Counts from two sub-
samples were averaged and expressed as number
per m"^.
Data Analysis
The microzooplankton (from water bottles)
data set consisted of six vertical profiles of four
sampling heights each. Principal components
analysis was used to look for vertical layering and
time-correlated changes in the makeup of these
assemblages. A list of taxa present in three or
more samples from at least one sampling height
was chosen. Abundances were log-transformed
[logio(jc + D], and principal components com-
puted from the covariance matrix. Component
scores for each of the 24 samples were used to
make plots in which two- and three-dimensional
groupings were sought that could be clearly re-
lated to sampling height or to the sequence in
which the samples were taken. The taxa having
high loadings on axes (components) identified
with time and vertical trends were subsequently
scrutinized individually. A similar analysis was
done for phytoplankton, but omitted here in the
interest of brevity.
Gut contents were conveniently analyzed by
lumping taxa into the 10 categories: dinoflagel-
late, tintinnid, rotifer, polychaete larva, lamelli-
branch larva, crustacean nauplius, copepodite
and adult copepod, amphipod, invertebrate egg,
and "other". Unidentifiable matter was ignored
in all comparisons. To test for differences in diet
between subsets of larvae, we used an adaptation
of the "bootstrap" (Efron 1982). The test criterion
was the percentage of prey comprised by a major
item in one of the two groups of guts. The null
hypothesis that two sets were not different was
simulated by combining the two data sets and
then, through repeated sampling, determining
the probability of observing the criterion percent-
age from such a mixture.
253
FISHERY BULLETIN: VOL. 86, NO 2
RESULTS
General Observations
The water column was very weakly stratified,
with temperatures of 12.9 ± 0.1°C at 1 m, 13.0
± 0.2°C at 4 m and 14.1 ± 0.1°C 8 m above the
bottom during the time of biological sampling.
Near the bottom, a turbid suspension limited vis-
ibility to arm's length; the surface of the sediment
was never clearly seen on any of the seven de-
scents during the hours of 0930-1630. The mid-
waters below about 3 m from the surface were
densely populated with larvaceans (visually esti-
mated and later confirmed to be about 10 L"^).
Total diatom cell counts (principally Nitzschia
spp.) were of order 10^ L~^ in all samples, bloom
quantities suggestive of recent upwelling (cf.
Tont 1981).
Currents and Plankton
During the hours of biological sampling, cur-
rents ran steadily alongshore to the southeast,
being deflected counterclockwise near the bottom
and ranging from about 14 cm s"^ at 8 m to 6 m
s"^ at 1 m above the seabed. At these current
speeds, one may expect that the approximately
5-h period from beginning to end of biological
sampling should correspond to a minimum spa-
tial spread of 1-2.5 km. Distances of this order
were previously found to be an important length
scale of variation in larval fish abundance ( Jahn
and Lavenberg 1986). Because the spatial dimen-
sion of interest regarding distribution of larval
fish prey was the vertical, we needed to quantify,
at least partially, the effects of time (vertical
migration?) and distance (advection) on the com-
position and vertical dispostion of the plankton.
Component 4 (11%)
0
2 -
1 -
CM
CM
C
0)
c
o
o.
E
o
o
0 -
1 -
■2 -
FIGURE 1. — Projections of microzooplankton
samples onto the first and fourth principal com-
ponent axes. The initial digit represents profile
number, M = midwater, B = near-bottom, final
digit is proximity to bottom (1 = 25 cm, 2 = 50
cm, 3 = 100 cm), see Figure 3.
254
JAHN ET AL.: FOOD-SEEKING LARVAL WHITE CROAKER
Accordingly, the microplankton data set, repre-
senting six vertical profiles separated in time,
was reduced to principal components for exami-
nation of possible time effects.
Twenty-four taxonomic/size categories of mi-
crozooplankton were used to compute principal
component scores for the 24 samples. The first
four principal components accounted for 60% of
the variance. No clear separation of midwater
from near-bottom samples was seen. The first
component, which accounted for 22% of the vari-
ance, separated the near-bottom samples into two
groups, morning to midday and afternoon (Fig. 1),
leaving the midwater samples at intermediate
projections. The midwater samples were in turn
separated by the fourth component (11% of the
variance) into time groups corresponding to those
of the near-bottom set. No stratification by sam-
pling height was seen within the near-bottom
samples, and none of the other axes provided sep-
aration by time. The highest loading variables on
components 1 and 4 (Table 1) were various sizes of
rotifer and, for component 1, three genera of
tintinnids (Favella, Acanthocystis , and Da-
dayella ). Much of the time-correlated variance
structure depicted in Figure 1 thus appears to be
due to change in the size composition of rotifers,
described in a later section, and a decrease in
these three tintinnids near the bottom in late af-
ternoon (Table 2). An identical analysis of the
Table 1. — Loadings of important variables on tfie first and fourth
principal components of microplankton data.
Component 1
Component 4
Variable
Loading
Variable
Loading
Rotifer
150-200 ^Lm
200-300 ^JLm
Tintinnids
Favella sp.
Acanthocyctis
Dadayella sp.
sp.
0.572
-0.380
0.365
0.331
0.314
Rotifer, 100-150 M-m
Egg, 50-100 \i.m
Copepod nauplii,
150-200 ^JLm
0.388
0.349
-0.252
phytoplankton data found no trends in time or
depth.
Larval Fish Abundance
Of 1,125 total fish larvae taken in the six pump
samples, 666 (59%) were white croaker, a deep-
bodied, robust larva (Watson 1982). More than
half (338) of these had absorbed the yolk sac and
were thus of feeding size. The second most abun-
dant feeding-stage larva was an unidentified gob-
iid type (84 specimens), but this taxon was not
taken above 100 cm of the seabed and so was
excluded from the gut analysis. Feeding-stage
California sardine, northern anchovy, and Cali-
fornia halibut — all relatively abundant O0.2
m""^) in the area three weeks earlier — each repre-
sented <1% of the catch. Although the earlier
survey employed oblique bongo net tows, past
comparison of the Nielsen pump with bongo tows
found no significant differences in diversity or
abundance estimates based on similar-volume
samples (R. Schlotterbeck^). We therefore think
the differences between the February survey and
our March samples were due mainly to a real
change in the ichthyoplankton, from a typical
late winter assemblage (McGowen 1987; Walker
et al. 1987) to a more depauperate one.
Vertical Distribution and Feeding
Incidence of Larval White Croaker
White croaker free embryos ranged in abun-
dance from <0.1 m"'^ at 0.5 m to ~1 m"-^ at 1 m
to >2 m-3 at 6.7 m above the bottom. Of 61 free
embryos dissected, none had gut contents.
Feeding-stage larvae of white croaker were
only slightly more abundant at 6.7 m (1.9-2.2
m"3) than at 1 m and 0.5 m (1.1-1.6 m'^), but
3R. Schlotterbeck, Robert Schlotterbeck, Inc., 18842
Ridgeview Cr., Villa Park, CA 92667, pers. commun. April
1986.
Table 2. — Density (cells per liter) of three tintinnids as a function of time and
sampling height. Each set of three numbers gives the density of Favella spp. (F),
Acanthocystis spp. (A), and Dadayella spp. (D).
Time (PST)
Height
F
1030
A D
1130
1220
1313
1425
1454
(cm)
FAD
F
A D
FAD
FAD
FAD
750
100
50
25
6
2
0
2
4 20
4 24
6 18
6 4
0 4 22
30 18 24
26 12 44
14 12 44
0
6
0
0
4 12
18 48
2 24
22 54
2 0 12
0 2 26
0 0 4
10 2 26
46 12 14
8 4 24
0 0 4
0 0 0
42 6 18
0 0 26
0 6 12
0 0 0
255
there was a marked gradient in development with
proximity to the seabed (Fig. 2). All the larvae at
6.7 m had unflexed notochords, and most were
<2.5 mm NL. Feeding incidence (proportion of
larvae with nonempty guts) was 78% at this
height. At 1 m, modal larval length was 2.65 mm,
FISHERY BULLETIN: VOL. 86, NO. 2
with a single postflexion specimen (Fig. 2); feed-
ing incidence was 74%. At 0.5 m there were still
some preflexion larvae, but a second length mode
at 6.8 mm represented postflexion-stage larvae.
Feeding incidence was 90%) at 0.5 m above the
bottom, being somewhat greater among flexion
39 n
O
CO
>
c5
o 0 J
3
0 ^
13 -1
G. lineatus
GUTS WITH CONTENTS ■
GUTS EMPTY 0
(excludes yolksac stage)
670 cm
(n = 109)
2.0 2.5 3.0
in'i'i
100 cm
(n=70)
2.0 2.5 3.0
' I ' I n ' I ' I ' I
5 10
0 ^
fl#^
2.0 2.5 3.0
50 cm
(n=83)
Length (mm)
Figure 2. — Length frequencies of feeding-stage larvae of Genyonemus lineatus at three sampling heights.
256
JAHN ET AL.: FOOD-SEEKING LARVAL WHITE CROAKER
and postflexion larvae (95%) than among preflex-
ion larvae (82%).
Gut Contents
The white croaker larvae were divided into
three size classes for analysis of gut contents with
regard to height above the bottom: preflexion lar-
vae <2.7 mm (size 1), preflexion larvae >2.7 mm
(size 2), and flexion and postflexion larvae (size
E
o
o
84 guts
373 prey
E
o
o
o
3). The largest preflexion larva was 4.6 mm NL,
and the smallest flexion stage larva was 5.5 mm
FL. The division of preflexion larvae at 2.7 mm
retained all but one specimen at 6.7 m in size 1
while partitioning the preflexion larvae at 1 m
and 0.5 m about equally into sizes 1 and 2 (Figs.
2, 3). Besides the 2.75 mm specimen at 6.7 m, a
single flexion stage larva at 1 m was excluded by
these criteria from the comparisons.
31 guts
82 prey
21 guts
66 prey
DINOFLAGELLATES
TINTINNIDS
ROTIFERS
POLYCHAETE LARVAE
LAMELIBRANCH LARVAE
NAUPLII
COPEPODS
AMPHIPODS
INVERTEBRATE EGGS
OTHER
^
^
^
Hii
ESS
E
o
o
it)
8 guts
15 guts
52 guts
22 prey
26 prey
194 prey
SIZE 1
SIZE 2
SIZE 3
Figure 3. — Percentage contribution of 10 food categories to the diet of larval white croaker at three heights above the
bottom. Size 1 = preflexion larvae <2.7 mm NL; size 2 = preflexion larvae >2.7 mm NL; size 3 = flexion and
postflexion-stage larvae.
257
FISHERY BULLETIN: VOL. 86, NO. 2
Most identifiable prey items fit into the nine
categories: dinoflagellate, tintinnid, rotifer, poly-
chaete larva, lamellibranch larva, crustacean
nauplius, copepodite and adult copepod, am-
phipod, and invertebrate egg (Fig. 3). The "other"
category applied only to size-2 larvae at 1 m (1
Globigerina sp.) and to size-3 larvae (1 zoea, 3
larvaceans, and two large [1 mm] unidentified
spheres).
Guts of preflexion (sizes 1 and 2) larvae from
the three sampling heights contained an array of
small (<300 jxm) organisms that varied mainly in
proportions from mostly rotifers (88%) in size-1
larvae at 6.7 m to a diverse mix of prey numeri-
cally dominated by nauplii in size-2 larvae at 0.5
m (Fig. 3). Percent similarity (overlap) among the
5 groups of preflexion larvae ranged from 24 to
75%. The gut of the single size-2 larva at 6.7 m,
not included in Figure 3, contained two tintin-
nids.
Size-3 larvae had a diet consisting chiefly of
copepodite and adult copepods that overlapped
only 8-9% with size-2 larvae and 1% or less with
the three groups of size-1 larvae. The copepods
eaten by size-3 larvae were mostly Corycaeus an-
glicus (62% of all copepods), unidentified cope-
podites (cyclopoid and calanoid, 25%), and Para-
calanus parvus (9%). Polychaete larvae were
identified only from the presence of setae in the
guts, so the proportion (nominally 16% of all prey
items) of this taxon in the diet is more an indica-
tion of incidence than of numerical importance.
Amphipods, mostly in the length range 1-1.5
mm, were found in white croaker larvae ranging
from 6.5 mm FL to 10.3 mm SL. The gut of the
flexion-stage larva at 1 m, not included in Figure
3, contained three C. anglicus and traces of poly-
chaete setae.
While there can be no doubt that flexion and
postflexion larvae had a different diet than pre-
flexion larvae, the pattern of decreasing propor-
tion of rotifers with increasing size and proximity
to the bottom among preflexion larvae was of
questionable statistical significance. The first
question asked was whether the very high per-
centage of rotifers in the diet of size-1 larvae at
6.7 m was likely to have arisen by chance from a
random sampling of size-1 larvae. Formally
stated, Hq = "all size-1 larvae had the same per-
centage of rotifers". The 123 nonempty guts were
pooled, and random samples of 84 each were
drawn. In 1,000 iterations, <4% of the samples
had >88% rotifers, so it was concluded that lar-
vae at 6.7 m ate significantly more rotifers than
similar-sized larvae near the bottom. The remain-
ing 75 preflexion larvae (sizes 1 and 2) are divided
into 4 small groups at 0.5 and 1 m, so we next
tested for a size effect by pooling across sampling
height, such that the guts of the 39 near-bottom
size-1 larvae contained 64% rotifers, and the 36
size-2 larvae had 36% rotifers. Bootstrapping as
before, <2% of samples of 36 had <36% rotifers,
so it was concluded that size-1 and size-2 larvae
differed in this regard. Further testing (e.g., of a
height effect within sizes) was not done because of
small sample sizes and multiple testing consider-
ations.
Abundance and
Vertical Distribution of Prey
Rotifers, all identified as the brachionoid Tri-
chocerca sp., figured importantly both in the diet
of preflexion larvae and in the time-related vari-
ance structure of the microplankton. As shown in
Table 3, there was a change in the size spectrum
of these animals that coincided approximately
with the time of changing from near-bottom sam-
pling to midwater sampling with the fish pump. It
was only the largest category of rotifer (200-300
|xm, including the "toe") that was found in the
guts of the larvae. The relative abundance of total
rotifers in the plankton at the times and heights
of pump sampling differed very little (25-33% of
all organisms in the 100-300 ^JLm size class), but
the percentage of rotifers in the 200-300 [xm class
increased from 21% (near-bottom, morning) to
86% of all rotifers (midwater, afternoon). The
dominance of rotifers in the diet of size-1 larvae in
midwaters is thus likely related to the larger size
of rotifer resident in the water column when that
height was sampled.
The most notable dietary difference among the
larval size groups analyzed was the switch from
small (50-300 (xm) to larger (0.5-2.5 mm) prey,
principally the copepod Corycaeus anglicus (0.5-
0.8 mm), upon flexion of the notochord. The abun-
dance of Corycaeus from the 100 ixm mesh pump
samples (Table 4) shows that this prey item was
equally or more abundant in midwater than near
the bottom, where all the flexion and postflexion
larvae were captured. (Within the bottom meter,
the similar-sized but more transparent Para-
calanus parvus outnumbered C anglicus by a fac-
tor of 5-20.) The only prey found in numbers in
these larvae that was restricted to the 0.5 m sam-
ples was gammarid amphipods. Larger crus-
taceans— cumaceans, crab and shrimp zoea,
258
JAHN ET AL.: FOOD-SEEKING LARVAL WHITE CROAKER
Table 3. — Density (rotifers per liter) of Trichocerca sp. as a function of time and sampling heighit. Eachi set of three numbers gives the density
of 100-150 M-m, 150-200 txm, and 200-300 ^.m rotifers. Height of simultaneous pump sample is given.
Time (PST):
Pump height:
Sampling
height
(cm)
Time (PST):
Pump height:
Sampling
1030
0.5 m
1130
0.5 m
1220
1 m
100-
150 pim
150-
200 ^.m
200-
300 pirn
100-
150 \i.m
150-
200 M-m
200-
300 M-m
100-
150 pLm
150-
200 M.m
1313
1 m
1425
6.7 m
1454
6.7 m
200-
300 M-m
750
8
0
4
4
0
10
0
0
20
100
8
14
0
0
66
0
14
22
10
50
24
18
0
0
8
40
0
34
0
25
0
2
12
16
46
0
0
30
0
height
100-
150-
200-
100-
150-
200-
100-
150-
200-
(cm)
150 fim
200 (xm
300 M-m
150 ^.m
200 M.m
300 M-m
150 M-m
200 M-m
300 M-m
750
20
0
0
0
12
14
0
0
60
100
0
4
0
12
24
0
46
0
102
50
6
0
34
0
4
16
0
0
42
25
0
26
0
0
0
6
0
2
0
Table 4. — Abundance (animals m 3) of cope-
podite and adult Corycaeus spp. in 100 [im
mesh samples from the fish pump.
Height
(m)
First sample
Second sample
6.7
1
0.5
1,080
120
180
460
500
140
mysids, and euphausid furcilia larvae — were all
abundant OlO m""^) in the 0.5 m, 330 ixm mesh
samples but with the exception of the callianassa
zoea mentioned above (from the gut of an 11 mm
larva) were not found in these white croaker lar-
vae.
DISCUSSION
The chief drawback of the pumping system
used was its inability to obtain a simultaneous
vertical profile. The sampling sequence left the
possibility that differences among heights might
be confounded by trends in time, as discussed by
Jahn and Lavenberg (1986). Slight time effects
were found among the vertical profiles of mi-
croplankton, increasing the suspicion that the ap-
parent vertical distributions of fish larvae and
macrozooplanktonic prey might have horizontal
components. To contradict the argument that
food-seeking did not bring postflexion larvae near
the bottom, one would need to invoke either an
afternoon increase of some two orders of magni-
tude in copepod abundance (Table 4) or else the
presence of flexion and postflexion larvae
throughout the water column in morning and
midday followed by their sudden disappearance
in the afternoon.
A two-order-of-magnitude change in copepod
species abundance over a distance of roughly 1
km (2 hours at 14 cm s"M is certainly possible;
though zooplankton structures reported from the
southern California continental shelf are gener-
ally larger than this (Star and Mullin 1981; Bar-
nett and Jahn 1987), there is always the possibil-
ity of sampling the edge of a patch. Since no such
edge was evident in the abundance or overall
composition of microplankton or of phytoplank-
ton, it seems unlikely that a macrozooplankton
change of this order occurred. Moreover, the main
copepod eaten, Corycaeus anglicus, is generally
more abundant in midwater than near the bottom
over the shallow shelf (A. Barnetf^), in accord
with its apparent distribution in this study. As to
a possible midwater abundance of postflexion
white croaker larvae, no such concentration has
ever been reported. In some nine vertical profiles
taken in daylight over a 6-d period. Brewer and
Kleppel (1986) took virtually all specimens >3.5
mm in their near-bottom sampler. White croaker
appears similar to another abundant sciaenid,
queenfish, in this regard (cf Jahn and Lavenberg
1986).
4A. Barnett, Marine Ecological Consultants, 531 Encinitas
Blvd.. Encinitas, CA 92024, pers. commun. July 1987.
259
FISHERY BULLETIN: VOL. 86, NO 2
The only unequivocal instance in which a prey
item of larval white croaker was vertically dis-
tributed similarly to the larvae was the trace of
amphipods found in the guts of competent (flexion
and postflexion) larvae. At the lengths of larvae
sampled (<12 mm) the prey were all planktonic
and nearly all about equally abundant in mid-
waters as near the bottom. The small numbers of
amphipods eaten may indicate an incipient tran-
sition to larger, suprabenthic crustacean prey.
The size gap between the large prey of these com-
petent larvae and the smaller prey of preflexion
larvae is probably an artifact of the bimodal size
distribution of sampled larvae. Though all of the
prey eaten by size-1 (<2.7 mm) larvae were <300
|xm in length, the more varied diet of larger pre-
flexion larvae contained some copepods as big as
500 |xm. There is therefore nothing in these data
to suggest that the switch from microplanktonic
to macroplanktonic prey is anything but a grad-
ual transition as the larvae grow.
Brewer and Kleppel (1986) also reported a
change to copepod prey in white croaker larvae
>6 mm. Our findings are further similar to those
of Brewer and Kleppel in that there was no indi-
cation that food-seeking had anything to do with
the descent of larval white croaker from mid-
waters to the near-bottom zone. The other defin-
able dietary trend in this study (besides ontoge-
netic change) was the high percentage of rotifers
eaten by midwater preflexion larvae. This was
apparently related to subtle but important differ-
ences in the available planktonic prey — signifi-
cantly, to a greater abundance of suitable-size ro-
tifers— at the time the midwater stratum was
sampled.
It seems safest to conclude that white croaker
larvae descend toward the bottom for reasons
quite apart from seeking food (see discussions in
Barnett et al. [1984], Brewer and Kleppel [1986],
Jahn and Lavenberg [1986]) and simply eat what-
ever they find there that suits them. Many poten-
tial macroplanktonic prey also favor the near-
bottom layer (Jahn and Lavenberg 1986; Barnett
and Jahn 1987). Older larvae and their prey may
occupy the near-bottom layer for different rea-
sons, or it may be that a single advantage, or set
of pressures, underlies the behavior of these di-
verse planktonic and semi-planktonic taxa. Some
species need to remain near shore, and living in
the bottom boundary layer helps assure this. The
boundary layer also tends to be more turbid than
overlying waters and so may lessen an animal's
jeopardy to visual (biting) planktivores. (The gen-
erality of the latter explanation only holds if
suprabenthic fish larvae are less important
planktivores than other water-column inhabi-
tants— see Gushing 1983.)
Rotifers have never to our knowledge been re-
ported as an important food of ocean-caught fish
larvae, even though the genus Brachionus is com-
monly cultured for feeding larval fish in the labo-
ratory. Schmitt (1986) reported that small,
laboratory-reared larval northern anchovy read-
ily fed upon (unidentified) wild-caught rotifers.
Rotifers are only occasionally abundant in neritic
waters, and never in oceanic waters (J. Beers^).
Their rarity notwithstanding, rotifers have the
ability very rapidly to dominate marine mi-
croplanktonic assemblages (Hernroth 1983), and
their good food quality (Theilacker 1987) and
high secondary productivity for a period of weeks
might constitute a significant enhancement to
growth and survival of a larval fish cohort.
Our previous experience in handling larval
white croaker specimens agrees with the findings
of Brewer and Kleppel (1986) in that lamelli-
branch larvae, easily seen through the body wall,
are a common food for small white croaker larvae.
In our study, this taxon was a minor constituent
of the plankton and of the larval fish diet. We
cannot say how unusual were the circumstances
we encountered, but we know that in terms of
diatom numbers and larval fish diversity these
conditions were not typical of March on the south-
ern California continental shelf That white
croaker larvae appeared to find these conditions
salubrious may be one reason this species is so
successful in southern California (Love et al.
1984).
ACKNOWLEDGMENTS
Thanks go to S. Caddell, R. Feeney, T. Garrett,
R. Lavenberg, J. McGowen, J. Petersen,
J. Rounds, and Captain L. Nufer for able partici-
pation in the field work. D. Carlson-Oda,
J. Rounds, and S. Shiba helped process larval fish
samples, and H. Schwarz helped prepare the
manuscript. K. Zabloudil generously loaned the
current meters, and R. Erdman assisted in proc-
essing data therefrom. We also thank J. Beers
and R. Brusca for help in accessing literature on
rotifers. D. Cohen, R. Lavenberg, and J. Petersen
5J. Beers, Scripps Institution of Oceanography, La Jolla, CA
92093, pers. commun. November 1986.
260
JAHN ET AL.: FOOD-SEEKING LARVAL WHITE CROAKER
reviewed the manuscript. The Southern Califor-
nia Edison Company funded the study.
LITERATURE CITED
Arthur. D K
1976. Food and feeding of larvae of three fishes occurring
in the California Current, Sardinops sagax, Engraulis
mordax, and Trachurus symmetricus . Fish. Bull., U.S.
74:517-530.
Balon, E K.
1975. Terminology of intervals in fish development. J.
Fish. Res. Board Can. 32:1633-1670.
Barnett. a. M . AND A E Jahn.
1987. Pattern and persistence of a nearshore planktonic
ecosystem off southern California. Cent. Shelf Res. 7:1-
25.
Barnett, A M , A E Jahn, P D. Sertic, and W Watson.
1984. Distribution of ichthyoplankton off San Onofre,
California, and methods for sampling very shallow
coastal waters. Fish. Bull., U.S. 82:97-111.
Brewer, G D , and G S Kleppel
1986. Diel vertical distribution of fish larvae and their
prey in nearshore waters of southern California. Mar.
Ecol. Prog. Ser. 27:217-226.
Brewer, G D , G S Kleppel, and M Dempsey.
1984. Apparent predation on ichthyoplankton by zoo-
plankton and fishes in nearshore waters on southern
California. Mar. Biol. (Berl.) 80:17-28.
Brewer, G. D , R L 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.
Buckley, L J., and R G Lough
1987. Recent growth, biochemical composition, and prey
field of larval haddock tMelanogrammus aeglefinus) and
Atlantic cod iGadus morhua ) on Georges Bank. Can. J.
Fish. Aquat. Sci. 44:14-25.
Clutter, R I
1969. The microdistribution and social behavior of some
pelagic mysid shrimps. J. Exp. Mar. Biol. Ecol. 3:125-
155.
Gushing, D H
1983. Are fish larvae too dilute to affect the density of
their food organisms? J. Plankton Res. 5:847-854.
Efron, B
1982. The jackknife, the bootstrap and other resampling
plans. Soc. Ind. Appl. Math., Philadelphia, PA, 92 p.
GaDOMSKI, D M , AND G W BOEHLERT
1984. Feeding ecology of pelagic larvae of English sole,
Parophrys vetulus , and butter sole, Isopsetta isolepis, off
the Oregon coast. Mar. Ecol. Prog. Ser. 20:1-12.
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,
Leiostomus xanthurus, and Micropogonias undulatus.
Mar. Ecol. Prog. Ser. 13:189-199.
GovoNi, J J , A J Chester, D E Hoss. and P B Ortner.
1985. An observation of episodic feeding and growth of
larval Leiostomus xanthurus in the northern Gulf of
Mexico. J. Plankton Res. 7:137-146.
Hamner, W M , AND J. H Carleton
1979. Copepod swarms: attributes and role in coral reef
ecosystems. Limnol. Oceanogr. 24:1-14.
HERNROTH, L.
1983. Marine pelagic rotifers and tintinnids — important
trophic links in the spring plankton community of the
Gullmar Fjord, Sweden. J. Plankton Res. 5:835-846.
Hunter. J R
1981. Feeding ecology and predation of marine fish lar-
vae. In R. Lasker (editor). Marine fish larvae: morphol-
ogy, ecology, and relation to fisheries, p. 34-77. Wash-
ington Sea Grant Program, Seattle.
Jahn. A E.. and R. J Lavenberg.
1986. Fine-scale distribution of nearshore, suprabenthic
fish larvae. Mar. Ecol. Prog. Ser. 31:223-231.
Lasker. R
1975. Field criteria for survival of anchovy larvae: the
relation between inshore chlorophyll maximum layers
and successful first feeding. Fish. Bull., U.S. 73:453-
462.
1978. The relation between oceanographic conditions and
larval anchovy food in the California current: identifica-
tion of factors contributing to recruitment. Rapp. P. -v.
Reun. Cons. int. Explor. Mer 173:212-330.
Lavenberg, R. J., G. E McGowen, A E Jahn. J. H. Petersen,
AND T C Sciarrotta
1986. Southern California nearshore ichthyoplankton:
a study of abundance patterns. CalCOFI Rep. 27:53-
64.
Leis, J M
1986. Epibenthic schooling by larvae of the clupeid fish
Spratelloides gracilis . Jpn. J. Ichthyol. 33:67-69.
Love, M S , G E. McGowen, W Westphal, R. J Lavenberg, and
L Martin
1984. Aspects of the life history and fishery of the white
croaker, Genyonemus lineatus (Sciaenidael, off Califor-
nia. Fish. Bull., U.S. 82:179-198.
McGowen, G E
1987. Coastal zone ichthyoplankton assemblages off
Southern California. Ph.D. Thesis, Univ. Southern
California, Los Angeles, 304 p.
Owen. R W.
1981. Microscale plankton patchiness in the larval an-
chovy environment. Rapp. P. -v. Reun. Cons. int. Ex-
plor. Mer 178:364-368.
SaINTE-MaRIE, B . AND P BRUNEL.
1985. Suprabenthic gradients of swimming activity by
cold-water gammaridean amphipod Crustacea over a
muddy shelf in the Gulf of Saint Lawrence. Mar. Ecol.
Prog. Ser. 23:57-69.
Schlotterbeck, R E , and D W Connally
1982. Vertical stratification of three nearshore southern
California larval fishes (Engraulis mordax, Genyonemus
lineatus, and Seriphus politus). Fish. Bull., U.S.
80:895-902.
SCHMITT. P D
1986. Prey size selectivity and feeding rate of larvae of the
northern anchovy, Engraulis mordax Girard. CalCOFI
Rep. 27:153-161.
Star. J. L.. and M M Mullin.
1981. Zooplanktonic assemblages in three areas of the
North Pacific Ocean as revealed by continuous horizontal
transects. Deep-Sea Res. 28A:1303-1322.
Tanaka, M.
1985. Factors affecting the inshore migration of pelagic
larval and demersal juvenile red sea bream, Pagrus
major, to a nursery ground. Trans. Am. Fish. Soc.
114:471-477.
261
FISHERY BULLETIN: VOL. 86, NO. 2
Theilacker. G
1987. Feeding ecology and growth energetics of larval
northern anchovy, EngrauUs mordax. Fish. Bull., U.S.
85:213-228.
Theilacker, G , and K Dorsey
1980. Larval fish diversity, a summary of laboratory and
field research. FAO Intergov. Oceanogr. Comm. Work-
shop Rep. No. 28, p. 105-142.
TONT, S. A.
1981. Historical changes of diatom abundance off south-
ern California as reflected in sea surface temperature, air
temperature, and sea level. J. Mar. Res. 39:191-201.
Utermohl. N
1931. Neue Wege in der quantitativen Erfassung des
Planktons. Verb. Int. Ver. Limnol. 5:567-597.
Walker, H J . Jr , W Watson, and A. M Barnett.
1987. Seasonal occurrence of larval fishes in the near-
shore Southern California Bight off San Onofre, Califor-
nia. Estuarine Coastal Shelf Sci. 25:91-109.
Watson, W.
1982. Development of eggs and larvae of the white
croaker, Genyonemus lineatus Ayres (Pisces:Sciaenidae),
off the southern California coast. Fish. Bull., U.S.
80:403-417.
WiSHNER, K F
1980. The biomass of the deep-sea benthopelagic plank-
ton. Deep-Sea Res. 27:203-216.
262
NEW MARINE DECAPOD CRUSTACEANS FROM WATERS INFLUENCED BY
HYDROTHERMAL DISCHARGE, BRINE, AND HYDROCARBON SEEPAGE
Austin B. Williams^
ABSTRACT
Five species of decapod crustaceans new to science are described. These are caridean shrimps of the
family Bresiliidae — Alvinocaris markensis from a Mid-Atlantic Rift Valley hydrothermal field, A.
muricola from a cold brine seep at the foot of the West Florida Escarpment in the Gulf of Mexico, and
A. stactophila from a hydrocarbon seep on the continental slope of the northern Gulf of Mexico, with
a key to the species of Alvinocaris ; a squat lobster of the family Galatheidae — Munidopsis alvisca
from the Guaymas Basin and from the Juan de Fuca and Explorer ridges in the eastern Pacific; and
a brachyuran crab of the family Bythograeidae — Bythograea mesatlantica from a Mid-Atlantic Rift
Valley hydrothermal field. Species of both Alvinocaris and Bythograea are now known from the
eastern Pacific and Mid-Atlantic. Munidopsis species are widely represented in the world oceans.
Deep ocean hydrothermally active fields and
waters influenced by brine and hydrocarbon seeps
continue to yield species new to science. Such en-
vironments were unknown until explored with
the aid of submersible research vessels from
which observations and collections could be ac-
complished. The species of decapod crustaceans
reported here come from hydrothermal fields in
the Mid- Atlantic Rift Valley, the Guaymas Basin
in the Golfo de California, and Juan de Fuca and
Explorer Ridges in the northeastern Pacific, a
cold brine seep at the foot of the West Florida
Escarpment, and a hydrocarbon seep on the con-
tinental slope of the northern Gulf of Mexico.
These are scattered localities that exhibit diverse
environmental conditions but that are bound
together by the common thread of chemotrophic
food chains (Childress et al. 1986; Brooks et al.
1987).
The material from the Mid-Atlantic Rift Val-
ley, West Florida Escarpment, and Guaymas
Basin was observed and collected by scientists
working with the aid of the DSRV Alvin and RV
Atlantis II based at the Woods Hole Oceano-
graphic Institution. That from the northern Gulf
of Mexico came from the Minerals Management
Service Northern Gulf of Mexico Outer Continen-
tal Slope (MMS/NGOMCS) Regional Office Proj-
ect, involving observation and collection of mate-
rial by scientists from LGL Ecological Research
iSystematics Laboratory, National Marine Fisheries Service,
NOAA, National Museum of Natural History, Washington,
D.C. 20560.
Manuscript accepted January 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.
Associates and Texas A&M University, with the
aid of the submersible research vessel Johnson-
Sea-Link and its support vessels. Specimens from
Explorer and Juan de Fuca Ridges were collected
with the aid of the Canadian DSRV Pisces IV and
its support vessels.
All specimens studied are deposited in the
Crustacean Collection of the United States Na-
tional Museum of Natural History, Smithsonian
Institution, Washington, D.C. 20560.
CARIDEA: BRESILIIDAE
Alvinocaris Williams and Chace, 1982
Three species of bresiliid shrimps described
below as new to science are placed in the hereto-
fore monotypic genus Alvinocaris. Certain fea-
tures of these species necessitate minor changes
in the generic diagnosis by Williams and Chace
(1982) as follows: Rostrum with or without ven-
tral teeth. Telson with 2-5 pairs of principal
spines on posterior margin. Strong median ster-
nal spine between posterior pair of pereopods.
Moreover, the branchial formula seems uni-
formly fixed in this genus as well as in the genus
Rimicaris Williams and Rona, 1986. The arrange-
ment, figured in Williams and Chace (1982) and
Williams and Rona (1986) may be described as
follows:
Phyllobranchs extensively developed in 2 se-
ries; asymmetrically Y-branched pleurobranchs
with relatively short ventral and progressively
longer and more expansive dorsal ramus associ-
263
FISHERY BULLETIN; VOL. 86, NO. 2
ated with pereopods 1-5; smaller and more nearly
symmetrical U-branched arthrobranchs of more
nearly uniform size associated with third maxil-
liped and pereopods 1-4.
Key to Known Species oi Alvinocaris
1. Telson with terminal margin slightly con-
cave in midline and bearing 3-5 pairs of
principal spines
A. lusca Williams and Chace
Telson with terminal margin convex and
bearing only 2 pairs of principal spines ... 2
2. Rostrum with ventral margin bearing 0-1
subterminal ventral spines
A. stactophila new species
Rostrum with ventral margin bearing 4 or
more subterminal ventral spines 3
3. Abdominal segment 3 with pleural margin
entire A. markensis new species
Abdominal segment 3 with pleural margin
obscurely serrate
A. muricola new species
Alvinocaris markensis new species
Figures 1, 2, 7
Material— \JSnU 234286, Holotype 9
(crushed), USNM 234287, Paratypes, 2 9 (dam-
aged), Mid-Atlantic Rift Valley about 70 km
south of Kane Fracture Zone (see Leg 106 Ship-
board Scientific Party [1986]; Ocean Drilling Pro-
gram Leg 106 Scientific Party [1986]),
23°22.09'N, 44°57.12'W, 3,437 m, Alvin Dive
1683, MARK vent, Stn. 1, scoop, 30 May 1986,
pilot D. Foster, observers S. Humphris and J. Ed-
mond. From NSF Ocean Drilling Program-Leg
106, NSF Grant OCE-8311201 to J. F. Grassle,
Woods Hole Oceanographic Institution, Woods
Hole, MA.
Measurements in mm. — Holotype 9, postor-
bital carapace length 4.16, rostrum 2.3, maxi-
mum carapace height 3.3, total length about
15.6. Paratype 9, same 2.7, 1.6, 2.4, 10.9.
Description. — Integument extremely thin,
264
membranous, shining, with a few minute puncta-
tions. Rostrum (Fig. la, 6) almost straight,
slightly elevated above horizontal in distal half,
sharply pointed tip usually reaching to between
midlength of second and tip of third peduncular
articles of antennule; dorsal margin raised into
thin serrate crest containing 12-17 teeth,
strongest in central sector of row, with about 1/3
length of crest continued onto carapace; ventral
margin less prominent and armed with 5-8 sub-
terminal teeth; tooth formulas examined, 17/8
(holotype), 17/5, 12/5 (apparently some subtermi-
nal dorsal teeth fused); strong lateral carina
broadened proximally and confluent with orbital
margin. Carapace with acute antennal spine dis-
tinct; pterygostomian spine acuminate and
prominent. Indistinct antennal groove curving
ventrad to intersect associated indistinct groove
at about midlength of carapace and continuing
posteriad.
Abdomen of female (Fig. Ic) apparently
broadly arched dorsally (all specimens examined
are crushed), gradually tapering posteriorly, nar-
rowest part of sixth somite about 1/2 width of first
somite; pleura of 3 anterior somites broadly
rounded, that of fourth somite drawn posterolat-
erally to strong acuminate spine flanked dorsally
by 0 or 1 much more slender and smaller spine;
posterolateral corner of fifth pleuron strongly
acuminate to nearly right angled and flanked
dorsally by 0-2 spines of variable size analogous
to condition on somite 4, spine number possibly
age related; sixth somite with middorsal length
about 1.9 that of fifth, broad based midlateral
spine overlapping base of telson, smaller pos-
terolateral spine acute; fourth somite with small
erect spine on sternite and fifth with analogous
strong, posteriorly directed spine. Telson (Fig.
\d) elongate subrectangular, length about 3.5
anterior width, 5.8 posterior width, and about
1.75 length of sixth somite, not including poste-
rior spines; armed with 6-8 dorsolateral spines
of nearly uniform size, sometimes bilaterally
unequal in number; posterior margin convex,
armed with 2 principal spines at each corner
and 10-12 feathered strong setae on margin be-
tween.
Eyes (Fig. la , 6 ) with cornea imperfectly devel-
oped, unfaceted though diffusely pigmented, glob-
ular to ovate in outline and with prominent spine
on anterodorsal edge.
Antennular peduncle (Fig. la, 6) reaching end
of antennal scale; basal article 1.3 length of sec-
ond and about 3.0 length of third, stylocerite well
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
Figure 1. — Alvinocans markensis, holotype $ : a part of cephalothorax and anterior appendages, lateral; b rostrum, eye, anten-
nular peduncle, antennal scale, dorsal of left side; c abdomen, lateral; cf tail fan; cheliped, e mesial, /"lateral; g pereopod 2,
h chela. Scales = 2 mm: I ia, c , d, e , f, g, h)\2 ib).
265
FISHERY BULLETIN: VOL. 86, NO. 2
separated from peduncle, tapering to slender
elongate tip reaching about 1/4 length of second
article, basal article with distodorsal margin
flanked by transverse row of setae, extended into
strong lateral spine reaching level equal to that of
stylocerite and closely appressed to second article;
shorter second article with strong mesiodistal ap-
pressed spine. Dorsolateral flagellum about 1.5
length of carapace, thickened proximal half bear-
ing conspicuous ventral setae; ventromesial flag-
ellum somewhat more slender in lateral view and
shorter.
Antennal scale (Fig. la, 6) about 2.5 as long as
wide, distolateral tooth falling short of distome-
sial apex of broadly rounded distal margin of
blade; basal article with strong ventrolateral
spine; flagellum slightly exceeding length of body
(missing from holotype).
Mandibles (Fig. 2g) similar, with 2-segmented
palp, incisor process broad and armed with 8
marginal teeth, slender molar process simple,
divergent, its narrowly rounded tip minutely
setose.
First maxilla (Fig. 2h) with proximal endite
asymmetrically oval-triangular, distal margin
bearing many long setae; distal endite with nar-
rowed base but broadened distally, armed with
many (about 37) short spines on mesial margin
and with scattered longer spinules marginally
and submarginally beyond either end of spine
row; palp scarcely bifurcated, with long distal
spine on obsolescent proximomesial branch and 1
shorter submarginal spine on distal branch.
Second maxilla (Fig. 2i) with proximal endite
represented by 2 similar lobes; distal endite sub-
triangular, expanded mesiodistally and paral-
leled laterally by narrow somewhat twisted palp,
scaphognathite with anterior lobe rectangulo-
ovate, fringed with uniformly long, silky setae on
anterior and mesial borders, shorter setae along
entire lateral margin; posterior lobe narrowly
ovate-triangular, fringed on blunt tip and adja-
cent mesial margin by strikingly long, tangled,
strong setae preceded proximally by shorter setae
similar to those on lateral margin.
First maxilliped (Fig. 2j, partly flattened view)
with irregularly fusiform endite, short palp much
exceeded in length and size by leaflike exopod,
epipod obscurely bilobed.
Second maxilliped (Fig. 2k, I) somewhat pedi-
form but flattened, mesial margin of articles
bearing long, feathered setae, mesial surface of
terminal article densely setose, tip of exopod ex-
ceeding leaflike epipod.
Third maxilliped (Fig. 2m, n) slender, 5-
segmented, reaching beyond antennular pedun-
cle; terminal segment trigonal in cross section,
tapered distally, bearing 3 terminal spines,
oblique tracts of dense setae along mesial surface;
similar tract of setae on carpus and another less
conspicuous group on merus-ischium, latter with
distolateral spine at articulation with carpus; ex-
opod much reduced, ovate-triangular, without
lash.
First pereopods (Fig. le, f) chelate, subequal;
fingers curved ventrally and slightly laterad;
dactyl much more slender than and slightly
longer than fixed finger; mesial surface of each
finger convex, lateral surface deeply concave; pre-
hensile surfaces uniformly offset, closing without
gape, each armed with row of almost uniform
teeth so closely set as to be almost contiguous,
line of sensory hairs mesial to cutting edges,
acute tip of dactyl slightly spooned by elongate
teeth slanted distad and curving around its exter-
nal edge. Leg not reaching tip of third maxilliped
and exceeding antennal peduncle. Palm of holo-
type female inflated, length slightly greater than
height and shorter than fingers (0.60); low ridge
ending in small hooked spine on proximomesial
surface near articulation with carpus. Carpus
longer than palm; bearing oblique ventral crest
ending in strong distoventral spine and flanked
mesially by patch of setae on triangular raised
area; rectangular distal notch above spine fol-
lowed by oblique distomesial margin leading to
poorly defined spine at condyle articulating with
palm; distolateral margin with rounded ventral
corner leading to sinuous border above it bearing
2 lobes near articulation with palm. Merus some-
what swollen is distal half and bearing small dis-
tomesial spine, distinct from ischium but fused to
it.
Second pereopod (Figs. 1^, h; 11) shorter and
more slender than first, but reaching beyond an-
tennal peduncle by about length of fingers. Fin-
gers slightly shorter than palm, similar in size
and shape; opposed edges without gape, each
spineless proximally, but distal half pectinate
with single row of spines directed obliquely distad
and increasing slightly in size to end in notice-
ably stronger spine crossing opposite member
when closed. Carpus slender, about 0.9 length of
chela; merus and ischium unarmed.
Third to fifth pereopods (Fig. 2a-f) similar in
length and structure, third reaching distal edge of
antennal scale. Length articles of these legs in
holotype 9, mm:
266
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
Figure 2. — Alvinocaris markensis , holotype 9: a pereopyod 3, b dactyl; c pereopod 4, d dactyl; e p)ereopK)d 5, f
dactyl; g mandible; h maxilla 1; i maxilla 2; j maxilliped 1 partly flattened; maxilliped 2, k endopod, ventral, I part
of exopod dorsal; maxilliped 3, m ventral, n dorsal, distal articles only. Scales: Uk, I) = 1 mm; 2 (6, d,f,g, A) = 1 mm;
3(a,c,e,i,j,m,n) = 2 mm.
267
FISHERY BULLETIN: VOL. 86, NO. 2
3rd
4th
5th
ischio-
merus
6.14
5.76
5.44
carpus propodus
2.56
2.18
2.30
3.78
4.12
5.44
dactyl
0.64
0.70
0.64
Each short dactyl armed with 5 spines on flexor
surface, grading from small proximally to longest
and strongest distally, often a sensory seta on
extensor surface. Propodi with setae along flexor
surface progressively more crowded distally.
Carpi with distodorsal extension projecting as a
stop along proximal part of propodal extensor sur-
face. Third leg stronger, at least in merus-
ischium, than fourth and fifth; merus of third and
fourth with closely appressed ventral spine at 1/2
and 3/4 length, that of fifth with spines at 1/3 and
2/3 length, distal spine strongest in each case;
ischium of third and fourth leg with 2 spines in
line with those on merus.
Pleopods well developed, pair 1 with endopods
about half length of exopods, tapering to acute
tip; appendices internae simple, that of pair 5
with blunt tip.
Uropod (Fig. Id) with rami subequal in length,
slightly exceeding distal end of telson, lateral
ramus with movable spine mesial to smaller dis-
tolateral tooth, diaeresis sinuous.
Remarks. — Remarks are given in the account
for A. stactophila.
Etymology. — The name is taken from an
acronym for the site of collection in the Mid-
Atlantic Ridge Valley about 70 km south of an
area known as the Kane Fracture Zone, "MARK",
and the Latin genitive suffix "ensis".
Alvinocaris muricola new species
Figures 3, 4, 7
Material .—\JSnU 234288, Holotype c^ (ceph-
alothorax and abdomen broken apart), USNM
234289, Allotype 9, West Florida Escarpment,
26°01'N, 84°54.61'W, 3,277 m, Alvin Dive 1754,
15 October 1986, pilot W. Sellers, observers
R. Carney and B. Hecker. USNM 234290, Para-
type 9 , West Florida Escarpment, same locality,
Alvin Dive 1753, 14 October 1986, pilot P. Tib-
betts, observers R. Carney and G. Knauer. All
from Barbara Hecker, Lamont Geological Ob-
servatory, Columbia University, Palisades, NY.
Measurements in mm. — Holotype 6, postor-
bital carapace length 6.4, rostrum 4.4, maximum
carapace height 4.5. Allotype 9 , same, 6.4, ros-
trum broken, 5.6.
Description. — Integument thin, shining, mi-
nutely punctate. Rostrum (Fig. 3a, b) almost
straight to slightly upturned in distal half,
sharply pointed tip reaching to proximal part of
third peduncular article of antennule; dorsal mar-
gin raised into thin serrate crest containing 17-21
teeth varying from obliquely erect in proximal
part to nearly horizontal, shorter and more dis-
tant distally, about 1/3 length of crest continued
onto carapace; ventral margin much less promi-
nent and armed with row of 6 correspondingly
smaller subterminal teeth, sometimes obscure;
lateral carina broadened proximally and conflu-
ent with orbital margin. Carapace (Fig. 3a ) with
broad based but slender, acuminate antennal
spine; pterygostomian spine correspondingly
acuminate and prominent. Prominent anterior
antennal carina curving posteroventrally to in-
tersect obliquely with carina extending from
pterygostomian spine at about midlength of cara-
pace, associated groove continuing indistinctly
posteriad.
Abdomen (Fig. 3d,e,f) of both male and female
broadly arched dorsally, gradually tapering dis-
tally, narrowest part of sixth somite less than 2/3
(0.60) width of first somite; pleura of 3 anterior
somites broadly rounded, margin of third slightly
serrated, that of fourth somite drawn posterolat-
erally to strong spine flanked dorsally by 0-3
more slender and smaller spines and preceded on
ventral margin by 0-2 small spines; number, po-
sition, and shape of either lateral or ventral
spines may be asymmetrical; posterolateral cor-
ner of fifth pleura acuminate and flanked dorsally
by 1 or 2 spines analogous to those on somite 4;
sixth somite with middorsal length about 1.7 that
of fifth, broad-based midlateral spine overlapping
base of telson, smaller posterolateral spine acute;
fourth and fifth somites each with strong, posteri-
orly directed spine on sternite. Telson (Fig. 3^)
elongate subrectangular, length about 3.0 ante-
rior width, 6.8 posterior width, and about 1.4
length of sixth somite, not including posterior
spines; armed with 7 dorsolateral spines of nearly
uniform size; posterior margin convex, armed
with 2 principal spines at each corner and 10 or 11
feathered strong setae on distal margin between.
Eyes (Fig. 3a, b) with cornea imperfectly devel-
oped, unfaceted though diffusely pigmented.
268
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
Figure 3. — Alvinocaris muricola, holotype d: a cephalothorax and anterior appendages, lateral; b rostrum, eye, antennular
peduncle, antennal scale, dorsal of right side; c antennular peduncle, distal articles, mesial; abdomen with variations in spination
of pleural margins fof allotype 1 1, d lateral, e margin of pleuron 3 from opposite side, /"(of holotjT)e d ) segments 4 and 5; g tail
fan; h median sternal spine between fifth pereopods; i median sternal spine on abdominal segment 5; cheliped, j mesial, k
lateral; I, m, n, o pereopods 2, 3, 4, 5. Scales = 1 mm: 1 (a-g,j-o); 2 (h, i).
269
FISHERY BULLETIN: VOL. 86, NO. 2
ovate in outline, though fused to each other
mesially, and each with a short upturned spine on
anterior surface.
Antennular peduncle (Fig. 3a, 6, c) reaching
beyond end of antennal scale; basal article 1.1 as
long as second and about 2.5 as long as third, all
measured on ventral margin; stylocerite well sep-
arated from peduncle, tapering to slender elon-
gate tip reaching tip of distolateral spine on basal
article; latter exceeding distodorsal margin of ar-
ticle, fringed by transverse subdistal row of setae,
and closely appressed to second article, distome-
sial spine much smaller; shorter second article
with stronger mesiodistal spine. Dorsolateral
flagellum about length of carapace, thickened in
basal 2/3, with annulations, except at base, longer
in female than in male and much longer than in
whiplike distal part; ventromesial flagellum
somewhat shorter and with annulations of vari-
able but shorter length.
Antennal scale (Fig. 3a, 6) about twice as long
as wide, distolateral tooth strong, falling slightly
short of broadly rounded distal margin of blade;
basal article with acute distal spine ventrally;
flagellum (broken in material studied) probably
slightly exceeding length of body.
Mandibles (Fig. 4a) similar, with 2-segmented
palp, incisor process broad and armed with 8 mar-
ginal teeth, slender molar process simple, diver-
gent, its narrowly rounded tip minutely setose.
First maxilla (Fig. 46) with proximal endite
asymmetrically oval-triangular, distal margin
bearing many long setae; distal endite with nar-
rowed base but broadened distally, armed with
many (about 30) short spines on mesial margin
and with scattered longer spinules submarginally
and marginally beyond either end of spine row;
palp scarcely bifurcated, with long distal spine on
obsolescent proximomesial branch and 1 shorter
submarginal spine on distal branch.
Second maxilla (Fig. 4c, c?) with proximal en-
dite represented by 2 similar lobes; distal endite
subtriangular, expanded mesiodistally and paral-
leled laterally by narrow, somewhat twisted palp;
scaphognathite with anterior lobe rectangulo-
ovate, fringed with uniformly long, silky setae on
anterior and mesial borders, uniformly shorter
setae along entire lateral margin; posterior lobe
narrow and acuminate, fringed on blunt tip and
adjacent mesial margin by strikingly long,
strong, tangled setae preceded proximally by
shorter setae similar to those on lateral margin.
First maxilliped (Fig. 4e, f) with irregularly
fusiform endite, short palp concealed and much
exceeded in length and size by leaflike exopod,
epipod obscurely bilobed; indistinct mesial lobule
on exopod possibly representing incipient lash.
Second maxilliped (Fig. 4g, h) somewhat pedi-
form but flattened, mesial margin of articles
bearing long, feathered setae, mesial surface of
terminal article densely setose, exopod barely ex-
ceeding leaflike epipod.
Third maxilliped (Fig. 4i, j) slender, 5-
segmented, reaching beyond antennular pedun-
cle; terminal article trigonal in cross section, ta-
pered distally, bearing 3 spines, transverse tracts
of dense setae along mesial surface; similar tract
of setae on carpus and another conspicuous group
mesiodistally on merus-ischium, latter with stout
distolateral spine at articulation with carpus;
exopod much reduced, subtriangular, without
lash.
First pereopods (Figs. 3^, k; If-k) chelate,
subequal and sexually dimorphic, at least in fully
mature individuals; fingers curved ventrally and
slightly laterad; dactyl more slender than and
with level of tip slightly shorter than or equal to
that of fixed finger; mesial surface of each finger
convex, lateral surface concave, with opposed sur-
faces uniformly offset; closing without gape, each
armed on prehensile edge with row of almost uni-
form teeth so closely set as to be almost contigu-
ous, acute tip of dactyl slightly spooned by elon-
gate teeth slanted distad and curving around its
external edge; line of sensory hairs mesial to cut-
ting edges. Leg shorter than to almost equaling
third maxilliped. Palm of holotype male inflated
laterally, but apparently somewhat irregularly
concave mesially, length 1.4 greatest height and
longer than fingers; palm relatively shorter in
allotype female, 0.3 length of fingers. Carpus
shorter than palm, with oblique ventral crest end-
ing in strong distolateral spine, flanked mesially
by patch of setae on polygonal raised area. Merus
somewhat swollen in distal half, distinct from is-
chium but fused to it, neither armed.
Second pereopod (Figs. 3/ ; 7e ) shorter and more
slender than first, reaching about to end of anten-
nal peduncle; fingers slightly longer than palm,
similar in size and shape, opposed edges without
gape, each pectinate with single row of teeth in
distal half directed obliquely distad and increas-
ing slightly in size to end in noticeably stronger
tooth crossing opposite member when closed, but
spineless proximally; carpus slender, about 1.16
longer than chela; merus and ischium unarmed.
Third to fifth pereopods (Fig. 3m, n,o) similar
in length and structure, third reaching to tip of or
270
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
2.
Figure 4. — Aluinocaris muricola, holotype 6: a mandible; 6 maxilla 1; maxilla 2, c ventral, d palp dorsal; maxilliped 1,
e ventral, /"dorsal; maxilliped 2, g ventral, /i dorsal; maxilliped 3, j ventral, _/ dorsal; ^ endopod of pleopod 1; /appendix
masculina, pleopod 2. Scales: I (e, f, g,j , o, k, I) = 1 mm; 2 (c, d) = 1 mm; 3 (/) = 0.2 mm; 4 (a, 6) = 0.2 mm.
271
FISHERY BULLETIN; VOL 86, NO 2
slightly beyond antennal scale. Length articles of
these legs in holotype 6 , mm:
ischium merus carpus propodus dactyl
3rd
1.76
3.84
2.33
2.34
0.35
4th
1.60
3.52
1.98
3.04
0.06
5th
1.60
3.20
2.43
4.96
0.42
Each short dactyl armed with about 4-6 corneous
spines on flexor surface, grading from small prox-
imally to longest and strongest distally; carpus of
each with distodorsal extension projecting as a
stop along proximal part of propodal extensor
surface; third leg stronger, at least in merus-
ischium, than fourth and fifth, but propodus suc-
cessively longer from third to fifth; merus of each
with ventral spine at 1/3 and 2/3 length; ischium
of third, fourth, and fifth leg with 2 spines in line
with those on merus.
Pleopods well developed; first pair with en-
dopods about 1/2 length of exopods in both sexes,
narrowed into distomesial projection in male (Fig.
4k) but evenly tapered in female; appendix mas-
culina (Fig. 41) of second pair in male (holotype)
armed with 7 slender spines extending beyond
level of simple slender appendix interna; en-
dopods of third to fifth in male and second to fifth
in female with simple slender appendix interna,
but that of fifth blunt tipped.
Uropod (Fig. 3^) with rami subequal in length,
lateral ramus slightly exceeding distal end of tel-
son, and with movable spine mesial to smaller
distolateral tooth, diaeresis sinuous.
Remarks. — Remarks are given in the account
for A. stactophila.
Etymology. — The name is from the Latin
"muria", brine, and "cola", inhabiting, for associ-
ation of the species with cold brine seeping from
the base of the West Florida Escarpment.
Alvinocaris stactophila new species
Figures 5, 6, 7
Material .—\]SnM 234291, Holotype d,
USNM 234292, Allotype 9, USNM 234293,
Paratypes, 5 c?, 2 9; north central Gulf of Mexico
about 129 km (80 miles) S of Louisiana,
27°46.94'N, 9r30.34'W, 534 m, Johnson-Sea-
Link Dive 1879, 28 September 1986, Bush Hill
hydrocarbon seep. From Linda H. Pequegnat and
Randall Howard, LGL Ecological Research Asso-
ciates, Bryan, TX, supported by partial funding
for Minerals Management Service-Northern Gulf
of Mexico contract 14-12-0001-30212.
Measurements in mm. — Holotype 6, postor-
bital carapace length 7.0, rostrum 2.7, maximum
carapace height 5.3, total length about 25. Allo-
type 9 , same 6.8, 2.0, 5.9, 24. Paratype 6 , same
4.2, 1.9, 3.2, total length not measured; paratype
9, same, 4.9, 2.0, 4.1.
Description. — Integument thin, shining, mi-
nutely punctate. Rostrum (Fig. 5a, b) almost
straight, imperceptibly elevated above horizontal
in distal half, sharply pointed tip usually reach-
ing proximal level of second peduncular article of
antennule, but sometimes to proximal part of
third peduncular article; dorsal margin raised
into thin serrate crest containing 12-17 teeth,
strongest in central sector of row, with about 1/2
length of crest continued onto carapace; ventral
margin less prominent and armed with 0 or 1
subterminal tooth; sample tooth formulas 11/1,
12/0, 14/1, 17/1; lateral carina broadened proxi-
mally and confluent with orbital margin. Cara-
pace (Fig. 5a, b) with buttressed acuminate an-
tennal spine distinct; pterygostomian spine
acuminate and prominent. Prominent anterior
antennal carina curving posteroventrally to in-
tersect obliquely with carina extending from
pterygostomial spine at about midlength of cara-
pace, associated groove continuing indistinctly
posteriad.
Abdomen (Fig. 5d) of both male and female
broadly arched dorsally, gradually tapering dis-
tally, narrowest part of sixth somite less than 1/2
(0.44) width of first somite; pleura of 3 anterior
somites broadly rounded, that of fourth somite
drawn posterolaterally to acuminate spine
flanked dorsally by 0-3 much more slender and
smaller spines; posterolateral corner of fifth
pleuron varying from strongly acuminate to
nearly rectangular and flanked dorsally by 2-5
spines analogous to those on somite 4; sixth
somite with middorsal length about 1.8 that of
fifth, broad-based midlateral spine overlapping
base of telson, smaller posterolateral spine acute;
fourth and fifth somites each with strong, posteri-
orly directed spine on sternite. Telson (Fig. 5e)
elongate subrectangular, length about 2.8 ante-
rior width, 5.2 posterior width, and about 1.7
length of sixth somite, not including posterior
spines; armed with 5-8 dorsolateral spines of
272
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
Figure 5 — Alvinocaris stactophila, holotype 6: a cephalothorax and anterior appendages, lateral; 6 rostrum, eye, antennular
peduncle, antennal scale, dorsal of left side; c antennular peduncle, distal articles, mesial; d abdomen, lateral; e tail
fan. Allotype 9: cheliped, /"mesial, g lateral; h, i,j, k pereopods 2, 3, 4, 5. Scale = 2 mm.
273
FISHERY BULLETIN: VOL. 86, NO. 2
nearly uniform size, occasionally unequal in
number on either side; posterior margin convex,
armed with 2 principal spines at each corner and
8-12 feathered strong setae on distal margin be-
tween.
Eyes (Fig. 5a, 6) with cornea imperfectly devel-
oped; unfaceted though diffusely pigmented in
adults, but with internal facetlike pattern in
smaller individuals; ovate in outline though fused
to each other mesially, and each with an up-
turned spine on anterodorsal surface.
Antennular peduncle (Fig. 5a, b, c) reaching
beyond end of antennal scale; basal article 1.3
length of second and about 2.2 length of third, all
measured on dorsal margin; stylocerite well sepa-
rated from peduncle, tapering to slender elongate
tip variably reaching as far as midlength of sec-
ond article; basal article with distodorsal margin
exceeded by rostral tip though extended into
strong lateral spine reaching level equal to that of
stylocerite and closely appressed to second article,
much smaller distomesial spine slightly diver-
gent; shorter second article with stronger
mesiodistal spine. Dorsolateral flagellum about
twice length of carapace, thickened in basal half;
ventromesial flagellum somewhat shorter.
Antennal scale (Fig. 5a, b) about twice as long
as wide, distolateral tooth strong, falling short of
broadly rounded distal margin of blade; basal ar-
ticle with small distal spine ventrally; flagellum
slightly exceeding length of body.
Mandibles (Fig. 6a) similar, with 2-segmented
palp, incisor process broad and armed with 7 mar-
ginal teeth, slender molar process simple, diver-
gent, its narrowly rounded tip minutely setose.
First maxilla (Fig. 66) with proximal endite
asymmetrically oval-triangular, distal margin
bearing about 25 long setae; distal endite with
narrowed base but broadened distally, armed
with many short spines on mesial margin and
with scattered longer spinules submarginally and
marginally beyond either end of spine row; palp
scarcely bifurcated, with long distal spine on ob-
solescent proximomesial branch and shorter adja-
cent submarginal spine and tangled setae on dis-
tal branch.
Second maxilla (Fig. 6c) with proximal endite
represented by 2 similar lobes; distal endite sub-
triangular, expanded mesiodistally and paral-
leled laterally by narrow somewhat twisted palp;
scaphognathite with anterior lobe rectangulo-
ovate, fringed with uniformly long, silky setae on
anterior and mesial borders, uniformly shorter
setae along entire lateral margin, posterior lobe
narrow and acuminate, fringed on blunt tip and
adjacent mesial margin by strikingly long, tan-
gled strong setae preceded proximally by shorter
setae similar to those on lateral margin.
First maxilliped (Fig. 6d, e) with irregularly
fusiform endite, short palp concealed and much
exceeded in length and size by leaflike exopod,
epipod obscurely bilobed; indistinct mesial lobule
on exopod possibly representing incipient lash.
Second maxilliped (Fig. 6f, g ) somewhat pedi-
form but flattened, mesial margin of articles
bearing long, feathered setae, mesial surface of
terminal article densely setose, exopod barely ex-
ceeding leaflike epipod.
Third maxilliped (Fig. 6h, i, j) slender, 5-
segmented, reaching beyond antennular pedun-
cle; terminal article trigonal in cross section,
tapered distally, bearing 3 spines, transverse
tracts of dense setae along mesial surface; similar
tract of setae on carpus and another less conspic-
uous group on merus-ischium, latter with stout
distolateral spine at articulation with carpus;
exopod much reduced, subtriangular, without
lash.
First pereopods (Figs. 5f, g; 7c, d) chelate,
subequal and sexually dimorphic, at least in fully
mature individuals; fingers curved ventrally and
slightly laterad; dactyl more slender than fixed
finger, tips varying slightly in relative length;
mesial surface of each finger convex, lateral sur-
face concave; prehensile surfaces uniformly off-
set, closing without gape, each armed with row of
almost uniform teeth so closely set as to be almost
contiguous, line of sensory hairs mesial to cutting
edges, acute tip of dactyl slightly spooned by elon-
gate teeth slanted distad and curving around ex-
ternal edge. Leg exceeding third maxilliped by
length of fingers in holotype male, but shorter
than third maxilliped in other individuals. Palm
inflated in holotype male, length 1.4 greatest
height and longer than fingers; palm relatively
shorter in allotype female and other individuals
examined, 0.3 length of fingers. Carpus shorter
than palm in holotype but longer than palm in
remainder of specimens examined, bearing
oblique ventral crest ending in strong distolateral
spine and flanked mesially by patch of setae on
polygonal raised area; notch above spine
smoothly concave and opposing low ridge ending
in small rounded spine on heel of palm; shallowly
concave anteromesial margin of carpus leading
dorsally to 2 low rounded lobes. Merus somewhat
swollen in distal half, distinct from ischium but
fused to it, neither armed.
274
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
1.
2.
Figure 6. — Alvinocans stactophila, allotype 9: a mandible; b maxilla 1; c maxilla 2; maxilliped 1, d ventral, e dor-
sal; maxilliped 2, /"ventral, g dorsal; maxilliped 3, h ventral, ; dorsal, y exopod. Paratype 6: k endopod of pleopod 1; I appendix
masculina, pleopod 2. Scales: 1 id-g, k) = \ mm; 2 (a, 6) = 0.5 mm; 3 (/) = 0.3 mm; 4 (c) = 1 mm.
275
FISHERY BULLETIN: VOL 86, NO. 2
Figure 7. — Parts of Alvinocaris chelae viewed by SEM. A. stactophila: fingers of small chela, a mesial, b dorsal; fingers of large
chela showing finely toothed opposed edges near tips, c mesial, teeth flush with convex surface, d lateral, teeth marginal on spooned
tips, with points rounded on dactyl, acute on fixed finger. A. muricola: e fingers of small chela, lateral; /large chela and distal
part of carpus, lateral. Scales: 100 p.m, d; 200 ^.m, a-c; 500 |j.m, e; 1 mm, f.
276
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
Figure 7. — Continued. — Alvinocaris muricola : fingers of large chela showing finely toothed opposed edges neair tips, g teeth flush
with convex mesial surface, h spooned lateral surface of same, points rounded on dactyl, acute on fixed finger; close-up lateral view
of teeth and associated sensory setae, teeth of fixed finger in foreground and of dactyl in background, i near distal end of fingers, y
near midlength of fingers; k sensory seta showing 2 rows of sensillae on concave surface. A. markensis: I fingers of small chela,
mesial view of distal part. Scales: 3 jjim, k\ 20 (jim, ;; 30 \i^m,j; 50 p.m, h\ 100 M-m, /; 200 ^.m, g.
277
FISHERY BULLETIN: VOL. 86, NO. 2
Second pereopod (Figs. 5/i; 7a, b) shorter and
more slender than first, reaching to between mid-
length and end of antennal peduncle; fingers
slightly longer than palm, similar in size and
shape, opposed edges without gape, each pecti-
nate with single row of teeth in distal half di-
rected obliquely distad and increasing slightly in
size to end in noticeably stronger tooth crossing
opposite member when closed, but spineless prox-
imally; carpus slender, about 1.2 longer than
chela; merus and ischium unarmed.
Third to fifth pereopods (Fig. 5i,j,k) similar in
length and structure, third reaching beyond an-
tennal scale by about 0.3 length of propodus.
Length articles of these legs in allotype 9 , mm:
3rd
4th
5th
ischio-
merus
4.48
4.89
4.16
carpus propodus
2.30
2.18
2.24
3.20
3.39
4.22
dactyl
0.48
0.48
0.48
Each short dactyl armed with about 6 corneous
spines on flexor surface, grading from small prox-
imally to longest and strongest distally; carpus of
each with distodorsal extension projecting as a
stop along proximal part of propodal extensor sur-
face; third leg stronger, at least in merus-
ischium, than fourth and fifth; merus of third and
fourth with ventral spine at 1/3 and 2/3 length,
distal one strongest, fifth without spines; ischium
of third with 2 spines in line with those on merus,
that of fourth and fifth spineless.
Pleopods well developed, pair 1 with endopods
about half length of expods in both sexes, endopod
of male (Fig. 6j) with asymmetrical mesial exten-
sion, that of female tapering to acute tip; pair 2
with appendix masculina of male (Fig. 6^) bear-
ing distal cluster of about 9 strong straight spin-
ules extending beyond level of simple slender ap-
pendix interna.
Uropod (Fig. 5e) with rami subequal in length,
slightly exceeding distal end of telson, lateral
ramus with movable spine mesial to smaller dis-
tolateral tooth, diaeresis sinuous.
Etymology . — The name is from the Greek
"stactos", oozing out or trickling, and "philos", to
love, for association of the species with hydrocar-
bons seeping from the substrate.
Remarks. — Alvinocaris lusca and the three
new species of Alvinocaris described here exhibit
minor differences that are highlighted in the key
to species given above, but their similarities seem
far more significant; i.e., general body appear-
ance and strength of integument, shape of ros-
trum (although that of A. stactophila sometimes
lacks ventral teeth), shape and general armature
of tail fan, blindness, and general structure of
appendages, including mouthparts. Some minor
differences that may be mentioned are features
such as number of incisor teeth on the mandible,
number of spines on the first maxilla, shape of the
second maxilla, lack of meral spines on pereopod
5 in A. stactophila, unequal distribution of spines
on ischia of pereopods 3-5 in the three species,
and shape of the endopod of male pleopod 1 and
appendix masculina (though males of A. marken-
sis are not yet known).
Each of these species lives in a distinctive ben-
thic environment, but all share similarities that
suggest dependence on a chemotrophic bacteria-
based food chain (Childress et al. 1986). Van
Dover et al. (in press) provide evidence from mor-
phology, behavioral and gut content analyses of
the similar Rimicaris exoculata Williams and
Rona that indicates a bacterial diet grazed from
surfaces of hydrothermal chimneys, although di-
rect observations of bacteria within the digestive
tract could not confirm this. The distinctively
spoon-shaped chelae of the first pereopods of both
Alvinocaris and Rimicaris species, with unbroken
comb of exceedingly fine teeth on the prehensile
edges, could be an adaptation for scooping or
sweeping bacteria toward the mouthparts.
Williams and Chace (1982) described the first
chelae of A. lusca as convex on the extensor sur-
face and concave on the flexor surface, but they
also said (p. 142) that the outer surface of the
fingers is convex and the inner surface concave.
The latter is misleading because in full extension
the convex side of the chela is mesial and the
concave side lateral. It is not yet known how these
appendages are used, but certainly the chelae can
be folded compactly against the leg's proximal
articles, and in the related Rimicaris exoculata
and R. chacei (Williams and Rona 1986) these
legs seem very mobile. Sensillae flanking prehen-
sile surfaces of the fingers seem well adapted to
aid feeding on finely particulate matter. More-
over, the species of Rimicaris have exceedingly
setose mouthparts.
In species of both genera, the second pair of
pereopods have much smaller chelae with fingers
bearing long sensory setae and spines on the pre-
hensile edges that are seemingly adapted for
278
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
grasping. For mobile animals of this morphologi-
cal makeup, the most likely feeding methods in
the stated environments would seem to be bacte-
rial concentration, along with secondary preda-
tion and scavenging.
ANOMURA: GALATHEIDAE
Munidopsis alvisca new species
Figure 8
Materm/. —USNM 234294, 9 Holotype,
USNM 234301, 9 Paratype, Guaymas Basin,
Golfo de Cahfornia, 27°00'N, lir25'W, 2,008 m,
Aluin Dive 1616, 8 August 1985, pilots J. Hardi-
man and R. Wilkes, observer J. F. Grassle. From
J. F. Grassle, Woods Hole Oceanographic Institu-
tion, Woods Hole, MA.
USNM 234295, 6 Paratype, Explorer Ridge,
Magic Mountain, 49°45.6'N, 130°16.16'W, 1,818
m, Pisces IV Dive P-1494, Coll. No. 1877, Gulati
Gusher-base, 1 July 1984, pilots-observers. Wit-
combe, Johnson, Tunnicliffe. USNM 234296, 9
ovig. Paratype, Explorer Ridge, Upper Magic
Mountain, 49°45.5'N, 130°16.1'W, 1,812 m,
Pisces IV Dive P-1497, Coll. No. 1873, Lunch
Hour Vent, 4 July 1984, pilots-observers. Shep-
herd, Juniper, Johnson. USNM 234297, 9 ovig.
Paratype, same. Coll. No. 1875, Crab Vent.
USNM 234298, 9 ovig. Paratype, same. Coll. No.
1875. USNM 234299, 2 9 ovig. Paratypes, Juan
de Fuca Ridge, Limbo Vent ( = 3 m from Holland's
Hillock Axial Seamount), 45°55'N, 130°03'W,
1,545 m, Pisces IV Dive P-1732, Coll. No. 1934, 2
August 1986, pilots-observers, K. Shepherd,
R. Embley, J. Franklin. From Verena Tunnicliffe,
Biology Department, University of Victoria, B.C.,
Canada.
Measurements in mm. — Holotype 9, carapace
length including rostrum 23.7, margin of orbit to
posterior edge of carapace 18.6, maximum cara-
pace width 15.7; Paratype 9 234301, same, 27.9,
20.8, 17.3; Paratype 6 234295, same, 13.8, 10.2,
8.4.
Description. — Carapace (Fig. 8a, c) exclusive
of rostrum distinctly longer than broad, moder-
ately arched transversely; anterior and posterior
cervical grooves apparent, depression in anterior
part of cardiac region; short moderately devel-
oped rugosities on each anterior branchial region,
but more distinct and transversely developed
rugae on each posterior branchial region, with
tendency to being continuous across anterior and
central part of cardiac region; posterior margin
with median concavity. Rostrum narrowly tri-
angular, concave dorsal surface smoothly curv-
ing to upturned tip exceeding eyestalks by more
than twice their length, distinct carina bearing
almost imperceptible scalelike rugae diminish-
ing to obsolescence on gastric region. Frontal
margin with broad angle lateral to eyestalk
followed by concave raised and sparsely orna-
mented margin ending in antennal spine followed
in turn by almost rectangular but acute antero-
lateral angle. Lateral plate obliquely rugose, pro-
jecting anteriorly below antennal peduncle, its
rather angular tip minutely but bluntly bi-
spinose.
Abdomen (Fig. 86) unarmed; transverse ridge
of segments 2 and 3 smooth, that of segment 4
obsolescent; segments 5 and 6 smooth.
Eyes (Fig. 8a , c ) moderate in size; well exposed,
smoothly ovate cornea cupped within broad based
movable ocular peduncle extended into elongate
mesiodorsal spine, directed obliquely upward at
low angle and ornamented with tiny irregular
obsolescent spinules, and much shorter mesioven-
tral spine.
Basal article of antennular peduncle with dis-
tal margin irregularly crenulate, slender dorso-
lateral spine and broader lateral spine flanked by
cluster of irregular small spinules, an obsolescent
mesiodorsal spine present. Antennal peduncle
with fixed basal article extended into stout, flat
ventral spine and shorter crenulate lateral spine;
succeeding articles short, second bearing stout
lateral angle, third unarmed, fourth with scal-
loped distal margin, its dorsomesial projection
spinelike.
Third maxilliped (Fig. 8e ) with ischium shorter
than merus, bearing mesial crest armed with
finely uniform, evenly spaced corneous teeth.
Basis with 2 low spines in line with crest on is-
chium. Merus with obsolescent spine at pos-
teromesial corner, mesial margin usually with
another at level of propodo-carpal joint, followed
after an interval by an obscure tubercle, and then
by a more prominent spine at base of convex dis-
tal margin; stronger spine at anterolateral cor-
ner; lateral margin broadly arched. Carpus,
propodus, and dactyl folded on merus-ischium
and about as long as those two articles together,
dense setation on dorsal surface of each, and dis-
tally on prehensile surface of propodus and
dactyl. Sternite (Fig. 8d) at base of third maxil-
279
liped with convex crenulate anterior margin on
mesial lobe, lateral lobe angular.
Epipods absent from pereopods.
Chelipeds (Fig. SP subequal, ornamented with
variably ciliate rugosities tending to arrange-
ment in longitudinal tracts; ischium with mesial
FISHERY BULLETIN: VOL. 86, NO. 2
fidge bearing subterminal spine and obsolescent
irregular subsidiary spines; merus rough, bear-
ing 3 mesial spines, 1 distodorsal spine, and a
smaller distoventral spine; carpus with 2 spines
in dorsolateral row paralleled by less prominent
ventrolateral row; palm with spines on prominent
Figure 8. — Munidopsis alvisca , holotype 9 : a carapace, eyes and right antenna, dorsal; b abdomen, somites 2-4 in folded
position; c part of cephalothorax and anterior appendages, lateral; d stemites at base of third maxilliped and chelipeds; e
left third maxilliped, merus and ischium; g left second pereopod. Paratype 9 234301: /"right cheliped. Scales: 1 (a, 6) = 5
mm; 2 (c) = 5 mm; 3 if, g) = 3 mm; 4 (d, e) = 1 mm.
280
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
dorsal ridge, stronger on right than on left; fin-
gers longer than palm, spooned at tips, prehensile
edges close fitting, entire, but small basal tooth of
fixed finger opposed by notch in prehensile edge
of dactyl.
Walking legs rather long, first walking leg
(Fig. 8^) reaching almost to tip of chela, second
and third reaching about to base of dactyl on pre-
ceding leg; corresponding articles of respective
legs approximately equal in length except for
meri which decrease posteriorly; each merus with
rounded, rugose dorsal crest ending in distal
spine; each carpus with longitudinal dorsal and
dorsolateral rib ending in more or less well-
developed spine, and often with secondary
spine(s) on distal margin between them; each
propodus slender, bearing small movable spine
distolaterally at base of dactyl; each dactyl slen-
der, acute corneous tip preceded by row of 12 or
more movable spines on prehensile edge. Slender
fifth leg with well-developed cleaning brush on
more or less flattened dactyl opposed by similar
setae on distal end of propodus.
Variation . — There is minor variation in orna-
mentation of the specimens available for study,
but none of it is associated with the disjunct dis-
tribution in the Golfo de California and the north-
eastern Pacific.
Remarks. — The specimens reported here were
taken around hydrothermal vent sites discussed
by Canadian American Seamount Expedition
(1985), ASHES Expedition (1986), and Tunni-
cliffe et al. (1985, 1986). Munidopsis has been
sighted at three other sites along Juan de Fuca
Ridge, but the only specimens collected are those
listed above (V. Tunnicliffe^).
Comparisons of Munidopsis alvisca with previ-
ously described species of the genus are aided by
reference to A. Milne Edwards (1880), Milne Ed-
wards and Bouvier (1897), Chace (1942), Sivert-
sen and Holthuis (1956), and Ambler (1980). Lack
of epipods on the pereopods immediately sepa-
rates M. alvisca from species such as M. crassa
Smith, 1885 and M. subsquamosa Henderson,
1885 which it superficially resembles. Both of the
latter species have relatively prominent rugae
and spines on the cephalothorax and legs whereas
M. alvisca has fairly smooth ornamentation on
2Verena Tunnicliffe, Department of Biology, University of
Victoria, P.O. Box 1700, Victor, B.C., Canada V8W 2R2, pers.
commun. 1987.
these body parts, except for minor development of
spines on the lateral carapace margin anteriorly.
The rostrum of all of these species is narrowly
triangular, curves moderately upward to the tip
and bears a middorsal carina, but the carina in
M. alvisca bears almost imperceptible scalelike
rugae and diminishes to obsolescence on the gas-
tric region whereas in both M. crassa and M. sub-
squamosa the carina is varyingly rugose, rather
strongly so in the former, and maintains this or-
namentation onto the gastric region. Moreover,
M. crassa bears tiny irregular marginal spines on
the rostrum.
Spination of the merus of the third maxilliped
is far weaker in M. alvisca than in the other two
species discussed, and both the anterolateral
spine of the ischium and the crenulate margin of
the crest on the ischium are less developed than
in them. On the other hand, M. alvisca possesses
both mesiodorsal and mesioventral eye spines
whereas M. subsquamosa and M. crassa lack the
mesioventral spine.
More distant comparisons seem inappropriate
because of different body proportions and orna-
mentation, rostral width, length, elevation and
spination, and structure of the eye and third max-
illiped. The keys for identification by both Chace
(1942) and Pequegnat and Pequegnat (1970), for
example, though strictly applicable to species of
the Atlantic basin, would ally M. alvisca with
M. aries (A. Milne Edwards, 1880), a much larger
species with broader cephalothorax and rostrum,
eyes almost hidden from dorsal view, and with
less transverse ornamentation. The revised ver-
sion of this key by Pequegnat and Pequegnat
(1971) would place M. alvisca in the couplet space
occupied by M. sundi Sivertsen and Holthuis,
1956, a species with superficially similar shaped
cephalothorax, but densely clothed with short
setae.
Etymology. — The name is an acronym taken
from names of the deep submersible vessels used
in collecting the species, Alvin and Pisces IV.
BRACHYURA: BYTHOGRAEIDAE
Bythograea tnesatlantica new
species
Figures 9, 10
Materia/.— USNM 234300, Holotype 9, Mid-
Atlantic Rift Valley about 70 km south of Kane
281
FISHERY BULLETIN; VOL. 86, NO. 2
Fracture Zone (see: Kong et al. [19851; Leg 106
Shipboard Scientific Party 1 19861; Ocean Drilling
Program Leg 106 Scientific Party [1986]),
23°22.09'N, 44°57.12'W, 3,437 m, Aluin Dive
1683, MARK vent, Stn. 1, scoop, 30 May 1986,
pilot D. Foster, observers S. Humphris and J. Ed-
mond. From NSF-Leg 106-Ocean Drilling Pro-
gram, NSF Grant OEC-8311201 to J. F. Grassle,
Woods Hole Oceanographic Institution, Woods
Hole, MA.
Measurements in mm. —
Carapace
Length
13.8
Width
23.3
Depth of cephalothorax
8.1
Frontoorbital width
7.7
Propodus lower margin
R 15.5
L 15.2
Dactyl length
P':ilm
8.3
8.3
1 aini
Height
7.9
7.8
Thickness
4.9
5.1
Description. — General aspect similar to that of
B. thermydron, cancroid, depressed. Carapace
(Figs. 9, lOd) broad, transversely elliptical, its
rounded lateral angles displaced somewhat ante-
riorly; almost flat in middle dorsally, very
slightly arched from anterior to posterior and
near lateral margins; anterolateral region pro-
duced, margin not toothed; surface finely granu-
late anteriorly and laterally, smooth but
minutely punctate to unaided eye over posterior
2/3 to 3/4; regions indistinct. Frontoorbital width
ca 1/3 carapace width.
Front almost evenly rounded and somewhat de-
flexed, projecting over folded antennules, shallow
median depression continued onto protogastric
region giving faint suggestion of bilobation; mar-
gin beaded with line of fairly uniform granules,
closely packed on anterior and anterolateral parts
but diminishing almost to obsolescence near or-
bits. Arcuate tract of scattered punctations
sweeping across anterolateral, hepatic, orbital,
protogastric, and metagastric regions. Trans-
verse tract of rather prominent granules at rear
edge of protogastric region. Carapace with
smooth part behind these anterior areas micro-
scopically granular and punctate anteriorly,
grading posteriorly into almost featureless sur-
FlGURE 9. — Bythograea mesatlantica , holotype 9: dorsal. Scale = 3 mm.
282
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
Figure 10. — Bythograea mesatlantica . holotype 9: right chela, a frontal view, b fingers viewed from tips; c left chela, frontal
view; d left side of cephalothorax in frontal view showing anterolateral pigmented spot, eye, antennules, antennae, and mouthparts
in situ; e mouth field showing third maxilliped turned to side, second maxilliped, first maxilliped with lacinia bearing tiny
"portunid lobe" at its mesial comer, partly hidden mandibles, and palate with patch of fine setae to either side of midline; /"abdomen
showing somites 3-6 and telson; g oviducal openings and parts of associated stemites. Scales: 1 (a-d, f-g) = 2 mm; 2 (e) = 1 mm.
283
FISHERY BULLETIN; VOL. 86, NO. 2
face. Protogastric, mesogastric, metagastric, and
cardiac regions poorly indicated; epibranchial
line indicated by small, light colored blotches
originating posterior to lateral angle at each side
and arching anteromesially over branchial re-
gion, then posteromesially toward mesogastric
region. Posterior margin concave and paralleled
by obsolescent postmarginal groove becoming
more pronounced along posterolateral margin.
Subhepatic and subbranchial areas orna-
mented with small granules, coarsest along upper
part of hepatic region but becoming finer and
more numerous near base of chelipeds.
Orbits sunk into essentially smooth transverse
concavity in anterolateral region confluent later-
ally at either side with a prominent irregularly
oval tan colored spot having very finely granu-
late, shallowly concave surface; somewhat in-
flated and irregularly granular suborbital area
almost fully visible in dorsal view, reaching level
of front, tilted anteroventrally from frontal plane
lateral to and almost at same level as epistome.
Eyestalks projecting anterolaterad, barely
movable, depressed and broadened to fit snugly in
orbit; unpigmented cornea terminal, subcircular,
narrower than eyestalk and anterolaterally ori-
ented.
Epistome (Fig. lOe) projecting well beyond
front, its anterior margin cut into 6 unequal
lobes; rather narrow and advanced submedian
lobes, separated by narrow deep notch, much
broader intermediate lobes and somewhat less
broadened lateral lobes less advanced.
Antennules folding transversely, stouter than
antennae, large bulbous basal articles contigu-
ous, concealed beneath front, interantennular
septum represented by minute remnant at upper
and lower edge of antennular fossa; slender
penultimate and terminal articles of peduncle
nearly equal in length, former slightly hollowed
laterally, latter slightly longer and more slender.
Flagella short; mesial 7-segmented ramus slen-
der; slightly shorter lateral ramus curved, multi-
segmented, thick at base but tapering to point,
dense mesial brush of long sensory setae in chord
of curve.
Antennal insertion mesial to eyestalk; pedun-
cle mesial to eyestalk, extending anteriorly or
anterolaterally in situ; fixed article broad but
short; first free article slender, ca 1.7 length of
second article; latter broadened distally; terminal
article short, its diameter only slightly greater
than that of flagellar base; flagellar length ex-
ceeding midline of front.
Mouth field (Fig. lOe) divergent anteriorly,
sides of its frame broadest posteriorly and some-
what swollen and granular at anterolateral cor-
ners, maximal inside anterior width about 1.4
minimal inside posterior width. Third maxil-
lipeds filling mouth field except for narrow gap of
nearly uniform width between ischia of en-
dognaths and rather irregular gap anteriorly be-
tween meri-carpi of endognaths and epistome; ex-
ognaths overlapping sides of mouth frame.
Endognaths with exposed surface bearing sparse,
sometimes linear, setose punctations; exposed
surface of ischium nearly smooth; elongate polyg-
onal in outline but primarily rectangular, great-
est (distal) width 1.1 narrowed part ca 1/2 length
from base; mesial margin straight through
most of its length but curved at each end, tooth-
less, bearing many stifl" straight setae, submar-
ginal zone somewhat thickened and flanked later-
ally by shallow longitudinal groove; anterior
margin nearly perpendicular to mesial margin
except for anteriorly projecting truncate lobe at
inner corner; lateral margin concave; posterome-
sial margin obliquely convex; basi-ischial suture
line visible posterolaterally. Merus slightly
narrower than ischium; low granules with tips
directed anteromesially along distal margin; ir-
regularly quadrate perimeter flanked by submar-
ginal thickened zone and groove similar to mesial
counterpart on ischium except on straight proxi-
mal side, anterolateral angle broadly rounded,
anteromesial angle at insertion of palp oblique;
mesial margin doubled anteriorly for reception of
folded palp, its ventral (exposed) side broadly
angled proximal to carpopropodal articulation;
posteromesial corner fitted to projecting lobe of
ischium, dorsal (hidden) side produced behind
carpus, its margin setose. Palp large, dactyl
reaching posteriorly about 1/4 length mesial mar-
gin of ischium. Carpus expanded distally, nar-
rowed proximally, bent nearly at right angle near
insertion and obscurely crimped inside angle;
dense tuft of setae on distooral surface. Propodus
wider than carpus, longer than broad, asymmetri-
cally ovate in ventral view; distal (longest) mar-
gin convex, densely beset with rows of strong ser-
rated setae, longest distally; distal tuft of such
setae on dorsal surface. Linear dactyl slightly
bent away from midline in distal part and setose
as propodus, especially on prehensile edge. Ex-
ognath narrow, not extending to full length of
merus; ventral surface slightly curved mesially to
fit closely against lateral side of endognathal is-
chium, with dorsomesial flange (widest distally)
284
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
fitting beneath latter; palp conspicuous, flagel-
lum densely beset with setae in hollow of curve.
First maxilliped with lacinia of endopod broad,
its distal edge 3-lobed and conspicuously though
not heavily setose; oblique mesial margin of
strongly advanced anterolateral lobe confluent
with broader gradually rounded and much less
advanced intermediate lobe, latter in turn fol-
lowed by still less advanced tiny mesial lobe, sep-
arated by a notch and directed anteromesially;
tuft of setae preceding notch.
Endostome large, divided by low median sagit-
tal ridge bifurcated somewhat anteriorly and
merging into projecting endostome; each half of
palate shallowly concave, crossed by low longitu-
dinal ridge slightly offset at its midlength and
trending anteromesially from near base of large
mandibular palp; ridge flanked laterally by irreg-
ular patch of velvety pubescence; smooth lateral
2/3 of palate receiving large efferent branchial
channels.
Chelipeds (Figs. 9, 10a, b, c) heavy, subequal;
integument punctuate on upper and extensor sur-
faces, obscure granulation on upper surfaces of
palms and on ridges or raised areas elsewhere;
chelae inflated, lower margin of palm arched
downward, its rather pronounced keel merging
into fixed finger; swollen palm with shallow exca-
vation proximally for reception of carpus in flexed
position, inner surface glabrous but drawn into
moderate and slightly granular elevation slightly
in front of proximal excavation. Fingers tan col-
ored in preservation (70% ethanol) and darkest
proximally, color of fixed finger not extending
onto palm; fingers not gaping, prehensile edges
entire except for obsolescent proximal tooth on
fixed finger of each hand; dactyl longer than rela-
tively straight fixed finger, arching down distally
to close in distal notch of spooned tip of fixed fin-
ger.
Carpus with extensor surface inflated, right
carpus with internal margin rounded, that of left
obscurely angled. Merus broadened mesially into
cristate flange angled distally for reception of car-
pus, strong granules in single line along inner
margin, outer surface rounded, strewn with obso-
lescent punctations and granules, latter most
prominent along distoventral tract.
Walking legs rather long, flattened, length de-
creasing posteriorly in order 3, 2, 1, 4; each with
dense patches of short darkened setae inter-
spersed with sparer longer setae on extensor sur-
face of carpus and propodus (as well as its lateral
side on legs 1 and 2), distoventral corner of car-
pus, and more extensively on dactyl; fifth legs
somewhat more flattened than others, propodi
relatively broader and not densely setose later-
ally. Mean maximum length of propodi about
twice width. Dactyls slightly longer than propodi,
narrowly lanceolate, shallow longitudinal
grooves on anterior and posterior surfaces ob-
scured by dense setae, tip stout, corneous. Merus
of each with upper margin finely granular, ante-
rior lower margin present throughout length but
posterior lower margin obsolescent proximally.
Sternum broadest between legs 1 and 2, nar-
rower posteriorly, glabrous beyond outline of ab-
domen.
Abdomen (Fig. lOf) ovate in outline, fully seg-
mented and densely fringed with plumose setae;
somite 1 slightly arched dorsally to fit contour of
adjacent carapace, somites 2-4 of about equal
length, somites 5 and 6 progressively longer; ab-
domen with greatest width at 4; telson nearly as
broad as somite 6, outline broadly arched distally.
Somites 2-5 bearing large, well-developed bi-
ramous pleopods, outer curved branch lying near
edge of abdomen and heavily beset with short
setae laterally and mesially, inner branch more
sparsely equipped with ovigerous setae and
jointed.
Female openings (Fig. 10^^) large, obscurely
subtriangular in outline.
Color in preservation predominantly off-white
except for fingers, matted setal tracts laden with
brownish finely particulate matter.
Remarks. — Brachyuran crabs that resemble
Bythograea were observed and reported by Rona
et al. (1986).
Bythograea mesatlantica differs in several re-
spects from Pacific members of the genus,
B. thermydron Williams (1980) and B. microps de
Saint Laurent (1984). Among obvious differences
from B. thermydron, the new species has even
less ornamentation on the carapace; it lacks a
distinct suborbital plate separated by a suture,
and the suborbital area is inflated, not flat and
inclined; there is a transverse concavity lateral to
each eyestalk that terminates near the very dis-
tinctive brown spot in the cuticle at either side of
the carapace; the eyestalk itself is shorter and
thicker than in B. thermydron and the shape and
position of the cornea differs. The ischium of the
third maxilliped is relatively shorter than in
B. thermydron and bears only sparse setiferous
punctations on the external surface, it lacks tiny
granules on the truncate lobe at the anterolateral
285
FISHERY BULLETIN: VOL. 86, NO. 2
corner, the submarginal thickened zone and
groove are less distinct; the merus is not tilted
dorsally in normal position, and the palp is rela-
tively shorter and club shaped rather than curved
like a knife edge along the prehensile edge. The
lacinia of the first maxilliped is more angular
anterolaterally and has a smaller "portunid lobe".
The epistome is less lobulate than in B. thermy-
dron and the concave palatal area has much less
setose covering. The nearly toothless chelae have
brown fingers and there is no dense patch of setae
on the inner side of the palms.
Comparisons with B. microps are necessarily
less complete because of the brief description of
the latter. The eyes are certainly not slender and
retracted in B. mesatlantica ; the chelipeds are not
noticeably dimorphic, and they are relatively
smooth rather than strongly granular and pilose
on the external surface as in B. microps.
The distinctive exocular spots on the carapace
seem similar to those noted on the chelipeds of
Hypsophrys noar Williams (1974, 1976) and Mu-
nidopsis lentigo (Williams and Van Dover 1983).
Their function is unknown.
Etymology. — From the Greek "mesos", middle
and "Atlantic", with reference to the Mid-
Atlantic Rift habitat.
ACKNOWLEDGMENTS
Contributors of specimens and crew members of
submersible vessels who helped to make the col-
lections are owed a special debt of gratitude for
securing the rare material described here. Donors
are acknowledged individually in each of the spe-
cies accounts. I am indebted to Keiko Hiratsuka
Moore for rendering the excellent illustrations, to
Ruth Gibbons for helping to produce the SEM
micrographs, and to F. A. Chace, Jr., R. B. Man-
ning, and B. B. Collette for critical review of the
manuscript.
LITERATURE CITED
Ambler, J. W.
1980. Species of Munidopsis (Crustacea, Galatheidae)
occurring off Oregon and in adjacent waters. Fish.
Bull., U.S. 78:13-34.
ASHES Expedition
1986. Pisces submersible exploration of a high-
temperature vent field in the caldera of Axial Volcano,
Juan de Fuca Ridge. Eos 67(44):1027.
Brooks, J M., M C Kennicutt II, C. R Fisher, S. A Macko,
K. Cole, J. J. Childress, R. R. Bidigare, and R. D. Vetter.
1987. Deep-sea hydrocarbon seep communities: Evidence
for energy and nutritional carbon sources. Science
238:1138-1142.
Canadian American Seamount Expedition
1985. Hydrothermal vents on an axis seamount of the
Juan de Fuca Ridge. Nature 313:212-214.
Chace, F A . Jr
1942. Reports on the scientific results of the Atlantis
expeditions to the West Indies, under the joint auspices of
the University of Havana and Harvard University. The
Anomuran Crustacea. I. Galatheidae. Torreia, Havana,
11:1-106.
Childress, J. J , C R Fisher, J M Brooks, M C. Kennicutt II,
R Bidigare, and A E Anderson
1986. Methanotrophic marine molluscan (Bivalvia,
Mytilidae) symbiosis: mussels fueled by gas. Science
233:1306-1308.
DE Saint Laurent, M.
1984. Crustaces Decapodes d'un site hydrothermal actif
de la dorsale du Pacifique oriental (13° Nord), en
provenance de la campagne frangaise Biocyatherm.
C. R. Hebd. Stances Acad. Sci. Ser. 3 Sci. Nat. (Paris)
299(9):355-360, plate 1.
KoNG. L., W B. F. Ryan, L. A. Mayer, R. S. Detrick, P J. Fox, and
K Manchester.
1985. Bare-rock drill sites, O.D.P. Legs 106 and 109;
evidence for hydrothermal activity at 23°N on the
Mid-Atlantic Ridge. EOS 66(46):936.
Leg 106 Shipboard Scientific Party (R. S. Detrick, J.
Honnorez, A. C. Adamson et al.).
1986. Mid-Atlantic bare-rock drilling and hydrothermal
vents. Nature 321:14-15.
Milne Edwards, A
1880. Reports on the results of dredging, under the
supervision of Alexander Agassiz, in the Gulf of Mexico
and in the Caribbean Sea, 1877, '78, '79, by the United
States Coast Survey Steamer "Blake," ....
VIII. — Etudes pr6liminaires sur les Crustac6s. Bull.
Mus. Comp. Zool. Harvard 8:1-68, 2 plates.
Milne Edwards, A., and E. L. Bouvier.
1897. Reports on the results of dredging, under the
supervision of Alexander Agassiz, in the Gulf of Mexico
(1877-78), in the Caribbean Sea (1878-79), and along
the Atlantic coast of the United States (1880), by the U.S.
Coast Survey Steamer "Blake" .... XXXV.
Description des Crustac6s de la famille des Galatheides
recuellis pendant I'exp^dition. Mem. Mus. Comp. Zool.
Harvard 19:1-141.
Ocean Drilling Program Leg 106 Scientific Party.
1986. Drilling the Snake Pit hydrothermal sulfide deposit
on the Mid-Atlantic Ridge, lat 23°22'N. Geology
14(12):1004-1007.
Pequegnat, L H., and W. E. Pequegnat.
1970. Deep-sea anomurans of Superfamily Galatheoidea
with descriptions of two new species. In W. E.
Pequegnat and F. A. Chace (editors). Contributions on
the biology of the Gulf of Mexico. Tex. A&M Univ.
Oceanogr. Stud. 1(5):125-170.
Pequegnat, W. E , and L H. Pequegnat.
1971. New species and new records of Munidopsis
(Decapoda: Galatheidae) from the Gulf of Mexico and
Caribbean Sea. Contributions on the biology of the Gulf
of Mexico. Tex. A&M Univ. Oceanogr. Stud. 1
(Suppl.):l-24.
RoNA, P. A., G. Klinkhammer, T. a. Nelson. J. H. Tefrey, and
H Elderfield.
1986. Black smokers, massive sulphides and vent biota at
the Mid-Atlantic Ridge. Nature 321:33-37.
286
WILLIAMS: NEW MARINE DECAPOD CRUSTACEANS
Saint Laurent, M., de. See De Saint Laurent.
SiVERTSEN, E., AND L. B HOLTHUIS
1956. Crustacea Decapoda (The Penaeidea and
Stenopodidea excepted). Rep. Sci. Results "Michael
Sars" North Atl. Deep-Sea Exped. 1910, 5(12):l-54,
plates 1-4.
tunnicliffe. v , m botros, m e de burgh, a dinet, h p.
Johnson. S. K Juniper, and R E McDuff
1986. Hydrothermal vents of Explorer Ridge, northeast
Pacific. Deep-Sea Res. 33:401-412.
TUNNICLIFFE, V , S K JUNIPER, AND M E DE BURGH
1985. The hydrothermal vent communities on Axial
Seamount, Juan de Fuca Ridge. Bull. Biol. Soc. Wash.
6:453-464.
Van Dover, C. L., B Fry. J F Grassle, S Humphris, and P. A.
RONA.
In press. Feeding biology of the Mid-Atlantic Ridge
hydrothermal vent shrimp: Functional morphology, gut
content analyses, and stable isotopic composition. Mar.
Biol. (Berl.).
Williams, A B
1974. A new species of Hypsophrys (Decapoda:
Homolidae) from the Straits of Florida, with notes on
related crabs. Proc. Biol. Soc. Wash. 87:485-492.
1976. Integumental organs of unknown function on
chelipeds of deep-sea crabs, genus Hypsophrys. J.
Morphol. 150:889-900.
1980. A new crab family from the vicinity of submarine
thermal vents on the Galapagos Rift (Crustacea:
Decapoda: Brachyura). Proc. Biol. Soc. Wash.
93:443-472.
Williams, A B., and F A. Chace, Jr.
1982. A new caridean shrimp of the family Bresiliidae
from thermal vents of the Galapagos Rift. J. Crust. Biol.
2:136-147.
Williams, A B , and P A. Rona.
1986. Two new caridean shrimps (Bresiliidae) from a
hydrothermal field on the Mid-Atlantic Ridge. J. Crust.
Biol. 6:446-462.
Williams, A. B , and C. L. Van Dover.
1983. A new species of Munidopsis from submarine
thermal vents of the East Pacific Rise at 21°N
(Anomura:Galatheidae). Proc. Biol. Soc. Wash.
96:481-488.
287
THE MEGALOPA STAGE OF THE GULF STONE CRAB, MENIPPE ADINA
WILLIAMS AND FELDER, 1986, WITH A COMPARISON OF MEGALOPAE
IN THE GENUS MENIPPE
Joel W. Martin,' Frank M. Truesdale,^ and Darryl L. Felder^
ABSTRACT
The laboratory-reared megalopa stage of the Gulf stone crab, Menippe adina , is described and illus-
trated and compared with megalopae of three other species of Menippe . The megalopa of M . adina
differs from that of Af. nodifrons in having serrate spines on the ventral margin of the dactylus of
pereiopod 5 and from that of M. rumphii in having spines on the dactyli of pereiopods 2-5 and a more
quadrate carapace. The megalopa of the morphologically similar A/ . mercenaria was also reared in the
laboratory, and selected characters are described and compared with the megalopa of A/, adina;
megalopae of the two species differ only slightly. Megalopae of M. adina taken from field collections
made off South Texas, U.S.A., were compared with and were found to be consistent with laboratory-
reared M . adina megalopae.
Stone crabs of the genus Menippe are large xan-
thid crabs common along the eastern coasts of the
United States and Mexico from North Carolina to
Yucatan, the Bahamas, Cuba, and Jamaica
(Rathbun 1930; Felder 1973; Williams 1984;
Williams and Felder 1986). Recently the
"common" stone crab, Menippe mercenaria (Say,
1818), was divided into two species: Menippe mer-
cenaria (Say) (restricted), known from the east
coast of the United States, the Caribbean, and the
west coasts of Florida and Yucatan, and Menippe
adina Williams and Felder, 1986, known from the
northwestern Gulf of Mexico; hybridization of the
two species occurs in northwest Florida (see
Williams and Felder 1986). These two species
(primarily M. mercenaria ) support an important
stone crab fishery in the southern United States
and Mexico (Williams and Felder 1986) and con-
sequently have been the subject of numerous in-
vestigations. Despite this interest, the complete
larval developments of both commercial species of
Menippe remain unknown. For M. mercenaria
(Say), Hyman (1925) described a prezoea and first
zoeal stage, and Porter (1960) described six zoeal
stages reared in the laboratory. Unfortunately,
Porter did not describe the megalopa stage, pre-
iLife Sciences Division, Natural History Museum of Los An-
geles County, 900 Exposition Boulevard, Los Angeles, CA
90007.
2School of Forestry, Wildlife, and Fisheries, and Louisiana
Agricultural Experiment Station, Louisiana State University,
Baton Rouge, LA 70803.
3Department of Biology and Center for Crustacean Research,
University of Southwestern Louisiana, Lafayette, LA 70504.
sumably because he considered it a postlarva and
not a true larval stage. An unpublished but often-
cited report by Kurata"* included descriptions of
the zoeal stages of M. mercenaria and a brief
sketch of the megalopa; Kurata's description of
the megalopa did not include morphology of the
pleopods, pereiopods, or mouthparts.
Because of recent interest in the phylogenetic
significance of the brachyuran megalopa (see Rice
1981a, in press; Martin in press) and postlarval
stages (Martin et al. 1984; Felder et al. 1985), and
because of the potential importance of stone crab
larval biology to aquaculture, it is surprising that
the megalopae of M . mercenaria and M . adina
remain undescribed. The present paper describes
the laboratory-reared megalopa of the Gulf stone
crab, Menippe adina Williams and Felder, and
compares it with field collections of the same spe-
cies from south Texas, laboratory-reared megalo-
pae of M. mercenaria, and all previously de-
scribed megalopae of the genus Menippe: Menippe
mercenaria (Say, 1818) (as described by Kurata
fn. 4); Menippe nodifrons Stimpson, 1859 (as de-
scribed by Scotto 1979); and Menippe rumphii
(Fabricious, 1798) (as described by Kakati 1977).
MATERIALS AND METHODS
A large ovigerous M . adina was collected from
Manuscript accepted February 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.
4Kurata, H. 1970. Studies on the life histories of decapod
Crustacea of Georgia. Part IIL Larvae of decapod Crustacea of
Georgia. Unpubl. rep., 274 p. University of Georgia Marine
Institute, Sapelo Island, GA.
289
FISHERY BULLETIN: VOL. 86, NO. 2
shallow waters of the northern Gulf of Mexico
near Grande Terre, LA, in May 1982 and held in
a small aquarium at room temperature. After the
eggs hatched, the zoeal larvae were given fresh
seawater and newly hatched Artemia nauplii
daily. Exuviae as well as dead and some living
megalopae were preserved in 70% ethanol. Draw-
ings were made with the aid of a Wild^ M-5
stereoscope and a Wild M-11 compound stereo-
scope, both with camera lucida; accuracy was ver-
ified with a Nikon Optiphot. Measurements were
made with an ocular micrometer. Ten laboratory-
reared megalopae were examined, measured, dis-
sected, and compared with megalopae from field
collections made in 1973 off south Texas. Com-
parisons with M. mercenaria are based on
laboratory-reared M . mercenaria megalopae from
two females collected on 13 August 1987 from the
Indian River system, north of Ft. Pierce, FL. Eggs
of these two females hatched on 21 August 1987,
and the megalopa stage was first reached after 17
days in mass culture aquaria (30%o salinity,
25°C, 12h:12h light/dark regime). Descriptions of
setation for all appendages proceed from proximal
to distal. Specimens examined under the scan-
ning electron microscope (SEM) were prepared
according to procedures outlined by Felgenhauer
(1987) but without postfixation in osmium tetrox-
ide and with 100% ethanol, rather than amyl ac-
etate, as the transitional fluid. Sibling megalopae
and field collections have been deposited in the
U.S. National Museum of Natural History,
catalogue No. USNM 229962 (laboratory-
reared M. adina), USNM 229961 (field-collected
M. adina), and USNM 229963 (laboratory-reared
M. mercenaria).
RESULTS
Carapace (Figs., lA, B, C, 3 A).— Length 1.67
mm, width 1.45 mm (A'^ = 10). Subquadrate, with
2 lateral prominences on each side; dorsoven-
trally thick, with minute tubercle centrally lo-
cated. Posterior border fringed with numerous
short setae; lateral margin with few scattered
setae. Rostrum ventrally deflexed, nearly verti-
cal, with deep medial depression, rounded anteri-
orly. Angular interorbital prominences extend
ventrally between orbit and antennule. Chroma-
tophores variable in placement, but almost al-
ways found in areas indicated in Figure IB.
Eyes (Figs. lA, B, C, 3A). — Large, exposed; eye-
stalks sometimes with 2 or 3 short, simple ante-
rior setae, always with posterodorsal chroma-
tophore.
Abdomen (Fig. lA, B). — Subequal in length to
carapace. All pleura with rounded posterolateral
angles. All somites with sparse setae dorsally;
somites 2-5 always with elongated chroma-
tophores.
Telson (Fig. IG). — Broadly rounded with vari-
able setation, occasionally with pair of small pos-
terior spines (as in Figure IB).
Antennule (Fig. IK). — Biramous; peduncle 3-
segmented, with variable setation. Basal seg-
ment of peduncle large, bulbous, always with
large chromatophore; middle segment subcylin-
drical with 0-2 distal setae; distal segment ovoid
with scattered short setae. Lower ramus
1-segmented with 6-8 setae; upper ramus
5-segmented with aesthetascs arranged in tiers,
usually 0, 7, 8, 6, 4 subterminal plus 3 terminal,
with short setae sometimes present on segments
2 and 4 (note: all aesthetascs not illustrated).
Antenna (Fig. IJ). — Flagellum 12-segmented
(sometimes 11), with 3 peduncular articles and 8
or 9 flagellar articles (see Rice, in press, for cor-
rect number of antennal segments in megalopae);
setation variable, usually 2, 3, 2, 0, 0, 2, 4, 0, 4 or
5, 1, 4, 4.
Mandibles (Fig. 2F). — Asymmetrical, with
broadly rounded spade-shaped cutting edges; palp
2-segmented with setation 0, 11-14.
Maxillule (Fig. 2E). — Protopodite with 1 or 2
long plumose setae on posterodorsal margin; en-
dopodite 2-segmented with setation 1, 2 subtermi-
nal plus 2 terminal; basal endite with 29-35
spines and setae; coxal endite with 13-16 spines
and setae.
Maxilla (Fig. 2D).— Scaphognathite with 70-78
fringing setae and 0—6 setae on blade; endopodite
unsegmented with 0 or 1 distolateral seta and 4 or
5 basal plumose setae; basal endite bilobed with
setation variable, usually 8-10, 9-11; coxal en-
dite bilobed with setation usually 7, 9 or 10.
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
Maxilliped 1 (Fig. 20). — Exopodite 2-seg-
mented, with setation 2 or 3, 5-7. Endopodite
290
MARTIN ET AL.: MEGALOPAE OF STONE CRABS iMENIPPE)
A-C
D-K-
Figure l. — Megalopa of the Gulf stone crab, Menippe adina. A, entire animal, lateral view; B, same, dorsal view; C, frontal view
of rostrum and eyes; D, pleopod 1; E, pleopod 4; F, pleopod 5; G, telson and posterior part of sixth abdominal segment; H,
dactylus of pereiopod 3; I, dactylus of pereiopod 5; J, antenna; K, antennule. Both scale bars = 1.0 mm.
291
FISHERY BULLETIN: VOL. 86. NO. 2
Figure 2. — Megalopa oi Menippe adina, mouthparts. A,
third maxilliped; B, second maxilliped; C, first maxil-
liped; D, maxilla; K,- maxillule; F, mandible. Scale
bar = 0.5 mm.
292
MARTIN ET AL.: MEGALOPAE OF STONE CRABS (MENIPPE)
unsegmented with 6-8 setae arranged as shown.
Basal endite setation 28-33; coxal endite setation
15-17. Epipodite with 22 or 23 long, minutely
plumose setae, appearing simple under low mag-
nification.
Maxilliped 2 (Fig. 2B). — Exopodite 2-segmented,
with setation 3, 5-8. Endopodite 4-segmented,
with setation usually 5, 2 or 3, 5 or 6, 9 or 10;
distal segment with 4 or 5 stout serrate setae.
Epipodite with 9 or 10 long minutely plumose
setae.
Maxilliped 3 (Fig. 2A). — Exopodite 2-seg-
mented, with setation 0 or 1, 6-8. Endopodite
5-segmented, with variable setation, usually 18-
20, 15 or 16, 5-9, 6-8, 7-10; ischium with scal-
loped medial border. Epipodite with 18 long
minutely plumose setae on distal two-thirds plus
8-12 plumose setae on proximal one-third. Pro-
topodite setation variable.
Pereiopods (Figs. lA, B, H, I, 3B, C, D).— Che-
lipeds long, stout, subequal; dactylus with 4 irreg-
ular teeth; immovable finger with 3 teeth (Fig.
3B); tips of fingers overlap distally when approxi-
mated. No recurved hook on basi-ischium
(Fig. 3B). Second to fourth pereiopods similar;
dactylus with 5 (rarely 4) serrate spines ventrally
(e.g., Figs. IH, 3C, D); propodus with long ven-
trodistal spine (Fig. IH). Fifth pereiopod dactylus
(Fig. II) with 3 long pectinate setae, 1 markedly
toothed and concave (Fig. 3E), on distal ventral
border and 3 or 4 serrate spines ventrally.
Pleopods (Fig. ID, E, F). — Decreasing in size
posteriorly. Pleopod 1 (Fig. ID) with 19-22
plumose setae; endopodite with 3 or 4 hooked
setae. Pleopod 4 (Fig. IE) with 19-21 plumose
setae; endopodite with 3 or 4 hooked setae
(Fig. 3F). Pleopod 5 (uropod) (Fig. IF) with 12-14
plumose setae; basal segment lacking setae or
with 1 or 2 setae (field collections); endopodite
absent.
Color. — Overall coloration rose-orange, with
dark blue-black chromatophores located as shown
in Figure lA, B.
DISCUSSION
The genus Menippe de Haan, 1833, presently
contains about 8 species, only 3 of which occur in
North America. The megalopa stage is now known
for 3 species in the genus: M. rumphii (Fabricious,
1798), M. nodifrons Stimpson, 1859, and M. adina
Williams and Felder, 1986. In addition, selected
characters of M. mercenaria (Say, 1818) are pre-
sented here for comparison; some characters of
that species are also obtainable from an unpub-
lished report by Kurata (fn. 4) (see Table 1).
Laboratory-reared megalopae of M. adina were
virtually identical to megalopae presumed to be-
long to M. adina that were collected off south
Texas. Even meristic counts of the mouthpart se-
tation agreed exactly, with the only observed dif-
ferences being that field-collected megalopae
were slightly larger and occasionally bore 1 or 2
setae on the basal segment of the uropod. Thus,
we feel that our laboratory conditions have not
adversely affected development or introduced ab-
normal characters, and we have used these field
collections for the SEM figures of M. adina mega-
lopae (Fig. 3).
We expected to find that characters of the
megalopa of M. adina are similar to those de-
scribed by Kurata (fn. 4) for the morphologically
similar (in adulthood) M. mercenaria, a species
known to hybridize with M. adina (see Williams
and Felder 1986). In general this is true. How-
ever, some characters reported by Kurata differ
from our observations on M. adina and from our
laboratory-reared megalopae of M. mercenaria
(Fig. 4). Kurata mentioned (but did not illustrate)
Table 1 . — Comparison of characters in megalopae of the genus Menippe. Dash ( — ) indicates information not available from reference.
Size! (mm)
CL CW
Setation
Spinafion
Setation
pleopod 5
Palp of
mandible
Epipod of
Dactylus of
Menippe
Maxilliped 1
Maxilliped 2
Maxilliped 3
Pereiopods 2-4
Pereiopod 5
Reference
adina
mercenaria
mercenaria
nodifrons
rumphii
1.67 1.45
1.70 1.55
1.7-8 —
1.50 1.31
1.60 1.55
0, 11-14
0, 11-13
0, 10-13
0,0,9
22-23
20-23
12-20, 226
22
9-10
7-10
up to 10
8
18
18-20
18
18
5
4
4-5
5
0
4
4
0
0
0
12-13
11-13
11-12
11
12
Present study
Present study
Kurata3
Scotto 1979
Kakati 1977
^CL = carapace length; CW = carapace width.
2From a megalops hatched from a stage 6 (rather than the typical stage 5) zoea.
3See text footnote 4.
293
FISHERY BULLETIN: VOL. 86, NO. 2
Figure 3. — Scanning electron micrographs (SEM) of selected characters of Menippe megalopae (presumably M. adina ) collected in
south Texas. A, dorsal view of carapace (x 25); B, ventral view of chelipeds showing dentition of the fingers and lack of recurved
hook on ischium (x 37); C, dactyli of second (upper figure) and third pereiopods (x 230); D, higher magnification of ventral
dactylar spine indicated by arrow in C (x 1,900); E, endopod of third abdominal pleopod showing 4 dentate hooklike setae
(x 2,200); F, serrate setae (only 2 of 3 shown) of dactylus of pereiopod 5 (x 2,300).
294
MARTIN ET AL.: MEGALOPAE OF STONE CRABS iMENIPPE)
Figure 4. — Scanning electron micrographs (SEM) of selected characters of the laboratory-reared megalopa of Menippe merce-
naria. A, dorsal view of carapace (x 37); B, ventral view of chelipeds and third maxillipeds (x 55); C, dactylus of second
pereiopod showing ventral serrate spines (x 270); D, ventral spines on dactylus of fourth pereiopod with fewer spinules than
anterior pereiopod spines (x 2,000); E, dactylus of fifth pereiopod showing four serrate "sensory" setae with one (arrow) more
obviously serrate (x 190); F, higher magnification of concave serrate setae indicated by eutow in E (x 1,500).
295
FISHERY BULLETIN: VOL. 86, NO. 2
"about 9 small spines" on the ischium of the third
maxilliped; we found no spines on M. adina or M.
mercenaria (Figs. 3B, 4B), but it is possible that
Kurata was referring to the acute borders of the
scalloped medial margin (our Figure 2A), in
which case the 2 species are similar. Spination of
the ischium of pereiopods 1-3 differs also; Kurata
described 5 or 6, 2 or 3, and 1 small spine on the
ischia of pereiopods 1, 2, and 3, respectively,
whereas we did not notice this condition in M.
adina or M. mercenaria (see Figures 3B, 4B). Fi-
nally, Kurata (fn. 4: pi. 74, fig. E) illustrated no
spines on the ventral surface of the fifth pereiopod
dactylus of M. mercenaria; these spines are obvi-
ous on both species (Figs. II, 4E). We found few
differences between megalopae of M. adina and
M. mercenaria. General morphology of the cara-
pace and chelipeds, spination of the dactylus of
the pereiopods, and setation of the pleopods
agreed almost exactly (compare Figures 3 and 4).
Ventral dactylar spines on the posterior walking
legs of M. mercenaria were not so serrate as in M.
adina and were sometimes armed with only 2 or
3 large spinules rather than the numerous spin-
ules seen in M. adina (e.g., Fig. 3D) and in the
more anterior legs of M. mercenaria (see Figure
3C, D, 4C). Also, in all but 1 of the 9 megalopae of
M. mercenaria examined there were 4 (rather
than 3) long serrate setae on the dactylus of the
fifth pereiopod (Fig. 4E). As in M. adina, one of
these setae was more serrate and concave than
were the other long setae (Fig. 4E, F). However,
we have not examined mouthpart morphology of
M. mercenaria in the detail in which we described
M. adina, and so it is possible that additional
characters will be found to separate these 2
species at the megalopa stage.
The megalopa of M. adina is very similar to
that of M. nodifrons as described by Scotto (1979).
Although the 2 species differ in setation of some
of the mouthparts, this setation may differ from
side to side in a given individual. The salient
character that serves to separate megalopae of
these 2 species is the presence in M. adina of 4
stout serrate spines on the dactylus of pereiopod
5. Scotto (1979) figured only setae (and no spines)
on the dactylus of the fifth pereiopod in M. nod-
ifrons and the dactylar spines on other pereiopods
apparently are not serrate (Scotto 1979, fig. 9c,
pereiopod 3).
The megalopa of M. rumphii described by
Kakati (1977) differs from that of M. nodifrons,
M. mercenaria, and M. adina in having a more
ovoid carapace with the rostrum only slightly de-
flexed. Kakati did not describe the dactyli of
pereiopods 2-5 for M. rumphii, but his figure of
pereiopod 2 (1977:639, fig. 2, p. 50) does not show
stout ventral spines on the dactylus. Possibly
Kakati overlooked these spines; if not, the ab-
sence of these spines on pereiopod 2 would further
serve to separate the megalopa of M. rumphii
from those of A/, nodifrons, M. mercenaria, and M.
adina. All 4 species have been described as hav-
ing a rose-orange coloration in life.
Although Rathbun (1930) and Monod (1956)
synonymized M. rumphii with M. nodifrons, de-
scriptions of the zoeal stages of M. rumphii and M.
nodifrons by Kakati (1977) and Scotto (1979), re-
spectively, show that larvae of the 2 species differ
considerably. In the first zoeal stage, M. rumphii
exhibits elongated posterolateral processes on ab-
dominal segment 5 that extend posteriorly to
more than half the length of the telsonal furcae,
which lack spines. The first zoea of M. nodifrons
has similar posterolateral processes but these do
not extend posteriorly beyond the fork of the tel-
son; the telsonal furcae bear 1 dorsal and 2 lateral
spines each. These differences are not apparent in
later zoeal stages, but their presence in the first
zoeal stage and the differences noted in the mega-
lopa stage may be reason to question the syn-
onymy of these 2 species.
Xanthid larvae are known to be variable, and it
is often difficult to reconcile larval and adult
groupings based on morphology. Larvae of some
morphologically disparate (as adult) species are
very similar, whereas zoeal stages for species in
some genera differ markedly in their morphology
(see Martin 1984; Martin et al. 1984, 1985; Mar-
tin and Abele 1986). Because of the known mor-
phological variability of xanthid larvae, charac-
ters presented for taxonomic purposes here and
elsewhere (e.g., Martin 1984) must be used with
caution.
It is not our intent to promote descriptions of
single stages in the life cycles of brachyuran
crabs. However, in those cases where a descrip-
tion of a single stage adds appreciably to our
knowledge of phylogeny (e.g.. Rice 1981b) or fills
a gap in the larval biology of a commercially im-
portant species complex (present study), we feel
such a description is justified. A detailed compari-
son of zoeal stages of the two species is planned for
the near future.
ACKNOWLEDGMENTS
We are grateful to D. H. Wilber and A. B. This-
296
MARTIN ET AL.: MEGALOPAE OF STONE CRABS (MENIPPE)
tie for constructive criticism of the manuscript, to
B. E. Felgenhauer for help in preparing the fig-
ures, to N. N. Rabalais and B. Cole for assistance
in rearing larvae of M. mercenaria, and to L. G.
Abele for providing space and facilities for the
research. We thank S. Silvers and K. Riddle of the
Florida State University Electron Microscopy
Center for their expert assistance. We also thank
C. Dugas, Louisiana Department of Wildlife and
Fisheries, for collecting the ovigerous specimen of
M. adina. This work was supported in part by the
National Science Foundation grants No. BSR-
8414347 and BSR-8615018 to J. W. Martin and
L. G. Abele, and by a research grant to D. L.
Felder, N. N. Rabalais, F. M. Truesdale, and
D. W. Foltz from the Louisiana Education Quality
Support Fund under grant No. 86-LUM(2)-084-
09.
LITERATURE CITED
Felder, D L
1973. An annotated key to crabs and lobsters (Decapoda,
Reptantia) from coastal waters of the northwestern Gulf
of Mexico. La. State Univ. Cent. Wetland Resour., Publ.
LSU-SG-73-02. Baton Rouge, LA.
Felder, D. L., J. W. Martin, and J. W. Gov.
1985. Patterns in early postlarval development of deca-
pods. In A. M. Wenner (editor), Larval growth, p. 163-
225. Crustacean Issues, vol. 2. Balkema Press, Rotter-
dam.
Felgenhauer, B E.
1987. Techniques for preparing crustaceans for scanning
electron microscopy. J. Crustacean Biol. 7:71-76.
Hyman. O W
1925. Studies on the larvae of crabs of the family Xanthi-
dae. Proc. U.S. Natl. Mus. 67:1-22.
Kakati, V S.
1977. Larval development of the crab, Menippe rumphii
(Fabricious), as observed in the laboratory (Crustacea,
Brachyura). Proc. Symp. Warm Water Zooplankton,
UNESCO/NIO Spec. Publ., p. 634-641.
Martin. J W
1984. Notes and bibliography on the larvae of xanthid
crabs, with a key to the known xanthid zoeas of the west-
em Atlantic and Gulf of Mexico. Bull. Mar. Sci. 34:220-
239.
In press. Phylogenetic significance of the brachyuran
megalopa: evidence from the Xanthidae. In A. A. Fin-
cham and P. S. Rainbow (editors), Aspects of decapod
crustacean biology. Symp. Zool. Soc. Lond., Vol. 59.
Martin. J W.. and L G Abele
1986. Notes on male pieopod morphology in the
brachyuran crab family Panopeidae Ortmann, 1893,
sensu Guinot (1978) (Decapoda). Crustaceana 50:182-
198.
Martin, J W., D L. Felder, and F M Truesdale.
1984. A comparative study of morphology and ontogeny
in juvenile stages of four western Atlantic xanthoid crabs
(Crustacea: Decapoda: Brachyura). Philos. Trans. R.
Soc. Lond., Ser. B, 303:537-604.
Martin, J. W., F. M. Truesdale. and D. L. Felder.
1985. Larval development of Panopeus bermudensis
Rathbun, 1891 (Brachyura, Xanthidae) with notes on
zoeal characters in xanthid crabs. J. Crustacean Biol.
5:84-105.
MONOD, Th
1956. Hippidea et Brachyura ouest-africains. M6m.
Inst. Fr. Afr. Noire 45:1-674.
Porter, H. J
1960. Zoeal stages of the stone crab, Menippe mercenaria
Say. Chesapeake Sci. 1:168-177.
Rathbun, M. J.
1930. The cancroid crabs of America of the families Eu-
ryalidae, Portunidae, Atelecyclidae, Cancridae, and Xan-
thidae. U.S. Natl. Mus. Bull. 152:1-609.
Rice, A. L.
1981a. The megalopa stage in brachyuran crabs. The
Podotremata Guinot. J. Nat. Hist. 15:1003-1011.
1981b. The zoea of Acanthodromia erinacea A. Milne Ed-
wards; the first description of a dynomenid larva (Deca-
poda, Dromioidea). J. Crustacean Biol. 1:174-176.
In press. The megalopa stage in majid crabs, with a re-
view of spider crab relationships based on larval charac-
ters. In A. A. Fincham and P. S. Rainbow (editors). As-
pects of decapod crustacean biology. Symp. Zool. Soc.
Lond., Vol. 59.
Scotto. L. E.
1979. Larval development of the Cuban stone crab,
Menippe nodifrons (Brachyura, Xanthidae) under labora-
tory conditions with notes on the status of the family
Menippidae. Fish. Bull., U.S. 77:359-386.
Williams, A. B
1984. Shrimps, lobsters, and crabs of the Atlantic coast of
the eEistem United States, Maine to Florida. Smithson.
Inst. Press, 550 p.
Williams, A. B., and D. L. Felder.
1986. Analysis of stone crabs: Menippe mercenaria (Say),
restricted, and a previously unrecognized species de-
scribed (Decapoda: Xanthidae). Proc. Biol. Soc. Wash.
99:517-543.
297
OCEANOGRAPHIC ASSOCIATIONS OF NEUSTONIC LARVAL AND
JUVENILE FISHES AND DUNGENESS CRAB MEGALOPAE OFF OREGON
Jonathan M. Shenker'
ABSTRACT
The larval and juvenile fishes and crabs inhabiting the neustonic zone within 50 km of the coast were
sampled biweekly from April through July 1984, with a Manta net (mouth 1.0 m wide x 0.7 m deep),
and a large neuston trawl (mouth 3.5 m wide x 1.0 m deep). The Manta net was an efficient sampler
for larval fishes and crabs, while the neuston trawl collected larger juvenile fishes that had rarely
been observed in previous studies.
Nocturnal sampling accounted for nearly all the ichthyoneuston and zooplankton taken. Dunge-
ness crab megalopae were the most abundant species. Although present throughout the survey, the
great majority were found in very large aggregations along visible convergence zones or in association
with Velella velella. Discrete groups of abundant larval and juvenile fishes were found prior to
upwelling [Parophrys vetulu3 , Scorpaenichthys marmoratus , Hemilepidotus spinosus, Hexagrammos
sp., and A noplopoma fimbria ) and after its onset (Engraulis mordax and Sebastes spp.). These species
had distinct zonal (east-west) distribution patterns and were generally associated with, or affected by,
hydrographic characteristics such as convergences, upwelling, and the Columbia River plume.
Recent studies of the ichthyoplankton off the
northwest coast of the United States have con-
tributed new information on the temporal and
spatial occurrences of larvae of coastal and
pelagic fish species (Richardson 1973; Richardson
and Pearcy 1977; Laroche and Richardson 1979;
Richardson et al. 1980; Kendall and Clark 1982a,
b; Clark 1984; Bates 1984; Mundy 1983; Brodeur
et al. 1985; Boehlert et al. 1985). These surveys
focused on larvae occurring below the surface
layer of the ocean, although concurrent neustonic
samples were occasionally collected (Kendall and
Clark 1982 a, b; Clark 1984; Bates 1984; Richard-
son^). Comparison of simultaneous surface and
subsurface samples demonstrated that many spe-
cies were found in both depth strata, while an
additional group of species was collected only
from the neustonic zone. Brodeur et al. (1987)
examined the larval fish and invertebrate compo-
nents of the neuston in the northeast Pacific, and
determined that these organisms were frequent
prey items of juvenile salmonids.
^Oregon State University, College of Oceanography, Hatfield
Marine Science Center, Newport, OR 97365; present address:
University of California, Bodega Marine Laboratory, P.O. Box
247, Bodega Bay, CA 94923.
2Richardson, S. L. Oregon's coastal ichthyoneuston - a pre-
liminary report. Unpubl. rep. Presented at American Soci-
ety of Ichthyologists and Herpetologists, Williamsburg, VA,
June 1975.
Manuscript accepted November 1987.
FISHERY BULLETIN; VOL. 86, NO. 2, 1988.
Standard plankton and neuston nets used in
these studies were effective in collecting the rela-
tively slow-moving early larvae. However, net
avoidance by larger larvae and juvenile stages
(Barkley 1972; Murphy and Clutter 1972) sug-
gests that use of traditional collecting gear is in-
appropriate for more mobile fishes. This paper
describes the results of a neustonic survey con-
ducted off the Oregon coast in the spring and sum-
mer of 1984 that focused on the larger ichthy-
oneuston. Conventional sampling gear and a new
net designed specifically for sampling juvenile
fishes were used to characterize the temporal and
spatial distribution patterns of both larval and
juvenile ichthyoneuston and Dungeness crab,
Cancer magister, megalopae.
MATERIALS AND METHODS
Sampling was conducted at approximately
2-wk intervals from early April through July
1984, on an east-west transect along the 44°40'N
parallel, 3 km north of Newport, OR (Fig. 1). Sta-
tions were located at distances of 1, 5, 10, 15, 20,
30, 40, and 50 km from shore. The stations were
occupied twice during each 24-h cruise, once dur-
ing the day and once at night. On one cruise (8-10
June), the 50 km station was occupied for 27
hours to assess diel variation in abundance of
neustonic organisms. In response to several
299
FISHERY BULLETIN: VOL. 86, NO. 2
45°
00'
-1 1 1 r-
440
50'
90
80
70
60
50 40
30
15 5
• • • •)
20 10 It
YAQUINA
BAY
125°
00'
124°
30'
124°
00'
Figure 1. — Location of stations along the transect off Newport,
OR. The numbers indicate distance offshore in kilometers.
weeks of strong upwelling, during which very few
larvae were collected close to the coast, the sam-
pling scheme was modified for the two July
cruises by eliminating some of the inshore sta-
tions and extending the transect as far offshore as
90 km. Only night samples were collected during
July.
One of the primary considerations of the survey
was to collect samples while minimizing net
avoidance by the target organisms. Reduction of
vessel-induced disturbance was accomplished by
deploying the collecting gear from the ends of 12
m-long outriggers on the FV Cumberland Trail,
a chartered 23 m commercial scallop-fishing ves-
sel.
Two different nets were used to collect the sam-
ples. Larval fish and zooplankton were collected
with a Manta neuston net (Brown and Cheng
1981), modified to have a mouth 1.0 m wide x 0.7
m deep. The Manta frame was equipped with a
green-colored 0.333 mm mesh net, PVC cod end
bucket, and General Oceanics model 2030 digital
flowmeter'^. A two-point bridle was attached to
the upper corners of the frame. Drag on the net
while towing kept the entire bridle and towing
wire assembly out of the water.
The second net was designed to collect the
larger juvenile fishes that were assumed to avoid
the smaller Manta net. The neuston trawl was
constructed with a mouth 3.50 m wide x 1.05 m
deep (Fig. 2a). The frame consisted of 43 mm (out-
3References to trade names do not imply endorsement by the
National Marine Fisheries Service, NOAA.
side diameter) heavy-duty galvanized pipe, with
towing points welded at the four corners. Flota-
tion for the frame was provided by three inflat-
able spar buoy floats. These 40 cm diameter floats
had hollow tubes running through their centers
and were fitted onto the upper bar of the frame.
The 8.5 m-long net was made of 4.8 mm green-
colored woven mesh, with a 15 cm-wide cloth
collar around the mouth, PVC cod end bucket and
General Oceanics model 2030 flowmeter. The net
had a mouth slightly larger than the frame
(3.70 m X 1.20 m) so the net could be laced around
the outside of the frame and flotation buoys. The
ends of six 12 mm polypropylene rope riblines
running the length of the net were shackeled to
the frame for additional support. A 4-point bridle
of 6.4 mm wire was attached to the corners of the
neuston trawl (Fig. 2b). Drag forces kept the en-
tire bridle and towing wire (except for the two
short segments attached to the bottom of the
frame) out of the water. A length of 5 cm
polypropylene rope was attached between the
upper towing points on the frame, as a bridle for
use in deploying and retrieving the net. A line
attached to the retrieval bridle was passed to a
hydraulic capstan through a block tied in the rig-
ging over the deck. During a tow, this retrieval
line was slackened, but remained attached to the
vessel, and floated in a broad arc behind the net
mouth.
At each station, the neuston trawl and Manta
net were towed simultaneously from the port and
starboard outriggers, at approximately 1-1.5 m/
second. Tows were generally made either against
or with the direction of the prevailing swells. The
Manta net was usually fished for approximately
8-9 minutes/tow and filtered about 300-400 m^.
The neuston trawl generally filtered 2,000-3,000
m'^ during a 10-11 minute tow. Additional tows
with the neuston trawl were frequently made at
a station to assess small-scale patchiness and
to sample visible "structures" in the surface
layer, such as convergence zones marked by foam
lines and rafts of the pleustonic hydroid Velella
uelella.
On several cruises, onset of high winds and
rough sea conditions prevented use of the Manta
net during the daj^ime sampling of the transect
as we returned toward shore. However, we were
able to fish the neuston trawl in winds up to an
estimated 40—45 km/hour, and in white-capped
seas of 2-3 m. Neuston trawl samples were col-
lected at alternate stations during the periods of
adverse weather.
300
SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON
a)
0.4m
1.05m
3.50m
1.0m
6.0m
b)
Figure 2a. — Front view of the neuston trawl; B = inflated 40 cm diameter spar buoys;
C = cloth collar around mouth of net; R = polypropylene rib lines.
Figure 2b. — Side view of the neuston trawl showing construction of the towing bridle.
Floats were removed for clarity.
Several very large catches of Dungeness crab
megalopae in the neuston trawl were subsampled
by volume, with the majority of the megalopae
being returned alive to the ocean. Samples from
both nets were preserved in 10% formalin/sea-
water and sent to the laboratory for analysis. All
larval and juvenile fishes were identified to the
lowest practical taxonomic level. After extrapo-
lating catch sizes from subsample counts, the
data were normalized to produce density esti-
mates of organisms/1,000 m^^. As an indicator of
patchiness of abundant taxa, the coefficient of dis-
persion (CD. = S'^/X ) was calculated for the sam-
ples in which each taxon was present.
Hydrographic and meteorological data were
collected at each station. Surface water tempera-
ture was determined with a handheld bucket
thermometer. Water samples were collected for
laboratory analysis of surface salinity using a
Model 8400 Guildline Autosalinometer. A 200
mL surface water sample was filtered through a
0.3 fxm pore size Gelman A/E glass fiber filter for
determination of chlorophyll a concentration. The
filters were stored in dessicant over dry ice at sea,
and then in a freezer, before acetone extraction of
the chlorophyll a and analysis with a Turner De-
signs Model 10 Fluorometer. During the daytime,
Secchi depths were estimated, using a 30 cm disc,
and surface irradiance was measured with a
handheld General Electric Model 214 Light
Meter. Other data collected included weather and
sea state conditions at each station.
RESULTS
Hydrographic Data
The neustonic realm was characterized by dis-
tinct temporal and spatial patterns that reflect
the highly dynamic springtime oceanographic
301
FISHERY BULLETIN: VOL. 86, NO. 2
processes of the Oregon coastal zone. Daily up-
welling indices derived from barometric pressure
data (Fig. 3; A. Bakun"*) indicated that in April
and May, local winds varied in direction and
strength, with one extended period of stormy
weather with southwest winds (negative index
values). Temperature and salinity profiles during
these months showed relatively little variation
along the transect within a sampling period (Fig.
4a, b, c). Chlorophyll a peaks and low salinity
near shore may have been due to the influence of
water exiting Yaquina Bay. The highest chloro-
phyll a values were observed following periods of
northwest winds, and the highest temperatures
occurred after a period of very light winds and
high insolation.
The hydrographic regime was altered by the
onset of northwest winds in June (Fig. 3), which
induced the upwelling of cold, more saline
nutrient-rich water along the coast. Strong up-
welling [defined by Small and Menzies (1981) as
a daily index value >50] was sporadic in June but
nearly continuous in July. Temperature, salinity,
and chlorophyll a profiles rapidly changed in re-
sponse to the upwelling-favorable winds (Fig. 5a,
b, c).
During the 8-10 June cruise, low temperature
and high salinity water was found within 5 km of
the coast. Low salinity, warm water more than 15
km offshore apparently was the plume of the Co-
lumbia River. By 19-20 June, continued offshore
transport of the surface layer pushed the inshore
edge of the plume to 40 km off the coast. Low
chlorophyll a concentrations, low temperatures,
and high salinity near shore on 19-20 June indi-
cated the occurrence of active upwelling.
Upwelling persisted through the end of July.
Temperature and salinity profiles were similar to
those observed during the second June cruise,
with the patterns of increasing temperature and
decreasing salinity farther offshore continuing to
the limits of sampling of 90 km on 8 and 9 July
and to 70 km on 23 and 24 July. During the up-
welling in July, a dramatic phytoplankton bloom
occurred in the surface waters within 30 km of the
coast. Chlorophyll a concentrations here were
denser and broader in offshore extent than the
surface and subsurface biomass concentrations
previously observed off Oregon (Small and Men-
zies, 1981).
+ 200
«
T3
C
c
0)
a
3 -200-
U
I ||'|||l||l ll'l' f |'l|| '1 11 1 T
lll.i„lil
yt4 y^ /N /K yfs yT\ xTs xl\
"•A. Bakun, Pacific Environmental Group, National Marine
Fisheries Service, Monterey, CA 93940, pars, commun. August
1984.
M J J
1984
Figure 3. — Daily upwelling indices for spring-summer 1984.
Positive values >50 indicate occurrence of winds inducing
strong upwelling. Arrows indicate dates of sampling cruises.
Ichthyoneuston
A total of 107 Manta net and 142 neuston trawl
samples collected 48 taxa of larval, juvenile, and
adult fishes. Larvae <10 mm and a few juveniles
were collected by the Manta net, while large
numbers of juvenile fishes up to 60-70 mm were
taken by the neuston trawl. Size-frequency data
for three of the commonest species (Fig. 6 a, b, c)
illustrate the relative ability of the nets to cap-
ture different sizes of fishes.
Nighttime sampling (59% of the Manta net
tows and 62% of the neuston trawl tows) ac-
counted for 93.3% of all fishes in the Manta net
catch and 96.5% of the fishes taken in the neuston
trawl (Table 1). Four Manta net and 8 neuston
trawl tows made during twilight collected 5.4%
and 2.9%, respectively, of the total number of
fishes taken with each gear type (Table 2). Only
1.3% of the Manta net catch and 0.6% of the neu-
ston trawl catch was made during daytime (Table
2).
Only two species were not collected predomi-
nantly at night. Larval Pacific saury, Cololabis
saira, were taken in both day and night Manta
net samples. Larval northern lampfish, Steno-
brachius leucopsarus , were abundant only in the
twilight samples from early June (Tables 1, 2).
Arbitrary criteria on frequency of occurrence
(in >15% of the night samples of either net) or
abundance (peak densities >40/l,000 m"^) indi-
cated that nine taixa of larval, juvenile, and adult
fishes were dominant components of the ichthy-
oneuston. Most taxa were characterized by dis-
302
SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON
14
o
o
<
if)
E
>-
X
Q_
O
O
_J
X
o
O O 4/6
• • 4/16-17
A A 5/4-5
▲ A 5/18-19
10 20 30 40
DISTANCE FROM SHORE (km)
%
50
Figure 4a-c. — Nighttime environmental conditions observed on each cruise during the pre-upwelling period,
a = temperature; b = salinity; c = chlorophyll a concentration.
303
FISHERY BULLETIN: VOL. 86, NO. 2
o
o
a:
01
LJ
Q.
UJ
o
o
Z
_J
<
Dl
>-
X
Q_
O
q:
o
_j
o
30 40 50 60 70 80
DISTANCE FROM SHORE (km)
100
Figure 5a-c. — Nighttime environmental conditions observed on each cruise during the upwelling period,
a = temperature; b = salinity; c = chlorophyll a concentration.
304
SHENKER OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON
tinct temporal shifts in occurrence: Five taxa
were abundant only prior to the onset of continu-
ous upwelling in mid-June, three taxa were spo-
radically abundant following the beginning of up-
welling, and one species (Ronquilus jordani) was
present tliroughout the survey.
Pre-upwelling Species
The numerically dominant species in the
nearshore region in spring was larval English
sole, Parophyrys vetulus. These larvae were
patchily distributed (CD. = 32.5) within 30 km
offshore, reaching densities of 75/1,000 m'^ in
Manta net collections. Larval abundance declined
rapidly after April (Table 1). Size-frequency dis-
tributions indicated that the Manta net was more
effective than the neuston trawl for collecting fish
<15 mm (Kolmogorov-Smirnov test, P<0.01).
Eye migration in the larger fish (20-24 mm) was
nearly complete, indicative of their impending
shift to a benthic existence. Similar-sized juve-
niles were taken nearshore in April in benthic
tows (B. Mundy^).
Cabezon, Scorpaenichthys marmoratus , larvae
were most abundant in Manta catches in the
spring (Table 1), although newly hatched larvae
were collected near shore through June. The
smallest larvae (4-6 mm.) were taken only at the
1-15 km stations, while larvae up to 10 mm were
found along the entire transect. These larvae
were patchily distributed, with April densities
reaching 207/1,000 m^ (CD. = 92.7). Frequency
of occurrence and larval density declined through
early July. Large juvenile cabezon (26-38 mm)
were collected by the neuston trawl (Fig. 6a).
These juveniles were encountered all along the
transect from April to mid-June.
Two dominant taxa found at all stations prior
to upwelling were greenling (Hexagrammos sp.)
and brown Irish lord, Hemilepidotus spinosus . Ju-
venile greenlings were the most frequently col-
lected species, occurring in AA.9>% of the night-
time neuston trawl samples (Table 1), although
they were rarely taken by the Manta net (Fig. 6b).
They were the most evenly dispersed of the abun-
dant fish collected (CD. = 2.2), and never ex-
ceeded densities of 10/1,000 m^. The two largest
catches of greenling were made in association
with distinct hydrographic features. Nineteen ju-
a) Scorpaenichthys marmoratus
100 -,
50 -
I J Manta net
Neuston trawl
TT'
4 6 8 10 20 26 32 38
b) Hexagrammos sp.
40 —I
30 -
20 -
10 -
c) Anoplopoma fimbria
5B. C. Mundy, Southwest Fisheries Center Honolulu Labora-
tory, National Marine Fisheries Service, NOAA, 2570 Dole
Street, Honolulu, HI 96822, pers. commun. May 1984.
T r
10 20 30 40 50 60 70 80 90 100 IK) 120130
Standard Length (mm)
Figure 6a-c. — Length-frequency data of larval and juvenile
fishes collected by each net. a = Scorpaenichthys marmoratus ;
b = Hexagrammos sp.; c = Anoplopoma fimbria.
veniles were collected from a convergence zone
near shore in mid-May (along with 150,000 Dun-
geness crab megalopae) and 20 juveniles were
taken from Columbia River plume water 50 km
offshore in early June. Mean lengths of greenling
from the neuston trawl samples increased from
305
FISHERY BULLETIN: VOL. 86, NO. 2
o
o>
c
a>
£1
E
:d
Tr
c
CO
II
en
a^
c
0)
>>
o
c
0}
3
3
(/)
o
QJ
o
o
£
o
O
>.
o
c
0)
3
(T
0)
tn
o
a>
C/5
OJ
y- in <M
'-'to
1/5 in CD
CD
CO
••- r- CO r^
1- CO CM C\J
r-- Tt LO CO
d o o d
CO
C\J
CD
CM
00
d
CO CD
CO CD
O
CO
Cvl
O
CD
CM
C\J
CO
T- -^ r-
CM
CO
CO
t--
CM
CO T •* T-
cD r-- CO Tf
CO in Tt CO
d d d> d
IT) CD O)
>- t- CO
CO
CD
CO
CO
CO
T- •.- CD
CD
in
in
uo
CO
CD
d
1 s
CO
'J-
CD
1
(D
1
O
CO
Cvj
^
CO
C\J
PO
CD
- 1
00
in
-
CO
-
CM
;; 1
CM
in
r^
CM
'I-
o
CO
CD
CM
CO
CO
d
d
CM
d
■•-
'-
d
CO o
CD O
CD
CO
CO
CO
in
CD
CO CM CD
CM
t r^ r^
r- o tt
■^ I^ CD
CD
CO
T- O
r~ O) CO
o in CD
CM T-
CM
CO
CO
CD
CO
CO
T- CM CM •--
•r- CO O
in t^ CD
co
CD
in
o
•<t
CO
C\J 1- •.-
CO CO »-
Cvi T- CO
If)
CO
CO
CD
CO
CM
CO
•,- ,- TJ-
CO
CO
CM
o
CO
CO
CM
CO
CO
•<t
CM
CO
in
CO
in
CM
CO
CO
CO
CD
CD
in
t- CO •«t
o in CO
T- d d
CO
o
CO
CD
CO
00
CO
CM
CD
cri
in y- y-
o
in
n o
CD o
CO
in
■<J CM CO 1
CD ■<* CM 1
1 ?
CO
o
oq
CM
in
d d d
d
^
1-
d
in
CD
CO
in
1-
00 CO
CO 00
CO
CO
CO
CO
00 •■- '-
CO T-
1 <^J
1 CD
00
" 1
■^
CO
■^
t ■*
CO
in CM
CM CD
in
CO
O CO
CD CM
CO
■<}■
in CD I -r-
CO t^ I CO
^ 1
CD
o
in y-
00 CD
ai
CO
CO CM
CO
c
cb
(0
o
■^ cc
CtJ 13
0) ^
Q. ^
..3" '°
I
o
.CO
ID
^ "J
2 "°
[5 c
O) o
- c p
C CO
-ie
CO
3
I.
c
c:
c- ■fe 'I'
o
CD
CO
■O "D
,9> E
CD
to
O ™ ^,^
55 2
CO
3
C
CO <0
S iS
5 3
CD ^
O
CD
£
0)
CO
3
o
-c
5 c
£l ■^ "^
fc X: -o
O
CO
3
.E
O CD
CO o
-5 o
^ CO
CXJ CD
si
u
en
CO
3
'cB
-3J
3
CJ
CD
0)
CO m
? "C
Oi ^
P CO
^^
CO
CD
If
3 CO qj
C CO
O Q.
CO
c
CO
3
en
o
(0
><
<D
<0 (D ^ O)
^ -S t ™
s|tl
-J £ "^ 2
< in
o
c:
CD
c
ill
£ 03 ">
CO
(D
CO
£
(O
c
(D
>
_ 3
. . <D
Q. Q. <5
CO CO •;-
CO
e fi^ to CO
tf^ CD <D
CO
o a>
-Q
E
■^ 0)
(0 to
_ __ E
^.fe 9? S S §-■§
-5 c "o
t 3J to to
E S.
CO CO CO </) CO o
Cl Cl —
dj Cb Q.
CO
. 3
o o
CO
ia^l^
- CO- 9 .a
CD >< -C X
to CD Q. ><
I
306
SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON
13
c
o
O
CD
<
<£>
o
0)
CO
d
CO
CO
00
CVJ
1^
CO •* CO c\j r--
•■- C\J
CO
r^ (j> ia n
O r- <£> ID
un o '-
■^ CM
en
CO
■* CO
d d
CO
uo
I/)
T- uo ■>- OJ •-
■.- O) T- ^ ^
•^ 00 uo CD in
d oj d (b o
CO
CO
CO
o
uo
(D
(D
CM
UO
00
CO
LO O
O)
n
Tt
o
1 ^
Tl- CD
CO
CM
in
CM
CO
t-~^ If)
1^
CO
CM
■■- CO
O CD
CM 00
1— (D
CD
2i
1 1 '^
1 "^
■V
h-
1 "^
CM 1
1 '- ^ CM 1
LL
1
1
1
1
o
00
1 i ^
CO
CD
r^
1 •*
•^ 1
1 in 1- 1- 1
CM
^'
<
1 1 r^
r^
■*
in
1 c^
r-^ 1
1 CM in CD 1
CO
d
d
d
T-'
d
d d (6
- I - CO
in ■<r
CD I T-
CO I T-
d CD
CO
co
CD
in
00
CD
C\J
CD
CD
in
■^ CM
O CO
■■- CO
cvj d
OJ
O)
CO
■^ 00 CM
CO
CD
■^ I h- T- in
CD
CD CM CD
CO CM ••-
r- c\j in
■.- ^ Tf
in
o
1^
o
CO
CO CM CO •— o
■* CO in CD 05
Tf CD TJ- o ■>-
CO in uri CD r--
'- ••-■<«■
oo
CvJ
CD
CD
d d
CO
in
■•- CM
d ■*
o
CM
CO
CO
in
d
1
CO
00
CO
1
^~
Cvj
CM
'" CO
CO ■*
CO
CM
in
o
1 °°
CO
CM
CM
CO
CD
in
1 CO
■*
O)
CM
CM
CM
CD
r--
B
o
c
C3)
c
CO
O o CO
tT3
x:
5. t^ ,„
CO g «
CO nj cio <li o 3J
o
O
CO
CO
3
i.
CO
'?: OI Jjco o^ g_
CO
3
"O
?
o
CO
CO
is
O
ID C
(0 cq
■g -c
fC
CD
CO
a>
CB :9
E t;
U) O
I
CTJ
S '^
?15
O -C
^ Q.
m <i,
fc O
CD O
&§
Uj CD
CO
O
'co
o
■ c
a. CO
^ £ 5
CO
CD -g o O —
lllll
3
Q
o
CO
o
1
if
CD
O
o
c
^
CO
OJ
"O
^
c
E
c
3
CO
C
s
CO
o
(LI
307
FISHERY BULLETIN; VOL. 86, NO. 2
tu c
_l o ■*
CD C 00
,< ^ O)
H CO ^
5
(0
C7>
0)
c
3
-3
C
00
CO
o
E
Q
c
(0
CO
5
CO
05
c
c
fO
5
CO
CO
c
CO
CO
Q.
<
C
CO
en
0)
o
0)
d
CO
C\J
<£>
CM
■^
CD
LT)
CD
CD
CO
(N
T-
>-
<>i
•* y-
■■- •*
TT
CD
in
CO
C\j
T-
1-
c\i
CO
o
in
CO o
o ••- T-
O £31 O
in in in
o
CO
CO
CO
CVJ
CO
in
CO
CO
in
CM
CO
CJ5
CO
in
en
CO
IT)
CO '-
CO CO
h- ■.- ,-
r- CD
CO
'-CD CD
O CO
CO CJ5
CD -^
in
CO
o
■>*■
in in
CM Cvj ■■-■•-
CO
CO
co
in
o
CM
CO
in
in
o
CD
oo
CD
CM CM
■^ CM
N- in
in
CO
I
o
6
0) CO
CO -^
•D 3
CD Ol
a>uj
LU
i CD
to
2
<n
(O
&
u
— (D
CO CO
3 -g
£ o
- CJ O
CO CO en
■^ O QJ
Q. C: JD
2 Si E
"55 o
S CO
CO O)
CO TJ
CO ^
CO 2
CO CO
o E
o >-
CO
o -o
If
3 CO
"3" y
C CO
o
CO
c:
"cD
CO
CO
CU
-c
CO QJ
0) CO
u "9
§ o
^£
-^ E
<
s '^
C (0
CO jo
^^ . _ ,
O CO CO CO CO
CO
•Q
I
CO
E
o
CL
It
9^
E
~ ''J
CD X
CO (U
>< I:
I
CO
3
'cO
CD
c:
o
"S
c
O CD
Ob
o
CO
&
CJ
o
CO
3
o
o
o
c
in
cii
CO £
o <o
S E
CO £
CO
CD
,- CD C35 "-
CD r~ oo
»- CM 1-
CT>
00
CM
CO
CD
CO
^ 3
§o
- c 9f
E o ^ "a
li O i; F
ICO 2.i
CO o w
9 SJ Q.B
a? CD ■§_^;_
- S o ^
2 9- k <i>
3 .° ^ "O
co
t^
00
CM
0
CD
CJ>
'*
CM
00
CD
0
0
CD
in
CU
cO
CM
y-
0
d
■^
G>
CM
CO
•<r
1—
CD
in
CM
r~-
CD
in
CM
'—
d
CM
l~-
CM
oJ
n
E
Q
0
10
0
1
'-'
(ii
0
0
c
^
CO
<D
■0
n
c
E
3
c
3
CO
c
CO
CO
0
1-
cu
2
308
SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON
38.8 mm (A^ = 10) in early April to 51.2 mm
{N = 72) during the first June cruise.
Brown Irish lord juveniles were the numeri-
cally most abundant and the second most fre-
quently occurring species (Table 1). Early juve-
niles (10-20 mm) were taken with both nets in
April and May only at the offshore (40-50 km)
stations. Larger juveniles (20-30 mm) were
patchily distributed from April through mid-June
all along the transect (CD. = 14.1). Although
47% of the total number of trawl caught speci-
mens were taken in densities below 15/1,000 m-^,
repeated sampling in one large aggregation ac-
counted for the bulk of the catch. Three sequen-
tial tows at the 50 km station on 5 May had juve-
nile densities of 15-39/1,000 m*^ , but distinctive
hydrographic characteristics at the station were
not detected.
The only species with a characteristic offshore
distribution prior to upwelling was sablefish,
Anoplopoma fimbria. Small juveniles (10-45
mm) were effectively taken by the Manta net,
while the neuston trawl was more efficient at col-
lecting 30-70 mm specimens (Fig. 6c). The 10-50
mm fish were collected primarily at the 50 km
station from April through mid-June, although
several specimens were taken closer inshore. The
50-75 mm fish were captured only on 8-10 June
from the Columbia River plume water at the 50
km station. After the onset of upwelling, two indi-
viduals ( 120-130 mm) were taken 80-90 km from
shore.
Upwelling Species
Three taxa were abundant on only one of three
cruises made after the onset of upwelling in mid-
June (Table 1). Distributions of northern an-
chovy, Engraulis mordax, larvae; rockfish (Se-
bastes spp.) larvae; and adult blue lanternfish,
Tarletonbeania crenularis, overlapped at some
offshore stations on 8 and 9 July. The offshore
portion of the transect was characterized by a
drop in salinity and an increase in temperature
between 60 and 70 km offshore (Fig. 5a, b), indi-
cating a transition from coastal to Columbia
River plume water.
Of the 814 rockfish larvae collected throughout
the survey, 87% were taken at the 50-90 km sta-
tions on this cruise, with a peak density of the
3.5-7.0 mm larvae of 684/1,000 m^. Adult blue
lanternfish were present in densities up to 315/
1,000 m"^ at the 60-90 km stations. Lanternfish
from simultaneous Manta and neuston trawl tows
displayed no difference in length-frequency dis-
tributions, although a significant change in the
size structure of the catches between stations was
observed (K-S test, P < 0.01). Mean size of the
fish decreased approximately 10% between adja-
cent stations, from 50.9 mm at 60 km to 37.8 mm
at 90 km. Anchovy larvae (3-9 mm) were re-
stricted to the lower salinity plume water at the
70-90 km stations, with peak and mean densities
of 368 and 210/1,000 m^, respectively.
Persistant Species
Only northern ronquils, Ronquilus jordani,
were abundant in both pre-upwelling and up-
welling periods. The elongate larvae and juve-
niles were collected with the Manta net starting
in mid-April, with abundances peaking at 40/
1,000 m*^ in July. Mean lengths over this time
interval increased from 8.7 mm (n = 8) to 26.6
mm (n = 43). Ronquils were relatively dispersed
along the transect (CD. = 7.6), and size or sea-
sonal patterns of distribution were not detected.
Dungeness Crab Megalopae
Dungeness crab megalopae were the dominant
component of the catches throughout the survey,
with an estimated total of 350,000 megalopae col-
lected in 249 hauls. The megalopae were found at
most stations during all cruises (Fig. 7), although
they were rare seaward of 50 km in July. Megalo-
pae were present in 71.4% of the Manta net day-
time tows and 93.8% of the nighttime tows, and in
90.9% and 98.9% of the day and night neuston
trawl collections, respectively. All twilight hauls
captured megalopae.
Despite the higher frequency of occurrence of
megalopae in the neuston trawl, the Manta net
was a more accurate estimator of their abun-
dance. For 58 pairs of nighttime catches with den-
sities under 2,000/1,000 m-^, the Manta net caught
approximately 3 times as many megalopae per m^
as the neuston trawl. The 4.8 mm mesh of the
neuston trawl apparently enabled some megalo-
pae to pass through the mesh, and some escape-
ment was observed while the trawl was sitting on
deck after retrieval. More similar estimates be-
tween the nets were obtained when higher densi-
ties of megalopae were sampled, probably because
of clumping of the large (11 mm total length)
spiny megalopae into large masses which clogged
the mesh.
In general, densities of megalopae caught by
309
FISHERY BULLETIN: VOL. 86, NO. 2
D — D Day
A
1.000
6 April
X A
100
10
1
0-
/
O-D /
H 1 1 ffl 1
\
\
\
1 1 El
1.000
100
10 ■
A A Night
8-10 June
1
\
t >\ r
1 -
0
\
< 1 1 E — ffl-
-ffl-
-E-
-m
1.000 •■
16-17 April
1.000 ■•
100D
1.000-
8-9 July
/
A
^"^A
100-
■
/
A
10-
•
1
0^
■i — 1 1 1 1
— 1
1
1
1.000 ■
100•
10
1 +
0
n-
/
t I
H H
A 18-19 May
I.OOOt
100-
10
0 5 10 15 20 30 40
50
1 ■•
0
23-24 July
A— i
H 1-
0 5 10 15 20 30 40
Distance From Shore (km)
50
Figure 7.— Daytime and nighttime densities of megalopae collected by the Manta net. Rough weather prevented the use of the
Manta net during the day of 19 May; neuston trawl data for this portion of the cruise are presented. Circled letters refer to events
discussed in the text.
310
SHENKER; OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON
the Manta net in the daytime were lower than at
night for each station (Fig. 7). The differences
between day and night were particularly appar-
ent on bright sunny days and at the offshore sta-
tions. Only one of 13 samples of megalopae col-
lected when light intensity exceeded 4,000 lux
had a density over 200/1,000 m^, while 12 of 24
day samples collected at lower light levels ex-
ceeded this density. Highest daytime densities
(1,600-2,400/1,000 m^ in neuston trawl samples)
were observed on 19 May when a sudden shift in
weather before daybreak resulted in very dark,
overcast and rainy skies, with 40-45 km/hour
winds and 2-3 m breaking waves.
Water clarity may also affect the daytime
abundance estimates of megalopae in the neuston
along the transect. Daytime densities were al-
ways low seaward of 30 km where Secchi depths
exceeded 9 m, while higher densities were ob-
served in the more turbid near-coastal waters
with Secchi depths of 4-9 m.
The 27-h occupancy of the 50 km station on 8-9
June provided additional evidence of a distinct
diel pattern in utilization of the neustonic habitat
by megalopae (Fig. 8). Abundance of megalopae
peaked during the twilight periods of dawn and
dusk. A significant decrease in abundance was
noted in the middle of the night, while midday
samples did not collect any megalopae.
In addition to the wide distribution of megalo-
pae along the transect, sampling occasionally de-
tected the aggregation of very large numbers of
megalopae (denoted by letters over station blocks
n
1,000
o
o
o
0)
D
a.
o
o
0)
0)
E
100 --
10
o o Manta Net
• • Neuston //
Trawl //
I
//
1 -: .
0 I ' ' f^m^^^^^m=^ I i I . ii I T-rr '•' i
18:00 24:00 06:00 12:00 18:00
Time of Day
Figure 8. — Density of megalopae collected by h)oth nets during
a 27 h sampling p>eriod 50 km ofTshore on 8-9 June.
in Figure 7). These aggregations were typically
associated with hydrodynamic or biological struc-
tures in the surface layer.
Event A: Simultaneous tows with the Manta
net and neuston trawl caught megalopae in densi-
ties of 373 and 325/1,000 m^. The neuston trawl
was then redeployed and towed along a foam line
marking a surface convergence. An estimated
60,000 megalopae were collected from the conver-
gence zone (24,000/1,000 m^).
Event B: A series of five consecutive tows
with the trawl net was made. Each tow covered a
linear distance of approximately 1 km in a west-
ward direction, with an estimated distance of 300
m between tows. Megalopae were very sparse in
the simultaneous Manta and trawl tows at the
beginning of the series, with densities of 0 and
1.9/1,000 m"^, respectively. The subsequent neu-
ston trawl tows produced density estimates of 79,
3,555, 2,414 and 187/1,000 m^. Apparently, the
megalopae were aggregated over an area extend-
ing at least 5 km in the east-west direction.
Event C: Following a period of southwest
winds, a large raft of Velella velella was found
around the inshore station. Velella velella were
generally found at offshore stations but at densi-
ties far lower than the 36 kg/1,000 m^ level taken
here. Although the overall extent of the raft could
not be determined at night, the density of megalo-
pae was estimated by both nets at 21,500/1,000
Event D: The density of megalopae sampled
by the Manta net was 790/1,000 m^, while a
simultaneous neuston trawl tow along a conver-
gence zone only 30 m away collected 150,000
megalopae (46,400/1,000 m^).
Census estimates of megalopal abundance were
computed from nighttime samples for each date
(Fig. 9). A modification of Smith's regional census
estimate (1972) was used to compute the total
number of megalopae occurring in a 1 m
wide X 1 m deep track along the surface from the
coast to 55 km offshore:
S (^A)
j=i
where C^ = estimate of abundance of megalopae
per meter of coastline during cruise
k
Aj = volume of water in the track sur-
rounding station i, computed from
311
FISHERY BULLETIN: VOL. 86, NO. 2
10'
0)
o
Q.
_o
a
0)
0)
E
3
10
4 ..
-O
O
O Transect Census
• Convergence Zones
10-
4/6 4/16 5/5 5/18 6/10 6/20 7/8 7/23
Sampling Date
Figure 9. — Census estimates of Dungeness crab megalopae at
night from Manta net samples for each cruise in a 1 m wide x 1
m deep track extending from the coast to 55 km offshore (30-95
km offshore on 7/8). Single samples from convergence zones
sampled with the neuston trawl contained more megalopae
than the entire census estimates for the respective cruises.
the midpoints of the distances be-
tween the adjacent stations
Z), = density of megalopae at station i.
Although census estimates varied between
cruises from approximately 4,000-78,000, these
estimates ignored the concentrations of megalo-
pae in convergence zones or other aggregations
that were sampled by the neuston trawl. The con-
vergences alone contained 5-15 times more
megalopae than the entire census estimates for
their respective cruises.
DISCUSSION
The neustonic habitat off Oregon is a dynamic
environment that supports an abundant and di-
verse fauna including Dungeness crab larvae and
numerous fish species. The large mouth size of
the neuston trawl and minimal disturbance from
the vessel probably contributed to larger catches
of surface-dwelling ichthyofauna than have been
made in previous studies in the same region
(Boehlert et al. 1985).
Several species of neustonic organisms were as-
sociated with specific oceanographic features
such as convergence zones and water masses like
the Columbia River plume. These aggregations of
Dungeness crab megalopae and several fish spe-
cies accounted for the major portion of the total
catch of each taxon. Recent studies by Shanks
(1983, 1985) and Kingsford and Choat (1985,
1986) further emphasize the importance of aggre-
gation of neustonic organisms in oceanic conver-
gences, surface slicks, and around floating ob-
jects. This information clearly demonstrates that
randomized or grid sampling plans will often fail
to detect micro- or meso-scale features of the
environment. Surveys may thus severely under-
estimate the abundance of species, and impor-
tant data on ecological characteristics such as
predator-prey interactions may not be obtained.
This is particularly important in understanding
the role of spatial co-occurrence of patches of lar-
val fishes and their prey on larval survival and
growth (Lasker 1975, 1981; Grover and 011a
1986).
Diel patterns in abundance were striking. Most
of the neustonic organisms were collected at
night, and their absence from daytime collections
may be attributed to vertical migration out of the
surface layer or to visual avoidance of the nets.
The fact that very few larval or juvenile sablefish
and greenlings were collected in subsurface sam-
ples in earlier studies (e.g., Richardson and
Pearcy 1977; Richardson et al. 1980; Kendall and
Clark 1982a; Clark 1984), suggests that these
fishes are obligate inhabitants of the neuston,
and their low abundance in daytime hauls indi-
cates substantial visual avoidance of the nets.
Conversely, other species may be facultative
neuston that undergo diel migrations into the
surface layer (Hempel and Weikert 1972), where
they can most easily be assessed. Dungeness crab
megalopae, in particular, disperse to at least 50 m
during the day, but concentrate at the surface at
dawn and dusk (Booth et al. 1985). Still other
species (the pseudoneuston) may have depth
ranges that overlap with the surface layer.
Previous studies on northwest ichthyoplankton
have defined basic coastal, transitional, and off-
shore assemblages of species during different sea-
sons (Richardson and Pearcy 1977; Richardson et
al. 1980; Kendall and Clark 1984a, b; Clark
1984). In general, the transitional region roughly
parallels the shelf break (approximately 30-40
km offshore in the vicinity of Newport, OR).
Richardson et al. (1980) ascribed the consistency
of these zonal assemblages to the spawning habits
of adults and larval transport in the alongshore
coastal circulation. According to Parrish et al.
(1981), the spatial and temporal patterns of
spawning, and durations of pelagic larval stages
of these species, should correspond with the sur-
face drift patterns of the region to minimize lar-
val advection out of suitable habitats. In the
Pacific Northwest, species with larvae adapted to
the nearshore zone should spawn from fall
312
SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON
through early spring when net surface drift is
primarily northward and onshore. Oceanic lar-
vae, however, should be spawned following the
spring-summer transition when upwelling re-
sults in southward and offshore transport.
The onshore/offshore distributions of several
abundant fish species in 1984 were similar to
those described earlier, and their seasonal occur-
rence usually coincided with the constraints dis-
cussed by Parrish et al. ( 1981). However, four spe-
cies had distributions that crossed the previously
described zonal boundaries (Fig. 10). Early
cabezon larvae appeared to disperse offshore as
they grew. In contrast, early juvenile brown Irish
lords were collected offshore in the early spring,
while the larger juveniles spread inshore across
all stations. Larval greenlings have only been col-
lected close to the coast in late fall (Bates 1984),
while the juveniles collected in this study were
found at all stations prior to upwelling. Juvenile
ronquils were also distributed along the transect,
during both non-upwelling and upwelling condi-
tions.
The dispersion of organisms across the transi-
tion zone may have been accomplished by several
mechanisms. Physical transport of the organisms
by mesoscale hydrographic events (e.g., eddies,
offshore jets, and meanders in the alongshore cur-
rents) (Ikeda and Emery 1984; Mooers and
Robinson 1984; Abbott and Zion 1985; Davis
1985) undoubtedly play important roles in the
onshore/offshore dispersal of planktonic organ-
isms. Transport with these events may be acceler-
ated or hindered by diel vertical migration into
water layers with different zonal flow patterns.
Additionally, several species (especially green-
ling and sablefish) are rapid swimmers whose mo-
bility can contribute to their dispersal or aggrega-
tion.
The rapid change in the composition of the
ichthyoplankton fauna following the onset of up-
welling has not been previously observed. Fishes
abundant prior to upwelling may respond to the
change in the environment in different ways. The
disappearance of greenling and brown Irish lord
juveniles suggests upwelling triggered settle-
ment of these fishes to the demersal habitat uti-
lized by older juveniles and adults. A possible
stimulus for this transition is the breakdown of
the thermocline, which has been identified as a
1 '
90 8*0 70 60
UPWELLING
1 1 1 1
PRE-UPWELLING
V."
45°
nn'
Sablefish 1
10-20 nnm
20-30 mm Brown Irish Lord
S
(vy:
Greenlinq
p-.
/■■'■■
4-6
mm
1 -
^[Vaquina
A^y BAY
&
:■ 1
6-10 mm Cabezon
English Sole
• • • • • •
50 40 30 20 15 10
km
•
5
/
4 4"
_ Rockfish SDD.
3 0'
1 1 1 1
Blue Lanternfish
Northern Anchovy
1
125°
00'
124°
30'
124°
00'
Figure 10. — Summary of the zonal distributions of larval and juvenile neustonic fishes in
the pre- and early upwelling period (April through mid-June) and the strong upwelling
period (July).
313
FISHERY BULLETIN: VOL. 86, NO. 2
barrier to settlement to neustonic juvenile red
hake, Urophycis chuss, (Steiner et al. 1982;
Steiner and 011a 1985).
In contrast, juvenile sablefish apparently re-
main in the water column off Oregon following
the transition to upwelling conditions. Small ju-
veniles (<50 mm) have been captured as far as
250 km offshore in the spring (Kendall and Clark
1982a), and specimens up to 250 mm have been
collected from the surface waters in late summer
with a purse seine (Brodeur and Pearcy 1986).
The very rapid growth rates of young juveniles,
reaching 2 mm/day (Boehlert and Yoklavich
1985; Shenker and 011a 1986), undoubtedly re-
sulted in their increasing ability to avoid the
towed neuston nets used in this study.
The most abundant larvae occurring after the
onset of upwelling were northern anchovy and
rockfish. As observed in previous studies
(Richardson 1973, 1980; Richardson et al. 1980),
these larvae were primarily found offshore. The
occurrence of these species, along with a high
abundance of vertically migrating blue lantern-
fish, on the periphery of the Columbia River
plume 60-90 km offshore on only one cruise, fur-
ther illustrates the patchy nature of neustonic
distributions.
Although the fish fauna in 1984 was generally
characterized by discrete temporal and/or spatial
limitations, the occurrence of Dungeness crab
megalopae transcended these limits through the
4-mo study. Megalopae were the most abundant
organisms collected throughout the survey, but
their abundance varied widely between adjacent
stations, with occasional very dense patches. Sim-
ilar patchy distributions of megalopae were ob-
served off British Columbia by Booth et al. (1985),
who measured horizontal patch dimensions of
2-4 km. Several dense swarms of megalopae were
observed in Bodega Bay, CA, during 1985
(Shenker and Botsford^). These patches were
long, sinuous aggregations extending 10-20 m
along the surface, approximately circular in cross
section and about 1 m in diameter. Densities were
visually estimated to be on the order of thousands
of megalopae per m^.
The temporal occcurrence of crab larvae in the
plankton spans two distinctly different oceanic
regimes, and the larvae are potentially trans-
ported long distances by the seasonal currents.
^Shenker, J. M., and L. W. Botsford, University of California,
Bodega Marine Laboratory, P.O. Box 247, Bodega Bay, CA
94923, unpubl. data.
Larvae typically hatch in mid-winter, pass
through 5 zoeal stages in approximately 90 days,
and then a month-long megalops stage before set-
tling to the bottom (Reilly 1983a). After hatching,
zoeae are released into the northerly flowing
Davidson Current. Despite a general onshore
component of flow of the Davidson Current, older
zoeal stages have been found progressively far-
ther offshore (Lough 1976; Reilly 1983a).
About the time of the spring transition in the
alongshore currents from a northerly to southerly
direction (Huyer et al. 1975), zoeae metamor-
phose into megalopae. To survive, these megalo-
pae must be transported back toward the shore,
and settle to the bottom in depths shallower than
25 m (Reilly 1983b). Larvae have been found at
least 100 km from shore (Lough 1976), although
it is unclear if these larvae make it back to shore,
or are simply lost from the population.
Again, an apparent anomaly exists between the
directions of movement of megalopae and surface
waters, where the Ekman layer moves offshore in
response to upwelling winds. These discrepancies
may be explained by mesoscale mixing processes,
as cited earlier. Alternatively, the diel vertical
movements of the larvae can move them into dif-
ferent water masses with different directions of
zonal movement.
The data from this study and previous research
on vertical migration indicate that at least the
early zoeal stages and megalopae move to the sur-
face during twilight, and below the surface dur-
ing the day (Reilly 1983a). Surface abundance
estimates of megalopae obtained in the 27-h sam-
pling on 8-9 June (Fig. 8), and by Booth et al.
(1985), decreased during the middle of the night.
This movement away from the surface is an ex-
ample of "midnight scattering", perhaps result-
ing from the lack of a light cue to orient plank-
tonic organisms to the surface (Owen 1981). If
midnight scattering is typical for megalopae,
abundance estimates made by sampling along a
transect throughout the night (Figs. 7, 9) proba-
bly underestimate the actual abundance of mega-
lopae utilizing the surface habitat. Megalopae
have been collected as deep as 50-70 m during
the day (Booth et al. 1985; Shenker and Botsford
fn. 6). These observations correlate with Jacoby's
(1982) laboratory demonstration that megalopae
are positively phototactic to dim light, but avoid
bright light.
The present study indicates that the phototac-
tic response of megalopae may assist their return
to shore in several ways. Downward movement
314
SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON
during sunny days, when upwelling wind stress
and offshore Ekman transport is generally
strongest, can move the larvae into the slow on-
shore flow below the surface Ekman layer (Peter-
son et al. 1979). In contrast to the usual northerly
upwelling winds, occasional storms blow from the
southwest, driving the surface layer onshore. The
occurrence of megalopae at the surface during the
day was most pronounced on these dark stormy
days (especially on the second May cruise), and
thus may facilitate their onshore transport.
Observations of megalopae entering into em-
bayments and nearshore areas from Washington
to northern California in 1984 and 1985 indicated
a dramatic increase in the abundance of megalo-
pae and an extended seasonal occurrence in the
plankton, as compared to previous years (Shenker
and Botsford fn. 6; Armstrong^). This high abun-
dance of crab larvae may presage an upswing in
the cyclical crab fishery along these coasts.
Numerous mechanisms have been proposed as
causes of the 10-yr cycles in crab abundance. Al-
though some hypotheses have been discounted,
several models have survived scrutiny as possible
causes of the cycles (see Botsford 1986 for review).
Potential mechanisms of environmental forcing
of the cycles focus on larval transport and sur-
vival. Johnson et al. (1986) detected periodic 10-
yr cycles in the occurrence and strength of south-
ward stress during the late larval period that
significantly correlated with commercial crab
catch 4 and 5 years later. This lag corresponds to
the time between larval settlement and growth
into the adult fishery (Botsford 1984).
Model simulations by Botsford (1986) indicated
that nonlinear effects of wind on larval transport
can produce the cyclical swings in crab abun-
dance. However, the models do not preclude the
possibility that density-dependent phenomena
(such as cannibalism by adult crabs on newly set-
tled juveniles, and predation by nemertean
worms on egg masses) may act in concert with the
environmental forcing to produce the observed
cycles.
The water's surface is the only oceanic habitat
that is easily accessible to observation using tech-
niques ranging from satellite and aerial scanning
to shipboard visual sighting of targets and contin-
uous monitoring of environmental parameters.
Because of this accessibility, micro- and meso-
scale patterns in distribution of neustonic taxa
'^D. A. Armstrong, School of Fisheries, WH-10, University of
Washington, Seattle, WA 98195, pers. commun. October 1984.
and their associations with hydrographic and bio-
logical characteristics of the surface zone can be
determined more easily than in other environ-
ments. The neustonic realm thus offers an excel-
lent opportunity to investigate the mechanisms of
transport of the early stages into appropriate
nursery habitats, and the availability of food for
growth that are required for successful recruit-
ment into adult stocks (Hjort 1914; Lasker 1975,
1981; Frank and Leggett 1982; Sinclair et al.
1984).
ACKNOWLEDGMENTS
This research was supported by the Northwest
and Alaska Fisheries Center of the National
Marine Fisheries Service. 1 thank Captain Leland
Oldenberg and his crew on the FV Cumberland
Trail, and my colleagues R. Brodeur, A. Chung,
J. Fisher, J. Hennessey, L. Krasnow, B. Mundy,
C. Paczkowski, and E. Rexstad for their assis-
tance on the sampling cruises. W. Laroche helped
identify juvenile Sebastes. W. G. Pearcy provided
valuable advise on the design and operation of
this project, and comments on the manuscript.
Helpful suggestions on this paper were also given
by R. Brodeur, A. W. Kendall, and two anony-
mous referees.
LITERATURE CITED
Abbott. M. R., and P. M. Zion.
1985. Satellite observations of phytoplankton variability
during an upwelling event. Cont. Shelf Res. 4:661-680.
Barkley, R. A.
1972. Selectivity of towed net samplers. Fish. Bull., U.S.
70:799-820.
Bates. R. D.
1984. Ichthyoplankton off Washington, Oregon and
northern California, October-November, 1981. Natl.
Mar. Fish. Serv., Northwest and Alaska Fish. Cent. Proc-
ess. Rep. 84-12, 41 p.
Boehlert, G. W., D M. Gadomski, and B C. Mundy.
1985. Vertical distribution of ichthyoplankton off the Or-
egon coast in spring and summer months. Fish. Bull.,
U.S. 83:611-622.
Boehlert, G. W., and M. M. Yoklavich.
1985. Larval and juvenile growth of sablefish,
Anoplopoma fimbria, as determined from otolith incre-
ments. Fish. Bull., U.S. 83:475-481.
Booth. J., A. Phillips, and G S Jamieson.
1985. Fine scale spatial distribution of Cancer magister
megalopae and its relevance to sampling methodol-
ogy. In B. R. Melteff (editor), Proceedings of the sympo-
sium on Dungeness crab biology and management,
p. 273-286. Univ. Alaska Sea Grant Coll. Program,
Rep. No. 85-3. Fairbanks, AK 99701.
Botsford. L. W.
1984. Effect of individual growth rate on expected behav-
315
FISHERY BULLETIN: VOL. 86, NO. 2
ior of the northern California Dungeness crab (Cancer
magister) fishery. Can. J. Fish. Aquat. Sci. 41:99-107.
1986. Population dynamics of the Dungeness crab (Cancer
magister). In G. S. Jamieson and N. Bourne (editors),
North Pacific workshop on stock assessment and man-
agement of invertebrates. Can. Spec. Publ. Fish. Aquat.
Sci. 92:140-153.
Brodeur, R D. D M Gadomski, W G Pearcy. H P.
Batchelder, and C B Miller
1985. Abundance and distribution of ichthyoplankton
in the upwelling zone off Oregon during anomalous el
Nino conditions. Estuarine Coastal Shelf Sci. 21:365-
378.
Brodeur, R. D. B C Mundy, W G Pearcy, and R W.
WiSSEMAN.
1987. The neustonic fauna in coastal waters of the north-
east Pacific: Abundance, distribution, and utilization by
juvenile salmonids. Oreg. State Univ. Sea Grant Coll.
Program, Rep. ORESU-T-87-001. Corvallis, OR 97331.
Brodeur, R D . and W. G. Pearcy.
1986. Distribution and relative abundance of pelagic non-
salmonid nekton off Oregon and Washington, 1979-
1984. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 46,
85 p.
Brown, D M , and L Cheng
1981. New net for sampling the ocean surface. Mar.
Ecol. Prog. Ser. 5:225-227.
Clark, J B
1984. Ichthyoplankton off Washington, Oregon and
northern California, May-June, 1981. Natl. Mar. Fish.
Serv., Northwest and Alaska Fish. Cent. Process. Rep.
84-11, 46 p.
Davis, R E
1985. Drifter observations of coastal surface currents dur-
ing CODE: the method and descriptive view. J.
Geophys. Res. 90:4741-4755.
Frank, W T , and W C. Leggett.
1982. Coastal water mass replacement: its effect on
zooplankton dynamics and the predator-prey complex as-
sociated with larval capelin (Mallotus villosus ). Can. J.
Fish. Aquat. Sci 39:991-1003.
Grover, J J , AND B L Olla.
1986. Morphological evidence for starvation and prey size
selection of sea-caught larval sablefish, Anoplopoma fim-
bria. Fish. Bull., U.S. 84:484-489.
HEMPEL, G , AND H Weikert
1972. The neuston of the subtropical and boreal North-
eastern Atlantic Ocean. A review. Mar. Biol. (Berl.)
13:70-88.
Hjort, J
1914. Fluctuations in the great fisheries of northern
Europe viewed in the light of biological research.
Rapp. P. -v. Reun. Comm. int. Explor. Sci. Mer 20:1-
228.
HUYER, A , R D PiLLSBURY, AND R L. SMITH
1975. Seasonal variation of the alongshore velocity field
over the continental shelf off Oregon. Limnol.
Oceanogr. 20:90-95.
Ikeda, M . AND W J Emery
1984. Satellite observations and modeling of meanders in
the California Current system off Oregon and Northern
California. J. Phys. Oceanogr. 14:1434-1450.
Jacoby, C a.
1982. Behavioral responses of the larvae oiCancer magis-
ter Dana (1852) to light, pressure, and gravity. Mar.
Behav. Physiol. 8:267-283.
Johnson, D F , L W Botsford, R D Methot, Jr , and T C.
Wainwright
1986. Wind stress and cycles in Dungeness crab (Cancer
magister) catch off California, Oregon and Washing-
ton. Can. J. Fish. Aquatic Sci. 43:838-845.
Kendall, A W . Jr , and J Clark
1982a. Ichthyoplankton off Washington, Oregon, and
Northern California. April-May, 1980. Natl, Mar.
Fish. Serv., Northwest Alaska Fish. Cent. Process. Rep.
82-11, 44 p.
1982b. Ichthyoplankton off Washington, Oregon and
Northern California. August, 1980. Natl. Mar. Fish.
Serv., Northwest and Alaska Fish. Cent. Process. Rep.
82-12, 43 p.
Kingsford, M J , AND J H Choat
1985. The fauna associated with drift algae captured with
a plankton-mesh purse seine net. Limnol. Oceanogr.
30:618-630.
1986. Influence of surface slicks on the distribution and
onshore movements of small fish. Mar. Biol. (Berl.)
91:161-171.
Laroche, J L, AND S. L Richardson
1979. Winter-spring abundance of larval English sole,
Parophrys vetulus , between the Columbia River and
Cape Blanco, Oregon, during 1972-1975, with notes on
occurrences of three other pleuronectids. Estuarine
Coastal Mar. Sci. 8:455-476.
Lasker, R
1975. Field criteria for survival of anchovy larvae: the
relation between the inshore chlorophyll layers and suc-
cessful first feeding. Fish. Bull., U.S. 71:453-462.
1981. The role of a stable ocean in larval fish survival and
subsequent recruitment. In R. Laisker (editor), Marine
fish larvae. Morphology, ecology and relation to fisheries,
p. 80-87. Wash. Sea Grant Program, Univ. Wash.
Press, Seattle.
Lough, R G
1976. Larval dynamics of the Dungeness crab, Cancer
magister, off the central Oregon coast, 1970-71. Fish.
Bull., U.S. 74:353-375.
MooERS, C. N K , and a. R. Robinson.
1984. Turbulent jets and eddies in the California Current
and inferred cross-shore transport. Science 223:51-53.
MUNDY, B. C.
1983. Yearly variation in the abundance and distribution
of fish larvae in the coastal zone off Yaquina Head, Ore-
gon, from June 1969 to August 1972. M.S. Thesis, Ore-
gon State Univ., Corvallis, OR, 158 p.
MuRPHY, G I , AND R I Clutter
1972. Sampling achovy larvae with a plankton purse
seine. Fish. Bull., U.S. 70:789-798.
Owen, R W
1982. Fronts and eddies in the sea: Mechanisms, interac-
tions and biological effects. In A. R. Longhurst (editor).
Analysis of marine ecosystems, p. 197-233. Acad.
Press, N.Y.
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.
Peterson, W T , C B Miller, and A. Hutchinson.
1979. Zonation and maintenance of copepod jwpulations
in the Oregon upwelling zone. Deep-Sea Res. 26:467-
494.
Reilly, P N
1983a. Dynamics of Dungeness crab. Cancer magister.
316
SHENKER: OCEANOGRAPHIC ASSOCIATION OF NEUSTONIC MEROPLANKTON
larvae off central and northern California. In P. W.
Wild and R. N. Tasto (editors!, Life history, environment,
and mariculture studies of the Dungeness crab. Cancer
magister . with emphasis on the central California fish-
ery, p. 57-85. Calif Dep. Fish game, Fish Bull. 172.
1983b. Predation on Dungeness crabs, Cancer magister,
in central California. In P. W. Wild and R. N. Tasto
(editors), Life history, environment, and mariculture
studies of the Dungeness crab. Cancer magister, with
emphasis on the central California fishery, p. 155-
164. Calif Dep. Fish. Game, Fish Bull. 172.
Richardson. S L.
1973. Abundance and distribution of larval fishes in wa-
ters off Oregon, May-October 1969, with special empha-
sis on the northern anchovy, Engraulis mordax. Fish.
Bull., U.S. 71:697-711.
1980. Spawning biomass and early life of northern an-
chovy, Engraulis mordax, in the northern subpopulation
off Oregon and Washington. Fish. Bull., U.S. 78:855-
877.
Richardson, S L . J L Laroche, and M. D. Richardson.
1980. Larval fish assemblages and associations in the
northeast Pacific Ocean along the Oregon coast, winter-
spring, 1972-1975. Estuarine Coastal Mar. Sci.
11:671-699.
Richardson, S L , and W G Pearcy.
1977. Coastal and oceanic fish larvae in an area of up-
welling off Yaquina Bay, Oregon. Fish. Bull., U.S.
75:125-146.
Shanks, A L
1983. Surface slicks associated with tidally forced inter-
nal waves may transport pelagic larvae of benthic inver-
tebrates and fishes shoreward. Mar. Ecol. Prog. Ser.
13:311-315.
1985. Behavioural basis of intemal-wave-induced shore-
ward transport of megalopae of the crab Pachygrapsus
crassipes. Mar. Ecol. Prog. Ser. 24:289-295.
Shenker, J. M . AND B L Olla.
1986. Laboratory feeding and growth of juvenile
sablefish, Anoplopoma fimbria. Can. J. Fish. Aquat.
Sci. 43:930-937.
Sinclair, M. F.. M. J. Tremblay, and P. Bernal.
1984. El Nino events and variability in a Pacific mackerel
{Scomber japonicus ) survival index: supj)ort for Hjort's
second hypothesis. Can. J. Fish. Aquat. Sci. 42:602-
608.
Small. L. F . and D W Menzies.
1981. Patterns of primary productivity and biomass in a
coastal upwelling region. Deep-Sea Res. 28:123-149.
Smith, P E
1972. The increase in spawning biomass of northern an-
chovy, Engraulis mordax. Fish. Bull., U.S. 70:849-874.
Steiner, W W., J. J. LuzcKovicH. and B L. Olla.
1982. Activity, shelter usage, growth and recruitment of
juvenile red hake, Urophycis chuss. Mar. Ecol. Prog.
Ser 7:125-135.
Steiner, W W., and B L Olla
1985. Behavioral responses of prejuvenile red hake, Uro-
phycis chuss, to experimental thermoclines. Environ.
Biol. Fish 14:167-173.
317
WINTER-TIME DISTRIBUTION AND ABUNDANCE OF COPEPOD
NAUPLII IN THE NORTHERN GULF OF MEXICO
M J Dagg^, P B Ortner2, and F. Al-Yamani^
ABSTRACT
Copepod nauplii were collected from continental shelf waters in 3 regions of the northern Gulf of
Mexico during winters between 1981 and 1984, off Cape San Bias, Florida, off the Mississippi River
delta, and off of Galveston, Texas. Some statistically significant iP < 0.05) patterns in the abundance
and distribution of nauplii were observed: there was significant interannual variability in naupliar
concentrations within the region around the Mississippi River delta; naupliar concentrations in the
upper 10 m decreased in the onshore-offshore direction in 2 of 4 comparisons; naupliar concentrations
in the upper 10 m differed regionally in 2 of 3 comparisons; and naupliar concentration was correlated
with chlorophyll concentration in 4 of 5 comparisons.
Maximum concentrations of nauplii (number per m^) within a water column were 2-10 times
greater at stations influenced by the Mississippi River plumes than in the other 2 regions. This
condition in attributed to vertical stratification imparted to the water column by the inflowing low
salinity water from the Mississippi River. We conclude that the physical stratification provides a
mechanism for the establishment of high concentrations of nauplii that otherwise would not exist in
the winter months on the continental shelf.
Microzooplankton are important diet items for
larval fish (Arthur 1976; Gamble et al. 1981;
Checkley 1982; Govoni et al. 1983; Houde and
Lovdal 1984; Stoecker and Govoni 1984), and gut
analyses indicate that copepod nauplii are fre-
quently the dominant prey form found in the lar-
vae of many fish species (Duka and Gordina
1973). Available concentrations of micro-
zooplankton are considered an important deter-
minant of larval survival rates in the ocean be-
cause this relationship has been demonstrated in
the laboratory (Laurence 1974; Houde 1978) and
because field studies have demonstrated a rela-
tionship between regions or periods of high mi-
croplankton and high larval abundance (Arthur
1977; Lasker 1978). Because survival is enhanced
by increased food availability, oceanographic
processes that result in increased concentration
or production of prey items are important.
The gulf menhaden, Brevoortia patronus (Clu-
peiformes), supports the largest volume fishery in
the United States (U.S. Department of Commerce
1983). Spawning occurs in the wintertime, from
October to March, in continental shelf waters of
the northern Gulf of Mexcio (Fore 1970; Lewis
and Rothmayr 1981; Warlen 1988), primarily off
iLouisiana Universities Marine Consortium, Chauvin, LA
70344.
^Applied Oceanic Marine Laboratory, Ocean Coastal Divi-
sion, NOAA, 4301 Rickenbacker Causeway, Miami, FL 33149.
of Mississippi and Alabama to the east of the Mis-
sissippi River delta, and ofi" of Louisiana to the
west of the delta. During February and December
1982, concentrations of menhaden larvae in the
region near the Mississippi River delta were
greater in plume waters than outside and were
much higher at the plume front (Govoni and
Hoss"^). Furthermore, gut contents of these larvae
indicated that nauplii were the most abundant
prey items.
Much of the continental shelf water of the
northern and western Gulf of Mexico is vertically
unstratified during the winter, and winter is also
the season of minimum primary productivity in
these regions, as it is in shelf waters of the north-
east and southeast United States. However,
coastal waters and nearshore regions influenced
by freshwater inputs can be physically stratified
during winter when low salinity plumes disperse
over higher salinity shelf waters. We postulated
that shelf waters influenced by fresh water from
the Mississippi River plumes would be regions of
increased production and concentration of micro-
zooplankton. The purpose of this paper is to exam-
ine this hypothesis by describing the vertical and
horizontal distribution and abundance of copepod
nauplii in shelf waters of the northern Gulf of
Manuscript accepted December 1987.
FISHERY BULLETIN: VOL. 86, NO. 2, 1987.
^Govoni and Hoss. Unpubl data. Southeast Fisheries Center
Beaufort Laboratory, National Marine Fisheries Service,
NOAA, Beaufort, NC 28516.
319
FISHERY BULLETIN: VOL. 86, NO. 2
Mexico and by comparing these with oceano-
graphic and hydrographic indicators of water
column stratification and productivity.
METHODS
Samples were collected during 5 winter cruises
between November and March, during 1981-84,
(Table 1). Microplankton samples and hydro-
graphic data were collected from 3 general re-
gions of the northern Gulf of Mexico: off of Galve-
ston, TX, in the region of the Mississippi River
delta, and off of Cape San Bias, FL (Fig. 1). Sam-
ples were not collected from all 3 regions on each
cruise.
Water samples for chlorophyll and nutrient
analyses were collected with Niskin bottles. For
each chlorophyll analysis, between 25 and 150
mL of sea water was filtered onto a GF/F or GF/C
glass fiber filter and homogenized by grinding in
90% aqueous acetone. Fluorescence of the filtrate,
brought up to a volume of 10.0 mL, was deter-
mined before and after acidification with 2 drops
of 10% HCl using a Turner Designs'' Model 10
fluorometer. Chlorophyll and pheopigment con-
centrations as chlorophyll equivalents were de-
termined from
chlorophyll (fxg/L) =
Table 1.— Station information, 1981-84.
Kifo-
fa)
V
KiRfa
-fo)
pheopigment (|JLg/L)
where K is the machine calibration constant, [„
and fa are the fluorescence readings before and
after acidification, R is the acid ratio, and u is the
volume of seawater filtered, in mL (Strickland
and Parsons 1968).
Samples for nutrient analyses were frozen. Ni-
trate and nitrite were analyzed according to
method number 353.2 described by EPA publica-
tion number EPA 600/4-79-020 (Environmental
Protection Agency 1979).
Temperature and salinity were measured by
several methods. During cruise I, temperature
was measured by the temperature sensor on the
MOCNESS^ net, and salinity was measured with
a refractometer. During cruises II and III, tem-
perature was measured with expendable bathy-
Depth
Lat.
Long.
Station
Date
Time
(m)
(N)
(W)
1-1
2-13-81
0000
18
29 '05'
94°07'
1-2
2-13-81
1200
18
2906'
94' 07'
1-3
2-14-81
0001
90
28 04'
93"03'
1-4
2-14-81
1200
90
28 04'
93"03'
1-5
2-15-81
0000
180
27' 54'
92'^51 '
1-6
2-15-81
1200
180
25' 54'
92'50'
1-7
2-16-81
0001
180
28 34'
89"37'
1-8
2-17-81
1200
180
28 35'
89 38'
1-9
2-18-81
0001
90
2850'
89 "16'
1-10
2-18-81
1200
90
28 50'
89 "16'
1-11
2-19-81
0000
18
28^56'
89'29'
1-12
2-19-81
1200
18
28°56'
89'^29'
1-13
2-20-81
0000
90
29 09'
85' 56'
1-14
2-21-81
1200
90
29 08'
85"56'
1-15
2-22-81
0000
180
29°28'
8607'
1-16
2-22-81
1200
180
29°28'
86°07'
1-17
2-23-81
0001
18
29°36'
85°47'
1-18
2-23-81
1200
18
29^35'
85 47'
1-19
2-24-81
0001
18
30°12'
87"06'
11-1
12-05-82
0700
25
28°53'
89' 29'
11-2
12-05-82
1255
62
28°53'
89^32'
11-3
12-05-82
1945
33
28' 59'
89°34'
11-4
12-06-82
2120
887
28 20'
89°27'
11-5
12-07-82
0715
885
28°20'
89'27'
11-6
12-07-82
1220
834
28' 19'
89 25'
11-7
12-08-82
1820
443
28°32'
89°53'
11-8
12-09-82
0830
402
28°34'
89°53'
11-9
12-09-82
1215
461
28^32'
89 '53'
11-10
12-13-82
1245
44
29'26'
85°53'
11-11
12-15-82
1930
19
28°54'
89°29'
11-12
12-16-82
0900
68
28°51'
89°32'
III-1
11-19-83
1200
27
28°54'
89°29'
III-2
11-20-83
0800
35
28°47'
89°59'
III-3
11-20-83
1946
732
28°38'
89°00'
III-4
11-21-83
0555
27
28°54'
89°30'
III-5
1 1 -22-83
1100
48
28°51'
89°30'
III-6
1 1 -23-83
0500
194
28°31'
89'=37'
III-7
1 1 -24-83
1030
50
28=50'
89°31'
III-8
11-25-83
0820
45
28^55'
89°34'
III-9
11-25-83
1945
40
28°56'
89°36'
111-10
11-26-83
1200
16
28°54'
89°29'
111-11
11-27-83
0900
9
29'02'
89°30'
111-12
11-28-83
0840
29
28^48'
89^57'
111-13
11-28-83
1250
44
28°56'
89°51'
111-14
11-29-83
0950
24
29''03'
89°39'
111-15
1 1 -30-83
0930
44
28°47'
89°58'
111-16
12-01-83
0800
38
28°52'
89°29'
IV- 1
3-14-83
53
27' 58'
95°53'
IV-2
3-19-83
59
28=18'
90^41 '
V-1
2-21-84
0850
18
28 54'
90°25'
V-2
2-22-84
0830
18
28' '54'
90°25'
V-3
2-23-84
0910
18
28°54'
90°25'
V-4
3-1-84
0835
18
28°54'
90°25'
V-5
3-1-84
0830
18
28^54'
90°25'
V-6
3-3-84
0730
18
28'54'
90°25'
'^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
^Multiple opening-closing net and environmental sensing
system.
thermographs (Sippican Instruments). Salinity
was measured with a YSI Model 33-S-C-T sali-
nometer on cruise II, and with a Beckman Model
RS5-3 salinometer on cruise III. For both of these
cruises, bottle samples, analyzed in the labora-
tory with a Guildline Model 8400A Autosal, were
used to check the shipboard salinity measure-
320
DAGG ET AL.: DISTRIBUTION OF COPEPOD NAUPLII
c
o
c
o
C/2
ca
D
O
321
FISHERY BULLETIN: VOL, 86, NO. 2
ments. On cruise IV, a Plessey Model 9040 CTD
was used for both temperature and salinity. On
cruise V, salinity samples were stored in bottles
and analyzed in the laboratory using a Guildline
Model 8400A Autosal, and temperature was
measured with reversing thermometers.
Microzooplankton were collected using 5 L
(cruises I, II, and V) or 30 L (cruises III and IV)
Niskin bottles. During cruise I, 1 L samples of
untreated water were collected from 3 depths and
preserved for later analysis. During subsequent
cruises, samples were collected from up to 8
depths; four liters were poured gently through a
20 |xm sieve, backwashed with filtered seawater
into a sample jar, and preserved in a 5% formalin-
seawater solution. In the laboratory, all organ-
isms in each preserved sample were identified
and counted. In selected samples, the length and
width of the first 50 copepod nauplii were mea-
sured using an ocular micrometer. In addition,
during cruise I larger nauplii were collected with
0.1 m^ 64 |jLm mesh nets nested inside the 1 m^
333 ^JLm mesh nets of a MOCNESS net (Wiebe et
al. 1976). Estimates of naupliar abundance from
these collections were used separately in the
analyses for patterns in distribution and abun-
dance. Lastly, from cruise I, the total number of
copepods in the shallow-water samples, total
number of copepods from the 333 jxm nets, and the
zooplankton displacement volume from the 333
(xm nets were analyzed for patterns in distribu-
tion and abundance.
RESULTS
Samples from the Mississippi River delta re-
gion, collected during 1981 (cruise I), 1982 (cruise
II), and 1983 (cruise III), were compared to deter-
mine if there was significant interannual vari-
ability within a region. Whole water samples and
samples retained on 20 ixm mesh sieves were
pooled for this analysis under the assumption
that nauplii in a whole water sample would be
retained on a 20 jjim mesh screen. The 64 [im net
samples were excluded from this analysis. This
test was made using average naupliar concentra-
tions in the upper 10 m only; naupliar concentra-
tions were typically lower in deeper waters, and
we did not have the same number of deep and
shallow stations for each of the 3 sampling years.
A 1-way analysis of variance indicated significant
variation between years within the Mississippi
River delta region {P <0.01). All 3 years were
statistically different. These results indicated to
us that regional or other spatial comparisons can
only be made using samples collected during the
same year. Interannual comparisons for other re-
gions were not possible.
Samples were collected from all 3 regions dur-
ing cruise I. Four tests for regional differences
during this cruise were made. First, regions were
compared based on naupliar concentrations in the
upper 10 m only (Table 2); based on whole water
collections F was significant at P < 0.05 and
based on samples collected with 64 \xm mesh nets
F was significant at P < 0.09. Second, regions
were compared based on naupliar concentrations
from all depths (Table 2); based on whole water
collections F was significant at P < 0.01 and
based on samples collected with 64 (xm mesh nets
F was significant at P < 0.22. Thus, whole water
samples collected on cruise I indicated significant
regional differences in naupliar concentration
but samples collected with 64 fxm mesh nets did
not. In addition, there were significant regional
differences in the concentrations of total copepods
(P < 0.01) and in zooplankton displacement vol-
ume (P < 0.05) during cruise I (data not shown).
Regional comparisons were not possible during
other cruises because all 3 regions were not sam-
pled in years other than 1981.
Where possible, we also tested for onshore-
offshore gradients in naupliar concentrations in
the upper 10 m. Stations were categorized accord-
ing to bottom depth as shallow, <18 m, interme-
diate, 18-90 m, and deep, >90 m, and compari-
sons were made between these 3 depth categories.
A highly significant onshore-offshore difference
existed during cruise I for nauplii collected with
the 64 |xm MOCNESS nets (Table 3). During
cruise III samples were only collected in the Mis-
sissippi River delta region but onshore-offshore
differences within this region were significant
(Table 3). No significant onshore-offshore gradi-
ents were observed in naupliar concentrations de-
termined from whole water samples collected
from all 3 regions during cruise I, or in nauplii
from only the Mississippi River delta region dur-
ing cruise II (Table 3).
Although regional differences in average con-
centrations of nauplii were not strong, the verti-
cal distributions were different. For example, sev-
eral stations at approximately the 50 m isobath
are compared in Figure 2. At station IV- 1 (Fig.
2a) ofT the coast of Texas, there was no marked
vertical heterogeneity in naupliar concentra-
tions, the maximum concentration was low (23
nauplii/L), and the average concentration was
322
DAGG ET AL.: DISTRIBUTION OF COPEPOD NAUPLII
Table 2. — Regional comparisons of naupliar concentrations during
cruise I. Samples were collected by 2 methods: whole water sam-
ples were collected with Niskin bottles, and 64 (i.m mesh nets were
used to collect the net caught samples.
Cruise
1
Depths
Sample
type
Probability
of F
upper 10 m
whole water
28
<0.05
upper 10 m
64 ^m nets
126
<0.09
all depths
whole water
57
<0.01
all depths
64 |jLm nets
222
<0.22
Table 3. — Onshore-offshore comparisons of naupliar concentra-
tions from the upper 10 m of each station. Tests for cruise I are
based on naupliar concentrations from all 3 regions pooled then
separated into 3 depth categories. During cruises II and III. nauplii
were only collected from the Mississippi River delta region.
Cruise
I
Regions
Sample
type
Probability
of F
all
whole water
28
<0.14
all
64 ^m nets
126
<0.01
Mississippi R.
20 jjim sieve
28
<0.25
Mississippi R.
20 M-m sieve
49
<0.01
NAUPLIUS CONCENTRATION (number/ liter)
Q.
UJ
Q
I
»-
Q.
UJ
Q
c
25 50 75
1, . 1 1
10-
^ /
4
1
If 1
1
1
30-
*
^
50-
Hi
112
n3
(0
Figure 2. — Naupliar concentrations (number/liter) from selected stations near the 50 m isobath on the Texas continental
shelf (a, station IV-1), the Louisiana Shelf (b, station IV-2, c, stations II-l, II-2, II-3), and the northwest Florida shelf (d,
station 11-10).
323
FISHERY BULLETIN: VOL. 86, NO. 2
low (11 nauplii/L). The same pattern was ob-
served off Cape San Bias (Fig. 2d); there was no
marked vertical heterogeneity, and the maxi-
mum and average concentrations were low, 14
and 12 nauplii/L. In contrast, nauplii at stations
near the 50 m isobath in the region near the Mis-
sissippi River delta typically showed marked ver-
tical heterogeneity in abundance. At station IV-2
(Fig. 2b), the maximum concentration was 187
nauplii/L and the minimum was 24 nauplii/L. At
station II- 1, II-2, and II-3 in the Mississippi River
delta region (Fig. 2c) naupliar abundances were
also vertically heterogenous, although concentra-
tions were lower than at Station IV-2. In general,
based on all our vertical profiles (most not
shown), regions influenced by the Mississippi
River can contain high concentrations of nauplii
in the surface layer, while subsurface concentra-
tions are similar to concentrations in the other 2
regions.
Vertical structure in the distribution of nauplii
appears to be related to physical structure. At the
stations off the coasts of Texas and Florida (Fig.
2a, 2d), temperture and salinity were essentially
vertically homogenous (Fig. 3a, 3d) whereas off
the Mississippi River delta temperature and
20%. 24 28
32 36
24%. 28
16*0 18
Figure 3. — The vertical distribution of temperature (•) and salinity (x) at the stations represented in Figure 2.
324
DAGG ET AL.: DISTRIBUTION OF COPEPOD NAUPLII
salinity were vertically stratified (Fig. 3b, 3c).
The same pattern was seen in the chlorophyll a
distributions (Fig. 4) and the nitrate distributions
(Fig. 5).
Correspondence between chlorophyll a concen-
tration and naupliar abundance was computed;
the correlation coefficient for a linear regression
of chlorophyll concentration on naupliar abun-
dance for all 5 cruises in = 269) was 0.39, signifi-
cant at the 1% level. However, analysis of covari-
ance showed that the relationship between
chlorophyll and naupliar concentration was sig-
nificantly different between cruises, and there-
fore the data from the 5 cruises should not be
pooled. Although a significant relationship be-
tween chlorophyll concentration and naupliar
concentration typically exists, it is not consistent
from cruise to cruise. Regression parameters and
correlation coefficients for each cruise are shown
in Table 4. Pairwise tests indicate that the slope
of the regression for cruise V was significantly
different from all the others, cruise IV was differ-
0 I 0
• CHLOROPHYLL a (ug/liter)
2.0 3 0 4.0 10 2.0
Figure 4. — The vertical distribution of chlorophyll a at the stations represented in Figure 2.
NITRATE (ug-at/liter)
Figure 5. — The vertical distribution of nitrate at the stations represented in Figure 2.
325
FISHERY BULLETIN: VOL. 86, NO. 2
Table 4. — Parameters and coefficients for linear regressions of chlorophyll
concentration on nauplius concentration for each cruise. Y = a + b(X) where
y = naupliar concentration as number per liter, and X = chlorophyll concentra-
tion as fjig per liter. Significance level tested is 0.01.
Significance
Significance
of
Cruise
n
Intercept
Slope
of slope
/-2
correlation
1
57
12.9
9.7
<0.01
0.40
<0.01
II
70
5.4
15.4
-0.01
0.53
<0.01
III
80
3.7
9.2
<0.01
0.34
<0.01
IV
20
6.9
39.4
<0.01
0.94
<0.01
V
42
26.1
1.1
NS
0.04
NS
ent from all others, cruise II was different from
all the others, and cruises I and III were not dif-
ferent from each other but different from cruises
II, IV, and V. The slopes of all relationships ex-
cept cruise V were significantly different from
zero.
The size-frequency distribution of nauplii did
not vary in any systematic manner (Fig. 6). For
example, at station IV-1, the bulk of the nauplii
were between 20 and 100 \x.m at all depths sam-
pled during both the daytime and nighttime peri-
ods. The size-frequency distribution was essen-
tially identical at station IV-2, off Louisiana (data
not shown).
At some stations, copepodid stages <600 ixm in
length (50-200 \xm in width) were counted in ad-
dition to nauplii because they are potential prey
items for menhaden larvae. Copepodites were
usually not as abundant as nauplii, especially at
stations with high naupliar abundances. On occa-
sion, they were as abundant or, at specific depths,
more abundant than the nauplii (Table 5). The
ratio of nauplii to copepodites varied widely, be-
tween 16.3 and 0.3, so it is not possible to assign
a constant factor to naupliar abundances to esti-
mate the increase in available prey attributable
to copepodid stages.
Other microzooplankton were usually not as
abundant as nauplii or copepods. In our samples,
various forms of eggs reached a maximum density
of 6 eggs/L (cruise III station 6 at 3 m, data not
shown). Pelecypod larvae at one station (III-ll)
were abundant, reaching a maximum of 29 lar-
vae/L at 4 m, compared with 22 nauplii, 7 cope-
podites, and 2 eggs/L in the same sample. Only on
this one occasion were organisms other than cope-
podites or nauplii the dominant form of micro-
zooplankton.
Table 5. — Abundance (number/liter) of other microzooplankton in addition to nauplii at
selected stations during cruise III.
Station
Depth
(m)
Nauplii
Cope-
podids
<600^m
Cope-
podids
>600 ii.vn
Larva-
ceans
Eggs
Other
III-3
6
17.3
7.5
1.0
7.8
0.5
1.8
9
29.3
12.0
0.5
11.0
0
1.8
12
17.5
8.0
1.0
6.8
1.5
2.3
20
7.8
2.8
0
1 0
2.0
0.5
III-7
0
24.0
6.3
4.0
0
1.0
1.0
3
15.3
12.3
1.0
0
0.3
1.3
5
3.0
5.0
1.0
0
0.8
0.3
13
4.8
5.8
0.3
0
0
0.5
22
2.0
1.0
0.3
0
0
0
111-14
0
15.8
12.0
1.3
0
1.3
2.0
5
44.8
3.8
0.3
0.5
3.0
4.0
10
40.0
6.8
0.8
1.3
4.3
0.3
15
11.5
8.0
3.8
1.0
1.8
0.3
20
12.0
35.0
6.8
0.3
0.3
0.8
111-16
0
15.8
25.3
5.0
0
0
12.8
6
8.0
16.0
2.3
0
1.0
2.5
9
30.0
14.0
2.0
0
0.3
2.5
12
29.0
13.3
0.5
0
0.8
1.3
15
17.0
12.5
1.0
0
0.8
0.5
25
31.8
12.8
0.8
0
0
5.0
326
DAGG ET AL.: DISTRIBUTION OF COPEPOD NAUPLII
NIGHT
20-
^i=H
-T 1
DAY
0 m 20
-1 1 1 I
20-
0^— -^
20
-I 1 1 1
52
^ 0
-I 1 1
20-1
E
13
C
o
LlI
o
:z.
o
o
CO
_J
CL
0
20-
0
20
0
20n
r~| .
somple lost
-1 1 1 1
20-1
n
-I 1 r
20-
:i-^
20
5m
X4-
lOm
15m
20 m
25m
30m
20n
0
20-
0
20-
0
20-
0
20-
0
20-
^aXL
5=L
r.
35m
-I ' 1
-T r ■!
1 1 I
o o o o o o o
^ CD ^ (I> O ^
BODY WIDTH (mm)
40m
20
=F=L
o o o o o o o
•3- CD oj i£ o ^
BODY WIDTH (mm)
45 m
Figure 6.— The size-frequency distribution
of nauplii from station IV- 1.
327
FISHERY BULLETIN: VOL. 86, NO. 2
DISCUSSION
Menhaden larvae between 3 and 20 mm long
collected during wintertime from 3 regions in the
northern Gulf of Mexico contained a variety of
diet items, including dinoflagellates, tintinnids,
copepod eggs, nauplii, copepodids, juvenile pele-
cypods, and pteropods (Govoni et al. 1983, 1985).
Dinoflagellates and tintinnids constituted the
main diet items of larvae <5.0 mm long but were
replaced by copepod nauplii, then copepodids and
small adult copepods as larvae grew larger.
In another study, Stoecker and Govoni (1984)
found that copepod nauplii were the dominant
items in the diet of menhaden larvae 7.0-9.0 mm
long, although only 12 fish were examined. Later
studies, however, have verified that nauplii are
the dominant items in the diet of a wide size
range of larval menhaden collected in the
boundary between the plume of the Mississippi
River and oceanic waters (Govoni and Hoss fn. 3)
The maximum width of food items in menhaden
larvae <5 mm long was about 200 |xm (Govoni et
al. 1983). This width increased up to about 400
(xm for larvae 10 mm long. In another study
(Stoecker and Govoni 1984) maximum prey size
was calculated in a different manner; the average
width of the largest prey type was used. With this
index the estimated maximum width of prey
items was 50 ^JLm for larvae <5 mm long and
about 140 |jLm for larvae 9-11 mm long. In our
study, most of the nauplii were between 40 and 80
|jLm body width, and most of the copepodites were
between 50 and 200 p-m body width. These cope-
pod developmental stages were in the size range
of diet items typically found in guts of intermedi-
ate size (7-11 mm) larvae of the gulf menhaden
and we believe our samples are reasonably repre-
sentative of the prey concentrations available to
these larvae. Other prey items for larvae in this
size range were common only on rare occasions.
The importance of high prey concentrations to
successful feeding and survival offish larvae has
frequently been noted (Hunter 1981). Concentra-
tions required to give high survival in laboratory
experiments are seldom found in the ocean but
careful attention to culture techniques can result
in reasonably high survival at prey densities that
are close to or overlap the maximum natural con-
centrations (Houde 1978). At stations near the
river delta that are strongly affected by the river
plumes, maximum concentrations of nauplii were
typically in the range of 20-50 nauplii/L. At sta-
tions farther down plume, away from the delta.
maximum naupliar concentrations were higher,
up to 187 nauplii/L at station IV-2 for example. It
is not clear whether this is a seasonal pattern, a
pattern due to the down-plume development of
the food web, or a pattern attributable to some
other factors. However, the concentrations ob-
served in waters with salinities lowered by the
river plumes are mostly within the range of 10-
100 nauplii/L that is frequently reported for
oceanic and coastal waters for microzooplankton,
or the 50-100 nauplii/L range for coastal and es-
tuarine areas (see Houde 1978; Hunter 1981 for
summaries). The addition of copepodite stages
and other potential prey items to the naupliar
abundances in this study would increase the esti-
mates of available prey somewhat but usually not
more than 20%. At least during the wintertime,
the waters surrounding the Mississippi River
delta do not appear to contain exceptionally high
concentrations of copepod nauplii, compared to
other coastal and estuarine areas.
The vertical distribution of nauplii was fre-
quently similar to that of chlorophyll. Regres-
sions of naupliar concentration on chlorophyll
concentration were significant for 4 of the 5
cruises although not the same in each case. As-
suming chlorophyll is a reasonable indicator of
phytoplankton abundance and thus food abun-
dance for nauplii, then nauplii appear to be ag-
gregated at the depth of highest food availability.
Year to year variability in the relationship indi-
cates that other factors are also important in de-
termining the abundance and distribution of nau-
plii. Because eggs released from adult female
copepods would sink out of the surface water be-
fore hatching, it is probable that active swimming
by nauplii plays a part in the aggregation process,
perhaps enchanced by physical convergence proc-
esses.
In our study we noted that there were interan-
nual differences in naupliar concentrations, there
were sometimes significant regional differences,
and there were sometimes significant onshore-
offshore differences. Although we anticipated
finding higher concentrations of nauplii in the
region near the Mississippi River delta, this pat-
tern was not always observed. This finding might
indicate that there is not an important difference
between the 3 regions as far as larval food
availability is concerned. Alternatively, it might
indicate that there are subregions within the
larger Mississippi River delta region that contain
a food environment for fish larvae that allows for
enhanced survival and growi;h, but that we failed
328
DAGG ET AL.: DISTRIBUTION OF COPEPOD NAUPLII
to clearly identify and segregate these regions.
The overall region did not always appear signifi-
cantly better than the other regions because only
parts of it are better.
On the continental shelf in the Gulf of Mexico,
as in other shelf regions of the United States, the
typical pattern is for the water to be vertically
well-mixed during the wintertime (Parker 1968).
Autumn cooling breaks down the thermal strati-
fication that has existed throughout the summer
and allows the isothermal water column to be
easily mixed. The major exception to this general
pattern in the Gulf of Mexico is the shelf region
influenced by the large volume of freshwater
runoff from the Mississippi River. During this pe-
riod, the freshwater influx is of sufficiently large
volume (average flow for December to March in
1975 through 1979 was 17,290 m^/second (U.S.
Army Engineer District, New Orleans Corps of
Engineers, 1980)) to physically stratify the shelf
waters hundreds of km downstream from the
delta. We suggest that this salinity-induced strat-
ification is a vital component of the recruitment
success of Gulf menhaden because it provides an
environment in which prey aggregations can
occur. It has been suggested that vertical stratifi-
cation by other small-scale physical phenomena
(e.g., Langmuir circulation) allows significant
patchiness of prey items to exist (Lasker 1975).
This patchiness provides small regions of compar-
atively high food concentrations for fish larvae,
and results in improved feeding success. In this
study, copepod nauplii were aggregated vertically
at stations with physical stratification and were
nearly homogenously distributed at stations lack-
ing physical stratification. Maximum naupliar
concentrations (no. per m^) at stratified stations
were typically 2-10 times greater than at non-
stratified stations.
In conclusion, we believe that the large fresh-
water inflow of the Mississippi River during the
wintertime spawning period of the Gulf men-
haden contributes to the feeding success and
survival of larval fish by providing physical strat-
ification which in turn results in a vertical strat-
ification of phytoplankton and microzooplankton
in layers or patches of relatively high concentra-
tions.
ACKNOWLEDGMENTS
Part of this work was done in cooperation with
personnel from the National Marine Fisheries
Service at Beaufort, NC and at Pascagoula, MS.
In particular, we would like to thank D. Hoss for
providing time and space in his field programs,
and J. Govoni for discussion of several portions
of the manuscript. S. Cummings, L. Hill,
E. Walser, J. White, J. Turner, P. Morgan,
C. Neill, and E. Turner assisted in collecting and
analyzing samples. Part of this work was done
in cooperation with D. Checkley and A. Amos at
the University of Texas; their contributions are
greatly appreciated. Thanks also to the captains
and crews of the F.R.V. Oregon II, the RV
Longhorn, and the R. J. Russell. This work was
supported by the Louisiana Universities Marine
Consortium, by NSF Grant OCE-81 19848, and by
the Ocean Assessment Division, National Ocean
Services, NOAA.
LITERATURE CITED
Arthur, D. K
1976. Food and feeding of larvae of three fishes occurring
in the California Current, Sardinops sagas, Engraulis
mordax, and Trachurus symmetricus. Fish. Bull., U.S.
74:517-530.
1977. Distributions, size, and abundance of microcopte-
pods in the California Current system and their possible
influence on survival of marine teleost larvae. Fish.
Bull., U.S. 75:601-611.
Checkley, D M, Jr.
1982. Selective feeding by Atlantic herring (Clupea
harengus ) larvae on zooplankton in natural assem-
blages. Mar. Ecol. Prog. Ser. 9:245-253.
DUKA, I A , AND A. D GORDINA.
1973. Abundance of ichthyoplankton and feeding of fish
larvae in the western Mediterranean and adjacent areas
of the Atlantic Ocean. Hydrobiol. J. 9:54-59.
Environmental Protection Agency.
1979. Methods of chemical analysis of water and wastes.
Nitrogen, nitrate-nitrate. U.S. Environ. Prot. Agency,
Cincinnati, OH 45268.
Fore, P L
1970. Oceanic distribution of eggs and larvae of the Gulf
menhaden. Rep. Bur. Comm. Fish. Biol. Lab., Beaufort,
NC, Fish. Wildl. Serv. Circ. 341.
Gamble, J C , P Maclachlan, N T Nicholl. and I. G. Baxter.
1981. Growth and feeding of Atlantic herring larvae
reared in large plastic enclosures. Rapp. P. -v. R6un.
Cons. int. Explor. Mer 178:121-134.
Govoni, J J , A J Chester. D E Hoss, and P B Ortner
1985. An observation of episodic feeding and growth of
larvae Leiostomus xanthrus in the northern Gulf of Mex-
ico. J. Plankton Res. 7:137-146.
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,
Leiostomus xanthrus, and Micropogonias undula-
tus. Mar. Ecol. Prog. Ser. 13:189-199.
Houde, E D
1978. Critical food concentrations for larvae of three spe-
cies of subtropical marine fishes. Bull. Mar. Sci.
28:395-411.
329
FISHKRY BULLKTIN: VOL. 86, NO 2
HOUDE. E D . AND J. A LOVDAL.
1984. Seasonality of occurrence, foods and food prefer-
ences of ichthyoplankton in Biscayne Bay, Florida.
Estuarine Coastal Shelf Sci. 18:403-419.
Hunter. J R
1981. Feeding ecology and predation of marine fish lar-
vae. In R. Lasker (editor). Marine fish larvae.
Morphology, ecology, and relation to fisheries, p. 34-
77. Univ. Wash. Press, Seattle.
Laurence, G C
1974. Growth and survival of haddock (Melanogrammus
aeglefinus) larvae in relation to planktonic prey concen-
tration. J. Fish, Res. Board Can. 31:1414-1419.
Lasker, R.
1975. Field criteria for survival of anchovy larvae: The
relation between inshore chlorophyll maximum layers
and successful first feeding. Fish. Bull., U.S. 73:453-
462.
1978. The relation between oceanographic conditions and
larval anchovy food in the California Current: Identifica-
tion of factors contributing to recruitment failure.
Rapp. P.-v. Reun. Cons. int. Explor. Mer 173:212-230.
Lewis, R. M., and C M Rothmayr
1981. Spawning and sexual maturity of gulf menhaden,
Brevoortia patronus. Fish. Bull., U.S. 78:947-951.
Parker. C A
1968, The effect of a cold air outbreak on the continental
shelf waters of the northwestern Gulf of Mexico. Tech.
Rep. 68-3T, 89 p. Off. Nav. Res.
Stoecker. D K , and J J Govoni.
1984. Food selection by young larval gulf menhaden (Bre-
voortia patronus). Mar. Biol. 80:299-306.
Strickland, J D H , and T R. Parsons.
1968. A practical handbook of seawater analysis. Bull.
Fish. Res. Board Can. 167, p. 201-203.
US Department of Commerce
1983. Fisheries Statistics for the United States, 1982.
US Army Engineer District, New Orleans Corps of Engi-
neers.
1980, Stages and discharges of the Mississippi River and
its tributaries and other watersheds in the New Orleans
District, 1975-1979.
Warlen, S. M.
1988. Age and growth of larval menhaden, Brevoortia pa-
tronus, in the northern Gulf of Mexico. Fish. Bull., U.S.
86:77-90.
WiEBE, P H , K. H Burt, S H Boyd, and A W Morton.
1976. A multiple opening/closing net and environmental
sensing system for sampling zooplankton. J. Mar. Res.
34:313-326.
330
THE EFFECTS OF SILTATION ON RECRUITMENT OF SPINY LOBSTERS,
PANULIRUS ARGUS
William F Herrnkind.' Mark J. Butler IV,' and
Richard A. Tankersley^
ABSTRACT
Several surveys in the Florida Keys indicated fewer juvenile spiny lobsters, Panulirus argus, in an
area where their primary habitat, stands of benthic algae Laurencia spp., was heavily silted as
compared with similar, less silted habitat. We tested several hypotheses explaining this relationship:
1) planktonic postlarval lobster abundances are lower in the silted area, 2) siltation of algae impedes
postlarval settlement or subsequent juvenile habitat selection, or 3) siltation increases mortality at
the time of metamorphosis. We also compared the time-to-metamorphosis for settling pueruli within
silted and nonsilted algae, analyzed the physical character of algal silt in low-silt and high-silt
regions and measured the abundances of epifauna constituting prey of juvenile spiny lobsters. Plank-
tonic postlarval abundances were substantially higher in the high-silt area thus rejecting hypothe-
sis 1. Likewise, results from laboratory experiments testing the effect of algal siltation on postlarval
time-to-metamorphosis and early postsettlement survival showed no short-term increase in mortal-
ity. Limited postlarval settlement and avoidance of silted algal habitats by juveniles, as determined
in substrate choice experiments, probably accounts for the paucity of young spiny lobsters in heavily
silted localities. In addition, although juvenile spiny lobsters are nonselective predators, lower prey
availability in silted algae probably promotes transciency which, in turn, causes increased mortality
by predation while juveniles are exposed. Large-scale siltation exacerbated by human activity must
be viewed as potentially deleterious to spiny lobster recruitment.
The western Atlantic or Florida spiny lobster,
Panulirus argus, is the focus of an intense com-
mercial and recreational fishery in south Florida,
particularly the Florida Keys. Besides severe
fishing pressure, spiny lobster populations are
subject to a variety of other factors that poten-
tially limit population size. For example, habitat
degradation, like that resulting from chronic sil-
tation, may affect not only adult lobsters but the
postlarval settlement stage as well. During 1983
and 1984 we sampled numerous sites in a region
of about 40 km^ east of Big Pine Key which was
chronically heavily silted and held low numbers
of newly settled spiny lobsters despite extensive
benthic algal growth typical of settlement habi-
tat. We hypothesized that postlarval spiny lob-
sters either do not settle in silted habitat or settle
there but do not survive. In either case, we sup-
posed that the heavy siltation reduced the carry-
ing capacity of otherwise suitable habitat, poten-
tially reducing regional recruitment where
siltation is widespread.
•Department of Biological Science, The Florida State Univer-
sity, Tallahassee, FL 32306-3050.
^Department of Biological Science, The Florida State Univer-
sity, Tallahassee, FL; present address: Department of Biology,
Wake Forest University, Winston-Salem, NC 27109.
Sediment particle size, composition, and stabil-
ity influence larval settlement in a variety of
marine benthic invertebrates (Crisp 1974, 1976;
Gray 1974; Rhoads 1974; Pearson and Rosenberg
1978). For example, heavy siltation and sediment
instability, created by natural biogenic rework-
ing of the substrate and subsequent resuspension
of sediments by turbulence reduces the abun-
dance of suspension feeding infauna (Rhoads and
Young 1971; Aller and Dodge 1974). In addition,
siltation from human activities (e.g., dredging,
shoreline development, boat traffic, etc.) can de-
grade benthic community structure via anaero-
biosis, direct burial, toxic poisoning, or increased
turbidity (Morton 1977; Allen and Hardy 1980;
Jones and Candy 1981; Cortes and Risk 1985).
Most available information concerns sessile or in-
faunal species, but little information exists for
mobile, epibenthic forms (Pearson and Rosenberg
1978). Although the habitat selection and bur-
rowing behavior of some shrimps and juvenile
clawed-lobsters has been investigated in relation
to substrate character (Ruello 1973; Howard and
Bennett 1979; Aziz and Greenwood 1982; Botero
and Atema 1982; Pottle and Elner 1982; Roach
1983; Herrnkind and Butler 1986), we know of no
research describing the effect of siltation on deca-
Manuscnpt accepted February 1988.
FISHERY BULLETIN: VOL. 86, NO 2, 1988.
331
FISHERY BULLKTIN: VOL. 86, NO. 2
pod recruitment. Our research focuses on the im-
pact of siltation on spiny lobster postlarvae and
early benthic juveniles, stages that are morpho-
logically and behaviorally distinct from adults.
Late stage P. argus phyllosome larvae drift in
the oceanic plankton for 6-9 months after hatch-
ing and metamorphose offshore into nonfeeding
pueruli (postlarvae) that swim inshore and settle
in benthic vegetation (Marx 1986). Nev^ly settled
pueruli metamorphose into cryptically colored
benthic juvenile instars after about one week.
Pueruli preferentially settle in highly architec-
tured benthic algal assemblages where subse-
quent survival and growth depend upon available
prey and physical refuge from predators
(Herrnkind and Butler 1986). Ubiquitous, widely
distributed stands of bushy red algae, Laurencia
spp., provide these essential conditions and prob-
ably serve as the most important regional settle-
ment and nursery habitat for juvenile spiny lob-
sters (Marx and Herrnkind 1985a, b; Herrnkind
and Butler 1986; Marx 1986). The early instars
remain within the algae for several months until
attaining about 20 mm carapace length (CL)
(Andree 1981; Marx and Herrnkind 1985a) when
they begin to occupy crevices in rubble or under
sponges, coral, and exposed seagrass rhizome
mats. In Florida, postlarval settlement is year-
round with vernal, autumnal, newmoon, and oc-
casionally aperiodic peaks (Little 1977; Little and
Milano 1980; Marx 1986). The spatial pattern of
settlement is poorly known although new recruits
are widely dispersed within algal habitats; diver
surveys have yielded estimates of one juvenile per
36 m^ of profuse algal growth (Marx and Herrn-
kind 1985a). Yet because postlarvae settle con-
tinuously and juveniles grow rapidly, a single
hectare of the above habitat is estimated to nur-
ture about 1,000 spiny lobsters annually (Marx
and Herrnkind 1985a; Marx 1986). There is no
compelling evidence suggesting that benthic
stage lobsters immigrate into Florida waters from
other Caribbean areas, although their planktonic
larvae presumably do so (Lyons 1980; Marx
1986). Recruitment is thus primarily limited to
postlarval influx. Therefore, precise knowledge of
the factors influencing postlarval recruitment
and recruit mortality is essential to managing the
intensive Florida spiny lobster fishery.
Here we report on studies undertaken to inves-
tigate the impact of the observed algal siltation
on spiny lobster recruitment. We compared field
abundances of both pueruli and early juveniles in
a representative silted and unsilted area, exam-
ined the relationship between siltation and avail-
able epifaunal prey, determined the impact of silt
load on puerulus survival from settlement
through metamorphosis, and tested the prefer-
ence of settling pueruli and algal-dwelling juve-
niles for silted and unsilted algae.
METHODS
Postlarval-Juvenile Abundance in
Silted and Unsilted Habitats
During June through August 1985, we com-
pared the natural abundance of newly settled ju-
venile spiny lobsters (6-20 mm CL) in previously
sampled, chronically silted and unsilted areas.
The silted site (No Name Key) was located ap-
proximately 30 m off the western shore of No
Name Key (Monroe County, FL, U.S.A.) and the
unsilted site (Burnt Point) 30 m off the northwest
shore of Grassy Key at Burnt Point (Fig. 1). The
benthic habitat at both sites was similar and
characterized by nearly contiguous stands of
algae (Laurencia spp.) at depths of 1.5-3.0 m. In-
tensive visual search in algal clumps by divers
was used to estimate the relative number of ben-
thic juveniles. Because newly settled spiny lob-
sters are almost exclusively found associated with
algal clumps (Marx and Herrnkind 1985a), catch
per unit effort (CPUE) as search time within
algae, gives a more suitable estimate for our pur-
poses of comparing abundance than density/area
per se. Modified Witham-type postlarval collec-
tors (Witham et al. 1964, 1968; Little and Milano
1980; Marx and Herrnkind 1985a) were used to
compare postlarval abundance among sites.
Twelve collectors were initially deployed at both
sites and visited approximately every 2 weeks for
3 months. Collector results are reported in CPUE
to standardize catch records biased by the loss of
collectors and different sampling durations.
Algal Silt Content and Prey Content
To determine the amount of silt and macro-
fauna contained in algal clumps at the silted and
unsilted sites, we bagged individual clumps (ap-
proximately 25 cm diameter) of Laurencia in the
field for subsequent laboratory processing. Care
was taken to ensure that loose silt present on the
surface of the algae was not disturbed during col-
lection. Ten clumps, ranging in displacement vol-
ume from 55 to 300 mL, were collected at each
site. Algal samples were rinsed through a series
332
HERRNKIND ET AL RECRUITMENT OK Sl'lNY LOBSTERS
Florida Keys ^
-r«*.
^B''
LITTLE PINE
KEY
/
NAME
KEY
r BAHIA
' HONDA
KEY
^NEWFOUND
HARBOR KEYS
BURNT POINT
GRASSY
KEY
Figure l. — Map of field sites in the middle and lower Keys, Monroe County, FL, U.S.A. Insets provide more detail of the areas
surrounding our field sites (•). Diagonal lines depict approximate extent of the heavily silted area we surveyed.
of sieves (500 |x, 250 |jl, and 63 jx; U.S. Standard
Sieve Series), but only the two smallest size frac-
tions were retained because subsamples >500 |x
consisted entirely of shell and algae fragments.
Silt samples were dried for 48 hours at lOO'C and
then weighed. The amount of algal-entrained silt
at the two sites was compared using a two-sample
^-test. Organic weight of the silt was derived by
digesting three silt samples in SO^f hydrogen per-
oxide for 1 week, then oven drying the remaining
silt at 60°C for 48 hours (Cortes and Risk 1985).
The fraction of carbonates in the silts was deter-
mined by dissolving the three samples in 5% hy-
drochloric acid for 1 week, then drying the sam-
ples as above (Cortes and Risk 1985). Percent
organics and carbonates (by weight) in the silts at
the two sites were compared in two-sample ^-tests
on arcsin transformed data.
We counted the number of epifaunal prey in
silted and unsilted clumps to determine the possi-
ble influence of siltation on juvenile spiny lobster
food abundance. The reported estimates of prey
abundance are means of two separate counts per
clump; 5 clumps per treatment were processed.
The volume of each Laurencia clump was deter-
mined by water displacement, and all silt load
333
FISHERY BULLETIN; VOL. 86, NO. 2
and prey abundance estimates standardized by
clump volume. Prey abundance data were ana-
lyzed using a two-way fixed-effects ANOVA on
log transformed data and Bonferroni pairwise
multiple comparisons.
Habitat Selection/Settling Experiments
We tested postlarval settlement and juvenile
habitat selection in laboratory experiments using
clumps of Laurencia spp. with high- and low-silt
loads (referred to hereafter as silted and un-
silted); the null hypothesis being equal selection
of both habitats. Experiments were conducted in
fourteen 75.7 L aquaria with subgravel filters
and circulating current of 3 cm s ^ Light was
provided by skylights and fluorescent lights with
a photoperiod of approximately 14L:10D. Two 20
cm diameter algal clumps, one silted and one un-
silted, were situated 25 cm apart at opposite ends
of each aquarium and at least 5 cm from aquar-
ium walls. The number of natural prey in both
silted and unsilted clumps far exceeded the num-
ber eaten daily by a juvenile. To further control
food availability in experiments with juveniles
we added equal amounts (10 mg) of Tetramin^
fish food to each clump, providing an overabun-
dance of food available ad libitum. If juveniles
chose one type of algal clump over the other, then
their selection was most likely based on the pres-
ence or absence of silt, because food abundance
and quality were similar, if not strictly identical,
in both types of algal clumps. Pueruli neither feed
nor respond to the differential abundance of po-
tential prey (Herrnkind and Butler 1986). Silted
algae was collected from the No Name Key site
(see section on Algal Silt Content and Prey Con-
tent); unsilted algae was collected just offshore of
the Sea World Marine Science and Conservation
Center on Long Key. Fresh algal clumps were
used in each experimental replicate. An experi-
ment was initiated by introducing a single
puerulus or juvenile spiny lobster to the center of
an aquarium through a 5 cm diameter PVC pipe.
Once a spiny lobster settled to the substrate, the
pipe was slowly withdrawn allowing the lobster
to move freely about the aquarium. This tech-
nique prevented "tailflipping" by lobsters and fa-
cilitated active selection of habitats. Twenty-four
hours later we located the lobsters and recorded
their positions, as in previous experiments
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
(Herrnkind and Butler 1986). Fourteen spiny lob-
sters were tested on day 1, 14 more on day 2, and
so on until our stock of animals was depleted.
Each lobster was used only once. All pueruli were
collected on the incoming tide from the plankton
in interisland channels. Pueruli were either
tested immediately or allowed to metamorphose
for later use in experiments requiring juveniles.
Data were analyzed with log-linear Goodness-of-
fit tests.
Metamorphosis Experiment
The effect of siltation on the survival and time-
to-metamorphosis of pueruli was tested experi-
mentally in an outdoor, flow-through seawater
system. One freshly collected puerulus was placed
in each of 46 seawater-filled 1 L plastic beakers,
23 containing 5 cm diameter clumps of unsilted
algae and the other 23 an equal amount of silted
algae. Each container was independently sup-
plied with flowing, filtered seawater. Algal
clumps were replaced daily. Pueruli do not feed
and their habitat selection operates independent
of food availability (Herrnkind and Butler 1986),
thus no food was added to the containers. Sea-
water temperature in the beakers remained be-
tween 26° and 28°C; photoperiod was approxi-
mately 14L:10D. Pueruli were monitored daily
and their survival and time-to-metamorphosis
recorded.
Pueruli were collected as transparent postlar-
vae from the plankton in interisland channels
which concentrate oceanic postlarvae as they
move into Florida Bay nursery areas. Time-to-
metamorphosis values represent the elapsed time
(in days) from puerulus collection until metamor-
phosis into the first benthic stage. Values are
likely to differ among collections as different co-
horts of pueruli arrive inshore. There are cur-
rently no techniques available to determine the
actual age of pueruli (i.e., time since metamor-
phosis from the phyllosoma stage), but estimates
of duration of the puerulus stage range from
2 weeks to 1 month (Lyons 1980; Calinski and
Lyons 1983). Differences in time-to-
metamorphosis between the two treatments were
analyzed via a two-sample ^-test.
Juvenile Spiny Lobster Prey Selection
Experiments
Laboratory experiments were conducted to de-
termine juvenile spiny lobster prey preference
334
HERRNKIND ET AL.: RECRUITMENT OF SPINY LOBSTERS
and rate of consumption of algal epifauna. Prey
were obtained by rinsing large clumps oiLauren-
cia through a 100 |jl sieve. Prey included small
gastropods, amphipods, isopods, and ostracods
ranging in size from 1 to 9 mm (Marx and Herrn-
kind 1985a). Prey were individually counted and
placed in 1 L plastic beakers containing 750 mL of
seawater. One starved (24-h) juvenile lobster (6-
8 mm CL) was introduced to each container, al-
lowed to feed for 12 hours, and was then removed;
the remaining prey were counted. Fifteen spiny
lobsters were tested in each experiment; each lob-
ster was used once. Three experiments were con-
ducted using different prey combinations (Table
1), but the total number of prey available re-
mained similar and exceeded the amount a single
lobster could consume in 12 hours. Electivity in-
dices calculated for each experiment were used in
multiple comparison tests to determine whether
juvenile spiny lobsters fed preferentially or ran-
domly (Johnson 1980).
Table 1 . — Relative prey availabilities (A) and predator usage (U)
values (percentages) in the three juvenile spiny lobster prey selec-
tion experiments. N = ^5 lobsters per experiment. F-values calcu-
lated from Johnson (1980) indicate whether prey choice differed
significantly from random in each trial: none of the tests were signif-
icant at P = 0.05.
Experiment
1
2
3
Prey type
A
U
A U
A
U
Gastropoda
Tncolia sp.
Batlillaria spp.
Tegula spp.
Atys spp.
Amphipoda
Isopoda
Copepoda
Ostracoda
Decapoda
F-values
df
43.5 615 714 83.0 58.0 54.5
8.7 10 4 3
6.5 00 —
— — 14.3
32.6 33.9 10.0
0.2
7.2
9.6
8.3
0.46
4,11
36
0.40
3,14
7.0
17.4
11.6
6.0
0.09
4,11
7.1
20.0
13.1
5.3
RESULTS
Postlarvae were more abundant on collectors at
the silted site than at the unsilted site. Diver
surveys revealed that higher numbers of algal-
dwelling juveniles (<20 mm CL) resided at the
unsilted site despite greater influx of pueruli into
the silted area (Table 2). Only one juvenile spiny
lobster collected at the silted site was <20 mm
CL, most were considerably larger (25—35 mm
CL) than those at the unsilted site, and some were
possibly large enough to have immigrated there
from adjacent unsilted areas.
Silt recovered from the algal clumps was
largely calcareous and formed a cohesive cast
around the algal filaments. Classifying algal-
bound silts by particle size would yield irrelevant
values because sieving caused fragmentation of
aggregated particles. Therefore we report only
the total dry weight of the silts. Algae at No
Name Key carried a higher silt load than algae at
Burnt Point ix = 125 vs. 65 g/L algae, respec-
tively; t = 2.90, df = 18, P <0.01). Silt at No
Name Key was characterized as 12.5 ± 7.4% or-
ganic and 28.1 ± 27.6% carbonate by weight,
whereas Burnt Point silt was 16.9 ± 1.06% or-
ganic and 66.3 ± 14.5% carbonate (means ± 1 SD).
There was no significant difference in silt compo-
sition between sites (organics: t =0.83, df=4,
P > 0.05; carbonates: t = 1.97, df = 4, P > 0.05),
although sample sizes at each site were small and
sample variance substantial. Algae at both sites
contained primarily gastropods, amphipods, and
isopods, although significantly more gastropods
and echinoderms occupied unsilted clumps
(Fig. 2, Table 3; P < 0.05 in Bonferroni pairwise
multiple comparisons).
Significantly more pueruli settled in unsilted
algal clumps than in silted clumps during labora-
tory settlement choice experiments (38 vs. 11, re-
spectively; G = 15.72, P < 0.001). Juvenile spiny
lobsters responded similarly in the habitat selec-
tion experiment (54 vs. 24; G = 11.78, P < 0.001).
We excluded the open sand habitat in aquaria
from our analysis because 1) juvenile spiny lob-
sters are never found residing on open sand in the
field, presumably due to a lack of food and refuge
Table 2. — (A) Postlarval lobster abundances at silted and unsilted
Florida Bay study areas. Postlarval catch per unit effort (CPUE)
was estimated from Witham collector catches. (B) Juvenile lob-
ster (8-20 mm CL) abundances at the two study sites in 1985 and
adjacent areas sampled dunng 1983 and 1984. Juvenile CPUE
was estimated via diver surveys.
A. POSTLARVAL ABUNDANCE
CPUE
1
Location Description
June July
August
No Name Key silted
Burnt Point unsilted
0.08 0.32
0.01 0.05
0.15
0.10
B. JUVENILE ABUNDANCES
Location Description
Diving
hours
CPUE2
No Name Key (1983-85) silted
Burnt Point (1983-85) unsilted
28
20
0.04
0.50
ippiip - no ot postlarvae no. o( collectors
no. of days between collections
2CPUE
no. of lobsters collected
no. of diver hours
335
FISHERY BULLETIN: VOL. 86, NO. 2
<
o
z
LU
CC
Z)
<
_l
u.
O
<
9
>
CC
Hi
m
1200 -
1000 -
800 -
600 -
- 400 -
200 -
GASTRO* OSTRA
AMPHI
ISO
DECA ECHINO*
Figure 2. — Abundances (x ± 1 SE) of the six most common prey found in algal clumps at
the silted No Name Key and unsilted Burnt Point sites. Five clumps were collected at each
site. Values are standardized by clump volume. Asterisks denote significant difference in
Bonferroni multiple comparison tests (P = 0.051. Abbreviation key: GASTROpod, OSTRAcod,
AMPHlpod, isopod, DECApod, ECHlNOderm.
Table 3. — Two-way fixed-effects ANOVA testing for differences in
the total number of individuals among six prey categones (see
Table 1) at two sites, one silted (No Name Key) and one unsilted
(Burnt Point). Data were log transformed.
Source
df
SS
Site 1 0.841 18.78 0.001
Prey type 5 6 242 27.89 0.001
Site X prey type 5 0.847 3.78 0.006
Error 48 2.149
(Marx and Herrnkind 1985b; Herrnkind and But-
ler 1986), 2) most spiny lobsters recovered from
open sand were actually in corners, indicating
edge-seeking behavior rather than selection for
sand per se, and 3) only 10.6% of 142 spiny lob-
sters tested were found on sand even though it
constituted 68% of the exposed substrate in
aquaria. Silt had no effect on puerulus survival
through metamorphosis to the first benthic instar
(13% vs. 9% mortality in silted and unsilted
algae, respectively), or time-to-metamorphosis
(Fig. 3; t = 0.37, P > 0.05).
The total number of prey items consumed in the
juvenile prey selection experiments ranged from
19 to 57 prey per lobster per 12 hours. Juveniles
fed randomly from the three different prey combi-
nations and frequencies offered to them (Table 1).
O
z
C/) 20
O
I
Ql
CC
o
10 "
LU
y^
— O O
' //^
SILTED
" U
%
. // UNSILTED
•
0 1 2 3 4 5 6
DAYS
FlGt_iRE 3. — Cumulative number of spiny lobster postlarvae
metamorphosing as a function of time in one of two treatments:
silted algae (Laurencia spp.) or unsilted algae. Twenty-three
postlarvae were tested in each treatment; two postlarvae died in
the unsilted treatment and three in the silted treatment.
DISCUSSION
The relative paucity of newly settled spiny lob-
sters in the heavily silted region around No Name
Key over a 3-yr period indicates that low recruit-
ment to benthic habitat is typical there. The ab-
sence of juveniles was apparently not due to a
336
HERRNKIND ET AL : RECRUITMENT OF SPINY LOBSTERS
lack of postlarval influx, which was higher than
that at the unsilted Burnt Point site, but instead
to low rates of postlarval settlement. Results from
our habitat selection experiments support this
hypothesis because settlement was significantly
lower in silted algal clumps than in unsilted
clumps. Previous studies showed that postlarvae
selectively settled in highly architectured materi-
als, like algal clumps (Herrnkind and Butler
1986). Thus, heavy silt covering an otherwise pre-
ferred habitat either masks the stimuli triggering
settlement or contains stimuli that elicit rejection
by pueruli. This question remains for further
study. We cannot conclusively ascertain from our
laboratory experiments the mechanisms govern-
ing habitat choice in the field where silted and
unsilted habitats may not be adjacent, as they
were in our aquaria. Yet for many species with
planktonic larvae, these kinds of experiments,
coupled with field observations of more general
patterns of behavior and abundance, provide
valuable insights into natural processes (Sulkin
1986).
Twenty percent of the pueruli we tested in lab-
oratory tanks settled in silted clumps despite the
general rejection of this habitat. Pueruli settling
in silted algae probably metamorphose normally
into the first benthic instar, as indicated by the
equivalent time-to-metamorphosis and early post-
settlement survival in both silt levels tested.
However, subsequent residency by juveniles pre-
sumably is limited, given their great mobility and
preference for unsilted algae with high food levels
(Marx and Herrnkind 1985b; Herrnkind and But-
ler 1986). We found that prey abundances were
significantly lower in silted algae. Thus, even if
pueruli settle in the silted habitat the subsequent
juveniles may leave to obtain adequate food. Fre-
quent interclump movement by juvenile spiny
lobsters, searching either for food or unsilted
habitat, would predictably result in increased
predatory mortality. Susceptibility to predation
is much greater for juveniles in the open, than it
is for individuals amidst algal clumps or dense
seagrass (Herrnkind and Butler 1986). Thus, ju-
venile residency patterns and susceptibility to
predation may, in addition to locally low settle-
ment, contribute to the paucity of lobsters in the
silted habitat.
The algal-bound silt load at No Name Key was
roughly twice that at Burnt Point where spiny
lobster recruitment was considerable. Our cur-
sory surveys from Key Largo to Boca Chica Key
indicate that similar silt levels are common.
though geographically variable in Florida Bay.
Benthic algae, including La//re/icm spp., serve as
sediment traps (Scoffin 1970) and demonstrate a
remarkable resistance to siltation, growing pro-
fusely even in heavily silted areas. Silt in these
areas is primarily calcareous, most of it probably
a byproduct of sediment processing by deposit
feeding shrimp (particularly Callianassa), an-
nelids, and sea cucumbers.
We did not evaluate the geographic extent of
siltation relative to spiny lobster settlement in
Florida Bay. However, the demonstrated aver-
sion to settling in naturally silted algae, charac-
teristic of the region around No Name Key,
strongly suggests that low postlarval recruitment
and juvenile abundances would occur in similar
conditions elsewhere. The sparse juvenile popula-
tion at our silted site, one-tenth that of the un-
silted site, suggests deleterious impact of high
chronic silt levels in areas of potential recruit-
ment. Human activities also cause siltation (Mor-
ton 1977; Allen and Hardy 1980). We noted that
algal stands adjacent to heavily trafficked boat
channels typically were more heavily silted than
adjacent areas. We suspect the effect of manmade
siltation to be similar to that from natural causes.
Although it is now generally accepted that Flor-
ida Bay shallows serve as the main nursery
grounds for the south Florida spiny lobster popu-
lation (Marx 1986), the regional distribution of
settlement and early juvenile habitation remains
to be mapped. Future wide-area surveys by con-
cerned researchers and agencies should include
sampling of new spiny lobster recruits as well as
silt levels. Meanwhile, sizable human activities
such as channel construction, dredging, spoil
dumping, coastal development, and mineral min-
ing must be viewed as potentially deleterious to
spiny lobster recruitment.
ACKNOWLEDGMENTS
We thank the Sea World Center for Marine Sci-
ence and Conservation on Long Key, FL, for logis-
tical support and the use of their facilities. John
Hunt and Jim Marx of the Florida Department of
Natural Resources Marine Research Laboratory
in Marathon provided field assistance and helpful
advice throughout the project. Comments by
P. Greenwood, D. Wilber, and two anonymous re-
viewers substantially improved the manuscript.
This research was supported by a Sea Grant
award (R/LR-B-16) to W. F. Herrnkind. Addi-
tional support was provided M. J. Butler via
337
FISHERY BULLETIN: VOL. 86, NO. 2
Marine Science Fellowships from the Aylesworth
Foundation and the International Women's Fish-
ing Association (IWFA).
LITERATURE CITED
Allen, K D . and J W Hardy
1980. Impacts of navigational dredging on fish and
wildlife: a literature review. U.S. Fish Wild. Serv., Biol.
Serv. Program FWS/OBS-80/07, 81 p.
Aller. R C , AND R E Dodge
1974. Animal-sediment relations in a tropical lagoon,
Discovery Bay, Jamaica. J. Mar. Res. 32:209-232.
Andree. S W
1981. Locomotory activity patterns and food items of ben-
thic postlarval spiny lobsters, Panulirus argus. MS
Thesis, Florida State University, Tallahassee, FL, 50 p.
Aziz. K A . and J G Greenwood
1982. Response of juvenile Metapenaeus bennettae Racek
& Doll 1975 (Decapoda: Penaeidae) to sediments of differ-
ing particle size. Crustaceana 43:121-126.
BOTERO, L , AND J ATEMA
1982. Behavior and substrate selection during larval set-
tling in the lobster Homarus americanus. J. Crust.
Biol. 2:59-69.
Calinski, M D , AND W G Lyons
1983. Swimming behavior of the puerulus of the spiny
lobster Panulirus argus (Latrielle, 1804) (Crustacea: Pal-
inuridae). J. Crust. Biol. 3:329-335.
Cortes, J N . and M J Risk
1985. A reef under siltation stress: Cahuita, Costa
Rica. Bull. Mar. Sci. 36:339-356.
Crisp, D J
1974. Factors influencing the settlement of marine inver-
tebrate larvae. In P. T. Grant and A. M. Mackie (editor),
Chemoreception in marine organisms, p. 177-265.
Acad. Press, N.Y.
1976. Settlement responses in marine organisms. In
R. C. Newell (editor). Adaptation to environment: essays
on the physiology of marine animals, p. 83-124.
Butterworths, Lend.
Gray, J S
1974. Animal-sediment relationships. Oceanogr. Mar.
Biol. Annu. Rev. 12:223-261.
HERRNKIND, W F , AND M J BUTLER IV.
1986. Factors regulating postlarval settlement and juve-
nile microhabitat use by spiny lobsters Panulirus
argus. Mar. Ecol. Prog. Ser. 34:23-30.
Howard, A E., and D B Bennett
1979. The substrate preference and burrowing behavior of
juvenile lobsters (Homarus gammarus (L.)). J. Nat.
Hist. 13:433-438.
Johnson, D H
1980. The comparison of usage and availability measure-
ments for evaluating resource preference. Ecology
61:65-71.
Jones, G , and S Candy.
1981. Effects of dredging on the macrobenthic infauna of
Botany Bay. Aust. J. Mar. Freshwater Res. 32:379-398.
Little, E J
1977. Observations on recruitment of postlarval spiny
lobsters, Panulirus argus, to the South Florida coast.
Fla. Mar. Res. Publ. No. 29, 35 p.
Little. E J , Jr , and G R Milano
1980. Techniques to monitor recruitment of postlarval
spiny lobsters, Panulirus argus, to the Florida
Keys. Fla. Mar. Res. Publ. No. 37, 16 p.
Lyons, W. G
1980. Possible sources of K'lorida's spiny lobster popula-
tion. Proc. Gulf Caribb. Fish. Inst. 33:253-266.
Marx. J M
1986. Recruitment and settlement of spiny lobster pueruli
in south Florida. Can. J. Fish. Aquat. Sci. 43:2221-
2227.
Marx, J M , and W F Herrnkind
1985a. Macroalgae (Rhodophyta: Laurencia spp.) as a
habitat for young juvenile spiny lobsters, Panulirus
argus. Bull. Mar. Sci. 36:423-431.
1985b. Factors regulating microhabitat use by young ju-
venile spiny lobsters, Panulirus argus: food and shel-
ter. J. Crust. Biol. 5:650-657.
Morton, J. W.
1977. Ecological effects of dredging and dredge spoil dis-
posal: a literature review. U.S. Fish Wild. Serv. Tech.
Pap. No. 94, Wash., DC.
Pearson, T H , and R Rosenberg
1978. Macrobenthic succession in relation to organic en-
richment and pollution of the marine environ-
ment. Oceanogr. Mar. Biol. Annu. Rev. 16:229-311.
Pottle, R A . and R W Elner
1982. Substrate preference behavior of juvenile American
lobsters, Homarus americanus, in gravel and silt-clay
sediments. Can. J. Fish. Aquat. Sci. 39:928-932.
Rhoads, D C
1974. Organism-sediment relations on the muddy
seafloor. Oceanogr. Mar. Biol. Annu. Rev. 12:263-300.
Rhoads, D C , and D K Young
1971. Animal-sediment relations in Cape God Bay, Mas-
sachusetts. II. Reworking by Molpadia oolitica
(Holothuroidea). J. Mar. Biol. 11:266-261.
Roach, S G
1983. Survivorship, growth, and behavior of juvenile lob-
sters Homarus americanus Milne-Edwards in controlled
environments in nature. Nova Scotia Dep. Fish. Mar.
Res. Tech. Rep. Ser. 83-02, 60 p.
RUELLO. N V
1973. Burrowing, feeding, and spatial distribution of the
school prawn Metapenaeus macleayi (Haowell) in the
Hunter region (Australia). J. Exp. Mar. Biol. Ecol.
13:189-206.
SCOFFIN, T p.
1970. The trapping and binding of subtidal carbonate sed-
iments by marine vegetation in Bimini Lagoon, Ba-
hamas. J. Sediment Petrol. 40:249-273.
SULKIN, S D
1986. Application of laboratory studies of larval behavior
to fisheries problems. Can. J. Fish. Aquat. Sci. 11:2184—
2188.
WiTHAM, R , R M iNGLE, AND H W SiMS jR.
1964. Notes on postlarvae o{ Panulirus argus. Q. J. Fla.
Acad. Sci 27:289-297.
WiTHAM, R , R M iNGLE, AND E A JOYCE jR
1968. Physiological and ecological studies of Panulirus
argus from the St. Lucie estuary. FL Board Conserv.
Tech. Ser. 53:1-31.
338
A LIMITED INFORMATION APPROACH FOR
DETERMINING CAPITAL STOCK AND INVESTMENT IN A FISHERY^
James E Kirkley^ and Dale E. Squires^
ABSTRACT
There have been few empirical studies on the level of capitalization and investment in fisheries
because the necessary data are often inadequate. Specifically, data on capital stock and investment
in a fishery are not routinely collected and compiled or are limited in scope. In this study, a method
is provided for estimating the aggregate capital stock and investment in a fishery utilizing the
available information. Data on acquisition and list prices and vessel characteristics for a sample of
New England vessels are obtained. The data are then used to estimate an hedonic cost function which
specifies the acquisition price as a function of vessel characteristics. The resultant equations are
subsequently used, with information on vessel characteristics for all New England vessels, to esti-
mate aggregate capital stock and investment. The results indicate that substantial investment oc-
curred in the otter trawl and scallop dredge fisheries, particularly since the Magnuson Fisheries
Conservation and Management Act. Moreover, the results demonstrate that the number and change
in the number of vessels are inadequate indicators of the level of capital stock and investment in a
fleet comprised of vessels with heterogeneous characteristics.
The common property nature of fisheries is recog-
nized as causing excess capitalization and har-
vesting capacity (Gordon 1954). The theoretical
argument is that since fishermen do not have to
pay for the utilization of common property fish
stocks, new vessels enter a fishery until net rev-
enue is driven to zero. This common-property fea-
ture results in more capital and investment than
is economically optimum. Public regulation of
fishing industries is usually advocated to redress
the excess entry, economic inefficiency, and loss
of economic rents in a common property fishery
(Scott 1979; Sissenwine and Kirkley 1982).
In essence, overcapitalization and excess in-
vestment are perceived to be the reasons for many
of the major fisheries problems (Hilborn 1983).
Economists argue that management of overcapi-
talization is necessary to realize the benefits of
fisheries (Cunningham et al. 1985). Alterna-
tively, gains from fisheries management require
control of overcapitalization and excess invest-
ment (Charles 1983a, b).
Clark et al. (1979) and Charles (1983a, b) ex-
tended the static theory of optimal fisheries in-
vestment as developed by Gordon (1954). They
^Virginia Institute of Marine Science (VIMS) Contribution
No. 1467.
^College of William and Mary, Schools of Marine Science and
Business Administration, Gloucester Point, VA 23062.
3Southwest Fisheries Center La Jolla Laboratory, National
Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 92038.
also demonstrated the relationships between in-
vestment, capital stock, overexploitation, and
fisheries management. They concluded that in-
vestment should be more conservative to prevent
overexploitation.
The economic literature provides substantial
justification for the need to solve the problems of
overcapitalization and excess investment. Yet,
few empirical studies document the level of capi-
talization and investment."* Moreover, there ap-
pears to be no attempt by any U.S. agency to
routinely collect and compile statistics on either
the stock of capital or the level of investment in
U.S. fisheries. Overcapitalization and excess in-
vestment, though, continue to be suggested as the
reasons for many of the economic problems of
fisheries.
Since data are often inadequate, many empiri-
cal studies on fisheries consider capital stock and
investment in terms of the number of vessels.
That is, capital stock and investment in a fishery
are tyically measured in terms of number of ves-
sels. In the absence of appropriate information,
the number of vessels may be the only basis for
examining overcapitalization and excess invest-
ment. Alternatively, if a fleet has identical-sized
vessels, fixed inputs or vessel characteristics, and
Manuscript accepted February 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.
'•Tettey et al. ( 1986) provided an exception, but their analysis
is restricted to the shrimp fishery and is based on cost data
obtained from boat builders.
339
FISHERY BULLETIN: VOL. 86, NO. 2
gear, the number of vessels may be used to indi-
cate the stock of capital and investment. How-
ever, Kendrick (1961) demonstrated that the total
number of operating units, such as the number of
plants, is an inadequate measure of capital stock.
Also, few fleets have identical-sized vessels, char-
acteristics, or gear. As a consequence, the number
of vessels does not provide adequate information
to indicate the level of capitalization and invest-
ment.
This study presents an approach for estimating
the aggregate or industry level of capital stock
and investment in a fishery comprised of hetero-
geneous vessels using available information.
Data available include acquisition and list prices
for some but not all vessels and boat characteris-
tics for all vessels in a fleet. The approach was
developed and used to obtain estimates of capital
stock in support of the United States and Cana-
dian maritime boundary dispute. It is viewed as
an initial step towards examining the problems of
excess capitalization by providing estimates of
aggregate capital stock and investment.
THE DEFINITION AND MEASUREMENT
OF CAPITAL AND INVESTMENT
Capital
The concept of capital has created problems for
economists for quite some time. The term has
been used in so many different contexts that it is
a source of enormous confusion (Hirshleifer
1970). Sloan and Zurcher (1968), in their
"Dictionary of Economics", define capital as "One
of the major factors of production consisting of
property from which an income is derived, ex-
pressed in terms of money. Popularly, the term is
frequently used interchangeably with capital
good. A distinction is sometimes made between
money capital, or that part of the capital held in
the form of money and bank deposits, and prop-
erty capital, or that part of the capital held in the
form of evidences of ownership such as stocks,
bonds, and mortages."
Hirshleifer (1970) presented three meanings of
capital: 1) real capital or capital stock, 2) capital
value, and 3) liquid capital. Real capital is de-
fined as a collection of capital goods or an aggre-
gate of heterogeneous capital inputs. It is one of
the major productive commodities or economic
factors of production. Capital value is the net dis-
counted value of expected future income streams
associated with a capital good. Liquid capital is
the level of current funds available or intended
for investment.
In the case of fisheries, real capital or capital
stock is the form of capital which should be exam-
ined with respect to the problems of overcapital-
ization and excess harvesting capacity. It is the
relevant measure of capital goods used or avail-
able for production (National Academy of Sci-
ences 1979). Moreover, capital stock is the con-
cept of capital used to define and measure the
services of capital inputs which are required to
harvest fish; that is, it is the concept of capital
required to define the economic production tech-
nology of a fishery. Thus, this study is concerned
with the concept of real capital stock. ^
The measurement of capital stock in a fishery,
however, presents several problems. First, capital
inputs are usually quite heterogeneous and can-
not be easily aggregated without restrictive as-
sumptions about the form of the catch equation or
fishermen's behavior. Conceptually, it should be
possible to combine all the different types of capi-
tal goods by weighting each type by its average
compensation (i.e., the rental price). However, in-
formation at this level of detail is typically not
available. Second, fisheries agencies, particularly
in the United States, generally do not collect and
compile information necessary to calculate capi-
tal stock and investment. Third, in order to com-
pare changes in capital stock over time, the meas-
ure of capital stock must be converted to some
base period value by deflation.
In empirical economic studies of traditional in-
dustries, the common practice is to measure the
capital input or stock by converting the purchase
or acquisition price or the book values of capital to
base period values by the use of a price index
(National Academy of Sciences 1979). Varian
(1984) noted: "The usual procedure is to measure
capital value and then deflate by a price index; in
some sense, this should measure the level of cap-
ital stock." In this study, the acquisition price is
used as the measure of the stock of capital or
capital value; the real stock of capital is obtained
by dividing the capital value by the producer
price index for heavy machinery.
Investment
The definition and measurement of investment
is more straightforward than is the definition and
5The market value of capital, acquisition price, and cost of
capital are used interchangeably in this study.
340
KIRKLEY AND SQUIRES: CAPITAL STOCK AND INVESTMENT IN A FISHERY
measurement of capital. In simple terms, invest-
ment is the exchange of money for some form of
property. Although there are several types of in-
vestment (Branson 1972), attention is restricted
in this study to an examination of net investment
(i„) which is defined as the difference between the
stock of capital in two periods of time (Baumol
1977). This is the same concept of investment ex-
amined by Tettey et al. (1986).
Replacement investment is another type of in-
vestment. The sum of replacement investment
and net investment equals total investment. Re-
placement investment is measured by multiply-
ing the rate of depreciation times the level of the
capital stock. Replacement investment, however,
is not estimated in this study since the method
and rate of depreciation must be arbitrarily as-
sumed.
METHOD OF ANALYSIS
The Hedonic Approach
While the definition and measurement of real
capital stock and investment are conceptually
straightforward, limited data and heterogeneous
capital inputs complicate the measurement of
capital stock and investment. In particular, data
are usually only available for a small number of
vessels which are often similar in their character-
istics; these data are inadequate for calculating
actual or observed capital stock and investment
in a fishery. However, data are available on the
acquisition price and characteristics of vessels
which permits estimation of the stock of capital
by an hedonic approach. In turn, estimates of cap-
ital stock from the hedonic approach can be used
to estimate net investment in a fishery.
The hedonic approach hypothesizes that the
price of a commodity is influenced by its charac-
teristics (Rosen 1974). The hedonic price or char-
acteristics function in market equilibrium re-
flects both the distribution of marginal rates of
substitution over households and the distribution
of marginal rates of transformation over firms. In
effect, the hedonic hypothesis states that goods
may be valued for their attributes and that im-
plicit or hedonic prices exist as a function of the
attributes.
Griliches (1971) estimated hedonic cost func-
tions of U.S. automobiles using advertised or list
prices and actual transaction prices. Griliches,
however, was concerned with explaining quality
differentials. Ladd and Martin (1976) examined
prices and demands for input characteristics
using a Neoclassical input characteristics model.
Triplett (1986) estimated quality adjusted price
indexes for computers. Thus, there is a history of
deriving prices of commodities as a function of
their characteristics.
The hedonic approach offers several attractive
properties for estimating capital stock in fishing
industries. First, the hedonic method incorpo-
rates changes in the quality of capital over time
(Triplett 1986). This is because changes in quality
should be reflected in vessel acquisition prices.
The hedonic approach, thus, permits a quality ad-
justed measure of capital stock to be obtained. In
contrast, vessel count ignores changes in the
quality of capital. Second, the hedonic approach
allows easy aggregation of the heterogeneous cap-
ital inputs frequently observed in a fishing fleet
because the heterogeneous capital is measured by
value rather than physical measures.
In the hedonic approach, cost (C) or input price
may be expressed as a function of the associated
characteristics of the commodity or input,
C = f(CHi, CH2, ...,CH„) ,
(1)
where C is cost, CH, is the tth characteristic
(Braeutigam et al. 1982). The characteristic cost
equation can be obtained either from a dual speci-
fication or by determining the reduced form equi-
librium equation.
The Dual and Reduced Form
The dual specifies cost as a function of input
prices and their characteristics. The partial
derivative of the dual cost function with respect to
input prices yields the demand for the inputs as a
function of prices and input characteristics. The
reduced form is obtained by solving structural
equations, equations which explain the behavior
and interrelations of endogeneous variables, of
demand and supply for the endogeneous variable,
price, in terms of exogeneous variables and dis-
turbance terms. ^
In this study, the reduced form is directly speci-
fied and estimated for two reasons: First, all of the
input price data necessary for estimating a dual
are not available. Second, estimation of the struc-
tural equations is complicated by the need to use
a limited dependent variable method of estima-
^See Rosen (1974) for additional information on obtaining
reduced form equations for hedonic prices.
341
FISHERY BULLETIN: VOL. 86, NO. 2
tion since demand is a discrete choice, and there
are individuals or firms which do not purchase a
vessel.
Rather than attempt to deal with these difTi-
culties which do not appear to have widely ac-
cepted solutions, it is assumed that there are
underlying structural equations of demand and
supply which permit the derivation of the reduced
form equation. Moreover, the direct specification
and estimation of the reduced form is the more
commonly used approach to estimate the value of
a commodity as a function of its characteristics
(Bockstael et al. 1986). Last, the use of the re-
duced form should not create a problem since the
hedonic function is a joint envelope in character-
istic space of demander's bid schedules and sellers
offer schedules, and thus, the observed implicit
price locus is a reduced form equilibrium vector
reflecting both supply and demand influences. As
a result, shifts in both market demand and supply
are incorporated into the reduced form equi-
librium vector.
The Capital Stock Model
As previously indicated, it is common practice
to use the acquisition price as a measure of capital
stock. By the hedonic approach, the capital stock
or acquisition price of a fishing vessel is postu-
lated to be a function of several vessel character-
istics. The following general specification is con-
sidered:
C = f(LN, CS, YEAR, FI, AGE, DNO, HPWR, GRT),
(2)
where C is the list or advertised price or the ac-
quisition price of a vessel, LN is length, CS is hull
type (1 = steel, 0 = other), FI is fishery or gear
tj^e,' AGE is age of vessel, DNO is a dummy
variable for new and used vessels (DNO = 1 for
new and 0 for used), HPWR is engine horsepower,
and GRT is gross registered tonnage. The vari-
able year equals the year of the observation. It is
included to incorporate technological changes
'^Gear type is an integer valued variable set equal to the gear
codes used by the Northeast Fisheries Center. This may influ-
ence the results similarly to the use of nonbinary dummy vari-
ables (Kmenta 1971). The use of this variable implies that the
cost difference between gear types is equal to a scalar multiple
of the lowest integer valued gear code. Thus, statistical rejection
of cost differences due to gear type is actually a rejection of
differences not being equal to scaleir multiples. However, the
large number of gear codes prohibits their separate treatment.
over time, such as the introduction of electronics
and stern trawling, and structural changes in the
industry such as changes in organization of capi-
tal and fish markets and changes in public regu-
lation.
Functional Form
Several functional forms could be specified for
Equation (2). Alternatively, generalized Box-Cox
(Box and Cox 1962) transformations could be used
to determine the functional form. However, the
selection of functional form is mostly concerned
with obtaining the relationship between the de-
pendent variable and individual independent
variables. This study is primarily concerned with
estimating the capital stock. Moreover, it may be
quite difficult to estimate capital stock from Box-
Cox transformations or other functional forms
since the conditional expectation is often of a com-
plex form (Smallwood and Blaylock 1986). Also,
Dadkhah (1984) noted that predictions based
upon some transformations and functional forms
result in biased predictions and asymmetric confi-
dence intervals.
Since this study is primarily concerned with
estimating capital stock and investment rather
than estimating the implicit prices for character-
istics, a linear functional form is proposed for
Equation (2). The capital stock equation is speci-
fied as a second-order polynominal in order to
provide an approximation to an unknown, under-
lying hedonic function. This is equivalent to
specifying a generalized quadratic flexible func-
tional form and imposing zero valued restrictions
on the cross product coefficients or interaction
terms (Lau 1978). Flexible functional forms are
widely used in economic studies concerned with
determining the underlying economic structure
because they impose very little structure on the
economic equations of concern (Blackorby et al.
1978).
Postulated Relations
The postulated relations between cost and each
of the explanatory variables is as follows: 1) the
coefficients for the first order terms of vessel ton-
nage, length, and engine horsepower should be
positive since costs should increase as the size
characteristics increase; 2) the coefficient for
steel-hulled vessels should be positive since steel
vessels are generally more expensive than wood
or fiberglass vessels; 3) the dummy variable coef-
342
KIRKLEY AND SQUIRES: CAPITAL STOCK AND INVESTMENT IN A FISHERY
ficient for new vessels should be positive since
new vessels tend to cost more because they are
newer and usually incorporate more recent tech-
nological advances; 4) the second-order coeffi-
cients for the size characteristics are expected to
be negative since cost is believed to increase at a
decreasing rate in response to increases in the
size of a vessel.
THE DATA
Data used to estimate the capital stock equa-
tion were obtained from three sources: First, in-
formation on vessel acquisition price and associ-
ated characteristics were obtained for 164 new
vessels from the Northeast Regional Office of the
National Marine Fisheries Service.^ Second, data
on list prices and vessel characteristics for 946
used vessels were obtained from various trade
magazines and vessel brokers. Third, detailed
vessel and fishery data were obtained from the
U.S. Coast Guard master vessel listing and the
Northeast Fisheries Center. Data were obtained
for the years 1965 through 1981.
The first two sets of data were used to estimate
the capital stock equations. The third data set
was used to estimate aggregate capital stock and
investment in selected New England fisheries
using the estimated capital stock equations; that
is, vessel characteristics for all of the vessels in
the New England fleet were inserted into the esti-
mated capital stock equations to obtain estimates
of capital stocks per vessel in a given year. The
estimated capital stock per vessel was summed
over all vessels in a year to obtain total or aggre-
gate capital stock. Net investment was then cal-
culated as the annual change in total capital
stock.
The use of both the list and acquisition prices,
however, presents a problem. The list price re-
flects the supply price or the price at which a
vessel is offered for sale. The acquisition price
reflects the price determined by the equilibrium
between demand and supply. As a consequence,
estimates of capital stock and investment may be
in error.
It is not known by how much the acquisition
price differs from the list price for used vessels.
Thus, the magnitude of the error cannot be calcu-
lated. Since the list price is not less than the ac-
quisition price, estimates of capital stock based on
Equation (2) are likely upwards biased. Counter
to this problem is the argument that by the time
a vessel is ready for fishing, required capital mod-
ifications or improvements may result in the cap-
ital stock being close in value to the list price.
There are no completely satisfactory economic
justifications for using both the list and acquisi-
tion prices. Two possible justifications are that
Griliches (1971) and others used both prices, and
capital improvements or modifications may be
nearly equal to the difference between the list
and acquisition prices. Another possible justifica-
tion is the accepted use of cost data obtained from
boat builders as in Griffin et al. (1978) which may
impose similar problems plus the statistical prob-
lem of measurement error. In the remainder of
this paper, the term "acquisition price" is used
although the estimation and calculation of capital
stock and investment are based on both price
series.
EMPIRICAL RESULTS
Equation (2) was estimated by ordinary-least-
squares with the dependent variable, the acquisi-
tion price, measured in both nominal and real
terms; the real price is the nominal price deflated
by the producer price index for heavy machinery.
As previously stated, deflation is necessary to es-
timate the real capital stock. The estimated coef-
ficients and statistical results are presented in
Tables 1 and 2.
Equation (2), however, was also estimated
using data for several groups of years between
1965 and 1981. The reason for considering differ-
ent time periods was that prior knowledge of New
England fisheries suggested that various changes
occurred in the fisheries during the selected time
periods.^ Moreover, economic studies of New Eng-
land fisheries have often been criticized for as-
suming stable relationships over time or over dif-
ferent cross sectional units. Failure to incorporate
changes in the estimated relationship between
cost and vessel characteristics over time may re-
sult in biased parameter estimates.
Changes which have possible ramifications for
the estimated relationships include 1) depressed
resource conditions and the presence of foreign
fishing between 1965 and 1971; 2) management
by the International Commission for Northwest
^Data are confidential and may not be available to other re-
searchers or the general public.
^Additional information on possible structural changes in
New England fisheries is available in Kirkley et al. (1982),
Dewar (1983), Doeringer et al. (1986), and Kirkley (1986).
343
FISHERY BULLETIN: VOL. 86. NO. 2
CO
y^
in
Tj-
O)
CO
CM
^
tr
Oi
O)
CO
CO
00
r~.
00
3
d
d
d
d
d
d
d
o
o r;^
•* —
o> ^
CO
r--
CO
t^
^
S
LU
to ^
CM <0
■* CO
en
in
CO
in
CM
O
T-
OO
^
<
CO ^
in in
tv. CO
<D
^ —
00
CM
CO
CM
CM
o
- s
£2.
' 2!.
CD
1
CO
in
1
d
C3>
r^
CM
1
3,
CD U)
> 05
>. >
E ?
CO
o
C:in
cop
5 P
CJ>
00
CM
Op
ay
CO
f
in
CM
CM
^
r^
o
CO
8
e£
Cvj o
CM Tt
CM CO
CO
T—
CO
^—
s
y—
in
•>*'
"g.
CO
in
« aJ
C\J —
T- --^
r^ ^-^
■*
CO
,_.
CO
CO
^
C/D ®
E
s^
o2
g^
in
05
d
op
in
CD
CO
8
CM
•<*
^ w
r- —
. Ol
S.9
CM ■•-
^ C
<^ C
C^J
d
T—
y—
y—
CD
^
1 ^^
1
'"
""^
^~^
"■"'^
§<
C\J ^_^
t^ ^^
in ^__
•<r
CM
(35
CO
!= CO
o
CM 05
CO o
C» CM
CM
T—
00
CO
CO
t-~
d
05
j_ (0 - —
o ■ - w
z
Q
05 00
r- o
CO CO
CO
o>
CJ>
CO
00
t^
in
o
in C.
5£i
O T-
CM
CM
d
5
CM
T-^
o
^
CD
in
CO
o
CM
CM
CM
•<r
» S2
2
o ^^
1^ ^^
in ^^
^
CD
<
S2 s
Q.
U
d CO
CM O
CO h-
00
T—
r-
o
CM O
CM Tt
<o ^
00
OO
1
1
CO
CM
> -g (0
LU
>
1 d
^ei
gS2
00
55
CM
1
1
<5
in
CO
rii s
To
c
CM
CO
1
o
(NJ
2 —
CO ^^
CO ^
CO
in
CM
CM
CO
oT
CM
In
5
CO
c
z
•,- CO
<75 r>J
(O o
5
O
d
■«T
CM
o
d
C\)
<o to
c II a
n y- c
_l
"" si
m ih
1 '*"'
1 "^^
CM
1
^
1
d
CO
1
CM
1
1
c ^^ —
c
o c S2
o
^_
o>
o
.^
(D
o
y—
CM
c o o
CD r: ^
n3
tf)
o> —
■* .~.
C\J ^^
CO
, ^
in
,,_^
q
,_^
T—
, ,
CJ c-
tn 2 '^
3
ST
J3
z
-I
§?
d°
CO '^
g^
o
in
in
O
CD
o
CD
O
in
0)
(0
CM O
CO CO
»- ■*
•*
d
o
CM
CM
CM
o
Cvj
CD 0
o
>
1 "^
CO ^-'
CD "-^
<o
CO
" —
•<t
^ —
o
"S
CVJ
,..
CO
CD ,^
TT
in
Y—
CO
^^
T •*
^ in
■<*
CO
o
CO
CM
in
in
in
"D -;-.' i -^
o in
o r--:
O <£>
d
o
d
CO
d
CO
d
CM
m nj © >
Ol => ss e
o
a.
CO
1 T—
T^
CM
T^
T^
ih
(/)
I
^-^
""^
^-^
' — '
' — '
""^
B
ra CT-g ®
^ « o x:
nj
1
oo
,—
<y>
in
05
CD
CO
S§oo
CC
CO —
CM O
o -~-
1-^
t^
OO
m
§
?;"
(5 CO
d
O
CD
■^
in
CO
CtD
o
CO
CM
in
S =* 0 o
(D
I
O -;
"^ ,^
to )-c
CM
(J)
CO
S « >, .
w
CM £.
1 2.
y-
d^
CM
^
^
^
2= CD S
Q)
1
1
« t: .c (0
5i c — c
OB
ffl ra .i" i;
E
Csj
® d
CM —
h-
§
C35
S"
in
d
■*
CM
« £ tr ^
CO
(0
Q.
1
S|
co
[
O)
d
i
CM
CM
d
1
in
CM
« 5 < —
2 S UJ m
IE 3) ^ CS
1
■ — '
2. w ^ 01
mo.-®
^
CNj x: -o >
QC c ra »
in
r^
■«r
CM
t
CO
CO
UJ
Tj-
Tt
CM ^^
to
Cvj
•<t
CO
^
™ £ i §
03
1-
«D CO
d CO
in o
CO
CM
T-
CM
iri
00
CO
in
<
tr
^ CO
•■- CM
CO CO
t^
r^
co
in
<J>
CM
o
H
o
CO 2.
72i
^-S
to
d
CM
.^
o
CM
?^
CM
1
o>
o
o
o
o
o
o
"" q! X) 0
c
■>s^ _^
CO
d ^
d
p^
LO
•<J-
0 ra ?>§
"5 - >,
«
GO - — -
5= CM
CM ®
00
CO
CO
<D
CM
o
in
o
CO
^
w
<ji r-
in >«
CO
CO
00
CM
o
CM
CD
o
c
o
■.- o
22 -*
iP CO
CM
OO
CM
in
y-i
CM
r^
C*J
Tj-
t-- ~^
1
in
O
1
CO
1
i
1
1
?% S E
1^ S E
t;: ^ — > ^
CO
CO
■V
O)
T—
h-
o
•S "^ m "C^
2
CO
CO
00
l~~
CO
(DPS"
CM
CM
T—
y—
(0 ^CM -«
in
t^
a>
ll^i
</)
^
r^
r^
1--.
5
5 a'l^
— CO -
(0
0)
1
in
1
CM
1
CD
1
00
O
^
1
in
CD
r^
r^
h-
00
00
CO
/>-
(j>
05
O)
<7>
O)
05
CD
o
o
CO
CD
O
o
a.
to
o
"(0
0)
c
o
CO
13
CT
(t>
0
'c
o
(D
B
E
CD
(0
eg
m
<
DC
(0
o
CD
CO
CO
00
d
05
00
corr- y- ^^ eg—, h-^^ t^ —
in"' "-o cMin -"jco •<rin
COT Tj-o 'tCD y- y- -^ <M
I CO I cvj I ^ C^ 06 ^ CD
CO
CM O
in CM
CO ,^
CO
CD
£2 =
si
CMd OOlh '-CO COCM
m ^ CD :;r
CD '^ 05 "^
••- -r- CM •.-
00 --.
CO
05 — .
■<3- —
^CM $2Jm SP 5^
CM--- CM'- incM ■>-o
I "^ I ~^ "^ I "^
CO — .
o
CO -^ CO — ~
" Eo «> ^
CO °^ CM ^
Tt -I- 1-
co
CO
CO
t^ 00
CO
••-t^ r^TT t^OO -^(D
■r-05 OOO) CMCD r^o
'- _ ■•- _ 00 _ o
00 C. CO C- CM —
- o
CM ^^ 1- ^^
t^ "tj- CD r^
^-05 in 00
in ^ CD .^
CD C CO C
CM
00
05
oco r^co yf (£> coin
o r^ •.- -^ eg Tj- ,- <d
"^(S '-(-i CDfvi CD_.
I S. I 2. CO 2i eg C-
CO,^ y- ^^ O^^ ■^^^ y- ^^ CO— ^
opco c^co ^co Tin ^. <y> Pm
coeg cb^ CM'— 05® 'i'^ -^o
'-COCO-q^CO.^'-.^ \ y^ jC3
— I — I -^ ^ ^ ^
000
CD ~ CO — . CM . — .
§s gte «s
CO o o 10 in '*•
I ^^ •* -^ CO -^
o o
CO ^^ CO . —
05 ^^ CO
o S o
CO
05 -^
CD "^ eg
CO o in eg T- CM
^-' eg "-' CO ^
^ CM
o r-
CO
P ^
'. 00
O CO O '-
'- _ CM _ CO
^ CO o CO '"' ~
d> y- O CO
CM '-
P in
O CO
05 fs.
eg £. n
CM CO
CO ^ — ^ o ^ —
7^ in • o
Si o r: CO
CO ■ CO •
P.
'- in CD
■^ 'X- f^ 'X- CD ;r-
■ in _; in ■ o
05 CD CO '^ ^ CO
in Z; CD ■ r- ■
I O '- '-
•t,^ CO. h^^ co,^ CM—, in,^
ooco '-r^ 'S-'- r^Tj- •^r^ cdcd
CO ,nCO rSI^ CO*" d"^ lri°
CD
CO
in
CO ^^
CM r^
in
CD
O I '-
CO CD O CO
CD^ in^ Tt^ 05^
ihcD COCO -^in oocM
cocD cocg cor^ oin
'- y- f-T*: C*'d ■^ y^
y- CO
■* -^ d
o ;r o r~
CO o 1^ '^
00 ■ '- T
■* '- 00 .r-
o
eg
CO ;;;-
r^ 00
eg CO
CO eg
f^ — -
I
o
T
in
r-~
r^ • —
CO eg
CO CN
h- CM
I
eg
in
in
CD 00
•«1- CM
'^ eg
05
06
in ---
00 CM
CO
CO
CO
CD
»- eg '-
05
00
1^
eg
1^
CO
'- in r-~ 05
t^ t^ t^ t^
I I I I
in CM CD CO o '-
CD r^ r- h~ CO 00
05 05 05 05 05 05
a
-C5 _:
(0 Q)
1- O)
m tfi
> <D
>, >
11
%o
™ §>
en to
CO ®
O S
— at
1^ —
..LU
^5.
S,|S
O _:_; 52
o S 5
5 i "P
eo e —
§5 £
>3 CO
I II S
3 — £
■^ £ c
^ .<» m
C ! CL
-O T- C
^ — ■-
o c <"
^ o £
™ B-g
M 2 ^
^^ o
10 o
n £
c o
o =:
£ 2 o o
■D 5
s —
03 to
o) 2
i? *
o o
(0 o
CM £
QC c
O cj)
<0 IS
I- re
QC t>j
CD CC
-§
. . Q.
«I
|c
o *^
re $
T3 Q.
? I
£^
re 0)
> 2
re
s
05
0)
?
n
(T>
r
0
0
re
<i>
0
>.
m
<
r
05
c
CL
,_
<
.^
Hi
en
>-
0
tn
TT
?
?
rr
T5
<1)
C
3
re
re
>
1
(ft
r
■n
0
r
re
9)
r
n
m
re
c
2
0)
F
(Tl
E
•V
05
■0
re
as
CO
2
0
■0
c
^
0
(0
Z
in
_i
j-
CO
'^
344
KIRKLEY AND SQUIRES: CAPITAL STOCK AND INVESTMENT IN A FISHERY
Atlantic Fisheries and increased fuel prices be-
tween 1972 and 1975; 3) the formulation and im-
plementaiion of the Fisheries Conservation and
Management Act in 1976 and 1977; 4) extremely
restrictive regulations in 1978 and 1979; and 5) a
long dock strike in New Bedford, frozen harbors,
and foreign purchases from American fishermen
in 1981. In addition, there was substantial entry
of new steel-hulled vessels between 1975 and
1977 and 1979 and 1980.
Chow (1960) tests of the equality of regression
equations were conducted to further examine the
possibility that the statistical relationships be-
tween cost and vessel characteristics changed
over time. The Chow test is an F-test of the stabil-
ity of the coefficients (Maddala 1977). The results
of these tests are presented in Table 3.
The results of the tests indicate that the esti-
mated relationships between cost and character-
istics were not the same for the selected time peri-
ods. Alternatively, the hypothesis of the equality
of the regression equations for different periods of
time could not be accepted at any reasonable level
of significance. However, these results only verify
that different models should be estimated for dif-
Table 3. — Results of tests for the equality of regression
equations.
Periods
Critical values'
tested
F-statistic2
0.05
0.01
1965-71
1972-75
1976-77
1978-79
1980-81
39.38
'^12,1050 = ''•^5
f^12,1050 =
2.18
1965-71
1972-75
8.45
1^12.325 = ■'•'75
f^12,325 ^
2.18
1965-75
1976-81
20.55
f^12.1086 = ''•^5
f^12.1086 =
2.18
1965-75
1976-79
21.33
1^12.668 = ■'■''5
'^12.668 =
2.18
1976-77
1978-79
3.60
'^12,329 = l-^S
'^12,329 =
2.18
1976-77
1978-81
3.22
^12.737 = l-^S
f^1 2.737 =
2.18
1976-79
1980-81
3.35
f^12.737 = ■'•75
f^12.737 =
2.18
1980
31981
2.76
f^12,384 = "I -^5
f^l2,384 =
2.18
'F-statistIc for denominator degrees of freedom equal to infin-
ity.
^Restricted residual sum of squares obtained by pooling data
over all time periods and estimating Equation (2). Unrestncted
residual sum of squares obtained by estimating Equation (2) for
each penod being examined and then adding the residual sum
of squares for each equation.
^Residual sum of squares obtained from estimating Equation
(2) without the time vanable
ferent periods of time. They are not conclusive
proof of the selection of the years for a particular
group of years. A more accurate determination of
the years to be included in each group requires
considerably more estimation and hypothesis
testing which is beyond the intent of this study.
As indicated by the ^-statistics in Tables 1 and
2, the age of the vessel, whether or not the vessel
is new or used, and the size characteristics appear
to be the more statistically significant explana-
tory variables. These results are consistent with
the results of Griffin et al. (1978). The statisti-
cally significant negative coefficients for the size
characteristics during the years 1972-75 are par-
ticularly interesting. These years coincide with
the Arab oil embargo when fuel prices increased;
expected increases in future operational costs
may have deterred new entry. The results also
indicate that the value of a vessel declines as it
becomes older. The coefficient for hull construc-
tion, CS, suggest that steel-hulled vessels are
more expensive than are those of other materials
such as fiberglass, ferro cement, wood, and alu-
minum. There does not appear to be a difference
in the value of a vessel based on gear type. This
may be a result of specification problems with the
dummy variable for gear t)^e (see footnote 7).
It is of further interest to examine the elastic-
ities of cost with respect to the vessel characteris-
tics. ^° These are presented in Table 4. As indi-
cated, vessel tonnage, length, and age have the
greatest influence on the acquisition price. The
elasticity for length is consistent with the manner
in which vessels are sold; that is, vessel prices are
largely stated in terms of their length.
There are, however, inconsistences in the esti-
mated capital stock equations and elasticities. In
'0 Elasticities indicate the p)ercentage change in cost result-
ing from a 1% change in the value of a characteristic.
Table 4. — Estimated elasticities of cost with respect to
vessel characteristics.'
Elasticities
Year
GRT2
HPWR
LN
AGE
1965-71
0.43
0.06
2.88
-0.32
1972-75
0.78
-0.04
-0.08
-0.08
1976-77
1.14
0.09
0.11
-0.11
1978-79
-0.14
0.30
2.38
-0.40
1980
0.16
0.30
1.43
-0.30
1981
-0.64
0.56
2.38
-0.41
' Elasticities based on nominal estimates and observed mean
values of cost and vessel charactenstlcs.
2GRT is gross registered tonnage; HPWR is engine horse-
power; LN is lengtti; AGE is age ol vessel.
345
FISHERY BULLETIN: VOL. 86, NO. 2
particular, not all of the estimated coefficiencts
have the desired or expected sign. For example,
the coefficient for tonnage in the 1981 estimate is
negative and that for the square of tonnage is
positive. In contrast, a positive coefficient for ton-
nage was expected since vessel cost should in-
crease with vessel tonnage, and the coefficient for
tonnage squared was expected to be negative be-
cause vessel costs are believed to increase at a
decreasing rate.
The incorrect signs could be the result of multi-
collinearity between vessel characteristics. Aux-
iliary regressions, though, do not indicate severe
multicollinearity. The R^ values are <0.35. If
multicollinearity is a problem, its major effect is
on the estimated relationships between cost and
individual characteristics, the estimated elastic-
ities, and the previously discussed Chow tests.
Multicollinearity is not thought to be a problem
for several reasons. First, Chow tests in the pres-
ence of multicollinearity usually result in the ac-
ceptance of the null hypothesis of the equality of
regression equations or that differences in the es-
timated coefficients are statistically insignificant
(Maddala 1977). Second, multicollinearity does
not pose a problem for estimating capital stock
and investment; the estimated equations still per-
mit estimation of the conditional mean.^^
11 For additional information on the problems of multi-
collinearity with respect to analysis and prediction, see Kmenta
(1971).
The equations for the real or deflated value of
capital were used to estimate total capital stock
and investment for the New England fleet be-
tween 1965 and 1981. The real capital stock equa-
tions were applied to vessel data available from
the Northeast Fisheries Center, and total capital
stock was calculated as the sum of the stock over
all vessels. Aggregate net investment was calcu-
lated as the difference in total capital stock be-
tween consecutive years.
Prior to estimating total capital stock and in-
vestment, it was necessary to define vessel type
by gear to avoid double counting which might
occur since vessels frequently switch port and
gear. Three categories of gear types were estab-
lished: 1) otter trawl, 2) scallop dredge, and 3) all
others, which include lobster trawl and pots, her-
ring gear, harpoons, etc. Vessels were assigned a
gear type based on a plurality of days at sea. Esti-
mates of capital stock and investment are pre-
sented in Table 5.
Several limitations should be considered when
evaluating the estimates. First, there is the previ-
ously discussed problem that estimates of capital
stock may be biased or overestimated. Second,
estimates are only for New England vessels or
those which are believed to be homeported in New
England. Third, the capital stock and investment
series pertain to New England and not the United
States; that is, estimates should not be inter-
preted as net changes in the capital stock and
investment in U.S. fisheries. Fourth, under-
Table 5. — Estimates of real capital stock and investment in New England harvest-
ing sector, 1965-81.1
Number
of
vessels
Capital stock
Investment
Index
of
investment
Year
Trawl
Dredge
Other
Trawl
Dredge
Other
- $ Millie
inc --_--.
Mio ■
1965
594
25.96
16.29
1.98
1966
619
28.84
14.75
2.82
2.88
-1.54
0.84
100
1967
628
39.37
11.42
1.63
10.53
-3.33
-1,19
276
1968
610
28.65
15.43
1.68
-10.72
4.01
0.05
-301
1969
602
29.22
12.92
1.29
0.57
-2.51
-0.39
-107
1970
607
25.61
8.29
1.37
-3.61
-4.63
0,08
-374
1971
620
23.65
12.80
1.47
-1.96
4.51
0,10
122
1972
655
27.01
7.74
4.70
3.36
-5.06
3,23
70
1973
666
28.42
7.12
5.13
1.42
-0.62
0,43
56
1974
695
21.26
4.19
6.40
-7.17
-2.93
1,27
-405
1975
737
29.04
5.15
8.48
7.78
0.96
2.08
496
1976
783
20.73
7.72
8.16
-8.31
2.57
-0.32
-278
1977
836
28.86
14.19
7.86
8.13
6.47
-0.30
656
1978
881
37,45
12.85
8.65
8.59
-1.34
0,79
369
1979
1,107
45.41
27.30
11.25
7,96
14.45
2,60
1147
1980
1,260
52.26
34.30
11.69
6.85
7.00
0,44
656
1981
1,246
43.35
30.15
11.60
-8.91
-4.15
-.09
-603
'Real Investment and capital are deflated with producer pnce Index for tieavy machinery,
1967 = 100.
346
KIRKLEY AND SQUIRES: CAPITAL STOCK AND INVESTMENT IN A FISHERY
tonnage vessels or those which are <5 gross
registered tons are excluded since data are
unavailable; thus, total capital stock and invest-
ment in New England fisheries are underesti-
mated.
Despite these limitations, the results provide
information on investment and capital stock in
New England fisheries which has not been avail-
able. As is characteristic of open-access fisheries,
investment in New England fisheries appears to
have increased over time. This is quite evident
subsequent to the passage of the Magnuson Fish-
eries Conservation and Management Act of 1976.
While possibly true, the observation that invest-
ment increased when the act was passed should
not be construed to imply that the act was respon-
sible for increased investment. Positive expecta-
tions regarding fish stocks or economic conditions
could also be responsible for increases in capital
stock and investment.
The largest increase in investment occurred in
1979 which coincided with the greatest influx of
vessels. It also coincided with increased fish
stocks in 1978 and 1979 and high vessel profits of
the previous year. In 1979, 226 vessels entered
the New England fisheries; 92 vessels were newly
constructed.
Disinvestment also occurred, but the reasons
are unknown. ^^ Disinvestment may be associated
with vessel sinkings, foreclosures on vessel mort-
gages, and exit from the region. Alternatively, if
vessel prices reflect the expected net discounted
present value of earnings from a fishery, the dis-
investment or even changes in investment may
be due to changes in economic expectations. Dis-
investment occurred between 1968 and 1970,
1974, 1976, and in 1981.
Disinvestment between 1968 and 1970 may re-
flect expectations of declining future earnings.
Disinvestment in 1974 may reflect the effects of
high fuel prices. Disinvestment in 1976 and 1981
may be associated with fishermen's expectations
of declining fish stocks, the effects of manage-
ment, and high fuel prices.
The reasons for investment are also unknown.
Increasing levels of investment between 1977 and
1980 may be due to economic expectations associ-
ated with fisheries management, fish stocks, and
future earnings.
The real capital stock increased approximately
6% per year between 1965 and 1981. The largest
increase occurred in 1978 and 1979 when there
was a substantial increase in the number of new
steel-hulled vessels and entry of vessels from the
mid-Atlantic region. Capital stock declined in
1981 when some of the mid-Atlantic vessels left
New England and the stocks of fish, particularly
sea scallops, declined (Northeast Fisheries Cen-
ter 1985).
The results also indicate why a vessel count
should not be used to indicate capital stock or
investment. The number of vessels increased be-
tween 1969 and 1970 and 1975 and 1976, but
investment decreased in 1970 and 1976 (Table 6).
Alternatively, the number of vessels increased in
each year relative to 1965, but the real capital
stock between 1969 and 1976 declined with re-
spect to 1965.
Table 6. — Indices of capital stock based on constant dollar value
and vessel count.
Capital stock
Constant dollar value
Vessel count
Year
Trawler!
Dredge'
Total
Trawler!
Dredge!
Total
1965
100
100
100
100
100
100
1966
111
91
105
102
94
104
1967
152
70
119
109
77
106
1968
110
95
103
104
106
103
1969
113
79
98
104
106
101
1970
99
51
80
108
74
102
1971
91
79
86
108
83
104
1972
104
48
89
108
74
110
1973
109
44
92
105
79
112
1974
82
26
72
107
49
117
1975
112
32
96
105
66
124
1976
80
47
83
109
202
132
1977
111
87
115
107
202
141
1978
144
79
133
116
189
148
1979
175
168
190
137
313
186
1980
201
211
222
158
440
212
1981
167
185
192
158
413
210
i2Hirshleifer (1970) provided additional information about
disinvestment.
'Gear type assigned by plurality of days absent.
The inadequacy of using vessel count is more
pronounced when comparing the indices based on
gear type. As shown in Table 6, the number of
trawl vessels in every year exceeded the number
of vessels in 1965 while capital stock was less
than that of 1965 for four years. The two scallop
dredge indices also indicate dissimilar changes
between vessel count and real capital stock. It
may, though, be possible than an index based on
vessel count and weighted by vessel size would
more closely compare to the real capital stock
index. The construction of these indices would re-
quire considerable additional analysis and deter-
mination of the weights necessary for aggrega-
tion.
347
FISHERY BULLETIN; VOL. 86, NO. 2
SUMMARY AND CONCLUSIONS
This study developed procedures for estimating
the capital stock and net investment using only
the information generally available. That is, data
on acquisition and list prices for some but not all
vessels and characteristics for most vessels in a
fleet were used to estimate capital stock and in-
vestment in a fishery. An hedonic approach which
specified the acquisition price as a function of ves-
sel characteristics was suggested as a possible
method for obtaining information necessary for
estimating capital stock and net investment. It
was argued that the approach was consistent with
procedures used by economists to estimate the
capital stock and investment in traditional indus-
tries.
Estimates were based on 1,110 observations ob-
tained from NMFS and classified advertisements
for the period 1965-81. Vessel characteristics
pertaining to size and age were shown to be the
more significant characteristics for explaining
the value of a vessel. The corresponding informa-
tion was then used to estimate the capital
stock and investment in the New England fish-
eries.
An interesting result was that although invest-
ment increased over time, there also was disin-
vestment in the fisheries. It was suggested that
this was possibly the result of vessel sinkings,
exit of vessels from the region, foreclosures, and
expectations of declining future net returns.
It also was demonstrated that a vessel count
should not be used to indicate capital stock or
investment. The number of vessels generally in-
creased over time, but the level of capital stock
did not coincide with these changes. This suggests
that more attention should be given to developing
economic measures of capital stock and invest-
ment.
A question still in need of attention, however,
cannot be answered from this study. That is,
given the level of investment, can an optimal uti-
lization and allocation of resources be deter-
mined. The answer is clearly no. This requires
management authorities to specify the objectives
of fisheries management and a detailed bioeco-
nomic model.
There is a need for more research on invest-
ment, and in particular, the determination of the
optimum rate of investment. This includes re-
search on the social discount rate, reasons for in-
vestment, and the marginal productivity of capi-
tal. More important, additional research is
required to better address the issues of fisheries
management, particularly whether or not public
expenditures should be allocated to managing
fisheries.
ACKNOWLEDGMENTS
Comments by William DuPaul, Wade Griffin,
Ray Hilborn, Maurice Lynch, Ivar Strand, and
two anonymous reviewers are gratefully ac-
knowledged. Any errors remain the responsibil-
ity of the authors. The article does not necessarily
reflect the opinion of either the National Marine
Fisheries Service or the Virginia Institute of
Marine Science.
LITERATURE CITED
Baumol, W J
1977. Economic theory and operations analysis.
Prentice-Hall, Englewood Cliffs, N.J.
Blackorby, D., D Primont, and R R. Russel
1978. Duality, separability, and functional structure: the-
ory and economic applications. North-Holland Publish-
ing Company, Amsterdam.
BocKSTAEL, N. E., W. M Hannemann, and I. E. Strand.
1986. Measuring the benefits of water quality improve-
ments using recreation demand models. Volume II of ben-
efit analysis using indirect or imputed mtirket methods.
EPA contract No. CR-811043-01-0. U.S. Environmen-
tal Protection Agency, Washington, D.C.
Box, G E P , and D R Cox
1962. An analysis of transformations (with discussion).
J. R. Stat. Soc, Ser. B:21 1-243.
Braeutigam, R R., a. F Daughety, and M. A. Turnqutst.
1982. The estimation of a hybrid cost function for a rail-
road firm. Rev. Econ. Stat. 64:394^04.
Branson, W H
1972. Macroeconomic theory and policy. Harp)er & Row,
Publishers, N.Y.
Charles. A. T.
1983a. Optimal fisheries investmentxomparative dy-
namics for a deterministic seasonal fishery. Can. J.
Fish. Aquat. Sci. 40:2069-2079.
1983b. Optimal fisheries investment under uncertainty.
Can. J. Fish. Aquat. Sci. 40:2080-2091.
Chow, G C
1960. Tests of equality between subsets of coefficients in
two linear regressions. Econometrica 28:591-607.
Clark, C W., F. H. Clarke, and G R Munro.
1979. The optimal exploitation of renewable resource
stocks: problems of irreversible investment.
Econometrica 47:25-47.
Cunningham, S., M. R. Dunn, and D. Whitmarsh.
1985. Fisheries Economics: an introduction. St. Mar-
tin's Press, N.Y.
Dadkhah, K. M
1984. Confidence interval predictions from a logarithmic
model. Rev. Econ. Stat. 66:527-528.
Dewar, M. E.
1983. Industry in trouble: the Federal government and
348
KIRKLEY AND SQUIRES: CAPITAL STOCK AND INVESTMENT IN A FISHERY
New England fisheries. Temple University Press,
Philadelphia, PA.
DOERINGER, P B.. P I Moss, AND D. G. TERKLA.
1986. The New England fishing economy. The Univer-
sity of Massachusetts Press, Amherst, MA.
Gordon, H S.
1954. The economic theory of a common property re-
source: the fishery. J. Polit. Econ. 62:124-142.
Griffin, W. L.J P Nichols, R. G. Anderson. J E Buckner. and
C M Adams.
1978. Costs and returns data: Texas shrimp trawlers.
Gulf of Mexico 1974-75. Texas A&M Univ., Dep. Agric.
Econ. tech. Rep. TAMU-SG-76-601.
Griuches, Z
1971. Hedonic price indexes for automobiles: an econo-
metric analysis of quality changes. In Price indexes and
quality changes, p. 103-130. Harvard University
Press, Cambridge, MA.
HILBORN. R.
1985. Fleet dynamics and individual variation: why some
people catch more fish than others. Can. J. Fish. Aquat.
Sci. 32:2-13.
Hirshleifer. J.
1970. Investment, interest and capital. Prentice-Hall,
Englewood Cliffs, NJ.
Kendrick, J
1961. Some theoretical issues in capital measurement.
Am. Econ. Rev. 51(2):102-111.
KMENTA, J.
1971. Elements of econometrics. Macmilliam Publish-
ing Co., Inc., N.Y.
KiRKLEY, J E , M P Pennington, and B. E. Brown.
1982. A short-term approach for analyzing the effects of
harvesting quotiis. J. Cons. int. Explor. Mer 40:173-
175.
KiRKLEY, J. E
1986. The relationship between management and the
technology in a multi-species fishery. Ph.D. Thesis,
Univ. Maryland, Collage Park, MD.
Ladd, G W . AND M B Martin.
1976. Prices and demands for input characteristics. Am.
J. Agric. Econ. 58:21-30.
Lau, L J
1978. Applications of profit fiinctions. In M. Fuss and
D. McFadden (editor), Production economics: a dual ap-
proach to theory and applications, p. 133-216. North-
Holland Publishing Company, Amsterdam.
Maddala. G S.
1977. Econometrics. McGraw-Hill Book Company, N.Y.
National Academy of Sciences
1979. Measurement and interpretation of productivity.
National Research Council, Washington, D.C.
Northeast Fisheries Center.
1985. Status of the fishery resources off the northeastern
United States for 1985. U.S Dep. Commer., NCAA
Tech. Memo. NMFS-F/NEC-42, 137 p. Northeast Fish-
eries Center, National Marine Fisheries Service, NOAA,
Woods Hole, MA.
Rosen, S.
1974. Hedonic Prices and implicit markets:product differ-
entiation in pure competition. J. Polit. Econ. 82:34-55.
Scorr, A.
1979. Development of economic theory of fisheries regula-
tion. J. Fish. Res. Board Can. 36:725-741.
SISSENWINE, M P , and J E. KiRKLEY
1982. Fishery management techniques:practical aspects
and limitations. Mar. Policy 6:22-42.
Sloan, H S., and A. J. Zurcher.
1968. Dictionary of economics. Barnes & Noble, Inc.,
N.Y.
Smallwood, D M., and J. R. Blaylock.
1986. Forecasting performance of models using the Box-
Cox transformation. Agric. Econ. Res. 38:14-24.
Tettey, E O . W L. Griffin, J. B. Penson, and J. R Stole.
1986. Implications of tax policy on investment in a com-
mon property resource. N. Am. J. Fish. Manage. 6:100-
104.
Triplett, J
1986. The economic interpretation of hedonic methods.
Surv. Curr. Bus. 66:36-40.
Varian, H. R.
1984. Microeconomic analysis. 2d, ed. W. W. Norton &
Company, N.Y.
349
ANALYSES OF THE RELATIONSHIP BETWEEN
THE DISTRIBUTION OF SEARCHING EFFORT, TUNA CATCHES, AND
DOLPHIN SIGHTINGS WITHIN INDIVIDUAL PURSE SEINE CRUISES
Tom Polacheck^
ABSTRACT
The fine scale distribution of searching effort within individual purse seine cruises in the eastern
tropical Pacific is analyzed in relationship to sightings of spotted dolphin, Stenella attenuata, and
tuna catches. The data for these analyses were derived from detailed observations made by National
Marine Fisheries Service observers aboard U.S. purse seiners. A clustering algorithm is developed
which separates the activity of a vessel into areas where sets were common and areas where they are
infrequent. Within clusters of high set densities, vessels tend to concentrate their searching effort.
Vessels searched proportionately greater distances relative to the physical distances between sets
while within clusters than when outside clusters. Encounter rates with schools of spotted dolphins
tend to be either much higher or much lower within defined clusters than outside them. Clusters with
low encounter rates were clusters in which non-dolphin associated tuna sets predominated. Because
of this dichotomy in the magnitude of the dolphin encounter rates within clusters, overall encounter
rates appeared to have relatively small biases if the concentration of searching effort within clusters
is ignored. The average catch of tuna per set was higher within the defined clusters than between
them. The overall results suggest that fine scale geographic effects need to be considered when using
data from purse seiners to examine changes in relative abundances of either dolphins or tuna.
Catch and effort data underlie most indices of
abundances used for assessing the status of com-
mercially exploited fish stocks. The validity of
using catch and effort data from commercial har-
vests has long been questioned because of the
likelihood that fishermen concentrate their effort
in areas of high fish densities (Helland-Hansen
1909). Yet, almost nothing is known about the
allocation of effort by individual vessels. If data
are available on the activity of individual vessels,
these data have been mainly used for standard-
ization of effort. The catch and effort activities
within individual cruises have not been examined
in detail (in part because the data for such an
examination generally do not exist). The purpose
of the present paper is to examine the searching
behavior of tuna purse seiners in the eastern trop-
ical Pacific (ETP) based on detailed data compiled
by the National Marine Fisheries Service
(NMFS). The main questions addressed are
whether seiners concentrate their effort and what
is the relation between searching behavior, en-
iSouthwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038 and Department of Biology, University of Oregon, Eu-
gene, OR 97403; present address: Northeast Fishery Center
Woods Hole Laboratory, National Meirine Fisheries Service,
NOAA, Woods Hole, MA 02543.
Manuscript accepted January 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.
counter rates for dolphins, and tuna catches.
The maximization of profit is presumably a
strong influence on the behavior of commercial
fishermen. Upon leaving port, the catch rate real-
ized by a fisherman is probably the most impor-
tant factor affecting his profits. Given this orien-
tation, it is reasonable to assume that fishermen
have developed strategies of when and where to
fish that increase their catch rates beyond that
achieved by random search. There is little empir-
ical information to support this assumption other
than correlations, which have been noted for
some fisheries, between the spatial distribution of
catch rates and effort for the fleet as a whole (Gul-
land 1955; Calkins 1963).
In a fishery where the detection of fish depends
upon visual cues, searching would be expected to
be located in the vicinity of previously located
fish, if there is a tendency for the underlying pop-
ulation to be spatially clustered. The search path
for a vessel in such a fishery might be expected to
look something like the hypothetical one depicted
in Figure 1. The amount of crisscrossing or
zigzagging in the vicinity of a catch and the area
over which the search extends would be expected
to vary between fishermen. The solution to the
optimal searching strategy for such a situation is
nontrivial and depends upon information on the
351
FISHERY BULLETIN; VOL. 86, NO. 2
Figure l. — A hypothetical cruise track for a vessel searching
for clustered prey. The location of catches are indicated by
xs.
underlying distribution of the fish population
(Koopman 1980). The question of optimal search-
ing strategies for fishermen has been receiving
increased attention (Pazynich 1966; Salia and
Flowers 1979; Clark and Mangel 1983). These
studies are primarily theoretical at this time and
their application to actual fisheries requires
knowledge of the spatial distribution of the fish
population.
Locating schools of tuna (e.g., yellowfin, Thun-
nus albacares, and skipjack, Katsuwonus
pelamis) in the purse seine fishery in the ETP
depends on visual cues. Fishermen use a variety
of cues including birds which feed on the same
prey as tuna, disturbances on the surface of the
water, floating debris which frequently have as-
sociated tuna, and schools of dolphins which are
often associated with tuna (primarily yellowfin).
Fishermen have names for the different types of
sets depending upon what is associated with the
tuna school. They refer to sets associated with
floating debris as log sets, sets associated with
dolphins as porpoise sets, and sets not associated
with other animals (except possibly birds) as
school sets. Fishermen when not engaged in a set
usually spend their day actively searching for
signs of tuna. They use 25 x binoculars to scan the
water while the boat cruises at speeds generally
between 10 and 12 knots.
The distribution of schools of tuna, as well as
schools of the most commonly associated dolphins
(i.e., the spotted dolphin, Stenella attenuata), ap-
pears to be spatially and temporally clustered
within the ETP considered as a whole (Calkins
and Chatwin 1967, 1971; Blackburn and Wil-
liams 1975; Suzuki et al. 1978; Au et al. 1979^;
Polacheck 1983). At finer geographic scales, there
is little available information although Au et al.
(fn. 2) suggested that schools of spotted dolphin
tend to be locally concentrated in areas of conver-
gences and fronts.
Given the above observation, it is not surpris-
ing that the detection or encounter process for
tuna or dolphins does not, in general conform to a
Poisson process when the distribution of search-
ing times, searching distances, or physical dis-
tances between nearest encounters are analyzed
(Polacheck 1983; Allen and Punsely 1984). In
such analyses it is impossible to separate or dis-
tinguish the effects of nonrandom search from
nonrandom distributions of tuna or dolphins.
In the harvesting of tuna schools associated
with dolphins, fishermen chase and capture the
associated dolphins (Perrin 1968, 1969), and some
dolphins may be incidently killed. NMFS, as part
of its responsibility under the Marine Mammal
Protection Act of 1972 for managing and monitor-
ing the status of dolphin populations, placed
trained observers aboard tuna purse seiners.
From the data collected by these observers, ap-
proximate cruise tracks can be drawn by connect-
ing all positions that were recorded. Many of
these approximate cruise tracks (e.g.. Figure 2)
have superficially a strong similarity to the hypo-
thetical one depicted in Figure 1. It was this sim-
ilarity that provided the impetus for the analyses
presented below.
METHODS
NMFS observers aboard tuna purse seiners col-
lected a wide variety of information, both on the
sightings of marine mammals and fishing opera-
2Au, D. W. K., W. L. Peryman, and W. F. Per-
rin. 1979. Dolphin distribution and the relationship to envi-
ronmental features in the eastern tropical Pacific. Natl. Mar.
Fish. Serv., Southwest Fish. Cent., Adm. Rep. LJ-79-43, 59 p.
352
POLACHECK: DISTRIBUTION OF SEARCHING EFFORT
Figure 2. — Examples of the approximate cruise track for two tuna purse seiners. Diamonds represent the
location of a set or chase. No geographic coordinates are given, and the orientation of figures were rotated at
random so as not to compromise any proprietary fishing information. The distance between tick marks equals
300 nautical miles.
353
FISHERY BULLETIN: VOL. 86, NO. 2
tions. Included in the recorded data are informa-
tion on all changes in a vessel's activity; the loca-
tion, tjqae, and catch of all purse seine sets; and
the location and identity of all marine mammal
sightings. In these data, a vessel's activity is clas-
sified into one of five mutually exclusive cate-
gories: searching, running, setting, chasing, or
resting. Searching is defined to be whenever a
vessel is moving and the crew are actively search-
ing for signs of tuna; running, anytime the vessel
is moving but not actively searching for signs of
fish (e.g., moving locations at night); chasing,
anytime schools of dolphins are being pursued
before the net has begun to be set. More detailed
descriptions of the available data, collection pro-
cedures and their preparation for the analyses
below can be found in Polacheck (1983, 1984^).
The analyses in this paper were part of a larger
project on the use of these observer data for as-
sessing the relative abundances of dolphin stocks.
As such, the emphasis in this paper is on the
encounter rate for the most important dolphin
species for the fishery (spotted dolphin), although
catch rates for tuna are also considered. The re-
sults presented in this paper are based on two
different approaches for analyzing the data. The
first method is a set of nearest neighbor calcula-
tions, and the second is a cluster analysis.
The nearest neighbor calculations were per-
formed in order to get an indication whether ves-
sels tend to search in the vicinity of a previous
encounter (either a sighting of marine mammals
or a set on tuna). In these calculations, the phys-
ical distance between either the next or preceding
encounter is compared with the distance to the
nearest other encounter made within the entire
cruise. Also, the proportion of times in which the
nearest encounter is not either the next or preced-
ing one is calculated. For a vessel that never re-
turned to the area of an encounter, this propor-
tion would equal 1. Similarly, if a vessel never
returned to the area of an encounter, the ratio of
the distance between either the next or preceding
encounter and the distance to the nearest other
encounter within an entire cruise would also
equal 1. Note that the expected values for these
proportions with random search are not necessar-
ily 1. The expected value will be dependent both
on the distribution of potential encounters and
3Polacheck, T. 1984. Documentation of the time sequen-
tial files created ft-om the tuna boat observer data bases for
analyzing relative abundances. Natl. Mar. Fish. Serv., South-
west Fish. Cent., Adm. Rep. LJ-84-33, 26 p.
the definition of random search (see Discussion).
These calculations were performed separately for
sets and chases for tuna and for the sightings of
spotted dolphin. In performing these nearest
neighbor calculations, the first and last encounter
during a cruise were not included.
The other main approach used for examining
the data is a form of cluster analysis. When the
sequences of distances between sets and chases
within any cruise were examined, they appeared
to be spatially and temporally clustered in the
sense that sets and chases in which the distance
to the next set or chase was small tended to be
clumped sequentially. This observation led to the
development of an algorithm for clustering sets
and chases that were spatially and temporally
related. Standard clustering algorithms were not
appropriate in this situation because of the prob-
lem of scaling spatial and temporal distances
within a common metric (i.e., how much time
should be equal to a given distance).
Note that the term "clustered" or "clustered
distribution" is used in this paper to refer to any
distribution in which high- and low-density areas
are more frequent than would be expected if the
distribution was generated by a Poisson process.
The term is not meant to refer to any particular
nonhomogeneous process. A cluster is considered
as an area of high density and should not be con-
strued as referring to a discrete unit.
The primary purpose of the clustering al-
gorithm was to define areas which a fisherman
might have thought to have a high density of
potential fishing targets so that the searching be-
havior of a vessel could be compared between
these areas and outside them. This analysis ex-
ploits the fact that the physical distance between
events is partially independent of the distance
that a vessel travels to locate them. Since the
purpose of the algorithm was to define areas of
potentially good fishing, chases of dolphin, as well
as sets, have been included as events in the clus-
tering algorithm. (Sets made for the purpose of
washing the net were not used.) The clustering
algorithm began with consideration of the dis-
tance between the first and second set and/or
chase. If this distance was less than a specified
amount, then these two events were placed in the
same cluster, and the distance between these two
and the third event were examined. This specified
amount will be referred to as the clustering
parameter. If the distance between the third
event and either of the events within the cluster
was less than the value of the clustering parame-
354
POLACHECK: DISTRIBUTION OF SEARCHING EFFORT
ter, this third event was included in the cluster
and the fourth event was examined. Whenever a
set or chase was found for which the distance be-
tween it and all members of the last defined clus-
ter exceeded the clustering parameter, a new
cluster was formed. This process was repeated
until all sets and chases within a cruise were
placed in a cluster.
Using this algorithm, all the activities within a
cruise could be considered to occur either between
or within clusters. Isolated sets or chases (i.e.,
clusters containing only a single event) were con-
sidered as occurring between clusters. Distances
travelled (i.e., distance searched and distance
run) within a cluster were defined as the dis-
tances travelled after the first set or chase until
the last set or chase in that cluster. Distances
travelled between clusters were defined from the
last event of the previous cluster to the first event
of the subsequent cluster. The distances travelled
until the location of the first set or chase and after
the last one were not included because of the large
distances involved in reaching the fishing
grounds.
The location of a cluster was estimated by cal-
culating the centroid for all sets and chases
within it. The size of a cluster was estimated by
determining the radius of the smallest circle with
a center at the centroid that encompassed all sets
and chases within it.
The sensitivity of this algorithm to the value of
the clustering parameter was examined for val-
ues of 50, 75, 100, and 150 miles. For most of the
results, only clusters with at least three members
are considered as clusters. Clusters with only two
members have been excluded from most of the
summary results describing a cluster and also in
the comparisons of results between and within
clusters. This was done because two physically
close events did not seem to warrant being called
a cluster. Yet, given the relative difficulty in lo-
cating potential sets, two close events might be
considered as areas of potentially good fishing.
Clusters with only two members contained 18% of
all sets and chases when the cluster parameter
equalled 50 miles and 6% of all sets when the
value equalled 150 miles. The overall results and
conclusions are robust to whether or not clusters
with two members are included or excluded.
An average intercluster distance for a cruise
was calculated in order to get an indication of the
stability of the clustering algorithm to the value
of cluster parameters. The intercluster distance
was defined as the distance from the nearest
member of a cluster or isolated set to the next set
and represents the minimum value that the clus-
ter parameter would have to be for a cluster or
isolated set to be combined in a single cluster with
the next set.
Encounter rates for schools of spotted dolphin
for each cruise were calculated between and
within clusters as the total number of sightings
divided by the total distance searched. In these
rates, if the first chase or set within a cluster was
based on a sighting of spotted dolphin, this sight-
ing was included in the encounter rates between
and not within a cluster. For the analysis of these
encounter rates, clusters were classified accord-
ing to the percentage of the total number of sets
and chases within a cluster that involved schools
of dolphins. In calculating the number of events
that occurred within a cluster, sequential non-
dolphin sets in which no searching was done be-
tween them were counted as a single event. This
was done to reduce the effect of multiple sets on
the same floating object counting as a large clus-
ter.
In order to see whether the searching behavior
within the defined clusters resulted in biased esti-
mates of encounter rates for dolphins if the clus-
ters were ignored in the estimates, two different
estimates for the overall encounter rate for a
cruise were calculated and compared. The first
estimate, which will be referred to as the unad-
justed rate, was simply the total number of en-
counters divided by the total distance searched
for an entire cruise. This would be an unbiased
estimate if search was in fact random. The second
estimate, which will be referred to as the adjusted
encounter rate, was calculated as the weighted
average of the encounter rate within and between
clusters. The weights for the encounter rate
within clusters were equal to the diameter of the
cluster. The weights between clusters equalled
the total distance searched between clusters. In
effect, this adjusted encounter rate is an estimate
of what the encounter rate would have been if a
vessel had made a straight line crossing of each
cluster. (In the calculation of these adjusted en-
counter rates, clusters with two members were
treated the same as other clusters to simplify the
calculations.)
The analyses in this paper are based on 35
cruises. This represents a subset of the cruises in
1979 with NMFS observers, which in turn is a
subset of all purse seine cruises for tuna in the
ETP during 1979. Analyses were restricted to
1979 because only for this year have the data
355
FISHERY BULLETIN: VOL. 86, NO. 2
been carefully edited for positional errors.
Cruises from 1979 were excluded either because
of insufficient positional data or because the ves-
sel made port stops during the middle of the
cruise. Inclusion of the time spent searching on
the way in and out of port could distort the results
on searching and encounter rates between clus-
ters. Preliminary analyses suggested that there is
little difference in the results between vessels
that went into port and those that did not.
However, to avoid further complicating the
analyses, these cruises have not been included
(see Polacheck 1983 for more detail).
RESULTS
Is Effort Concentrated?
The percentage of sets or chases for which the
nearest one was not the preceding or next set or
chase ranged from 12 to 77 among cruises (Fig. 3).
The mean percentage was 41 (SE = 1.9, n = 35).
The average ratio of the physical distances be-
tween the next or preceding set or chase com-
pared to the distance to the nearest set or chase
within a cruise was 1.45 (SE = 0.044, n = 35) and
ranged from 1.00 to 2.24. Consideration of the
same statistics for the distances between sight-
ings of spotted dolphin indicates an even more
concentrated pattern. The mean percentage of
sightings for which their nearest neighbor was
not either the next or preceding one equalled 80
(SE = 2.8, n = 34, note one cruise recorded no
sightings of spotted dolphins) and ranged from 25
to 100 (Fig. 4). The average ratio of the distance
between the next or preceding sighting compared
with the nearest sighting was 4.05 (SE = 0.261,
n = 34) and ranged from 1.14 to 7.82. For these
sighting statistics, low percentages and ratios
near 1.00 are found in vessels with few sightings
(Fig. 4). These results suggest that in general ves-
sels return to the area of a previous sighting and/
or set and search in that area at least 41% of the
time.
There appears to be large differences among
vessels in their ability and success at locating
potential fishing targets. Thus, the average dis-
tance searched between sets or chases varies by
about a factor of 4 among cruises, while the aver-
age physical distance between sets or chases
varies by about a factor of 6 (Fig. 5). There is little
relationship between the average distance
searched between sets and chases and the aver-
age distance to the next one. However, vessels
15
>
u
z
U4
3
O
m
a
u.
10
0 I 1 1 1 1 1 1 1
10 20 30 40 50 60 70 80
PERCENT OF THE SETS AND CHASES
IN WHICH THE NEAREST ONE WAS
NEITHER THE PRECEDING OR NEXT ONE
Figure 3. — The frequency distribution for the percentage of the
sets and chases within a cruise for which the nearest other set
or chase was neither the preceding or next one.
15
10
UJ
O
111
cc
u.
5 -
— I —
60
100
30 40 50 60 70 80 90
PERCENT OF SIGHTINGS IN WHICH
THE NEAREST ONE WAS NEITHER
THE PRECEDING OR NEXT ONE
Figure 4. — The frequency distribution for the jjercentage of
sightings of schools of spwtted dolphins, Stennella attenuata ,
within a cruise for which the nearest other sighting was neither
the preceding or next one. Shaded portions represent cruises
with less than 15 sightings (note one cruise had no sightings of
spotted dolphins).
356
POLACHECK: DISTRIBUTION OF SEARCHING EFFORT
with the largest distances between sets also have
large average search distances, but the converse
is not true (Fig. 5). The importance of dolphin
fishing as compared to log and school fishing ap-
pears not to be related to these differences among
cruises (Fig 6). It should be noted that the points
in Figure 5 must lie above the straight line with
a slope of 1.0. The expected relationship between
the variables in Figure 5 depends upon both the
underlying searching process and the spatial dis-
tribution of potential sets. While a positive rela-
tionship would be expected if both of these are
random, a more precise definition is beyond the
scope of this paper. The purpose in presenting
Figures 5 and 6 is to display the range of differ-
ences in the success of vessels in locating poten-
tial fishing targets. The large variation among
cruises suggests that all vessels may not be using
the same searching strategy and is an important
factor to keep in mind when considering the re-
sults from the clustering algorithm.
^ 220
1
■5 200
S 180
u
oc
o
»-
Ui
(A
X
u
UI
160
140
120
O 100
I-
D
UI
<
UI
<A
80
60
<
« 40
UI 20„
S 0
_L
_L
10
20
30
40
_L
50
JL
60
70
MEAN DISTANCE TO THE NEAREST
SET OR CHASE (nautical miles)
Description of Clusters
The clustering algorithm grouped between 30
and 100% of all sets and chases within a cruise
into clusters with at least three members (Table
1). The percentage of all sets and chases that are
included in clusters increases with the value of
the cluster parameter so that 60-100% of all sets
and chases within a cruise occur within 150 miles
of another one. These percentages can only in-
crease with increases in the value of the cluster
parameter (i.e., a set that is included within a
cluster for a lower value of the cluster parameter
will always be included in a cluster at a higher
value). Similarly, the average cluster size and the
total percentage of the distance searched that oc-
curs within clusters must be nondecreasing func-
tions of the clustering parameter (Table 1). How-
ever, even when the clustering parameter equals
150 miles so that most sets occur within clusters,
the percentage of the total distance searched
within clusters averages only 59% of the total
distance searched during the cruise. This indi-
cates that substantial searching activity occurs
far from any set or chase.
The average intercluster distance for a cruise
ranged from 134 to 425 miles for clusters defined
220
200
z
UJ -^
an
< w
Ui UJ
««
liJ ^
O X
ro
^i
<S<
o«
Z H
< UJ
UJ V>
s
180
160
140
120
100
80
60
40
20
0
• .
• •
•• •
'.
10 20 30 40 50 60 70 80
PERCENT OF THE SETS AND
CHASES INVOLVING DOLPHINS
90
Figure 5. — The relationship between the average distance
searched between sets and chases within a cruise and the aver-
age distance to the nearest one.
Figure 6. — The relationship between the average distance
searched between sets and chases and the proportion of sets and
chases in which schools of dolphins were associated for a cruise.
357
FISHERY BULLETIN: VOL. 86, NO. 2
Table 1. — Statistics summarizing the results of the clustering algorithm showing the
sensitivity to a range of values for the clustering parameter. Standard deviations are
given as opposed to standard errors to provide an indication of the variation between
cruises. All means are the simple average for 35 cruises.
Cluster parameter
50
75
100
150
Percent of sets within
clusters with at least
3 members
Mean
SD
Range
64
14.1
30-90
75
13.5
31-97
81
13.3
31-97
88
9.2
60-100
Average cluster radius
for clusters with at least
3 members (nautical
miles)
Mean
SD
Range
35
8.5
19-56
55
19.7
27-142
77
40.4
34-225
116
66.0
38-390
Percent of the search-
ing within clusters with
at least 3 members
Mean
SD
Range
25.8
12.6
9-61
39
16.4
16-91
47
16.1
19-91
59
10.6
31-91
Average intercluster
distance (nautical
miles)
Mean
SD
Range
210
65.9
134-425
261
75.9
155-465
297
82.2
173-513
356
100.9
196-691
30 r
20
z
UJ
O
10
A
N = 767
30r
20 -
10
B
N = 570
50
40 r
30 -
UJ
O 20
oc
Ul
a.
10
75 100 125 150 175 200 225 250 >250 75 100 125 150 175 200 225 250 275 >275
40r
C
N = 448
_L
I
I
_L
_L
_L
30
20
10
D
N = 309
_L
_L
_L
_L
_L
_1_
100 125 150 175 200 225 250 275 300 >300
INTER-CLUSTER DISTANCE (nautical mNaa)
150 175 200 225 250 275 300 325 350 >350
MTEH-CLUSTCR MSTANCC (nauUcal nritM)
Figure 7. — The frequency distribution of intercluster distances. Distances from all cruises have been pooled. The values of the
clustering parameter are A) 50, B) 75, C) 100, and D) 150.
358
POLACHECK: DISTRIBUTION OF SEARCHING EFFORT
by a value of 50 for the clustering parameter and
from 196 to 691 miles for a value of 150 miles
(Table 1). The frequency distribution of these in-
tercluster distances (Fig. 7) is an indication of the
stability of the clusters to the value of the cluster-
ing parameter. Thus, for a value of 100, over 65%
of the clusters have an intercluster distance
exceeding 175 miles. This suggests that 65% of
clusters will be stable up to a value of 175 miles
for the clustering parameter. (Note, this is not
strictly true. If the set preceding the first mem-
ber of a cluster was less than 175 miles away and
this set was also less than 175 miles from the first
set of the next cluster, this set plus these two
clusters would be combined in a single cluster
for a value of the cluster parameter less than
175.)
These statistics describing the characteristics
of the defined clusters suggest that the algorithm
used to create them successfully separates the ac-
tivities of a cruise into areas where sets are com-
mon and areas where they are infrequent. The
major differences in the clusters with different
values for the clustering parameters result from
the merging of two relatively close clusters or the
inclusion of an isolated set or chase near the
boundary of a cluster (e.g., for 80% of the cruises,
the actual number of clusters decreases or re-
mains the same over a range of 50-150 miles
for the clustering parameter). However, the fact
that many of these descriptive statistics vary
continuously with the value of the clustering
parameter suggests that these defined clusters
do not represent distinct units, but areas of
high concentration in a continuously grading
system.
Cruises vary greatly with respect to the
amount of variability they exhibit in response to
changes in the value of the clustering parameter
(Table 2). Such variability is to be expected since
no single searching strategy is used by all vessels
and vessels may change their strategy during the
course of a cruise. In addition, the spatial distri-
bution of potential sets probably also varies with
time and space. These sources of variability
among cruises, combined with the relatively
small sample sizes within a cruise, may be part of
the reason that the descriptive statistics charac-
terizing clusters vary continuously with the value
of the clustering parameter.
The lack of any sharp demarcation in the clus-
ters as a function of the clustering parameter,
combined with the large amount of variability
exhibited among different cruises, creates a prob-
lem in presenting results based on the clustering
algorithm. Consequently, whenever summary
statistics are presented, results are given for a
range of values for the clustering parameter.
Table 2. — Examples of the effect of changes in the value of the clustering parameter for 5 arbitrarily selected
cruises.
Average
Percent of
cluster radius
Percent of
Number of
sets and
for clusters
the search-
Average
Value
clusters
chases with
with at least
ing within
Intercluster
of the
with
clusters with
3 members
clusters with
distance
cluster
at least
at least
(nautical
at least
(nautical
Cruise
parameter
3 members
3 members
miles)
3 members
miles)
1
50
4
90
32
61
171
75
1
96
142
91
298
100
1
96
142
91
298
150
1
96
142
91
298
2
50
10
85
27
34
210
75
7
97
61
60
294
100
4
97
105
68
397
150
4
98
126
71
443
3
50
5
30
27
9
402
75
5
31
36
16
450
100
5
31
46
19
479
150
5
60
68
36
614
4
50
4
78
30
36
425
75
4
80
39
41
465
100
4
82
38
45
513
150
4
82
38
45
513
5
50
1
65
45
38
302
75
3
79
45
43
362
100
3
79
45
43
362
150
3
79
45
43
362
359
FISHERY BULLETIN: VOL. 86, NO. 2
Whenever it seemed important to display the re-
sults by cruise, a value of 100 miles for the clus-
tering parameter was used. It should be empha-
sized that the general patterns appear to be
independent of the exact value of the clustering
parameter.
Is Searching DiflFerent in Clusters?
To determine whether vessels tended to in-
crease their searching effort in the vicinity of
sets, the ratio of the total distance travelled (i.e.,
the distance searched plus the distance run) to
the actual physical distances between sets was
compared between and within clusters. This ratio
is about 1.7 times greater in clusters than be-
tween clusters (Table 3). Also for no more than 7
cruises is the ratio within clusters less than the
ratio between clusters (Table 3). By a sign test
(Snedecor and Cochran 1967), these results imply
that this ratio is significantly greater than 1.00
(P < 0.005).
In addition, the proportion of the total distance
travelled that is devoted to searching is much
greater when a vessel is between than when it is
within clusters (Table 4). The fact that vessels
run, and are not actively looking for tuna, propor-
tionately more between than within clusters is an
indication that clusters are areas in which a ves-
sel has decided to remain. Vessels tend only to
run at night. Usually, as long as the vessel is
moving during daylight, the crew will be search-
ing. Thus, large amounts of running tend to occur
when vessels are actively moving to new areas.
Table 3. — Comparison between and within clusters of the ratio of the total distance
travelled (i.e., distance searched plus distance run) to the actual physical distance
between sets and chases. Means are the average values of the ratio for the 35 cruises
being considered.
Value of the cluster parameter
50
75
100
150
Between
Mean
1.79
1.70
1.65
1.59
SD
0.064
0.062
0.060
0.065
Range
1.24-2.70
1.13-2.71
1.12-2.68
1.11-2.81
Within
Mean
3.08
2.84
2.79
2.68
SD
0.343
0.254
0.250
0.239
Range
1.13-11.03
1.16-8.67
1.09-8.67
1.12-8.67
Number of cruises out
128
131
132
131
of 35 in which the ratio
is greater within than
between clusters
ip < 0.05 of observing this many ratio greater within clusters than between if they were in fact
equal based on a sign test (Snedecor and Cochran 1967).
Table 4. — Comparison between and within clusters of the ratio of the distance run without
searching to the distance searched. Means are the average values of this proportion for the
35 cruises being considered.
Value of the cluster parameter
50
75
100
150
Between
clusters
Mean
ratio
SE
Range
0.85
0.063
0.35-1.94
0.93
0.068
0.26-1.94
0.97
0.067
0.26-2.05
1.06
0.095
0.26-3.45
Within clusters
Mean
ratio
SE
Range
0.29
0.040
0.01-0.88
0.33
0.036
0.04-0.98
0.38
0.042
0.04-1.13
0.46
0.048
0.05-1.06
Number of cruises out
of a total of 35
in which the propor-
tion is greater be-
tween clusters than
within them
134
134
134
134
ip < 0 001 of observing the portion being greater between clusters than within 34 out of 35 cruises, if
in fact the proportion was equal based on a sign test (Snedecor and Cochran 1 967).
360
POLACHECK: DISTRIBUTION OF SEARCHING EFFORT
However, some running can still be expected
within clusters because vessels mark favorable
logs with radio transmitter in order to return to
them at a latter time. It should be noted that
three or more sets or chases rarely occur on the
same day (Fig. 8). Therefore, the proportionately
smaller amount of running that occurs within
clusters is not an artifact resulting from clusters
that do not span more than a single day.
Are Clusters Areas of
High Spotted Dolphin Densities?
Encounter rates {total number of sightings of
spotted dolphins divided by the total distance
searched) tend to be much greater within clusters
dominated by sets in association with dolphins
than either within clusters dominated by non-
dolphin sets or while searching between clusters
(Table 5). The clustering algorithm would tend to
ensure that such differences are likely to occur,
but the magnitude of the differences is large and
900
800
700
600
>
o
500
LLI
o
lu 400
a.
300
200
100
° 1 2 3 4 5
NUMBER OF SETS OR CHASES
Figure 8. — The frequency distribution of the number of sets
or chases during a day for all days on which there was at
least one chase or set.
Table 5. — Comparison of encounter rates (number of schools of spotted dolphins, Stenella
attenuata, per 100 miles searched) while searching within clusters classified according to the
percentage of sets involving dolphins, and while searching between clusters. Unweighted
means are the average value of the encounter within each cruise. Weighted means are based
on the encounter rate weighted by the distance searched by a cruise and are equivalent to the
total number of sightings divided by the total distance searched pooled across all vessels.
Within
clusters in which the percent
of the cluster parameter
of sets on dolphins ranged
from
Between
Value
0-25
25-75
75-100
clusters
50
Unweighted mean
0.17
1.45
2.74
0.52
SE
0.070
0.182
0.219
0.046
Weighted mean
0.10
1.04
2.46
0.50
SE
0.047
0.120
0.205
0.046
Number of cruises
27
22
30
35
Total distance searched
17.9
10.0
23.1
148.2
(thousands of miles)
75
Unweighted mean
0.15
1.32
2.32
0.47
SE
0.051
0.191
0.188
0.044
Weighted mean
0.12
0.88
2.16
0.46
SE
0.039
0.111
0.166
0.044
Number of cruises
25
22
31
35
Total distance searched
28.2
15.1
33.0
125.4
(thousands of miles)
100
Unweighted mean
0.14
1.15
2.08
0.46
SE
0.046
0.181
0.174
0.045
Weighted mean
0.15
0.81
1.82
0.45
SE
0.043
0.117
0.148
0.044
Number of cruises
25
20
31
35
Total distance searched
35.1
16.0
45.0
106.7
(thousands of miles)
150
Unweighted mean
0.10
0.73
1.83
0.43
SE
0.040
0.106
0.159
0.049
Weighted mean
0.11
0.68
1.63
0.40
SE
0.037
0.096
0.114
0.041
Number of cruises
24
18
32
35
Total distance searched
33.9
27.2
57.1
85.7
(thousands of miles)
361
FISHERY BULLETIN: VOL. 86, NO. 2
relatively insensitive to the value of the cluster-
ing parameter. However, the fact that a large per-
centage of the clusters tend to be dominated by
either dolphin or non-dolphin sets is not a neces-
sary consequence of the clustering algorithm and
suggests that the two types of methods for locat-
ing and catching tuna tend to be spatially and
temporally distinct. Encounter rates are substan-
tially lower in clusters dominated by non-dolphin
sets than when a vessel is searching between
clusters (Table 5). This result is also not a neces-
sary consequence of the clustering procedure and
suggests that these clusters not only define areas
of high densities of spotted dolphin schools but
also areas of low densities.
Encounter rates could be lower in non-dolphin
areas because of differences in detectability not
related to the density of schools. For example,
Hammond'* suggested that the crew may scan
closer to the vessel when searching in non-
dolphin areas (see also Polacheck 1983). It seems
unlikely that such factors could account for all of
the differences between the encounter rates in
Table 5.
Differences in detectability due to differences
in weather conditions between and within clus-
ters could also affect the results in Table 5. En-
counter rates do decrease at higher Beaufort sea
states (Polacheck 1983). However, little search-
ing occurs above Beaufort state 4. The difference
in encounter rates at Beaufort 0-2 compared with
Beaufort 3-4 (about a factor of 1.28) is insuffi-
cient to explain the difference in Table 5
(Polacheck 1983). Moreover, areas of non-dolphin
sets tend to be in nearshore areas with calmer
seas and fishermen do not consider Beaufort 4
conditions as being too rough to fish.
Do Tuna Catches Differ Between
and Within Clusters?
The average tons of tuna caught per set tend to
be greater for sets which occur within clusters
than sets between clusters (Fig. 9). For all values
of the clustering parameter, the average catch per
set was greater within than outside of clusters in
approximately 70% of the cruises (Table 6). Para-
metric statistical comparisons of the average tons
per set are not appropriate because of the large
differences in the average catch per set among
vessels (Fig. 9). A nonparametric sign test (Sned-
cor and Cochran 1966) suggests that the differ-
ences in catch per set between and within clusters
are significant at least at the 0.05 probability
level for all values of the clustering parameters
that were considered.
60 t-
56
CO
QC
liJ
t-
V)
u
_l
o
z
UJ
UJ
I-
LLI
UJ
Ui
QC
LU
a
I-
X
o
<
O
</)
z
o
I-
z
<
LU
2
36
\
•
32
-
/
/
/
/
28
/
/
/
/
• /
24
/
/
/
/
20
•
/
/•
• /
/
16
•
/
/
/
/
/
12
• •
/
• /
/
/
• •
•
8
"
/
/
/
•
•
•
/
•
•
4
/
t
•
/
•
•
/ •
, • •
•
0
(■ 1
- i
1 1 1 1 1
0 4 8 12 16 20 24 28 32
MEAN TONS CAUGHT PER SET WITHIN CLUSTERS
Figure 9. — The average tons of tuna caught per set while a
vessel was travelling between clusters versus the average tons
per set within clusters with at least three members. The dashed
line represents the expected value if the catch rates were equal.
The value of the cluster parameter equals 100.
Table 6. — The ratio of the tons of tuna caught per set within clus-
ters with at least 3 members compared to tons caught per set
between clusters. The means are the average values for the ratio
within a cruise. Cruises in which 100% of the sets were within
clusters are not inlcuded in the results.
■♦Hammond, P. S. 1981. Some problems in estimating the
density of dolphin populations in the eastern tropical Pacific
using data collected aboard tuna purse seiners. Inter-Am.
Trop. Tuna Comm. Intern. Rep.
Value of the cluster parameter
50 75
100 150
Mean
173 2.02
298 2.76
SE
0.22 0.379
0.476 0.499
Range
0.41-6.83 0.23-11.18
0.39-11.68 0.43-8.43
n
35 34
34 25
Number of
125 125
125 119
cruises in
which ratio
was greater
than 1
' P < 0 05 of getting the observed number if the ratio equaled 1 by a sign test
(Snedecor and Cochran 1966).
362
POLACHECK: DISTRIBUTION OF SEARCHING EFFORT
Does Searching Behavior Bias
the Overall Estimate of
Dolphin Encounter Rates?
When the two different methods of estimating
the overall encounter rates within a cruise are
compared, the adjusted encounter rates tend to be
smaller (Table 7). However, the differences are
not large, not because the biases are necessarily
small, but because positive and negative effect of
concentrating searching effort tend to cancel each
other. Positive biases would be expected in the
unadjusted rate due to concentrating of searching
effort in clusters dominated by sets made in asso-
ciation with dolphins, since these appear to be
areas of high dolphin densities (Table 5). Simi-
larly, negative biases would be expected due to
searching in non-dolphin clusters. For the data
considered here, substantial and roughly equal
amounts of searching occurred in both types of
clusters. The effects of concentrating searching
effort within the two types of clusters tended to
cancel each other. This suggests that the major
effect of nonrandom searching on measures of
relative abundance of dolphins will change in re-
lationship to the importance of dolphin and non-
dolphin fishing. Such changes tend to occur with
changes in the relative abundance of large yel-
lowfin tuna compared to skipjack and small yel-
lowfin tuna.
DISCUSSION
The fundamental question in interpreting the
results of this paper is whether the clusters that
have been defined bear any relationship either to
the searching strategy of the vessels or to the
underlying distribution of dolphin and tuna
schools. Two factors hamper answering this ques-
tion: The first is the large variability among ves-
sels; the second is the lack of appropriate null
hjqjotheses by which to test the results. The large
variability is to be expected and is inescapable.
Not only is there a large stochastic element in the
catch and encounter process, but large differences
can be expected in searching strategies among
vessels. Thus, a large range exist in skill and
experience among fisherman. In addition, the
amount of information, which is shared among
vessels varies and some vessels may be acting as
scouts for other vessels (Orbach 1975).
The specification of null hypotheses is difficult
because an infinite number of searching models
are compatible with the definition of random
search (i.e., defining a random search as one in
which the search path is independent of the dis-
tribution of the objects being sought). In order to
actually model a random searching process, the
probability of changing the direction of the search
path needs to be specified. A random search could
encompass an)^hing fi-om Brownian motion to
random straight line crossings of an area. With a
finite amount of searching, these will not neces-
sarily yield the same results.
Most of the results from this paper are compat-
ible with a model of clustered searching for clus-
tered prey, and some of them seem hard to explain
unless the searching and the schools of dolphins
are nonrandomly distributed. That the distance
travelled relative to the actual distances between
sets tends to be greater in areas where the density
of sets is high, as is the proportion of this distance
which is spent searching, are unlikely results un-
less searching is concentrated in these areas.
Also, the higher encounter rates in clusters dom-
inated by dolphins, the comparisons of the dis-
tances and frequency of the nearest set with the
preceding or next set, and the high percentage of
sets which fall into clusters are results that would
be expected if searching, tuna, and dolphins were
Table 7. — Comparison between the unadjusted overall encounter rate (nunnber of
schools of Stenella attenuata per 1 00 miles of searching) for a cruise and the encounter
rate adjusted for possible bias due to the concentration of searching within clusters.
Number of cnjises
out of 34 in which
the corrected esti-
Cluster
mate was less than
parameter
Mean
SE
Range
the uncorrected one^
Unadjusted
0.80
0.078
0.00-1.77
Adjusted
50
0.73
0.074
0.00-1.66
27
75
0.74
0.074
0.00-1.60
24
100
0.75
0.077
0.00-1.73
21
150
0.76
0.078
0.00-1.78
21
'One cruise had no recorded sighting of spotted dolphin schools.
363
FISHERY BULLETIN: VOL 86, NO. 2
spatially clustered. Perhaps the most surprising
result in this context is that the average catch per
set tends to be greater within the defined clusters
than outside of them. There is nothing in the clus-
tering algorithm that would tend to produce this
result. If these larger catches per set reflect large
schools of tuna, this argues that these defined
clusters are areas where tuna tend to concentrate
and not just areas of high densities of sets that
could be found in any random distribution.
Within many of the cruises, the locations of two
or more of the defined clusters overlap spatially
(Fig. 10), indicating that vessels often return to
an area after a period of searching elsewhere. In
addition, clusters from different cruises overlap
spatially and temporally. This overlapping of
clusters indicates that the overall searching be-
havior is even more nonrandom than the results
from this paper suggest. The overlapping both
within and among cruises is a dimension that
should be considered in future extensions to the
present work.
Orbach (1975), in a nonquantitative, anthropo-
logical study of the purse seine fishery, included
a general qualitative description of the searching
behavior of the fishermen which supports many of
Figure 10. — Examples of the spatial relation between clusters for the two cruises depicted in Figure 2. Open circles indicate the
position of all clusters with at least three members. The radius of each circle is scaled to the estimated radius of a cluster. The
associated numbers are the number of sets and chases within a cluster. Solid circles indicate the position of clusters with only two
members. Isolated sets are indicated by an x. The value of the clustering parameter equals 100. Arrows indicate the order of
movement between clusters. Note A and B are drawn to different scales. Distance between tick msirks equals 300 miles. No geographic
coordinates are given and the orientations were rotated at random so as not' to compromise any proprietary fishing information.
364
POLACHECK: DISTRIBUTION OF SEARCHING EFFORT
the quantitative results in this paper. He stated
that fishermen perceive two kinds of areas: The
first, in which fishermen refer to their activity as
"scratching", are regions where isolated schools of
fish are encountered; the second, which fishermen
refer to as an "area", are portions of the ocean
where schools of tuna are congregated. He also
reported that isolated tuna schools tend to be
small. Fishermen prefer to concentrate in an
"area" and actively search for them. I did not be-
come aware of this study by Orbach until the final
calculations of this paper were completed. Thus,
his description is an independent indication that
the results are not an artifact of the clustering
algorithm.
The method used for estimating the location
and size of a cluster could be refined. Such refine-
ments were beyond the scope of the present work
and would not affect any of the main results.
However, such refinements might be important if
the method used for calculating the adjusted en-
counter rates (Table 7) was used to develop rela-
tive abundance indices for dolphins in the ETP.
As pointed out by one reviewer, the problem of
estimating the shape, size, and location of a clus-
ter is analogous to the problem of determining the
home range for an animal from a set of observed
positions over time (Sanderson 1966; Cooper
1978; Schoener 1981; Swihart and Slade 1985)
and is part of the more general problem of how to
estimate the limits and size of clusters from any
clustering algorithm. The methods used for esti-
mating home ranges cannot be directly applied to
the tuna boat observer data but might provide a
basis for developing a better estimator for these
statistics relating to cluster size in any extension
to the present work.
To the extent that the results from this paper
indicate that searching is clustered, they suggest
that when these data are used for estimating rel-
ative densities of dolphin schools, or when catch
and effort data from purse seiners are used to
assess tuna stocks, rather fine geographic stratifi-
cations are needed to avoid biases from the non-
random searching within a cruise. The estimates
of the cluster radius could be considered as a
guide to appropriate levels of stratification. For
example, from 31 to 46% of all clusters with at
least three members had a radius of less than 60
miles (Fig. 11). This suggests that at a minimum,
50 r
40-
5 30
Ui
u
oc
111
Q. 20
10 -
1
A
N=230
_
t t
1 1 1 1 1
50
40 -
30 -
20 -
B
N=230
-
1
1 1 1 1 1
30
60
90
120
150
180 >180
30
60
90
120
150
180 >180
40
30
lU
O 20
c
10
C
N=214
X
_L
X
X
_r
30 60 90 120 150 180 >180
CLUSTER RADIUS (nautical mllaa)
40r-
30-
20 -
10 ~
-
]
D
N=184
1
1
1
gj;;:
1
1 t
1
I
30 60 90 120 150 180 >180
CLUSTER RADIUS (nautical miles)
Figure 11. — The frequency distribution for estimated radii for all clusters with at least three members. Estimates from
all cruises have been pooled. The values of the clustering parameter are A) 50, B) 75, C) 100, and D) 150.
365
FISHERY BULLETIN. VOL. 86, NO. 2
a 2° stratification would be necessary to avoid
biases in these areas.
ACKNOWLEDGMENTS
I wish to thank Peter Frank and Tim Smith for
their generous support and useful comments dur-
ing the research presented here. This paper would
not exist without the data collected by the Na-
tional Marine Fisheries Service observers. I wish
to acknowledge these observers and the staff of
the Southwest Fisheries Center of the National
Marine Fisheries Service for collecting, editing,
and initially preparing these data. A number of
the staff of the Southwest Fisheries Center pro-
vided support and technical assistance at various
stages in this research. The number of people in-
volved is too large to acknowledge all of them
individually. Special thanks go to Phil Tarentino,
Michael Lichter, and Susan Boyer.
The art department of the Southwest Fisheries
Center were responsible for the final preparation
of some of the figures. Veronica van Kouwen
helped with the preparation of the final
manuscript.
The manuscript is a portion of a Ph.D. Thesis
submitted to the University of Oregon, Eugene,
OR.
LITERATURE CITED
Allen, R., and R. Punsley.
1984. Catch rates as indices of abundance of yellowfin
tuna, Thunnus albacares , in the eastern Pacific Ocean.
Inter-Am. Trop. Tuna Comm. 18:303-379.
Blackburn, M., and F. Williams.
1975. Distribution and ecology of skipjack tuna, Katsu-
wonus pelamis , in an offshore area of the eastern tropical
Pacific. Fish. Bull., U.S. 73:382-411.
Calkins, T. P.
1963. An examination of fluctuations in the "concen-
tration index" of purse seiners and bait boats in the fish-
ery for tropical tuna in the eastern Pacific. Inter-Am.
Trop. Tuna Comm. Bull. 8:255-316.
Calkins. T. P., and B M Chadwin.
1967. Geographical distribution of yellowfin tuna and
skipjack catches in the eastern Pacific Ocean, by quarters
of the year 1963-1966. Inter-Am. Trop. Tuna Comm.
Bull. 12:435-508.
1971. Geographical catch of yellowfin and skipjack tuna
in the eastern tropical Pacific Ocean, 1967-1970, and
fleet and total catch statistics. Inter-Am. Trop. Tuna
Comm. Bull. 15:285-377.
Clark. C. W., and M. Mangel.
1983. Search theory in ecology and resouce manage-
ment. In M. Freedman (editor). Proceedings of the In-
ternational Conference on Population Biology, p. 380-
388. Springer Lecture Notes in Biomathematics, Vol.
52. Springer Verlag, N.Y.
Cooper. E
1979. Home range criteria based on temporal stability of
areal occupation. J. Theor. Biol. 73:687-695.
GULLAND, J. A
1955. Estimation of growth and mortality in commerical
fish fxipulations. Fish. Invest. Minist. Agric. Fish. Food
(G.B.), Series 2. 18:1-46.
Helland Hansen, B
1909. Statistical research into the biology of the haddock
and cod in the North Sea. Rapp. P-v R6un 10:1-62.
Koopman, B O
1980. Search and screening. Pargamon Press, N.Y.,
509 p.
Orbach, M K
1975. The cultural system of the tuna seinermen of San
Diego, California. Ph.D Thesis, Univ. California, San
Diego, CA, 411 p.
Pazynich, G I
1966. Primery primeneniya teorii statisticheskiskh resh-
enii k zadacham taktiki promysla (Examples of the appli-
cation of statistical-solution theory to problems in the
fishing tactics). Noe Khoz. 6:80-81. (Translation
available in the library of National Marine Fisheries
Service Northeast Fishery Laboratory, Woods Hole, MA.)
Perrin, W. F.
1968. The porpoise and the tuna. Sea Frontiers 14:166-
174.
1969. Using porpoise to catch tuna. World Fishing
18:42-45.
POLACHECK, T
1983. The relative abundance of dolphins in the eastern
tropical Pacific based on encounter rates with tuna purse
seiners. Ph.D Thesis, Univ. Oregon, Eugene, OR, 444 p.
Sall\, S. B., AND J M. Flowers.
1979. Elementary applications of search theory to fishing
tactics as related to some aspects of fish behavior.
F.A.O. Conference on Fish Behavior in Relation to Fish-
ing Techniques and Tactics, p. 343-355.
Sanderson, G.
1966. The study of mammal movements - a review. J.
Wildl. Manage. 30:215-235.
SCHOENER, T W
1981. An empirically based estimate of home range.
Theor. Popul. Biol. 20:281-325.
Snedecor, G W , AND W C. Cochran
1967. Statistical methods. 6th ed. Iowa State Univ.
Press, Ames, 593 p.
Suzuki, Z , P K. Tomlinson, and M Honma.
1978. Population structure of Pacific yellowfin tuna.
Inter-Am. Trop. Tuna Comm. Bull. 17:277-411.
SwiHART, R. K . and N. a. Slade.
1985. Influence of sampling interval on estimates of home
range size. J. Wildl. Manage. 49:1019-1025.
366
A COMPREHENSIVE THEORY ON THE ETIOLOGY OF BURNT TUNA
Cheryl Watson,' Robert E. Bourke,^ and Richard W Brill^
ABSTRACT
Over the past 14 years, the Hawaii handline fishery has experienced phenomenal growth in the catch
for large yellowfin tuna, Thunnus albacares, and bigeye tuna, Thunnus obesus. These fish are
primarily caught for the sashimi (raw consumption) market but have been continually plagued with
a product quality problem known as "burnt tuna" or, in Japanese, yake niku. Not only does this
problem significantly reduce the value of the catch, it also limits export markets and expansion of this
low-capital, high-return fishery to other areas of the Pacific. Previous research and suggestions for
mitigating burnt tuna have centered on the hypothesis that it is caused by high muscle temperature
and low pH, which is the result of a violent struggle during capture.
A new, more comprehensive hypothesis is presented: Burnt tuna is actually caused by the p)ost-
mortem activation of enzymes known as calcium-activated proteases and by the enhancement of the
effect of these enzymes by high blood catecholamine levels. Previously unexplainable observations,
such as the propensity of female fish to become burnt more often during the summer months, the
efficacy of brain destruction in preventing burnt tuna, and the lack of effect of cooling on the incidence
of burnt tuna, are explainable in light of this new hypothesis.
One of the largest fisheries in Hawaii is the hand-
line fishery for large (>50 kg) yellowfin tuna,
Thunnus albacares, and bigeye tuna, T. obesus,
caught primarily for raw consumption as sashimi.
Yearly landings increased from 89 short tons (ex-
vessel value, $131,000) in 1973 to 615 short tons
(ex-vessel value, $2.1 million) in 1984 (Yuen
1979; Hudgins and Pooley 1987). The total eco-
nomic value of the fishery has been estimated as
high as $5 million yearly (Ikehara'*). In Hawaii,
the night handline fishery is known as ika shibi
from the Japanese words for squid and tuna, and
the daytime fishery is known as palu ahi from the
Hawaiian words for chum and yellowfin tuna.
There is also growing international interest in
this tjTDe of fishing because of its low initial capi-
tal investment, low operating and fixed expenses,
strong export markets, and high profitability
(Strong 1979; Gibson 1981; Jerrett 1984). Boats
can be as small as 6 m and require only one- or
two-man crew. Catch rates in Hawaii have
ranged from two fish per hook per night (Yuen
'Pacific Gamefish Research Foundation, 74-425 Kealakehe
Parkway, #15, Kailua-Kona, HI 96740; present address: De-
partment of Physiology, John A. Bums School of Medicine, Uni-
versity of Hawaii, Honolulu, HI 96822.
2465B Kawailoa Road, Kailua, HI 96734.
3Southwest Fisheries Center Honolulu Laboratory, National
Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu,
HI 96822-2396.
^Ikehara, W. N. 1981. A survey of the ika-shibi fishery in
the State of Hawaii, 1980. Southwest Fish. Cent. Adm. Rep.
H-82-4C, 12 p. Southwest Fisheries Center Honolulu Labora-
tory, National Marine Fisheries Service, NOAA, Honolulu, HI
96822-2396.
1979) to one-half fish per hook per night
(Bourke^), illustrating the profitability for the in-
dividual fisherman.
Unfortunately, handline (as well as primarily
recreational trolling) fishermen are plagued by a
product quality problem known as "burnt tuna"
or, in Japanese, as yake niku (literally translated
as "cooked meat"). When fish are intended for raw
consumption, product quality is of utmost impor-
tance. Prime quality tuna flesh should be red,
translucent, and firm and have a delicate flavor.
Burnt tuna is pale, exudes a clear fluid, and has
a soft texture and a slightly sour taste. Although
perfectly palatable when cooked or canned, burnt
tuna is considered unsuitable for raw consump-
tion and commands only a fraction of the price of
prime quality fish. Fish are usually exported
whole to preserve freshness, so burnt tuna often is
not detected until shipping costs have been in-
curred. This discourages exports of tuna caught
from areas or via fishing techniques with high
incidences of the problem.
Burnt tuna aflects from 5 to 100% of the tissue
from an individual fish and ranges from mild to
severe. Approximately 25% of the fish caught by
the Hawaii handline fishery are burnt, as are
50% of the large yellowfin tuna caught by com-
mercial and recreational trollers (Bourke fn. 5).
^Bourke, R. E. 1985. Hilo ikashibi fishery 1984 survey:
Problems of a maturing fishery and a potential solution to the
burnt tuna problem. Poster presentation at 1985 Tuna Confer-
Manuscript accepted January 1988.
fishery BULLETIN: VOL. 86, NO. 2, 1988.
367
FISHERY BULLETIN: VOL. 86, NO. 2
Annual losses to the handline fishery are an esti-
mated 169^ of the value of the total catch (Cramer
et al. 1981). The problem has also been reported
in fish intended for raw consumption caught by
purse seine (Nakamura et al. 1977). Surprisingly,
the problem rarely occurs in fish caught by
longlining (Williams 1986).
PAST RESEARCH ON BURNT TUNA
The Japanese were the first to investigate the
causes of burnt tuna and possible mitigating
strategies (Itokawa 1968, 1969). The first con-
trolled laboratory investigations were those of
Nakamura et al. (1977) and Konagaya and Kona-
gaya (1978, 1979). Nakamura et al. (1977) con-
cluded that high muscle temperature and low
muscle pH caused myofibrillar protein denatura-
tion and also noticed that, once denaturation
began, it continued even if the tissue was kept at
0°C. Because of the relatively high thermostabil-
ity of tuna myofibrillar protein and because yake
niku occurs in species (e.g., frigate mackerel and
sardine) that do not generate high muscle tem-
peratures during struggling, Konagaya and Kon-
agaya (1978) concluded that acid denaturation of
myofibrillar proteins at moderate temperatures
was the underlying cause.
Cramer et al. (1981) studied handline-caught
yellowfin tuna in Hawaii and found that the oc-
currence of burnt tuna did not correlate with
muscle temperature at time of landing and corre-
lated only loosely with extracellular muscle pH.
Ikehara^ conducted an engineering study to de-
velop methods to cool large yellowfin tuna more
rapidly, the presumption being that rapid cooling
would prevent muscle degradation. Although suc-
cessful in developing a technique to increase cool-
ing of deep muscle temperature, as shown in Fig-
ure 1, there was no apparent correlation between
rate of cooling and incidence of burnt flesh. In
spite of this lack of directly observed correlation,
some publications designed for fishermen still
stress that high muscle temperatures and low
muscle pH are the prime causes of burnt tuna
(Gibson 1981). Others have expressed doubt as to
the validity of this hypothesis (Jerrett 1984).
The high muscle temperature-low pH hypothe-
sis appears to fit with what is known about tuna
physiology, in that these fishes are capable of pro-
ducing muscle temperatures significantly above
ambient (Carey et al. 1971; Carey 1973) and ex-
hibit some of the highest rates of muscle glycoly-
sis (production of muscle lactate and concomitant
production of acidity) observed in nature
(Hochachka et al. 1978; Hochachka and Momm-
sen 1983). Yet some observations do not fit this
hypothesis. For example, burnt tuna occurs more
frequently in summer, more frequently in female
fish, and more frequently in fish fought for short
periods of time than in fish fought for long periods
(>7 minutes or <2 hours) (Davie and Sparksman
1986; Nakamura^). Furthermore, burnt tuna oc-
curs rarely in longline-caught fish and in fish
subjected to brain or spinal column destruction
immediately following capture (Nakamura et al.
1977; [Suisan Sekai] 1977; Cramer et al. 1981;
Davie and Sparksman 1986; Nakamura fn. 7).
The hypothesis that burnt tuna is caused by high
muscle temperature and low muscle pH does not
seem to directly fit with any of these observa-
tions.
A NEW ANALYSIS OF THE BURNT
TUNA PROBLEM
At the biochemical level, the high muscle
temperature-low pH h3rpothesis would predict
that the observed drop in extracellular pH would
be accompanied by a similar drop in intracellular
pH and activation of lysosomal proteases. These
proteases would then degrade actin and myosin,
the dominant muscular proteins, resulting in the
'^R. Nakamura, Department of Animal Sciences, University of
Hawaii, Honolulu, HI 96822, pers. commun. October 1987.
o
LJ
I—
<
a:
o
I/)
Q.
LiJ
□ UKN 1
30.0
^^=!5--
NOT BURNT
^^>
"~^^^^^
•-«
.^^_^^6.2)
20.0
\
s
s
\
~~^~~^ (64.4)
(M.9)
\s._^ N.
(103.0) ^ - ^
~(62.1)
"(61.7)
10.0-
■
X,,^^
9.0
8.0
(99.8 )
7.0-
1
— 1 1-
— 1 1
H 1 1 1 1
10 11
ELAPSED TIME (h)
^Ikehara, W. N. 1981. Development of a small-boat chill-
ing system for the reduction of burnt tuna. Final Report for a
Pacific Tuna Development Foundation contract, var. pag.
Figure 1. — Data are from Ikehara (text fn. 6). Initial and final
deep muscle temperatures are plotted on semilogarithmic axes
to linearize the rate of temperature change. Numbers in paren-
theses are the body weight (kg) for each fish.
368
WATSON ET AL : ETIOLOGY OF BURNT TUNA
undesirable changes seen in burnt tuna. The lyso-
somal proteases, specifically cathepsins B, D, and
L, have an optimum pH of about 5 (Dahlmann et
al. 1984). However, recent work by Abe et al.
(1985, 1986) has shown that, intracellularly, tuna
muscle is well buffered and that a fall in extracel-
lular pH is not necessarily accompanied by an
equivalent fall in intracellular pH. Using is-
chemic rat gastrocnemius muscle, Hagberg
(1985) found that a 1 pH unit drop (7.30-6.36) in
extracellular pH was accompanied by a drop in
intracellular pH of only 0.4 unit (7.00-6.60).
Thus, the acidic intracellular environment that
would favor the action of the lysosomal proteases
probably is not present in burnt tuna. Also, with
the exception of calcium-activated neutral
protease, all other muscle proteases (cathepsins
B, D, L; alkaline serine protease; neutral trypsin-
like protease; and alkaline cysteinyl protease) de-
grade myosin (Dahlmann et al. 1984). Yet
Hochachka and Brill (1987) found burnt muscle
had no increase in 3-methyl-histidine, a specific
marker for myosin breakdown. Decomposition of
other myofibrillar proteins is, therefore, impli-
cated.
Electronmicrographs of postmortem burnt tuna
muscle (Davie and Sparksman 1986) showed a
consistent, rapid disintegration of Z-discs and ir-
regularities in the sarcoplasmic reticulum (SR).
The changes in burnt muscle were not different in
kind from postmortem changes seen in unburnt
tissue, but were a result of a significant increase
in the rate of disintegration. Selected destruction
of the Z-discs, troponin and tropomyosin, and the
SR is characteristic of a pair of proteases known
as calcium-activated neutral proteases (CANP's)
(Sugita et al. 1984; Suzuki et al. 1984). These
proteases are cytoplasmic, ubiquitous, activated
by increased intracellular calcium levels, and ac-
tive at pH 5.5-8.0 (Sakamoto and Seki 1985;
Koohmaraie et al. 1986; Seki and Kimura 1986;
Zeece et al. 1986b). The intracellular pH most
likely found in burnt tuna muscle is, therefore,
more consistent with CANP action than with
lysosomal proteases whose activity requires a pH
closer to 4.5 (Hochachka and Brill 1987). In addi-
tion, while cathepsin D's activity is greatly re-
stricted at 15°C, CANP is still active at 5°C
(Koohmaraie et al. 1986; Zeece et al. 1986a).
A new etiology of burnt tuna proposed by
Hochachka and Brill (1987) is summarized in Fig-
ure 2. Their hypothesis predicts that low intracel-
lular ATP concentrations lead to the leaking of
Ca"^^ into the cell and increases in intracellular
INTENSE MUSCLE ACTIVITY
I
LACK OF 02 + ATP
i
METABOUC COLLAPSE OF MEMBRANE
i
RISE IN INTRACELLULAR CALCIUM
i
02
i
PSE
i
LLU
i
IN 0
i
ACTIVATION OF CANP
i
UNDESIRABLE MUSCLE TEXTURE
Figure 2. — Biochemical reac-
tions involved in the development
of burnt tuna as proposed by
Hochachka and Brill (1987).
Ca"^^ concentrations. These increases, in turn, ac-
tivate CANP, which specifically attacks troponin,
tropomyosin, SR, and mitochondria. The break-
down of the latter two intracellular organelles
releases more calcium into the cytoplasm, thus
further increasing the activity of CANP.
The effect of brain and spinal cord destruction
on reducing the incidence of burnt tuna and of
similar muscle degradation seen in other fish spe-
cies can be explained by this new hypothesis,
which assumes the initial drop in ATP is the root
cause of the elevated intracellular calcium
(Amano et al. 1953; Fujimaki and Kojo 1953;
Konagaya and Konagaya 1978; Ikehara fn. 6).
Brain destruction maintains elevated muscle
ATP levels after capture (Boyd et al. 1984). Data
recently collected on the use of brain and spinal
cord destruction in large yellowfin tuna (Fig. 3)
40
I-
I 35
m
g 30
m
I 20
o
Sl5
z
o
UJ
Q.
10--
5-
1 4 FISHERMEN NOT USING BRAIN DESTRUCTION
A
--
^ ^\
/\
SINGLE FISHERMAN
^
■^
\
/ \
^
r
/ \
1
\ r~^^^^ \
1
\ / \
1
\ / \
1
\ 1 \
1 l\
\ J \
1 / \
\ 1 \
/ 1 ^
\
\ 1 \
1
\
--^ \
1
/ 1
U
"^
"■--. ^' \
1 I
BRAIN DESTRUCTION USED \
1
1 /
\
(■ 1 1—
+-
H
1 1 1 1 1 \
JAN
MAR
MAY JUL
MONTH
SEP NOV
Figure 3. — Incidence of burnt tuna occurring in large yellowfin
tuna caught by 14 commercial handline fishermen operating
out of Hilo, HI, January-September 1984, and by 1 fisherman
who began using brain destruction in May. His incidence offish
becoming burnt dropped dramatically in spite of the normal
summer increased incidence seen in the catch of 14 other fisher-
men.
369
FISHERY BULLETIN: VOL. 86, NO. 2
clearly show the effectiveness of this technique in
preventing burnt tuna. A drop from 30 to < 10% of
the tuna becoming burnt was noted when the
fisherman changed his killing technique from
shooting to the use of a brain spike.
A MORE COMPREHENSIVE
HYPOTHESIS ON THE ETIOLOGY
OF BURNT TUNA
The hypothesis presented by Hochachka and
Brill (1987) does not explain all the observed fac-
tors leading to tunas' different propensity to be-
come burnt. It cannot explain why tunas strug-
gling for <7 minutes rarely become burnt
(Nakamura fn. 7), nor can it account for the obser-
vations of Davie and Sparksman (1986) and
Bourke (unpubl. data), who found that tuna
struggling for extended (>1 hour) periods of time
have a lower probability of becoming burnt. It
also cannot explain why female fish become burnt
more often than males (Nakamura et al. 1977), or
why fish caught in summer become burnt more
often (Fig. 3).
Our revised hypothesis is presented in Figure
4. We have incorporated into the Hochachka and
Brill (1987) hypothesis, the action of the neuro-
transmitters-hormones norepinephrine (NE) and
epinephrine (E) (collectively referred to as cate-
cholamines). Periods of intense physical activity
or capture stress increase the levels of circulating
catecholamines by about 10-200 times in fish
other than tunas (Nakano and Tomlinson 1967;
Ling and Wells 1985). Most likely a similar situa-
tion occurs in tuna during hooking, fighting, and
capture. The importance of elevated circulating
catecholamines in this particular schema comes
INTENSE MUSCLE ACTIVITY
i
LACK OF 02 + ATP
i
METABOUC COLLAPSE OF MEMBRANE
i
RISE IN INTRACELLULAR CALCIUM
i
ACTIVATION OF CANP
STRESS OF CAPTURE
NE & E RELEASE
■INCREASE OF CANP ACTION ■
from their potentiating effect on the action of
CANP. Catecholamines cause the phosphoryla-
tion of troponin, resulting in a more rapid and
prolonged proteolysis of this muscle structure
(Toyo-oka 1982); in other words, high circulating
levels of catecholamines greatly increase the ef-
fectiveness of the enzymes responsible for a tuna
becoming burnt. ^
The gills are the organ primarily responsible
for the degradation of catecholamines, and circu-
lating catecholamines are rapidly cleared from
the blood (Nekvasil and Olson 1986). We estimate
circulating catecholamines in tuna have a half-
life of 30 minutes or less. Therefore, we hypothe-
size that, in a longline-caught fish remaining in
the water for several hours after hooking (Davie
and Sparksman 1986), plasma catecholamines
are reduced to low levels before the fish is landed
and killed, thus resulting in a low percentage of
burnt fish. Similarly, fish caught by rod and reel
or by handline and fought for several hours would
have the low circulating catecholamine levels and
low propensity to become burnt. Indeed, Gibson
(1981) recommends that handline-caught fish be
attached to a buoy and left for an hour prior to
being brought on board and killed, as a measure
to prevent burnt tuna. That this technique pre-
sumably lowers blood catecholamine levels prior
to death could explain its efficacy. On the other
hand, we hypothesize that a fish landed during
the peak of its blood catecholamine concentra-
tions would have the greatest propensity for be-
coming burnt. This indeed appears to be the case.
The mechanism that ties catecholamines to the
observed seasonality and increased number of fe-
male fish becoming burnt lies in the biochemical
structural similarity of catecholamines and re-
productive steroids, particularly estrogen. These
steroids reach a peak during spawning season,
which is May through October for yellowfin tuna
in Hawaii (June 1960). A corresponding increase
occurs in the percentage of the tuna catch that
becomes burnt during this season (Fig. 3). In the
presence of tyrosine hydroxylase, estrogen is con-
verted to catecholestrogen and consequently com-
petes for the same degradative enzymes in the
gills, therefore slowing the clearance of cate-
cholamines from the blood (Nekvasil and Olson
1986). Female fish, with high circulating estro-
gen levels, would then be expected either to reach
UNDESIRABLE MUSCLE TEXTURE
Figure 4. — Comprehensive theory of the etiology of burnt tuna
proposed in this paper.
^Although Cortisol is also known to be released during stress,
its mechanism of action is much slower than NE or E, making
it unlikely that it could exert catabolic effects within 15 min-
utes.
370
WATSON ET AL.: ETIOLOGY OF BURNT TUNA
higher maximum circulating catecholamine lev-
els or to maintain high levels for longer periods,
thus explaining the greater number of female fish
that become burnt and the seasonality of the oc-
currence of burnt tuna.
As pointed out by Davie and Sparksman (1986)
and Hochachka and Brill (1987), burnt tuna is a
quantitative change in tissue decomposition, not
a qualitative one. High blood catecholamine lev-
els at the time of death act as accelerators in this
inevitable metabolic cascade. However, because
we are dealing with a change in rate, the action of
catalysts can make all the difference as to
whether or not a fish becomes burnt due to CANP.
FUTURE RESEARCH
Under normal conditions, CANP is inhibited by
calpastatins. This class of proteins is distinct from
the other cysteine protease inhibitors, the cys-
tatins. Although the cystatins can inhibit cathep-
sins B and H, they are unable to alter CANP
activity. Conversely, calpastatin is only effective
against calcium-activated neutral protease (Bar-
rett et al. 1986; Parkes 1986). Unfortunately, al-
most nothing else is known about the relationship
of CANP to its endogenous inhibitor, either struc-
turally or physiologically. Certainly investiga-
tions into the role of calpastatin would prove
valuable for this research.
Immediate future research will concentrate on
tracking the specific action of CANP in burnt
tuna and attempting to stop this action with Ca^ ^
chelating agents such as EGTA (ethyleneglycol-
bis-(aminoethyl ether )-N,N,N',N'-tetraacetic
acid) or by use of its intracellular inhibitor, cal-
pastatin. Also, blood catecholamine levels of
stressed and unstressed fish will be measured,
along with metabolic clearance rates of nore-
pinephrine and epinephrine.
When viewed as a process of metabolic deregu-
lation of CANP, the rapid deterioration of tuna
muscle ceases to be an isolated muscular phe-
nomenon. The Z-disc disintegration characteris-
tic of burnt tuna is also present in cardiac muscle
injury due to ischemia and muscular dystrophy
(Sugita et al. 1984). Given the highly conserved
nature of muscle tissue, an understanding of
burnt tuna may also provide insights into the
metabolic processes of human disease.
ACKNOWLEDGMENTS
This research was supported by a grant to
Pacific Gamefish Research Foundation from the
Department of Land and Natural Resources, Di-
vision of Aquatic Resources, State of Hawaii.
LITERATURE CITED
Abe, H , R W Brill, and P W Hochachka.
1986. Metabolism of L-histidine, camosine, and anserine
in skipjack tuna. Physiol. Zool. 59:439-450.
Abe, H., G P Dobson, U Hoeger, and W. S. Parkhouse.
1985. Role of histidine-related compounds to intracellular
buffering in fish skeletal muscle. Am. J. Physiol.
249:R449-R454.
Amano, K., M Bito, and T Kawabata
1953. Handling effect upon biochemical change in the fish
muscle immediately after catch — I. Difference of gly-
colysis in the frigate mackerel killed by various meth-
ods. Bull. Jpn. Soc. Sci. Fish. 19:487-498.
Barrett, A J . N D Rawlings, M E. Davies. W. Machleidt,
G Salvesen, and V Turk
1986. Cysteine proteinase inhibitors of the cystatin super-
family. In A. Barrett and G. Salvesen (editors),
Proteinase inhibitors, p. 515-569. Elsevier Sci. Publ.
BV, North Holland.
Boyd, N S , N D Wilson, A R Jerrett, and B I Hall
1984. Effects of brain destruction on post harvest muscle
metabolism in the fish kahawai (Arripis trutta ). J. Food
Sci. 49:177-179.
Carey, F. G
1973. Fishes with warm bodies. Sci. Am. 228:36-44.
Carey, F G , J M Teal, J W Kanwisher, and K D Lawson.
1971. Warm-bodied fish. Am. Zool. 11:137-145.
Cramer, J L, R. M Nakamura, A E Dizon, and W N. Ikehara.
1981. Burnt tuna: Conditions leading to rapid deteriora-
tion in the quality of raw tuna. Mar. Fish. Rev.
43(6):12-16.
Dahlmann, B , L Kuehn, and H. Reinauer
1984. Proteolytic enzymes and enhanced muscle protein
breakdown. In W, H. Horl and A. Heidland (editors),
Proteases, p. 505-517. Plenum Press, N.Y.
Davie, P S , and R. I. Sparksman.
1986. Burnt tuna: An ultrastructural study of post-
mortem changes in muscle of yellowiin tuna iThunnus
albacares) caught by rod and reel and southern bluefin
tuna (Thunnus maccoyii) caught on handline or long-
line. J. Food Sci. 51:1122-1128, 1168.
FUJIMAKI, M , AND K KOJO
1953. Handling effect upon biochemical change in the fish
muscle immediately after catch — II. Changes of acid-
soluble phosphorus compounds of frigate mackerel mus-
cle. Bull. Jpn. Soc. Sci. Fish. 19:499-504.
Gibson, D J M
1981. A handbook on processing southern bluefin tuna for
the fresh chilled sashimi market in Japan. Minist.
Agric. Fish., Wellington, var. pag.
Hagberg, H
1985, Intracellular pH during ischemia in skeletal mus-
cle: Relationship to membrane {wtential, extracellular
pH, tissue lactic acid and ATP. Pfluegers Arch.
404:342-347.
Hochachka, P. W., and R W Brill
In press. Autocatalytic pathways to cell death: A new
analysis of the tuna bum problem. Fish Physiol.
Biochem. 3.
371
FISHERY BULLETIN: VOL. 86, NO. 2
HOCHACHKA, P W . W C HULBERT, AND M. GUPPY.
1978. The tuna power plant and furnace. In G. D. Sharp
and A. E. Dizon (editors), The physiological ecology of
tunas, p. 153-174. Acad. Press, N.Y.
HOCHACHKA, P. W., AND T. P MOMMSEN.
1983. Protons and anaerobiosis. Science (Wash. D.C.)
219:1391-1397.
HUDGINS. L L.. AND S. G. POOLEY.
1987. Growth and contraction of domestic fisheries:
Hawaii's tuna industry in the 1980s. In D. J. Doulman
(editor). Tuna issues and perspectives in the Pacific is-
lands region, p. 225-241. East-West Center Press, Hon-
olulu.
ITOKAWA, S
1968. On maintaining the freshness of tuna seine
catches. Reports of the Mie Prefecture Owase Fishery
Experiment Station, p. 1-15.
1969. On experiments in prevention of burnt flesh in yel-
lowfin tuna. Rep. Mie Pref Owase Fish. Exp. Stn. p.
1-9.
Jerrett. a R.
1984. Detection and prevention of burning in southern
bluefin tuna. Part A: Research results and discus-
sion. Fish Processing Bull. 4, 22 p. + app.
June, F C
1960. Spawning of yellowfin tuna in Hawaiian
waters. Fish. Bull., U.S. 54:47-64.
KONAGAYA, S., AND T KONAGAYA
1978. Denaturation at moderate temperatures of myo-
fibrillar protein of red-meat fish: A possible cause of
yake-niku. Bull. Tokai Fish, Res. Lab. 96:67-74.
1979. Acid denaturation of myofibrillar protein as the
main cause of formation of "yake-niku", a spontaneously
done meat, in red meat fish. Bull. Jpn. Soc. Sci. Fish.
45:245.
KOOHMARAIE. M . J. E SCHOLLMEYER, AND T R DUTSON.
1986. Effect of low-calcium-requiring calcium activated
factor on myofibrils under varying pH and temperature
conditions. J. Food Sci. 51:28-32, 65.
Ling, N., and R. M. G Wells
1985. Plasma catecholamines and erythrocyte swelling
following capture stress in a marine teleost fish. Comp.
Biochem. Physiol. 820:231-234.
NaKAMURA, K , Y FUJII. AND S ISHIKAWA
1977. Experiments on the prevention of "burning" of
tunas — L An examination of causes of occurrence. Bull.
Tokai Reg. Fish. Res. Lab. 90:39-43. (Engl, transl. by
W. G. Van Campen, 1980, 8 p., Transl. No. 46; available
Southwest Fish. Cent., Natl. Mar. Fish. Serv., NOAA,
Honolulu, HI 96822-2396.)
NaKANO. T . AND N TOMLINSON
1967. Catecholamine and carbohydrate concentrations in
rainbow trout iSalmo gairdneri ) in relation to physical
disturbance. J. Fish. Res. Board Can. 24:1701-1715.
Nekvasil, N P , AND K. R. Olson
1986. Extraction and metabolism of circulating cate-
cholamines by the trout gill. Am. J. Physiol. 250:R526-
R531.
Parkes, C.
1986. Calpastatins. In G. Barrett and A. Salvesen (edi-
tors). Proteinase inhibitors, p. 571-587. Elsevier Sci.
Publ. BV, North Holland.
Sakamoto, S . and N Seki
1985. Limited hydrolysis of carp myosin by cal-
pains. Bull. Jpn. Soc. Sci. Fish. 51:1551-1557.
Seki, N , and Y Kimura
1986. Degradation of troponin and tropomyosin by cal-
pain. Bull. Jpn. Soc. Sci. Fish. 52:1673-1680.
Strong. R D
1979. 1978-79 PTDF funded feasibility study to establish
an ika-shibi fishery on Guam — Final Report. Pac. Tuna
Dev. Found., May 1979, var. pag.
SuGiTA. H., S Ishiura. K Kamakura, H. Nakase, K. Hagiwara.
I Nonaka, and K Tomomatsu.
1984. Ca-activated neutral protease in physiological and
pathological conditions. In S. Ebashi, M. Endo, K. Ima-
hori, S. Kakiuchi, and Y. Nishizuka (editors). Calcium
regulation in biological systems, p. 243-256. Acad.
Press, Tokyo.
[Suisan Sekai]
1977. How to effectively kill tunas in order to maintain
quality and higher prices. Suisan Sekai 9:52-
57. (Engl, transl. by T. Otsu, 1978, 4 p., Transl. No.
25; available Southwest Fish. Cent. Honolulu Lab.,
Natl. Mar. Fish. Serv., NOAA, Honolulu, HI 96822-
2396).
Suzuki. K., S. Kawashima, and K. Imahori.
1984. Structure and function of Ca2+-activated
protease. In S. Ebashi, M. Endo, K. Imahori, S. Kaki-
uchi, and Y. Nishizuka (editors). Calcium regulation in
biological systems, p. 213-226. Acad. Press, Tokyo.
Toyo-OKA, T
1982. Phosphorylation with cyclic adenosine 3:5'
monophosphate-dependent protein kinase renders bovine
cardiac troponin sensitive to the degradation by calcium-
activated neutral protease. Biochem. Biophys. Res.
Commun. 107:44-50.
Williams, S C
1986. Marketing tuna in Japan. Queensl. Fish. Ind.
Train. Comm., Brisbane, Queenl., Aust., 60 p.
Yuen, H S H
1979. A night handline fishery for tunas in
Hawaii. Mar. Fish. Rev. 41(8):7-14.
Zeece, M G , K. Katoh, R M Robson, and F. C Parrish, Jr.
1986a. Effect of cathepsin D on bovine myofibrils under
different conditions of pH and temperature. J. Food Sci.
51:769-772, 780.
Zeece, M G , R M Robson. M L Lusby, and F C Parrish, Jr.
1986b. Effect of calcium activated protease (CAF) on
bovine myofibrils under different conditions of pH and
temperature. J. Food Sci. 51:797-803.
372
REPRODUCTIVE BIOLOGY OF
THE SPOTTED SEATROUT, CYNOSCION NEBULOSUS , IN SOUTH TEXAS^
Nancy Brown-Peterson,^ Peter Thomas,^ and Connie R. Arnold'
ABSTRACT
The spotted seatrout, Cynoscion nebulosus, spawns from April through the end of September in
shallow bays in South Texas. All females were sexually mature at 300 mm SL and males at 200 mm
SL. Histological examination of the testes showed that spermatogenesis began in January and contin-
ued until September, although spermatozoa remained in the testes until November. Spermatogenesis
was more active in the peripheral lobules of the testes than in the central lobules during the latter
half of the spawning season. The sequence and timing of final oocyte maturation (FOM) was investi-
gated for the first time in this species. Lipid coalescence began at dawn and germinal vesicle migra-
tion started by midmoming. By early afternoon, oocj^tes were hydrated and spawning occurred at
dusk.
Evidence of multiple spawning was examined. Morphological and histological data showed that
oocytes were continually recruited from March through the end of September, and the percentages of
vitellogenic and fully yolked oocytes did not decline during the spawning season. An average of only
15.5% of the vitellogenic oocytes underwent FOM and hydration during a single spawn. Postovulatory
follicles were found in many fish with mature ovaries throughout the reproductive season. Laboratory
studies showed that this species is capable of repeated spawns. Batch fecundity was best predicted by
ovary-free body weight of the fish and averaged 451 ±43 eggs/g ovary-free body weight. Estimates
of spawning frequency ranged from every other day to once every three weeks.
The spotted seatrout, Cynoscion nebulosus , sup-
ports important commercial and recreational
fisheries throughout its range (Chesapeake Bay,
Virginia to Tampico, Mexico; Tabb 1966).
Whereas aspects of the reproductive biology of
this sciaenid species have been documented
throughout its range (i.e., Chesapeake Bay:
Brown 1981; Georgia: Mahood 1975; Florida:
Moody 1950, Klima and Tabb 1959, Tabb 1961;
Mississippi: Overstreet 1983; Louisiana: Hein
and Shepard 1979; Texas: Pearson 1929, Miles
1950, 1951), a comprehensive study of reproduc-
tion in spotted seatrout does not exist for any
area. An extensive knowledge of the reproductive
life history of a species is necessary to understand
certain aspects of its reproductive physiology and
endocrinology. Most previous studies have con-
centrated on the size at sexual maturity and the
extended spawning season of this species (Pear-
son 1929; Moody 1950; Tabb 1961; Mahood 1975).
A protracted spawning season is generally char-
iContribution No. 696, University of Texas at Austin, Marine
Science Institute, Port Aransas, TX 78373.
2Marine Science Institute, University of Texas at Austin,
Port Aransas, TX; present address; Florida Department of Nat-
ural Resources, Indian River Lagoon Aquatic Preserves, 4842 S.
U.S. Highway 1, Fort Pierce, FL 34982.
3Marine Science Institute, University of Texas at Austin,
Port Aransas, TX 78373.
Manuscript accepted February 1988.
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.
acteristic of multiple spawners (Nikolskii 1969).
In addition, Overstreet's (1983) histological data
suggest that C. nebulosus may be a multiple
spawner. However, the possibility that spotted
seatrout are multiple spawners has not been thor-
oughly discussed in the literature, and previous
estimates of fecundity (Sundararaj and Suttkus
1962; Overstreet 1983) have not considered the
multiple spawning nature of C. nebulosus or have
included estimates of batch fecundity.
In the present study, the reproductive biology
of C. nebulosus in South Texas was investigated,
and particular attention was given to evidence for
multiple spawning and estimates of batch fecun-
dity and spawning frequency. The spawning sea-
son, spawning sites, time of spawning, and per-
centage of running ripe fish were documented. In
addition, the temporal pattern of final oocyte
maturation was determined. The histological ap-
pearance of the gonads was examined, and partic-
ular attention was given to the presence of post-
ovulatory follicles in ovarian tissue, an indicator
of multiple spawning in the northern anchovy,
Engraulis mordax (Hunter and Goldberg 1980).
The size-frequency distribution of vitellogenic
oocytes was examined, since this can indicate
multiple spawning (deVlaming 1983). Batch fe-
cundity was determined and spawning frequency
373
FISHERY BULLETIN: VOL. 86, NO. 2
was estimated from both field-caught and labora-
tory fish.
MATERIALS AND METHODS
Collection of Samples
Spotted seatrout were collected in Redfish Bay
or Lydia Ann Channel near Port Aransas, TX,
U.SA. at depths of 1-3 m (Fig. 1). Fish were cap-
tured using a 300 m gill net (82 mm stretch) or a
300 m trammel net (outer panels, 178 mm
stretch; inner panel, 89 mm stretch) over shallow
beds of turtle grass, Thalassia testudinum or
shoal grass, Halodule wrightii, bordered by a 1-2
m drop-off into a channel. Fish <300 mm SL were
captured with hook and line in 1-2 m of water.
Samples were collected weekly or twice
monthly from March 1982 through early May
1985. No samples were taken in November and
December 1983, October and December 1984, and
February 1985. During 1982, samples were col-
lected at dusk only; in 1983, 1984, and 1985, sam-
ples from dawn, midday, and midnight were also
taken. During each sampling period, salinity and
temperature were recorded as well as the time of
capture.
Analysis of Fish and Gonads
Total length (TL) and standard length (SL)
were measured to the nearest mm for each speci-
men and total body weight (WT) was determined
to the nearest 10 g. Gonads were removed and
weighed to the nearest 0.1 g (gonad weight, GW)
and the gonadosomatic index (GSI) was calcu-
lated, using the formula: GSI = (GWAVT) x 100.
Reproductive stage of the gonads was assessed
macroscopically using the criteria in Table 1. The
macroscopic criteria used were similar to those
used by Overstreet (1983) and Macer (1974). A
small portion of tissue was removed from the
anterior or midsection of one gonad from each fish
and preserved in Davidson's fixative for histolog-
ical analysis (Jones 1966). Tissues were dehy-
drated and embedded in paraffin. Seven micron
sections were cut and stained with Harris' hema-
toxylin and counterstained with eosin. Reproduc-
tive stage of each sample was assessed microscop-
ically using the criteria in Table 1. Fish with
^Area enlarged
Gulf of Mexico
Figure 1. — Location of spotted seatrout sampling sites in South Texas. Asterisks denote sampling sites.
374
BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT
mature gonads were considered sexually mature.
The microscopic criteria used were similar to
those described by Yamamoto (1956), Hyder
(1969), and Macer (1974). The percentages of each
stage of oocyte, atretic structures, and postovula-
tory follicles were determined from histological
slides of the ovaries of females >305 mm SL by
lected fields of view. An oocyte was counted if
>50% of the cell was in the field of view. Two
hundred and fifty to 500 oocytes were counted for
each histological specimen.
A small portion of tissue (<0.5 g) was removed
from the center of each ovary of females >305 mm
SL collected in 1984 and 1985, placed in clearing
slides of the ovaries of females >305 mm SL by SL collected in 1984 and 1985, placed in clearing
counting all the oocytes in three randomly se- solution (6 parts ethanol, 3 parts formalin, 1 part
Table 1 . — Criteria used to describe gonadal reproductive stages of male and female spotted seatrout collected in South Texas.
Stage
Sex
Macroscopic appearance
Microscopic appearance
Immature
Regressed
M
Early
developing
F
M
Developing
F
Mature
Ripe
Running
ripe or
spawning
Partially
spent
Spent
F Ovary small, thin, light pink, slight vasculariza-
tion. GSI range: 0.39-0.55.
M Testes small, thin, light grey, appearance similar to
mesentery.
F Ovary small, light pink, vascularization more obvious
than in immature fish. GSI range: 0.39-1.16.
Testes small, thin, white, slightly larger than imma-
ture fish. GSI range: 0.04-0.20.
Ovary similar to regressed fish but slightly
larger. GSI range: 0.80-1.26.
Testes similar to regressed fish but slightly
larger. GSI range: 0.16-0.48.
Ovary visably enlarged, light yellow, highly vascular,
approximately 60% of the length of the body cav-
ity. GSI range: 0.95-2.6.
M Testes thickened, creamy white, no free milt ex-
pelled when cut. GSI range: 0.21-0.68.
F Ovary large, brilliant yellow orange, oocytes visible
to naked eye, vascularization prominent. Ovaries
80-90% length of body cavity. GSI range: 2.5-
9.6.
M Testes creamy white, thicker, more firm and elon-
gated than developing testes. A small amount of
milt expelled when cut. GSI range: 0.54-1.42.
F Oocytes hydrated, ovary looks clear, takes up al-
most entire body cavity, highly vascular-
ized. GSI range: 12.5-19.9.
M Testes white, swollen, milt does not flow with light
pressure, but flows freely when cut. GSI range:
1.07-1.62.
F Hydrated oocytes expelled with little to no pressure,
ovary fluid, fills almost entire body cavity. GSI
range: 7.7-17.6.
M Testes creamy white, milt freely flowing with slight
pressure. GSI range: 1.41-3.84.
F Ovary looks similar to mature condition but more
flacid; occupies smaller percentage of body cav-
ity. GSI range: 2.0-5.3.
M Tests looks the same as running ripe but
smaller. GSI range: 0.85-1.77.
F Ovary flacid but still highly vasculahzed, no longer
than 50% of body cavity, pinkish in color. GSI
range: 1.33-2.63.
M Testes flacid, width reduced, white. GSI range:
0.13-0.80.
Only primary oocytes present; no atretic oocytes.
Lamellar margin thin.
Only primary spermatogonia present.
Primary chromatin nucleolar and early perinuceolar
stage oocytes. Some atretic oocytes. Lamellar
margin thicker than immature, more convoluted.
Primary and secondary spermatogonia present.
Appearance of late perinucleolar oocytes and ph-
mary cortical alveoli stage oocytes. No atretic
oocytes present.
Many secondary spermatogonia and primary sper-
matocytes.
Oocytes in secondary cortical alveolar stage and
yolk granule stage. Many oocytes still in primary
cortical alveolar stage. Early yolk globular stage
present.
All stages of spermatogenesis present, with few pri-
mary spermatogonia and free spermatozoa. Pri-
mary and secondary spermatocytes predominate.
Oocytes in yolk globular stage most common. Yolk
and oil globules begin to encroach on nucleus.
Largest oocytes range from 300 to 375 y.m.
No primary spermatogonia present, some second-
ary spermatogonia. Secondary spermatocytes
and spermatids most numerous. Spermatozoa in
central lobules.
Many hydrated oocytes, irregularly shaped,
eosinophilic. Other oocytes in yolk granular and
yolk globular stages.
Spermatozoa, spermatids and secondary spermato-
cytes predominate. Few to no spermatogonia.
Central lobules filled with spermatozoa.
Same as ripe ovaries but fewer hydrated oocytes.
Atretic and postovulatory follicles may be present.
Same as ripe fish, but spermatozoa evident in sperm
ducts.
Similar to mature ovary, except atretic oocytes in
several stages of degeneration always evident.
Postovulatory follicles occasionally present.
No spermatogonia, few primary spermatoocytes.
Most active spermatogenesis in peripheral lob-
ules. Spermatozoa partially filling lobules and
abundant in sperm ducts.
Massive atresia of all remaining vitellogenic oocytes.
Many primary and chromatin nucleolar oocytes
present. Lamellar membrane highly convoluted.
Spermatozoa present in some lobules. Most lobules
small, with only primary and secondary sper-
matogonia.
375
FISHERY BULLETIN: VOL, 86, NO. 2
glacial acetic acid), and vigorously shaken for 30
seconds. Within a few minutes the cytoplasm
cleared and the germinal vesicle could be easily
observed microscopically. Ovarian fragments
were taken from females collected at the spawn-
ing site over a 24-h period, placed in clearing solu-
tion and then examined under low-power magni-
fication to determine the stage of final oocyte
maturation.
Oocyte Size-Frequency Distributions
and Estimates of Batch Fecundity
To determine fecundity and the frequency dis-
tribution of oocyte diameters, a 2-15 g piece of
tissue was removed from the midsection of the
ovaries of 57 fish and weighed to the nearest
0.01 g. The tissues were placed in a modified
Gilson's solution (Bagenal 1966) for 3-12 months
and periodically shaken to separate the oocytes
from connective tissues. Ovaries containing hy-
drated oocytes were examined after three months
since hydrated oocytes of spotted seatrout began
to disintegrate when left in Gilson's solution for a
longer period of time.
The volumetric method was used to estimate
fecundity (Bagenal and Braum 1971). The oocyte
samples were suspended in 500-1,500 mL of
water and three replicate 0.5 or 1 mL subsamples
were taken. All the oocytes >30 ^JLm were
counted, and those >80 |jim in diameter (the
growing oocytes) in each sample were measured
to the nearest 15 ^m using an ocular micrometer.
A total of 556-1,110 growing oocytes were mea-
sured in each sample. The number of resting
oocytes (oocyte diameter 30-80 ^JLm) was deter-
mined by diluting the original oocyte suspension
1:10, and counting three replicate subsamples.
Altogether, the frequency distributions of oocytes
from 48 fish were analyzed (3 in developing stage;
9 in mature, spawning not imminent stage; 14 in
mature, just prior to spawning stage; and 22 in
running ripe stage). Fecundity was calculated fol-
lowing Macer's (1974) formula and expressed as
relative fecundity of number of eggs per gram
ovary-free body weight. Batch fecundity (BF) is
defined as all oocytes >350 ixm which were
undergoing final oocyte maturation that formed a
distinct batch, and all hydrated oocytes. This defi-
nition of batch fecundity is in agreement with
Hunter and Macewicz's (1985) statement that
oocytes undergoing final oocyte maturation may
be included as hydrated oocytes when hydration
occurs very rapidly.
Spawning of Fish in the Laboratory
Four female and two male spotted seatrout
were maintained in a 30,000 L recirculating sys-
tem. The tank, filtration system and feeding
regime of the fish has been described previously
(Arnold et al. 1976). The salinity ranged from 25
to 30%(. Spawning was induced by increasing the
temperature and photoperiod from wintertime
settings of 13°C, 9L:15D to 26°C and 15L:9D
(Arnold et al. 1976). The filter boxes were checked
daily for the presence of buoyant, newly fertilized
eggs.
Statistical Analysis
Simple linear regression, oneway analysis of
variance, and analysis of covariance were com-
puted for the data using SPSS packaged programs
(SPSS 1981).
RESULTS
Size at Maturity
Some female spotted seatrout were sexually
mature after they reached 231 mm SL and >90%
of the females had reached sexual maturity at 271
mm SL (Table 2). By 300 mm SL, all female spot-
ted seatrout were sexually mature. Fish 300 mm
SL or larger made up 85.4%, and immature fish
comprised 6.5%, of all the females sampled.
Male spotted seatrout reached sexual maturity
at a much smaller size than females. The size at
Table 2. — Number and percentage of mature female and male
spotted seatrout by 10 mm size categories collected in South)
Texas, April 1982-N^ay 1985. Maturity was judged by histological
and macroscopic inspection.
Standard
Female
Male
length
(mm)
N
% mature
N
% mature
201-210
5
0
5
100
211-220
4
0
4
100
221-230
4
0
1
100
231-240
4
50
3
100
241-250
6
83
3
67
251-260
6
100
5
100
261-270
5
80
7
100
271-280
11
91
10
100
281-290
14
100
11
100
291-300
24
96
20
100
301-310
60
100
40
100
311-320
94
100
68
100
321-330
115
100
80
100
>330
945
100
507
100
Total
1,297
764
376
BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT
which most male spotted seatrout attain sexual
maturity could not be determined, since the
smallest fish collected by the sampling methods
(200 mm SL) were all sexually mature (Table 2).
Season and Time of Spawning
Histological and macroscopic examination of
the gonads (see Table 3 and Figure 3) and mean
GSI values (Fig. 2) show that spotted seatrout
have an extended reproductive season in South
Texas. Mean GSI values of males and females
increased by April 1982-85 and remained ele-
vated through the end of September. A 5°C in-
crease in water temperature at the sampling sites
to 23°C during the first week of April 1982-85
was paralleled by an increase in GSI to 2.1 in
males and 4.5 or greater in females. The pattern
of seasonal changes in mean GSI values of males
was relatively consistent during the three-and-a-
half years of sampling. Mean GSI values began to
increase in mid-February, reached a maximum of
1.9-2.4 by April, and slowly declined during the
spawning season until they dropped rapidly to
regressed levels of 0.2 by the first half of October
(Fig. 2). The seasonal patterns in mean GSI val-
ues of females were also similar from 1982 to
1985. In all four years, mean GSI increased in
April and subsequently declined in May. In 1982-
84, mean GSI increased again later in the season
prior to the final decrease to regressed levels in
October (Fig. 2). Thus, mean GSI values appear to
be bimodal, with one period of peak spawning
activity in April, and the second period of peak
spawning activity varying between August 1982
and July 1983 and 1984.
Male and female spotted seatrout in spawning
condition (males with freely flowing milt, females
with ovulated oocytes) were consistently captured
during a 2-h period around dusk over shallow (1
m) beds of Thalassia testudinum or Halodule
wrightii bordered by a channel 2 m deep. The
salinity at the spawning sites ranged from 20 to
37%c. Although actual spawning was not ob-
served, collection of newly fertilized eggs from the
spawning area at dusk confirmed that spawning
was taking place (Holt and Holf*). Spawning fish
were not captured over beds of scattered H.
wrightii or T. testudinum that were not immedi-
ately adjacent to a channel. Females with freely
^S. A. Holt and G. J. Holt, University of Texas at Austin,
Marine Science Institute, Port Aransas, TX 78373, pers. com-
mun. 1983.
t
33
1-25
17
1-9
I I I I I I I I I I I I |'i''rV;'i^|"r|T| I M I M ' M 1 M ' I M I I ly'V' I I I I M M I I I Ml I I I I I I I I I ' i ' i ' i
°C
MM J SNJMMJ
1982 1983
N J M M' J S N
1984
J M M
1985
Month
Figure 2. — Bi-monthly mean water temperature and mean gonadosomatic index (GSI) of male and female spotted seatrout collected
in South Texas from March 1982 to May 1985, including ± 1 SE of the mean. Sample size for each data point: 3-61 for females (sample
size <15 in only 9 cases) and 2-52 for males (sample size <9 in only 9 cases).
377
FISHERY BULLETIN: VOL. 86, NO. 2
flowing oocytes were only captured at dusk and
milt appeared to flow more freely in males at dusk
than at other times of the day.
Gonadal Development in Males
Testicular recrudescence began in January
with the appearance of primary spermatocytes
(Fig. 3), and by February the majority of the
males were in the early developing or developing
reproductive stage. In March, 89% of the males
had testes containing free spermatozoa (Fig. 3),
although only 27% were running ripe. Over 94%
of all males captured from April until the end of
August were running ripe. Spent males first ap-
peared in August and all the males during Octo-
ber and November were either in the spent or
regressed condition. In December, all males cap-
tured had regressed testes containing only pri-
mary and secondary spermatogonia (Fig. 3).
Histological observations revealed that the
testes are the common unrestricted spermatogo-
nia type, as described by Grier (1981). Spermato-
genesis occurred throughout the entire testis dur-
ing the majority of the year. However, from late
February through early August, spermatogenesis
was more advanced in the central lobules then in
the peripheral lobules. By August there was no
spermatogenic activity in the central lobules, al-
though spermatogenesis continued in the periph-
eral lobules until mid-September. The only period
of the year when there was no active spermato-
genesis was from October through late January,
although primary and secondary spermatogonia
were present from October through the end of
April (Fig. 3). Primary spermatocytes were com-
mon from late January until March, although
they did not disappear from the testes entirely
until the end of July. Secondary spermatocytes
first appeared in mid-February, were common
through the end of July and did not disappear
until early September (Fig. 3). Spermatids and
spermatozoa were present in the central lobules
by late February, and from March through mid-
August the central lobules were swollen with
spermatozoa (Fig. 3). In late August and Septem-
ber many of the central lobules appeared to be
partially empty of spermatozoa, although some
spermatozoa were still present up to mid-Novem-
ber (Fig. 3). The peripheral lobules contained ac-
tive spermatogenic cysts of primary and second-
ary spermatocytes during June and July and the
lumens of these lobules began to fill with sperma-
tozoa. The peripheral lobules were swollen with
spermatozoa during August and September, but
by mid-October no spermatozoa remained.
Gonadal Development in Females
Gonadal recrudescence was observed in a small
percentage of the females captured in January
and by March 94% of the females were undergo-
ing ovarian development (Table 3). Fish with ma-
ture ovaries were found from March until the end
of September, while running ripe fish were cap-
tured fi-om April through the end of September
(Table 3). Partially spent fish (females that ap-
peared to have spawned at least once but still
contained vitellogenic oocytes) were captured
from May through the end of September. No com-
1° Spermatogonia
2° Spermatogonia
1° Spermatocytes
2° Spermatocytes
Spermatids
Spermatozoa
A
-| r
J J
Month
M
M
A
S
— r
N
D
Figure 3. — Seasonal cycle of spermatogenesis in spotted seatrout collected in South Texas, as
determined by histological and macroscopic observations. Data from April 1982 through
mid-May 1985 are combined. 1° - primary, 2° - secondary.
378
BROWN-PETERSON ET AL : REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT
Table 3. — Percentage of female spotted seatrout In seven repro-
ductive stages by month, as assessed by histological and macro-
scopic examination of the ovaries. Data from April 1982 to May
1985 are combined. REG ^ Regressed, E DEV = Early Develop-
ing, DEV = Developing, MAT = Mature, RR = Ripe and Running
Ripe, P SP = Partially Spent, SP = Spent.
N
Percent in
each
reproductive
stage
Month
REG
EDEV
DEV
MAT
RR
PSP
SP
January
70
98
2
February
51
55
43
2
March
124
4
56
31
9
April
372
4
6
78
12
May
220
4
84
5
7
June
104
1
56
19
24
July
114
1
48
30
20
1
August
100
58
32
7
3
September
37
14
37
30
11
8
October
8
87
13
November
27
100
December
51
98
2
Total
1,278
tions: 150-190 (xm) varied from a lov^^ of 7.3% in
June to a late season high of 16.1%, and averaged
13% of the total number of oocytes in the ovary.
Yolk globular oocytes, the largest oocytes present
(diameter: 200-375 p.m), were most common. The
percentage of oocytes in the yolk globular stage
ranged from a high of 30.8% in April to an end-of-
season low of 22.6% in September and averaged
26.9% during the reproductive season. In addition
to the actively growing oocytes in the ovary, a
small percentage of atretic oocytes was always
found in fish with mature ovaries. Atretic oocytes
are defined here as vitellogenic oocytes undergo-
ing alpha stage atresia. The percentage of atretic
oocytes ranged from a low of 2.2% in March to a
midsummer high of 8% in July and averaged
5.4%. Additionally, postovulatory follicles (POF)
were observed in the ovaries offish captured from
pletely spent fish were captured before July and
few were captured during the remainder of the
reproductive season.
Histological observations offish with regressed
ovaries collected from late September to mid-
February showed only primary chromatin nucle-
olar and early perinucleolar oocytes. Atretic
oocytes were present from September until mid-
December; no atretic oocytes were observed from
mid-December through the end of February. The
appearance of late perinucleolar and primary cor-
tical alveoli stage oocytes in late January, Febru-
ary, and early March represented the initial
stages of ovarian recrudescence. Ovarian devel-
opment was proceeding rapidly by early March
and oocytes in the secondary cortical alveoli and
yolk granule stages were common.
During the reproductive season, histological
observations of the ovary showed a heterogeneous
morphology. Oocytes in all stages of growth (from
resting to the fully grown yolk globular stage)
were distributed throughout the ovaries of all ma-
ture, running ripe, and partially spent fish col-
lected from March through the end of September.
Vitellogenesis was probably continuous from
February until the end of September, as shown by
the continual presence of large numbers of
oocytes in the cortical alveoli and yolk granule
stages. The relative percentages of three types of
post cortical alveolar stage oocytes in fish in the
mature reproductive stage did not change
markedly from March through the end of Septem-
ber (Fig. 4). The percentage of oocytes in the yolk
granule stage (diameter in histological prepara-
Q
o
<
1-
z
m
o
DC
LU
Q.
LU
>
I-
<
_l
LU
IT
O
z
LU
=)
o
LU
DC
P771 Yolk granule frr^ Yolk globule ^g Atretic
N I 5 I 58 I 21 I 17 I 27 I 29 I 9 I
36-1
32
28
24H
20
16
12-
8-
4-
M
M J J
MONTH
Figure 4. — Frequency of three different types of oocytes in fe-
male spotted seatrout ovaries in the mature reproductive stage
(MAT, Table 3) as assessed by histological observation. Fish
were collected from South Texas during the reproductive sea-
son. Frequency refers to the percentage of each type of oocyte
relative to the total number of all types of oocytes counted. Data
from April 1982 through the end of September 1983 are com-
bined and the number (A^) of individuals examined each month
is indicated. Bars indicate mean percentage of oocytes ±1 SE of
the mean. Yolk granule oocyte diameters ranged from 150 to
190 p-m, and yolk globule oocyte diameters ranged from 200 to
375 fim in histological preparations. Atretic oocytes are defined
here as oocytes in alpha stage atresia.
379
FISHERY BULLETIN: VOL. 86, NO. 2
April through September, although the percent-
age of POF relative to the total number of oocytes
was always small (<5%).
The histological appearance of the zona radiata
and follicle layers changed as the oocytes devel-
oped. The zona radiata (chorion) was thin and
nonstriated in perinucleolar and cortical alveoli
oocytes. The zona radiata started to take on its
characteristic striated appearance when oocytes
reached the yolk granular stage and became no-
ticeably thicker and more striated as oocytes
grew into the yolk globular stage. In contrast, the
granulosa and thecal layers appeared to decrease
in thickness as oocyte development proceeded.
Oocytes in the perinucleolar and cortical alveoli
stages had thick, well-developed granulosa and
thecal layers. In yolk granular and yolk globular
oocytes, both follicle layers were noticeably thin-
ner and the thecal layer was not always continu-
ous around the oocyte.
Frequency distributions of growing oocytes (di-
ameters in Gilson's solution >80 fxm; perinucle-
olar to hydrated stage oocytes) from ovaries of 48
fish were analyzed, and Figure 5 shows typical
patterns in fish from four reproductive stages. In
a fish captured in March with developing ovaries,
growing and vitellogenic oocytes (oocytes during
the phase of active vitellogenin uptake) ranging
from 80 to 185 jxm in diameter were present (Fig.
5A; 625 oocytes counted). Fish in the mature re-
productive stage in which spawning did not ap-
pear to be imminent, as indicated by GSI values
<4, had growing and vitellogenic oocyte diame-
ters ranging from 100 to 320 M-m (Fig. 5B; 556
oocytes counted), comprising 21% of the total
number of oocytes >30 fxm in the ovary. Fish just
prior to spawning, as indicated by GSI values >7,
had a distinct batch of oocytes with diameters
Figure 5. — Frequency distributions of growing oocyte diame-
ters O80 p.m) in female spotted seatrout collected from South
Texas. Growing oocytes comprised 21% of the total number of
oocytes >30 ^.m in the ovary. Oocyte diameter refers to the
diameters of oocytes after preservation in Gilson's solution. Fre-
quency refers to the percentage of each size of oocyte relative to
the total number of oocytes in the subsample. Each graph repre-
sents data from a single fish at a different reproductive stage.
The arrow at 350 jim indicates the minimum size necessary for
final maturation to occur. A. Developing stage in March,
GSI = 1.6. B. Mature stage in which spawning is not immi-
nent, GSI = 3.7. Fish in this stage were collected from April
through the end of September. C. Mature stage just prior to
spawning, GSI = 7.4. Fish in this stage were collected from
April through the end of September. D. Running ripe stage,
GSI = 17.6. Fish in this stage were collected from April through
the end of September.
>350 fxm that were undergoing final oocyte mat-
uration (Fig. 5C; 585 oocytes counted). However,
there were no other distinct modes of vitellogenic
oocytes. In running ripe fish, the batch of large
oocytes hydrated to a diameter of 520 ixm or
greater (Fig. 5D; 1,110 oocytes counted). The re-
sults (Figs. 5A-D) clearly demonstrate that C.
nebulosus has a continuous distribution of grow-
ing and vitellogenic oocytes. The oocyte frequency
distribution of the vitellogenic oocytes <320 \i.vc\
in diameter remaining in running ripe fish ap-
I- 2-
lU
o
LU
LU
DC 2-J
80 200 320 440 560 680
OOCYTE DIAMETER (nm)
380
BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT
peared similar to the oocyte distribution in non-
spawning fish (Fig. 5B, D). The percentage (21%)
of growing and vitellogenic oocytes in the ovary
remained constant throughout the reproductive
season in fish in the mature and running ripe
stages.
Final Oocyte Maturation
Final oocyte maturation (FOM) was highly syn-
chronized in spotted seatrout and occurred only in
oocytes >400 ixm ^. Figure 6A shows a photomi-
crograph of a histological section of a spotted
seatrout ovary in the mature reproductive stage
that was not undergoing final oocyte maturation.
Many oocytes were in the yolk globular stage and
appeared to be fully grown. The first readily ob-
servable stage of FOM in "cleared" oocytes was
lipid coalescence (Fig. 6B). The oil droplets in the
oocytes began to coalesce around the germinal
vesicle (nucleus) and subsequently formed one to
three large oil droplets. This stage was not always
observed in histological preparations since many
of the oocytes were not sectioned through their
centers. The yolk globules remained discrete dur-
ing lipid coalescence. After the lipids had coa-
lesced, the germinal vesicle (GV) began to mi-
grate to the periphery of the oocyte (germinal
vesicle migration, or GVM). GVM could be seen
in both histological sections (Fig. 6C) and in
"cleared" oocytes (Fig. 6D). The oil droplet occu-
pies the center of the oocytes shown in Figure 6C,
D. Histological observation of this stage (Fig. 6C)
showed that the yolk globules were not coalesced,
the oil droplets had coalesced to form one or two
large droplets and the GV had begun to lose its
integrity and often appeared semicircular. At the
completion of GVM, the nuclear (germinal vesi-
cle) membrane broke down (GVBD) and the nu-
clear material intermingled with the cytoplasm of
the oocyte. Hydration occurred shortly thereafter,
followed by ovulation and spawning of the fully
mature oocyte.
Final oocyte maturation occurred within 10
hours in spotted seatrout in the natural environ-
ment (Fig. 7). A total of 209 fish were collected
over eight 24-h periods from April through Au-
gust in 1984 and 1985. Forty-three percent of the
fish collected between the hours of 0500 and 1500
were undergoing FOM. Lipid coalescence was
first observed at dawn (0545), and GVM started
at 0900. By 1430, all fish undergoing final matu-
ration had hydrated oocytes and ovulation and
spawning commenced at dusk (1830) and contin-
ued until 2100. None of the fish collected from
2100 to 0500 were undergoing FOM.
Batch Fecundity
The significant positive relationship {P <
0.001) between BF and ovary-free body weight
can be best described by the following equation:
BF = 459WT - 56,066, r^ = 0.56 (Fig. 8), while
curvilinear equations best described the relation-
ship between BF and SL and TL. The coefiicients
of determination in all cases were <0.56.
A one-way analysis of variance showed that
mean BF (number of eggs per gram ovary-free
weight) did not vary significantly during the
April through September spawning season. Mean
relative batch fecundity was highest in Septem-
ber, lowest in May, and varied little during April,
June, and July (Table 4).
A prominent batch of oocytes was present only
in females that were in all stages of final oocyte
maturation or were running ripe (Figs. 5C, D; 7).
The average batch size calculated from 14 fish
containing hydrated oocj^es and no postovulatory
follicles was 451 ± 43 eggs/g ovary-free body
weight. This number averaged 15.5 ± 2.5% of the
number of growing and vitellogenic oocytes in the
ovary.
Table 4. — Monthly mean batch fecundity
expressed as number eggs/g ovary-free
body weight of spotted seatrout in South
Texas. All means were not statistically dif-
ferent.
Month
N
Mean fecundity ±1 se
April
19
477 ± 42
May
2
320 ± 72
June
3
435 ± 109
July
5
409 ± 76
August
3
361 ± 60
September
3
560 ± 79
5A 400 (xm live oocyte equals a 350 ^.m oocyte preserved in
Gilson's solution. Both measurements represent oocytes begin-
ning FOM.
Spawning Frequency
To estimate the spawning frequency of spotted
seatrout in South Texas, the percentage of run-
ning ripe females captured monthly from April
through the end of September in 1982 through
1985 was examined. Only fish captured at dusk
and >305 mm SL were included in this analysis.
The percentage of spawning females ranged fi-om
381
FISHERY BULLETIN: VOL. 86, NO. 2
B
«Q
•#«,
>
«l
^IIf
^ Qy
»:
382
BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT
Figure 6. — Photomicrographs of oocytes from spotted seatrout
in the mature reproductive stage collected from South
Texas. A. Histological section of an ovary in the mature stage
that is not undergoing final oocyte maturation (FOM). Oocyte
development is continuous, with fully grown yolk globular
oocytes co-occurring with oocytes in earlier stages of develop-
ment (Magnification 120x). B. "Cleared" oocytes in the lipid
coalescence stage, the first stage of FOM (magnification 40 x).
C. Histological section of an oocyte undergoing germinal vesicle
migration (GVM). (Magnification 160 x ). D. "Cleared" oocytes
in the GVM stage (magnification 40 X). Key: GV = germinal
vesicle or nucleus, L = Lipid droplets, YG = yolk globules,
YGO = yolk globular oocyte, ZR = Zona radiata.
Lipid Coalescence
KV\\\\\\\N
GVM
14L:10D
Hydration
K\\\\\\1
ion and Spawning
kWWN
^ ^^^^^M
0000
04'00
III
0800 1200
Time
I — r^
1600 2000
Figure 7. — Time-course of final oocyte maturation in spotted seatrout collected from
South Texas. Lipid coalescence is the initial stage in final oocyte maturation. Data
obtained from 46 fish undergoing final oocyte maturation in April and May
1985. GVM = germinal vesicle migration. 14L: lOD = hours of light and dark.
Figure 8.— Relation between batch fecundity (BF)
and ovary-free body weight (WT) of spotted seatrout
from April 1984 to May 1985. Thirty-three fish with
oocytes >350 \i.m that were undergoing final oocyte
maturation and formed a distinct batch of oocytes or
had hydrated oocytes are included.
Weight (g)
383
FISHERY BULLETIN: VOL. 86, NO. 2
a low of 8% in May to a high of 45% in September
(Table 5). An average of 27.5% of the females
captured during the spawning season were run-
ning ripe, corresponding to an average spawning
frequency of once every 3.6 days or 50 times dur-
ing the 6-mo spawning period.
The spawning frequency was also estimated
from females that were undergoing final oocyte
maturation (FOM) between the hours of 0600 and
1400 in April, May, July, and August 1984 and
1985 (Table 5). For comparison, the actual per-
centage of running ripe females captured during
the same months is also presented. An average of
42.8% of the females examined for final oocyte
maturation from April through August were
undergoing FOM. Therefore, average spawning
frequency was once every 2.3 days, or 80 times
during the 6-mo spawning season.
Spawning frequency was also estimated from
the percentage of females captured with ovaries
containing postovulatory follicles (POF). An av-
erage of 13.1% of the females captured had
ovaries which contained POF ranging from 12
hours to 2 days old. This would correspond to a
spawning frequency of once every 7.6 days or 24
times during the spawning season.
Finally, the spawning frequency of four female
C. nebulosus in the laboratory under conditions of
controlled temperature and photoperiod were ex-
amined (Table 6). The fish spawned from 1 to 10
times each month for 17 months, an average of 17
spawns per individual over a 12-mo period (Table
6). This would correspond to a spawning fre-
quency of once every 21 days.
DISCUSSION
Sexual Maturity and Spawning Season
Male spotted seatrout reached sexual maturity
at a smaller size than females which appears to be
a fairly common phenomenon in spotted seatrout
throughout its range (see Mercer 1984 for a re-
view). Both male and female spotted seatrout in
South Texas reached sexual maturity at a size
similar to that reported for other groups of C.
nebulosus along the Gulf Coast (Moody 1950;
Klima and Tabb 1959; Overstreet 1983) and at a
smaller size than along the East Coast (Tabb
1961; Brown 1981).
Running ripe females were captured only be-
tween one hour before and two hours after sunset
in South Texas, suggesting a high degree of syn-
chrony in spawning fish. Collections of newly fer-
Table 5. — Percentage of spotted seatrout spawning (RR) or under-
going final oocyte maturation (FOf^) in South Texas. Data on
spawning fish collected April 1982-f^ay 1985. Data on fish under-
going FOM collected April 1984-IVIay 1985.
Spawning
fish
Fish undergoing FOM
1982-85
1984-85
Month
N
# RR
% RR
N
# FOM
% FOM
% RR
April
256
42
16
33
14
42
37
May
140
11
8
21
10
48
13
June
101
20
20
—
—
—
—
July
82
33
40
7
3
43
52
August
101
36
36
8
3
38
—
September
20
9
45
—
—
—
—
Mean
27.5
42.8
25.5
Table 6. — Number of spawns and average
water temperature by month for four female
spotted seatrout contained in a 30,000 L tank in
Porl Aransas, TX, under conditions of con-
trolled temperature and photoperiod, July
1974-November 1975. Expenmental proce-
dures descnbed in Arnold et al. (1976).
Month
Number of spawns
°C
July
8
26.75
August
7
26.75
September
4
26.0
October
6
25.5
November
4
24.5
December
8
24.0
January
9
23.5
February
10
24.0
March
10
24.0
April
3
25.75
May
1
25.0
June
0
22.0
July
6
25.75
August
2
25.0
September
2
23.0
October
9
23.5
November
10
24.0
Total
99
Average: 1 7
spawns/ 1 2 months/ female
tilized eggs during a 25-h period after sunset (Holt
et al. 1985) provides supporting evidence for this
spawning synchrony.
Both histological data and GSI values showed
spotted seatrout have an extended spawning sea-
son in South Texas. Gronadal recrudescence began
in January in male and in February in female
spotted seatrout. Spawning commenced in April
and continued until the end of September. Other
studies of C. nebulosus along the Gulf and East
coasts of the United States have also reported
long spawning seasons (Mercer 1984). Addition-
ally, other sciaenids have extended spawning sea-
sons (Merriner 1976; White and Chittenden 1977;
DeVries and Chittenden 1982; Love et al. 1984),
384
BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT
suggesting a prolonged spawning season is a com-
mon reproductive strategy among sciaenids liv-
ing in temperate and subtropical waters.
The duration of the spawning season may be
related to water temperature. Perhaps this appar-
ent association with temperature is related to the
viability and development of spawned spotted
seatrout eggs. Twenty-three degrees may be the
minimum temperature necessary for successful
spawning, as indicated by both the failure to cap-
ture running ripe fish at lower temperatures and
by data from spotted seatrout induced to spawn
under laboratory conditions (Arnold et al. 1976;
Table 6). However, since spawning ceased in Sep-
tember when water temperatures were well
above 23°C, possibly a decrease in photoperiod in
combination with a decrease in temperature pro-
vides the necessary cue for termination of spawn-
ing. Hein and Shepard (1979) suggested photope-
riod may be an important regulating factor in
C. nebulosus spawning. Data from laboratory-
spawned spotted seatrout (Arnold et al. 1976) also
support this speculation.
The seasonal pattern of mean GSI values dur-
ing the spawning season was relatively consistent
for both males and females over the Sj-yr period,
1982-85 (Fig. 2). The bimodality of the female
GSI data suggests the possibility of two peaks in
spawning activity, although the timing of the sec-
ond peak varies from year to year, thus demon-
strating the need for several consecutive years of
data. Bimodal spawning peaks have been previ-
ously reported for the species by Hein and Shep-
ard (1979) in Louisiana, Stewart (1961) in
Florida, and Brown (1981) in Chesapeake Bay,
VA. However, mean GSI values should be used
with caution when attempting to predict actual
peaks in spawning activity (deVlaming et al.
1982).
Fecundity
Accurate annual fecundity measurements are
difficult to determine for multiple spawning
fishes with an extended spawning season. Meth-
ods of calculating annual fecundity from mea-
surements of the total number of growing oocytes
at the beginning of the spawning season (Bagenal
1966), or other approaches based on the total
number of oocytes at the beginning of the spawn-
ing season minus egg retention at the end of the
spawning season (Conover 1985), are inappropri-
ate for multiple spawning species such as spotted
seatrout which show continuous recruitment of
oocytes during the reproductive season. Thus, the
previous estimates of annual fecundity in spotted
seatrout, which did not take continuous recruit-
ment of oocytes into consideration and measured
either total fecundity (Overstreet 1983) or fecun-
dity of growing and vitellogenic oocytes (Sun-
dararaj and Suttkus 1962), probably underesti-
mated annual fecundity. Furthermore, BF has
not previously been calculated for this species.
Although no monthly differences in BF were
apparent (Table 4), sample sizes were too small to
draw any definite conclusions from these data.
The relatively low coefficient of determination
(0.56) is similar to values reported by Conover
(1985) for Atlantic silversides, Menidia menidia,
another multiple spawning species. Perhaps a
more accurate estimate of annual fecundity than
previously reported for spotted seatrout can be
obtained by multiplying BF by the number of
spawns during the reproductive season. Unfortu-
nately, as discussed later, estimates of spawning
frequency vary considerably, so it is not possible
to make an accurate estimation of the annual fe-
cundity. However, available data indicate that
average annual fecundity may be greater than 10
million eggs.
Multiple Spawning
Histological examination of the testes revealed
that spermatogenesis ceased earlier in the central
lobules, although they contained spermatozoa
one-and-one-half months longer than the periph-
eral lobules. It is possible that the same central
lobules act as storage areas for spermatozoa pro-
duced by the more spermatogenically active pe-
ripheral lobules during the second half of the
spawning season, as suggested by Hyder (1969)
for Tilapia . This may represent a strategy in mul-
tiple spawning fish with a prolonged spawning
season that allows for a constant supply of sper-
matozoa while investing a minimal amount of en-
ergy into spermatogenesis.
Several lines of evidence indicate that female
C. nebulosus also spavni several times during the
reproductive season. The relatively high percent-
age of running ripe females, fish undergoing
FOM and partially spent fish captured through-
out the spawning season, and the absence of com-
pletely spent fish until the last third of the spawn-
ing season, suggest that an individual does not
spawn all the vitellogenic eggs in the ovary at one
time. Indeed, oocyte size-frequency analysis
shows a continuous distribution of growing and
385
FISHERY BULLETIN: VOL. 86, NO, 2
vitellogenic oocytes (Fig. 5) and fecundity esti-
mates show that only about 15% of the growing
oocytes undergo FOM prior to a spawn. Histolog-
ical data shows that the percentage of vitellogenic
oocytes in the ovary remains constant throughout
the spawning season (Fig. 4), which suggests that
new oocytes may be recruited into the vitellogenic
phase as rapidly as mature oocytes are released.
Convincing histological evidence of multiple
spawning is the presence of postovulatory follicles
(POF) from May through the end of September in
ovaries containing many vitellogenic oocytes.
Hunter and Goldberg (1980) characterized pos-
tovulatory follicles in laboratory-spawned En-
graulis mordax, a multiple spawning fish, and
found POF in all females that had spawned in the
laboratory one or two days previously. Finally,
laboratory studies also show that spotted seatrout
are capable of multiple spawning under relatively
constant environmental conditions (Table 6).
Tucker and Faulkner (1987) also found that six
female fish kept in raceways outdoors at the am-
bient summer temperature and photoperiod
spawned repeatedly.
Spawning Frequency
It is especially difficult to determine the
spawning frequency of wide-ranging, multiple-
spawning marine fishes such as C. nebulosus that
are not group-synchronous spawners. One
method to estimate spawning frequency is to
count the number of distinct batches of vitel-
logenic oocytes in the ovary (Shackley and King
1977). However, only one distinct batch of vitel-
logenic or hydrated oocytes can be distinguished
in spotted seatrout ovaries at any one time (Fig.
5) and the reliability of this method has been
questioned (deVlaming 1983). Therefore, three
techniques were used to estimate spawning fre-
quency in spotted seatrout.
Spawning frequency was estimated to be once
every 3.6 days from the percentage of running
ripe fish caught on the spawning grounds. Al-
though this is probably an overestimate owing to
sampling bias, the error may not be susbtantial,
since the spawning grounds are also the feeding
grounds for this species (Moody 1950) and many
nonspawning individuals were captured. The
time of sample collection did not significantly in-
fluence the estimate of spawning frequency. High
spawning frequencies were also obtained (every
2.3 days) when fish were captured 6-12 hours
prior to spawning and examined for signs of final
oocyte maturation (Table 5). The spawning fre-
quency of other sciaenids fishes has been esti-
mated by this technique (DeMartini and Foun-
tain 1981; Love et al. 1984). Additionally, Hunter
and Macewicz (1985) suggested that this method
produces a useful first approximation of spawning
frequency.
The proportion of fish having POF in the ovary
has also been used to determine spawning fre-
quency (Hunter and Goldberg 1980; Albeit et al.
1984; Hunter and Macewicz 1985). Spotted
seatrout were found to have POF throughout the
spawning season, although the age of the POF
was often difficult to determine. Furthermore, de-
tailed laboratory studies have not been under-
taken to accurately age POF in spotted seatrout.
However, the once a week estimate of spawning
frequency obtained using this method is similar
to spawning frequencies reported for two other
sciaenids, the queenfish, Seriphus politus, (De-
Martini and Fountain 1981) and the white
croaker, Genyonemus lineatus, (Love et al. 1984).
Spawing frequency estimates from POF are prob-
ably more reliable than estimates based on the
number of spawning fish since sampling bias
is less likely to occur when capturing fish with
POF.
Another method used to quantify spawning fre-
quencies in various species is direct observation
of spawning in the laboratory or in "controlled"
field situations, such as impoundments (Gale and
Deutsch 1985; Hubbs 1985; Heins and Rabito
1986). Spotted seatrout spawned an average of
once every three weeks per individual under con-
trolled temperature and photoperiod in the labo-
ratory (Table 6). Tucker and Faulkner (1987)
found that six female spotted seatrout kept out-
doors at ambient temperature and photoperiod
averaged one spawn per individual every 2.3
weeks. The same spawning frequency was noted
for an individual female, although that same in-
dividual later spawned three times in four days
(Tucker and Faulkner 1987). Thus, spotted
seatrout appear to be capable of the high spawn-
ing fi"equencies estimated from field-caught fish
with hydrated ovaries, although this frequency is
probably not sustained throughout the entire
spawning season. In general, the spawning fre-
quencies in both laboratory studies are lower
than those estimated from actual spawning fish
in the field. However, it is unclear whether this is
due to an overestimation of spawning frequency
in the field or to a decline in spawning frequency
owing to confinement in the laboratory.
386
BROWN-PETERSON ET AL.: REPRODUCTIVE BIOLOGY OF THE SPOTTED SEATROUT
ACKNOWLEDGMENTS
We thank John Trant, John Gourley, John
Smith, and Wayne Wofford for their help with the
field work and other aspects of the research.
Funding was provided by grants from the Sid
Richardson Foundation and the Caesar Kleberg
Foundation.
LITERATURE CITED
Alheit. J . V H. Alaron, and B. J. Macewicz.
1984. Spawning frequency and sex ratio in the peruvian
anchovy, Engraulis ringens. CalCOFI. Rep. 25:43-52.
Arnold. C R . T D Williams, W. A. Fable, Jr., J. L Lasswell,
AND W H Bailey.
1976. Methods and techniques for spawning and rearing
spotted seatrout (Cynoscion nebulosus ) in the laboratory.
Proc. 30th Conf Southeast. Assoc. Game Fish Comm.
30:167-178.
Bagenal. T B.
1966. The ecological and geographical aspects of the fe-
cundity of the plaice. J. Mar. Biol. Assoc. U.K. 46:161-
186.
Bagenal, T. B., and E. Braum.
1971. Eggs and early life history. In W. E. Ricker (edi-
tor), Methods for assessment of fish production in fresh
waters, p 159-181. IBP (Int. Biol. Prgramme). Handb.
3, 2d ed., Blackwell Sci. Publ. Oxford, England.
Brown, N J.
1981. Reproductive biology and recreational fishery for
spotted seatrout, Cynoscion nebulosus, in the Chesa-
peake Bay area. MA Thesis, College of William and
Mary, Gloucester Point, VA, 120 p.
Conover, D O.
1985. Field and laboratory assessment of patterns in fe-
cundity of a multiple spawning fish: The Atlantic silver-
side, Menidia menidia. Fish. Bull., US. 83:331-341.
DeMartini, E. E . AND R K. Fountain.
1981. Ovarian cycling frequency and batch fecundity in
the queenfish, Seriphus politus: attributes representa-
tive of serial spawning fishes. Fish. Bull., U.S. 79:547-
560.
deVlaming, V.
1983. Oocyte development patterns and hormonal in-
volvements among teleosts. In J. C. Rankin, T. J.
Pitcher, and R. T. Duggan (editors). Control processes in
fish physiology, p. 176-199. Wiley & Sons, N.Y.
deVlaming, V , G Grossman, and F Chapman.
1982. On the use of the gonosomatic index. Comp.
Biochem. Physiol. 73A:31-39.
DeVries, D a., and M. E Chittenden, Jr.
1982. Spawning, age determination, longevity and mor-
tality of the silver seatrout, Cynoscion nothus , in the Gulf
of Mexico. Fish. Bull., U.S. 80:487-500.
Gale, W. F., and W G Deutsch.
1985. Fecundity and spawning fi-equency of captive tesse-
lated darters - fractional spawners. Trans. Am. Fish.
Soc. 114:220-229.
Grier, H J.
1981. Cellular organization of the testis and spermatoge-
nesis in fishes. Am. Zool. 21:345-357.
HEIN, S , AND J SHEPARD.
1979. Spawning of spotted seatrout in a Louisiana estuar-
ine ecosystem. Proc. Annu. Conf. Southeast. Assoc.
Fish. Wildl. Agencies 33:451-465.
Heins, D. L , AND F. G. Rabito, Jr.
1986. Spawning performance in North American min-
nows: Direct evidence of the occurrence of multiple
clutches in the genus Notropis. J. Fish Biol. 28:343-
357.
Holt, G. J., S. A. Holt, and C. R Arnold.
1985. Diel periodicity of spawning in sciaenids. Mar.
Ecol. Prog. Ser. 27:1-7.
HUBBS, C.
1985. Darter reproductive season. Copeia 1985:56-68.
Hunter, J R , and S. R Goldberg.
1980. Spawning incidence and batch fecundity in north-
em anchovy, Engraulis mordax. Fish. Bull., U.S.
77:641-652.
Hunter J. R., and B. J. Macewicz.
1985. Measurement of spawning frequency in multiple
spawning fishes. In R. Lasker (editor). An egg produc-
tion method for estimation spawning biomass of pelagic
fish: application to the northern anchovy, Engraulis mor-
dax, p. 79-94. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS 36.
Hyder, M.
1969. Histological studies on the testis of Tilapia leucos-
ticta and other species of the genus Tilapia (Pisces:
Teleostei). Trans. Am. Microsc. Soc. 88:211-231.
Jones, R. McC.
1966. Basic microscopic techniques. 5th ed. Chicago
Univ. Press. (See p. 249).
Klima, E F , AND D. C. Tabb.
1959. A contribution to the biology of the spotted weak-
fish Cynoscion nebulosus from northwest Florida, with a
description of the fishery. Fla. State Board Conserv.
Tech. Ser. 30, 24 p.
Love, M S , G E McGowen, W Westphal, R J Lavenberg, and
L. Martin
1984. Aspects of the life history and fishery of the white
croaker, Genyonemus lineatus (Sciaenidae), off Califor-
nia. Fish. Bull., U.S. 82:179-198.
Macer, C. T.
1974. The reproductive biology of the horse mackerel,
Trachurus trachurus (L.) in the North Sea and English
Channel. J. Fish Biol. 6:415-438.
Mahood, R K.
1975. Spotted seatrout in coastal waters of Geor-
gia. Proc. Annu. Conf Southeast. Assoc. Game Fish
Comm. 29:195-207.
Mercer, L. P
1984. A biological and fisheries profile of spotted seatrout,
Cynoscion nebulosus. N.C. Dep. Nat. Res. Comm. Dev.
Div. Mar. Fish., Spec. Sci. Rep. No. 40, 87 p.
Merriner, J V.
1976. Aspects of the reproductive biology of the weakfish,
Cynoscion regalis (Sciaenidae), in North Carolina. Fish.
Bull., U.S. 74:18-26.
Miles, D W
1950. The life histories of the spwtted seatrout, Cynoscion
nebulosus and the redfish, Sciaenops ocellatus. Tex.
Game Fish Comm. Mar. Lab. Annu. Rep. 1949-1950, 38
P-
1951. The life histories of the seatrout, Cynoscion nebulo-
sus, and the redfish, Sciaenops ocellatus: Sexual develop-
ment. Tex. Game Fish Comm. Mar. Lab. Annu. Rep.
1950-1951, 11 p.
387
FISHERY BULLETIN: VOL. 86, NO. 2
Moody, W. D.
1950. A study of the natural history of the spotted
seatrout, Cynoscion nebulosus, in the Cedar Key, Florida
area. Q. J. Fla. Acad. Sci. 12:147-172.
NiKOLSKn, G. V
1969. Theory offish population dynamics as the biological
backgtound for rational exploitation and management of
fishery resources. Oliver & Boyd, Edinb., 323 p.
OVERSTREET, R. M.
1983. Aspects of the biology of the spotted seatrout,
Cynoscion nebulosus, in Mississippi. Gulf Res. Rep.
Suppl. 1:1-43.
Pearson, J. C
1929. Natural history and conservation of redfish and
other commercial Sciaenids on the Texas coast. Bull.
U.S. Bur. Fish. 44:129-214.
Shackley, S E , AND P E King
1977. The reproductive cycle and its control; frequency of
spawning and fecundity in Blennius pholis L. J. Exp.
Mar. Biol. Ecol. 30:73-83.
SPSS Inc.
1981. SPSS user's guide. 2d ed. SPSS, Inc., 988 p.
Stewart, K W
1961. Contributions to the biology of the spotted seatrout
(Cynoscion nebulosus) in the Everglades National Park,
Florida. M.S. Thesis, Univ. Miami, Coral Gables, FL,
103 p.
Sundararaj, B I , and R D Suttkus
1962. Fecundity of the spotted seatrout Cynoscion nebulo-
sus (Cuvier) from Lake Borgne area, LA. Trans. Am.
Fish. Soc. 91:84-88.
Tabb, D C
1961. A contribution to the biology of the spotted seatrout,
Cynoscion nebulosus of East Central Florida. Fla. State
Board Conserv. Tech. Ser. 35, 21 p.
1966. The estuary as a habitat for spotted seatrout,
Cynoscion nebulosus. Am. Fish. Soc. Spec. Publ. 3:59-
67.
Tucker, J W , Jr , and B E Faulkner.
1987. Voluntary spawning patterns of captive spotted
seatrout. Northeast Gulf Sci. 9:59-63.
White. M L . and M E Chittenden, Jr
1977. Age determination, reproduction and population
dynamics of the Atlantic croaker, Micropogonias undula-
tus. Fish. Bull., U.S. 75:109-123.
Yamamoto, N.
1956. Studies on the formation of fish eggs. I. Annual
cycle in the development of ovarian eggs in the flounder,
Liopsetta obscura. J. Fac. Sci. Hokkaido Univ. Ser. VI,
Zool. 12:362-373.
388
NOTES
NOTES ON REPRODUCTION IN
THE SCALLOPED HAMMERHEAD, SPHYRNA
LEWINI, IN NORTHEASTERN TAIWAN WATERS
The reproductive mode of hammerhead sharks
(family Sphyrnidae) is placental viviparity, fol-
lowing Teshima's (1981) designations. Three spe-
cies of hammerhead sharks are known from
northeastern Taiwanese waters: Sphyrna lewini
(Griffith and Smith), S. mokarran (Riippel), and
S. zygaena (Linnaeus). The scalloped hammer-
head, S. lewini, commonly found from Pung-Chia
Island to Guei-Shan Island (Fig. 1) is one of the
most abundant shark species in this area. Based
on data from the Nan Fan Ao Fish Market (lo-
cated in northeastern Taiwan), 460-510 t/year
from 1982 through 1984 were landed, represent-
ing one-fourth of the total catches of sharks in
this area (the total landing being 1,760-2,240 t/
year). Catches peaked in spring and winter and
were lower in summer and autumn.
This species is also distributed in western and
southern waters of Taiwan; however, the catch
from those areas were smaller. The employees of
the southern and western fish markets explained
that the scalloped hammerhead is occasionally
seen in small numbers in those areas, but there
are no landing data. It is also common in coastal
warm temperate and tropical seas throughout
much of the world (Compagno 1984). Although
26°N
Figure 1. — Sampling area of
Sphyrna lewini.
FISHERY BULLETIN; VOL. 86, NO. 2, 1988
25°N
the scalloped hammerhead is one of the most
valuable food resources in Taiwan, many facets of
its life history, particularly reproduction, are not
well known. This study provides information
about certain aspects of reproduction in the scal-
loped hammerhead in northeastern Taiwan
waters.
Materials and Methods
From September 1982 to June 1983 and from
December 1983 to September 1985, shark speci-
mens were examined and material collected
monthly at Nan Fan Ao Fish Market. These
sharks had been caught by drift longlines set near
the surface to around 100 m, or by surface har-
poon. A total of 674 scalloped hammerhead
sharks were examined at the fish market (Table
1). Data recorded included total length measured
from the tip of the snout to the tip of the upper
lobe of the caudal fin (straight line measure),
body weight, clasper length (measured from
cloaca to the tip of claspers), ovarian egg diameter
and number, and condition of the uteri. In addi-
tion, uterine embryos were counted, sexed, and
measured for total length. Counts of litter size
included any uterine eggs as well as embryos.
Stages of maturity for both females and males
were categorized simply as "mature" or "imma-
ture". Females having threadlike uteri and tiny
ovarian eggs were called immature, while those
with eggs larger than 25 mm in diameter, with
uteri containing embryos or eggs or with empty
but expanded and flaccid uteri, were designated
as mature. Males with rigid claspers were classi-
fied as being mature. Clasper length relative to
total length also gave an indication of maturity.
Table 1 . — The number of specimens examined in
this study.
Number of
specimens
Immature
Mature
Month
Male
Female
Male
Female
Jan.
13
11
18
93
Feb.
3
3
17
43
Mar,
4
13
4
44
Apr.
8
34
18
68
May
3
6
9
33
June
3
2
4
20
July
4
3
11
60
Aug.
3
1
6
19
Sept.
0
7
3
15
Oct.
0
1
1
3
Nov.
0
6
1
38
Dec.
1
8
1
10
Total
42
95
91
446
Results
The reproductive organs of scalloped hammer-
head closely resemble those of the bonnethead
shark, S. tiburo (described by Schlernitzauer and
Gilbert 1966).
As with the bonnethead shark, only the right
ovary of the scalloped hammerhead is functional,
supplying both oviducts. As the ovarian eggs ma-
ture (>25 mm diameter), they pass through the
common ostium into the oviducts, where they are
fertilized. The eggs then become encased in the
embryonic membrane as they pass through the
nidamental gland and descend into the uterus. In
the uterus, embryonic development proceeds,
nourished by a yolk sac. After a period, the uter-
ine compartments develop, which enclose the em-
bryo, and a yolk-sac placenta is implanted. After
the yolk is exhausted, the embryo is nourished
until birth by the placenta through the umbilical
stalk.
Based on the condition of the uterus and ovary,
female scalloped hammerheads became mature
at a larger size, around 210 cm, than males. All
females over 230 cm were mature.
Based on the rigidity of the claspers, male scal-
loped hammerheads reached their first maturity
at a total length of 198 cm, while all those over
210 cm were mature.
The clasper length of males increases rapidly
relative to total length, until the sharks reach
around 200 cm, at which size the clasper length/
total length relationship plateaus, suggesting
that sexual maturity has been attained (Fig.
2). Also, this size marked the approximate transi-
tion point from fiaccid to rigid claspers (Fig.
2).
It takes roughly 10 months of development
from egg formation to ovulation (Fig. 3). In Octo-
ber and November, eggs were very small, measur-
ing about 2 mm in diameter. By July they had
increased in size to about 30-38 mm and num-
bered 40-50 per female, and by August and Sep-
tember they had grown to about 40-45 mm. In
one mature female, in September we counted 28
uterine eggs and 4 ovarian ones measuring 45
mm in diameter. This suggested that the 4
ovarian eggs were ready for ovulation. Because
ovarian eggs larger than 30 mm in diameter
seemed near ovulation, we concluded that ovula-
tion occurred between July and October.
The parturition season lasts from May to July,
and the gestation period of this species was esti-
mated to be roughly 10 months (Fig. 4). No em-
390
30
26
22
X
I 18
LU
5 14
<
•• •
• • •• • •
• •• •
o
o
CO
CDO
ooo
o o
o
10
6 -
o
J.
_L
• : rigid
O flaccid
■ ■
120 140 160 180 200 220 240
TOTAL LE^4CTH (cm)
260
280
Figure 2. — Relationship between total length and clasper length in Sphyrna lewini.
Jan heh Mtr Apf M«y Jun lul Au* Sep < >cl No* Dec
Figure 3. — Monthly increase in the ovarian egg diameter of Sphyrna lewini. Error
bars represent 2 SD; numerals, sample size.
bryos were found in July and August, though
some mature sharks had large ovarian eggs
measuring about 30-40 mm in diameter during
those months. The first uterine eggs, and embryos
measuring around 2 cm were found in September.
Uterine eggs were observed until November.
After nine more months of rapid growth, the em-
bryos attained a total length of around 45 cm in
the period between May and June. These embryos
were regarded as full term because they were eas-
ily separated from placenta implying that partu-
rition would occur soon.
391
O
X
UJ
50
40
30
20
10
-i_
-L.
M
M J J
MONTH
N [)
Figure 4. — Monthly growth of embryos of Sphyrna lewini.
Solid triangles indicate the uterine eggs. Error represent 2 SD;
numerals, sample size.
Counts of the number of uterine embryos as
well as eggs of 110 gravid females (230-320 cm
TL) ranged between 12 and 38 (mean 25.8). The
relationship between the total number of uterine
embryos or eggs and the size can be described by
the regression equations (Fig. 5):
N = -26.105 + 0.179L,
where N is litter size, and L is total length in cm
of the female. There is considerable variation in
number of embryos with length, and the correla-
tion coefficient r is low (0.567), but obviously fe-
cundity is related to the size of the parent.
Examination of all females carrying developing
embryos showed that occasionally some uterine
eggs failed to develop. The number of nondevelop-
ing uterine eggs carried by a single female was
1-4.
As with litter size, the number of ovarian eggs
increased with the length of the adult female.
The embryos from 51 gravid females were
sexed; of a total of 1,281 embryos, 637 were fe-
males. The sex ratios of embryos differed by litter;
for instance, some individuals had predominantly
male (13:5) or predominantly female (23:10) lit-
ters. But, as a whole, sex ratio was about 1:1.
The ratio of males to females for immature
sharks in northeastern Taiwan waters was about
1:2 (42:91), but decreased to about 1:5 (95:446) for
mature individuals (Table 1). There was no ap-
parent increase in the relative number of males
in the catch during the parturition season.
Discussion
Scalloped hammerhead sharks are abundant in
the coastal seas around northeastern Taiwan, es-
pecially during the spring and winter, and are
captured by harpoon or drift longlines at or near
the surface. Rarely individuals smaller than 120
cm TL are captured because fishing gear and
strategy are likely selective for larger fish.
Clarke (1971) reported that in Hawaii, scalloped
hammerhead pups usually stay close to the bot-
tom.
Approximately equal numbers of male and fe-
male scalloped hammerhead sharks are born, al-
though a much higher proportion of females than
males are caught in the studied area. We are un-
certain whether scalloped hammerhead females
are more vulnerable to the fishing gear or are
simply more numerous in this area. Similar pre-
dominances in females in the catch have been
found in the scalloped hammerhead in the Gulf of
California (Klimley 1981; Klimley and Nelson
1981). During the summer in 1979 off Baja Cali-
fornia, females outnumbered males by 1.6 x at
Isla Cerrolvo, 3.8 x at the El Bajo Seamount, and
40
38-
32
28 -
o
>
a:
00
UJ
24
S 20
z
18
-
•t
-
•
•
•
•
•
• • •
-
•
• •
/
•
• • •• • X
• • •/ •
•
~
•
■
•
• •/ • • •••
•
"
•
•
"
y
• • \ •
~
•
/
•
•• • \
•
X
••
N = -26.105 + 0.179L
/
(r=0.5665)
- y
•
•
, (n=110)
•
•
•
•
•
:
J l_
•
12-
240 280 280 300 320
TOTAL LENGTH (cm)
Figure 5. — Relationship between fecundity (number of
uterine eggs or embryos) and total length of Sphyrna
lewini females.
392
3.1 X farther north at Isla Las Animas (Klimley
1981; Klimley and Nelson 1981). Similar dispari-
ties in sex ratios were also observed in sandbar
sharks by Springer ( 1960). He felt that adult male
sandbar sharks live over a large geographical and
depth range, perhaps in deep cool oceanic waters
inaccessible to the fishermen's gear, while fe-
males occur in warmer inshore water where they
are more accessible to fishermen. He suggested
that the males move inshore only to mate. We had
insufficient data to test this suggestion in our
sample area.
Castro (1983) reported that S. lewini from
North American waters probably mature at about
180 cm; he did not mention whether this length
referred to male, female, or both. Bass et al.
(1975) reported that male scalloped hammerhead
at Mozambique matured between 140 and 165
cm, reaching a maximum length of at least 295
cm; females matured at about 212 cm, reaching at
least 309 cm. Compagno (1984) reported "maxi-
mum" sizes ranging from 370 to 420 cm. Our
largest males were 305 cm TL, and the smallest
mature male was 198 cm TL. The largest female
observed was 324 cm, and the smallest mature
female was 210 cm. It seems that in our sample
area, males reached maturity at a somewhat
larger size than in the other studied areas while
females were about the same.
The close relationship of growth pattern of
uterine embryos and ovarian eggs implies that
eggs are transferred into the uterus and fertilized
immediately after parturition. If the estimate of
10 month gestation period is correct, adult fe-
males give birth once each year.
The length at birth of scalloped hammerheads
has been reported to be around 50 cm from Natal
and southern Mozambique coastal waters (Bass et
al. 1975), 43 cm from northeastern United States
to Chesapeake Bay waters (Casey 1964), 38-45
cm in North American waters (Castro 1983), and
42-55 cm combined from all oceans (Compagno
1984). Our largest uterine embryos measured
about 47 cm TL.
Pupping season appears to be during the sum-
mer in Taiwan as well as in Mozambique (Bass et
al. 1975) and North America (Castro 1983). In
Hawaii, Clarke (1971) found newborn pups
throughout the year but with increased numbers
in summer.
Fisheries Center, National Marine Fisheries Ser-
vice, NOAA, for reading the first draft of this
manuscript and offering useful comments. We
also thank Jin-Jehn Wu, Department of Fish-
eries, National Taiwan College of Marine Science
and Technology, for his help in collecting speci-
mens. Financial support was received from Na-
tional Science Council (contract: NSC 72-0409-
B019091 and NSC 73-0409-B019-02).
Literature Cited
Bass, A. J . J D. D' Aubrey, and N Kistnasamy.
1975. Sharks of the east coast of southern Africa. III. The
families Carcharinidae and Sphymidae. Oceanogr.
Res. Inst., Durban, Invest. Rep. No. 38, 100 p.
Casey, J G
1964. Angler's guide to sharks of the northeastern United
States, Maine to Chesapeake Bay. Bur. Sport Fish.
Wildl., Circ. No. 179, 32 p.
Castro, J I.
1983. The sharks of North American waters. Texas
A&M Univ. Press, 180 p.
Clarke, T A
1971. The ecology of the scalloped hammerhead shark,
Sphyrna lewini, in Hawaii. Pac. Sci. 25:133-144.
Compagno, L J. V.
1984. FAO species catalogue. Vol. 4. Sharks of the world.
An annotated and illustrated catalogue of shark species
known to date. Psirt 2. Carchariniformes. FAO Fish.
Synop. No. 125, 4(2):251-655.
Klimley, A. P
1981. Grouping behavior in the scalloped hammer-
head. Oceans 24(4):65-71.
Klimley. A P and D R. Nelson
1981. Schooling of scalloped hammerhead shark, Sphyrna
lewini, in the Gulf of California. Fish. Bull., U.S.
79:356-360.
Schlermitzauer, D a , and P W. Gilbert.
1966. Placentation and associated aspects of gestation in
bonnethead shark, Shyrna tiburo. J. Morph. 120:219-
231.
Springer, S.
1960. Natural history of the sandbar shark, Eulamia mil-
berti. U.S. Fish. Wildl. Serv., Bull. 61:1-38.
Teshima, K.
1981. Studies on the reproduction of Japanese smooth
dogfish, Mustelus manazo, and M. griseus. J. Shi-
monoseki Univ. Fish. 29:113-119.
Che-Tsung Chen
Tzyh-Chang Leu
Shoou-Jeng Joung
Graduate School of Fisheries
National Taiwan College of Marine Science
and Technology
Keelung, Taiwan
Republic of China
Acknowledgments
We sincerely thank Susumu Kato, Southwest
393
OCCURRENCE OF YOUNG-OF-THE-YEAR KING,
SCOMBEROMORUS CAVALLA , AND SPANISH,
5. MACULATUS, MACKERELS IN
COMMERCIAL-TYPE SHRIMP TRAWLS ALONG
THE ATLANTIC COAST OF
THE SOUTHEAST UNITED STATES^
King mackerel, Scomberomorus cavalla, and
Spanish mackerel, S. maculatus , are migratory
scombrids that support large recreational and
commercial fisheries along the southeast coast of
the United States (Manooch 1979). Recent evi-
dence indicates that both species may be overex-
ploited in portions of their range, prompting the
South Atlantic Fishery Management Council to
impose catch limits and landing quotas^. Many
aspects of the biology and ecology of adult mack-
erels in this region have been studied (Manooch
et al. 1978; Collette and Russo 1984), and the
larval stages have also received attention (Fahay
1975; Collins and Stender 1987). However, little
is known concerning the distribution and occur-
rence of juvenile (young-of-the-year) mackerels
along the Atlantic coast of the southeastern
United States, nor does it seem to be widely
known that large numbers of these young fishes
may be included in the bycatch of a major fishery.
This report provides preliminary information on
both of these topics.
Methods
During 1980-82 and 1985-86 the Marine Re-
sources Monitoring and Assessment Program
(MARMAP) at the South Carolina Marine Re-
sources Research Institute conducted trawl sur-
veys of the nearshore fish fauna in the South At-
lantic Bight (Cape Hatteras, NC, to Cape
Canaveral, FL). Before 1986, trawl gear consisted
of two semiballoon shrimp trawls with an 18.3 m
footrope, a 12.2 m headrope, 4.1 cm stretch mesh
in the cod end, and 1.5 x 0.9 m doors towed at 2.5
knots (4.6 m/second). In 1986, paired "tongue"
trawls with a 22.9 m footrope, 4.1 cm mesh, and
3.0 X 1.0 m doors were towed at approximately
2.5 knots. Sampling strategy and length of tow
(20 minutes to 1 hour) varied between cruises.
Station depths were 3-18 m in 1980-81 and 3-9
m in subsequent years. In each sample, all mack-
erels were identified to species and measured
(fork length), and number and total weight were
recorded for each species. We conducted two addi-
tional cruises in 1986 to test tongue trawl nets
equipped with trawl efficiency devices (TED's). A
National Marine Fisheries Service (NMFS) TED
equipped with finfish deflector"^ was installed in
one net, a Georgia TED^ in a second net, and a
third net acted as a control. The TED's were
fished against each other and against the control
for a total of 30 tows on the first cruise (July-Au-
gust 1986) and 15 tows on the second (September-
October 1986).
Tongue nets, also known as bib, falcon, cobra,
or mongoose trawl nets, have become widely used
by commercial shrimpers in some areas during
recent years (Edwards 1987). The major differ-
ence between these and other common towed
gears used in the penaeid shrimp fishery is a mod-
ified and elongated headrope that is held well
above the footrope by attachment to the trawl
warp with a third bridle. A large float, usually
attached to the center of the headrope, produces a
high, vertical mouth opening. The result is a net
that fishes a larger portion of the water column
than other common nets with similar footrope
and door configurations.
Results
During years when semiballoon nets were used,
catch per unit effort of mackerels was relatively
low (king mackerel: 0.2-0.4 individuals/net-hour;
Spanish mackerel: 0.3-2.2 individuals/net-hour).
Tongue trawl nets on four 1986 cruises gave
cruise-specific average catches of 2.5-8.7 individ-
uals/net-hour for king mackerel and 12.4-115.2
individuals/net-hour for Spanish mackerel, using
catches of control [unmodified] nets for TED
cruises (Table 1). Both species were taken as late
as 30 October, except for two king mackerel
caught in December 1982. Approximately 79% of
king and 91% of Spanish mackerel were caught in
depths <9 m during the two years in which sam-
ple depths extended to 18 m.
Mackerels taken in shrimp trawls were almost
entirely juvenile fishes. Rather than pool catches
between cruises that often differed in time of
year, geographic area, and sampling strategy, ex-
amples of length frequencies of Spanish and king
iContribution No. 244 of the South Carolina Marine Re-
sources Center.
^South Atlantic Fishery Management Council, Council Meet-
ing Summary, 27-29 April 1987, Charleston, SC.
3Described and illustrated in the Federal Register, vol. 52, no.
124; Monday, June 29, 1987 - Rules and Regulations; p. 24244-
24262.
394
FISHERY BULLETIN: VOL. 86, NO 2, 1988.
Table 1. — Results of nearshore cruises using two types of shrimp trawls.
Cruise dates
Net type
Sampling!
strategy
Sample area2
Total
net-hour
No.
Spanish
No.
kings
July-Sept. 1980
semiballoon
SR + NR
NC,
SC, GA, FL
114
250
40
Apr.-June 1981
semiballoon
SR
NC.
SC. GA. FL
77
21
13
Sept. 1982-
Jan. 1983
semiballoon
SR
NC,
SC, GA, FL
73
27
27
Aug. -Sept. 1985
semiballoon
SR + NR
NC,
SC, GA, FL
59
120
36
Aug. -Sept. 1986
tongue
NR
NC,
SC, GA, FL
41
1,421
104
Oct. 1986
tongue
SR
NC,
SC, GA
55
681
481
July-Aug. 1986
tongue
NR
SC
20
2,303
84
Sept.-Oct. 1986
tongue
NR
SC
20
327
80
^SR = stratified random sampling; NR = nonrandom sampling.
2At least some portions of the waters of these states were sampled.
mackerels taken in tongue nets on two cruises are
presented in Figures 1 and 2, respectively. Com-
parisons of the catch rates of Spanish mackerel
during the TED evaluations in July- August 1986
showed that significantly fewer were taken in the
tongue net equipped with a NMFS TED than the
control net (Mann-Whitney: P < 0.001) or the
Georgia TED-net (P < 0.05) (Table 2). There were
no significant differences in catches of king mack-
erel (Table 3), or in catches of either species be-
tween nets on the September-October cruise, per-
haps due to the smaller sample sizes.
Figure 1. — Length-frequency distribution of
Spanish mackerel taken in shrimp tongue trawl nets
during July-August 1986 along the southern At-
lantic coast of the United States in = 2,303).
FORK LENGTH (cm)
395
130 -I
24 28 32 36
FORK LENGTH (cm)
r
40
I
44
Figure 2. — Length-frequency distribution of king mackerel taken in shrimp tongue trawl nets
during October 1986 along the southern Atlantic coast of the United States (n = 481).
Table 2. — Catches of Spanish) mackerel during evaluations of tongue nets equipped with trawl efficiency devices
(TED'S). C = 22.9 m footrope length tongue net; NMFS = 22.9 m footrope length tongue net with a NMFS TED;
GA = 22.9 m footrope length tongue net with a Georgia TED. Tow times = 1 hour.
No. of
Net
No. of Spanish mackerel
Net
No. of Spanish mackerel
Cruise
Comparison
tows
type
Total
X ±SD
Range
type
Total
X ±SD
Range
July-Aug.
C-NMFS
10
C
1,104
110 ±27
68-144
NMFS
519
52 ±23
12-73
C-GA
10
C
1,199
120 ±65
40-276
GA
1,219
122 ± 74
44-300
NIVIFS-GA
10
NMFS
880
88 ±53
35-176
GA
1,650
165 ±142
67-533
Sept.-Oct.
C-NMFS
5
C
104
21 ± 15
2-41
NMFS
53
11 ± 11
2-27
C-GA
5
C
206
41 ±38
1-103
GA
155
31 ±28
0-76
NI^FS-GA
5
NMFS
108
22 ±23
4-58
GA
149
30 ±25
3-59
Table 3. — Catches of king mackerel during evaluations of tongue nets equipped with trawl efficiency devices
(TED'S). C = 22.9 m footrope length tongue net; NMFS = 22.9 m footrope length tongue net with a NMFS TED;
GA = 22.9 m footrope length tongue net with a Georgia TED. Tow times = 1 hour.
Comparison
No. of
tows
Net
type
No.
of king
mackerel
Net
type
No.
of king
mackerel
Cruise
Total
X ±
SD
Range
Total
X ±
SD
Range
July-Aug.
C-NMFS
10
C
41
4±
4
1-13
NMFS
46
5±
9
0-
-30
C-GA
10
C
43
4±
5
0-15
GA
41
4 ±
4
0-
-13
NMFS-GA
10
NMFS
23
2±
3
0-8
GA
26
3±
5
0-
-17
Sept.-Oct.
C-NMFS
5
C
26
5±
7
0-15
NMFS
14
3±
3
0-
-5
C-GA
5
C
40
8±
8
4-22
GA
22
4 ±
5
1-
-14
NMFS-GA
5
NMFS
24
5±
2
2-8
GA
20
4±
4
2-
-11
Discussion
Although it is possible that juvenile mackerels
were more abundant in 1986 than in previous
years, the increased catches of these fishes in
tongue nets over semiballoon nets suggests that
the former are much more efficient in capturing
these fishes. Preliminary data from a gear com-
parison cruise in 1987 indicate that tongue nets
do catch more pelagic fishes than semiballoon
nets even after adjusting for differences in
footrope lengths (G. Sedberry"*). Unfortunately,
■*G. Sedberry, Marine Resources Research Institute, South
Carolina Wildlife and Marine Resources Department, P.O. Box
396
these tests were conducted during March-April
when juvenile mackerels are rare in the coastal
waters of South Carolina.
In 1986 we collected juvenile mackerels during
July through October. Because of incomplete tem-
poral sampling, we do not know if they were
present earlier and later in the year in this re-
gion. Based on the occurrence of early larval
stages, spawning of both mackerels in the South
Atlantic Bight extends from May through at least
September (Collins and Stender 1987). If growth
rate estimates of ca. 3 mm/day for juveniles are
correct (M. R. Collins, unpubl. data), king mack-
erel spawned in early May could be recruited into
the bycatch of the shrimp fishery in June. Late-
spawned fish from the previous year may also be
present at this time. In South Carolina, the open
season for commercial trawling for penaeid
shrimps in state waters usually extends from
June through December, which coincides with the
presence of juvenile mackerels in the heavily
fished nearshore waters. In addition, mackerels
were much more abundant in tows made in
depths <9 m, which includes the preferred
shrimping areas, than in deeper waters. This may
be due either to greater abundance in these
depths or to greater catchability in response to
the fact that the trawl nets fish a larger portion of
the water column in shallower areas.
It is difficult to accurately estimate the bycatch
of mackerels in the commercial shrimp fishery
owing to lack of current, detailed information
from throughout the region on number of vessels,
effort expended, gears used, and areas fished.
However, our catch rates suggest that the impact
of tongue nets on mackerel stocks may be signifi-
cant. As the current status of these stocks is such
that strong restrictions have been imposed on
both the recreational and commercial fisheries, it
is unfortunate that the situation may be exacer-
bated by a potentially large bycatch of juvenile
mackerels in the shrimp fishery. More informa-
tion is needed on the ecology and behavior of
young mackerels, and their vulnerability to vari-
ous gears, in order to resolve this conflict.
Acknowledgments
We wish to express our appreciation to the
many people who assisted in the tedious tasks of
collecting and processing trawl samples over the
years, including the captains and crews of the
vessels involved. This work was sponsored by the
National Marine Fisheries Service (Southeast
Fisheries Center), the South Carolina Wildlife &
Marine Resources Department, and the Sport
Fishery Research Foundation.
Literature Cited
COLLETTE. B B . AND J L RUSSO
1984. Morphology, systematics, and biology of the
Spanish mackerels (Scomberomorus , Scombridae).
Fish. Bull., U.S. 82:545-692.
Collins, M R . and B W Stender.
1987. Larval king mackerel (Scomberomorus cavalla),
Spanish mackerel (S. maculatus), and bluefish iPo-
matomus saltatrix ) off the southeast coast of the United
States, 1973-1980. Bull. Mar. Sci. 41:822-834.
Edwards. M L.
1987. The shrimp trawl. Natl. Fisherman 67(13):89-92.
Fahay, M P
1975. An annotated list of larval and juvenile fishes cap-
tured with surface-towed meter net in the South Atlantic
Bight during four RV Dolphin cruises between May 1967
and February 1968. U.S. Dep. Commer., NOAA Tech.
Rep. NMFS SSRF-685, 39 p.
Manooch. C S , III
1979. Recreational and commercial fisheries for king
mackerel, Scomberomorus cavalla , in the South Atlantic
Bight and Gulf of Mexico, U.S.A. In E. L. Nakamura
and H. R. Bullis (editor), Proceedings of the mackerel
colloquium, p. 33-41. Gulf States Mar. Fish. Comm.,
Brownsville, TX.
Manooch. C S , III. E L. Nakamura, and A B Hall
1978. Annotated bibliography of four Atlantic scombrids:
Scomberomorus brasiliensis , S. cavalla, S. maculatus,
and S. regalis. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS Circ. 418, 166 p.
Mark R Collins
Charles A. Wenner
Marine Resources Research Institute
South Carolina Wildlife and Marine
Resources Department
P.O. Box 12559
Charleston, SC 29412
12559, Charleston, SC 29412, pers. commun. May 1987.
STOMACH CONTENTS OF COMMERCIALLY
CAUGHT HUDSON RIVER STRIPED BASS,
MORONE SAXATILIS , 1973-75
The Hudson River estuary is a detritus-driven
ecosystem. Only a few of the 100 or more reported
fish species function as tertiary piscivores more
typical of a grazing food chain. Of these few spe-
cies, which include the American eel, Anguilla
rostrata, and the summer-transient juvenile
FISHERY BULLETIN: VOL. 86, NO. 2. 1988.
397
bluefish, Pomatomus saltatrix, striped bass, Mo-
rone soxatilis, are probably the most important in
terms of biomass and commercial value. How-
ever, commercial fishing for the American eel and
striped bass has been banned in the Hudson since
1976 because both species are contaminated with
higher levels of toxic polychlorinated biphenyls
(PCB's) than most other Hudson River fish spe-
cies.
To date, only one published paper has touched
upon the diet of prespawning Hudson River
striped bass larger than 400 mm TL (Gardinier
and Hoff 1982), and their results for fish of that
size are based on less than 10 samples for which
fish remains were identified to species or family
level. The present paper describes the findings of
stomach content analysis for 510 striped bass,
most of which were prespawning adults larger
than 400 mm TL (Fig. 1) collected during the
spring months of 1973-75 by commercial gill nets
in the Tappan Zee region of the Hudson River.
Description of Sampling Area and Methods
The Tappan Zee gill-net fishery is located ap-
proximately 90 m south of the Tappan Zee Bridge
at river km 43.5 (measured from the Manhattan
Battery) in a relatively shallow section that is 4.5
km wide, a location through which many of the
spawners move on the way to upstream spawning
grounds. The staked gill nets were set perpendic-
ularly to the north-south current in water depths
of 2.4-9.8 m and spanned a distance of approxi-
mately 1.6 km out to the dredged channel on the
east side of the river. Stretched mesh sizes were
11.4, 11.8, 12.1, 12.7, 13.3, and 14.0 cm. Water
temperatures during the sampling period ranged
from approximately 6° to 19.5°C, and the salinity
was essentially that of freshwater, rarely rising
above 300 mg/liter. Sampling dates included the
following periods in each year: 28 March-9 May
1973, 8 April-20 May 1974, and 6 April-19 May
1975. Fish were collected from the nets at 8-12 h
intervals, placed on ice, and taken directly to a
laboratory for total length (TL) measurements
and stomach content analysis. Prey items were
identified with the aid of a binocular dissecting
scope and frequency of occurrence was noted.
Results
Stomach Content Analysis
A summary table of stomach content data,
pooled for the three years of analysis (Table 1),
shows that adult, prespawning striped bass were
highly piscivorous and that 89% of the stomachs
containing identified food items contained fish
and 21% contained invertebrate remains, primar-
ily those ofCrangon species. Clupeid species were
the most prevalent fish, with blueback herring,
Alosa aestivalis, predominating. Most of the clu-
peids were adult, prespawning herring, approxi-
mately 200 mm TL.
NUMBER OF FISH = 510
140
400 450 500 550 600 650 700 750 800 850 900 950 1000 1050
TOTAL LENGTH (mm)
Figure 1. — Length distribution of Hudson River striped bass examined for stomach con-
tents.
398
Table 1— Stomach contents of adult prespawning striped bass collected by commercial gill net during
spring 1973, 1974, and 1975.
% Frequency of
% Frequency of occurrence
Number of
occurrence based on
based on number of
stomachs
total stomachs
stomachs containing
in which
analyzed
identified food items
Food item
item occurred
(n = 510)
(n =201)
Fish
179
35.1
89.1
Clupeidae
73
14.3
36.3
Alosa pseudoharengus
13
2.5
6.5
Alosa aestivalis
29
5.7
14.4
Brevoortia tyrannus
3
0.6
1.5
Clupeidae
28
5.5
13.9
Morone spp.
9
1.8
4.5
Morone americana
6
1.2
3.0
Morone saxatilis'^
1
0.2
0.5
Morone spp.
2
0.4
1.0
Other fish
29
5.7
14.4
Ammodytes americanus
14
2.7
7.0
Microgadus tomcod
7
1.4
3.5
Osmerus mordax
2
0.4
1.0
Syngnathus fuscus
2
0.4
1.0
Anchoa mitciiilii
2
0.4
1.0
Ictalurus catus
1
0.2
0.5
Scomber scombrus
1
0.2
0.5
Unidentified fish remains
68
13.3
33.8
Invertebrates
43
8.4
21.4
Crangon spp.
38
7.5
18.9
Polyctiaeta
5
1.0
2.5
Unidentified contents
11
2.2
Not applicable
Total stomachs analyzed
510
Number (percent) containing identified food items
201 (39.4)
Number (percent) containing food material (identified and unidentified)
212 (41.6)
Number (percent) empty
296 (58.4)
'Two yearling striped bass in female collected 15 April 1974.
The next most abundant fish prey species in the
composite 1973-75 sample was the American
sand lance, Ammodytes americanus. Sand lance
were present in striped bass stomachs only during
1975, when the frequency of occurrence was about
ll9c (Fig. 2). Atlantic tomcod, Microgadus tom-
cod, were found in striped bass stomachs only
during 1973 when the frequency of occurrence
was about 20%, somewhat higher than the fre-
quency of clupeid species for that year (Fig. 2).
These fish were yearling, postspawning Atlantic
tomcod averaging about 125 mm TL. The Morone
species, white perch, M. americana, and striped
bass, had a frequency of occurrence of 4.5% for the
composite sample and were present in striped
bass stomachs every year (Fig. 2).
Discussion
Similar to findings in this study, Trent and
Hassler (1966) found that blueback herring were
predominant in the diet of adult, prespawning
striped bass from the Roanoke River, NC, and
Manooch (1973) found that herring ranked second
after Atlantic menhaden, Brevoortia tyrannus , in
the spring diet of adult striped bass from Albe-
marle Sound, NC. Data from Gardinier and HofF
(1982) do not indicate that blueback herring or
clupeids were important in the diet of prespawn-
ing Hudson River striped bass collected during
1976 and 1977, but fish remains in bass stomachs
were generally not identified to the family or spe-
cies level in their study.
American sand lance are rarely found over
muddy bottoms (Bigelow and Schroeder 1953;
Leim and Scott 1966) or within estuaries (Mass-
man 1960; Norcross et al. 1961), but they ranked
second only to blueback herring in the diet of
Hudson River striped bass. Sand lance were not
reported from the Hudson River prior to 1975
when adults and larvae were relatively abundant
in February-April ichthyoplankton collections
south of the Tappan Zee Bridge (Dew and Hecht
1976). Data from the present study (Fig. 2) and
from early spring ichthyoplankton sampling dur-
ing 1976 (Dew, unpub. data), indicate that sand
lance are not abundant every year in the Hudson,
but it is probable that during some years adult
sand lance serve as important alternate prey for
prespawning striped bass.
399
o
o
o
o
>-
o
3
o
8?
100
80
.. 60
40
20
^ CLUPEIDS
[^ MORONE SPP.
^ ATLANTIC TOMCOD
□ SAND LANCE
□ UNIDENTIFIED FISH REMAINS
N = 36
i
it
-iii
1973
N = 39
i
Fl
i
I
i
A
N =126
1974
1975
N = 201
P
i
i
^±.„L^
1973- 1975
Figure 2. — Percent frequency of occurrence offish prey categories based on number of stomachs containing identified food.
The fact that Atlantic tomcod were found only
during 1973 may be due to a sampling artifact
because 1973 was the only year in which March
samples were taken, and all tomcod prey were
found during March 1973. In March, gravid
alewives, Alosa pseudoharengus , and blueback
herring have not yet moved into the estuary and,
prior to the arrival of these apparently preferred
prey species, Atlantic tomcod may serve as an
alternate forage fish. Thus the winter-spawning
Atlantic tomcod may be a regular diet item at a
time when alternate prey species are not avail-
able.
It is evident from Table 1 that there are several
alternate fish prey for striped bass in the Hudson
River (e.g., blueback herring, sand lance, alewife,
Atlantic tomcod, white perch, and striped bass),
and it is expected that striped bass predation
upon these species would be of a compensatory
nature. In a simple predator-prey situation where
there are no alternate prey, predation is often
depensatory (Neave 1953) because a population of
predators would tend to seek out and consume a
high proportion of small year classes and a lower
proportion of strong year classes. In a more com-
plex community where there are several species
of prey, predation may be compensatory if preda-
tors change their feeding habits in response to the
availability of food (Ivlev 1961; Forney 1971).
That is, if several equally suitable prey species
were available within a system, the predator spe-
cies would tend to feed more heavily on the most
abundant species, thus acting in a compensatory
manner.
The ratio of white perch to striped bass in stom-
ach samples identified to the species level is 6:1
(Table 1). This ratio is based on a small sample
size, but it closely approximates the long-term
average ratio of 5.9:1 for annual CPUE values for
white perch and striped bass collected in several
hundred bottom trawls from the Haverstraw Bay
region (Milepoint 36) during 1971-77 (Lawler,
Matusky and Skelly Engineers unpubl. data). In
other words, yearling and older white perch and
yearling striped bass may be equally attrac-
tive prey species, and their frequency of oc-
currence in adult striped bass stomachs may
depend primarily upon the frequency of ran-
dom encounters rather than active prey selec-
tion. If this were true, the frequency of can-
nibalism should be greatest during those years
when the abundance of yearling striped bass rela-
tive to white perch (and other species) was great-
est.
Acknowledgments
Financial support for this project came from
Lawler, Matusky and Skelly Engineers; Con-
solidated Edison of New York; Central Hudson
Gas and Electric; and Orange and Rockland
Utilities.
400
Literature Cited
BiGELOW. H. 3., AND W. C. SCHROEDER
1953. Fishes of the Gulf of Maine. Fish. Bull. 53:1-
577.
Dew. C B . and J H Hecht
1976. Observations on the papulation dynamics of At-
lantic tomcod Microgadus tomcod in the Hudson River
estuary. Proc. Fourth Symposium On Hudson River
Ecology, Paper 25. Hudson River Environmental Soci-
ety, Bronx, N.Y.
Forney, J L
1971. Development of dominant year classes in a yel-
low perch population. Trans. Am. Fish. Soc. 100:739-
749.
GaRDINIER. M N , AND T B HOFF
1982. Diet of striped bass in the Hudson River estuary.
New York Fish Game J. 29(2):152-165.
IVLEV, V S.
1961. Experimental ecology of the feeding of fishes.
Yale Univ. Press, New Haven, 302 p.
Leim. a H . AND W B Scott
1966. Fishes of the Atlantic coast of Canada. Fish. Res.
Board Can. Bull. 155, 485 p.
Manooch, C S.
1973. Food habits of yearling and adult striped bass, Mo-
rone saxatilis (Walbaum), from Albemarle Sound, North
Carolina. Chesapeake Sci. 14(2):73-86.
Massmann, W H
1960. Additional records for new fishes in Chesapeake
Bay. Copeia 1960:1-70.
Neave, F.
1953. Principles affecting the size of pink and chum
salmon populations in British Columbia. J. Fish. Res.
Board Can. 9:450-491.
Norcross, J J , W H Massmann, and E. B. Joseph
1961. Investigations of inner continental shelf waters off
lower Chesapeake Bay. Part II. Sand lance larvae,
Ammodytes americanus. Chesapeake Sci. 2(l-2):49-
59.
Trent, L., and W. W. Hassler.
1966. Feeding behavior of adult striped bass, Roccus sax-
atilis , in relation to stages of sexual maturity. Chesa-
peake Sci. 7(4):189-192.
C. Braxton Dew
Lawler, Matusky, and Skelly Engineers
One Blue Hill Plaza
Pearl River, NY 10965
Present address;
Northwest and Alaska Fisheries Center Kodiak Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 1638
Kodiak, AK 99615
ACCUMULATION OF AGE PIGMENTS
(LIPOFUSCIN) IN
TWO COLD-WATER FISHES
In fisheries management, age structured models
are the preferred method for meeting the key ob-
jectives of estimating optimal yields and deter-
mining the effect of fishing on stock structure
(Gulland 1978). However, fev^^ species of commer-
cial marine fishes exist in which age can be deter-
mined with certainty (Boehlert 1985). The con-
centration of age pigments (lipofuscin)
(Ettershank 1984) in fish tissues (Agius and
Agbede 1984) may be a measure of fish age that
could be used to validate other ageing techniques
and might also improve estimates of age of long-
lived species where other techniques are difficult
to apply.
Lipofuscin originates in biological membranes
through lipid peroxidation (Tappel 1975). Lipo-
fuscin accumulation has been documented for a
wide variety of animals, from mammals to the
bread-mold Neurospora (Ettershank et al. 1983
and references therein). The rate of accumulation
has been shown to be constant during the lifetime
of laboratory-raised mice (Reichel 1968; Miquel et
al. 1978), dogs (Munnell and Getty 1968), fiesh-
flies (Ettershank et al. 1983); man (Strehler et al.
1959); and also wild populations of mice (Dapson
et al. 1980). On the other hand, the rate of lipofus-
cin accumulation has been shown to vary with
level of activity and lifespan (Sohal and Donato
1978). It is expected that the rate in which lipo-
fuscin accumulates with age in a natural popula-
tion of fishes, or stock, with its free genetic inter-
change and likely common habitat, would be
fairly uniform (but see Smith 1987). In addition,
if measurements are made in nonmitotic and
constantly metabolizing tissues such as brain
or myocardium, the variation in concentra-
tion due to environmental effects is least like-
ly.
In this paper we present the results of a prelim-
inary study designed to assess the usefulness
of extracted lipofuscin as a method of ageing
fishes. Two species of cold-water fishes are in-
cluded: the rainbow trout, Salmo gairdneri,
reared in captivity, and of known age; and the
Dover sole. Microstomas pacificus, a long-lived
fish, in which age is not known with certain-
ty. We present spectral characteristics of the ex-
tracted lipofuscin from several tissues and the
change in concentration of lipofuscin with fish
age.
FISHERY BULLETIN: VOL. 86, NO. 2
401
Materials and Methods
Specimens of Dover sole were collected on De-
cember 1986 ofTPoint Conception. Ages were esti-
mated by J. Butler, E. Lynn, and M. Drawbridge
of the Southwest Fisheries Center using otolith
sections. Ages ranged from 2.7 to 44.7 years (av-
erage of three readings). Specimens of the species
Salmo gairdneri were collected in July 1986 at
Hot Creek Hatchery. Ages ranged from 3 months
to 3 years.
Samples were frozen after collection and kept
at -80°C until analysis. Studies by Nicol (1987)
showed that this form of preservation yielded the
lowest fluorescence when compared with ethanol
and formalin-preserved samples. There was no
indication that the level of fluorescence changed
as a function of time of preservation or by inter-
action with the extracted lipofuscin, as in the
case for formalin. To excise the brain, the top of
the skull was opened and the four brain lobes
were removed as a unit, without the optic nerve.
To excise the heart, we separated the muscle
from connective tissue, blood vessels, and fat de-
posits.
Three methods were compared for maximum
lipofuscin extraction. The first two methods (Tap-
pel 1975; MacArthur and Sohal 1982) were specif-
ically developed for lipofuscin extraction and
employed chloroformrmethanol (2:1) as the ex-
tractive solvent; they differ in the optimal
volume-to-weight ratio (30:1 and 20:1, respec-
tively), temperature of extraction, and number of
times the chloroform phase is washed with water.
The third method was developed for lipid extrac-
tion in fishes (Bligh and Dyer 1959) and uses chlo-
roform:methanol:water (1:2:0.8) as the extractive
solvent. Subsamples (n = 3) of cerebellum of
Stenella sp. collected at the eastern tropical
Pacific and kept at -30°C were defrosted, dried
with lint-free paper, weighed, and extracted fol-
lowing the three methods as described originally.
Tissues from Dover sole were extracted basi-
cally following the MacArthur and Sohal (1982)
technique, with two additional steps to ensure a
complete washing out of flavoproteins and pho-
tooxidation of retinol when present (Fletcher et
al. 1973). Tissue was freeze dried prior to analy-
sis. Ten mL of chloroform: methanol (2:1, v/v) was
added to a homogenizer containing the sample for
a final solvent:sample ratio of about 120:1 (vol/
dry weight). The sample was ground with a teflon
pestle attached to an electric drill and later sub-
merged in a 40°C water bath for 1 minute. A 2 mL
subsample was taken from the homogenate.
Three mL of deionized water was added, the sam-
ple was shaken, and the emulsion centrifuged 10
minutes at 1,912 g and 0°C. After centrifugation,
the hyperphase was decanted and a second rinse
performed. After decanting the hyperphase a sec-
ond time, 1 mL of the hypophase (chloroform con-
taining lipofuscin) was sampled and transferred
to a polypropylene tube. Three mL of chloroform
were added to the sample (for a total of 4 mL)
which was then exposed to UV irradiation (254
nm) for 1-2 minutes to photooxidize retinol. This
last step was routinely performed for liver tissue.
Samples were then transferred to glass tubes,
sealed, and kept in the refrigerator until analysis.
Sample fluorescence was measured at the emis-
sion peak (430-440 nm) in a quartz cuvette with
a Perkin Elmer Fluorescence Spectrophotometer^
Model MFP-44A. The sample was excited at the
peak of fluorescence excitation (—360 nm). The
intensity of the fluorescent emission (at 430 nm)
was normalized to the intensity of the quinine
solution standard (1 mg L"^ in IN sulphuric acid)
and expressed in fluorescence units.
Lipofuscin in Dover sole is expressed as 1) total
lipofuscin content per organ and 2) weight-
specific lipofuscin concentration, calculated by di-
viding the total lipofuscin content by the dry
weight of the entire organ.
Lipofuscin ft-om rainbow trout tissues was ex-
tracted following the Bligh and Dyer (1959) tech-
nique, without modifications. Whole tissues were
ground in water with a tissue homogenizer to give
a final concentration of 100 mg of tissue (wet
weight) in 0.8 mL of homogenate. A sample of 0.8
mL was taken and solvents were added to give a
final ratio of 1:2:0.8 (chloroform:methanol:water)
with a solvent to sample ratio of 20:1 (vol/w). The
sample was then filtered through a 2.4 cm glass
fiber filter (Whatman GF/C), the tissue re-
extracted with 1 mL of chloroform and refiltered.
The extract was then washed with 3 mL of water,
shaken, and centrifuged for 10 minutes at 1,912 g
and 0°C. One mL of the chloroform hypophase
was subsampled and fluorescence estimated as
described above.
Lipofuscin in rainbow trout is expressed as
1) total lipofuscin content per organ and
2) weight-specific lipofuscin concentration (total
lipofuscin content of the organ divided by its wet
weight).
iReference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
402
Results
EfiBciency of Extraction
Lipofuscin was successfully extracted and
quantified from three different fish tissues: brain,
heart, and liver. Wavelengths of fluorescence ex-
citation and emission maxima of extracted lipo-
fuscin in chloroform are presented in Figure 1.
All maxima are within the range cited by Shel-
dahl and Tappel (1973); fluorescence excitation
maxima were between 340 and 370 nm and fluo-
rescence emission maxima between 420 and 470
nm.
Retinol (wavelengths of maximum fluorescence
excitation at 325-340 nm and maximum fluores-
cence emission at 475 nm) was photooxidized by
exposing the chloroform extract to UV irradiation
(254 nm). As expected, retinol was found in liver
and sometimes it was present in brain and heart
tissues. For example, wavelengths of fluorescence
excitation and emission maxima in liver of Dover
sole shifted from 352 to 365 nm and from 470 to
440 nm, respectively, after UV irradiation (see
Figure IC). Mullin and Brooks (1988) also found
that this UV irradiation is effective in oxidizing
retinol although they did not find significant
retinol interference in fish tissue. It seems liver
tissues may require UV irradiation to oxidize
retinol while brain and heart tissue should be
checked for retinol presence before irradiation.
The extractive efficiency of the three methods
tested are compared in Table 1. All methods ex-
tracted similar fluorescent compounds from the
brain oi Stenella sp., as the fluorescence excita-
tion and emission spectra were similar. The
MacArthur and Sohal (1982) method extracted
Table 1. — Comparison of the three methods of lipofuscin extrac-
tion in brain tissue Fluorescence excitation at 360 nm and fluores-
cence emission at 440 nm. Results are presented as fluorescence
units per mg of wet tissue, where the fluorescence emission signal
is normalized to the intensity of emission of a standard quinine
sulfate solution (1 mg L Hn IN sulphuric acid). The three methods
were significantly different from each other, P < 0.05, Newman-
Keuls range test (Zar 1974). ANOVA: F ratio = 21.81, 2 df,
P < 0.05.
Fluorescence
units
Fluorescence
maxima
Average
10-3
SD
10-3
n
Excitation
(nm)
Emission
(nm)
f^acArthur and
Sohal (1982)
Tappel (1975)
Bligh and Dyer
(1959)
4.30
2.83
1.98
0.39
0.37
0.45
3
3
3
362
362
360
445
440
445
_L
_L
_L
3E0 360 400 440 480 520
>-
(/)
Z
LiJ
UJ
U
z:
UJ
o
UJ
(r
o
Z)
B
365
1
435
1 1 1
A
1
AZ)
1
1
320 360 400 440 480 520
C
365
430
450
1
u
^l
r
t(l) //
\(3) N
/
\ r
\ \
\ N
/
W/
\ \
\ S
/
A
\ X
X X
\^ X
1 1
it
1
1 1 1
320 360 400 440 480 520
WAVELENGTH (nm)
Figure 1. — Spectral characteristics of lipofuscin extracted from
Dover sole, Microstomus pacificus, tissues (uncorrected spec-
tra). In chloroform. (A) Brain: (1) Lipofuscin fluorescence exci-
tation spectrum (emission wavelength 440 nm); (2) lipofuscin
fluorescence emission spectrum (excitation wavelength 365
nm). (B) Heart: (1) Lipofuscin excitation spectrum (emission
wavelength 430 nm); (2) lipofuscin fluorescence emission spec-
trum (excitation wavelength 365 nm ). (C ) Liver: ( 1 ) Lipofuscin
fluorescence excitation spectrum (emission wavelength 430
rmi); (2) lipofuscin emission spectrum (excitation wavelength
365 nm) before UV radiation; (3) Same as (2) after UV radiation
(254 nm).
significantly more fluorescent pigment, than did
either the Tappel (1975) or the Bligh and Dyer
(1959) method which extracted 66% and 46% of
maximum extraction, respectively). Thus all
three methods are useful for quantitative estima-
tion of extractable lipofuscin in fish tissue but
403
comparisons of amount of concentrations between
tissues or species cannot be made if different ex-
traction methods are used.
Lipofuscin Concentration
The total concentration of extracted lipofus-
cin in Dover sole brain tissue was positively
correlated with fish length (Fig. 2A)
{Y = -15.4 + 0.516X,r2 = 0.43;6 ^0,P<0.01,
t - 5.1, df=34). Furthermore, total extracted
lipofuscin content in the brain was positively cor-
related with age in organisms estimated by annu-
5 300
>-
cr
Q
LlI
o
CO
UJ
cc
o
200
100
-A
o
o
o
-
(7
^
-
o
/
CTTL
o
5
o
o
-
^
o
o O
-
o
~
Y = -l54+05 X
1
1
1
1
1 1
0 08
B
o
o
0
[ct
-
^X
<"
X 006
-
oo yU
UJ
^ 004
Q
<
m 002
-
^ o
0
7
o
CD
O y
/ o
%3
oo
o
-
Y
= -19 1
10"' +0.2
10"^ X
0
1
1
1
lar increments in otoliths to be from 2 to 15 years
old (Fig. 3A) (Y = -2.5 + 1.0 X, r'^ = 0.75; 6^0,
P < 0.01, t - 5.5, df = 10). On the other hand, in
older Dover sole (estimated age from 20 to 45
years; n = 26) the total extracted lipofuscin in the
brain did not increase with their estimated age.
The concentration of lipofuscin in the brain (ex-
tracted lipofuscin concentration per unit of g dry
weight) followed a similar pattern (Fig. 3B). The
concentration in the brain increased linearly with
age for fish 2-15 years old (7 = 30.39 + 14.85 X,
r2 = 0.79, b i=0, P < 0.01, t = 6.66, df = 10) but
fishes older than 15 years did not show an in-
crease in pigment concentration with age.
The total extracted lipofuscin content extracted
ft-om rainbow trout brains increased with age
(Fig. 4A) (Y = 0.14 + 0.18 X, r^ = 0.37; 6^0,
P < 0.01, t = 3.75, df = 23), as did the content of
the heart (Fig. 4C) (Y = -0.54 + 0.82 X,
r2 = 0.67; 6 ^ 0, P < 0.01, t = 7.36, df = 23), and
liver (Fig. 4E) (Y = -2.86 + 5.81 X, r^ - 0.62;
b i=0,P < 0.01, t = 7.54, df = 34) from 3 months
to 3 years. The concentration of extracted lipo-
fuscin per unit wet weight of heart tissue did
not change with age (Fig. 4D) whereas that of
brain and liver tissue decreased with age (Fig.
4B, F).
<
cc
CD
CC
UJ
CL
UJ
O
111
o
(O
UJ
cc
o
Z)
10
200 400
LENGTH (mm)
600
oo o
Figure 2. — Dover sole, Microstomas paciftcus, brain:
(A) Extracted lijwfuscin per unit of dry weight as a function of
fish length (mm) (n = 36); (B) brain dry weight (g) as a func-
tion of fish length (mm). Fluorescence units: the intensity of
fluorescence normalized to a standard solution of quinine sul-
fate (1 mg L-i in IN sulphuric acid).
Figure 3. — Lipofuscin in Dover sole. Microstomas
pacificas, brain tissue as a function of age in = 36).
(A) Total lipofuscin content per brain; (B) weight-
specific (g dry weight) lipofuscin in brain.
I
a
UJ
>-
cc
Q
to
I-
UJ
to
UJ
cc
o
300
200
100
o B
o o°
Yn.,<s = 30 4 +I4 85X
-J I I I L
10 20 30
AGE (years)
40
404
<
o
cc
o
(r
a
UJ
o
CO
UJ
on
o
3
B
(
1
: J
}
1
1 1
AGE (yeors)
Figure 4. — Lipofuscin in rainbow trout, Salmo gairdneri, as a function of age (n = 21). (A) Total lijxjfuscm in
brain; (B) weight-specific (mg wet weight) lipofuscin in brain; (C) total lipofuscin in heart; (D) weight-specific
(mg wet weight) lipofuscin in heart; (E) total lipofuscin in liver; (F) weight-specific (mg wet weight) lipofuscin in
liver.
Discussion
Our results indicate that lipofuscin accumu-
lated in the brain of Dover sole with time (Fig.
2A). The concentration of lipofuscin increased
over a wide range of fish lengths and with esti-
mated age of 15 years but did not increase with
older fish. Several explanations exist for the lack
of change in concentration in older fish: 1) older
fish were incorrectly aged; 2) growth of brain
tissue masked the actual rate of accumulation;
and 3) the rate of accumulation changed during
the lifespan of this species due to changes in
metabolic activity. Either one or several of these
factors may cause the lack of lipofuscin accumula-
tion in older fish. We consider each of these issues
below.
We do not know the accuracy of the age deter-
mination in Dover sole but believe it is unlikely
that 3 readers would confuse fish aged 30-40
years with those of 20 years. Although future re-
search will shed more light on this controversial
subject, we think that grossly inaccurate age de-
termination is the least likely an explanation.
A key difference between fishes and other or-
ganisms in which lipofuscin accumulation has
been clearly documented as a function of chrono-
logical age (mammals and invertebrates) is that
405
fishes have indeterminate growth. For example,
Figure 3B indicates that the brain of Dover sole
continues to grow long after this fish begins re-
production (450 mm). Thus, unlike mammals and
invertebrates which have determinate growth,
the total content of extractable lipofuscin in any
fish organ cannot be used as a measure of age but
rather concentration (content per unit of weight)
must be used. Furthermore, even the concentra-
tion of lipofuscin in the brain, a slow growing
tissue, is a function of both the rate of lipofuscin
accumulation and the rate of tissue growth. It
follows then, that if the rate of brain growth is
such that it masks the actual rate of lipofuscin
accumulation in older organisms, lipofuscin con-
centration will not increase with time.
A critical assumption underlying the use of
lipofuscin as a determinant of age is that
metabolic rate, and hence lipofuscin accumula-
tion rate, remains fairly constant over the portion
of the life history of interest. The rainbow trout
were reared in a hatchery where no net change in
the environment occurred from 1983 to 1985.
Dover sole, on the other hand, gradually migrate
into deeper, colder, and less oxygenated waters as
they age. Presumably the metabolism of fish
under these conditions would be lower as might
be the rate of lipofuscin accumulation in the
brain.
The literature on fishes provide no conclusive
evidence that lipofuscin is an accurate index of
age over the entire lifespan. Hill and Radtke (in
press) reported that the extracted lipofuscin per
unit dry weight in the brain of the tropical fish
Dascillus albisella accumulates exponentially
with age. The relationship is driven by a single
point for an 11-year-old individual and it would
probably be linear without that single point, if
only fish 1-7 years old were considered. In
hatchery-reared Cyprinus carpio the total ex-
tracted lipofuscin per unit of dry weight in the
brain of fish of the same age (6 years) increased
with weight over 8-fold range in weight (Griven
et al.^). About 45% of the lipofuscin content could
be explained by difference in weight among fish.
Thus, in fish of the same age there was a strong
size effect on lipofuscin concentration. These re-
sults are very similar to those found in this study.
Aloj Totaro et al. (1985) found that lipofuscin in-
creased over a range of 0-2 years. But the method
2Girven, R. J,, R. W. Gauldie, Z. Czochanska, A. D. Wool-
house. Manuscr. in prep. A critical tests of the lipwfuscin
technique of age estimation in fish.
was different as they measured lipofuscin gran-
ules present in the electric lobe of Torpedo mar-
morata brains rather than using extractable lipo-
fuscin of the entire brain. Thus, all studies to date
seem to indicate that accumulation occurs in the
brains of fishes, but results are not conclusive
owing to small sample sizes, limited age ranges,
and failure to identify the effects of brain growth
on rates of accumulation.
Additional research is required to evaluate
lipofuscin as a method of age determination in
fishes. The effect of brain growth on lipofuscin
accumulation rates must be considered in such
studies. A promising approach in this regard may
be to estimate lipofuscin on a per cell basis in-
stead of on a weight basis. This could be accom-
plished by either expressing extracted lipofuscin
relative to DNA concentration or by histological
techniques.
Acknowledgments
This work was funded by Sea Grant contract
RyNP-l-15C and by NOAA contract 43ABNF6
1987. The authors would like to thank A. Dizon,
M. M. Mullin, E. Brooks, and J. Butler for helpful
discussions; J. Butler, E. Lynn, and M. Draw-
bridge for age determination by otoliths of Dover
sole specimens; and M. Rowan of California Fish
and Game for providing the rainbow trout sam-
ples.
Literature Cited
Agius, C and S. a. Agbede.
1984. An electron microscopical study on the genesis of
lipofuscin, melanin, and haemosiderin in the haemopoi-
etic tissues offish. J. Fish Biol. 24:471-488.
Aloj Totaro. E., F. A Pisanti, P Russo, and P Brunetti.
1985. Evaluation of aging parameters in Torpedo mar-
morata. Ann. See. R. Zool. Belg. 115(2):203-209.
BuGH, E G . AND W J Dyer.
1959. A rapid method of total lipid extraction and purifi-
cation. Can. J. Biochem. Physiol. 37:911-917.
BOEHLERT, G. W.
1985. Using objective criteria and multiple regression
models for age determination in fishes. Fish. Bull, U.S.
83:103-107.
Dapson, R W , A T. Feldman. and G Pane.
1980. Differential rates of ageing in natural populations
of old-field mice (Peromyscus polionotus). J. Gerontol.
35:39-44.
Ettershank. G.
1984. A new approach to the assessment of longevity in
the Antarctic krill Euphausia superba. J. Crustacean
Biol. 4:295-305.
Ettershank, G , I MacDonnell, and R Croft.
1983. The accumulation of the age pigment by the fleshfiy
406
Sarcophaga bullatta Parker (Diptera: Sarcophagidae).
Aust. J. Zool. 31:131-138.
Fletcher. B. L . C J Dillard. and A L Tappel,
1973. Measurement of fluorescent lipid peroxidation
products in biological systems and tissues. Anal.
Biochem. 52:1-9.
Gulland, J. A.
1978. Analysis of data and development of models. In
J. A. Gulland (editor), Fish population dynamics, Ch.4.
Wiley.
Hill, K. T., and R L Radtke.
In press. Gerontological studies of the damselfish, Dascyl-
lus albisella. Bull. Mar. Sci.
MacArthur, M. C, and R. S Sohal.
1982. Relationship between metabolic rate, aging, lipid
peroxidation, and fluorescent age pigment in milkweed
bug, Oncopeltus fasciatus (Hemiptera). J. Gerontol.
37:268-274.
MlQUEL, J., P. LUNDGREN, AND J E. JOHNSON, JR.
1978. Spectrofluorometric and electron microscopic study
of lipofuscin accumulation in the testis of ageing mice.
J. Gerontol. 33:5-19.
MuLLiN, M. M., AND E Brooks.
1988. Extractable lipofuscin in larval marine fish. Fish.
Bull., U.S. 86:407-415.
Munnell, J. F., AND R. Getty.
1968. Rate of accumulation of cardiac lipofuscin in the
ageing canine. J. Gerontol. 23:154-158.
NiCOL, S.
1987. Some limitations on the use of the lipofuscin ageing
technique. Mar. Biol. (Berl.) 93:609-614.
Reichel, W.
1968. Lipofuscin pigment accumulation and distribution
in five rat organs as a function of age. J. Gerontol.
23:145-153.
Sheldahl, J A , AND A. L. Tappel.
1973. Fluorescent products from aging Drosophila
melanogaster: an indicator of free radical lipid peroxida-
tion damage. Exp. Gerontol. 9:33-41.
Smith. P J
1987. Homozygous excess in sand flounder, Rhombosolea
plebeia, produced by assortive mating. Mar. Biol.
(Berl.) 95:489-492.
Sohal, R. S., and H. Donato.
1978. Effects of experimentally altered life spans on the
accumulation of fluorescent age pigment in the housefly,
Musca domestica. Exp. Gerontol. 13:335-341.
Strehler. B L , D D Mark, A S Mildvan, and M V Gee.
1959. Rate and magnitude of age pigment accumulation
in the human myocardium. J. Gerontol. 14:430-
439.
Tappel, A L.
1975. Lipid peroxidation and fluorescent molecular dam-
age to membranes. In B. F. Trump and A. V. Arstila
(editors), Pathobiology of cell membranes. Vol. 1. Acad.
Press, N.Y.
Zar.J H
1974. Biostatistical analyses. Prentice-Hall, Englewood
Clifis, NJ, 620 p.
Maria Vernet
John R. Hunter
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 271
La Jolla, CA 92037
Russell D. Vetter
Marine Biology Research Division
Scripps Institution of Oceanography
University of California San Diego
La Jolla, CA 92093-0202
Institute of Marine Resources
Scripps Institution of Oceanography
University of California San Diego
La Jolla, CA 92093-0202
EXTRACTABLE LIPOFUSCIN IN LARVAL
MARINE FISH
The "age pigment", lipofuscin, is thought to be a
biochemically heterogeneous byproduct of the
peroxidation of polyunsaturated lipids which ac-
cumulates in dense, intracellular inclusions
called ceroid bodies. The material has been stud-
ied both microscopically and biochemically in tis-
sues of several species (Miquel et al. 1977; Shi-
masaki et al. 1980; Dowson 1982), and some
portion of it is quantitatively extractable with
organic solvents (Fletcher et al. 1973).
Flies prevented from flying by putting baffles
in the bottles in which they were raised, accumu-
lated lipofuscin (as assessed by solvent extrac-
tion) more slowly than did free-flying flies but
had a longer lifespan, so that at the ends of the
respective lifespans the body contents of lipofus-
cin were similar in the two groups (Sohal and
Donato 1978). Extractable lipofuscin thus ap-
pears to accumulate as a function of cumulative
oxidative metabolism; it could be an indicator of
physiological (rather than strictly chronological)
age.
Additionally, if lipofuscin represents an inte-
gral of oxidative metabolism since birth and
weight represents an integral of growth over the
same period, the ratio of lipofuscin to organic
weight should be proportional to the reciprocal of
cumulative net growth efficiency [K2 = growth/
assimilation = growth/( growth + respiration),
hence I/K2 = 1 + respiration/growth].
Ettershank (1984a) introduced the fluoromet-
ric measurement of extractable lipofuscin as a
measure of physiological age in growing marine
crustaceans, based on the work with insects, and
(1984b) recommended a simple method for rou-
tine use in marine work. He also argued (without
presenting extensive evidence) that preservation
FISHERY BULLETIN: VOL. 86, NO. 2, 1988.
407
of tissue in formalin-seawater did not invalidate
the analysis; thus, historical samples of zooplank-
ton (and, presumably, larval fish) appeared to be
usable for study of intraspecific geographic or in-
terannual variability.
Although most studies of lipofuscin (especially
by histology) have concerned postmitotic cells,
such as nervous tissue, we evaluated the assay as
an estimate of the relative efficiency of growth of
larvae analyzed whole. That is, we were less in-
terested in lipofuscin as an indicator of age (for
which, as we show, other measures of mass are
useful) than as an indicator of health whose rela-
tion to mass would reflect environmental condi-
tions over time. We therefore investigated the im-
portance of interfering fluorescing pigments
(Csallany and Ayaz 1976), tested the effect of
preservation in formalin (Nicol 1987), and mea-
sured the accumulation of extractable lipofuscin
in three species of larval fish reared in the labora-
tory— California grunion, Leuresthes tenuis;
white seabass, Atractoscion nobilis; and Califor-
nia halibut, Paralichthys californicus (hereafter
referred as grunion, seabass, and halibut respec-
tively).
Analytical Considerations
We analyzed lipofuscin by a method first de-
scribed by Fletcher et al. (1973), as modified by
Ettershank (1984b). The tissue to be analyzed
(usually whole larvae) was frozen (-15° or
-70°C) and later freeze-dried, and a 1-5 mg sam-
ple was homogenized in at least 2 mL of 2:1
chloroform: methanol in a Wheaton glass tissue
homogenizer. After extracting for 3-4 hours at
4°C, 100 mM MgCl2 (25% of the solvent volume)
was added, and the solutions were thoroughly
mixed and then centrifuged for 20 minutes at
3000 rpm at -4°C. The lower, chloroform layer
was withdrawn for fluorometric analysis after
reaching 20°C in a water bath, and the fluores-
cence was measured on a Turner 111^ fluorometer
using a CS 7-60 filter (approximately 360 nm) for
excitation and a CS 47B filter (approximately 430
nm) for emission. A known concentration of
quinine sulfate was the standard, and results are
therefore reported as "fluorescence units" or FU/
mg.
A stock solution of quinine sulfate (2 mg/L in 1
N H2SO4) was stored in a light-tight reagent bot-
tle. To compare analyses done at different times,
sets of standards were prepared from this stock in
distilled water (0.02, 0.04, 0.06, 0.08, and 0.10
|jLg/mL). Some of the sets prepared over a year's
time are shown in Figure 1; the overall reproduci-
bility is good.
60 1-
UJ
o
45
UJ
o
CO
UJ 30
tr
o
Z)
15
0
iReference to trade names does not imply endorsement by the
National Marine Fisheries Service, NCAA
0 .03 .06 .09 .12
QUININE SULPHATE (/ig-mr')
Figure 1. — Reproducibility of different sets of quinine sulfate
standards prepared from a stock solution over a year.
Precision of the method, including extraction
from tissue, was estimated by comparing repli-
cate samples of liver and white muscle from adult
halibut. The coefficients of variation for 5 deter-
minations on each of 3 sets of tissue were for mus-
cle, 0.2, 0.35, and 0.57, and for liver, 0.1, 0.15, and
0.18. The difference in variability between the
two kinds of tissues is probably due to the greater
difficulty in homogenizing muscle tissue.
Quinine sulfate standards were compared on a
Farrand Spectrofluorometer, with excitation at
350 nm and emission at 420 nm, and on the
Turner 111 fluorometer. The correlation coeffi-
cient for measurements on the two instruments
was 0.97. We therefore used the Turner 111 rou-
tinely, so that analyses could be done easily in a
hood.
Csallany and Ayaz (1976) described interfer-
ence by retinol in the analysis of organic-solvent-
soluble lipofuscin in mammalian tissues, and rec-
ommended a chromatographic step to remove this
contaminant. We extracted a variety of fish tis-
sues in 2:1 chloroform: methanol, and after addi-
tion of MgCl2 and centrifugation, dried the chlo-
408
roform layer gently (35°C) in an open vial
overnight in a hood on a sand bath. The dried
extract was reconstituted with 1:9 chloro-
form: methanol and chromatographed on a 30 cm
Sephadex LH20 column at a flow rate of 3-4 mL/
hour. A fraction was collected by hand every 20
minutes, its volume measured, and its fluores-
cence determined on the Turner 111 at the same
wavelengths used for routine analysis of lipofus-
cin. To determine where a retinol peak would ap-
pear, we chromatographed a tissue extract to
which a commercial preparation (Sigma) of
retinol had been added. Also, extracts of larval
fish in the pre-eyed stage were compared with
extracts of larvae that had pigmented eyes, as-
suming that the eyed larvae would have more
retinol.
A UV irradiation step had been proposed to
degrade retinol where it may interfere with mea-
surement of extracted lipofuscin, but Csallany
and Ayaz (1976) reported that this procedure was
ineffective. We therefore determined the time
course of degradation of commercial retinol in 1:9
chloroform: methanol by UV irradiation in quartz
tubes.
At the wavelengths we used, retinol was not an
important interfering substance in a variety of
larval fish and in the adult grunion muscle and
adult halibut liver tissues (Fig. 2). There also was
little difference between extracts of eyed and pre-
FlGURE 2. — Chromatograms (fluorescence vs. fraction number)
for various fish, compared with fish tissue plus retinol.
'50- HALIBUT LIVER
100
LARVAL GRUNION (whole)
LARVAL SEABASS (whole)
20 5
FRACTION (ml)
409
eyed larvae. We therefore did not use UV irradia-
tion routinely. If UV irradiation is used to de-
grade retinol, the exposure must be kept short (1
or 2 minutes), or else another compound fluoresc-
ing at these wavelengths appears (Fig. 3), giving
the spurious impression that UV did not affect
retinol; this may explain the negative results of
Csallany and Ayaz (1976).
To evaluate the effect of preservation in forma-
lin, we froze subsamples from stocks of larvae,
and preserved other subsamples in 10% formalin
in a glass container, using formalin from a glass
reagent bottle. We analyzed frozen and preserved
80,-
4 6
MINUTES
Figure 3. — Time course of fluorescence resulting from UV
irradiation of duplicate batches (two symbols) of retinol.
larvae after 3 months, and after an additional
year we analyzed more preserved larvae.
Nicol (1987) recently demonstrated that forma-
lin preservation significantly increases the fluo-
rescence of lipofuscin extracted in chloroform-
methanol, and our results show similar analytical
problems (Table 1). There was significant en-
hancement in fluorescence in the larvae pre-
served for over a year, when compared with those
preserved for 3 months. Quinine sulfate stan-
dards run at the same time as the samples for
different periods were very similar (Fig. 4), indi-
cating real changes with time of preservation.
Thus, this method will not permit using historical
collections of formalin-preserved animals to de-
termine their relative physiological states until
more is known about the time course and nature
of the effect of formalin on the extract of pre-
served tissue.
Larval Growth Experiments
Seabass larvae were obtained from Hubbs Re-
search Institute and halibut larvae from the Los
Angeles County Natural History Museum; larval
grunion were obtained by stripping adults and
bringing fertilized eggs into the laboratory for
hatching in an aerated, 10 L container of sea-
water at room temperature.
Larval grunion were reared in 40 L Nalgene
tubs with spigots at the bottom (which facilitated
emptying and cleaning), initially stocked with
100—120 newly hatched grunion per tub, with
replicated "high" and "low" food concentrations.
Larvae were fed newly hatched Artemia nauplii
and the rotifer, Brachionus , which was cultured
on Dunaliella tertiolecta in 100 L, lighted
Table 1 . — Lipofuscin (as fluorescence units, FU) per animal and per unit dry weight (DW) of frozen vs. formalin-
preserved larval fish. N = number of larvae per analysis.
Frozen (3 months)
Preserved (3 months)
Preserved (1 year)
FU (larva) -1
FU
(mg DW)-i
FU (larva) -1
FU (mg DW)-1
FU (lar^a)i
FU
(mg DW)-'
Grunion
16.0
47.4
24.3
60.1
44.3
98.1
A/ = 15
16.7
47.3
31.0
76.7
45.0
93.1
17.3
47.7
23.6
60.4
48.0
100.1
17.0
49.7
24.0
26.0
63.0
65.3
54.3
51.0
112.7
106.2
X =
16.8
48.0
25.7
65.3
48.5
102.0
Seabass
0.68
18.6
0.82
13.6
1.72
22.3
A/ = 30
0.47
13.0
0.71
12.0
2.05
26.3
0.50
14.5
0.46
8.0
2.03
27.2
0.43
12.7
0.52
8.7
1.63
22.2
0.33
9.4
0.60
10.5
2.05
28.9
X =
0.48
13.6
0.62
10.6
1.89
25.4
410
80
60
UJ
o
UJ
o
CO
UJ
en
o
3
40
20
OL
1
0 .02 .04 .06 .08 .10
QUININE SULPHATE {/xg -mr')
Figure 4. — Fluorescence of various concentrations of
quinine sulfate standards read at 10 x gain (lower pair
of lines) and 30x gain (upper pair) in 1985 (lower line of
each pair) and 1986 (upper of each pair).
polyethylene tubs augmented with f/2 phyto-
plankton nutrients (May 1971; Theilacker and
McMaster 1971). One spring and one summer ex-
periment were completed at ambient tempera-
tures (spring = 17.5°-19°C, summer = 21°-23°C),
with one of the "high food" containers in the sec-
ond experiment kept at 26°C with an aquarium
immersion heater.
We were unable to maintain absolute high and
low food concentrations because of oscillations in
the supply of food organisms. Thus, the high con-
centration was kept at 3 times the low concentra-
tion, although absolute amounts varied. The
mean initial food concentrations in the spring ex-
periment were 50 jxg C/L (= "low") and 154 (xg
C/L (= "high"); in summer, 122 and 394 ^JLg C/L,
respectively. Tubs were censused every 2 days
through each experiment to determine how much
food was uneaten and how many larval fish had
died (estimated by counting and removing
corpses), and to add fresh food. In the low food
containers, it was not unusual to find little un-
eaten food, particularly as the larvae grew. Once
a week the tubs were emptied and the remaining
larvae counted directly. It was clear from this
direct census that all dead larvae were not ac-
counted for by searching for corpses every 2 days
because of cannibalism, necrophagy, or decay.
Every 4 days a known number of larvae was re-
moved; their lengths were measured and they
were frozen for future analyses.
We could not estimate larval ingestion pre-
cisely because of the uncertainty in how many
fish were alive through a 2-d interval. This prob-
lem was exacerbated as the larvae within each
tub diverged in size, so that variance in individ-
ual ingestion increased. Although 3 times more
food was offered in the high food containers, these
larvae actually ingested about twice the amount
of food as did those in the low food containers
(Table 2). This difference was due to better sur-
vival in the high food containers, which affected
the ratio between available ration and number of
larvae. To compensate for this, we routinely har-
vested more animals from the high food contain-
ers than from the low food containers.
Freeze-dried animals or tissues were weighed
on a Cahn electrobalance. Protein was deter-
mined by a method of Dorsey et al. (1977) on an
aliquot of tissue homogenized in cold 1 M NaCl.
DNA was measured by an ethidium bromide tech-
nique (Bentle et al. 1981, as modified by M. S.
Lowrey). Basic measures of size — dry weight,
protein, and DNA — were strongly and linearly
correlated (Fig. 5), so that comparing lipofuscin
with any of these measures would give similar
patterns.
Lipofuscin accumulated as the larval fish grew
(Fig. 6), but at quite different rates for the 3 spe-
cies, grunion accumulating most rapidly (relative
to gain in weight) and seabass much the slowest.
Table 2. — Estimated average ingestion, size, composition, and
growth efficiency for 20-day-old, laboratory-reared larval California
grunion. Compare with Figure 5.
Sphng
Summer
experiment
experiment
Low
High
Low
High
food
food
food
food
M-g C ingested'
per individual
701
1,723
1,379
2,792
p.g dry weight per
individual
755
1,305
901
1,895
^.g DNA per mg
protein
84.8
59.4
48.1
31.3
\xg C2 per
individual
245
603
482
977
Growth from
hatching (|j.g C)
104
462
341
836
Gross growth
efficiency
15%
27%
25%
30%
'Calculated from measured carbon In Brachionus and Anemia.
^Estimated from literature values.
411
o
>
cr>
Ld
2-
o
cr
CL
0
100
1
80
o
>
k.
n
60
CP
:L
40
<L
Z
n
ZO
LARVAL
GRUNION
•
-
. /Y=0.43X-0.06
1 1 1
LARVAL SEABASS
0
4 6 8 10 0 2 4
DRY WEIGHT (mg- larva"')
Figure 5. — Relations of protein (upper) and DNA (lower) to dry weight for laboratory-reared larval grunion (left)
and white seabass (right).
This may reflect differences in the Hfespans and
metabolic rates; grunion are small fish, reaching
maturity in a year and living less than 5 years
(Frey 1971), while seabass and halibut grow to
very much larger size and can live more than 20
years (Frey 1971; Thomas 1968).
Figure 7 shows the variability in the amount of
lipofuscin in larvae of the same age. At time "0",
all fish were newly hatched. The variance was
low in the first few days, but increased dramati-
cally with time because larvae within each tub
grew (and, presumably, respired) at very different
rates. There were differences in averages between
high food and low food conditions, and between
experiments (Table 2), but the variance in dry
weight, protein, DNA, or lipofuscin was so large
that the overlap obscured any differences be-
tween conditions of rearing.
Grunion and halibut larvae start life with
greater concentrations of lipofuscin than do sea-
bass, and though the concentration decreases
rapidly as grunion and halibut age (Fig. 8), they
still have almost a 10-fold greater concentration
than do larval seabass when all are 20 days old.
All three species increased in weight faster than
they increased in lipofuscin, so the concentration
of lipofuscin was "diluted" by growth. Because
protein was a constant fraction of dry weight
(Fig. 5, upper), this dilution was not due to skele-
tal growth alone. If the rate of weight-specific
growth exceeds the rate of weight-specific respi-
ration during early life, the concentration of lipo-
fuscin should decrease, as observed, and only
when growth ceases or slows considerably should
lipofuscin accumulate relative to weight. Alter-
natively, lipofuscin could change chemically with
time, becoming more difficult to extract (Vernet
et al. 1988), so that the rate of accumulation of
lipofuscin would be underestimated.
We attempted to stop growth by starving the
412
200
160
r^ 120 -
GRUNION
10
8
4 -
0
HALIBUT
Y=I4.5X +0.8
0
.15
.30
.45
.60
.75
6 8 10
DRY WEIGHT ( mg • larva"')
Figure 6. — Relation of lipofuscin fluorescence per larva to dry weight for larval grunion (upper left) white seabass (lower left), and
California halibut (right). Note different axes.
larvae to see if we could detect an increase in
lipofuscin. Table 3 shows results for seabass
starved for the final 20% of the rearing period;
there was no significant change in the concentra-
tion of lipofuscin. This is not what one would ex-
pect if accumulation of lipofuscin is proportional
to physiological age, unless no lipid is metabo-
lized during starvation. However, this result also
could reflect slow transformation of lipofuscin
from a more to a less soluble pool (Vernet et al.
1988).
Table 3. — Lipofuscin fluorescence (FU) per
unit dry weigfif of larval white seabass starved
for various periods after age 29 days.
N = number of analyses; ranges in parenthe-
ses.
FU (mg dry weight) -1 N
Initial
29 days old
Starved
2.4
(1.6-3.0)
4
2 days
3.1
(2.3-4.0)
2
4 days
1.3
(0.54-1.7)
6
6 days
1.8
(1.4-2.3)
2
8 days
2.5
1
Conclusions
Our intent was to evaluate the utility of mea-
suring extracted lipofuscin fluorometrically as an
indicator of the integrated metabolic health of
fish, especially preserved ones, and of relative net
efficiency of growth. We conclude that this tech-
nique is unlikely to be useful in these ways, at
least within the larval period. Although the accu-
mulation of total body burden of lipofuscin was
demonstrated, the variability among individuals
grown under the same conditions became so large
over time that we were unable to calibrate the
method in an ecologically meaningful sense. The
variability was evident in all measures of growth,
413
lUU
"GRUNION
60
•
120
-
80
•
•
•
• •
40
• •
•
• -1 • 1
•
•
•
•
•
1
•
• •
!:•
! 1 1
o
>
LU
O
LU
CJ
LD
Ld
a:
o
20
15 -
10
5 -
0
SEABASS
(5).
'OfHALIBUT
8 -
6 -
2 -
0
I
1
0 10 20 30
DAYS AFTER HATCHING
Figure 7. — Lipofuscin fluorescence per larva vs. age since
hatching for laboratory-raised larval grunion (top), white sea-
bass (center), and California halibut (bottom).
200
160
120
80
40
0
100
75
50
25
0
GRUNION
Y=-0.99X + 55
SEABASS
. Y=-0.48X + 13
HALIBUT
Y = -I.7X + 67
0 10 20 30
DAYS AFTER HATCHING
Figure 8. — Lipofuscin fluorescence per unit dry weight vs. age
since hatching for laboratory-reared larval grunion (top), white
seabsiss (center), and California halibut (bottom). Linear fits Eire
for comparative purposes only; relations for seabass and halibut
look curvilinear.
414
and is common in culture of larval fishes. Such
variability may be reduced in nature, where
runts may be subject to intense predation.
Lipofuscin analysis may be usefijl only when
applied to postmitotic tissue, such as nervous tis-
sue in mature fish, or to whole organisms whose
mitotic growth has essentially ceased, such as
adult copepods or insects (since the methods ap-
pear usefiil in arthropods — Ettershank et al.
1983; Sohal and Donato 1978). In these organ-
isms, the vagaries of growth are reduced, and the
accumulation of lipofuscin during starvation or
exercise might show that lipofuscin concentration
is interpretable as a measure of physiological age,
habitat quality, and net growth efficiency.
Lipofuscin is known to be a polytypic sub-
stance, probably variable in composition among
different organisms. We have assumed that a con-
stant proportion of the same substance is ex-
tracted. This may not be true (Vernet et al. 1988),
and considerable work remains to be done on the
basic biochemistry of the component substance(s).
Though the extraction and fluorometric measure-
ment is tantalizingly simple, it may well be that
the microscopical method used to quantify "ceroid
bodies" is the best approach. Fluorescent tech-
niques used in histochemical research (Brizzee
and Jirge 1981), combined with automatic imag-
ing procedures, might decrease the tedium of
staining and visual microscopy.
Acknowledgments
We thank Refik Orhun and Steve Cadell for the
larval white seabass and halibut, and Maria Ver-
net and two referees for helpful comments. Re-
search was supported by California Department
of Fish and Game project C-921 and National Sci-
ence Foundation grant OCE86-00742.
Literature Cited
Bentle, L a.. S Dutta, and J Metcoff.
1981. The sequential enzymatic determination of DNA
and RNA. Anal. Biochem. 116:5-16.
Brizzee, K R , and S K Jirge.
1981. Fluorescent microscope techniques for the visual-
ization and histological quantification of autofluorescent
lipofuscin bodies in brain tissues. In J. E. Johnson (edi-
tor). Current trends in morphological techniques, Vol.
III. CRC Press.
CSALLANY, A. S , AND K L AYAZ
1976. Quantitative determination of organic solvent solu-
ble lipofuscin pigments in tissue. Lipids 11:412-417.
DoRSEY, T E , P W McDonald, and O A Roels
1977. A heated Biuret-Folin protein assay which gives
equal absorbance with different proteins. Anal. Bio-
chem. 78:156-164.
DOWSON. J H
1982. The evaluation of autofluorescence emission spec-
tra derived from neuronal lipopigment. J. Microsc.
128:261-262.
Ettershank, G.
1984a. A new approach to the assessment of longevity in
the Antarctic krill Euphausia superba. J. Crust. Biol.
4(Spec. No. l):295-305.
1984b. Methodology for age determination of Antarctic
krill using the age pigment lipofuscin. Biomass Handb.
No. 26. SCAR/SCOR/LABO/ACMRR.
Ettershank, G., I MacDonnell, and R Croft.
1983. The accumulation of age pigment by the fleshfly
Sarcophaga bullata Parker (Diptera: Sarcophagiidae).
Aust. J. Zool. 31:131-138.
Fletcher, B L . C J Dillard. and A L Tappel.
1973. Measurement of fluorescent lipid peroxidation
products in biological systems and tissues. Anal. Bio-
chem. 52:1-9.
Frey. H W
1971. California's living marine resources and their uti-
lization. Calif. Dep. Fish Game, Resour. Agency, 148 p.
May, R C
1971. Effects of delayed initial feeding on larvae of the
grunion Leuresthes tenuis (Ayres). Fish. Bull., U.S.
69:411-425.
MiQUEL, J , J Org, K G. Bensch, and J E Johnson, Jr.
1977. Lipofuscin: fine-structural and biochemical studies.
In W. A. Pryor (editor). Free radicals in biology, p. 133—
182. Acad. Press, N.Y.
NICOL, S
1987. Some limitations on the use of the lipofuscin ageing
technique. Mar. Biol. (Berl.) 93:609-614.
Shimasaki, H., N Veta, and O. S. Privett
1980. Isolation and analysis of age-related fluorescent
substances in rat testes. Lipids 15:236-241.
Sohal, R. S , and H Donato.
1978. Effects of experimentally altered life spans on the
accumulation of fluorescent age pigment in the housefly,
Musca domestica. Exp. Gerontol. 13:335-341.
Theilacker, G H , AND M F. McMaster.
1971. Mass culture of the rotifer Brachionus plicatilis and
its evaluation as food for larval anchovies. Mar. Biol.
(Berl.) 10:183-188.
Thomas, J C
1968. Management of the white seabass (Cynoscion no-
bilis ) in California waters. Calif Dep. Fish Game, Fish.
Bull. 142, p. 1-34.
Vernet, M , J R Hunter, and R. D. Vetter
1988. Accumulation of age pigments in two cold-water
fishes. Fish. Bull., U.S. 86:401-407.
M M. Mullin
E. R. Brooks
Institute of Marine Resources
Scripps Institution of Oceanography
University of California, San Diego
La Jolla, CA 92093-0218
415
NOTICES
NOAA Technical Reports NMFS published during last 6 months of 1987
55. Proximate composition, energy, fatty acid, sodium,
and cholesterol content of fmfish, shellfish, and their
products. By Judith Krzynowek and Jenny Mur-
phy. July 1987, iii + 53 p., 2 tables.
56. Some aspects of the ecology of the leatherback turtle,
Dermochelys coriacea, at Laguna Jalova, Costa
Rica. By Harold F. Hirth and Larry H.
Ogren. July 1987, iii + 14 p., 12 tables, 13 figures.
57. Food habits and dietary variability of pelagic nekton
off Oregon and Washington, 1979-1984. By
Richard D. Brodeur, Harriet V. Lorz, and William G.
Pearcy. July 1987, iii + 32 p., 32 tables, 1 figure.
58. Stock assessment of the gulf menhaden, Brevoortia
patronus, fishery. By Douglas S. Vaughan.
September 1987, iii + 18 p., 19 figures, 14 tables.
59. Atlantic menhaden, Brevoortia tyrannus , purse seine
fishery, 1972-84, with a brief discussion of age and
size composition of the landings. By Joseph W.
Smith, William R. Nicholson, Douglas S. Vaughan,
Donnie L. Dudley, and Ethel A. Hall. September
1987, iii + 23 p., 3 figures, 12 tables, 14 appendix
tables.
60. Gulf menhaden, Brevoortia patronus, purse seine
fishery, 1974-85, with a brief discussion of age and
size composition of the landings. By Joseph W.
Smith, Eldon J. Levi, Douglas S. Vaughan, and Ethel
A. Hall. December 1987, iii + 8 p., 1 figure, 8 ta-
bles, 2 appendix tables.
61. Manual for starch gel electrophoresis: a method for
the detection of genetic variation. By Paul B.
Aebersold, Gary A. Winans, David J. Teel, George B.
Milner, and Fred M. Utter. December 1987, iii + 19
p., 8 figures, 1 table, appendices.
62. Fishery publication index, 1980-85; Technical Memo-
randum index, 1972-85. By Cynthia S. Martin,
Shelley E. Arenas, Jacki A. Guffey, and Joni M.
Packard. December 1987, iii + 149 p.
Some NOAA publications are available by purchaise from the Superinten-
dent of Documents, U.S. Government Printing Office, Washington, DC
20402.
416
ERRATA
Fishery Bulletin: Vol. 86, NO. 1
Butler, John L., and Darlene Pickett, "Age-specific vulnerability of Pacific sardine,
Sardinops sagax, larvae to predation by northern anchovy, Engraulis mor-
dax," p. 163-167.
Pages 165 and 166 have incorrect figures for the legends, please replace with the
following:
o
z
o
z
o
a.
Vi
HI
a
t-
z
LLI
o
cc
LJ
Q.
uu
—
J
80
-
1
..- X
60
-
ANC
1
HOVY ..'
- V.--
'■''
.^
—
J
' J
40
-
1
t ,
- ■
uf'
-
-
-l
""
'/
r^
^^SARDINE
20
~ <
■' «
--^
K^
n
1 1
1
1 1 1 1 1 1 1 1
8 12 16
LENGTH (mm)
20
24
Figure 1. — Increase by size of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, responding to attack by adult northern an-
chovy and 95% confidence intervals. Data on anchovy larvae
from Folkvord and Hunter (1986).
100
O 80
z
a
O 60
z
UJ
o
40
20
ANCHOVY
v.-
SARDINE
8 12 16
LENGTH (mm)
24
Figure 3. — Increase by size of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, escaping attack by adult northern anchovy and
95% confidence intervals. Data on anchovy larvae from
Folkvord and Hunter (1986).
100
O
z
a
z
o
a.
(A
HI
o
lij
Q.
80
60
40
20
.1
ANCHOVY If
SARDINE
10
20
AGE (days)
30
40
Figure 2. — Increase by age of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax, responding to attack by adult northern an-
chovy and 95% confidence intervals. Size categories of reared
larvae have been converted to ages using growth rates esti-
mated from the field. Data on anchovy larvae from Folkvord
and Hunter (1986).
100
z
UJ
o
80 -
z
Q.
<
O 60
V)
40 -
20
-
T
-
ANCHOVY , -'-
-
1
V
-
..'
F
.^
r' J
L
SARDINE
_
•••"f*Y -T-
.^
, — t
<^
, t-^
,1 F= 1 f^ 1 ] 1 1
10
20
AGE (days)
30
40
Figure 4. — Increase by age of the percentage of Pacific sardine
larvae, Sardinops sagax, and northern anchovy larvae, En-
graulis mordax , escaping attack by adult northern anchovy and
95% confidence intervals. Size categories of reared larvae have
been converted to ages using growth rates estimated from the
field. Data on anchovy larvae from Folkvord and Hunter (1986).
303
j 2 .q
INFORMATION FOR CONTRIBUTORS TO THE FISHERY BULLETIN
Manuscripts submitted to the Fishery Bulletin will reach print faster if they conform to the following instructions.
These are not absolute requirements, of course, but desiderata.
CONTENT OF MANUSCRIPT
The title page should give only the title of the paper,
the author's name, his/her affiliation, and mailing ad-
dress, including ZIP code.
The abstract should not exceed one double-spaced
page.
In the text, Fishery Bulletin style, for the most part,
follows that of the U.S. Government Printing Office
Style Manual. Fish names follow the style of the Amer-
ican Fisheries Society Special Publication No. 12, A List
of Common and Scientific Names of Fishes from the
United States and Canada, Fourth Edition, 1980.
Text footnotes should be typed separately from the
text.
the literature cited. The accuracy of the literature
cited is the responsibility of the author. Abbreviations
of names of periodicals and serials should conform to
Serial Sources for the BIOSIS Data Base. (Chemical Ab-
stracts also uses this system, which was developed by
the American Standards Association.)
Common abbreviations and symbols, such as mm,
m, g, L, mL, mg, °C (for Celsius), %, %c, and so forth,
should be used. Abbreviate units of measure only when
used with numerals. Periods are only rarely used with
abbreviations.
We prefer that measurements be given in metric
units; other equivalent units may be given in parenthe-
ses.
FORM OF THE MANUSCRIPT
Figures and tables, with their legends and headings,
should be self-explanatory, not requiring reference to
the text. Their placement should be indicated in the
right-hand margin of the manuscript.
Preferably figures should be reduced by photography
to 53/4 inches (for single-column figures, allowing for
50% reduction in printing), or to 12 inches (for double-
column figures). The maximum height, for either width,
is 14 inches. Photographs should be printed on high
quality or glossy paper.
Do not send original drawings to the Scientific Editor;
if they, rather than the photographic reductions, are
needed by the printer, the Scientific Publications Office
will request them.
Each table should start on a separate page. Consis-
tency in headings and format is desirable. Vertical rules
should be avoided, as they make the tables more expen-
sive to print. Footnotes in tables should be numbered
sequentially in arable numerals. To avoid confusion
with powers, they should be placed to the left of numer-
als.
Acknowledgments, if included, are placed at the
end of the text.
Literature is cited in the text as: Lynn and Reid
(1968) or (Lynn and Reid 1968). All papers referred to in
the text should be listed alphabetically by the senior
author's surname under the heading "Literature Cited."
Only the author's surname and initials are required in
The original of the manuscript should be typed,
double-spaced, on white bond paper. Please triple space
above headings. We would rather receive good dupli-
cated copies of manuscripts than carbon copies. The se-
quence of the material should be:
TITLE PAGE
ABSTRACT
TEXT
LITERATURE CITED
TEXT FOOTNOTES
APPENDIX
TABLES (Each table should be numbered with an
arable numeral and heading provided).
LIST OF FIGURES (Entire figure legends)
FIGURES (Each figure should be numbered with
an arable numeral; legends are desired)
ADDITIONAL INFORMATION
Send the ribbon copy and two duplicated or carbon
copies of the manuscript to:
Dr. Andrew E. Dizon, Scientific Editor
Fishery Bulletin
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 271
La Jolla, CA 92038
Fifty separates will be supplied to an author free of
charge and 50 supplied to his organization. No covers
will be supplied.
Contents — Continued
WATSON, CHERYL, ROBERT E. BOURKE, and RICHARD W. BRILL. A compre-
hensive theory on the etiology of burnt tuna 367
BROWN-PETERSON, NANCY, PETER THOMAS, and CONNIE R. ARNOLD.
Reproductive biology of the spotted seatrout, Cynoscion nebulosus, in South
Texas 373
Notes
CHEN, CHE-TSUNG, TZYH-CHANG LEU, and SHOOU-JENG JOUNG. Notes
on reproduction in the scalloped hammerhead, Sphyrna lewini, in northeastern
Taiwan waters 389
COLLINS, MARK R., and CHARLES A. WENNER. Occurrence of young-of-the-
year king Scomberomorus cavalla, and Spanish, S. maculatus, materials in
commercial-type shrimp trawls along the Atlantic coast of the southeast United
States 394
DEW, C. BRAXTON. Stomach contents of commercially caught Hudson River
striped bass, Morone saxatilis , 1973-1975 397
VERNET, MARIA, JOHN R. HUNTER, and RUSSELL D. VETTER. Accumula-
tion of age pigments (lipofuscin) in two cold-water fishes 401
MULLIN, M. M., and E. R. BROOKS. Extractable lipofuscin in larval marine
fish 407
Notices: NOAA Technical Reports published during the last 6 months of 1987 . . . 416
GPO 791-008
.<*<■': ">».
BuUetin
^^AT£S O^
Marine Biological Laboratory
LIBRARY
JAN 2 0 1989
Vol. 86, No. 3
Woods Hole, Mass.
-^
July 1988
BARLOW, JAY. Harbor porpoise, Phocoena phocoena, abundance estimation for Califor-
nia, Oregon, and Washington: I. Ship surveys 417^
BARLOW, JAY, CHARLES W. OLIVER, TERRY D. JACKSON, and BARBARA L.
TAYLOR. Harbor porpoise, Phocoena phocoena, abundance estimation for California,
Oregon, and Washington: IL Aerial surveys 433-'
SIGLER, MICHAEL R, and JEFFREY T FUJIOKA. Evaluation of variability in sable-
fish, Anoplopoina fimbria, abundance indices in the Gulf of Alaska using the bootstrap
method 445
BOEHLERT, GEORGE W, and TAKASHI SASAKI. Pelagic biogeography of the armor-
head, Pseudopentaceros wheeleri, and recruitment to isolated seamounts in the North
Pacific Ocean 453
SAVOY, THOMAS F, and VICTOR A. CRECCO. The timing and significance of density-
dependent and density-independent mortality of American shad, Alosa sapidissima. . 467
HOSS, DONALD E., LINDA COSTON-CLEMENTS, DAVID S. PETERS, and PATRICK.
A TESTER. Metabolic responses of spot, LeiosUymus xanthurus, and Atlantic croaker,
Micropogonias undulatiis, larvae to cold temperatures encountered following recruitment
to estuaries 483
WATANABE, YOSHIRO, JOHN L. BUTLER, and TSUKASA MORI. Growth of Pacific
saury, CololaMs saira, in the northeastern Pacific Ocean 489
DAVIES, N. M., R. W GAULDIE, S. A. CRANE, and R. K THOMPSON. Otolith
ultrastructure of smooth oreo, Pseudocyttits maculatus, and black oreo, Allocyttus sp.,
species 499
BECKER, D. SCOTT. Relationship between sediment character and sex segregation in
English sole, Parophrys vetulus 517
JAMIESON, GLEN S., and ANTAN C. PHILLIPS. Occurrence of Cancer crab (C.
magister and C. oregonensis) megalopae off the west coast of Vancouver Island, British
Columbia 525
STONER, ALLAN W, and ROGER J. ZIMMERMAN. Food pathways associated with
penaeid shrimps in a mangrove-fringed estuary 543^
WILSON, ELIZABETH A., ERIC N. POWELL, and SAMMY M. RAY The effect of the
ectoparasitic pyramidellid snail, Boonea impressa, on the growth and health of oysters,
Crassostrea virginica, under field conditions 553
BROUSSE AU, DIANE J., and JENNY A BAGLIVO. Life tables for two field populations
of soft-shell clam, Mya arenaria, (Mollusca: Pelecypoda) from Long Island Sound . . . 567
(Contimied on back cover)
Seattle, Washington
U.S. DEPARTMENT OF COMMERCE
C. William Verity, Secretary
NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION
William E. Evans, Under Secretary for Oceans and Atmosphere
NATIONAL MARINE FISHERIES SERVICE
James W. Brennan, Assistant Administrator for Fisheries
Fishery Bulletin
The Fishery Bulletin carries original research reports and technical notes on investigations 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 Na 1103. Beginning 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 instead of being issued
individually. 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 ntunbers to libraries, research institutions, State and Federal agencies, and in exchange for other scientific
publications.
SCIENTIFIC EDITOR, Fishery Bulletin
Dr. Andrew E. Dizon
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 271
La Jolla, CA 92038
Editorial Committee
Dr. Jay Barlow
National Marine Fisheries Service
Dr. William H. Bayliff
Inter-American Tropical Tftma Commission
Dr. George W. Boehlert
National Marine Fisheries Service
Dr. Bruce B. CoUette
National Marine Fisheries Service
Dr. Robert C. Francis
University of Washington
Dr. James R. Kitchell
University of Wisconsin
Dr. William J. Richards
National Marine Fisheries Service
Dr. Tim D. Smith
National Marine Fisheries Service
Mary S. Fukuyama, Managing Editor
The Fishery BvMetin (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. Second class postage is paid in Seattle; Wash., and additional offices.
POSTMASTER send address changes for subscriptions to Fishery Bulletin, Superintendent of Documents, U.S. Government Printing
Officci Washington, DC 20402.
Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated.
The Secretary of Commerce has determined that the publication of this periodical 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 Manage-
ment and Budget
Fishery Bulletin
CONTENTS
Vol. 86, No. 3 July 1988
BARLOW, JAY. Harbor porpoise, Phocoena phocoena, abundance estimation for Califor-
nia, Oregon, and Washington: I. Ship surveys 417
BARLOW, JAY, CHARLES W OLIVER, TERRY D. JACKSON, and BARBARA L.
TAYLOR. Harbor porpoise, Phocoena phocoena, abundance estimation for California,
Oregon, and Washington: IL Aerial surveys 433
SIGLER, MICHAEL P., and JEFFREY T FUJIOKA. Evaluation of variability in sable-
fish, Ancyplopoma fimbria, abundance indices in the Gulf of Alaska using the bootstrap
method 445
BOEHLERT, GEORGE W, and TAKASHI SASAKI. Pelagic biogeography of the armor-
head, Pseudopentaceros wheeleri, and recruitment to isolated seamounts in the North
Pacific Ocean 453
SAVOY, THOMAS F, and VICTOR A. CRECCO. The timing and significance of density-
dependent and density-independent mortality of American shad, Alosa sapidissima. . 467
HOSS, DONALD E., LINDA COSTON-CLEMENTS, DAVID S. PETERS, and PATRICIA
A. TESTER. Metabolic responses of spot, Leiostomus xanthurus, and Atlantic croaker,
Micr&pogonias undidatvs, larvae to cold temperatures encountered following recruitment
to estuaries 483
WATANABE, YOSHIRO, JOHN L. BUTLER, and TSUKASA MORI. Growth of Pacific
saury, Cololabis saira, in the northeastern Pacific Ocean 489
DAVIES, N. M., R. W GAULDIE, S. A. CRANE, and R. K. THOMPSON. Otolith
ultrastructure of smooth oreo, Pseudocyttits maculatus, and black oreo, Allocyttus sp.,
species 499
BECKER, D. SCOTT. Relationship between sediment character and sex segregation in
English sole, Parophrys vetulus 517
JAMIE SON, GLEN S., and ANTAN C. PHILLIPS. Occurrence of Cancer crab (C.
magister and C. oreganensis) megalopae off the west coast of Vancouver Island, British
Columbia 525
STONER, ALLAN W, and ROGER J. ZIMMERMAN. Food pathways associated with
penaeid shrimps in a mangrove-fringed estuary 543
WILSON, ELIZABETH A., ERIC N. POWELL, and SAMMY M. RAY. The effect of the
ectoparasitic pyramidellid snail, Boonea impressa, on the growth and health of oysters,
Crassostrea virginica, under field conditions 553
BROUSSE AU, DIANE J., and JENNY A. BAGLIVO Life tables for two field populations
of soft-shell clam, Mya arenaria, (MoUusca: Pelecypoda) from Long Island Sound . . . 567
{Contimied on next page)
Seattle, Washington
1988
For sale by the Superintendent of Documents, U.S. Government Printing Offi
DC 20402— Subscription price per year; $16.00 domestic and $20.00 foreign
issue; $9.00 domestic and $11.25 foreign.
Marine Biological Laboratory
LIBRARY
e, Washingti
'ost per si
t^m 2 0 1989
li
Woods Hole, Mass. j
Contents— ConiiriMed
FARLEY, C. AUSTIN, PETER H. WOLF, and RALPH A. ELSTON. A long-term study
of "microcell" disease in oysters witii a description of a new genus, Mikrocytos (g. n.),
and two new species, Mikrocytos mackini (sp. n.) and Mikroq^tos roughleyi (sp. n.) . 581
Notes
WILLIAMS, AUSTIN B. Cojoined twin adult shrimp (Decapoda: Penaeidae) 595
GOULD, EDITH, DIANE RUSANOWSKY, and DONNA A. LUEDKE. Note on
muscle glycogen as an indicator of spawning potential in the sea scallop, Placopecten
magellanixrus 597
JAMIESON, GLEN S, and ELLEN K PIKITCH. Vertical distribution and mass mortality
of prawns, Pandalvs platyceros, in Saanich Inlet, British Columbia 601
YANG, MEI-SUN. Morphological differences between two congeneric species of pleuro-
nectid flatfishes: Arrowtooth flounder, Atheresthes stomias, and Kamchatka flounder,
A. evermanni 608
McINTYRE, JOHN D, REGINALD R. REISENBICHLER, JOHN M. EMLEN, RICHARD
L. WILMOT, and JAMES E. FINN. Predation of Karluk River sockeye salmon by coho
salmon and char 611
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 used or purchased because of this NMFS publication.
HARBOR PORPOISE, PHOCOENA PHOCOENA,
ABUNDANCE ESTIMATION FOR CALIFORNIA, OREGON, AND
WASHINGTON: I. SHIP SURVEYS
Jay Barlowi
ABSTRACT
The density and total population size of harbor porpoise along the coasts of California, Oregon, and
Washington are estimated from ship surveys using line transect methods. Surveys were completed
between September 1984 and May 1986 using teams of 3-5 observers. Data include 852 porpoise groups
sighted during 6,590 km of transects. Sighting rates varied more due to effects of sea state than due
to the presence of rain, fog, or sun glare. Experiments using additional observers indicate that approx-
imately 22% of trackline groups were missed by a team of 5 observers. Harbor porpoise density is
calculated from transects along the 18 m isobath and is extrapolated to other depth zones based on a
model of porpoise abundance as a function of depth. Total population size is estimated as 45,713 (SE
= 7,865) animals.
Approximately 200-300 harbor porpoise are taken
annually in central California set net fisheries (Dia-
mond and Hanan^; Hanan, et al.^). Little is known
about porpoise abundance in this area. Dohl et al.*
estimated that 1,600-3,000 porpoise reside in cen-
tral and northern California based on their aerial
surveys of coastal cetaceans. However, because har-
bor porpoise are frequently missed in aerial surveys
(Kraus et al. 1983), this estimate is probably low.
More information is needed on abundance, distribu-
tion, and population structure to determine the
significance of harbor porpoise mortality in set
nets.
Beginning in 1984, the National Marine Fisheries
Service (NMFS) has conducted ship and aerial
surveys of harbor porpoise abundance in California,
Oregon, and Washington. This report presents
results from four ship surveys. Results of the aerial
surveys are presented by Barlow et al. (1988).
'Southwest Fisheries Center La JoUa Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038.
^Diamond, S. L., and D. A. Hanan. 1986. An estimate of har-
bor porpoise mortality in California set net fisheries: April 1, 1983
through May 31, 1984. Adm. Rep. SWR-86-15, 40 p. Available
from National Marine Fisheries Service, Southwest Region, 300
S. Ferry Street, Terminal Island, CA 90731.
^Hanan, D. A., S. L. Diamond, and J. P. Scholl. 1986. An esti-
mate of harbor porpoise mortality in California net fisheries April
1, 1984 through March 31, 1985. Adm. Rep. SWR-86-16, 38 p.
Available from National Marine Fisheries Service, Southwest
Region, 300 S. Ferry Street, Terminal Island, CA 90731.
^Dohl, T. P., R. C. Guess, M. L. Duman, R. C. Helm. 1983.
Cetaceans of central and northern California, 1980-83: status,
abundance, and distribution. Report prepared for U.S. Minerals
Management Service, Contract #14-12-0001-29090, 284 p.
Porpoise density is estimated from survey data
using line transect models (Burnham et al. 1980).
Total abundance is estimated by extrapolating from
density observed along transect lines to the entire
area inhabited. Abundance in offshore regions is
based on a model of porpoise density as a function
of water depth. In addition to abundance estimation,
survey data are used to examine the effect of envi-
ronmental conditions on sighting efficiency and the
possibility of temporal changes in harbor porpoise
distribution.
METHODS
Ship Survey Methods
Surveys were conducted from two National
Oceanic and Atmospheric Administration (NOAA)
research vessels, the 52 m RV David Starr Jordan
(Surveys 1, 3, and 4) and the 54 m RV McArthur
(Survey 2)^. Both vessels were of similar design with
viewing stations located on top of the pilothouse
(viewing height was approximately 10 m above sea
surface). Transect lines followed as close as possi-
ble to the 18 m isobath (roughly 2-4 km from the
coast), although the actual depth along the transect
varied from approximately 15-45 m, depending on
the presence of local navigational hazards. The areas
Manuscript accepted May 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
^Cruise reports available from the Southwest Fisheries Center,
P.O. Box 271, La Jolla. CA 92038.
417
FISHERY BULLETIN: VOL. 86, NO. 3
surveyed are shown in Figure 1. Survey design
va^ried among the four trips:
Surveys 1 and 3 were in September 1984 and 1985,
respectively. Both were designed to survey harbor
porpoise density and abundance from Point Concep-
tion, CA to Cape Flattery, WA. An attempt was
made to survey the entire coastline on each of these
cruises, but several sections of the coast were missed
(Fig. 1) because of fog and heavy weather. Five
observation positions were used on these two
surveys.
Survey 2 was from 24 January to 9 February 1985
and was primarily designed to examine seasonal
changes in harbor porpoise distribution between
Point Conception, CA and Cape Flattery, WA. Data
from this cruise were not used for density or abun-
dance estimation.
1 \ \ \ 1 n 49°N
48°
- 47°
10 PORPOISE/KM
I I
II
SURVEY 1 SURVEY 2 SURVEY 3
SEPTEMBER JANUARY/FEBRUARY SEPTEMBER
1984 1985 1985
SURVEY 4
APRIL/MAY
1986
- 46°
- 45°
- 44°
41°
- 38°
37°
36°
- 35°
34°
- 33°
32°
123°
121°
119°
117°W
Figure 1.— Relative sightings per kilometer based on 30-minute latitudinal strata. Lines parallel to the coast indicate areas that were
surveyed. Histograms indicate relative numbers of harbor porpoise seen per kilometer of transect, with bars to the left indicating rela-
tive numbers in calm seas (Beaufort 0, 1, and 2) and bars to the right indicating relative number in rough seas (Beaufort 3, 4, and
5).
418
BARLOW: SHIP SURVEYS OF HARBOR PORPOISE
Survey ^ was from 24 April to 5 May 1986 and
was designed to investigate factors which affect har-
bor porpoise density estimation. The surveys con-
centrated on several areas of high porpoise density
in central California. The vessel's activities were
coordinated with a helicopter to gather information
on the avoidance of the ship by harbor porpoise. Ex-
periments were also conducted on survey 4 to deter-
mine whether an independent team of 3 observers
would sight any porpoise that were missed by the
primary team of 5 observers. Data from this survey
were not used for density or abundance estimation.
Typically, 8-10 observers were used on each
survey, with a rested observer starting every half
hour and rotating through 5 primary observation
positions at half-hour intervals. The 5 positions con-
sisted of port and starboard inboard observers, port
and starboard outboard observers, and a recorder
positioned amidship. The inboard observers
searched with 7 power (7 x ) binoculars from straight
ahead to 90° (survey 1) or to 45° (surveys 2, 3, and
4) on their respective sides of the vessel. On survey
1 the outboard observers searched with 25 x,
pedestal-mounted binoculars. Although sightings
could be made at great distances from the vessel
using the 25 x binoculars, these distant sightings
contributed little to the estimation of trackline den-
sity, and use of 25 x binoculars was discontinued.
On subsequent surveys, both the inboard and out-
board observers used 7 x binoculars. The outboard
observers searched from straight ahead to 90° on
their respective sides of the vessel. The recorder
searched in the immediate vicinity of the ship using
unaided eyes and (intermittently) 7x binoculars.
On survey 2, only 3 observation positions were
used from Point Conception to Point Sur, CA and
from Point Reyes, CA to Cape Flattery, WA. When
effort was reduced to 3 observers, the inside obser-
vation positions were eliminated.
On survey 4, a second team of 3 observers was
added to monitor the effectiveness of the principal
team. This monitor team searched using unaided
eyes and (intermittently) 7x binoculars from the
pilothouse deck (viewing height approximately 7 m
from sea surface). The principal team and the moni-
tor team did not communicate sighting information,
and independent records were kept.
Data were noted by the recorder on data coding
forms. Data on search effort included the beginning
and ending times and positions for continuous legs
of effort, the ship's heading and speed, personal
identification codes for the observers, sea surface
temperature, water depth, Beaufort sea state, sun
position relative to the ship, and codes indicating
the presence of rain or fog within 5 km. The ship
position was determined from a Loran navigational
system or by triangulation using coastal landmarks
and dead reckoning. Ship speed was recorded direct-
ly from the Omega/Loran system or was calculated
based on time and distance traveled between suc-
cessive position fixes along straight transect lines.
Water depth was measured using a 38 kHz acoustic
depth sounder.
Data for sightings consisted of the above ele-
ments, plus estimated group size, distance to shore,
an estimate of the angle between the trackline of
the ship and the group, and an estimate of the
distance from the ship to the group. Group size
refers to all the individuals associated with a sight-
ing event. In most cases, groups were closely asso-
ciated individuals that surfaced together (mean =
2.92, median = 2.0). In two cases (Point Arena and
Monterey Bay, CA), groups consisted of 50-80 loose-
ly associated individuals that were organized in
subgroups of 4-10). Group size was estimated and
recorded independently by each observer; the mean
of these estimates was used in subsequent analyses.
The angle from the trackline to the porpoise was
estimated visually with the aid of a pelorus mounted
in front of the observer stations, or, when 25 x
binoculars were used, from a calibrated collar on the
pedestal mount. On surveys 1 and 2, distances to
harbor porpoise were estimated visually using the
radar distance-to-shore as a reference, or, when 25 x
binoculars were used, distances were estimated
using calibrated reticles in the oculars. On surveys
3 and 4, distances were estimated using calibrated
reticles in the oculars of 7 x binoculars. Data were
also collected on the porpoises' direction of travel
relative to the ship.
The length of a transect was estimated as the
product of ship speed and elapsed time. To stratify
density estimates by sea state, rain, and fog, the
effort record was divided into segments during
which the sea state, rain, and fog codes did not
change.
In five areas, information was collected on varia-
tion in harbor porpoise density with water depth.
During survey 3, three sections of the coast were
surveyed intensively (Fort Bragg to Cape Vizcaino,
CA; Cape Blanco to Coquille Point, OR; and Cape
Lookout to Tillamook Head, OR), with transect lines
following the 18, 56, 92, and 185 m isobaths. On
survey 4, the 18 and 46 m isobaths were surveyed
in Monterey Bay, CA and in the vicinity of the Rus-
sian River, CA. These data formed the basis of a
model (below) to extrapolate porpoise density from
419
FISHERY BULLETIN: VOL. 86, NO. 3
the usual transect lines (along the 18 m isobath) to
deeper waters.
Helicopter Observations
During survey 4, a Hughes 500-D helicopter was
used to collect information on harbor porpoise
behavior in response to the survey ship. The heli-
copter flew approximately 10 km ahead of the
vessel, and 3 observers in the helicopter looked for
harbor porpoise. Once a group of harbor porpoise
was sighted, the helicopter hovered at 200-300 m
while observers made behavioral observations and
periodically recorded the helicopter's position using
an on-board Loran system. Fluorescein dye pack-
ages were dropped in the water to allow the heli-
copter to maintain its position when harbor porpoise
were diving. Radio communication was maintained
with personnel on the ship who also kept records
of the helicopter position using radar distances and
bearings based on returns from an X-band radar
transponder in the helicopter. The ship changed
course, when necessary, to ensure that it passed in
close proximity to the porpoise that were being
observed. Porpoise observers on the ship were not
aware of the helicopter's activities and were not told
of sightings made by the helicopter observers (al-
though they were able to see dye patches in some
cases). Behavioral observations from the helicopter
included time spent at the surface, time spent div-
ing, and direction of porpoise movement.
Density Estimation
Line transect methods were used to estimate the
density of harbor porpoise from sightings. The
assumptions of these methods are considered in
detail in the discussion. The usual formula for
estimating density (D) based on line transect
surveys of small cetaceans is given by
D =
m ■ n- G
2 ■ L
(1)
where /(O)
n
G
L =
the probability density function for
sightings evaluated at zero perpen-
dicular distance,
number of sightings of groups,
average group size calculated as the
total number of individuals in all
groups divided by the number of
groups (iN/n), and
length of the transect.
(Holt and Powers 1982; Hammond and Laake 1983;
Holt in press). I did not use mean group size explicit-
ly in abundance estimation, and density of harbor
porpoise individuals, D, was estimated as
D = /(O) ■ iRI2)
where R = the number of individuals
length of transect (XN/L).
(2)
seen per
Equation (2) is functionally equivalent to Equation
(1), but it simplifies variance estimation. Typically
when using Equation (1), variances (and possibly
covariances) must be estimated for/(0), G, and n.
Using Equation (2), variances are needed only for
/(O) and R, and covariance between mean group size
and number of groups is handled implicitly. Sight-
ing distributions appear to be independent of group
size, G, (Results section), hence no adjustments were
made to /(O) for group size bias.
The parameter /(O) is, in effect, a measure of sight-
ing efficiency and should not vary with porpoise
abundance. Sighting efficiency is, however, likely
to change with sighting conditions, such as Beau-
fort sea state. Given these expectations and because
relatively large sample sizes are needed to estimate
/(O) accurately, values for/(0) were estimated for
each survey by pooling all sightings within defined
sea state categories. In order to estimate density
on a finer scale, estimates of R were stratified by
geographic region and multiplied by the pooled esti-
mate of /(O).
The sighting probability density function evalu-
ated at zero distance, /(O), was determined
empirically by fitting curves to the frequency
distribution of sightings as a function of perpen-
dicular distance from the trackline (Burnham and
Anderson 1976). Differences in distributions of
perpendicular distance were tested using the
Kolmogorov-Smirnov 2-sample test. To avoid bias
due to rounding error, angle and radial distance data
were "smeared" (Butterworth 1982; Hammond and
Laake 1983). Angles were smeared by adding a
uniformly distributed random number between - 5°
and -1-5° to angle estimates. Radial distances were
smeared by adding a uniformly distributed random
number between 0.2 and -(-0.2 times the estimated
distance. These smearing levels were based on the
degree of rounding that was apparent from the data
(Barlow^).
^Barlow, J. 1987. Abundance estimation for harbor porpoise
(Phocoena phocoena) based on ship surveys along the coasts of
Cahfornia, Oregon, and Washington. Adm. Rep. LJ-87-05.
420
BARLOW: SHIP SURVEYS OF HARBOR PORPOISE
Several models were investigated for estimating
/(O) from sighting distributions. The FORTRAN pro-
gram Transect (Laake et al. 1979) was used to fit
2-, 3-, 4-, and 5-parameter Fourier series and
2-parameter exponential power series models. The
FORTRAN programs Hazard and Hermite (S. Buck-
land'') were used to fit the 2-parameter hazard rate
model (constrained such that parameter P > 2,
Buckland 1985) and the 1-, 2-, 3-, and 4-parameter
Hermite polynomial model (Buckland 1985). Of these
models, the 2-parameter hazard rate model was
selected based on its ability to fit the observed
distributions and its lack of dependence on group-
ing criteria (Buckland 1985).
Perpendicular distances were grouped into strata,
the size of which increased with perpendicular
distance: 0-25 m, 25-50 m, 50-100 m, 100-200 m,
200-400 m, 400-800 m, 800-1,600 m, and 1,600-
3,200 m. Several alternative groupings were inves-
tigated, and the choice of outpoints made very little
difference in estimates of /(O). The above strata
(increasing with distance) gave lower variances in
/(O) than when each stratum was of equal size
(possibly because the hazard rate model assumes a
distinct shoulder in the sighting distribution, and
that shoulder is lost if the first distance strata are
large).
No established criteria exist for choosing an appro-
priate perpendicular distance at which to truncate
sighting distributions. Burnham et al. (1980) recom-
mend that no more than 1-3% of sightings be
eliminated by truncation. Using this recommenda-
tion, models were not able to adequately fit the
observed sighting distributions. In this report, trun-
cation distance was chosen in four ad hoc steps:
1) The hazard rate model was fit to perpendicular
distance data truncated at distances of 400, 800,
1,600, and 3,200 m. 2) Truncation distances were
identified which gave acceptable x^ values (P >
0.1). 3) Of the acceptable truncation distances, the
standard error in/(0) was estimated empirically by
randomly drawing 10 samples (of the same size as
the original sample) from the observed distribution
of perpendicular distances and by calculating the
standard deviation of/(0) estimated from each ran-
dom sample. 4) Truncation distances were chosen
as those which gave the lowest coefficient of varia-
tion in/(0).
Variance in R, the number of porpoise seen per
kilometer, was estimated using jackknife statistics
(Efron 1982). Jackknife estimates were calculated
by first estimating the value of i? using all data. The
value, Rj^, was again estimated excluding the A;th
segment of search effort. This process was repeated
for each effort segment. To ensure that each kth seg-
ment was of equivalent length, effort segments with
the same sea state, rain, and fog codes were com-
bined in a linear array and were then divided into
10 segments of approximately equal length. The
variance in the estimate of R was calculated as
10
s2 =
~ • 1 (i?, - Rf
10 ^=1
(3)
Avail, from Southwest Fisheries Center, P.O. Box 271, La Joila,
CA 92038.
^S. Buckland, Inter-American Tropical Tuna Commission, P.O.
Box 271, La Jolla, CA 92038, pers. commun. July 1986.
The variance of D was estimated using the Good-
man (1960) product variance formula (assuming no
covariance) using this jackknife variance for R and
the above Monte Carlo variance for/(0).
Fraction of Missed Animals
On survey 4, a second, independent team of 3
observers were used to estimate the fraction of
harbor porpoise that are missed by the primary
team of 5 observers. The fraction of missed animals
in a sighting survey is analogous to the fraction of
unmarked animals in a mark/recapture experi-
ment (Pollock and Kendall 1987). This fraction
was estimated using the Chapman (1951) modifica-
tion of the Petersen (or Lincoln) index method
(Pollock and Kendall 1987). Confidence limits were
estimated using Adams' (1951) method, which
assumes a binomial sampling distribution. Standard
error was estimated using standard binomial
formulas.
Abundance Estimation
A model was used to estimate the number of har-
bor porpoise along the entire coastline based on the
density that was observed along the 18 m isobath.
In shallow areas, such as the Bering Sea and
Georges Bank, harbor porpoise are found a con-
siderable distance from land (Gaskin 1984), hence
offshore distribution is better modelled as a func-
tion of depth than as a function of distance from
shore. (Although harbor porpoise are also found in
very deep water in fjords and inland waterways of
Alaska [Taylor and Dawson 1984], this represents
a special case that is not applicable to coastal waters
considered here.) The model used to estimate abun-
dance was based on data collected on surveys 3 and
4 and on data from a ship surveys by La Barr and
421
FISHERY BULLETIN: VOL. 86, NO. 3
Ainley^ and Szczepaniak^ in central California. The
number of harbor porpoise seen per kilometer of
transect was taken as an index of relative density
along each isobath. A simple descriptive model was
then constructed to give relative density as a func-
tion of water depth.
Fifteen depth strata were used in abundance
estimation: 0-10, 10-20, 20-30, . . . , and 140-150 m.
The surface area within the strata was calculated
from digitized bathymetric data. Kelp beds were
assumed to be unsuitable as harbor porpoise habitat;
hence, kelp bed area was subtracted from the total
area within the 0-10 m stratum. Kelp bed areas for
the entire west coast were taken from Crandall
(1915). More recent estimates for limited areas in
central California are in good agreement with these
previous values (G. Van Blaricom^").
For each of 15 depth strata, the abundance of har-
bor porpoise was estimated as the product of their
density along the survey line (the 18 m isobath), the
density in that depth strata relative to that along
the survey line, the surface area included within that
depth strata, and the inverse of the estimated frac-
tion of trackline animals that were seen. Since
survey effort and harbor porpoise density both
varied geographically, abundance estimates were
made for each of 8 geographic regions (Fig. 2).
Areas within the depth strata were estimated from
NOAA bathymetric data. The estimate of total abun-
dance along the coast, A/'y, is therefore given by
15
^- = ^ I «^ .?. <^' ■ ^^.'>
(4)
where Dj = density of individuals observed on the
transect line in the jth geographic
strata,
4 = ratio of density in depth strata k to
that on transect line (see Figure 4),
Aji^ = area in geographic region j' and depth
strata k, and
F = the estimated fraction of trackline
animals seen by the usual team of 5
observers.
'LaBarr, M. S., and D. G. Ainley. 1985. Depth distribution of
harbor porpoise off central California: A report of cruises in April
and May-June 1985. Report to U.S. National Marine Fisheries
Service, Northwest and Alaska Fishery Center, 7600 Sand Point
Way N.E., Seattle, WA. Contract No. 41-USC252.
'Szczepaniak, I. D. 1987. Abundance and distribution of har-
bor porpoise (Phocoena phocoena) in the Gulf of the Farallones
National Marine Sanctuary. Contract report prepared for
National Park Service, Point Reyes National Seashore, Point
Reyes, CA 94956.
'"G. Van Blaricom, U.S. Fish and Wildlife Service, University
of California, Santa Cruz, CA 93106, pers. commun. August 1986.
Equation (4) was applied independently to the dif-
ferent surveys and, within surveys, to different sea
state strata. When combining estimates from differ-
ent sea states or different cruises, abundance was
calculated as the mean of the densities in each of
the stratum, weighted by the length of the transect
line within that stratum.
In estimating standard error for total abundance,
variances of products were calculated using the
Goodman (1960) product variance formula, and vari-
ances of ratios were estimated using a Taylor
approximation (Yates 1953, p. 198). Area was
assumed to be known without error. Statistical error
in the indices of abundance for the depth strata could
not be estimated given the paucity of available in-
formation. To account for uncertainty in the model
of depth distribution, three versions of the model
are proposed to span a range of possibilities.
RESULTS
On the four surveys, 852 groups of harbor por-
poise were sighted (an estimated 1,818 individuals).
A distance of 6,590 km was surveyed during 56
days. The number of sightings per kilometer sur-
veyed varied geographically and these geographic
patterns appeared to change appreciably between
cruises (Fig. 1).
Sighting Distributions
The number of sightings on the inshore and off-
shore sides of the vessels were approximately equiv-
alent (383 and 392, respectively). The cumulative
distributions of perpendicular sighting distances
were not significantly different for these two sides
(P = 0.06). Therefore, sighting distributions were
assumed to be symmetrically distributed about the
trackline, and the distributions of perpendicular
sighting distances from both sides of the vessel were
pooled for subsequent analyses.
The distributions of perpendicular sighting dis-
tances for the first three surveys were significant-
ly different from one another (P < 0.01 for all). This
was probably the result of the modifications in
survey methods between these cruises. Surveys 3
and 4 used the same methods, and sighting distribu-
tions were not significantly different (P = 0.39).
Given that changes in methods result in differences
in sighting distributions, all surveys were treated
separately in subsequent analyses.
Distributions of perpendicular distance were not
significantly different between individuals sighted
422
BARLOW: SHIP SURVEYS OF HARBOR PORPOISE
48°
46°
45»
44*
43"
420 .
41* -
40°
39°
38*
37°
36°
35°
34° -
33° -
32°N
132°W 131° 130° 129° 128" 127" 126" 125° 124° 123° 122° 121" 120° 119° 118° 117»W
Figure 2.— Geographic regions used as strata in abundance estimation. Broken line indicates the 200 m
isobath and delineates likely harbor porpoise habitat.
423
FISHERY BULLETIN: VOL. 86, NO. 3
alone (group size of 1) and larger groups (group size
of 3 and greater) (P = 0.56).
Environmental Conditions Affecting
Sightings
Sighting efficiency was not significantly affected
by rain, fog, or sun glare. Rain/fog conditions were
considered "poor" if rain or fog were present within
5 km of the vessel and "good" if neither were pres-
ent. The distributions of perpendicular sighting
distances were not significantly different between
these two strata (P = 0.32, 0.44, 0.78, and 0.64,
respectively, for surveys 1, 2, 3, and 4), and the
number of porpoise per kilometer surveyed was
higher in the "poor" category for two of the surveys.
Sun glare from the water's surface was considered
to contribute to "poor" sighting conditions if the sun
was within 45° of the trackline in front of the ship.
Conditions were considered "good" when the sun
was in other positions or was obscured by clouds.
As with rain/fog conditions, the distributions of
perpendicular sighting distances were not signifi-
cantly different between these "good" and "poor"
sun glare categories (P = 0.87, 0.47, 0.30, and 0.55,
respectively, for surveys 1, 2, 3, and 4). The number
of harbor porpoise per kilometer surveyed were
slightly higher in the poor category for three of the
surveys. In paired comparisons when glare was pres-
ent on only one side of the bow, approximately equal
numbers of sightings were made on the sides with
and without glare (60 vs. 59, respectively). All
categories of rain, fog, and glare are included in
subsequent analyses.
Sea state did have a significant effect on porpoise
sightings. Sea state was categorized as calm
(without white-caps, Beaufort sea states 0, 1, and
2) or rough (with white-caps, Beaufort sea states 3,
4, and 5) following the classification used by Holt
and Cologne (1987). Distributions of perpendicular
distances were not significantly different between
these categories for any of the surveys (P > 0.05);
however for all surveys combined, the number of
harbor porpoise detected per kilometer was much
lower during rough seas (0.32 km"^) than during
calm seas (1.22 km"^). There were insufficient
sightings to estimate density for rough seas separ-
ately; therefore, rough sea data were excluded in
subsequent analyses. For all three surveys, the
numbers of harbor porpoise detected per kilometer
was higher at Beaufort 0 & 1 than at Beaufort 2,
and for survey 3, the distributions of perpendicular
sighting distance were significantly different be-
tween these categories {P = 0.03). Porpoise density
is, therefore, estimated separately for Beaufort 0
& 1 and for Beaufort 2 conditions. (For comparison,
harbor porpoise abundance was also estimated pool-
ing Beaufort sea states 0, 1, and 2. Estimated abun-
dance was approximately the same by both methods,
but the variance was slightly lower using the
stratified sea state categories. For this reason, only
the stratified estimates are presented here.)
Helicopter Observations
Helicopter observation of the behavior of harbor
porpoise in response to the survey ship were made
on only 6 groups of animals. Plots of vessel tracks
and movements of the groups are given in the cruise
report (see footnote 5). Only in one case was a
distinct behavioral change noted in response to the
ship. In that case, when the vessel was within 800
m, the group moved rapidly, perpendicular to the
path of the vessel and then parallel to and in the op-
posite direction of the vessel. Observers on the ship
saw this harbor porpoise group as they moved rapid-
ly out of the path of the vessel. Observers on the
ship also saw 2 of the other 5 groups. Although this
sample of behavior is small, movement in response
to the survey vessel appeared limited to within 1 km
of the vessel and, when it occurred, animals did not
travel far from their original positions.
Porpoise Density
The probability density distributions of perpen-
dicular sighting distances are shown in Figure 3 for
surveys 1 and 3 and for Beaufort sea states 0 & 1
and 2. The hazard rate model gave acceptable fits
for all sighting distributions {P > 0.1) when the trun-
cation criteria was set at 400 m (Table 1). For survey
1, the optimum truncation points were chosen as 400
m for Beaufort 0 & 1 and 800 m for Beaufort 2; for
survey 3, this distance was 400 m for both Beaufort
sea state categories. The fits of these models are
shown in Figure 3. Estimates of density and stand-
ard errors are given in Table 2.
Depth Distribution Model
The model of harbor porpoise depth distribution
was based on the relative densities of harbor por-
poise at different water depths. Ship survey data
were pooled into five depth ranges: 18-37 m (10-20
fathoms), 37-55 m (20-30 fathoms), 55-73 m (30-40
fathoms), 73-91 m (40-50 fathoms), and 91-110 m
(50-60 fathoms). Ship surveys are generally not
practical inshore of the 18 m isobath, but estimates
424
BARLOW: SHIP SURVEYS OF HARBOR PORPOISE
10
0.0
0.1 0.2
PERPENDICULAR DISTANCE (km)
8
\
SURVEY 1, BEAUFORT 2
6 -
\
V,
4 -
\
\
2 -
^
0 -
■
0.4 00
10
0.4 0.0
0.2
0.4
0.6
0.8
SURVEY 3, BEAUFORT 2
-I r
0.1
0.2
1 r-
0.3
0.4
PERPENDICULAR DISTANCE (km)
Figure 3.— Probability density distributions for perpendicular sighting distances. Histograms indicate observed distributions, and solid
lines indicate the best fit of the hazard rate model to these data.
Table 1 .—Estimated values of the probability density functions
evaluated at zero perpendicular distance, ^(0). Estimates are based
on the hazard rate model and were made for truncation distances
of 400, 800, 1,600, and 3,200 m. Estimates are given only if the
model gave an acceptable fit to the data (P > 0.1). Asterisks in-
dicate f(0) values with the lowest coefficient of variation (paren-
theses).
Survey
Beaufort
sea state
Truncation distance
400 m 800 m 1 ,600 m 3,200 m
1
1
3
3
0 & 1
2
0 & 1
2
7.85 •
(0.23)
10.48
(0.59)
4.51 •
(0.22)
6.97 •
(0.19)
5.31
(0.29)
8.15 *
(0.21)
3.10
(0.66)
4.31
(0.24)
7.09
(0.33)
2.69
(0.31)
5.78
(0.51)
from aerial surveys (Barlow et al. 1988) show
roughly equal density at 0.61 and 1.85 km from the
shore (the latter corresponding approximately to the
18 m isobath). Relative density from ship surveys
was measured in the number of sightings per kilo-
meter of searching effort. Relative densities at
Table 2.— Density estimates, D, for harbor porpoise (km"^) along
the 18 m isobath in each of eight geographic strata. Density was
calculated per Equation (2), using estimates of f(0) (Table 1) which
had the lowest coefficients of variation. Values are not adjusted
for missed animals. Standard errors are in parentheses.
Geo-
graphic
region
1
2
3
4
5
6
7
8
Total
Survey 1
Survey 3
Beaufort
0 & 1
Beaufort
2
Beaufort
0 & 1
Beaufort
2
0.0
(0.0)
1.1
(1.0)
1.4
(0.8)
7.9
(3.4)
1.8
(0.8)
2.5
(1.1)
0.0
(0.0)
0.0
(0.0)
0.0
(0.0)
2.5
(1.0)
2.1
(1.3)
2.0
(0.6)
0.8
(0.6)
0.0
(0.0)
1.3
(0.6)
0.0
(0.0)
0.6
(0.5)
0.1
(0.1)
0.0
(0.1)
0.8
(0.6)
2.5
(0.8)
2.5
(0.6)
0.8
(0.4)
0.1
(0.1)
0.1
(0.1)
0.0
(0.0)
6.7
(7.2)
0.6
(0.3)
2.6
(0.9)
1.3
(0.7)
1.2
(0.5)
Pooled
estimates
0.04
(0.02)
0.51
(0.30)
0.03
(0.03)
2.83
(1.69)
0.91
(0.32)
2.22
(0.40)
2.64
(0.78)
1.09
(0.45)
1.33
(0.30)
425
FISHERY BULLETIN: VOL. 86. NO. 3
18-37 m show no consistent relationship to those
at 37-55 m or 55-73 m (Table 3), but on average
these appear to be approximately equal. Relative
densities at 18-37 m are, however, consistently
higher than densities at 73-110 m in all areas
(Table 3). A total of 236 km were searched in waters
deeper than 110 m and no harbor porpoise were
seen.
Despite high variability in patterns of depth
distribution and lack of ship coverage in shallow
waters, some generalizations can be made about the
depth distribution of harbor porpoise along the west
coast. The relative abundance of harbor porpoise ap-
pears to be roughly constant from shore to 55 m,
to be markedly lower at 73-110 m, and to be very
low in waters deeper than 110 m.
Based on the above relationships, I propose the
following preliminary model for the depth distribu-
tion of harbor porpoise along the coasts of Califor-
nia, Oregon, and Washington: constant abundance
from the coast to the 80 m isobath, linearly decreas-
ing abundance from the 80-120 m isobaths, and zero
abundance in waters deeper than 120 m (Fig. 4a).
Because considerable uncertainty exists in this
model, I propose two alternative models (Fig. 4b,
c). Alternative models b and c are less likely than
the primary model given because both conflict with
some of the available data. The alternative models
do, however, encompass the likely range of relative
density values and provide a means to evaluate the
sensitivity of the abundance estimate to different
models of depth distribution.
Fraction of Missed Animals
The experiment on survey 4 indicates that some
trackline groups were seen by 1 group of observers
and were missed by the other. A total of 103 sight-
ings was made by both teams, 33 of which were
estimated to be within 100 m perpendicular distance
from the transect line. Of the 103 total sightings,
85 were detected only by the 5 principal observers,
6 were detected only by the 3 monitor observers,
and 12 were detected by both teams. Of the 33
trackline sightings, 20 were detected only by the
principal observers, 3 were detected only by the
monitor observers, and 10 were detected by both
teams. The Petersen estimate of the fraction of
trackline porpoise seen by the primary team of 5
observers is thus 0.780 (SE = 0.117, 95% C.L. =
0.45-0.95). This indicates that approximately 22%
of trackline sightings are missed by the principal
teams of 5 observers.
Porpoise Abundance
Estimates of porpoise abundance in each of the
eight geographic strata are given in Table 4 for the
primary model of offshore distribution. Independent
estimates are given for survey 1 and for survey 3
Table 3. — Relative harbor porpoise abundance observed within the specified depth
ranges at a variety of study sites. Relative abundance is measured as number
of porpoise sightings made per kilometer. Numbers in parentheses indicate
kilometers surveyed.
Depth range
Location
18-37 m
37-55 m
55-73 m
73-91 m
91-110 m
Central California'
0.02
(89)
0.08
(172)
0.03
(403)
0.03
(279)
0.01
(166)
Gulf of the
Farallones, CA^
0.29
(181)
0.00
(159)
0.30
(133)
0.00
(7)
0.00
(7)
Fort Bragg, CA
0.05
(41)
—
0.05
(43)
—
0.03
(35)
Coquille Pt, OR
0.24
(43)
—
0.35
(52)
—
0.08
(50)
Tillamook
Head, OR
0.50
(57)
—
0.12
(52)
—
0.00
(17)
Monterey Bay, CA
0.46
(220)
0.29
(76)
—
—
—
Russian River, CA
0.00
(26)
0.30
(33)
—
—
—
'Data taken from LaBarr and Ainley (see text footnote 8) assuming an average survey speed
of 9 knots.
^Data taken from Szczepaniak (see text footnote 9) assuming an average survey speed of
9 knots.
426
BARLOW: SHIP SURVEYS OF HARBOR PORPOISE
DEPTH DISTRIBUTION MODELS
2
18
16
14
12
1
0.8
0.6
0 4
0.2
0
2
1.8
16
1.4
1.2
1
0.8
0 6
0.4
0.2
0
Primary Model
5 15 25 35 45 55 65 75 85 95 105 115 125 135 145
LU
>
LU
CC
CO
z
lU
Q
LJJ
>
I-
<
_J
UJ
CC
CO
z
lU
o
LU
>
LU
CC
5 15 25 35 45 55 65 75 85 95 105 115 125 135 145
WATER DEPTH (meters)
Figure 4.— Proposed models for the depth distribution of harbor porpoise expressed as relative den-
sities, 4, within 10 m depth intervals. Density at 20-30 m is assumed to be knowm and is given a
relative value of 1. Figures represent a) a primary model of offshore distribution, b) a high estimate
of offshore range, and c) a low estimate.
-
—
High Model
—
' F " '
1
5 15 25 35 45 55 65 75 85 95 105 115 125 135 145
in each area. Both surveys show similar patterns,
with higher abundances in the northern strata (4-8)
and very low abundance in strata 1 and 3. Despite
similar patterns, differences between the paired
estimates are in some cases, large and statistically
significant (^tests, P < 0.05). Because region 8 was
not covered on the third survey, it is not possible
to compare estimates of total abundance for the en-
tire coast between surveys. The total abundances
for regions 1-7 (Point Conception to the Columbia
River) are 46,550 (SE = 10,932) animals and 32,029
(SE = 10,906) animals for surveys 1 and 3, respec-
tively. The difference between these estimates is not
statistically significant (^test, P > 0.05). Pooling the
results of the two surveys, the estimate of harbor
porpoise abundance between Point Conception and
Cape Flattery in September of 1984 and 1985 is
45,713 (SE = 7,865) animals (Table 4). The same
estimate using the alternate models of offshore
distribution ranges from 28,769 to 78,019 (Table 5).
427
FISHERY BULLETIN: VOL. 86, NO, 3
Table 4.— Estimated abundance of harbor porpoise in each of the
eight geographic strata based on the primary model of offshore
distribution. Estimates for Beaufort 0 & 1 and for Beaufort 2 were
computed separately and then averaged, weighting by transect
length. Pooled estimates for the eight strata were obtained as an
average of the two surveys, weighting by transect length. All esti-
mates are adjusted for missed animals. Standard errors are in
parentheses.
Geographic
region
Survey 1
Survey 3
Pooled
estimates
1
2
3
4
5
6
7
8
Totals
Regions 1-3
Regions 1-7
Regions 1-8
2,401
(2,180)
0
6,909
(2,959)
11,245
(6,943)
9,061
(2,724)
16,934
(7,097)
9,808
(4,311)
2,401
(2,180)
46,550
(10,932)
56,358
(11,751)
126
(68)
932
(646)
153
(158)
9,096
(9,855)
3,296
(1,410)
12,786
(3,676)
5,641
(2,424)
1,210
(669)
32,029
(10,906)
96
(52)
1,459
(885)
112
(116)
7,909
(4,784)
4,806
(1,745)
11,107
(2,363)
10,416
(3,311)
9,808
(4,311)
1,667
(895)
35,904
(6,578)
45,713
(7,865)
DISCUSSION
Distribution
Harbor porpoise are not uniformly distributed
between Cape Flattery and Point Conception. Al-
though there are no obvious discontinuities within
this range, density varies geographically and tem-
porally. The most dramatic temporal changes are
between the two September surveys and the Janu-
ary-February survey (Fig. 1). The coasts of Wash-
ington and northern Oregon were found to have
relatively high densities of harbor porpoise in Sep-
tember, but, despite excellent sighting conditions,
very few porpoise were seen there in January. High
densities of harbor porpoise were also seen in
Monterey Bay on both September cruises and on
survey 4 in May. This area was intensively surveyed
in February, and few harbor porpoise were seen. As
can be seen in Figure 1, adjacent areas tended to
have similar densities within a survey. Less consis-
tency is found when the same areas are compared
between different surveys.
Table 5.— Estimated abundance of harbor porpoise in central
California (regions 1-3) and along the entire coast (regions 1-8)
based on two alternate models of offshore distribution. All
estimates are adjusted for missed animals. Standard errors are
in parentheses.
Survey 1 Survey 3
Pooled
estimates
Alternate Model b
Regions 1-3
Regions 1-8
Alternate Model c
Regions 1-3
Regions 1-8
3,966
(3,602)
95,132
(19,515)
1,505
(1,367)
35,736
(7,550)
1,986
(1,104)
770
(421)
2,744
(1,478)
78,019
(13,356)
1,054
(561)
28,769
(4,995)
The apparent changes in distribution could be
caused by small changes in depth distributions. The
majority of survey effort was along the 18 m iso-
bath. A large fraction of animals could be missed
if their depth distribution changed by 10 m or less.
More information on depth distributions is needed
before the apparent temporal changes in geographic
distribution can be interpreted.
Porpoise Density
Estimates of harbor porpoise density ranged from
0.03 to 2.8 animals/km^ along transect lines in the
eight geographic regions (pooled estimates, Table
2). In another study, Szczepaniak (fn. 9) estimated
0-1.9 porpoise/km^ in four study areas in the Gulf
of the Farallones, CA. Taylor and Dawson (1984)
found 1.2-5.9 porpoise/km- at study sites in Glacier
Bay, AK. Flaherty and Stark^^ estimated 0.8-1.6
porpoise/km^ in Washington Sound. Densities in
the present study are therefore within the range of
densities found in other areas along the same coast.
Harbor porpoise density was estimated for
California, Oregon, and Washington based on aerial
surveys that were concurrent with the present study
(Barlow et al. 1988). The overall estimate of harbor
porpoise density from that study (corrected for
missed animals) was 1.79 porpoise/km^. The overall
estimate from the ship survey (1.33 porpoise/km^)
can be corrected for missed animals to yield an
estimate of 1.73 porpoise/km^. Given that the coef-
I'Flaherty, C, and S. Stark. 1982. Harbor porpoise {Phocoena
phocoena) assessment in "Washington Sound". Final Report
#80-ABA-3584 submitted to National Marine Mammal Laboratory,
National Marine Fisheries Service, NOAA, 7600 Sand Point Way,
NE, Seattle, WA 98115. 84 p.
428
BARLOW: SHIP SURVEYS OF HARBOR PORPOISE
ficient of variation in the pooled ship estimates is
nearly 25%, these estimates are in very close agree-
ment. However, because the aerial estimates are
based only on the small fraction of the coastline that
was surveyed under optimal conditions, the ship
estimates are probably a better representation of
porpoise density for the entire coast.
Of the areas surveyed, harbor porpoise density is
highest in northern California and Oregon. The
highest density was seen in northern Oregon (region
7) during survey 1. The second highest density was
observed in northern California between Bodega
Head and Cape Mendocino (region 4) on survey 3.
Two areas in central California (regions 1 and 3)
were found to have very low densities. Region 1 in-
cludes the Big Sur coastline from Point Conception
to Point Sur. This area is characterized by steep
depth gradients and hence has little habitat that is
suitable for harbor porpoise. Region 1 was relatively
well covered, with 378 km of trackline surveyed at
Beaufort sea states 0-2. In contrast, region 3 in-
cludes the Gulf of the Farallons with its broad
coastal shelf within the 100 m isobath. Based on
surveys of 764 km, Szczepaniak (fn. 9) estimated
1,033 harbor porpoise are found in the Gulf of the
Farallones alone. This is much greater than my
estimate of 112 animals in region 3 based on only
175 km of survey effort. Because of his greater
amount of search effort in this area, I believe that
Szczepaniak' s estimates for region 3 are more ac-
curate than mine. Although regions 1 and 3 were
both identified as low density areas, more confidence
can be placed on this conclusion for region 1 than
for region 3.
Abundance
The size and behavioral characteristics of harbor
porpoise make estimating their abundance difficult.
Harbor porpoise are small, occur in groups of only
a few individuals, and surface without conspicuous
splashes; their distribution is extremely patchy.
Even v^th 5 observers, the effective path width that
can be searched from a ship is <1 km, and that path
width decreases very rapidly in rougher sea states.
All of these factors contribute to high variability in
the abundance estimates presented here. Seasonal
and year-to-year changes in the distribution of har-
bor porpoise may also contribute to the variability
seen within geographic strata. These are, however,
the best (and, for some regions, the only) estimates
of harbor porpoise abundance for the study area.
Although there are no prior estimates for Oregon
or Washington coasts, Dohl et al. (fn. 4) estimated
harbor porpoise abundance in central and northern
California. Their estimates range from 3,000 har-
bor porpoise in autumn to 1,600 in summer, which
correspond (approximately) to the pooled estimate
of 11,457 for regions 1-4 based on the present study.
There are, however, several problems with the ap-
plication of their methods to the estimation of har-
bor porpoise abundance. In a direct comparison with
shore counts, Kraus et al. (1983) showed that
observers on aircraft saw only 10-20% of harbor
porpoise groups. Dohl et al. (fn. 4) did not apply a
correction to account for harbor porpoise groups
that are submerged at the time the aircraft passed.
Also, Dohl et al. did not stratify estimates by
distance from shore or depth. Although most of their
harbor porpoise sightings were within 0.5 km (0.25
nmi) of shore, their density estimates were extrap-
olated to an area extending 166 km from the coast.
Estimates from the current study are based on
better methodology than previous estimates.
In addition to exposed coastal habitats, harbor
porpoise are also found in bays along the coasts of
California, Oregon, and Washington. Goetz (1983)
reported that harbor porpoise are found throughout
the year in Humboldt Bay, CA. Harbor porpoise
have been seen in San Francisco Bay, but are
described as rarely present^^. Abundance of harbor
porpoise in inland waters may, however, vary
seasonally (Taylor and Dawson 1984). No estimates
exist for the total number of harbor porpoise in-
habiting bays. Survey effort in the present study
was limited to exposed coastal areas (including
Monterey Bay, but excluding San Francisco Bay,
Humboldt Bay, Coos Bay, Yaquina Bay, the mouth
of the Columbia River, Willapa Bay, and Grays Har-
bor). If harbor porpoise density in bays were the
same as that which was observed along the 18 m
isobath, population sizes presented here could be in-
creased by approximately 3.1% to account for por-
poise inhabiting 900 km"^ (the approximate com-
bined area of Humboldt Bay, Coos Bay, Yaquina
Bay, the mouth of the Columbia River, Willapa Bay,
and Grays Harbor).
Line Transect Assumptions
Biases in abundance estimates can be an even
greater problem than high variability. In the case
of estimates presented here, biases could be intro-
duced if the assumptions of line transect sampling
'^Szczepaniak, I. D., and M. A. VV^ebber. 1985. Status of the
harbor porpoise {Phocoena phocoena) in the eastern North Pacific,
with an emphasis on California. Contract report to the Center
for Environmental Education, Washington, D.C., 52 p.
429
FISHERY BULLETIN: VOL. 86, NO. 3
are not met (Burnham et al. 1980; Hammond and
Laake 1983). Of these assumptions, the most rele-
vant to this study are 1) the area must be sampled
randomly or the animals must be randomly distrib-
uted within the area; 2) all groups on the trackline
must be detected; and 3) group size must be esti-
mated without error. These assumptions will be
addressed below.
To address the first assumption (random distribu-
tion), cruise tracks were chosen to systematically
cover the coast from Point Conception to Cape Flat-
tery. Because the surveys were designed to cover
the entire longshore range of harbor porpoise in this
area, randomly placed survey tracks were deemed
unnecessary. vVlthough some areas of the coast were
missed, these locations were determined by weather
and were presumably not correlated with porpoise
abundance. Surveys were, however, limited to a
very narrow strip along the 18 m isobath. Initially,
the choice of this survey track was based on the
observation that, in aerial surveys, harbor porpoise
were usually found within 0.5 km (0.25 nmi) of the
shoreline in California (Dohl et al. fn. 4). The 18 m
isobath was simply the shallowest reasonable work-
ing depth for the NOAA survey ships. In the course
of these surveys, it was found that harbor porpoise
are commonly distributed much further from the
coast than 0.5 km and that one survey track could
not adequately cover their habitat. The offshore
distribution of harbor porpoise is not random, but
is related to water depth, distance from shore, or
both. The model from which I extrapolated density
at 18 m to density at other depths was based on a
rather limited sample at a few locations along the
coast. The assumption of random search in offshore
areas was not met. Additional work is required to
evaluate the effect of this.
The second assumption is that 100% of the
animals in the immediate vicinity of the trackline
were detected. Animals near the trackline can be
missed because they move away from the path of
the ship, because they do not surface within the
visual range of the observers, or because the
observers fail to detect animals that do surface. Any
of these would result in a negative bias and an
underestimation of porpoise abundance using line
transect methods. These three problems are con-
sidered in more detail.
West-coast harbor porpoise are commonly said to
avoid vessels (Flaherty and Stark fn. 11; Szczepa-
niak and Webber fn. 12) and may be missed or not
counted in the proper perpendicular distance
category for this reason. On the surveys, the major-
ity of harbor porpoise were oriented roughly parallel
to the ship at the time they were sighted and were
swimming parallel to the ship and in the opposite
direction (see footnote 5). This was also observed
in one instance from the helicopter; however, in that
case the group first moved perpendicular to the path
of the ship. These observations indicate that harbor
porpoise are reacting to the ship before they are
seen by observers. Reaction to and avoidance of the
ship does not necessarily mean that estimates of
trackline density are biased if animals are detected
before they travel an appreciable distance from the
trackline. In several instances, harbor porpoise sur-
faced within 50 m of the ship and directly in its path.
These animals appeared startled and quickly moved
to avoid the ship. In these cases, the rapid move-
ment of the animals and splashes associated with
that movement made the animals more visible to
observers. Because avoidance behavior may make
harbor porpoise more visible and because the
distributions of perpendicular distance show only a
single mode (at the origin), vessel avoidance prob-
ably does not introduce a large bias in harbor por-
poise abundance estimation. More work is needed
in this area.
Harbor porpoise near the trackline may also be
missed if they either inadvertently or intentionally
do not surface within the visual range of the ob-
servers. Typical mean dive times for harbor porpoise
have been measured as 1.5-2.3 minutes (Glacier
Bay, AK; Taylor and Dawson 1984), 1.8 minutes
(northern Oregon; B. Taylor^^), and 0.4-1.4 minutes
(Bay of Fundy; Watson and Gaskin 1983). The ships'
speed during surveys was approximately 18.5 km/h
or 310 m/min; thus, in 2 minutes the ship would
travel 620 m. The average distance at which animals
were first seen was 704 m from the ship. If in-
dividual dive times were appreciably longer than 2
minutes, some trackline individuals would not be
detected by observers. In data collected in north-
ern Oregon, 16% of dive times were greater than
2.5 minutes (B. Taylor fn. 13). In addition, harbor
porpoise have been reported to increase dive times
up to 7 minutes in the presence of boat traffic
(Flaherty and Stark fn. 9). (This latter estimate is
considerably longer than any other published esti-
mate, and it is possible that those researchers missed
one or more surfacings). Helicopter observations in
Monterey Bay indicated that porpoise groups did not
extend dive times in the presence of the survey
vessel (see footnote 5). This area might not be repre-
sentative, however, because harbor porpoise may
''B. Taylor, Department of Biology, University of California, San
Diego, CA 92093, pers. commun. August 1986.
430
BARLOW: SHIP SURVEYS OF HARBOR PORPOISE
be more accustomed to vessel traffic there than
along the majority of the coast. It is likely that some
harbor porpoise are missed because they do not sur-
face near the vessel; however, it is not possible to
quantify this source of bias without additional study.
Trackline animals may be missed even if they do
not avoid the ship and do surface within visual range
of the observers if their surfacing is not detected.
In another study comparing ship surveys to aerial
and shore surveys, Kraus et al. (1983) found that
observers on ships saw only about 50% of the har-
bor porpoise in an area. In that study, however, ship
observers stood only 2.5 m above the sea surface
(versus 10 m in this study), and the estimate of 50%
was based on all groups, not just on trackline
animals. Based on the experiment using monitor
observers in the present study, an estimated 22%
of harbor porpoise that surface on the trackline are
missed by the usual team of 5 observers. If this is
underestimated by some percentage, population size
would be underestimated by the same percentage.
The third critical assumption is that group size is
estimated without error. In the case of harbor por-
poise, group size is small and estimates are typical-
ly based on actual counts. For tropical dolphins,
which school in groups of several hundreds, the
problem of group size estimation is more acute (Holt
and Powers 1982; Hammond and Laake 1983). Only
in two instances did harbor porpoise group size
exceed 20: in Monterey Bay and near Point Arena,
both in California. Excluding these two sightings,
mean group sizes are 2.05, 2.33, 2.03, and 1.59 for
surveys 1, 2, 3, and 4 (respectively); including the
two sightings, means are 2.30 and 2.26 for surveys
1 and 3. These values are comparable to other esti-
mates of mean group size for coastal populations of
harbor porpoise: 2.2 based on aerial surveys in
California (Dohl et al. fn. 4), 2.6 based on ship
surveys in the Gulf of the Farallons (Szczepaniak
and Webber fn. 12), 2.3 based on shore surveys in
northern Oregon (see footnote 5), and 2.75-3.23
based on aerial surveys along California, Oregon,
and Washington (Barlow et al. 1988). The consis-
tency of all these estimates from different platforms
indicates that group size estimation from ships is
not likely to be a major source of bias in abundance
estimation.
Variance Estimation
Although the estimates of standard error for
abundance and density are very high, these may still
be underestimates because the choice of a trunca-
tion criterion was based on minimizing variance and
because all possible sources of sampling errors were
not considered. The model upon which relative abun-
dance in the various depth strata was based is too
crude to allow reasonable estimates of its variabil-
ity. Estimates based on alternate models of depth
distribution indicate that abundance estimation is
relatively sensitive to the choice of models. Addi-
tional field work may help refine this model and
allow estimation of variance for the parameters 4
in Equation (4).
ACKNOWLEDGMENTS
Surveys of this magnitude could not be executed
without the help of many people. I thank the survey
crews for their many hours of labor: S. Beavers,
P. Boveng, S. Bragg, S. Chivers, S. Diamond,
V. DoUarhide, J. Flanders, B. Goetz, S. Hawes,
S. Heimhch-Boran, A. Hohn, S. Kruse, S. Mizroch,
F. Mann, M. Newcomer, R. Rasmussen, A. Read,
A. Robles, M. Scott, K. Sechiguchi, S. Sexton,
G. Silber, I. Szczepaniak, B. Taylor, B. Troutman,
M. Webber, J. Wexler, and K. Wynne. The heli-
copter crew included R. Holt, C. Oliver, and
B. Taylor. The initial survey was planned in col-
laboration with D. DeMaster and T. Jackson. I thank
S. Buckland, R. Holt, and S. Sexton for aid in
analysis and interpretation of line transect data.
B. Taylor provided unpublished data from shore
observations of harbor porpoise in northern Oregon.
Geographic areas were computed by K. Forney.
R. Allen prepared the figures presented here. This
manuscript was improved by the critical reviews of
P. Boveng, R. Brownell, D. Chapman, D. DeMaster,
S. Diamond, D. Goodman, D. Hanan, A. Hohn,
R. Holt, J. Lecky, S. Sexton, G. Smith, and
B. Taylor.
LITERATURE CITED
Adams, L.
1951. Confidence limits for the Petersen or Lincoln index
used in animal population studies. J. Wildl. Manage. 15:
13-19.
Barlow, J., C. Oliver, T. D. Jackson, and B. L. Taylor.
1988. Harbor porpoise, Phocoena phocoena, abundance esti-
mation for California, Oregon, and Washington: II. Aerial
surveys. Fish. Bull., U.S. 86:433-444.
Buckland, S. T.
1985. Perpendicular distance models for line transect sam-
pling. Biometrics 41:177-195.
BURNHAM, K. P., AND D. R. ANDERSON.
1976. Mathematical models for nonparametric inferences
from line transect data. Biometrics 32:325-336.
BuRNHAM, K. P., D. R. Anderson, and J. L. Laake.
1980. Estimation of density from line transect sampling of
biological populations. Wildl. Monogr. No. 72, 202 p.
431
FISHERY BULLETIN: VOL. 86, NO. 3
BUTTERWORTH, D. S.
1982. On the functional form used for g{y) for Minke whale
sightings, and bias in its estimation due to measurement in-
accuracies. Rep. int. Whal. Comm. 32:883-888.
Chapman, D. G.
1951. Some properties of the hypergeometric distribution
with applications to zoological censuses. Univ. Calif. Publ.
Stat. 1:131-160.
Crandall, W. C.
1915. II. The kelp beds from lower California to Puget Sound.
In F. W. Cameron (editor), Potash from kelp, 122 p. U.S.
Dep. Agric. Rep. No. 100.
Efron, B.
1982. The jackknife, the bootstrap and other resampling
plans. CBMS Regional Conference Series in Applied
Mathemetics 38, 92 p. Society for Industrial and Applied
Mathematics, Phila.
Gaskin, D. E.
1984. The harbour porpoise (Phocoena phocoena L.): regional
populations, status, and information on direct and indirect
catches. Rep. int. Whal. Comm. 34:569-586.
GOETZ, B. J.
1983. Harbor porpoise (Phocoena phocoena L.) movements
in Humboldt Bay, California and adjacent ocean waters.
MS Thesis, Humboldt State University, Humboldt, CA, 118
P-
Goodman, L. A.
1960. On the exact variance of products. J. Am. Stat. Assoc.
55:708-713.
Hammond, P. S., and J. L. Laake.
1983. Trends in estimates of abundance of dolphins (Stenella
spp. and Delphinics delphis) involved in the purse-seine
fishery for tunas in the eastern tropical Pacific Ocean,
1977-81. Rep. int. Whal. Comm. 33:565-588.
Holt, R. S.
In press. Estimation of abundance of dolphin stocks taken
incidentally in the eastern tropical Pacific yellowfin tuna
fishery. Mar. Mamm. Sci.
Holt, R. S., and J. Cologne.
1987. Factors affecting line transect estimates of dolphin
school density. J. Wildl. Manage. 51:836-843.
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. U.S. Dep. Commer.,
NOAA Tech. Memo. NOAA-TM-NMFS-SWFC-23, 95 p.
Kraus, S. D., J. R. Gilbert, and J. H. Prescott.
1983. A comparison of aerial, shipboard, and land-based
survey methodology for the harbor porpoise, Phocoena pho-
coena. Fish. Bull., U.S. 81:910-913.
Laake, J. L., K. P. Burnham, and D. R. Anderson.
1979. User's manual for program TRANSECT. Utah State
University Press, Logan, Utah, 26 p.
Pollock, K. H., and W. L. Kendall.
1987. Visibility bias in aerial surveys: A review of estimation
procedures. J. Wildl. Manage. 51:502-510.
Taylor, B. L., and P. K. Dawson.
1984. Seasonal changes in density and behavior of harbor por-
poise (Phocoena phocoena) affecting census methodology in
Glacier Bay National Park, Alaska. Rep. int. Whal. Comm.
34:479-483.
Watson, A. P., and D. E. Gaskin.
1983. Observations on the ventilation cycle of the harbour
porpoise (Phocoena phocoena L.) in the coastal waters of the
Bay of Fundy. Can. J. Zool. 61(1):126-132.
Yates, F.
1953. Sampling methods for censuses and surveys. Hafner
Publishing Co. Inc., N.Y., 401 p.
432
HARBOR PORPOISE, PHOCOENA PHOCOENA,
ABUNDANCE ESTIMATION FOR CALIFORNIA, OREGON, AND
WASHINGTON: II. AERIAL SURVEYS
Jay Barlow,! Charles W. Oliver,^ Terry D. Jackson,^ and
Barbara L. Taylor^
ABSTRACT
We conducted aerial surveys in September 1984 and September and October 1985 to determine the abun-
dance of harbor porpoise along the coasts of California, Oregon, and Washington. Two observers and
a recorder searched along predetermined transect lines at 0.61 and 1.85 km offshore. Strip transect
methods were used. A total of 366 groups of harbor porpoise were seen in the 9,500 linear kilometers
that were surveyed. Apparent density was significantly affected by sea state and cloud cover. Using
observations made during optimal conditions (clear skies and calm seas), apparent harbor porpoise den-
sity averaged 0.56 animals km'". Behavioral observations from shore and from a helicopter indicated
that porpoise are near the surface only 23.9% of the time. To account for this, porpoise density was
multiplied by a factor of 3.2, resulting in an adjusted estimate of 1.79 animals km"". Only a small percent-
age of the total area inhabited was surveyed under optimal sighting conditions, hence density estimates
were not extrapolated to estimate total porpoise abundance. Harbor porpoise density showed similar
patterns to those measured from ship surveys, and adjusted aerial estimates are approximately equal
to ship estimates.
Harbor porpoise, Phocoena phocoena, are subject to
mortality in the halibut set net fishery in central
California (NMFS^; Diamond and Hanan^). To evalu-
ate the significance of this mortality, an estimate
of population size is needed. Two aerial surveys and
three ship surveys were conducted from 1984 to
1986 to gather information on harbor porpoise abun-
dance along the coasts of California, Oregon, and
Washington. Observations were also made from
shore-based stations and from a helicopter to pro-
vide ancillary information needed for population
estimation. Results and population estimates from
'Southwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038.
^National Marine Fisheries Service, Southwest Region, 300
South Ferry Street, Terminal Island, CA 90731; present address:
Pacific Marine Center, NOAA, 1801 Fairview Avenue East,
Seattle, WA 98102.
^National Marine Fisheries Service, National Marine Mammal
Laboratory, 7600 Sand Point Way, N.E., Seattle, WA 98115;
present address: University of California, San Diego, Department
of Biologj', La Jolla, CA 92093.
^NMFS. 1980. A report based on the workshop on stock
assessment and incidental take of marine mammals involved in
commercial fishing operations. January 1980. Available from
National Marine Fisheries Service, National Marine Mammal
Laboratory, 7600 Sand Point Way NE, Seattle, WA 98115.
^Diamond, S. L., and D. A. Hanan. 1986. An estimate of har-
bor porpoise mortality in California set net fisheries: April 1, 1983
through March 31, 1984. Adm. Rep. SWR-86-15, 40 p. Available
from National Marine Fisheries Service, Southwest Region, 300
S. Ferry Street, Terminal Island, CA 90731.
the ship surveys are reported by Barlow (1988).
Preliminary results from the 1984 and 1985 aerial
surveys were presented by Oliver and Jackson^ and
Oliver^ respectively. Here we present population
density estimates based on the aerial surveys and
on shore and helicopter observations.
The aerial surveys were flown in September of
1984 and in September and October of 1985 from
Point Conception, CA to Cape Flattery, WA. Sur-
veys were coordinated by the National Marine Fish-
eries Service (NMFS) in collaboration with the
California Department of Fish and Game, the Ore-
gon Department of Fish and Wildlife, and the
Washington Department of Wildlife. Survey design
was based on information given by Dohl et al.^
regarding harbor porpoise distribution in California.
They reported that harbor porpoise were usually
^Oliver, C. W., and T. D. Jackson. 1987. Occurrence and
distribution of marine mammals at sea from aerial surveys con-
ducted along the U.S. west coast between December 15, 1980 and
December 17, 1985. Adm. Rep. LJ-87-19, 189 p. Available from
National Marine Fisheries Service, Southwest Fisheries Center,
P.O. Box 271, La Jolla, CA 92038.
^Oliver, C. W. 1986. Trip report: 1985 harbor porpoise aerial
survey, September 9 to October 15, 1985. Adm. Rep. LJ-86-21,
29 p. Available from National Marine Fisheries Service, South-
west Fisheries Center, P.O. Box 271, La Jolla, CA 92038.
»Dohl, T. P., R. C. Guess, M. L. Duman, R. C. Helm. 1983.
Cetaceans of central and northern California, 1980-83: status,
abundance, and distribution. Report prepared for U.S. Minerals
Management Service, contract #14-12-0001-29090.
Manuscript accepted May 1988. ,
fishery BULLETIN: VOL. 86, NO. 3, 1988. V//' ^Yj
433
FISHERY BULLETIN: VOL. 86, NO. 3
found within 0.25 nautical miles (nmi) of the shore-
line. We therefore designed our aerial surveys to
cover a very narrow coastal band. Subsequent in-
formation from ship surveys (Barlow 1988) has
shown their distribution to extend considerably far-
ther from the coast. Therefore, estimates of porpoise
density from aerial surveys apply to a relatively
small portion of harbor porpoise habitat. For this
reason we do not estimate population size by ex-
trapolating aerial density estimates to the entire
area inhabited. The density estimates presented
here are used to corroborate estimates based on ship
surveys and to estimate density for areas that were
too shallow to be surveyed by ship.
Based on previous studies of dive times (Watson
and Gaskin 1983; Taylor and Dawson 1984) we ex-
pected a proportion of the harbor porpoise to be
diving and therefore missed by aerial observers.
Shore-based studies were conducted in September
1985 from cliffs in northern Oregon to determine
average dive times for west-coast harbor porpoise.
Helicopter observations were made in April and May
1986 in Monterey Bay and near Bodega Head, CA
to gather dive time information and obtain a direct
measure of the fraction of time that harbor porpoise
groups are visible from the air. These two samples
did not differ significantly from previous samples
of harbor porpoise in Alaska, so all samples were
pooled to adjust estimates of porpoise density from
aerial surveys to account for the probability of miss-
ing submerged animals.
METHODS
Aerial Survey Methods
Strip transect methodology (Seber 1973) was used
during the aerial surveys. This method assumes that
all individuals within a transect strip are detected.
Transect lines were flown parallel to the coast line
at distances of 0.61 and 1.85 km (0.33 and 1.0 nmi)
offshore. Transect strips of equal width were sur-
veyed on both sides of the aircraft. The margins of
the strips were denoted by tape marks or streamers
on the wing struts. Strips were divided into inside
and outside swaths of unequal width (Fig. 1) by a
third tape mark or streamer between the other two.
When porpoise were sighted within the transect
strip, the pilot was directed to leave the transect
line and circle over the porpoise to obtain an ac-
curate count of the number within the original group
that was sighted. If additional groups or individuals
were sighted during this circling, they were ex-
cluded from density estimates. Porpoise density, x,
was calculated as the number of individuals sighted
within a transect, n, divided by the product of the
transect width, w, times the distance, d, that was
flown:
X = n/iW ■ d)
(1)
We used both single and twin-propeller, high-
wing, 4-passenger aircraft in our surveys. The
search team consisted of two observers seated in the
right and left passenger seats. A data recorder sat
in the copilot's seat and did not search. If the
recorder sighted animals that were missed by the
observers, these were noted but were not included
in density estimates. The planes were flown at an
altitude of 213 m (700 feet) and at an airspeed of
158-167 km/h (85-90 knots). The original survey
plan called for all sections of the coast to be covered
twice on each survey. This was accomplished in
1984, but poor weather in 1985 resulted in the
Washington coast and part of the Oregon coast be-
ing covered only once. The dates flown and areas
covered are given in Table 1.
Table 1.— Dates, areas covered, and observer teams during
aerial surveys for harbor porpoise. Geographic regions refer
to those shown in Figure 2. Observer team refers to a pair of
individuals.
Regions
Observer
Distance
Date
covered
team
to shore (km)
9/09/84
7,8
A
0.61
9/10/84
5,7
A
0.61
9/11/84
5,6
A
0.61
9/1 3/84
1,2,3,4,5
B
0.61
9/14/84
1,2,3,4,5
B
0.61
9/17/84
1,2,3,4
B
0.61
9/18/84
1,2,3,4
B
0.61
9/11/85
6,7
C
0.61
9/16/85
1,2,3
D
0.61
9/17/85
4
D
0.61
9/18/85
4,5
D
0.61,1.85
9/1 9/85
4,5
D
0.61,1.85
9/20/85
2,3
D
1.85
10/04/85
7,8
E
0.61
10/14/85
1,2,3
D
1.85
10/15/85
4,5
D
1.85
Data gathered on both the 1984 and 1985 surveys
were similar in format. Recorded data on sighting
conditions included Beaufort sea state, a measure
of cloud cover, a code indicating the presence of haze
or fog, sun position relative to the aircraft, and a
subjective measure of the observers' ability to see
into the water through turbidity, surface reflection
and diffraction. The latter was called surface pene-
tration and was recorded separately for each of the
434
BARLOW ET AL.: AERIAL SURVEYS OF HARBOR PORPOISE
STRIP TRANSECT
300 400
DISTANCE (meters)
500
600
700
E
CO
o
H
\ Inside
\ Swatt
\. c
1 ^
)utsi(
vath
--.
1985
\ '-y/yy/y//.
•I-;-;-;-!-;-!";'!
Wk
|------^-N ^ 1
100 200 300 400
DISTANCE (meters)
500
600
700
Figure 1.— Configuration of transect strip widths and distributions of perpendicular sighting distances for aerial
surveys in 1984 and 1985. Angles are given as declinations from horizontal. Histograms indicate the relative number
of porpoise seen in the given distance interval.
four swaths (inside and outside on both sides of the
aircraft). In severe sun glare conditions, searching
was discontinued for one or both swaths on one side
of the aircraft. Additional data included individual
observer numbers, date, time, and position (meas-
ured to lOths of minutes of latitude and longitude).
All of these data were recorded at the beginning and
end of continuous transects and whenever condi-
tions changed or a sighting was made. Additional
data were recorded for marine mammal sightings,
including a code for the species of animal seen, an
estimate of the number of individuals, a code in-
dicating on which side of the aircraft the animals
were seen, and a code indicating in which of the
swaths the animals were found (inside, outside, or,
if the animals were not within the designated strips,
neither).
An attempt was made to gather the above infor-
mation in a similar manner for both surveys. Some
differences in subjective measures of sighting con-
ditions could, however, be expected because there
was no overlap in observers between years. In addi-
tion, there were some differences in design between
the 1984 and 1985 surveys. In 1984, the inside and
outside swaths were from 123 to 305 m and from
305 to 620 m, respectively (as measured from the
midline of the transect). The margins of these
swaths corresponded to declinations angles of 19°,
35°, and 60° (Fig. 1). In 1985, the swaths were
91-294 m and 294-503 m and corresponded to
declination angles of 23°, 36°, and 67° (Fig. 1). In
1984, effort was concentrated on the inside swath
and the outside swath was not intended for abun-
dance estimation. The change in swath size was in-
tended to reduce the total area being searched, thus
potentially allowing the outer swath to be used for
density estimation. In 1985, effort was divided
equally between the two swaths. During the 1985
435
FISHERY BULLETIN: VOL. 86, NO. 3
surveys the declination angle to marine mammals
was measured when the animals were perpendicular
to the aircraft using hand-held inclinometers.
Declination angles were not measured during the
1984 survey.
Distance to the coast was monitored using
declination angles (19° for 0.61 km and 6° for 1.85
km). On the 1984 survey, this distance was 0.61
km for the entire survey. During the 1985 survey,
we surveyed at both 0.61 and 1.85 km from the
coast. In 1984, the coast was taken to be the outer
limit of the surf zone. In 1985, the coast was taken
to be the outer limit of the surf zone or, if kelp
beds were present, the outer margin of those
beds.
Shore Observation Methods
Observations of harbor porpoise diving behavior
were made from rocky headlands in northern Ore-
gon (Tillamook Head, Neahkahnie Mountain, Cape
Meares, and Cape Lookout) immediately before the
second aerial survey (7-11 September 1985). Ob-
servers were equipped with 7 x 50 binoculars with
compasses and ocular reticles and a single 20 x 120
binocular. Ventilation data were collected whenever
possible and included the number of animals at the
surface and the length of time spent at the surface.
Observations were recited aloud by the observer and
were written down by a second person or were
recorded onto magnetic tape. The ventilation cycle
typically consisted of a period with several surfacing
rolls and breaths (which we call a surfacing series)
followed by a much longer period of submergence
(which we call a dive). This dive cycle corresponds
to ventilation pattern B as described by Watson and
Gaskin (1983) for harbor porpoise in the Bay of
Fundy and the pattern described by Taylor and
Dawson (1984) for porpoise in Glacier Bay.
Helicopter Observation Methods
Behavioral observations were also made by three
observers in a 4-passenger, jet-turbine helicopter.
Upon locating a group of harbor porpoise, a fluores-
cein dye marker was dropped and the helicopter
hovered or circled slowly above the group at an
altitude of approximately 300 m. The number of
animals, the time they were visible at the surface,
and the dive times were recorded, along with infor-
mation on cloud cover, sea state, and water turbid-
ity. Each behavioral session was given a subjective
rating based on how well the observers could follow
the group and obtain accurate dive times. Only ses-
sions with good or excellent ratings were included
in analyses.
Probability of
Missing Submerged Animals
Given that a porpoise would be within the visual
range of an observer, the probability that it will be
at the surface during the passage of the aircraft is
related to the average time it spends at the surface,
s, the average time spent below the surface, d, and
the window of time during which it is within the
visual range of an observer, t. This probability was
calculated as
Pr (being visible)
s -I- t
s + d
(2)
The probability of missing a submerged animal is
equal to the complement of this value.
Density Estimation
Density of harbor porpoise was estimated as the
number of animals seen divided by the area searched
(Equation (1)). This raw density estimate was ad-
justed by dividing by the probability that an animal
would be visible from the air at any given instant
(Equation (2)). The area searched was estimated as
the swath widths times the lengths of the transects.
Transect lengths were calculated as the sum of the
great circle distances between successive position
fixes. Densities were calculated for each of the eight
statistical regions used by Barlow (1988) (Fig. 2).
The statistical difference in harbor porpoise
density between different sighting conditions or
different areas was tested using the raw density
estimates. Density estimates for short transects
were frequently zero, thus violating the parametric
assumptions of normally distributed, homoscedastic
error. Nonparametric tests were therefore chosen
for density comparisons. In discussing statistical
tests, a transect segment refers to the length of
transect line between two successive position fixes
and are typically <20 km. The measured variables
relating to sighting conditions are constant within
a segment, and because each sighting is accom-
panied by a new position fix, a segment will contain
at most, one sighting.
Whenever applicable, the Wilcoxon paired-sample
test (Wilcoxon 1945) was used to test one factor
while controlling for as many other factors as possi-
ble. Ten paired measures of density were created
436
BARLOW ET AL.: AERIAL SURVEYS OF HARBOR PORPOISE
49°N
48°
47°
46°
43°
-T — \ 1 ^-y
40°
38°
37°
36°
35°
32^
-l_
REGION
REGION
REGION
REGION
REGION 4
REGION 3
_i_
UNITED STATES
Bodega Head
San Francisco
REGION 2
REGION 1
Point Sur
\ i Point Conception
\?^pa
_i_
132°W 131° 130° 129° 128° 127° 126° 125° 124° 123° 122° 121° 120° 119° 118° 117°W
Figure 2.— Geographic regions used as strata in density estimation.
by adding all appropriate transect segments to a
linear array and dividing that array into 10 equal
parts.
When paired tests were not applicable, nonpara-
metric ANOVA models were used. For simple com-
parisons, the Kruskal-Wallis single-factor analysis
of variance was used (Kruskal and Wallis 1952). For
two-way comparisons, we used a two-factor exten-
sion of the Kruskal-Wallis test (Scheirer et al. 1976).
For both tests, three replicate measures of density
were created for each cell by randomly assigning
transect segments as replicate 1, 2, or 3.
437
FISHERY BULLETIN: VOL. 86, NO. 3
A principal assumption of strip transect methods
is that all individuals within the designated strip are
counted. Assuming that the fraction of diving
animals does not vary with sighting conditions, any
effect of sighting conditions on apparent density is
likely due to missed animals. In any instance where
we show that poorer sighting conditions result in
a significant decrease in apparent density (one-tailed
for paired tests, tw^o-tailed for ANOVA tests), we
eliminate the category of sighting conditions which
resulted in that lower estimate. If it is not possible
to predict which category would be worse a priori
in paired tests (e.g., right swath vs. left), two-tailed
probabilities are used. Although we cannot be sure
of eliminating all biases using these methods, this
pattern of data paring should avoid much of the bias
due to missed animals.
RESULTS
In 1984, 247 groups of harbor porpoise (680 in-
dividuals) were seen within transect strips which
covered a linear distance of 5,763 km. In 1985, we
saw 119 groups (384 individuals) in surveys of 3,715
km. Mean group sizes were 2.75 and 3.23 in-
dividuals, respectively, for 1984 and 1985. For 1984,
the relative frequencies of individuals seen within
the inside and outside swaths are illustrated in
Figure 1. For 1985, the perpendicular distances
from the trackline to the animals were calculated
from declination angles, and the relative distribu-
tion of sightings is shown in Figure 1 as a function
of perpendicular distance.
Inside vs. Outside Swath
For 1984 data, only the inside swaths were used,
but for 1985 both inside and outside swaths were
considered for density estimation. For 1985, we
tested whether the density in the inside swaths was
greater than the density of the outside swaths. We
only considered cases when the water surface pene-
tration codes were equal in both the inside and out-
side swaths. Data for inside and outside were thus
paired, with all other sighting factors equal. For
1985, the density in the inside was greater (0.09 vs.
0.06 porpoise/km^), but this difference was not
significant (P > 0.10).
Surface Penetration
Observers used a subjective coding system to
describe their ability to see through the sea surface.
Cloud cover, haze, and water turbidity contributed
to poor surface penetration. In 1984, observers used
codes to indicate good and poor conditions; in 1985,
observers used codes for excellent, good, and poor.
There were frequent cases when observers recorded
different codes for the inside swaths on opposite
sides of the plane; hence, paired tests were again
appropriate. For 1984, we tested whether density
in the "good" category was higher than density in
the "poor" category. When surface penetration was
different in the inside swaths on opposite sides of
the plane, mean density in the "good" category was
higher than in the "poor" category (0.16 vs. 0.12
porpoise/km^), but this difference was not signifi-
cant (P > 0.25). For 1985, no tests were necessary
because the mean density in "excellent" category
was lower than in the "good" category, and like-
wise, density in the "good" was lower than in the
"poor" category. All categories of water surface
penetration were included in subsequent analyses.
Effects Due to Observers and
Side of the Plane
Sightings were classified based on which observer
made the sighting and on whether the sighting was
on the inshore or offshore side of the aircraft. In
fact, these two classifications were confounded in
1985 because the two principal observers were sit-
uated on the same sides of the aircraft for most of
this survey. Effects of these classifications on den-
sity estimation were considered together. Survey
teams were defined as pairs of observers who
worked together. There were two such teams for
1984 and three for 1985 (Table 1). Only one of the
teams in 1985 had sufficient numbers of sightings
to be considered here. Statistical tests were based
on paired cases during which both members of the
sighting team were searching.
For 1984, porpoise density on the offshore side
of the airplanes was greater than on the inshore side
for both team A and team B (Table 2). The differ-
ence in density estimates between observers was
less than the difference between inshore and off-
Table 2.— Relative harbor porpoise densities (km"^) for teams
of observers. Density estimates are stratified by inshore and
offshore sides of the aircraft and by individual observers.
Observer
team
Inshore Offshore
Observer
1
Observer
2
A
B
D
0.275
0.168
0.281
0.425
0.244
0.164
0.320
0.232
0.300
0.380
0.179
0.146
438
BARLOW ET AL.: AERIAL SURVEYS OF HARBOR PORPOISE
shore (Table 2), and neither was statistically signif-
icant. For 1985, the opposite was seen; the density
on the inshore side was greater than on the offshore
side and this difference was less than the difference
between observers (Table 2), but again, neither dif-
ference was statistically significant. In 1984, both
observers had previous experience doing aerial
surveys, but in 1985, the observer with the lower
density estimates had no previous experience in
cetacean surveys. Experience may be a factor in
density estimates from strip transects, but since the
inexperienced observer was always on the same side
of the plane, it was not possible to test this with a
factorial design.
Area and Sea State
Analysis of research vessel data (Barlow 1988) in-
dicated two geographic regions in California with
low porpoise density (regions 1 and 3 in Figure 2).
We tested whether density observed from aircraft
are also lower in these areas. Because observation
conditions may have differed in the two areas, we
included sea state as a second factor in a two-way
ANOVA. Mean values are presented in Table 3. For
1984, three categories of sea state were used:
Beaufort 0 & 1, Beaufort 2, and Beaufort 3 and
greater. The effect due to area was significant (P
< 0.001), with the area that showed low density in
the ship surveys also showing lower density in the
aerial survey. For 1985, only the first two categories
of sea state were used due to insufficient data at
Beaufort 3 and greater. Again the effect of area was
significant (P < 0.025), and the same trends were
seen. In neither case were the effects of sea state
Table 3. — Uncorrected harbor porpoise densities (km' ) for
the two-way comparison of area and sea state. Low-density
areas refer to two regions in California that were found to have
much lower than average density in previous ship surveys (see
text). High-density areas include all other regions. Only inside
swaths were included. Numbers in parentheses refer to area
(km^) surveyed under the given condition. Densities in
brackets were excluded from two-way comparisons due to the
small area covered in one cell.
1984
survey
1985
survey
Beaufort
sea state
Low-
density
area
High-
density
area
Low-
density
area
High-
density
area
0 & 1
0.027
(183)
0.579
(264)
0.089
(124)
0.762
(126)
2
0.081
(308)
0.312
(1012)
0.089
(214)
0.183
(425)
3-1-
0.102
(79)
0.191
(252)
[0.000]
(28)
[0.049]
(566)
or the interaction effects significant. To eliminate
area effects from confounding statistical results,
only data for the larger, high-density area were in-
cluded in subsequent tests. The low-density area was
included in later estimates of overall harbor porpoise
density.
Sea State and Cloud Cover
Both sea state and cloud cover can affect sighting
conditions. Because both are affected by local
weather, the effects of these are likely to be con-
founded. These two factors were therefore tested
simultaneously in a two-way ANOVA. We con-
sidered only the inside swath and excluded the
two low-density regions. We used the same sea
state categories as above. The sky was categorized
as clear if cloud cover was <25% and cloudy if
>25%. Mean porpoise densities for each category
are given in Table 4. It was necessary to exclude
the Beaufort 0 & 1 category for 1985 because only
52 km were surveyed in cloudy conditions for these
sea states. The effect due to cloud cover was signifi-
cant for 1984 (P < 0.025) and 1985 (P < 0.05). The
effect of sea state and the interaction effect of
sea state and cloud cover were not significant for
either survey (P > 0.10). Transect segments with
>25% cloud cover were excluded from subsequent
analyses.
Table 4. — Uncorrected harbor porpoise densities (km"^)
for the two-way comparison of sea state and cloud cover.
Clear refers to <25% cloud cover, and cloudy refers to
>25%. Data include only inside swaths in high-density
areas. Numbers in parentheses refer to area (km^) sur-
veyed under the given condition. Densities in brackets were
excluded from two-way comparisons due to the small area
covered in one cell.
Beaufort
sea state
1984
survey
1985
survey
Clear
Cloudy
Clear
Cloudy
0& 1
1.340
(81)
0.240
(183)
[0.832]
(115)
[0.000]
(11)
2
0.371
(572)
0.236
(440)
0.271
(232)
0.078
(193)
3 +
0.266
(162)
0.056
(90)
0.088
(272)
0.014
(294)
Sea State
The effect of sea state was tested alone using only
the transect segments which occurred under clear
skies (<25% cloud cover) and within the high-density
areas. For 1984, only the inside swath was included.
439
FISHERY BULLETIN: VOL. 86, NO. 3
For 1985, both inside and outside swaths were in-
cluded. As in previous stratifications, apparent den-
sity clearly decreased with increasing sea state
(Table 5). This effect was significant for surveys in
both 1984 (P < 0.05) and 1985 (P < 0.005). Transect
segments surveyed during sea states of 2 or greater
were excluded in subsequent analyses.
•Uncorrected harbor porpoise den-
^) for the stratification based on sea
Table 5.-
sities (km
state. Data include only high-density areas sur-
veyed when cloud cover was <25%. Data for
1984 include inside swaths only; data for 1985
include inside and outside swaths. Numbers in
parentheses refer to area (km^) surveyed
under the given condition.
Beaufort
sea state
1984 survey
1985 survey
0 & 1
1.340
0.807
(81)
(234)
2
0.371
0.193
(572)
(471)
3 +
0.266
0.058
(162)
(553)
Between Survey Differences
We considered the 1984 survey and the 1985
surveys at 0.61 km and 1.85 km from shore as three
independent estimates of harbor porpoise density.
Because apparent density was shown to vary greatly
with sighting conditions and because sighting con-
ditions varied between surveys, it was necessary to
compare these three under similar conditions.
The highest (and presumably least biased) den-
sities were obtained when sea state was Beaufort
0 & 1 and when cloud cover was <25%. Between
survey comparisons imder these conditions are given
in Table 6 for the eight geographic regions given
in Figure 2. For 1985, there were no transect
segments at 0.61 km from shore under the condi-
tions Beaufort 0 & 1 and clear skies. For 1984, only
three regions contained more than 10 km- of
searching effort at 0.61 km from shore. For 1985,
only four areas had any searching effort at 1.85 km
from shore. The only direct density comparisons
with reasonable sample sizes are for region 1 (0.000
vs. 0.048 porpoise/km^) and region 3 (0.111 vs.
0.110 porpoise/km^) (respectively for 1984 and
1985). The densities for all regions pooled (0.671 and
0.510 porpoise/km^) are similar, but because of the
small sample size and geographic variation in sam-
pling, a statistical test of this difference is mean-
ingless.
In comparing surveys, sample size and regional
coverage improved slightly when Beaufort 2 was
considered (still allowing a maximum of 25% cloud
cover) (Table 7). For 1984, coverage was relatively
complete in all regions. For 1985, coverage at 0.61
km from shore was limited to regions 2-4, and
coverage at 1.85 km was limited to regions 1-5.
Comparing the two surveys in 1985, distance from
shore made little difference in overall density for
all regions combined, and neither survey had con-
sistently higher values than the other. Comparing
the 1984 survey to the two 1985 surveys, the former
had a higher overall density for all regions combined,
but again this difference was not consistent among
regions. Sample size and regional coverage were
again too poor for meaningful statistical tests.
Table 6.— Uncorrected harbor porpoise densities (km"^) in
eight geographic regions surveyed during Beaufort 0 & 1 con-
ditions. Data for 1984 are based on inside swaths of transects
flown 0.61 km from the coast. Data for 1985 are based on in-
side and outside swaths of transects flown at 0.61 and 1.85
km from the coast. Only those segments surveyed when cloud
cover was <25% are included. Numbers in parentheses refer
to area (km^) surveyed under the given condition.
Geographic
region
1984 survey
density
0.61 km
1985 survey
density
0.61 km
1985 survey
density
1.85 km
1
0.000
0.048
(43)
(0)
(103)
2
0.000
—
0.609
(1)
(0)
(51)
3
0.111
—
0.110
(45)
(0)
(55)
4
0.953
—
0.862
(6)
(0)
(183)
5
1.120
—
—
(9)
(0)
(0)
6
1.374
—
—
(64)
(0)
(0)
7
5.745
—
—
(1)
(0)
(0)
8
(0)
(0)
(0)
All regions
0.671
—
0.510
(170)
(0)
(392)
Ventilation Patterns
Harbor porpoise did not appear to react to the
helicopter during aerial observations; they were
visible throughout a surfacing series and were not
visible during dives. Knowing this, we were able to
use data on surfacing series and dive times to deter-
mine the fraction of time harbor porpoise would be
visible from the air. Data on ventilation patterns
440
BARLOW ET AL.: AERIAL SURVEYS OF HARBOR PORPOISE
were available from the helicopter study in Califor-
nia (13 groups were observed, mean group size was
2.7), the Oregon shore observations (11 groups,
mean group size was 4.2), and a previous shore study
in Glacier Bay, AK (28 solitary individuals). The
mean times spent in surfacing series were 34.2, 24.6,
and 30.0 seconds (respectively for the three studies).
The corresponding mean dive times did not differ
significantly between study sites (P > 0.05 using
pairwise ^tests). Using the pooled data set (n = 52),
the mean time spent in a surfacing series was 30.02
seconds (SE = 1.95) and the mean time spent in a
dive was 95.81 seconds (SE = 5.32). The percent-
age of time spent at the surface is 23.9%.
Average and Adjusted
Density Estimates
Harbor porpoise densities under optimal condi-
tions (Beaufort 0 & 1 and <25% cloud cover) were
averaged for the two surveys, weighting by transect
length (Table 8). Given an average survey speed of
160 km/h and assuming that the window for harbor
porpoise observation is 400 m long, the time win-
dow during which a point would be visible is 9.0
Table 7— Uncorrected harbor porpoise densities (km"^) in
eight geographic regions surveyed during Beaufort 2 condi-
tions. Data for 1984 are based on inside swaths of transects
flown 0.61 km from the coast. Data for 1985 are based on in-
side and outside swaths of transects flown at 0.61 and 1.85
km from the coast. Only those segments surveyed when cloud
cover was <25% are included. Numbers in parentheses refer
to area (km^) surveyed under the given condition.
Geographic
region
1984 survey
density
0.61 km
1985 survey
density
0.61 km
1985 survey
density
1.85 km
1
2
3
4
5
6
7
8
All regions
0.000
(107)
0.092
(109)
0.196
(117)
0.562
(180)
0.489
(119)
0.439
(57)
0.114
(70)
0.268
(37)
0.295
(796)
(0)
0.102
(59)
0.062
(48)
0.282
(110)
0.459
(6)
(0)
(0)
(0)
0.192
(224)
0.184
(103)
0.103
(68)
0.287
(52)
0.070
(57)
0.234
(171)
(0)
(0)
(0)
0.188
(452)
seconds. Using this estimate and the surface and
dive times estimated above, the probability that a
porpoise will be seen is estimated as 0.310 from
Equation (2). An instantaneous count would there-
fore underestimate porpoise abundance by a factor
of 3.2. Average values were therefore multiplied by
this factor (Table 8).
DISCUSSION
Results indicate that sighting conditions must be
very good in order to estimate harbor porpoise abun-
dance from aerial strip transects. Both sea state and
cloud cover had very large and significant effects
on apparent density. Limiting observations to the
best categories of sea state (Beaufort 0 & 1) and
cloud cover (<25%) can be used to minimize the bias
due to missed animals. These conditions are, how-
ever, rare and only occurred during 5.3% and 10.3%
of the transects in 1984 and 1985. The actual occur-
rence of these conditions is even more rare if one
considers flights that were cancelled due to bad
weather.
The effects on sea state and cloud cover on sight-
ing conditions were predicted by observers before
analysis of survey data was begun. Most harbor por-
poise were first seen when submerged a small
distance below the surface. Surfacings were rela-
tively inconspicuous to aerial observers and were not
an important cue in sighting porpoise. Both sea state
and cloud cover affect the ability of observers to see
through the water's surface and to spot submerged
animals. Increasing sea state causes more refrac-
tion of light at the water's surface, increases glit-
Table 8.— Harbor porpoise densities (km"^) in eight geographic
regions from a) a weighted average of uncorrected estimates from
1984 and 1985 aerial surveys, b) the same average adjusted by
a factor of 3.2 to account for submerged porpoise that were missed
by aerial observers, and c) a weighted average for the 1 984 and
1985 ship surveys. Aerial estimates are based on observations
made under conditions of Beaufort 0 & 1 and with <25% cloud
cover. Ship estimates include Beaufort sea states of 0, 1, and 2.
Geographic
region
Area
surveyed
(km*)
Aerial estimates
Uncorrected
a
Research
vessel
Corrected estimates
b c
1
146
0.03
2
1
—
3
100
0.11
4
189
0.86
5
9
1.12
6
64
1.37
7
1
—
8
0
—
All regions
562
0.55!
0.10
0.05
—
0.66
0.35
0.04
2.75
3.68
3.58
1.18
4.38
2.88
—
3.43
—
1.42
1.79
1.73
441
FISHERY BULLETIN: VOL. 86, NO. 3
tery reflection of sunlight, and causes whitecaps
which obscure subsurface observation. Cloud cover
decreases penetration of sunlight into the sea and
causes its surface to appear dark and glazed. It is
therefore not surprising that calm seas and clear
skies result in higher apparent densities.
It is more surprising that apparent density did not
vary with observer's subjective appraisal of surface
penetration. This may have been because surface
penetration was only tested in paired cases for which
sea state and cloud cover were identical. In these
cases, differences in surface penetration may have
been due primarily to subjective differences in the
way individual observers were coding it. If all cases
are considered, surface penetration is very highly
correlated with sea state and cloud cover and prob-
ably could be used as an alternative measure of
sighting conditions. We prefer, however, to use sea
state and cloud cover because their measure is less
subjective than surface penetration and could more
easily be used by other researchers.
Missed Animals
A principal assumption of strip transect methods
is that all individuals within the strip are counted.
We cannot necessarily meet this assumption just by
eliminating the categories of sighting conditions
with significantly lower density. In fact, it is possi-
ble that this method of selectively eliminating data
could overestimate density by eliminating a category
of sighting conditions which (by random chance) had
a significantly lower density. We do not believe that
this is likely in the cases of sea state or cloud cover
because the trends were the same for both surveys
and because the categories that were eliminated
were judged a priori as being poorer sighting con-
ditions. We believe that porpoise density is more
likely underestimated due to missed animals. Kraus
et al. (1983) found that observers in aircraft saw only
14% of the harbor porpoise groups known to be pres-
ent based on shore-based observations. Missed
animals may include some individuals that were near
the surface and visible but were not seen, as well
as others that were diving and were too deep to be
seen.
We infer that some near-surface animals were
missed based on three reasons. First, apparent den-
sity decreased with increasing sea state and cloud
cover, hence near-surface animals must be missed
in (at least) the poorer conditions. (An alternative
explanation is that porpoise spend less time near the
surface when sighting conditions are poor.) Second,
of the two principal observers in 1985, the less ex-
perienced observer may have missed more porpoise.
Third, in five instances in 1984 and one instance in
1985, the data recorder saw harbor porpoise in the
inside swath that were missed by the observer. In
three of these cases, conditions were Beaufort 1 with
<25% cloud cover. Recorders searched only occa-
sionally as conditions permitted, so it is not possi-
ble to use these data to quantify how many near-
surface animals were missed under good conditions.
Based on behavioral observations, we can also be
certain that some harbor porpoise are missed be-
cause they are too deep to be seen. Water visibility
was typically only 2-5 m during the surveys. Aerial
observers have seen harbor porpoise dive out of view
during the passing of the plane or while circling.
Harbor porpoise were not visible from the helicopter
during dives, even in very calm water. Some frac-
tion of porpoise must be missed because they are
too deep to be seen.
We have tried to account for the fraction of diving
animals by dividing density estimates by the frac-
tion of time harbor porpoise are known to be near
the surface (i.e., within surfacing series). Uncor-
rected estimates of harbor porpoise density are
based on the assumption that porpoise are never too
deep to be seen; this undoubtedly results in an
underestimate of porpoise density. Our method of
adjusted estimates of porpoise density assumes that,
when diving, porpoise are always too deep to be
seen; this has been corroborated by helicopter ob-
servations. The latter estimate should therefore be
closer to the true value of porpoise density (if biases
due to other factors have been eliminated).
Offshore Distribution of
Harbor Porpoise
In 1985, aerial transects were flown at 0.61 and
1.85 km from the shore. The latter value was chosen
to correspond approximately to the ship transects
along the 10-fathom (18.3 m) isobath. The intent was
to directly compare aerial estimates of density to
estimates made from ships at the same distance
from shore (considered below) and to provide a
means to extrapolate ship estimates to regions that
were too shallow to survey by ship. In regard to the
latter, we wish to know whether harbor porpoise
density at 0.61 km from shore is different from that
at 1.85 km.
In 1985, surveys at 0.61 km were never flown
under good sighting conditions (Beaufort 0 & 1 and
<25% cloud cover), so direct comparisons between
0.61 and 1.85 km are not possible. Considering only
the best category of sighting conditions, it is possi-
442
BARLOW ET AL.: AERIAL SURVEYS OF HARBOR PORPOISE
ble to compare 1985 transects at 1.85 km with 1984
transects at 0.61 km in geographic regions 1 and
3 (Table 6). Sample sizes are very small, but den-
sities are roughly comparable (0.070 vs. 0.057,
respectively for 1.85 and 0.61 km from shore). Sam-
ple sizes in 1985 can be increased if Beaufort 2 is
considered instead of Beaufort 0 & 1 (still with <25%
cloud cover). Based on geographic regions 1 through
5, porpoise densities are virtually identical at 1.85
and 0.61 km from shore (0.188 vs. 0.192, respective-
ly, Table 7).
In a preliminary model of harbor porpoise depth
distribution. Barlow (1988) hypothesized that har-
bor porpoise density is constant from the shore to
the 75 m isobath. Ship data did not, however, in-
clude any transects inshore of 18.3 m depth. In this
shallow area, the model was not based on any data.
Although sparse, data from aerial surveys show that
densities are similar in areas that were surveyed by
ship (1.85 km from shore) and in areas that were
too shallow to be surveyed by ship (0.61 km from
shore). These data are consistent with the model pro-
posed by Barlow.
Comparison of Ship
and Aerial Density Estimates
Estimates of harbor porpoise density from 1984
to 1985 ship surveys (Barlow 1988) can be compared
with adjusted and unadjusted estimates from the
aerial surveys (Table 8). The overall density for all
regions is higher for the ship surveys than for the
unadjusted aerial. Adjusted estimates from the
aerial surveys are very close to the overall estimates
from ship surveys.
Previous comparisons have been made of sighting
efficiency from aerial and surface vessel platforms
(Kraus et al. 1983). They found that observers on
boats saw 52% of the harbor porpoise groups seen
by shore-based observers, whereas aerial observers
saw only 14%. Based on this, density from aerial
surveys might be expected to be only 27% of that
from ship surveys. In the present study, unadjusted
density based on aerial surveys is 32% of the den-
sity from ship surveys. The two studies are not
directly comparable, however. Weather conditions
are not reported by Kraus et al. (1983) and may have
included less than optimal sighting conditions. It
should be noted that the fraction of harbor porpoise
groups seen by aerial observers in their study (14%)
is even lower than the fraction of harbor porpoise
we assumed would be in surfacing series and hence
near the surface (23.9%). It is possible that
behavioral differences between harbor porpoise
from the two coasts (such as travelling or behavior
mode "A" noted by Watson and Gaskin (1983)) could
account for some of the differences noted above.
Estimates based on ship surveys cannot, of course,
be considered the true density of harbor porpoise.
The overall estimate based on ship surveys is rela-
tively imprecise (C.V. = 49%) and may be biased
(Barlow 1988). The ship survey estimate is, however,
superior to current estimates from aircraft for sev-
eral reasons. Line transect methods were used on
the ship surveys, and the principal assumption of
this method (that 100% of the animals in the imme-
diate vicinity of the trackline are seen) is more easily
met than the comparable strip transect assumption
(that 100% of the animals within a strip are seen).
Acceptable sighting conditions for ship surveys in-
cluded Beaufort 0 & 1 , and 2 and were not restricted
by cloud cover (Barlow 1988). This allowed more
complete geographic coverage than did aerial
surveys. Also, the ship travelled much slower than
the aircraft (10 knots vs. 80-90 knots), thus the prob-
ability of missing a diving individual was much less.
Barlow (1988) calculated that diving animals located
near the trackline would be missed by observers on
ships only if dive times exceeded 2 minutes. Final-
ly, estimates of the correction factor to account for
submerged animals is relatively imprecise. Addi-
tional observations on ventilation patterns may
allow further refinements in density estimates based
on aerial surveys.
RECOMMENDATIONS
The design of future surveys for harbor porpoise
could be improved based on the results obtained
from our aerial surveys. We found that sighting con-
ditions deteriorated rapidly with both increasing
cloud cover and rougher sea states. To the extent
that is possible, aerial surveys for harbor porpoise
should only occur on clear days with little wind. Ob-
servations made by the data recorders indicate that
some harbor porpoise will be missed even in good
sighting conditions. If strip transects are used, ex-
periments with two independent teams of observers
searching at the same time could be used to quan-
tify the fraction of animals that are missed by using
just one team. Given that fewer harbor porpoise
were seen between 400 and 500 m of the track line,
we also suggest that, when surveying at 213 m
altitude, the strip widths should be decreased to only
include the area between 100 and 400 m.
We believe, however, that the problem of miss-
ing harbor porpoise could be reduced if line transect
methods were used in place of strip transects. Line
443
FISHERY BULLETIN: VOL. 86, NO. 3
transect methods assume that 100% of all animals
are seen directly along the trackline and use
statistical techniques to estimate the number of
animals that are missed as a function of the distance
from this tracWine. Line transects would require use
of an aircraft with unobstructed downward visibil-
ity through a belly window and use of a third ob-
server who could view animals directly under the
aircraft.
Harbor porpoise are now known to occur further
from the shoreline than was believed at the begin-
ning of this study (Barlow 1988). In future surveys,
transect lines should be placed so as to cover a
greater fraction of the harbor porpoise habitat.
ACKNOWLEDGMENTS
Support, personnel, and flight time were provided
by the Southwest Fisheries Center (NMFS), the
Southwest Regional Office (NMFS), the California
Department of Fish and Game, the National Marine
Mammal Laboratory (NMFS), the Washington State
Department of Wildlife, and the Oregon Department
of Fish and Wildlife. The observers and data
recorders were R. Brown, S. Diamond, D. Hanan,
R. Holt, S. Jeffries, W. Ferryman, and B. Troutman
(also Jay Barlow, Terry D. Jackson, and Charles W.
Oliver). The pilots were R. Anthes, J. Cain, L. Heitz,
J. York, and B. Zorich. Shore observations were
made by M. Herter, T. LoughHn, D. Merrick,
R. Rowlett, D. Rugh, and B. Yerman (also Barbara
L. Taylor). Unpublished porpoise data from Alaska
were collected by P. Dawson (and Barbara L.
Taylor). The method of adjusting for submerged
animals was suggested by D. Chapman based on an
unpublished report by R. A. Davis. Drafts of this
paper were reviewed by R. Brown, D. DeMaster,
K. Forney, D. Hanan, R. Holt, S. Jeffries, and two
anonymous reviewers.
LITERATURE CITED
Barlow, J.
1988. Harbor porpoise Phocoena phocoena, abundance estima-
tion for California, Oregon and Washington; \. Ship
surveys. Fish. Bull., U.S. 86:417-432.
Kraus, S. D., J. R. Gilbert, and J. H. Prescott.
1983. A comparison of aerial, shipboard, and land-based
survey methodology for the harbor porpoise, Phocoena pho-
coena. Fish. Bull., U.S. 81:910-913.
Kruskal, W. H., and W. a. Wallis.
1952. Use of ranks in one-criterion analysis of variance. J.
Am. Stat. Assoc. 47:583-621.
Scheirer, C. J., W. S. Ray, and N. Hare.
1976. The analysis of ranked data derived from completely
randomized factorial designs. Biometrics 32:429-434.
Seber, G. A. F.
1973. The estimation of animal abundance and related param-
eters. Hafner Press, N.Y., 506 p.
Taylor, B. L., and P. K. Dawson.
1984. Seasonal changes in density and behavior of harbor por-
poise (Phocoena phocoena) affecting census methodology in
Glacier Bay National Park, Alaska. Rep. int. Whaling
Comm. 34:479-483.
Watson, A. P., and D. E. Gaskin.
1983. Observations on the ventilation cycle of the harbour
porpoise Phocoena phocoena (L.) in the coastal waters of the
Bay of Fundy. Can. J. Zool. 61(1):126-132.
Wilcoxon, F.
1945. Individual comparisons by ranking methods. Biom.
Bull. 1:80-83.
444
EVALUATION OF VARIABILITY IN
SABLEFISH, ANOPLOPOMA FIMBRIA, ABUNDANCE INDICES IN
THE GULF OF ALASKA USING THE BOOTSTRAP METHOD
Michael F. Sigler and Jeffrey T. Fujioka^
ABSTRACT
Relative population numbers (RPN's) and length compositions were computed for sablefish, Anoplopoma
fimbria, in the Gulf of Alaska from the results of the Japan-United States cooperative longline survey
from 1979 to 1986. A statistical evaluation of annual changes in the RPN's using the bootstrap method
is demonstrated and showed that sablefish abundance increased significantly from 1979 to 1986, an in-
crease likely due to recruitment of two strong year classes. The effect of missing data on the bootstrap
calculations was examined and found to be negligible.
Early in this century, United States and Canadian
fishermen began harvesting sablefish, Anoplopoma
fimbria, in nearshore waters from California north
to southeastern Alaska, but sablefish were not
heavily exploited until Japanese longline vessels
began fishing in the Bering Sea in 1958. Japanese
catches off both the U.S. and Canadian coasts rose
dramatically in the following two decades. After
passage of the Fishery Conservation and Manage-
ment Act, foreign catches were reduced and the
domestic allocation was increased. The domestic
catch eventually increased nearly fivefold, thereby
replacing the foreign fishery, and in 1985, sablefish
in the Gulf of Alaska were harvested entirely by
domestic fishermen.
Before the reduction of the foreign fishery, infor-
mation on sablefish abundance consisted of statistics
on catch per unit effort from the Japanese longline
fishery. In 1978, the Fisheries Agency of Japan and
the U.S. National Marine Fisheries Service (NMFS)
began a cooperative longline survey along the con-
tinental slope of Alaska to assess the abundance of
sablefish and Pacific cod, Gadus macrocephalus. The
survey, conducted annually, has provided eight con-
secutive years (1979-86) of data for the Gulf of
Alaska, seven years (1980-86) of data for the Aleu-
tian region, and five years (1982-86) of data for the
eastern Bering Sea. The first year of the survey,
1978, was experimental.
Relative population numbers (RPN's) and length
compositions from the longline survey results from
'Northwest and Alaska Fisheries Center Auke Bay Laboratory,
National Marine Fisheries Service, NOAA, P.O. Box 210155, Auke
Bay, AK 99821.
1979 to 1985 have been estimated previously by
Sasaki (1986). In this study, the observed increase
in the RPN's for sablefish in the Gulf of Alaska is
evaluated statistically and an explanation of the
probable source of the increase is discussed.
Statistical analysis of the survey results is based
on the bootstrap method (Efron 1982; Efron and
Gong 1983). This method is a relatively new statis-
tical procedure that has been little used in fisheries
analysis. Thus, this paper also demonstrates an
application of the bootstrap method to statistical
evaluation of fishery survey data.
SURVEY METHODS
The Gulf of Alaska portion of the Japan-U.S.
cooperative longline survey, conducted annually
each summer from 1979 to 1986, covered five
International North Pacific Fisheries Commission
(INPFC) statistical areas: Shumagin, Chirikof,
Kodiak, Yakutat, and Southeastern (Fig. 1). One of
a total of 47 stations each ranging in depth from
about 100 to 1,000 m was sampled daily by longline.
The longline was 16 km long and consisted of 160
hachis (the Japanese word for "skate" or length of
longline); each hachi was 100 m long and consisted
of 45 hooks baited with squid. Soak time, the time
between setting and retrieval, varied from 3 hours
at the beginning of the longline gear to 7 or 8 hours
at its end. The depth at which the fish were caught
was estimated by measuring the depth of water
under the vessel with an echo sounder every fifth
hachi. The fish caught were tallied by species and
hachi as the longline was brought aboard, then they
were weighed and their length was measured. Most
Manuscript accepted May 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
445
FISHERY BULLETIN: VOL. 86, NO. 3
Figure 1.— International North Pacific Fisheries Commission statistical areas sampled during the Gulf of Alaska
portion of the Japan-U.S. cooperative longline survey, 1979-86.
sablefish were sexed, some were tagged and re-
leased, and others were sampled for otoliths and
scales. Detailed survey methods are described in
Sasaki et al. (1983).
STATISTICAL METHODS
The catch data were stratified because of differ-
ences in the catch rate by depth. Assignment of the
150 -
U
z
o
lU
c
100-
0.5
5.5
10.5 15.5
CATCH PER HACHI
20.5
25.5
446
Figure 2.— Frequency distribution of sablefish catch per hachi (numbers) by year, station, and strata,
1979-86.
SIGLER and FUJIOKA: VARIABILITY IN SABLEFISH
catch data to a stratum was based on the recorded
depth of every fifth hachi or the interpolated depth
of each of the intervening four hachis. The number
of strata totaled nine, with the first stratum repre-
senting 101-200 m, the second representing 201-
300 m, and so on. A catch per hachi value was
calculated for each stratum of a station (Fig. 2), and
the resultant values within each statistical area were
averaged.
Not all strata were sampled at each station and
as a result, sometimes no stations were sampled
within certain strata (Table 1). Although not sam-
pled one year, these strata generally were sampled
the previous year, because of slight annual differ-
ences in the depths sampled at each station. Catch
per hachi values CPUE,j were estimated for sta-
tions within the unsampled strata from the catch per
hachi values for sampled stations from the previous
year and the annual change in the catch per hachi
values in the adjacent stratum. For each station:
CPUE, J = CPUE,_i J * CPUE ,,j /CPUEi_i__,.
where i = yed.r,j = stratum, a.ndj' = the adjacent
stratum. Catch per hachi values were estimated only
for stations within unsampled strata. If one or more
stations within a stratum were sampled, then no
catch per hachi values were estimated for the un-
sampled stations. The effect of these missing values
on the survey results vdll be discussed later.
Relative population numbers were calculated from
the catch per hachi values to index annual changes
in sablefish abundance. As in Sasaki (1985), the aver-
age catch per hachi was multiplied by the areal size
of each stratum and statistical area to calculate an
RPN for each stratum and statistical area. The areal
sizes used to calculate the RPN's are for the con-
tinental slope only (Table 2), which corresponds to
the area covered by the survey. The resultant RPN's
were summed across strata to calculate an RPN for
each statistical area, and these RPN's were summed
Table 1.— Number of stations sampled by year, INPFC area, and depth during the Japan-U.S. cooperative longline survey, 1979-86.
Depth
INPFC Statistical area
Depth
INPFC statistical area
Shu-
Chiri-
South-
Shu-
Chiri-
South-
Year
(m)
magin
kof
Kodiak
Yakutat
east
Year
(m)
magin
kof
Kodiak
Yakutat
east
1979
101-200
7
4
6
6
2
1983
101-200
9
5
7
8
3
201-300
7
5
7
10
5
201-300
10
6
9
11
9
301-400
7
5
8
6
301-400
10
5
9
11
8
401-500
7
5
9
6
401-500
10
6
9
11
8
501-600
6
4
9
6
501-600
10
5
9
11
8
601-700
6
4
8
5
601-700
9
5
9
11
7
701-800
6
4
7
5
701-800
8
4
8
11
7
801-900
5
1
4
9
4
801-900
5
4
4
10
7
901-1,000
4
1
2
4
1
901-1,000
2
0
1
5
2
1980
101-200
9
5
8
7
4
1984
101-200
10
7
7
8
2
201-300
10
6
8
10
6
201-300
10
7
9
11
9
301-400
10
6
8
10
8
301-400
10
6
9
11
8
401-500
10
6
9
10
8
401-500
10
6
9
11
8
501-600
10
5
9
10
7
501-600
10
5
9
11
8
601-700
9
5
9
9
7
601-700
8
5
8
11
8
701-800
6
4
8
9
7
701-800
7
5
6
10
8
801-900
6
2
3
7
4
801-900
3
0
3
7
5
901-1,000
5
1
2
6
1
901-1,000
0
0
0
6
1
1981
101-200
9
6
7
8
1
1985
101-200
9
6
6
7
1
201-300
10
6
9
11
9
201-300
10
7
8
11
9
301-400
10
6
9
11
9
301-400
10
6
9
11
8
401-500
10
6
9
11
9
401-500
10
6
9
11
7
501-600
10
6
9
11
9
501-600
10
6
9
11
8
601-700
9
4
9
11
9
601-700
9
5
9
11
8
701-800
7
2
7
11
9
701-800
7
4
9
10
8
801-900
6
1
5
7
7
801-900
4
4
4
11
6
901-1,000
2
0
1
2
1
901-1,000
0
0
0
5
2
1982
101-200
9
5
7
10
4
1986
101-200
9
5
7
7
2
201-300
9
6
9
11
9
201-300
10
6
9
11
9
301-400
10
5
9
11
8
301-400
10
6
9
11
8
401-500
10
6
9
11
8
401-500
10
6
9
11
8
501-600
10
5
9
11
8
501-600
10
6
9
11
8
601-700
10
5
9
11
8
601-700
10
4
9
11
8
701-800
9
4
8
11
7
701-800
10
3
8
11
8
801-900
7
4
6
9
6
801-900
5
2
7
11
8
901-1,000
3
3
5
8
3
901-1,000
1
1
3
5
1
447
FISHERY BULLETIN: VOL. 86, NO. 3
Table 2.— Slope area sizes (km^.
2\1
Depth
(m)
Slope area (km'')
Shumagin Chlrikof Kodiak Yakutat Southeast
201-400
4,001
2,350
3,106
2,988
1,781
401-
-600
2,269
1,766
2,255
1,666
822
601-
-800
1,629
1,955
1,923
1,470
1,006
801-
-1,000
1,248
2,012
2,296
1,489
1,165
201-
-1.000
9,147
8,083
9,580
7,613
4,774
'Shumagin, Chirikof, and Kodiak areas and Yakutat area from long.
147-I440W; data from E. Brown (NWAFC Seattle Laboratory, NMFS. NOAA,
7600 Sand Point Way NE. Seattle, WA 98115, pers. commun. December
1985). Yakutat area from long. 144-137°W. and Souttieastern area; data from
R. Haight (NWAFC Auke Bay Laboratory, NMFS, NOAA, P.O. Box 210155,
Auke Bay, AK 99821, pers. commun. 1986).
across statistical areas to calculate an RPN for the
Gulf of Alaska. The bootstrap method (Efron 1982;
Efron and Gong 1983) then was applied to the re-
sultant RPN's to test the statistical significance of
annual changes in the RPN's.
The bootstrap method is a nonparametric statis-
tical procedure based on Monte Carlo methods (see
Shreider [1966] for a description of Monte Carlo
methods). The bootstrap method is a new technique
not common in the fisheries and ecology literature,
but examples of its application to survey design and
biomass estimation can be found in Kimura and
Balsiger (1985) and Haslett and Wear (1985), re-
spectively. In addition, Rao and Wu (1984) proved
the applicability of the bootstrap method to strati-
fied sampling, which is the sampling method used
in the longline survey. The bootstrap method is
useful when parametric assumptions are difficult to
justify; no parametric estimate is readily available
for the accuracy of a statistic, e.g., a sample median;
or the procedure to compute the statistic of interest
is complicated. Simply described, the bootstrap
method works as follows: Given the observed data
set <Xi,X2,. . .,X^>, the sample <Xi*,X2*,. . .,
Xn*> is drawn by independent random sampling
with replacement from the observed data set and
the desired statistic (e.g., a median) is computed
from the sample. The resultant statistic is termed
the bootstrap replicate. In the next step, the sam-
ple is drawn and the bootstrap replicate is computed
some large number B times. The resultant B boot-
strap replicates form the bootstrap distribution. An
estimate of the accuracy of the median, a standard
error, then can be calculated from the bootstrap
distribution by standard methods.
In this study, we used the bootstrap method to test
the null hypothesis that the difference RPN^j;. -
RPN^jr. = 0, for any RPNjjt , where i = year
(1979-86), i' = any subsequent year, and k = sta-
tistical area (Shumagin, Chirikof, Kodiak, Yakutat,
and Southeastern). The bootstrap method was used
because parametric assumptions are difficult to
justify for the longline survey data and the pro-
cedure to compute the statistic of interest, the
variance of the RPN, is tedious and error prone. In
our application of the bootstrap method, stations
were randomly sampled with replacement within
each area. A value denoted RPN^ ;^* was computed
from the catch per hachi values of the sampled sta-
tions by the method of RPN calculation described
previously. Stations then were sampled with re-
placement from year i' within each area, a second
value denoted RPNj ;^* was computed, and the
difference RPNj j^* - RPN^j^* was found. Sampl-
ing with replacement from the 2 years and the com-
putation of the difference were repeated 1,000 times
producing a bootstrap distribution of 1,000
differences.
Efron and Tibshirani (1986) outlined three
methods for setting an approximate confidence
interval from a bootstrap distribution for a statistic
of interest, here the difference RPNj /^ - RPN^ /£.
Use of the simplest method, the percentile method,
is considered correct if the bootstrap distribution of
the statistic of interest is described by a normal
distribution (Efron and Tibshirani 1986). The nor-
mality of the bootstrap distribution for the dif-
ference was tested using the D'Agostino D Test
(D' Agostino and Stephens 1986) and found to be nor-
mal, thus justifying the use of the percentile method.
The statistical significance of the difference
RPNj /£ - RPNj /£ then was evaluated by the
following criteria. If the 95% confidence interval for
the difference did not include zero, then the null
hypothesis was rejected, the annual change in the
RPN was considered statistically significant, and the
change in sablefish abundance was considered real.
Conversely, if the 95% confidence interval for the
difference included zero, the null hypothesis was ac-
cepted and the change in sablefish abundance was
not considered significant.
RESULTS
The RPN for the Gulf of Alaska increased 111%
from 1979 to 1986 (Fig. 3). The 95% confidence
interval for this increase did not include zero and
therefore was judged statistically significant (alpha
= 0.05; Table 3), showing that the difference was
not due to random error in the survey and that
sablefish abundance in the Gulf of Alaska has in-
creased markedly since 1979. The difference con-
sists of significant increases from 1980 to 1981, 1981
448
- KODIAK
A.
-
/
/
-
/
/
I 1
1
1 1 1
30
1979 80
81 82 83 84
85 86 1979 80 81 82 83
YEAR
84 85 86
Figure 3.— Relative population number (RPN) for the Gulf of Alaska and the Shumagin, Chirikof, Kodiak, Yakutat,
and Southeastern areas, 1979-86. Dashed lines (— -) signify that the annual change was statistically significant.
to 1982, and 1984 to 1985 (Fig. 3). Differences be-
tween other years were not significant.
The RPN for each of the statistical areas of the
Gulf of Alaska generally increased from 1979 to
1986 (Fig. 3); the differences between 1979 and 1986
were statistically significant for all areas (Table 3),
showing that sablefish abundance has increased
throughout the Gulf of Alaska. The sharp jump in
the RPN for the Gulf of Alaska from 1980 to 1982
was caused by significant increases in four areas,
Shumagin, Chirikof, Kodiak, and Yakutat. The
sharp jump from 1984 to 1985 was caused by signif-
icant increases in two areas, Kodiak and Yakutat.
DISCUSSION
As noted earlier, sometimes not all strata were
sampled at a station (Table 1). This shortcoming was
a consequence of the length of the sampling gear
and the topography at a station. At stations where
the bottom gradient was slight for all or part of the
station, the 16 km of longline gear sometimes was
not long enough to sample all strata. As a result,
sometimes no stations were sampled within certain
strata. For these strata, catch per hachi values were
estimated using combined data from the current and
previous year's data. The resultant values are esti-
449
FISHERY BULLETIN: VOL. 86, NO. 3
Table 3.— Statistical significance of annual changes in
relative population number of sablefish, Gulf of Alaska,
Japan-U.S. cooperative longline survey, 1979-86. The sym-
bols used are defined as follows; + signifies a significant
increase in RPN; - signifies a significant decrease in
RPN; O signifies no significant change.
Area
Year
80
81
82
83
84
85
86
Gulf of Alaska
1979
O
+
+
+
+
+
+
1980
+
+
+
+
+
+
1981
+
+
+
+
+
1982
O
o
+
+
1983
o
+
+
1984
+
+
1985
o
Shumagin
1979
O
o
+
+
+
+
+
1980
o
+
+
+
+
+
1981
+
+
+
+
+
1982
o
o
+
+
1983
o
o
o
1984
o
o
1985
o
Chirikof
1979
o
o
+
+
+
+
+
1980
o
+
+
+
+
+
1981
+
+
+
+
+
1982
o
o
o
+
1983
o
o
+
1984
o
o
1985
o
Kodiak
1979
-
o
+
+
+
+
+
1980
o
+
+
+
+
+
1981
+
+
+
+
+
1982
o
o
+
+
1983
o
+
+
1984
+
o
1985
o
Yakutat
1979
o
+
+
+
+
+
+
1980
+
+
o
+
+
+
1981
+
o
o
+
+
1982
-
-
o
+
1983
o
+
+
1984
+
+
1985
+
Southeastern
1979
o
+
+
o
+
+
+
1980
o
o
o
o
+
+
1981
o
o
o
o
+
1982
o
o
o
+
1983
o
o
+
1984
o
+
1985
o
mates of catch per hachi as if the strata had been
sampled. Because the estimated values mimic the
observed data, the probability of rejecting the null
hypothesis is somewhat higher than the nominal
value. In general an inaccurate estimate should have
a small total effect because catch per hachi values
were missing for only 9 of the 320 strata available
during the survey (8 years x 5 areas x 9 strata).
The only comparison where the estimated catch per
hachi values might have caused an incorrect conclu-
sion is the 1984-85 comparison where 7 strata were
unsampled. This comparison was retested using only
the observed catch per hachi values. The annual
change, like the comparison using the estimated
catch per hachi values, was found to be statistically
significant.
Another potential effect of incomplete sampling
at a station is bias in the bootstrap calculation. Each
bootstrap replicate RPN^j^* was computed from a
bootstrap sample selected by sampling with replace-
ment from the stations within an area. If one or
more stations were missing an observation for a
strata, then it was possible for the bootstrap sam-
ple to have all missing values for the strata. In this
case, no catch per hachi values were available to
calculate a value for the strata and the value for the
strata was treated as zero. This treatment may
negatively bias the resultant RPN, ;,*. The extent
of the negative bias was tested by comparing the
RPN to a bootstrap estimate of the RPN, denoted
RPNfc, where RPN^ is the mean of the 1,000
RPNijt* from an area and year. If RPN^ generally
was smaller than RPN, then the bootstrap replicates
were negatively biased. The percentage difference
between the resultant RPNj's compared to the
RPN's ranged from - 0.7 to + 0.5 and averaged only
-1-0.03, showing that the negative bias had little
effect on the RPN,^*.
Other fishery and survey data substantiate the
significance of the overall increase in the RPN and
also the marked increases from 1981 to 1982 and
from 1984 to 1985. The CPUE in the Japanese long-
line fishery showed a similar pattern to the RPN's
for the years the data overlapped, from 1979 to
1983; the fishery CPUE increased from the late
1970's, with the largest increase from 1981 to 1982,
and decreased from 1982 to 1983 (Fig. 4).
Examination of length compositions for depths
101-200 m indicate that the strong 1977 year class
(Sasaki 1982; McFarlane and Beamish 1983; Funk
and Bracken 1984) was responsible for the RPN in-
crease from 1980 to 1982 and that the RPN increase
from 1984 to 1985 was due to a strong 1980 year
class. Sablefish recruiting to the survey area first
appear at depths 101-200 m, and strong year classes
are more distinguishable in the length compositions
at these depths. The length compositions for depths
101-200 m show the initial appearance and subse-
quent increase in length of fish of the strong 1977
year class (Fig. 5). The mode at 47 cm FL in 1979
indicates the first year that the 1977 year class was
available to the survey gear. The rightward move-
ment of the mode in succeeding years illustrates the
increase in fish length for the 1977 year class. The
movement of the mode slowed in 1982 which we in-
terpret to be due mainly to the movement of larger
450
SIGLER and FUJIOKA: VARIABILITY IN SABLEFISH
460
420
■o 380
a
(0
§340
JC
300
260
220
Q.
180
- \
\
FISHERY CPUE /
v..
J
1977 78
Figure 4.— Sablefish catch per hachi (kg/10 hachis) for the Japanese longline fishery, 1977-83 (Fujioka
1986) and relative population number (RPN), 1979-86.
Figure 5.— Length-frequency distributions of sablefish
in the Gulf of Alaska for depths 101-200 m, 1979-86.
The shaded bands illustrate the progression of the modes
of two strong year classes, 1977 and 1980.
T r
55 63
LENGTH (cm)
r
71
T
79
451
FISHERY BULLETIN: VOL. 86, NO. 3
fish to depths >200 m. The timing of this movement
corresponds to the year that the RPN significantly
increased.
The pattern of length compositions from 1983 to
1985 parallels those due to the strong 1977 year
class and suggests that the 1980 year class is also
strong (Fig. 5). The shoulder at 49-51 cm FL in
1983, not present in 1982, indicates the first year
that the 1980 year class was available to the survey
gear. This shoulder is due to distinct modes at 47-51
cm FL in the Chirikof, Yakutat, and Southeastern
areas. The rightward movement of the mode in suc-
ceeding years illustrates the increase in fish length
for the 1980 year class. The mode at 53 cm FL in
1984 is similar to the modes at 50-57 cm FL found
in the Gulf of Alaska for depths 101-200 m during
a trawl survey conducted in 1984 by the Northwest
and Alaska Fisheries Center (Brown 1986) and is
further evidence for a strong 1980 year class. The
rightward movement of the mode slowed in 1985,
presumably due to movement of larger fish to depths
>200 m, and again corresponds to the year of a
significant increase in RPN.
ACKNOWLEDGMENTS
Thanks to Evan Haynes and Leslie Williams for
thoroughly editing the manuscript, Jerry Fella for
statistical advice, and Takashi Sasaki for courteously
providing cooperative survey data.
LITERATURE CITED
Brown, E.
1986. Preliminary results of the 1984 U.S. -Japan cooperative
bottom trawl survey of the central and western Gulf of
Alaska. In R. L. Major (editor), Condition of groundfish
resources of the Gulf of Alaska region as assessed in 1984,
p. 259-296. U.S. Dep. Commer., NOAA Tech. Memo.
NMFS F/NWC-106.
D'Agostino, R. B., and M. A. Stephens.
1986. Goodness-of-fit techniques. Marcel Dekker, N.Y.
Efron, B.
1982. The jackknLfe, the bootstrap and other resampling
plans. Appl. Math. 38, 92 p. Soc. Ind. Appl. Math., Phila.
Efron, B., and G. Gong.
1983. A leisurely look at the bootstrap, the jackknife, and
cross-validation. Am. Stat. 37:36-48.
Efron, B., and R. Tibshirani.
1986. Bootstrap methods for standard errors, confidence in-
tervals, and other measures of statistical accuracy. Stat.
Sci. 1:54-77.
FUJIOKA, J.
1986. Sablefish. In R. L. Major (editor). Condition of
groundfish resources of the Gulf of Alaska region as assessed
in 1985, p. 79-108. U.S. Dep. Commer., NOAA Tech.
Memo. NMFS F/NWC-106.
Funk, F., and B. E. Bracken.
1984. Status of the Gulf of Alaska sablefish (Ancyplopoma fim-
bria) resource in 1983. Alaska Dep. Fish Game, Info. Leafl.
235, 55 p.
Haslett, S. J., AND R. G. Wear.
1985. Biomass estimation of Artemia at Lake Grassmere,
Marlborough, New Zealand. Aust. J. Mar. Freshwater Res.
36:537-557.
KiMURA, D. K., AND J. W. BALSIGER.
1985. Bootstrap methods for evaluating sablefish pot index
surveys. N. Am. J. Fish. Manage. 5:47-56.
McFarlane, G. a., AND R. J. Beamish.
1983. Overview of the fishery and management strategy for
sablefish (Ancrpkrpoma fimbria) off the west coast of Canada.
In Proceedings of the International Sablefish Symposium,
p. 13-35. Alaska Sea Grant Rep. 83-8, Univ. Alaska,
Fairbanks.
Rao, J. N. K., and C. F. J. Wu.
1984. Bootstrap inference for sample surveys, /n J. E.
Gentle (editor). Proceedings of the section on survey
research methods, p. 106-112. Am. Stat. Assoc.
Sasaki, T.
1982. Condition of sablefish stocks in the North Pacific. Far
Seas Fish. Lab. Fish. Agency Jpn., Shimizu, 17 p. (Document
submitted to the International North Pacific Fisheries
Commission.)
1985. Studies on the sablefish resources in the North Pacific
Ocean. Far Seas Fish. Res. Lab., Bull. 22, 107 p. Fisheries
Agency of Japan, Shimizu.
1986. Stock assessment of sablefish in the eastern Bering
Sea, Aleutian Islands region, and the Gulf of Alaska. Far
Seas Fish. Res. Lab., Fish. Agency Jpn., Shimizu, 33 p.
Sasaki, T., D. Rodman, and K. Funato.
1983. Preliminary report on Japan-U.S. joint longline survey
by Ryusho maru No. 15 in the eastern Bering Sea, Aleutian
region and Gulf of Alaska, 1982. Far Seas Fish. Res. Lab.,
Fish. Agency Jpn., Shimizu, 116 p.
Shreider, Y. a.
1966. The Monte Carlo method. Pergammon Press, N.Y.
452
PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD, PSEUDOPENTACEROS
WHEELERI, AND RECRUITMENT TO ISOLATED SEAMOUNTS
IN THE NORTH PACIFIC OCEAN
George W. Boehlerti and Takashi Sasaki^
ABSTRACT
The pelagic armorhead, Pseudopentaceros wheeleri, occurs widely in the North Pacific Ocean. Benthic
specimens have been taken from Japan, the Hawaiian Archipelago, and the west coast of North America,
but the main reproductive populations are located on southern Emperor-northern Hawaiian Ridge sea-
mounts between lat. 29° and 35°N. The period between spawning and recruitment to the seamounts
is apparently between 1.5 and 2.5 years, suggesting an extended pelagic existence. We describe the
distributional patterns in the North Pacific based upon over 30 years of published and unpublished records.
The majority of pelagic specimens are captured in the subarctic water mass in the northeast Pacific.
Based upon the distributional patterns and the oceanography of the North Pacific, we propose migratory
paths for both the main population and for the individuals that occur rarely in other locations. The long
pelagic period and variability in ocean conditions may play an important role in recruitment to seamounts
and the variability in year-class strength for this species.
The pelagic armorhead, Pseudopentaceros wheeleri,
is a member of the boarfish family Pentacerotidae.
Until recently, considerable confusion existed about
its taxonomy and distribution. Originally described
as Pentaceros richardsoni Smith 1844, it was
thought to be distributed virtually worldwide, with
centers of abundance in the North Pacific and in the
South Atlantic near South Africa (Fujii 1986). Oc-
currences were typically sporadic; most records
were from pelagic captures, hence the name pelagic
armorhead. Several authors, including Welander et
al. (1957), Clemens and Wilby (1961), and Wagner
and Bond (1961), used the name Pseudopentaceros
richardsoni, but more recent studies used the
original binomial following Follett and Dempster
(1963).
The 1967 discovery of large concentrations of
armorhead in the mid-Pacific, over southern
Emperor-northern Hawaiian Ridge (SE-NHR)
seamounts (from lat. 29° to 35°N) (Sasaki 1974;
Uchida and Tagami 1984) stimulated increased
interest in this species. Over the next 10 years,
nearly 1 million metric tons (t) were taken by Soviet
and Japanese trawlers (Borets 1975; Takahashi and
Sasaki 1977). The increased availability of specimens
allowed Hardy (1983) to revise the family. He first
'Southwest Fisheries Center Honolulu Laboratory, National
Marine Fisheries Service, NOAA, 2570 Dole Street, Honolulu, HI
96822-2396.
^Far Seas Fisheries Research Laboratory, 5-7-1 Orido, Shimizu,
Shizuoka 424, Japan.
separated the Pentaceros richardsoni complex
from other members of the genus by referring it to
Pseudopentaceros Bleeker, based on morphological
considerations. Second, he separated the complex
into three species: P. richardsoni confined to the
Southern Hemisphere, and P. wheeleri and P.
pectoralis in the North Pacific. The features used
to distinguish P. wheeleri and P. pectoralis were
largely morphological. Pseudopentaceros pectoralis
is more robust and deeper bodied and its dis-
tribution is typically pelagic; P. wheeleri was
known only from benthic specimens on sea-
mounts. Hardy (1983), however, lacked transitional
specimens; subsequent work based on morpho-
logical and electrophoretic grounds revealed that
P. wheeleri was a morphological derivative of
P. pectoralis, with changes occurring after settle-
ment. The deeper bodied pelagic form ceases growth
in length after recruiting to the seamounts and
instead transforms body shape to a leaner form.
"Fat" and "lean" specimens of the same length thus
differ significantly in appearance (Humphreys et al.
in press).
Pseudopentaceros wheeleri has an unusual life
history. Spawning occurs at the benthic population
centers at the SE-NHR seamounts between Novem-
ber and March (Sasaki 1974; Borets 1976; Bilim et
al. 1978). Larvae and juveniles are pelagic (Honma
and Mizusawa 1969; Fedosova and Komrokov 1975;
Borets and Sokolovsky 1978) and widely distributed
in the North Pacific. Recruitment to the seamounts
Manuscript accepted April 1988.
Fishery Bulletin; Vol. 86, No. 3, 1988.
453
FISHERY BULLETIN: VOL. 86, NO. 3
occurs between ages 1.5 and 2.5 years, but in the
major spawning population centers, over 95% of the
fish are <2 years (Uchiyama and Sampaga^). This
lack of older specimens and presence of fish in poor
physiological condition has led to suggestions that
this species is semelparous, with death after a single
spawning (Humphreys and Tagami 1986; Hum-
phreys et al. in press). The long interval between
spawning and recruitment for this species suggests
that oceanographic conditions in the North Pacific
regulate its distribution (Boehlert 1986).
Published records of the pelagic distribution of
P. wheeleri are too few and varied to understand its
biogeography. In different years, specimens have
been caught in the Gulf of Alaska, near the Aleu-
tian Islands, and off the coasts of Japan, Oregon,
California, and British Columbia. In this paper, we
document the pelagic and benthic occurrences from
'Uchiyama, J. H., and J. Sampaga. In review. Age and growth
of the pelagic armorhead, Pseudopentaceros wheeleri, at Hancock
Seamounts. Southwest Fisheries Center Honolulu Laboratory,
National Marine Fisheries Service, NOAA, 2570 Dole Street,
Honolulu, HI 96822-2396.
both published and unpublished accounts, consider
this distributional pattern in light of general ocean-
ographic conditions in the North Pacific, propose
hypotheses about the migration and recruitment of
armorhead to the seamounts of the central North
Pacific, and assess interannual variation in abun-
dance and how it may relate to oceanographic
variability.
MATERIALS
The data sources used include published and un-
published collection records from throughout the
North Pacific Ocean. Published records on occur-
rences in Japan (Abe 1957, 1969; Zama et al. 1977;
Okamura et al. 1982), the eastern North Pacific
(Welander et al. 1957; Larkins 1964; Honma and
Mizusawa 1969; Chikuni 1970; Ignell et al. 1986),
and the west coast of North America (Wagner and
Bond 1961; Follett and Dempster 1963; Smith 1965)
were typically of single or a few specimens. Larger
numbers of unpublished collection records were in
data from large-scale surveys (Table 1). These data
Table 1 .—Data bases searched for pelagic armorhead, Pseudopentaceros wheeleri, captures excluding single
collections and literature reports. Areas and seasons are averages over the years covered.
Years, season effective
Area
Data source
Lat. N
Long.
Salmon drift gill-net surveys^
1972-86, spring-fall
38°-65<'
140°E-166°W
University of Washington (FRI)
purse seining, longlining^
Canada salmon longlining^
1956-82, spring-summer
42°-60°
165°E-125°W
1962-67, spring-summer
42°-54°
120°-145°W
Bureau of Commercial Fisheries
salmon gill net^'^
1955-71, spring-summer
40°-60°
165'»E-125°W
Oshoro Maru salmon
1963-66, 1968, 1969, 1971,
gill-net-longline surveys'*
1974-86, spring-summer
36°-66°
163°E-124°W
Hokusei Maru squid and salmon
gill-net surveys'*
1973-86, summer
18°-58°
141°E-156°W
Japan Marine Fishery Resource
Research Center pomfret surveys*
1979-82, all months
22°-49°
143°E-124°W
Japanese commercial whaling reports*
1952-79, summer
40°-58°
Pacific Biological Station
1983, 1985-86,
gill-net surveys'^
spring-summer
47°-54°
138«'-130<'W
Taiwan Fisheries Research Institute
squid surveys®
1986, July-August
38°-46°
153°E-175°W
National Marine Fisheries Service
Auke Bay cooperative
squid gill-net surveys^
1985-86
43°-55°
145°-175°W
Japan Fisheries Agency
squid gill-net surveys
1986-87, July-August
36°-47°
150°-165°W
'Salmon research data file. North Pacific Ocean, 1972-86. Salmon Division, Far Seas Fisheries Research Latxjratory, Shimizu,
Japan.
2t^acy et al 1978
sLarklns 1964.
♦Data record of oceanographic observations and exploratory fishing. No. 1 (1957)-No. 28 (1985) Faculty of Fisheries, Hokkaido
University.
sjapan Marine Fishery Resourse Research Center 1980, 1983a, 1983b, 1985
^Biological data file of whales In the North Pacific Ocean, 1952-79. Whale Section, Far Seas Fisheries Research Laboratory,
Shimizu. Japan. See also ChikunI (1970).
'Sloan 1983; Robinson and Jamieson 1984.
'Z. Shyu, Taiwan Fisheries Research Institute, Keelung 20220, Taiwan, pers. commun. May 1987.
^Ignell et al. 1986.
454
BOEHLERT and SASAKI: PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD
are seasonally limited to primarily spring, summer,
and early fall months, but cover 1952-87 and pro-
vide broad geographic coverage between lat. 18°
and 66 °N. Greater detail and a summary of many
of these surveys are provided by Macy et al. (1978).
Since armorhead were an incidental catch, it is dif-
ficult to assess the relative value of the different
surveys with respect to sampling effort.
The most typical sampling gears used were long-
lines, gill nets, and purse seines targeting salmon,
squid, or pomfret; rarer collections were made with
plankton nets, dip nets, or hook and line. Thus,
sampling was typically in surface waters, most likely
in the upper 50 m; the true depth distribution of the
pelagic animals, however, is unknown. The only
other "gear" was the sei whale, Balaenoptera bore-
alis, from which stomach contents were studied for
about 27 years; Chikuni (1970) reported armorhead
in the stomachs.
RESULTS AND DISCUSSION
Distributional Patterns
The known center of distribution and spawning
for benthic P. wheeleri is in the SE-NHR seamount
region, bounded by lat. 29°-35°N and long. 171°E-
179°W (Takahashi and Sasaki 1977; Humphreys and
Tagami 1986). These reproductive fish are typical-
ly between 23.0 and 28.5 standard length (SL)
(Sasaki 1986) and range in age from 1.5 to 2.5 years
(Uchiyama and Sampaga fn. 3). They are found at
depths between 200 and 500 m (Takahashi and
Sasaki 1977). Other benthic occurrences have been
recorded but are sporadic (Fig. 1). Several speci-
mens were captured off Japan in 1957 (Abe 1957),
1969, 1971-73, 1976 (Zama et al. 1977), and 1979
(Okamura et al. 1982). These fish were typically the
same size as those captured on the seamounts, but
no mention of reproductive status was made. Fish
of similar length occur rarely on the west coast of
North America, but only between lat. 37°00' and
44°25'N (Wagner and Bond 1961; Follett and Demp-
ster 1963; Smith 1965). Again, reproductive status
is unknown for these specimens. No benthic occur-
rences off British Columbia or Alaska have been
reported, despite extensive trawling surveys which
included seamounts (Hughes 1981; Alton 1986).
Large specimens have been taken rarely in the
Hawaiian Archipelago (Fig. 1; Randall 1980;
Humphreys et al. in press) and are typically 4-5
years old and appear to be in reproductive condi-
120'
150
Figure 1.— Known distribution of benthic occurrences of pelagic armorhead, Pseudopentaceros wheeleri. Squares represent the main
population centers at central North Pacific seamounts. All others are rare or sporadic occurrences.
455
FISHERY BULLETIN: VOL. 86, NO. 3
tion. Their relationship with the SE-NHR seamount
populations is unknown.
The most common pelagic occurrences (1955-87)
were fish ranging from 18 to 26 cm SL; smaller sizes
are poorly represented, probably because of the
nature of the sampling gear. Smaller specimens
were generally captured with plankton nets (Fedo-
sova and Komrakov 1975; Borets and Sokolovsky
1978), dip nets (Honma and Mizusawa 1969; Ran-
dall 1980; Fujii'*), and in one case, in a whale stomach
(Kawamura 1982). The sea-surface temperatures at
which the specimens were captured ranged from
8.6° to 15.0°C (average 12.1°C). For all years,
pelagic captures were restricted to the eastern
North Pacific, with the exceptions of larvae or early
juveniles captured near the spawning centers at the
seamounts (Komrakov 1970; Table 2), over 20
specimens in a Bryde's whale stomach in 1979
(Kato^, Fig. 2B), and two individuals captured in
1986 at long. 155°E (Fig. 2). This is despite exten-
sive sampling in the western North Pacific, par-
ticularly by the Japan Marine Fishery Resource
Research Center (JAMARC) and the Hokkaido
University, and in the Bering or Okhotsk Seas (Table
1).
Pelagic captures in the 1950's were sporadic, but
several fish were taken north of lat. 50°N, especially
in 1958 (Fig. 2 A). Relatively few were captured in
the 1960's, with the exception of July through early
August 1969, when the armorhead was an impor-
*E. Fujii, Nippon Luther Shingaku Daigaku, Tokyo, Japan, pers.
commun. July 1987.
^H. Kato, Whales Research Institute, Tokyo, Japan, pers.
commun. December 1987.
Table 2. — Pelagic specimens of armorhead, Pseudopentaceros wheeleri, for which sizes were
available. The growth curve from Uchiyama and Sampaga (text footnote 3), based upon
enumeration of daily growth increments, was used to deduce age from length. For the youngest
fish (bottom), average readings from daily growth increments, rather than the growth curve,
were used. Data are from various published sources and collections listed in Table 1 . (SST
= sea-surface temperature.)
Standard
Lat.
length
Year
Date
N
Long.
SST
(mm)
Age (yr)
No.
Gear
1955
9/27
45°41'
165°5'W
108-122
0.38-0.42
2
Gill net
1955
9/27
45°35'
165°5'W
—
200
0.79
Dip net
1956
7/31
45°49'
160°3'W
12.2
245
1.29
Gill net
1956
9/15
45°00'
155°00'W
12.1
245
1.29
2
Gill net
1956
9/17
49°00'
150°00'W
13.2
245
1.29
Gill net
1956
9/17
51°00'
150°00'W
11.0
245
1.29
Gill net
1958
8/14
44°39'
174°48'W
—
64-91
0.26-0.33
2
Dip net
1958
8/25
49°43'
146°10'W
11.8
256
1.50
Gill net
1963
8/24
49°0'
162°0W
12.9
249
1.40
Gill net
1967
6/15
54°38'
150°1'W
8.8
121-134
0.42-0.46
4
Dip net
1968
7/
41°30'
165°30'W
—
285
>2.00
1
Handline
1969
4/1
29°48'
179°5'E
—
8-23
0.08-0.25
?
Plankton
1969
4/1
31 °6'
176°0'E
—
8-23
0.08-0.25
—
Plankton
1969
4/1
32°6'
173°0'E
—
8-23
0.08-0.25
—
Plankton
1972
5/21
38°30'
175°3W
14.3
42-50
0.21-0.23
5
Whale stomach
1984
7/25
45°30'
155°0'W
13.0
262-265
1.81-2.02
3
Gill net
1984
7/25
45°26'
154°58'W
12.0
278
>2.00
2
Longline
1985
7/9
46°57'
129°4W
—
272
>2.00
1
Gill net
1985
7/10
47°30'
128°36W
—
272-298
>2.00
2
Gill net
1985
7/11
47°59'
129°19'W
—
277
>2.00
1
Gill net
1985
7/13
46°23'
131°29W
—
188-219
0.71-0.94
5
Gill net
1985
7/14
46°40'
131°8W
—
183-233
0.68-1.1
9
Gill net
1985
7/15
46°35'
131°5W
—
206
0.83
1
Gill net
1985
7/19
46°54'
131°28'W
—
190-217
0.72-0.92
2
Gill net
1985
7/21
47°18'
130°23'W
—
191-220
0.73-0.95
3
Gill net
1985
7/22
47°35'
130°48'W
—
190-201
0.72-0.79
2
Gill net
1985
7/23
47°37'
130°3rW
—
198-200
0.77-0.79
2
Gill net
1985
7/12
47°0'
155°1'W
11.6
263-265
1.87-2.02
2
Longline
1985
7/13
45°28'
155°5W
11.8
238-262
1.17-1.81
12
Longline
1985
7/11
47'='0'
155°0'W
11.6
248-262
1.35-1.81
5
Gill net
1985
7/12
45°30'
155°0W
11.8
235-240
1.12-1.20
4
Gill net
1985
7/13
44°0'
155°0'W
12.8
248-258
1.35-1.63
4
Gill net
1985
2/23
30° 16'
181°18W
—
8-15
0.08-0.16
15
Neuston
1985
2/24
29°27'
180°56'W
—
5-15
0.04-0.16
15
Neuston
1986
8/3
43°51'
164°54'W
13.0
50-100
0.23-0.35
5
Dip net
456
BOEHLERT and SASAKI: PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD
150°
120'
Figure 2.— Records of pelagic occurrences of armorhead, Pseudopentaceros wheeleri, in the North Pacific. Most data were obtained
fronn either published records or unpublished data described in Table 1: Each symbol represents a collection taking armorhead,
rather than single specimens. (A) 1955-69. Here, note that the area in the box contains the locations of 195 sei whales captured
in July-August 1969 that had fed extensively upon armorhead. The samples from the central Pacific were larvae and early juveniles
(Fedosova and Komrakov 1975). (B) 1970-87.
457
FISHERY BULLETIN: VOL. 86, NO. 3
tant diet component of sei whales taken by the
Japanese commercial fleet. This species normally
feeds on zooplankton and micronekton (Kawamura
1982), but in this year, 195 whales were recorded
to have fed extensively on armorhead. Although
data records for whale stomach contents extend
from 1952 to 1979, 1969 was the only year when
large numbers fed on armorhead, with the excep-
tion of two occurrences each in 1971 and 1972.
Based on the known feeding mode of sei whales,
Chikuni (1970) suggested that the armorhead ap-
parently schools in surface waters. In the 1970's,
catches were relatively small, but in the 1980's,
specimens have been captured in each year, with the
greatest catches in 1985. In that year, fish occurred
in more easterly areas as compared to other years.
Relationship of Distribution Patterns
to North Pacific Oceanography
The SE-NHR seamounts are located in the mid-
Pacific transition zone, between subarctic and sub-
tropical water masses (Roden 1970). The spawning
season of armorhead typically lasts from late
November through March (Sasaki 1974; Bilim et al.
1978), and larvae are neustonic, at least through the
first few months of life (G. W. Boehlert unpubl.
data). Near surface drift in this region is largely
wind driven (McNally 1981); as shown by Lagran-
gian drifters, mean surface flow in the SE-NHR sea-
mount region during winter months is typically
eastward (Kirwan et al. 1978; Emery et al. 1985)
or southeastward (McNally et al. 1983), and long-
term mean Ekman transport has a southeastward
component (Favorite et al. 1976). Based on mean
ship-drift data, progressive vector diagrams over a
3-mo period, with starting points at four SE-NHR
seamounts (Bakun^), show this typical eastward flow
with a southward component (Fig. 3). If pelagic
armorhead followed such a pattern, individuals
would remain in the subtropical gyre, continuing
eastward and then turning south, entering the
California Current system. The scarcity of speci-
mens in this region (Figs. 1, 2) suggests instead that
most fish move northeastward in some manner,
entering subarctic waters and residing within the
Alaska gyre; surface drift in winter is northeastward
in anomalous years (McNally 1981), and average
'Compilation by the NMFS Pacific Fisheries Environmental
Group from historic ship drift files assembled by the U.S. Naval
Oceanographic Office (A. Bakun, Pacific Fisheries Environmen-
tal Group, National Marine Fisheries Service, NOAA, P.O. Box
831, Monterey, CA 93942, pers. commun. April 1987.
36°N
35°N
30*N
25°N
17rE
175°E
180'
178°W
Figure 3.— Progressive vector diagrams of mean surface currents (starting date, 15
December; duration, 90 days), from long-term mean ship drift data from 1° squares,
with distances calculated on a daily basis. Triangles indicate starting locations at four
seamounts with armorhead spawning populations. Each mark on the vector between
the origin and arrow represents a 5-d period. See text footnote 6.
458
BOEHLERT and SASAKI; PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD
surface flow is in this direction in summer months
(McNally et al. 1983).
The two available studies on larval distribution of
armorhead are conflicting and suggest that patterns
of movement may differ from year to year. Komra-
kov (1970) conducted surveys in March-April 1969
and observed larvae only at northern seamounts and
in the open ocean several hundred kilometers north-
east of the SE-NHR seamounts. Borets (1979), how-
ever, observed 5-20 mm larvae remaining mostly
around the seamounts, with highest abundance
south of lat. 33°N in 1976. Because spawning oc-
curs at seamounts northwest of this area, this pat-
tern is consistent with southeastward drift during
the first few months of life (Fig. 3). If larvae remain
in the seamount region, they could be transported
to the northeast in summer months, or they may
actively migrate northeastward after reaching the
juvenile stage. A similar migration apparently
occurs for the pomfret, Bramajaponica, which prior
to spawning moves into the North Pacific current
region in a broad latitudinal band between the sub-
tropical convergence and the subarctic boundary;
immature fish and spent adults then move north-
ward into the subarctic region to feed during April-
May (Shimazaki and Nakamura 1981). The pomfret
differs from the armorhead, however, in that it
remains pelagic throughout its life and also is abun-
dant in the western North Pacific. Larval and
juvenile Brama sp. do occur in the same area as lar-
val armorhead (Borets and Sokolovsky 1978; G. W.
Boehlert unpubl. data).
A possible scenario of movements after the lar-
val stage can be deduced from the ages of individuals
captured in different regions. Based upon the
growth of pelagic armorhead (Uchiyama and Sam-
paga fn. 3), most of the larger specimens indicated
in Figure 2 are from 1 to 2 years old, and the pelagic
duration for this species is most often from 1.5 to
2.5 years. Using daily growth increments for lar-
val and small juveniles and fitted growth curves for
larger juveniles and adults, one can convert lengths
to ages. The pelagic specimens of known length
range in estimated age from 0.04 to >2 years (Table
2). The youngest animals are typically found in the
region of the seamounts (Komrakov 1970; Fedosova
and Komrakov 1975; Borets and Sokolovsky 1978),
but intermediate-sized fish are found north and east
of the seamounts. Different age groups seem to be
distributed in different areas in the eastern North
Pacific (Fig. 4). Armorhead spawn in the seamount
region (as area A in Figure 4) during November to
March, and juveniles 5-25 mm long are found in the
SE-NHR seamounts area from February to early
April (Fedosova and Komrakov 1975; Borets 1979).
Subsequent occurrences of fish older than 0.25 year
(Table 2) are found only in the region northeast of
the seamounts and in the Gulf of Alaska (area B in
Figure 4). Northeastward movements are contrary
to mean ocean currents, but Roden et al. (1982)
described intense northward flow along the axis of
the southern Emperor Seamounts that could con-
ceivably play a role in transport of fish at the young-
est stages.
Recruitment of later stage armorhead to the sea-
mounts requires movement to the west. The only
significant westward circulation in the North Pacific
is found in the Alaskan Stream (Favorite 1967). This
narrow current occurs north of lat. 51 °N and ex-
tends westerly to near long. 170 °E, where it splits
to two branches, one turning northerly into the
Bering Sea and the other southerly under the influ-
ence of the northern Emperor Seamounts or Kor-
mandorskie Ridge (Favorite 1967; dashed line in
Figure 4). Damitsky et al. (1984) described southerly
movement of this water to at least lat. 40 °N and
suggested that it played a role in transport of
planktonic food for armorhead on the seamounts.
Unless returning armorhead are deep in the water
column and therefore not sampled, distributional
data (Fig. 2) do not support this route for westerly
movement. Fish near recruitment size and age are
instead captured in the region marked C on Figure
4, south of the area where younger fish are captured.
These fish apparently migrate back to the region of
the seamounts by some unknown, but probably
active, mechanism.
Two less important, but plausible, movement
patterns may explain the rarer occurrences in south-
ern Japan, off Oregon and California, and in the
Hawaiian Archipelago (Fig. 4). Chelton (1984) dis-
cussed possible interannual changes in the lati-
tudinal position of the West Wind Drift which may
lead to differing magnitudes of transport in the
Alaska and California Currents. The rare benthic
occurrences of the pelagic armorhead off the west
coast of North America (area D in Figure 4) may
come from specimens that drift southward in the
California Current in years of greater southern
transport. Some of these may not settle out and re-
main within the subtropical gyre, possibly recruit-
ing to the Hawaiian chain far to the south of the
normal reproductive population. This route, because
of its distance, apparently takes considerably longer
than that in the subarctic gyre; based upon satellite
drifters, McNally et al. (1983) suggested that a full
circuit of the subtropical gyre takes 4.5 years. The
armorhead collected in the Northwestern Hawaiian
459
FISHERY BULLETIN: VOL. 86, NO. 3
Figure 4.— Distributions of benthic and pelagic specimens of armorhead, Pseudopentaceros wheeleri, by age group with possible
patterns of movement. Area A represents the major spawning area, and area E the region inhabited by the large, older specimens.
Area B is the region in which specimens younger than 1 year have been captured, and area C where fish from 1 to 2 years old
have been captured. The arrow through this region represents the main population movement pattern. Other arrows represent
possible movements for stray fish. Areas marked D and F represent locations of benthic collections from California and Japan,
respectively.
Islands (area E in Figure 4) are typically larger than
those in the normal part of the range and are corre-
spondingly older, with ages estimated at 4 or 5 years
(Humphreys et al. in press), similar to the time sug-
gested for a circuit of the gyre by McNally et al.
(1983).
The other plausible movement pattern can be used
to explain the existence of specimens in Japan. Ben-
thic specimens of armorhead are rarely captured in
Japan (Fig. 1); a report of 3,000 t landed in 1969
by Soviet fishermen in this region (Abe 1969) may
be in error, since the Soviet fishery on the SE-NHR
seamounts developed heavily that year (Borets 1975)
and SE-NHR fish were likely sold in Japan. The
reproductive condition of these animals is not
known, so the source of the spawning population
supplying them is at question. The large, reproduc-
tive individuals in area E (Fig. 4) have been found
as far south as lat. 23°43'N (Humphreys et al. in
press). It is possible that larvae and juveniles are
transported to Japan in the northern part of the
North Equatorial Current (Uda and Hasunuma
1969). This is consistent with the capture of over
20 fish in a Bryde's whale stomach in September
1979 at lat. 26°N (H. Kato fn. 5; Fig. 2B). Subse-
quent captures of this species have not been made
in Japanese waters, where their presence may re-
quire a rare recruitment event from an upstream
population source, as has been suggested for sea-
mount populations of lobsters (Lutjeharms and
Heydorn 1981). These animals may also remain
within a gyral circulation, however, as suggested by
the presence of two specimens captured in the
Kuroshio in 1986 (Fig. 2B).
Recruitment to the Seamounts
The seasonal timing of the recruitment of the
pelagic armorhead to the SE-NHR seamounts is
unknown. That no records of larger pelagic fish cap-
tured near the seamounts (Fig. 2) exist may be an
artifact of sampling in inappropriate seasons or lack
of sampling in deeper water. The youngest benthic
specimens from the seamounts were slightly in ex-
460
BOEHLERT and SASAKI: PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD
cess of 1.5 years in early summer (Uchiyama and
Sampaga fn. 3) and 2 years in winter, suggesting
that some recruitment occurs in spring.
In most fisheries, temporal recruitment patterns
can be discerned from length-frequency analyses,
because the smaller recruits represent a larger pro-
portion of the population during the season of
recruitment. For the armorhead, no increase in
smaller size classes is apparent in monthly length-
frequency samples. This may be due to cessation of
somatic growth after recruitment to the seamount
(Humphreys and Tagami 1986; Uchiyama and Sam-
paga fn. 3). Early recruits, with their deeper bodies
and greater fat content, differ morphologically from
longer seamount residents. The transition from
so-called "fat" to "lean" morphotypes occurs in
association with development of gonads (Humphreys
et al. in press). Thus, an index of morphological
change can be used to detect recruitment patterns,
much the same as length frequencies can be used
for other species. Condition factor is frequently used
to assess "fatness" of fish; it is normally expressed
as weight divided by length to a power (typically 3)
multiplied by some scaling factor (Ricker 1975). For
armorhead, there is a marked change in condition
factor with morphological change; newly recruiting
fish have high condition factors relative to inter-
mediate or lean fish. By considering the monthly
proportion of fish above an arbitrary value of con-
dition factor, we can estimate the influx of new
recruits to the seamounts. During May 1972-
December 1973, the major influx of new recruits
with a high condition factor occurred in April-May
1973 (Fig. 5). The data for 1972 are incomplete, but
some recruitment apparently occurred in August
and September (Fig. 5). Differences in seasonality
of recruitment between 1972 and 1973 may be
indicative of interannual variability in temporal
recruitment patterns.
An influx of fish to the seamounts may also be
reflected in the catch per unit effort (CPUE), an
index of stock abundance. Monthly averages of
CPUE (in metric tons per hour) by Japanese
trawlers for armorhead at the SE-NHR seamounts
were highest in March and April (Sasaki 1986).
Overall armorhead CPUE decreased continuously
from 54.1 t/h in 1972 to only 0.3 t/h in 1982 (Sasaki
1986) and further decreased to 0.06-0.07 t/h in
1983-85 (T. Sasaki unpubl. data). The estimated
CPUE in 1986, however, was 0.31 t/h, an increase
of four or five times that in the previous year. The
CPUE of trawlers operating in May-Septem.ber
1986 was highest in May, decreased in June and
July, but increased slightly in August and Septem-
40
35-
30-
H 25
z
HI
O 20-
oc
UJ
°- 15H
10
5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
1972
MONTH
1973
Figure 5.— Seasonality of the proportion of armorhead, Pseudopentaceros wheeleri, with
condition factors >2.0 from four seamounts of the northern Hawaiian Ridge (Southeast
Hancock, Northwest Hancock, C-H, and Colahan) for 1972-73. Squares, males; triangles,
females. Data are based upon 2,104 males and 1,704 females.
461
FISHERY BULLETIN: VOL. 86, NO. 3
ber. Based on all the information, the major recruit-
ment may occur in spring with some additional
recruitment in August and September (Fig. 5); data
are not sufficient to determine the recruitment
pattern in detail.
A broad geographic separation exists between the
location of pelagic captures of larger armorhead and
the location of the spawning populations (Figs. 2,
4). If recruitment to the seamounts occurs predom-
inantly in spring (as suggested for 1973 in Figure
5), then temporal sampling patterns may have
missed these fish, although JAMARC pomfret
surveys covered this area in some seasons (Table 1).
Locating the seamounts, which have small (2-5 km)
summits must be a formidable task given the wide
ocean areas over which armorhead are distributed.
A similar situation exists for rock lobster, Jasus
tristani, in the South Atlantic, that recruits from
an upstream population some 2,000 km away and
in sufficient numbers to support a fishery on Vema
Seamount in some years (Lutjeharms and Heydorn
1981).
Open-ocean migrations of fishes may depend
upon many potential cues, including electric fields
(McCleave and Power 1978), magnetic fields
(Walker 1984), gyres (Williams 1972), and phero-
mones (Nordeng 1977). Certain characteristics of
these isolated, open-ocean seamounts may promote
their detection by armorhead. First, current-topog-
raphy interactions may create significant signals in
physical and biological features. The region of the
SE-NHR seamounts is active in front development
(Roden and Paskausky 1978); upwelling, eddies, and
other aspects of flow complexity also occur around
these seamounts (Roden et al. 1982) and down-
stream from them (Royer 1978). The biological sig-
nals may include increased chlorophyll in response
to upwelling or doming of isotherms (Genin and
Boehlert 1985), or aggregations of various organ-
isms and the larger animals which prey upon them
around seamounts (see review in Boehlert and Genin
1987). Gravity anomalies associated with seamounts
may also play a role; positive gravity anomalies exist
at the summit and slopes, and negative anomalies
are seen in the surrounding "moat" regions (Wedge-
worth and Kellogg 1987). Seamounts often have
strong magnetic dipoles associated with them, and
the dipole might serve as a landmark for magnetic
orientation by fish (Klimley^). While fish have been
shown to have magnetoreceptors (Walker et al.
1985), their use of magnetic maps remains specula-
T. Klimley, Scripps Institution of Oceanography, La JoUa, CA
92038, pers. commun. 30 June 1987.
tive but possible (Gould 1985). Although we cannot
postulate the mechanism that armorhead use for
recruitment, it is clear that the effects of seamounts
may be detected at distances greater than their area
alone would suggest.
Interannual Variations in
Recruitment Strength
The year-class strength of armorhead recruiting
to seamounts appears to be independent of the
parent stock size (Wetherall and Yong 1986). As an
example, Borets (1975) estimated that on the SE-
NHR seamounts from 1968 to 1973, the stock size
varied by a factor of <1.8 while recruitment varied
by >5.5 times. The relative abundance of armorhead
at the SE-NHR seamounts area increased in 1986
after a long period at a very low level. This increase
probably corresponds to the high abundance of
pelagic specimens captured in the northeastern
Pacific in 1985 (Fig. 2B; Table 2) that consisted of
two age groups. The increased recruitment at the
SE-NHR seamounts in 1986 suggests that environ-
mental conditions were favorable to the survival of
young armorhead in the 1984 and 1985 winter
seasons.
A wide variety of factors, both biotic and physical,
can affect survival and ultimate year-class strength
in fishes (Lasker 1978). During the 2 years between
spawning and recruitment for armorhead, an ex-
tended migration through varied pelagic environ-
ments occurs (Fig. 4). Feeding conditions for larval
and juvenile stages are characterized by interannual
variability; Fedosova (1980) suggested that warm
years were more productive for zooplankton prey
and, thus, favorable to the survival of young armor-
head. Interannual variation in atmospheric systems
(Seckel 1988) or large-scale ocean currents of the
kind described by Mysak et al. (1982) may also play
a role in armorhead recruitment strength. Changes
in the position of the Alaska gyre by up to 700 km
southwest of its normal position may have occurred
from 1981 to 1985, with an associated increase in
seawater temperature (Royer and Emery 1987).
Large-scale atmospheric phenomena, such as the
longitudinal position of the Aleutian Low, may
create definite interannual variations in winter wind
systems that may be seen in surface current pat-
terns (Seckel 1988). These patterns may, in turn,
be related to the latitudinal position of the subtrop-
ical front, which varies interannually between lat.
28° and 32°N (Roden 1970). Variability in these
features influences surface drift (McNally 1981),
which in turn affects the neustonic young of armor-
462
BOEHLERT and SASAKI: PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD
head. If flow in certain years results in transport
of armorhead to other regions (such as the western
Pacific or the southern part of the northeast Pacific;
Fig. 4), it is possible that large-scale mortalities of
armorhead occur later in life, resulting in weak year
classes at the SE-NHR seamounts.
ACKNOWLEDGMENTS
We appreciate assistance from many individuals
who provided access to different data sources used
in this paper, including Akito Kawamura, Colin Har-
ris, Sigeiti Hayasi, Genji Kobayashi, and Glen Jamie-
son. We also thank Ted Pietsch, Gunter Seckel, and
Peter Klimley and an anonymous reviewer for com-
ments on an earlier draft of the manuscript.
LITERATURE CITED
Abe, T.
1957. New, rare or uncommon fishes from Japanese [waters].
VI. Notes on the rare fishes of the family Histiopteridae.
Jpn. J. Ichthyol. 6:35-39, 71-74.
1969. Notes on some edible marine fishes collected between
the Bonin Islands and the mouth of Sagami Bay— III. Bull.
Tokai Reg. Fish. Res. Lab. 60:5-8.
Alton, M. S.
1986. Fish and crab populations of Gulf of Alaska seamounts.
In R. N. Uchida, S. Hayasi, and G. W. Boehlert (editors),
Environment and resources of seamounts in the North
Pacific, p. 45-51. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS 43.
BiLIM, L. A., L. A. BORETS, AND L. K. Platoshina.
1978. Characteristics of ovogenesis and spawning of the boar-
fish in the region of the Hawaiian Islands. In Fisheries
oceanography, hydrobiology, biology of fishes and other
denizens of the Pacific Ocean, p. 51-57. Izv. Pac. Ocean
Sci. Res. Inst. Fish. Oceanogr. (TINRO), Vladivostok, 102.
(Engl, transl. by W. G. Van Campen, 1986, 9 p., Transl. No.
106; available Southwest Fish. Cent. Honolulu Lab., Natl.
Mar. Fish. Serv., NOAA, Honolulu, HI 96822-2396.)
Boehlert, G. W.
1986. Productivity and population maintenance of seamount
resources and future research directions. In R. N. Uchida,
S. Hayasi, and G. W. Boehlert (editors). Environment and
resources of seamounts in the North Pacific, p. 95-101.
U.S. Dep. Commer., NOAA Tech. Rep. NMFS 43.
Boehlert, G. W., and A. Genin.
1987. A review of the effects of seamounts on biological
processes. In B. Keating, P. Fryer, R. Batiza, and G. W.
Boehlert (editors), Seamounts, islands, and atolls, p. 319-
334. Geophys. Monogr. 43.
BORETS, L. A.
1975. Some results of studies on the biology of the boarfish
{Pentaceros richardsoni Smith). Invest. Biol. Fishes Fish.
Oceanogr., TINRO, Vladivostok 6:82-90. (Engl, transl. by
W. G. Van Campen, 1984, 9 p., Transl. No. 97; available
Southwest Fish. Cent. Honolulu Lab., Natl. Mar. Fish. Serv.,
NOAA, Honolulu, HI 96822-2396.)
1976. The state of the boarfish, Pentaceros richardsoni
(Smith) stock in the Hawaii submersive range areas. Invest.
Biol. Fishes Fish. Oceanogr., TINRO, Vladivostok 7:58-62.
(Engl, transl. by W. G. Van Campen, 1987, 5 p., Transl.
No. 112; available Southwest Fish. Cent. Honolulu Lab.,
Natl. Mar. Fish. Serv., NOAA, Honolulu, HI 96822-
2396.)
1979. The population structure of the boarfish, Pentaceros
richardsoni, from the Emperor Seamounts and the
Hawaiian Ridge. J. Ichthyol. 19:15-20.
BORETS, L. A., AND A. S. SOKOLOVSKY.
1978. Species composition of the ichthyoplankton of the
Hawaiian submarine ridge and the Emperor Seamounts.
In Fisheries oceanography, hydrobiology, biology of fishes
and other denizens of the Pacific Ocean, p. 43-50. Izv. Pac.
Ocean Sci. Res. Inst. Fish. Oceanogr. (TINRO), Vladivostok,
Vol. 102. (Engl, transl. by W. G. Van Campen, 1986, 10
p., Transl. No. 105; available Southwest Fish. Cent. Hono-
lulu Lab., Natl. Mar. Fish. Serv., NOAA, Honolulu, HI
96822-2396.)
Chelton, D. B.
1984. Commentary: Short-term climatic variability in the
northeast Pacific Ocean. In W. G. Pearcy (editor). The in-
fluence of ocean conditions on the production of salmonids
in the North Pacific, p. 87-99. Oreg. State Univ. Sea Grant,
ORESU-W-83-001.
Chikuni, S.
1970. On gregarious fish, Pseudopentaceros richardsoni
(Histiophoridae). Enyo (Far Seas) Fish. Res. Lab. News
3:1-4. (Engl, transl., The "phantom fish," "kusakari tsu-
bodai"— an outline, by J. H. Shohara, 7 p.; available Natl.
Mar. Fish. Serv., NOAA, Terminal Island, CA 90731.)
Clemens, W. A., and G. V. Wilby.
1961. Fishes of the Pacific coast of Canada. 2d ed. Bull.
Fish. Res. Board Can. 68, 443 p.
Darnitsky, V. B., B. L. Boldyrev, and A. F. Volkov.
1984. Environmental conditions and some ecological char-
acteristics of fishes from the central North Pacific sea-
mounts. In P. A. Moiseev (editor). Proceedings, conditions
of formation of commercial fish concentrations, p. 64-77.
Minist. Fish., U.S.S.R., All-Union Res. Inst. Mar. Fish.
Oceanogr. VNIRO. (Engl, transl. by W. G. Van Campen,
1987, 12 p., Transl. No. 114; available Southwest Fish. Cent.
Honolulu Lab., Natl. Mar. Fish. Serv., NOAA, Honolulu, HI
96822-2396.)
Emery, W. J., T. C. Royer, and R. W. Reynolds.
1985. The anomalous tracks of North Pacific drifting buoys
1981 to 1983. Deep-Sea Res. 32:315-347.
Favorite, F.
1967. The Alaskan Stream. Int. N. Pac. Fish. Comm. Bull.
21, 20 p.
Favorite, F., A. J. Dodimead, and K. Nasu.
1976. Oceanography of the subarctic Pacific region, 1960-
1971. Int. N. Pac. Fish. Comm. Bull. 33, 187 p.
Fedosova, R. a.
1980. Inter-year changes in the mesoplankton of the waters
of the submarine banks of the Hawaiian Ridge (Northwest-
ern Pacific Ocean). [In Russ.] Izv. Pac. Ocean Sci. Res.
Inst. Fish. Oceanogr. (TINRO), Vladivostok 104:84-87.
Fedosova, R. A., and 0. E. Komrakov.
1975. Feeding of Pentaceros richardsoni frys in the Hawaiian
region. [In Russ.] Invest. Biol. Fishes Fish. Oceanogr.,
TINRO, Vladivostok 6:52-55. (Engl, transl. by W. G. Van
Campen, 1985, 4 p., Transl. No. 100; available Southwest
Fish. Cent. Honolulu Lab., Natl. Mar. Fish. Serv.. NOAA,
Honolulu, HI 96822-2396.)
FOLLETT, W. I., AND L. J. DEMPSTER.
1963. Relationships of the percoid fish Pentaceros richard-
soni Smith, with description of a specimen from the coast
463
FISHERY BULLETIN: VOL. 86, NO. 3
of California. Proc. Calif. Acad. Sci., 4th Ser. 32(10):315-
338.
FUJII, E.
1986. Zoogeographical features of fishes in the vicinity of sea-
mounts. In R. N. Uchida, S. Hayasi, and G. W. Boehlert
(editors). Environment and resources of seamounts in the
North Pacific, p. 67-70. U.S. Dep. Commer., NOAA Tech.
Rep. NMFS 43.
Genin, a., and G. W. Boehlert.
1985. Dynamics of temperature and chlorophyll structures
above a seamount: An oceanic experiment. J. Mar. Res.
43:907-924.
Gould, J. L.
1985. Are animal maps magnetic? In J. L. Kirschvink, D. S.
Jones, and B. J. MacFadden (editors), Magnetite biomineral-
ization and magnetoreception in organisms, p. 257-268.
Plenum Press, N.Y.
Hardy. G. S.
1983. A revision of the fishes of the family Pentacerotidae
(Perciformes). N.Z. J. Zool. 10:177-220.
HONMA, Y., AND R. MiZUSAWA.
1969. A record of the young of boar fish, Pseudopentaceros
richardsoni, from North Pacific. Jpn. J. Ichthyol. 15:134-
136.
Hughes, S. E.
1981. Initial U.S. exploration of nine Gulf of Alaska sea-
mounts and their associated fish and shellfish resources.
Mar. Fish. Rev. 43(l):26-33.
Humphreys, R. L., Jr., and D. T. Tagaml
1986. Review and current status of research on the biology
and ecology of the genus Pseudopentaceros. In R. N.
Uchida, S. Hayasi, and G. W. Boehlert (editors). Environ-
ment and resources of seamounts in the North Pacific, p.
55-62. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 43.
Humphreys, R. L., Jr., G. Winans, and D. T. Tagaml
In press. Synonomy and proposed life history of the North
Pacific pelagic armorhead, Pseudopentaceros wheekri Hardy
(Pisces: Pentacerotidae). Copeia 1988.
Ignell, S., J. Bailey, and J. Joyce.
1986. Observations on high-seas squid gill-net fisheries, North
Pacific Ocean, 1985. U.S. Dep. Commer., NOAA Tech.
Memo. NMFS F/NWC-105, 52 p.
Japan Marine Fishery Resource Research Center.
1980. Report of exploratory fishing of Pacific pomfret. Jpn.
Mar. Fish. Resour. Res. Cent. Rep. 53-16.
1983a. Report of exploratory fishing of Pacific pomfret.
Jpn. Mar. Fish. Resour. Res. Cent. Rep. 54-18.
1983b. Report of exploratory fishing of Pacific pomfret.
Jpn. Mar. Fish. Resour. Res. Cent. Rep. 55-16.
1985. Report of exploratory fishing of Pacific pomfret. Jpn.
Mar. Fish. Resour. Res. Cent. Rep. 57-12.
Kawamura, a.
1982. Food habits and prey distributions of three rorqual
species in the North Pacific Ocean. Sci. Rep. Whales Res.
Inst. 34:59-91.
KiRWAN, A. D., Jr., G. H. McNally, E. Reyna, and W. J.
Merrell, Jr.
1978. The near-surface circulation of the eastern North
Pacific. J. Phys. Oceanogr. 8:937-945.
KOMRAKOV, 0. E.
1970. Distribution and fishery of the boarfish (Pentaceros
richardsoni Smith) in the Hawaiian region. In the collection:
The present state of biological productivity and the volume
of biological resources of the world ocean and prospects for
their utilization. Kaliningrad, p. 155-163. (Engl, transl.
by W. G. Van Campen, 1987, 9 p., Transl. No. 117; available
Southwest Fish. Cent. Honolulu Lab., Natl. Mar. Fish. Serv.,
NOAA, Honolulu, HI 96822-2396.)
Larkins, H. a.
1964. Some epipelagic fishes of the North Pacific Ocean,
Bering Sea, and Gulf of Alaska. Trans. Am. Fish. Soc.
93:286-290.
Lasker, R.
1978. The relation between oceanographic conditions and lar-
val anchovy food in the California Current: Identification
of factors contributing to recruitment failure. Rapp. P. -v.
R6un. Cons. int. Explor. Mer 173:212-230.
Lutjeharms, J. R. E., and A. E. F. Heydorn.
1981. Recruitment of rock lobster on Vema Seamount from
the islands of Tristan da Cunha. Deep-Sea Res. 28A: 1 237.
Macy, p. Y., J. M. Wall, N. D. Lampsakis, and J. E. Mason.
1978. Resources of non-salmonid pelagic fishes of the Gulf
of Alaska and eastern Bering Sea. Parts 1-3, Final Rep.
BLM/OCSEAP Contracts R7120811 and R7120812, North-
west and Alaska Fish. Cent., NMFS, NOAA, Seattle, Wash.
McCleave, J. D., AND J. H. Power.
1978. Influence of weak electric and magnetic fields on turn-
ing behavior in elvers of American eel Anguilla rostrata.
Mar. Biol. (Berl.) 46:29-34.
McNally, G. J.
1981. Satellite-tracked drift buoy observations of the near-
surface flow in the eastern mid-latitude North Pacific. J.
Geophys. Res. 86:8022-8030.
McNally, G. J., W. C. Patzert, A. D. Kirwan, Jr., and A. C.
Vastano.
1983. The near-surface circulation of the North Pacific using
satellite tracked drifting buoys. J. Geophys. Res. 88:7505-
7518.
Mysak, L. a., W. W. Hsieh, and T. R. Parsons.
1982. On the relationship between interannual baroclinic
waves and fish populations in the northeast Pacific. Biol.
Oceanogr. 2:63-103.
Nordeng, H.
1977. A pheromone hypothesis for homeward migration in
anadromous salmonids. Oikos 28:155-159.
Okamura, I., K. Amaoka, and F. Mitani (editors).
1982. Fishes of the Kyushu-Palau Ridge and Tosa Bay. Jpn.
Fish. Res. Conserv. Assoc, Tokyo, 435 p.
Randall, J. E.
1980. New records of fishes from the Hawaiian Islands. Pac.
Sci. 34:211-232.
Ricker, W. E.
1975. Computation and interpretation of biological statistics
of fish populations. Fish. Res. Board Can. Bull. 191, 382 p.
Robinson, S. M. C, and G. S. Jamieson.
1984. Report on a Canadian commercial fishery for flying
squid using drifting gill nets off the coast of British Colum-
bia. Can. Ind. Rep. Fish. Aquat. Sci. 150, 25 p.
Roden, G. I.
1970. Aspects of the mid-Pacific transition zone. J. Geophys.
Res. 75:1097-1109.
Roden, G. I., and D. F. Paskausky.
1978. Estimation of rates of frontogenesis and frontolysis in
the North Pacific Ocean using satellite and surface meteor-
ological data from January 1977. J. Geophys. Res. 83:4545-
4550.
Roden, G. I., B. A. Taft, and C. C. Ebbesmeyer.
1982. Oceanographic aspects of the Emperor Seamounts
region. J. Geophys. Res. 87:9537-9552.
ROYER, T. C.
1978. Ocean eddies generated by seamounts in the North
Pacific. Science (Wash., D.C.) 199:1063-1064.
464
BOEHLERT and SASAKI: PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD
RoYER, T. C. AND W. J. Emery.
1987. Circulation in the Gulf of Alaska, 1981. Deep-Sea Res.
34:1361-1377.
Sasaki, T.
1974. Kita Taiheiyo no kusakari tsubodai (The pelagic armor-
head, Pentaceros richardsoni Smith, in the North Pacific).
Bull. Jpn. Soc. Fish. Oceanogr. 24:156-165. (Engl, transl.
by T. Otsu, 1977, 13 p., Transl. No. 16; available Southwest
Fish. Cent. Honolulu Lab., Natl. Mar. Fish. Serv., NOAA,
Honolulu, HI 96822-2396.)
1986. Development and present status of Japanese trawl
fisheries in the vicinity of seamounts. In R. N. Uchida,
S. Hayasi, and G. W. Boehlert (editors). Environment and
resources of seamounts in the North Pacific, p. 21-30. U.S.
Dep. Commer., NOAA Tech. Rep. NMFS 43.
Seckel, G. R.
1988. Indices for mid-latitude North Pacific winter wind
systems; an exploratory investigation. GeoJournal 16:97-
111.
Shimazaki, K., and S. Nakamura.
1981. Ecological studies of the pomfret (Brama japonica
Hilgendorf). I. The seasonal distribution pattern and eco-
logical considerations. Res. Inst. N. Pac. Fish., Spec. Vol.,
p. 91-103.
Sloan, N. A.
1983. Canadian-Japanese experimental fishery for oceanic
squid off British Columbia, summer 1983. Can. Ind. Rep.
Fish. Aquat. Sci. 152, 42 p.
Smith, J. G.
1965. A second record of the pelagic armorhead, Pentaceros
richardsoni Smith, 1849, in California waters. Calif. Fish
Game 51:213-214.
Takahashi, Y., and T. Sasaki.
1977. Kita Taiheiyo chubu kaizan ni okeru tororu gyogyo
(Trawl fishery in the central North Pacific seamounts).
Northern Waters Groundfish Resour., Far Seas Fish. Res.
Lab., 45 p. (Engl. Transl. by T. Otsu, 1977, 49 p.; Transl.
No. 22; available Southwest Fish. Cent. Honolulu Lab., Natl.
Mar. Fish. Serv., NOAA, Honolulu, HI 96822-2396.)
Uchida, R. N., and D. T. Tagami.
1984. Groundfish fisheries and research in the vicinity of sea-
mounts in the North Pacific Ocean. Mar. Fish. Rev. 46(2):
1-17.
Uda, M., and K. Hasunuma.
1969. The eastward subtropical countercurrent in the west-
ern North Pacific Ocean. J. Oceanogr. Soc. Jpn. 25:201-
210.
Wagner, E. J., and C. E. Bond.
1961. The percoid fish Pseudopentaceros richardsoni from
Oregon waters. Fish. Comm. Oreg., Res. Briefs 8(l):71-73.
Walker, M. M.
1984. Magnetic sensitivity and its possible physical basis in
the yellowfin tuna, Thunniis albacares. In J. D. McCleave,
G. P. Arnold, J. J. Dodson, and W. H. Neill (editors).
Mechanisms of migration in fishes, p. 125-142. Plenum
Press, N.Y.
Walker, M. M., J. L. Kirschvink, and A. E. Dizon.
1985. Magnetoreception and biomineralization of magnetite.
Fish. In J. L. Kirschvink, D. S. Jones, and B. J. MacFad-
den (editors). Magnetite biomineralization and magneto-
reception in organisms, p. 417-437. Plenum Press, N.Y.
Wedgeworth, B., and J. Kellogg.
1987. A 3-D gravity-tectonic study of Ita Mai Tai Guyot: an
uncompensated seamount in the East Mariana Basin. In
B. Keating, P. Frj-er, R. Batiza, and G. W. Boehlert (editors),
Seamounts, islands, and atolls, p. 73-84. Geophys. Monogr.
43.
Welander, a. D., R. C. Johnson, and R. A. Hajny.
1957. Occurrence of the boarfish, Pseudopentaceros richard-
soni, and the zeid, Allocyttus verrucosus, in the North
Pacific. Copeia 1957:244-246.
Wetherall, J. A., and M. Y. Y. Yong.
1986. Problems in assessing the pelagic armorhead stock on
the central North Pacific Seamounts. In R. N. Uchida,
S. Hayasi, and G. W. Boehlert (editors), Environment and
resources of seamounts in the North Pacific, p. 73-85. U.S.
Dep. Commer., NOAA Tech. Rep. NMFS 43.
Williams, F.
1972. Consideration of three proposed models of the migra-
tion of young skipjack tuna {Katsuwoniis pelamis) into the
eastern Pacific Ocean. Fish. Bull., U.S. 70:741-762.
Zama, a., M. Asai, and F. Yasuda.
1977. Records of the pelagic armorhead, Pentaceros richard-
soni from Hachijo Island and the Ogasawara Islands. Jpn.
J. Ichthyol. 24:57-60.
465
THE TIMING AND SIGNIFICANCE OF DENSITY-DEPENDENT AND
DENSITY-INDEPENDENT MORTALITY OF AMERICAN SHAD,
ALOSA SAPIDISSIMA
Thomas F. Savoy and Victor A. Creccoi
ABSTRACT
We used stock-recruitment, pre- and postrecruitment mortality data for American shad, Alosa
sapidissima, in the Connecticut River to estimate density-dependent and density-independent mortality
rates at the prejuvenile (age 1-100 days) and postjuvenile (age 101 days to 5 years) phases. Total post-
juvenile mortality rates from 1967 through 1982 were dominated by density-independent mortality, with
only 6% (^opost = 0.30) of the mean total (Z = 4.85) being ascribed to density-dependent processes. By
contrast, 23% {Z^ = 1.13) of the total prejuvenile mortality was compensatory, of which nearly all
took place during the embryonic and early larval periods. Egg and early larval mortality rates from 1979
through 1987 were positively correlated to June river flows, and inversely related to mean June
temperature, whereas mortality rates of all other life stages showed no such relationships. Daily egg
and early larval mortality rates from 1979 through 1987 were usually higher and more variable than
mortality rates during later stages. There were significant negative correlations between egg and early
larval mortality rates and the strength of 1979-87 year classes in the adult stock, whereas mortality
rates of late larvae and juveniles were independent of year-class strength. Density-dependent mortality
during the egg and early larval stages comprised over 40% of the total mortality at those stages, resulting
in the number of midlarvae and juveniles being positively correlated to adult recruitment. These data
suggest that year-class strength of American shad in the Connecticut River is established after the egg
and larval stages.
It is generally accepted that year-class strength of
most fishes is established prior to the juvenile stage
from density-independent (climatic factors) and
density-dependent (competition, predation, cannibal-
ism) processes (Gushing 1974; May 1974; Goodyear
1980). Although density-dependent mortality is
believed to be the mechanism that keeps fish popu-
lations stable under low to moderate exploitation
(McFadden 1977; Gushing 1980), density-dependent
processes such as inter- and intraspecific competi-
tion, predation, and cannibalism are difficult to
measure. This is particularly so for many highly
fecund fishes, whose population sizes undergo wide
fluctuations that often cannot be predicted by con-
ventional stock-recruitment models (Parrish and
MacGall 1978; Bakun 1984). Despite the acknowl-
edged importance of density-dependent mortality to
stock-recruitment theory (Ricker 1975; Ware 1980),
few studies have been able to quantify density-
dependent mortality, or determine the life stages
at which compensation occurs.
Density-dependent mortality may be confined to
the prejuvenile stage (Gushing 1974, 1980) or the
'Connecticut Department of Environmental Protection, Marine
Fisheries Office, Waterford, CT 06385.
postjuvenile oceanic stage (Peterman 1978, 1982),
or may occur throughout the prerecruitment period
(Gulland 1965). The testing of these hypotheses has
proceeded slowly because of environmental noise
surrounding many stock-recruitment relationships,
measurement errors associated with recruitment
estimates (Ludwig and Walters 1981), and the lack
of long-term mortality data on eggs and larvae.
One species, for which accurate and long-term
stock-recruitment and prerecruitment mortality
data exist, is the American shad, Alosa sapidissima,
an anadromous clupeid that spawns in many Atlan-
tic coast rivers (Walburg and Nichols 1967). Pre-
vious studies on American shad in the Gonnecticut
River have demonstrated that larval and juvenile
mortality rates decline with age (Grecco et al. 1983,
1986) and that growth and survival rates among
discrete larval cohorts are significantly affected by
short-term hydrographic and meteorological events
(Grecco and Savoy 1985b, 1987a). Although pre-
vious stock-recruitment studies of Gonnecticut River
shad (Leggett 1977; Lorda and Grecco 1987) found
evidence of density-dependent mortality before the
juvenile stage, no studies have attempted to esti-
mate density-dependent mortality during the egg
and larval stages, or quantify density-dependent and
Manuscript accepted March 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
467
FISHERY BULLETIN: VOL. 86, NO. 3
density-independent mortality after the juvenile
stage.
In this study we used stock-recruitment, pre- and
postrecruitment mortality data for American shad
in the Connecticut River from 1967 through 1987
to 1) estimate egg mortality rates for the 1979-87
year classes; 2) measure the contribution of density-
dependent and density-independent mortality before
and after the juvenile stage; and 3) determine the
life stage(s) at which most of the density-dependent
mortality takes place.
METHODS
Data Source
Estimates of adult recruitment and parent stock
size (± SE) from 1967 to 1982 (Table 1) were based
on annual population estimates derived by earlier
mark-recapture studies (Leggett 1976; Crecco and
Savoy 1987^) and by the annual mean number of
American shad lifted over the Holyoke Dam (Crecco
and Savoy 1985a). The parent stock size (PAR J
from 1967 through 1987 was the annual population
estimate of female shad minus that year's commer-
cial catch of female shad. Parent stock (PAR^) was
separated into the parent stock that spawned above
the Holyoke Dam (PJ (Moffitt et al. 1982; O'Leary
and Booke 1986^, 1987^) and the spawning stock
that spawned below the dam (P,, = PAR, - P„).
Since female American shad mature between ages
four and six, female shad recruitment from the
1967-82 year classes was the sum of virgin 4-, 5-,
and 6-yr-old female shad in the 1970-87 runs based
on the age-class structure from previous studies
(Jones et al. 1976; Leggett 1976; Crecco et al.
1984^). Since the sex ratio of mature progeny from
each year class is close to 1:1 (Leggett 1976), total
recruitment (Rf) from the 1967-82 year classes was
estimated by doubling the female shad recruitment
estimates. Direct estimates of male shad recruit-
ment are biased by gill net selectivity and differen-
tial culling practices of commercial fishermen
(Crecco et al. 1984^).
^Crecco, V. A., and T. F. Savoy. 1987b. Fishery Management
Plan for American shad in the Connecticut River. Unpubl.
manuscr., 117 p. Connecticut Department of Environmental Pro-
tection, Hartford, CT 06106.
'O'Leary, J., and H. E. Booke. 1986. Connecticut River anad-
romous fish investigations. Mass. Coop. Fish. Res. Unit Proj. Per-
formance Rep., F-45-R-2, 37 p.
"O'Leary, J., and H. E. Booke. 1987. Connecticut River anad-
romous fish investigations. Mass. Coop. Fish. Res. Unit Proj. Per-
formance Rep., 32 p.
^Crecco, V. A., T. Savoy, and L. Gunn. 1984. Population
dynamics studies of American shad in the Connecticut River. CT
Dep. Environ. Prot. Final Rep. AFC 13, 76 p.
•^Crecco, V. A., L. Gunn, and T. Savoy. 1981. Connecticut
River shad study, a progress report. Unpubl. manuscr., 87 p.
Connecticut Department of Environmental Protection, Hartford,
CT 06106.
Table 1.— Estimates of total female American shad parent stock, female spawn-
ing stock above and below the Holyoke Dam, mean June flows from 1967 to 1987,
and total number of adult recruits from the 1967-82 year classes. SE = stand-
ard errors about the estimates x 10^.
Total
Spawning
Spawning
spawning
stock
stock
Recruit-
June
stock
above
below
ment
flow
Year
X 10^
SE
X 10^
SE
X 10^
SE
X 10^
SE
m^/s
1967
167
35
4
0.8
163
34
444
112
437
1968
202
38
5
0.9
197
37
236
59
603
1969
384
70
10
1.8
374
68
490
116
375
1970
413
95
12
2.8
401
92
550
126
243
1971
424
111
14
3.7
410
107
982
205
203
1972
167
40
8
1.9
159
38
430
87
616
1973
111
30
3
0.8
108
29
308
70
534
1974
306
62
10
2.0
296
60
652
139
334
1975
247
65
16
4.2
231
61
560
101
379
1976
435
80
166
31.0
269
49
650
105
286
1977
207
39
112
21.0
95
18
1,240
191
250
1978
210
54
45
12.0
165
42
714
110
488
1979
248
51
87
18.0
161
33
882
127
445
1980
341
58
196
33.0
145
25
1,256
196
201
1981
293
50
143
24.0
150
26
758
132
316
1982
501
67
109
15.0
392
52
282
49
643
1983
423
77
185
34.0
238
43
386
1984
610
67
245
27.0
365
40
661
1985
555
96
219
38.0
336
58
311
1986
392
71
175
32.0
217
39
439
1987
235
37
136
14.0
146
23
250
468
SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD
American shad larvae (10-30 mm TL) were sam-
pled during daylight hours in the Connecticut River
with a 6.1 m plankton bag seine (2.4 m deep, wing
and bag mesh of 0.505 mm) and 30 m lead ropes
from 15 May to 20 July 1979-84, 1986, and 1987.
No larval sampling was conducted in 1985. One seine
haul was taken weekly at 8-12 fixed stations located
throughout the major spawning areas (Fig. 1). Fur-
ther details on sampling and methods of estimating
larval abundance from net samples are contained
in Crecco et al. (1983).
Juvenile American shad (40-90 mm TL) were col-
lected weekly from 20 July through 15 October
1967-87 at 7-14 fixed stations located above and
below the Holyoke Dam (Fig. 1) based on weekly and
biweekly seine surveys (Scherer 1974; Foote 1976;
Marcy 1976; Crecco and Savoy 1984). The annual
juvenile index of relative abundance (IND() from
1967 to 1987 was the mean juvenile catch per seine
haul ( ± SE) from all stations and collection dates
(Table 2).
The Timing of
Density-Dependent Mortality
Analysis
One of the primary objectives of this study was
to determine the magnitude and timing of density-
dependent mortality for American shad. Peterman
(1978, 1982) found that density-dependent mortality
for some stocks of the anadromous sockeye salmon,
Oiworhynchus nerka, was confined mainly to the 4-5
yr oceanic postjuvenile phase. Given that the Ameri-
STATIONS
X - L A RVA L
0-JUVENILE
km 1 90
SUN D ERLU N D
HATFIELD
HO LYOK E
km 140
X AG AW AM
ENFIELD
Figure 1.— Location of larval and juvenile
American shad sampling stations on the Con-
necticut River.
MASS
CONN
WILSON O
GLASTON BURY
km 25
O SALMON
RIVER
DEEP
RIVER
ESSEX
k m 1 2
469
FISHERY BULLETIN: VOL. 86, NO, 3
Table 2.— Mean juvenile indices, scaled juvenile abundance from
Equation (2) and total egg production for American shad in the Con-
necticut River from 1967 through 1987. SE = standard errors
about the estimates.
Juvenile
Total egg
Juvenile
abundance
SE
production
SE
Year
index
20.2
SE
9.4
X 10^
X 10^
X 10«
X 10^
1967
71,070
33,072
334
70
1968
11.1
1.7
39,053
5,981
404
76
1969
19.0
3.8
68,848
13,770
768
140
1970
27.8
8.2
97,809
28,850
826
190
1971
65.7
14.1
231,151
49,608
848
222
1972
15.3
2.9
53,830
10,203
334
80
1973
12.7
3.4
44,682
11,962
222
60
1974
21.4
6.3
75,292
22,165
612
124
1975
23.7
5.7
83,384
20,054
494
130
1976
22.4
5.9
78,810
20,760
870
160
1977
--NOT AVAILABLE- -
414
78
1978
27.2
5.9
95,698
20,758
420
108
1979
19.6
3.2
68,959
11,259
496
102
1980
42.7
11.0
150,231
38,701
682
116
1981
16.0
2.7
56,293
9,499
586
100
1982
4.7
1.1
16,536
3,870
1,002
134
1983
26.3
6.6
92,531
23,221
846
154
1984
13.0
2.1
45.738
7,388
1,220
134
1985
17.8
2.7
62,626
9,499
1,110
192
1986
17.0
2.8
59,811
9,851
784
142
1987
44.7
15.1
157,268
53,126
470
74
can shad has a similar 4-6 yr postjuvenile phase as
sockeye salmon, we tested Peterman's hypothesis
for American shad with key factor analysis (Bellows
1981; Rosenberg and Doyle 1986). We fixed the
postjuvenile period for American shad in the Con-
necticut River between 101 days and 5 years, corre-
sponding to the average age (101 days) at which
juvenile shad leave the river (Crecco and Savoy
1985b) and the average age (5 years) when they
return to the river as mature adults. The total post-
juvenile (age 101 days to 5 years) mortality rates
(ZA() for the 1967-82 year classes were related to
the scaled juvenile indices (J<) for those year
classes (Table 2) in a linear model:
ZA, =
a +
b{Jt),
(1)
where ZAf = -logg(i2(/J<). If significant density-
dependent mortality is present during the post-
juvenile stage, the slope (6) of Equation (1) would
be positive and differ significantly (P < 0.05) from
zero.
The juvenile indices were scaled to thousands of
fish (Jf) so that the y-axis intercept in Equation (1)
directly estimates the mean density-independent
mortality (Zj^g^i), and the slope (b) times the geo-
metric mean juvenile abundance (GM) from 1967
through 1987 is the mean density-dependent mor-
tality rate (Z^post). Total juvenile abundance (J()
was estimated by multiplying the juvenile indices
(IND,) by a scalar (SC):
SC = [GR exp{-EZA) ■ GM] = 3,518.3, (2)
where GR is the geometric mean total adult
recruitment for 1967 through 1982; GM is the geo-
metric mean juvenile index from 1967 through 1987;
EZA (4.85) is the mean total instantaneous mortality
among postjuveniles from 1967 through 1982. We
estimated EZA as the sum of mortality during the
late juvenile period (age 101-365 days), and the
subadult stage (age 1-5 years). The mean total in-
stantaneous mortality rate during the late juvenile
stage was estimated as 2.65 (0.01 • 265 days), using
a mean daily mortality rate of 0.01 (SE = 0.002)
extrapolated from the 1979-84 larval and juvenile
survivorship curves (Crecco and Savoy 1985b). The
mean total instantaneous mortality rate of subadult
shad was 2.2, based on an annual instantaneous
natural mortality rate of 0.45 from the method of
Pauly (1980) plus 0.10 to reflect oceanic fishing mor-
tality (2.2 = 4(0.45 -I- 0.10)). We estimated the an-
nual natural mortality rate (0.45) by substituting the
;^(0.25, SE = 0.03) and L (55 cm FL, SE = 3 cm)
parameters of the von Bertalanffy equation for male
and female shad combined and preferred ocean tem-
perature (14 °C) of American shad (Leggett and
Whitney 1972) into Pauly' s multiple regression
model. The oceanic fishing mortality estimate (0.10)
was based on tagging studies in Delaware Bay
(White et al. 1969; Zarbock 1969) and off the New
York-New Jersey coast (Nichols 1958).
To determine if density-dependent mortality takes
place during the egg and larval stages as Cushing
(1980) hypothesized, we related total prejuvenile
mortality rates {ZEJt) to annual egg production
(EggSj ), and to both egg production (Table 2) and
mean June river flow (JFLOW) from 1967 to 1987
(Table 1) in linear regression models:
ZEJt = a -H 6 (Eggs,)
(3a)
and
ZEJt = a + ^(Eggs,) -H c(JFLOW,), (3b)
where ZEJt = -log^(J( /Eggs,). Mean June river
flows (m^/s) were included in Equation (3b) because
previous studies (Crecco and Savoy 1985b, 1987a)
have shown that high June flows reduced prejuvenile
survival rates, leading to a significant inverse cor-
relation (r = - 0.74, P < 0.01) between the juvenile
indices of abundance from 1967 through 1980 and
470
SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD
mean June river flow (Crecco and Savoy 1984,
1985b). Mean June flows (m^/s) were measured
within the major spawning areas (Leggett 1977) by
the United States Geological Survey (U.S. Geo-
logical Survey 1967-84). We estimated total egg
production (Eggs,) in Equations (3a) and (3b) as the
product of the mean fecundity of a female American
shad times that year's parent stock (PARf) of
female shad (Table 1). The average fecundity was
reduced from 269,000 ova (Leggett 1969) to 200,000
to reflect the average rates of egg retention and in-
complete fertilization (Watson 1970; Reed and Russo
1976^). Since Leggett (1969) showed that the aver-
age fecundity of American shad varied by less than
10% from 1966 through 1973, we were justified in
using an average fecundity for all years.
The mean density-independent mortality rate
(Z/pre) present during the prejuvenile stage was the
y-axis intercept (a) in Equation (3b) plus the slope
(c) times the overall geometric mean June flow
'Reed, R. J., and A. Russo. 1976. American shad research
Connecticut River, Massachusetts, 1976. 1. Fecundity, egg reten-
tion, sex ratio, and age class composition. Unpubl. manuscr., 16
p. Massachusetts Cooperative Fisheries Research Unit, Univer-
sity of Massachusetts, Amherst, MA 01003.
(GJFLOW) from 1967 through 1987:
Zipre = a + c(GJFLOW),
(4)
The mean density-dependent mortality rate (Zq^^^)
was the slope (6) of Equations (3a) and (3b) times
the geometric mean egg production (GEgg) from
1967 through 1987. A positive and statistically sig-
nificant slope (6) of prejuvenile mortality on egg
abundance would support the Gushing hypothesis
that density-dependent mortality takes place before
the juvenile stage. If the b estimates of Equations
(1), (3a), and (3b) were all positive and significant,
our results would support Gulland's (1965) hypoth-
esis that density-dependent mortality occurs over
the entire prerecruitment phase.
Results
Although total juvenile abundance varied 14-fold
from 1967 through 1982 (Table 2), the total post-
juvenile mortality rates {ZA() exhibited relatively
low variability (95% G.I.: 4.70-5.00) about the esti-
mated mean {EZA = 4.85) (Fig. 2). The slope of the
linear regression between total postjuvenile mortal-
6.6
5.4
5.2
5.0
<
N 4.8
4.6
4.4
4.2
4.0
50
1 1 1
100 160 200
Juvenie Abundance (thousands)
260
300
Figure 2.— Relationship between American shad postjuvenile mortality (ZA,) and relative abundance of juveniles
(J() from 1967 through 1982.
471
FISHERY BULLETIN: VOL. 86, NO. 3
ity rates (ZAt) and scaled juvenile abundance (J,)
was positive and statistically significant (P < 0.04)
(Table 3). Since the average density-dependent
mortality (Z^post = 0.30) present during the post-
juvenile stage comprised only 6% of the mean total
postjuvenile mortality rate {ZA = 4.85), density-
dependent mortality is not large enough to alter the
significant linear correlation (r = 0.78, P < 0.01)
betv^^een the juvenile indices (J^ ) and adult recruit-
ment (Rf) for the 1967 through 1982 year classes
(Fig. 3).
The prejuvenile mortality rates (ZEJ,) from 1967
through 1987 v^ere positively correlated (r = 0.70,
P < 0.0004) to total egg production (EggS() (Fig. 4).
Egg production alone explained 49% of the varia-
tion in prejuvenile mortality (Table 3) and the slope
Table 3.— Estimates of density-dependent (Zp), density-independent mortality (Z,) and the fraction of density-
dependence (ZqIZi^Zq) during the postjuvenile (1967-82) and prejuvenile (1967-87) mortality phases for
American shad in the Connecticut River. SE = standard error, and numbers in parentheses = 95% C.I. for
Zq,ZEJ, = prejuvenile mortality, Z4, = postjuvenile mortality.
Model
Parameters
SE
Z, Zol{Z,,Z^)
Prejuvenile
ZEJ, = a + b(Eggs,)
ZEJt = a + b(Eggs,)
+ c (J Flow,)
Postjuvenile
ZA, = a + b(J,)
a = 5.466
b = 1.88 X 10"^
r^ = 0.49
a = 4.287
b = 1.97 X 10"^
c = 0.0028
r^ = 0.80
a = 4.53
b = 3.49 X 10"
r^ = 0.275
0.313
4.43 X 10"^
0.303
2.87 X 10"^
0.0005
0.15
1.51 X 10"^
1.13 5.47 0.17
(0.862-1.392) (0.136-0.203)
1.18 5.34 0.18
(1.007-1.351) (0.158-0.202)
0.30 4.53 0.062
(0.035-0.560) (0.008-0.110)
1300
« 1025
c
<0
(0
3
o
I 750
u
3
< 475
200
— n
10
—J-
20
— I—
30
— r-
40
— I—
50
— I—
60
70
Juvenle Index
Figure 3.— Relationship between total adult recruitment (R,) and the juvenile abundance indices (/ATZ), ) of American
shad from 1967 through 1982.
472
SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD
of the regression differed significantly from zero,
indicating the presence of significant density-
dependent mortaHty during the prejuvenile stage.
When prejuvenile mortality rates were related to
both egg production (Eggs,) and mean June river
flow, (JFLOW,), the multiple regression model ac-
counted for 80% of the variability in prejuvenile
mortality (Table 3), and the slope estimates for egg
production (6) and June flow (c) were positive and
highly significant. Note that the standard error (SE)
about the slope estimate (6) was reduced by 60%
when June flow effects were considered. These
results suggest that prejuvenile mortality rates are
affected by a combination of density-dependent (egg
production) and density-independent (June river
flow) factors.
The mean density-dependent mortality rate dur-
ing the prejuvenile phase (Z^pj.^ = 1.18) from the
multiple regression model was four times greater
than the mean Z^post value (0.30) for postjuveniles
(Table 3), suggesting that 80% (1.18/1.18 + 0.30)
of the total compensatory reserve for American shad
occurs before the juvenile stage. Whereas compen-
satory density-dependent mortality may play a sig-
nificant role in regulating egg and larval abundance
during years of high egg production, it is clear that
most of the variability (82%) in egg and larval abun-
dance is ascribed to density-independent factors.
Egg, Larval, and Juvenile
Mortality, 1979-87
Analysis
To examine how larval and juvenile mortality
rates varied with year-class strength, survivorship
curves for 19,000-180,000 American shad larvae
and 800-3,500 juveniles were developed annually
from 1979 to 1987 following the techniques of Lough
(1976) and Hewitt et al. (1985). Mortality rates could
not be estimated directly for prolarvae because
American shad yolk-sac larvae (7-9 mm TL) remain
in deep water (Marcy 1976) and were only partially
susceptible to the plankton seine.
All larvae and juveniles collected annually were
separated into four length intervals: 1) 10-13 mm
TL, reflecting first- feeding larvae with undeveloped
pelvic fins (Wiggins et al. 1984); 2) 14-19 mm TL,
associated with the onset of pelvic fin development
(Lippson and Moran 1974); 3) 20-28 mm TL, repre-
senting larvae approaching metamorphosis, char-
acterized by invagination of the gut (Maxfield 1953);
9.0
ao
W 7.0
6.0
6.0 i
T-
20
— r-
30
— r-
40
I
60
— I—
60
— r-
70
I
80
— I—
90
100
no
120
130
Total Eggs (bMons)
Figure 4.— Relationship between American shad prejuvenile mortality (ZEJ^) and the number of eggs produced
(Eggs,) from 1967 through 1987.
473
FISHERY BULLETIN: VOL. 86, NO. 3
and 4) 29-80 mm TL, corresponding to the juvenile
stage and the onset of scale development (Marcy
1976). These four groups are hereafter referred to
as early larval, midlarval, late larval, and juvenile
stages, respectively.
We used otolith increment counts (Savoy and
Crecco 1987) to age a subsample of 100-400 Ameri-
can shad larvae and 80-200 juveniles annually from
1979 to 1987. The age-length relationship for lar-
vae and juveniles in all years was well described by
the Gompertz growth equation (Crecco et al. 1983;
Crecco and Savoy 1985b) so we used a pooled
Gompertz equation to describe larval and juvenile
growth. The average age (t) of each of the four
length groups was estimated by rearranging the
Gompertz equation:
t = logAKIK - log,(L,(Lo))]/a
(5)
where L( = total length (mm); Lq = length at
hatching (8.0 mm); K = growth rate at the inflec-
tion point; and a = rate of exponential decay.
The daily instantaneous mortality rates (d) for
each stage and their standard errors from 1979
through 1987 were estimated by an exponential
model,
Nt = A exp{-dt),
(6)
that related abundance (A^^) and age (t) by non-
linear regression methods (SAS 1982). The total
instantaneous mortality rates of early (ZELf), mid
(ZMLf), and late (ZLLt) larvae and juveniles (ZJf)
were estimated by multiplying the corresponding
daily mortality rate (d) by the average duration (Ds
= days) within each stage. Previous studies (Crecco
and Savoy 1985a; Savoy and Crecco 1987) found
that the growth rates of early (10-13 mm) and mid-
larvae (14-19 mm) were positively correlated with
June water temperatures, whereas the growth of
late larvae and juveniles were independent of river
temperatures. As a result, we adjusted the stage
duration {Ds ) of early and midlarvae with the mean
June temperatures (U.S. Geological Survey 1979-
87) from 1979 through 1987 and the larval growth-
subtracting the sum of larval {ZEL, , ZMLf , ZLL,),
juvenile (ZJ^) and postjuvenile (ZA^) mortality rates
from the total prerecruitment mortality rate
(ZTotal,):
ZEt = ZTotalt - [ZELi + ZML, + ZLLt
+ ZJt + ZAt), (7)
where ZTotal^ = -logp(/2,/Eggs,). However, since
the 1983 through 1987 year classes have not been
fully recruited to the spawning population, we esti-
mated total adult recruitment {Rp^ ) for those year
classes by the following environment-dependent
stock-recruitment model:
Rpt = 24.29 (PARt) exp(- 0.0052 • P,,)
exp(- 0.0032 • PJ
exp(- 0.0025 • JFLOW()
(8)
Adult recruitment was estimated independently
of the juvenile indices by substituting each
year's parent stock size (PAR^), mean June flow
(e/FLOW), female parent stock lifted over the Hol-
yoke Dam (P„ ), and female parent stock below the
Holyoke Dam (P^) (Table 1) into the model. This
nonlinear model was shown (Lorda and Crecco 1987)
to be a good predictor (r^ = 0.81, P < 0.001) of
adult recruitment (P^) for the 1966 through 1982
year classes. Moreover, the predicted recruitment
levels (P() of the 1983 through 1987 year classes
from Equation (8) were closely correlated (r = 0.92,
P < 0.01) with the corresponding juvenile abundance
(J() for those years (Table 4) which is consistent
with the positive correlation (r = 0.78, P < 0.01)
between adult recruitment and the 1967-82 juvenile
indices (Fig. 2). This justifies the use of Equation
(8) to predict adult recruitment, total mortality, and
postjuvenile mortality rates for the 1983-87 year
classes.
The standard errors about the egg mortality rates
(ZEt) were derived as the sum of the variances of
all other terms (Cochran 1965):
SEze' = ^SEzel' + SE^ml' + S^zll' + ZJ^ + SE^/ + SE^xotai' 0)
temperature equations (Crecco and Savoy 1985b:
table 7).
We estimated total egg and prolarval mortality
rates (ZE^) indirectly from 1979 through 1987 by
The standard errors about the total prerecruitment
(ZTotal,) and postjuvenile (ZA^) mortality rates
(Table 5) were based on the same principle as Equa-
tion (9) (App. 1).
474
SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD
Table 4— Predicted total adult recruitment of
American shad from the environmental dependent
stock-recruitment model (Equation (8)), 95% con-
fidence limits about the recruitment values and
juvenile abundance for the 1983-87 year classes.
Predicted
adult
Juvenile
recruitment
abundance
Year
X 10^
95% C.I.
X 10^
1983
634
362-906
92,531
1984
196
112-280
45.738
1985
555
317-793
62,626
1986
587
335-839
59,811
1987
955
547-1,363
157,268
Table 5.— Estimates of the total (ZTotal,) and
postjuvenile (Z4,) instantaneous mortality rates
and their standard errors (SE) for American
shad from 1979 through 1987.
Year
ZTotal,
SE
ZA,
SE
1979
10.94
0.26
4.36
0.23
1980
10.90
0.24
4.78
0.32
1981
11.26
0.25
4.31
0.25
1982
13.00
0.27
4.77
0.34
1983
11.80
0.30
4.98
0.36
1984
13.34
0.26
5.45
0.28
1985
12.21
0.29
4.72
0.28
1986
11.80
0.30
4.62
0.29
1987
10.80
0.28
5.10
0.47
Given that ZE represents total mortality through-
out the egg and prolarval period, daily egg mortal-
ity rates were determined by dividing ZE by the
average duration (Ds) of the egg and prolarval
stages of American shad in the Connecticut River.
Watson (1968) reported an inverse relationship be-
tv^^een the incubation period (D ) of shad eggs and
water temperature (T) by the expression:
D = 120.95 exp(-0.154 • T).
(10)
To determine the duration {Ds) of the egg stage
from 1979 through 1987 we substituted the mean
June temperatures for the Connecticut River
(U.S.G.S. Annual Water Year Reports 1979-87) into
Equation (10).
The total number of prolarvae (age 2 days) and
early larvae (age 10 days) for the 1979 through 1987
year classes was estimated by
NE^ = Eggs, • exp(-Z£;,) (11)
and
NFt = Eggs, • exipi-ZEt - ZELi) (12)
respectively, where NEt is the estimated number
of prolarvae larvae and NFt is the number of early
larvae. The abundance of older larvae and juveniles
from each year class was estimated by adding their
respective total mortality rates to Equation (12). To
determine the life stage(s) at which year-class
strength is established, we related the stage-specific
total mortality rates and abundance estimates to the
number of adult recruits (i?, or Rpt) from the
1979-87 year classes in several linear models. If
year-class strength is established early, there should
be a significant positive correlation between the
number of prolarvae (NE) and early larvae (NF) and
adult recruitment, and total mortality rates during
these early stages {ZE and ZEL ) should be inverse-
ly related to adult recruitment.
To determine the extent of density-independent
mortality at the egg, larval, and juvenile stages, we
correlated the stage-specific total mortality rates
from 1979 through 1987 to mean May and June river
flows (m^/s) and water temperatures (°C) in sev-
eral linear models. May and June hydrographic
and meteorological parameters were used because
they coincide with egg, larval, and juvenile develop-
ment in the Connecticut River (Leggett 1977) and
were the only monthly abiotic variables that were
significantly linked to adult shad recruitment from
1966 to 1980 (Crecco and Savoy 1984, 1987c). May
and June water flows and temperatures were re-
corded by the U.S. Geological survey (U.S. Geo-
logical Survey 1966-1980) within the major spawn-
ing areas (river km 89) of American shad (Leggett
1977).
Results
The mean daily mortality rates from 1979 through
1987 declined by an order of magnitude from the
egg through the juvenile stages (Table 6). Total egg
mortality rates were relatively high (mean ZE =
2.584, cv = 18.5%) and were inversely correlated
(r = - 0.76, P < 0.03) with adult recruitment (Rt or
Rpt) from those year classes (Table 7). These data
indicate that shad eggs and prolarvae (age 1-2 days)
from 1979 through 1987 experienced high (85-96%)
mortality that was directly linked to year-class
success.
The total mortality rates (ZEL) among early lar-
vae were slightly lower and more variable (mean
ZEL = 1.608, cv = 31.5%) than the egg mortality
rates (Table 6), and were also inversely related (r
= - 0.83, P < 0.01) to adult recruitment from 1979
through 1987. By contrast, the total mortality rates
of older larvae {ZML and ZLL) and juveniles (ZJ)
475
FISHERY BULLETIN: VOL. 86, NO. 3
Table 6.— Stage-specific total instantaneous mortality rates of egg, larval, and juvenile
American sfiad from 1979 to 1987. SE = stage-specific standard errors; cv = coeffi-
cient of variation, Days = duration of each stage.
Instantaneous total mortality
Egg + pro-
Early
Mid-
Late
larvae
larvae
larvae
larvae
Juvenile
Year
or
(day 2)
(day
3-9)
(day 10-18)
(day 19-29)
(day 30
1-100)
statistic
mean
SE
mean
SE
mean
SE
mean
SE
mean
SE
1979
2.481
0.554
1.323
0.245
1.116
0.189
0.330
0.110
1.330
0.280
1980
1.853
0.819
1.218
0.186
1.096
0.336
0.693
0.099
1.260
0.280
1981
2.580
0.684
1.836
0.108
0.408
0.328
0.726
0.154
1.400
0.210
1982
2.940
0.501
2.144
0.152
1.160
0.210
0.726
0.187
1.260
0.210
1983
3.012
0.785
1.400
0.546
0.522
0.234
0.836
0.143
1.050
0.140
1984
2.706
0.601
2.520
0.352
0.640
0.330
0.484
0.242
1.540
0.280
1985
1986
N 0
0.196
DATA
1.206
1.330
1.120
0 pao
3.158
0.652
1.323
0.261
0.374
0.066
0.280
1987
1.944
0.852
1.098
0.180
0.864
0.128
0.464
0.088
1.330
0.210
2
2.584
1.608
0.877
0.579
1.286
SE
0.169
0.179
0.111
0.066
0.051
cv
0.185
0.315
0.359
0.325
0.120
Days
7.2
6.9
8.9
11.0
70.0
Daily Z
0.359
0.233
0.099
0.053
0.018
Table 7.— Correlations between stage-specific total mortalities and
adult recruitment and several abiotic factors from 1979 to 1987 for
American shad. P = probability levels.
Stage
Adult
recruit-
ment
May
flow
May
tempera-
ture
June
flow
June
tempera-
ture
Egg
p
-0.76
0.03
0.38
0.35
-0.08
0.85
0.66
0.08
-0.59
0.13
Early
P
-0.83
0.01
0.42
0.29
-0.21
0.62
0.81
0.01
-0.64
0.09
Mid
P
0.16
0.71
-0.41
0.32
0.49
0.22
0.09
0.83
-0.34
0.42
Late
P
-0.02
0.96
0.20
0.64
-0.51
0.20
-0.13
0.75
0.18
0.68
Juv
P
-0.19
0.65
0.05
0.91
0.02
0.97
0.26
0.53
-0.08
0.86
were highly variable, but showed no relationship to
adult recruitment.
Given that high (94-99%) egg and early larval
mortality rates were inversely related to year-class
strength, the abundance of first-feeding larvae (age
10 days) was closely correlated (r = 0.84, P < 0.01)
with adult recruitment (Rpt) from 1979 through
1987 (Table 8). Because year-class strength from
1967 through 1987 was independent (r = - 0.15, P
< 0.46) of egg production (EggS(), our results sug-
gest that year-class strength of American shad is
determined by the number of shad that survive the
embryonic and early larval stages.
Early larval mortality rates from 1979 to 1987
were positively correlated with mean June river
flows. No such relationships were evident for any
other stage-specific mortalities (Table 7). The high
egg and early larval mortality rates (Table 6) and
relative failure of the 1982 and 1984 year classes
coincided vdth major storm events in mid- June 1982
and late May 1984 (Crecco and Savoy 1984) which
increased river flows and kept water temperatures
below 17°C until late June. The high river flows in
1982 and 1984 were also coupled with the highest
parent stocks on record since 1967 (Table 1). By con-
trast, the relatively low egg and early larval mor-
tality rates for the dominant 1980 and 1987 year
classes were associated with low June river flows,
a steady rise in June water temperatures (Crecco
and Savoy 1984) and low to moderate size parent
stocks (Table 1). These results suggest that domi-
nant year classes of American shad are most likely
to occur when lower than normal June flows are
coupled with relatively small spawning stocks.
Prejuvenile Density-Dependent
Mortality: A Closer Look
Analysis
Having estimated that 82% of the density-
dependent mortality for American shad takes place
before the juvenile stage (Table 3), we attempted
to estimate the relative contribution of density-
dependent mortality during the egg, early, mid-, and
late larval and juvenile stages from 1979 through
1987. We related egg mortality rates (ZEi) from
476
SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD
Table 8.— Relationship between the number (x 10®) of American shad at each life history
stage and adult recruitment from each year class, r = correlation coefficient and P =
probability levels from 1979 to 1987.
Stage
Year-
Pro-
Early
Mid-
Late
class
Egg
larvae
larvae
larvae
larvae
Juvenile
Adult^
1979
496
41.5
11.0
3.6
2.6
0.7
882
1980
682
106.9
31.6
10.6
5.3
1.5
1,256
1981
586
44.4
7.1
4.7
2.3
0.6
758
1982
1,002
53.0
6.2
1.9
0.9
0.3
282
1983
846
41.6
10.3
6.1
2.6
0.9
634
1984
1,220
81.5
6.6
3.5
2.1
0.5
198
1985
1986
1,110
784
NO DATA-
555
587
33.3
8.9
2.7
1.8
0.6
1987
470
67.3
22.4
9.5
5.9
1.6
955
r =
-0.15^
0.33
0.84
0.80
0.79
0.82
P =
0.46
0.43
0.01
0.02
0.02
0.01
'Predicted or observed recruitment in thousands.
^Correlation coefficient for egg abundance and recruitment was based on the 1967-87 data (Tables
1.2).
1979 to 1987 to total egg production (Eggs^) in a
linear model:
ZEt = a + 6(Eggs,).
(13)
As before, significant density-dependence would be
shown if the slope (6) in Equation (13) was positive
and differed significantly from zero.
The cumulative amount of density-dependent mor-
tality during the egg and early larval stages com-
bined was estimated by summing the instantaneous
total mortality rates of early larvae (ZELf) and
eggs (ZEt) and then regressing the total {ZEf +
ZELf) against egg production as in Equation (13).
This procedure was repeated for each subsequent
stage by adding their respective instantaneous total
mortality rates. We then estimated the relative
magnitude of density-dependent mortality (Z^) at
each stage by multiplying the slope (6) of each
regression equation by the geometric mean egg pro-
duction (GEgg) from 1967 through 1987. The
density-independent mortality rate {Zj) at each
stage was expressed by the y-a.xis intercept (a) of
each regression equation. The percentage contribu-
tion of density-dependent mortality {%Zf)) for each
period was the ratio of Z^ to total mortality {Z^ +
Zj) times 100.
Results
Our results showed that significant density-
dependent mortality first occurs during the early
larval stage and persists for all stages thereafter
(Table 9). The percentage contribution of density-
dependent mortality i%Z^) rose from 24% of the
total during the egg stage to 41% during the early
larval stage, and then declined during the mid- and
late larval stages as the magnitude of density-
independent mortality increased. The mean density-
dependent mortality rate for the egg and early larval
Table 9.— Relationship between the cumulative mortality rates and parent stock sizes (Eggs,).
The cumulative mortality rate was egg mortality (ZE) plus early larval mortality (Z2), midlarval
mortality (Z3), late larval mortality (Z4), and juvenile mortality (Z5). Zg = density-dependent
mortality rate, Z, = density-independent mortality rate
o/oZo =
percentage density-dependent
mortality, %Z, = percentage density-independent mortality, t = student /-statistic.
Model
b
parameter
SE(b)
t
Zo
Zi
o/oZo
o/oZ,
ZE
= a
+ b(Eggs,)
9.89
X
10-3
6.40 X
10"
-9
1.55
0.59
1.83
24
76
Z2
= a
+ fa (Eggs,)
2.57
X
io-»
7.76 X
10"
-a
^3.31
1.54
2.24
41
59
Z3
= a
+ b(Eggs,)
2.44
X
io-«
8.16 X
10"
-y
'2.99
1.46
3.21
31
69
Z4
= a
+ 6(Eggs,)
2.60
X
io-«
7.74 X
10"
-y
'3.36
1.55
3.67
30
70
Z5
= a
+ b(Eggs,)
2.69
X
10"®
7.50 X
10"
-y
'3.59
1.61
4.88
25
75
'Significant student f-statistic at the P < 0.01 level and SE(b) = standard error of b.
477
stages combined (Zq = 1.54) between 1979 and
1987 was not significantly different (Table 9) from
the total density-dependent mortality rate (Zq^j.^ =
1.18) estimated during the prejuvenile period from
1967 to 1982 (Table 3), suggesting that nearly all
of the density-dependent mortality for prejuvenile
American shad occurs before the midlarval stage.
DISCUSSION
Although American shad eggs and early larvae
experience high mortality (15-40%/day) in the Con-
necticut River, the average density-dependent mor-
tality rate (Z^p^g = 1.18) during those stages com-
prised a relatively small percentage (18%) of the
total prerecruitment mortality. This suggests that
most of the annual variability in American shad
recruitment is explained by density-independent fac-
tors, which is consistent with the significant positive
correlation between mean June flow and egg and
early larval mortality rates, and with the significant
inverse correlation (r = - 0.74, P < 0.001) between
mean June river flow and adult recruitment from
the 1967 through 1982 year classes (Crecco and
Savoy 1984, 1987b). Whereas these data illustrate
that density-dependent mortality plays a minor role
in governing annual variability in American shad
recruitment, this does not mean that compensatory
processes are trivial. As pointed out by Ricker (1954)
and others (Gushing 1974; Garrod and Horwood
1984), only a small amount of density-dependent
mortality is required to stabilize the growth poten-
tial of fish populations because density-dependent
effects become progressively more effective at
higher egg and larval densities (Shepherd and
Gushing 1980; Murray 1982).
Although our results on American shad support
the Gulland (1965) hypothesis that density-depend-
ent mortality persists throughout the prerecruit-
ment period, most (82%) of the density-dependent
mortality occurs during the egg and early larval
stages. Our average estimate of density-dependent
mortality {Zq = 1.54) during the early larval
periods does not differ significantly from the mean
Zd value (1.34) estimated among the 5-d larval
cohorts in 1983 and 1984 (Grecco and Savoy 1987a),
or from the mean density-dependent mortality rate
Zj) (1.21) estimated by stock-recruitment methods
for the entire prerecruitment period (Lorda and
Grecco 1987). That year-class strength is established
early in the ontogeny of American shad is supported
further by the significant positive correlation (r =
0.84, P < 0.009) between the relative abundance of
early larvae and all subsequent stages from 1979
FISHERY BULLETIN: VOL. 86, NO. 3
through 1987 and adult recruitment from those year
classes (Table 8).
The main causes of density-dependent mortality
are thought to be predation, competition, and canni-
balism (Ricker 1954). Since adult American shad are
not thought to feed in freshwater (Walburg and
Nichols 1967), we can probably eliminate cannibal-
ism as a mechanism for significant density-depend-
ent mortality during the egg and early larval stages.
Therefore, density-dependent mortality among early
American shad larvae is most likely caused by intra-
specific competition for food or space and predation.
The exact underlying density-dependent mortality
mechanisms are difficult to quantify because the out-
come of competition and predation may depend on
June flow effects shown here (Table 3) and else-
where (Grecco and Savoy 1987a, b) to be the prin-
cipal density-independent factor. High June river
flows have been shown to reduce June river tem-
peratures (Grecco and Savoy 1984) and the growth
rates of shad eggs (Watson 1968) and early larvae
(Grecco and Savoy 1985b). Because slower growing
larvae may be susceptible to predation for a longer
period of time (Hunter 1976), periods of high flow
may indirectly enhance egg and larval predation.
Additionally, since high flows reduce the spatial
patchiness (Grecco and Savoy 1987a), abundance
and availability of river zooplankton (Whitton 1980;
Threlkeld 1986), high flows may result in increased
levels of competition among American shad larvae
for available prey, especially if shad larvae are
capable of depleting local aggregations of edible
zooplankton. Periods of high runoff that coincide
with peak larval production, such as in June 1982
and 1984, may advect larvae and their zooplankton
prey from eddies and backwaters where they are
normally found (Gave 1978) to areas of high pred-
ator abundance. In light of the many ways in which
June flows potentially mediate larval mortality, it
is unlikely that a single compensatory mechanism
is responsible for the relatively high density-depend-
ent mortality rate (Z^pre) among early larvae.
Since egg and prolarval shad have endogenous
food reserves, density-dependent mortality of these
stages is likely due to predation and competition for
sites among spawning adults. As spawning stocks
reach high densities, such as in 1982-84, crowding
of adult fish on the spawning grounds may result
in a reduction in the number of eggs released ((jood-
year 1980). In addition, since the amount of spawn-
ing habitat in the Connecticut River can be con-
sidered fixed from year to year, larger spawning
stocks are more apt to deposit an increasing per-
centage of eggs in unfavorable areas. Layzer (1974)
478
SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD
noted that the highest survival rates of American
shad eggs occurred over gravel and rubble sub-
strates. Since these areas generally have higher
water velocities, eggs lodged within the interstices
of the gravel and rubble are less likely to be eaten
by predators or covered by silt. Dense aggregations
of eggs may also be highly susceptible to fungal
agents (Leach 1925), which could then act as effi-
cient density-dependent predators. However, since
predators of American shad eggs and larvae in the
Connecticut River have not yet been identified,
density-dependent losses are difficult to quantify.
Also, predation mortalities in general may be either
density-dependent or density-independent, depend-
ing on the functional response between the predator
and the prey (Hassell 1978), and whether predators
actively search for discrete aggregations of eggs and
larvae or prey on randomly encountered eggs and
larvae (Gulland 1987).
In the absence of direct egg mortality estimates
of American shad, the accuracy of our egg mortal-
ity estimates {ZEt) from Equation (7) are difficult
to evaluate, particularly since the standard errors
about the estimates are highly variable (Table 6). To
provide an independent estimate of daily egg mor-
tality, we used the multiple regression method of
McGurk (1987) to estimate the average daily egg
mortality rate. This method requires an estimate of
the patchiness and the mean dry weight of a shad
egg. We derived Lloyd's patchiness index (x =
7.26, SE = 0.376) for American shad eggs in the
Connecticut River from the 1974-75 egg surveys in
the Holyoke Dam impoundment (NUSCo 1977). Sub-
stituting the mean weight of a shad egg (0.00025
g, Leggett 1969) and the mean patchiness index into
McGurk's equation 4, yields an instantaneous daily
egg mortality rate of 0.34 which closely approx-
imates our average estimate (0.36) from 1979 to
1987 (Table 6).
Lastly, since it is presently unknown if younger
(smaller) virgin female American shad produce
smaller, less viable eggs than older shad as was
reported for cod, Gadus morhua, (Knutsen and
Tilseth 1985), the contribution of the age structure
to the parent progeny relationship (Rosenberg and
Doyle 1986) should be considered. This issue and an
examination of whether egg deposition rates are
density-dependent should be addressed by monitor-
ing egg retention rates among postspawning Ameri-
can shad.
ACKNOWLEDGMENTS
We would like to thank Rita Lorenzetti-Langan,
Mary Payette, Thomas Stanford, and all the other
people who have contributed field and laboratory
time or helpful advice. We also thank the two
anonymous reviewers for their comments on the
manuscript.
LITERATURE CITED
Bakun, a.
1984. Report of the working group on environmental studies
and monitoring. In J. Csirlce and G. Sharp (editors),
Reports of the expert consultation to examine changes in
abundance and species composition of neritic fish resources,
p. 41-54. FAO Fish. Rep. 291 Vol. 1.
Bellows, T. S., Jr.
1981. The descriptive properties of some models for density-
dependence. J. Anim. Ecol. 50:139-156.
Cave, J. R.
1978. American shad, Alosa sapidissima, larval distribution,
relative abundance and movement in the Holyoke Pool, Con-
necticut River, Mass. M.S. Thesis, Univ. Massachusetts,
Amherst, MA, 65 p.
Cochran, W. G.
1965. Sampling Techniques. 2d ed. John Wiley & Sons,
Inc., N.Y., 413 p.
Crecco, V. A., AND T. Savoy.
1984. Effects of fluctuations in hydrographic conditions on
year-class strength of American shad {Alosa sapidissima)
in the Connecticut River. Can. J. Fish. Aquat. Sci. 41:
1216-1223.
1985a. Density-dependent catchability and its potential
causes and consequences on Connecticut River shad, Alosa
sapidissima. Can. J. Fish. Aquat. Sci. 42:1649-1657.
1985b. Effects of biotic and abiotic factors on growth and
relative survival of young American shad in the Connecticut
River. Can. J. Fish. Aquat. Sci. 42:1640-1648.
1987a. Effects of climatic and density-dependent factors on
intra-annual mortality of larval American shad. Am. Fish.
Soc. Symp. 2:69-81.
1987b. Recruitment mechanisms of the American shad, Alosa
sapidissima: test of the critical period and match-mismatch
hypothesis. Am. Fish. Soc. Symp. 1:455-468.
Crecco, V. A., T. Savoy, and L. Gunn.
1983. Daily mortality rates of larval and juvenile American
shad {Alosa sapidissima) in the Connecticut River with
changes in year-class strength. Can. J. Fish. Aquat. Sci.
40:1719-1728.
Crecco, V., T. Savoy, and W. Whitworth.
1986. Effects of density-dependent and climatic factors on
American shad, Alosa sapidissima, recruitment: a predic-
tive approach. Can. J. Fish. Aquat. Sci. 43:457-463.
Gushing, D. H.
1974. The possible density dependence of larval mortality and
adult mortality in fishes. In J. H. S. Blaxter (editor), The
early life history of fish, p. 21-38. Springer- Verlag, Berlin.
1980. The decline of the herring stocks and the gadoid out-
burst. J. Cons. Int. Explor. Mer 39:74-85.
FOOTE, P. S.
1976. Blood lactic acid levels and age structure of American
shad {Alosa sapidissima, Wilson) utilizing the Holyoke Dam
fish lift, MA, 1974 and 1975. M.S. Thesis, Univ.
Massachusetts, Amherst, MA, 97 p.
GaRROD, D. J., AND J. W. Horwood.
1984. Reproductive strategies and the response to exploita-
479
FISHERY BULLETIN: VOL. 86, NO. 3
TION. In G. W. Potts and R. J. Wooton (editors), Fish
reproduction: strategies and tactics, p. 367-384. Acad.
Press, London.
Goodyear, P. C.
1980. Compensation in fish populations. In C. H. Howatt
and J. R. Stauffer, Jr. (editors). Biological monitoring of fish,
p. 253-280. Lexington Books, Lexington, MA.
GULLAND, J. A.
1965. Survival of the youngest stages of fish, and its rela-
tion to year-class strength. Spec. Publ. ICNAF 6, p.
363-373.
1987. Natural mortality and size. Mar. Ecol. Prog. Ser. 39,
p. 197-199.
Hassell, M. p.
1978. The dynamics of arthropod predator-prey systems.
Princeton Univ. Press, Princeton, NJ, 237 p.
Hewitt, R. P., G. H. Theilacker, and N. C. H. Lo.
1985. Causes of mortality in young jack mackerel. Mar.
Ecol. Prog. Ser. 26, p. 1-10.
Hunter, J. R.
1976. Culture and growth of northern anchovy, Engraulis
mordax, larvae. Fish. Bull., U.S. 74:81-88.
Jones, R. A., P. Minta, and V. A. Crecco.
1976. A review of American shad studies in the Connecticut
River. In Proceedings Workshop on American Shad, p.
135-164. U.S. Natl. Mar. Fish. Ser.
Knutsen, G. M., and S. Tilseth.
1985. Growth, development, and feeding success of Atlantic
cod larvae Gadiis rrwrhua related to egg size. Trans. Am.
Fish. Soc. 114:507-511.
Layzer, J. B.
1974. Spawning sites and behavior of American shad, Alosa
sapidissima, (Wilson), in the Connecticut River between
Holyoke and Turners Falls, MA, 1972. M.S. Thesis, Univ.
Massachusetts, Amherst, MA, 46 p.
Leach, G. C.
1925. Artificial propagation of shad. Rep. U.S. Comm.
Fish., 1925, app. VnL459-486. (Doc. 981.)
Leggett, W. C.
1969. Studies on the reproductive biology of the American
shad (Alosa sapidissima). A comparison of populations
from four rivers of the Atlantic seaboard. Ph.D. Thesis,
McGill Univ., 125 p.
1976. The American shad, Alosa sapidissima, with special
reference to its migrations and population dynamics in the
Connecticut River. In D. Merriman and L. M. Thorpe
(editors). The Connecticut River ecological study: the impact
of a nuclear power plant, p. 169-225. Am. Fish. Soc.
Monogr. 1, 252 p.
1977. Density-dependent, density-independent, and recruit-
ment in the American shad (Alosa sapidissima) population
of the Connecticut River. In W. Van Winkle (editor). Pro-
ceedings of the conference on assessing the effects of jx)wer-
plant-induced mortality on fish populations, p. 3-17.
Pergamon Press, N.Y.
Leggett, W. C, and R. R. Whitney.
1972. Water temperature and the migrations of American
shad. Fish. Bull., U.S. 70:659-670.
Lippson, A. J., and R. L. Moran.
1974. Manual for the identification of early developmental
stages of fishes of the Potomac River estuary. MD Dep.
Natl. Resour., PPSR-MP-13, 282 p.
LoRDA, E., AND V. A. Crecco.
1987. Stock-recruitment relationship and compensatory mor-
tality of American shad in the Connecticut River. Am. Fish.
Soc. Symp. 1:469-482.
Lough, R. G.
1976. Mortality and growth of Georges Bank - Nantucket
Shoals herring larvae during the 1975-1976 winter period.
Int. Comm. Northwest At!. Fish. Resident Doc. 76/6/123,
Ser. No. 4004, 30 p.
LuDWiG, D., AND C. Walters.
1981. Measurement errors and uncertainty in parameters
estimates for stock and recruitment. Can. J. Fish. Aquat.
Sci. 38:711-720.
Marcy, B. C, Jr.
1976. Early life history studies of American shad in the lower
Connecticut River and the effects of the Connecticut Yankee
plant. In D. Merriman and L. M. Thorpe (editors). The Con-
necticut River ecological study: the impact of a nuclear power
plant, p. 141-168. Am. Fish. Soc. Monogr. I. 252 p.
May, R. C.
1974. Larval mortality in marine fishes and critical period
concept. In J. H. S. Blaxter (editor). The early life history
of fish, p. 3-19. Springer- Veriag, N.Y.
Maxfield, G. H.
1953. The food habits of hatchery-produced pond-cultured
shad Alosa sapidissima, reared to a length of two inches.
Chesapeake Biol. Lab. Publ. No. 98, 38 p.
McFadden, J. T.
1977. An argument supporting the reality of compensation
in fish populations and a plea to let them exercise it. In
W. Van Winkle (editor), Proceedings of the conference on
assessing the effects of power-plant-induced mortality on fish
populations, p. 153-183. Pergamon Press, N.Y.
McGuRK, M. D.
1987. Natural mortality and spatial patchiness: reply to
Gulland. Mar. Ecol. Prog. Ser. 39:201-206.
Moffitt, C. M., B. Kynard, and S. G. Rideout.
1982. Fish passage facilities and anadromous fish restoration
in the Connecticut River. Fisheries 7:2-11.
Murray, B. G., Jr.
1982. On the meaning of density dependence. Oecologia
(Beri.) 53:370-373.
Nichols, P. R.
1958. Effect of New Jersey-New York pound-net catches on
shad runs of Hudson and Connecticut Rivers. U.S. Fish
Wildl. Serv., Fish Bull. 58:491-500.
NUSCo (Northeast Utilities Service Company).
1977. Entrainment analyses of shad eggs in the Connecticut
River. Northeast Util. Serv. Co., 167 p.
Parrish, R. H., and a. D. MacCall.
1978. Climatic variation and exploitation in the Pacific
mackerel fishery. Calif. Dep. Fish. Game Fish Bull. 167,
110 p.
Pauly, D.
1980. On the relationships between natural mortality, growth
parameters, and mean environmental temperature in 175
fish stocks. J. Cons. Explor. Mer 39:175-192.
Peterman, R. M.
1978. Testing for density-dependent marine survival in
Pacific salmonids. J. Fish. Res. Board Can. 35:1434-1450.
1982. Nonlinear relation between smolts and adults in Babine
Lake sockeye salmon (Onx^orhynchus nerka) and implications
for other salmon populations. Can. J. Fish. Aquat. Sci.
39:904-913.
RiCKER, W. E.
1954. Stock and recruitment. J. Fish. Res. Board Can. 11:
559-623.
1975. Handbook of computations for biological statistics
of fish populations. Bull. Fish. Res. Board Can. 119, 382
P-
480
SAVOY AND CRECCO: MORTALITY OF AMERICAN SHAD
ROSENBURG, A. A., AND R. W. DOYLE.
1.986. Analysing the effect of age structure on stock-recruit-
ment relationships in herring (Clupea harengus). Can. J.
Fish. Aquat. Sci. 43:674-679.
Savoy, T. F., and V. A. Crecco.
1987. Daily increments on the otoliths of larval American
shad and their potential use in population dynamics studies,
/n R. C. Summerfelt (editor), Age and growth of fish, p.
413-431. Iowa State Univ. Press, Ames.
SAS.
1982. Statistical analysis system user guide. SAS Institute
Inc. Gary, NC, 584 p.
SCHERER, M. D.
1974. Analysis of factors affecting passage of American shad
{Alosa sapidissima; Wilson) at Holyoke Dam, MA, and
assessment of juvenile growth and distribution above the
dam. Ph.D. Thesis, Univ. Massachusetts, Amherst, MA,
224 p.
Shepherd, J. G., and D. H. Gushing.
1980. A mechanism for density-dependent survival of larval
fish as the basis of a stock-recruitment relationship. J.
Gons. Int. Explor. Mer 39:160-167.
Threlkeld, S. T.
1986. Lifetable reponses and population dynamics of four
cladoceran zooplankton during a reservoir flood. J. Plank-
ton Res. 8:639-647.
U.S. Geological Survey.
1966-87. Water resources data, Connecticut, water year
reports. U.S. Geol. Surv., var. pagination.
Walburg, G. H., and P. R. Nichols.
1967. Biology and management of the American shad and
status of the fisheries. Atlantic coast of the United States,
1960. U.S. Fish Wildl. Serv. Spec. Sci. Rep. 550, 105
P-
Ware, D. M.
1980. Bioenergetics of stock and recruitment. Gan. J. Fish.
Aquat. Sci. 37:1012-1024.
Watson, J. F.
1968. The early life history of the American shad, Alosa
sapidissima (Wilson), in the Connecticut River above Hol-
yoke, Massachusetts. M.S. Thesis, Univ. Massachusetts,
Amherst, MA, 55 p.
1970. Distribution and population dynamics of American
shad, Alosa sapidissima (Wilson), in the Connecticut River
above Holyoke Dam, Massachusetts. Ph.D. Thesis, Univ.
Massachusetts, Amherst, MA, 105 p.
White, R. L., J. T. Lane, and P. E. Hamer.
1969. Population and migration study of major anadromous
fish. Final Project Report AFSC-1-1, 1-2. NJ Dep. Gonserv.
Econ. Dev. Misc. Rep. No. 3M, 6 p.
Whitton, B. a.
1980. River Ecology. Univ. Calif. Press, Berkeley, 711 p.
Wiggins, T. A., T. R. Bender, V. A. Mudrak, and J. A. Coll.
1984. The development, feeding, growth and survival of
cultured American shad larvae through the transition from
endogenous to exogenous nutrition. Spec. Publ. Penn. Fish.
Comm., Benner Spring Fish Research Station, 34 p.
Zarbock, W. M.
1969. Annual progress report, Delaware River Basin Anad-
romous Fishery Project, AFS-2-2. U.S. Bur. Sport Fish.
Wildl., 20 p.
481
FISHERY BULLETIN; VOL. 86, NO. 3
APPENDIX 1
The total prerecruitment (ZTotal,) and postjuvenile (ZAi) mortality
rates from 1979 through 1987 were estimated by
ZTotal^ = -\og,iPart * 200,000) - \og,{Rt)
and
ZAt = log,(J,) - \og,iR,),
respectively. The standard errors about ZTotal^ and ZAi were estimated
as the sum of the variances (Cochran 1965):
■'ZTotal, -
C.2
SE^Tntal. = V +
nt n^
and
respectively. The variance estimates (5|^, Spavf) of adult recruitment
and spawning stock size were calculated from the log transformed
estimates from the daily lift rates at the Holyoke Dam, where Uf is the
total number of days in which 99% of the American shad were lifted.
The variance (Sj) about the juvenile abundance estimates (J^) was based
on the log transformed catches per seine haul from all stations and col-
lection dates, where rij is the total number of seine hauls made in that
year.
482
METABOLIC RESPONSES OF SPOT, LEIOSTOMUS XANTHURUS, AND
ATLANTIC CROAKER, MICROPOGONIAS UNDULATUS, LARVAE
TO COLD TEMPERATURES ENCOUNTERED FOLLOWING
RECRUITMENT TO ESTUARIES
Donald E. Hoss, Linda Coston-Clements, David S. Peters,
AND Patricia A. Tester^
ABSTRACT
The larvae of marine fishes that spawn during fall-winter in coastal North Carolina waters experience
a decrease in temperature as they enter estuarine nursery areas. To determine the effect of changes
in temperature on larval metabolism, the oxygen consumption of spot, Leiostomus xanthurus, and Atlantic
croaker, Micropogonias undulatus, was measured and their QO2 and Q^g values were determined. Atlan-
tic croaker respiration decreased with temperature at rates that would be expected if no compensation
or stress were involved. Spot showed unexpectedly high respiration rates at low temperature. The in-
creased respiration is apparently due to stress. Based on laboratory feeding and growth data, we con-
cluded that spot are subject to an energy deficit at <10°C. We infer the timing of larval immigration
corresponds with environmental temperatures reaching tolerable levels. Atlantic croaker larvae immigrate
earlier in the winter and are exposed to cold water for longer periods than spot larvae. Our conclusion
is that stress and energy loss experienced by early immigrating spot larvae may result in increased
mortality.
The larvae of fishes that spawn during fall and
winter in offshore North Carolina waters experience
a decrease in both temperature and salinity as they
enter estuarine nursery areas (Fig. 1). The spot,
Leiostomus xanthurus, and Atlantic croaker, Micro-
pogonias undulatus, two sympatric species of Sciae-
nidae, are representative of winter spawning species
off the North Carolina coast.
Previously, we have examined the effects of in-
creased temperature on the oxygen and food con-
sumption of the postlarval stages of these two
species (Hoss et al. 1971, 1974; Peters and Kjelson
1975). In this paper we continue our research on the
early life history of these species and evaluate how
decreasing water temperature, encountered follow-
ing recruitment into estuarine waters, might affect
oxygen consumption, food consumption, and ulti-
mately survival.
Crawshaw et al. (1981) stated that young fish
typically select warm shallow water because 1) it
permits more rapid growth owing to higher metab-
olism, given an adequate food supply, and 2) the
predation by larger fish is less in shallow water. For
some fish this explanation is plausible, but this does
not apply to the larvae of many winter-spawning
'Southeast Fisheries Center Beaufort Laboratory, National
Marine Fisheries Service, NOAA, Beaufort, NC 28516.
marine fishes which begin life in relatively warm
coastal waters and then enter colder estuarine
waters. For these species we expect that metabolism
and growth of the estuarine immigrants should be
reduced (Brett 1956). The specific objective of this
paper is to describe how decreasing temperature
affects the metabolism of larval fish as they are
moved from warm to cold water by a combination
of passive and active transport mechanisms that are
not, as yet, completely understood. Oxygen con-
sumption is a common method of estimating meta-
bolic activity, which frequently changes in response
to environmental conditions (O'Hara 1968). In this
study we measured routine oxygen consumption
which is the amount used by fish whose only move-
ments are spontaneous.
STUDY AREA AND METHODS
Spot and Atlantic croaker spawn off the North
Carolina coast inshore of the Gulf Stream over the
continental shelf (Hildebrand and Cable 1930;
Dawson 1958; Powles and Stender 1978). Here, spot
spawn from October to February, but principally
from December to January while Atlantic croaker
spawn from September to May but principally be-
tween October and December (Lewis and Judy
1983). After between 30 and 60 days in coastal
Manuscript accepted May 1988.
FISHERY BULLETIN: VOL. 86. NO. 3, 1988.
483
North
Carol
^^6%o
10 km
21°C
HOSS ET AL.: METABOLIC RESPONSES OF SPOT AND ATLANTIC CROAKER
waters (Warlen 1982; Warlen and Chester 1985), the
larvae enter estuaries where they develop into
juveniles. In the spawning area, water temperatures
are usually between 18° and 25°C (Fahay 1975;
Hettler and Powell 1981). The fish encounter de-
creasing temperatures as they move inshore to the
estuarine nursery areas. In the lower Newport River
estuary, for example, mean water temperatures be-
tween November and March may range from 14°
to 6°C with the highest temperatures during this
period occurring in November and the lowest in
January (Hoss 1974).
Larvae of Atlantic croaker and spot were obtained
from both field collections and eggs spawned in the
laboratory. Older larvae were captured in a bridge
net (Hettler 1979) and held in the laboratory for no
more than a week prior to use. First feeding larvae
were obtained from spawned fish, reproduced by the
methods of Hettler and Powell (1981), and then were
reared at experimental temperatures.
Oxygen consumption was measured with a differ-
ential respirometer (Umbreit et al. 1964), following
procedures used by Hoss, Hettler, and Coston
(1974). Fish were transferred to 15 mL respiration
flasks and, following a 2-h acclimation period, their
oxygen consumption was measured. Numbers of lar-
vae per flask varied between 1 and 30, depending
on the size of the larvae. Acclimation temperatures
were 10°, 15°, and 20°C. Notochord or standard
lengths and dry weights were obtained for individual
fish.
The metabolic equation Q = aW^, was used to
describe the relation between oxygen consumption
and dry weight of fish acclimated at 10°, 15°, and
20°C. In this equation, W is the weight of the fish,
and a and k are constants for the species obtained
from least-squares regression of the log of oxygen
consumption on the log of weight (Winberg 1956).
A k value of 0.67 implies that oxygen consumption
varies in proportion to surface area, whereas a value
of 1 indicates that respiration varies in proportion
to weight.
We used the metabolic equation to estimate oxy-
gen consumption of larvae of equal weight at differ-
ent temperatures. We compared larvae of 4 mg dry
weight because this is the realistic estimate of their
weight as they are transported from coastal to estu-
arine waters (Warlen 1982; Warlen and Chester
1985).
Growth and feeding rates of small spot (^20 mm
SL) collected from the Newport River were calcu-
lated from data collected in the laboratory at several
temperatures. Wet weights (^15-30 mg) were re-
corded to the nearest milligram, and 10 fish were
randomly assigned to 4 L test and control con-
tainers. Control fish were dried to determine the
dry/wet weight ratio, which was then used to esti-
mate initial dry weight of experimental fish. One
experiment was conducted at 6°, 8°, 10°, 12°, and
16°C, and two experiments were conducted at 18°C.
In all cases fish were fed newly hatched brine shrimp
several times a day to assure an ad libitum food
supply. After 4-6 days all food was removed; lar-
vae were allowed time to clear their guts and then
were dried and weighed.
Growth and feeding rates were expressed as per-
cent of body weight per day. Growth rate was
calculated from the expression:
Growth rate = 100 [(WwAVo)
l/n
1]
where Wn = dry weight of all fish in a tank at day
n
Wo = estimated original dry weight of fish
n = number of days fed.
Calculation of feeding rates required the assump-
tion of constant growth rates. Using original dry
weights and calculated growth rates we determined
the total dry weight of fish in each container at the
beginning of each day. Dry weights of brine shrimp
eaten each day divided by the calculated dry weights
of fish gives proportion of body weight ingested.
These proportions were then summarized as aver-
age daily percent of dry body weight ingested.
In order to compare metabolic parameters, i.e.,
oxygen consumption, feeding, and growth rates, the
following conversion factors were used: 1.0 mg dry
wt = 5.0 cal (Thayer et al. 1973; Paffenhoffer 1967);
1.0 mg O2 = 3.38 cal (Phillipson 1966) and 0.7 mg
O2 = 1 mL O2 at STP. One tenth calorie per fish
per day was added to all the rates so that measured
zeros could be shown on a log scale.
RESULTS AND DISCUSSION
The regression equations relating oxygen con-
sumption to weight at several temperatures are
shown in Table 1. Higher coefficient of determina-
tion (R^) values were found at higher temperatures,
a trend best explained by differences in the size
range of fish measured at different temperatures.
Values for k, were generally comparable to values
reported by other investigators for fish of a similar
size and at comparable temperatures— Hoss (1974),
pinfish; Houde and Schekter (1983), bay anchovy,
sea bream, and lined sole; Almatar (1984), herring;
and Laurence (1978), cod and haddock.
485
FISHERY BULLETIN: VOL. 86, NO. 3
Table 1.— Metabolic equations relating oxygen consumption to
body size. Q = aW* where Q = oxygen consumption (piL Oj ■
h"'), a and k are regression coefficients, W is dry wt in mg, N is
tfie number of observations, T the temperature, and Rj the coef-
ficient of determination.
Species
T
N
W
a
k
R2
Atlantic
10
52
1.50 - 6.43
1.80
0.66
0.42
croaker
15
65
1.90 -11.71
2.00
1.02
0.65
20
81
0.013- 9.31
3.59
0.86
0.94
Spot
10
37
1.82 - 8.01
3.63
0.70
0.54
15
48
2.18 -10.27
3.15
0.64
0.63
20
79
0.014- 8.91
4.07
0.92
0.97
Comparing between measured oxygen consump-
tion rates at 10° and 15°C and those predicted from
Van't Hoff's equation (Vernberg and Vernberg
1972), we conclude that Atlantic croaker show no
sign of regulating their oxygen consumption as
water temperature is decreased. The difference be-
tween oxygen consumption rates based on Q^q
values of 2 and 3 (Fig. 2) is an expected range. For
every 10°C change in temperature, the rate of a
chemical reaction typically changes by a factor of
2 to 3. A Qio value of appreciably <2 or more than
a-
z
O
h-
o.
ID
CO
z
o
o
X
o
<iU
Ta
15
'-
Atlantic croaker
-
.-.-5
10
-
^ — ^ ^ -^
8
-
.-''5'' ,,'-'
-
_^--'' ,'''
6
-
-'''' ,--'
2 ^''''
4
i 1 1
20
15
10
8
Spot
-'i
10 15
20
TEMPERATURE (°C)
Figure 2.— Oxygen consumption rates for 4 mg Atlantic croaker
(A) and spot (B) at three temperatures estimated from equations
in Table 1. The bars indicate standard errors from regressions in
Table 1. Broken lines are estimates of the rates expected based
on Van't Hoff's equation, the rate measured at 20°C, a Q^ of 2
(upper line) and a Qjq of 3 (lower line).
3, indicates that some process other than a chemical
one is involved (e.g., a change in cell membrane
permeability). A Qjq of one indicates temperature
independence (Vernberg and Vernberg 1972). Our
conclusion that Atlantic croaker did not display
thermal stress is based on the fact that measured
respiration rates at reduced temperatures (10° and
15°C) were within the range expected (Fig. 2A).
For spot a decrease in temperature from 20° to
15°C resulted in a decrease in oxygen consumption
of approximately the amount expected for a Qjq of
3. A further decrease in temperature to 10°C, how-
ever, caused an increase in the respiration rate. The
changes in oxygen consumption at low temperatures
could be interpreted either as adaptive, i.e., main-
taining a high metabolic rate even at the lower tem-
perature, or inadaptive, i.e., a metabolic breakdown.
Based on feeding, growth, and survival data, how-
ever, we think the increase in respiration by spot
at 10° C is a result of cold stress, not adaptation.
In Figure 3 we present three measures of metabolic
rate, ad libitum feeding rate, maximum growth rate,
and routine oxygen consumption for spot, all as a
function of temperature. Feeding, growth, and oxy-
gen consumption rates decrease with decreasing
temperature and the rates are similar over a limited
range of the conditions tested (Fig. 3). The rates of
decline in feeding and growth from 18° to 12°C ap-
proximates that of oxygen consumption from 20°
to 15° C. At lower temperatures stress appears to
become important. For example, it was not possi-
ble to measure growth at 10° C or below because
only a fraction of the larvae survived. At 8° and
10° C some of the larvae did not eat and at 6°C none
of them did. This agrees with Dawson (1958) who
concluded that the lethal minimum temperature for
spot is in the 4.0°-5.0°C range and probably varies
with size. The intersection of ad libitum feeding rate
and routine oxygen consumption occurs at approx-
imately 10°C (Fig. 3). At this temperature there is
just enough energy available for routine metabolism.
Below this temperature there is not enough energy
available even at the ad libitum feeding rate to
maintain the larvae, and spot held at this
temperature for any length of time would be unlikely
to survive.
We conclude from our data on metabolic responses
to temperature that spot and Atlantic croaker lar-
vae differ in their response to cold temperatures
which prevail at the time of their recruitment to the
estuary and that this difference may have impor-
tant implications for their survival. Both species
spawn in warm waters of the continental shelf
where the eggs hatch. As the larvae grow they are
486
HOSS ET AL.: METABOLIC RESPONSES OF SPOT AND ATLANTIC CROAKER
<
I
W
Ll.
w
UJ
cr
o
<
o
Figure 3.— Ad libitum feeding rate, growth rate
at ad libitum feeding rate, and routine oxygen con-
sumption for 4 mg dry wt spot at various
temperatures between 6° and 20°C.
10. Op
80-
60-
4 0
20-
10
0 8
0 6
0.4
0.2
Spot
• = AD LIBITUM FEEDING RATE
A = GROWTH RATE AT AD LIBITUM FEEDING
O - ROUTINE OXYGEN CONSUMPTION
n 1' ^ ' 'I
10 14 18 22
TEMPERATURE (°C)
transported from warm coastal waters into cold
estuarine waters. Atlantic croaker are capable of
enduring low winter temperatures with decreased
metabolic rates that allow for balanced energy in-
take. Spot, in contrast, show signs of thermal stress
manifested as increased respiration rate (at 10°C).
This increased metabolism along with no attendant
increase in feeding results in an energy deficit and
in eventual mortality of the larvae. Species specific
differences in the time of entry to the estuary serves
as ecological evidence supporting our contention
that spot are more susceptible to cold weather. Most
spot enter the estuary after the peak in Atlantic
croaker immigration and generally after the coldest
weather.
Our findings have important implications with
respect to recruitment of estuarine-dependent fish
which spawn in the ocean during winter. It may be
that during severe winters, many of the larvae of
cold sensitive species (e.g., spot) that reach the
estuary early are killed by cold water temperatures
(10°C or less). Thus, only the late arriving larvae
survive to recruit into the fishery. The difference
in survival between severe and normal winters may
help to explain in part the difference between good
and poor year classes of certain fish.
Acknowledgments
We thank W. Hettler for providing eggs of labor-
atory spawned fish, and W. Hettler and J. Govoni
for providing critical reviews of the manuscript. This
research was funded in part by a cooperative agree-
ment between the National Marine Fisheries Service
and the Department of Energy E(49-7)5.
LITERATURE CITED
Almatar, S. M.
1984. Effects of acute changes in temperature and salinity
on the oxygen uptake of larvae of herring (Clupea harengus)
and plaice (Pleuronectes platessa). Mar. Biol. (Berl.) 80:117-
124.
Brett, J. R.
1956. Some principles in the thermal requirements of fishes.
Q. Rev. Biol. 31:75-87.
Crawshaw, L. I., B. P. MoFFiTT, D. E. Lemons, and J. A.
Downey.
1981. The evolutionary development of vertebrate thermo-
regulation. Am. Sci. 69:543-550.
Dawson, C. E.
1958. A study of the biology and life historj' of the spot,
Leiostomus xanthurus Lacepede, with special reference to
South Carolina. Bears Bluff Lab. Contrib. No. 28, 48 p.
Fahay, M. p.
1975. An annotated list of larval and juvenile fishes captured
with surface-towed meter net in the South Atlantic Bight
during four RV Dolphin cruises between May 1967 and
February 1968. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS SSRF-685, 39 p.
Hettler, W. F.
1979. Modified neuston net for collecting live larval and
juvenile fish. Prog. Fish.-Cult. 41:32-33.
Hettler, W. F., and A. B. Powell.
1981. Egg and larval fish production at the NMFS Beaufort
Laboratory, Beaufort, N.C., U.S.A. Rapp. P.-v. R6un Cons,
int. Explor. Mer 178:501-503.
487
FISHERY BULLETIN: VOL. 86, NO. 3
HiLDEBRAND, S. E., AND L. E. CABLE.
1930. Development and life history of fourteen teleostean
fishes at Beaufort, N.C. Bull. U.S. Bur. Fish. 46:383-488.
Hoss, D. E.
1974. Energy requirements of a population of pinfish, Lago-
don rhomboides (Linnaeus). Ecology 55:848-855.
Hoss, D. E., L. C. CosTON, and W. F. Hettler, Jr.
1971. Effects of increased temperature on postlarval and
juvenile estuarine fish. Proc. Annu. Conf. Southeast. Assoc.
Game Fish. Comm. 25:635-642.
Hoss, D. E., W. F. Hettler, Jr., and L. C. Coston.
1974. Effects of thermal shock on larval estuarine fish - eco-
logical implications with respect to entrainment in power
plant cooling systems. In J. H. S. Blaxter (editor), The early
life history of fish, p. 357-371. Springer- Verlag, N.Y.
HouDE, E. D., and R. C. Schekter.
1983. Oxygen uptake and comparative energetics among
eggs and larvae of three subtropical marine fishes. Mar.
Biol. (Berl.) 72:283-293.
Laurence, G. C.
1978. Comparative growth, respiration and delayed feeding
abilities of larval cod (Gadv^ morhua) and haddock (Melano-
grammus aeglefinus) as influenced by temperature during
laboratory studies. Mar. Biol. (Berl.) 50:1-7.
Lewis, R. M., and M. H. Judy.
1983. The occurrence of spot, Leiostomus xanthunis, and
Atlantic croaker, Micropogcmias undulatus, larvae in Onslow
Bay and Newport River Estuary, North Carolina. Fish.
Bull., U.S. 81:405-412.
LiPPSON, A. J., AND R. L. MORAN (editors).
1974. Manual for the identification of early developmental
stages of fishes of the Potomac River estuary. Power Plant
Siting Program, Md. Dep. Nat. Res. PPSP-MP-13, p. 220-
222.
O'Hara, J.
1968. The influence of weight and temperature on the meta-
bolic rate of sunfish. Ecology 49:159-161.
Paffenhofer, G.-A.
1967. Calorie content of larvae of the brine shrimp Artemia
salina. Helgol. Wiss. Meeresunters. 16:130-135.
Peters, D. S., and M. A. Kjelson.
1975. Consumption and utilization of food by various postlar-
val and juvenile fishes of North Carolina estuaries. In L. E.
Cronin (editor), Estuarine research, vol. 1, p. 448-472.
Acad. Press, N.Y.
Phillipson, J.
1966. Ecological energetics. E. Arnold, Lond., 57 p.
Powell, A. B., and H. R. Gordy.
1980. Egg and larval development of the spot Leiosto-
mus xanthunis (Sciaenidae). Fish. Bull., U.S. 78:701-
714.
POWLES, H., AND B. W. STENDER.
1978. Observations on composition, seasonality and distribu-
tion of ichthyoplankton from MARMAP cruises in the South
Atlantic Bight in 1973. S.C. Mar. Res. Cent. Tech. Rep.
Ser. 11, 47 p.
Thayer, G. W., W. E. Schaaf, J. W. Angelovic, and M. W.
LaCroix.
1973. Caloric measurements of some estuarine organisms.
Fish. Bull., U.S. 71:289-296.
Umbreit, W. W., R. H. Burris, and J. F. Stauffer.
1964. Manometric techniques. Burgess Publ. Co., Minneap.,
305 p.
Vernberg, W. B., and F. J. Vernberg.
1972. Environmental physiology of marine animals.
Springer-Verlag, N.Y., 346 p.
Warlen, S. M.
1982. Age and growth of larvae and spawning time of Atlan-
tic croaker. Proc. Annu. Conf. Southeast. Assoc. Fish
Wildl. Agencies 34:204-214.
Warlen, S. M., and A. J. Chester.
1985. Age, growth, and distribution of larval spot, Leio-
stomus xanthums, off North Carolina. Fish. Bull., U.S.
83:587-599.
WiNBERG, G. G.
1956. Rate of metabolism and food requirements of fishes.
Can. Fish. Res. Board, Transl. Ser. 194. 202 p.
488
GROWTH OF PACIFIC SAURY, COLOLABIS SAIRA, IN
THE NORTHEASTERN AND NORTHWESTERN PACIFIC OCEAN'
YosHiRO Waia.nabe,2 John L. Builer,^ and Tsukasa Mori^
ABSTRACT
Growth of the Pacific saury, Cololabis saira, from the northeastern and northwestern Pacific Ocean was
studied using otolith growth increments. We found that growth of Pacific sauries from the western Pacific
was higher than that from the eastern Pacific. Assuming that otolith growth increments are deposited
daily, average growth rates from hatching up to 1 year old were 0.62 mm/d in the eastern and 0.85 mm/d
in the western Pacific. Because the growth rate changes at around 100 mm, two curves were used to
model the growth of Pacific saury in the western Pacific: one for fish up to 100 mm and the other for
fish larger than 100 mm. Based on counts of daily increments. Pacific sauries may be short lived. The
oldest specimen examined was only 14 months old.
The Pacific saury, Cololabis saira (Brevoort), is
distributed throughout the North Pacific Ocean and
is one of the most important commercial fishes in
the northwestern Pacific. The average annual catch
of Pacific saury in Japan has been approximately
200,000 1 (metric tons) in the last 20 years (Statistics
and Information Department, Japan 1985). The
catch has varied by an order of magnitude in the
last 20 years from a minimum of 63,000 t in 1969
to a maximum of 406,000 t in 1973. Fluctuation in
stock size is a major factor in catch variability
although economic factors such as fish price may
also affect total landings. However, the causes of
stock fluctuation in the western Pacific remain
unknown. In the eastern Pacific, the Pacific saury
has not been exploited but is recognized as a poten-
tial fishery resource (Ahlstrom 1968; Smith et al.
1970).
Investigations of the Pacific saury have mainly
been devoted to such subjects as systematics, abun-
dance, distribution, migration, and formation of
fishing ground in relation to oceanographic condi-
tions (e.g., Hubbs and Wisner 1980; Smith et al.
1970; Odate 1977; Fukushima 1979; Sablin and
Pavlychev 1982; Gong 1984). Age determination and
growth, however, remain controversial (Hatanaka
1955; Hotta 1960; Novikov 1960; Sunada 1974; Kim
"Contribution No. 429 from Tohoku Regional Fisheries Research
Laboratory.
^Tohoku Regional Fisheries Research Laboratory, Fisheries
Agency, Shiogama, Miyagi 985, Japan.
^Southwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038.
*Facu]ty of Fisheries, Hokkaido University, Hakodate, Hokkaido
041, Japan.
and Park 1981), notwithstanding their critical im-
portance for fish stock assessment.
The discovery of daily increments in the otoliths
of fishes (Pannella 1971) has made it possible to
estimate age and growth of larval and juvenile fishes
accurately. Daily increments have been used to age
many species of fishes (Jones 1986). Nishimura et al.
(1985) reported the presence of growth increments
in Pacific saury otoliths observed by scanning elec-
tron microscopy and suggested that it is possible to
estimate age and growth of Pacific saury by using
daily increments in the otolith. The purpose of this
paper is to determine the age of Pacific sauries from
the eastern and western North Pacific using daily
increments and to compare the growth rates in these
areas.
MATERIALS AND METHODS
We read otoliths of 75 Pacific sauries from the
northeastern and 172 from the northwestern Pacific
Ocean. Details of sampling and methods of reading
otoliths are summarized in Table 1 and Figure 1.
Additional samples from the western Pacific were
used to determine the relation between otolith size
and fish length. Fish from the eastern Pacific were
fixed and preserved in 80% alcohol after capture,
and those from the western Pacific were stored
frozen and thawed when processed. Because speci-
mens frequently have damaged upper jaws, knob
length (the distance from the tip of the lower jaw
to the posterior end of the muscular knob at the base
of a caudal peduncle) is the standard measure of
body size in Pacific saury. All body lengths in this
paper are knob length.
Manuscript accepted March 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
489
FISHERY BULLETIN: VOL. 86, NO. 3
Table 1.— Collection records of saury samples from the eastern (# 1-10) and the western (# 11-22)
North Pacific. N = Neuston net; G = Gill net; D = Dip net.
Location
Size range
Micro-
Sample
Date
Lat.
Long.
No.
(KnL mm)
Gear
scope
1
80 06 29
36°02'N
124°04'W
10
19.8-109.0
N
LM
2
80 06 29
36°07'N
123°55'W
10
19.8- 66.0
N
LM
3
80 06 29
36°07'N
123°45'W
10
15.3- 96.0
N
LM
4
81 10 26
46°36'N
127°49'W
5
47,8-235.0
N
LM
5
81 10 27
47°19'N
126°09'W
6
38.0-142.0
N
LM
6
81 10 28
47°20'N
124°30'W
12
23.9-216.0
N
LM
7
81 10 29
46°21'N
127°38'W
1
209.0
N
LM
8
81 10 30
46°38N
125°54'W
14
20.6- 85.0
N
LM
9
81 11 01
45°21'N
127°38'W
4
27.0- 71.0
N
LM
10
81 11 02
45°38'N
124°51W
3
70.0-206.0
N
LM
11
84 05 26
37°01'N
164°02'E
19
145.0-230.0
G
SEM
12
84 06 02
37°00'N
158°00'E
114
38.0-125.0
N
SEM
13
84 07 16
42°00'N
172°00'E
7
213.0-282.0
G
SEM
14
84 10 06
43°09'N
153°14'E
5
300.0-330.0
D
SEM
15
85 05 20
36°00N
150°00'E
3
21.0- 69.5
N
LM
16
85 05 21
38°15'N
149°59'E
2
8.3- 12.5
N
LM
17
85 05 23
38°30'N
152°00'E
2
26.5- 45.5
N
LM
18
85 05 24
38°00'N
152°00'E
9
29.0- 85.0
N
LM
19
85 05 29
38°45'N
156°00'E
3
16.0- 33.5
N
LM
20
85 05 29
38°30N
ise-'oo'E
2
27.0- 33.0
N
LM
21
85 05 29
38015N
156°00'E
3
24.6- 65.0
N
LM
22
85 05 30
38°00'N
156°00'E
3
31.5- 68.5
N
LM
140°E
150°E 160°E 170°E
Figure 1.— Locations of Pacific saury collection in the North Pacific. Figures by
490
WATANABE ET AL.: GROWTH OF SAURY
Sagittae were dissected out from fish and left to
dry after removing tissues and membranes. We used
a dissecting microscope with a polarizing filter to
dissect otoliths from small larvae and juveniles. The
otoliths were read either by light microscopy (LM)
or by scanning electron microscopy (SEM). Otoliths
that were to be read by LM were mounted in
EUKITT^ after dissection. Otolith radius was
measured from the focus to posterior margin and
the increments were counted along the same tran-
sect using the otolith reading system, which was
developed by the Southwest Fisheries Center of
the National Marine Fisheries Service, NOAA, and
which consists of a light microscope, a video
monitor, a micro-computer, and a digitizer (Methot
1981).
For SEM, otoliths were mounted in epoxy or
methacrylate resin. The otolith radius was measured
from the focus to the posterior margin with an
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
optical comparator. The otoliths were ground
oblique to the sagittal plane parallel to the long
axis of the otolith in order to have a flat plane
through the otolith nucleus. The polished surface
was washed in xylene, using an ultrasonic washer,
then dried and etched for 50 seconds with 0.2M
EDTA-2Na (disodium ethylenediaminetetraacetic
acid). The etched surface was coated with palladium
platinum and observed under an SEM (JSM-25) at
15 kV.
The three authors of this paper read saury oto-
liths independently: the senior author read fish from
the western Pacific up to 85 mm by LM, the second
author read otoliths from the eastern Pacific by LM,
and the third read otoliths from the western Pacific
larger than 38 mm by SEM. To confirm that we
were all interpreting the same structure as growth
rings by SEM and LM, we compared 50 data points
read by SEM and 14 points read by LM for west-
ern Pacific sauries between 38 and 85 mm. The
distribution of increment number versus knob length
was the same. We also checked for possible biases
for the two readers using LM by having each read
the same set of otoliths independently.
130°W 120°W
dots indicate sample numbers in Table 1.
RESULTS
The nucleus of a Pacific saury sagitta is approx-
imately 20 ^m in diameter and is composed of four
to six small dense bodies which appear to be assem-
blages of calcareous spherules (Fig. 2a). These dense
bodies are separated from one another and each is
surrounded by a small concentric ring.
We observed the nucleus areas of otoliths from
Pacific sauries collected in the western Pacific in
1985 (sample # 15-22) and found that most of
them had a distinct ring of about 27 fim in radius.
Between the nucleus and this distinct ring, four
(or five) indistinct growth rings were detected (Fig.
2b).
We measured knob lengths of 27 larval and juven-
ile Pacific sauries before freezing and after thaw-
ing, and found that the ratio of these two measure-
ments was 0.997. There was virtually no shrinkage
by freezing and thawing. Theilacker (1980) found
that preservation of larval northern anchovy,
Engraulis mordax, in 80% alcohol did not cause
additional shrinkage of the body after net treatment.
Thus, knob lengths after both 80% alcohol preser-
vation and freezing are comparable to each other,
and this measurement corresponds to the size after
net treatment. Since shrinkage factors by net treat-
ment are not known for the saury, lengths are un-
corrected for net shrinkage.
491
FISHERY BULLETIN: VOL. 86, NO. 3
if
.^1
^
a
Figure 2.— Light micrographs of Pacific saury sagittae. a) OtoHth nucleus composed of 5 or 6 separate dense bodies with surrounding
cores, b) Assumed 4 embryonic and 1 hatching (arrow) rings.
The daily periodicity of growth increment forma-
tion in the Pacific saury has not been verified. For
that reason we plotted the number of increments
versus knob length instead of age versus length. We
used the Laird-Gompertz equation to describe the
relations of increment number and length as growth
curves for both the eastern and western Pacific
saury. Hatching size of artificially fertilized and in-
cubated Pacific saury from the western Pacific was
reported to be 7.19 mm in average live total length
(Yusa 1960). From the drawing of a newly hatched
larva in Yusa's paper, we estimated live knob length
to be 6.60 mm. Shrinkage factors of northern an-
chovy in the size range from 6.00 to 7.99 mm were
0.90 for a 5-min net treatment and 0.85 for a 10-min
net treatment (Theilacker 1980). Using these values,
the capture size of a newly hatched larva of Pacific
saury after a 5-min net treatment was estimated to
be 5.95 mm and a 10-min treatment to be 5.61 mm.
We fixed the hatching size from 5.85 to 5.95 mm
in the growth curve, because the Pacific saury lar-
vae at this size are in a more advanced develop-
mental stage and shrank less by net treatment than
northern anchovy.
The resulting growth equation for the eastern
Pacific saury was
KnL = 5.85 exp((0.0427/0.115)(l - e(-o.oii5(/-5))))
and the equation for the western Pacific saury was
KnL = 5.95 exp((0.0504/0.0128)(l - e(-ooi28(/-5))))
where KnL is a knob length in mm and / is the total
number of increments observed in an otolith. The
term, 7-5, indicates that five increments were pre-
sumed to have been present at hatching. Data from
the western Pacific saury appear to consist of two
curves separated around 100 mm in KnL. Two
Laird-Gompertz curves fit much better than one
curve. The intersection of the two curves was at 1 14
increments and 100 mm. The growth equation for
fish smaller than 100 mm was
492
WATANABE ET AL.: GROWTH OF SAURY
KnL = 5.90 exp((0.0865/0.0293Xl - e(-oo293(/-5))))
and for fish larger than 100 mm KnL was
KnL = 3.01 exp((0.0592/0.0126)(l - e(-o.oi26(/-5))))
The estimated mean square error, 215.7, of the two
curves was smaller than that for a single curve,
351.7. The two-curve model fits much better for the
smaller size range up to 100 mm. The estimated
mean square error of the two-curve model for this
size range, 75.7, was much smaller than that of the
one-curve model, 240.8.
The growth rate of Pacific saury in the eastern
Pacific was slower than that in the western Pacific
(Figs. 3, 4). The knob length of saury in the eastern
Pacific would be about 75 mm at 100 rings, 170 mm
at 200 rings, and 220 mm at 300 rings, whereas in
the western Pacific knob length would be about 100
mm at 100 rings, 230 mm at 200 rings, and 300 mm
at 300 rings. Assuming that the rings are formed
daily, overall growth rates of the first one year
of their life were 0.62 mm/d and 0.85 mm/d for
the eastern and western Pacific Ocean, respectively.
The largest specimen examined was 330 mm from
the western Pacific, which had 328 increments, and
the oldest fish, also from the western Pacific, which
measured 320 mm, had only 418 increments. These
fish would be classified as very large or large by
Novikov's categories offish size composition (Novi-
kov 1960, 1973). The largest fish examined from the
eastern Pacific was 235 mm and had 241 incre-
ments. Hughes (1974), however, reported larger fish
from the eastern Pacific.
In Pacific saury from the western Pacific, the rela-
tion between otolith radius in f^m (OR) and knob
length in mm (KnL) was linear on logarithm-
logarithm coordinates (Fig. 5). The equation com-
puted by the geometric mean regression (Ricker
1973) was
In (OR) = 2.33 + 0.749 In (KnL) (r = 0.979).
The otolith radius at hatching (5.9 mm KnL) cal-
culated by this formula was 38.9 ^im, which was 12
/im larger than the radius of the presumed hatch-
ing ring.
200 -
E
E
I
H
o
z
UJ
_J
CO
o
100 -
00427 f ^(-0.0115(l-5))\
0.0115 ^ ^® '
KnL: KNOB LENGTH
I: NUMBER OF INCREMENTS
100 200
NUMBER OF INCREMENTS
Figure 3.— Growth curve of the eastern Pacific saury.
300
493
FISHERY BULLETIN; VOL. 86, NO. 3
300
£ 200
£
X
I-
o
z
111
_J
m
O
z
^ 100
50
KnL = 5.95e
0.0504 / ^(-0.0128(l-5))\
0.0128^^"® '
I <114
0.0865 / {-0.0293{l-5))\
KnL = 5.90e 0.0293 ^^'^ '
0.0592 / (-0.0126(1-5)))
KnL = 3.01e 0.0126^^"®
KnL: KNOB LENGTH
I: NUMBER OF INCREMENTS
50
100
200
300
400
NUMBER OF INCREMENTS
Figure 4.— Growth curve of the western Pacific saury.
DISCUSSION
The microstructure of otolith growth increments
of the Pacific saury is similar to that of daily incre-
ments in some other fishes (Nishimura et al. 1985).
Thus, the following discussion is based on the
assumption that the increments are daily growth
rings. A rearing experiment of larval sauries is
under way in the senior author's laboratory to verify
daily periodicity of the increment formation.
Formation of a few embryonic growth rings or a
lamellar structure has been reported in California
grunion, Leuresthes tenuis, (Brothers et al. 1976);
mummichog, Fundulus heteroclitus, (Radtke and
Dean 1982); and walleye pollock, Theragra chalco-
gramma, (Nishimura and Yamada 1984). Radtke
and Dean (1982) mentioned that deposition of
growth rings in the embryonic stage might be
related to a long incubation period. Pacific saury has
a long incubation period— about 17 days under
13.5°-15.7°C (Yusa 1960). At this temperature, eye
pigmentation begins 7 or 8 days before hatching,
and pectoral fins show constant movement from 5
or 6 days before hatching (Yusa 1960). Notochord
flexion occurs about midway through embryonic
development at 14°-22°C (Uchida et al. 1958). Thus
saury is more advanced at hatching than killifish
based on the embryonic development of killifish
reported by Armstrong and Child (1965). Observ-
ing the central area of otoliths, we found four faint
rings and a dark ring immediately outside of those
rings. We assumed therefore that the four faint
rings are embryonic rings and the dark ring is the
hatching ring. This assumption may be confirmed
by examining otoliths of late embryos and newly
hatched larvae of saury.
494
WATANABE ET AL.: GROWTH OF SAURY
1000 I-
500
CO
Q
<
CC
o
I-
o
100
40
In OR =2.33 +0.749 In KnL
r = 0.979
OR: OTOLITH RADIUS
KnL: KNOB LENGTH
J \ \ \
10
50 100
KNOB LENGTH (mm)
200 300
Figure 5.— Knob length and otolith radius relationship of the western Pacific saury.
Previous studies on age and growth of the Pacific
saury have based on annuli on scales and/or otoliths.
Sunada (1974) found five age groups in Pacific
sauries off southern Oregon, California, and Baja
California. Mean fork lengths of age groups were
171, 220, 246, 270, and 268 mm for age 0, 1, 2, 3,
and 4, respectively. Hughes (1974) examined age
composition of 5,248 sauries collected in waters off
California up to Vancouver Island. He found spring-
and autumn-bom fish in his samples, but little differ-
ence was noted in growth rates between two groups.
Approximate knob lengths of 1.0- to 5.0-year-old fish
were 180, 230, 255, 290, 310 mm. The growth rates
given in these two papers are not very much dif-
ferent. The saury grows at 0.5-0.6 mm/d up to 1
year old, which is almost equal to our growth rate
in the eastern Pacific, 0.62 mm/d. Hatanaka (1955)
found five age groups in the western population of
the saury, 0-4 years old, and estimated mean body
length of age groups to be 80, 160, 230, 265 mm for
1- to 4-year-old fish, respectively. Novikov (1960,
1973) divided sauries captured in autumn into five
size groups, very small ( - 200 mm), small (201-240),
medium (241-290), large (291-320), and very large
(321-(-), and assigned the small, medium, and large
to 1-, 2-, and 3-year-old with maximum 5-year-old
fish.
A different model of Pacific saury growth in the
western Pacific was proposed by Hotta (1960) based
upon a hypothesis of two subpopulations. He separ-
ated the saury into spring-spawning and autumn-
spawning populations based upon the observations
of fish size composition, scales, otoliths, and num-
bers of vertebrae. He assigned four ages of half year
intervals, 1.0, 1.5, 2.0, 2.5 years old, to fish 210-240,
260-280, 290-300, and 310-330 mm, respectively.
The growth rate up to 1 year was 0.6-0.7 mm/d. Kim
and Park (1981) examined Pacific sauries from
Korean waters and found four size groups of four
different ages of half-year intervals as well. They
presented two growth models for each of two sub-
populations, spring and autumn spawning, based
upon the von Bertalanffy equation. The sizes at ages
were almost identical to those of Hotta (1960). How-
ever, the hypothesis of two saury subpopulations in
the western Pacific is not supported by electro-
phoretic analyses of genetic separation (Numachi
1971; Hara et al. 1982).
The average growth rate of the western Pacific
saury in this paper was 1.1 mm/d from 0 to 8 or
495
FISHERY BULLETIN: VOL. 86, NO. 3
9 months old. It was still faster than Novikov's
growth rate of the corresponding age period (0.83
mm/d), which was the highest rate of all the previous
reports. The fish would become 316 mm in one year
according to our model.
Support for the fast growth rate of Pacific saury
presented in this paper in the western Pacific can
be found in rearing experiments. Hotta (1958)
reared young sauries caught by a set net. He reared
them in a crawl and fed them minced anchovy and
mackerel twice a day. Young sauries 116 mm in
mean length became 172 mm in the rearing period
after 72 days. The growth rate was 0.78 mm/d. The
sauries fed three times a day grew 130-143% faster
than the group fed twice a day. Thus growth rates
of young sauries may be higher than 1.1 mm/d (0.78
X 1.4 mm/d) when food is readily available. Our
growth rate of sauries in this size range was approx-
imately 1.5 mm/d in the western Pacific. For Atlan-
tic saury, Scomberesox saurus scombroides, reared
by Brownell (1983), the average growth rate of the
larvae was 0.62 mm/d from hatching (7.5 mm SL)
to 47-day old (36.8 mm SL). The growth rates of
Cololabis saira in a corresponding period were 0.48
mm/d in the eastern Pacific and 1.0 mm/d in the
western Pacific.
Our results indicate that the growth rate of Pacific
saury in the western Pacific is much higher than in
the eastern Pacific. This could be due to a differ-
ence in food availability between the two areas.
However, mean zooplankton standing stock in
1951-66 was 34.8 g/m^ in the California Current
region in June (Smith and Eppley 1982), whereas
that of Kuroshio water off southern Japan was 4.7
g/m^ and of Oyashio area off northern Japan was
25.7 g/m2 in May to July (Odate 1986). Thus, differ-
ences in zooplankton standing stock do not explain
the difference in growth rates.
On the other hand, there seems to be a reasonable
explanation for the reacceleration of growth rate at
around 100 mm in the western Pacific saury. The
western sauries hatch out mainly in offshore water
of the Kuroshio Current (lat. 31-33°N) off Japan
in winter. They migrate north to the Oyashio area
(up to 46-50 °N) where copepods are highly avail-
able. Young and adult sauries feed actively and gain
fat. They are in the northward migration stage in
early summer when they are about 100 mm, and are
moving from poor Kuroshio water to rich Oyashio
water (Fukushima 1979). High zooplankton stand-
ing stock in the Oyashio water and its derivatives
might be responsible for the reacceleration of
growth rate in fish older than 100 days.
The growth rates of Pacific saury in the western
Pacific may differ from year-to-year due to environ-
mental factors and may result in changes in size
composition of the fish. Between 1968 and 1972,
mean knob length of exploited sauries in the west-
ern Pacific was 170-250 mm, whereas in the 1980's
the major mode in the size composition was 290-310
mm (S. Kosaka''). This increase could have been
due to an acceleration of growth rate or a shift of
spawning season to early months or both in recent
years. The high growth rate of western Pacific saury
presented in this paper has come from specimens
collected in 1984 and 1985. The growth rate in the
late 1960's and early 1970's may have been lower
than that presented in this paper. Investigation of
the interannual variation in growth rates using daily
increments would distinguish between these two
hypotheses.
We used three different gears to collect Pacific
saury samples in the western Pacific. Knob lengths
of sauries collected were from 8.3 to 125 mm by ring
net, 145 to 282 mm by gill net, and 300 to 330 mm
by stick-held dip net. Sauries of 125-145 mm might
not be available either to the ring net or to the gill
net. Further, the ring net may select small juveniles
of a cohort in the size range over 100 mm, and this
may have produced the two growth curves. This
problem needs to be examined further with data on
gear selectivity.
We do not know how long Pacific sauries survive
after becoming adult. The oldest specimen in our
sample was about 14 months old after hatching (418
increments). The largest saury aged (330 mm),
which had 328 growth rings, is close to the max-
imum size. Although the maximum known length
of the Pacific saury was reported to be about 400
mm (Hubbs and Wisner 1980), the largest fish ex-
ploited in Japan is about 340 mm. Therefore, the
lifespan of the Pacific saury is about one year in the
western Pacific. Our results are more consistent
with those of Kosaka (1979) who found two age
groups (0 and 1 year) than those (Sablin 1979) who
found three age groups (0, 1, and 2 years).
In Japan, fishing efforts of the Pacific saury is
regulated by fishing season as well as by the num-
ber of fishing boats. The fishing season starts in mid-
August. Pacific sauries hatched in the main spawn-
ing season are about 250 days old at this time of year
(approximately 270 mm) and are grovdng at the rate
of 0.8 mm/d. Thus a 2-wk postponement at the
beginning of fishing season would result in an 11
mm (10-15 g body weight) increase of average fish
*S. Kosaka, Tohoku Regional Fisheries Research Laboratory,
Fisheries Agency, Shiogama, Miyagi 985, Japan, pers. commun.
496
WATANABE ET AL.; GROWTH OF SAURY
length. The biomass and yield of Pacific saury need
to be reestimated based on faster growth rates
presented here.
ACKNOWLEDGMENT
We thank A. Nishimura of Hokkaido Salmon
Hatchery; J. Yamada of Hokkaido University; R.
Lasker, S. Tsuji, and N. C. H. Lo of Southwest
Fisheries Center of National Marine Fisheries Ser-
vice (NMFS), NOAA; and S. Kosaka and T. Wata-
nabe of the Tohoku Regional Fisheries Research
Laboratory for valuable suggestions and reading the
manuscript. We also thank A. W. Kendell Jr. of the
Northwest and Alaska Fisheries Center, NMFS,
NOAA, for providing us with some specimens from
the eastern Pacific, and D. Abramenkoff of the
Southwest Fisheries Center for help in reading Rus-
sian papers. Y. Watanabe would like to thank the
Science and Technology Agency of Japan for fund-
ing his stay at the Southwest Fisheries Center.
LITERATURE CITED
Ahlstrom, E. H.
1968. An evaluation of the fishery resources available to
California fishermen. In The future of the fishing industry
in the United States, p. 65-80. Univ. Wash. Publ. Fish.,
New Ser. IV.
Armstrong, P. R, and J. S. Child.
1965. Stages in the normal development of Fundulxis hetero-
clitus. Biol. Bull. 128:143-168.
Brothers, E. B., C. P. Mathews, and R. Lasker.
1976. Daily growth increments in otoliths from larval and
adult fishes. Fish. Bull., U.S. 74:1-8.
Brownell, C. L.
1983. Early growth rate and feeding of a small group of
laboratory-reared saury, Scomberesox saurus scombroides
(Pisces: Scomberesocidae). S. Afr. J. Mar. Sci. 1:245-248.
FUKUSHIMA, S.
1979. Synoptic analysis of migration and fishing conditions
of saury in the Northwest Pacific Ocean. (In Jpn.) Bull.
Tohoku Reg. Fish. Res. Lab. 41:1-70.
Gong, Y.
1984. Distribution and movements of Pacific saury, ColoJubis
saira (Brevoort), in relation to oceanographic conditions in
waters off Korea. Bull. Fish. Res. Dev. Agency 33:1-156.
Hara, M., a. Kijima, and Y. Fujino.
1982. Genetic study on population structure of Pacific saury.
(In Jpn.) Bull. Tohoku Reg. Fish. Res. Lab. 45:1-18.
Hatanaka, M.
1955. Biological studies on the population of the saury, Colo-
labis saira (Brevoort). Part I. Reproduction and growth.
Tohoku J. Agric. Res. 6:227-269.
Hotta, H.
1958. On the growth of the young saury, Cololabis saira, in
the rearing experiment. (In Jpn.) Bull. Tohoku Reg. Fish.
Res. Lab. 11:47-64.
1960. On the analysis of the population of the saury (Colo-
labis saira) based on the scale and the otolith characters.
and their growth. (In Jpn.) Bull. Tohoku, Reg. Fish. Res.
Lab. 16:41-64.
HUBBS, C. L., AND R. L. Wisner.
1980. Revision of the sauries (Pisces, Scomberesocidae) with
description of two new genera and one new species. Fish.
Bull., U.S. 77:521-566.
Hughes, S. E.
1974. Stock composition, growth, mortality, and availability
of Pacific saury, Cololahis saira, of the northeastern Pacific
Ocean. Fish. Bull., U.S. 72:121-131.
Jones, C.
1986. Determining age of larval fish with the otolith incre-
ment technique. Fish. Bull., U.S. 84:91-103.
Kim, Y. M., and Y. J. Park.
1981. A study on the growth of saury, Cololabis saira (Bre-
voort), based on the length composition in the Korean
waters. (In Korean) Bull. Fish. Res. Dev. Agency 27:
59-70.
Kosaka, S.
1979. Sanma no nenrei seicho ni tsuite (Age and growth of
saury). (In Jpn.) Rep. Japan-Soviet Coop. Res. Meet.
Saury Scomber 11:118-122.
Methot, R. D.
1981. Growth rates and age distributions of larval and juven-
ile northern anchovy, Engraulis mordax, with inferences on
larval survival. Ph.D. Thesis, Univ. California, San Diego,
209 p.
Nishimura, A., Y. Watanabe, and J. Yamada.
1985. Daily growth increment-like microstructure in otoliths
of the Pacific saury Cololabis saira. (In Jpn.) Bull. Tohoku
Reg. Fish. Res. Lab. 47:33-35.
Nishimura, A., and J. Yamada.
1984. Age and growth of larval and juvenile walleye pollock
Theragra chalcogramma (Pallas), as determined by otolith
daily growth increments. J. Exp. Mar. Biol. Ecol. 82:191-
205.
NoviKOV, Yu.V.
1960. Oprjedjenije vozrasta po chjeshchuje i vozrastnoi sostav
sairy (Cololabis saira (Brevoort)) v raionje ziuzhnykh
kuril'skikh ostrovov (Determination of age by scales and age
composition of saury (Cololabis saira (Brevoort)) in region
of southern Kuril Islands). (In Russ.) Izv. TINRO 46:233-
241.
1973. Izuchjenije populiatsionnoi struktury i vozrastnogo
sostava sairy sjevjero-zapadnoi chasti tikhogo okjeana (The
study of population structure and age composition of saury
of north-western part of the Pacific Ocean). (In Russ.) Izv.
TINRO 87:149-154.
NUMACHI, K.
1971. Kouso no takei ni yoru sanma keitougun ni kansuru
kenkyu (Study on the population of the Pacific saury based
upon analyses of enzymatic polymorphism). (In Jpn.)
Ocean Res. Inst., Univ. Tokyo, Tokyo, 57 p.
Odate, S.
1977. On the distribution of Pacific saury in the North Pacific
Ocean. (In Jpn.) Res. Inst. North Pac. Fish. Sp. Vol.,
p. 353-382. Fac. Fish., Hokkaido Univ., Hakodate, Ja-
pan.
Odate, K.
1986. Automatic data processing for estimation of abundance
of zooplankton in the waters off northeastern Honshu,
Japan, 1951-1976. an Jpn.) Bull. Tohoku Reg. Fish. Res.
Lab. 48:31-47.
Pannella, G.
1971. Fish otoliths: daily growth layers and periodical pat-
terns. Sci. 173:1124-1127.
497
Radtke, R. L., and J. M. Dean.
1982. Increment formation in the otoliths of embryos, larvae,
and juveniles of the mummichog, Fundulus heteroclitus.
Fish. Bull., U.S. 80:201-215.
RiCKER, W. E.
1973. Linear regressions in fishery research. J. Fish. Res.
Board Can. 30:409-434.
Sablin, V. V.
1979. Sanma no nenrei ni kansuru mondai ni tsuite (Problems
on age composition of saury). (In Jpn.) Rep. Japan-Soviet
Coop. Res. Meet. Saury Scomber 11:114-117.
Sablin, V. V., and V. P. Pavlychev.
1982. Dependence of migration and catch of Pacific saury
upon thermal conditions. Bull. Tohoku Reg. Fish. Res. Lab.
44:109-117.
Smith, P. E., E. H. Ahlstrom, and H. D. Casey.
1970. The saury as a latent resources of the California Cur-
rent. Calif Coop. Oceanic Invest. Rep. 14:88-130.
Smith, P. E., and R. W. Eppley.
1982. Primary production and the anchovy population in the
FISHERY BULLETIN: VOL. 86, NO, 3
Southern California Bight: Comparison of time series.
Limnol. Oceanogr. 27:1-17.
Statistics and Information Department, Japan.
1985. Fisheries statistics of Japan 1984. Minist. Agric, For.
Fish., Jpn., 288 p.
Sunada, J. S.
1974. Age and growth of the Pacific saury, Cololabis saira.
Calif Fish Game 60:64-73.
Theilacker, G. H.
1980. Changes in body measurements of larval northern
anchovy, Engraulis mordax, and other fishes due to handling
and preservation. Fish. Bull., U.S. 78:685-692.
UCHIDA, K., S. IMAI, S. Mito, S. Fujita, M. Ueno, Y. Shojima,
T. Senta, M. Tahuku, and Y. Dotsu.
1958. Studies on the eggs, larvae and juvenile of Japanese
fishes. (In Jpn.) Kyushu Univ., Fukuoka, Japan, 89 p.
YusA, T.
1960. Embryonic development of the saury Cololabis saira
(Brevoort). Bull. Tohoku Reg. Fish. Res. Lab. 17:1-14.
498
OTOLITH ULTRASTRUCTURE OF SMOOTH OREO, PSEUDOCYTTUS
MACULATUS, AND BLACK OREO, ALLOCYTTUS SR, SPECIES
N. M. Davies,' R. W. Gauldie/2 S. A. Crane,' and
R. K. Thompson^
ABSTRACT
The ultrastructure of sagittal otoliths from 14 Pseudocyttus maculatiis and 25 Allocyttus sp. individuals
were examined to determine their suitability for estimating age in these two species. Scanning electron
microscopy revealed high levels of complexity in both external surface topography and internal struc-
tural organization in the sagittae of both species. Many different crystal forms were found, including
calcite-like prisms. A close similarity in otolith structure exists between the two species. Deposition of
check rings analagous to annual and daily growth increments was found to be irregular with the underlying
complexity of crystalline growth obscuring the finer (analogous to daily) growth rings, making their
periodicity difficult to validate and implying that with present techniques the sagittal otoliths of the oreo
species Pseudocyttus maculatus and Allocyttus sp. are not suitable for age estimation.
The smooth oreo, Pseudocyttus maculatus, and
the black oreo, Allocyttus sp., are two related
species of the family Oreosomatidae. They are both
important commercial species in New Zealand. The
black oreo is the most commonly caught oreo in
New Zealand waters, while the smooth oreo is the
second-most commonly caught oreo. Little is known
about the biology of these fish. The black oreo is
endemic to New Zealand while the smooth oreo
occurs in New Zealand, South Australian, South
African, and South American waters (Last et al.
1983). In the waters south of New Zealand, the
distributions of the two species overlap (McMillan
1985). The habitat range of the smooth oreo is be-
tween 650 and 1,200 m, and that of the black oreo is
between 600 and 1,200 m (McMillan 1985).
A preliminary examination of the ultrastructure
of otoliths [sagittae] of these fish was undertaken
as part of a study to establish an ultrastructural
basis for a suitable ageing technique. This study
describes the external and internal structure and
organization of the otoliths in terms of the suitabil-
ity of the various check rings of the sagittae for age
estimation.
MATERIALS AND METHODS
Three otoliths (the sagitta, astericus, and lapillus)
'Fisheries Research Centre, Ministry of Agriculture and Fish-
eries, Greta Point, Evans Bay Parade, P.O. Box 297, Wellington,
New Zealand.
^To whom reprint requests should be sent.
'Electron Microscope Unit, Kirk Building, Victoria University
of Wellington, Private Bag, Wellington, New Zealand.
are contained in the endolymphatic sac (Fig. la). The
sagitta is the largest otolith and is located in the
most ventral position in the sac. The arrangement
of the three otoliths in the endolymphatic sac ranges
between the primitive where a large astericus and
sagitta with no lapillus is present (Gauldie et al.
1986) and the typical teleost arrangement where a
small astericus is located close to the sagitta, and
an even smaller lapillus is displaced into the atrium
of the semi-circular canal. The oreosomatids are
primitive fishes taxonomically, lying in the order
Beryciformes (Nelson 1976), and the arrangement
of otoliths reflects the taxonomic position of the fish.
The orientation of otoliths described here refers to
the orientation in situ. The lateral face is the out-
ward (antisulcal) surface; the medial face is the in-
ward (sulcal) surface. Investigation was restricted
to the sagitta primarily because of the difficulties
in establishing homologies for daily and annual type
check rings in the astericus and lapillus.
Sagittae were dissected from 14 smooth and 25
black oreo individuals caught in bottom trawls off
the east coast of New Zealand. These individuals
ranged in length from 24.5 to 40.1 cm (black oreo)
and 35.1 to 51.2 cm (smooth oreo).
Whole otoliths were photographed at 6 x to 20 x
using a WILD'* photomicroscope. The sagittae were
embedded on glass slides in epoxy resin with the
antisulcus surface uppermost and finely ground on
a Struers Planapol-2 petrographic grinder. The
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
Manuscript accepted April 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
499
FISHERY BULLETIN: VOL. 86, NO. 3
500
DAVIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO
Figure la. —Relative positions of the otoliths in the right endolymphatic sac (ENS) of the smooth oreo in lateral view: astericus (AS),
lapillus (LA), and sagitta (SA). Anterior (A), dorsal (D), posterior (P), and ventral (V) surfaces are indicated. Magnifi-
cation = 6.3 X.
b. —Dorsal (regular) and ventral (irregular) lobes of the otolith in lateral view, nucleus (N), sulcus (S), and rest as in Figure
la. Magnification = 12.5 x.
c. — SEM of the lateral surface topography of the otolith. Scale bar = 1 mm.
ground surface was polished with 2000 grit wet and
dry paper to yield a smooth surface for etching. The
most successful results were obtained by etching
with a 0.1 M solution of disodium salt of EDTA. The
otoliths were immersed in this solution for 15 to 20
minutes. Other suitable etching solutions employed
were 1) a 1% solution of HCl for 20 to 30 seconds
and 2) a 2% solution of Histolab RDO (a commer-
cial etching solution comprising a mixture of HCl
and EDTA) for 5 minutes. A cellulose acetate peel
was made of the etched surface to obtain an exact
replica of the surface features. The peel was placed
on a microscope slide under a cover slip, cleared with
ethanol or distilled water, viewed, and photographed
using a Zeiss photomicroscope. Direct observations
of thin sections (about 20 ^m) of otoliths did not
show any more information than that observed in
acetate peels. Acetate peels had the advantage of
allowing successive grinds to be examined thereby
avoiding the problem of losing information that
might be located only in very narrow layers within
the otolith.
Scanning electron microscope (SEM) photographs
were taken of otoliths using a Phillips 505 SEM.
Whole otoliths were glued on to SEM pin type
mounts, cemented in position with contact cement,
and sputter-coated with gold at approximately 5
Torr. The external surface topography of both the
medial and lateral faces of the otoliths was photo-
graphed. Selected pieces of otoliths broken by thumb
501
FISHERY BULLETIN: VOL. 86, NO. 3
pressure were examined and photographed to ob-
tain internal structural information. Finely ground
cross-section surfaces were also polished and etched
for examination with the SEM.
RESULTS
Smooth Oreo Otolith
The sagitta is clearly divided into two distinct
structures: a small, smooth, regular dorsal lobe and
a larger, irregular ventral lobe (Fig. lb). The irreg-
ular lobe has branched crystal formations and clefts
at the ventral edge. The lateral face topography is
complex (Fig. Ic). The central bumpy area contains
the nucleus between the two lobes. Radiating out-
wards from the nucleus are concentric ridges on the
antisulcul surface.
The crystal morphology of the lateral surface of
the sagitta is variable. Over much of the central
parts of the sagitta, large and variably oriented
crystals give a coarse appearance to the surface of
the otolith (Fig. 2a) which presumably obstructs, by
diffraction, potentially clear zones in the whole
otolith viewed by transmitted light. At the edge of
the regular lobe, the crystal type alters to form slabs
of crystal layers (Fig. 2b). The analogous area on
the irregular lobe yields variable crystalline struc-
tures with complex alignments (Fig. 2c). Deep
troughs and branching furrows break up the crystal
forms at the edge of the irregular lobe.
The medial surface has three distinct parts: the
central sulcal area, the edge of the irregular lobe,
and the edge of the regular lobe (Fig. 3a). Raised
ridges and two prominent knobs are found in the
central sulcal area. Crystals compacted into a petal-
like growth pattern are found in this part of the
otolith (Fig. 3b). Contrasting to this, more porous
crystal structures occur in the edge areas of the
regular lobe (Fig. 3c). Further variety is found on
the irregular lobe, where a very porous, honeycomb-
like crystal arrangement exists (Fig. 3d). At the
Figure 2a. —Coarse crystal structure of the central lateral surface of the smooth oreo otolith. Scale bar = 0.1 mm.
b. —Split-screen SEM of the transition of the crystal-type at the edge of the lateral surface of the regular
lobe. Magnification = 163 x and 652 x.
c. —Haphazard crystal alignments at the edge of the lateral face of the irregular lobe. Scale bar = 0.1 mm.
502
DAVIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO
50iim30.1kU 1.63E2 6648/99 SS07A
503
FISHERY BULLETIN: VOL. 86, NO. 3
SejinSe.lkU 1.63E2 6719/99 88676
Figure 3a. —Medial view of the smooth oreo otolith showing the irregular lobe (I), regular lobe (R), and sulcus
(S). Scale bar = 1 mm.
b. —Split-screen SEM of petallike crystal growth in the sulcus on the medial surface. Magnification
= 163 X and 1141 x.
504
DA VIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO
c
* -'-^
m-
\4
^
r*^. >\. ^ *^^ -^r ^9^ K^"^ yur
*H
\,
,?■ -■
18jiii3e.lkU 1.31E3 6717/99 88676
d a»
10jiii30.1kU 1.62E3 6749/99 SS014B
Figure 3c. —Porous crystals on the medial surface of the regular lobe. Scale bar = 10 tixn.
d.— Honeycomblike crystal structure on the medial surface of the irregular lobe. Scale bar = 10 fim.
505
FISHERY BULLETIN: VOL. 86, NO. 3
Figure 3e.— A laminar pattern of crystalline growth on the medial face of the irregular lobe. Scale bar = 10 fxm.
edge of this lobe, a laminar crystalline growth pat-
tern develops, adding to the overall variation of
crystal structures on the medial surface (Fig. 3e).
Other studies of otolith ultrastructure present the
otolith as homogeneous in crystal form, composed
almost entirely of monoclinal aragonite crystals
(Degens et al. 1969). The complex crystallinity of
the oreo otoliths resembles that of the mollusc shell,
v^hich, although aragonitic, often has a pattern of
complex variation in crystal habit (Carriker et al.
1980).
Within the broken otolith, the nucleus lies at the
center of a spherical primordium (Fig. 4a). Crystals
grow outwards from the primordium and epitaxial
(Degens 1976) growth patterns exist (Fig. 4b). Com-
plex leaf-shaped crystals occur in areas directly
beneath the lateral surface of the fractured irregular
lobe (Fig. 4c). Beneath the medial surface, a remark-
able series of hexogonal crystals of calcite occur as
large rectangular blocks embedded within the
otolith (Fig. 4d). Calcitic prisms have been described
in molluscs as resulting from the regeneration of
broken shells (Watabe 1983). It is difficult to imagine
otoliths being broken and regenerated in situ.
Major and minor check rings similar to those
described (Gaul die 1987) for otoliths from the orange
roughy, a deepwater species from the same habitat,
occur (Fig. 4e). When polished and etched, the check
rings become clearly visible with deep etched checks
occurring between the less deeply etched checks
analagous to microscopic growth increments in
other species (Fig. 4f). At a higher magnification the
finer increments become obscured by the coarseness
of the underlying crystal type (Fig. 4g). Those fine
increments that are visible occur irregularly and
have varying widths.
Large-scale rings analogous to opaque/hyaline an-
nual zones were observed in the regular lobe. The
mean width of these zones, measured using
transmitted light (Fig. lb), was 0.34 mm (±0.06).
The concentric ridges observed by SEM on the
lateral surface of the regular lobe (Fig. Ic) have a
mean width of 0.27 mm (±0.06). These two struc-
tures, opaque/hyaline zones and surface ridges, have
about the same width with no statistically signifi-
cant differences between them.
Examination of otolith cross sections reveals
widely spaced, large rings with many finer rings in
between (Fig. 4h). When examined in greater detail,
the demarcation between the large and finer rings
506
DA VIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO
a
Figure 4a. —Broken smooth oreo otolith reveals the spherical primordium (PM) and nucleus (N) and the surround-
ing radial crystal growth. Scale bar = 0.1 mm.
b.— Epitaxial crystal development in the broken otolith. Scale bar = 0.1 mm.
507
FISHERY BULLETIN: VOL. 86, NO. 3
10iim30.lkU 1.49E3 6828/99 SSOl
Figure 4c.— Leaf-shaped crystals beneath the lateral surface of the broken otolith. Scale bar = 10 ^m.
d.— Calcitic prism crystals beneath the medial surface of the broken irregular lobe. Scale bar = lOji.
508
DAVIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO
e -
Figure 4e.— Discontinuous uniform crystal growth forming rings in the broken otolith. Scale bar = 0.1 mm.
f. —Fine and deep (arrowed) increments on the polished and etched otolith surface. Scale bar = 0.1 mm.
509
FISHERY BULLETIN: VOL. 86, NO. 3
g '
i . V pit
10iiin30.1kU 1.31E3 6725/99 SSeSB
Figure 4g.— Fine increments obscured by coarse crystallinity on the polished and etched otolith surface. Scale bar = 10 jim.
h.— Fine increments (arrowed) found between widely spaced large rings in the otolith cross-section. Magnification = 72x .
510
DAVIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO
j r , - ' • . • . v-
.4,
o:^^"^^i*i ^.^
X \,* •^-.^r ^ ^:j-j
5^-
,-* «
^
■^-
Figure 4i. —Variation of increment widths (arrowed) in the otolith cross-section. Magnification = 160 x.
j.— Cellulose-acetate peel showing the intermittency of fine increment sequences (arrowed), often obscured by a coarse underly-
ing crystallinity. Magnification = 1 60 x .
511
FISHERY BULLETIN: VOL. 86, NO. 3
becomes difficult to determine because of the inter-
mittent nature of the growth increments and the
variety of width-sizes (Fig. 4i). In large areas of the
otolith, increments appear to be absent or indeter-
minate, with a coarse underlying crystal structure
(Fig. 4j) making accurate determination of incre-
ment sequences difficult. However, the fine incre-
ments of the oreo otolith are 3 to 5 ^m wide, which
is within the range of daily growth increments
described for other species (Jones 1986; Gauldie in
press).
Black Oreo Otolith
The black oreo otolith is almost identical to that
found in the smooth oreo in overall shape, propor-
tion, structure, topography, surface and internal
crystallinity, and increment pattern. Some minor
differences do, however, exist.
In the medial sulcus, the prominent knobs found
in the smooth oreo otolith are smaller than in the
black oreo. On the surface of the sulcus, large leaf-
like crystals having various orientations occur (Fig.
5a). Also present in the sulcus are porous, sponge-
like crystals adjacent to membranous structures
(Fig. 5b). Smooth patches, where crystals appear ab-
sent, occur on the lateral surface of the otherwise
coarsely crystalline irregular lobe (Fig. 5c). At
higher magnification the smooth patches are seen
to be smaller growth forms of the larger adjacent
crystals.
DISCUSSION
Despite some minor differences in topography and
crystallinity, the sagittae of both species are essen-
tially identical. The otoliths are structurally complex
v^ath a great variety of crystalline forms. The coarse
Figure 5a. —Split-screen SEM of large leaflike crystals in the black oreo otolith sulcus. Magnification = 356 x and 979 x .
b.— Porous, sponge-like crystals in the otolith sulcus. Scale bar = 10 ^^m.
c. —Smooth patches on the central lateral surface of the otolith. Scale bar = 0.1 mm.
512
DA VIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO
513
FISHERY BULLETIN: VOL. 86, NO. 3
crystals on the central lateral surface are com-
parable with those in the oyster shell described as
individual laths (Carriker et al. 1980). The leaflike
crystals in the sulcus of the black oreo otolith are
similar to the chalky crystal forms in the oyster shell
(Carriker et al. 1980). Such a variety of crystalline
forms is uncommon in teleost otoliths. The low legi-
bility of structures of various kinds in the otolith may
reflect this complex crystallinity. However, the com-
plex crystallinity of the mollusc shell is thought to
reflect changes in both the external and internal
milieu of the organism (Wilbur and Saleuddin 1983).
Thus the difficulties of reading the oreo otolith in
the conventional sense may be offset by the life
history record (albeit difficult to translate) provided
by its complex crystallinity.
The broken sections of the otolith reveal the in-
ternal structure organization, and development of
crystals. Epitaxial crystal grov^h in the oreo otolith
results in columnar, monoclinal crystals of ara-
gonite. However, the presence of calcite-like prisms
has not been described for other otoliths. Calcite
occurs on the antisulcal surface of some otoliths
apparently by simple crystallization out of the fluid
of the endolymphatic sac (Morales-Nin 1985), but
calcite has never been described from within an
aragonite otolith (Carlstrom 1963). In molluscs,
calcite replacement of aragonite results in an orderly
alignment of calcite crystals following the alignment
of the original aragonite crystals. The disorderly
appearance of the calcite-like hexagons in the
smooth oreo otolith may be due to a diagenetic
transformation of aragonite to calcite with depth.
The compensation depth for the aragonite/calcite
transformation is about 3,000 m (Fyfe and Bischoff
1965) well beyond the known range of the smooth
oreo which has a maximum recorded depth of 1,300
m. However, there may be enough variation in
either the kind or amount of stabilizing protein in
the smooth oreo otolith to allow crystal changes to
occur at shallower depths than 3,000 m.
The complex, and often coarse, crystal structure
of the oreo otolith obscures the sequences of incre-
ments when they do occur. As a result, large rings
observed at low magnification become indistinct at
higher magnification when many finer increments
appear. The large rings could be assumed to be
annual check rings, but the difficulties in differen-
tiating between the fine and large rings create am-
bivalence in one's interpretation. The finer micro-
scopic growth increments, analagous to daily growth
rings, have no uniform width and occur intermit-
tently making accurate counting impossible.
The suitability of an otolith for determining the
age of a fish depends on the pattern of both annual
and daily check rings inferred from the structure
of the otolith. The hyaline/opaque zones observed
in the regular lobe (using transmitted light) had a
similar mean width to the concentric ridges found
by SEM on the lateral surface. The ^statistic we
obtained accepts the null hypothesis that no sig-
nificant difference exists between the two means.
Mel'nikov (1981) regarded these opaque/hyaline
zones as annual check rings in the otolith of Allo-
cyttus verrucosus. However, because no evidence ex-
ists for a relationship between the surface ridges and
fish age, it is possible that Mel'nikov's (1981) ages
are incorrect. Furthermore, the width of the opaque/
hyaline zones (0.34 mm) would indicate daily growth
increments less than 1 \xm wide. There are no
reports in the literature of validated daily growth
rings of such small size. In addition, the microscopic
growth increments which we have observed in the
oreo otolith are 3 to 5 ^m wide, which is a size range
commonly observed in other species.
With the techniques available we have been unable
to use either annual- or daily-type structures to
develop a technique for age estimation for Pseudo-
cyttus maculatus and Allocyttu^ sp. The reasons for
these difficulties may lie in the crystal morphology
of the otoliths which are more complex than any so
far described in the literature.
ACKNOWLEDGMENTS
All otoliths were supplied by Peter McMillan (Fish-
eries Research Centre). All 35 mm photographs
were processed by Alan Blacklock (Fisheries Re-
search Centre). SEM photographs were taken at the
SEM Unit, Zoology Department, Victoria Univer-
sity of Wellington.
LITERATURE CITED
Carlstrom, D.
1963. A crystallographic study of vertebrate otoliths. Biol.
Bull. (Woods Hole) 125:441-463.
Carriker, M. R., R. E. Palmer, and R. S. Prezant.
1980. Functional ultra-morphology of the dissoconch values
of the oyster Crassostrea virginica. Proc. Natl. Shellfish.
Assoc. 70:139-183.
Degens, E. T.
1976. Molecular mechanisms of carbonate, phosphate and
silica deposition in the living cell. Top. Curr. Chem. 64:
1-112.
Degens, E. T., W. G. Deuser, and R. L. Haedrich.
1969. Molecular structure and composition of fish otoliths.
Mar. Biol. 2:105-113.
Fyfe, W. S., and J. L. Bischoff.
1965. The caJcite-aragonite problem. In L. C. Pray and R. C.
Murray (editors), Dolomitization and limestone diagenesis,
514
DA VIES ET AL.: OTOLITH ULTRASTRUCTURE OF SMOOTH AND BLACK OREO
p. 3-13. Symp. Soc. Econ. Minist. Spec. Publ. 13.
Gauldie, R. W.
1987. The fine structure of check rings in the otolith of the
New Zealand orange roughy. NZ J. Mar. Freshwater Res.
21:267-274.
In press. A study of otolith daily growth rings in the orange
roughy {Hoplostethus atlanticus) aimed at resolving prob-
lems in age, growth, recruitment and otolith architecture.
NZ Fish. Res. Bull.
Gauldie, R. W., D. Dunlop, and J. Tse.
1986. The remarkable lungfish otolith. NZ J. Mar. Fresh-
water Res. 20:81-92.
Jones, C.
1986. Determining age of larval fish with the otolith incre-
ment technique. Fish. Bull., U.S. 84:91-103.
Last, P. R., E. 0. G. Scott, and F. H. Talbot.
1983. Fishes of Tasmania. Hobart, Tasmanian Fisheries
Development Authority.
Mel'nikov, Y. S.
1981. Size-age composition and growth pattern oiAUocyttus
verrucosus (Oreosomatidae). Ichthyol. (Engl, transl. Vopr.
Ikhtiol.) 21:178-184.
Morales-Nin, B.
1985. Caracteristicas de los otolitos cristalinos de Genypterus
capensis, (Smith, 1847) (Pisces: Ophidiidae). Invest. Pesq.
49:379-386.
McMillan, P. J.
1985. Black and smooth oreo dories, /n J. A. Colman, J. L.
McKoy, and G. G. Baird (editors), Background papers for
the 1985 total allowable catch recommendations. (Report
held in the Fisheries Research Centre Library, P.O. Box 297,
Wellington, New Zealand.)
Nelson, J. S.
1976. Fishes of the world. John Wiley and Sons, N.Y., 416 p.
Watabe, N.
1983. Shell repair. In A. S. M. Saleuddin and K. M. Wilbur
(editors). The mollusca, No/4, p 289-316. Acad. Press, N.Y.
Wilbur, K. M., and A. S. M. Saleuddin.
1983. Shell formation. In A. S. M. Saleuddin and K. M.
Wilbur (editors), The mollusca, No/4, p 236-288. Acad.
Press, N.Y.
515
RELATIONSHIP BETWEEN SEDIMENT CHARACTER AND
SEX SEGREGATION IN ENGLISH SOLE, PAROPHRYS VETULUS'
D. Scott Becker^
ABSTRACT
English sole, Parophrys vetulus, were collected by otter trawl in the nearshore zone of Puget Sound,
Washington, during two surveys conducted in 1981-82 and 1984. Stations in both surveys were distributed
across a broad range of sedimentary environments. Sex segregation by English sole was strongly
associated with the grain-size characteristics of bottom sediments. This association was persistent across
a variety of sampling conditions, including different years, seasons, embayments, and depths. Fish age
did not appear to influence the observed association. Results of this study provide the first documenta-
tion of the influence of sediment character on sex segregation by a pleuronectid, and suggest that this
relationship should be considered in future studies of English sole and, perhaps, other pleuronectids as well.
Sex segregation by fishes of the family Pleuronec-
tidae (i.e., righteye flounders) has been documented
for numerous species from the United States (Alver-
son and Chatwin 1957; Fadeev 1970; Alton 1972),
Europe (Rae 1965), and Asia (Moiseev 1953;
Kovtsova 1982). These patterns have been related
to such factors as geographic location, depth,
season, and age of fish. However, because most of
these observations have been made while fishes
were being collected for other purposes, the under-
lying reasons for the observed patterns are poorly
understood.
Several studies have documented sex segregation
in English sole, Parophrys vetulus, a pleuronectid
found in nearshore areas along the west coasts of
the United States and Canada (Hart 1973). In de-
scribing the catch of the English sole fishery in
Hecate Strait, British Columbia, Ketchen (1956)
noted that considerable sex segregation takes place,
particularly during the summer months. In a study
of the population characteristics of this species in
Puget Sound, Holland (1969) found a difference of
18% between the percentages of males in the pop-
ulations from two locations. Because both of these
accounts were largely anecdotal, no evaluations of
the observed patterns were made.
One factor that has yet to be evaluated with
respect to sex segregation by pleuronectids is the
character of bottom sediments. Because pleuronec-
tids live in close association with the sea floor, it
'Contribution No. 759, School of Fisheries, University of Wash-
ington, Seattle, WA 98195.
^School of Fisheries, University of Washington, Seattle, WA
98195; present address: PTI Environmental Services, 3625 132nd
Ave. S.E., Bellevue, WA 98006.
might be surmised that the characteristics of that
environment exert a differential influence on the
distributions of the two sexes. Additional support
for this hypothesis comes from the fact that the
influence of sediment character on the overall
distribution (i.e., both sexes pooled) of several
pleuronectids has been well documented (Ketchen
1956; Alverson et al. 1964; Feder et al. 1974; Pearcy
1978; Scott 1982; Becker 1984).
In this paper, I evaluate patterns of sex segrega-
tion by English sole in relation to the character of
bottom sediments in nearshore areas of Puget
Sound, WA. This study provides the first determina-
tion of how sediment character influences sex segre-
gation in a pleuronectid. In relating sediment
character to sex segregation in English sole, I also
evaluate the potential influences of depth, season,
and fish age on the observed relationships.
METHODS
English sole were sampled during two independ-
ent surveys conducted in several embayments of
Puget Sound in 1981-82 and 1984. The relationship
between sex segregation and sediment character
was evaluated separately for each survey. Although
the two surveys differed with respect to such vari-
ables as station locations, sampling times, and sam-
ple sizes, the basic collection and analytical methods
used in both surveys were similar.
In both surveys, the sex ratio of the English sole
captured at each station was represented by the pro-
portion of the sample comprised of males (i.e., male
proportion). Sediment character at each station was
represented by the percentage (by weight) of the
Manuscript accepted April 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
517
FISHERY BULLETIN: VOL. 86, NO. 3
grain-size distribution comprised of fine-grained
material (i.e., <0.063 mm in size). This variable was
used as an index of sediment character because
many physical, chemical, and biological character-
istics of sedimentary environments covary with the
grain-size distribution (Buchanan 1984).
During 1981-82, English sole were collected from
15 May to 6 July 1981 at 12 transects in Commence-
ment Bay, Elliott Bay, and Sinclair Inlet (Fig. 1).
Fishes also were collected during three additional
time periods (23-24 November 1981; 8-9 March
1982; 30-31 August 1982) to evaluate seasonal
patterns at three stations in Commencement Bay
(CB-1, CB-4, and CB-5; Fig. 1). These stations were
selected to represent a gradient of sediment grain
size. Although the seasonal sampling occurred over
a 2-yr period, it was assumed that interannual varia-
tion was not sufficiently large to obscure major
seasonal patterns.
During the May- July sampling, English sole were
collected at each transect during morning (07:00-
09:30), midday (10:00-13:30), evening (20:30-23:30),
and midnight (00:30-04:00). During the remaining
sampling periods, fish were collected only during
midday and evening. For each transect, results ob-
tained for different periods of the diel cycle were
pooled before male proportions were calculated. Sex
determinations and evaluations of spawning condi-
tion were made for all English sole larger than 160
mm total length (TL). Age was not determined.
During 1984, English sole were collected from 4
to 9 June 1984 at 15 transects in Commencement
Bay and at 2 transects in Carr Inlet (Fig. 1). Sam-
pling was conducted during daylight hours (06:30-
17:00) at all transects. Sex determinations and
evaluations of spawning condition were made for all
individuals larger than 225 mm TL. Age was deter-
mined by otolith analysis for all fish.
During both surveys, English sole were collected
using a 7.6 m (headrope) Marinovich otter trawl
having a body mesh of 3.2 cm (stretched) and a cod
end liner mesh size of 0.8 cm (stretched). Trawling
was conducted along isobaths at a constant vessel
speed of approximately 2.5 knots.
During both surveys, bottom sediments were sam-
pled using a modified van Veen bottom grab and the
grain-size distribution of the top 2 cm of sediment
was determined using standard sieve and pipette
techniques (Folk 1968). During 1981-82, sediments
were collected at five sampling points located at
approximately equal distances along each transect.
During 1984, sediments were collected at a variable
number of stations located within 300 m from each
transect.
The association between percent fine-grained
sediment and male proportion of English sole was
tested using Spearman's rank correlation coefficient
{Vg). The association between station depth and
male proportion was tested in the same manner.
Differences in male proportion among age-classes
of fish and among the seasonal sampling transects
were tested using the G-test (Sokal and Rohlf
1981).
RESULTS
Station depth, percent fine-grained sediment,
sample size, and male proportion at each transect
from the 1981-82 and 1984 surveys are presented
in Tables 1 and 2, respectively. During 1981-82,
4,430 English sole were sampled from stations
ranging in depth from 8 to 36 m. Percent fine-
grained sediment ranged from 5.8 to 92.1% and
male proportion ranged from 0.20 to 0.91. During
1984, 1,007 English sole were collected from
stations ranging in depth from 5 to 20 m. Per-
cent fine-grained sediment ranged from 8.6 to
87.5% and male proportion ranged from 0.13 to
0.98.
The correlation between percent fine-grained sedi-
ment and male proportion of English sole was
significant (P < 0.01; r, = 0.73) for the May- July
sampling of the 1981-82 survey (Fig. 2). This cor-
relation was also significant (P < 0.001; r^ = 0.80)
for the 1984 survey (Fig. 2). Correlations between
depth and male proportion were not significant {P
> 0.05) for either the 1981-82 (r, = -0.10) or 1984
(Vg = -0.04) survey.
For the stations sampled seasonally in 1981-82
(CB-1, CB-4, and CB-5), heterogeneity of male pro-
portion among stations was significant {P < 0.05)
during all four sampling periods (Fig. 3). In all cases,
male proportion was lowest at CB-5, highest at
CB-4, and intermediate in magnitude at CB-1. This
gradient of increasing male proportion was consis-
tent with the gradient of increasing percent fine-
grained sediment at the three stations (Table 1). The
mean abundance of English sole at the three sta-
tions rose from a minimum level in March to a peak
in May-July, and then declined through August to
a level in November that was similar to that in
March (Fig. 3).
For both surveys combined, 64 (1.2%) English sole
were found to be in spawning condition. Twenty-
four of those individuals were collected during
March, to represent a spawning component of 14.8%
during that month. The remaining 42 individuals
were sampled between May and July to represent
518
BECKER: SEDIMENT AND SEX SEGREGATION IN ENGLISH SOLE
r^
^^ Elliott Bay
^^EB-1
/^
'^^EB-2
SEATTLE
I-
EB-4
LEGEND
CR-Carr Inlet
E B - Elliott Bay
SN - Sinclair Inlet
10
ii NALmCAL MLES
n
POINT DEFIANCE
LEGEND
HY
Hyielxjs Waterway
BL
Blair Waterway
SI
Silcum Waterway
Ml
Milwaukee Waterway
SP
St Paul Waterway
MD
Middle Waterway
CI
City Waterway
SS
Souttiwest Shoreline
CB
Commencement Bay
KILOMETERS
Figure 1.— Locations of transects sampled in 1981-82 (denoted by CB, EB, and SN) and 1984 (all remaining
transects).
519
FISHERY BULLETIN: VOL. 86, NO. 3
Table 1. — Characteristics of transects and samples of English sole during 1981-82.'
Mean percent
fine-grained
SampI
e size^
Male proportion
Depth
May-
Novem-
May-
Novem-
Station^
(m)
sediment^
July
August
ber
March
July
August ber
March
CB-1
10
27.1
655
249
65
90
0.83
0.74 0.58
0.57
CB-2
33
47.0
554
0.78
CB-3
36
44.2
160
0.79
CB-4
14
71.6
589
146
174
38
0.91
0.82 0.77
0,87
CB-5
13
7.6
181
67
68
35
0.55
0.64 0.44
0.20
CB-6
35
11.2
128
0.66
EB-1
14
56.0
82
0.68
EB-2
31
76.2
17
0.76
EB-3
16
5.8
90
0.53
EB-4
32
8.8
49
0.49
SN-1
8
92.1
683
0.84
SN-2
10
13.7
310
0.30
'Sampling in May-July and November was conducted in 1981 , whereas sampling in August and March was conducted in 1982.
^Locations of transects are presented in Figure 1 .
'Each value Is based on five replicate samples.
<AII English sole were larger than 160 mm TL.
Table 2.— Characteristics of transects and samples of English sole
during 1984.
Mean percent
Male
Depth
fine-grained
Sample
propor-
Station'
(m)
sediment^
size^
tion
CR-1
18
14.5 (2)
60
0.28
CR-2
12
8.6 (2)
60
0.13
SS-1
9
18.2 (2)
59
0.29
SS-2
10
22.9 (6)
60
0.48
SS-3
20
8.7 (6)
60
0.35
CI-1
5
68.8 (4)
57
0.75
CI-2
9
58.2 (4)
52
0.73
MD-1
6
40.4 (2)
59
0.42
SP-1
5
48.9 (3)
60
0.77
MI-1
10
87.5 (5)
60
0.87
SI-1
12
73.8 (5)
60
0.98
BL-1
11
55.2 (1)
60
0.77
BL-2
12
74.3 (9)
60
0.88
BL-3
10
55.4 (6)
60
0.88
HY-1
9
68.1 (5)
60
0.45
HY-2
9
79.8 (7)
60
0.80
HY-3
9
61.1 (12)
60
0.73
'Locations of transects are presented in Figure 1.
^Number of stations used to calculate mean percent fine-grained sediment
at each transect is given in parentheses.
'All English sole were larger than 225 mm TL.
Z
g
I-
o
Q.
o
DC
Q.
LU
Figure 2.— Comparison of percent fine-grained sediment and male
proportion of English sole populations using Spearman's rank
correlation coefficient (rj. *• = P < 0.01, *** = P < 0.001.
i.u -
•
0.8 -
•
••
•
•
•
•
0.6 -
•:
0.4 -
•
n? -
MAY- JULY 1981
rs=0.73**
n
1
1 1
20
40
60
80
100
i.u -
•
• • •
0.8-
•
0.6-
0.4-
•
•
•
•
••
n? -
JUNE 1984
•
rs = 0.80***
n
1
1
1 1
20 40
60 80 100
PERCENT FINE-GRAINED
SEDIMENT
520
BECKER: SEDIMENT AND SEX SEGREGATION IN ENGLISH SOLE
CB-5 illij CB-1
CB-4
***
***
o
o
Q.
o
oc
a.
UJ
B
MARCH MAY-JULY AUGUST NOVEMBER
1982 1981 1982 1981
SAMPLING PERIOD
Figure 3.— Comparisons of male proportion of English sole populations at three Commencement
Bay stations during four sampling periods using the G-test. Mean abundance of English sole at
all three stations is plotted for each sampling period. The values of percent fine-grained sedi-
ment at Stations CB-5, CB-1, and CB-4 were 7.6%, 27.1%, and 71.6%, respectively. ' = P <
0.05, ••• = P< 0.001.
a spawning component of 0.9% during that period.
No fish were found to be in spawning condition dur-
ing August and November.
For English sole aged 3-10 years old during the
1984 study, male proportion for each age-class
ranged from 0.50 to 0.66 (Fig. 4). However, the
heterogeneity among age-classes was not significant
(P > 0.05). Nineteen fish aged 2 or >10 years old
were not included in this analysis because sample
sizes by age-class were too small.
OC
o
Q.
o
DC
Q.
LU
0.8 -1
0.6-
0.4
0.2-
126 217 261 209 97 40 24 14
Figure 4.— Comparison of male proportion among age-
classes of English sole collected in 1984. Sample size
is presented above each bar.
T 1 \ \ \ 1 I r
3456789 10
FISH AGE (yr)
521
FISHERY BULLETIN: VOL. 86, NO. 3
DISCUSSION
Results of this study demonstrate that sex segre-
gation by English sole in nearshore areas of Puget
Sound is strongly associated with the grain-size
distribution of bottom sediments. In addition to the
results of statistical analyses, the strength of this
association is demonstrated by its persistence across
a variety of sampling conditions. For example, the
association was found in two independent surveys
conducted 2-3 years apart. Furthermore, the 1981-
82 survey represented an analysis of large-scale
patterns across three embayments, whereas the
1984 survey focused on small-scale patterns within
a single embayment. The association between sex
segregation and sediment grain size also persisted
across seasons and across a range of depth.
Although a significant association was found be-
tween sediment grain size and male proportion of
English sole, a variety of factors that covary with
sediment grain size may have been partly or totally
responsible for this association (Lagler et al. 1977;
Moyle and Cech 1982). Examples of physical/chem-
ical variables that often covary with grain size and
may influence fish distributions include depth, cur-
rent speed, turbidity, and dissolved oxygen. Ex-
amples of biological factors include the composition
of prey assemblages and the distribution of pred-
ators. Although results of this study demonstrated
that depth did not influence the observed associa-
tion between sediment grain size and male propor-
tion, further experimentation is required to identify
the factor or suite of factors that directly influences
sex segregation in this species.
Despite the fact that depth did not influence the
patterns of sex segregation identified for English
sole, the range of depth considered in this study
(5-36 m) is a small fraction of the total range
occupied by this species (0-550 m; Hart 1973). It
therefore is possible that depth-related sex segrega-
tion in this species may be observed if a wider range
of depths were examined. For example, considerable
sex segregation by depth was found for Dover sole,
Microstomus pacificus, off the west coast of the
United States when a large depth range (110-440
m) was evaluated (Hagerman 1952; Alton 1972).
Based on results of the 1984 survey, no relation-
ship between age and male proportion in English
sole was evident. It therefore is unlikely that age
influenced the observed relationship between male
proportion and sediment grain size. A decline in
male proportion with increasing age has been found
for several pleuronectids (Rae 1965; Kovtsova 1982).
Because most of the fish examined in this study were
relatively young (<8 years old) compared with the
maximum age of 15-17 years sometimes reached by
individuals of this species (Ketchen 1956; Holland
1969; Van Cleve and El Sayed 1969), consideration
of older fish may alter the patterns observed in this
study. Although sample sizes for fish aged 9-10
years were relatively small in this study, male pro-
portion in those age groups (0.50) was lower than
that observed in all younger age groups (0.57-0.66).
The season of sampling did not alter the observed
relationship between male proportion and sediment
grain size. This relationship remained intact despite
the fact that the mean abundance of English sole
changed substantially throughout the seasonal cycle.
The observed seasonal changes in mean abundance
are consistent with the seasonal pattern of migra-
tion exhibited by adults of this species and most
other pleuronectids. That pattern includes a migra-
tion to deeper water in the fall for overwintering
and spawning, and a return migration to shallower
water in spring for feeding throughout the summer
(Ketchen 1956; Alverson et al. 1964; Roff 1982). It
is unlikely that spawning substantially influenced
the observed seasonal patterns of sex segregation
in English sole, because the largest percentage of
fish in spawning condition during any season was
<15%. However, sex segregation in relation to
spawning in pleuronectids has been documented for
several species (Moiseev 1953; Alverson and Chat-
win 1957), and may influence sex segregation in
English sole when a substantial percentage of the
population is in spawning condition.
Results of this study suggest that the influence
of sex segregation should be considered when
characteristics of English sole populations are com-
pared among habitats that differ with respect to
sediment character. Patterns based on population
characteristics capable of exhibiting a dependence
on sex (growth, condition, disease prevalence, tissue
contamination) could be strongly biased if differ-
ences in the sex ratio among stations are not con-
sidered. This concern should probably be extended
to studies of other pleuronectids, because the rela-
tionship between sex segregation and sediment
character is unknown for most species.
Although the reasons for the observed patterns
of sex segregation in English sole are unknown, a
potential explanation is that different energetic
needs induce male and female fish to utilize differ-
ent habitats. Female English sole grow to a larger
size than males and mature at an older age (Ketchen
1956; Holland 1969; Van Cleve and El-Sayed 1969;
Hart 1973). In addition, fecundity in females is pro-
portionate to individual size (Ketchen 1956; Hart
522
BECKER: SEDIMENT AND SEX SEGREGATION IN ENGLISH SOLE
1973). These patterns are found for numerous
pleuronectids and are thought to allow females to
outgrow predators and maximize fecundity (Roff
1982). Female English sole may therefore select
different habitats than males to support their differ-
ential reproductive requirements. Because many
characteristics of benthic macroinvertebrate assem-
blages (i.e., the primary prey of English sole) are
related to sediment character (Gray 1974; Rhoads
1974; Pearson and Rosenberg 1978), selection of
habitats that differ in sediment grain size provides
females with different prey spectra and energetic
potentials than are available to males. In addition,
differential habitat selection would reduce the poten-
tial of competition for food between sexes.
In summary, sex segregation by English sole was
strongly associated with the grain-size character-
istics of bottom sediments. This association was
persistent across a variety of sampling conditions,
including different years, seasons, embayments, and
depths. Fish age did not appear to influence the
observed association. Results of this study provide
the first documentation of the influence of sediment
character on sex segregation in a pleuronectid, and
suggest that this relationship should be considered
in future studies of English sole and, perhaps, other
pleuronectids as well.
ACKNOWLEDGMENTS
The 1981-82 survey was supported by the
National Oceanic and Atmospheric Administration
(Contract NA80RAD00050), and was part of a
dissertation submitted to the School of Fisheries of
the University of Washington (Seattle) in partial
fulfillment of a Ph.D. degree for the author; A. J.
Mearns was the Project Officer. The 1984 survey
was supported under Cooperative Agreement No.
CX810926-01-0 between the U.S. Environmental
Protection Agency and the State of Washington
Department of Ecology; J. D. Krull of Ecology was
the Project Manager. I thank R. A. Pastorok and
two anonymous reviewers for their helpful sugges-
tions on the manuscript.
LITERATURE CITED
Alton, M. S.
1972. Characteristics of the demersal fish fauna inhabiting
the outer continental shelf and slope off the northern Oregon
coast. In A. T. Pruter and D. L. Alverson (editors), The
Columbia River estuary and adjacent ocean waters, p. 586-
634. Univ. Wash. Press, Seattle.
Alverson, D. L., and B. M. Chatwin.
1957. Results from tagging experiments on a spawning stock
of petrale sole, Eopsetta jordani (Lockington). J. Fish. Res.
Board Can. 14:953-974.
Alverson, D. L., A. T. Pruter, and L. L. Ronholt.
1964. A study of demersal fishes and fisheries of the north-
western Pacific Ocean. H. R. MacMillan Lect. Fish., Inst.
Fish., Univ. B.C., Vancouver, 190 p.
Becker, D. S.
1984. Resource partitioning by small-mouthed pleuronectids
in Puget Sound, Washington. Ph.D. Thesis, Univ. Wash-
ington, Seattle, 120 p. + appendices.
Buchanan, J. B.
1984. Sediment analysis. In N. A. Holme and A. D.
Mclntyre (editors), Methods for the study of marine benthos,
p. 41-65. Blackwell Scientific Publications, Boston, MA,
387 p.
Fadeev, N. S.
1970. The fishery and biological characteristics of yellowfin
soles in the eastern part of the Bering Sea. Transl. in Soviet
Fisheries Investigations in the northeastern Pacific (Israel
Program Sci. Transl), Part V, p. 332-396.
Feder, H. M., C. H. Turner, and C. Limbaugh.
1974. Observations on fishes associated with kelp beds in
southern California. Calif. Fish Game, Fish Bull. 160, 138 p.
Folk, R. L.
1968. Petrology of sedimentary rocks. Univ. Texas, Austin,
170 p.
Gray, J. S.
1974. Animal-sediment relationships. Oceanogr. Mar. Biol.
Annu. Rev. 12:223-261.
Hagerman, F. B.
1952. The biology of the Dover sole, Microstomus pacificvs
(Lockington). Calif. Fish Game, Fish Bull. 85, 48 p.
Hart, J. L.
1973. Pacific fishes of Canada. Fish. Res. Board Can., Bull.
180, 740 p.
Holland, G. A.
1969. Age, growth, and mortality of races of English sole
{Parophrys vetulus) in Puget Sound, Washington. Pac.
Mar. Fish. Comm., Bull. 7:36-50.
Ketchen, K. S.
1956. Factors influencing the survival of the lemon sole
(Parophrys vetulus) in Hecate Strait, British Columbia. J.
Fish. Res. Board Can. 13:647-694.
KOVTSOVA, M. V.
1982. Size-age structure and sex ratio of the Barents Sea
population of the plaice, Pleuronectes platessa (Pleuronec-
tidae). J. Ichthyol. 22:68-79.
Lagler, K. F., J. E. Bardach, R. R. Miller, and D. R. M.
Passino.
1977. Ichthyology. John Wiley and Sons, N.Y., 506 p.
MOISEEV, P. A.
1953. Cod and flounder of the Far Eastern Seas. Fish. Res.
Board Can. Transl. Ser., No. 119, 576 p.
MOYLE, P. B., AND J. J. CeCH.
1982. Fishes: an introduction to ichthyology. Prentice-Hall,
Englewood Cliffs, NJ, 593 p.
Pearcy, W. G.
1978. Distribution and abundance of small flatfishes and other
demersal fishes in a region of diverse sediments and bathym-
etry off Oregon. Fish. Bull., U.S. 76:629-640.
Pearson, T. H., and R. Rosenberg.
1978. Macrobenthic succession in relation to organic enrich-
ment and pollution of the marine environment. Oceanogr.
Mar. Biol. Annu. Rev. 16:229-311.
Rae, B. B.
1965. The lemon sole. Fishing News Ltd., Lond., 106 p.
523
FISHERY BULLETIN: VOL. 86, NO, 3
Rhoads, D. C. shelf. Can. J. Fish. Aquat. Sci. 39:943-947.
1974. Organism-sediment relations on the muddy sea floor. SoKAL, R. R., AND F. J. Rohlf.
Oceanogr. Mar. Biol. Annu. Rev. 12:263-300. 1981. Biometry. W. H. Freeman and Co., San Francisco,
ROFF, D. A. CA, 859 p.
1982. Reproductive strategies in flatfish: a first synthesis. Van Cleve, R., and S. Z. El Saved.
Can. J. Fish. Aquat. Sci. 39:1686-1698. 1969. Age, growth, and productivity of an English sole
Scott, J. S. {Parophrys vetuliis) population in Puget Sound, Washing-
1982. Selection of bottom type of groundfishes of the Scotian ton. Pac. Mar. Fish. Comm., Bull. 7:51-71.
524
OCCURRENCE OF CANCER CRAB (C MAGISTER AND
C. OREGONENSIS) MEGALOPAE OFF THE WEST COAST OF
VANCOUVER ISLAND, BRITISH COLUMBIA
Glen S. Jamieson and Antan C. Phillips^
ABSTRACT
The seasonal and cross-shelf occurrences of Cancer crab (C. magister and C. oregonensis) megalopae
in 1985 along a transect line perpendicular to the coast off Tofino, British Columbia, are presented.
Megalopae of both species were generally absent from surface waters during daylight hours. The two
species may have slight temporal differences in nocturnal surface abundance, with C. magister occur-
ring later in the evening and earlier in the morning than C. oregonensis. Their relative occurrence at
the surface during the night was used to calibrate cross-shelf megalopal abundance data. Cross-shelf
megalopal intermolt stage proportions were calculated, relating degree of megalopal development to
proximity of nearshore habitat required for successful settlement of larvae.
Cancer magister megalopae were present from April to August, with peak abundance in May and
June. Megalopae were abundant in a broad band 37-148 km from shore, with peak abundance (2,871
10 m"^) 56 km offshore in June. In May, some late stage megalopae were collected in coastal inlets but
settlement appeared low in coastal study areas. Megalopal abundance decreased abruptly shoreward of
28 km from the coast.
Cancer oregonensis megalopae were also present from April to August, with their pattern of cross-
shelf abundance basically similar to that of C. magister. However, in contrast to C. magister, abundance
of late stage megalopae in coastal inlets was relatively high (313 10 m"^) in June, indicating that a signifi-
cant settlement of megalopae of this species could have occurred.
Evidence for cross-shelf movement of Cancer megalopae is discussed.
Dungeness crab, Cancer magister, range from the
Aleutian Islands to northern Mexico in the eastern
Pacific (Hart 1982) and are commercially exploited
from northern California to Kodiak Island, AK. As
part of an ongoing study of Dungeness crab recruit-
ment off the west coast of Vancouver Island, the
abundance and distribution of larvae off Tofino,
British Columbia, are being studied to determine
how variability in annual recruitment is affected by
larval settlement.
Dungeness crab larvae are planktonic and pass
through five zoeal stages and one megalopal stage
before settling to the sea bottom. Studies of larvae
prior to the 1970s primarily involved descriptions
of larval morphology (Mir 1961; Poole 1966), and it
was not until Reed (1969) developed laboratory
culture methods that larval environmental require-
ments were first described. Optimal ranges of tem-
perature and salinity for laboratory-cultured zoeae
were 10.0°-13.9°C and 25-30%o respectively, but
their survival was not significantly affected by the
temperature and salinity ranges occurring in the
waters off Oregon, where Reed's study was con-
ducted, at the time of year when larvae are com-
monly found. Lough (1976) suggested offshore lar-
val movement would allow larvae to avoid lower
nearshore salinities, and that normal oceanic salinity
levels probably favor survival over the long term.
The temporal occurrence of larvae in open coast
oceanic waters varies somewhat according to lati-
tude, with larvae present earliest in the season in
the southern part of the species' range. Seasonal
occurrence has not been well documented in the in-
shore waters of Georgia Strait, Puget Sound, and
Juan de Fuca Strait, but appears to differ signifi-
cantly from that in open coast waters. Larval settle-
ment, which typically occurs in May and June off
the outer coast of Washington, can occur as late as
mid-September in northern Puget Sound (D. Arm-
strong2).
Temporal and spatial distributions of crab larvae
have been documented for the years 1975-80 in the
Gulf of Farallones and the San Francisco-San Pablo-
Suisun Bay complex in central California (Reilly
'Department of Fisheries and Oceans, Fisheries Research
Branch, Pacific Biological Station, Nanaimo, B.C., Canada V9R
5K6.
^D. Armstrong, University of Washington, Seattle, WA 98195,
pers. commun. December 1986.
Manuscript accepted March 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
525
FISHERY BULLETIN: VOL. 86, NO. 3
1983) and for the years 1969-70 along a transect
off Newport, OR (Lough 1975, 1976). These studies
suggest that while larvae are hatched in the near-
shore, shallow-water environment preferred by
adult Dungeness crab, they subsequently move
alongshore and offshore, and then inshore, often
over considerable distances. Larvae hatched off
California and Oregon in late winter have been
hypothesized to move northward with the Davidson
Current before its reversal in March- April, and off-
shore as a result of upwelling, Ekman transport,
estuarine runoff, and geostrophic flow, depending
on location and season (Lough 1976; Wild and Tasto
1983). Later stage zoeae are typically found pro-
gressively further offshore, and it seems to be the
megalopal stage that returns inshore (Lough 1976;
Reilly 1983). The megalopal stage is the strongest
swimming stage (Jacoby 1982), but there is no direct
evidence to indicate that their inshore movement is
active.
There are five species of crabs of the genus Cancer
in British Columbia (Hart 1982), and larvae of C.
oregonensis also occur in abundance in offshore
waters along with C. magister (Lough 1975). Cancer
oregonensis has no commercial importance, but since
it occurs with C. magister, data on both species of
crabs are included in this study. Comparisons of
occurrence between the two species may provide
insight into possible environmental mechanisms or
processes that influence transport from offshore
areas to the inshore juvenile habitats, which the lar-
vae of both species must reach.
This study establishes the seasonal and spatial
occurrences of Cancer megalopae along a transect
over and beyond the continental shelf off the west
coast of Vancouver Island (Fig. 1). The geograph-
ical area is of particular interest because of the loca-
tion of a major regional crab fishery near Tofino
(Jamieson 1985; Noakes and Jamieson 1986) and the
resulting importance of understanding factors influ-
encing the magnitude of local larval crab settlement.
The oceanography off Vancouver Island has been
relatively well studied (Freeland et al. 1984; Thom-
son 1981), and because of the intrusion of Juan de
Fuca Strait waters into the general longshore ocean-
ographic regime and the relative increase in topo-
graphical complexity of the continental shelf, it is
considerably more complicated than that found
Figure 1.— The survey transect line, with stations, over which this study was conducted.
526
JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE
south of Cape Flattery (Hickey 1979). Along the
eastern Pacific coast, Cape Flattery marks a sig-
nificant change in nearshore oceanography, and
this study to investigate larval crab distribution in
open coast waters is the first one north of this
landmark.
GENERAL OCEANOGRAPHY OFF
VANCOUVER ISLAND
A seasonal surface current moving northward
from about lat. 32°N to 51 °N (northern Vancouver
Island) typically exists in nearshore waters from
October to March (Fig. 2). South of Cape Flattery,
this is called the Davidson Current (Hickey 1979)
whereas off British Columbia, it appears to consist
of two components. Nearshore, there is the year-
round, northward flowing Vancouver Island Coastal
Current, about 20-30 km wide, and further offshore,
there is the seasonal Shelf-Break Current, perhaps
a continuation of the Davidson Current, which
reverses direction in response to changes in the
large-scale wind field (Freeland et al. 1984). The
Davidson Current off Oregon has an average
northward flow of 50 cm/s for 30% of the time
(Boisvert 1969), sufficient to transport larvae sig-
nificant distances northward during their develop-
mental period if they remained continuously in the
current. The California Current is a seasonal, south-
ward moving, surface current of similar magnitude
and location to the Davidson Current. It occurs off
Vancouver Island (Hickey 1979) in the spring and
summer, along with the seasonal, southward flow-
ing, Shelf-Break Current on the outer continental
shelf (Freeland et al. 1984). These currents could
transport larvae located off Vancouver Island dur-
ing the spring and summer southwards (Fig. 2).
North of Vancouver Island, outer continental shelf
currents are poorly described, but off Cape St.
James (southern tip of the Queen Charlotte Islands),
the surface flow is strongly southward almost year-
round, turning northward only in March and April
(Freeland et al. 1984). The Vancouver Island Coastal
Current originates at the mouth of Juan de Fuca
Strait, and so could transport larvae out of the Puget
Sound-Georgia Strait complex.
MATERIALS AND METHODS
General Methodology
Sampling was largely done along an offshore
transect line perpendicular to the coast off Tofino,
British Columbia (Fig. 1), with stations located at
0, 9.3, 18.5, 27.8, 37, 46, 55.6, 74.1, 92.3, 111.2,
148.2, and 185.3 km from shore. Stations were more
closely spaced nearshore where larval abundance
was expected to be greatest, but the transect ex-
tended well beyond the shelf break to determine
what the seaward distribution of megalopae might
be.
Six offshore samplings were made along the tran-
sect from February through August 1985, but only
three extended the full 185 km. Dates of sampling
were 27 February-4 March, 17 and 18 April, 30 and
31 May, 14 and 15 June, 6 and 7 July, and 24 and
25 August. February-March, June, and July sam-
plings were terminated 74 km, 130 km, and 111 km
offshore, respectively, because of poor weather and
time constraints.
At each station, a neuston tow and an oblique
bongo tow to 250 m, or to within 5 m of bottom,
were made, and a temperature profile was recorded
by means of an expendable bathythermograph. On
all cruises, bongo tows were done during daylight
hours and neuston tows at night; late-stage crab
larvae congregate near the water surface at night
(Booth et al. 1985). On 17 June 1985, an hourly, noc-
turnal series of neuston tows was made at one loca-
tion, 9.3 km offshore, to identify patterns of night-
time abundance of megalopae in surface waters.
To supplement offshore sampling along the tran-
sect, neuston tows were made after dusk in the in-
lets around Tofino biweekly from early May through
August 1985 to monitor presence and movement of
larvae. In addition, a beam trawl and epibenthic sled
were used in inshore waters in July 1985 to sample
for newly settled larvae in an effort to establish time
and magnitude of larval settlement.
The RV G. B. Reed was used for all offshore
sampling except for the June sample, which was
collected from a 27 m charter vessel. Inshore sam-
pling was conducted from a 7 m aluminum herring
skiff except for mid-June, when the charter vessel
was again used.
Gear
The neuston sampler was a modified otter surface
sampler (Mason and Phillips 1986) with a square
mouth opening 45 cm on each side; under calm sea
conditions, it sampled the top 35 cm of the water
column. A General Oceanics flowmeter^ in the
mouth of the net was used to establish volume of
water filtered. Netting was black, 500 pt Nitex, and
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
527
FISHERY BULLETIN: VOL. 86, NO. 3
WINTER
54°
51'
48°
CONFUSED AND
^;i^;S^ VARIABLE
CURRENTS
54°
51°
48°
Figure 2.— Prevailing surface circulation off the British Columbia-Washington coast in winter and
summer. Broken arrows indicate uncertain currents. Numbers give speeds (cm s"^) (modified from
Thomson 1981).
528
JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE
was the same size as that on the bongo gear. The
neuston sampler was towed at 4 kn approximately
10 m from the side of the vessel to reduce hull
avoidance by the larvae. Tow duration was usually
15 minutes but was shortened when crab larvae
occasionally became very abundant at dawn and
dusk.
Bongo gear was a modified SCOR design (Mason
et al. 1984), having a mouth opening of 25 cm (each
side) with the outboard (left) net of 500 ^a Nitex and
the inboard (right) net of 230 ^ Nitex. A General
Oceanics flowmeter was mounted in the mouth of
each net. Sampling procedure followed that de-
scribed by Smith and Richardson (1977*). At stations
<100 m water depth, the bongo was fished in an un-
dulating fashion from the bottom to the surface in
order to filter a standard volume of water (about
300 m^).
The epibenthic sled (Phillips and Mason 1986) used
in inshore waters had a 60 cm mouth opening and
1 mm Nitex netting; it was towed for 10 minutes
at 2-3 kn. Nine tows were made over bottoms of
unknown characteristics. The modified beam trawl
(Gunderson and Ellis 1986), also used in inshore
waters, had a 3 m mouth opening, a 7 mm mesh net,
and a 3 mm mesh cod end; it was towed for 10
minutes at about the same speed as the sled. For
both gear types, distance towed was calculated by
radar triangulation to reference points on land.
Data Analysis
All plankton samples were preserved initially in
a 4% formaldehyde solution of saltwater. In the
laboratory, settled volume was determined, and
general composition of the plankton noted. Samples
were then sieved, and Cancer larvae picked by hand
and returned to a 2% formaldehyde solution before
their identification. Cancer magister megalops are
readily identifiable by their larger size (Trask 1970;
Lough 1975), but the currently used key (Lough
1975) was not always effective in separating C. ore-
gonensis and C. jyroductus. Presence of lateral spines
is a subjective criterion, and morphological dimen-
sions and counts of setae were two variable to
distinguish species. These smaller megalops were
finally concluded to be C. oregonensis after rearing
hundreds of larvae to the juvenile stage and finding
no C. productus.
With beam trawl samples, the catch was sorted
*Smith, P. R., and S. L. Richardson. 1977. Manual of methods
for fisheries resource survey and appraisal. Part 4. Standard tech-
niques for pelagic fish egg and larval surveys. Southwest Fish.
Cent., Natl. Mar. Fish. Serv., NOAA, Adm. Rep. No. 77-11.
on deck to remove juvenile crabs, and the megalopae
were preserved as above. Species composition of the
catch was noted.
For the bongo tows, the volume of water filtered
was used to calculate a haul factor (Mason et al.
1984), which accounted for tow depth and allowed
expression of the data as the integrated number of
organisms beneath 10 m^ of sea surface.
For neuston tows, relative abundance was ex-
pressed as area swept rather than as volume filtered
and again expressed as number of animals under 10
m^ of sea surface.
Calculations of larval abundance are necessarily
conservative, and direct comparisons between dif-
ferent gear types are not presently possible, given
our current understanding of gear efficiency and lar-
val catchability, which varies with both depth and
time of day. Numbers of larvae reported here are
thus directly comparable only within each gear type
used. For bottom gear, numbers of crabs were cal-
culated with no consideration of gear efficiency.
Molt Staging
State of development within the megalopal stage
was determined for all Cancer megalopae collected
at a station, or 25 randomly selected individuals of
each species if the number collected exceeded 25.
The sequence of epidermal changes occurring dur-
ing this intermolt period has been described for
Dungeness crab by Hatfield (1983), and her proce-
dures and staging criteria were applied for both
species. Whole megalopae were stained with 0.25%
Fast Green dye in either water or polyvinyl lacto-
phenol for 2-18 hours, and then the second maxil-
lipeds were removed and mounted. Megalopae were
identified to 1 of the 13 intermolt stages recognized
by Hatfield (1983); these were then combined to
form 3 groups: early (stages 1-4), middle (stages
5-8), and late (stages 9-13) megalopae. Temporal
durations of these three groups in laboratory studies
were 5.8, 15.4, and 6.3 days, respectively (Hatfield
1983). Corresponding durations for C. oregonensis
are unknown. Stages were grouped into the three
categories described for simplification of analysis;
many of Hatfield's stages were of <48-h duration,
and this was considered too fine a resolution for our
purposes.
RESULTS
Water Temperature
Water temperature at the surface ranged from
529
FISHERY BULLETIN: VOL. 86, NO. 3
8.4° to 14.4°C, and from 7.6° to 8.8°C at 50 m, dur-
ing the period 17 April to 7 July (Table 1). Warmest
surface temperatures were in May and July, and at
50 m, were warmest in June.
Species Occurrence
Three species of crab larvae predominated in the
plankton collected. The porcellanid crab, Petrolisthes
cinctipes, predominated in waters within 8 km of
shore, while C. oregonensis and C. magister domi-
nated in more offshore waters. Other crab species
were present but at much lower abundance than
these three species. Megalopae of C. oregonensis
were of comparable abundance to C. magister at
most stations sampled.
Surface Abundance
Presence of crab megalopae at the surface was
dependent on time of day, with slight differences
in timing of maximal abundance occurring between
the two Cancer species (Fig. 3). At 9.3 km offshore
in June, C. oregonensis was several orders of mag-
nitude more abundant and also seemed to peak in
abundance a little earlier in the evening and later
in the morning than did C. magister. Relative abun-
dance of both species declined in the middle of the
night.
Data in Figure 3 was smoothed by eye (Fig. 4)
to allow megalopal abundance data from the
transect to be weighted by time of capture at
night. Multiplier values were determined, and these
were used to adjust actual abundance data by
sampling time for the May and June sampling
periods (Tables 2, 3). Only these data were weighted,
and since day length changes with time of year, ad-
justment of data collected more than two weeks on
either side of 14 and 15 June was considered
inappropriate.
In hindsight, some data (indicated by asterisks in
Tables 2, 3) was found to have been collected at a
time when megalopae were just beginning to reach,
or had just left, the surface. The calibration slope
(Fig. 4) is very steep at both these times and it was
impossible to estimate accurately a meaningful scal-
ing multiplier for these data. Our estimated mega-
lopal abundance at these stations should be viewed
with particular caution; a zero value may not in-
dicate that megalopae were absent, but only that
they were not at the surface.
Bongo tows during the day poorly sampled the
abundance of C. magister megalopae (Table 4)
shown to be present by nighttime neuston tows
(Tables 2, 3). This was possibly due to sampling the
total column below only a few square meters of sea
surface, because of integration of the results, or to
avoiding the net by the megalopae. It was unlikely
because of failure to sample deeply enough, since
on the continental shelf, tow depth approached bot-
tom depth. Megalopae (maximum per tow was 4)
were primarily collected by bongo gear at stations
where they were abundant in neuston samples at
night, but relative abundance estimates obtained
from bongo nets are considered only useful in a
general qualitative sense because of the low absolute
numbers caught.
Temporal Occurrence
(Time of Year)
No Cancer megalopae were found in samples col-
lected shortly after dusk on 27 February-4 March.
Cancer megalopae were first observed on the next
Table 1 .-
-Watei
' temperatures (°C) by depth (m) at the stations sampled off Tofino, B.C. on
17 April, 31 May, 14 June, and 6 and 7 July 1985.
Depth (m)
Distance
offshore
(km)
April May June July
0
20 50 0 20 50 0 20 50 0 20 50
0
9
19
28
37
56
74
93
111
130
148
185
8.8
8.8
8.7
8.8
8.6
8.7
8.7
8.5
8.5
8.5
8.4
8.8
8.7
8.1
8.8
8.6
8.7
8.6
8.5
8.4
8.5
8.6
7.9
8.2
8.3
8.2
8.2
8.2
8.0
8.0
8.4
12.2
13.4
13.4
13.1
13.8
13.3
11.8
11.9
11.5
11.0
11.4
9.2
9.5
8.6
9.6
9.3
9.8
10.0
10.3
10.5
10.0
10.0
7.6
7.9
7.9
8.2
8.2
8.2
8.1
8.7
8.4
12.2
12.2
12.3
12.8
12.7
12.9
12.9
12.8
12.7
12.7
11.4
9.0
10.4
12.3
9.4
10.5
12.0
12.2
12.2
12.4
8.0
8.0
8.3
8.1
8.7
8.5
8.5
8.6
13.5
14.0
13.3
13.3
14.4
14.4
11.0
9.0
10.0
10.7
13.3
13.3
7.6
7.7
7.7
8.6
8.6
530
JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE
gsijung-
jq6i|iM J I lAio - —
)L|6l|IM| |D3||nDU
in
O
to
g ID
<1> 00
00
mB|||Mj |DO!jnDu -
jgBjiiMj |iA|0 -
jasuns-
p— I — I — r
O
O
o
[III — I — I r
O
O
-r-[ — I — I — r
B
o
O
o
•q
(O
S
o
to
()
a
n
-4->
rf
JS
O
*
bo
<
C
Q
c
o
o
o
^
(M
bb
o
o
^
^
C\J
"3.
V
rt
V)
M
o
-)
5?
o
n
c
OJ
u,
f\j
bserved du
.C.
^^
O pQ
H
2 -
5 o
to
9- c
Q
° iG
CL
0! O
8SI Jung —
mbi |iMt I IMO
-mbl|IM| |D3UnDU
(q6i|iM( I Dounou
|L|&||IM4 |IAI3
i8suns-
I I I
o
T — I — r
o
o
o
o
to
z
i §
o
(-
o ^
( )
z
<
. 33
o
O <=>
o
s^ «^
o
»5
o
CM
O)
o
b
o
o
CVI
^
o
CO
a>
o
3
c
o
Q
c«
(M
-13
CJ
C
1
o
o
o
'S
o
o
CJ
CO
OS
D
o
(aoD^jns oas j.ojoI-N) 33NVaNnaV
531
Figure 4.— Smoothed abundance of Cancer
magister and C. oregonensis megalopae ob-
served during sampling through one night at
a station 9 km off the coast off Tofino, B.C.
1000
o
o
Z!
c/)
O
0)
'E
O
LlI
O
z
<
Q
3
QD
<
^1
Q
UJ
I
t-
O
o
CO
FISHERY BULLETIN: VOL. 86, NO. 3
10
JUNE 1985
o
in
O
10
100
10-
'E
O
UJ
u
<
Q
CD
<
6>
51
^•|
Q
LU
I
f-
o
o
CO
1800
2000
2200
2400
0200
0400
0600
TIME (POST)
Table 2.— Scaling values used to standardize the actual number of megalopae caught
(unsealed) to the estimated number which would have been caught at the times of peak
nocturnal abundance, 22:40 and 23:40 hours, for Cancer oregonensis and C. magister. respec-
tively, on 30 and 31 May 1985. * = Value uncertain because of time of sampling. Values
rounded to one decimal place.
c.
magister
C. oregonensis
Distance
(km)
(A/
10 m-2)
(N 10 m-2)
Time
Multiplier
Unsealed
Scaled
Multiplier Unsealed
Scaled
18:12
0
oo
2.3
?*
oo
0
0*
19:02
9
36.7
0.2
7.2*
CO
4
?'
19:43
19
26.8
0.0
1.1*
00
0
0*
20:26
28
20.0
0
0
oo
0
0*
21:09
37
11.0
0.4
4.7
oo
0
0*
21:52
46
3.9
14.4
56.4
5.2
2.2
11.5
22:35
56
1.6
29.9
47.0
1.0
0.1
0.1
23:37
74
1.0
17.0
17.0
—
0
0
00:47
93
3.3
36.0
119.9
24.8
1.0
25.7
01:55
111
5.0
2.6
13.3
40.0
1.8
72.1
03:56
148
110.0
1.1
118.4*
2.5
0.1
0.2
05:57
185
OO
0
0*
oo
0
0*
532
JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE
sampling date, 16 and 17 April, but maximum abun-
dance (unweighted as to diel time of sampling)
occurred in late May and the second week of June
(Fig. 5). Cancer magister megalopae were collected
as late as 24 August, the last sampling date, but
relative abundance along the transect was low after
the beginning of July.
Cancer oregonensis megalopae were caught in
April but occurred in abundance only in the late May
and mid- June surveys, with peak abundance in June.
1000
100 120 140 160 180 200 220 240 260
CALENDAR DAYS
Figure 5.— Neustonic abundance of megalopae observed along the transect line during April to August 1985 surveys.
Table 3.— Scaling values used to standardize the actual number of megalopae caught
(unsealed) to the estimated number which would have been caught at the times of peak
nocturnal abundance, 22:40 and 23:40 hours, for Cancer oregonensis and C. magister, respec-
tively, on 14 and 15 June 1985. * = Value uncertain because of time of sampling. Values
rounded to one decimal place.
C.
magister
C. oregonensis
Distance
{N
10m-2)
(N 10 m"^)
Time
(km)
Multiplier
Unsealed
Sealed
Multiplier Unsealed
Sealed
01:42
0
0
0
37.1
8.8
327
00:58
9
3.6
0.1
0.4
27.4
170.5
4,671
00:09
19
1.2
0.1
0.1
16.3
0.1
2.3
23:19
28
1.0
0.2
0.2
2.7
7.6
20.5
05:36
37
OO
0.1
?*
OO
0
0*
03:52
56
55.0
52.2
2,871*
2.6
15.7
40.9
02:02
74
5.2
3.7
19.2
40.0
4.8
192
00:25
93
2.4
8.9
21.6
20.8
1.5
30.4
22:45
111
1.2
8.4
10.3
1.0
12.5
12.5
21:10
130
—
0
0
—
0
0
533
FISHERY BULLETIN: VOL. 86, NO. 3
Table 4.— Tow characteristics and number of Cancer magister megalopae (10 m^^ of sea surface) caught by 0.25 m
bongo tows off Tofino, B.C., on 31 May and 14 June 1985.
May
June
Distance
Water
Max.
Water
Max.
offshore
Time
volume
depth
No.
Megalopae
Time
volume
depth
No.
Megalopae
(km)
(h)
(m3)
(m)
33
megalops
(A/
10 m-'')
(h)
K)
(m)
megalops
(A/
10 m"^)
0
19:27
210
1
1.6
08:00
240
36
3
4.5
9
18:52
182
55
1
3.0
08:55
220
55
0
0
19
18:16
114
72
0
0
09:51
222
75
0
0
28
17:37
166
120
0
0
10:50
209
115
0
0
37
16:55
199
151
0
0
11:47
284
146
1
5.1
56
15:30
352
250
4
28.4
13:32
381
192
0
0
74
14:07
327
250
1
7.6
15:21
338
200
1
5.9
93
12:42
326
250
3
23.0
17:04
296
200
2
13.5
111
11:17
360
250
1
6.9
18:56
269
200
0
0
130
21:12
327
200
2
12.2
148
08:55
400
250
0
0
—
185
06:18
319
250
0
0
—
Cross-Shelf Spatial Distribution
With both species, no clear pattern in cross-shelf
distribution was evident from the April data, per-
haps because of relatively low overall megalopal
abundance. In late May (Table 2), scaled abundance
(weighted as to time of night) of C. magister was
highest at 93 km offshore, but megalopae were
generally abundant (>10 10 m"^ sea surface) from
46 to 148 km offshore. In mid-June, the basic pat-
tern observed in late May was still evident, although
megalopal abundance near shore had declined (Table
3). Highest estimated C. magister abundance was
at 56 km offshore.
With C. oregonensis, scaled abundance in late May
peaked (>20 10 m' 2) 93-111 km offshore (Table 2).
An exceptionally large number of megalopae (4,671
10 m"^) was found 9 km offshore in June (Table 3),
while from 56 to 111 km offshore, abundance re-
mained high (>20 10 m- 2).
Sampling of inlet waters around Tofino with
neuston gear showed that a few Cancer megalopae
were present but that no substantial (average was
<1 10 m~2) numbers occurred. Maximum C. ma-
gister megalopal abundance was 2.9 megalopae 10
m"^ on 5 June, with megalopae observed only be-
tween 29 May and 19 June. Maximum C. oregonen-
sis abundance was 2.3 megalopae 10 m"^ on June
19, with megalopae observed only between 4 and
19 June.
Intermolt Stage
All C. magister megalopae collected in April were
of early developmental stage (Fig. 6), whereas all
those collected in August were mid-stage mega-
lopae. In May, June, and July, the general pattern
of offshore distribution by molt stage was for late
stage larvae to be nearest inshore and early stage
larvae to be furthest offshore.
Cancer oregonensis megalopae had a similar
developmental distribution pattern to those of C
magister (Fig. 7); late molt megalopae were most
abundant closest to shore. However, in late May,
most megalopae were early stage, in contrast to C.
magister, whereas by mid-June, all three intermolt
groups were present.
When the scaled megalopal abundance at distance
offshore (Tables 2, 3) is multiplied by the percent-
age at each molt stage at a specific location (Figures
6, 7, for each species, respectively), the actual abun-
dance by molt group with distance offshore is deter-
mined (Figures 8, 9, respectively). In late May, late
stage C. magister megalopae were relatively abun-
dant 9 km offshore, indicating that some Dungeness
crab settlement may have occurred. However, lar-
val sampling in the inlets and bays showed few
megalopae present. In contrast to previous years,
no recently settled juvenile crabs (0 age-class crabs)
were found by local fishermen in intertidal areas or
on floating objects, confirming that the magnitude
of megalopal settlement in 1985 was relatively small.
In mid- June, there were few C. magister megalopae
of any intermolt stage present within 40 km of the
coast and little evidence of late stage megalopae off-
shore. Early and mid-stage megalopae were most
abundant from 56 to HI km offshore.
For C. oregonensis, the opposite was observed
(Fig. 9). Few late stage megalopae were present in
late May, and unlike C. magister, megalopae were
not concentrated in nearshore waters. However,
megalopae in all three intermolt stages were abun-
dant inshore in mid-June, with late stage megalopae
dominating at the coast. The timing of occurrence
534
JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE
Figure 6.— Percentage of the three Cancer
magister megalopal intermolt groups observed
at each station during the April to August 1985
surveys.
LlI
o
UJ
CL
LU
O
CC
LlI
CL
(a) APRIL 16/17, 1985
2 33 2
(-
z
UJ
o
tr
LU
CL
(b) may 30/31
100
80-
60-
40
20
0
3-
^25
25
231
I
)!^
1
'1 f
X
'■'i
25
25
25
o— o..
(c) JUNE 16
®
5 4 25 24
17 25 25
(e) august 23/26
100 -,
80
60
40
20
0
— t \ 1 1
0 40 80 120 140
DISTANCE (Km)
JULY 6/7
63 3 2
0 40 80 120
DISTANCE (Km)
C. magjster
INTERMOULT STAGES
o o
■ a
1-4
5-8
9-13
n= 10 unless otherwise
specified above graph
of late stage megalopae at the coast thus differed
between the two species.
Inshore Benthic Sampling
Survey of commercial crab habitat with beam
trawls and an epibenthic sled yielded few newly
settled megalopae or juvenile C magister (Table 5).
A maximum of 83 crab/10,000 m^ was collected,
well below densities which might be expected if
settlement had been substantial. No C. oregonensis
were caught, perhaps because of the habitat being
sampled, and no larval settlement of either species
was observed when intertidal mud flat areas were
searched by foot. In a concurrent study, no signifi-
cant abundance of 0 age-class C. magister was
observed during monthly sampling throughout the
remainder of the year.
DISCUSSION
Megalopal Source
Temporal duration of C. magister larval stages has
been estimated both by laboratory rearing (Poole
1966; Reed 1969; Brugman 1972; Gaumer 1973;
Ebert et al. 1983) and the first appearances of lar-
val stages in field studies (Poole 1966; Lough 1976;
Reilly 1983). The length of the total larval period
has been estimated as 105-125 days by Reilly (1983),
535
FISHERY BULLETIN: VOL. 86, NO. 3
(a) may 30,31
LlI
(J
cr
UJ
Q-
100-1
80-
60
40
20
0
3 C, oregonensis
INTERMOULT STAGES
• . 1-4
o <, 5-8
n = 25 unless otherwise
specified above
graph
^■,..,r,.£.,,---,-°— °-
1 — f r»— ,
0 20 40 60 80 100 120 140 160
(b) JUNE 13-17
II
C) JULY 6,7
15 15
0 20 40 60 80 100 120
100-1
80-
60
40
20
0
0 20 40 60 80 100
DISTANCE OFFSHORE (km)
Figure 7.— Percentage of the three Cancer oregonensis megalopal intermolt groups
observed at each station during the May to July 1985 surveys.
Table 5. — Beam trawl and epibenthic sled catches from nearshore waters around Tofino, 16-18 June
1985.
Area
Depth
(m)
Gear
type
Catch of crabs
No. 10" m-2
Area
swept
(m ) Juveniles Megalops Juveniles Megalops
Templar Channel
7
trawl
2,650
sled
375
Chesterman Beach
5
trawl
—
sled
300
Cox Bay
8
trawl
2,500
7
sled
400
13
trawl
2,600
15
sled
420
30
trawl
2,260
23
sled
360
Lennard Island
25
trawl
sled
350
Long Beach
7
trawl
2,700
sled
400
13
trawl
2,650
sled
400
23
trawl
sled
220
0
0
0
0
0
0
0
0
3
0
0
0
1
1
26
0
0
4
75
27
48
0
0
0
18
50
23
50
0
0
0
0
0
0
0
0
83
0
0
0
4
25
45
536
JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE
5?
o
CO
CD
\a
00
<3>
i-H
8
s
^
•-s
o
1
bo
c
'C
C
o
JS
o
a)
-a
>
O
bo
c
6
o
e
3
o
■kJ .5
■— +J
^g
0) c
C3 3
o a>
^■^
bo "S
I":
§>
I
o
V
o
c
-a
c
3
-3
<a
o
CO
CX3
D
O
(aoDjjns D9S g.LUQ
33NvaNnav aanvDS
537
FISHERY BULLETIN: VOL. 86, NO. 3
<
. <
Q UJ
§5
tjl
in
m
o
oo
o
to
O
o
S
0)
o
o
bX)
C
T3
o
CD
c
o
(33
o
^ — .
o
E
01
a.
o
JC
H
T
■ —
c«
S
LU
-o
CC
0)
>
o
o
(\i
X
1-.
Si
CU
6
U)
X!
oo
>
<
o
o
(VI
O
in
O
o
— r~
O
in
o
CO
o
O
O
O
- O
O
CD
O
ID
O
o
CO
Q
T^O
§■ °
o <u
t/1
Ci3
-^ o
o aj
s -^
.s t;
>> s
-^ c
(U 3
ft 2
O ^
bd >
P II
Cn
o
O
<a
c
CIS
c
=1
-a
-2
Ci3
I
D
(9DD)jns D3S 2-^OIN) BONVQNnav 031^735
538
JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE
130 days (range: 89-143) by Lough (1976), 45 and
108 days at 17.8° and 10.0°C, respectively, by Reed
(1969), and 128-158 days under natural conditions
by Poole (1966). In laboratory studies, Poole (1966)
found the total time required was 111 days at
10.5°C. Since water temperatures typically range
from about 8° to 13°C off the British Columbia-
California coast during January-February (Thom-
son 1981), the total larval period in the study area
is assumed to be about 110 days.
Off Vancouver Island, megalopae can be collected
in abundance in open coast waters from mid- April
to August. Ovigerous Dungeness crab, collected at
Tofino and held in ambient temperature seawater
at Nanaimo, hatched their eggs from January to
March, with most hatching occurring in February
(G. Jamieson, unpub. data). This suggests that mega-
lopae collected between April and June could be
from local populations whereas most megalopae
collected in July and August may have largely
originated elsewhere.
Lough (1975) stated that the hatching period of
C. oregonensis off Oregon extends from January to
July, with two broods indicated, one primarily hatch-
ing in February and the other in May-June. He
estimated larval duration of a brood at 155 days
(range: 123-203 days) from field sampling, but there
was possible intermixing of the larvae from differ-
ent broods, as well as population variability in the
timing of hatching. We have no data on the hatching
period of this species in British Columbia.
Given the currents off the west coast of Vancouver
Island, then, Cancer larvae present there in the
spring could theoretically have originated anywhere
between northern California to perhaps southern
Alaska. It seems very unlikely that the larvae are
entirely the progeny of adult crabs on the west coast
of Vancouver Island.
Onshore Movement
Geographical location of larval settlement is very
dependent on currents. Adult Cancer of both species
are largely found in nearshore, shallow-water
habitats (Hart 1982), and such environments are
apparently both the origin and preferred destina-
tion of larval crabs (Butler 1956). Along the open
coast, Dungeness crab larvae are known to settle
in both estuarine and nearshore areas (Wild and
Tasto 1983; Stevens and Armstrong 1984; Arm-
strong and Gunderson 1985). In waters largely sur-
rounded by land, such as Puget Sound, Georgia and
Queen Charlotte Straits, and, to a lesser extent,
Hecate Strait, Dixon Entrance, and southeastern
Alaska, larval crabs may perhaps remain nearshore
throughout their entire developmental period. How-
ever, studies suggest that while larvae may be
hatched nearshore along the outer coast, they subse-
quently move offshore and then inshore (Lough
1976; Reilly 1983). The extent to which this may oc-
cur can profoundly affect the degree of dispersal of
a local region's progeny.
Evidence for an offshore-onshore movement of C.
magister during the larval development period is
ambiguous. In the field, three studies of the offshore
spatial pattern of larval distribution have been
undertaken: off central and northern California
(Reilly 1983), off Newport, Oregon (Lough 1976),
and off Tofino, British Columbia (this study). The
California study extended to about 185 km from
shore from San Francisco north to Cape Mendocino,
with some eariier CALCOFI data (1949-75) extend-
ing to about 275 km offshore included. Sampling
consisted of discrete-depth and oblique plankton
tows and was mostly during the day. The sampling
gear (0.5 m diameter opening) most often used was
preceded by the towing cable and bridle (P. Reilly^),
and gear avoidence by same megalopae may have
occurred. Because of the gear and protocol used, it
is difficult to interpret Reilly's (1983) results in a
quantitative sense. However, offshore movement of
larvae during zoeal stages II-V was indicated, as
was the later presence of megalopae in nearshore
waters following a period when stage V zoeae were
generally absent from within 40 km of shore, but
a mechanism to explain the onshore transport of
megalopae was not established. As part of the over-
all study (Wild and Tasto 1983), Hatfield (1983)
determined the intermolt stage of many larvae col-
lected by Reilly (1983). Earlier stage megalopae
were in general collected further offshore, earlier
in the year, and at a lower latitude.
The Oregon study (Lough 1976) was on one track-
line extending 110 km offshore off Newport, with
mostly daytime, oblique bongo samples collected at
specified stations. Although early stage zoeae of
both species were abundant nearshore, late stage
zoeae were not, and were largely collected at the
offshore stations. Lough (1975) noted that C. ore-
gonensis megalopae were found further offshore and
later during the summer upwelling season than were
C. magister megalopae. In 1970, large numbers of
C. magister megalopae, which were not intermolt
staged, were found inshore. In 1971, virtually no C.
magister larvae older than stage III were collected
^P. Reilly, California Department of Fish and Game, Menio Park,
CA 94025, pers. commun. February 1988.
539
FISHERY BULLETIN: VOL. 86, NO. 3
at any of the stations, and Lough suggested that a
mass mortahty of C. magister larvae might have oc-
curred. In contrast, C. oregonensis larvae were in
greater abundance during 1971 than 1970.
Results from the current study and Booth et al.
(1985) suggest that the sampling gear and proce-
dures used by Reilly (1983) and Lough (1976) gave
poor abundance estimates of megalopae. The num-
ber of megalopae caught during daylight is relatively
low, and the presence or absence of only a few
megalopae can greatly influence the estimated
megalopal density per unit of sea surface. This sen-
sitivity to bias can only be compensated for by many
replicate samplings, and it is logistically easier to
estimate absolute megalopal abundance by sampling
at night when megalopae are concentrated in a
relatively easily sampled, narrow depth range,
namely the neuston (Booth et al. 1985). Reilly (1983)
sampled once both day and night at 12 stations, and
noted no significant differences in megalopal den-
sity between day and night at the surface and in
oblique tows. However, his surface tows were not
neuston tows, and consisted of a 0.5 m diameter net
towed about 0.25-0.5 m below the water surface,
or below the wave troughs if the surface was rough
(P. Reilly fn. 5). This qualifies his observations, and
may explain the suggested low abundance of mega-
lopae in his study area (maximum density was 43
100 m"^ at the surface).
On occasion, megalopae have been observed to be
in association with Velella velella at the surface dur-
ing daylight (Wickham 1979; Reilly 1983; G. Jamie-
son unpub. data), but we have generally observed
relatively few megalopae in surface waters during
the day in comparison to the number observed
present at night. Accepting that relative megalopal
abundance is best determined at night from neuston
tows, megalopae off British Columbia were shown
to be abundant in specific areas offshore and at these
locations, early and mid-stage megalopae predomi-
nated.
All three studies indicate that 1) later stage C.
magister larvae are found offshore and, since they
are hatched at inshore locations, some mechanism
must be transporting them away from shore, 2)
megalopae are found inshore, with Hatfield (1983)
and this study showing that earlier stage megalopae
predominate furthest offshore, and 3) while long-
shore drift of larvae may occur, it is difficult to
establish its significance in the absence of simul-
taneous current and larval distribution data, both
geographically and vertically in the water column.
None of these studies demonstrates conclusively
that larvae which ultimately settle inshore were ever
very far offshore, and while it is shown that larvae
can be carried offshore in abundance, this may well
represent larval wastage, in that relatively few of
them, if any, may ultimately return inshore to sur-
vive at settlement. The same is assumed to apply
for C. oregonensis.
The California data on larval distribution, which
is the most extensive spatially, are somewhat am-
bigious in that they may be interpreted either as
onshore-offshore-onshore movement, or as a general
offshore dispersal coupled with northwards-south-
wards drift of those larvae remaining inshore. To
account for an absence of middle stage larvae in the
nearshore waters of Reilly's (1983) study area, the
latter scenario assumes that larval crab production
south of the Gulf of the Farallones is relatively low,
and that the water mass entering the sample area
from the south as part of the Davidson Current con-
tains few larvae. Little data appear to exist to refute
this possibility; the commercial fishery for Dunge-
ness crab only exists at a few locations south of San
Francisco, and then only sporadically (Warner
1985). Our data and Lough's (1976) data are inade-
quate to investigate longshore drift because bio-
logical samples in each study were only collected
along one offshore transect and no data on currents
was simultaneously obtained.
Off the west coast of Vancouver Island, we found
that megalopae may be concentrated in areas
between surface currents flowing in opposite direc-
tions, specifically in the shear between the Van-
couver Island Coastal Current and the outer Shelf-
Break Current. Megalopae were abundant 40-70 km
from shore, with C. magister megalopae particularly
concentrated about 50 km from shore. In 1985, there
was no major crab settlement observed on the outer
coast near Tofino, suggesting that the Coastal Cur-
rent may sometimes be an effective barrier to on-
shore movement and successful settlement. Under-
standing the horizontal and vertical distributions of
crab larvae in the water column now seems essen-
tial if transport mechanisms of larvae and their
movement from hatching to settlement are to be
understood.
ACKNOWLEDGMENTS
We wish to thank the officers and crew of the
Canadian Research Vessel G. B. Reed for their en-
thusiastic assistance in collecting data; D. Hartley,
skipper of the charter vessel Beatrice, for his skill
and persistence in meeting the objectives of the
charter; and the students and term employees who
participated in collecting and analyzing data. Glen
540
JAMIESON and PHILLIPS: OCCURRENCE OF CANCER CRAB MEGALOPAE
Brown and Dorothy Young, in particular, contrib-
uted substantially.
LITERATURE 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 Pro-
ceedings of the Symposium on Dungeness Crab Biology and
Management, p. 145-170. Alaska Sea Grant Rep. No. 85-3.
BOISVERT, W. E.
1 969. Major currents off the west coasts of North and South
America. Nav. Oceanogr. Off. Tech. Rep. 221, 34 p.
Booth, J., A. Phillips, and G. S. Jamieson.
1985. Fine scale spatial distribution of Cancer magister
megalopae and its relevance to sampling methodology. In
Proceedings of the Symposium on Dungeness Crab Biology
and Management, p. 273-286. Alaska Sea Grant Rep. No.
85-3.
Brugman, a.
1972. The effects of temperature on the growth of Dungeness
crab. Cancer magister Dana. MS Thesis, Humbolt State
Univ., Areata, CA, 34 p.
Butler, T. H.
1956. The distribution and abundance of early postlarval
stages of the British Columbia commercial crab. Fish. Res.
Board Can., Prog. Rep. 107, p. 22-23.
Ebert, E. E., a. W. Haseltine, J. L. Houk, and R. 0. Kelly.
1983. Laboratory cultivation of the Dungeness crab. Cancer
magister. In P. W. Wild and R. N. Tasto (editors), Life
history, environment, and mariculture studies of the Dunge-
ness crab, Cancer magister, with emphasis on the central
California fishery resource, p. 259-309. Calif. Dep. Fish
Game Fish Bull. 172.
Freeland, H. J., W. R. Crawford, and R. E. Thomson.
1984. Currents along the Pacific Coast of Canada. Atmos.-
Ocean 22, p. 151-172.
Gaumer, T. F.
1973. Controlled rearing of Dungeness crab larvae and the
influence of enwonmental conditions to their survival.
Oreg. Fish Comm., Portland, NOAA 73082801, NMFS 1 5
R, 53 p.
Gunderson, D. R., and I. E. Ellis.
1986. Development of a plumb staff beam trawl for sampling
demersal fauna. Fish. Res. 4:35-41.
Hart, J. F. L.
1982. Crabs and their relatives of British Columbia. Br.
Columbia Prov. Mus. Handb. 40, 267 p.
Hatfield, S. E.
1983. Intermoult staging and distribution of Dungeness crab.
Cancer magister, megalopae. In P. W. Wild and R. N. Tasto
(editors). Life history, environment, and aquaculture studies
of the Dungeness crab, Cancer magister, with emphasis on
the central California fishery resource, p. 85-96. Calif. Dep.
Fish Game Fish Bull. 172.
HiCKEY, B. M.
1979. The California Current system - hypotheses and facts.
Prog. Oceanogr. 8:191-279.
Jacoby, C. a.
1982. Behavioral responses of the larvae of Cancer magister
Dana (1852) to light, pressure and gravity. Mar. Behav.
Physiol. 8:267-283.
Jamieson, G. S.
1985. The Dungeness crab (Cancer magister) fisheries of
British Columbia. In Proceedings of the Symposium on
Dungeness Crab Biology and Management, p. 37-60.
Alaska Sea Grant Rep. No. 85-3.
Lough, R. G.
1975. Dynamics of crab larvae (Anomura, Brachyura) off the
central Oregon coast, 1969-1971. Ph.D. Thesis, Oregon
State Univ., Corvallis, 299 p.
1976. Larval dynamics of the Dungeness crab, Cancer
magister, off the central Oregon coast, 1970-71. Fish.
Bull., U.S. 74:353-375.
Mason, J. C, and A. C. Phillips.
1986. An improved otter surface sampler. Fish. Bull., U.S.
84:480-484.
Mason, J. C, A. C. Phillips, and 0. D. Kennedy.
1984. Estimating the spawning stocks of Pacific hake
(Merluccius prodiLctus) and walleye pollock (Theragra chal-
cogramma) in the Strait of Georgia, B.C. from their released
egg production. Can. Tech. Rep. Fish. Aquat. Sci. 1289,
51 p.
Mir, R. D.
1961. The external morphology of the first zoeal stages of
the crabs. Cancer magister Dana, Cancer antennarius
Stimpson and Cancer anthonyi Rathbum. Calif. Fish Game
47:103-111.
NOAKES, D., AND G. S. JaMIESON.
1986. Preliminary analysis of British Columbia commercial
landing statistics for 1979 to 1984 inclusive: a multispecies
approach. Can. MS Rep. Fish. Aquat. Sci. 1882, 167 p.
Phillips, A. C, and J. C. Mason.
1986. A towed, self-adjusting sled sampler for demersal fish
eggs and larvae. Fish. Res. 4:235-242.
Poole, R. L.
1966. A description of laboratory-reared zoea of Cancer
magister Dana and megalopae taken under natural condi-
tions (Decapoda Brachyura). Crustaceana 11:83-97.
Reed, P. H.
1969. Cultiu-e methods and effects of temperature and salin-
ity on survival and growth of Dungeness crab (Canxer
magister) larvae in the laboratory. J. Fish. Res. Board Can.
26:389-397.
Reilly, p. N.
1983. Dynamics of Dungeness crab. Cancer magister, larvae
off central and northern California. In P. W. Wild and R. N.
Tasto (editors). Life history, environment, and mariculture
studies of the Dungeness crab. Cancer magister, with em-
phasis on the central California fishery resource, p. 57-84.
Calif. Dep. Fish Game Fish Bull. 172.
Stevens, B. G., and D. A. Armstrong.
1984. Distribution, abundance and growth of juvenile Dunge-
ness crabs. Cancer magister, in Grays Harbor estuary,
Washington. Fish. Bull., U.S. 82:469-483.
Thomson, R. E.
1981. Oceanography of the British Columbia coast. Can.
Spec. Publ. Fish. Aquat. Sci. 56, 291 p.
Trask, T.
1970. A description of laboratory-reared larvae of Cancer pro
diictus Randall (Decapoda, Brackyura) and a comparison to
larvae of Cancer magister Dana. Crustaceana 18:133-146.
Warner, R. W.
1985. Overview of the California Dungeness crab. Cancer
magister, fisheries. In Proceedings of the Symposium on
Dungeness Crab Biology and Management, p. 11-25.
Alaska Sea Grant Rep. No. 85-3.
Wickham, D. E.
1979. The relationship between megalopae of the Dungeness
crab. Cancer mugister, and hydroid, Velella velella, and its
541
FISHERY BULLETIN: VOL. 86, NO. 3
influence on abundance estimates of C. magister megalopae.
Calif. Fish Game 65:184-186.
Wild, P. W., and R. N. Tasto (editors).
1983. Life history, environment, and mariculture studies of
the Dungeness crab. Cancer magister, with emphasis on the
central California fishery resource. Calif. Dep. Fish Game
Fish Bull. 172, 352 p.
542
FOOD PATHWAYS ASSOCIATED WITH PENAEID SHRIMPS IN
A MANGROVEFRINGED ESTUARY
Allan W. Stoneri and Roger J. Zimmerman^
ABSTRACT
High abundance and production of juvenile Penaeus spp. in tropical estuaries has been attributed to high
concentrations of mangrove-derived detritus in the nursery habitats. Examination of the diets o{ Penaeus
notialis, P. subtilis, and P. brasiliensis in the mangrove-fringed Laguna Joyuda, Puerto Rico showed
that even the smallest juveniles are predators consuming capitellid polychaetes (20-38% of diets) and
amphipods (20-76%). Less than 25% of the diets was detritus. Ontogenetic variation in diets was greater
than interspecific variation, and there was no evidence for dietary separation among the sympatric species.
Seasonal shifts in foods reflected abundance patterns of macrobenthic prey species. Despite the con-
sumption of prey organisms generally classified as detritivorous, stable carbon isotope ratios in the
penaeids (-18.1 to -15.0"/oo), their food items (-18.8 to -17.7''/oo), and primary producers indicated
that shrimps and the majority of sediment dwellers in Laguna Joyuda obtain most of their carbon from
benthic algae (-14.4°/oo) and not from mangrove detritus (-25.0 to -22.9°/oo).
The juveniles of penaeid shrimps are abundant in
many tropical and subtropical estuaries of the world,
particularly where wetland habitats such as marsh
grasses or mangroves are prominent coastal fea-
tures (Edwards 1978; Staples 1980; Stoner 1988).
The relationship between wetlands and commercial
shrimp has been attributed to at least two factors:
high food abundance and shelter from predators, the
relative importance of which is still debated (Boesch
and Turner 1984). Early wetland studies (Odum and
Heald 1972, 1975) suggested that a large variety of
fishes and invertebrates including Penaeus spp.
were directly dependent upon detritus from the
vascular wetland plants. More specific studies of
penaeid diets have since revealed that the shrimps
are omnivorous or carnivorous in many shallow-
water habitats (Moriarty 1976, 1977; Chong and
Sasekumar 1981; Moriarty and Barclay 1981) and
that some species may have a direct influence on
the abundance of small macrofauna (Leber 1983,
1985).
The most important commercial shrimp species
in the Caribbean Sea and along the north coast of
South America are Penaeus notialis, P. subtilis, P.
brasiliensis, and P. schmitti, all of which have
nurseries in coastal wetland habitats (Stoner 1988).
'Center for Energy and Environment Research, University of
Puerto Rico, Mayaguez, Puerto Rico 00708; present address; Carib-
bean Marine Research Center, 100 E. 17th Street, Riviera Beach,
FL 33404 and Lee Stocking Island, Exuma Cays, Bahamas.
^Southeast Fisheries Center Galveston Laboratory, National
Marine Fisheries Service, NOAA, 2700 Avenue U, Galveston, TX
77550.
Although the biology of tropical Atlantic penaeids
has been studied in a few localities (Neiva 1969;
Nikolic and Ruiz 1969; Lindner 1971; Garcia 1974;
Garcia et al. 1985), we have found no information
on their diets.
Examinations of shrimp diets reported here were
conducted in a small coastal lagoon in Puerto Rico
where Penaeus populations are large and where four
species coexist (Stoner 1988). In this report, diets
of the three most abundant Penaeus species are
described with particular reference to ontogenetic
variation, similarities among the species, the sig-
nificance of detritus, and seasonality in diets as
related to the abundance of foods in the field.
Because foods are retained and assimilated differ-
entially in the gut, examination of gut contents does
not necessarily give a true indication of the relative
importance of foods. For example, detritus has been
reported as a significant component of the gut con-
tents of Penaeus spp. (Odum and Heald 1972), but
detritus is known to be indigestible compared with
soft-bodied prey organisms such as polychaetes.
Furthermore, in the case of predators, the primary
source of carbon is not revealed through gut anal-
ysis. For these reasons, stable carbon isotope ratios
were measured for Penaeus spp. and most other
organisms common in the lagoon.
METHODS AND MATERIALS
Shrimps were collected in Laguna Joyuda on the
west coast of Puerto Rico Gat. 18°07'N, long.
Manuscript accepted March 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988, ^t^^-ST )
543
FISHERY BULLETIN: VOL. 86, NO. 3
67°irW). The study site and shrimp and fish pop-
ulations in the lagoon have been described by Stoner
(1986, 1988). Briefly, Laguna Joyuda is a polyhaline
basin with a total surface area of 1.21 km^ and an
average depth of approximately 1.5 m. One narrow
channel 0.5 km long represents the only connection
to the Puerto Rico shelf. Lagoon sediments are
mostly fine mud and sand with very high organic
content derived from red, white, and black man-
groves which surround approximately 75% of the
shoreline. Over the last 20 years the lagoon has had
a salinity range of 4 to 44''/oo, depending upon pre-
cipitation and degree of channel closure (A. W.
Stoner, unpubl. data).
Four penaeids utilize the lagoon as a nursery area.
Penaeus notialis and P. subtilis are approximately
equal in abundance and together comprise over 92%
of the total penaeid assemblage (Stoner 1988).
Penaeus brasiliensis made up 7.0% of the total col-
lection in 1984 to 1985, and P. schmitti was rela-
tively uncommon, making up <0.3% of the total
(Table 1). For this study, we examined the diets of
the three most abundant shrimp species.
Penaeids were collected with a 5 m otter trawl
with 2.5 cm wings and body, and 5 mm cod end liner.
All collections were made between nautical twilight
and midnight at three sites: a northern muddy bot-
tom arm of the lagoon (station 5), a sandy mud site
in the central basin (station 3), and a sandy mud site
near the channel (station 1). Collections were made
monthly, during the last quarter of the moon, from
July 1985 to June 1986.
Shrimps were divided into size classes on the basis
of carapace length (CL) for gut content analyses.
Penaeid shrimps between 3 and 6 mm CL could not
be identified to species and were simply identified
as Penaeus juveniles (all were of the grooved vari-
ety and, therefore, did not include P. schmitti).
Shrimps larger than 6 mm CL could be identified
and were examined by individual species in 4 mm
size classes, up to 26 mm CL in the case of P. sub-
tilis. To yield sufficient numbers of individuals in
Table 1 . — Composition of the penaeid shrimp assemblage at three
stations in Laguna Joyuda, Puerto Rico, during the 12-mo study
period. Values are total numbers collected in 72 trawl samples and
percentages of totals at the individual sites (parentheses).
Station
Species
1
2
3
Penaeus notialis
P. subtilis
P. brasiliensis
P. schmitti
Totals
282 (49.1)
236 (41.1)
54 (9.4)
2 (0.4)
574
271 (53.6)
185 (36.5)
50 (9.9)
0 (0)
506
229 (48.6)
217(46.1)
22 (4.7)
3 (0.6)
471
all size classes for each monthly collection, members
of individual species were pooled from all sampUng
areas. In the case of P. brasiliensis, collections from
2-mo intervals were pooled to analyze seasonal
variation in the diets of this less abundant species.
Food items taken from the proventriculus of up to
25 shrimp were pooled for each sampling date and
size class, and preserved with 70% isopropanol and
a dilute solution of rose bengal stain.
We used the gravimetric sieve fractionation pro-
cedure developed by Carr and Adams (1972) to
analyze gut contents of the shrimp. This procedure
has been widely used for juvenile fishes (Sheridan
1979; Stoner 1980; Livingston 1984) and a variety
of decapod crustaceans, including Penaeus spp.
(Laughlin 1982; Leber 1983). Gut contents were
washed through a series of six sieves of decreasing
mesh size (2.0-0.075 mm mesh) and each sieve frac-
tion was examined with a dissecting microscope.
Because all of the items in a particular sieve frac-
tion were of approximately equal size, the relative
proportion of the gut contents made up of each food
type was measured directly by counting. After ex-
amination, each sieve fraction was dried overnight
at 80°C and the total contribution of each food type
to total dry weight was calculated.
With few exceptions, each food particle was placed
in a mutually exclusive category (Table 2). In most
cases, food items or fragments could be identified
to major taxonomic group such as Amphipod or
Polychaete. The classification "Animal Remains"
was applied where fragments were unidentifiable
to taxon, but where the tissue was stained by rose
bengal. The major food categories were used for
statistical interpretation of diets; however, when-
ever an animal or plant could be identified to a lower
taxonomic level, this information was recorded.
Similarities between and among the diets of
various shrimp species and size classes were mea-
sured with Czekanowski's coefficient (Bray and
Curtis 1957; Field and McFadane 1968). Dendo-
Table 2. — List of the general food categories encountered in the
foreguts of Penaeus species and the codes employed in histograms
for shrimp diets.
AM Amphipod OS
CC Calanoid copepod PM
CY Cyclopoid copepod PO
CZ Crab zoea RU
DE Detritus SA
FO Foraminifera TA
FR Fish remains TH
HC Harpacticoid copepod
GA Gastropod MS
IE Invertebrate egg
NE Nematode
Ostracod
Plant material (green)
Polychaete
Ruppia maritimia
Sand
Tanaidacean
Thalassia testudinum
Miscellaneous— used in
histograms for all food
items making up <4% of
the total dry weight.
544
STONER and ZIMMERMAN: FOOD PATHWAY ASSOCIATED WITH PENAEID SHRIMPS
grams were then constructed by complete linkage
classification.
Plant and animal materials were collected from
the lagoon for stable carbon isotope analyses on
several occasions between 1981 and 1984. These
were taken to give an indication of the range of
d^^C values between organisms associated with
mangroves, plankton, and benthos and to test the
efficiency of using the technique to identify food
pathways. Materials were collected with trawl,
sieve, plankton net, and by hand. All materials were
fresh and not exposed to chemical preservatives. In
the field, samples were placed in plastic bags on ice
as temporary storage. Within 24 hours, these sam-
ples were flushed free of salt using deionized water,
dissected to acquire tissue uncontaminated by gut
contents and outside shells, treated for 5 to 10
minutes with 5% phosphoric acid to remove carbon-
ates, double rinsed in deionized water, and oven-
dried at 70°C until brittle. Dried samples were
ground to a fine powder with a mortar and pestle,
packaged in plastic bags, labeled, and stored for
later mass spectrometer analyses. The homogenized
samples were combusted at 550 °C for 24 hours in
the presence of CuO in evacuated sealed pyrex tubes
using techniques modified from Stofer (1980). CO2
gas was analyzed from the combusted samples on
a Finnigan - MAT 25 P isotope ratio mass spectrom-
eter. d^^C was calculated according to Craig (1957)
and methods and definitions generally followed
those reviewed by Fry and Sherr (1984). For small
animals and all plants, the entire organism (with
guts removed where possible) was used in analyses.
For large animals, muscle tissue was removed from
the body and used separately for analyses. In most
cases, organisms were pooled to acquire mean d^^C
values for n individuals. Unlike more complex
marine systems (Fry et al. 1982), our d^^C were
well separated between groups of species and dis-
tinctions were relatively clear-cut. Since diversity
was also relatively low in the lagoon, d^^C signa-
tures in food pathways based on the dominant
primary producers were easily detected. Some
organisms with known restricted diets, that were
closely associated with particular plants in the
system, were selected as controls to follow trophic
fractionation of d^^C. These included Uca vocator
from the intertidal forest floor (a mangrove detritus
feeder), Haminoea antillarum from the middle of
the lagoon (an algae grazer), and Balanns ebumetcs
and Isognomon alatus (filter feeders). Based on
'Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
results of Fry et al. (1984), we assumed that ^^C
enriched the isotope ratio by about l^/oo from one
trophic level to the next.
RESULTS
The Shrimp Community
The shrimp community composition during the
study period was similar at the three sampling sites,
except that Penaeus schmitti were not collected at
station 3 (Table 1). Penaeus notialis dominated the
shrimp assemblage in the lagoon (50.4%), followed
by P. subtilis (41.1%), and P. brasiliensis (8.1%).
Only five individuals of P. schmitti (0.3%) were
collected in 12 months of sampling. Catch per unit
effort over time was highly variable for each of the
three primary species, with no distinct seasonality
(Fig. 1). Penaeus subtilis, however, showed a
general increase in abundance over the sampling
period. Penaeus brasiliensis populations remained
relatively low and stable throughout the year ex-
amined. Population and community structure of the
penaeids in Laguna Joyuda has been examined in
detail elsewhere (Stoner 1988) and the important
point to be made here is that the three most abun-
dant Penaeus spp. in Laguna Joyuda are sympatric
in both time and space.
Q
UJ
I-
O
o
o
w
D.
1
DC
I
CO
_l
<
H
O
7 8 9 10 11 12 1 2 3 4 5 6
MONTHS
Figure 1.— Abundance of the three dominant Penaeus species in
Laguna Joyuda, Puerto Rico, during the study period. The unit
of effort for each sampling date is 18 trawl samples, pooled over
all three stations. P.n. = Penaeus notialis; P.s. = P. subtilis; P.b.
= P. brasiliensis.
Shrimp Diets
Over 1,300 Penaeus guts were examined for the
present study. The proventriculi of most individuals
contained freshly consumed food items; 89% of all
545
FISHERY BULLETIN: VOL. 86, NO. 3
P. notialis, 87% of all P. subtilis, and 78% of all
P. brasiliensis contained food. At least 50% of all
animals in any one size class contained food on a
given sampling date, but no seasonal or ontogenetic
pattern in empty guts w&s evident.
The primary dietary components of all three
penaeids were amphipods, polychaetes, harpacticoid
copepods, and detritus (Fig. 2). For all species in-
crease in shrimp size was correlated with decrease
in the relative importance of harpacticoid copepods
in the guts. Other small taxa such as nematodes and
foraminifera also decreased with shrimp size. Detri-
tal components of the diets remained relatively con-
stant at approximately 20 to 25% of the gut con-
tents. Abundance of polychaetes and amphipods
changed relatively little with size in P. notialis and
P. subtilis, except that amphipods increased in im-
portance with size in P. subtilis and large quantities
of amphipods were taken by the largest size class.
Cluster analyses for the diets of individual species
revealed little ontogenetic variation for these two
species, except with the distinct separation of the
largest P. subtilis (Fig. 2). For both P. notialis and
P. subtilis, all size classes between 7 and 22 mm CL
were clustered within similarity indices of 0.78.
Although only three size classes were represented
for P. brasiliensis, ontogenetic variation in P.
brasiliensis was greater than in the other two
species (Fig. 2). Polychaete consumption decreased
from 47% of the diet in the 7 to 10 mm class to 22%
in the 15 to 18 mm class, while amphipod consump-
tion increased from 0 to 61%. Detritus consumption
also decreased with size in P. brasiliensis, contrib-
uting to the low similarity indices among the size
classes.
Four major clusters of shrimp feeding types were
Penaeus juvenile
3-6| AM I PO
HC
I
DE
^(235)
0 10 20 30 40 50 60 70 80 90 100
E
E,
X
I-
(3
LU
LD
O
<
CL
<
<
o
7-10
11-14
15-18
Penaeus notialis
7-10
AM 1
PO 1 HC
°^ VA
11-14
AM
1 PO
HC
DE
%
15-18
AM
PO 1
OE V/y
19-22
AM
PO
1 DE Y/^
(206)
(215)
(50)
(8)
0 10 20 30 40 50 60 70 80 90 100
J-
X
J-
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3
Penaeus subtilis
10 20 30 40 50 60 70 80 90 100
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3
Penaeus brasiliensis
PO
AM
HC
DE
PO
AM
HC
009)
DE
PO
DE
(82)
](21)
0 10 20 30 40 50 60 70 80 90 100
X
X
X
J
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3
PERCENT OF DIET
(N)
SIMILARITY
Figure 2.— Diets of the three dominant Penaeus species in Laguna Joyuda, Puerto Rico, shown as a function
of shrimp size. Food categories are identified in Table 2. The similarity index used in the cluster diagrams is
Czekanowski's coefficient.
546
STONER and ZIMMERMAN: FOOD PATHWAY ASSOCIATED WITH PENAEID SHRIMPS
revealed when the full matrix of similarity indices,
including all species and size classes, was incor-
porated into a cluster analysis (Fig. 3). The greater
ontogenetic variation in diets of P. brasiliensis
resulted in the three size classes for the species fall-
ing into three distinct clusters, while all four size
classes of P. notialis were included under two
clusters. The largest P. subtilis were clustered with
the largest P. brasiliensis. Except for P. brasili-
ensis, all shrimps between 7 and 14 mm CL were
found in one trophic group and all between 15 and
22 mm CL were found in a second group.
Species or generic level identification of prey
organisms indicated no species-specific differences
among the diets of the three shrimp species. For all
three species, all amphipods identifiable to species
were Grandidierella bonnieroides and all identifi-
able polychaetes were nereids (probably Nereis
occidentalis and Steninonereis martini). In P. sub-
tilis and P. brasiliensis, all harpacticoid copepods
were Euterpina spp. In P. notialis, 86% were Euter-
pina spp. and 14% were Microsetella sp. The cala-
noid copepods were a mixture ofAcartia tonsa and
Pseudodiaptomiis spp.
Animals classified in groups 2 and 3 of the multi-
species cluster (Fig. 3) contained sufficient numbers
of individuals to make seasonal analyses of diets
(Figs. 4, 5). Although harpacticoid copepods were
taken in lower amounts by group 3 shrimps, by and
large both groups showed similar seasonal trends
in diet. Amphipods were taken in large numbers
from July to October 1985 and from March through
June 1986. Polychaetes and harpacticoid copepods
were consumed most abundantly in November
through March. Consumption of detritus was rela-
tively constant in both groups, with slightly higher
detrital intakes in group 2 individuals during
November and December 1984.
SIMILARITY
1.0 0.8 0.6 0.4 0.2
I 1 1 1 1 1 1 1 1 —
B-7
Juv.
N-7
S-7
B-11
S-11
N-11
N-15
S-15
S-19
N-19
B-15
S-23
3
Figure 3.— Cluster diagram for the diets of
shrimps incorporating all species and size classes.
The cluster strategy is the same as that in Figure
2. Food categories are identified in Table 2.
100
10 11 12 1 2 3 4 5 6
MONTHS
Figure 4.— Diets of shrimp trophic group 2 by sampling date.
Carbon Isotopes
The d^^C values for three different samples of
tissues from Penaeus spp. ranged from -15.0 to
-18.1"/oo (Table 3). These values were much higher
than the plankton fraction <35 /i (comprised primar-
ily of dinoflagellates; -26.8 to -27.2''/oo) or the
primary copepod species in the lagoon, Acartia
tonsa (-24.0 to -25.9°/oo). The d^^C values for the
shrimps are also much higher than the values for
either mangrove leaves (green or dead), or detritus
particles from the sediment, mostly of mangrove
origin. The only primary producers with d^^C
values within the range of Penaeus spp. were the
seagrass Thalassia testudinum (- 16.1°/oo) and the
100
7 8 9 10 11 12 1 2 3 4 5 6
MONTHS
Figure 5.— Diets of shrimp trophic group 3 by sampling date.
547
FISHERY BULLETIN: VOL. 86. NO. 3
algal mat comprised primarily of the filamentous
blue-green alga Spirulina sp. Because there are only
small patches of T. testudinum in Laguna Joyuda
and the blue-green algae dominated the lagoon floor,
the data suggest that the main source of carbon for
penaeid shrimps in Laguna Joyuda is benthic blue-
green algae. Blue-green algae also appear to be the
primary source of carbon for other decapod crusta-
ceans in the lagoon including Callinectes spp. and
hermit crabs (Paguridae) (Table 3).
The carbon isotope data, suggesting algal sources
of carbon for Penaeus spp., are consistent with the
results of dietary analyses. The primary prey of
penaeids were the amphipod Grandidierella bon-
nieroides and capitellid polychaetes. The d^^C value
for G. bonnieroides was within the range for
Penaeus spp., -YI.TIoo. Similarly, when the bodies
of capitellids were analyzed after removal of the
guts, the d^^C value of - 18.8°/oo also agrees with
shrimps after adjusting for trophic fractionation of
rVoo less (Fry et al. 1984). Other organisms with
d^^C values similar to that of blue-green algae were
the bubble snail Haminoea antillarum and the
bivalve Macoma brevifrons. The carbon isotope
ratios of these benthic feeders were more negative
than Spirulina (-1 A. V'/(m), implying modification of
d^^C values from other sources; however, the major
input appears to be from blue-green alga.
By contrast, organisms associated with the man-
grove forest, Uca vocator (-23.0Voo) and Aratus
pisonii (-23.3*'/oo), closely resembled carbon
isotope ratios of detritus (-22.9 to -25.0°/oo).
Moreover, the <35 fi plankton fraction (-26.8 to
-27.0"/oo) and associated grazers, Acartia (-24.0
to -25.9%o), Balanus (-24.0''/oo), and Isognomon
(-24.4°/oo), were not separable from the mangrove
group. These data indicate that mangrove carbon
is likely being incorporated into plankton-based food
Table 3. — Carbon isotope ratios in organisms associated with the penaeid shrimp food web in Laguna
Joyuda, Puerto Rico.
Number
Tissue or
of pooled
Date of
Organisms
composition
indiv.
collection
6'^C
Plankton
<35fi fraction
—
9/29/83
-26.8 to -27.2
Acartia tonsa
whole bodies
500
1/21/81
-25.9
Acartia tonsa
whole bodies
500
9/8/83
-25.2
Acartia tonsa
whole bodies
500
9/27/83
-24.0
Macroalgae
Spirulina sp.
entire filaments
—
2/14/83
-14.2
Caulerpa sertularoldes
entire alga
—
9/29/83
-20.2
Seagrass
Thalassia testudinum
green leaves
—
9/29/83
-16.1
Mangroves
Avicennia germinans
roots
—
2/14/84
-24.0
Rhizophora mangle
green leaves
—
9/24/81
-29.2
Rhizophora mangle
dead leaves
—
2/14/84
-27.3
Detritus
Particules from the sediment
—
9/24/81
-22.9 to -25.0
Polychaetes
Capitellidae
body without gut
13
9/24/81
-18.8
Capitellidae
whole bodies
10
2/14/81
-23.1
Capitellidae
whole bodies
10
9/30/81
-22.0
Capitellidae
whole bodies
10
2/14/84
-22.9
Nereidae
whole bodies
2
2/14/84
-21.6
t^ollusca
Haminoea antillarum
whole bodies
1
9/24/81
-15.8
Isognomon alatus
whole bodies
6
9/24/81
-24.4
Macoma brevifrons
whole bodies
6
9/24/81
-18.8
Crustacea
Aratus pisonii
leg muscle
10
9/30/81
-23.3
Balanus eburneus
whole bodies
20
9/24/81
-24.0
Callinectes danae
claw muscle
2
2/14/84
-15.4
Callinectes sapidus
claw muscle
1
2/14/84
-16.4
Grandidierella bonnieroides
whole bodies
52
2/14/84
-17.7
Pagurus sp.
leg muscle
3
2/14/84
-16.2
Penaeus notialis
tail muscle
12
9/31/81
-18.1
Penaeus notialis
tail muscle
10
2/14/84
-15.4
Penaeus subtilis
tail muscle
10
2/14/84
-15.0
Petrolisthes sp.
whole bodies
12
9/24/81
-23.5
Uca vocator
leg muscle
10
9/30/81
-23.0
548
STONER and ZIMMERMAN: FOOD PATHWAY ASSOCIATED WITH PENAEID SHRIMPS
chains in the lagoon. This can be supported using
Acartia as a link, since the copepod is known to feed
on detritus particles and because the growth of
Acartia is improved with detritus is included in the
diet (Roman 1984). Suspended detritus particles
from mangroves were a component of our <35 fi
plankton fraction, suggesting availability to plank-
tonic consumers.
DISCUSSION
Early studies of penaeid diets led to the general
conclusion that the shrimps were largely detritivor-
ous with incidental amounts of animal or plant
material in the guts (Williams 1955; Darnell 1958;
Odum and Heald 1972). Dall (1968) concluded that
several Australian penaeids were not predators, but
consumers of small-sized and disabled animals. Al-
though penaeid shrimps are known to sort sediments
for organic particles with their delicate pereiopods
(Dall 1968; Lindner and Cook 1970) it is now ap-
parent that penaeid shrimps are capable of taking
prey organisms such as large polychaetes, as well
as the more difficult prey such as gastropods and
bivalves, caridean shrimps, crabs, echinoderms, and
even fishes (Moriarty 1977; Marte 1980; Leber
1983). Lindner and Cook (1970) reported that
Penaeus setiferus is cannibalistic at times. The rela-
tively constant low amount of detrital particles in
the guts of shrimp from Laguna Joyuda suggests
that the less readily digestible detritus is taken in-
cidentally with animal prey or as a response to low
prey abundance. In addition, Gleason and Zimmer-
man (1984) showed that nematodes, oligochaetes,
polychaetes, and copepods were stripped from
detritus by P. aztecits during feeding. Reports of
shrimp filled with unrecognizable debris assumed to
be detritus may be a consequence of actual detritus
consumption or incomplete development of methods
sufficient to make animal remains identifiable.
Despite considerable attention given to feeding in
the commercially significant penaeids, relatively few
investigators have examined ontogenetic variation
in diets. In this study, even within juvenile size
classes, there were clear patterns of dietary change
with shrimp size. All three of the subject species
abandoned smaller prey organisms such as foramini-
fera, nematodes, and harpacticoid copepods in favor
of amphipods, polychaetes, and shrimp as the pred-
ators became larger. Similar findings were reported
by Leber (1983) for P. duorarum. These ontogenetic
shifts in prey are undoubtedly related to increasing
size of the chelae and mouth parts and must be con-
sidered in any ecological interpretation of trophic
position; Moriarty (1977) suggested that the great
opportunism of penaeid shrimps precluded their
being placed in one trophic group throughout their
life cycle.
On the other hand, with the exception of acceler-
ated ontogenetic shifts in the diets of P. brasilien-
sis, interspecific variation in feeding in Laguna
Joyuda were relatively minor. Despite high abun-
dance and great temporal and spatial overlap in the
Penaeus spp. of the lagoon (Stoner 1988), coupled
with very low biomass values for macrofauna (<2
g dry wt/m^), there is no evidence for resource par-
titioning among the grooved shrimps.
Seasonal diets in Penaeus spp. were coincident
with seasonal trends in the abundance of major prey
organisms. In the lagoon, the one important amphi-
pod species, Grandidierella bonnieroides, demon-
strated maxima between July and September and
again from May to June; this corresponded with
maximum amphipod consumption between July and
October and March through June in the same year
as benthic studies. Polychaete consumption was
highest between November and March (particular-
ly in the group 3 shrimps), corresponding with the
October to April peak in capitellid abundance in the
lagoon. Consequently, the oscillation in polychaete
and amphipod feeding appears to be related to the
availability of foods.
The hypothesis that mangrove estuaries are fueled
primarily by carbon from mangrove detritus has
become established from the pioneering work on
mangrove-associated food webs conducted in the
North River estuary of south Florida (Odum and
Heald 1972, 1975). Mangrove litter inputs to Laguna
Joyuda are high (Levine 1981), the sediments are
rich in organic content, and detritus comprised a
portion of the gut contents of juvenile shrimp. It is
unlikely, however, that a large amount of carbon
derived from detritus or detritus-associated
microbes contributes in a large way to the tissues
of the shrimp in the lagoon. In fact, the only organ-
isms which had carbon isotope ratios similar to that
of detritus were those normally found in direct
association with the trees such as fiddler and man-
grove crabs. The d^^C values for shrimp in Laguna
Joyuda were, in fact, very similar to values for
penaeids from the open waters of the Gulf of Mex-
ico (Fry and Parker 1979).
Benthic algal primary production in Laguna
Joyuda is probably highly significant because car-
bon isotope ratios in the algae and shrimp were
similar. Gleason (1986) also found that juvenile P.
aztecu^ penaeids in a Galveston Bay salt marsh
derived their carbon from blue-green algae, green
549
FISHERY BULLETIN: VOL. 86, NO. 3
algae, and epiphytes of salt marsh grasses, not from
Spartina detritus. Algal foods also proved to be the
primary source of carbon for penaeids in Georgia
marshes (Hughes and Sherr 1983) and in P. duora-
rum inhabiting seagrass beds in the Gulf of Mexico
(Leber 1983) where food webs have long been con-
sidered detritus based. The significance of algal pro-
duction in mangrove areas has been pointed out by
Rodelli et al. (1984). They concluded that both
mangrove and algal carbon were utilized by most
of animals in a Malaysian swamp, but virtually no
animals collected at distances >2 km from the
swamp contained mangrove-derived carbon. To this
date, there is little evidence to suggest that natant
species such as fishes and decapod Crustacea can
use detritus as a primary food source even if
microbially enriched (Boesch and Turner 1984).
Localities where mangroves grow often support
a variety of other primary producers, and variation
in the relative significance of detrital and algal car-
bon sources may be associated with tidal amplitude
and flushing. Laguna Joyuda, with microtidal influ-
ences, appears to be fueled by algal carbon where-
as shrimp-producing mangrove areas in Ecuador
have meso- and macrotidal regimes and no apparent
algal growth (Zimmerman, pers. observ.). That man-
groves provide critical substratum and protective
cover for a large number of organisms is undisputed;
however, the assumed significance of mangrove-
derived detritus should be examined further.
ACKNOWLEDGMENTS
This research was supported by the Office of Sea
Grant, NOAA (Grant no. R/A-01-2), a grant from
the National Science Foundation (No. R-II-8610677),
and funding from the University of Puerto Rico to
the Center for Energy and Environment Research.
D. Corales participated in the field work, along v^ath
G. Owen and 1. Sanders. B. A. Buchanan identified
the shrimps collected and assisted in the prepara-
tion of figures. L. L. Cruz conducted all of the gut
analyses and T. Robles assisted in manuscript prep-
aration. W. J. Richards and B. A. Buchanan pro-
vided helpful criticisms of the manuscript.
LITERATURE CITED
Boesch, D. F., and R. E. Turner.
1984. Dependence of fishery species on salt marshes: the role
of food and refuge. Estuaries 7:460-468.
Bray, J. R., and J. T. Curtis.
1957. An ordination of the upland forest communities of
southern Wisconsin. Ecol. Monogr. 27:325-349.
Carr, W. E. S., and C. a. Adams.
1972. Food habits of juvenile marine fishes: evidence of the
cleaning habit in the leatherjack, Oligoplites saurus, and the
spottail pinfish, DiplodxLS holbrooki. Fish. Bull., U.S. 70:
1111-1120.
Chong, V. C, AND A. Sasekumar.
1981. Food and feeding habits of the white prawn Penaeiis
merguiensis. Mar. Ecol. Prog. Ser. 5, p. 185-191.
Craig, H.
1957. Isotopic standards for carbon and oxygen and correc-
tion factors for mass-spectrometric analysis of carbon diox-
ide. Geochim. Cosmochim. Acta 12:133-149.
Dall, W.
1968. Food and feeding of some Australian penaeid shrimp.
FAO Fish. Rep. 2:251-258.
Darnell, R. M.
1958. Food habits of fishes and larger invertebrates of Lake
Ponchartrain, Louisiana. Publ. Inst. Mar. Sci., Univ. Texas.
5:353-416.
Edwards, R. R. C.
1978. The fishery and fisheries biology of penaeid shrimp on
the Pacific coast of Mexico. Oceanogr. Mar. Biol. Annu.
Rev., p. 145-180.
Field, J. G., and G. McFarlane.
1968. Numerical methods in marine ecology. I. Quantitative
"simUarity" analysis of rocky shore samples in False Bay,
South Africa. Zool. Air. 3:119-138.
Fry, B., R. K. Anderson, L. Entzeroth, J. L. Bird, and P. L.
Parker.
1984. '^C enrichment and oceanic food web structure in the
northwestern Gulf of Mexico. Contrib. Mar. Sci. 27:49-63.
Fry, B., R. Lutes, M. Northam, and P. L. Parker.
1982. A '^C/'^C comparison of food webs in Caribbean
seagrass meadows and coral reefs. Aquat. Bot. 14:389-398.
Fry, B., and P. L. Parker.
1979. Animal diet in Texas seagrass meadows: d^^C evidence
for the importance of benthic plants. Estuarine Coastal
Mar. Sci. 8:499-509.
Fry, B., and E. B. Sherr.
1984. d^^C measurements as indicators of carbon flow in
marine and freshwater ecosystems. Contrib. Mar. Sci. 27:
13-47.
Garcia, S.
1974. Biologie de Penaeus duorarum notialis en Cote
D'lvoire, IV. Relations entre la repartition et les conditions
du milieu - etude des variations du sex-ratio. Doc. Sci. Cent.
Rech. Oceanogr. Abidjan. 5(3-4): 1-39.
Garcl\, S., M. Lemoine, and E. Lebrun.
1985. Seasonal and long-term variability of recruitment in
French Guiana shrimp fishery on Penaevs suhtilis. FAO
Fish. Rep. No. 327, SuppL, p. 242-250.
Gleason, D. F.
1986. Utilization of salt marsh plants by postlarval brown
shrimp: carbon assimilation rates and food preferences.
Mar. Ecol. Prog. Ser. 31:151-158.
Gleason, D. F., and R. J. Zimmerman.
1984. Herbivory potential of postlarval brown shrimp associ-
ated with salt marshes. J. Exp. Mar. Biol. Ecol. 84:235-
246.
Hughes, E. H., and E. B. Sherr.
1983. Subtidal food webs in a Georgia estuary: d^^C anal-
ysis. J. Exp. Mar. Biol. Ecol. 67:227-242.
Laughlin, R. a.
1982. Feeding habits of the blue crab, Callinectes sapidus
Rathbun, in the Apalachicola Estuary, Florida. Bull. Mar.
Sci. 32:807-822.
550
STONER and ZIMMERMAN: FOOD PATHWAY ASSOCIATED WITH PENAEID SHRIMPS
Leber, K. M.. III.
1983. Feeding ecology of decapod crustaceans and the in-
fluence of vegetation on foraging success in a subtropical
seagrass meadow. Ph.D. Thesis, Florida State Univ.,
Tallahassee, 166 p.
1985. The influence of predatory decapods, refuge, and
microhabitat selection on seagrass communities. Ecology
66:1951-1964.
Levine, E. a.
1981. Nutrient cycling by the red mangrove, Rhizophora
mangle L., in Joyuda Lagoon on the west coast of Puerto
Rico. M.S. Thesis, Univ. Puerto Rico, Mayaguez, 103 p.
Lindner, M. J.
1971. Shrimp resources of the Caribbean Sea and adjacent
regions. FAO Fish. Rep. 2:149-156.
Lindner, M. J., and H. L. Cook.
1970. Synopsis of biological data on the white shrimp Penaeus
setiferus (Linnaeus) 1767. FAO Fish. Rep. 4:1439-1469.
Livingston, R. J.
1984. Trophic response of fishes to habitat variability in
coastal seagrass systems. Ecology 65:1258-1275.
Marte, D. L.
1980. The food and feeding habit of Penaeus monodon
Fabricus collected from Makato River, Aklan, Philippines
(Decapoda, Natantia). Crustaceana 38:225-236.
MORIARTY, D. J. W.
1976. Quantitative studies on bacteria and algae in the food
of the mullet Mugil cephal-us L. and the prawn Metapenaeus
bennettae (Racek & Dall). J. Exp. Mar. Biol. Ecol. 22:131-
143.
1977. Quantification of carbon, nitrogen and bacterial
biomass in the food of some penaeid prawns. Aust. J. Mar.
Freshwater Res. 28:113-118.
MORIARTY, D. J. W., AND M. C. BARCLAY.
1981. Carbon and nitrogen content in food and assimilation
efficiencies of penaeid prawns in the Gulf of Carpentaria.
Aust. J. Mar. Freshwater Res. 32:245-251.
Neiva, G. deS.
1969. Observations on the shrimp fisheries of the central and
southern coast of Brazil. FAO Fish. Rep. 3:847-858.
NiKOLic, M., and M. E. Ruiz de Quevedo.
1969. Notas biologico-pesqueras sobre camaron bianco
Penaeus schmitti Burkenroad 1936. FAO Fish. Rep. 3:
1107-1117.
Odum, W. E., and E. J. Heald.
1972. Trophic analyses of an estuarine mangrove community.
Bull. Mar. Sci. 22:671-738.
1975. The detritus-based food web of an estuarine mangrove
community. In L. E. Cronin (editor), Estuarine research
Vol. L p. 265-286. Acad. Press, N.Y.
RoDELLi, M. R., J. N. Gearing, P. J. Gearing, N. Marshall,
AND A. SaSEKUMAR.
1984. Stable isotope ratio as a tracer of mangrove carbon in
Malaysian ecosystems. Oecologia 61:326-333.
Roman, M. R.
1984. Utilization of detritus by the copepod, Acartia txmsa.
Limnol. Oceanogr. 29:949-959.
Sheridan, P. F.
1979. Trophic resource utilization by three species of sciaenid
fishes in a northwest Florida estuary. Northeast Gulf Sci.
3:1-15.
Staples, D. J.
1980. Ecology of juvenile and adolescent banana prawns,
Penaeus merguiensis, in a mangrove estuary and adjacent
offshore area of the Gulf of Carpentaria. L Immigration and
settlement of postlarvae. Aust. J. Mar. Freshwater Res.
31:635-652.
Stofer, Z.
1980. Preparation of carbon dioxide for stable carbon isotope
analysis of petroleum fractions. Anal. Chem. 52:1389-1391.
Stoner, a. W.
1980. The feeding ecology of Lagodon rhovnhoides (Pisces:
Sparidae): Variation and functional responses. Fish. Bull.,
U.S. 78:337-352.
1986. Community structure of the demersal fish species of
Laguna Joyuda, Puerto Rico. Estuaries 9:142-152.
1988. A nursery ground for four tropical Penaeus species:
Laguna Joyuda, Puerto Rico. Mar. Ecol. Prog. Ser.
42:133-141.
Williams, A. B.
1955. A survey of the North Carolina shrimp nursery
grounds. J. EHsha Mitchell Sci. Soc. 71:200-207.
551
THE EFFECT OF THE ECTOPARASITIC PYRAMIDELLID SNAIL,
BOONEA IMPRESSA, ON THE GROWTH AND HEALTH OF
OYSTERS, CRASSOSTREA VIRGINICA, UNDER FIELD CONDITIONS
Elizabeth A. Wilson,^ Eric N. Powell/ and Sammy M. Ray^
ABSTRACT
Boonea (= Odostomia) impressa are contagiously distributed on oyster reefs so that some oysters are
parasitized more than others. The parasite's mobihty and the abihty of oysters to recover from snail
parasitism may be important in assessing the impact of parasitism on oyster populations. During a 4-week
exposure period in the field, B. impressa reduced American oyster, Crassostrea virginica, growth rate
and increased the intensity of infection by the protozoan, Perkinsus ( = Dermocystidium) marinus, but
produced few changes in the oyster's biochemical composition because, although net productivity was
reduced, the oysters retained a net positive energy balance (assimilation > respiration). During a 4-week
recovery period, growth rate returned to normal (control) levels, but infection by P. ynarinv^ continued
to intensify in previously parasitized oysters kept B. impressa-tree. Most changes in biochemical com-
position during recovery, including increased lipid and glycogen contents, could be attributed to the con-
tinuing increase in infection intensity of P. marinus. Consequently, the temporal stability and size of
snail patches, particularly as they regulate infection by P. marinus, may be the most important factors
influencing the impact of B. impressa on oyster reefs.
Parasitism can be an important factor affecting the
population dynamics (Wickham 1986; Brown and
Brown 1986; Rabat 1986) and health (Brockelman
1978; Mohamed and Ishak 1981; Ford 1986) of host
species. Three parasites are known to be especially
important in oysters. Perkinsus (= Dermocysti-
dium) marinus, Haplosporidium nelsoni (MSX), and
Boonea ( = Odostomia) impressa detrimentally af-
fect oyster growth, health, and biochemical composi-
tion (Mengebier and Wood 1969; Feng et al. 1970;
Soniat and Koenig 1982; White et al. 1984; Ford
1986; Ward and Langdon 1986; White et al. 1988,
in press).
The pyramidellid gastropod, Boonea impressa, is
one of a widely distributed group of parasitic, marine
opisthobranchs (Fretter and Graham 1949; Fretter
1951; Allen 1958). Boonea impressa removes nutri-
ents directly from its host by piercing the flesh with
a hollow stylet and sucking the body fluids using a
buccal pump (Fretter and Graham 1949; Fretter
1951; Allen 1958). The most common host of 5. im-
pressa is Crassostrea virginica (Hopkins 1956; Allen
1958; Wells 1959) but, like other odostomians, it is
not entirely host specific (Wells 1959; Robertson
1978; Robertson and Mau-Lastovicka 1979). Found
'Department of Oceanography, Texas A&M University, College
Station, TX 77843.
^Department of Marine Biology, Texas A&M University at
Galveston, Galveston, TX 77550.
Manuscript accepted March 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
abundantly on oyster reefs from Massachusetts to
the Gulf of Mexico, B. impressa has been reported
in numbers as high as 100 per oyster (Hopkins 1956).
Under laboratory conditions, B. impressa reduced
oyster growth rates as the result of both direct
removal of assimilated carbon from the oyster and
direct interference with the oyster's ability to feed
(White et al. 1984, 1988; Ward and Langdon 1986).
Not surprisingly, parasitism by B. impressa produc-
ed changes in the biochemical composition of oyster
tissue. Parasitism by 15 snails, a relatively high field
density found in dense snail patches decreased car-
bohydrate and free amino acid content and increased
lipid content of mantle tissue. Reproduction was af-
fected as well (White et al. in press).
Perkinsus marinus is an important cause of mor-
tality of oysters in the Gulf of Mexico (Mackin 1962;
Ray 1966a; Hofstetter 1977). Sublethal effects in-
clude reduced growth (Menzel and Hopkins 1955;
Ray et al. 1953) and changes in biochemical com-
position (Soniat and Koenig 1982; White et al. in
press). Perkinsus marinus can be transmitted from
one oyster to another through the water (Ray 1954;
Mackin 1962; Andrews 1965) or by B. impressa
feeding (White et al. 1987). The intensity of infec-
tion by P. marinus is also increased in B. impressa-
parasitized oysters (White et al. 1987).
Usually stress affects organisms by altering nor-
mal metabolic activity. Many organisms may be able
553
FISHERY BULLETIN: VOL. 86, NO. 3
to return to normal metabolic condition after the
stress is removed (Portner et al. 1979; Pickering et
al. 1982; Kendall et al. 1984). This ability, termed
recovery, is an important adaptation to changes in
the natural environment. Recovery is not always
complete nor does it occur over relatively short time
scales in all cases. The time required for recovery
varies depending on the parameter being studied,
the stress that was applied, and the species involved
(Kendall et al. 1984; Haux et al. 1985; Neff et al.
1985). Apparently deleterious effects occurring after
exposure, during the so-called recovery period, are
well described (e.g. Kendall et al. 1984; Powell et
al. 1984). White et al. (1984) showed that C.
virginica could attain normal growth rates within
one week after B. impressa were removed.
However, growth rates frequently return to normal
more rapidly than other metabolic parameters (Ken-
dall et al. 1984). Boonea impressa are extremely
mobile and are more contagiously distributed than
their hosts, so that some oysters are highly para-
sitized while others remain parasite free (Powell et
al. 1987). Individual B. impressa change hosts often
but typically move between existing aggregates
(Wilson et al. in press). Therefore, refugia from
parasitism may exist and recovery may be impor-
tant in assessing the overall impact of B. impressa
on oysters.
Stress, produced by laboratory conditions, fre-
quently accompanies laboratory experimentation
(e.g., Koenig et al. 1981; Powell et al. 1984; Kukal
and Kevan 1987). The effect of B. impressa on C.
virginica has been assessed primarily through
laboratory experimentation. Consequently, we ex-
amined the effect of B. impressa on C. virginica
under field conditions and assessed the ability of
oysters to recover normal growth rates and bio-
chemical composition once snail parasitism ceased.
MATERIALS AND METHODS
Field Study
Oysters and snails used in this study were col-
lected at Goose Island State Recreation Area near
Rockport, TX. The oysters were weighed on a
Mettler balance using the underwater m.ethod of
Andrews (1961). Fifteen oysters (precontrols) were
sacrificed to define the biochemical composition of
the oysters and Perkinsus marinus levels that ex-
isted naturally at the collection site. The rest of the
oysters were placed in semi-enclosed plexiglass
domes (see figure 2 in Kendall et al. 1984 for descrip-
tion) in a tidal creek near the Aransas Pass
Lighthouse on Lydia Ann Channel, near Port Aran-
sas, TX. The domes allowed water to circulate
over the oysters, while excluding such large
predators as oyster drills and crabs. Boonea im-
pressa, however, could readily move into or out of
the domes.
Two of the four domes (20 oysters per dome) con-
tained oysters exposed to B. impressa at a concen-
tration of 10 snails per oyster, a level of parasitism
commonly observed on reefs in the collection area
(White et al. 1984). The domes were positioned so
that the probability of snails moving from the ex-
posure domes with parasitized oysters to the con-
trol domes was minimized (Fig. 1). Snails on each
experimental oyster were counted twice weekly for
4 weeks. No B. impressa were ever found on con-
trol oysters. Boonea impressa did emigrate from the
exposure domes, however, so snails were added as
needed to maintain the 10:1, snaihoyster ratio. Con-
trol oysters were handled the same as experimentals
each week in an effort to minimize differential ef-
fects caused by handling stress (see Pickering et al.
1982; Andrews and Hewatt [1957] were unable to
find any effect of handling on P. marinus infection
in oysters).
At the end of 4 weeks, the oysters were reweighed
and half from each dome were sacrificed. The re-
maining oysters were replaced in the field for a 4-wk
recovery period without B. impressa. All B. im-
pressa were removed from the previously para-
sitized oysters by hand prior to replacement. The
domes were visited twice weekly and each oyster
handled as before. No B. impressa were found on
the oysters during the recovery period.
Laboratory Analysis
Perkinsus marinus infection was measured in
each oyster by incubating a small piece of mantle
tissue in thioglycollate medium by the method of Ray
(1966b). Intensity of infection was based on a semi-
quantitative 0 to 5 rating assigned during micro-
scopic inspection of the tissue after treatment with
Lugol's solution (Mackin 1962). Small pieces of
gonadal tissue were preserved in Bouin's fixative,
sectioned and stained in 0.5% toluidine blue (Preece
1972) for gonadal analysis. Mantle cavity volume
was determined by filling the shells with silicon
caulking. Oyster growth, as expressed by shell de-
position, was measured using the underwater weigh-
ing method. Condition index was derived by divid-
ing the total lyophilized dry weight by the mantle
cavity volume (Lawrence and Scott 1982).
The mantle and adductor muscle from each oyster
554
WILSON ET AL.: EFFECT OF SNAIL ON OYSTERS
TREATMENTS: 4 WEEK EXPERIMENTAL PERIOD
Unparasitized Controls
Parasitized
4 WEEK RECOVERY PERIOD
Recovery Control
Previously Parasitized
I 1
Figure 1.— Position of plexiglass domes during 4-wk experimental period and during 4-wk
recovery period.
were frozen immediately on dry ice and kept at
-40°C until lyophilized. Prior to biochemical anal-
ysis, lyophilized tissues were weighed and homog-
enized on ice. Total lipids were isolated from a por-
tion of the homogenate using the water:methanol:
chloroform method of Folch et al. (1957) with a
modified ratio of 0.8:2.1. This isolation produced a
two phase system and an insoluble pellet after cen-
trifugation in the cold. The pellet was used for pro-
tein and glycogen assays. The water phase was used
for amino acid analysis. The organic phase was
divided and dried under Ng gas at 40°C. One half
was used to determine lipid phosphate concentra-
tion by the spectrophotometric method of White et
al. (1979). The other half was dissolved in chloro-
form to which nonadecanoic acid was added as an
internal standard. The redissolved lipids were frac-
tionated on a silicic acid (Unisil,^ 100-200 mesh)
column. Neutral lipids were recovered by eluting
with 10 times the column volume of chloroform. The
chloroform was removed under N2 at 40°C ((rehron
and White 1982). The resulting chloroform fraction,
containing fatty acid methyl esters, was analyzed
by gas chromatography using a capillary, nonpolar,
methyl silicone high performance column and flame
ionization detector.
The amino acids were analyzed on a Dionex 3000
amino acid analyzer. Because residual chloroform
interfered with the analysis, a second extraction was
performed on the sample in the cold using a chloro-
form:water ratio of 1:3; a-amino-n-butyric acid (2.5
fimole • mL~^ sample) was used as an internal
standard. The amino acids accounting for the bulk
of the free amino acid (FAA) pool, taurine, hypo-
'Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
taurine, aspartic acid, serine, threonine, glutamine,
glutamate, glycine, and alanine, were separated us-
ing a lithium citrate buffer and measured using
o-phthalaldehyde as the detecting compound.
Total soluble protein was estimated by Peterson's
(1977) modification of the Lowry method. Protein
was precipitated using a final concentration of 10%
trichloroacetic acid in the cold. After centrifugation
at 4°C the resultant pellets were resuspended
in 50:50, 10% SDS:0.8N NaOH for 30 minutes.
Replicate samples were centrifuged and analyzed
spectrophotometrically for protein concentration.
Bovine serum albumin (Sigma) was used as the
standard.
Glycogen was degraded to glucose enzymatically
with amyloglucosidase (Carr and Neff 1984). After
preincubation with amyloglucosidase, glucose was
measured using the glucose oxidase-peroxidase pro-
cedure (Sigma glucose kit) (Roehrig and Allred
1974). Rat liver glycogen (Sigma) was used as a
standard.
Statistical Analysis
To assess the effect of Boonea impressa on the
oysters during the 4-wk exposure period, the level
of snail parasitism for each oyster had to be deter-
mined. Because the number of snails on each oyster
was counted only every third day, the total number
of snails that parasitized each oyster was estimated
by assuming that the same number of snails were
present on days between counts as found on the
previous visit. The total number of snail-days per
oyster, the sum of the number of snails present on
each day during the 4-wk exposure, will be referred
to as the snail scale. For example, during the first
week of the exposure period, one oyster had 15
555
FISHERY BULLETIN: VOL. 86, NO. 3
snails on Monday, 10 on Wednesday, and 4 on Satur-
day; therefore Tuesday was assigned 15 snails,
Thursday and Friday 10 snails, and the total for the
week was 64 snail-days.
Except where noted in the text, results were
analyzed by multiple analysis of covariance
(MANCOVA) using ranked data followed by Dun-
can's multiple range tests (a = 0.05) to locate sig-
nificant differences within the MANCOVA. Mantle
cavity volume was used as the covariate for com-
paring levels of P. marinus infection intensity, dry
weight, and shell weight gain between treatment
groups (e.g., parasitized and control oysters).
Similar tests comparing biochemical parameters
between treatment groups included the intensity of
P. marinus infection and tissue dry weight as co-
variates. When the requirement of parallelism was
not met, the MANCOVA was modified as described
by Smith and Coull (1987). Perkinsus marinus in-
fection intensity was used as both a dependent and
independent variable because, although B. impressa
can influence the intensity of P. marinus infection
(White et al. 1987), P. marinus itself can affect the
biochemical composition of oysters (Soniat and
Koenig 1982; White et al. in press). Only parasitized
oysters (or those recovering from parasitism) were
used in analyses of the effect of snail scale (intensity
of parasitism).
The FAA pool was analyzed three ways: 1) the
entire pool, 2) the pool minus taurine and hypo-
taurine because these are the only amino acids not
found in protein, and 3) after exclusion of the major
components (taurine, hypotaurine, glycine, and ala-
nine) so that changes in the lesser constituents of
the pool could be examined.
All analyses for the 4-wk exposure period used
MANCOVA analyses with nested variables, which
took into account which of the 4 domes the oysters
were in during the exposure period. Overall, signif-
icant differences in biochemical composition (i.e.,
amino acid content, glycogen, protein, etc.) between
the 2 nonparasitized domes or between the 2 snail
parasitized domes did not occur more frequently
than expected by chance (Binomial Test, a = 0.05).
Therefore, the two equivalently treated domes of
each treatment group (Fig. 1) were lumped together
for comparison with precontrol and recovery
oysters. Nevertheless, the experimental design
represents a case of pseudoreplication; hence the
caveats of Hurlbert (1984) should be considered
when reviewing the statistical analysis.
RESULTS
Growth, Perkinsus marinus Intensity,
and Reproduction
Boonea impressa affected oyster growth rate, the
intensity of infection by P. marinus, and reproduc-
tive state. The mean initial shell weights of the
precontrol, control, and parasitized groups were not
significantly different from each other. By the end
of the 4-wk exposure period, both parasitized and
unparasitized oysters had gained weight, but oysters
parasitized by B. impressa gained significantly
less weight than unparasitized oysters (Table 1).
Oysters with more snails typically gained less weight
than oysters with fewer snails. The relationship
between the intensity of B. impressa parasitism
(snail scale) and weight gain among parasitized
oysters was significant (Spearman's rank, P = 0.03;
weight gain normalized to initial weight, P = 0.001,
Fig. 2).
Mantle cavity volume did not vary among any of
the treatment or recovery groups. Condition index
was not significantly different between treatment
groups or between recovery groups. Condition in-
dex was significantly higher in the two recovery
Table 1 .—Mean and standard deviations for initial shell weight (g) and average shell weight gain per oyster during
the 4-wk exposure period and 4-wk recovery period. Significance levels (sig.) from Duncan's multiple range test
(o = 0.05). A test was restricted to a single column. Groups having the same letter within a column are not significantly
different.
Initial weight
Weight gained
during 4-wk exposure
Mean ± SD Sig.
Weight gained
during 4-wk recovery
Group
Mean ± SD Sig.
Mean ± SD Sig.
Precontrol (n = 15)
Treatment
Control (n = 40)
Parasitized (n = 40)
Recovery
Control (n = 6)
Previously parasitized (n =
15)
22.82 ± 12.16 A
17.16 ± 8.48 A
16.84 ± 7.47 A
3.49 ±1.21 A
2.54 ±1.31 B
2.16 ± 0.73 B
2.89 ±1.05 A
556
WILSON ET AL.: EFFECT OF SNAIL ON OYSTERS
Parasitized
U.4J)
"
•
0.40
-
0.35
-
O)
0.30
•
S
•
c
CO
O
0.25
•
0.20
•
•
o
0.15
•
•
•
• •
•
0.10
• •
• • •
0.05
• •
I 1 I
1 '
Previously Parasitized
0.45
•
•
0.40
0.35
•
O)
0.30
•
•
2
c
CO
(3
*-*
SI
0.25
•
•
(f)
o
0.20
0.15
•
0.10
•
•
• •
0.05
•
1 1 1 1 1
100 200 300
Snail Scale
400
500
100 200 300
Snail Scale
400
500
Figure 2.— Left: Weight gain (g) per g initial weight of individual snail-parasitized oysters after the 4-wk experimental period as a func-
tion of the level of Boonea impressa parasitism, or snail scale. Right: Weight gain (g) per g initial weight of individual oysters after
the 4-wk recovery period as a function of the level of B. impressa (snail scale) during the 4-wk treatment period.
groups than in the treatment and precontrol groups
(Table 2).
During the recovery period, when all oysters were
parasite-free, the previously parasitized oysters
(those with snails during the treatment period)
gained significantly more weight than the recovery
controls (Table 1). Oysters which were previously
parasitized by fewer B. impressa during the treat-
ment period gained more weight during the
recovery period than those that were previously
parasitized by more snails (Spearman's Rank corre-
lation, P = 0.07, Fig. 2), but the relationship was
considerably weaker than during the treatment
period.
The intensity of infection by P. marinus increased
throughout the experiment in parasitized oysters so
that, after recovery, previously parasitized oysters
had higher prevalences and intensities of infection
than they did after 4 weeks of parasitism; these
values in turn were higher than the precontrol
values at the experiment's inception (Table 3).
Exactly the opposite trend occurred in control
oysters. Recovery controls had the lowest values of
prevalence and intensity. Consequently, after the
recovery period, previously parasitized oysters had
significantly higher intensities of infection than
recovery controls. The proportion of infected in-
dividuals (33% of the controls, 93% of the previously
parasitized oysters) was significantly higher as well
(x^, P < 0.05). Within the parasitized oysters, the
intensity of parasitism (snail scale) did not correlate
with the increase in intensity of P. marinus infec-
tion during either the treatment or recovery period
(Fig. 3, Spearman's Rank, P > 0.05; the interaction
557
FISHERY BULLETIN: VOL. 86, NO. 3
Table 2.— Means and standard deviations for volume (mL), condition index (g mL"')
and dry weight (g). Significance levels (sig.) from Duncan's multiple range test (o = 0.05).
A test was restricted to a single column. Groups having the same letter within a column
are not significantly different, n, see Table 1 .
Volume
Condition index
Dry weight
Group
Mean ± SD
Sig.
Mean ± SD
Sig.
Mean ± SD Sig.
Precontrol
10.4 ± 4.1
A
3.4 ± 0.8
C
0.35 ± 0.14 AS
Treatment
Control
11.2 ± 3.6
A
5.4 ± 1.0
B
0.34 ± 0.11 AB
Parasitized
9.2 ± 1.7
A
5.5 ± 1.0
B
0.28 ± 0.12 B
Recovery
Control
9.5 ± 4.6
A
7.1 ± 1.5
A
0.32 ± 0.14 B
Previously
parasitized
10.8 ± 3.8
A
7.7 ± 1.5
A
0.51 ± 0.24 A
Table 3.— Mean and standard deviations for Perkinsus marinus
infection intensity among groups. Significance levels from Dun-
can's multiple range test (o = 0.05) apply to the infection intensity
data. Groups with the same letter are not significantly different.
Percent incidence is the number of oysters infected divided by the
number of oysters in the sample. Infection intensity was calculated
using the 5-point scale of Mackin (1962). n, see Table 1.
Table 4. — Results of histological examination of gonadal samples
for each group. Significance levels (sig.) from Duncan's multiple
range test (a = 0.05) are for differences In the number of eggs
present per female in a histological section. Percent of females
spawning is the number of females with eggs observed in the
gonoducts divided by the number of females in the sample.
Perkinsus marinus
Percent
of
females
Infection Intensity
Signifi-
Preva-
Sex
Eggs present
Group
Mean ± SD
cance
lence
Group
Female
Male
Mean ± SD
Sig.
spawning
Precontrol
1.47 ± 1.45
AB
66.6
Precontrol
7
7
31.1 ± 4.4
A
86.0
Treatment
Treatment
Control
1.10 ± 1.06
AB
65.0
Control
6
4
36.6 ±11.0
A
100.0
Parasitized
1.63 ± 1.32
AB
79.0
Parasitized
4
6
21.6 ± 3.9
B
100.0
Recovery
Recovery
Control
0.83 ± 1.60
B
33.3
Control
5
1
37.6 ± 6.0
A
80.0
Previously
Previously
parasitized
2.21 ± 1.38
A
92.9
parasitized
11
4
42.9 ± 9.8
A
81.0
term in the MANCOVA was also nonsignificant).
The number of eggs per microscopic field esti-
mated from histological sections, was significantly
lower in parasitized oysters after the 4-wk treatment
period (Table 4). No differences between controls
and previously parasitized oysters were present
after the 4-wk recovery period. The proportion of
oysters ready to spawn (female oysters which had
eggs visible in the gonoducts) was not significantly
different in parasitized and unparasitized oysters
during treatment or recovery (x^, a = 0.05).
Biochemical Composition
Mean levels of the various biochemical compo-
nents measured in adductor muscle and mantle
tissue are given in Tables 5 and 6. A comparison of
precontrol, treatment control, and recovery control
oysters documents the changes in biochemical com-
position produced by handling stress and natural
environmental changes that occurred during the
experiment. Glycogen increased significantly in the
mantle tissue during the 4-wk treatment period,
then decreased during the recovery period (Table
7). Glycogen in the adductor muscle also decreased
during the recovery period (Table 7). Hypotaurine
increased in both tissues during the treatment
period (Table 8). Most other FAA and the total pool
dropped in concentration in the mantle tissue in the
first 4 weeks, but then stabilized. In contrast, in the
adductor muscle, significant increases in glycine and
hypotaurine during the treatment period were off-
set by a significant decrease in alanine, so the total
pool changed little. Again, the FAA pool stabilized
during the first 4 weeks. Hence, there was little
difference in treatment and recovery controls in any
measured biochemical component.
Snail parasitized and control oysters did not differ
significantly in the concentration of any of the
biochemical components in either the mantle or
adductor muscle (Tables 7, 8). Only one biochemical
parameter differed between the recovery control
558
WILSON ET AL.: EFFECT OF SNAIL ON OYSTERS
Parasitized
4.5 r •
4.0
3.5
B 3.0
_c
c
o
CO
3
§ 2.0
CD
CO
^ 1.5
0)
1.0
0.5
• • •
100
-••-
200 300
Snail Scale
400
500
Previously Parasitized
4.5
4.0
3.5
CO
r
0
3.0
■*-*
c
c
o
.^^
o
? h
0)
c
CO
3
c
?0
l~
CO
E
to
3
eo
C
1.5
*
>-
0)
Q.
••
1.0
0.5
• •
100 200 300
Snail Scale
400
500
Figure 3.— Left: Intensity oiPerkinsus marinus infection in snail -parasitized oysters as a function of the level of snail parasitism after
the 4-wk treatment period. Right: Intensity of P. marinus infection in previously snail-parasitized oysters as a function of the level
of snail parasitism after the 4-wk recovery period.
and previously parasitized oysters. Recovery con-
trols had significantly less glycogen in the mantle
tissue (Table 7). No biochemical component was
significantly correlated with the intensity of snail
parasitism (snail scale) in parasitized oysters dur-
ing the 4-wk treatment period except taurine in the
adductor muscle (Table 9). A similar comparison,
using the previously parasitized oysters after the
4-wk recovery yielded only two significant correla-
tions with snail scale; soluble protein and hypo-
taurine content. In addition, lipid phosphate in the
mantle tissue and fatty acid content in the adductor
muscle were significantly correlated with the inten-
sity of Perkinsus marinus infection after the
recovery period (all recovery oysters, controls and
previously parasitized, were included in the analysis.
Table 9).
DISCUSSION
Oysters averaged as few as about 3 and as many
as 14 snails per day during the 4-wk treatment
period; however, only 11% of the oysters had 10 or
more snails per day and 76% had fewer than 9 per
day. This range of parasitism is typical for many
reefs in the experimental area (Copano Bay- Aransas
Bay, TX-White et al. 1984; Powell et al. 1987) and
corresponds to the lower parasite levels used by
White et al. (1988, in press) in laboratory studies
on these animals.
The effects of snail parasitism were minor on a
biochemical level, but substantial on an organismal
level. Growth, reproductive capacity, and health, as
measured by Perkinsus marinus infection, were
significantly affected. Other odostomians also re-
559
FISHERY BULLETIN: VOL. 86, NO. 3
Table 5.— Means and standard deviations for all biochemical components in mantle tissue.
FAA = total of the 9 free amino acids measured; FAA- = FAA less taurine and hypotaurine; FAA-
= FAA less taurine, hypotaurine, glycine, and alanine; amino acids in ^moles mL" \ other tissue
components in mg • g dry wt'\
Treatment
Recovery
Previously
Precontrol
Control
Parasitized
Control
parasitized
(n
=
15)
{n
=
10)
(n
=
10)
(n
=
6)
(n
= 15)
X
±
SD
X
±
SD
X
±
SD
X
±
SD
X
± SD
FAA
351
±
94
240
±
39
280
±
38
294
±
64
270
± 64
FAA-
220
±
67
116
±
14
121
±
15
134
±
50
117
± 43
FAA-
57
±
18
42
±
6
46
±
7
49
±
26
42
± 13
Taurine
102
±
33
95
±
27
119
+
24
108
±
15
99
± 21
Hypotaurine
27
±
10
39
±
10
50
±
23
49
±
10
54
± 20
Aspartic acid
22
±
5
15
±
2
19
±
3
22
±
13
18
± 5
Serine
8
±
5
5
±
2
6
±
2
5
±
3
5
± 3
Threonine
3
±
1
1
±
0.2
1
±
0.7
1
±
0.4
1
± 0.5
Glutamine
8
±
5
2
±
1
3
±
0.8
4
±
3
5
± 3
Glutamic acid
14
±
5
16
±
3
18
±
5
14
±
5
12
± 4
Glycine
71
±
23
33
±
5
38
±
10
44
±
31
38
± 17
Alanine
87
±
33
36
±
8
38
±
10
31
±
10
34
± 16
Protein
62
±
45
28
±
39
60
±
57
45
±
40
23
± 16
Glycogen
28
±
22
111
±
59
66
±
28
37
±
13
88
± 48
Lipid phosphate
(as PO4-3)
0.62
±
0.41
0.48
±
0.27
0.82
±
0.39
0.5
±
0.2
0.87
± 0.54
Fatty acids
20.7
±
33.1
33.2
±
49.1
48.7
±
58.1
9.4
±
10.5
6.3
± 3.2
Table 6.— Means and standard deviations for biochemical components of adductor muscle tissue.
See Table 5 for additional information.
Treatment
Recovery
Previously
Precontrol
Control
Parasitized
Control
parasitized
{n
=
11)
{n
=
10)
(n
=
10)
("
=
6)
{n
= 15)
X
±
SD
X
±
SD
X
±
SD
X
±
SD
X
± SD
FAA
328
±
119
306
±
56
289
±
47
277
±
110
347
± 81
FAA-
272
±
121
227
±
54
207
±
35
210
±
95
266
± 78
FAA-
57
±
16
57
±
8
44
±
5
75
±
25
84
± 24
Taurine
52
±
17
53
±
25
66
±
25
42
±
16
47
± 13
Hypotaurine
8
±
4
22
±
7
23
±
3
24
±
10
33
± 6
Aspartic acid
17
±
7
18
±
4
14
±
4
18
±
6
21
± 8
Serine
5
±
4
7
±
3
5
±
2
27
±
21
28
± 21
Threonine
3
±
1
2
±
0.9
1
±
0.5
2
±
1
3
± 1
Glutamine
6
±
2
4
±
2
2
±
1
6
±
4
10
± 8
Glutamic acid
20
±
6
18
±
4
21
±
6
21
±
4
20
± 5
Glycine
54
±
15
124
±
37
119
±
42
85
±
54
121
± 44
Alanine
142
±
63
49
±
19
39
±
10
49
±
30
60
± 26
Protein
41
±
13
49
±
40
45
±
56
45
±
35
36
± 22
Glycogen
6
±
3
7
±
3
6.8
±
0.3
4
±
2
4
± 2
Lipid phosphate
(as PO;3)
0.1
±
0
0.2
±
0
0.34
±
0.3
0.25
±
0.28
0.2
± 0.1
Fatty acids
15.0
±
10.0
17.5
±
16.7
15.3
±
24.1
6.9
±
11.5
8.6
± 16.9
duce host growth rate (Nishino et al. 1983). Oysters
parasitized by Boonea impressa gained significant-
ly less weight than nonparasitized oysters. Para-
sitized female oysters had significantly fewer eggs
than controls (the effect on males was not quan-
tified).
Reduced growth and impaired reproduction in the
host are commonly associated with marine parasites
(e.g., Menzel and Hopkins 1955; Cheng et al. 1983;
Hawkes et al. 1986). Starvation produces a similar
phenomenon (Fair and Sick 1982; Pipe 1985; Wright
and Hetzel 1985; Devi et al. 1985). Several lines of
evidence suggest that a reduction in net productiv-
ity, but not a negative energy balance, produced the
results observed here. White et al. (1988) developed
an energy flow model for oysters and snails. Using
that model, oysters of 4 to 7 cm long, the size we
used, would not incur a negative energy balance un-
560
WILSON ET AL.: EFFECT OF SNAIL ON OYSTERS
Table 7.— Results of Duncan's multiple range test for biochemical components of the mantle and adductor muscle. Each test considered
data for one biochemical component (one vertical group of 5 means) for each tissue. Different letters within a vertical group indicate
significant differences at a = 0.05 within that group.
Lipid
Fatty
Lipid
Fatty
Protein
Glycogen
phosphate
acids
Protein
Glycogen
phosphate
acids
Mantle
Adductor muscle
Precontrol
A
B
A
A
Precontrol
A
AB
A
A
Treatment
Treatment
Control
B
A
A
A
Control
A
A
A
A
parasitized
AB
A
A
A
Parasitized
A
AB
A
A
Recovery
Recovery
Control
AB
B
A
A
Control
A
B
A
A
Previously
Previously
parasitized
B
A
A
A
parasitized
A
B
A
A
Table 8. — Results of Duncan's multiple range test for free amino acids in the mantle tissue and
adductor muscle. Each test considered data for one biochemical component (one vertical group
of 5 means) for each tissue. Different letters within a vertical group indicate significant differences
at a = 0.05 within that group. Tau = taurine; Hyp = hypotaurine; Asp = aspartic acid; Ser
= serine; Glu = glutamic acid; Gin = glutamine; Gly = glycine; Ala = alanine; FAA = total
for the amino acids measured; FAA- = FAA less taurine and hypotaurine; FAA- = FAA less
taurine, hypotaurine, glycine, and alanine.
Tau
Hyp
Asp
Ser
Thr
Gin
Glu
Gly
Ala
FAA
FAA-
FAA-
Mantle
Precontrol
A
B
A
A
A
A
AB
A
A
A
A
A
Treatment
Control
A
A
B
A
B
B
AB
B
B
B
B
A
Parasitized
A
A
AB
A
AB
B
A
B
B
AB
B
A
Recovery
Control
A
A
AB
A
B
AB
B
B
B
AB
B
A
Previously
parasitized
A
A
AB
A
B
AB
B
B
B
B
B
A
Adductor muscle
Precontrol
A
C
AB
B
A
A
A
C
A
A
A
BC
Treatment
Control
A
B
AB
B
AB
A
A
AB
B
A
A
CD
Parasitized
A
B
B
B
B
A
A
AB
B
A
A
D
Recovery
Control
A
AB
AB
A
AB
A
A
BC
B
A
A
AB
Previously
parasitized
A
A
A
A
A
A
A
AB
B
A
A
A
Table 9.— Significant P-values obtained by MANCOVA from comparison of snail scale and Perkinsus mahnus infection intensity. Tests
using snail scale considered only the parasitized oysters after the treatment period and the previously parasitized oysters after the recovery
period. Tests using P. mahnus considered all oysters, controls and snail parasitized. Abbreviations same as Table 8. M = mantle
tissue; A = adductor muscle; — = not significant in either tissue at o = 0.05.
Lipid
Gly- phosphate
cogen (as PO4 ^)
Fatty
acids
Tau Hyp Asp Thr Ser Glu Gly Ala FAA FAA- FAA- Protein
Treatment
Snail scale
Perkinsus mahnus
Recovery
Snail scale
Perkinsus mahnus
— 0.02(A) — — — — ___— — —
— — _ _ _ 0.004(A) — — ——— —
— 0.03(M) 0.002(A)
— 0.04(A) — — — — — ———— 0.004(M)
561
FISHERY BULLETIN: VOL. 86, NO. 3
til fed upon by at least 25 average-sized (1.75 mm
maximum width) snails. Seven snails, a typical value
in our experiments, would reduce net productivity
by only 5 to 30%, on a daily basis. In addition, con-
dition index and mantle glycogen levels increased
during the treatment period in both control and
parasitized oysters and the effect of snail parasitism
on all biochemical components was small (oysters
can regulate some biochemical components even
during starvation, Swift and Ahmed 1983; but see
Riley 1976).
Consequently, both reduced growth and impaired
reproductive capacity can be attributed to a reduc-
tion in assimilated carbon available to the host, as
a result of either a reduction in filtration rate result-
ing in less energy being assimilated (Ward and
Langdon 1986) or the direct removal of assimilated
carbon by the snail. Neither effect was permanent.
Growth rate resumed and reproductive state re-
turned to control levels during recovery. In both
cases compensatory adjustments occurred during
the recovery period so that previously parasitized
oysters gained weight and increased egg number
faster than the controls. Loosanoff and Nomejko
(1955) and Eagle and Chapman (1953) also noted
compensatory shell growth in oysters.
Perkinses marirms infection is an important cause
of mortality in oysters (Mackin and Sparks 1962;
Mackin 1962; Hofstetter 1977). Boonea impressa can
transmit P. marinus from one oyster to another and
can also increase the intensity of infection (White
et al. 1987). Continued deterioration after a stress
is removed occurs frequently in "recovery" experi-
ments (Kendall et al. 1984; Powell et al. 1984)
demonstrating the necessity of examining recovery
capacity in acute (vs. chronic) stresses. Growth rate
typically recovers more rapidly than most biochem-
ical parameters. In snail-parasitized oysters, both
prevalence and intensity of P. marinus infection in-
creased during the recovery period. Hence, in con-
trast to growth rate, no recovery from P. marinus
actually occurred. By the end of the 8-wk period,
infection intensity had increased by about 1 unit on
Mackin's (1962) 5 point scale. Growth still occurred
and reproductive capacity returned to control levels,
however, during this period. These results contra-
dict those of Menzel and Hopkins (1955) who showed
that P. marinus retarded growth in proportion to
the intensity of infection and Mackin (1953) who
observed decreased fecundity in heavily infected
oysters. Haven (1962) obtained results analogous to
ours. Possibly, the oysters in our study were not
infected heavily enough to retard growth and
reproduction. Mean infection intensity in previous-
ly parasitized oysters after the recovery period was
only 2.2, a light to moderate infection.
Changes observed at the biochemical level among
the previously parasitized oysters after the recovery
period, particularly in fatty acid and lipid phosphate
content, were related to increased infection inten-
sity of P. marinus (Table 9). Stein and Mackin
(1955), Mackin (1962), and White et al. (in press)
also noted changing lipid levels related to infection
intensity. Lipid phosphate is predominantly a struc-
tural component whereas fatty acids, usually as tri-
glycerides, are storage materials in many marine
invertebrates (Gabbott 1976; Trider and Castell
1980; Gehron and White 1982). The increased lipid
phosphate content in mantle tissue, however, prob-
ably indicates an increase in structural material,
noted histologically by Stein and Mackin (1955,
1957) to occur in conjunction with P. marinus
infections.
Glycogen is the primary storage material in most
bivalves (Beninger and Lucas 1984). Parasitism fre-
quently affects carbohydrate metabolism (Cheng
1963; Mohamed and Ishak 1981; Thompson and
Binder 1984; Thompson 1986). White et al. (in press)
suggested that P. marinus alters oyster metabolism
favoring gluconeogenesis. Our results support this
hypothesis. Glycogen levels dropped only in recovery
control oysters in which P. marinus infection inten-
sity also declined. Changing fatty acid content might
be similarly explained. In contrast. Stein and Mackin
(1957) noted decreased glycogen levels in heavily in-
fected oysters (3 to 5 on Mackin's scale). Few of our
oysters were this heavily infected, however. An
alternative explanation, that slower reproductive
development in parasitized oysters was responsible
for variation in glycogen and fatty acid content, can-
not be completely excluded because P. marinus in-
fection intensity did not correlate with glycogen
levels in recovery oysters. Parasitized oysters had
fewer eggs than control oysters, however. Addi-
tionally, neither the number of eggs present nor the
number of oysters spawning differed significantly
between control and previously parasitized oysters
during the recovery period.
Results of previous workers suggest that signif-
icant decreases in storage compounds, whether
caused by B. impressa or P. marinus, only occur
in heavily infected animals (e.g.. Stein and Mackin
1957; White et al. in press). This, too, is true for the
amino acid pool where significantly decreased levels
are usually associated with more severely stressed
animals (e.g., Powell et al. 1982, 1984). The few
significant effects on amino acids in this study, like
all the other biochemical components measured.
562
WILSON ET AL.: EFFECT OF SNAIL ON OYSTERS
were produced by increases in concentration. Soniat
and Koenig (1982) observed significant changes in
the free ammo acid pool due to P. marinus, par-
ticularly in taurine concentration. We noted changes
in taurine during the treatment period and hypo-
taurine during recovery in the adductor muscle but
these were related to snail parasitism.
Biochemical components, though rarely signifi-
cantly affected, were affected not just in the mantle
tissue but also in the adductor muscle. One possi-
bility, that the snail's effect is localized at the point
of feeding, is not supported by the data. Snail para-
sitism produces systemic effects.
CONCLUSIONS
Complementary results of White et al. (1984,
1988, in press) and this study permit a general
description of the impact of snail parasitism on
oysters at normal field levels. Both growth rate and
reproductive development slow significantly, but
recover rapidly once the snails are removed. Hence,
the temporal stability of snail patches must deter-
mine the cumulative effect on field populations of
oysters. The prevalence and intensity of infection
by Perkinsus marinus is significantly increased, but
recovery does not occur. That is, Boonea impressa
probably facilitates and encourages the normal
spread and intensification of P. marinus from which
oysters, if they recover, only do so the following
winter when low temperatures typically reduce in-
fection levels (Hewatt and Andrews 1956; Burrell
et al. 1984; Soniat 1985). This effect, then, is long
term. Most changes in biochemical components were
due to infection by P. marinus. Snail feeding
reduces net productivity but, at normal field levels,
starvation is an unlikely result. Increased infection
by P. marinus typically raises glycogen and lipid
levels, at least in light to moderate infections. Of
the free amino acids, taurine and hypotaurine have
been shown to be affected by P. marinus and B. im-
pressa. Little change in the remaining FAA or the
total pool has been observed singly or in concert.
Feng et al. (1970) noted increased taurine levels in
oysters parasitized by Bucephalus sp. and Minchinia
nelsoni. Hence, an increase in taurine and hypo-
taurine levels apparently is a general response to
parasitism in oysters.
Yuill (1987) emphasized the importance of subtle
effects produced by parasites on host populations.
Perkinsus marinus is an important source of mor-
tality in oyster populations (Mackin and Sparks
1962; Mackin 1962; Hofstetter 1977). Data suggest
that one of the most important aspects of parasitism
by B. impressa is to encourage this second parasitic
organism. To this extent, over the year, B. impressa
at normal field densities could be responsible for a
substantial amount of mortality in oyster popula-
tions.
ACKNOWLEDGMENTS
We would like to thank M. White, D. Davies, and
L. Priest for laboratory and statistical assistance.
T. J. McDonald graciously provided his expertise in
GC analysis, and M. C. Kennicutt and A. Vastano
ran the fatty acid analyses. Suggestions by J. Par-
rack, T. Bright, and an anonymous reviewer im-
proved the manuscript. We thank R. Covington for
typing the manuscript and tables. R. Pratt, care-
taker of the Aransas Pass Lighthouse, provided
space for the field portion of the study. This research
was supported by part of an institutional grant
NA85AA-D-FG128 to Texas A&M University by the
National Sea Grant Program, National Oceanic and
Atmospheric Administration, U.S. Department of
Commerce to E. N. Powell and S. M. Ray. We ap-
preciate this support.
LITERATURE CITED
Allen, J. F.
1958. Feeding habits of two species of Odostomia. Nautilus
72:11-15.
Andrews, J. D.
1961. Measurement of shell growth in oysters by weighing
in water. Proc. Natl. Shellfish Assoc. 52:1-12.
1965. Infection experiments in nature with Dermocystidium
marinum in Chesapeake Bay. Chesapeake Sci. 6:60-67.
Andrews, J. D., and W. G. Hewatt.
1957. Oyster mortality studies in Virginia. II. The fungus
disease caused by Dermocystidium Tnarinum in oysters of
Chesapeake Bay. Ecol. Monogr. 27:1-25.
Beninger, p. G., and a. Lucas.
1984. Seasonal variations in condition, reproductive activity,
and gross biochemical composition of two species of adult
clam reared in a common habitat: Tapes decussatus L.
(Jeffreys) and Tapes phillippinarum (Adams & Reeve).
J. Exp. Mar. Biol. Ecol. 79:19-37.
Brockelman, C. R.
1978. Effects of parasitism and stress on hemolymph protein
of the African giant snail, Achatinafulica. Z. Parasitenkd.
57:137-144.
Brown, C. R., and M. B. Brown.
1986. Ectoparasitism as a cost of coloniality in cliff swallows
(Hirundo pyrrhonota). Ecology 67:1206-1218.
Burrell, V. G., M. Y. Bobo, and J. J. Manzl
1984. A comparison of seasonal incidence and intensity of
Perkinsus marinus between subtidal and intertidal oyster
populations in South Carolina. J. World Mariculture Soc.
15:301-309.
Carr, R. S., and J. M. Neff.
1984. Quantitative semi-automated enzymatic assay for tissue
563
FISHERY BULLETIN: VOL. 86, NO. 3
glycogen. Comp. Biochem. Physiol. B Comp. Biochem.
77:447-449.
Cheng, T. C.
1963. The effects of Eckinoparyphium larvae on the struc-
ture of and glycogen deposition in the hepatopancreas of
Helisoma trivolvis and glyconeogenesis in the parasite lar-
vae. Malacologia 1:291-301.
Cheng, T. C, J. T. Sullivan, K. H. Rowland, T. F. Jones,
AND H. J. MORAN.
1983. Studies on parasitic castration: soft tissue and shell
weights of Ilyanassa obsoleta (MoUusca) parasitized by larval
trematodes. J. Invertebr. Pathol. 42:143-150.
Devi, V. U., Y. P. Prabhakara Rao, and D. G. V. Prasada Rao.
1985. Body component indices of an intertidal gastropod
Morula granulata (Daclos) subjected to starvation. Indian
J. Mar. Sci. 14:44-45.
Eagle, J. B., and C. R. Chapman.
1953. Oyster condition affected by attached mussels. Natl.
Shellfish Assoc. Conv. Addendum 1951, p 70-78.
Fair, P. H., and L. V. Sick.
1982. Serum amino acid concentrations during starvation in
the prawn, Macrobrachium rosenbergii, as an indicator of
metabolic requirements. Comp. Biochem. Physiol. B Comp.
Biochem. 73:195-200.
Feng, S. Y., E. A. Khairallah, and W. J. Canzonier.
1970. Hemolymph-free amino acids and related nitrogenous
compounds of Crassostrea virginica infected with Bucepha-
lus sp. and Minchinia nelsoni. Comp. Biochem. Physiol.
34:547-556.
Folch, J., M. Lees, and G. H. S. Stanley.
1957. A simple method for the isolation and purification of
total lipides from animal tissues. J. Biol. Chem. 226:497-
509.
Ford, S. E.
1986. Comparison of hemolymph proteins from resistant and
susceptible oysters, Crassostrea virginica, exposed to the
parasite Haplosporidium nelsoni (MSX). J. Invertebr.
Pathol. 47:283-294.
Fretter, V.
1951. Turbonilla elegantissima (Montagu), a parasitic opis-
thobranch. J. Mar. Biol. Assoc. U.K. 30:37-47.
Fretter, V., and A. Graham.
1949. The structure and mode of life of the Pyramidellidae,
parasitic opisthobranchs. J. Mar. Biol. Assoc. U.K. 28:
493-532.
Gabbott, P. A.
1976. Energy metabolism. In B. L. Bayne (editor). Marine
mussels: their ecology and physiology, p. 293-356. Cam-
bridge Univ. Press, N.Y.
Gehron, M. J., AND D. C. White.
1982. Quantitative determination of the nutritional status of
detrital microbiota and the grazing fauna by triglyceride
glycerol analysis. J. Exp. Mar. Biol. Ecol. 64:145-158.
Haux, C, M. Sjobeck, AND A. Larsson.
1985. Physiological stress responses in a wild fish population
of perch (Perca fluviatilis) after capture and during subse-
quent recovery. Mar. Environ. Res. 15:77-95.
Haven, D. S.
1962. Seasonal cycle of condition index of oysters in the York
and Rappahannock Rivers. Proc. Natl. Shellfish. Assoc.
51:42-66.
Hawkes, C. R., T. R. Meyers, and T. C. Shirley.
1986. Length-weight relationships of blue, Paralithodes
platypus, and golden, Lithodes aequispina, king crabs para-
sitized by the rhizocephalan, Briurosaccus callosus Boschma.
Fish. Bull., U.S. 84:327-332.
Hewatt, W. G., and J. D. Andrews.
1956. Temperature control experiments on the fungus
disease, Dermocystidium marinum, of oysters 1. Proc.
Natl. Shellfish. Assoc. 46:129-133.
HOFSTETTER, R. P.
1977. Trends in population levels of the American oyster,
Crassostrea virginica Gmelin on public reefs in Galveston
Bay, Texas. Tex. Parks Wildl. Dep. Tech. Ser. No. 24, 90 p.
Hopkins, S. H.
1956. Odostomia impressa parasitizing southern oysters.
Science (Wash. D.C.) 124:628-629.
Hurlbert, S. H.
1984. Pseudoreplication and the design of ecological field
experiments. Ecol. Monogr. 54:187-211.
Kabat, a. R.
1986. Effects of trematode parasitism on reproductive out-
put of the bivalve Transennella tantilla. Can. J. Zool. 64:
267-270.
Kendall, Jr., J. J., E. N. Powell, S. J. Connor, T. J. Bright,
AND C. E. Zastrow.
1984. The importance of monitoring metabolic recovery in
the coral Acropora cervicomis after short-term exposure to
drilling muds: calcification rate and protein concentration.
Coral Reefs. 2:215-225.
Koenig, M. L., E. N. Powell, and M. R. Kasschau.
1981. The effects of salinity change on the free amino acid
pools of two nereid polychaetes, Neanthes succinea and
Leonereis culvert. Comp. Biochem. Physiol. A Comp.
Physiol. 70:631-637.
KUKAL, 0., AND P. G. KEVAN.
1987. The influence of parasitism on the life history of a high
arctic insect, Gynaephora groenlandica (Wocke) (Lepidop-
tera:Lymontriidae). Can. J. Zool. 65:156-163.
Lawrence, D. R., and G. I. Scott.
1982. The determination and use of condition index of
oysters. Estuaries 5:23-27.
LOOSANOFF, V. L., AND C. A. NOMEJKO.
1955. Growth of oysters with damaged shell-edges. Biol.
Bull. (Woods Hole) 108:151-159.
Mackin, J. G.
1953. Incidence of infection of oysters by Dermocystidium
in the Barataria Bay area of Louisiana. Natl. Shellfish.
Assoc. Conv. Addendum 1951, p 22-36.
1962. Oyster diseases caused by Dermocystidium marinum
and other microorganisms in Louisiana. Publ. Inst. Mar.
Sci., Univ. Tex. 7:132-229.
Mackin, J. G., and A. K. Sparks.
1962. A study of the effect on oysters of crude oil loss from
a wild well. Publ. Inst. Mar. Sci., Univ. Tex. 7:230-261.
Mengebier, W. L., and L. Wood.
1969. The effects of Minchinia nelsoni infections on enzyme
levels in Crassostrea virginica - II. serum phosphohexose
isomerase. Comp. Biochem. Physiol. 29:265-270.
Menzel, R. W., and S. H. Hopkins.
1955. The growth of oysters parasitized by the fungus Dermo-
cystidium marinum and by the trematode Bucephalus
cuculus. J. Parasitol. 41:333-342.
Mohamed, a. M., and M. M. Ishak.
1981. Growth rate and changes in tissue carbohydrate dur-
ing schistosome infection of the snail Biomphalaria alexan-
drina. Hydrobiologia 76:17-21.
Neff, J. M., P. D. BoEHM, AND W. E. Haensly.
1985. Petroleum contamination and biochemical alterations
in oysters (Crassostrea gigas) and plaice (Pleuronectes
platessa) from bays impacted by the Amoco Cadiz crude oil
spill. Mar. Environ. Res. 17:281-283.
564
WILSON ET AL.: EFFECT OF SNAIL ON OYSTERS
NiSHINO, T., S. NOJIMA, AND T. KiKUCHL
1983. Quantitative studies of the life history and interspecific
relationship of two gastropod species, Odostomia sp. (ecto-
parasite) and Umbonium (Suchium) moniliferum (Lamarck)
(host). I. Life history and population dynamics oi Odostomia
sp. Kyushu Univ., Publ. Amakusa Mar. Biol. Lab. 7:61-
79.
Peterson, G. L.
1977. A simplification of the protein assay method of Lowry
et al. which is more generally applicable. Anal. Biochem.
83:346-356.
Pickering, A. D., T. G. Pottinger, and P. Christie.
1982. Recovery of the brown trout, Salmo trutta L., from
acute handling stress: a time-course study. J. Fish Biol.
20:229-244.
Pipe, R. K.
1985. Seasonal cycles in and effects of starvation on egg
development in Mytilus edulis. Mar. Ecol. Prog. Ser. 24:
121-128.
POrtner, H., B. Surholt, and M. Grieshaber.
1979. Recovery from anaerobiosis of the lug worm, Arenicola
marina L.: changes of metabolite concentrations in the
body-wall musculature. J. Comp. Physiol. 133:227-231.
PovreLL, E. N., S. J. Connor, J. J. Kendall, Jr., C. E. Zastrow,
AND T. J. Bright.
1984. Recovery by the coral Acropora cervicomis after drill-
ing mud exposure. The free amino acid pool. Arch. Environ.
Contam. Toxicol. 13:243-258.
Powell, E. N., M. Kasschau, E. Chen, M. Koenig, and J.
Pecon.
1982. Changes in the free amino acid pool during environmen-
tal stress in the gill tissue of the oyster, Crassostrea vir-
ginica. Comp. Biochem. Physiol. A Comp. Physiol. 71:
591-598.
Powell, E. N., M. E. White, E. A. Wilson, and S. M. Ray.
1987. Small-scale spatial distribution of a pyramidellid snail
ectoparasite, Boonea impressa, in relation to its host, Crasso-
strea virginica, on oyster reefs. Pubbl. Stn. Zool. Napoli
I Mar. Ecol. 8:107-130.
Preece, a.
1972. A manual for histologic technicians. Little, Brown and
Company, Boston, 428 p.
Ray, S. M.
1954. Experimental studies on the transmission and patho-
genicity of Dermocystidium marinum, a fungus parasite of
oysters. J. Parasitol. 40:235.
1966a. Notes on the occurrence oi Dermocystidium marinum
in the Gulf of Mexico coast during 1961 and 1962. Proc.
Natl. Shellfish. Assoc. 54:45-54.
1966b. A review of the culture method for detecting Dermo-
cystidium marinum, with suggested modifications and pre-
cautions. Proc. Natl. Shellfish. Assoc. 54:55-69.
Ray, S. M., J. G. Mackin, and J. L. Boswell.
1953. Quantitative measurement of the effect on oysters of
disease caused by DerTnocystidium murinum. Bull. Mar.
Sci. Gulf Caribb. 3:6-33.
Riley, R. T.
1976. Changes in the total protein, lipid, carbohydrate, and
extracellular body fluid free amino acids of the Pacific oyster,
Crassostrea gigas, during starvation. Proc. Natl. Shellfish.
Assoc. 65:84-90.
Robertson, R.
1978. Spermatophores of six eastern North American pyra-
midellid gastropods and their systematic significance (with
the new genus Boonea). Biol. Bull. (Woods Hole) 155:360-
382.
Robertson, R., and T. Mau-Lastovicka.
1979. The ectoparasitism of Boonea and Fargoa (Gastropoda:
Pyramidellidae). Biol. Bull. (Woods Hole) 157:320-
333.
Roehrig, K. L., and J. B. Allred.
1974. Direct enzymatic procedure for the determination of
liver glycogen. Anal. Biochem. 58:414-421.
Smith, L. 0., and B. C. Coull.
1987. Juvenile spot (Pisces) and grass shrimp predation on
meiobenthos in muddy and sandy substrata. J. Exp. Mar.
Biol. Ecol. 105:123-136.
SONIAT, T. M.
1985. Changes in levels of infection of oysters by Perkinsus
marinus, with special reference to the interaction of tem-
perature and salinity upon parasitism. Northeast Gulf Sci.
7:171-174.
SoNiAT, T. M., AND M. L. Koenig.
1982. The effects of parasitism by Perkinsus marinus on the
free amino acid composition of Crassostrea virginica mantle
tissue. J. Shellfish Res. 2:25-28.
Stein, J. E., and J. G. Mackin.
1955. A study of the nature of pigment cells of oysters and
the relation of their numbers to the fungus disease caused
by DerTnocystidium marinum. Tex. J. Sci. 7:422-429.
1957. A histochemical study of the glycogen content of
oysters infected by Dermocystidium marinum. Tex. A&M
Res. Found., Tech. Rep. No. 23, Proj. No. 23, 5 p.
Swift, M. L., and M. Ahmed.
1983. A study of glucose, Lowry-positive substances, and
triacylglycerol levels in the hemolymph of Crassostrea vir-
ginica (Gmelin). J. Shellfish Res. 3:45-50.
Thompson, S. N.
1986. Effects of dietary carbohydrate on the nutritional
physiology and blood sugar level of Trichoplusia ni para-
sitized by the insect parasite, Hyposoter exiguae. Parasit-
ology 92:25-30.
Thompson, S. N., and B. F. Binder.
1984. Altered carbohydrate levels and gluconeogenic enzyme
activity in Trichoplusia ni parasitized by the insect para-
site, Hyposoter exiguae. J. Parasitol. 70:644-651.
Trider, D. J., AND J. D. Castell.
1980. Influences of neutral lipid on seasonal variation of total
lipid in oysters, Crassostrea virginica. Proc. Natl. Shell-
fish. Assoc. 70:112-117.
Ward, J. E., and C. J. Langdon.
1986. Effects of the ectoparasite Boonea ( = Odostomia) im-
pressa (Say) (Gastropoda: Pyramidellidae) on the growth rate,
filtration rate, and valve movements of the host Crassostrea
virginica (Gmelin). J. Exp. Mar. Biol. Ecol. 99:163-180.
Wells, H.
1959. iiotes on Odostomia impressa (Say). Nautilus 72: 140-
144.
White, D. C, M. W. Davis, J. S. Nickels, J. D. King, and
R. J. Bobbie.
1979. Determination of the sedimentary microbial biomass
by extractible lipid phosphate. Oecologia (Berl.) 40:51-62.
White, M. E., E. N. Powell, and C. L. Kitting.
1984. The ectoparasitic gastropod Boonea ( = Odastomia) im-
pressa: Population ecology and the influence of parasitism
on oyster growth rates. Pubbl. Stn. Zool. Napoli I Mar.
Ecol. 5:283-299.
White, M. E., E. N. Powell, and S. M. Ray.
1988. Effect of parasitism by the pyramidellid gastropod
Boonea impressa on the net productivity of oysters
(Crassostrea virginica). Estuarine Coastal Shelf Sci. 25:
359-377.
565
FISHERY BULLETIN: VOL. 86, NO. 3
White, M. E., E. N. Powell, S. M. Ray, and E. A. Wilson.
1987. Host-to-host transmission of Perkinsus marinus in
oyster {Crassostrea virginica) populations by the ectopara-
sitic snail Boonea impressa (Pyramidellidae). J. Shellfish
Res. 6:1-5.
White, M. E., E. N. Powell, S. M. Ray, E. A. Wilson, and
C. E. Zastrow.
In press. Metabolic changes induced in oysters {Crassostrea
virginica) by the parasitism of Boonea impressa (Pyramidel-
lidae:Gastropoda). Comp. Biochem. Physiol.
WiCKHAM, D. E.
1986. Epizootic infestations by nemertean brood parasites on
commercially important crustaceans. Can. J. Fish. Aquat.
Sci. 11:2295-2302.
Wilson, E. A., M. E. White, E. N. Powell, and S. M. Ray.
In press. Patch formation by the ectoparasitic snail, Boonea
impressa, on its oyster host, Crassostrea virginica. Veliger.
Wright, D. A., and E. W. Hetzel.
1985. Use of RNA:DNA ratios as an indicator of nutritional
stress in the American oyster Crassostrea virginica. Mar.
Ecol. Prog. Ser. 25:199-206.
YuiLL, T. M.
1987. Diseases as components of mammalian ecosystems:
mayhem and subtlety. Can. J. Zool. 65:1061-1066.
566
LIFE TABLES FOR TWO FIELD POPULATIONS OF
SOFT-SHELL CLAM, MYA ARENARIA, (MOLLUSCA : PELECYPODA)
FROM LONG ISLAND SOUND
Diane J. Brousseau' and Jenny A. Baglivo^
ABSTRACT
Life tables were constructed for two populations oiMya arenaria from Long Island Sound, USA, based
on schedules of age-specific fecundity and mortality determined under natural conditions. Mya arenaria
shows a basic conservatism in general life history pattern. In both populations, fecundity increases with
increasing female size; sexual maturity is attained at 1 year of age; a single annual breeding season occurs
and survivorship curves approximate the type III of Deevey, which is characterized by extremely heavy
mortality early in life followed by relatively constant mortality thereafter. Differences in the age-specific
parameters for the two populations exist, however; for clams greater than 1 year of age, both age-specific
fecundity and survivorship are significantly higher in the Stonington population. These differences in
the structure and dynamics of the two populations may be due to environmental heterogeneity. Reduced
body size due to slower growth in coarse substrate, as well as the increased maintenance demands resulting
from burrowing and valve activity in large-grained sediment, may account for the lower egg production
and lower survival rates found in the Westport population.
The life history pattern of a species has been defined
as the way in which that species partitions the
limited resources of time or energy among the three
basic biological processes of growth, maintenance,
and reproduction. Efforts to generate empirical
values for life history parameters, age-specific fecun-
dity and survivorship, have only recently allowed the
construction of life tables for field populations.
Determining life history parameters for a commer-
cially important species such as the soft-shell clam,
Mya arenaria, is particularly useful since they can
be used in theoretical models which are designed to
analyze the effect of changing survival and fecun-
dity values on the growth rate of the population.
Life tables now exist for a number of benthic
marine invertebrates: the ha,rna,c\e—Chathamulus
stellatus (Connell 1961); the prosobranchs—
Dicathais orbita (Phillips and Campbell 1974),
Nucella (= Thais) lapillus (Frank 1969), Nodilit-
torina tuber culata (Doran 1968), and Conns pen-
naceus (Perron 1983); the coelenterates— Mwr-icea
califomica (Grigg 1977), M.fruticosa (Grigg 1977),
and Balanophyllia elegans (Fadlallah 1983); the
bivalves— Mi/a arenaria (Brousseau 1978), Tapes
phillipinarum (Yap 1977), and Gemma gemma
'Department of Biology, Fairfield University, Fairfield, CT
06430.
^Department of Mathematics and Computer Science, Fairfield,
University, Fairfield, CT 06430; present address: Mathematics
Department, Boston College, Chestnut Hill, MA 02167.
(Weinberg 1985). These studies, however, examine
life history parameters for a single species popula-
tion or for successive cohorts within a population.
The present study examines age-specific fecundity
and survivorship in two geographically separated
populations of the soft-shell clam, Mya arenaria,
using identical methodology. Since differences in
methodology can influence estimates of demo-
graphic parameters (Fadlallah 1983) uniformity of
approach is necessary for interpopulation com-
parisons. This study is the first reported examina-
tion of species- specific traits in two naturally occur-
ring populations and was carried out as part of a
broader study of the population dynamics of this
species along the Connecticut shore of Long Island
Sound.
MATERIALS AND METHODS
Field Study Areas
Field studies were conducted at two intertidal
sites in Long Island Sound, one located at Barn
Island in Stonington, CT Oat. 41°20'N; long.
71°53'W) and the other in the Saugatuck River in
Westport, CT Gat. 41°06'N; long. 73°23'W)(Fig. 1).
The Stonington site is a narrow intertidal sandflat
which extends approximately 10 m shoreward to a
coarse sand beach. At low tide the Westport site ex-
tends 30 m (at its widest point) shoreward to a Spar-
Manuscript accepted April 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
567
FISHERY BULLETIN: VOL. 86, NO. 3
Figure 1.— Map showing locations of the two study sites: Barn Island in Stonington, CT (A) and Saugatuck River in Westport, CT (B).
tina altemijlora marsh. The substrate there is
coarser than at the Stonington site; 50% by weight
is >4 mm grain size diameter, much of that classi-
fiable as either cobble or boulder (Brousseau and
Baglivo 1987). Both study sites are closed to shellfish
harvesting to high levels of bacterial contamination.
Although annual salinity and temperature profiles
for the two sites are not available, monthly surface
and bottom salinity and temperature readings for
the period March 1979- June 1981 and October
1985-August 1986 are available for four sites in
western Long Island Sound: Norwalk, CT (lat.
41°02.5'N; long. 73°27.2'W), Bridgeport, CT Gat.
41°08.7'N; long 73°11.1'W), Stratford, CT (lat.
41°07.6'N; long. 73°07.6'W), and New Haven, CT
(lat. 41°14.4'N; 72°54.2'W)(Tettlebachetal., 1984;
Blogoslawski, pers. comm.^). All water sampling was
done at spring low tide in approximately 20 ft of
water. The annual average range of surface temper-
atures for the four sites was 0.5°-24.2°C. The
lowest surface temperature, 0.0 °C, was recorded at
the New Haven station in February 1980; the high-
est was 25.0°C recorded at the Stratford station in
August 1986. Annual average range of bottom
temperatures was 0.6°-23.4°C. Surface sahnities
ranged from 16.9 to 33.3°/oo with a low of 15.97oo
reported for Stratford and a high of 35.77oo re-
corded at the Norwalk station. Mean bottom salin-
ities during that period ranged between 17.0 and
34.4«/oo.
Fecundity
Oocyte production by female clams collected dur-
ing the summer spawning seasons of 1984 and 1985
was estimated using a histological technique (Brous-
seau 1976). One hundred twenty-five gravid females
from the Westport population and 123 from the
Stonington population were examined. Size-specific
fecundity rates were converted to age-specific rates
using age-size information for M. arenaria obtained
from analysis of internal shell growth bands (Brous-
seau and Baglivo 1987). Age-specific fecundity
estimates for x-yr-old clams, m^., were calculated
using the formula:
Mx = 0.5 ^ Wj P(size-class i | x yr-old)
(1)
'W. Blogoslawski, Northwest Fisheries Center Milford Labora-
tory, National Marine Fisheries Service, NOAA, Roger Avenue,
Milford, CT 06460, pers. commun. March 1987.
where m^ is the mean fecundity for size-class i,
P(size-class i \ x-yr-old) is the conditional probabil-
ity that an x-yr-old clam is in size-class i, and the
sum is taken over 10 mm size-classes i. The condi-
tional probabilities P (size-class i \ x-yr-old) were
derived empirically (see Appendix Tables 1, 2). Ten
millimeter size-classes were used because of sparse-
ness in the data. By convention, m^ represents the
568
BROUSSEAU and BAGLIVO: FIELD POPULATIONS OF MYA ARENARIA
total number of eggs produced by a female of age
X which may be considered to be "female eggs".
Because the sex ratio of both populations is 1:1
(Brousseau 1987a), only half of the oocytes produced
will eventually become females. Accordingly, each
rrijc is one-half the total annual fecundity per female
of age X.
Mortality
To determine size-specific l-yr survival rates of
adult clams (>1 year of age), M. arenaria from
Stonington and Westport were collected, individual-
ly tagged, and returned to their original sites (Ston-
ington: 31 May-1 June 1985; Westport (Plot A): 1
May 1984 and Westport (Plot B): 22 May 1984).
Clams were measured to the nearest 0.1 mm antero-
posteriorly and marked for identification using a
method described previously (Brousseau 1979).
Tagged clams were replanted in parallel furrows ex-
cavated to a depth of 20-30 cm. All plots were 0.3 m
apart and located at midtide level (-I-3.0 m). At the
end of the test period, the clams were collected. All
numbered individuals, both alive and dead, were
returned to the laboratory for measurement. A total
of 1,049 clams from the Stonington population were
tagged, of which 78% were recovered. One-year
recovery rate at the Westport site was 40% of the
1,977 clams initially planted. Mortality rates were
determined on the basis of the number of clams
recovered. This method gives a more accurate
estimate of mortality since it does not consider clams
that were not recaptured in these estimates.
If, on recapture, a dead clam showed no evidence
of growth, either death had occurred naturally in
a slow-growing individual or premature death had
occurred as a result of trauma due to the marking
procedure. In order to correct for premature mor-
tality caused by handling, the proportion of live
clams which showed an increase in shell length over
the year was calculated (Table 1). For the Westport
population all live clams <40 mm shell length grew,
while 94.7% of live clams over 40 mm grew. For the
Stonington population all live clams <50 mm shell
length had grown during the year, while 91.9% of
Table 1 . — Distribution of Mya arenaria recovered alive showing
growth and no growth during the test period.
Westport
(clams >40 mm)
Stonington
(clams >50 mm)
live clams over 50 mm grew. In order to estimate
size-specific survivorship then, the following rules
were applied:
Westport:
Classes <40 mm: P(surviving one year) (2)
= NJiN^ + Nog)
Classes >40 mm: F(surviving one year)
= NJ{N^ + Nog/OMI)
Stonington:
Classes <50 mm: ^(surviving one year) (3)
= NJiN^ + Nog)
Classes >50 mm: P(surviving one year)
= NJiN^ + Ar^G/0.919)
where A^^ is the number alive and Nog is the num-
ber dead with growth. In the smaller classes (West-
port, <40 mm; Stonington, <50 mm), the sum A^^
+ Nog represents all recovered individuals, while
in the larger classes, the adjustment factor allows
us to add a number of dead without growth back
into the total recovered in the same proportion as
the live with no growth.
Age-specific survival rates, P_j, were based on the
same age-size information used to determine age-
specific fecundity rates. These rates were calculated
using the formula:
Px = -t.pj P(size-class i \ a;-yr-old)
(4)
Number
Percent
Growth No growth Growth No growth
196 11 440 39
94.7 5.3 91.9 8.1
where p^ is the probability that a clam in size-class
i survives one year, P(size-class i \ x-yr-old) is the
conditional probability that an a;-yr-old clam is in
size-class i, and the sum is taken over 10 mm size
classes i.
The probability of surviving from settlement to
one year (a period of about 10 months) is calculated
from the estimates of the density of the population
in 2 consecutive years. Both populations were sam-
pled during periods of maximum settlement in 1985.
This period occurred approximately 3 weeks earlier
in the Westport population. At each sampling, 18
to 35 samples (0.1 m^ x 30 cm deep) were taken
along transects running from mean low water shore-
ward to the mean high water mark. Samples were
wet-seived in the laboratory (2 mm mesh) and the
shell lengths of the clams retained by the seive were
measured. Clams of size 2-20 mm in the first year
of sampling are assumed to be the spat; those of size
20-50 mm the following year are assumed to have
come proportionately from the settlement of the
previous year in the same ratios as the empirical
probability distributions (Appendix Tables 1, 2).
569
FISHERY BULLETIN: VOL. 86, NO. 3
RESULTS
Fecundity
Fecundity of Mya arenaria increased with in-
creasing female body size (Fig. 2). Size-specific
fecundity schedules, however, differed at the two
study sites. Comparison of regression lines for
log(fecundity) vs. shell length by analysis of covari-
ance indicated that the relationships were signifi-
cantly different {P = 0.004). Size-specific oocyte pro-
duction in the Stonington clams was larger than in
clams from Westport. The smallest gravid females
observed were 34 and 27 mm in shell length in the
Stonington and Westport populations, respective-
ly. Age at first reproduction is 1 year in clams from
both sites. Sex ratios in both populations did not
differ significantly from 1:1 (Brousseau 1987a).
Recruitment and Mortality
Both temporal and spatial variations in annual
recruitment were observed during the study period.
A substantial settlement occurred at the Westport
site in 1985, as evidenced by the high densities of
2-20 mm clams present in late July of that year
(Table 2). In contrast, spat densities at the Stoning-
ton site were low in early August of 1985, when
settlement occurred. Nonetheless, persistence of the
1985 year class at Westport was poor, whereas ap-
proximately one-half of those set in the Stonington
population during that year were alive one year
later. The coarse substrate at the Westport site may
have aided attachment of the byssal stage juveniles,
resulting in the higher recruitment rate during the
summer of 1985. Direct estimates of M. arenaria
survivorship during the postsettlement to 1-yr
1 7
30
35
46
52
Stn
57
63
68
- + + -
79
16
D
O 16
LLI
« «
e « «
• «
* « «
a * « • *
* • * • * * «
* * « « 2 *
3* »••««•«
»«*«*«« « «*
' * 22 * • •
2 • * •
2 •2**
O ,5
o
15
13
27
33
38
- + + -
49
55
60
66
82
SHELL LENGTH (mm)
Figure 2.— Log of fecundity versus shell length for Mya arenaria from
570
BROUSSEAU and BAGLIVO: FIELD POPULATIONS OF AfTA ARENARIA
Table 2.— Densities per 0.1 m^ of Mya arenaria in size-classes
2-50 mm in the Westporl and 2-60 mm in the Stonington popula-
tions in 1985 and 1986.
Location
Size-class
and date Samples
(mm)
/V/O.l m^
(range)
Westport, Connecticut
23 July 1985 28
2-20
23.36
(0-84)
09 June 1986 18
10-20
0.11
(0-1)
20-30
0.44
(0-5)
30-40
0.89
(0-4)
40-50
0.50
(0-3)
Stonington, Connecticut
15 Aug. 1985 35
2-20
0.37
(0-7)
24 June 1986 24
20-30
0.04
(0-1)
30-40
0.13
(0-1)
40-50
0.33
(0-2)
50-60
0.75
(0-3)
period (10 months) obtained by dividing the number
of 1-yr-olds alive in the summer 1986 by the esti-
mated cohort size in the previous year were 0.0283
and 0.5869 for the Westport and Stonington popu-
lations, respectively.
Size-specific survival rates of adult clams {>1 year)
at both sites are shown in Tables 3 and 4. These
represent empirical estimates from mark and recap-
ture studies with survival adjustments in the larger
classes made according to Equations (2) and (3).
Statistical comparison of the survival distributions
of adult clams, however, was limited to those re-
covered live or recovered showing some growth. At
the Westport site survival was constant across the
10 mm size classes: 20-29 through 50-59 and 60 -i- .
At Stonington, however, the size-specific survival
for classes 30-39 through 70-79 and 80 -i- increased
slightly with size (P < 0.05, x^ test). The probability
18
30
35
Wp 1
46
52
57
63
68
74
79
17
16
O 16
LU
U-
O 15
O
« • »
« * « « 4 • *
* • * *
* * V ■« • •
• * * • *
•3 ;•
2* *»2* *♦ **?
* * 2 • *
* « «
* ««* 22***
• • 2 • * *
* 2 * *
• •*223* •
15
13
27
33
38
44
49
55
6C
66
7 1
82
SHELL LENGTH (mm)
Figure 2.— Continued— the Stonington (STN) and Westport (WPl) populations.
571
FISHERY BULLETIN: VOL. 86, NO. 3
Table 3.— Size-specific survival rates for marked
Mya arenaria from the Stonington population. P
= probability of an individual from size-class / sur-
viving one year; N^ = number of live clams; Nqq
= number of dead clams sfiowing growth.
Size-class
Size (mm)
Na
Nog
P
2
20-30
3
30-40
23
3
0.8846
4
40-50
106
18
0.8548
5
50-60
134
7
0.9462
6
60-70
161
11
0.9308
7
70-80
159
9
0.9420
8
eo+
25
0
1 .0000
Table 4.— Size-specific survival rates for marked
Mya arenaria from the Westport population. P =
probability of an individual from size-class / surviv-
ing one year;
Na =
number of live clams; N,
DG -
number of dead clams showing growth.
Size-class Size (mm)
N^
N.
DG
2
20-30
15
5
0.7500
3
30-40
87
56
0.6084
4
40-50
150
76
0.6515
5
50-60
48
19
0.7052
6
60-70
9
5
0.6303
7
70-80
0
0
0.0000
8
80-1-
of surviving one year for adult clams at Stonington
(30 -1- mm in shell length) is significantly higher than
at Westport (20+ mm in shell length) (P < 0.001,
X^ test).
Age-specific survivorship estimates (Table 5) were
calculated by assuming that 20-30 mm clams at
Stonington survivied with the same probability as
30-40 mm clams and that 90-100 mm clams at Ston-
ington survived with the same probability as 80-90
mm clams and application of Equation (4) using the
empirical distributions in Appendix Tables 1 and 2.
Table 5. — Age-specific survival rates for marked
Mya arenaria from the Westport and Stonington,
CT populations. P^ = probability of surviving
from age-class x to age-class x + ^.
Age-class
(yr)
Westport
1
2
3
4
5
6
7
8
9
0.6390
0.6442
0.6625
0.6680
0.6595
0.6422
0.6128
0.5991
0.5021
Stonington
P.
0.8843
0.9307
0.9390
0.9550
0.9661
0.9746
0.9746
0.9746
0.9746
Population Dynamics
The age-specific survivorship and fecundity sched-
ules for M. arenaria are combined in Hfe table form
in Tables 6 and 7. Since rates of larval survivorship
are difficult to measure for species with planktonic
larvae, simplifying assumptions are necessary for
estimating survivorship probabilities of larvae (age
0). In order to complete the life tables the equilib-
rium settlement rates (rj were calculated using
the method of Brousseau et al. (1982). Our choice
of r^ implies that the net reproductive rate, Rq,
which is defined by Rq = ^l^m^, were Ij = sur-
vivorship and m_c = fecundity, equals one for both
populations. This is not meant to imply however,
that the populations studied here are considered to
be in equilibrium. Rather, it is used simply as a
theoretical construct in which to examine possible
consequences of differing fecundity and mortality
schedules on the two populations of M. arenaria.
Table 6.— Life table for the Westport, CT Mya arenaria
population, assuming equilibrium conditions, i^ =
survivorship to beginning of age interval x, or /^_, x
Pj, . , when X > 1 ;
m^ = fecundity during age interval x.
Age (yr)
'x
m^
'x^x
0
1.0
1
^1.0599 X
10-'
1485615
0.1575
2
6.7728 X
10"^
2989512
0.2025
3
4.3630 X
io-«
3867846
0.1688
4
2.8905 X
io-«
5192053
0.1501
5
1.9309 X
10-8
5892286
0.1138
6
1.2734 X
10-8
6464925
0.0823
7
8.1780 X
lo-'^
6478050
0.0530
8
5.0115 X
10-*^
6842903
0.0343
9
3.0024 X
io-«
8390977
0.0252
10
1.5075 X
10-*^
8390977
0.0126
1 /, = r, X (probability of surviving from 2 months to 1 year).
Table 7.— Life table for the Stonington, CT Mya are-
naria population, assuming equilibrium conditions. 1^
= survivorship to beginning of age interval x, or /,_,
X P^_, when x > 1; m^ = fecundity during age inter-
val X.
Age (yr)
'x
m.
'x^x
0
1.0
1
'2.014 X
10-8
1109952
0.0224
2
1.7811 X
10-8
3712746
0.0661
3
1.6577 X
10-8
4947569
0.0820
4
1.5566 X
10-8
7125027
0.1109
5
1.4866 X
10-8
8287563
0.1232
6
1.4362 X
10-8
8724158
0.1253
7
1.3997 X
10-8
8724158
0.1221
8
1.3642 X
10-8
8724158
0.1190
9
1.3296 X
10-8
8724158
0.1160
10
1.2958 X
10-8
8724158
0.1130
' /, = r^ X (probability of surviving from 2 months to 1 year).
572
BROUSSEAU and BAGLIVO: FIELD POPULATIONS OF MYA ARENARIA
The equilibrium settlement rate (r,) for the West-
port population is 3.7453 x 10'^ and that for the
Stonington population is 3.4318 x 10"^. The equi-
librium first year survival rates (i.e., the probability
of surviving from egg to year 1 necessary to main-
tain a population at equilibrium) then, become
1.0599 X 10-^ and 2.014 x lO'^ for the Westport
and Stonington populations, respectively.
Under the assumption of equal rates of larval im-
port and export, it is possible to calculate empirical
settlement rates by estimating the total number of
eggs released per unit area by spawning females in
1985 and dividing this value into the densities of
2-20 mm spat settled per unit area 2 months later.
Size-frequency data for the populations was used to
determine the number of clams in each 10 mm size
class (Fig. 3). The number of animals in each class
(assuming that one-half the population is female) was
multiplied by the mean size-specific annual fecun-
dity value for that size class. Summing over all size
classes of reproducing females gives an estimate of
the 1985 oocyte production per unit area. Empirical
settlement rates of 8.4589 x 10"*^ and 7.1953 x
10"^ were calculated for the Westport and Stoning-
ton populations. These represent values of 2.2585
and 2.0967 times the estimated equilibrium settle-
ment rate of these populations. Multiplying the em-
pirical settlement rate by the probability of surviv-
ing the remainder of the year give actual first year
actual first year survival rates of 2.3939 x 10"^
and 4.2229 x 10"^ for Westport and Stonington,
respectively.
The heavy mortality evident by the low juvenile
survivorship rates {l{) for M. arenaria take into ac-
count the losses incurred during fertilization, meta-
morphosis and recruitment, and subsequent survival
through the first year of life. Table 8 compares the
first year survival rates of representative marine
invertebrate species with planktonic and non-
planktonic modes of development. Early survival in
species lacking planktonic larvae is on average,
three orders of magnitude higher than that of
species which pass through a planktonic larval stage.
DISCUSSION
General comparisons of life history traits are use-
ful but quantitative comparisons are possible only
from the more detailed information found in life
tables (age-specific fecundity and survivorship). The
difficulty in generating such information, especial-
ly for marine bivalves, many of which have plank-
tonic larval stages during their life cycles, has
resulted in the construction of few complete life
tables. Moreover, no reported field study has ex-
amined life history traits for more than one popu-
lation of a species simultaneously. Consequently, the
extent to which quantitative differences in life
history parameters are characteristic of the life
history of a single species is unknown.
Table 8.— Empirical estimates of first year survival rates of marine invertebrate species with and
without planktonic larvae (adapted from Perron 1983).
Class
Species
Survivorship
Developmental
mode
References
Anthozoa
Muricea californica
2.6
X
10-^
planktonic
Grigg (1977)
Muricea fruticosa
3.7
X
io-«
planktonic
Grigg (1977)
Balanophyllia elegans
5.0
X
10-2
nonplanktonic
Fadlallah (1983)
Bivalvia
Gemma gemma
Cohorts 1978-1981
2.3
1.2
2.0
1.1
X
X
X
X
10-2
10-^
10-^
10-^
nonplanktonic
Weinberg (1985)
Mya arenaria
Westport population
2.4
=
10-^
planktonic
Present study
Stonington population
4.2
X
10-8
Tapes phillipinarum
1.8
X
io-«
planktonic
Yap (1977)
Gastropoda
Thais lamellosa
^9.0
X
10-3
nonplanktonic
Spight (1975)
Conus pennaceus
7.0
X
10-"
nonplanktonic
Perron (1983)
Lacuna pallidula
2.0
X
10-*
nonplanktonic
Smith (1973)
Aplysia Juliana
28.4
X
10-^
planktonic
Sarver(1979)
Crustacea
Balanus glandula
6.0
X
10-5
planktonic
Connell (1970)
Barnacles (3 spp.)
2.0
X
10-5
planktonic
Mines (1979)
'Does not include prehatching mortality.
^Survival through 30-d planktonic period only.
573
FISHERY BULLETIN: VOL. 86, NO. 3
WP1 1985
7.7-
a.3 -
2.2-
I
I
i-^
I
-I 1 k * 1 1 1-
2 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
SHELL LENGTH (MM)
STN 1985
11.0
B.a--
7.7--
in
5.5--
9.S-
2.2-
I
II
i i
a a ^ a a ^
ll
I i
i ^ i .
-t — ♦ — —I
2 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
SHELL LENGTH (MM)
Figure 3.— Size-frequency distributions of Mya arenaria from the Stonington (STN)
574
BROUSSEAU and BAGLIVO: FIELD POPULATIONS OF MYA ARENARIA
WP1 1986
11.0
9.9--
a.s
7.7-
3.3 -
2.2 -
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
SHELL LENGTH (MM)
STN 1986
11.0
a.s
7.7 -
CO
3.3 -
2.2-
ii
ll
^
a^
»ii
» i
^
II
^
1 1
I
^
t^ i
U
I
Im
2 5 1 0 1 5 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
SHELL LENGTH (MM)
Figure Z— Continued— a.nd. Westport (WPl) populations in 1985 and 1986.
575
FISHERY BULLETIN: VOL. 86, NO. 3
The two populations of M. arenaria from Long
Island Sound appear to show a basic conservatism
in general life history pattern. In both populations,
fecundity increases rapidly in young females with
individuals reaching sexual maturity in time to
reproduce at the beginning of their second year of
life. Similarly, larval and adult survivorship sched-
ules follow the type III survivorship curve of Deevey
(1947). Extremely heavy mortality early in life is
followed by roughly constant mortality thereafter.
This pattern is similar to that described for a popu-
lation of M. arenaria from Gloucester, MA (Brous-
seau 1978) except that in the latter, age of first
reproduction occurs at the end of the second year.
It is interesting to note, however, that some of the
major life history features of M. arenaria show con-
siderable latitudinal variation within the species.
Frequencies of spawnings during the year increases
(for review see Ropes and Stickney 1965; Brousseau
1987a) and length of life and body size show a ten-
dency to decrease with decreasing latitude (Belding
1930; Newcombe 1935). Details of the ways in which
such variations affect the life history parameters,
however, remain to be studied.
The possibility of gene flow between populations
of animals with planktonic larval stages always
exists. Nevertheless, the amount of genetic overlap
is effectively reduced as the geographical distance
between the populations increases. The significant
quantitative differences in the age-specific demo-
graphic parameters for the two populations studied
here suggest that v^ithin the framework of a general
life history strategy, a response to the biotic and
abiotic components of the immediate environment
is possible. Evidence from this study indicates that
environmental conditions at the Westport site may
be less optimal for the growth and maintenance of
M. arenaria than are those at the Stonington site.
The higher equilibrium settlement rate calculated
for the Westport population indicates that on aver-
age, a larger annual spatfall is needed to maintain
that population than is required at the Stonington
site.
The biotic factors most often cited as agents capa-
ble of altering the survival and fecundity of in-
dividuals are predation, competition, disease, and
parasitism. It is unlikely that predation was a major
source of adult mortality at either site for reasons
previously discussed. Moreover, the effects of preda-
tion would be limited to the small, surface-dwelling
clams, since 1) crabs, fish, and birds are unable to
capture deep burrowing adults and 2) it has been
demonstrated that M. arenaria exhibits a "size
refuge" from naticid predators (Edwards and Hueb-
ner 1977). Hancock (1973) has suggested that com-
petition for food or space between the spat and
adults may contribute to lowered survival in newly
settled clams. If this were the case, one would ex-
pect lower juvenile survivorship rates in the Stoning-
ton population where population densities were
greater (Fig. 3).
Trematode infestations have been demonstrated
to cause castration and high mortalities in a popula-
tion of venerid clams, Transenella tantilla, in
California (Obrebski 1968). Although both trematode
infections (Stunkard 1938; Uzmann 1951) and fungal
parasites have been reported in M. arenaria
(Andrews 1954), no evidence of either was observed
in the 2,826 histologically prepared clams (1,583
from Stonington; 1,243 from Westport) examined
in a study of the reproductive cycle in the two popu-
lations (Brousseau 1987a). Sarcomatous neoplasia,
a proliferative disorder characterized by increased
number of "leukemia-like" cells in the tissues and
organs (Farley 1969; Brown et al. 1977; Cooper et
al. 1982), has been identified in samples of clams
from both Stonington and Westport. Prevalence of
the disorder in the Stonington population ranges
from 0 to 46%, whereas 0-69% of the M. arenaria
from Westport were neoplastic, depending on the
collection date (Brousseau 1987b). At present it is
not known if neoplasia is a source of mortality in
field populations of clams. It seems reasonable to
assume that it is, however, since neoplastic cells are
invasive and at times cause the destruction of organs
and tissues in infected animals (Yevich and Barszcz
1977). If sarcomatous neoplasms prove to be malig-
nant, this disease could be responsible to some
degree for the higher mortality rates reported in the
Westport population.
The abiotic factors with the greatest effect on the
biotic potential of estuarine organisms are temper-
ature, salinity, substrate, and food availability. Adult
M. arenaria typically inhabit the intertidal zone and
are adapted to a vvide range of fluctuations in water
temperature and salinity. In addition, sediments
tend to buffer temperature and salinity fluctuations
(Sanders et al. 1965; Johnson 1965, 1967). There-
fore, infaunal organisms like the soft-shell clam are
subject to less extreme environmental fluctuations
than are exposed organisms living on or attached
to the surface. Belding (1930) noted that M.
arenaria withstands extreme variation in salinity,
being able to adjust to changing tides every six
hours. Shaw and Hamons (1974) found that lethal
conditions for burrowed clams were met only when
temperatures persisted in the high 20 °C range and
salinities were 2°/oo or lower. In the laboratory.
576
BROUSSEAU and BAGLIVO: FIELD POPULATIONS OF MYA ARENARIA
adult survival is not altered by salinities of 2.5°/oo
(Chanley 1957; Pfitzenmeyer and Drobeck 1963;
Castagna and Chanley 1973) to 35"/oo (Castagna
and Chanley 1973). Annual water temperature and
salinity patterns in Long Island Sound fall well
within the range of conditions tolerated by M.
arenaria (see Materials and Methods). If tempera-
ture and salinity differences did occur, their effects
would be minimal.
Substrate differences at the Stonington and West-
port sites appear to be the most immediate cause
of the observed differences in age-specific fecundity
and survivorship. Several investigators have found
that sediment type is important in controlling
growth rate and shell allometry in M. arenaria
(Belding 1930; Swan 1952; Newell and Hidu 1982).
Clams grown on coarse sediments (gravel, cobble),
such as that at the Westport site are slow-growing
and more globose in shape than clams from sand or
mud environments. Although differences in growth
rate could be due to differences in food availability
at the two sites, the allometric variations (Brous-
seau and Baglivo 1987) strongly support the hypoth-
esis that substrate effect is the factor controlling
growth. Reduced body size in Westport clams may
have an indirect effect on fecundity by restricting
egg production. In addition, both Glude (1954) and
Pfitzenmeyer and Drobeck (1967) demonstrated that
M. arenaria burrows fastest in fine-grained sedi-
ments since physical resistance to burrowing in-
creases with increasing particle size (Trueman 1954).
Hence, the high daily maintenance requirements of
Westport clams may result in less energy available
for reproduction, long-term maintenance, and sur-
vival. The Westport population may be an example
of a population inhabiting a marginal environment.
The underlying pattern of life history (reproduc-
tive effort, patterns of recruitment, survival profile,
and growth schedule) for both populations of M.
arenaria is very similar. This is not surprising since
population parameters are viewed as evolved,
species-specific traits. Nevertheless, our evidence
for local differentiation in different habitats demon-
strates the degree to which individual populations
of a widespread species can respond in different
ways to their immediate environments; environmen-
tal heterogeneity can be reflected in the structures
and dynamics of local populations.
ACKNOWLEDGMENTS
We wish to thank K. Schellinkhout, J. Smeriglio,
J. Trautman, and J. Wachter for technical assist-
ance in the field. Working space and running sea-
water raceways for marking and holding animals
were made available at the National Marine Fish-
eries Laboratory, Milford, CT (J. E. Hanks and A.
Calabrese, Directors). Temperature and salinity data
for Long Island Sound was provided by W. Blogos-
lawski. Financial support for this study was provided
under grant number NA82AA-D-00018 of the Con-
necticut Sea Grant Program.
LITERATURE CITED
Andrews, J. D.
1954. Notes on fungal parasites of bivalve molluscs in Chesa-
peake Bay. Proc. Natl. Shellfish. Assoc. 45:157-163.
Belding, D. L.
1930. The soft-shelled clam fishery of Massachusetts. Dep.
Conserv., Div. Fish. Game, Commonw. Mass. Mar. Fish. Ser.
1, 65 p.
Brousseau, D. J.
1976. Spawning cycle, fecundity and recruitment in a popula-
tion of Mya arenaria (soft-shell clam) from Cape Ann,
Massachusetts. Fish. Bull., U.S. 76:155-166.
1978. Population dynamics of the soft-shell clam, Mya
arenaria. Mar. Biol. (Berl.) 50:63-71.
1979. Analysis of growth rate in Mya arenaria using
von Bertalanffy equation. Mar. Biol. (Berl.) 51:221-
227.
1987a. A comparative study of the reproductive cycle of the
soft-shell clam, Mya arenaria in Long Island Sound. J.
Shellfish Res. 6:7-15.
1987b. Seasonal aspects of sarcomatous neoplasia in Mya
arenaria (Soft-shell clam) from Long Island Sound. J.
Invertebr. Pathol. 50:269-276.
Brousseau, D. J., and J. Baglivo.
1987. A comparative study of age and growth in Mya
arenaria (soft-shell clam) from three populations in Long
Island Sound. J. Shellfish Res. 6:17-24.
Brousseau, D. J., J. Baglivo, and G. E. Lang.
1982. Estimation of equilibrium settlement rates for benthic
marine invertebrates: its application to Mya arenaria
(Mollusca:Pelecypoda). Fish. Bull., U.S. 80:642-644.
Brown, R. S., R. E. Wolke, S. B. Saila, and C. W. Brown.
1977. Prevalence of neoplasia in 10 New England populations
of the soft-shell clam (Mya arenaria). Ann. N. Y. Acad. Sci.
298:522-534.
Castagna, M., and p. Chanley.
1973. Salinity tolerance of some marine bivalves from
estuarine environments in Virginia waters on the western
Mid-Atlantic coast. Malacologia 12:47-96.
Chanley, P. E.
1957. Survival of some bivalves in water of low salinity. Proc.
Natl. Shellfish. Assoc. 48:52-65.
CONNELL, J. H.
1961. The effects of competition, predation by Thais lapillus
and other factors on natural populations of the barnacle
Balanus balanoides. Ecol. Monogr. 31:61-104.
1970. A predator-prey system in the marine intertidal region
I. Balanus glandula and several predatory species of Thais.
Ecol. Monogr. 40:49-78.
Cooper, K. R., R. S. Brown, and P. W. Chang.
1982. The course and mortality of a hematopoietic neoplasm
in the soft-shell clam, Mya arenaria. J. Invertebr. Pathol.
39:149-157.
577
FISHERY BULLETIN: VOL. 86. NO. 3
Deevey, E. S.
1947. Life tables for natural populations of animals. Q. Rev.
Biol. 22:283-314.
DORAN, L. D.
1968. Hermit crabs and gastropod survivorship. Can. J.
Zool. 46:719-722.
Edwards, D. C, and J. D. Heubner.
1977. Feeding and growth rates of Polinices duplicatns prey-
ing on Mya arenaria at Barnstable Harbor, Massachusetts.
Ecology 58:1218-1236.
Fadlallah, Y. H.
1983. Population dynamics and life history of a solitary coral,
Balanophyllia elegans from Central California. Oecologia
58:200-207.
Farley, C. A.
1969. Sarcomatoid proliferative disease in a wild population
of blue mussel (Mytilus edulis). J. Natl. Cancer Inst.
43:509-516.
Frank, P. W.
1969. Growth rates and longevity of some gastropod molluscs
on the coral reef at Heron Island. Oecologia (Berl.) 2, p.
232-250.
Glude, J. B.
1954. Survival of soft-shell clams, Mya arenaria buried at
various depths. Maine Dep. Sea Shore Fish. Res. Bull. 22,
260 p.
Grigg, R. W.
1977. Population dynamics of two gorgonia corals. Ecology
58:278-290.
Hancock, D. A.
1973. The relationship between stock and recruitment in ex-
ploited invertebrates. Rapp. P. -v. R6un. Cons. Perm. int.
Explor. Mer 164:113-131.
HiNES, A. H.
1979. The comparative reproductive ecology of 3 species of
intertidal barnacles. In S. E. Stancyk (editor), Reproduc-
tive ecology of marine invertebrates, p. 213-234. Belle W.
Baruch Library in Marine Science, Vol. 9. Univ. South
Carolina Press, Columbia, SC.
Johnson, R. G.
1965. Temperature variation in the infaunal organisms of a
sandflat. Limnol. Oceanogr. 10:114-120.
1967. Salinity of intertidal water in a sandy beach. Limnol.
Oceanogr. 12:1-7.
Newcombe, C. L.
1935. Growth of Mya arenaria in the Bay of Fundy region.
Can. J. Res. 13:97-137.
Newell, C. R., and H. Hidu.
1982. The effects of sediment type on growth rate and shell
allometry in the soft-shell clam Mya arenaria. L. J. Exp.
Mar. Biol. Ecol. 65:285-195.
Obrebski, S.
1968. On the population ecology of two intertidal in-
vertebrates and the paleoecological significance of size-
frequency distributions of living and dead shells of the
bivalve Transenella tantilla. Ph.D. Thesis, Univ. Chicago,
102 p.
Perron, F. E.
1983. Growth, fecundity and mortality of Cornis pennaceus
in Hawaii. Ecology 64:53-62.
Phillips, B. F., and N. A. Campbell.
1974. Mortality and longevity in the whelk, Dicathais orbita
(Gmelin). Aust. J. Mar. Freshwater Res. 25:25-33.
Pfitzenmeyer, H. T., and K. G. Drobeck.
1963. Benthic survey for populations of soft-shelled clams,
Mya arenaria in the lower Potomac River, Maryland. Ches-
apeake Sci. 4:67-74.
1967. Some factors influencing reburrowing activity of soft-
shell clam Mya arenaria. Chesapeake Sci. 8:193-199.
Ropes, J. W., and A. P. Stickney.
1965. The reproductive cycle of Mya arenaria in New
England. Biol. Bull. (Woods Hole) 128:315-327.
Sanders, H. L., P. C. Mangelsdorf, Jr., and G. R. Hampson.
1965. Salinity and faunal distribution in the Pocasset River,
Massachusetts. Limnol. Oceanogr. 10 (Suppl.), p. R216-
R229.
Sarver, D. J.
1979. Recruitment and juvenile survival in the sea hare
ylpii/sra JM^iana (Gastropoda:Opisthobranchia). Mar. Biol.
(Berl.) 54:353-361.
Shaw, W. N., and F. Hamons.
1974. The present status of the soft-shell clam in Maryland.
Proc. Natl. Shellfish. Assoc. 64:38-44.
Smith, D. A. S.
1973. The population biology of Lacuna pallidula (DaCosta)
and Lacuna vincta (Montagu) in Northeast England. J.
Mar. Biol. Assoc. U.K. 53:493-520.
Spight, T. M.
1974. Sizes of populations of marine snail. Ecology 55:712-
729.
Stunkard, H. W.
1938. The morphology and life cycle of the trematode Hinms-
thla quissentensis (Miller and Northrop, 1926). Biol. Bull.
(Woods Hole) 75:145-164.
Swan, E. F.
1952. The growth of the clam, Mya arenaria, as affected by
the substratum. Ecology 33:530-534.
Tettlebach, S., L. Potti, and W. Blogoslawski.
1984. Survey of Vibrio associated with a New Haven shell-
fish bed, emphasizing recovery of larval oyster pathogens.
In R. Colwell (editor). Vibrios in the environment. John
Wiley and Sons, Inc., N.Y.
Trueman, E. R.
1954. Observations on mechanisms of the opening of the
valves of a burrowing lamellibranch, Mya arenaria. J. Exp.
Biol. 31:291-305.
UZMANN, J. R.
1951. Cercaria myae sp. nov., a fork-tailed larva from the
marine bivalve, Mya arenaria. J. Parasitol. 38(2):161-164.
Weinberg, J. R.
1985. Factors regulating population dynamics of the marine
bivalve Gemma gemma: intraspecific competition and salin-
ity. Mar. Biol. (Berl.) 86:173-182.
Yap, W. G.
1977. Population biology of the Japanese little-neck clam.
Tapes phillipinaT^m in Kaneohe Bay, Oahu, Hawaiian
Islands. Pac. Sci. 31:223-244.
Yevich, p. p., and C. a. Barszcz.
1977. Neoplasia in soft-shell clams (Mya arenaria) collected
from oil-impacted sites. Ann. N.Y. Acad. Sci. 298:409-426.
578
BROUSSEAU and BAGLIVO: FIELD POPULATIONS OF MYA ARENARIA
APPENDIX
Appendix Table 1.— Empirical distributions for age-size relationships for the Westporl, CT Mya
arenaria population. Percentage in each size category per age are indicated in parentheses.
Size
(mm)
Age (yr)
1
2
3
4
5
6
7
8
9 +
10-
-20
1
(3.8)
20-
-30
9
(34.6)
3
(2.0)
30-
-40
13
(50.0)
55
(37.2)
19
(12.7)
2
(2.9)
40-
-50
3
(11.5)
70
(47.3)
65
(43.3)
15
(21.4)
7
(15.6)
4
(8.7)
3
(8.3)
3
(7.9)
50-
-60
19
(12.8)
60
(40.0)
40
(57.1)
24
(53.3)
23
(50.0)
16
(44.4)
17
(44.7)
6
(23.1)
60-
-70
1
(0.7)
5
(3.3)
12
(17.1)
13
(28.9)
17
(37.0)
14
(38.9)
14
(36.8)
14
(53.8)
70-
-80
1
(1.4)
1
(2.2)
2
(4.3)
2
(5.6)
4
(10.5)
6
(23.1)
80-
-90
1
(2.8)
Appendix Table 2.— Empirical distributions for age-size relation-
ships in the Stonington, CT Mya arenaria population. Percentage
in each size category per age are indicated in parentheses.
Size
(mm)
Age (yr)
1
2
3
4
5
6 +
20-30
15
(26.8)
30-40
22
(39.3)
4
(5.1)
40-50
13
(23.2)
5
(6.4)
2
(1.1)
50-60
6
(10.7)
31
(39.7)
37
(20.2)
5
(3.8)
1
(7.1)
60-70
31
(39.7)
94
(51.4)
26
(19.8)
2
(7.1)
1
(7.1)
70-80
7
(9.0)
41
(22.4)
66
(50.4)
14
(50.0)
4
(28.6)
80-90
9
(4.9)
31
(23.7)
11
(39.3)
6
(42.9)
90-100
3
(2.3)
1
(3.6)
2
(14.3)
579
A LONG-TERM STUDY OF "MICROCELL" DISEASE IN OYSTERS
WITH A DESCRIPTION OF A NEW GENUS, MIKROCYTOS (G. N.), AND
TWO NEW SPECIES, MIKROCYTOS MACKINI (SR N.) AND
MIKROCYTOS ROUGHLEYI (SR N.)
C. Austin Farley,' Peter H. Wolf,^ and Ralph A. Elston^
ABSTRACT
Continuing long-term studies of oyster disease problems have been carried out over the past 26 years
using field monitoring, gross, histologic, and ultrastructural pathologic methods.
A microorganism of uncertain taxonomy was discovered in 1963 by J. G. Mackin in association with
lesions and mortalities of Japanese oysters, Crassostrea gigas, from Denman Island, British Columbia,
Canada. Mackin coined the term "microcell" for this organism and described the parasite as 1-3 fim
cells with small nuclei which occurred within vesicular connective tissue cells adjacent to characteristic
abscesses. We are describing this organism as Mikrocytos mackini sp. n. in his honor. Similar appearing
organisms were seen by the senior author in flat oysters, Ostrea edulis, from Milford, Connecticut, on
three different occasions: 1) in oysters transferred from Milford, Connecticut, to Chincoteague Bay,
Virginia; 2) in oysters transferred from Milford to Elkhom Slough, California; and 3) in oysters trans-
ferred from Milford to Oxford, Maryland, and held in recirculated sea water. The causative organism
in these three episodes has been shown by electron microscopy to be Bonamia ostreae, the parasite that
was implicated in recent mortalities in flat oysters in Europe. Similar organisms have also been seen
in Olympia oysters, Ostrea lurida, from Oregon and in the Sydney rock oyster, Saccostrea commercialis,
from Australia. Presence of the organism in the latter species is associated with the winter mortalities
originally described by T. C. Roughley, and the pathogen is here described as Mikrocytos roughleyi (sp.
n.) in his honor.
"Microcell" type parasites of oysters are associated
with a complex of diseases that occur in Japanese
oyster, Crassostrea gigas; Sydney rock oyster, Sac-
costrea commercialis; flat oyster, Ostrea edulis;
and Olympia oyster, 0. lurida, in North America,
Europe, and Australia. Severity of disease varies
from an acute, highly lethal form to a chronic,
seasonally recurring disease that does not produce
massive mortalities. The etiologic agents are small,
morphologically simple, and very difficult to com-
pare and characterize taxonomically at light micro-
scope levels of resolution. Associated lesions vary
according to species affected and provide some of
the differences that may be used to distinguish the
agents involved. The complexity of this group and
the difficulties involved in achieving an understand-
ing regarding whether we are dealing with one or
a group of organisms and how they were transferred
to new locations, the long time span involved in
'Northeast Fisheries Center Oxford Laboratory, National
Marine Fisheries Service, NOAA, Oxford, MD 21654.
^62 MacKenzie Street, Bondi Junction, New South Wales,
Australia 2022.
'Center for Marine Disease Control, Battelle/Marine Research
Laboratory, Sequim, WA 93282.
answering these questions, and the continuing dis-
semination of unpublished privileged information
shared in informal workshop gatherings of scientists
with common interests, make it necessary to use un-
published anecdotal information in order to provide
as complete a story as possible.
The first oyster mortality known to be associated
with "microcell" disease was reported in C. gigas
from Denman Island, British Columbia, Canada by
Quayle (1961). Quayle's report documents the epi-
zootic aspects from 1956 to 1960 and demonstrates
the gross appearance of the disease in the Pacific
oyster. A causative agent was not identified until
several years later, when the late J. G. Mackin"*
(unpubl. data) discovered a small intracellular
organism intimately associated histologically with
tissue abscesses in diseased oysters (C gigas) from
Denman Island and called this organism "micro-
cell". He demonstrated this material at the 1963
Shellfish Mortality Conference held at Oxford, MD.
Mackin's demonstration provided us with the in-
sight to identify similar organisms in histologic sec-
^J. G. Mackin, deceased, Texas A&M University, College Station,
TX.
Manuscript accepted May 1988.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
581
FISHf:RY BULLETIN: VOL. 86, NO. 3
tions of Denman Island oysters in our laboratory
collection and further led to the discovery of micro-
cell disease in 0. edulis, 0. lurida, and 5. commer-
cialis from Australia ("Australian winter disease")
described by Roughley (1926). In 1979, the char-
acteristics of microcell disease were demonstrated
at a microscopic diagnosis workshop held at the
Ministry of Agriculture and Fisheries Laboratory
in Weymouth, England. This workshop was
attended by molluscan pathologists from several
European countries including France. In the late
1970s, serious mortalities of 0. edulis in France,
associated with microcell infections, were described
in a paper by Pichot et al. (1979), in which the
organism was named Bonamia ostreae. Reference
to the earlier work on microcell by Mackin and
others (Katkansky et al. 1969) was not included in
their report.
Since the taxonomic relationship and status of
these similar parasites have not been described, it
is the purpose of this paper to present a complete
background on microcell related epizootics and mor-
phological information which, in North America,
preceded the French report (Pichot et al. 1979). Fur-
thermore, microscopic and ultrastructural compari-
sons of the microorganisms are provided and a new
genus, Mikrocytos g. n., and two new species, Mikro-
cytos mackini sp. n. and Mikrocytos roughleyi sp.
n., are described.
MATERIALS AND METHODS
General Procedures
Oyster tissues were collected from a variety of
sources as follows:
Code WWC were C. gigas collected from Henry Bay,
Denman Island, British Columbia, Canada on a
periodic basis from April 1968 to June 1969 by D. B.
Quayle (spring 1969 samples were collected by
N. Bourne). Live oysters were sent by air freight
to the Oxford Laboratory where clinical and gross
features were recorded and they were processed for
histological and, in some cases, ultrastructural
studies.
Code S-124-A were C. gigas from Hawaii collected
in September 1972.
Code S-41 were S. commercialis collected by Peter
Wolf from 22 July to 23 July 1965 from the Georges
River, Woolooware Bay, New South Wales and ship-
ped to the Oxford Laboratory for processing.
Code FK were progeny of 0. edulis from Boothbay
Harbor, ME that were spawned at Milford, CT in
April 1961. Seed oysters were transplanted to
Chincoteague Bay, VA in May 1961 and processed
at subsequent intervals (FK-1-1 to FK-2-5, August
1961; FK-3-1, August 1961; FK-4-1 and FK-4-2,
February 1962).
Code WAC were 0. edulis bred in the Milford, CT
hatchery from 1963 to 1965, and introduced into
California bays as follows:
WAC-1, Milford 1963 seed oysters planted in
Morro Bay, CA in December 1964 and sampled on
7 December 1965 during a heavy mortality.
Sample WAC-21-28 consisted of 38 oysters from
a Milford 1963 stock shipped to California in 1964
and held at Pigeon Point Laboratory, Pigeon Point,
CA until heavy mortality occurred and sampled on
11 May 1966.
WAC-3-1-10 were 1963 Milford oysters placed in
Morro Bay in 1964. Heavy mortality was noted and
samples were taken. WAC-3-11-15 were 1963
Milford stock placed in Morro Bay in 1964. Mortality
was low. Oysters were necropsied and fixed on 1
May 1966. WAC-3-16-26 were Milford 1962 stock
placed in Morro Bay in 1963. They experienced 40%
mortality and were examined and fixed on 1 May
1966. Oysters WAC-4-1 through 4-5 were from
Milford 1963 stock placed in Tomales Bay, CA in
1964; low mortality was observed. They were ex-
amined and fixed on 1 May 1966.
Code FMT were 0. edulis used in an experimental
holding study at the Oxford Laboratory. Ten 2-yr-
old 0. edulis from the Milford hatchery were placed
in each of three tanks receiving 0.45 /um membrane
filtered 26°/oo seawater on 23 February 1968; pH,
temperature, salinity, and mortality were monitored
daily until 10 October 1968. Crassostrea virginica
from the Mispillian River, Delaware Bay, were
placed in each tank on 28 March. Tanks were desig-
nated A (control), B (fed tissues of moribund 0.
edulis from Pigeon Point, CA), and C (fed tissues
from Denman Island C. gigas infected with Denman
Island disease). Seawater in each tank was recir-
culated through a glass wool, charcoal, calcium flow-
through filter via an airlift system moving from C
to B to A, respectively. Oyster codes were FMT-A-
1-20, FMT-B-1-20, FMT-C-1-20, numbered as they
were fixed (species were designated at time of
fixation).
Code WAO-A were 50 specimen samples of native
oysters (0. lurida) from Yaquina Bay, OR, sampled
582
FARLEY ET AL.: MICROCELL DISEASE IN OYSTERS
monthly from February 1969 to January 1971.
Samples were coded from WA0-7A to WAO-22A.
Tissues were fixed in Zenker's acetic, Davidson's
fluid (Shaw and Battle 1957), or McDowell's fixative
(McDowell and Trump 1976) as modified by Farley
et al. (1986) (1% glutaraldehyde/4% formaldehyde;
pH 7.2-7.4 in one-half ambient seawater). Six /im
sections were stained with Harris' hematoxylin-
eosin, Ziehl's fuchsin, periodic acid Schiff reagent
(PAS) with malt diastase digestion, Feulgen picro-
methyl blue (Farley 1969), or Giemsa (Howard and
Smith 1983).
Electron Microscopy Procedures
Lesions from C. gigas from Denman Island were
fixed in 2% glutaraldehyde in pH 7.2 seawater, post-
fixed in 1% osmium tetroxide in phosphate buffer,
pH 7.2, and embedded in Epon-Araldite (Feng et al.
1971). Fifty to 100 nm sections, selected on the basis
of interference color (silver), were cut and stained
with lead citrate and uranyl acetate. Ostrea edulis
sections from the WAG material (WAC-2-19) were
deparaffinized, postfixed in 2% glutaraldehyde and
1% osmium tetroxide, plastic embedded, ultrasec-
tioned, and stained in the same fashion as the C.
gigas materials. Sections were examined in a Zeiss^
EM 9 electron microscope.
RESULTS
condition index remained high (most oysters in
medium to fat condition). Mantle recession occurred
most commonly from April through June and was
most prevalent in June. Pale digestive gland was
present in up to 24% of the oysters in the spring
and was also seen in up to 16% of the oysters in fall
samples. Shell pustules (Fig. 1), abscesses, and
ulcers as described by Quayle (1961) (Fig. 2) were
present from April through June.
Microscopical examination (Table 2) revealed
granular hemocyte infiltration of vesicular connec-
tive tissue (VCT) in most samples with high prev-
alences of microcell infections occurring sporadically
throughout the years. Abscesses (Fig. 3) in the VCT
of the mantle and gonad consisted of apparently
viable granular hemocytes at the periphery with
phagocytosis of moribund cells deeper in the lesion
and coagulative necrosis in the center. Microcell
organisms were associated with these abscesses.
The parasite (Fig. 4) was 1-3 p<m in diameter, con-
tained a small (1 f^m) Feulgen-positive nucleus, and
occurred cytozoically in VCT cells and extracellu-
larly adjacent to and within abscesses. Microcells oc-
curred in 60% of the abscesses found in histologic
sections. However, microcells were never found out-
side of abscesses.
A similar disease was discovered by F. Kern^ in
C. gigas from Hawaii. Microcell parasites were
similar in morphology and size (Figs. 5, 6) to the
Denman Island organism, but infections were more
"Denman Island Disease" Studies
Table 1 presents the seasonal prevalence of gross
features of samples collected in this study. Visual
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
^F. G. Kern, Northeast Fisheries Center Oxford Laboratory,
National Marine Fisheries Service, NOAA, Oxford, MD 21654,
pers. commun.
Table 1.— Gross pathology in oysters from Denman Island, B.C.
Number
% pale
in
% fat
% medium
% watery
% mantle
digestive
% shell
% with
Code
Date
sample
condition
condition
condition
recession
gland
pustules
abscesses
WWC-1
22 May
1967
27
88
11
0
11
0
7
7
WWC-2
5 June
1967
28
100
0
0
11
7
14
7
WWC-3
10 Apr.
1968
25
88
4
8
78
24
0
16
WWC-4
22 Apr.
1968
25
100
0
0
28
0
4
12
WWC-5
24 May
1968
25
100
0
0
0
0
0
4
WWC-6
24 June
1968
25
92
8
0
8
0
4
0
WWC-7
5 Aug.
1968
25
48
48
7
8
0
0
0
WWC-8
14 Oct.
1968
25
100
0
0
0
0
0
0
WWC-9
18 Nov.
1968
25
76
24
0
0
16
0
0
WWC-1 0
27 Jan.
1969
25
44
56
0
0
4
0
0
WWC-11
26 Feb.
1969
18
33
61
6
0
0
0
0
WWC-1 2
21 Mar.
1969
25
96
7
0
0
0
0
0
WWC-1 3
11 Apr.
1969
25
100
0
0
4
16
16
WWC-1 4
12 May
1969
25
72
28
0
0
0
12
12
WWC-1 5
20 June
1969
25
88
12
0
12
0
4
0
WWC-1 8
9 June 1980
42
95
2
2
19
0
17
17
583
FISHERY BULLETIN: VOL. 86, NO. 3
.A
100
%«i"-
m-
""■^^0
IS
#
\^
^
m
•%
19 ^ <«.
^9m
1 #
n #
*
4 '
10
n
t
10
«
^ ' • • • 10
♦ t
584
FARLEY ET AL.: MICROCELL DISEASE IN OYSTERS
Table 2.— Percentage of prevalance of lesions and parasites in Denman Island oysters.
Hyaline
Cilif^tp^
Granulocyte
Tissue
hemocyte
Microcell
\^\ 1 laico
infiltration
abscess
infiltration
Neoplasms
infection
Mytilicola
Gill Other
Code
Date
(0/0)
(%)
(%)
(%)
(%)
(%)
{%) (%)
WWC-1
22 May 1967
33
7
0
0
7
7
4 4
WWC-2
5 June1967
4
18
0
0
14
4
4 0
WWC-3
10 Apr. 1968
40
16
0
0
4
0
0 0
WWC-4
22 Apr. 1968
0
16
0
0
16
0
0 4
WWC-4
24 May 1968
8
4
0
0
4
0
0 4
WWC-6
24 June1968
32
4
24
0
0
0
0 0
WWC-7
5 Aug. 1968
0
0
0
0
0
0
0 0
WWC-8
14 Oct. 1968
36
0
0
4
0
0
0 0
WWC-9
18 Nov. 1968
12
7
36
8
0
0
0 0
WWC-10
27 Jan. 1979
20
0
8
0
0
0
0 0
WWC-11
26 Feb. 1969
4
0
0
0
0
0
0 0
WWC-1 2
21 Mar. 1969
8
4
0
0
4
8
0 0
WWC-1 3
11 Apr. 1969
20
20
8
0
12
0
4 4
WWC-14
12 May 1969
24
32
0
8
16
0
0 0
WWC-1 5
30 June1969
8
12
0
0
0
0
0 0
WWC-18
9 June1980
4
20
0
0
16
8
0 0
systemic with diffuse inflammatory infiltration of
connective tissue (Fig. 5) associated with the pres-
ence of microcell parasites (Fig. 6). Microcells were
cytozoic in hemocytes and VCT. Focal abscesses
were present but not nearly as prominent in the
Hawaiian oysters as the lesions in Denman Island
oysters.
Australian Winter Disease Studies in
Saccostrea commercialis
Australian winter disease was characterized by
pustules, ulcerations, and abscesses (Fig. 7). Seven
F'iGURE 1.— Right and left valves of Crassostrea gigas from
Denman Island, British Columbia. The darker rounded lesions
are conchiolinous shell pustules characteristic of the Denman
Island disease.
Figure 2.— Scale units in Figures 2-15 are in micrometers. Macro-
photograph showing shell pustule (A) and adjacent tissue ulcer
(B).
Figure 3.— Histologic section from Denman Island oyster. A large
abscess-type lesion is apparent in the connective tissue-gonad
region of the section. 100 x . Feulgen picromethyl blue (PPM)
stain (specimen WWC-2-7).
Figure 4.— Higher magnification photomicrograph taken at the
edge of the lesion in Figure 3. Many microcell protistan para-
sites (Mikrocytos mackini sp. n.) are evident in vesicular con-
nective tissue cells adjacent to necrotic inflammatory cells deeper
in the lesion. 1,000 x . Harris' hematoxylin and eosin (HHE)
stain (specimen WWC-2-7).
Figure 5.— Lesion caused by microcell infection in C. gigas from
Hawaii. 100 x. HHE stain (specimen S-124A-45).
Figure 6.— High magnification photomicrograph showing micro-
cell parasites intracellular in hemocytes in vesicular connective
tissue of the Hawaiian oyster. 1,000 x . HHE stain (specimen
S-124A-45).
individuals, consisting of six females and one male,
displayed ulcerations in the gonad and mantle.
Ulcerations of the gills were also common and fre-
quently occurred near the adductor muscle. Im-
paired adductor muscle contraction was character-
istic of the disease.
Histologically, the animals contained abscesses
(Fig. 8) with intense phagocytic infiltrations in the
connective tissue and varying degrees of necrosis.
The abscesses contained a small (1-2 ^m) organism
(Figs. 9, 10) which contained a nucleus >1 ^m that
was spherical with bipolar or eccentric nucleolar
structures. The size and cytozoic location of these
organisms suggest a strong similarity to other
microcell type parasites seen in other species of
oysters. Four of the six females had gonads in a
state of resorption and digestive diverticular epithe-
lium was slightly metaplastic in two of the seven
oysters.
Microcell Disease Study in Ostrea edulis
Episode 1
Microcell disease in 0. edulis from progeny from
Boothbay Harbor, ME brood stock spawned at
Milford, CT and transferred to Chincoteague Bay,
MD (Code FK).
Two of the 13 oysters fixed between August 1961
and July 1962 had heavy infection of microcells (Fig.
13) which were 1-3 ^m in diameter with a Feulgen-
positive nucleus (1 yim. diameter) and found intra-
cellularly within hemocytes. In three of the other
oysters, moderate infiltrations of hemocytes were
observed but without parasites.
585
FISHERY BULLETIN: VOL. 86, NO. 3
r
If*
1
* ♦ t
f
^ j^
%
*
♦ V*
•J, •'• »
8
• •♦v.
."»
.» jnl..
>'i'-.
1,
100 j:
#
^* idir
* JiiJf
ft
i . • «^.
tl* ♦
•
\'^'
m
9 «
10
10
10
Figure 7.— Ulceration in the mantle of Saccostrea commercialis with Australian winter disease.
Figure 8.— Large abscess-type lesion in the connective tissue-gonad region of Saccostrea commercialis (code S-41-7)
from Australia with Australian winter disease. 100 x . HHE stain (specimen S-41-7).
Figure 9. -Higher magnification photomicrograph from lesion in Figure 8. Note intracellular microcell parasites
(Mikrocytos roughleyi sp. n.) with hemocytes. 1,000 x. HHE stain (specimen S-41-7).
Figure 10.— Heavy infection of Australian winter disease; microcell parasites in the gonad of S-41-4. 1,000 x . HHE
stain.
586
FARLEY ET AL.: MICROCELL DISEASE IN OYSTERS
Episode 2
Microcell disease in 0. edulis transferred from
Milford, CT to Elkhorn Slough, CA (Code WAC).
Epizootiological gross and histopathologic data
are presented in Table 3. Clear association is ap-
parent between high mortality, emaciation (watery
condition), mantle recession, hemocytic infiltration,
and heavy microcell infection.
Infections were particularly intense in gill and GI
tract epithelia which contained dense infiltrations
of infected hemocytes (Fig. 11). Infections were
characterized by the intracellular (in hemocytes) and
extracellular presence of "microcell" organisms
(Fig. 12) similar to those seen in the Chincoteague
Bay (FK) samples.
Episode 3
Microcell disease in 0. edulis transferred from
Milford, CT to aquaria in Oxford, MD (Code FMT).
The experiment ran from 20 February 1968 until
9 October 1968. No transmission was noted. How-
ever, tank C drained accidentally on 19 March and
the filtered seawater was replaced with unfiltered
seawater from Chincoteague Bay, VA. A combined
total of 9 of 11 animals died between 22 March and
27 March and the remaining 2 live oysters were
fixed on 28 March.
The first six that died were from tank C and had
moderate to heavy cases of microcell disease with
associated hemocytic infiltration. Histologic exam-
ination of both species, either after mortalities, or
when they were sacrificed in October, failed to
reveal any more infections. The parasites and
disease characteristics were identical to the pre-
viously described cases (Fig. 14).
Microcell Disease Study in Ostrea lurida
Microcell disease in 0. lurida from Yaquina Bay,
OR (Codes WA0-7-A and WA0-12-A).
Monthly to bimonthly samples of 50 specimens
were examined grossly and for histopathology for
a 2-yr period. A 24% prevalence of microcell disease
was seen in WA0-7-A samples collected in February
1969, and 12% prevalence was seen in the WAO-
12-A samples collected in February 1970. No micro-
cell infections were seen in any of the other samples.
However, neoplasms and Mytilicola orientalis in-
fections were seen commonly throughout the study.
Microcell infections tended to be less intense than
in 0. edulis, but organisms were similar in size and
appearance to the organisms seen in flat oysters
(Fig. 15). Intracellular infections were in VCT cells
and not hemocytes.
Ultrastructural Studies
Electron microscopy revealed 0. edulis parasites
that were usually intracellular with up to three
organisms per cell. Parasites were 1-3 f^m in diam-
eter, and contained an eukaryotic nucleus about 1
/im in diameter. A crescent-shaped, peripheral
nucleolus was evident in most organisms (Fig. 16).
The cytoplasm contained numerous ribosome-like
organelles and spherical dense bodies, presumably
membrane bound, which were 90-130 nm in diam-
eter, and generally resembled the "haplosporo-
somes" (Fig. 17) described in other haplosporidan
Table 3. — Epizootiology of microcell disease in O. edulis from California.
Stock
origin
Sample
location
Date
fixed
Mortality
N
Condition
Mantle
reces-
sion
Hemocyte
infil-
tration
Microcell
infection
Code
Fat
Med
Watery
WAC-1
(1-20)
[Gilford
1964
Morro Bay
7 Dec.
1965
Heavy
20
no gross
data
18
18
WAC-2
(1-25)
Milford
1964
Pigeon Pt.
1 May
1966
Heavy
38
1
8
19
22
25
25
WAC-3
(1-10)
Milford
1964
Morro Bay
1 May
1966
Heavy
10
2
6
2
6
8
8
WAC-3
(11-15)
Milford
1963
Morro Bay
1 May
1966
Low
5
4
1
0
0
0
0
WAC-3
(16-26)
Milford
1963
Morro Bay
1 May
1966
40%
10
6
4
0
1
6
3
WAC-3
(27-31)
Milford
1965
Morro Bay
1 May
1966
0
5
5
0
0
0
0
0
WAC-4
(1-5)
Milford
1964
Tomales Bay
1 May
1966
Low
5
5
0
0
1
0
0
587
FISHERY BULLETIN: VOL. 86, NO. 3
Figure 1 1.— Low magnification view of Ostrea edulis from California (WAC-2-29). Note diffuse inflammatory infiltrate
in vesicular connective tissue. 100 x. HHE stain.
Figure 12.— Microcell infection in Ostrea edulis (specimen WAC-2-19) from population of oysters introduced into Califor-
nia waters from Milford, CT. 1,000 X. HHE stain.
Figure 13.— Microcell infection in hemocytes in Ostrea edulis transported from Milford, CT to Chincoteague Bay, VA
in 1962 (first diagnosed case oi Bonamia ostreae). 1,000 x. HHE stain (specimen FK-5-1).
Figure 14.— Microcell infection in Ostrea edulis. Transferred from Milford, CT to Oxford, MD recirculated seawater
aquarium. Cells of the vesicular connective tissue are infected. 100 x . HHE stain (specimen FMT-B-1-3).
Figure 15.— Microcell infection in vesicular connective tissue and hemocytes of Ostrea lurida from Yaquina Bay,
OR. 1,000 X. FPM stain (specimen WAO-7A-41).
588
FARLEY ET AL.: MICROCELL DISEASE IN OYSTERS
parasites (Perkins 1979). Larger "membrane"
bound structures (500 nm) consistent with the ap-
pearance of mitochondria were also present in the
cytoplasm; however, cristi could not be distin-
guished.
Ultrastructural studies were also performed on
positively diagnosed oysters (C. gigas) from Denman
Island, British Columbia (sample WWC-18, collected
9 June 1980). Microcells were always associated with
focal abscesses, but their occurrence was restricted
to the periphery. Parasites were found as cytozoic
organisms primarily in VCT cells (Fig. 18). Micro-
cells were 3-4 p<m in diameter, had nuclei 1 ^m in
diameter, and nucleoli 250-300 nm in diameter.
Nucleoli were spherical, eccentrically located within
the nucleus, but never peripheral (Fig. 19). One to
many parasites occurred within the cytoplasm of
vesicular cells. None were ever found within hemo-
cytes. Figure 20 shows an organism possibly under-
going division. The cytoplasm was densely packed
with free ribosomes and contained a variety of
organelles as follows: double membrane bound dense
bodies 50-185 nm in diameter (Fig. 19A); double
membrane bound, dumbbell-shaped structures (Figs.
20, 21h), approximately 37 x 18 nm to 85 x 260
nm; and dense bodies 40-45 nm in diameter that
appeared to be membrane bound, and a suggestion
of six- and five-side angularity (Figs. 19, 20, 21v).
Endoplasmic reticulum was extremely sparse if
present at all. The plasma membrane complex con-
sisted of possibly two membranes with the external
membrane containing dense material. An electron-
lucent zone was present around the cell, suggestive
of a glycocalyx.
Taxonomic Descriptions
The information acquired on these diseases and
the organisms associated with them and outlined
previously allows us to propose taxonomic descrip-
tions of them.
Mikrocytos g. n. (Protista incerta sedis) Gittle cell)—
Definitive life cycle stages that would permit
higher classification of this protistan parasite have
not been observed. Named after the term micro-
cell as coined by the late John G. Mackin. Small
(1-4 ^m), unicellular, protistan, cytozoic parasite
normally infecting VCT cells of oysters. Always
associated with abscess-type focal inflammatory
lesions.
Type species - Mikrocytos mackini sp. n.
Bonamia, the other closely similar genus, infects
hemocytes of ostreid oysters only and is associated
with systemic non-abscess type disease manifes-
tations.
Mikrocytos mackini sp. n. — Named in honor of the
late John G. Mackin who discovered this parasite
in the early 1960s.
Type specimen - A 6 /im thick hematoxyhn-and-
eosin-stained histologic section of an infected
oyster, C. gigas (WWC-2-7), was deposited in the
Registry of Marine Pathology, Northeast Fish-
eries Center, Oxford, MD 21654.
Host - Crassostrea gigas
Type locale - Henry Bay, Denman Island, British
Columbia, Canada.
Range - Occurrence confined to above site. (A closely
related candidate for inclusion within this species
was found in C. gigas from Hawaii.)
Morphologic characteristics - Small, 1-4 nm intra-
cellular parasites of VCT cells; infections always
associated with focal inflammatory tissue ab-
scesses. Parasites are unicellular and contain a
small 1 p/m nucleus that has an eccentric nucleolus
250-300 nm in diameter. The cytoplasm contains
dense, double membrane bound, dumbbell-shaped
haplosporosome-type organelles 50-180 nm in
diameter and 40-45 nm membrane bound five-
and/or six-sided dense bodies.
Mikrocytos roughleyi sp. n. — Named in honor of
T. C. Roughley who published the initial study of
the Australian winter disease in the 1920s.
Type specimen - A 6 fim thick hematoxylin-and-
eosin-stained section of an infected oyster, S. com-
mercialis (S-41-7), was deposited in the Registry
of Marine Pathology, Northeast Fisheries Center,
Oxford, MD 21654.
Host - Saccostrea commercialis
Type locale - Georges River, Woolooware Bay, New
South Wales, Australia.
Range - Known only from the above location and
other high salinity estuaries in this region of New
South Wales.
Morphologic characteristics - Infections occur in
hemocytes and are associated with focal abscess-
type lesions in the gill, connective, and gonadal
tissues. Organisms are small 1-3 ptm cells that con-
tain an eccentric nucleus and a cytoplasmic
vacuole. Ultrastructural characteristics are not
known.
Comparisons — Mikrocytos g. n. is always associated
with focal abscesses and occurs in crassostreid
oysters. Bonamia is always associated with general-
ized infections and only occurs in ostreid oysters.
589
is:
n
fi
B
FISHERY BULLETIN: VOL. 86, NO. 3
y
"^^
16
S?'5- ...
200
17
-vs^ 100
2000
■VM
B
D
A'
18
19
200
i^
\
20
200
21
100
590
FARLEY ET AL.: MICROCELL DISEASE IN OYSTERS
Mikrocytos mackini sp. n., which has eccentric
nucleoH, and M. roughleyi sp. n., which has a cyto-
plasmic vacuole and a nucleus that is displaced to
the periphery of the cell, are morphologically distinct
from B. ostreae (the only closely related species). All
three species occur in separate host species and all
appear to be host specific.
DISCUSSION
Denman Island Disease
Quayle's original study (1961) documented heavy
mortality (40%) clearly associated with surface
tissue pustules in C. gigas in an area of British
Columbia in May. Studies of C. gigas by Mackin
resulted in the discovery of the "microcell" organ-
ism and its association with tissue abscesses. Sub-
sequent histological examinations done by the senior
author, in cooperation with D. Quayle and N.
Bourne, confirmed the association of mortality,
pustules, tissue abscesses, and microcell infections
that have continued to occur during each May and
June to the present time.
Histopathologically, this disease (caused by Mikro-
cytos mackini) is characterized by acute inflam-
matory abscesses which remain focal until the oyster
dies or resolution occurs. While microcell organisms
are not always found in abscesses, they are never
found in oysters that do not have abscesses, in-
dicating at least an associational relationship.
Electron microscopy has demonstrated only one
stage of a small protistan organism that contains
organelles resembling haplosporosomes. No clear
demonstration of mitochondria has been accom-
plished. The haplosporosome-like organelles often
IGURE 16.— Scale units in Figures 16-21 are in nanometers. Elec-
tron micrograph of deparaffinized Ostrea edulis from Califor-
nia. Two intracellular microcells showing prominent nuclei with
peripheral nucleoli. 36,000 x.
IGURE 17.— Higher magnification of California microcells show-
ing haplosporosome-like bodies (A) and a probable mitochondria
(B). 108,000 X.
IGURE 18.— Electron micrograph of several microcells in vesicular
connective tissue of Crassostrea gigas (A) from Denman Island,
British Columbia. Probable degenerate microcells can also be
seen in the cytoplasm of phagocytic hemocytes (B). 7,320 x .
IGURE 19.— Electron "micrograph" of microcell from Denman
Island Crassostrea gigas. Note the eccentric nucleolus (A), myelin
bodies (B), haplosporosome-like bodies (C), and virus-like entities
(D). 36,900 x.
IGURE 20.— Electron micrograph of Crassostrea gigas microcell
with structure suggestive of division. 36,900 x.
IGURE 21.— Higher magnification of electron micrograph show-
ing haplosporosomes and virus-like dense bodies. 108,000 x.
tend to be elongated and contain layers of mem-
branes. Internal structure of these organelles is not
nearly as dense as that seen in Bonamia or other
haplosporidans.
The small 45 nm dense bodies also seen in the
cytoplasm have characteristics suggestive of virus
structure; namely, uniform size, abundant occur-
rence in the cytoplasm only, and a suggestion of
icosahedral symmetry. The cytoplasmic occurrence,
46 nm size, icosahedral symmetry, and the presence
of an envelope are characteristics of the family
Togaviridae. The lack of paracrystalline arrays,
strategy of development, empty capsids, and extra-
cellular occurrence prevent conclusive identification
of these particles as virus at this time, and these may
prove to be an exclusive organelle that is a char-
acteristic of this protistan group. The presence of
a lytic virus in the parasite could also explain the
self -limiting nature of the focal abscesses character-
istic of the Denman Island disease.
Australian Winter Disease
Since Roughley described this disease in an Aus-
tralian S. commercialis in 1926, little progress has
been made toward identification of the etiologic
agent. Careful examination of the tissues of affected
animals, collected by Peter Wolf, has revealed ap-
parent small cytologic and histozoic organisms
associated with abscesses. These organisms have
features such as size, morphology, and tissue loca-
tion remarkably similar to organisms present in
other oyster microcell diseases. Peter Wolf (unpubl.
data) has stated that this disease thrives in high
(30-357oo) salinity (and is unknown in lower salin-
ities); the incubation period is about 2V2 months and
mortality does not occur in animals less than 3 years
old. The occurrence of shell lesions and focal tissue
abscesses appears to be common features in Sac-
costrea and Crassostrea. This may indicate a degree
of resistance characterized by the ability of the
animal to isolate parasites in a focal lesion and to
eliminate them by either rupture of the abscesses
or diapedesis through the mantle epithelium to the
shell surface, with subsequent calcification via shell
pustule formation, or it may indicate a host parasite
relationship response. This mechanism was reported
in oysters that had acquired resistance to Haplo-
sporidium nelsoni (Farley 1968).
Kern (fn. 6) found microcell infections in C. gigas
from Hawaii (Fig. 6) that were not always asso-
ciated with focal abscesses. General systemic
infection and inflammatory infiltration were noted
in these cases (Fig. 5), but the organisms appear
591
FISHERY BULLETIN: VOL. 86, NO. 3
to be identical to the Denman Island microcell
organisms.
VCT while B. ostreae is primarily a disease of
hemocytes.
Microcell Infections in Ostrea edulis
Microcell infections were first seen in 0. edulis
in moribund oysters which had been transferred
from Milford, CT to Chincoteague Bay, VA in 1962.
Two animals from the FK sample had developed
clinical manifestations of the disease. Other cases
appeared in animals transferred from Milford to
California, and Milford to Oxford. Morphology of the
infectious organisms was identical in all of these
episodes and the histopathology always consisted of
acute inflammatory infiltration and systemic in-
volvement. All episodes were associated with move-
ment of oyster stocks originating in Milford, and all
experienced severe mortality. No differences were
noted between any of the 0. edulis epizootics in
regard to morphology of the organisms. Compari-
son of the American 0. edulis infections vdth tissues
of French oysters experiencing B. ostreae infections
(Pichot et al. 1979; Balouet et al. 1983) revealed no
morphologic or histologic differences at the light or
electron microscope level. Indeed, ultrastructural
comparisons demonstrate close similarities. Size
comparisons of organisms are identical and the
nucleus contains a peripheral nucleolus, and iden-
tical haplosporosome-like organelles are present in
the cytoplasm. The major difference is in epizootic
occurrences. The French epizootic occurred in
natural or feral populations of European flat
oysters. Since introduction of oysters to French sites
from locations outside of France was a common
event in the past, the source of the index case may
have originated from an introduction of oysters from
Elkhorn Slough, CA in the late 1970s (Elston et al.
1986). Contagious spread (Tige et al. 1981) is well
documented in many locations in France and also
the Netherlands. With the exception of the estab-
lished breeding populations of flat oysters in the cen-
tral Maine coast region, no natural or feral popu-
lations of 0. edulis exist in the United States. The
State of Maine carefully controls imports into the
state; the disease has not been established in this
population. The discovery of microcell infections in
0. lurida in Oregon suggests that this may be a
naturally occurring disease in that species. Infec-
tion intensity and prevalences suggest that some
animals may die from the disease. The disease ap-
pears to be enzootic in the Oregon location. The lack
of ultrastructiiral studies prevents close comparison
of the 0. lurida disease with the 0. edulis disease.
However, the disease in 0. lurida tended to infect
CONCLUSIONS
There is a complex of oyster diseases caused by
a group of protistan parasites of several species.
These small intracellular and extracellular organ-
isms designated originally as microcells have been
found in association with serious disease in two
species of Crassostrea and two species of Ostrea.
It appears that the disease in C. gigas and S. com-
mercialis, whOe exhibiting some similarities in types
of lesions, are probably caused by different species
of microcell type parasites.
A new genus, Mikrocytos g. n., and two new
species have been described for the organisms caus-
ing disease in oysters: Mikrocytos mackini sp. n. in
C. gigas from British Columbia, Canada, and Mikro-
cytos roughleyi sp. n. in S. commercialis from
Australia.
Disease that has struck 0. edulis in France is iden-
tical to the microcell disease seen in 0. edulis in
three episodes in the United States. The organism
causing the disease in 0. edulis is Bonamia ostrea£
and is clearly different from the microcell organism
found in C. gigas in British Columbia and Hawaii.
Finally, additional ultrastructural studies are needed
for more complete characterization of the organisms
from 0. lurida and S. commercialis.
ACKNOWLEDGMENTS
We thank Cecelia Smith, Dorothy Howard, and
Gretchen Roe for preparation of histologic ma-
terials; Jane Wade for preparation of the ultra-
structural material; and Muriel McNeils, Karen
Hayman, and Jane Swann for manuscript prepara-
tion. The senior author would like to acknowledge
the help of the late John Mackin for his expert ad-
vice through the years; Daniel B. Quayle and Neil
Bourne for assistance with samples of oysters from
British Columbia; Fred Kern for allowing us to use
his Hawaiian material; and Albert K. Sparks and
Inke Sunila for critical review of the manuscript
without implying agreement with interpretations
herein. Partial support from the Department of
Energy under Contract DE-AC06-76RLO 1830 to
Battelle Memorial Institute is acknowledged.
LITERATURE CITED
Balouet, G., M. Poder, and A. Cahour.
1983. Haemocytic parasitosis: morphology and pathology of
592
FARLEY ET AL.: MICROCELL DISEASE IN OYSTERS
lesions in the French flat oyster, Ostrea edulis L. Aqua-
culture 34:1-14.
Elston, R. a., C. a. Farley, and M. L. Kent.
1986. Occurrence and significance of bonamiasis in European
flat oysters, Ostrea edulis, in North America. Dis. Aquat.
Org. 2:49-54.
Farley, C. A.
1968. Minchinia nelsoni (Haplosporida) disease syndrome in
the American oyster, Crassostrea virginica. J. Protozool.
15:585-599.
1969. Probable neoplastic disease of the hematopoietic
system in oysters (Crassostrea virginica and Crassostrea
gigas). Natl. Cancer Inst. Monogr. 31:541-555.
Farley, C. A., S. V. Otto, and C. L. Reinisch.
1986. New occurrence of epizootic sarcoma in Chesapeake
Bay soft-shell clams, Mya arenaria. Fish. Bull., U.S.
84:851-857.
Feng, S. Y., C. N. Burke, and L. H. Kharallah.
1971. Light and electron microscopy of the leukocytes of
Crassostrea virginica (Mollusca: Pelecypoda). Z. ZeUforsch.
120:222-245.
Howard, D. H., and C. S. Smith.
1983. Histological techniques for marine bivalve moUusks.
U.S. Dep. Comm., NOAA Tech. Memo NMFS-F/NEC-25,
97 p.
Katkansky, S. C, W. a. Dahlstrom, and R. W. Warner.
1969. Observations on survival and growth of the European
flat oyster, Ostrea edulis in California. Calif. Fish Game
55:69-74.
McDowell, E. M., and B. F. Trump.
1976. Histologic fixatives suitable for diagnostic light and
electron microscopy. Arch. Pathol. Lab. Med. 100:405-413.
Perkins, F. 0.
1979. Cell structure of shellfish pathogens and hyperparasites
in the genera Minchinia, Urospondium, Haplosporidium,
and Marteilia—taxonomic implications. Mar. Fish. Rev.
41(l):25-37.
PiCHOT, Y., M. CoMPS, G. TiGE, H. Grizel, and M. a. Rabouin.
1979. Recherches sur Bonamia ostreae gen. n., sp. n., para-
site nouveau de I'huitre plate Ostrea edulis L. Rev. Trav.
Inst. Peches Marit. 43:131-140.
QUAYLE, D. B.
1961. Denman Island oyster disease and mortality, 1960.
Fish Res. Board Can. Ms. Rep. Ser., No. 713, p. 1-9.
ROUGHLEY, T. C.
1926. An investigation of the cause of an oyster mortality on
the Georges River, New South Wales, 1924-25. Proc. Linn.
Soc. N.S.W. 51:446-491 (-i- plates).
Shaw, B. L., and H. I. Battle.
1957. The gross and microscopic anatomy of the digestive
tract of the oyster, Crassostrea virginica (Gmelin). Can.
J. Zool. 35:325-347.
TiGE, G., H. Grizel, A. G. Martin, A. Langlade, and M. A.
Rabouin.
1981. Situation epidemiologique consecutive a la presence du
parasite Bonamia ostreae en Britagne. Evolution au cours
de I'annee. 1980. Sci. Peches Bull. Inst. Peches Marit.
315:13-20.
593
NOTES
CONJOINED TWIN ADULT SHRIMP
(DECAPODA: PENAEIDAE)
A two-headed roughback shrimp, Trachypenaeiis
similis (Smith), caught at the entrance to Galveston
Bay, TX 10 May 1987 by Harold Fraley, together
with a color photograph of the specimen taken sub-
sequent to capture, was sent to me for identifica-
tion, morphological examination, and deposit in the
crustacean collection of the National Museum of
Natural History (USNM 234419), Smithsonian
Institution.
The specimen (Fig. 1) is composed of two cephalo-
thoraxes (heads) perfectly aligned with the median
sagittal plane and conjoined posteriorly to an un-
paired, normally segmented abdomen. The cephalo-
thoraxes and abdomen were disarticulated when the
preserved specimen reached me in October 1987, but
were restored to normal position easily with aid of
the photograph as a guide. The lower cephalothorax
is that of an adult female, carapace length including
rostrum 32 mm, short carapace length (orbital
margin to posteromedian edge of carapace) 21 mm;
respective measurements for the upper carapace are
34.4 mm and 21.7 mm. The abdomen is flexed and
twisted to the left, and the fourth and fifth segments
are damaged, hence its length cannot be measured
accurately. Comparison of the specimen with sper-
matophore bearing females of the species in the
USNM crustacean collection indicates that it is adult
in size, about 85 mm total length.
Shrimps are sometimes caught and preserved
while in the act of molting. In that event the cara-
pace being molted tends to be loosened at the
thoraco-abdominal juncture so that its posterior end
can be flipped dorsally and away, freeing the husk-
like old carapace from the underlying soft new cara-
pace. A first impression that this specimen was
caught and preserved while in the act of molting was
not borne out by the structures observed.
The integument of each carapace is firm, as is that
of the other exoskeletal parts. Both carapaces are
similar in shape and structure, including the part
of the lower carapace that is hidden by the upper
one. It is noteworthy that the upper carapace is
larger than the lower, just the opposite of what
would be expected if the upper one represented an
ecdysial discard. Eyes in both heads have normal
dark corneal pigment, though the corneal surfaces
are shriveled by preservation. Antennules, anten-
FlGURE 1.— Conjoined twin Trachypenaeus similis in diagrammatic lateral view, distal parts of appendages inten-
tionally deleted, except for those of uropods. Scale = 1 cm.
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
595
nae, mandibles and associated palps, first and sec-
ond maxillae, and first, second, and third maxilli-
peds are present and apparently normal on both
heads, though the antennal flagella are broken
and the right third maxilliped has apparently been
lost from the upper head. Similarity in the ceph-
alothoraxes is limited to features pointed out
above.
The lower cephalothorax has normal appendages
and internal organs, or traces of them, except for
some broken and missing articles on the pereopods.
The heart, hepatopancreas, and gonads apparently
disintegrated during the interval of time between
capture and fixation, hence lost all traces of their
conformation in life. Fluids from the disrupted cir-
culatory system were fixed as irregular clots. The
mouth, esophagus, stomach, and a fragment of the
gut are present, though the latter is connected
neither to the pyloric stomach nor to the sector of
intestine that courses through the abdomen to the
anus. However, one can visualize that the path of
the intestine in the lower cephalothorax was func-
tionally normal before it was disrupted by break-
down of the other internal organs that surrounded
it in life.
The upper cephalothorax lacks a mouth, the ster-
nal plate between the mouthparts is not perforated,
and there is no esophagus, stomach, or fragment of
intestine. It seems therefore that the upper cepha-
lothorax, though equipped with normal head append-
ages, could not function in feeding. Moreover, the
rear part of the upper cephalothorax fitted over the
rear part of the lower one like a firmly pulled down
cap, with its branchiostegites deeply overlapping
those of the lower one. In this arrangement the
posterior part of the lower cephalothorax filled
the space that would normally have been occupied
by pereopods, external reproductive structures,
thoracic endophragmal system and body wall,
gills, and internal organs of the upper cepha-
lothorax. As a result there was little or no room for
development of these structures in the upper
cephalothorax, although there may have been a
heart. The membrane that lined the branchiostegites
and body wall of the upper cephalothorax seems to
have extended backward from the region of the
cervical groove to merge with its counterpart in the
posterior region of the lower cephalothorax, and
with the normal integumental lining of the abdomen
in order to have maintained confluence in the blood
sinuses.
The abdomen, though crushed at the level of the
fourth and fifth segments, bears normal pleopods,
uropods, and telson. The anterior end of the ab-
dominal muscle mass is preserved in a shape that
fits the posterior end of both cephalothoraxes, but
the main connection extended into the functional
lower one in which complete organ systems were
located.
There is a large literature treating malformations
of decapod crustaceans, primarily lobsters, fresh-
water crayfishes, and crabs, but there is little pub-
lished information of this sort on shrimps (Bateson
1894; Johnson 1968; Johnson and Chapman 1969;
Pauley 1974), aside from the subject of disease which
is not at issue here (Couch 1978). The most ex-
haustive account is that of Bateson who, along with
many others before and after, discussed duplication
of parts, intersexes, and malformations that occur
during molting, Perez Farfante (1980), for example,
noted anomalous intersexes in the Indo-West Pacific
needle shrimp, Penaeopsis rectacuta (Bate). The
majority of these accounts treat malformed limbs
or their parts (for a well-illustrated example see
Shuster et al. 1963). Fewer studies are concerned
with teratology.
Monsters with fused double cephalothoraxes,
though rare, have long been known among larvae
of the lobsters Homarus americanus H. Milne
Edwards and H. gammartis (Linnaeus) (Herrick
1896, 1911). Ryder (1886) noted four forms of con-
joined twins in larval American lobsters: lateral
fusion of cephalothoraxes that demonstrated
absence of eyes, possession of a single median eye
or paired eyes representing the right eye of the right
larva and left eye of the left larva, while the ab-
domens of each type were separate and divergent
at a wide angle, and cephalothoraxes of two embryos
fused together along their dorsal surfaces, with full
complement of eyes, appendages, and separate ab-
domens, but with internal organ systems fused.
Ryder attributed all of these twinnings to fusion
coincident with the process of gastrulation and
gradual formation of the embryos. Herrick (1896,
1911) discussed and figured some of these cases also
but thought that fusion came later in development
than gastrulation.
I have found no account of conjoined twinning in
shrimps, and no report of twinning that parallels the
case presented here. What is amazing is that an
animal so bizarre could molt at all, let alone progress
through a series of molts to attain mature size.
Whether the deformity resulted from embryonic
malformation or from subsequent injury cannot now
be determined, although angle of divergence and
median sagittal alignment of the cephalothoraxes
suggests that the malformation resulted from aber-
rant molting.
596
Acknowledgments
I thank R. J. Zimmerman and K. N. Baxter,
Southeast Fisheries Center Galveston Laboratory,
National Marine Fisheries Service, for directing the
specimen to me. J. C. Harshbarger aided with
sources of information on pathology, and I. Perez
Farfante with B. B. Collette critically reviewed the
manuscript. Keiko Hiratsuka Moore rendered the
illustration.
Literature Cited
Bateson, W.
1894. Materials for the study of variation treated with
especial regard to discontinuity in the origin of species.
MacMillan and Co., London and New York, xvi + 598 p.
Couch, J. A.
1978. Diseases, parasites, and toxic responses of commercial
penaeid shrimps of the Gulf of Mexico and south Atlantic
coasts of North America. Fish. Bull., U.S. 76:1-44.
Herrick, F. H.
1896. The American lobster. A study of its habits and devel-
opment. Bull. U.S. Fish Comm. 15(for 1895):l-252, pis.
A-J, 1-54.
1911. Natural history of the American lobster. Bull. U.S.
Bur. Fish. 29(for 1909):147-408, pis. 28-32.
Johnson, P. T.
1968. An annotated bibliography of pathology in inverte-
brates other than insects. Burgess Pub. Co., Minneapolis,
MN, xiii + 322 p.
Johnson, P. T., and F. A. Chapman.
1969. An annotated bibliography of pathology in inverte-
brates other than insects. Suppl. Cent. Pathobiol., Univ.
Calif. Irvine, Misc. Publ. No. l:i-ii, 1-76.
Pauley, G. B.
1974. A bibliography of pathology in invertebrates other than
insects from 1969-1972. NOAA-NMFS Middle Atlantic
Coastal Fisheries Center Pathology Investigations, Oxford,
MD, Informal Rep. No. 24:i-ii, 1-122.
PEREZ Farfante, I.
1980. Revision of the penaeid shrimp genus Penaeopsis
(Crustacea: Decapoda). Fish. Bull., U.S. 77:721-763.
Ryder, J. A.
1886. The monstrosities observed amongst recently hatched
lobsters. Am. Nat. 20(8):742-743.
Shuster, C. N., Jr., D. H. B. Ulmer, Jr., and W. A. Van Engel.
1963. A commentary on claw deformities in the blue crab.
Univ. Del. Estuarine Bull. 7(2&3):15-23.
Austin B. Williams
Systematics Laboratory
National Marine Fisheries Service, NOAA
National Museum of Natural History
Smithsonian Institution
Washington, DC 20560
NOTE ON MUSCLE GLYCOGEN AS
AN INDICATOR OF SPAWNING POTENTIAL
IN THE SEA SCALLOP,
PLACOPECTEN MAGELLANICUS
During the reproductive cycle of the Atlantic sea
scallop, Placopecten magellanicus, glycogen levels
rise and fall in the hemolymph (Thompson 1977) and
in the adductor muscle (Robinson et al. 1981; Gould
1983), reflecting the buildup of glycogen reserves
in the muscle and their later transfer to the gonad.
Muscle glycogen normally rises to a yearly peak in
spring after the phytoplankton blooms, then is trans-
ferred to the gonad for gamete differentiation and
maturation (Robinson et al. 1981). The glycogen
transfer is followed by an increase in size of the
maturing gonad and a loss of muscle weight (Gould
1983). During the autumnal algal blooms, glycogen
levels in the muscle rise again slightly and drop
thereafter to an annual low during the winter
months, when the small energy reserves are used
for basal maintenance and to initiate gametogenesis.
Glycogen reserves from the muscle and lipid
reserves from the digestive gland are the major
sources of stored energy supplied to the scallop
gonad. High spring glycogen levels most drama-
tically indicate the degree of buildup of energy
stores used to fuel gamete differentiation and
maturation, whereas low winter muscle glycogen
levels correspond to the postspawning exhaustion
of reserves. Winter values higher than the normal
range for any given population, therefore, could in-
dicate an unusually large and extended period of
nutrient availability, but more probably would sug-
gest resorption of gametes.
We suggest, therefore, that the spring peak and
the winter ebb of muscle glycogen be used as meas-
ures of the relative spawning potential and spawn-
ing success, respectively, for Placopecten. Sampling
during these two seasons may readily provide infor-
mation on the recruitment contribution of different
scallop populations.
Timing of the seasonal high and low values for this
metabolic parameter can vary by several weeks from
year to year, reflecting the timing and intensity of
phytoplankton blooms (themselves dependent on
other environmental variables), and the time and
degree of success in spawning. To obtain a practical
data base for this major measure of seasonal energy
reserves, therefore, we sampled a single bed of sea
scallops off Asbury Park, NJ, on a year-round
monthly basis for 3V2 years. In examining mean an-
nual high and low muscle glycogen values for these
scallops, data were averaged for animal collections
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
597
from mid- April through June each year to arrive at
a general value for the spring buildup and peak. Al-
though such values do not indicate the maximal
glycogen values, which may be reached either gradu-
ally or quickly in any of those months, an average
better enables year-to-year comparisons. Data for
collections from late November through February
were similarly averaged to obtain a value for the
postspawning winter period of low muscle glycogen.
We used this same parameter for specimens col-
lected from random sites in the Gulf of Maine, and
found a different seasonal pattern in scallops from
depths greater than ca. 110 m. Because these deep-
water sea scallops came from many different sites
in the Gulf of Maine, data were averaged for each
collection date for each site. This report presents
our data for spring and winter adductor muscle
glycogen in a single subtital sea scallop population
for the years 1981-84, and in deepwater sea scallops
from the Gulf of Maine for 1980-82.
Materials and Methods
Asbury Park sea scallops were collected by trawl
from a site 31m deep on the southern shelf of the
Hudson River Canyon off Asbury Park, approx-
imately 37 km NNE of Manasquan Inlet (ca. 40° 13'
X 73° 47'). Collections were made at monthly inter-
vals from spring 1981 through late July 1984 and
during two intensive weekly sampling periods from
early May through mid- June in 1983 and 1984, to
monitor the spring buildup of muscle glycogen. For
some months, particularly in the spring of 1982,
collections were not available. Each collection
comprised 6 males and 6 females (shell height
95-110 mm). The sea scallops were held overnight
in 5°C aerated seawater at the Northeast Fisheries
Center's (NEFC) laboratory at Sandy Hook, NJ, and
transported the following day in a cooler to the
Milford, CT, NEFC laboratory. For transport, the
sea scallops were placed on top of paper toweling
that had been soaked in seawater, then wrung out
and layered over ice enclosed in a sealed plastic bag.
The animals were dissected the same day, and all
tissue specimens were stored at -80°C until test-
ing, (jonad volumes were also noted. Deepwater sea
scallops were dissected on shipboard immediately
after collection by trawl, and the muscle tissue held
at -40°C while at sea, then transferred on dry ice
to the -80°C freezer in Milford. Because we relied
on volunteer help for many of these collections,
gonad data were not always available for shipboard
samples of adductor muscle from deepwater sea
scallops.
Muscle dissection, tissue preparation, and the pro-
cedure for glycogen analysis are described in detail
elsewhere (Gould et al. 1985); glycogen levels are
presented as /ig of glucose per gram of wet tissue
(pig g"0- Because there were no detectable differ-
ences between sexes for muscle glycogen levels,
data for males and females were combined.
Results and Discussion
Asbury Park Sea Scallops
In the spring of 1981, the Asbury Park sea scallops
had muscle glycogen levels averaging higher than
2,000 fig g"^ (Table 1). Such levels are not uncom-
mon in well-fed scallop populations, as observed
during several years of monitoring activity on the
continental shelf off New England and the mid-
Atlantic states (Gould 1981, 1983) during the
NEFC's Ocean Pulse/Northeast Monitoring Pro-
gram (NEMP) and the NEFC's Resource Assess-
ment surveys. The mean annual low levels in the
Table 1.— Seasonal high and low levels in adductor muscle glycogen for both males and
females in a single population of sea scallops off Asbury Park, NJ. Values were averaged
for mean seasonal highs during and after spring phytoplankton blooms (April, May, June) and
for mean seasonal lows after spawning (December, January, February). Gonad volumes are
also shown for the same time periods.
Muscle glycogen
Gonad volume
Sample
{>igg"
')
Am
nual ratio
(mL)
Year
Season
N
X
SE
spring:winter
X
SE
1981
Spring
34
2,217
283
no
data
1982
Spring
24
544
32
5.42
0.74
1983
Spring
84
610
15
6.23
0.29
1984
Spring
72
254
12
4.51
0.24
1981-82
Winter
22
203
18
10.9
6.50
0.47
1982-83
Winter
24
226
14
2.4
3.00
0.46
1983-84
Winter
12
490
57
1.2
2.75
0.39
598
Asbury Park sea scallops the following winter were
within the normal range (200-300 /ig g^^) observed
for sea scallop populations over those same years
of monitoring.
In 1982 and 1983, the Asbury Park sea scallops
were apparently adequately fed during the spring
months, although glycogen levels were less than one
third of those seen in 1981 and had no discernible
peak. Muscle glycogen in the winter months of 1983
and 1984, however, was sufficiently high in both
sexes (as compared with postspawning levels for this
population, and with mean winter levels in other
populations) to prompt the suspicion that gamete
resorption had taken place, and that the 1983
spawning season had not been very successful.
Moreover, although an intensive weekly sampling
was performed in May and June 1984, we did not
observe the normal seasonal increase in muscle
glycogen; instead, the values resembled those of a
typical winter low. When spring values for each year
are compared with the subsequent winter's post-
spawning values (Table 1), a picture emerges of
declining nutritional status from 1981 to the end of
the study.
It is possible, of course, that the spring values in
1982 and 1983 were more typical for this popula-
tion and that spring values for 1981 may have been
unusually high. The latter phenomenon could have
been the result of especially heavy phytoplankton
blooms, or of oceanic currents favorable to the bot-
tom settlement of planktonic nutrients. Certainly
the most important single variable is nutrient
availability.
The 1984 glycogen levels indicated either that
little or no food was available to the sea scallops (at
30 m), or that they were not assimilating normally
any food that was available. This phenomenon has
yet to be explained satisfactorily, because the phyto-
plankton bloom in the area that year was extensive
(J. O'Reillyi). Steven K. Cook^ had suggested that
some oceanographic event, such as the inshore intru-
sion of an offshore water mass known as the "cold
pool" (e.g. Hopkins and Garfield 1979) may have
caused an unusually early formation of a thermo-
cline, one that effectively prevented settlement of
planktonic detritus to the bottom. Whatever the
reason, the Asbury Park sea scallops showed dimin-
ishing glycogen reserves for spawning from 1981
to the end of the study in 1984. Either planktonic
nutrients were not reaching that population, or food
was available but the sea scallops were not feeding
or assimilating properly.
If the latter should be the case, it is perhaps rele-
vant that the Asbury Park sea scallop population
lies approximately 24 km downstream from the
Christiaensen Basin, where general current patterns
are southwesterly. Several active dumpsites are
located in the Christiaensen Basin, including those
for New York's sewage sludge and dredge spoils,
where copper is a major contaminant (see Steimle
et al. 1982). Moreover, as little as 10 ^g L"^ copper
in the water column has been shown to interfere
with gamete production and maturation (resorbing
gametes) and probably also with feeding or nutri-
ent assimilation in Placopecten magellanicus (Gould
et al. 1985, 1988). Chemical analysis of tissues from
these same Asbury Park sea scallops is under way,
to determine whether metal levels were sufficient-
ly elevated to induce this effect.
Deepwater Scallops
A data pattern for muscle glycogen similar to that
seen in the Asbury Park sea scallops for 1984 has
been observed in deepwater sea scallops taken from
various sites in the Gulf of Maine (Table 2). These
sea scallop beds were sampled randomly during the
NEFC trawl survey cruises, and one fixed station
was sampled seasonally during NEFC NEMP
cruises (Gould 1981, 1983). Sea scallops taken from
waters >110 m deep routinely showed very low
glycogen levels throughout the year, the highest an-
nual levels being reached in December. In the fall,
vertical mixing of the subsurface and intermediate
water increases to as deep as 150 m (McLellan et
al. 1953; Colton 1968; Hopkins and Garfield 1979;
Mountain and Jessen 1987), with the disappearance
of any strong thermocline. In a recent comparison
of food resources in shallow (20 m) and in deepwater
(180 m) populations, Shumway et al. (1987) observed
that a number of intact planktonic algal species
reached the deepwater sea scallops after the fall
phytoplankton bloom; this late annual food source
"may provide just enough energy to sustain the
population." On the whole, however, nutrient avail-
ability is very low at such depths, as indicated by
the absence of chlorophyll in the deeper water
column (J. O'Reilly^).
Deepwater sea scallops are visibly undernourished
U. O'Reilly, Northeast Fisheries Center Sandy Hook Laboratory,
National Marine Fisheries Service, NOAA, P.O. Box 428, High-
lands, NJ 07732, pers. commun. October 1984.
^Steven K. Cook, National Weather Service, 2980 Pacific High-
way, San Diego, CA 92101, pers. commun. March 1984.
^J. O'Reilly, Northeast Fisheries Center Sandy Hook Laboratory,
National Marine Fisheries Service, NOAA, P.O. Box 428, High-
land, NJ 07732, pers. commun. May 1985.
599
Table 2.— Mean seasonal high and low levels of adductor muscle glycogen in deepwater sea
scallop populations of the Gulf of Maine.
Station
coordinates
Lat.
Long.
Date of
collection
Depth
(m)
Bottom
temperature
CC)
Sample
N
Mean
gonad
volume
(mL)
Muscle
glycogen
0^9 g"^)
43'='21'
'69O03'
01/29/82
160
7.0
12
2.25
153
43023'
69055'
04/23/80
168
4.2
5
303
43°25'
69022
04/23
180
4.9
5
—
164
43" 16'
6904O'
04/23
159
4.0
5
—
248
43''20'
69O03'
05/07
155
5.0
5
—
282
43021'
67O06'
05/07
155
5.0
4
—
274
42°49'
68049'
05/10/81
202
6.0
2
145
43°07'
68042'
05/10
174
5.5
8
—
199
43°30'
6903O'
05/24
144
4.2
12
—
178
43»33'
69=07'
05/24
152
8.4
4
—
180
44°21'
67021'
08/03/82
137
10.4
12
4.70
178
42°56'
70O16'
08/16/80
155
4.9
8
—
226
43°26'
69057'
08/16
134
5.3
9
—
120
43°21'
^69003'
08/24/82
156
5.5
(at 100 m)
12
8.67
361
43021 '
'69O03'
09/07/80
155
6.0
18
9.09
133
43017'
690 18'
11/04/80
161
6.0
10
—
437
430 18'
69O03'
11/04
167
7.2
10
—
330
430OO'
6902O'
11/04
181
5.8
10
—
318
43037'
69032'
11/04
113
8.1
5
—
582
43021 '
^69003'
12/06/82
160
7.5
12
2.42
587
43025'
69022'
12/06
162
10.0
12
3.75
495
430 18'
70=03'
12/07
158
7.2
12
2.67
409
42059'
70=10'
12/07
183
7.8
13
2.58
424
'Deepwater station off Toothaker Ridge that was sampled whenever possible; other stations were selected
randomly during resource survey cruises.
(thin shells and small adductor muscles), lack the
necessary glycogen reserves for successful spawn-
ing, and very probably resorb gametes. Moreover,
bottom temperatures seldom reach 10°C (Moimtain
and Jessen 1987), the lowest temperature at which
Placopecten magellanicus has been observed to
spawn (Culliney 1974). In a recent study of a single
deepwater sea scallop population in the Gulf of
Maine, Barber et al. (1988) report reduced fecun-
dity, followed by gamete resorption and a possible
minor spring spawning, in turn followed by redevel-
opment, continued resorption, and an abrupt fall
spawning attempt. Almost certainly, deepwater sea
scallops do not spawn successfully. Recruitment to
these beds, therefore, would be haphazard and
originate both from populations on nearby ledges
and from spatfall out of the Gulf of Maine gyre, from
upstream spawning populations.
We have reported here that glycogen levels in
Placopecten magellanicus adductor muscle, meas-
ured during the annual peak period in late spring
and during the annual low period in winter, can in-
dicate scallop populations with little energy reserves
for successful spawning. In the case of the deep-
water sea scallops in the Gulf of Maine, lack of avail-
able nutrients is undoubtedly the reason for their
low muscle glycogen. Still to be clarified are the
events leading to the 1984 failure of the Asbury
Park sea scallops to develop the necessary energy
reserves for spawning.
Acknowledgment
We are grateful to Vincent S. Zdanowicz and
Anthony L. Pacheco (NMFS, NEFC, Sandy Hook
Laboratory, NJ) for making possible the Asbury
Park collections, and to Thomas R. Azarovitz, Don-
ald D. Flescher, Henry W. Jensen, Malcolm J. Siver-
man, and their trawl-survey colleagues (NMFS,
NEFC, Woods Hole Laboratory, MA) for the deep-
water scallop collections. We also thank Laure A.
Devine, Todd M. Welch, and Beth C. Marks for
technical assistance.
Literature Cited
Barber, B. J., R. Getchell, S. Shumway, and D. Schick.
1988. Reduced fecundity in a deep-water population of the
600
giant scallop, Placopecten magellanicus (Gmelin), in the Gulf
of Maine, U.S.A. Mar. Ecol. Prog. Ser. 42:207-212.
COLTON, J.
1968. Recent trends in subsurface temperatures in the Gulf
of Maine and contiguous waters. J. Fish. Res. Board Can.
25:2427-2437.
CULLINEY, J. L.
1974. Larval development in the giant scallop Placopecten
magellanicus (Gmelin). Biol. Bull. 147:321-332.
Gould, E.
1981. Field stress in the scallop Placopecten nuigellanieus.
I.C.E.S. (MEQC), C.M.1981/E:7, 16 p.
1983. Seasonal biochemical patterns for a single population
of sea scallops, Placopecten magellaniais, and their use in
interpreting field data. I.C.E.S. (MEQC), C.M.1983/E:57,
17 p.
Gould, E., R. A. Greig, D. Rusanowsky, and B. C. Marks.
1985. Metal-exposed sea scallops, Placopecten magellanicus
(Gmelin): A comparison of the effects and uptake of cad-
mium and copper. In F. J. Vernberg, F. P. Thurberg, A.
Calabrese, and W. B. Vernberg (editors), Marine pollution
and physiology: recent advances, p. 157-186. Acad. Press,
N.Y.
Gould, E., R. J. Thompson, L. J. Buckley, D. Rusanowsky,
AND G. R. Sennefelder.
1988. Uptake and effects of copper and cadmium in the gonad
of the scallop Placopecten magellanicus: Concurrent metal
exposure. Mar. Biol. 97:217-223.
Hopkins, T.S., and N. Garfield III.
1979. Gulf of Maine intermediate water. J. Mar. Res. 37:
103-139.
McLellan, H. J., L. Lauzier, and W. B. Bailey.
1953. The slope water off the Scotian shelf. J. Fish. Res.
Board Can. 10(4):155-176.
Mountain, D. G., and P. F. Jessen.
1987. Bottom waters of the Gulf of Maine, 1978-1983. J.
Mar. Res. 45:319-345.
Robinson, W. E., W. E. Wehling, M. P. Morse, and G. C.
McLeod.
1981. Seasonal changes in soft-body component indices and
energy reserves in the Atlantic sea scallop, Placopecten
magellanicus. Fish. Bull. 79:449-458.
Shumway, S. E., R. Selvin, and D. F. Schick.
1987. Food resources related to habitat in the scallop Placo-
pecten magellanicus (Gmelin, 1791): A qualitative study.
J. Shellfish Res. 6(2):89-95.
Steimle, F. J. Caracciolo, and J. B. Pearce.
1 982. Impacts of dumping on New York Bight Apex benthos.
In G. F. Mayer (editor). Ecological stress and the New York
Bight: science and management, p. 213-223. Estuarine
Res. Fed., Columbia, SC.
Thompson, R. J.
1977. Blood chemistry, biochemical composition, and the an-
nual reproductive cycle in the giant scallop, Placopecten
magellanicus, from southeast Newfoundland. J. Fish. Res.
Board Can. 34:2104-2116.
Edith Gould
Diane Rusanowsky
Donna A. Luedke
Northeast Fisheries Center Milford Laboratory
National Marine Fisheries Service, NOAA
212 Rogers Avenue
Milford, CT 06460
VERTICAL DISTRIBUTION AND
MASS MORTALITY OF
PRAWNS, PANDALUS PLATYCEROS, IN
SAANICH INLET, BRITISH COLUMBIA
Prawn, or spot prawn, Pandalus platyceros Brandt,
1851, British Columbia's largest shrimp species, is
extensively fished in Canada and is of considerable
economic importance (Noakes and Jamieson 1986).
The species ranges from California to Alaska and,
being largely associated with rocky terrain, is fished
with traps in many of the region's coastal inlets
(Butler 1980).
This study reports a fortuitous observation of
catastrophic mortality of prawn in Saanich Inlet,
noted during a series of observations on the vertical
distribution of prawn on the walls of this fjord using
a submersible. These observations are important
because, under the circumstances involved, these
mobile benthic organisms had ample opportunity to
avoid the apparent rapid intrusion of lethal environ-
mental conditions by moving upwards, and thereby
remaining in a favorable environment.
Well-documented sudden mass deaths of adult
marine invertebrates in subtidal environments have
usually been associated with man-induced environ-
mental perturbation, such as an oil spill, pollutant
discharge, entrainment of organisms into a lethal
environment (e.g., dredge or power plant cooling
water intake), or the entrapment of benthic organ-
isms by some lethal environmental event (Tulkki
1965). However, the selective high mortality of one
or only a few species in a subtidal community, with
no associated physical habitat perturbation and with
apparent opportunity for escape, has been infre-
quently described in documented catastrophic mor-
talities (Brongersma-Sanders 1957; Swanson and
Sindermann 1979; Levings 1980a, b; Tunnicliffe
1981; Burd and Brinkhurst 1984, 1985; Renaud
1986). It is known that species differ in their rela-
tive tolerances to environmental stress (e.g., Renaud
1986), but for subtidal invertebrates, the proximity
to lethal conditions of the majority of a population
for extended time periods has not been generally
noted. This study shows that prawn may occur close
to lethal environmental conditions, and that abrupt
mortality results if lethal water conditions sudden-
ly intrude. In certain locations, such mortality may
be more frequent than previously recognized and
may justify a unique exploitation strategy.
Materials and Methods
This study was conducted between 6 and 10
FISHERY BULLETIN: VOL. 86, NO. 3, 1988. gQl
October 1986, at Bamberton (lat. 48°35'N, long.
123°3rW), located half-way down Saanich Inlet, a
24 km long fjord on the southeast coast of Van-
couver Island. Because of its close proximity to
regional oceanographic laboratories, this inlet has
been well studied oceanographically (Carter 1934;
Herlinveaux 1962; Richards 1965; Anderson and
Devol 1973; Pickard 1975; Emerson et al. 1979;
Thomson 1981) and biologically (Tunnicliffe 1981;
Burd and Brinkhurst 1984, 1985). At its mouth, it
has a submerged (75 m) sill behind which there is
a deep (maximum depth = 234 m) basin, and water
below the sill depth is typically isolated, oxygen-
deficient, and may contain hydrogen sulphide. The
basin is flushed only when water above the sill is
sufficiently dense to cascade into it. Throughout the
year, there typically remains one or more oxyclines
in the water column structure. The study area is
generally characterized by a 20-30° slope to about
70 m, and a 30-60° slope from this depth to the
bottom of the inlet (209 m).
Observations were made by two people from the
submersible Pisces IV, which has been previously
described by Mackie and Mills (1983). Two 3-h dives
were made each day— one typically starting at 1300
and the other at 1900, about one hour after sunset.
Dives to the bottom of the inlet were conducted at
random locations over a 3 km length of shoreline
and began with a direct descent to the bottom of
the inlet. The submersible was then moved upwards,
horizontal and about 1 m from the cliff, from the
bottom of the inlet to a depth of 20 m. After sur-
facing, the submersible was then towed to another
dive location, where the process was repeated. A
total of 17 ascents were completed during the
8 dives.
Ascent speed varied according to habitat complex-
ity, slope, and crustacean abundance, but was slow
enough to permit recording of the species observed
except in areas where extremely dense concentra-
tions of animals were found. Observers were sta-
tioned on opposite sides of the submersible, with no
overlap in their visual field. Each observer was able
to scan approximately 90° on one side of the sub-
mersible's path.
As discussed by Richards and Schnute (1986), a
general problem with use of submersibles has been
the quantification of species abundance. Both the
presence of lights and the submersible itself might
affect animal behavior and hence bias observations.
Prawns tend to be cryptic and were frequently found
in association with bottom debris, and occasionally
in holes. After extensive observations, including
moving at specific locales with lights on or off and
moving at variable speeds, it was concluded that
prawns and other demersal crustaceans were not
apparently affected by Pisces's presence, allowing
them to be effectively studied. On the approach of
Pisces, demersal crustaceans would occasionally
swim a short distance with rapid flicks of their
abdomens, but in most cases, they would simply
assume an aggressive stance towards the submers-
ible and continuously face it as it passed by. They
did not retreat under cover.
Visibility of benthic animals varied somewhat
because of changes in bottom topography and its
effect on distance of the submersible from the sub-
strate. Water clarity was generally good (>7 m), but
resolution was poor at the perimeter of the illum-
inated area (about 5 m maximum).
Data were initially recorded into handheld tape
recorders, and tapes were transcribed shortly after
each dive. Prawns were individually counted and
their depths of occurrence noted over a vertical
transect of about 3 m. However, this procedure
could not be used when live prawn abundance ex-
ceeded approximately 6 m"^, because of their high
density and because the prawns stirred up the sub-
strate by their movements. This situation only oc-
curred over a narrow depth range (between 70 and
85 m depth, depending on date of observation). Dead
prawns, which often only consisted of exoskeleton
fragments, were also difficult to count at a density
greater than about 2 m"^. In both situations, num-
ber per meter of depth was conservatively estimated
by multiplying the respective minimum average den-
sity per square meter by 3 m, the transect width
over which live and dead prawn abundances were
being assessed.
Qualitative notes on abundance of munids. Muni-
da quadraspina Benedict, 1902, the dominant ben-
thic crustacean present, were recorded by depth
interval. Observers noted other invertebrate and
fish species present in each transect.
Movement of observed crustaceans was sufficient-
ly slow, relative to the submersible' s movement, to
prevent their crossing the submersible's path and
possible double counting by the observers. Body size
of some individuals was estimated by comparing
them to a 30 cm rod, marked in 10 cm intervals, that
hung in front of the left viewport. Sizes were later
confirmed by measurement of carapace lengths of
specimens collected using the extendable arm of the
submersible.
Water samples, and on some occasions crusta-
ceans, were collected at selected depths by pump-
ing water through jars attached to the exterior of
the submersible until they had been thoroughly
602
flushed, and then sealing them. Samples were fixed
with manganous sulphate reagent and alkaline
iodide solution as soon as the submersible had sur-
faced and was recovered at the end of each dive,
and dissolved oxygen concentrations were deter-
mined later using a modified Winkler titration pro-
cedure (Strickland and Parsons 1972). Ambient
water temperature was recorded continuously
during each dive. On 14 October 1986, a more com-
prehensive set of oceanographic measurements
(temperature, salinity, and dissolved oxygen [DO2])
was obtained by University of British Columbia
(UBC) oceanographers 1 km from the study
area.
Results
Substrate type was a soft, light-brown flocculent
ooze at the bottom of the inlet up to a depth of about
60 m, when it became more gravelly. The floor of
the inlet was relatively flat, changing to a slope of
30-60° at the walls up to a depth of about 100 m,
at which point 5-10 m vertical rock cliffs often oc-
curred. The slope then lessened at approximately
70 m depth to 20-30°, with frequent rock outcrop-
pings observed up to the minimum depth (20 m)
surveyed.
Salinity and temperature below 20 m ranged from
30.0 to 31.4^00 and from 8.0° to 11.2°C, respec-
tively, throughout the study. During the first two
days of observations, measured DO2 levels were
>1 mL • L"^ at depths above 20 m. Between the
afternoon and evening dives of the third day (8 Octo-
ber), an abrupt decrease in DO2 concentration to
between 0.76 and 0.92 mL • L"^ was detected at
depths of 75-77 m. However, at nearby depths of
71-73 m, DO2 levels remained above 1 mL • L~^
(1.29-2.00 mL • L'^).
Dissolved oxygen (DO2) measurements obtained
during the last two days of our study show a similar
profile to those obtained four days later in the center
of the inlet by the UBC team (Fig. 1). In both cases,
a region of low DO2 (<1 mL • L"^) was seen in
waters of intermediate depth, although the depth
at which this lens of low DO2 occurred differed by
about 20 m. Minimum DO2 level recorded during
our dives was 0.76 mL ■ L"' at 77 m on 8 October,
while on 14 October in the center of the inlet, the
lowest value noted was 0.44 mL • L"^ at 110 m.
Since low DO2 levels normally occur in the deepest
waters of the inlet (Pickard 1975; Burd and Brink-
hurst 1984), these results indicated that there had
recently been intrusions of denser, more oxygenated
water over the sill into the deeper regions of the
inlet, displacing the low DO2 layer upwards or ad-
vecting low DO2 water into the study area.
There were clear differences in depth ranges in-
habited by species commonly observed in the study
area (Fig. 1). The species we observed were pri-
marily benthic in habit, although some epibenthic
species such as spotted ratfish, Hydrolagus colliei
(Lay and Bennet, 1839); spiny dogfish, Squalus
acanthias Linnaeus, 1758; and Pacific cod, Gadus
macrocephalus Tilesius, 1810, were periodically
observed, usually at depths below 80 m. Various
rockfish (Sebastes sp.) were observed around rock
outcroppings.
The most abundant benthic invertebrate species
observed below 60 m were four species of shrimp
{Spirontocaris holmesi Holthuis, 1947; S. sica Rath-
bun, 1902; pink shrimp, Pandalus jordani Rathbun,
1902; and prawn) and munids. A few Dungeness
crab, Cancer magister (Dana, 1851) were observed
at 40-80 m depth. Greatest densities of prawns
generally occurred between 70 and 85 m depth,
although their observed range was from 20 to 159 m
(Table 1). There was no obvious difference in the
depth range of major prawn concentration (70-85
m) between afternoon and night dives on the same
day. However, more prawns were observed in the
depth range of 20-70 m at night than during the
day, but these were relatively few in comparison to
those at 70-85 m depth (Table 1). During the latter
part of the study, most prawns were in a narrow
band between 70 and 79 m depth. An amphipod, Or-
choTnene ohtusa (Sars, 1890), was common on the
substrate from 80-210 m water depth. Major con-
centrations of pelagic amphipods were observed at
depths of 45-75 m and 23-90 m during the day and
night, respectively. Euphausiids were most abun-
dant at 90-135 m during the day and at 50-90 m
at night.
Munids were not observed above the main prawn
concentration at 70-85 m. A wide size range of
munids was observed, ranging from recently settled
juveniles to adults of about 3 cm carapace length.
In general, large individuals were found below
100 m whereas small munids were found from 80
to 120 m depth. Munids were observed down to the
deepest depth surveyed (209 m).
On the evening dive of the third day, 8 October,
dying and dead prawns, the latter covered with
swarms of amphipods, were observed between 82
and 90 m depth, with most between 85 and 90 m
(Table 1). Some live prawns showing disoriented
behavior were also observed. Examination of col-
lected live prawns observed to be in poor condition
when sampled showed no evidence of disease or
603
200-
DISSOLVED
Figure 1.— Dissolved oxygen levels (mL • L'') at depth during each of the 4 days of diving
(lines were drawn by eye) and for the UBC sampling, and relative depth distributions of the
main abundances of the four large crustaceans observed. Oblique bar = zone of prawn mortal-
ity; stipple = area devoid of pink shrimp after low DOg water intrusion. Numbers = dates in
October 1986. A ( ) = 6 October, B (•••) = 7 October, C ( — ) = 8 October, D (-•-•)
= 9 October, £(•—•)= 14 October (UBC profUe).
604
Table 1 .—Average prawn abundance per observer per 5 m depth increnrient observed in the depth range
20-124 m during the 4 days of observations, 6-9 October 1986. A = afternoon dive, E = evening dive,
numbers in brackets = no. of ascents, no. of observers, < > = dead or dying.
6 October
A E
7 October
A E
8 October
j
9 October
Depth
A
E
A
E^
(m)
(2.1)
(2,2)
(1.1)
(2.1)
(1.1)
(3.2)
(3.2)
(3.2)
20-24
0
0
0
1.5
0
0
0
..^
25-29
0
0
0
1.5
0
0.5
0
—
30-34
0
0
0
3.5
0
0
0
—
35-39
0
0.25
0
4.5
0
0.5
0
—
40-44
0
1.00
0
2.5
0
2.0
0
—
45-49
0
2.0
0
2.0
0
0
0
—
50-54
0
2.5
0
0
0
0
0
—
55-59
0
7.0
0
0.5
0
0
0
—
60-64
0
10.0
1.0
1.0
0
0.5
0
—
65-69
0
14.25
0
5.0
0
1.0
0.6
3.0
70-74
1.0
^54.0
38.0
9.0
^80.0
2.0
^36.0
^80.0
75-79
^80.0
^80.0
0
^80.0
0
^80.0
^80.0 <
8>
^80.0
80-84
^80.0
^36.0
0
0
0
^16.0 <^18>
0 <^12>
^36.0 <^18>
85-89
0.5
1.0
0
0
0
0 <^30>
0 <^30>
0 <^30>
90-94
0.5
1.5
0
0
0
0 < ^6>
0 <0.3>
0 <h2>
95-99
2.5
0.75
0
0
0
0
0
0 < 4>
100-104
2.0
0.75
0
0
0
0.3
0
0
105-109
2.5
2.75
0
0
0
0
0.25
0
110-114
3.0
4.25
0
0
0
0
0
0
115-119
2.0
0.25
1.0
0
0
1.0
0.25
0
120-124
0.3
1.5
0
0
0
0
0.25
0
Total alive
174.3
219.0
40.0
111.0
80.0
103.8
117.4
199.0
Total dead
0
0
0
0
0
54.0
50.3
64.0
Estimated
mortality (%)
0
0
0
0
0
34
30
24
'No observations were made above 65 m on the evening of 9 October because of the entanglement of Pisces in submerged
rope. Recorder equipment failure prevented inclusion of one of the observer's counts of prawn on some of the dives.
^Live prawn abundance was estimated >€ m'^, but for purposes of analyses, 6 m'^ x 3 m' visibility was assumed,
giving 18 prawn per meter of water depth
3Dead prawn abundance was estimated
giving 6 dead prawn per meter of water depth
'Dead prawn abundance was estimated >2 m"', but for purposes of analyses, 2 m ^ x 3 m' visibility was assumed.
parasitism (S. Bower^). Dead prawns, sometimes
consisting only of exoskeleton remains, were ob-
served at a density >2 m"^ in some areas. Ap-
parently healthy prawns were concentrated in the
depth range of 75-80 m at densities >6 m"^. Dead
and dying prawns were also observed during both
dives on 9 October. By comparing the estimated
numbers of dead and living prawns observed in a
vertical transect, it was conservatively estimated
that approximately 25% of the prawn population
observed may have died during this 24-h period, with
observations taken over 9 ascents along about 1.5
km of shoreline (Table 1).
Prawn was the only species observed to be dying.
Apparently healthy amphipods, munids, and flatfish
were observed around the dying prawns. However,
coincident with the onset of prawn mortality, the
vertical distribution of pink shrimp separated into
two groups, one above and one below the depth
range of prawn mortality (Fig. 1). No dead pink
IS. Bower, Pacific Biological Station, Nanaimo, B.C., V9R 5K6,
pers. commun. October 1986.
shrimp were observed. Only amphipods were ob-
served eating the dying or dead prawns.
Discussion
Our observations indicate that the tolerance of
prawn to low DO2 levels may be less than that re-
ported from laboratory experiments. The tolerance
of prawn in a sealed chamber (10°C, SO^/oo salinity)
to low DO2 levels has been experimentally ex-
amined by Whyte and Carswell (1982). Under their
experimental conditions, prawns exhibited a reduced
metabolic rate at DO2 levels below approximately
2.5 mL • L"^ and died at approximately 0.35 mL
• L"^ They did not determine how long prawns
would survive at dissolved oxygen levels below 1 mL
• L"^ since their experimental design included only
a fixed amount of oxygen. Our study suggests that
minimum tolerance occurs at around 1 mL ■ L~^
since at levels below this, death occurred. Oxygen
stress in munids has been reported (Burd 1983) to
coincide with a loss of equilibrium similar to the
disoriented locomotor behavior we observed for
605
prawns. Based on our observations, we further note
that prawns appear to be less tolerant to low DO2
levels than many of the other species found in the
same depth range. However, published data on
tolerance to hypoxic conditions for the species pres-
ent exists only for munids, which have been shown
to tolerate hypoxic conditions as low as 0.1-0.15 mL
• L-i (Burd 1983; Burd and Brinkhurst 1984,
1985).
All benthic species observed, except prawn, had
a relatively large depth range over which individuals
were found in abundance. Those species apparent-
ly more tolerant to low DO2 levels were found from
about 85-210 m. It is unknown why in contrast to
other species observed, prawns were concentrated
in a narrow depth range at 70-85 m water depth,
so close to lethal water conditions. Most prawns
were apparently prevented from going deeper by
intolerance to low DO2 concentrations, although a
few individuals were below this low DO2 layer and,
for the short term at least, were apparently sur-
viving. At night, there was little change in the ob-
served general depth preference of the main prawn
concentration, although more prawns were observed
at shallower depths. Prawns were not observed
moving vertically on the cliffs in a directed manner,
and so prawns observed at shallower depths at night
may have been hidden there during the day.
With the sudden movement of low DO2 water
into the depth range occupied by prawns, it is
unknown why prawns did not simply walk upwards
on the cliff, away from the low DO2 area, and stay
in a tolerable environment, as did the pink shrimp.
The distance prawns would have had to travel was
<10 m in the 70-75 m depth range. Some vertical
movement of prawns may have occurred, since in
the afternoon dive prior to the evening dive in which
dead prawns were first observed, the depth range
in which prawns were abundant was narrow (5 m)
and at its shallowest depth (70-75 m).
Two oceanographic factors apparently caused the
observed prawn mortality: the existence of a low
DO2 water mass in close proximity to the prawns
and some event which caused this water mass to in-
trude suddenly into the prawn habitat. As indicated
earlier, the presence of oxyclines in Saanich Inlet
is well documented, although the close proximity of
prawns to this lethal environment had not previously
been described.
We offer two possible explanations that could
account for sudden intrusion of the anoxic layer:
1) change in the amplitude of oscillations of the
oxycline, or 2) an overall change in level of the mean
oxygen surfaces, perhaps related to a change in
subsurface properties. With respect to the first,
Thomson et al. (in press) showed that in Saanich
Inlet, there are regular, peak-to-peak oscillations in
DO2 level of the order of 2.5 mL • L' at 100 m
depth. These oscillations were found to occur over
a period of hours, with a standard deviation and
range of effective vertical isopycnal displacement
estimated to be 2.0 and 9.6 m, respectively. Thom-
son et al. (in press) collected their data in April 1987,
when the mean DO2 level at 100 m was 4.8 mL ■
L"^ If similar oscillations of the oxycline occurred
during our observations, when DO2 levels were
much lower, then with a moderate change in oscilla-
tion amplitude, prawn could suddenly experience
lethal DO2 levels for time periods up to approx-
imately 6 hours. The causal mechanism generating
the oscillations and changes in amplitude of oscil-
lation of the pycnocline, and hence oxycline, is
currently unknown, but is probably due to internal
gravity waves propagating within the inlet (R.
Thomson^).
The second explanation involves a rapid change
in average depth of the oxycline caused by changes
in vertical density profile of the water column. Inter-
mediate depth waters outside Saanich Inlet are most
dense in the fall, and intrusion of this denser water
over the sill into the inlet typically occurs at this time
(Pickard 1975). Such intrusions are often caused by
strong tidal influxes, and fluctuations in depth of
the pycnocline and oxycline subsequently propagate
down the inlet as a density intrusion (Holbrook and
Halpern 1982). Any intrusion has the potential of
suddenly altering oxygen concentrations at various
depths. The observation of mortality beginning in
late afternoon on 8 October is in agreement with
that expected based on the daily and hourly timing
of tidal action seen during the study period.
Prawn mortality as described would thus appear
to be an episodic, but perhaps not an uncommon,
event in Saanich Inlet. It is probably a fall phenom-
enon, for the oceanographic reasons described above
and since in other years, this was when the hypoxic
layer was shallowest (Richards 1965; Tunnicliffe
1981; Burd and Brinkhurst 1984). Our observations
clearly demonstrate that sudden catastrophic mor-
tality can occur on a scale which may noticeably
affect species abundance in an area. If undocu-
mented, such episodic mass mortality may confound
an understanding of species population dynamics.
For fishermen in the area, mortality of prawn is a
concern and may explain why seasonal landings may
2R. Thomson, Institute of Ocean Sciences, Sidney, B.C. V8L 4B2,
pers. commun. May 1987.
606
not meet expectations. Understanding the timing
and likelihood of such events can improve manage-
ment, and since the geographical occurrence of such
mortality may often be quite localized, increased
harvest in specific locations prior to high natural
mortality events might be justified.
Acknowledgments
The assistance of F. Chambers and the pilots and
support staff of Pisces IV is gratefully acknowl-
edged, since without their professional, enthusiastic
assistance, the study could not have been conducted.
D. Mackas and D. Latrouite, staff of the Department
of Oceanography at UBC, T. Butler, and J. Fulton
assisted in data collection, analysis and/or species
identification, and R. Thomson assisted in interpre-
tation of oceanographic events. C. Levings, F. Ber-
nard, S. Bowers, J. Boutillier, and two anonymous
reviewers provided constructive reviews. The par-
ticipation of E. Pikitch was sponsored by NOAA Of-
fice of Sea Grant, Department of Commerce, under
grant No. NA85AA-D-SG095, project #R/E5-7.
Literature Cited
Anderson, J. J., and A. H. Devol.
1973. Deepwater renewal in Saanich Inlet, an intermittent-
ly anoxic basin. Estuarine Coastal Mar. Sci. 1:1-10.
Brongersma-Sanders, M.
1957. Mass mortality in the sea. Mem. Geol. Soc. Am. 67:
941-1010.
BURD, B. J.
1983. The distribution, respiration and gills of a low oxygen
tolerant crab, Munida quadrispina (Benedict, 1902) (Gala-
theidae, Decapoda) in an intermittently anoxic fjord. MS.
Thesis, Univ. Victoria, Victoria, B.C., 151 p.
BURD, B. J., AND R. 0. Brinkhurst.
1984. The distribution of the galatheid crab Munida quadra-
spina (Benedict, 1902) in relation to oxygen concentrations
in British Columbia fjords. J. Exp. Mar. Biol. Ecol. 81:1-20.
1985. The effect of oxygen depletion on the galatheid crab
Munida qvxidraspina in Saanich Inlet, British Columbia. In
J. S. Gray and M. E. Christiansen (editors), Marine biology
of polar regions and effects of stress on marine organisms,
p. 435-443. John Wiley & Sons Ltd.
Butler, T. H.
1980. Shrimps of the Pacific coast of Canada. Can. Bull.
Fish. Aquat. Sci. 202, 280 p.
Carter, N. M.
1934. Physiography and oceanography of some British
Columbia fjords. Proc. 5th Pac. Sci. Cong., 1933, Vol. 1,
p. 721-733.
Emerson, S., R. E. Cranston, and P. S. Liss.
1979. Redox species in a reducing fjord: equilibrium and
kinetic considerations. Deep-Sea Res. 26:859-878.
GiLMARTIN, M.
1962. Annual cyclic changes in the physical oceanography of
a British Columbia fjord. J. Fish. Res. Board Can. 19:
921-974.
Herlinveaux, R. H.
1962. Oceanography of Saanich Inlet in Vancouver Island,
British Columbia. J. Fish. Res. Board Can. 19:1-37.
HoLBROOK, J. R., and D. Halpern.
1982. Winter time near-surface currents in the Strait of Juan
de Fuca. Atmos.-Ocean 20:327-339.
Levings, C. D.
1980a. Demersal and benthic communities in Howe Sound
Basin and their responses to dissolved oxygen deficiency.
Can. Tech. Rep. Fish. Aquat. Sci. 951, 27 p.
1980b. Benthic biology of a dissolved oxygen deficiency event
in Howe Sound, B.C. In H. J. Freeland, D. M. Farmer, and
C. D. Levings (editors), Fjord oceanography, p. 515-522.
Plenum Pub. Co., N.Y.
Mackie, G. 0., AND C. E. Mills.
1983. Use of the PISCES IV Submersible for zooplankton
studies in coastal waters of British Columbia. Can. J. Fish.
Aquat. Sci. 40:763-776.
NOAKES, D., AND G. S. JaMIESON.
1986. Preliminary analysis of British Columbia commercial
landing statistics for 1979 to 1984 inclusive: a multispecies
perspective. Can. MS Rep. Fish. Aquat. Sci. 1882, 191 p.
PiCKARD, G. L.
1963. Oceanographic characteristics of inlets of Vancouver
Island, British Columbia. J. Fish. Res. Board Can. 20:
1109-1120.
1975. Annual and longer term variations of deepwater
properties in the coastal waters of southern British Colum-
bia. J. Fish. Res. Board Can. 32:1561-1587.
Renaud, M. L.
1986. Hypoxia in Louisiana coastal waters during 1983:
implications for fisheries. Fish. Bull., U.S. 84:19-26.
Richards, F. A.
1965. Anoxic basins and fjords. In J. P. Riley and 0. Skir-
row (editors), Chemical Oceanography, Vol. 1, p. 611-645.
Academic Press, N.Y.
Richards, L. J., and J. T. Schnute.
1986. An experimental and statistical approach to the ques-
tion: Is CPUE an index of abundance? Can. J. Fish. Aquat.
Sci. 43:1214-1227.
Strickland, J. D. H., and T. R. Parsons.
1972. A practical handbook of seawater analysis. Fish. Res.
Board Can. Bull. 167, 309 p.
SwANSON, R. L., AND C. J. Sindermann (editors).
1979. Oxygen depletion and associated benthic mortalities in
New York Bight, 1976. NOAA Porf. Rep. No. 11, 345 p.
Rockville, MD.
Thomson, R. E.
1981. Oceanography of the British Columbia coast. Can.
Spec. Publ. Fish. Aquat. Sci. 56, 291 p.
Thomson, R. E., T. A. Curran, M. C. Hamilton, and R.
McFarlane.
In press. Time series measurements from a moored fluores-
cence-based dissolved oxygen sensor. J. Atmos. Oceanic
Technol.
Tulkki, p.
1965. Disappearance of the benthic fauna from the Basin of
Bomholm (southern Baltic) due to oxygen deficiency. Can.
Biol. Mar. VI:455-463.
TUNNICLIFFE, V.
1981. High species diversity and abundance of the epibenthic
community in an oxygen deficient basin. Natiu-e (Lond.)
294:354-356.
Whyte, J. N. C, AND B. L. Carswell.
1982. Determinants for live holding the spot prawn Pandalus
607
platyceros, Brandt. Can. Tech. Rep. Fish. Aquat. Sci. 1 129,
29 p.
Glen S. Jamieson
Department of Fisheries and Oceans
Fisheries Research Branch
Pacific Biological Station
Nanaimo, British Columbia V9R 5K6, Canada
Ellen K. Pikitch
Department of Fisheries and Wildlife
Oregon State University
M. 0. Hatfield Marine Science Center
Newport, OR 97365
Present address:
School of Fisheries WH-10
University of Washington
Seattle, WA 98195, USA
MORPHOLOGICAL DIFFERENCES BETWEEN
TWO CONGENERIC SPECIES OF
PLEURONECTID FLATFISHES: ARROWTOOTH
FLOUNDER, ATHERESTHES STOMIAS, AND
KAMCHATKA FLOUNDER, A. EVERMANNI
The two flatfishes of the genus Atheresthes (family
Pleuronectidae) are commonly caught in the eastern
Bering Sea commercial trawl fishery. From 1977 to
1983, they comprised an estimated 10.03% of the
total flatfish catch by the foreign trawl vessels in
the Bering Sea/ Aleutian Islands region (data com-
piled from U.S. Foreign Fisheries Observer Pro-
gram, Northwest and Alaska Fisheries Center).
However, these two species, the arrowtooth
flounder, A. stomias, and Kamchatka flounder,
A. evermanni, are morphologically similar and
hence difficult to distinguish. As a result, fisheries
workers in the field often lump the two species or
misidentify them. Because the two species may have
biological differences not presently known, it is im-
portant for management considerations and stock
assessments to distinguish the species in fisheries
surveys. The objective of this paper is to describe
gross morphological differences between the two
species more explicitly, so that the two can be ac-
curately identified in the field.
Norman (1934) thought that these two species of
Atheresthes were so similar that eventually they
were shown to be identical. However, based on elec-
trophoretic evidence, Ranck et al. (1986) concluded
that A. stomias and A. evermanni are valid species.
Wilimovsky et al. (1967) previously had reached this
same conclusion by using a special morphological
character index to separate the two species. This
index is a function of caudal vertebrae number, gill
raker number, distance from anterior eye margin
to dorsal origin, and eye diameter. Unfortunately,
the index is too complex to use in the field because
it is based partly on characters that cannot easily
be evaluated by gross external examination. This
study describes a simpler method for differentiating
the two species based on previously described ex-
ternal morphological characteristics and two new
morphological characters.
Methods
Collections were made in the eastern Bering Sea
in an area between lat. 54° and 59°N, long. 163°
and 174°W (Fig. 1) aboard the National Marine
Fisheries Service RV Chapman in summer 1984.
Specimens were unselectively sampled in the field
from trawl catches containing A. stomias and A.
evermanni. The fork length and sex of the fish along
with location of sample were recorded, and each
specimen was preserved in 3.7% seawater/formal-
dehyde solution.
In the laboratory, the following characteristics
were examined:
1) Upper eye position: Specimens were first
classified according to the position of the upper eye,
following Norman (1934). If the orbit of the upper
eye interrupted the profile of the head (Fig. 2A), the
specimen was classified as A. stomias. If the upper
eye did not interrupt the profile of the head (Fig.
2B) and was completely on the right side of the
head, the specimen was classified as A. evermanni
(Norman 1934; Wilimovsky 1967).
2) Gill raker counts: After initial separation of
the specimens on the basis of the upper eye posi-
tion, the four gill arches of the eyed side were
removed and the gill raker count of each of the four
arches recorded. Counts of the upper and lower
limbs were recorded separately and the two counts
were separated by a plus sign (for example, 4 + 12
means 4 rakers on the upper limb and 12 on the
lower limb). If a gill raker straddled the angle of the
arch, it was included in the count of the lower limb.
In this study, only the lath-shaped structures were
counted as gill rakers; the rudiments were not
counted.
Results
A total of 251 fish was examined. Based on the
upper eye position, 170 specimens were classi-
608
FISHERY BULLETIN: VOL. 86, NO. 3, 1988.
63 OON
61 00
59 00
57 00
- 55 00
- 53 00
51 00
179 OOE
1 76 OOW 171 00 166 00 161 00
Figure 1.— Bering Sea collecting localities.
156 00
Arrowtooth flounder
(Atheresthes stomias)
Kamchatka flounder
(Atheresthes evermanni)
Figure 2.— Blind side view of head profiles oi Atheresthes stomias (A) and A. evermanni (B).
609
fied as A. stomias, and 81 as A. evermanni. Gill
rakers decreased in number from the first to the
third gill arches in both species. Neither species had
gill rakers on the fourth gill arch (though they did
have rudiments).
First Gill Arch
In general, A. stomias had more gill rakers (on
both the upper and lower limbs) on the first gill arch
than did A. evermanni; however, the counts did
overlap both in arrangement of gill rakers and total
numbers (Table 1). Of the A. stomias examined,
75.2% had 4 or more rakers on the upper limb and
none had fewer than 3. On the other hand, no A.
evermanni had 4 or more gill rakers, and 49.4% had
2 rakers on the upper limb on the first gill arch.
Second Gill Arch
As was the case of the first gill arch, A. stomias,
in general, had more gill rakers on the second gill
arch than did A. evermanni. There were minor over-
laps (1.8%) between these two species in the total
gill raker counts and in the distribution of the gill
rakers (Table 2). In addition to having different gill
raker counts and patterns, all of the A. everTnanni
examined had 1 gill raker on the upper limb of the
second gill arch, whereas 98.2% of the A. stomias
had 2 gill rakers on the upper limb of the second
gill arch (Table 2).
Third Gill Arch
In general, A. stomias had more gill rakers on the
third gill arch than did A. evermanni. The range of
total raker counts for A. stomias was 3 to 7, with
most (85.3%) having 4 or 5. Total raker counts for
A. evermanni ranged from 2 to 5, with 64.2% hav-
ing 3. The most frequent arrangements of gill rakers
on the third gill arch oiA. stomias were 0 -t- 4 (43.5%)
and 0-1-5 (41.2%), and the most frequent arrange-
ment for A. evermanni was 0 -t- 3 (64.2%). Only three
(1.8%) A. stomias had one gill raker on the upper
limb of the third gill arch. The rest (98.2% A.
stomias and all A. evermanni) had no gill rakers on
the upper limb of the third gill arch.
Anomalous Specimens
Three specimens had the upper eye interrupting
the profile of the head but had only 1 gill raker on
the upper limb of the second gill arch. These fish
were probably A. stomias with anomalous gill raker
Table 1 .—Gill-raker arrangements and their per-
centages (%) of frequency of occurrence (FO) on
the first gill arches (eyed side) of Atheresthes
stomias and Atheresthes evermanni collected from
eastern Bering Sea.
stomias
evermanni
Pattern
FO
%
FO
%
5+11
2
1.2
0
0
4+13
1
0.6
0
0
4+12
59
34.7
0
0
4+11
65
38.1
0
0
4+10
1
0.6
0
0
3+13
1
0.6
0
0
3+12
11
6.5
0
0
3+11
29
17.1
4
4.9
3+10
1
0.6
32
39.5
3 + 9
0
0
5
6.2
2+11
0
0
3
3.7
2+10
0
0
29
35.8
2 + 9
0
0
8
9.9
Total
170
100
81
100
Mean
15.1
12.4
Table 2.— Gill-raker arrangements and their per-
centages (%) of frequency of occurrence (FO) on
the second gill arches (eyed side) of Atheresthes
stomias and Atheresthes evermanni collected from
the eastern Bering Sea.
stomias
even
FO
manni
Pattern
FO
%
%
3 + 10
1
0.6
0
0
2+11
3
1.8
0
0
2+10
54
31.8
0
0
2 + 9
108
63.4
0
0
2 + 8
1
0.6
0
0
1+9
3
1.8
8
9.9
1+8
0
0
61
75.3
1+7
0
0
11
13.6
1+6
0
0
1
1.2
Total
170
100
81
100
Mean
11.3
8.9
2 or more
rakers on
upper limb
167
98.2
0
0
1 raker on
upper limb
'3
1.8
81
100
'All of these A. stomias have two gill rakers on the upper
limbs of the second gill arches of the blind side. Their pat-
terns are either 2+10 or 2 + 9.
counts. One of the three had gill raker patterns of
4 + 11, 1 + 9, and 0 + 4 on the first, second, and third
gill arches, respectively. The gill raker patterns on
the blind side of this specimen were 3-1-11 on the
first gill arch, 2 + 9 on the second, and 0 + 4 on the
third. Thus, because this specimen had 4 gill rakers
on the upper limb of the first gill arch on the eye
610
side and 2 gill rakers on the upper limb of the sec-
ond gill arch on the blind side, it is referred to A.
stomias. The other two anomalous specimens also
had 2 gill rakers on the upper limbs of the second
gill arch of the blind side and were also recorded
as A. stomias.
Discussion
From this study, it is evident that the two species
of Atheresthes can most easily be distinguished by
eye position. The number of gill rakers on first and
second gill arches can be used to assist and verify
identification.
When identifying specimens, eye position should
be examined first. If the upper eye interrupts the
profile of the head, this specimen is A. stomias; if
the upper eye does not interrupt the profile of the
head, the specimen is A. evermanni. If the head is
in bad shape (e.g., damaged during the trawl opera-
tion) or if the examiner has difficulty using eye posi-
tion and head profile to identify a specimen, the gill
arches must be examined. Two or more gill rakers
on the upper limb of the second gill arch indicates
that the specimen is A. stomias; if there is only 1
gill raker, the specimen is A. everm,anni.
The number of gill rakers on the first gill arch has
generally been used to distinguish the two species
of Atheresthes. However, this study demonstrated
a greater overlap between the two species in number
of rakers on the first gill arch than the second gill
arch (Tables 1, 2), indicating that the second gill arch
is a better character for assigning individuals to the
species.
The study also suggests that the number of gill
rakers on the upper limb of the first gill arch is
species specific. If there are 4 or more gill rakers,
the specimen is A. stomias; 2 or fewer gill rakers
indicate the specimen is A. evermanni.
The uncertainty in examining the first gill arch
is when there are 3 gill rakers on the upper limb.
Approximately 25% of A. stomias and 50% of A.
evermanni samples had 3 gill rakers on the upper
limb of the first gill arch. Thus, when 3 gill rakers
are present on the upper limb of the first gill, the
second gill arch must also be examined to distinguish
the two species.
Acknowledgments
I want to thank Jean Dunn and James Allen for
their comments and suggestions. Richard Bakkala
and Patricia Livingston reviewed my earlier manu-
script; their help is also appreciated. I also want
to thank the anonymous reviewers for their com-
ments.
Literature Cited
Norman, J. R.
1934. A systematic monograph of the flatfishes (Hetero-
somata). Vol. I. Psettodidae, Bothidae, Pleuronectidae.
Br. Mus. (Nat. Hist.), Lond., 459 p. [Reprinted, 1966, by
Johnson Reprint, N.Y.]
Ranck, C, F. Utter, G. Milner, and G. B. Smith.
1986. Genetic confirmation of specific distinction of arrow-
tooth flounder, Atheresthes stomias, and Kamchatka
flounder, A. evermanni. Fish. Bull., U.S. 84:222-226.
WiLiMOVSKY, N. J., A. Peden, and J. Peppar.
1967. Systematics of six demersal fishes of the North Pacific
Ocean. Fish. Res. Board Can., Tech. Rep. 34, 95 p.
Mei-Sun Yang
Northwest and Alaska Fisheries Center
National Marine Fisheries Service, NOAA
7600 Sand Point Way NE
Seattle, WA 98115-0070
PREDATION OF KARLUK RIVER
SOCKEYE SALMON BY COHO SALMON
AND CHAR
The number of sockeye salmon, OncorhynchiLS
nerka, in Alaska's Karluk River (Fig. 1) declined
from millions to thousands during the early part of
the present century. Rounsefell (1958) discussed
alternative explanations for the decline including a
general loss of fertility of the system as the number
of salmon carcasses declined, competition, over-
fishing, subtle changes in climate, and predation; he
concluded that the combined effect of predation and
fishing was the most probable explanation. Later,
Van Cleave and Bevan (1973) suggested that the
weir constructed in the river each year to facilitate
counting the fish as they entered the system was
the most probable cause of the decline. It prevented
free movement of both adults and juveniles in the
river. All of these hypotheses remain as potential
explanations for the decline.
Fredin et al. (1974) described a relation that
showed two equilibrium regions between the spawn-
ing stock and the resultant run for sockeye salmon
in the Kodiak area. We developed a stock-recruit-
ment curve (Fig. 2) for sockeye salmon in the Karluk
River basin that also showed two equilibrium
regions, and suggested that the population had "col-
FISHERY BULLETIN VOL. 86, NO. 3, 1988
611
i
1^
>
xk ^
^ ^
8
i
i*: Ci
O
\
n^S^
N
^ CO
b
b«: —
"\V%:
CS
^\ (
^^
^
\ \
"V^
^
+ <
612
Spawning Stock (millions)
Figure 2.— Stock-recuit relation for sockeye salmon in the Karluk River basin.
Squares are the running geometric mean (by 9) of stock and recruit estimates
for the 1922-77 broods. The curved, solid line was described by R = 1.83
(lO*^) + 7.73 P + 1.29(10"^) p2 - 5.58(10-^2)P^ where, i? = recruits and
P = stock. Ages of fish in the escapement (1922-36 from Barnaby 1944;
1937-69 from the Northwest and Alaska Fisheries Center, National Marine
Fisheries Service, Auke Bay, AK; 1980-85 from Alaska Department of Fish
and Game, Kodiak, AK) were used to estimate the recruits produced by each
brood. The diagonal lines show how the replacement line changes as the ex-
ploitation rate increases from 0 to 0.3.
lapsed" into the lower of the two. Fishing can cause
such a collapse (Peterman 1977) and recovery
becomes impossible unless exploitation rates are
reduced to levels substantially lower than the rate
that caused the collapse.
Multiple equilibria in an exploited population can
be caused by depensatory mortality— the loss of a
relatively greater fraction of the population when
it is small than when it is large (Neave 1953). Several
functional responses (Ricker 1954; Rolling 1973)
have been used to describe relations between prey
density and predation rate— one of which (Type III
relation) can produce multiple equilibria in the stock-
recruitment curve of a prey population (Peterman
1977). The Type III or S-shaped functional response
is characteristic of predators that consume a small
fraction of the prey at low prey population density;
as prey population density increases, however, the
predators rapidly increase the fraction consumed
through learning or aggregation. The concave. Type
II functional response holds when the fraction con-
sumed is high at low prey density.
The apparent potential for stock collapse, as
depicted in the Karluk sockeye stock-recruitment
curve, could be the consequence of Type III preda-
tion mortality. A preliminary survey of the food
habits of fish in the system showed that coho salmon.
0. kisutch, and two chars— the Dolly Varden,
Salvelinics malma, and Arctic char, S. alpinus—
were predators of juvenile sockeye salmon. We set
out to determine whether the functional responses
for coho salmon and for char were of Type II or Type
III. Our approach was to describe the relation be-
tween the number of prey eaten per predator and
the index of prey abundance provided by the annual
counts of adult sockeye salmon that entered the
system for spawning. Unfortunately, the study had
to be terminated after five years because of a man-
agement decision to enhance the productivity of
Karluk Lake with commercial fertilizer; we could
not eliminate the possibility that the effects of fer-
tilization would confound predation responses. We
describe the data that were accumulated during five
field seasons and our tentative conclusions concern-
ing the role of predation mortality in the dynamics
of these sockeye salmon.
Methods
Sampling sites were established at locations
around the littoral zone of Karluk Lake at the outlets
of spawning streams and at beach spawning areas
(Fig. 1) in 1982. The Karluk River was sampled from
the outlet at the lake to about 100 m downstream.
613
In each of the five subsequent years, a field crew
sampled each location at weekly intervals, from late
April to October.
Juvenile coho salmon were collected with beach
seines and minnow traps, and chars with beach
seines, floating gill nets, and hook and line. Only
coho salmon larger than 80 mm were found to be
predators of sockeye salmon during the April-
October sampling period. Since char were captured
with hook and line or in gill nets, our samples con-
tained only fish that were large enough to consume
sockeye salmon fry. Coho salmon were preserved
in formalin for examination later, when the contents
of the stomachs were removed and the sockeye
salmon fry and fingerlings were counted. The chars
were tagged and released after the contents were
flushed from their stomachs and preserved in
formalin.
Results
Juvenile coho salmon were aggregated around
tributary outlets and in the littoral areas of Karluk
Lake. The chars were found almost exclusively
around the tributary outlets and in the Karluk River.
Few coho salmon or chars were captured by sein-
ing in the pelagic areas of the lake. We did not
distinguish between the two chars.
Predation rates for the chars did not increase as
the abundance of sockeye salmon increased (Table
1), but the average number of sockeye salmon fry
consumed by each coho salmon did increase and ap-
peared to be depensatory (Fig. 3). A general equa-
tion (Real 1979) was used to describe the relation
between the predation rate (Y) by coho salmon and
the index of prey abundance (X),
Y = bX'-lil + aJC-).
(1)
A value of c = 1 provides a classic. Type II func-
tional response curve, while values of c exceeding
1.0 provide a sigmoid, depensatory shape. A value
of c = 2 gives the classic Type III curve.
Best (least squares) fit values for a, b, and c were
obtained by transforming Equation (1) into the form
ln(l/y - a/b) = In(i-i) - c\n(X).
(2)
Using trial values of a/b, we regressed the left side
of Equation (2) on \nX until we identified the value
of a/b giving the lowest residual variance. The
Table 1 . — Predation on juvenile sockeye salmon by predatory coho
salmon (i.e., juvenile coho salmon longer than 80 mm) and chars
in Karluk Lake, AK during June and July from 1982 to 1986.
Coho salmon
Chars
Sockeye
salmon
Number
Predation
Number
Predation
Year
examined
rate
examined
rate
escapement
1982
252
0.475
220,000
1983
3.132
0.076
95
0.98
164,000
1984
250
0.661
128
19.45
436,000
1985
956
0.452
485
10.06
420,000
1986
423
0.740
571
4.50
996,000
o
■a
5n
0 0.2 0.4 0.6 0.8
Index of Prey Density (Millions of Spowners)
1.0
Figure 3.— Functional response curve for coho salmon greater than 80 mm
(predators) and sockeye salmon (prey) in Karluk Lake. The index of prey abun-
dance was divided by 1,000,000 (e.g. index 0.8 = 800,000 adults in the
escapement).
614
results are alb = 1.35, c = 3.74 (standard error =
0.79), R^ = 0.88, ln(6-i) = 46.98.
Hence Equation (1) becomes
Y = 6X3^4/(1 + 1.356X3^") . ft ^ g -46.98
Inasmuch as Y ranges between zero and bla, it
seems unlikely that Y, or a corresponding statistical
error term for Equation (1) would be approximated
by a normal distribution. On the other hand, so long
as Y does not rise above b/a (to which Equation (1)
constrains it), the left side of Equation (2) lies
between - «> and + <», and the error term is more
likely approximated by a normal distribution. Con-
sequently, we cautiously used the standard error
associated with c, to test whether c > 1.0; that is,
whether the functional response curve is sigmoid
(Type III). In fact, c lies more than 3.5 standard
errors above 1 (Type II response) and more than 2.2
standard errors above 2 (Type III response). Al-
though the power of a test involving only 5 data
points is weak, we feel that a tentative conclusion
of depensatory predation by juvenile coho salmon
is justified.
Discussion
Many adult salmon as they attempt to get to the
spawning grounds, and as they spawn, are killed by
Kodiak brown bears. Card (1971) reviewed the
available literature concerning predation of salmon
by bears at Karluk and, in years when fish were not
abundant, noted that bears had been observed to
leave the salmon spawning areas to feed on berries
in the local area, indicating that their predation also
may be depensatory. We used data from Card's
summary to approximate the relation between the
number of adult sockeye salmon in a nm (X) and
the number of unspawned adults (Y) estimated to
have been killed by bears (Fig. 4). Although R^ was
only 0.424, a depensatory relation was indicated as
Y = X2io6/(3io2.85 + 0.00435X2106). c, from
Equation (1), was 2.106, with a standard error of
1.098. Because predation of sockeye salmon by
bears, as well as predation by coho salmon, appeared
to be depensatory, it is unlikely that predation by
coho salmon alone was the sole cause of the com-
plex stock-recruitment curve for sockeye salmon.
Prudent management of these salmon, and of
salmon in systems similar to Karluk, may require
regulation of harvest to prevent collapse of popula-
tions into relatively low equilibrium regions. Harvest
levels that would have prevented collapse of the
Karluk population can be estimated from the stock-
recruitment curve (Fig. 2). An exploitation rate
between 30 and 35% of the recruits should have
maintained stock sizes associated with the upper
equilibrium region. Exploitation at a constant rate
of 0.40 increases the slope of the replacement line
to the point that collapse of the population into the
lower equilibrium region becomes inevitable (see
Peterman 1977 for a description of the relation
between the size of stability regions and exploita-
tion rate). When depensatory mortality is potentially
high for economically important populations, it may
be necessary to limit exploitation to less than 35%
of the recruits to prevent collapse.
X)
V
V
E
Z3
0.2 0.4 0.6
Millions of Adult Salmon
Figure 4.— Functional response curve for predation of sockeye salmon by
bears. Number killed is thousands of unspawned salmon.
615
Acknowledgments
We thank Jon Nelson for his support, and all
others that participated in the project.
Literature Cited
Barnaby, J. T.
1944. Fluctuations in abundance of red salmon, Oncorhyn-
chus nerka (Walbaum), of the Karluk River, Alaska. Bull.
U.S. Bur. Fish. 50:237-295.
Fredin, R. a., S. Pennoyer, K. R. Middleton, R. S. Roys,
S. C. Smedley, and a. S. Davis.
1974. 5. Information on recent changes in the salmon fish-
eries of Alaska and the conditions of the stocks. Int. North
Pac. Fish. Comm. Bull. No. 29, p. 37-142.
Card, R.
1971. Brown bear predation on sockeye salmon at Karluk
Lake Alaska. J. Wildl. Manage. 35:193-204.
Rolling, C. S.
1973. Resilience and stability of ecological systems. Annu.
Rev. Ecol. Syst. 4:1-23.
Neave, F.
1953. Principles affecting the size of pink and chum salmon
populations in British Columbia. J. Fish. Res. Board Can.
9:450-491.
Peterman, R. M.
1977. A simple mechanism that causes collapsing stability
regions in exploited salmonid populations. J. Fish. Res.
Board Can. 34:1134-1142.
Real, L. A.
1979. Ecological determinants of functional response.
Ecology 60:481-485.
RiCKER, W. E.
1954. Stock and recruitment. J. Fish. Res. Board Can. 11:
559-623.
Rounsefell, G. a.
1958. Factors causing decline in sockeye salmon of Karluk
River, Alaska. Bull. U.S. Bur. Fish. 58:83-169.
Van Cleave, R., and D. E. Bevan.
1973. Evaluation of the causes for the decline of the Karluk
sockeye salmon runs and recommendations for rehabilita-
tion. U.S. Fish Wildl. Serv., Fish. Bull. 71:627-649.
John D. McIntyre
Reginald R. Reisenbichler
John M. Emlen
National Fishery Research Center
U.S. Fish and Wildlife Service
Building 20Jt, Naval Station
Seattle, WA 98115
Richard L. Wilmot
James E. Finn
National Fish and Wildlife Research Center
U.S. Fish and Wildlife Service
1101 E. Tudor Road
Anchorage, AK 99503
616
INFORMATION FOR CONTRIBUTORS TO THE FISHERY BULLETIN
Manuscripts submitted to the Fishery Bulletin will reach print faster if they conform to the following instructions.
These are not absolute requirements, of course, but desiderata.
CONTENT OF MANUSCRIPT
The title page should give only the title of the paper,
the author's name, his/her affiliation, and mailing address,
including ZIP coda
The abstract should not exceed one double-spaced paga
In the text, Fishery Bulletin style, for the most part,
follows that of the U.S. Government Printing Office Style
Manual. Fish names follow the style of the American
Fisheries Society Special Publication Na 12, A List of Com-
mon and Scientific Narfies of Fishes from the United States
and Canada, Fourth Edition, 1980.
Tfext footnotes should be typed separately from the text.
Figures and tables, with their legends and headings,
should be self-explanatory, not requiring reference to the
text. Their placement should be indicated in the right-hand
margin of the manuscript.
Preferably figures should be reduced by photography to
5% inches (for single-column figures, allowing for 50%
reduction in printing), or to 12 inches (for double-column
figures). The maximum height, for either width, is 14
inches. Photographs should be printed on high quality or
glossy paper.
Do not send original drawings to the Scientific Editor;
if they, rather than the photographic reductions, are needed
by the printer, the Scientific Publications Office will request
them.
Each table should start on a separate paga Consistency
in headings and format is desirable Vertical rules should
be avoided, as they make the tables more expensive to print.
Footnotes in tables should be numbered sequentially in
arable numerals, lb avoid confusion with powers, they
should be placed to the left of numerals.
Acknowledgments, if included, are placed at the end of
the text.
Literature is cited in the text as: Lynn and Reid (1968)
or (Lynn and Reid 1968). All papers referred to in the text
should be listed alphabetically by the senior author's sur-
name under the heading "Literature Cited." Only the
author's surname and initials are required in the literature
cited. The accuracy of the literature cited is the respon-
sibility of the author. Abbreviations of names of periodicals
and serials should conform to Serial Sources for the BIOSIS
Data Base. {Chemical Abstracts also uses this system, which
was developed by the American Standards Association.)
Common abbreviations and symbols, such as mm, m,
g, L, mL, mg, °C (for Celsius), %, "/oo, and so forth,
should be used. Abbreviate units of measure only when used
with numerals. Periods are only rarely used with abbrevia-
tions.
We prefer that measurements be given in metric units;
other equivalent units may be given in parentheses.
FORM OF THE MANUSCRIPT
The original of the manuscript should be typed, double-
spaced, on white bond paper Please triple space above
headings. We would rather receive good duplicated copies
of manuscripts than carbon copies. The sequence of the
material should be:
TITLE PAGE
ABSTRACT
TEXT
LITERATURE CITED
TEXT FOOTNOTES
APPENDIX
TABLES (Each table should be numbered with an
arable numeral and heading provided).
LIST OF FIGURES (Entire figure legends)
FIGURES (Each figure should be numbered with an
arable numeral; legends are desired)
ADDITIONAL INFORMATION
Send the ribbon copy and two duplicated or carbon copies
of the manuscript to:
Dr. Andrew E. Dizon, Scientific Editor
Fishery Bulletin
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 271
La Jolla, CA 92038
Fifty separates will be supplied to an author free of
charge and 50 supplied to his organization. No covers will
be supplied.
Contents— ConHntted
FARLEY, C. AUSTIN, PETER H. WOLF, and RALPH A. ELSTON. A long-term study
of "microcell" disease in oysters with a description of a new genus, Mikrocytos (g. n.),
and two new species, Mikrocytos mackini (sp. n.) and Mikroaytos rcmghleyi (sp. n.) . 581
Notes
WILLIAMS, AUSTIN B. Cojoined twin adult shrimp (Decapoda: Penaeidae) 595
GOULD, EDITH, DIANE RUSANOWSKY, and DONNA A. LUEDKE. Note on
muscle glycogen as an indicator of spawning potential in the sea scallop, Plac&pecten
magellanicus 597
JAMIESON, GLEN S., and ELLEN K. PIKITCH. Vertical distribution and mass mortality
of prawns, Pandalus platyceros, in Saanich Inlet, British Columbia 601'"
YANG, MEI-SUN. Morphological differences between two congeneric species of pleuro-
nectid flatfishes: Arrowtooth flounder, Atheresthes stomias, and Kamchatka flounder,
A. evermanni 608
McINTYRE, JOHN D, REGINALD R. REISENBICHLER, JOHN M. EMLEN, RICHARD
L. WILMOT, and JAMES E. FINN. Predation of Karluk River sockeye salmon by coho
salmon and char 611
• GPO 791-008
^^ATES O^ ^
ulletin
LIBRARY
JUL 7 1989
Woods Hole, Mass.
Vol. 86, No. 4
October 1988
MENDELSSOHN, ROY. Some problems in estimating population sizes from catch-
at-age data 617
BAYLIFF, WILLIAM H. Integrity of schools of skipjack tuna, Katsuwonus pelamis,
in the eastern Pacific Ocean, as determined from tagging data 631
CASTRO, MARGARIDA, and KARIM ERZINI. Comparison of two length-frequency
based packages for estimating growth and mortality parameters using simulated
samples with varying recruitment patterns 645
LIN, BIING-HWAN, and NANCY A. WILLIAMS. Specifying a functional form for
the influence of hatchery smolt release on adult salmon production 655
BEACHAM, TERRY D., CLYDE B. MURRAY, and RUTH E. WITHLER. Age, morphol-
ogy, developmental biology, and biochemical genetic variation of Yukon River fall chum
salmon, Oncorhynchus keta, and comparisons with British Columbia populations .... 663
YANG, M. S., and R A. LIVINGSTON. Food habits and daily ration of Greenland halibut,
Reinhardtius hippoglossoides, in the eastern Bering Sea 675
CROSS, JEFFREY N. Aspects of the biology of two scyliorhinid sharks, Apristums
brunnetLS and Parmatums xaniurus, from the upper continental slope off southern
California 691
SHANKS, ALAN L. Further support for the hypothesis that internal waves can cause
shoreward transport of larval invertebrates and fish 703
HOBSON, EDMUND S., and JAMES R. CHESS. Trophic relations of the blue rockfish,
Sebastes mystinus, in a coastal upwelling system off northern California 715
GRIMES, CHURCHILL B., CHARLES F IDELBERGER, KENNETH W ABLE, and
STEPHEN C. TURNER. The reproductive biology of tilefish, Lopholatilus chamae-
leonticeps Goode and Bean, from the United States Mid- Atlantic Bight, and the effects
of fishing on the breeding system 745
HARGREAVES, N. B. A field method for determining prey preferences of predators . . . 763
FORD, RICHARD F, BRUCE F PHILLIPS, and LINDSAY M. JOLL. Experimental
manipulation of population density and its effects on growth and mortality of juvenile
western rock lobsters, Panulirus cygnus George 773
BARSHAW, DIANA E., and DONALD R. BRYANT-RICH. A long-term study on the
behavior and survival of early juvenile American lobster, Homarus americanus, in
three naturaUstic substrates: eelgrass, mud, and rocks 789
{Continued on back cover)
Seattle, Washington
U.S. DEPARTMENT OF COMMERCE
C. William Verity, Secretary
NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION
William E. Evans, Under Secretary for Oceans and Atmosphere
NATIONAL MARINE FISHERIES SERVICE
James W. Brennan, Assistant Administrator for Fisheries
Fishery Bulletin
The Fishery Bulletin carries original research reports and technical notes on investigations 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. Beginning 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 instead of
being issued individually. Beginning with volume 70, number 1, January 1972, the Fishery Bulletin became a periodical, issued quarter-
ly. 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.
SCIENTIFIC EDITOR, Fishery Bulletin
Dr. Andrew E. Dizon
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 271
La Jolla, CA 92038
Editorial Committee
Dr. Jay Barlow
National Marine Fisheries Service
Dr. William H. Bayliff
Inter-American Tropical Tuna Commission
Dr. George W. Boehlert
National Marine Fisheries Service
Dr. Bruce B. Collette
National Marine Fisheries Service
Dr. Robert C. Francis
University of Washington
Dr. James R. Kitchell
University of Wisconsin
Dr. William J. Richards
National Marine Fisheries Service
Dr. Tim D. Smith
National Marine Fisheries Service
Mary S. Fukuyama, Managing Editor
The Fishery Bulletin (ISSN 0090-0656) is published quarterly by the Scientific PubHcations Office, National Marine Fisheries Service,
NOAA, 7600 Sand Point Way NE, BIN C15700, Seattle, WA 98115. Second class postage is paid in Seattle, Wash., and additional
offices. POSTMASTER send address changes for subscriptions to Fishery Bulletin, Superintendent of Documents, U.S. Government
Printing Office, Washington, DC 20402.
Although the contents have not been copyrighted and may be reprinted entirely, reference to source is appreciated.
The Secretary of Commerce has determined that the publication of this periodical 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.
Fishery Bulletin
CONTENTS
Vol. 86, No. 4
October 1988
MENDELSSOHN, ROY. Some problems in estimating population sizes from catch-
at-age data 617
BAYLIFF, WILLIAM H. Integrity of schools of skipjack tuna, Katsuwonus pelamis,
in the eastern Pacific Ocean, as determined from tagging data 631
CASTRO, MARGARIDA, and KARIM ERZINI. Comparison of two length-frequency
based packages for estimating growth and mortality parameters using simulated
samples with varying recruitment patterns 645
LIN, BIING-HWAN, and NANCY A. WILLIAMS. Specifying a functional form for
the influence of hatchery smolt release on adult salmon production 655
BEACHAM, TERRY D., CLYDE B. MURRAY, and RUTH E. WITHLER. Age, morphol-
ogy, developmental biology, and biochemical genetic variation of Yukon River fall chum
salmon, Oncorhynchus keta, and comparisons with British Columbia populations .... 663
YANG, M. S., and R A. LIVINGSTON. Food habits and daily ration of Greenland halibut,
Reinhardtius hippoglossoides, in the eastern Bering Sea 675
CROSS, JEFFREY N. Aspects of the biology of two scyliorhinid sharks, Apristurus
brunneus and Parmaturus xaniurus, from the upper continental slope off southern
California 691
SHANKS, ALAN L. Further support for the hypothesis that internal waves can cause
shoreward transport of larval invertebrates and fish 703
HOBSON, EDMUND S., and JAMES R. CHESS. Trophic relations of the blue rockfish,
Sebastes mystinus, in a coastal upwelling system off northern California 715
GRIMES, CHURCHILL B., CHARLES F IDELBERGER, KENNETH W ABLE, and
STEPHEN C. TURNER. The reproductive biology of tilefish, Lopholatilus chamae-
leonticeps Goode and Bean, from the United States Mid-Atlantic Bight, and the effects
of fishing on the breeding system 745
HARGRE AVES, N. B. A field method for determining prey preferences of predators . . . 763
FORD, RICHARD F, BRUCE F PHILLIPS, and LINDSAY M. JOLL. Experimental
manipulation of population density and its effects on growth and mortality of juvenile
western rock lobsters, Panulirus cygnus George 773
BARSHAW, DIANA E., and DONALD R. BRYANTRICH. A long-term study on the
behavior and survival of early juvenile American lobster, Homarus americanus, in
three naturalistic substrates: eelgrass, mud, and rocks 789
{Continued on next page)
Seattle, Washington
1988
For sale by the Superintendent of Documents, U.S. Government Printing 01 ice,
DC 20402— Subscription price per year: $16.00 domestic and $20.00 foreigr
issue: $9.00 domestic and $11.25 foreign.
Marine Biological Laboratory
LIBRARY
Cost pe.
JUL 7 1989
^jts Hole, Mass.
(Contents— C'owimMerf)
BLAYLOCK, ROBERT A. Distribution and abundance of the bottlenose dolphin,
Tursiops truncatus (Montagu, 1821), in Virginia 797
Notes
SEKI, MICHAEL R, and MICHAEL W. CALLAHAN. The feeding habits of two deep
slope snappers, Pristipomoides zonatus and P. auricilla, at Pathfinder Reef, Mariana
Archipelago 807
DITTY, JAMES G., GLEN G. ZIESKE, and RICHARD F. SHAW. Seasonality and depth
distribution of larval fishes in the northern Gulf of Mexico above latitude 26°00'N ... 811
ROGERS, CHRISTOPHER W, DONALD R. GUNDERSON, and DAVID A. ARMSTRONG.
Utilization of a Washington estuary by juvenile English sole, Parophrys vetulus. . . 823
SAFRIT, GLEN W, and FRANK J. SCHWARTZ. Length-weight relationships for gulf
flounder, Paralichthys albigutta, from North Carolina 832
WIEBE, PETER H. Functional regression equations for zooplankton displacement
volume, wet weight, dry weight, and carbon: a correction 833
GRAVES, JOHN E., MARIE A. SIMOVICH, and KURT M. SCHAEFER. Electro-
phoretic identification of early juvenile yellowfin tuna, Thunnus albacares 835
CAHOON, LAWRENCE B., and CRAIG R. TRONZO. A comparison of demersal zoo-
plankton collected at Alligator Reef, Florida, using emergence and reentry traps . . . 838
Index 847
Notice 860
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 used or purchased because of this NMFS publication.
SOME PROBLEMS IN ESTIMATING POPULATION SIZES
FROM CATCH-ATAGE DATA
Roy Mendelssohn'
ABSTRACT
A new method for estimating population sizes from catch-at-age data is given. The method treats the
observed population sizes as missing data and uses a combination of the Kalman filter and the EM
algorithm to derive maximum likelihood estimates of the parameters and minimum mean square error
estimates of the population sizes. The algorithm does not assume that the observation errors and the
errors in the population dynamics are uncorrected with equal variances, which is a common assumption
of existing techniques. A new parameterization for both recruitment and fishing mortality is given, based
on smoothness priors. Recruitment (or fishing mortality) is estimated as a nonparametric function of
time by calculating an "optimal" tradeoff between goodness-of-fit and smoothness of the function. The
algorithm allows for multiple sources of observations (fishing, surveys, etc.) and allows for missing data
in the observations, which can arise if the different sources of the observations occur on different time
scales. An example suggests that the new algorithm may better capture variation that is important when
using the population estimates to study the role of the environment (or other exogenous variables) on
the population dynamics.
I can address the motivation of this paper by con-
sidering a slightly modified version of a model pro-
posed by Colhe and Sissenwine (1983). Assume that
the underlying population dynamics satisfy
A^(a + 1, ^ + 1) = [Nia,t) - C(a,0] m + w{a,t)
a
1,A
(1)
where N{a,t) is the number of fish age a at time t,
C{a,t) is the catch of age a fish at time t, m =
exp{-m) is the mortality rate and the vector w{t)
= iw{l,t), . . .,w{A,t)y is a sequence of indepen-
dent, identically distributed normal random vectors
with mean 0 and covariance matrix Q, and for any
vector a, the notation a^ denotes the transpose of
the vector. We assume that the initial population
vector A'^(O) is gaussian with a mean of pi and
covariance 1.
The population itself is not observed. Instead we
observe that
n{a,t) = q(t)Nia,t) + v{a,t) (2)
where q is an unknown parameter and the vector
'Southwest Fisheries Center Pacific Fisheries Environmental
Group, National Marine Fisheries Service, NOAA, P.O. Box 831,
Monterey, CA 93942.
v{t) = (v(l,t), .. .,v (A, 0)^ is a sequence of indepen-
dent, identically distributed normal random vectors
with mean 0 and covariance matrix R. It is assumed
that E{v{t) w{ty) = 0; that is, the observation
error and the underlying randomness in the popula-
tion are uncorrelated. There is some interest in the
value of the estimates of the q{t) (or if mortality is
to be estimated, in the estimate of m) but the major
interest lies in estimating the unobserved popula-
tion sizes N{a,t). The estimates of the N{a,t) should
reflect not only the trend in the population, much
as a regression might, but also the period-to-period
variation of the population, such as might be related
to environmental changes. This will be the emphasis
throughout the paper.
The model described in Equations (1) and (2) dif-
fers from that of Collie and Sissenwine (1983) in that
I do not assume that the mortality rate is known;
here I allow the underlying population dynamics to
be random, and the observation errors in Equation
(2) to be additive rather than multiplicative. For
known in, Collie and Sissenwine (1983) suggested
minimizing
T A
1 1 (v(a,0' + wia,tf)
t=\ a=l
(3)
Manuscript accepted July 1988.
FISHERY BULLETIN: VOL. 86, NO. 4, 1988.
over the parameters (in our notation) 9 = (q,N{t))
617
FISIIEKV BLLLETIN: VOL. 86, NO. 4
where N{t) is the vector of unobserved population
sizes. The concern here is not whether the multi-
plicative or additive form of the model is more cor-
rect, but rather what elements of the model are data,
what are parameters, and how to calculate appro-
priate estimates of each.
If there were all-seeing observers, both the popu-
lation process A/^(0 and the observation process n{t)
would be available as data. Equations (1) and (2)
imply a sequence of conditional probability distribu-
tions/,v(A^(^ + 1) |A^(0, e,) and f,Mt)\ Nit), d^),
where d^, Q., are parameters of the distributions to
be estimated. Assuming additive gaussian errors,
then 01 would include the mean vector FN{t) and
a covariance matrix that can be calculated recursive-
ly (see below). The mortality rate m serves as a con-
straint on the form of the estimates of the mean vec-
tor, much as in the theory of regression. Similarly,
the parameters for the observation process are a
mean vector H(t)N{t) and a covariance matrix,
where H{t) in this case depends on the parameters
q{t) that constrain the estimates of the mean vec-
tor. With an all-seeing observer, both. n{t) and N{t)
are realized values of random vectors and hence, are
the data to be used to estimate the unknown param-
eters of the distributions, 6-^ and 02-
Thus, the unobserved population sizes are most
appropriately treated as missing data. The estima-
tion scheme proposed by Colhe and Sissenwine
(1983) treats the N{t)as parameters to be estimated.
Little and Rubin (1983, 1987: sec. 5.4) showed that
treating missing data as parameters in likelihood
equations does not produce maximum likelihood
estimates of the parameters unless the proportion
of missing data approaches zero as the sample size
increases. This is because much of the asymptotic
theory of maximum likelihood estimation depends
on the number of observations becoming large,
relative to the number of parameters. Little and
Rubin (1983, 1987) showed that for a regression-like
situation, the bias due to treating data as param-
eters can be quite large.
The alternate approach discussed by Little and
Rubin (1983, 1987) is to integrate out the missing
data from the complete data likelihood and maximize
this function over the parameters as usually defined
in estimation theory. This is the approach taken by
Shumway and Stoffer (1982a), who used the EM
(expectation-maximization) algorithm of Dempster
et al. (1977) and Kalman filtering to derive maxi-
mum likelihood estimates for the parameters of the
model and minimum mean square error estimates
of the missing data.
I can explain some of the problems with esti-
mating Equations (1) and (2), using the likelihood
of Equation (3). Under the gaussian assumptions of
the model, the complete data log-likelihood is given
by (Shumway and Stoffer 1982a)
log I I
1
(iV(0) - M(0))^ S
- 1
(iV(0) - M(0)
^loglQI -I I(iV(0 - FN{t - l)y
2 2 / = i
Q-i (Nit) - FN(t - 1))
rp 1
- log I i? I - - J.{n{t) - H{t)N{t)y R-^ in{t)
2 2 < = i
H{t)N{t))
(4)
whereF = ifil a.ndH{t) = qit)I. Similarly, the com-
plete data log-likelihood in Equation (3) by substi-
tution is
-111 Nit) - FNit - 1) I |- + I I nit)
t=\
- Hit)Nit)
(5)
Collie and Sissenwine (1983) noted that their estima-
tion scheme assumes that the process and observa-
tion errors have the same variance. However, from
Equations (4) and (5) it can be seen that they make
the stronger and unlikely assumption that both the
errors in the population dynamics and the errors in
the observation process are uncorrected. Further,
we can see from Equation (4) that when the Nit)
are treated as parameters, the estimates of the Nit)
depend on the observed data nit) for t = 1,T.
Following Shumway and Stoffer (1982a), the ex-
pected log-likelihood conditioned on the observed
data comprises three parts: a term due to estimating
the expected value of the initial population size.
618
MENDELSSOHN: ESTIMATING POPULATION SIZES
^log|I| - ltr{^~'[P(0\T) + (NiO) - m)
X (iV(0) - H)^]}, (6)
a term due to the unobserved dynamics,
^ 1 log I Q(0
2 (=1
- ^ 1 «r{Q(0-' [(A'(<|r) - FNit - 1\T))
2 ^=1
X (A^(^|T) - FNit\TV
+ Pit\T) + FP{t - 1\T)F' - Pit - 1\T)
X F^ - FP(t, t - 1\T)]}, (7)
and a term due to the observation process,
J I log 1 Rit) I
- - 1 tr{R{ty'[{nit) - H{t)Nit\T))
2 (=1
X {n{t) - Hit)Nit\T)y
+ Hit)P{t\T)Hit)]} (8)
where N{t\s) denotes
Nit\s) = E[Nit)\n{l),...Ms)l (9)
P{t\s) denotes
P(t|s) = E[iNit) - Nit\s))
X (Nit) - N{t\s)y I nil),
■■■Ms)], (10)
and Pit, t - l|s) denotes
Pit,t - lis) = EliNit) - Nit\s))
X iNit - 1) - Nit - l\s)y
X ln(l),...,n(s)]. (11)
As shown in Equations (6) through (8), the proper
estimates of the Nit) and the related covariance
matrices should be conditional expectations based
on all of the data rather than on only the data up
to time t. Assuming that all quantities are calculated
properly, estimates that include only the data up to
time t - 1 are termed "predicted" estimates, esti-
mates that include only the data up to time t are
termed "filtered" estimates, while estimates given
all the data are termed "smoothed" estimates. I
shall show below that the appropriate formulas for
predicted, filtered and smoothed estimates differ
significantly. Thus using Equation (5) as the likeli-
hood and treating the Nit) as parameters does not
produce proper estimates of the Nit).
In the rest of this paper, I review state-space
models and methods for estimating both the param-
eters and the unobserved components of the model.
A very readable background for what follows is
chapter 3 in Shumway (1988). The estimation
scheme described does not require that the compo-
nents of R, Q, and 2 have equal variance and are
uncorrected. Explicit estimates of these matrices
are given. Then I show that a variety of age-based
models proposed in the literature can be formulated
as a state-space model, but that the formulations as
presented make the same error of treating the un-
observed components as parameters, and assume
zero covariance in the errors. Auxiliary information
as in Deriso et al. (1985) can be put into this for-
mat. And I show that the state-space formulation
can include multiple observations of the population,
but where some of the observations are missing.
This can arise when the population is observed from
fishing and from a variety of surveys, but some of
the surveys are not done every year. This is essen-
tially the problem discussed in Methot^. I give true
maximum likelihood estimates for this model, allow-
ing the different observation processes to have dif-
ferent error structures and estimate the relative
weight that should be given each. This is a sig-
nificant advance over the procedure in Methot
(fn. 2).
A related paper is the analysis of Brillinger et al.
(1980) who use a modified Kalman filter and max-
imum likelihood estimation to estimate the average
birth and death rates and population structure of
Nicholson's blow-fly data when only total numbers
^Methot, R. 1986. Synthetic estimates of historical abundance
and mortahty for northern anchovy, Engraulis mordax. Adm.
Rep. LJ-86-29. Southwest Fisheries Center La JoUa Laboratory,
National Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla,
CA 92038.
619
FISHERY BULLKTIN; VOL. 86. NO. 4
of births and deaths at any time are available. In
their model, knowing the births is equivalent to
knowing the recruitment at each time period and
the deaths are observed directly without error. Also,
they appear to use the filtered estimates of the
population structure, while the smoothed estimates
are the minimum mean squared estimates given all
the data. P. Sullivan^ in his Ph.D. dissertation in-
dependently developed a length-based fishery model
using Kalman filtering and maximum likelihood
estimation.
I reiterate that the models considered here assume
additive errors, while much of the existing literature
prefers multiplicative errors, particularly for the
observation equation. The basis of this preference
appears to be that models with multiplicative errors
have given a better "fit" to the data for other esti-
mation schemes. In these estimation schemes the
errors are assumed to have equal variances and are
assumed to be uncorrelated. The better fit found for
multiplicative errors in these algorithms may be due
to these assumptions on the error variances. In ef-
fect, the errors are being scaled by the observed data
which suggests that the assumption of equal vari-
ances is incorrect. In the present formulation I
assume additive errors, but I can allow for errors
with unequal variances, for errors that depend on
the size of either the observed catch n{a,t) or on the
unobserved population N{a,t), and for missing
observations, and I can obtain simple to compute
standard errors of the estimated underlying popu-
lation sizes. The algorithm is also simple to program.
If multiplicative errors are assumed, it is more dif-
ficult to calculate exact maximum likelihood esti-
mates. Approximate likelihood methods (such as the
extended Kalman filter) have known undesirable
properties. When the full richness of the assump-
tions allowed in additive error models is used, it is
an open question if multiplicative errors are to be
preferred.
STATE SPACE MODELS
The State-Space Model can be written in the form
(Jazwinski 1970; Anderson and Moore 1979; Ljung
and Soderstrom 1983):
xit) = F{t)x{t - 1) + B{t)u{t -1) + Gw{t) (12)
y{t) = H{t)xit) + v{t) (13)
^P. Sullivan, Center of Quantitative Sciences, University of
Washington, Seattle, WA 98195, pers. commun. 1988.
where x(t) = (xiit), . . .,x^,(0)' is the p-dimensional
unobserved state of the system; u{t) = (Wi(0. • • •-
UjXf))' is a p-dimensional vector of deterministic in-
puts; ^(0 = iy\{t). . ■ ■,y,f{t)y is the observed data
of the system; w{t) = {w^{t), . . .,w^,{t)y is a se-
quence of zero mean normal vectors with common
covariance matrix Q; v{t) = {v^{t), . . .,v,^{t)y is a
sequence of zero mean normal vectors with common
covariance matrix R\ and F, B, G, and H are ap-
propriately dimensioned matrices that may depend
on an unknown parameter vector B. Note that q, the
dimension of the observation vector, can be larger
than p, the dimension of the state vector. Thus
several different observation processes of the under-
lying dynamics are allowed.
Using the same notation as in Equations (9) and
(10), the predicted, filtered, and smoothed estimates
of the state vector and the covariance matrices can
be calculated recursively as follows: for prediction
and filtering,
x{t\t - 1) = F{t)x{t - l\t - 1) + Bu{t - 1) (14)
P{t\t - 1) = F{t)P{t - l\t - l)Fity
+ GQ{t)G^ (15)
Kit) = P{t\t - l)H{ty
X {H{t)P{t\t - l)H{ty + 7^(0) ' (16)
xit\t) = xit\t - 1) + Kit)
X iyit) - Hit)xit\t - 1)) (17)
Pit\t) = Pit\t - 1) - Kit)Hit)Pit\t - 1) (18)
where x(0|0) = ^i and P(0|0) = I, and for smoothing.
Jit - 1) = Pit - l\t - l)HityiPit\t - l))-i (19)
xit - 1\T) = xit - l\t - 1) + Jit - 1)
X ixit\T) - xit\t - 1)) (20)
Pit - 1\T) = Pit - l\t - 1) + Jit - 1)
X iPit\T) - Pit\t - l))Jit- 1)^(21)
The predicted state variable xit\t - 1) differs from
the quantity often used as the predicted value in the
fisheries literature in that it is based on the last
period's filtered estimate rather than on the last
620
MENDELSSOHN: F.STIMATIN(; I'OPILATION SIZES
period's predicted estimate. (For example, Deriso
et al. (1985) suggested using backward VPA or
cohort analysis to obtain predicted values.) The
filtered estimate x{t |0 is a weighted average of the
predicted value of x{t\t - 1) and the observed er-
ror in estimating y{t), where the weighting term
K{t) (the Kalman gain matrix) is regression-like.
Similarly, the covariance of the estimate, as meas-
ured by P{t \t - 1), increases due to the prediction,
but decreases by the amount K{t)H{t)P{t\t - 1)
after the observation has been made. The smoothed
estimate x{t\T), which is the correct estimate of the
underlying population (it satisfies the conditional ex-
pectation), is found by a backward recursion on the
filtered estimates, where the filtered estimate is ad-
justed by a regression on the error between the
smoothed and predicted estimates of the following
period. Thus the smoothed estimates correct the
predicted estimates both by the error in predicting
the observed data as well as by the error in pre-
dicting the underlying population when using the
filtered estimates of the underlying population. The
square roots of the diagonal terms of the various
P matrices produced by the Kalman recursions are
the standard errors of the predicted, filtered, and
smoothed estimates of the population.
Equations (12) and (13) are the basic form of the
state-space model. It is a simple extension to the
model to allow any of the matrices F, B, G, or H
to be nonlinear functions of the past values of the
yit) (see, for example, Shiryayev (1984), section
VI. 7), to allow the error vectors w{t) and v{t) to de-
pend on past values of the y{t) (Shiryayev 1984), or
to allow the v{t) to depend on the underlying state
vector x{t) (Zehnwirth 1988).
In a typical fisheries problem, the matrix H{t)
represents fishing. If some age-specific measure of
effort E{a,t) is known, then H(t) is a diagonal
matrix with E{a,t) on the diagonals. Or it may be
assumed that the exploitation rate is of the form
s{a)E{t), where E{t) is known and the s(a) values
are to be estimated. Then for given values of s(a),
the matrix H(t) has s{a)E{t) on its diagonals. The
matrix F{t) is formed in a similar manner to repre-
sent the population dynamics.
In some parameterizations, it is assumed that a
known vector is subtracted from the state vector
either before or after the effect of F on the popu-
lation. For example, the known vector might be the
catch from the previous time period. The extension
to the Kalman filter in this case is straightforward,
an example of which can be found in Jazwinski
(1970). Essentially, all predicted estimates of the
state are corrected by the constant amount. The
covariance and gain calculations are unaffected by
the known vector.
The Kalman filter. Equations (14) through (21),
assumes that the matrices F, B, G, H, R, Q, and I,
and the vector jj. are known. For fisheries problems,
the matrices F and H usually depend on a set of
parameters to be estimated (e.g., F = ml), and R,
Q, and ^ are to be estimated. Let 0 be a vector con-
taining the parameters that F and H depend on, and
let 0 = (0,i?,Q,/i) be the total parameters of the
model. Shumway and Stoffer (1982a) showed that
conditional on 0, the complete data likelihood is
given by Equation (4). They apply a result of Demp-
ster et al. (1977), which shows that maximum like-
lihood estimates of the parameters can be obtained
by finding the conditional expectation (the E-step)
of the complete data likelihood with respect to the
missing "data" (in this case the missing data are the
sufficient statistics of the normal distribution) and
alternately estimating the expected value of the
missing data, and then maximing the likelihood (the
M-step) using the completed data. Shumway and
Stoffer (1982a) showed that the expected conditional
log-likelihood is given by Equations (6) through (8).
All of the terms in this likelihood, for a given value
of 0, can be found by the Kalman filter. Moreover,
given these values, the maximization problem is a
deterministic one.
If we assume that the matrix F is independent of
time and unrestricted, then Shumway and Stoffer
(1982a) showed that the maximization step is accom-
plished by setting
F = 5,(l)5,_i(0)-i
Q = (5,(0) - S,{l)Sf_\{0)St(iy)/T
T
R = T-'^ [{y{t) - H{t)x{t\T)
t=\
X {y{t) - H{t)x{t\T)y
+ H{t)P{t\T)H{ty]
M = x{0\T).
T
where 5,(j) = I iPit,t - j\T)
+ xit\T)xit -JlTY).
(22)
(23)
(24)
(25)
(26)
621
FISHKRY BULLETIN: VOL. m. NO. 4
Shumway and Stoffer (1982a) gave a recursive for-
mula for calculating P(<,^ - 1|T) while performing
the backward smoothing recursion.
If we assume that F is constrained to be of the
form F = TfiD where m is a constant and D is a
known matrix, then Shumway and Stoffer (1982b)
showed that
Q = (C - rhBD' - mDB' + rh'^DAD') (27)
m =
tr{Q-^DAD')
(28)
Equations (27) and (28) can be solved by taking
an initial guess for m, then iteratively solving for
Q and m until the values converge.
Finally, we can make explicit the effect of assum-
ing equal variances and no co variances for both w{t)
and v(t). For given estimates of Q and R, since both
are square, symmetric matrices, they can be fac-
tored as
Q = UDU'
R = LL'
where U'ls an upper triangular matrix, Z) is a diag-
onal matrix, and L is the lower triangular square
root of R. To obtain an underlying population
dynamic that has an uncorrelated error vector w{t)
and uncorrelated observations with variances of 1,
we make the following transformations:
G = GU
y{t) = L-'yit)
H{t) = L-'Hit)
v{t) = L-h{t)
and replace G, w, y, H, and v in Equations (14)
through (21) with the transformed values. Then v{t)
has covariance matrix 7, and 'w{t) has co variance
matrix D. The assumption that both the error in the
dynamics and in the observations are equal, further
constrains the values of D to be identical. This is
a very strong assumption.
AN EXAMPLE
As an example of these methods, I use the data
622
for Pacific mackerel published in Parrish and
MacCall (1978). I emphasize that I am only using
these data for illustrative purposes and do not claim
to be making a careful, thorough reexamination of
the problem. Though m can be estimated using
Equation (28), I assume that the value of m is known
a priori. If I were to use a different value of m, it
would be difficult to judge to what extent the new
estimates differ solely due to the different mortal-
ity rate, rather than due to the estimation scheme.
I assume, as in the reference, that the mortality rate
m is equal to 0.5, so that the F matrix in my nota-
tion is a matrix with a value of 0.6065 in position
(i, i - 1), i = 2, . . . ,7, corresponding to the under-
lying dynamics for age groups 1 through 6.
Recruitment in Pacific mackerel is highly variable.
I want to obtain estimates of recruitment that ac-
curately reflect this variability while still being con-
sistent with the observed data. Also, I do not want
to a priori assume a functional relationship between
recruitment and population size. To this end, I
assume that the recruitment time series, after tak-
ing differences of a given order, is a random vari-
able, i.e.,
V'^r(0 = w{t)
(29)
where w{t) is a. normal random variable with a mean
of zero and with an unknown variance o" and V'''
denotes A;th order finite differencing. Akaike (1979)
originally showed that this formulation is the dis-
crete equivalent of fitting a spline to the data (in
this case as a function of time), where the estimate
of the variance o" expresses the tradeoff between
the degree of smoothness in the fitted curve with
fidelity to the observed data. In this approach, k and
the variance of w{t) are treated as hyperparameters
of the model. A fitting criterion such as AIC is then
used to determine the best value of k given the data.
Following Kitagawa and Gersch (1984), I could use
smoothness priors to more generally decompose
recruitment as
r{t) = Tit) + Sit) + 4(0 + wit)
(30)
where Tit) is a trend term (as in Equation (29)), Sit)
is a seasonal term, and |(0 is an irregular stationary
term. A decomposition such as Equation (30) would
be useful, for example, in modeling the monthly an-
choveta recruitment considered in Mendelssohn and
Mendo (1987). However, for convenience in this
paper, I restrict recruitment to be of the form in
Equation (29).
MENDELSSOHN: ESTIMATING POPULATION SIZES
The relationship between this "smoothness
priors" approach, "smoothing spHnes", and other
penalized likelihood methods is discussed further for
a variety of contexts in Brotherton and Gersch
(1981), Kitagawa and Gersch (1984, 1985, 1988),
Ansley and Kohn (1986), and Kohn and Ansley
(1987, 1988). Wahba (1977) and O'Sullivan (1986)
discussed the relationship between generalized
cross-validation, penalized likelihood functions, and
determining the tradeoff between smoothness and
fit.
For k = 0, Equation (29) models recruitment as
a random variable around a fixed but unknown mean
value. For k = 1, Equation (29) models recruitment
as a random walk with unknown mean level and drift
(variance). Higher values of /c have similar interpre-
tations. Values of A' higher than two or three rarely
need to be considered, since these include the dis-
crete equivalent of cubic splines. Cubic splines can
approximate most functionals (in this case of time)
to a reasonable degree of accuracy.
For this example, I assume k = 1, so that
r(0 = r{t - 1) + w{t)
(31)
which is a random walk with unknown variance. (A
more complete analysis of this data would probably
also include an irregular stationary term as in Equa-
tion (30) and determine the "best" order of differ-
encing using a given criterion.) Equation (31) can
be incorporated into the state space model by let-
ting the (1,1) element of the matrix F be equal to
1. The matrices H{t) are diagonal matrices whose
values are calculated from table 13 in Parrish and
MacCall (1978). Because I am assuming that the
estimates of F are known, then the value of Q for
the M step is maximized as
T-\S,iO) - S,{1)F^ - FS,il) + FS,_,{0)F^].
(32)
As in Parrish and MacCall (1978), I treated age
groups 4 through 6 as fully selected by the fishery,
and will refer to these age groups as "adults".
Similarly, I refer to the number of age-1 fish at the
start of the season as the number of recruits. I
assume that F and H{t) are known, so the estima-
tion problem is reduced to determining the means
of the initial population sizes and the values of the
two covariance matrices Q and R.
The resulting maximum likelihood estimates of Q
and R (Tables 1, 2) show that the variances of the
error terms differ by up to two orders of magnitude,
hardly meeting the usual assumption of equal vari-
ances. Moreover, the covariances (expressed as cor-
relations in the tables) are quite high, so that using
- as a weighting factor will not be adequate. The
a
predicted, filtered, and smoothed estimates of the
adults (Fig. 1) are very similar, reflecting that the
errors have been "filtered out" over time by the
population dynamics.
Table 1 .—Estimated values of the matrix O presented as a variance-correlation matrix. The diagonal
terms are the variances, and the off-diagonal terms are the cross-correlations.
Age 0
Age 1
Age 2
Age 3
Age 4
Age 5
Age 6
Age 0
8.72E + 09
-0.9075
0.2429
0.0592
0.1785
-0.2096
0.0249
Age 1
4.59E + 09
-0.1750
-0.0574
-0.1658
0.1803
-0.0432
Age 2
1.07E-I-09
0.0396
0.2416
0.1109
0.0675
Age 3
5.28E + 08
0.7030
0.5964
0.3140
Age 4
1.90E-t-08
0,6717
0.3944
Age 5
3.08E + 07
0.4854
Age 6
2.18E + 06
Table 2. — Estimated values of the matrix R presented as a variance-correlation matrix.
Age 0
Age 1
Age 2
Age 3
Age 4
Age 5
Age 6
Age 0
7.30E + 07
-0.0497
0.8505
0.7367
0.6594
0.5120
-0.4019
Age 1
2.62E + 07
0.1594
0.0896
-0.0544
-0.3411
0.4325
Age 2
4.95E + 06
0.6764
0.6412
0.3830
-0.2418
Age 3
3.43E + 06
0.7286
0.4908
-0.2260
Age 4
7.11E + 05
0.5708
-0.2030
Age 5
2.05E-I-05
-0.1331
Age 6
5.40E + 04
623
FISHERY BULLETIN: VOL. 86, NO. 4
on
Q
<
O
oi
w
CQ
C/5
w
50000
40000
30000 -
20000
10000 -
0
1930
—
- FILTERED
PREDICTED
; 1
All /
1
1 ^
r 1
1940
1950
1960
1970
TIME
Figure 1.— Estimated number of adults, 1939-65, using values from Parrish and MacCall.
The predicted, filtered, and smoothed estimates
of the recruits (Fig. 2), unlike those of the adults,
are not similar. The filtered and smooth estimates
are often indistinguishable, but there are some years
(such as 1941 and 1943) where there are significant
differences. The predicted values are very smooth,
tending to emphasize the trend in the recruitment.
In an analysis of recruitment estimates produced by
standard cohort analysis of the anchoveta off Peru,
Mendelssohn and Mendo (1987) found the estimates
600000
SMOOTHED
FILTERED
PREDICTED
1930
1940
1950
1960
1970
TIME
Figure 2.— Estimated number of recruits, 1939-65, using values from Parrish and MacCall.
624
MENDELSSOHN: ESTIMATING POPULATION SIZES
to be far too smooth. The results of this present ex-
ample may explain their observation. If these values
were used in a subsequent analysis, say to determine
the role of the environment on recruitment, a totally
false picture of this relationship could emerge. I em-
phasize that the predicted estimates are calculated
from the previous season's filtered values, whereas
it is often true in the fisheries literature that the
predicted values are estimated from the previous
predicted values, rather than from the filtered
values. Direct comparison with the estimate in
Parrish and MacCall (1978) are difficult because
restricting recruitment to be of the form (Equation
(29)) with k fixed, rather than the more general form
(Equation (30)) with k variable, may not be appro-
priate for the Pacific mackerel data. But overall,
their estimates tend to resemble the smoother trend
of my predicted estimates.
RELATIONSHIP TO OTHER LITERATURE
AND A NEW PARAMETERIZATION
If the models are restricted to additive errors,
then most of the simpler difference equation models
proposed in Collie and Sissenwine (1983), Deriso
et al. (1985), Fournier and Archibald (1982), Four-
nier and Doonan (1987), and among others can be
formulated as I did. Some of these models assume
recruitment is a nonlinear function of the underly-
ing population, which cannot be handled in this
model without some modifications (suggested
below). However, all of these authors treat the
underlying population sizes as parameters of the
data rather than as missing data. As discussed
earlier, it is questionable whether this will produce
proper estimates of the underlying populations.
Biases from treating missing data as parameters in
a regression setting are explicitly discussed in Little
and Rubin (1983, 1987).
A very broad class of possible models that can be
selected to model catch-at-age data are given by
Schnute (1985). He correctly identified the values
that are parameters of the difference equations he
discusses, and these are sufficient for estimating the
likelihood (if evaluated properly). However, if we
assume observation error, then it can be shown (see,
for example, Shumway 1988) that the innovations
are determined by predictors calculated from the
previous period's filtered, rather than predicted
values. Moreover, the minimum mean squared error
estimates of the underlying populations, as dis-
cussed earlier, are the smoothed estimates. It ap-
pears that Schnute (1985) used the predicted or.
at best, the filtered estimates of the underlying
populations.
A popular parameterization that appears to have
been first suggested by Doubleday (1976) is to
assume that the observation matrix H{t) is of the
form H{t) = {s{a)f{t)} where s{a) is an age-depen-
dent selectivity factor and/(0 is a time-dependent
exploitation rate. These values can be found by using
a minimization routine during the M-step of the
algorithm. However, it should be noted that the
estimate of/(0 for each t will depend on R and that
the estimate ofR will depend on both s(-) and/(-),
so that either R, s, and / should be solved for
together, or else they should be successively solved
for using Equation (7) while holding the other
parameter values fixed.
Alternatively, Equation (7) can be differentiated.
Then for given values of/(0, the optimum value of
the vector s(a) at each iteration are the diagonals
of the matrix S given by
S = AB'
(33)
where A = 1 y{t)x{t\Tyfit), and 5 = 1 f^{t)
t = i
( = 1
{P{t\T) + x{t\T)xit\Ty. However, this is an un-
constrained estimate and does not guarantee that
s(a) is between (0,1). It can be shown that the op-
timal solution is to set s(a) at zero if s (a) is negative
or to 1 if s(a) is greater than 1.
For fixed values of sia) and R, Equation (7) is
maximized when/(0 takes the value
fit)
tr{R-^y{t)x{t\TYS')
tr{R-^S{P{t\T) + xit\T)xit\Ty)S')
(34)
where the matrix iS is as above. This is the uncon-
strained solution. The constrained solution again is
to force the estimate to He within the closed inter-
val (0, 1) as with the estimate of the s{a). Since the
estimates oiR, s{a), and/(0 are interrelated, I have
found it to be a workable procedure to first estimate
Q as given above and then for a given number of
iterations, iteratively solve for /(f) then s(a). When
these values stabilize, estimate R using the formula
given above. While this procedure does not neces-
sarily maximize Equation (7), it is sufficient for the
generalized EM algorithm the new values increase
the function given in Equation (7).
As with the original estimates, the smoothed and
filtered estimates of recruitment (Fig. 3) are close
625
FISHERY BI'LLKTIN: VOL. 86, NO, 4
400000 r-
g 300000
U
w
p-
2 200000
w
CQ
IS
00
W
100000
0
SMOOTHED
HLTERED
PREDICTED
1930
1940
1950
1960
1970
TIME
Figure 3.— Estimated number of recruits 1939-65, assuming q(a, t) = f(t)s(a).
to each other, while the predicted values are too
smooth. The estimated values of/(f ) (Fig. 4) general-
ly decline with time. All three estimates are signifi-
cantly different from the previous estimates. Years
where relative highs and lows occur differ, show-
ing the sensitivity of this class of models to the
assumed form of the H matrix. Since the estimates
oi R, s{a) and/(0 are interrelated, I suspect that
the parameter estimates are highly correlated and
hence unstable. While I have not calculated the
parameter covariance matrix, it should be checked
for any serious analysis using these techniques.
0.4
^ 0.3
0.2
0.1
H
<
Z
o
<
O
J
CL
X
PU
O
H
<
Ol,
pL.
on
U
0.0
1930
1940
1950
1960
1970
TIME
Figure 4.— Estimated values of /(O when q(a, t) = f(t)s(a).
626
MENDELSSOHN: ESTIMATING POPULATION SIZES
While the parameterization H{t) = {f{t)s{a)}
greatly reduces the number of parameters, I still
face the same problem as when I treated the un-
observed population as parameters: each new period
adds another parameter to the model. The model
still appears to be overparameterized, and the
asymptotic theory for maximum likelihood estima-
tion may be invalid.
As with the recruitment estimates, the number
of effective parameters can be reduced by adding
a smoothness prior on/(0- However, /(^ is a pro-
portion and not likely to be a normal variable. The
f{t) are also constrained to lie in the interval (0, 1),
so a transformation to an unconstrained variable
would also be desirable. If/(0 is a binomial random
variable then arcsin {\Jf{t )) is approximately normal
with nearly equal variance. This suggests the trans-
formation of variables
y{a,t) = s{a)sin'^{eit))x{a,t).
(37)
fit) = sinHeit))
(35)
where eit) is an unconstrained normal variable.
Then the smoothness prior becomes
Vh{t) = w(t)
(36)
(a smoothness prior that includes a seasonal com-
ponent or irregular stationary part, as in Equation
(30), can also be used), and for any age class, the
observation equation becomes
The underlying population dynamics must also be
expanded to include the smoothness prior constraint
(Equation (36)).
Unfortunately, the observation equation is no
longer linear in the state vector. The smoothness
prior is a prior distribution on the f{t), and a full
Bayesian analysis can be done to obtain the overall
distribution. The variance o{w(t) is then treated as
a hyperparameter in the analysis.
A simpler approach is to evaluate the filter equa-
tions approximately by using any one of a number
of nonlinear filters (see Anderson and Moore 1979).
One that is easy to implement, given the nonlinear-
ities in this problem, is the extended Kalman filter
(EKF), which at each time period just linearizes all
the nonlinear terms around the value of the pre-
dicted state vector. The EKF, however, can have
divergence problems and is not guaranteed to find
the true penalized likelihood estimates.
When using the EKF, it works to make a forward
and backward pass of the filter given the current
estimates of f{t) and x{t\T), and then to estimate
s{a) and R as before. I tested the algorithm on the
mackerel data with k = 1 in the constraint (Equa-
tion (36)). The resulting estimates (Fig. 5) are similar
to the previous estimates, but the estimated values
of /(O (Fig. 6) are less variable with a stronger trend
than before. It is clear from Figure 5 that the
400000
00
§ 300000
u
w
O
a: 200000
w
CO
H 100000
W
0
SMOOTHED
FILTERED
PREDICTED
1930
1940
1950
1960
1970
TIME
Figure 5.— Estimated number of recruits using a first order spline for estimating /{<).
627
FISHERY BULLETIN: VOL. 86, NO. 4
X
u
<
U
O
a;
u
z
o
:s
o
u
IX,
H
W
.5 r-
1.0
0.5
0.0
1930
1940
1950
1960
1970
TIME
Figure 6.— Spline estimates of f{t.).
smoothness prior estimate has shifted some of the
variation in the observed data to variations in the
underlying population dynamics rather than varia-
tions in/(^). Further research needs to be done to
see which of these estimates is the most "valid".
The EM algorithm in general can be sensitive to
the initial values given the parameters (see Wu
1983), and I have found that the fixed value of 1 can
also affect the estimates. Initial combinations of p<,
f{t), and s{a) that are totally inconsistent v^ith the
observed catch data can cause the algorithm to find
a local maximum. This can be avoided by ex-
perimenting with several, very different starting
values and determining if they converge to the same
estimates.
If recruitment is thought to be a linear function
of the previous population size, then there is no
problem including this in the Kalman filter. If
recruitment is a nonlinear function of the previous
population, then the EKF can again be used to ap-
proximately determine the conditional expectations
needed for the EM algorithm.
If information is available from a variety of
sources, say from fishing and from surveys, as in
Methot (fn. 2), then each of the vectors and matrices
can be partitioned to represent this situation. For
example, let y^it) be the observed catches from a
survey and yf{t) the observed catches from fishing.
Let y{ty = {y sit), y fit))', and partition the H matrix
similarly. Then the diagonal blocks of H will con-
tain the observation dynamics for the survey and
for fishing, while the off-diagonal blocks will be zero.
Given these modifications, all the algorithms de-
scribed previously in this paper can be used to derive
estimates for this situation.
Research or other surveys of the fishery usually
occur less frequently than does commercial fishing,
causing part of the vector y{t) to be missing at given
times periods. Shumway and Stoffer (1982a) and
Shumway (1988) gave a straightforward modifica-
tion of the Kalman filter for this case.
DISCUSSION
I have introduced a new method for estimating
population sizes from catch-at-age data that in-
cludes, if additive errors can be assumed, many of
the previous difference equation models. I show that
it is incorrect to treat the unobserved population
sizes as parameters to be estimated rather than as
missing data. I also show that the minimum mean
square estimates of the population sizes are the
smoothed estimates rather than the predicted
estimates suggested in many papers. The model
assumes neither equal variances in the errors in the
population dynamics nor in the observation errors
and does not require that the errors be uncorrected.
For Pacific mackerel, the smoothed estimates are
shown to be much more variable than the predicted
estimates.
628
MENDELSSOHN: ESTIMATING POPULATION SIZES
I also suggest a new parameterization for the age-
specific exploitation rate q{a,t). If it is assumed that
q{a,t) = s{a)f{t), then the model is over determined.
I put a smoothness prior on f(t) in order to obtain
a tradeoff between the degree of smoothness mf{t)
as a function of time versus fidelity to the data. The
degree of differencing can be treated as a hyper-
parameter of the model to determine the optimal
amount of differencing given the data.
An advantage of the approach of this paper is that
the calculations are straightforward and simple to
program, and explicit formulas are given for the op-
timal parameter values at each iteration of the EM
algorithm. Additional optimization software is not
required to perform the calculations. Moreover,
properties of the Kalman filter and the EM algo-
rithm are well known. There is a large literature giv-
ing variants of the filter to calculate sensitivity to
model misspecification, and recursive formulas for
the derivatives with respect to a given parameter
of the model also exist in the literature.
It is also simple to include environmental variables
into the formulation, either as additional state vari-
ables or as fixed effects in the observation equation
or both (see Sallas and Harville [1988] on how to
estimate the fixed effect parameters within the con-
text of Kalman filtering). Thus the influence of the
environment can be modeled directly, rather than
resorting to the conventional practice of obtaining
population estimates first and correlating these
estimates with the environmental variables second.
A disadvantage of my approach is that there is
no guarantee that any of the estimates of the under-
lying population sizes will be positive. The popula-
tion sizes are treated as normal random variables,
and it is quite possible for the additive corrections
in the filtered or smoothed estimates to make small
population sizes negative if the observation error is
large. P. Sullivan (fn. 3) has found that for a length-
based model the Kalman filtering approach works
best when there are pulses in the recruitment, that
is, when the population is not in equilibrium. The
likelihood surface is such that without recruitment
pulses it is difficult to estimate the parameters of
the growth-curve. Most fisheries are not in equilib-
rium, however. As the models in this paper do not
contain a growth-curve, it is unclear if a similar find-
ing will be valid.
Some of my results suggest that the estimates are
sensitive to the form of the model chosen for the
population dynamics. This is not surprising, because,
unlike most missing data problems, the missing part
of the data is never observed directly, but only
through the presumed form of the dynamics. For
example, when modeling catch (or catch per unit ef-
fort) against an environmental variable, catch data
often are not available for all periods. But there are
at least some periods when both variables are ob-
served, which can be used to estimate the relation-
ship between the two sets of variables. This rela-
tionship is used to produce the smoothed estimates
of the missing data. For estimating population sizes
from catch-at-age data, the a priori estimate of the
form of the observation equation replaces this em-
pirically derived estimate.
In many of the references cited, multiplicative
errors are preferred in the observation equation
because variances appear to change with the size
of the population. My experience is that relaxing the
assumption of equal, uncorrelated errors appears to
at least partially take into account the observed
differences. If the model estimates are not satisfac-
tory, assuming additive, gaussian errors, then the
regular EM algorithm can be used to properly esti-
mate the smoothed estimates of the underlying
population. However, the EM algorithm requires the
complete data likelihood as well as the expectation
of the log-likelihood with respect to {y{l), . . .,y(T)).
In multiplicative models, assumptions about the
error structure can lead to very complicated multi-
variate distributions for the complete data due to
the Jacobian of the transformation. The conditional
expectation of the log-likelihood may have to be
evaluated by numerically integrating a nontrivial
multiple integral. Certainly, as a first pass, the
simpler techniques of this paper would appear to
have a lot to offer as an alternative.
LITERATURE CITED
Akaike, H.
1979. Likelihood and the Bayes procedure. In J. M. Ber-
nardo, M. H. DeGroot, D. V. Lindley, and A. F. M. Smith
(editors), Bayesian statistics, p. 143-166. Univ. Press,
Valencia.
Anderson, B. D. 0., and J. B. Moore.
1979. Optimal filtering. Prentice Hall, Englewood Cliffs,
NJ, 357 p.
Ansley, C. F., and R. Kohn.
1986. On the equivalence of two stochastic approaches to
spline smoothing. In J. Gani and M. B. Priestly (editors),
Time series and allied processes, p. 391-405. J. Appl. Prob.
23A.
Brillinger, D. R., J. Guckenheimer, p. Guttorp, and G.
OSTER.
1979. Empirical modeling of population time series data: The
case of age and density dependent vital rates. In Some
mathematical questions in biology, p. 65-90. Lectures on
Mathematics in the Life Sciences 13, American Mathe-
matical Society, Providence.
629
FISHERY BULLETIN: VOL. 86, NO. 4
Brotherton, T., and W. Gersch.
1981 . A data analytic approach to the smoothing problem and
some of its variations. Proc. 20th IEEE Conf. Decis. Con-
trol, p. 1061-1069.
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.
Dempster, A. P., N. M. Laird, and D. B. Rubin.
1977. Maximum likelihood from incomplete data via the EM
algorithm. J. R. Stat. Soc. 39(Ser. B):l-38.
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.
DOUBLEDAY, W. G.
1976. A least squares approach to analyzing catch at age data.
Res. Bull. Int. Comm. Northwest Atl. Fish. 12:69-81.
FOURNIER, D. A., AND C. ARCHIBALD.
1982. A general theory for analyzing catch at age data. Can.
J. Fish. Aquat. Sci. 39:1195-1207.
FOURNIER, D. A., AND I. J. DOONAN.
1987. A length-based stock assessment method utilizing a
generalized delay-difference model. Can. J. Fish. Aquat.
Sci. 44:422-437.
Jazwinski, a. H.
1970. Stochastic processes and filtering theory. Acad.
Press, N.Y., 376 p.
Kitagawa, G., and W. Gersch.
1984. A smoothness priors - state space modeling of time
series with trend and seasonality. J. Am. Stat. Assoc. 79:
378-389.
1985. A smoothness priors time varying AR coefficient
modeling of nonstationary time series. IEEE Trans.
Autom. Control AC-30:48-56.
1988. Smoothness priors in time series, /w J. Spall (editor),
Bayesian analysis of time series and dynamic models, p.
431-476. Marcel Dekker, N.Y.
Kohn, R., and C. F. Ansley.
1987. A new algorithm for spline smoothing and interpola-
tion based on smoothing a stochastic process. SIAM J. Sci.
Stat. Comput. 8:33-48.
1988. The equivalence between Bayesian smoothness priors
and optimal smoothing for function estimation. In J. Spall
(editor), Bayesian analysis of time series and dynamic
models, p. 393-430. Marcel Dekker, N.Y.
Little, R. J. A., and D. B. Rubin.
1983. On jointly estimating parameters and missing data by
maximizing the complete data likelihood. Am. Stat. 37:
218-220.
1987. Statistical analysis with missing data. Wiley, N.Y.,
278 p.
Ljung, L., and T. Soderstrom.
1983. Theory and practice of recursive identification. M.I.T.
Press, Cambridge, MA, 529 p.
Mendelssohn, R., and J. Mendo.
1987. Exploratory analysis of anchoveta recruitment off Peru
and related environmental series. In D. Pauly and I.
Tsukayama (editors). The Peruvian anchoveta and its up-
welling ecosystem: three decades of change, p. 294-306.
ICLARM, Manila.
O'SULLIVAN, F.
1986. A statistical perspective on ill-posed problems. Stat.
Sci. 4:502-527.
Parrish, R. H., and a. MacCall.
1978. Climatic variation and exploitation in the Pacific
mackerel fishery. Calif. Dep. Fish Game, Fish Bull. 167,
110 p.
Sallas, W. M., and D. a. Harville.
1988. Noninformative priors and restricted maximum likeli-
hood estimation in the Kalman filter, hi J. Spall (editor),
Bayesian analysis of time series and dynamic models, p.
477-508. Marcel Dekker, N.Y.
Schnute, J.
1985. A general theory for analysis of catch and effort data.
Can. J. Fish, Aquat. Sci. 42:414-429.
Shiryayev, a. N.
1984. Probability. Springer-Verlag, N.Y., 577 p.
Shumway, R. H.
1988. Applied statistical time series analysis. Prentice-Hall,
Englewood Cliffs, NJ, 379 p.
Shumway, R. H., and D. S. Stopfer.
1982a. An approach to time series smoothing forecasting
using the EM algorithm. J. Time Ser. Anal. 3:253-264.
1982b. An algorithm for parameter estimation and smooth-
ing in space-time models with missing data. U.S. Dep. Com-
mer., NOAA, Final Rep. P.O. 82-ABA-1190.
Wahba, G.
1977. A survey of some smoothing problems and the method
of generalized cross-validation for solving them. In R.
Krishnaiah (editor). Applications of statistics, p. 507-
524. North Holland Publishing Co., Amsterdam.
Wu, C. F.
1983. On the convergence properties of the EM algorithm.
Ann. Stat. 11:95-103.
Zehnwirth, B.
1988. A generalization of the Kalman filter for models with
state-dependent observation variance. J. Am. Stat. Assoc.
83:164-167.
630
INTEGRITY OF SCHOOLS OF SKIPJACK TUNA,
KATSUWONUS PELAMIS, IN THE EASTERN PACIFIC OCEAN,
AS DETERMINED FROM TAGGING DATA
William H. Bayliff'
ABSTRACT
Little information concerning the integrity of schools is available for any species of fish. In this study
the integrity of schools of skipjack tuna, Katsuwonus pelamis, was analyzed with data for returns of
tagged fish which had been in the same schools when originally tagged. Two methods, the first using
Chi-square contingency tests and the second using binomial homogeneity tests, were employed. From
the results obtained with the first method it appears that after 1 month at liberty the tagged and un-
tagged fish were randomly mixed with one another in some cases and after 3 to 5 months at liberty
they were randomly mixed with one another in nearly all cases. The results obtained with the second
method indicate somewhat less rapid mixing of the tagged and untagged fish.
Schooling occurs in many species of fish, and many
studies have been made of the reasons for school-
ing and the behavior of the fish in the schools (e.g.,
Parr 1927; Shaw 1970; Pitcher 1986). Almost
nothing has been written, however, about the integ-
rity of schools over extended periods of time.
Parr (1927) stated that "apparently permanent"
schools are formed by pelagic fishes such as mack-
erel, sprat, and herring, and Sharp (1978) reported
that, "From the genetic sample data for the east-
ern Pacific yellowfin [{Thunnus albacares)] and the
Pacific-wide skipjack [{Katsuwonus pelamis)] there
is evidence for a cohesiveness of related fishes in
schools .... What is observed is that where more
than one very rare allele (overall expected occur-
rence <.01) is encountered in a large sample, the
individuals exhibiting the rare alleles are often the
same length or within 1 cm of each other. This is
highly unlikely unless they are related."
On the other hand, Helfman (1981) reported "ag-
gregations [ of freshwater fish] that disbanded dur-
ing twilight," and Moyle and Cech (1982) stated that
"most schools break up at night." Observers on
fishing vessels and aircraft have reported that the
schools of tunas frequently break up and reform.
Scott and Flittner (1972), for example, stated that
"the relatively large nighttime schools [of bluefin
tuna, Thunnus thynnus,] break down into several
smaller foraging schools and begin their search for
ilnter-American Tropical Tuna Commission, P.O. Box 271, La
Jolla, CA 92038.
food .... The gradual increase in school size during
the daylight hours may be due to regrouping of the
smaller schools through random encounters." Such
observations might lead to the conclusion that there
is considerable mixing of fish from different schools
and that fish of the same species and same approx-
imate size in the same areas would mix thoroughly
with one another within a period of a few days or
weeks.
Anonymous (1960) stated that "Tag returns from
individual schools [of skipjack tuna] suggest that,
normally, the skipjack in Hawaiian waters remain
within a school for one month or less, then at least
some of the school break off, move into new areas,
and regroup with other fish or schools. From the
releases off Hilo and Mexico, however, it is evident
that there are situations, possibly environmentally
conditioned, where the schools remain intact ... for
at least 2 or 3 months." Lester et al. (1985) studied
the occurrence of various parasites in skipjack tuna
of the same and different schools and concluded
that "school half-life is likely to be in terms of at
least weeks rather than days." They stated, how-
ever, that their data did "not support the hypothesis
[of Sharp (1978)] that fish stay in the same school
for life."
Examination of data for fish tagged and released
at the same location and time shows that some have
been recaptured weeks or months later in the same
purse seine set or baitboat stop, and others have
been recaptured weeks or months later on the same
date in widely separated locations (Hunter et al.
Manuscript accepted July 1988.
FISHERY BULLETIN: VOL. 86. NO. 4, 1988.
631
FISHERY BULLETIN: VOL. 8(i. NO. 4
1986). (Henceforth in this report, to save space, both
sets and stops will be referred to as sets.) The former
results might be due merely to chance. The latter
results indicate that not all the fish remain together
at all times, but this information is of limited value.
A much more powerful method of analysis is needed.
Turner (1986) employed statistical tests to show that
tagged bluefin tuna caught during the calendar year
after release had mixed considerably with the un-
tagged population, but knowledge about shorter
term mixing is necessary for short-lived species,
such as skipjack tuna. The present report describes
what are believed to be new and useful methods of
analysis of the integrity of schools of fishes, using
data for tagged skipjack tuna released and recap-
tured in the eastern Pacific Ocean.
MATERIALS AND METHODS
The methods of tagging the fish are described by
Bayliff and Holland (1986).
Tagged fish, or tags unaccompanied by the fish
to which they were attached, are recovered and
returned by fishermen, unloaders, and cannery
workers, accompanied by information which is used
to assign them to specific sets. Additional details
regarding this aspect of the study are discussed later
in this report.
The methods of collecting and processing the catch
statistics are discussed by Shimada and Schaefer
(1956) and Joseph and Calkins (1969). Hennemuth
(1957) and Shingu et al. (1974) described the
methods of sampling the fish and the calculations
employed to determine the size composition of tunas
in the catches. The areas shown in Hennemuth's
figure 1 have been changed several times since that
report was published, however; the areas used cur-
rently are shown by Peterson (1982: fig. 30).
ANALYSES AND RESULTS
If tagged fish released at the same location on the
same day mix thoroughly with the population of un-
tagged fish in the same area, schools of fish caught
in that area will have approximately equal ratios of
tagged to total fish, whereas if they do not mix
thoroughly some of the schools will have much
higher ratios than the others. In this report the
numbers of tagged fish recaptured in sets made at
various intervals after release are compared with
the numbers of tagged fish which would be expected
in those sets if the tagged fish had mixed thoroughly
with the rest of the population during the interval
Table 1 .—Data used for analysis of integrity of schools of skipjack tuna. The ranges
of values of sets and average weight in pounds are explained in the text.
Month
Numbers
Average
Tagging
of
of
weight
cruise
Dates of release
recapture
returns
Sets
in pounds
1042
2, 3, 4, 17, 18,
June
439
60-228
5.86-6.84
19, 23, 24 June
July
152
174
5.85
1962
August
44
333
4.96
1043
5, 9, 20, 27, 28
June
130
11-89
6.10-6.27
June; 1 July 1963
July
248
265-282
6.08-6.10
August
108
385
5.74
September
39
388
6.99
October
13
247
4.98
1070
12, 13, 14, 30
June
7
118-129
5.21-5.22
June 1973
July
32
144
5.08
August
15
120
4.59
September
27
45
5.35
October
33
205
5.26
November
10
229
5.69
1075
26 June; 20,
July
13
42-215
5.56-6.12
21 (two releases)
August
55
274
6.83
July 1975
1079
9, 10, 17, 18, 19
June
158
232-297
4.31-4.32
June 1976
July
261
386
4.29
August
15
47
4.24
September
118
542
4.88
October
61
546
4.56
November
6
262
4.12
632
BAYLIFF: INTEGRITY OF SKIPJACK TUNA SCHOOLS
between release and recapture. A description of how
this was done is given below. The data used in the
analyses are summarized in Table 1, and a more
detailed summary of the data for tagging cruise
1079, which will be discussed in more detail than the
other cruises, is given in Table 2.
First, the units of time to be employed were
selected. Tagged skipjack tuna have been recaptured
after as long as 3 years at liberty, but the great
majority of recaptures have been made within 6
months after release. It was decided to examine the
data by monthly intervals because, for the experi-
ments with sufficient total numbers of returns, that
would produce several intervals vdth sufficient num-
bers of returns of tagged fish for each interval. Also,
the statistical and size-frequency data of the Inter-
American Tropical Tuna Commission (lATTC) are
routinely calculated by month (and quarter and
Table 2— Data from cruise 1079 used for analysis of integrity of
schools of skipjack tuna.
Numbers
Average
Date of
Month of
of
weight
release
recapture
returns
Sets
in pounds
9 June 1976
June
13
297
4.32
July
13
386
4.29
August
1
47
4.24
September
5
542
4.88
October
3
546
4.56
November
0
262
4.12
10 June 1976
June
26
294
4.32
July
17
386
4.29
August
0
47
4.24
September
8
542
4.88
October
5
546
4.56
November
1
262
4.12
17 June 1976
June
80
252
4.32
July
120
386
4.29
August
9
47
4.24
September
48
542
4.88
October
31
546
4.56
November
1
262
4.12
18 June 1976
June
33
240
4.32
July
88
386
4.29
August
4
47
4.24
September
43
542
4.88
October
18
546
4.56
November
2
262
4.12
19 June 1976
June
6
232
4.31
July
23
386
4.29
August
1
47
4.24
September
14
542
4.88
October
4
546
4.56
November
2
262
4.12
Total
June
158
297
4.32
July
261
386
4.29
August
15
47
4.24
September
118
542
4.88
October
61
546
4.56
November
6
262
4.12
year), so no special calculations are required.
Second, the areas of study were selected. If the
tagged fish are released at a particular location, the
vessels fishing near that location would catch more
tagged fish, at least during the first few months
after release, than would vessels fishing several hun-
dred or more miles away. Thus only data for the
vessels fishing near the location of release should
be considered. The areas of study were selected by
examining charts of the distributions of fishing ef-
fort and recaptures of tagged fish, by 1 -degree areas
and by months, and arbitrarily excluding those with
lower recaptures per unit of fishing effort. This was
quite simple, as during the periods in question prac-
tically no tagged fish were recaptured south of lat.
20°N, and there was little fishing effort north of lat.
20 °N outside the area- time strata selected. The
areas of study for the data for tagging cruise 1079
are shown in Figure 1.
Third, a list of sets for each area-time stratum was
prepared. This included the weights of skipjack tuna
caught (Table 3, column 2) and the numbers of
tagged fish returned (Table 3, column 4). (Through-
out this report the weights are expressed in short
tons [0.907 metric tons] and pounds [0.454 kg]. The
lATTC uses this system because the fishermen
estimate the weights of the fish caught in individual
sets in short tons, and these estimates are an im-
portant component of its data base.) In a few cases
the catches were recorded only as weights of mixed
skipjack tuna and some other species, and in those
cases the weights were divided by 2, assuming that
they consisted of equal weights of skipjack tuna and
the other species. (The rest of this table will be dis-
cussed later.)
It can be seen in Table 1 that ranges instead of
individual values are given for "Sets" for cases when
the month of recapture is the same as the month
of release. This is because sets made before the date
of release were not considered for the analyses, and
it was also decided not to use data for tagged fish
recaptured on the date of release because these
could not have mixed with the rest of the fish in the
area to any appreciable extent. Therefore for cmise
1079, for example, there were 297 June sets after
9 June, 294 after 10 June, 252 after 17 June, 240
after 18 June, and 232 after 19 June (Table 2).
Fourth, an average weight for each month of re-
capture for each experiment was selected. Monthly
average weight data for purse seine- and baitboat-
caught fish from area 1 in Peterson (1982: fig. 30)
were used for this purpose because they closely cor-
respond to the strata selected for study. The aver-
633
FISHERY BULLETIN: VOL. 86, NO. 4
120'W 115 110'^ lOS'^ 120°W 115" 110>^ 105° 12a'W 115" 110" 105"
30P
25°
20^
Th
1
JULY
\
0
^ -
S
K
\
^
"^
VK
«sj
^^
\
^
K
s
"y
^
s
Aj
K
1
<
[S
U
\
1
\
)
\
X
h
X
f
1 1
TV AUGUST
— K^
^ " IvV.
" jOt^ \ ^
V \
X -K
^ ^ X „ _S
I
««^
"V,
^
1
SEPTEMBER
V
h
^ -
N
L
I,
<
\
v\
^
^N
V 1
'
>^.
^
:5
A
^ '
t
N
u
\ 1
\>
\
X
\\
X
\
M
-\.
x^
^
\
^
p^
1
OCTOBER
\
Wv
\
s
^
^
^ 1
^ 1
-
-
-
—
-
M,
h
\
^
IN^
I
N
U
\
1
\
}
\
X
\\
-
"
X
-
-
.J
-
35°N
30°
25°
Figure L— Locations of release (X's) and areas of recapture selected for tagging cruise 1079 (areas delineated by
heavy lines).
age weights were estimated by
w^ =
5! w,
ii = l
ijPS
X w
'JPS
+
J.W,
1 = 1
jBB >< '^jBB
1=1 i=l
1, 2) because only the sets after the dates of release
are considered, and the portions of the total catches
which are from purse seines and baitboats differ in
accordance with the dates considered.
The data were analyzed by 1) Chi-square con-
tingency tests and 2) the binomial homogeneity test
described by Kendall and Stuart (1961:578-579).
where Wj = average weight of skipjack
tuna in stratum j,
Wijps and W^jp^ = weights of skipjack tuna
caught in purse seine set i
and baitboat set i, respec-
tively, made in stratum j,
n = number of purse seine or
baitboat sets in stratum j,
and
Wjps and Wj^b = average weights of skipjack
tuna caught in stratum j by
purse seiners and baitboats,
respectively.
These are listed in Tables 1 and 2.
The average weights usually differ within strata
which correspond to the months of release (Tables
Chi-Square Contingency Tests
A computer program, SCHOOL, was written to
analyze the data. For a given release date and month
of recapture, this program estimates the number of
fish caught in each set from the weight of fish caught
and the average weight of the fish. It then sums the
estimates of the numbers of fish caught and num-
bers of tagged fish returned for all sets and calcu-
lates the tagged to total ratio. This ratio is then used
with the equation for the binomial distribution to
estimate the probabilities of 0, 1, 2, 3, . . .tagged fish
appearing in each set if the fish are randomly mixed.
The sums of the probabilities for all the sets for 0,
1, 2, 3, . . . tagged fish are then calculated so that
these can be compared with the observed data, as
described in the next paragraph. A sample output
634
BAYLIFF: INTEGRITY OF SKIPJACK TUNA SCHOOLS
from this program is shown in Table 3. It can be seen
that the sum of 39.40 + 6.41 + 1.00 + 0.16 + 0.02
(last line) is equal to 47, the number of sets, and that
the sum of (39.40 x 0) + (6.41 x 1) + (1.00 x 2)
+ (0.16 X 3) is equal to 9, the number of tagged fish
returned. (The numbers of tagged fish returned for
each set had been entered in Table 3, column 4, as
explained above. If the total, 9, had been entered
for the first set, or any other set, however, the result
in the bottom line would have been the same.)
The bottom lines from all the outputs from
SCHOOL for tagging cruise 1079 are listed in the
"exp." (expected) lines of Table 4. Just below these
are listed the observed ("obs.") numbers of sets with
0, 1, 2, 3, . . . tagged fish. At the bottom of each sec-
tion of this table the sums of the expected and ob-
served values are listed. Chi-square tests were run,
using MINITAB (Ryan et al. 1985), on the expected
Table 3 —Probabilities of 0, 1, 2, 3, 10 tagged fish in sets made in August in the area shown in Figure 1 from fish released on 17
June 1976. This table is similar to the output from program SCHOOL except that (1) normally the lines for the individual sets are not
printed and (2) the output does not include the numbers of fish.
Tons
Fish
Tag
IS
Set
P(0)
P(1)
P(2)
P(3)
P(4)
P(5)
P(6)
P(7)
P(8)
P(9)
P(10)
Total
1
12.0
5,660
0
0.6921
0.2546
0.0468
0.0057
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
2
15.5
7,311
1
0.6217
0.2954
0.0702
0.0111
0.0013
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
3
15.0
7,075
0
0.6313
0.2903
0.0667
0.0102
0.0012
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
4
5.0
2,358
0
0.8578
0.1314
0.0101
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
5
3.0
1,415
0
0.9121
0.0838
0.0038
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
6
1.0
472
1
0.9698
0.0296
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
7
2.0
943
0
0.9405
0.0575
0.0018
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
8
1.0
472
0
0.9698
0.0296
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
9
25.0
1 1 ,792
0
0.4646
0.3561
0.1365
0.0349
0.0067
0.0010
0.0001
0.0000
0.0000
0.0000
0.0000
0.9998
10
25.0
11,792
0
0.4646
0.3561
0.1365
0.0349
0.0067
0.0010
0.0001
0.0000
0.0000
0.0000
0.0000
0.9998
11
1.0
472
0
0.9698
0.0296
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
12
1.0
472
3
0.9698
0.0296
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
13
1.0
472
0
0.9698
0.0296
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
14
2.0
943
0
0.9405
0.0575
0.0018
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
15
5.0
2,358
0
0.8578
0.1314
0.0101
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
16
5.0
2,358
0
0.8578
0.1314
0.0101
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
17
9.0
4,245
0
0.7588
0.2093
0.0289
0.0027
0.0002
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
18
5.0
2,358
0
0.8578
0.1314
0.0101
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
19
5.0
2,358
0
0.8578
0.1314
0.0101
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
20
22.0
10,377
1
0.5093
0.3436
0.1159
0.0260
0.0044
0.0006
0.0001
0.0000
0.0000
0.0000
0.0000
0.9998
21
2.0
943
0
0.9405
0.0575
0.0018
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
22
1.5
708
0
0.9550
0.0438
0.0010
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
23
2.5
1,179
0
0.9262
0.0709
0.0027
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
24
2.0
943
0
0.9405
0.0575
0.0018
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
25
8.0
3,774
0
0.7824
0.1919
0.0235
0.0019
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
26
10.0
4,717
0
0.7359
0.2256
0.0346
0.0035
0.0003
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
27
13.0
6,132
0
0.6712
0.2675
0.0533
0.0071
0.0007
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
28
4.5
2,123
0
0.8711
0.1201
0.0083
0.0004
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
29
6.0
2,830
0
0.8319
0.1530
0.0141
0.0009
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
30
4.0
1,887
0
0.8846
0.1084
0.0066
0.0003
0.0000
0.0000
0,0000
0.0000
0.0000
0.0000
0.0000
0.9999
31
3.0
1,415
0
0.9121
0.0838
0.0038
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
32
6.5
3,066
0
0.8193
0.1632
0.0162
0.0011
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
33
7.0
3,302
0
0.8068
0.1731
0.0186
0.0013
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
34
10.0
4,717
0
0.7359
0.2256
0.0346
0.0035
0.0003
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
35
4.0
1,887
0
0.8846
0.1084
0.0066
0.0003
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
36
1.5
708
1
0.9550
0.0438
0.0010
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
37
1.5
708
0
0.9550
0.0438
0.0010
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
38
4.0
1,887
0
0.8846
0.1084
0.0066
0.0003
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
39
4.0
1,887
0
0.8846
0.1084
0.0066
0.0003
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
40
10.0
4,717
0
0.7359
0.2256
0.0346
0.0035
0.0003
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
41
8.0
3,774
0
0.7824
0.1919
0.0235
0.0019
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
42
5.0
2,358
0
0.8578
0.1314
0.0101
0.0005
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
43
2.5
1,179
0
0.9262
0.0709
0.0027
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
44
2.5
1,179
0
0.9262
0.0709
0.0027
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
45
7.5
3,538
2
0.7945
0.1827
0.0210
0.0016
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
46
2.0
943
0
0.9405
0.0575
0.0018
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9998
47
0.5
236
0
0.9848
0.0150
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9999
Total
293.5
138,443
9
39.40
6.41
1.00
0.16
0.02
0.00
0.00
0.00
0.00
0.00
0.00
46.99
635
FISHERY HI'LLKTIN: VOL. 86, NO. 4
Table 4.— Data for analysis, by Chi-square contingency tests, of the integrity of schools of skipjack tuna released during tagging cruise
1079. [exp. = expected; obs. = observed.]
Date of
release
Month of
recapture
Tags
Occurrences of tagged fish
1
value
df
Proba-
bility
0
1
2
3
4
5
>5
Total
9 June
1976
June
13
13
exp.
obs.
284.87
285
11.38
11
0.63
1
0.09
0.02
297.00
297
0.001
1
>0.05
10 June
1976
26
26
exp.
obs.
271.06
281
20.61
8
1.81
2
0.32
1
0.11
0.04
1
0.01
1
294.00
294
4.672
1
<0.05
17 June
1976
80
80
exp.
obs.
194.53
218
44.10
20
9.32
6
2.23
4
0.68
0.31
0.76
4
252.00
252
16.032
2
<0.01
18 June
1976
33
33
exp.
obs.
212.49
220
23.70
13
2.82
3
0.56
2
0.22
2
0.11
008
240.00
240
2316
1
>0.05
19 June
1976
6
6
exp.
obs.
226.26
226
5.50
6
0.21
0.02
232.00
232
0.012
1
>0.05
Total
158
158
exp.
obs.
1,189.21
1,230
105.29
58
14.79
12
3.22
7
1.03
2
0.46
1
0.85
5
1,315.00
1,315
38.280
3
<0.01
9 June
1976
July
13
13
exp.
obs.
373.43
375
12.14
9
0.40
2
0.01
386.00
386
0.203
1
>0.05
10 June
1976
17
17
exp.
obs.
369.73
373
15.56
10
0.66
2
0.03
1
386.00
386
0.686
1
>0.05
17 June
1976
120
120
exp.
obs.
294.44
339
70.30
20
15.83
8
3.93
8
1.05
5
0.28
4
0.09
2
386.00
386
80.519
3
<0.01
18 June
1976
88
88
exp.
obs.
314.41
342
58.43
22
10.48
9
2.08
8
0.43
2
0.09
2
0.02
1
386.00
386
31.073
2
<0.01
19 June
1976
23
23
exp.
obs.
364.32
368
20.43
14
1.16
3
0.07
1
386.00
386
0.662
1
>0.05
Total
261
261
exp.
obs.
1,716.33
1,797
176.86
75
28.53
24
6.12
18
1.48
7
0.37
6
0.11
3
1,930.00
1,930
141.566
3
<0.01
9 June
1976
August
1
1
exp.
obs.
46.02
46
0.96
1
0.02
47.00
47
10 June
1976
0
0
exp.
obs.
47.00
47
47.00
47
17June
1976
9
9
exp.
obs.
39.40
41
6.41
4
1.00
1
0.16
1
0.02
47.00
47
0.402
1
>0.05
18 June
1976
4
4
exp.
obs.
43.30
44
3.42
2
0.26
1
0.02
47.00
47
19 June
1976
1
1
exp.
obs.
46.02
46
0.96
1
0.02
47.00
47
Total
15
15
exp.
exp.
221.74
224
11.75
8
1.30
2
0.18
1
0.02
235.00
235
0.408
1
>0.05
and observed values, and the results are shown in
the last four columns of Table 4. The categories were
combined so that none had an expected value of less
than 5. For the first test in Table 4 (releases on 9
June 1976, and recaptures during June), for exam-
ple, the expected values are 284.87 and (11.38 -i-
0.63 + 0.09 + 0.02 = 12.12) and the observed values
are 285 and (11 -i- 1 = 12). When only one category
had an expected value of 5 or greater no test was
run.
The two total lines for each section of Table 4 are
listed in Table 5, with the equivalent values for the
other four experiments. These were summed for
each cruise and month of recapture, and Chi-square
tests were run with these sums. Those results are
shown in the last four columns of Table 5.
The results of the tests are summarized as follows:
Months of recapture
Category
df
6
7
8
9
10
11
releases made on
1
3/11
2/17
0/12
0/10
0/6
individual dates
2
3
4
1/1
1/1
2/2
3/4
2/2
0/1
totals for cruises
1
0/1
0/2
0/4
0/2
0/3
0/2
2
1/1
0/1
1/1
3
1/2
2/2
4
1/1
overall totals
1
2
3
4
1/1
1/1
1/1
1/1
0/1
0/1
636
HAYLIFF: INTEGRITY OF SKIFMACK TUNA SCHOOLS
Table ^.—Continued.
Date of
release
Month of
recapture
September
Tags
Occurrences of tagged fish
value
df
Proba-
bility
0
1
2
3
4 5
>5 Total
9 June
1976
5
5
exp
obs
537.05
537
4.90
5
0.05
542.00
542
10 June
1976
8
8
exp
obs
534.12
534
7.75
8
0.12
542.00
542
0.002
1
>0.05
17 June
1976
48
48
exp
obs
498.05
509
40.23
20
3.35
12
0.30
0.03
1
542.00
542
2.969
1
>0.05
18 June
1976
43
43
exp
obs
502.28
508
36 68
26
2.77
7
0.22
1
0.02
542.00
542
0.889
1
>0.05
19 June
1976
14
14
exp
obs
528.37
529
13.26
12
0.35
1
0.01
542.00
542
0.030
1
>0.05
Total
118
118
exp
obs
2,599.87
2,617
102.82
71
6.64
20
0.53
1
0.05
1
2,710.00
2,710
39.481
2
<0.01
9 June
1976
October
3
3
exp
obs
543.02
544
2.96
1
0.02
1
546.00
546
10 June
1976
5
5
exp
obs
541.06
541
4.88
5
0.06
546.00
546
17 June
1976
31
31
exp
obs
517.04
524
27.07
15
1.74
5
0.13
2
0.01
546.00
546
1.766
1
>0.05
18 June
1976
18
18
exp
obs
528.71
530
16.60
14
0.65
2
0.03
546.00
546
0.099
1
>0.05
19 June
1976
4
4
exp
obs
542.04
542
3.93
4
0.04
546.00
546
Total
61
61
exp
obs
2,671.87
2,681
55.44
39
2.51
8
0.16
2
0.01
2,730.00
2,730
1.465
1
>0.05
9 June
1976
November
0
0
exp
obs
262.00
262
262.00
262
10 June
1976
1
1
exp
obs
261.00
261
0.99
1
262.00
262
17 June
1976
1
1
exp
obs
261.00
261
0.99
1
262.00
262
18 June
1976
2
2
exp
obs
260,02
260
1.96
2
0.02
262.00
262
19 June
1976
2
2
exp
obs
260.02
260
1.96
2
0.02
262.00
262
Total
6
6
exp
obs
1,304.04
1,304
5.90
6
0.04
1,310.00
1,310
0.000
1
>0.05
This means that for individual dates for recaptures
during June (i.e., during the month of release), 3 of
11 tests equivalent to those in Table 4 with 1 degree
of freedom were significant at the 5% level; for
recaptures during July, 2 of 17 tests equivalent to
those in Table 4 with 1 degree of freedom were sig-
nificant; and so on. From these tests it appears that
after 1 month at liberty the tagged and untagged
fish are randomly mixed with one another in some
cases, and after 3 to 5 months at liberty they are
randomly mixed with one another in nearly all cases.
r= I
(T, - F,Pf
t=i F,P (1 - P)
where T, = number of tagged fish in set z,
Fj = number of tagged and untagged fish in
set i,
n = number of sets, and
P = probability of a fish being tagged =
« n
I (T,)/I (F,)
i = l
! = 1
Binomial Homogeneity Tests
For the binomial homogeneity test the formula
is used. Computer program SCHOOLA was used for
this purpose.
The results are given in Table 6. In general, the
637
FISHERY BULLETIN: VOL. 86, NO. 4
Table 5-
-Data for analysis, by (
Dhi-square contingency tests, of the integrity of schools
= expected; obs. = observed.]
of skip
ack tuna released during all cruises
[exp.
Tagging
cruise
Month of
recapture
Tags
Occurrences of tagged fish
value
df
Proba-
bility
0
1
2
3
4
5
>5
Total
1042
June
439
439
exp.
obs.
971.84
1,114
193.13
74
53.10
26
19.75
11
8.53
4
3.89
7
4.76
19
1,255.00
1,255
121.553
4
<0.01
1043
130
130
exp.
obs.
203.70
212
39.03
27
11.32
11
4.42
8
2.08
3
1.16
3
4.29
2
266.00
266
5.428
3
>0.05
1070
7
7
exp.
obs.
345.25
345
6.51
7
0.23
0.01
352.00
352
0.009
1
>0.05
1079
158
158
exp.
obs.
1,189.21
1,230
105.29
58
14.79
12
3.22
7
1.03
2
0.46
1
0.85
5
1,315.00
1,315
38.280
3
<0.01
Total
734
734
exp.
obs.
2,710.00
2,901
343.96
166
79.44
49
27.40
26
11.64
9
5.51
11
9.90
26
3,188.00
3,188
149.523
6
<0.01
1042
July
152
152
exp.
obs.
1,265.04
1,307
107.21
55
15.36
22
3.20
4
0.75
0
0.17
1
0.27
3
1,392.00
1,392
32.137
2
<0.01
1043
248
248
exp.
obs.
1,479.32
1,536
158.72
96
26.60
18
6.72
12
2.19
6
0.79
2
0.44
5
1,675.00
1,675
50.425
3
<0.01
1070
32
32
exp.
obs.
545.75
546
28.57
28
1.54
2
0.09
576.00
576
0.002
1
>0.05
1075
13
13
exp.
obs.
338.04
343
11.01
6
0.84
0.09
1
350.00
350
2.130
1
>0.05
1079
261
261
exp.
obs.
1,716.33
1,797
176.86
75
28.53
24
6.12
18
1.48
7
0.37
6
0.11
3
1,930.00
1,930
141.566
3
<0.01
Total
706
706
exp.
obs.
5,344.48
5,529
482.37
260
72.87
66
16.22
34
4.42
13
1.33
9
0.82
12
5.923.00
5,923
231.819
4
<0.01
1042
August
44
44
exp.
obs.
2,613.04
2,618
41.88
33
0.98
4
0.03
1
2,656.00
2,656
0.582
1
>0.05
1043
108
108
exp.
obs.
2,211.08
2,219
90.80
80
7.13
7
0.78
3
0.10
0.01
1
2,310.00
2,310
2.334
2
>0.05
1070
15
15
exp.
obs.
465.42
465
14.16
15
0.39
480.00
480
0.012
1
>0.05
1075
55
55
exp.
obs.
1,044.57
1,050
48.13
39
2.92
5
0.26
2
0.03
1,096.00
1,096
0.602
1
>0.05
1079
15
15
exp.
obs.
221.74
224
11.75
8
1.30
2
0.18
1
0.02
235.00
235
0.408
1
>0.05
Total
237
237
exp.
exp.
6,555.85
6,576
206.72
175
12.72
18
1.25
7
0.15
0.01
1
6,777.00
6,777
14.206
2
<0.01
1043
September
39
39
exp.
obs.
2,290.28
2,291
36.42
35
1.18
2
0.05
2,328.00
2,328
0.014
1
>0.05
1070
27
27
exp.
obs.
156.88
160
19.77
14
2.88
5
0.41
1
0.05
180.00
180
0.483
1
>0.05
1079
118
118
exp.
obs.
2,599.87
2,617
102.82
71
6.64
20
0.53
1
0.05
1
2,710.00
2,710
39.481
2
<0.01
Total
184
184
exp.
obs.
5,047.03
5,068
159.01
120
10.70
27
0.99
2
0.10
1
5,218.00
5,218
36.868
2
<0.01
1043
October
13
13
exp.
obs.
1,481.38
1,482
12.26
11
0.35
1
0.01
1,494.00
1,494
0.031
1
>0.05
1070
33
33
exp.
obs.
788.27
789
30.49
29
1.17
2
0.04
820.00
820
0.017
1
>0.05
1079
61
61
exp.
obs.
2,671.87
2,681
55.44
39
2.51
8
0.16
2
0.01
2,730.00
2,730
1.465
1
>0.05
Total
107
107
exp.
obs.
4,941.52
4,952
98.19
79
4.03
11
0.21
2
0.01
5,044.00
5,044
1.094
1
>0.05
1070
November
10
10
exp.
obs.
906.14
906
9.69
10
0.14
916.00
916
0.002
1
>0.05
1079
6
6
exp.
obs.
1,304.04
1,304
5.90
6
0.04
1,310.00
1,310
0.000
1
>0.05
Total
16
16
exp.
obs.
2,210.18
2,210
15.59
16
0.18
2,226.00
2,226
0.002
1
>0.05
638
BAYLIFF: INTEGRITY OF SKIPJACK TUNA SCHOOLS
X"/df values do not appear to decrease consistent-
ly with time, as would be expected if the fish tend
to mix gradually with time. If, however, only the
values corresponding to date of release-month of
recapture strata with more than 10 tag returns are
considered the x"/df values tend to decrease with
time. The x~ values corresponding to these strata
are summed at the bottoms of the first five sections
of Table 6, and these sums are divided by the sums
of the degrees of freedom to obtain total x"/df
values. It can be seen that these also tend to
decrease with time. In addition, the x" values and
degrees of freedom for the strata with more than
10 tag returns for all the experiments are summed
in the last section of Table 6; the x"/df values again
tend to decrease with time. It thus appears that the
fish were gradually mixing as time passed.
One thousand Monte Carlo simulations were run
for each stratum with more than 10 tags, using com-
puter program MONTCARL, to determine the prob-
ability of obtaining an equal or greater value of
X-/df, if the tagged and untagged fish were ran-
domly mixed with one another. The results are
shown in the last column of Table 6. In most cases
these indicated that the tagged and untagged fish
were not randomly mixed. These data tend to in-
dicate less rapid mixing with time than do the data
for the Chi- square contingency tests.
DISCUSSION
There is a fundamental difference between the
two methods. The Chi-square contingency tests test
whether the tagged fish occurred in many or few
of the schools, whereas the binomial tests test
whether the ratios of tagged to total fish are con-
sistent among schools. For example, if there was
a total of 15 returns of tagged fish obtained from
15 different sets from a total of 20 25-ton sets and
20 1-ton sets, it would make no difference for the
Chi-square contingency tests which sets the tagged
fish occurred in. For the binomial homogeneity tests,
however, the x'/df value would be much greater
if the tagged fish occurred in 15 of the 20 1-ton
sets than if they occurred in 15 of the 20 25-ton
sets.
The Chi-square contingency tests are adversely
affected by the small numbers of tag returns, which
makes them rather low powered. For example, for
the releases of 19 June 1962, and 28 June 1963,
Table 6. — Results of binomial homogeneity tests with skipjack tuna tag return
data.
Date of
Montfi of
Proba-
Year release
recapture
Sets
Tags
x'
x'/df
bility
1962 2 June
June
228
33
714.44
3.15
<0.01
July
174
15
236.33
1.37
>0.05
August
333
10
1,467.22
4.42
3 June
June
223
26
880.38
3.97
<0.01
July
174
18
272.03
1.57
<0.05
August
333
10
304.10
0.92
4 June
June
212
32
4,395.24
20.83
<0.01
July
174
15
327.12
1.89
<0.05
August
333
9
666.62
2.01
17 June
June
162
129
7,862.99
48.84
<0.01
July
174
6
333.26
1.93
August
333
1
81.91
0.25
18 June
June
152
15
435.92
2.89
<0.01
July
174
2
466.73
2.70
August
333
0
—
—
19 June
June
141
95
543.94
3.89
<0.01
July
174
28
575.73
3.33
<0.01
August
333
5
236.97
0.71
23 June
June
77
95
392.78
5.17
<0.01
July
174
40
744.76
4.30
<0.01
August
333
7
533.24
1.61
24 June
June
60
14
99.18
1.68
<0.05
July
174
28
1,592.40
9.20
<0.01
August
333
2
299.98
0.90
Total
June
1,255
439
15,324.87
12.29
July
1,044
144
3,748.37
3.61
639
FISHERY BULLETIN: VOL. 86. NO. 4
Table 6.— Continued.
Date of
Month of
Proba-
Year release
recapture
Sets
Tags
x'
x'/df
bility
1963 5 June
June
89
26
141.07
1.60
>0.05
July
282
25
702.74
2.50
<0.01
August
385
16
1,250.94
3.26
<0.01
September
388
6
645.62
1.67
October
249
1
2,110.73
8.51
9 June
June
89
28
293.56
3.34
<0.01
July
282
50
1,036.42
3.69
<0.01
August
385
24
514.06
1.34
>0.05
September
388
6
779.54
2.01
October
249
2
314.69
1.28
20 June
June
53
12
56.66
1.09
>0.05
July
282
15
532.52
1.90
<0.05
August
385
10
681.99
1.78
September
388
6
1,105.22
2.86
October
249
2
192.00
0.78
27 June
June
24
33
197.47
8.59
<0.01
July
282
101
1,683.31
5.99
<0.01
August
385
48
1,019.23
2.65
<0.01
September
388
19
843,91
2.18
<0.01
October
249
8
258.17
1.05
28 June
June
11
31
401.40
40.14
<0.01
July
282
43
1,071.63
3.81
<0.01
August
385
5
85.27
0.22
September
388
1
268.06
0.69
October
249
0
—
—
1 July
July
265
14
1,272.38
4.82
<0.01
August
385
5
1,000.85
2.61
September
388
1
447.33
1.16
October
249
0
—
—
Total
June
266
130
1,090.16
4.16
July
1,675
248
6,299.00
3.77
August
1,155
88
2,784.23
2.42
September
388
19
843.91
2.18
1973 12 June
June
129
4
125.55
0.98
July
144
9
190.46
1.33
August
120
4
90.50
0.76
September
45
5
77.04
1.75
October
205
7
289.43
1.42
November
229
1
78.43
0.34
13 June
June
118
3
156.86
1.34
July
144
8
118.45
0.83
August
120
4
256.57
2.16
September
45
4
24.50
0.56
October
205
13
214.15
1.05
>0.05
November
229
3
116.14
0.51
14 June
June
105
0
—
—
July
144
6
333.14
2.33
August
120
2
78.99
0.66
September
45
8
41.23
0.94
October
205
10
387.81
1.90
November
229
5
477.96
2.10
30 June
July
144
9
344.95
2.41
August
120
5
50.35
0.42
September
45
10
21.62
0.49
October
205
3
324.07
1.59
November
229
1
197.57
0.87
Total
October
205 13
214.15 1.05
640
BAYLIFF: INTEGRITY OF SKIPJACK TUNA SCHOOLS
Table 6.— Continued.
Date of
Month of
Proba-
Year release
recapture
Sets
Tags
x'
x'/df
bility
1975 26 June
July
215
5
192.62
0.90
August
274
15
536.17
1.96
<0.01
20 July
July
51
1
13.23
0.26
August
274
8
246.67
0.90
21 July
July
42
7
64.87
1.58
August
274
10
1,255.33
4.60
21 July
July
42
0
—
—
August
274
22
1,036.53
3.80
<0.01
Total
August
548
37
1,572.70
2.88
1976 9 June
June
297
13
468.83
1.58
>0.05
July
386
13
874.92
2.27
<0.01
August
47
1
38.13
0.83
September
542
5
628.52
1.16
October
546
3
1,830.07
3.36
November
262
0
—
—
10 June
June
294
26
664.08
2.27
<0.01
July
386
17
744.88
1.93
<0.01
August
47
0
—
—
September
542
8
578.70
1.07
October
546
5
354.91
0.65
November
262
1
410,61
1.57
17 June
June
252
80
977.05
3.89
<0.01
July
386
120
1,283.42
3.33
<0.01
August
47
9
359.63
7.82
September
542
48
1,448.02
2.68
<0.01
October
546
31
1,002.48
1.84
<0.01
November
262
1
153.34
0.59
18 June
June
240
33
1,749.76
7.32
<0.01
July
386
88
947.11
2.46
<0.01
August
47
4
399.34
8.68
September
542
43
2,018.78
3.73
<0.01
October
546
18
704.02
1.29
>0.05
November
262
2
105.17
0.40
19 June
June
232
6
249.12
1.08
July
386
23
611.28
1.59
<0.05
August
47
1
292.31
6.35
September
542
14
594.33
1.10
>0.05
October
546
4
771.34
1.42
November
262
2
242.38
0.93
Total
June
1,083
152
3,859.72
3.58
July
1,930
261
4,461.61
2.32
September
1,626
105
4,061.13
2.50
October
1,092
49
1,706.50
1.57
all
June
2,604
721
20,274.75
7.84
July
4,649
653
14,508.98
3.13
August
1,703
125
4,356.93
2.56
September
2,014
124
4,905.04
2.44
October
1,297
62
1,920.65
1.48
there were sets with more than five tagged fish in
them, and yet the results of the Chi-square tests
were not significant. This is caused by the require-
ment that each category have an expected value of
5 or greater. Accordingly, it is likely that if there
had been more tag returns, there would have been
more categories for many of the tests and signifi-
cant results for more of them. The practice of com-
bining the results for fish of the same experiment
released on different days ("Total" lines in Table
4) and for fish of different experiments ("Total"
lines of Table 5) helps to overcome this weakness.
It can be seen in the text table above that the ratios
increased with the degrees of freedom. If there were
more tag returns, there would be fewer tests with
one degree of freedom and more with three or more
degrees of freedom, and the ratios of significant
tests would almost certainly increase.
641
FISHERY BULLETIN: VOL. 86, NO. 4
On the other hand, the way that the tag return
data are handled causes it to appear that there were
more sets with more than one tagged fish than was
actually the case. Ideally, all tagged fish would be
recovered by fishermen as soon as they are caught,
and then set aside for later examination by an
lATTC employee, and the tag numbers, locations,
and dates of recapture would be recorded so that
each fish could be assigned to the proper set. In
reality, however, less than half the tagged fish
recaptured ax^e recovered by the fishermen, and
since the fish from different sets are mixed in the
wells of the vessels, the chance of assigning tagged
fish to specific sets is lost, except when enough fish
are caught in one set to fill an entire well. Virtually
all of the tagged fish which are not recovered by
fishermen are recovered later by unloaders and can-
nery workers. The unloaders and cannery workers
usually inform the lATTC employee to whom they
return the tagged fish (or the tag without the fish
attached to it) that the fish was found in or had come
from a particular well or pair of wells of a particular
boat. The lATTC employee who receives the tagged
fish or tag records this information, and later
another lATTC employee compares this information
with an abstract of the vessel's logbook and assigns
the fish to the set which contributed the greatest
weight of fish of the species in question to that well
or pair of wells. For example, if a particular well
contained fish from sets with 12, 15, 20, and 13 tons
made on 1, 2, 3, and 4 June, respectively, and each
included one tagged fish, all recovered by unloaders
and cannery workers, all four would be assigned to
the 3 June set. This would make it appear that the
tagged fish tend to remain together more than is
actually the case. Another way to handle a situation
such as this would be to allocate the four fish among
the four sets in proportion to the weights of fish in
them, in this case one to each set. This is not done,
however, because tagged fish from the same trip of
the same vessel are often returned to the lATTC
over a considerable period of time, and it is not feas-
ible to keep the tags for long periods waiting for all
of them to be returned before processing them. Fur-
thermore, allocation of tagged fish in the way just
described would tend to make it appear that the
tagged fish remain together less than they actually
do if they are not randomly mixed with the untagged
ones.
In addition to the problems created by failure to
recover all the tagged fish as soon as they are
caught, there are probably problems created by false
data. Sometimes the persons recovering tagged fish
keep the tags they have recovered over a period of
several days, weeks, or even longer, and then return
them to an lATTC employee, telling him that they
were all recovered that day in the well or pair of
wells that was unloaded that day. The likelihood that
the data are false can often, but not always, be
detected by an alert lATTC employee. When the
likelihood that the data are false is not detected, it
will appear that the tagged fish remain together
more than is actually the case. It is believed that,
in spite of attempts to detect data which are likely
to be false, some false data are included in the
analyses and make it appear that the tagged fish
remain together more than is actually the case.
The fact that the numbers of tag returns were
small, coupled with the requirement for the Chi-
square contingency tests that the categories for the
expected numbers of tagged fish be equal to or
greater than 5, tends to make it appear that the
tagged fish mix more rapidly with the untagged ones
than is actually the case. The biases resulting from
the mixing of the fish caught in different sets in the
same wells and from false data tend to make it ap-
pear that the tagged and untagged fish mix less
rapidly than is actually the case. Thus the two fac-
tors tend to cancel each other out, at least partial-
ly, although the first bias may be stronger than the
second. If so, the tentative conclusion made above
that the tagged and untagged fish mix thoroughly
within about 3 to 5 months may be incorrect; that
time could be somewhat longer. For the binomial
homogeneity tests only the second bias exists, so the
rate of mixing of the tagged and untagged fish in-
dicated by these tests is probably somewhat slower
than is actually the case.
Sharp (1978) stated that it is likely that skipjack
tuna in the same school are "related," but it seems
unlikely that tunas could school together for their
entire lives, an implication attributed to Sharp (1978)
by Lester et al. (1985). During the egg and larval
stages, the fish are at the mercy of their environ-
ment, and individuals which were together at one
time would often be separated by the currents. Fur-
thermore, large tunas of the genus Thunnus occur
mostly in subsurface waters at depths to nearly 300
m (Suzuki et al. 1977). Although there are areas of
greater and lesser concentrations of large, subsur-
face-dwelling fish, there is no evidence that they
form concentrated schools such as those which oc-
cur at the surface. The present study indicates that
skipjack tuna in the size range of about 3.4 to 7.0
pounds (about 43 to 53 cm in length) of the same
school mix randomly with those of other schools
642
BAYLIFF: INTEGRITY OF SKIPJACK TUNA SCHOOLS
within about 3 to 5 months, or possibly somewhat
longer. This is in agreement with the findings of
Anonymous (1960) and Lester et al. (1985), but not
those of Sharp (1978), discussed at the beginning
of this report.
ACKNOWLEDGMENTS
Appreciation is expressed to Gayle Ver Steeg, who
compiled the catch and effort data; to Stephen T.
Buckland, Richard G. Punsly, and Patrick K. Tom-
linson, who contributed some ideas for analysis of
the data; and to Stephen T. Buckland, Andrew E.
Dizon, Ashley J. Mullen, Richard G. Punsly, Michael
D. Scott, and three anonymous reviewers, who made
some useful suggestions for improvement of the
manuscript.
LITERATURE CITED
Anonymous.
1960. Tagging returns indicate that skipjack is not a wide-
ranging species. U.S. Natl. Mar. Fish. Serv., Commer.
Fish. Rev. 22(ll):25-26.
Bayliff, W. H., and K. N. Holland.
1986. Materials and methods for tagging tunas and billfishes,
recovering the tags, and handling the recapture data. FAO
Fish. Tech. Pap. 279, 36 p.
Helfman, G. S.
1981. Twilight activities and temporal structure in a fresh-
water fish community. Can. J. Fish. Aquat. Sci. 38:1405-
1420.
Hennemuth, R. C.
1957. An analysis of methods of sampling to determine the
size composition of commercial landings of yellowfin tuna
{Neothunnus macropterus) and skipjack (Kat.suwonus pela-
mis). Inter-Am. Trop. Tuna Comm., Bull. 2:171-243.
Hunter, J. R., A. W. Argue, W. H. Bayliff, A. E. Dizon,
A. Fonteneau, D. Goodman, and G. R. Seckel.
1986. The dynamics of tima movements: an evaluation of past
and future research. FAO Fish. Tech. Pap. 277, 78 p.
Joseph, J., and T. P. Calkins.
1969. Population dynamics of the skipjack tuna (Katsuwonus
pelamis) in the eastern Pacific Ocean. Inter-Am. Trop. Tuna
Comm.. Bull. 13:1-273.
Kendall, M. G., and A. Stuart.
1961. The advanced theory of statistics, Vol. 2. HafnerPubl.
Co., N.Y., 676 p.
Lester, R. J., A. Barnes, and G. Habib.
1985. Parasites of skipjack tuna, Katsuwonus pelamis:
fishery implications. Fish. Bull., U.S. 83:343-356.
MoYLE, P. B., and J. J. Cech, Jr.
1982. Fishes: an introduction to ichthyology. Prentice-Hall,
Inc., Englewood Cliffs, NJ, 593 p.
Parr, A. E.
1927. A contribution to the theoretical analysis of the school-
ing behavior of fishes. Occas. Pap. Bingham Oceanogr.
Collect., 32 p.
Peterson, C. L. (editor).
1982. Annual report of the Inter- American Tropical Tuna
Commission, 1981. Inter-Am. Trop. Tuna Comm., 303 p.
Pitcher, T. J.
1986. Functions of schooling behaviour in teleosts. In T. J.
Pitcher (editor), The behavior of teleost fishes, p.
294-337. Johns Hopkins Univ. Press, Baltimore.
Ryan, B. F., B. L. Joiner, and T. A. Ryan, Jr.
1985. MINITAB handbook. 2d ed. Duxbury Press, Boston,
374 p.
Scott, J. M., and G. A. Flittner.
1972. Behavior of bluefin tuna schools in the eastern north
Pacific Ocean as inferred from fishermen's logbooks,
1960-67. Fish. Bull, U.S. 70:915-927.
Sharp, G. D.
1978. Behavioral and physiological properties of tunas and
their effects on vulnerability to fishing gear. In G. D. Sharp
and A. E. Dizon (editors). The physiological ecology of tunas,
p. 297-449. Acad. Press, N.Y., San Franc, and Lond.
Shaw, E.
1970. Schooling in fishes: critique and review. In L. R. Aron-
son, E. Tobach, D. S. Lehrman, and J. S. Rosenblatt
(editors), Development and evolution of behavior, p.
452-480. W. H. Freeman and Company, San Franc.
Shimada, B. M., and M. B. Schaefer.
1956. A study of changes in fishing effort, abundance, and
yield for yellowfin and skipjack tuna in the eastern tropical
Pacific Ocean. Inter-Am. Trop. Tuna Comm., Bull. 1:347-
469.
Shingu, C, p. K. Tomlinson, and C. L. Peterson.
1974. A review of the Japanese longline fishery for tunas and
billfishes in the eastern Pacific Ocean, 1967-1970. Inter-
Am. Trop. Tuna Comm., Bull. 16:65-230.
Suzuki, Z., Y. Warashina, and M. Kishida.
1977. The comparison of catches by regular and deep longline
gears in the western and central equatorial Pacific. Bull.
Far Seas Fish. Res. Lab. (Shimizu), 15:51-89.
Turner, S. C.
1986. An analysis of recaptures of tagged bluefin with respect
to the mixing assumption. Int. Comm. Conserv. Atl. Tunas,
Collect. Vol. Sci. Pap. 24:196-202.
643
COMPARISON OF TWO LENGTH-FREQUENCY BASED PACKAGES FOR
ESTIMATING GROWTH AND MORTALITY PARAMETERS USING
SIMULATED SAMPLES WITH VARYING RECRUITMENT PATTERNS
Margarida Castro and Karim Erzini'
ABSTRACT
Length-frequency distributions were simulated for species with recruitment patterns characteristic of
many tropical fish; 1 ) one recruitment peak per year, fast growth and very high mortality, 2) one recruit-
ment peak per year, slow growth and moderate to high mortality, 3) two recruitment peaks per year,
slow growth and moderate to high mortality, and 4) random recruitment, slow growth and moderate
to high mortality. Two microcomputer program packages— one incorporating the ELEFAN I & II pro-
grams and the other implementing a form of Modal Progression Analysis— were used to estimate growth
and mortality parameters, and these were compared with the initial parameters used to generate the
simulated samples. The results, while generally encouraging, suggest that multiple recruitments per year
make it difficult to estimate growth and mortality parameters using these two packages.
Information concerning growth, mortality, and
recruitment patterns is of great importance in
lengtli-frequency analysis. The purpose of this paper
was to evaluate two sets of methods used in length-
frequency analysis in terms of their ability to
produce accurate estimates of growth and mortal-
ity parameters in the absence of such biological
information.
The methodology chosen consisted of generation
of length-frequency distributions with known pa-
rameters to which the length-frequency methods
were applied. The results obtained with the method
were compared with the initial conditions. This pro-
cedure has been used in other studies (Hampton and
Majkowski 1987; Jones 1987).
The development of the program for simulating
length frequencies was guided by assumptions im-
plicit in the length-frequency methods and by known
factors concerning the biology offish. These include
1) average individual growth in accordance with the
von Bertalanffy growth curve, 2) little variation in
natural mortality throughout the exploited phase,
3) exponential decline in the numbers of a cohort,
4) length distributions normal for each age class,
and 5) recruit numbers random. Some other fea-
tures of the program, such as the selectivity of the
gear (logistic type) are not standard assumptions of
length-frequency methods but are options for pa-
rameters necessary to describe the effect of fishing.
'Graduate School of Oceanography, Narragansett Bay Campus,
University of Rhode Island, Narragansett, RI 02882.
The authors believe that the simulated length fre-
quencies accurately reflect the assumptions of the
length-frequency methods and therefore the gener-
ated samples can be used to test, correct, and pos-
sibly improve these methods. The simulated sam-
ples might also help to define a range of situations
when a specific length-frequency method can or can-
not be used.
Traditionally, length-frequency analysis methods
have been used as validation methods for age deter-
minations made independently. Recently, these tech-
niques have grown in importance and frequency of
use, in particular in tropical fisheries, where age
determinations based on direct reading of check
marks in hard parts of the fish are difficult, and in
crustaceans, which do not have permanent hard
structures. As a result, length-frequency analysis
has been used in situations where very little is
known about the biology of the species.
It is the purpose of this work to contribute to the
understanding of the possible errors that are made
when length-frequency analysis is used without
biological information on mortality levels, growth
parameters, and, in particular, recruitment pat-
terns. It might be argued that such methods of
length-frequency analysis are particularly useful in
the situations described above, precisely because
they do not require a priori knowledge of biological
information. The question then becomes, is it legit-
imate to use length-frequency analysis techniques
in the absence of minimum biological information?
And if the answer to this question is no, then what
Manuscript accepted July 1988.
FISHERY BULLETIN: VOL. 86. NO. 4, 1988.
645
FISHERY BULLETIN: VOL. 8(5, NO. 4
are the minimum requirements for each one of the
length-frequency analysis methods? We hope this
paper can contribute to providing some understand-
ing of this problem.
MATERIALS AND METHODS
The Simulation Program: SIMULPOP^
SIMULPOP was developed in BASIC, for IBM
microcomputers and compatibles. Populations are
^The program SIMULPOP is available from Margarida Castro
upon request.
simulated by following cohorts in time and the gen-
eral characteristics and assumptions of the program
include 1) individual growth described by a von
Bertalanffy growth curve, no seasonality consi-
dered; 2) recruitment: different patterns and ran-
dom in numbers; 3) natural mortality: random and
normally distributed; 4) fishing mortality: random
and normally distributed, and corrected for incom-
plete selectivity in younger ages; 5) selectivity:
logistic equation considered to represent selectiv-
ity of the fishing gear; and 6) length distribution
for each age normally distributed, with a mean given
by the von Bertalanffy growth curve. The shapes
of the frequency distributions are independent of
CO
0)
0)
0)
c
0)
20.004
15.00
10.004
5.00
0.00
20.00
15.00
10.00
5.00
0.00
25.00
20.00
15.00-
10.00
5.00
0.00
25.00
20.00
15.00
10.00
5.00
0.00
30.00
25.00
20.00
15.00
10.00
5.00
0.00
Itl^
Jan
Jlllllu^
Mar
May
Jul
Sep
JllUu
Nov
3 5 7 9 11 13 15 1719
Length classes (1 cnn)
10.00
4 7 10 13 16 19 22 25 28 31
Length classes (1 cm)
Figure la.— Example of simulated length-frequency distributions for situation 1, the sardine-type species with one recruitment peak
a year. Only 6 of the 12 months are represented.
Figure lb.— Example of simulated length-frequency distributions for situation 2, the sparid/lutjanid-type species with one recruitment
peak a year. Only 6 of the 12 months are represented.
646
CASTRO AND ERZINI: COMPARISON OF LENGTH-FREQUENCY PACKAGES
catch size and can be regarded as unbiased. Exam-
ples of simulated length frequencies for each of the
four situations are presented in Figures la-d and
examples of component distributions contributing
to the composition of a particular distribution are
given in Figures 2a-d. In what follows, the word
cohort refers to the fish recruited in a particular
period. For one spawning peak a year cohort and
age class are equivalent. However, in the situa-
tions where multiple recruitment periods were
simulated, more than one cohort will contribute to
a given age class. In multiple recruitment situations
the word cohort does not have its traditional
meaning.
The Choice of Parameters
The following four situations were simulated and
the parameters are given in Table 1: 1) A sardine-
like species, characterized by small size, fast growth,
high mortality, very intense fishing mortality, and
with one recruitment peak per year (situation 1);
2) a small sparid/lutjanid type species, with larger
size, slower growth with moderate to high fishing
mortality, and one recruitment peak per year (situa-
tion 2); 3) a species with the same characteristics
as the one described previously, but with two re-
cruitment peaks per year (situation 3); and 4) a
species with the same characteristics as the two
0}
Oi
cr
a:
05
c
3 5 7 9 11 13 15 17 19 21 23 25 27 29
Length classes (1 cm)
^
10.00-
5.00-
0.00
10.00-
5.00-
0.00
2 10.00-
jiiiiiii..
Mar
■.■llllllllllli.,,. -1-^
Jul
Sep
lov
3 5 7 9 11 13 15 17 19 21 23 25 27 29
Length classes (1 cm)
Figure Ic— Example of simulated length-frequency distributions for situation 3, the sparid/lutjanid-type species with two recruit-
ment peaks a year. Only 6 of the 12 months are represented.
Figure Id.— Example of simulated length-frequency distributions for situation 4, the sparid/lutjanid-type species with random recruit-
ment. Only 6 of the 12 months are represented.
647
FISHERY BULLETIN: VOL. 8«. NO. 4
AU -
(/)
.
0)
•^ -^
S5
20 -
2i o
3 ^
cr
0) 0)
L. f
V- -c
x: >^
♦^ o
in -
^^
Qi ^
■•> Age class 1
-♦- Age class 2
-B- Age class 3
-■- Total of all ages
4 5 6 7 8 9 10 11 12 13 14
Length classes (1 cm)
-f— # — I — I — I-
15 16 17 18 19 20
-o- Cohort 12
-♦- Cohort 1 1
-D- Total age 1
-o Cohort 10
-»- Cohort 9
■9- Total age 2
-n-
Cohort 8
Cohort 7
/
\
2 -
-o-
Total age 3
J^
2 -
-D-
I 1 ■*'
Cohort 6
1
-♦-
Cohort 5
ir^
^X
-»-
Total age 4
.^
'^a.
0 -
1 1
1 T-n^BP 1
Figure 2a.— Example showing component distributions
for one length-frequency sample (month of November of
the same simulation represented in Figure la) for situa-
tion 1. Age class 4 is not represented due to low fre-
quencies.
Figure 2b.— Example showing component distributions
for one length-frequency sample (month of November of
the same simulation represented in Figure lb) for situa-
tion 2. Age class 6 is not represented due to low fre-
quencies.
Figure 2c.— Example showing component distributions for
one length-frequency sample (month of November of the
same simulation represented in Figure Ic) for situation
3. There are two cohorts contributing to each age class.
Ages 5 and 6 are not represented due to low frequencies.
Figure 2d.— Example showing component distributions
for one length-frequency sample (month of November of
the same simulation represented in Figure Id) for situa-
tion 4. In this case there are multiple cohorts contributing
to each age class. Age classes 1 and 6 are not represented
due to low frequencies.
5 10 15 20 25
Length classes (1 cnn)
30
648
CASTRO AND ERZINI; COMPARISON OF LENGTH-FREQUENCY PACKAGES
Age class
Age class 2
Age class 3
Age class 4
Age class 5
All age classes
7 9 11 13 15 17 19 21 23 25 27 23
Length classes (1 cm)
31
10 -1
Cohort 30
Cohort 29
Cohort 28
Cohort 27
Cohort 26
Cohort 25
Total age 2
15 20 25
Length classes (1 cm)
30
Cohort 24
Cohort 23
Cohort 22
Cohort 21
Cohort 20
Cohort 19
Cohort 18
Total age 3
Cohort 17
Cohort 1 6
Cohort 15
Cohort 14
Cohort 1 3
Cohort 12
Cohort 1 1
Total age 4
Cohort 10
Cohort 9
Cohort 8
Cohort 7
Cohort 6
Total age 5
649
FISHERY BULLETIN: VOL. Sfi, NO. 4
Table 1 .—Parameters chosen for the simulations.
Situation:
1
2
3
4
Oldest age present in catch (years)
4
6
6
6
Age of recruitment to the area
of adult stock (months)
6
6
6
6
Growth parameters
L„ (cm)
20
35
35
35
K
0.3
0.2
0.2
0.2
to (years)
0
0
0
0
Instantaneous annual
natural mortality rate
Mean
0.4
0.25
0.25
0.25
SD
0.015
0.01
0.01
0.01
Instantaneous annual
fishing mortality rate
l\/1ean
1.7
0.8
0.8
0.8
SD
0.05
0.02
0.02
0.02
Selectivity parameters (cm)
Mesh
1.5
3.5
3.5
3.5
Length 25% retension
3.25
11.25
11.25
11.25
Length 75% retension
8.75
16.75
16.75
16.75
Standard deviations of
length-at-age
Age 0
0.5
1.0
1.0
1.0
Age 1
1.0
1.2
1.2
1.2
Age 2
1.5
1.5
1.5
1.5
Age 3
1.5
2.0
2.0
2.0
Age 4
1.5
2.0
2.0
2.0
Age 5
2.0
2.0
2.0
Age 6
2.0
2.0
2.0
Recruitment pattern
once/yr
once/yr
twice/yr
stochastic
May-June
May-June
Mar. -Apr.
Sept. -Oct.
previous ones, but with stochastic recruitment, and
random recruitment intensities assigned to random-
ly selected months (situation 4).
The choice of these recruitment patterns was
based on the knowledge that tropical species have
different types of recruitment periodicities with
varying temporal and spatial variation (Thresher
1984). The same species may have different patterns
(Sale et al. 1984). Two principal peaks and year-
round recruitment have been reported in coral reef
fish from Hawaii (Walsh 1987) and Curasao (Luck-
hurst and Luckhurst 1977). One single recruitment
per year has also been reported (Li 1960; Gladstone
and Westoby 1988).
Mortality values, both natural and fishing, were
chosen to have small standard deviations (Table 1)
because the main objective was to examine the ef-
fects of different recruitment patterns, and we
wanted to keep the possible effects of variations in
other factors as small as possible.
The standard deviations of length-at-age used
(Table 1) are representative of values for species
with life history characteristics similar to those of
the cases considered for this study (K. Erzini, work
in progress).
The Length-Frequency Analysis
The two techniques chosen to represent length-
frequency analysis were ELEFAN (Electronic
Length Frequency Analysis) (Brey and Pauly 1986)
and the package entitled "Length Frequency Based
Fish Stock Assessment Microcomputer Programs"
(LFSA package) by Sparre (1986; adapted to MS
DOS by K. Erzini). ELEFAN has been widely used
in the analysis of tropical fish stocks, and its non-
parametric basis for determination of K and L^
makes it a unique methodology for the analysis of
length frequencies.
In ELEFAN, the length-frequency samples are
restructured in order to emphasize peaks. Details
of the restructuring methodology are given in Brey
and Pauly (1986). Growth curves are generated for
values of K and L^ within specified ranges and fit
to the reconstructed length-frequency data. The best
curves are considered to be the ones that pass
through the most peaks and the least troughs.
The LFSA package uses a method of a different
nature— the Bhattacharya method (Bhattacharya
1967), to separate normal curves, under the assump-
tion that the length distributions for each age are
650
CASTRO AND ERZINI: COMPARISON OF LENGTH-FREQUENCY PACKAGES
normal. The decomposition of each length-frequency
sample into component distributions is carried out
by plotting a logarithmic transformation of the dif-
ferences between successive length frequencies. A
normal distribution appears as a series of values
making up a straight line with a negative slope. In
the LFSA package, which implements the Bhat-
tacharya (1967) method, the user selects the points
believed to make up a normal distribution, and the
mean, standard deviation and various other
statistics are computed. In the next step, the means
of all distributions are plotted against time and the
mean lengths thought by the user to reflect the pro-
gression of a cohort are linked. Finally, growth
parameters are computed from the linked modes by
a method referred to as a Gulland and Holt plot
(Gulland and Holt 1959).
In both packages, the growth parameters are used
to create age based catch curves for estimation of
instantaneous annual total mortality, Z. Therefore,
except for the estimation of Z, the methodologies
of the two packages are quite independent. However
they are both characterized by a certain degree of
subjectivity.
Methodology
Following the suggestion of Hampton and Maj-
kowski (1987), two different teams were formed.
One (team A) created the simulated samples (10
cases of 12 monthly samples for each situation), and
another (team B) ran the length-frequency analysis.
The 40 cases were given arbitrary filenames and
were mixed by team A prior to analysis by team B.
This was done to avoid influencing the choice of ini-
tial values or parameter ranges, required by some
of the methods applied. In estimating the growth
parameters using the LFSA package, constraints
on the limit of acceptable estimates of L^ were
guided by the value of the midpoint of the largest
size class in each particular case. For analysis by
ELEFAN I & II, the size of the largest length class
also helped guide the choice of range of potential
values of L^. Team A provided information to
team B in different phases. In phase 1, samples were
provided to team B with information on mesh size,
and only broad descriptions of the type of species,
and indications of fishing mortality levels. Team B
analyzed the length-frequency samples with both
packages to the best of his ability. In Phase 2, exact
information on growth, mortality, and number of
age classes was provided and new estimates of Z
were obtained using both packages. The results pro-
duced by team B are presented in Table 2.
It should be noted that expected values for Z in
Table 2 are less than the sums of F and M in Table
1. This is because Table 1 values are inputs, and F
is subsequently corrected for selectivity.
RESULTS
In situation 1, the sardine-type species with one
recruitment peak per annum, the samples were
simulated using growth parameters typical of a
small clupeid with high fishing effort expressed by
a high value oiF and small mesh size. Thus a typical
length-frequency sample consists of 4 component
distributions or 4 cohorts (Figure 2a). While esti-
mates of the growth parameters by ELEFAN were
very good, the LFSA package estimates of i(^ were
surprisingly high.
Close examination of the length frequencies, the
Bhattacharya method and Gulland and Holt plot im-
plemented by Sparre revealed a number of factors
Table 2.— Results of estimation of growth and mortality parameters (mean and standard deviation) using
ELEFAN and LFBFSA packages. Z, and Zj are total mortalities calculated using estimated and actual K
and L„ values.
Parameters
ELEFAN
LFBFSA
Expected v
alues
Situation
K
/-»
^1
^2
K
/.»
^1
^2
Z
1
Mean
0.30
21.3
2.09
1.78
0.46
21.0
3.01
1.96
0.3
20.0
1.32
SD
0.03
1.30
0.41
0.26
0.14
3.55
0.67
0.25
2
Mean
0.18
38.8
1.35
1.22
0.18
33.2
0.89
1.17
0.2
35.0
1.02
SD
0.02
3.10
0.27
0.14
0.06
2.97
0.33
0.22
3
Mean
0.19
37.9
1.24
1.17
0.14
36.1
0.79
1.09
0.2
35.0
1.02
SD
0.03
2.33
0.27
0.18
0.05
3.84
0.38
0.13
4
Mean
0.19
37.2
1.24
1.15
0.20
36.2
1.15
1.14
0.2
35.0
1.02
SD
0.04
2.93
0.40
0.20
0.06
4.35
0.20
0.11
651
FISHERY Hl'LLETIN: VOL. 86, NO. 4
contributing to tlie high estimates of K. First, the
relatively fast growth rate and high mortality re-
sulted in early overlapping or accumulation of dis-
tributions and in large fish being rare so that team
B could never identify more than 3 out of 4 modes
corresponding to fished age classes using the Bhat-
tacharya method in any sample (Fig. 2a). Second,
the third mode was consistently overestimated
because the distributions for age classes 3 and 4
were merged together. Third, the young-of-the-year
fish do not appear in the samples as a well-defined
distribution until late in the monthly time series of
samples because recruitment does not take place un-
til June. Finally, we found that estimates of /iT using
the Gulland and Holt plot in the LFSA package were
very sensitive to small deviations in the estimates
of modal lengths obtained using the Bhattacharya
method.
Total mortality estimates for situation 1 using
the estimated K and L^ values were not good for
either package. Estimates of Z using the actual K
and L^ values used in the simulations were within
35% and 45% of the expected Z (Table 2).
Situation 2 was the sparid/lutjanid type, character-
ized by a single recruitment peak per year and 5
distributions corresponding to the 5 fished cohorts
in the catch (Fig. 2b). Both methodologies gave
similar estimates of K, close to the actual value.
However, L^ was overestimated by ELEFAN and
underestimated by the LFSA package. The mean
estimate of Z was within 32% of the expected Z for
the ELEFAN catch curve analysis and within 13%
for the LFSA package analysis. Mean Phase H
estimates of Z were 20% and 15% above the ex-
pected Z (Table 2).
Situation 3, the sparid/lutjanid type with two re-
cruitment peaks per year (Fig. 2c) produced good
results using ELEFAN. However, modal progres-
sion estimates of K were low, with corresponding
underestimates of Z (Table 2). Component distribu-
tions were poorly defined compared to situation 2;
age classses 4 and 5 were often obscured by the age
class 3 distribution. Incorrect separation of distribu-
tions and bad estimates of growth parameters were
therefore not unexpected.
For the last situation, the sparid/lutjanid type with
stochastic recruitment (Fig. 2d), estimates of K,
L^, and Z were generally good for both packages.
However, as shown by the standard deviations, the
range of estimates for certain parameters was
quite high. This was the case for K estimated by
ELEFAN and L^ estimated by modal progression
analysis in the LFSA package.
DISCUSSION
This preliminary study has shown that, as ex-
pected, the structure of the data has a big effect on
the estimates derived using length-frequency pack-
ages. In general, the results were encouraging.
However, it should be noted that the simulated
length-frequency distributions can be regarded as
representing high-quality samples of the hypothe-
tical populations in terms of lack of bias, sample
size, and frequency of sampling. In other words,
real life length-frequency data is seldom of this
quality.
The modal progression analysis implemented by
Sparre (1986) was more sensitive to the structure
of the length-frequency samples. Worst results in
terms of estimation of growth parameters were ob-
tained under multiple recruitment (situation 3) and
fast growth and high mortality (situation 1). A fun-
damental problem with the Bhattacharya method
is the inability to identify modes at the upper end
of the size spectrum, particularly when there is fast
growth or many age groups. Identification of modes
using the Bhattacharya method might have been im-
proved by using smaller size class intervals, par-
ticularly for situation 1. However, even when there
was little ambiguity in the selection of modes using
the Bhattacharya method, it was found that the
Gulland and Holt plot for estimating K and L^ was
very sensitive to small underestimates and over-
estimates of the modes considered to represent
growth over time.
Length converted catch curve estimates of total
mortality are highly dependent on the estimated
growth parameters. Consequently, estimates of Z
generally paralleled estimates of K and L^ and
were not as good as estimates of Z obtained using
the actual simulation values of K. These latter esti-
mates of Z were generally close to actual Z values
for all situations despite the fact that the length-
frequency data necessarily did not meet steady-state
assumptions because of variable recruitment and
mortality. However, the variability of mortality
rates was deliberately kept small because the pri-
mary objective was to examine the effects of differ-
ent recruitment patterns.
ELEFAN, the Bhattacharya, and the modal pro-
gression method of the LFSA package all require
subjective decision making by the user. It would
seem that ELEFAN is less subjective or that poor
choices are less likely to be made by the user than
in the selection of modes by the Bhattacharya
method and in the choice of modes for the modal
652
CASTRO AND ERZINI: COMPARISON OF LENGTH-FREQUENCY PACKAGES
progression analysis implemented in the LFSA
package.
We feel that length-frequency analysis should not
be used in the complete absence of information on
growth and recruitment patterns or with very small
data sets. Other important information includes data
on migration and seasonal patterns in distribution,
and such information should be used to guide sam-
pling programs. Irregular recruitment both in terms
of level and pattern may strongly affect the results.
Clearly, length-frequency analysis can be a useful
tool when used in conjunction with other methods.
However, it seems unreasonable to expect such tech-
niques to produce reliable information when the
classical methods of fisheries fail or cannot be used.
For example, traditionally, growth parameters have
been estimated from age-length keys and mortal-
ities derived from the age structure of the catch.
In cases where the age-length key cannot be ob-
tained, there is a temptation to obtain growth
parameters at any cost using length-frequency anal-
ysis. If this is done, great care should be taken to
ensure that a minimum amount of biological infor-
mation exists. The use of length-frequency analysis
as a "black box" where a length-frequency distribu-
tion goes in from one side and a whole set of
biological parameters emerge does not seem correct.
If as a first step, the data are plotted and there is
no visual evidence of progressing modes, then even
if biological information is available, length-fre-
quency techniques should perhaps not be applied at
all.
To have a more complete picture of the limitations
and usefulness of length-frequency techniques, a
much wider range of conditions must be tested. For
example, the effect of variations in individual
parameters particularly M and F, and in combina-
tions of parameters must be tested. The effects of
size class width on length-frequency analysis is also
an area which should be investigated. It is the in-
tention of the authors to continue this work in order
to examine as wide a range of situations as possible.
ACKNOWLEDGMENTS
The authors would like to express their gratitude
to Saul Saila for his guidance, advice, and support
during the course of this work. The authors would
also like to thank D. Pauly, P. Sparre, and an anony-
mous reviewer for their comments. This work was
sponsored in part by USAID Grant No. DAN-
4146-G-SS-5071-00 (Fisheries Stock Assessment
CRSP).
LITERATURE CITED
Bhattacharya, C. G.
1967. A simple method of resolution of a distribution into
Gaussian components. Biometrics 23:115-135.
Brey, T., AND D. Pauly.
1986. Electronic length frequency analysis. A revised and
expanded user's guide to ELEFAN 0, 1 and 2. Ber. Inst.
Meereskd. Christ. -Albrechts-Univ. Kiel, No. 149, 77 p.
Gladstone, W., and M. Westoby.
1988. Growth and reproduction in Canthigaster valentini
(Pisces, Tetraodontidae): a comparison of a toxic reef fish
with other reef fishes. Environ. Biol. Fishes 21:207-221.
GULLAND, J. A., AND S. J. HOLT.
1959. Estimation of growth parameters for data at unequal
time intervals. J. Cons. Int. Explor. Mer 25:47-49.
Hampton, J., and J. Majkowskl
1987. An examination of the accuracy of the ELEFAN com-
puter programs for the length-based stock assessment.
In D. Pauly and G. R. Morgan (editors), Length-based
methods in fisheries research, p. 203-216. ICLARM Conf.
Proc. 13, Manila.
Jones, R.
1987. An investigation of length composition analysis using
simulated length compositions. In D. Pauly and G. R.
Morgan (editors). Length-based methods in fisheries
research, p. 217-238. ICLARM Conf. Proc. 13, Manila.
Li, K.-W.
1960. On the biology of the Hong Kong Golden Thread,
Nemiptems inrgatus (Houttyun). Hong Kong Univ. Fish.
J. 3:89-109.
Luckhurst, B. E., and K. Luckhurst.
1977. Recruitment patterns of coral reef fishes on the fring-
ing reef of Curasao, Netherlands Antilles. Can. J. Zool.
55:681-689.
Sale, P. F., P. J. Doherty, G. F. Eckert, W. A. Douglas, and
D. J. Ferrel.
1984. Large scale spacial and temporal variation in recruit-
ment to fish populations on coral reefs. Oecologia (Berl.)
64:191-198.
Sparre, P.
1986. Length frequency based fish stock assessment micro-
computer programs (basic program package for the micro-
computer Apple II), Part I, User's manual. Dan. Inst. Fish.
Mar. Res., Charlottenlund Slot, 2920 Charlottenlund, Den-
mark.
Walsh, W. J.
1987. Patterns of recruitment and spawning in Hawaiian reef
fishes. Environ. Biol. Fishes 18:257-276.
Thresher, R. E.
1984. Reproduction in reef fishes. Trop. Fish Hobby. Nep-
tune, NJ.
653
SPECIFYING A FUNCTIONAL FORM FOR THE INFLUENCE
OF HATCHERY SMOLT RELEASE ON ADULT SALMON PRODUCTION
BiiNG-HwAN Lin' and Nancy A. Williams^
ABSTRACT
The hypothesis of density independent marine survival of salmon has been tested extensively with con-
flicting results. Unduly restrictive functional form and data deficiency have been suggested as the major
contributing factors to the mixed results. Focusing on the issue of functional form selection, this paper
utilizes the extended Box-Cox flexible functional form which allows the data to determine the statistical
relationship between smolts and adult production without a priori restrictions. The model is applied to
Hokkaido chum salmon, Oncorhynchus keta, and Oregon coho salmon, 0. kisutch. Empirical results sug-
gest the existence of density dependence for both Hokkaido chum and Oregon coho salmon. Further,
an increasing variability of adult production with respect to an increase in smolts is found for Hokkaido
chum salmon but not for Oregon coho salmon.
Two issues pertaining to the relationship between
the number of hatchery smolts released and the
number of adult salmon returned have been inves-
tigated recently in the literature. First, the hypoth-
esis of density independence in the relationship be-
tween salmon adults and smolts has been tested. The
null hypothesis is a linear relationship between
adults returned and smolts released, such that the
additional adult salmon produced from an increase
in smolts released is constant. The second issue is
the relationship between the variability of adult pro-
duction and the number of smolts released. If, in
fact, increases in smolts increase the variability of
adults produced, fishery management strategies can
be improved by considering the trade-off between
the mean and variance of adult salmon returned
(Walters 1975; McCarl and Rettig 1983).
The empirical results of the test of density in-
dependence have been mixed. Nickelson (1986) pro-
vided an excellent discussion of previous results
pertaining to the test of density independence for
Oregon coho salmon. In short, this hypothesis for
marine survival of Oregon coho salmon is rejected
by McCarl and Rettig (1983) and Peterman and
Routledge (1983), but accepted by Peterman (1981),
Clark and McCarl (1983), and Nickelson (1986). In
addition, biologists in the Oregon Department of
Fish and Wildlife have examined several model
specifications and manipulations in data sets and
'Department of Agricultural Economics and Rural Sociology,
University of Idaho, Moscow, ID 83843.
^Department of Economics, Loyola College, 4501 North Charles
St., Baltimore, MD 21210-2699.'
have drawn conflicting conclusions about density in-
dependence. This led McCarl and Rettig to suggest
that conflicting conclusions are caused by the use
of different functional form specifications and to
suggest further that resolution of the issue of den-
sity independence in Oregon coho salmon requires
more refined data. In the case of Hokkaido chum
salmon, the null hypothesis of density independence
fails to be rejected by McCarl and Rettig.
Regarding the estimation of the variability of adult
salmon production, Peterman (1981) pointed out the
importance of functional form specification and
argued for the use of the multiplicative-error model
rather than an additive-error model. McCarl and
Rettig (1983) demonstrated that the specification of
a multiplicative-error model imposes unwarranted
restrictions on the estimation of the variability in
adult production. McCarl and Rettig utilized the
specification developed by Just and Pope (1978,
1979). As a result, the variability in adult salmon
production is estimated and conflicting conclusions
of the test of density independence emerged.
It is apparent that functional form specification
is critical in the test of density independence and
the estimation of variability in adult production. The
purpose of this paper is to reexamine these two
issues by using the extended Box-Cox flexible func-
tional form. Both Hokkaido chum salmon, Oncor-
hynchus keta, and Oregon coho salmon, 0. kisutch,
data are used in this study.
The next section of the paper discusses the im-
portance of the functional form specification and
demonstrates the superior flexibility of the Box-Cox
Manuscript accepted July 1988.
FISHERY BULLETIN: VOL. 86, NO. 4, 1988.
655
functional form compared with the Just-Pope spe-
cification used by McCarl and Rettig (1983). The em-
pirical results of the Box-Cox functional form are
then discussed and compared with those of McCarl
and Rettig.
METHODS
Previous findmgs of Peterman (1978, 1981) in-
dicate the importance of the assumption made
regarding error term in testing the hypothesis of
density independence. In the process of examining
the effect of the number of released smolts on the
production of adults and its variability, Peterman
(1981) employed two alternative (additive-error and
multiplicative-error) model specifications of the
error term:
^1 = CjS^^i + Vi . . .Additive-error model (1)
Ag = C2S''2 expiVo) ...Multiplicative-error
model (2)
where A, = adult production of salmon using spe-
cification i,
S = number of smolts,
C, = survival rate parameter in model i,
/c, = density dependence parameter in
model i,
Vi = error term for model i.
By applying these two models to several sets of
salmon data, Peterman (1981) concluded that the
multiplicative-error model appears to generate
better statistical results than its counter model. In
addition, the results of the multiplicative-error
model suggest that an increase in the number of
smolts will increase the variation in total adult
returns. Because the variability in adult production
is not only influenced by the number of smolts but
also other factors affecting the survival of smolts
such as the body size of released smolts, Peterman
(1981) suggested that model (2) should be modi-
fied by including more explanatory variables. By
following Peterman's suggestion a third model can
be specified with the additional variable body size,
B:
A3 = CgS^sS'^s exp(y3).
(3)
The mean and variance of adult production for this
model can be expressed as
E{A^) = CsS^sBdsEiexpiVs))
FISHERY BULLETIN: VOL. 86, NO. 4
Var(A3) = (C3S^'3B''3)2 Var(exp(y3)).
The instantaneous rates of change in mean and
variance of adult production with respect to smolt
body size can be expressed as
dE{A^)ldB = d^EiA^yB
dVar{As)/dB = 2iYar{A._i)/EiA:^))dE{A:^)/dB.
If body size of smolts is enlarged, one would ex-
pect higher yields, dE{As)/dB > 0. Since both
mean and variance are positive, the above model im-
poses a restriction that the smolt body size has a
positive effect on variability, 3Var(A3)/a5 > 0.
This restriction is unwarranted because of lacking
theoretical support; rather the effect (positive, nega-
tive, or zero) of body size on variability of adult
return should be tested empirically. For this reason,
McCarl and Rettig (1983) adopt a model developed
by Just and Pope (1978, 1979) which can be ex-
pressed as
A4 = C,S''iB''i + C^S'^^B''^ expiV^)
= fiS,B) + h'HS,B) exviV,)
(4)
where h'''{S,B) is a component of the standard
deviation of adult production as shown below.
The mean and variance of adult production for this
model can be expressed as
EiA^) = C.S'^iB'^i + C^S''^B''5E{exp{V^))
Var(A4) = (C5S^-55<5)2Var[exp(l/4)].
Because the signs of d^ and d^ are to be deter-
mined in the estimation, the advantage of model (4)
over model (3) is that it allows for body size to have
a positive effect on mean return and unknown
(positive, negative, or zero) effect on the variabil-
ity of adult production.
There are problems inherent in model (4), how-
ever, the first being that this specification produces
a constant percentage change (^4) in adult produc-
tion when the number of smolts released changes
by 1%, a constant output elasticity, £„,. Output
elasticity is an economic term which is widely used
in measuring the relationship between input (smolt
release) and output (adult production) and has the
advantage of being unit free. An output elasticity
of 1.0 means that an increase in smolt release by
1% will result in the same percentage increase in
656
LIN an-i WILLIAMS: INFLUENCE OF SMOLT ON ADULT SALMON
adult production, implying density independence.
When the hypothesis of density independence is re-
jected, output elasticity is less than 1.0. Therefore,
there is an one-to-one correspondence between
the hypothesis of density independence and the
value of output elasticity. The purpose of using
the concept of output elasticity here is to facilitate
the discussion of the restriction inherent in model
(4).
There is no theoretical support for imposing the
restriction of constant output elasticity, rather it
should be treated as a hypothesis to be tested. More
important, body size (B) is likely to have a positive
effect on the output elasticity, i.e., de^JdE > 0. In
other words, when body size of smolts is enlarged,
the improved ability of enduring unfavorable envi-
ronmental conditions should increase the incre-
mental return rate of adult salmon. But, a constant
output elasticity implies that body size and the out-
put elasticity are independent.
The comparison between model (3) and model (4)
centers around the role of body size in the variability
of adult production. However, data on body size is
unavailable so that the comparison becomes em-
pirically irrelevant. Consequently, the difference
between these two m.odels, in essence, rests on
model specification. It should also be pointed out
that the estimate of h'''{S,B) is influenced by the
functional form of f{S,B) and vice versa, because
h'-{S,B) is the heteroscedastic error term to be
handled by the weighted least squares method. It
is, therefore, important to select a more general
functional form for the mean and variance of adult
production in testing the hypothesis of density in-
dependence and in estimating the variability in adult
production.
The Box-Cox flexible functional form developed
by Box and Cox (1964) and extended by Zarembka
(1974) has been a popular tool for both discrimi-
nating among alternative functional forms and pro-
viding added flexible form in model specification
(Moschini and Meilke 1984). The extended Box-Cox
functional form for relating adult salmon produc-
tion to smolts and other explanatory variables X
(such as upwelling) can be expressed as
^<^' = a. + a,SW + a.X'S' +
where A*'*' =
S^^^ =
(A^ -
In A
1)/A
(5f^^ - I)/ IX
In 5
(5)
for A ?t 0
for A = 0
for ^ 9^ 0
for /.i = 0
v(e) ^ i i^Q ■
\\nX
e ~ NID(0, o2).
1)1 B
for 0 ^ 0
for 0 = 0
Model (5) includes the linear (A = /.< = 0 = 1),
multiplicative-error or log-log (A = f^ = 0 = 0), and
log-linear (A = 0, /i = 0 = 1) functional forms as
special cases. Therefore, models (3) and (4) are
special cases of model (5), which allows both non-
constant output elasticity and nonzero effect of X
on the output elasticity.^
When the variability of adult salmon production
is affected by the values of its explanatory variables,
the error term has nonconstant variance, i.e.,
heteroscedasticity. Zarembka (1974) demonstrated
that while the Box-Cox model is fairly robust to
departures from normality, it is sensitive to hetero-
scedasticity. Failure to correct for this problem can
generate misleading results (Lahiri and Egy 1981).
When heteroscedasticity is present, we can also
specify a Box-Cox functional form for the variance
of the error term in model (5) as the following:
e ~ NID{Q, h{S;X) o|)
where h{S^) = p^S^'^ + [i^X^^K
(6)
The parameters, a,, /3,, A, ^x, 0, t, and ^ can be
estimated by maximum-likelihood algorithms (see
Appendix for a discussion of the log-likelihood func-
tion and estimation methods). The hypothesis of den-
sity independence can be tested by estimating the
model with the restriction that A = fi = 1 against
the unrestricted model.
'The superior flexibility of model (5) compared with models (3)
and (4) is an important consideration in testing the hv'pothesis of
density independence in light of the following remarks on the com-
parison of models (1) and (2) in Peterman (1981, p. 1117):
"This is not to say that model 2 is the 'true' form of natural
variability, because there are numerous other models that were
not tested here (many of these alternatives cannot be tested in
practice). ..."
We also cannot claim that the extended Box-Cox functional form
can produce the "best" or "true" functional form. There exist other
flexible functional forms, such as Fourier (Gallant 1984), and the
literature is silent in the comparison of these flexible functional
forms.
Even though the Box-Cox functional form was first proposed
in 1964, its application and investigation of its statistical proper-
ties have not received much attention until recently. Therefore,
there are numerous aspects of transformations that merit further
study (Box and Cox 1982). Lacking software support also makes
its application difficult. Nevertheless, the superior flexibility of the
Box-Cox functional form compared with other functional forms
used traditionally is evident and its application should be encour-
aged.
657
FISHERY BULLETIN: VOL. 86. NO. 4
RESULTS
In order to test the hypothesis of density inde-
pendence for salmon utilizing the extended Box-Cox
flexible functional form, the two data sets analyzed
by McCari and Rettig (1983) were also used here.
The first data set contains total Hokkaido hatchery
chum salmon fry releases and brood year adult
returns for the years 1950 through 1969 (Moberly
and Lium 1977). The second data set pertains to
Oregon coho salmon for the years 1960 through
1980 (Oregon Department of Fish and Wildlife
1982). This latter data set was also analyzed by
Clark and McCarl.
Hokkaido Chum Salmon Results
Due to the lack of data on body size and other
factors affecting the survival rate of fry, adult pro-
duction (in thousands) is estimated with the single
explanatory variable, number of fry released (in
millions). As explained in the Appendix, the depen-
dent variable is divided by its geometric mean of
3,332,440. The iterated weighted least squares
method produces the following maximum log-likeli-
hood results with the ^statistics given in parenthe-
ses below the coefficients:
^(-0.4) ^ _ 1,254 -l,756.3S'-i-4) (7)
(-4.07) (4.07)
R- = 0.48, Durbin-Watson = 1.5, Log-likelihood
= -7.89.
The weight used to remove heteroscedasticity is
5(1.06) ^ (51.06 _ i)/i.o6. This implies that as the
number of fry is increased by 1%, the standard
deviation of the adult production increases by 1.06%,
which is much smaller than the 2.5% reported by
McCarl and Rettig (1983).
The above results suggest that the output elas-
ticity" of fry is 1,756.3A''^S "i-^. When evaluated at
the mean values oiA and S (which are 1.1278 and
288.36, respectively), a percentage increase in the
number of fry increases the adult production by
0.66%, implying density dependence. In addition,
^Equation (7) can be written as
{A.-^-^ - l)/(-0.4) = -1254 + 1756.3(S-'-^
A-"^" = 0.8 -h 501.85-^'*.
l)/(-1.4), or
Output elasticity £„^ = (%M)/(%A5) = d.A\dS) {SI A) =
l.TSe.SS-^M"" dtJdS = -2,458.8S-^-M"'^ < 0.
658
this output elasticity is a decreasing function of fry
releases. In contrast to these results, McCarl and
Rettig (1983) reported a constant output elasticity
of 1.09 which was not found to be statistically dif-
ferent from 1.0, supporting density independence.
The hypothesis of density independence was for-
mally tested by estimating the linear relationship
between A and S (i.e., the power transformations
for A and S are restricted to be one) by using the
weighted least squares method (S was treated as the
weight) with the following results:
A = 0.18 + 0.0034S
(0.6) (2.84)
(8)
R" = 0.31, Durbin-Watson = 1.04, Log-likelihood
= -12.56.
The weighted least squares method produces a
Durbin-Watson value of 1.04 which is below the
lower limit of its critical value, suggesting the possi-
ble existence of autocorrelation. However, the
Durbin-Watson value is well known to be below the
lower limit (or above the upper limit) which could
be the cause by model misspecification or autocorre-
lated error terms. Because the use of improper
functional form is a model misspecification, the
extended Box-Cox functional form needs to be
explored before assuming the existence of an
autocorrelation problem in light of low (or high)
Durbin-Watson statistics. The extended Box-Cox
results have a Durbin-Watson statistic of 1.50 (im-
plying no autocorrelation), and hence, it is concluded
that the low Durbin-Watson value in Equation (8)
is a result of incorrect functional form. Since the
Durbin-Watson statistic is for detecting first-order
autocorrelation, the least squares procedure de-
scribed in Pagan (1974) was applied to test higher-
order autocorrelation. It is concluded that the Box-
Cox results are free from autocorrelation problems,
first or higher orders.
The hypothesis of density independence can be
tested by comparing the log-likelihood values of
Equations (7) and (8). The test statistic of twice the
difference between the log-likelihood functions
under the two specifications follows a chi-square
distribution with the number of degrees of freedom
equal to the number of restrictions (Theil 1971). This
test procedure is similar to the Akaike Information
Criterion (Akaike 1974) and has the advantage of
testing the significance of the difference between
the log-likelihood functions of different model spe-
cifications. It is concluded that the density-indepen-
LIN and WILLIAMS: INFLUENCE OF SMOLT ON ADULT SALMON
dence hypothesis can be rejected at a 1% level
with a critical value of 9.21 at 2 degrees of free-
dom.
For comparison purposes, the Hokkaido chum
salmon data was used to fit the multiplicative-error
model (model (2)) by applying the weighted least
squares method with the following results:
ln(A) = -3.27 + 0.583 ln(5)
(2.06) (2.06)
(9)
R- = 0.19, Durbin-Watson = 1.15, Log-likelihood
= -11.22.
By comparing the values of the log-likelihood func-
tion of Equations (7) and (9), it can be concluded that
the multiplicative-error model can be rejected a 5%
significance level. Even though the multiplicative-
error model produces a bigger log-likelihood value
than the linear model, the difference between these
two log-likelihood values is not statistically signifi-
(0
c
o
c
CD
(0
T3
<
3 -
3
s
-
Linear
Q
B
Multiplicative-
-
Q
a
Error i . . j__.
/-
-^
,.^'-"^^-'
.-- '
Box-Cox
Q
^^:<:^
-
3 ^
^
a
a
a
/ a
I
s
1
1
a
1
1
Q
1 1 1
1
140 190
240 290 340 390 440
Fry Released (millions)
490 540
Figure 1.— Plot of Hokkaido chum salmon data and estimated relationships: 1950-1969. Dots are actual obser-
vations, linear model is Equation (8), multiplicative-error model is Equation (9), and Box-Cox model is Equation (7).
659
FISHERY BULLETIN: VOL. Sfi. NO. 4
cant. Figure 1 shows the data set with estimated
relationships, Equations (7)-(9), superimposed. It is
evident that the Box-Cox specification produces a
relationship of much bigger curvature and better fit
than the multiplicative-error specification.
Oregon Coho Salmon Results
McCarl and Rettig (1983) suggested that aggre-
gated (wild and hatchery) adult coho salmon produc-
tion (in thousands) is affected by smolt releases (in
millions), S, and upwelling index, U. The flexible
functional form for Oregon coho salmon can be ex-
pressed as Equation (5). The model was estimated
by iterated ordinary least squares with the follow-
ing maximum log-likelihood results:
A(-o*5) =-50.08 + 93.715(-20) + 0.59[/«'<" (10)
(-1.17) (1.08) (3.37)
R^ = 0.51, Durbin-Watson = 2.07, Log-likelihood
= -1.66.
To detect any violations of the assumption regard-
ing the homoscedastic error term, a series of tests
were conducted by running regressions of squared
residuals (e-) or logs of (e-) on the predicted values
of A or the explanatory variables S and U. The
regression of e^ on all explanatory variables is
known as the Breusch-Pagan-Godfrey test and the
regression of log(e") on all explanatory variables
are known as the Harvey test (White 1987). Five
tests were conducted using the chi-square distribu-
tion, and the assumption of homoscedastic error fails
to be rejected at a 5% significance level. The same
conclusion is reached when model (2) was fitted by
Peterman (1981). The Box-Cox results are also
found to be free from autocorrelation problems, first
or higher orders.
Empirical results as summarized in Equation (10)
indicate that the number of smolt released con-
tributes positively to adult production at a 15%
significance level. Upwelling also positively affects
adult production at a 1% significance level. The Box-
Cox results produce a nonlinear relationship be-
tween adult production and smolt release and an
output elasticity of less than one, suggesting that
the null hypothesis should be rejected. To formally
test the hypothesis of density independence, the
power transformations for A and S are restricted
to be 1.0 and the Box-Cox functional form is re-
estimated with the following results:
A = -0.58 + 0.000655 + 0.084[/(''^^' (11)
(-1.2) (0.1) (3.6)
R'^ = OAl, Durbin-Watson = 2.09, Log-likelihood
= -6.54.
By comparing the values of the log-likelihood
functions for Equations (10) and (11) and follow-
ing a chi-square test with 2 degrees of freedom, it
is concluded that the hypothesis of density in-
dependence for Oregon coho salmon can be rejected.
The same conclusion was reached by Peterman
and Routledge (1983) and McCarl and Rettig
(1983).
CONCLUSION
The findings of testing the hypothesis of density-
independent marine survival for salmon and of the
effect of the number of smolts released on the vari-
ability of adult production have important implica-
tions for fishery managers as noted in the literature.
If the hypothesis of density independence fails to
be rejected, there is no technical maximum^ for the
adult production from releasing smolts. A technical
maximum of adult production exists when the num-
ber of smolts has a positive and decreasing effect
on adult production. If the variability of adult pro-
duction is positively affected by the number of
smolts, it will be useful for fishery managers and
the fishing industry to know the form of variability
to evaluate the effectiveness of salmon hatchery
operations. Further, fishery managers can improve
management strategies by considering the trade-off
between the mean and variance of adult production.
The hypothesis of density independence has been
tested extensively for different sets of data with con-
flicting results. Functional form selection and data
deficiency have been suggested as the causes of con-
flicting findings.
Results of this study confirm that functional form
selection is critical in testing the hypothesis of den-
sity independence and estimating the form of the
variability of adult production. By using the ex-
tended Box-Cox functional form, it is concluded that
there exists a density-dependent relationship be-
^A "technical maximum" refers to the maximum adult produc-
tion in physical terms. This may not be an appropriate objective
for fishery managers to achieve, because the release of smolts at
technical maxima may not generate maximum benefits to the
fishing industry. Maximization of the return to hatchery operations
appears to be a more suitable objective of a single-attribute model
to be accomplished without considering the risk factor.
660
LIN and WILLIAMS: INFLUENCE OF SMOLT ON ADULT SALMON
tween the adult production and the number of chum
salmon fry released in Hokkaido. Also, as the num-
ber of fry increases, the variability in adult produc-
tion increases as well. The results reported by
McCarl and Rettig (1983), using the Just-Pope
specification (a special case of extended Box-Cox),
conclude that the hypothesis of density indepen-
dence fails to be rejected and the effect of the
number of fry on the variability of adult production
is more than twice that of this study. The Box-Cox
results of aggregated Oregon coho salmon also in-
dicate density dependence, and the same conclusion
is also reached by McCarl and Rettig (1983) and
Peterman and Routledge (1983). Because Nickelson
(1986) reached a different conclusion using disag-
gregated data, the use of the extended Box-Cox
specification to analyze disaggregated data for
Oregon coho salmon is, therefore, recommended by
the authors as a possible research need. However,
partitioning the data set according to high and low
upwelling, for example, will lead to the problem of
insufficient degrees of freedom, as pointed out by
an anonymous reviewer. This can be overcome only
after a sufficient number of years of data collection
have transpired. Finally, data reliability needs to be
secured before the selection of functional form can
improve our understanding of this issue.
ACKNOWLEDGMENTS
This work is the result of research sponsored
by the Alaska Sea Grant College Program, coop-
eratively supported by the U.S. Department of
Commerce, NOAA Office of Sea Grant and Extra-
Mural Program, under grant number NA86AA-
D-SG041, project number R/14-09; and by the
University of Idaho with funds appropriated by the
state.
LITERATURE CITED
Akaike, H.
1974. A new look at the statistical model identification. IEEE
Trans. Autom. Control AC-19:716-723.
Box, G. E. P., AND D. R. Cox.
1964. An analysis of transformations. J. Royal Stat. Soc.
26(Ser. B):211-243.
1982. An analysis of transformations revisited, rebutted.
J. Am. Stat. Assoc. 77:209-210.
Clark, J., and B. A. McCarl.
1983. An investigation of the relationship between Oregon
coho salmon (0. kisutch) hatchery releases and adult produc-
tion utilizing the law of the minimum regression. Can. J.
Fish. Aquat. Sci. 40:516-523.
Gallant, R.
1984. The fourier flexible form. Am. J. Agric. Econ. 66:
204-208.
Just, R. E., and R. D. Pope.
1978. Stochastic specification of production functions and
economic implications. J. Econ. 7:67-86.
1979. Production function estimation and risk related con-
siderations. Am. J. Agric. Econ. 61:276-284.
KOUTSOYIANNIS, A.
1977. Theory of econometrics. 2ded. The McMillan Press,
Lond.
Lahiri, K., and D. Egy.
1981. Joint estimation and testing for functional form and
heteroscedasticity. J. Econ. 15:299-307.
McCarl, B. A., and R. B. Rettig.
1983. Influence of hatchery smolt releases on adult salmon
production and its variability. Can. J. Fish. Aquat. Sci.
40:1880-1886.
Moberly, S. a., and R. Lium.
1977. Japan salmon hatchery review. Alaska Dep. Fish
Game, Div. Fish. Rehabil. Enchance. Dev., 124 p.
Moschinl G., and K. D. Meilke.
1984. Parameter stability and the U.S. demand for beef.
West. J. Agric. Econ. 9:271-282.
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.
Oregon Department of Fish and Wildlife.
1982. Comprehensive plan for production and management
of Oregon's anadromous salmon and trout. Part II - coho
salmon plan. Oreg. Dep. Fish. Wildl., Portland, OR, 188
P-
Pagan, A. R.
1974. A generalized approach to the treatment of autocorre-
lation. Aust. Econ. Pap. 13:267-280.
Peterman, R. M.
1975. "Ocean effects" in salmon. PR-3 Inst. Resour. EcoL,
37 p. Univ. Br. Columb., Vancouver, B.C.
1978. Testing for density dependent marine sur\'ival in
Pacific salmonids. J. Fish. Res. Board Can. 35:1434-
1450.
1981 . Form of random variation in salmon smolt-to-adult rela-
tions and its influence on production estimates. Can. J.
Fish. Aquat. Sci. 38:1113-1119.
Peterman, R. M., and R. D. Routledge.
1983. Experimental management of Oregon coho salmon (On-
corhynchus kisutch): designing for yield of information.
Can. J. Fish. Aquat. 40:1212-1223.
Seaks, T. G., and S. K. Layson.
1983. Box-Cox estimation with standard econometric prob-
lems. Rev. Econ. Stat. 65:160-164.
Spitzer, J. J.
1982. A primer on Box-Cox estimation. Rev. Econ. Stat.
64:307-313.
1984. Variance estimates in models with the Box-Cox trans-
formation: implications for estimation and hypothesis test-
ing. Rev. Econ. Stat. 66:605-653.
Theil, H.
1971. Principles of econometrics. J. Wiley and Sons, N.Y.
Walters, C. J.
1975. Optimal harvest strategies for salmon in relation to en-
vironmental variability and uncertain production param-
661
KlSllKKV HULLETIN; VOL. 86. NO. 4
eters. J. Fish. Res. Board Can. 32:1777-1784. Zarembka, P.
White, K. J. 1974. Transformation of variables in econometrics. In P.
1987. Shazam: A general computer program for econometric Zarembka (editor), Frontiers in econometrics, p. 81-104.
methods (Version 5). Am. Stat. 41:80. Academic Press, N.Y.
APPENDIX
Because the estimation of the extended Box-Cox functional form is carried out by max-
imum likelihood procedures, the log-likelihood function for the extended Box-Cox func-
tional form and estimation methods are briefly presented in this appendix.
Under the assumptions that S and B are nonstochastic and the error term is normally
and independently distributed with zero mean and constant variance, aj, the log-likelihood
function of model (5) can be expressed following Spitzer (1982):
Lia, A, M, e, o\) = -T/2(ln 2n + In o^) - i2af)-He'e) + (A - 1)2 In A (12)
where T is the number of observations. To reduce the dimension of the estimation prob-
lem, the parameter Oj can be eliminated from Equation (12) to derive the concentrated
log-likelihood function as follows:
L{q, a, ^i, 6) = -r/2(ln 2n + In d^) + (A - 1)1 In A
where o'l = {llT)e e.
When heteroscedasticity is present, the concentrated log-likelihood function for model
(5) and the error term expressed in model (6) can be expressed as
L{a, A, M, Q) = -T/2(ln 2ti + In d^ + 1)
- I In (/?i + /J.S'^' + /?35'«') + (A - 1)2 In A (14)
where d'^ = (l/T)e'V~^e and F is a nxn matrix (n is the number of observations) in which
off-diagonal elements are zeros and diagonal elements are /?i -i- P-yS^^^ + P^B'^l
The maximum log-likelihood parameter estimates for (a, A, /u, 6, and ft) can be obtained
by nonlinear least squares methods or iterated ordinary (weighted) least squares procedures.
Seaks and Layson (1983) provide an example of the iterated ordinary (weighted) least
squares method using the Time Series Processor (TSP) computer package for estimating
Box-Cox flexible functional form with standard econometric problems; i.e., heteroscedas-
ticity and autocorrelation.
As Spitzer (1984) pointed out, the ordinary least squares method underestimates the
variance of the error term while the first derivative only gradient estimation methods (e.g.,
Marquardt) overestimate the variance. In order to compress the range of under- and
overestimation of the error variance, Spitzer suggested that the dependent variable be
divided by its geometric mean. This scaling process will then eliminate the last term in
the concentrated log likehhood function in Equations (12)-(14).
662
AGE, MORPHOLOGY, DEVELOPMENTAL BIOLOGY, AND
BIOCHEMICAL GENETIC VARIATION OF YUKON RIVER FALL
CHUM SALMON, ONCORHYNCHUS KETA, AND COMPARISONS WITH
BRITISH COLUMBIA POPULATIONS
Terry D. Beacham, Clyde B. Murray, and Ruth E. Withler^
ABSTRACT
Fall chum salmon, Oncorhynchtis keta, populations spawning in the Yukon River drainage were surveyed
for variation in age, size and shape at maturity, developmental biology, and biochemical genetics. Yukon
River fall chum salmon matured at older ages and smaller sizes than chum salmon in British Columbia.
They also had proportionately smaller heads, thinner caudal peduncles, and smaller fins than British
Columbia chum salmon, perhaps illustrating morphometric adaptation to long distance freshwater migra-
tion. Yukon River chum salmon were less fecund and had smaller eggs than those in British Columbia,
and they also tended to have faster development to alevin hatching and fry emergence than most British
Columbia populations. Maximum alevin and fry size of Yukon River salmon occur at lower water
temperatures during development than for most British Columbia populations, possibly indicating a
developmental adaptation to low winter water temperatures. Genetic differentiation among chum salmon
populations in the Yukon River drainage was observed.
The Yukon River is a major North American river,
originating in British Columbia and flowing over
3,200 km through the Yukon Territory and Alaska
to the Bering Sea, and draining an area of approx-
imately 860,000 km". Five species of Pacific salmon
(Oncorhynchus) occur in the Yukon River (Gilbert
1922), but chinook, 0. tshawytscha, and chum
salmon, 0. keta, are the most abundant and are ex-
ploited in commercial and subsistence fisheries
(McBride et al. 1983). Chum salmon in the Yukon
River are characterized by distinct seasonal races
(Gilbert 1922). The early-maturing or "summer"
chum salmon return to the Yukon River between
early June and mid-July and spawn in the lower 800
km of the drainage (Buklis 1981). Later-maturing
or "fall" chum salmon enter the Yukon River from
mid-July through late August and spawn in the
up-river portions of the drainage, migrating as far
as 2,800 km upstream (Milligan et al. 1986). Fall
chum salmon are also larger than summer chum
salmon (Buklis 1981; Buklis and Barton 1984), and
generally have higher fecundity and younger age
compositions than summer chum salmon (Sano
1966).
'Department of Fisheries and Oceans, Biological Sciences
Branch, Pacific Biological Station, Nanaimo, British Columbia,
Canada V9R 5K6.
Chum salmon generally spawn in rivers only a
short distance from salt water (<200 km), a trait
very different from the long distance freshwater
migrations of Yukon River fall chum salmon. Yukon
River fall chum salmon are noted for their high oil
content upon entering the river (Gilbert 1922), an
adaptation necessary to provide sufficient energy
reserves for the freshwater migration, as in Amur
River chum salmon in the Soviet Union (Nikol'skii
1961). Adaptations in other biological characters
may reflect the environmental conditions experi-
enced by Yukon River fall chum salmon. Thus we
examined the variation in life history traits of Yukon
River fall chum salmon, and compared this varia-
tion to that found in chum salmon in British
Columbia.
In 1984, we began a survey of variation in bio-
logical characters of fall chum salmon in the Yukon
River. Regional biochemical genetic variation had
previously been reported for chum salmon in British
Columbia (Beacham et al. 1987), and we investigated
the biochemical genetic differentiation of Yukon
River fall chum salmon. We had previously exam-
ined the adaptive nature of the variation in some
morphometric and life history traits of chum salmon
in British Columbia (Beacham and Murray 1987) and
used this for comparison for Yukon River fall chum
salmon.
Manuscript accepted July 1988.
FISHERY BULLETIN; VOL. 86, NO. 4, 1988.
663
FISllKKV Kl'LLETIN: VOL. Hti. NO. 4
METHODS AND MATERIALS
Age and Morphology
Chum salmon were collected with gill nets. After
capture in 1984, postorbital-hypural length (Vlady-
kov 1962), caudal peduncle depth, and length of the
base and height of the anal and dorsal fins were all
recorded on the left side in the field to the nearest
millimeter. The number of gill rakers on the left
anterior arch was recorded, as well as the number
of branchiostegal rays on the left side. Five scales
and two otoliths were collected from each individual
for age determination, and the sex of an individual
was confirmed by internal inspection. When the age
of an individual estimated from scales and from
otoliths disagreed, the age determined from scales
was assigned. In 1985, only postorbital-hypural
length and sex were recorded for individuals, and
only scales were collected for age determination.
Population differences and sexual dimorphism in
the meristic characters were examined by two-way
analysis of variance, with population and sex as the
indices. Population is used in the manner as de-
scribed by Ricker (1972), as a group of fish spawn-
ing in a particular river at a particular season, and
there is no substantial interbreeding with another
group spawning in a different river. Morphometric
measurements of both males and females were stan-
dardized to a postorbital-hypural length of 520 mm
by the method outlined by Beacham and Murray
(1983):
M, = M„
L..
L =
L =
where M, = size of standardized morphometric
character,
M„ = observed character size,
length that characters are standard-
ized to (520 mm),
observed postorbital-hypural length,
and b is the regression coefficient of
log^ M„ on logp Lj, (stocks and sexes
separate).
Two-way analysis of variance was again used to ex-
amine population differences and sexual dimorphism
in the morphometric characters.
Developmental Biology
We estimated fecundity of Yukon River chum
664
salmon by collecting and freezing both ovaries from
14 females in the Dawson City commercial fishery
in 1985 (we also collected length and scales); we
subsequently thawed both ovaries and made tot;vl egg
counts for each female. The methodology of the
survey of developmental biology was similar to that
outlined by Beacham and Murray (1987). Gametes
were collected from five male and five female Kluane
River chum salmon on 17 October 1985. The gametes
were then shipped to the laboratory on ice, the eggs
fertilized at 8°C, and then subsequently reared in
controlled water temperatures of 4°, 8°, and 12°C
in vertical stack incubators. Five full-sib families
were obtained from the crosses, with each family
replicated in each incubator. Water temperatures
were recorded daily, and mean temperatures in the
incubators during the study were 4.1° (SD = 0.29),
8.0° (SD = 0.44), and 12°C (SD = 0.40), respectively.
Egg diameter (millimeters) and weight (milli-
grams) for each female were determined from 30
water-hardened eggs preserved for at least 3 months
in 10% formalin. During incubation, dead eggs were
removed from each family, stored in Stockard's solu-
tion, and later inspected to remove unfertilized eggs.
Egg survival rates were then calculated based upon
the number of eggs initially fertilized. Once hatch-
ing began in each family, we recorded the number
of newly hatched alevins daily, and within 1 day of
50% hatching we anesthetized and preserved 30 per
family in 10% formalin for subsequent determina-
tion of alevin length and weight. Fork length was
recorded to the nearest 0.1 mm, total weight re-
corded to the nearest milligram, the yolk separated
from the rest of the body and weighed (milligrams),
and then tissue weight (milligrams) determined by
subtraction. Dead alevins were also removed and
counted in order to determine alevin survival rates.
The timing of fry emergence (swim-up) for each
family was determined by placing the alevins in an
emergence trap modified from Mason (1976), where
the alevins were classified as newly emergent fry
only when they became neutrally buoyant and
positively phototactic When 50% of the fry from a
particular family had emerged, all of the family re-
maining in the incubator was anesthetized and then
preserved in 10% formalin, and 30 fry were random-
ly drawn from the preserved samples and fry length
and weight determined as for alevins.
Variation in egg size was analyzed with a one-way
(female was the index) analysis of variance. Varia-
tion in survival rates was analyzed by first determin-
ing survival rates for each group as proportions and
then transforming them to radians with the arcsine
BEACHAM ET AL.: YLKON RIVER CHUM SALMON POPULATION
square root transformation to normalize the data.
We then used an analysis of variance model:
Y,jki = ^^ + T, + Fj + TF,j + R^ji, + e.^ki
where i^,jA-/ = transformed survival rate,
yi = overall mean,
Tj = fixed effect of temperature {i =
1-3),
Fj = random effect of family (j = 1-5),
TFij = random interaction between tem-
perature and family,
R^Jl^ = random effect of replicate {k =
1-2), and
e-j./ = error term for l\h observation in
subgroup ijk.
Variation in alevin and fry size characters was ana-
lyzed with the same model. Satterthwaite's (1946)
approximation was necessary to calculate an appro-
priate mean square to test the effect of family.
Biochemical Genetics
Biochemical genetic sampling began on Canadian
populations in 1984 and was expanded in 1985 to
include United States populations (Fig. 1). Method-
ology of sample collection has been outlined by
Beacham et al. (1985). Summarized briefly, heart,
liver, and muscle samples were collected from adult
chum salmon, frozen, and stored at - 20°C for later
electrophoretic analysis by a consultant. We verified
the scoring of all gels from a complete photographic
record. Horizontal starch gel electrophoresis, de-
scribed by Utter et al. (1974), was used to detect
protein variation. The loci and buffer systems used
n —
150°
180°
170°
160°
140=
130°
70°
BEAUFORT
SEA
0 100 500
kilometres
-65°
-60°
NORTON
SOUND
J_
Figure 1.— Locations in Yukon River drainage where chum sahiion were sampled during 1984-1986. Listed in ascending
order are (l)Teshn River. (2) Kluane River, (3) Koidern River, (4) Yukon River at Minto, (5) Dawson City, (6) Fishing
Branch River, (7) Porcupine River at Old Crow, (8) Sheenjek River, (9) Chandalar River, (10) Delta River, (ll)Toklat
River, (12) Yukon River at Emmonak.
665
FISHERY BULLETIN: VOL. 86, NO. 4
are outlined in Beacliam et al. (1985). Lgg was re-
named as Tapep and was scored as a four-allele locus
(L. Seeb^) (although only three were present in the
stocks we surveyed) on both a Tris-boric acid-EDTA
buffer described by Markert and Faulhaber (1965)
and an amine citrate buffer described by Clayton
and Tretiak (1972). Me was renamed as MdhP, 6-Pg
as Pgdh, Pmi as Mpi, and Agp as GSpdh.
We determined allelic frequencies for each locus
by summing the numbers of each allele and dividing
by the total number of alleles counted. Genotypic
frequencies at each polymorphic locus in each popu-
lation were tested for departures from Hardy-
Weinberg equilibrium by chi-square. We used the
log-likelihood ratio statistic (G-test) (Sokal and RohLf
1969) to test equality of allelic frequencies between
countries, among populations within countries, and
between years for samples taken from the same
population. An approximate i^-ratio (G-statistic
summed over all loci/degrees of freedom) was used
to test the relative magnitude of the sources of varia-
tion. We calculated genetic distance among popula-
tions using Nei's (1978) statistic and the 7 loci in-
dicated in Table 7 and constructed a denogram from
the matrix of the distances using the unweighted
pair group mean method of Sneath and Sokal (1973).
RESULTS
Age and Morphology
Age of Maturity
The dominant age of maturity for fall chum
salmon returning to Canadian rivers during 1984
and 1985 was four years, with age 3 and age 5 chum
salmon each comprising less than 15% of the total
return (Table 1). The proportion of chum salmon
returning at three or five years of age varied an-
nually within a population, possibly reflecting dif-
ferent production from the respective brood years.
Yukon River chum salmon matured at a significantly
older age than chum salmon in British Columbia
(Xz^ = 193.9, P < 0.01), but four years was the
dominant age of maturity in both areas. Of 227 chum
salmon sampled in 1984 from which age could be
determined from both otoliths and scales, the same
age was recorded in 184 (83%) of the cases. For the
43 other fish, the age estimated from scales was one
year older than that estimated from otoliths in 22
cases, whereas in the remaining 21 fish, the age
estimated from otoliths was either one year (19
cases) or two years (2 cases) older than that esti-
mated from scales. In 1984, age could not be deter-
mined from 16% (48 fish) of the scales collected and
9% (26 fish) of the otoliths collected. In 1985, age
could not be determined from 18% (124 fish) of the
scales examined due to scale resorption. Age of
maturity was more likely to be determined from
otoliths than from scales.
Table 1 . — Percentage of chum salmon returning at ages 3-5 years
and mean age of return for Yukon River populations sampled on
thie spawning grounds during 1984-85. The mean for British
Columbia chum salmon was derived from Beacham and Murray
(1987).
Year
N
Age (yr)
Mean
Population
3
4
5
age
(yr)
Yukon River
Kluane
1984
100
16.0
78.0
6.0
3.90
1985
96
29.2
70.8
0.0
3.71
Total
196
22.4
74.5
3.1
3.81
Minto
1984
100
13.0
69.0
18.0
4.05
1985
99
25.3
72.7
2.0
3.77
Total
199
19.1
70.9
10.0
3.91
Fishing Branch
1984
100
14.0
82.0
4.0
3.90
1985
77
1.3
72.7
26.0
4.25
Total
177
8.5
78.0
13.5
4.05
Porcupine
1985
67
1.5
85.1
13.4
4.12
Teslin
1985
64
0.0
56.3
43.7
4.44
Koidern
1985
81
0.0
74.1
25.9
4.26
Total
Mean
784
12.5
73.7
13.8
4.01
British Columbia
Total
Mean
1 1 ,749
32.4
62.1
5.5
3.73
^University of Idaho, Moscow, ID 83843, pars, commun.
December 1985.
Meristics
No sexual dimorphism was observed in the num-
ber of gill rakers or branchiostegal rays (P > 0.05).
No population differences in gill raker number were
observed (P > 0.05), but significant differences
among populations were observed in branchiostegal
ray number (^2,297 = 4.36, P < 0.05). The popula-
tion with the greatest number of gill rakers (Fishing
Branch) also had the greatest number of branchi-
ostegal rays (Table 2). Yukon River chum salmon
had more gill rakers and branchiostegal rays than
the average British Columbia chum salmon.
Morphometry
Yukon River male chum salmon were longer than
666
BEACHAM ET AL.: VIKON FilVKR CHUM SALMON POPULATION
females at all ages examined (P < 0.05) (Table 3).
Although there could be significant variation in
mean length-at-age among populations (P < 0.05),
Yukon River chum salmon were substantially
smaller than chum salmon of the same age spawn-
ing in British Columbia {P < 0.05). Size differences
between Yukon River and British Columbia chum
salmon increase with age, with age 3 Yukon River
males 94% of the length of age 3 British Columbia
males, but age 5 Yukon River males are 86% of the
length of their British Columbia counterparts. Sim-
ilar results were also observed for females.
We examined whether there was any differen-
tiation of selected morphometric characters with
respect to sexual dimorphism or distance of fresh-
water migration. With the morphometric measure-
ments of both males and females standardized to a
postorbital-hypural length of 520 mm, males had
longer postorbital head lengths, thicker caudal
peduncles, longer base length of the dorsal fin, and
longer dorsal and anal fins (all P < 0.05) (Table 4).
No sexual dimorphism in base length of the anal fin
was observed. Significant population differences in
the relative sizes of the morphometric characters
were also observed, with the Kluane River popula-
tion having the relatively smallest characters, and
the P'ishing Branch River population having the pro-
portionately largest characters.
Substantial differences exist in length of fresh-
water migration between Yukon River fall chum
salmon and chum salmon in British Columbia. The
relative sizes of the morphometric characters ex-
amined are all smaller for Yukon River chum salmon
than for chum salmon in British Columbia (Table 4),
perhaps illustrating morphometric adaptation of
Yukon River chum salmon to the long freshwater
migration.
Developmental Biology
Fecundity and Egg Size
The fecundity-length relationship for 14 female
chum salmon is described by
Table 2.— Mean number of gill rakers on left anterior gill arch
and mean number of branchiostegal rays on left side for Yukon
River fall chum salmon sampled during 1984. Standard error
of mean is in parentheses. One hundred fish were sampled
per population. Data for 9,206 British Columbia chum salmon
are from Beacham and Murray (1987).
F = 0.100 L
L60
Population
Gill rakers
Branchiostegal rays
Yukon River
Kluane
22.88 (0.10)
13.72 (0.07)
Minto
22.90 (0.09)
13.58 (0.08)
Fishing Branch
23 13 (0.09)
13.89 (0.07)
Mean
22.97 (0.05)
13.73 (0.04)
British Columbia
Mean
22.62 (0.01)
13.42 (0.01)
where F = number of eggs and L is postorbital-
hypural length (mm). Mean fecundity of age 4
females was 2,271 eggs (SD = 208, n = 7), and that
of age 5 females was 2,451 eggs (SD = 106, n = 3).
Mean egg weight of the five Kluane River females
used in the study of developmental biology ranged
from 145 mg (SD = 6 mg) to 210 mg (SD = 5 mg).
Mean diameters ranged from 6.66 mm (SD = 0.21
mm) to 7.25 mm (SD = 0.18 mm). Mean egg size
of Yukon River chum salmon is substantially smaller
than the mean weight of 290 mg and mean diam-
eter of 8.39 mm for egg size of chum salmon in
Table 3. — Mean postorbital-hypural length-at-age (mm) for chum salmon sampled on the spawning grounds during
1984-85. Sample sizes are given in parentheses. The mean length-at-age for British Columbia chum salmon was
derived from Beacham and Murray (1987).
Male
! age
Female age
Population
3
4
5
3
4
5
Yukon River
Kluane
509
(19)
516
(59)
556
(4)
490
(25)
501
(87)
526
(2)
Minto
525
(13)
528
(46)
541
(8)
509
(25)
516
(95)
525
(12)
Fishing Branch
477
(5)
524
(64)
548
(5)
487
(10)
504
(74)
519
(19)
Porcupine
541
(41)
546
(4)
462
(1)
514
(16)
531
(5)
Teslln
521
(17)
521
(14)
492
(19)
506
(14)
Koidern
510
(29)
514
(10)
468
(31)
503
(11)
Mean
510
(37)
524
(256)
531
(45)
497
(61)
505
(322)
515
(63)
British Columbia
Mean
543
(1974)
595
(3186)
621
(344)
532
(1357)
579
(2879)
605
(255)
667
FISHERY BILLETIN: VOL. 88. NO. 4
Table 4.— Mean measurements (mm) for six morphometric characters for cfium salmon
sampled during 1984. Measurements for each individual were standardized to a postorbital-
hypural length of 520 mm. Standard deviations are indicated in parentheses The size
(%) of each standardized character relative to body length is indicated for both Yukon
and British Columbia chum salmon. Data for 4,862 male and 4,044 female British Colum-
bia chum salmon were standardized to 600 mm postorbital-hypural length and were derived
from Beacham and Murray (1987).
Fishing
% of length
Kluane
Minto
Branch
Yukon
B.C.
Males
Head length
79.4 (4.3)
80.7 (3.3)
81.3 (2.9)
15.48
16.02
Caudal peduncle
39.9 (2.8)
41.5 (2.3)
41.2 (2.5)
7.85
8.37
Anal fin base
67.4 (4.4)
70.0 (3.5)
69.3 (4.4)
13.25
13.68
Anal fin height
57.8 (3.9)
59.5 (3.1)
58.7 (4.5)
11.28
11.65
Dorsal fin base
62.9 (5.6)
66.1 (4.0)
63.9 (4.6)
12.34
13.01
Dorsal fin height
88.1 (6.2)
90.2 (5.4)
92.0 (5.6)
17.33
17.64
Sample size
37
32
39
Females
Head length
73.5 (4.3)
74.7 (4.0)
76.3 (3.5)
14.38
14.64
Caudal peduncle
37.7 (3.1)
38.9 (3.3)
39.5 (2.2)
7.44
8.06
Anal fin base
67.8 (5.2)
68.0 (4.8)
68.8 (4.8)
13.12
13.79
Anal fin height
56.0 (4.0)
57.6 (4.8)
58.0 (4.5)
11.00
11.65
Dorsal fin base
58.2 (4.1)
58.4 (4.4)
59.1 (4.4)
11.26
11.79
Dorsal fin height
79.8 (6.5)
82.7 (5.3)
84.4 (5.4)
15.83
16.24
Sample size
63
68
61
British Columbia (Beacham and Murray 1987).
Survival Rates
Embryo survival rates were lowest at a 4°C incu-
bation temperature and highest at 12 °C, with sig-
nificant differences observed among incubation
temperatures {P < 0.05) (Table 5). Significant dif-
ferences were also observed in embryo survival rates
among families (P < 0.0), but not between replicates
within families (P > 0.10). An interaction between
family and incubation temperature occurred (P <
0.01), illustrating that trends in embryo survival
with respect to incubation temperature were not the
same for all families.
Alevin survival rates were uniformly high (Table
5), with no significant difference observed among
temperatures, among families, or between replicates
within families. The interaction between family and
incubation temperature also occurred for alevin sur-
vival rates (P < 0.05).
Table 5. — Survival rates of embryos and alevins for Kluane River
chum salmon maintained at constant water temperatures of 4°,
8°, and 12°C. Time of 50% hatching (days) of the alevins and 50%
emergence (days) of the fry is also indicated. N is number of fer-
tilized eggs (for embryo survival rates) or alevins hatched (for alevin
survival rates). Standard deviations of 50% hatching and emer-
gence times are in parentheses.
Time to
Temperature
Survival
50% hatching
(°C)
N
rate
or emergence
Embryos
4.1 (0.28)
2,172
0.767
113.9 (2.3)
8.0 (0.21)
2,387
0.906
65.7 (0.6)
12.0 (0.35)
2,432
0.955
43.7 (0.9)
Alevins
4.1 (0.29
1,666
0.975
182.2 (1.9)
8.0 (0.20)
2,162
0.998
111.6 (1.2)
12.0 (0.35)
2,323
0.991
79.6 (1.1)
December or early January (Beacham and Murray
1986, 1987), nearly three months later than the
spawning time of the Kluane River population.
Hatching and Emergence Time
Warmer water temperatures during incubation
enhanced the development rate of Kluane River
chum salmon (Table 5). Timings of both alevin hatch-
ing and fry emergence at a specific incubation tem-
perature were comparable with that of a British
Columbia chum salmon population spawning in late
Alevin and Fry Size
Alevin and fry length and weight were influenced
by the water temperature during development. The
longest and heaviest alevins and fry were observed
at 4°C (Table 6). Significant differences in alevin
length and weight were observed among incubation
temperatures, among families, and between reph-
668
BEACHAM ET AL : VIKON RIVER CHUM SALMON POPILATION"
cates within families (all P < 0.05). Significant inter-
actions between temperature and family were also
observed for all alevin size characters (all P < 0.05),
again illustrating that alevin size characters did not
respond consistently among families to changes in
incubation temperature.
The effect of different incubation temperatures
on fry size characters was similar to that for the
alevin size characters. Significant differences in fry
length, total weight, and tissue weight were ob-
served among temperatures, among families, and
between replicates. The amount of yolk remaining
at the time of fry emergence was similar at all in-
cubation temperatures (P > 0.05) (Table 6). In-
teractions between family and incubation temper-
ature were present for all fry size characters (all P
< 0.05).
Biochemical Genetics
Regional differentiation of allelic frequencies for
chum salmon stocks in Alaska and the Yukon Terri-
tory was observed. Chum salmon from Alaska
generally had a lower frequency of Idh-1"^'^ and a
higher frequency of Tapep (Lgg)'^'^ than those from
the Yukon Territory (Table 7). Regional differen-
tiation in allelic frequencies was also observed for
the other loci examined. The Delta and Toklat River
populations, both tributaries of the Tanana River,
had higher frequencies of Idh-3'*"' than all other
populations except the Teslin River population. Only
one genotypic frequency was not in Hardy- Weinberg
equilibrium, that being Pgdh in the Toklat River
population. The disequilibrium was due to a hetero-
zygote deficiency.
Variation in allelic frequencies among populations
Table 6.— Fork length, total weight, yolk weight, and tissue weight
for Kluane River chum salmon alevins and fry maintained at con-
stant 4°, 8°, and 12°C water temperatures. N is the number of
alevins and fry measured and weighed for all families combined.
Standard deviations are in parentheses.
Temper-
ature
(°C)
N
Fork
length
(mm)
Total
weight
(mg)
Yolk
weight
(mg)
Tissue
weight
(mg)
Alevin
4
8
12
Fry
4
8
12
150
150
149
150
150
150
21.1 (1.4)
20.8 (0.7)
20.4 (0.9)
32.3 (1.8)
31.0 (0.9)
30.5 (1.1)
222.2 (38.2)
108.9 (18.9)
176.6 (20.1)
308.4 (52.9)
277.5 (27.4)
277.6 (28.1)
136.9(19.1) 85.3(22.4)
106.7(12.7) 74.2 (8.0)
109.8 (12.9) 66.8 (9.4)
30.0 (8.7) 278.4 (51.3)
28.0 (7.2) 249.5 (23.5)
30.7 (6.4) 247.5 (27.2)
within a region was greater than annual variation
in allelic frequencies within a population (Fgy 43 =
4.00, P < 0.01) (Table 8). Annual stability of allelic
frequencies was examined for the loci for which two
years of data was available in each population. Of
39 comparisons made, 3 (8%) were significant (Table
8).
Heterozygosity of populations from Alaska was
generally higher than that from the Yukon Terri-
tory. Mean heterozygosities were 0.156 (SD =
0.009) and 0.138 (SD = 0.019), respectively, for the
loci outlined in Table 7. Heterozygosity values are
dependent upon the number of polymorphic and
monomorphic loci included in the calculations, and
thus will vary among studies.
The analysis of genetic distance for all pairwise
combinations of the 10 populations sampled ranged
from 0.0000 to 0.0072. A dendogram based upon the
unweighted pair group mean analysis (UPGMA)
illustrated two main clusters of populations (Fig. 2).
Figure 2.— Dendogram produced from cluster analysis using Nei's
(1978) genetic distance value. Codes are C = Canadian and US
= United States.
KLUANE C
KOIDERN C
SHEENJEK US
MINTO C
PORCUPINE C
CHANDALAR US
L FISHING BRANCH
TESLIN C
TOKLAT US
DELTA US
1 1 1 1 1 1 1 1 1
0.0032 0.0028 0.0024 0.0020 O.OO16 0.0012 0.0008 0.0004 0.0000
GENETIC DISTANCE
669
FISHERY Bl'l.LETIN: VOL. 8(;, NO. 4
Table 7.— Observed allelic frequencies at polymorphic loci for 10 Yukon River drainage
and other alleles were assigned numbers
Year
ldh-1
ldh-3
MdhP
Stock
N
100
N
100
40
85
N
100
Kluane
1984
107
1.000
106
0.358
0.585
0.057
107
0977
1985
106
1.000
106
0 396
0.552
0.052
106
0991
Pooled
213
1.000
212
0.377
0.568
0.054
213
0.984
Minto
1984
130
1.000
130
0.481
0.508
0.012
130
0.904
1985
114
1.000
114
0.482
0.461
0.057
114
0 899
Pooled
244
1.000
244
0.482
0.486
0.033
244
0.902
Fishing Branch
1984
126
1.000
126
0.460
0.512
0.028
126
0.944
1985
99
1.000
98
0.444
0.510
0.046
99
0.914
Pooled
225
1.000
224
0.453
0.511
0.036
225
0.931
Teslin
1985
90
1.000
90
0.611
0.361
0.028
90
0.889
Koidern
1985
100
1.000
100
0.405
0.570
0.025
100
0.980
Porcupine
1985
81
1.000
81
0.463
0 506
0.031
81
0 938
Toklat
1985
120
0.946
120
0.579
0.350
0.071
120
0.887
1986
124
0.952
124
0.512
0.399
0.089
124
0 911
Pooled
244
0.949
244
0.545
0.375
0.080
244
0.900
Delta
1985
146
0.990
145
0.538
0.417
0.045
146
0.877
1986
147
0.990
146
0.490
0.449
0.062
147
0.884
Pooled
293
0.990
291
0.514
0.433
0.053
293
0.881
Sheenjek
1985
143
0.993
147
0.398
0.585
0.017
144
0.962
1986
150
0.987
124
0.415
0.528
0.056
150
0.907
Pooled
293
0.990
271
0.406
0.559
0.035
294
0.934
Chandalar
1986
147
1.000
143
0.448
0.497
0.056
147
0.939
Table 8.— Analysis of heterogeneity of allelic frequencies between countries, among
sampled during
ldh-1
dh-3
MdhP
df
Mpi
Source of variation
df
G
df
G
df
G
G
Between countries
1
45.4**
2
12.6**
1
9.6**
1
4.5*
Among populations within
8
35.7**
16
77.4**
8
59.2**
8
13.9
countries
Canada
5
0.0
10
34.7**
5
45.2**
5
10.0
United States
3
35.7**
6
42.7**
3
14.0
3
3.9
Between years within
6
0.3
12
21.5
6
8.7
6
4.5
populations
Kluane
0.0
2
0.6
0.6
1.3
Minto
0.0
2
8.7**
0.0
0.1
Fishing Branch
0.0
2
1.1
1.1
1.7
Toklat
0.0
2
2.3
0.5
1.2
Delta
0.2
2
1.8
0.0
0.2
Sheenjek
0.1
2
7.0
6.5"
0.0
The Toklat and Delta River populations were dis-
tinctive from other populations surveyed. The popu-
lation from Chandalar River, a tributary of the Por-
cupine River, was similar to the Porcupine River
population (sampled at the Old Crow fishery) and
to the Fishing Branch River population, also a trib-
utary of the Porcupine River. The populations from
Kluane and Koidern Rivers, tributaries of the White
River, were also similar to each other, as well as to
the Sheenjek River population in Alaska.
DISCUSSION
Yukon River fall chum salmon undertake the long-
est freshwater spawning migration of chum salmon
in North America and spawn in locations where
winter environmental conditions are very severe.
They are also relatively abundant, with an average
of almost 450,000 chum salmon harvested annually
during 1974-83 (Buklis and Barton 1984). Their
abundance indicates that they have adapted success-
670
HKACHAM KT Al„: VrKON KIVKK CIirM SALMON POPULATION
fall chum salmon populations during 1984-86. The most common allele at a locus was designated 100,
according to the mobility relative to that of the 100 allele.
Year
Mpi
Pgdh
N 100
G3pdh-2
N 100
Tapep
Stock
N
100
N
100
50
75
Kluane
1984
105
0.910
107
0.991
106
0.887
1985
106
0.943
105
0.990
104
0856
106
0.901
0.099
0.000
Pooled
211
0.927
212
0.991
210
0.871
106
0.901
0.099
0.000
Minto
1984
130
0.923
130
0.962
130
0.888
1985
114
0.934
114
0.956
114
0.842
114
0.842
0.158
0.000
Pooled
244
0.928
244
0.959
244
0.867
114
0.842
0.158
0.000
Fishing Branch
1984
126
0.893
126
0.960
113
0.854
1985
97
0.933
99
0.995
96
0.880
99
0.803
0.197
0.000
Pooled
223
0.910
225
0.976
209
0.866
99
0.803
0.197
0.000
Teslin
1985
90
0.972
90
0.994
89
0.983
90
0.861
0.139
0.000
Koidern
1985
100
0.945
100
1.000
99
0.879
97
0.845
0,155
0.000
Porcupine
1985
79
0.943
81
0.981
75
0.847
80
0.806
0.194
0.000
Toklat
1985
120
0.917
120
0.971
112
0.848
120
0.808
0.158
0.033
1986
124
0.883
123
0.976
111
0.833
122
0.852
0.119
0.029
Pooled
244
0.900
243
0.973
223
0.841
242
0.831
0.138
0.031
Delta
1985
141
0.922
146
0.949
134
0.888
146
0.818
0.147
0.034
1986
147
0.935
147
0.939
104
0.880
142
0.852
0.106
0.042
Pooled
288
0.929
293
0.944
238
0.884
288
0.835
0.127
0.038
Sheenjek
1985
150
0.913
144
0.972
129
0.876
150
0.847
0.147
0.007
1986
150
0.913
148
0.973
120
0.879
149
0.822
0.158
0.020
Pooled
300
0.913
292
0.973
249
0.878
299
0.834
0.152
0.013
Chandalar
1986
145
0.897
147
0.973
85
0.871
145
0.834
0.148
0.017
populations within countries, and between years within populations for Yukon River fall chum salmon
1984-86. P < 0.01; P < 0.05.
Pgdh
G3pdh-2
df G
Tapep
df G
Total
Standardized
Source of variation
df
G
df
G
F-statistIc
Between countries
1
7.9*
1
0.7
2
48.3"
9
129.0"
14.33
Among populations within
8
32.0--
8
34.6"
11
20.6*
67
273.4**
4.08
countries
Canada
5
22.6*
5
30.3*
5
10.3
40
153.1**
3.83
United States
3
9.4
3
4.3
6
10.3
27
120.3**
4.46
Between years within
6
5.1
6
3.0
6
5.8
48
48.9
1.02
populations
Kluane
0.2
0.6
Minto
0.0
1.9
Fishing Branch
4.8"
0.4
Toklat
0.0
0.1
2
1.7
Delta
0.1
0.0
2
1.8
Sheenjek
0.0
0.0
2
2.3
fully to a long distance freshwater migration and
extreme winter conditions.
Several biological characters differ between chum
salmon in the Yukon River and in British Colum-
bia. For Yukon River fall chum salmon, approx-
imately equal numbers of salmon mature, on aver-
age, at three and five years of age, although there
is annual variation (Buklis and Barton 1984). In the
Amur River, fall chum salmon mainly mature at four
and five years of age (Smirnov 1975). Similar results
were recorded in our study. Yukon River chum
salmon matured at older ages and at smaller mean
lengths-at-age than did chum salmon in British Co-
lumbia. A general trend of younger ages at matur-
ity and increased mean lengths-at-age in southern
as compared with northern chum salmon popula-
tions has been reported by Salo (in press). This
trend may be a result of the earlier timing of fry
emergence and later timing of adult spawning of
more southern populations, allowing more time for
671
FISHERY BULLETIN: VOL. 86, NO. 4
ocean growth, particularly in the year of maturity.
Body shape of Yukon River and British Colum-
bia chum salmon was different, with Yukon River
chum salmon having a shorter head, thinner caudal
peduncle, and smaller fins than British Columbia
chum salmon. This more fusiform body shape is pre-
sumably an adaptation to the long migration in fresh
water, as selection should result in a body shape that
would minimize energy consumption during migra-
tion. Morphological differentiation with respect to
distance of upstream migration has been reported
to occur in other Oncorhynchus species (Eniutina
1954; Taylor and McPhail 1985), as well as with
respect to river size (Hjort and Schreck 1982;
Beacham and Murray 1987).
The mean fecundity of 2,325 eggs per female for
Yukon River fall chum salmon reported in our study
is similar to other results. Elson (1975) reported
mean fecundities of 2,360 eggs and 2,513 eggs per
female for Porcupine River chum salmon sampled
in 1971 and 1973, respectively. Raymond (1981)
reported mean fecundities of Tanana River chum
salmon of 2,355 eggs in 1977 and 2,762 eggs in 1978.
Fecundities of Yukon River fall chum salmon are
less than those reported for many chum salmon
stocks in British Columbia (Beacham 1982), and also
less than fall chum salmon in the Amur River (3,200
to 4,300 eggs) (Smirnov 1975). Mean egg size of
Yukon River fall chum salmon is also less than that
of Amur River chum salmon (180 to 300 mg, 6.7 to
9.0 mm diameter) (Smirnov 1975).
The different fecundities and ages at maturity of
Yukon River and British Columbia chum salmon
present an interesting contrast in life history char-
acters. Yukon River fall chum salmon mature at an
average age of 0.28 years older than British Colum-
bia salmon, which means that they incur an addi-
tional 5% mortality if the instantaneous mortality
rate during the last year of life for chum salmon is
0.013 per month (Ricker 1976). The lower fecundity
and older age at maturity of Yukon River salmon
indicate that they are not as productive as chum
salmon in British Columbia or that mean survival
rates of the two groups are not equivalent. Egg-to-
fry survival rates for Yukon River chum salmon
have been reported as a mean of about 2.5% (Buklis
and Barton 1984), whereas those for British Colum-
bia chum salmon average about 10% (Bakkala 1970;
Beacham and Starr 1982). If Yukon River chum
salmon are as productive as those in British Colum-
bia, then ocean survival rates of Yukon River chum
salmon must be higher than those of British Colum-
bia chum salmon.
When incubated under the same water tempera-
tures, Yukon River chum salmon alevins hatch and
the fry emerge sooner than most chum salmon
populations in British Columbia (Beacham and Mur-
ray 1987). The faster development rates presumably
occur as a response to lower water temperatures
during the winter in the Yukon River tributaries
than in rivers in British Columbia. Yukon River
chum salmon alevins and fry are shorter and lighter
than those from British Columbia (Beacham and
Murray 1987), presumably reflective of smaller ini-
tial egg size of Yukon River chum salmon. At incu-
bation temperatures of 4°, 8°, and 12 °C, maximum
alevin and fry size for Yukon River chum salmon
was observed at 4°C, but for British Columbia
stocks, maximum alevin and fry size was generally
observed at 8°C. These results suggest that Yukon
River chum salmon are better adapted for develop-
ment under low water temperatures than are British
Columbia chum salmon.
Yukon River chum salmon are generally distinc-
tive in electrophoretic characteristics from chum
salmon in Cook Inlet in Alaska (Okazaki 1981) and
British Columbia (Okazaki 1981; Beacham et al.
1985, 1987). For example, the allelic frequency of
Idh-3^^ is 0.28 in Queen Charlotte Islands popula-
tions and 0.17 in populations in northern British
Columbia (Beacham et al. 1987), but this allele was
not detected in our study of Yukon River chum
salmon. Heterozygosity of Yukon River chum
salmon was lower than that observed for British
Columbia salmon (the same loci were included in the
analysis) (Beacham et al. 1987). Kijima and Fujio
(1984) reported that average heterozygosity is
related to effective population size in Japanese chum
salmon populations, with more abundant populations
having increased genetic variance. Abundance of the
Yukon River populations examined in our study is
unknown, but the catch data suggest that the in-
dividual Yukon River populations may not be as
abundant as major chum salmon populations in
British Columbia.
Allelic frequencies for most salmon populations
are reported to show little annual variation (Grant
et al. 1980; Utter et al. 1980; Beacham et al. 1985,
1987), allowing for pooling of samples from a par-
ticular population over several years. It should thus
not be necessary to conduct annual sampling in
order to characterize the populations contributing
to fisheries. Stock identification based on stable
traits, such as allelic frequencies, reduces annual
sampling costs for the baseline stocks. This differs
from scale analysis, in which variation in the char-
672
HKACIIAM ET AL : VIKON RIVf:R CHl'M SALMON POPULATION
acters used for stock identification makes annual
sampling of the baseline stocks necessary (e.g.,
Wilcock and McBride 1983; Wilcock 1984), and
restricts the incorporation of results into manage-
ment decisions during the fishing season.
ACKNOWLEDGMENTS
The extensive field sampling necessary for this
program was accomplished with the help of many
individuals, including Art Demsky, Peter Etherton,
Robbin Hunka, Tom Hurds, and especially Wally
Barner. Special thanks are due to Bill Arvey, Larry
Buklis, and other members of the Alaska Depart-
ment of Fish and Game for sampling of Alaskan
populations. Age determination of chum salmon
from scales was conducted by Yvonne Yole. Main-
tenance and supervision of the incubation laboratory
was provided by Wally Barner, Bill Andrews, and
Bruce Patten. July Dawes, Brenda Hoitsma, and
Tracey Briggs weighed and measured the chum
salmon eggs, ale\ins, and fry examined in this study.
The electrophoretic analysis was conducted by Helix
Biotech Ltd. of Richmond, B.C., under contract to
the Department of Fisheries and Oceans.
LITERATURE CITED
Bakkala, R. G.
1970. Synopsis of biological data on the chum salmon, On-
corhynchus keta (Walbaum) 1972. U.S. Fish Wildl. Sen-.
Circ. 315, 89 p.
Beacham, T. D.
1982. Fecundity of coho salmon (Oncorhynchus kisutch) and
chum salmon (0. keta) in the northeast Pacific Ocean. Can.
J. Zool. 60:1463-1469.
Beacham, T. D., A. P. Gould, R. E. Withler, C. B. Murray,
AND L. W. Barner.
1987. Biochemical genetic survey and stock identification of
chum salmon {Oncorhynchus keta) in British Columbia.
Can. J. Fish. Aquat. Sci. 44:1702-1713.
Beacham, T. D., and C. B. Murray.
1983. Sexual dimorphism in the adipose fin of Pacific salmon
(pncorhynckus). Can. J. Fish. Aquat. Sci. 40:2019-2024.
1986. Comparative developmental biologj' of chum salmon
(Oncorhynchus keta) from the Fraser River, British Colum-
bia. Can. J. Fish. Aquat. Sci. 43:252-262.
1987. Adaptive variation in body size, age, morphology, egg
size, and developmental biology of chum salmon (Oncorhyn-
chus keta) in British Columbia. Can. J. Fish. Aquat. Sci.
44:244-261.
Beacham, T. D., and P. Starr.
1982. Population biology of chum salmon, Oncorhynchus keta,
from the Fraser River, British Columbia. Fish. Bull., U.S.
80:813-825.
Beacham, T. D., R. E. Withler, and A. P. Gould.
1985. Biochemical genetic stock identification of chum salmon
(Oncorhynchus keta) in southern British Columbia. Can. J.
Fish. Aquat. Sci. 42:437-448.
BUKLIS, L. S.
1981. Yukon and Tanana River fall chum salmon tagging
studies, 1976-1980. Alaska Dep. Fish. Game Info. Leafl.
194, 40 p.
BuKLis, L. S., and L. H. Barton.
1984. Yukon River fall chum salmon biology and stock status.
Alaska Dep. Fish. Game Info. Leafl. 239, 67 p.
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.
Elson, M. S.
1975. Enumeration of spawning chum salmon (Oncorhynchus
keta) in the Fishing Branch River in 1971, 1972, 1973, and
1974. In L. W. Steigenberger, M. S. Elson, and R. T.
Delury (editors), Northern Yukon fisheries studies, 1971-
1974. Vol. 1, Ch. II, 45 p. North. Oper. Branch, Fish. Mar.
Serv., Dep. Environ., PAC/T-75-19. New Westminster,
British Columbia, Canada.
Eniutina, R. I.
1954. Local stocks of pink salmon in the Amur basin and
neighbouring waters. Vopr. Ikhtiol. 2:139-143. (Fish.
Res. Board Can. Transl. 284).
Gilbert, C. H.
1922. The salmon of the Yukon River. Bull. U.S. Bureau
Fish. 38:317-332. (Doc. 928.)
Grant, W. S.. G. B. Milner, P. Krasnowski, and F. M. Utter.
1980. Use of biochemical genetic variants for identification
of sockeye salmon (Oncorhynchus nerka) stocks in Cook
Inlet, Alaska. Can. J. Fish. Aquat. Sci. 37:1236-1247.
Hjort, R. C, and C. B. Schreck.
1982. Phenot^TDic differences among stocks of hatchery and
wild coho salmon, Oncorhynchus kisutch, in Oregon, Wash-
ington, and California. Fish. Bull., U.S. 80:105-119.
KlJIMA, A.. AND Y. FUJIO.
1984. Relationship between average heterozygosity and river
population size in chum salmon. Bull. Jpn. Soc. Sci. Fish.
50:603-608.
Markert, C. L., and I. Faulhaber.
1965. Lactate dehjTogenase isozyme patterns of fish.
J. Exp. Zool. 159:319-322.
Mason, J. C.
1976. Some features of coho salmon, Oncorhynchus kisutch,
fry emerging from simulated redds and concurrent changes
in photobehaviour. Fish. Bull., U.S. 74:167-175.
McBride, D. N., H. H. Hamner, and L. S. Buklis.
1983. Age, sex, and size of Y'ukon River salmon catch and
escapement, 1982. Alaska Dep. Fish Game Tech. Data Rep.
90, 141 p.
MILLIGAN, p. a., W. 0. RUBLEE, D. D.CORNETT, AND R. A. C.
Johnston.
1986. The distribution and abundance of chum salmon
(Oncorhynchus keta) in the upper Yukon River basin as
determined by a radio-tagging and spaghetti tagging pro-
gram: 1982-1983. Can. Tech. Rep. Fish. Aquat. Sci. 1351,
141 p.
Nei, M.
1978. Estimation of average heterozygosity and genetic
distance from a small number of individuals. Genetics 89:
583-590.
Nikol'skii, G. V.
1961. Special ichthyology. Published for the National Sci-
ence Foundation, Washington, D.C. and the Smithsonian
Institution, by the Israeli Program for Scientific Translation
Jerusalem, 538 p.
673
FISHERY Bl'LLKTIN: VOL. 8«. NO. 4
Okazaki, T.
1981. Geographic distribution of allelic variations of enzymes
in chum salmon, Oncorhynchutt keta. populations of North
America. Bull. Jpn. Soc. Sci. Fish. 47:507-514.
Raymond, J. A.
1981. Incubation of fall chum salmon (Oncorhynchus keta) at
Clear Air Force Station, Alaska. Alaska Dep. Fish. Game
Info. Leafl. 189, 26 p.
RiCKER, W. E.
1972. Hereditary and environmental factors affecting certain
salmonid populations. In R. C. Simon and P. A. Larkin
(editors). The stock concept in Pacific salmon, p. 27-160.
H. R. MacMillan Lectures in Fisheries, University of British
Columbia, Vancouver, B.C.
1976. Review of the rate of growth and mortality of Pacific
salmon in salt water, and noncatch mortality caused by
fishing. J. Fish. Res. Board Can. 33:1483-1524.
Salo, E. 0.
In press. The life history of chum salmon (Oncorhynchus
keta). Can. Fish. Aquat. Sci. Spec. Publ.
Sano, S.
1966. Salmon of the North Pacific Ocean Part III. A review
of the life history of North Pacific salmon. 3. Chum salmon
in the Far East. Int. North Pac. Fish. Comm. Bull. 18:
41-57.
Satterthwaite, F. E.
1946. An approximate distribution of estimates of variance
components. Biom. Bull. 2:110-114.
Smirnov, a. I.
1975. The biology, reproduction, and development of the
Pacific salmon. Moscow University, Moscow, USSR.
(Translated from Russian: Fisheries and Marine Services
Translation 3861, Ottawa, Canada.)
Sneath, p. H., and R. R. Sokal.
1973. Numerical taxonomy. W. H. Freeman and Co., San
Francisco, CA, 573 p.
SOKAL, R. R., AND F. J. ROHI.F.
1969. Biometry. W. H. Freeman and Co., San Francisco,
CA, 776 p.
Taylor, E. B., and J. D. McPhail.
1985. Variation in body morpliologj' among British ('olum-
bia populations of coho salmon, Oncorhipich ux kinutrh . Can.
J. Fish. Aquat. Sci. 42:2020-2028.
Utter, F. M., D. Campton, S. Grant, G. Milnek, ,J. Seeb. and
L. WlSHARD.
1980. Population structures of indigenous salmonid species
of the Pacific Northwest: I. A within and between species
examination of natural populations based on genetic varia-
tions of proteins. In D. C. Himsworth and W. J. McNeil
(editors), Salmonid ecosystems of the North Pacific,
p. 285-304. Oregon State University Press, Corvallis,
OR.
Utter, F. M., H. 0. Hodgins, and F. W. Allendorf.
1974. Biochemical genetic studies of fishes: potentialities and
limitations. In D. Malins (editor). Biochemical and bio-
physical perspectives in marine biology. Vol. 1, p. 213-237.
Academic Press, San Francisco, CA.
Vladykov, V. D.
1962. Osteological studies on Pacific salmon of the genus
Oncorhynchus. Bull. Fish. Res. Board Can. 136, 171
P-
WiLCOCK, J. A.
1984. Origins of chinook salmon (Oncorhynchui: tshaun/tschn
Walbaum) in the Yukon River fisheries, 1983. Alaska Dep.
Fish. Game Info. Leafl. 243, 30 p.
WiLCOCK, J. A., and D. N. McBride.
1983. Origins of chinook salmon {Oncorhynchus tshauytscha
Walbaum) in the Yukon River fisheries, 1982. Alaska Dep.
Fish. Game Info. Leafl. 226, 36 p.
674
FOOD HABITS AND DAILY RATION OF GREENLAND HALIBUT,
REINHARDTIUS HIPPOGLOSSOIDES, IN
THE EASTERN BERING SEA
M. S. Yang and P. A. Livingston'
ABSTRACT
This study shows that diet of Greenland halibut varies mainly by depth and size, and that size of prey
fish increases as the Greenland halibut increases in size. A total of 1,333 Greenland halibut (or turbot),
Reinhardtius hippoglossoid.es, stomachs were collected in the eastern Bering Sea from May 1983 to
November 1985 and analyzed. Stomach content data were divided into four groups based on sample loca-
tion (depth). Using length of the sample animals within each depth group, data were further divided
into five size groups. Walleye pollock, Theragra chalcogramma, was the most important prey (58% by
weight of the total stomach content). Squids (mainly Berryteuthis sp.) were the second most (20% by
weight) important food of Greenland halibut. Zoarcids and some deep-water fishes (e.g., bathylagids,
myctophids, macrourids) were also important food for Greenland halibut 30-69 cm long. Euphausiids
were only important as food (64% by weight) of the fish <20 cm collected in the continental shelf <200
m deep. Fishes >70 cm fed almost exclusively on fish in all depth areas.
Variation in mean stomach content weight throughout the day was used to determine the diel feeding
pattern; from this it appears that Greenland halibut is a continuous feeder. Daily rations (% of body
weight per day) of Greenland halibut were calculated using an exponential gastric evacuation rate model.
Fish >70 cm had a higher daily ration value (1.17% of body weight per day) than did those of the two
smaller size groups (0.66 and 0.64% of body weight per day for 30-49 and 50-69 cm size groups,
respectively).
Greenland halibut (or turbot), Reinhardtius hippo-
glossoides, is an amphiboreal fish, occurring in both
the North Atlantic and the North Pacific, but not
in the intervening Arctic Ocean (Hubbs and Wili-
movsky 1964). Within this range, the species has
been most extensively studied in the Atlantic Ocean.
In the Pacific Ocean, Greenland halibut has been
found from Baja California (Schmidt 1934), Oregon
(Niska and Magill 1967), and Vancouver B.C. (West-
rheim and Pletcher 1966), through the Bering and
Okhotsk Seas (Shmidt 1950), to Honshu Island,
Japan (Hart 1973), but the center of abundance is
in the eastern Bering Sea area.
Data of the resource assessment surveys (from
1979 to 1985) in the eastern Bering Sea performed
by the Northwest and Alaska Fisheries Center
(NWAFC), National Marine Fisheries Service
(NMFS), show that Greenland halibut ranked
between 5th and 12th place in terms of relative
abundance (kg/ha) among the groundfish species;
however, it is the most abundant species in continen-
tal slope areas (Bakkala 1986^).
These assessments suggest that Greenland halibut
is a key member of the eastern Bering Sea ecosys-
tem. The importance of this species in predator-prey
relationships of this ecosystem is poorly understood
since little is known about its food habits and food
consumption rate. Food habits of Greenland halibut
in the North Atlantic have been studied by Bower-
ing and Lilly (1985) and Haug and Gulliksen (1982).
In the eastern Bering Sea, Mikawa (1963), Mito
(1974). Smith et al. (1978), and Livingston et al.
(1986) reported stomach contents analysis of the
Greenland halibut, but the sample sizes in these
studies were small and the analyses were limited.
The objective of this study is to provide a descrip-
tion of the food habits of Greenland halibut in the
eastern Bering Sea, including diel, spatial, and
seasonal variations in stomach contents; influence
of predator size; and daily ration.
'Northwest and Alaska Fisheries Center. National Marine Fish-
eries Service. NOAA. 7600 Sand Point Way N.E., BIN C15700.
Building 4, Seattle. WA 98115.
^Bakkala, R. G. 1986. Greenland turbot— biological report.
Unpubl. manuscr., 21 p. Northwest and Alaska Fisheries Center,
National Marine Fisheries Service, NOAA. 7600 Sand Point Way
N.E.. Seattle, WA 98115.
Manuscript accepted June 1988.
FISHERY BULLETIN: VOL. 86. NO. 4. 1988.
675
I'lSllKin' HILLKTIN: \()1.. Si\. NO, 1
METHODS AND MATERIALS
Sample Collection and
Stomach Content Analysis
Stomachs from 1,333 Greenland halibut were
collected from May 1983 to November 1985 in the
eastern Bering Sea area (Fig. 1) from NMFS re-
search vessels and foreign commercial fishing
vessels (through the U.S. Foreign Fisheries Ob-
server Program). A bottom trawl was the only sam-
pHng gear used, and trawl samples were taken
throughout the day and night. Due to the low abun-
dance of Greenland halibut in the shelf area (<200
m), stomachs were taken from virtually all Green-
land halibut encountered in trawl catches. Random
size-stratified samples were obtained from the
catches in the slope area. Captured fish were
checked in ihv i'lvld for signs of regurgitation and
were discarded when there was evidence of food
items in the mouth or gill plates or of flaccid
stomachs. Stomachs from the sampled fish were ex-
cised and put into cloth bags with a specimen label
containing fork length, sex, and station information,
and were preserved in 4% formaldehyde solution.
Individual fish weights were not recorded at sea, but
were estimated by using the weight-length equation.
W{g) = 0.0060717 x L(cm)
i08K(;4
(1)
estimated from the Greenland halibut data base of
the Resource Assessment and Conservation Engi-
neering (RACE3) Division of the NWAFC.
•^Northwest and Alaska Fisheries Center, National Marine Fish-
eries Service, 7600 Sand Point Way N.E., BIN C1.5700, Building
4, Seattle, WA 98115.
63 OON
■- 61 OON
59 OON
■- 57 OON
-■- 55 OON
■■- 53 OON
51 OON
1 79 OOE
176 OOW
171 OOW
166 OOW
161 OOW
156 OOW
Figure 1. -Sampling locations for Greenland halibut in the eastern Bering Sea by four different depth strata, <200 m ( + ), 200-.399
m (O), 400-599 m (■), and >600 m (A).
676
YANG and LIVINGSTON; FOOD HABITS OF GREENLAND HALIBUT
Stomachs were analyzed individually in the labor-
atory. Prey items were identified to the lowest possi-
ble taxonomic level and counted. Wet weights of the
prey items were recorded to the nearest milligram
after blotting with paper towels. The fork lengths
of prey fish were also measured.
Diet Description
Since the depth distribution of Greenland halibut
in this study was broad (from 62 m to 891 m), stom-
ach content data were first subjectively divided by
100 m depth groups. For each 100 m depth class,
percent frequency of occurrence (%F0) of prey
items, percentage of total stomach content weight
(%W) by prey t}T)e, and the percentage of total prey
number (%A^) by prey type were calculated by using
ECO/INDEX, a computer program for calculating
feeding ecology indices (Vodopovich and Hoover
1981). Based on similarities of major prey items
(using percent by weight), stomach content data
were combined into four depth groups for analysis:
Depth 1 (<200 m). Depth 2 (200-399 m), Depth 3
(400-599 m), and Depth 4 (^600 m).
Within each of the four depth groups, data were
subjectively divided by fish length into 10 cm size
groups. By comparing percent by weight of the
major prey categories (e.g.. gadids, squids) for each
size group, the fish within each of the four depth
groups were finally lumped into five size groups: <20
cm, 20-29 cm, 30-49 cm, 50-69 cm, and >70 cm.
Seasonal breakdowns of stomach contents by depth
and predator si^^e group could not be performed due
to insufficient sample sizes.
Diel Feeding Pattern
Because of the small sample size of fish <30 cm
long, only data from three size groups (30-49 cm,
50-69 cm, >70 cm) were used for diel feeding anal-
ysis. For each size group, the stomach content
weights as percent of body weight were calculated
for each 3-h period of the 24-h day. Any possible
seasonal variations of the diel feeding pattern could
not be analyzed because of insufficient seasonal
samples.
Daily Ration
In this study, daily ration was calculated using
Elliott and Persson's (1978) model. The basic
assumption of this model are that the rate of gastric
evacuation (R) is exponential and temperature
dependent. If stomach samples are taken at fixed
intervals of t hours, the mean stomach content
weight as a percentage of fish weight (S,) in each
interval (i) is calculated for a total of m intervals
over the 24-h period. According to Elliott and
Persson (1978), the daily ration (D.R.) in terms of
percentage of body weight is therefore given by
D.R. =
Rt
1 - expi-Rt) ' = 1
24 SR
J. 8,(1 - expi-Rt))
(2)
where S = J. S,/m. Elliott (1972) found the gen-
eral relationship between R and temperature (T)
was exponential:
R = ae
hT
(3)
Based on data presented in the literature for the
normal temperature range of both freshwater and
marine fishes, Durbin et al. (1983) concluded that
the slope (b) is fairly constant for different prey
types and fish species (mean = 0.115), while the
intercept (a) changes with prey type and can be
estimated from gastric evacuation rate experiments.
Since there were no gastric evacuation rate data
available for Greenland halibut, results of gastric
evacuation experiments on walleye pollock, Thera-
gra chalcogramma, feeding on juvenile pollock and
squid were used (Dwyer et al. 1987). Although wall-
eye pollock is taxonomically very different from
Greenland halibut, these two species have some prey
in common. In addition, both species are active, off-
bottom feeders which could be expected to be more
similar in terms of metabolic rates than benthic
feeding, small mouth flounders whose food intake
has been studied more extensively. The intercept
"a" in Equation (2) was 0.0143 for juvenile walleye
pollock and 0.0079 for squid. For this study, the
intercept for walleye pollock prey was used to cal-
culate daily ration when fish was the main prey
(>70% of diet by weight), and the intercept for
squid was used when squid was the main prey. If
the diet was split evenly between fish and squid
prey, daily ration was calculated using both inter-
cept values to obtain a likely range of daily ration
values.
Average bottom temperatures for the eastern
Bering Sea for this study were estimated from
oceanographic data on the Bering Sea (Ingraham
677
I'lSllKKV HII.LKTIN: VOL. 8t;. NO. 4
ions'*). Because of the small differences in temper-
atures between different seasons (e.g., 2.90°C,
2.47°C, and 2.93°C were the average temperatures
for spring, summer, and autumn at locations where
fish 30-49 cm were collected), and the lack of sam-
ples for fish <30 cm, daily rations were calculated
for three size groups (30-49 cm, 50-69 cm, and >70
cm) with all seasons combined. Temperatures used
for each of the three size-groups were calculated by
matching the haul locations of each size group in our
study with the long-term monthly mean bottom
temperature at those positions and calculating the
average bottom temperature.
RESULTS
General Description of Diet
Stomachs from 1,333 Greenland halibut were
analyzed; of these, 610 stomachs (46%) were empty.
The size of the Greenland halibut ranged from 9 to
99 cm (fork length) with a mean of 56 cm. The sam-
pling depth ranged from 62 to 891 m with most of
the samples (55%) collected from the area 400-599
m deep.
Prey consumed included gastropods, cephalopods,
crustaceans, ophiuroids, and fish (Table 1). Twelve
families offish and at least 14 different fish species
were represented in the stomach contents. Fish
dominated the contents in terms of frequency of oc-
currence, number, and weight; walleye pollock was
the most important fish species consumed with
respect to all three measures of prey importance.
Three genera of squid were consumed (mainly
Berryteuthis sp.) and were the second most impor-
tant prey. The importance of the various prey
species or groups (e.g., gadids, squids) changes with
bottom depth and Greenland halibut size. Those
changes will be discussed in the following sections.
Spatial and Size Differences
Depth 1 (<200 m)
Gadids constituted more than 87% by weight of
the stomach contents of all but the smallest (<20 cm)
size group (Fig. 2A). Euphausiids comprised 64%
by weight (54% by number) of the diet of Greenland
halibut <20 cm long. In size group 20-29 cm, the
■•Ingraham, W. J. 1983. Temperature anomalies in the eastern
Bering Sea 1953-82. NWAFC Processed Rep. 83-21, 348 p.
Northwest and Alastca Fisheries Center, National Marine Fisheries
Service, NOAA, 7600 Sand Point Way N.E., Seattle, WA 98115.
percentage by number of euphausiids was still high
(52%), but the percentage by weight decreased
dramatically to only 2%. Gadids were the dominant
prey for size groups larger than 30 cm in Depth 1
in terms of percent of frequency of occurrence, per-
cent of total stomach contents weight, and the
percent of prey number. Cephalopods, clupeids,
osmerids, stichaeids, myctophids, and macrourids
were not important food items at this depth.
Depth 2 (200-399 m)
No Greenland halibut smaller than 30 cm were col-
lected at this depth range. Cephalopods, in terms
of the three diet measures shown in Figure 2B, were
the dominant food items found in the size groups
30-49 cm and 50-69 cm. On the other hand, gadids
were the dominant prey in size group >70 cm (Fig.
2B). When all sizes were combined, the cephalopods
were more important than gadids (Fig. 2B, lower
right); however, when the size groups were sepa-
rated (Fig. 2B), the different contribution of gadids
and cephalopods to different size groups is very clear
(cephalopods are most important for fish <70 cm and
gadids are important for fish ^70 cm). Miscellaneous
prey fishes found in the entire Depth 2 group
included zoarcids, bathylagids, myctophids, and
pleuronectids.
Depth 3 (400-599 m)
No stomachs of Greenland halibut smaller than 30
cm were collected in this depth range (Fig. 2C). For
Greenland halibut 30-49 cm long, zoarcids (38.6%)
were the most important prey item in terms of
percentage by weight of the stomach contents,
followed by cephalopods (30.7%) and gadids (17%).
In terms of percent of prey number, cephalopods
comprised 39% of the total prey in this size group,
followed by gadids (22%), bathylagids (17.3%), and
zoarcids (8.7%). For the size group 50-69 cm,
cephalopods, gadids, and zoarcids comprised 61, 21,
and 12% by weight of the stomach contents, respec-
tively. Cephalopods also comprised the highest
percentage (35%) by the number in this size group.
Gadids were the dominant prey of large Greenland
halibut (>70 cm). They comprised 87% by weight,
69% by number, and 82% by frequency of occur-
rence of the stomach contents of this size group (Fig.
2C). Other prey fishes found in the Depth 3 group
included stichaeids, myctophids, cottids, macroiu^ids,
cyclopterids, and pleuronectids. For the Depth 3
group (Fig. 2C, lower right), gadids were the domi-
678
YANG and LIVINGSTON: F-OOn HABITS OF GREENLAND HALIBIT
Table 1.— Prey items found in the stomachs of Greenland halibut collected in the eastern Bering Sea during summer 1983 through fall
1985. %F0 = percent frequency of occurrence, %/V = percent by number, %W = percent by weight, t = <0.01% W.
Prey item
%F0
%/V
o/oW
Prey item
%F0
o/oN
%W
Gastropoda
Ophiuroidea
1.1
1.81
t
Buccmum sp.
0.4
0.34
0.14
Ophiurida
1.0
1.72
t
Cephalopoda
31.9
22.67
20.28
Unidentified Ophiuroidea
0.1
0.09
t
Teuthoidea
30.8
21.98
19.65
Larvacea Copelata
0.4
0.69
t
Gonatidae
10.5
8.16
11.64
Gonatopsis sp.
Gonatus sp.
1.9
0.7
1.20
0.60
2.97
0.13
Teleostei
Clupeidae
71.7
61.36
77.93
Gonatus magister
Berryteuthis sp.
Berryteuthis magister
0.1
6.4
0.3
0.26
4.81
0.26
0.09
6.69
1.03
Clupea pallasii
Osmeridae
Bathylagidae
0.3
0.1
7.6
0.17
0.09
8.24
0.74
t
0.63
Unidentified Gonatidae
1.4
1.03
0.73
Leurogiossus stilbius
0.4
0.43
0.04
Unidentified Teuthoidea
20.5
13.82
8.01
Unidentified Bathylagidae
7.2
7.81
0.59
Octopoda
Unidentified Cephalopoda
0.1
1.0
0.09
0.60
0.51
0.12
Myctophidae
Stenobrachius leucopsarus
5.0
0.1
6.87
0.09
0.71
t
Unidentified fvlyctophidae
4.8
6.78
0.71
Crustacea
7.8
12.39
0.41
Gadidae
32.7
26.35
61.08
Mysidacea
2.2
1.64
0.01
Theragra chalcogramma
28.2
23.35
58.39
Gnathopausia gigas
0.7
0.52
t
Unidentified Gadidae
4.4
3.00
2.69
Holmesiella anomala
1.0
0.77
t
Zoarcidae
5.0
4.13
3.51
Pseudomma truncatum
0.1
0.09
t
Lycodes sp.
4.1
3.52
2.95
Unidentified Mysidacea
0.4
0.26
t
Lycodes diapterus
0.1
0.09
0.05
Cumacea
0.1
0.17
t
Lycodes palearis
0.1
0.09
0.11
Unidentified Zoarcidae
0.6
0.43
0.40
Amphipoda
1.2
1.54
t
Macrouridae
1.0
1.12
2.00
Gammaridea
Hyperiidea
Parathemisto libellula
1.0
0.77
t
Coryphaenoides sp.
0.7
0.43
0.39
0.3
0.77
t
Coryphaenoides filifer
Unidentified Macrouridae
0.1
0.8
0.09
0.60
0.04
1.57
Euphausiacea
1.8
4.90
t
Icelidae
Thysanoessa inermis
1.2
3.78
t
Icelus spiniger
0.1
0.09
0.03
Unidentified Euphausiacea
0.6
1.12
t
Cottidae
0.2
0.18
0.21
Decapoda
Dasycottus setiger
0.1
0.09
0.20
Caridea
2.4
3.79
0.15
Hemitripterus boline
0.1
0.09
0.01
Pasiphaeidae
Cyclopteridae
0.6
0.35
1.73
Pasiphaea pacifica
0.1
0.09
t
Aptocyclus venthcosus
0.1
0.09
0.55
Hippolytidae
0.4
2.58
0.06
Careproctus cypselurus
0.1
0.09
0.82
Eualus sp.
03
2.06
0.04
Unidentified Cyclopteridae
0.3
0.17
036
Eualus biunguis
0.3
0.52
0.02
Stichaeidae
0.7
0.43
0.09
Panadalidae
1.2
0.78
0.08
Lumpenus maculatus
0.3
0.17
0.04
Pandalus sp.
0.8
0.52
0.04
Unidentified Stichaeidae
0.4
0.26
0.05
Pandalopsis dispar
0.3
0.17
0.03
Pleuronectidae
Unidentified Pandalidae
0.1
0.09
0.01
Reinhardtius liippoglossoides
0.6
0.34
1.60
Crangonidae
Unidentified Teleostei
20.6
12.88
5.02
Crangon communis
0.3
0.17
0.01
Unidentified Caridea
0.3
0.17
t
Unidentified organic material
0.3
0.17
t
Reptantia
0.4
0.27
0.25
Total number of stomachs
1,333
Anomura
Total stomachs with food
723
Paralithodes sp.
0.1
0.09
0.14
Total prey weight (g)
47,713.52
Brachyura
0.3
0.18
0.11
Total prey number
1,165
Chionoecetes sp.
0.1
0.09
0.08
Chionoecetes opilio
0.1
0.09
0.03
Unidentified Crustacea
0.1
0.09
t
nant prey in the diet due to the consumption of
walleye pollock by Greenland halibut >70 cm; how-
ever, cephalopods and other fishes were more im-
portant than gadids for the two smaller size groups.
Depth 4 (^600 m)
No stomachs were collected for fish smaller than
30 cm in this depth group (Fig. 2D). For size groups
30-49 cm, cephalopods were the dominant prey
(56%) of Greenland halibut in terms of percent by
weight, followed by bathylagids at 29%. However,
the percent of number and the percent of frequency
of occurrence of bathylagids (FOG in Figure 2D, up-
per left) were higher than those of the cephalopods.
The stomach contents (by weight) of Greenland
halibut 50-69 cm long was composed of 57%
cephalopods, 22% macrourids, 12% bathylagids, 4%
679
KISHKkV mi.LKTIN: VOL. S(i. NO. -1
<20 cm
S= 37, NE = 20
DEPTH 1
(<200m)
200
20 29 cm
S= 23. NE
12
100-|
80-
%N 60-
40
20-{:
0
20-::!
60-:
80 4
100
0
~i — r
1 — I — r
100
%F.O.
200
30-49 cm
S= 120, NE = 82
%W
100
%F.O.
%N
%W
50-69 cm
S= 60, NE = 44
100 -,
100 —
I r
0
100
%F.O.
LEGEND
■:■>■>'■>
GAD
EUP
CEP
FOG
CRU
OTH
m
200
>70cm
S = 4, NE = 3
%N
100
80-):
60-
40
20
0
%W
60
80-
100-
0
20 -f::::::i::;:;:;:;x::;' "'■"■'
40
i 1 r
^n — ^-r
100
%F.O.
200
All sizes combined
S= 244, NE = 161
%N
100
80
60
40-1
20
0
%W
|«v«ti«ffi«'>^ia*i
•:::-:-:-::::::::::::::-t^
20 -•xi:;:;:;:;:;:::;:::::;:;:;:
60-:
80 -|:;:::::;:;:;:;:;:;:;:;:::;:;:;^
100-
0
I r
"I — 1 — r
100
%F.O.
— I
200
Figure 2. — Percent of frequency of occurrence (%F0), percent of prey number (%Af), and percent of stomach content weight
(%W) of major prey items in the stomach contents of Greenland halibut (by depth and by size). S, total number of stomachs;
NE, nonempties; %F0, GAD, Gadids; EUP, Euphausiids; CEP, Cephalopod; FOG, Fish other than gadids; CRU, Crustacean;
OTH, Others. A) Depth 1.
680
YANG and LIVINGSTON: FOOD HABITS OF GREENLAND HALIBUT
DEPTH 2
(200-399 m)
%N
100
80-
60-
40-
20-
0
20
30-49 cm
S=26, NE
%W
40-
60-
80
100
0
I I 1
100
%F.O.
> 70 cm
S= 19. NE = 9
100
80-1
%N
%W
.•.•.•.•.•.••■•■•I
60 -■:::::•:::
0-
40
80-
100-
I I I I
200
~^ — 1 — 1 — I — I — I — r
0 100
%.F.O.
200
50-69 cm
S- 71, NE = 33
B
%w
%N
%W
All sizes combined
S= 118, NE = 49
100-,
80-
60-
40-
20
0
20 -fJ
40-
60-
80-
100
lYtr'.'.'r
0
l — \ — I —
100
%F.O.
LEGEND
GAD
EUP
CEP
FOG
CRU
OTH
!*!'I*!'I*I*
P
■
200
Figure 2— Continued.— '^) Depth 2.
681
KlSllKKV HILLKTIN: VOL. Hti. NO. 4
DEPTH 3
(400 -599 m)
30-49 cm
S= 193, NE = 88
lOOn
%N
60-
40-
%W
20-
40-
60-
80-
100
20— T!v!v' :•:■:■:■:■:■: :':■
0
^^^^^^^
=L
0
1 — \ — \ — r
100
%F.O.
>7Gcm
S= 140, NE = 101
100-1
80
%N
„„ mill.
40-;:
20 4
0
%W
80 -t
100
t i i 1 i'i'iiwW'i
4o-g!:ixixix!x:::::x
'"-••t •'
1 — I — I — I — r~
0 100
%F.O.
1 — \ — r
200
200
50-69 cm
S = 400, NE = 239
%N
%W
luu— 1
80-
60-
40-
2U-
U
20-
40-
60-
80-
1 nn
IU(J-|
1
1 1 1
1
1 1 1 1 1
0
100
%F.O.
All sizes combined
S= 734, NE = 428
100-
80-1
o/^, BOH
%N
40-
20-|:i
0
20
%W
60-
80-
100-
■■ V ■ ■ ■ *
3
0
1 — I — r
100
%F.O.
200
LEGEND
GAD
EUP
CEP
FOG
CRU
OTH
'^^
200
Figure 2.-Continued.-C) Depth 3.
682
YAi\(; and LIVINGSTON: FOOD HABITS OF (;REKNLAND HALIBIT
DEPTH 4
(^600 m)
30-49 cm
S= 11, NE = 5
%N
%W
80-
60-
40-
20-
^ ( ,
n
: : : : : :
u
/ / ,
20-
40-
60-
80-
inn
1 1 1
I
1
1 1 1 1
0
100
%F.O.
%N
> 70 cm
S= 87, NE = 40
100-1
80-
60
40
2o-:ii:iiii:::ig:i:i:i:ii::i
X'X'X'.v.v.'.'.v
0
%w
60-
80-
100-
•:•:•:•:•:•:•:•:•:*.'.'.'.'.
:i:::::::x:S?x-ri
4oJfe:i:i:!:i:i:i:::x::::::
0
1 — I — r
100
%F.O.
200
1 — \ — r
200
50-69 cm
S= 139, NE = 40
100 -,
%w
%N
; I ; ; ; ; ; ; -^
%w
All sizes combined
S= 237, NE = 85
100-
80-
60-
40-
20-
0
90 — .v.v.v.v
'- ^ '.•.'.•.•.:•.•.:
40-;::::::::::i^
60-;;:;:;:;:;:;:;:;!
80-
100 1-^
D
200
LEGEND
/ / / A
• GAD
^ EUP
CEP
FOG
CRU
OTH
1 — \ — \ — r
100
%F.O.
200
Figure 1— Continued— Y)) Depth 4.
683
FISllKKV HULLETIN: VOL. 86. NO. 4
myctophids, and 0.6% stichaeids. In terms of
percent by number and the percent of frequency of
occurrence, the fishes other than gadids (FOG in
Figure 2D) were more important than the cepha-
lopods. No gadids were found in either the 30-49
cm group or the 50-69 cm group of fishes; however,
they were the dominant prey (75% by weight, 64%
by number, and 73% by frequency of occurrence)
of Greenland haUbut >70 cm (Fig. 2D, lower left)
as in the other three depth groups (Fig. 2A,B,C,
lower left). Other food in this size group included
macrourids, cyclopterids, and pleuronectids. Even
though gadids did not occur in the two smaller size
groups, they were important in the Depth 4 group
as a whole (Fig. 2D, lower right).
Trends in Fish Consumption
The size of the walleye pollock consumed by
Greenland halibut increased dramatically with pred-
ator size (Fig. 3). The relationship appears linear
with r^ = 0.835. Based on the age-length key for
walleye pollock (Halliday and Umeda 1986), the wall-
eye pollock eaten by Greenland halibut were approx-
imately age 0 and age 1 for smaller size fish (<50
cm), age 1 and age 2 for medium size fish (50-69
cm), and age 3 and age 4 for the fish >70 cm
(Fig. 4).
The importance of prey fishes in the diet of Green-
land halibut appears to be depth related (Table 2).
Clupeids disappear from the diet in waters >200 m
deep, while gadids are important in all depths but
occur most frquently (85%FO), and comprise most
of the diet both in number and weight (65% and
93%, respectively), in the area <200 m deep. Zoar-
cids appear in stomachs only in the area 200-600
m deep, and bathylagids start appearing at 200-399
m and increase in importance as the water depth
increases. Myctophids seem to be more important
in the area 400-599 m than in the other depths.
500
450
400-
'§ 350-
E
-P. 300
WALLEYE POLLOCK
Y =
= 5.5686 X-
-97.972
r^
= 0.835
n =
167
A A. A
20 40 60 80
Predator fork length (cm)
100
Figure 3.— Scatter plot of the fork length of walleye pollock that were consumed by Greenland
halibut of different sizes in the eastern Bering Sea.
684
XI-
n = 40
YANC; and LIVINGSTON: FOOD HABITS OF GREENLAND HALIBUT
16
1 2 — n = 64
8
4^
0
12-
8-
4
0
12
^ 8
OJ
3
Q) 4 —
0
12H
8
4-1
0
12-
8-
4
0
IH
nmH
rn n.n n
r^^^
n = 1
_n_
n-4
Predator
fork length
>70cm
50-69 cm
30-49 cm
20-29 cm
< 20 cm
0
50
100
150
200
250
300
350
400
450
Prey pollock fork length (mm)
FiGi'RE 4.— Length-frequency distributions of walleye pollock consumed by Greenland halibut in the eastern Bering Sea.
while macrourids are important prey in even deeper
waters (^600 m).
Trends in Stomach Fullness
Fewer empty stomachs were found in summer
than in spring and autumn except in the ^70 cm size
group (Fig. 5). The occurrence of empty stomachs
for all size groups was about 35% in summer and
50% in autumn. In spring, samples from the 30-49
cm size group had the greatest percentage of empty
stomachs (about 70%), followed by size groups 50-69
cm (about 58%) and >70 cm (about 12%).
There were no apparent diel trends in stomach
content weight in this study (Fig. 6). The stomach
content weight (expressed as percentage of body
weight) for large fish (^70 cm) was fairly constant
except for the 0900-1200 h time period. The
685
FISHERY BULLETIN: VOL. m. NO. J
Table 2 —Importance of the prey fish by depth found in the stomachs of Greenland halibut collected from the eastern Bering Sea.
= percent frequency of occurrence, %N = percent by number, %W = percent by weight.
%F0
Depth 1
Depth 2
Depth 3
Depth 4
«200 m)
(200-399 m)
(400-599 m)
(>600 m)
Prey fish
%F0
o/oN
%W
%F0
%/V
o/oW
%F0
o/oN
o/oW
%F0
00
%N
00
%W
Clupeidae
0.6
0.3
2.8
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Osmeridae
0.6
0.3
0.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Stichaeidae
1.2
0.7
0.4
0.0
0.0
0.0
0.5
0.3
0.1
1.2
0.7
0.1
Gadidae
85.1
64.9
92.7
36.7
26.5
36.2
45.3
31.4
64.4
42.4
25.2
679
Zoarcidae
0.0
0.0
0.0
8.2
8.8
7.0
7.4
6.5
4.8
0.0
0.0
0.0
Bathylagidae
0.0
0.0
0.0
6.1
5.9
0,1
9.8
9.4
0.5
11.8
21.7
2.1
Myctophidae
0.6
0.3
0.1
4.1
4.4
0.2
7.2
11.1
1.3
2.4
2.8
0.4
Cottoidei
0.0
0.0
0.0
0.0
0.0
0.0
0.6
0.5
0.4
0.0
0.0
0.0
lylacrouridae
0.6
0.3
0.5
0.0
0.0
0.0
0.9
0.8
1.1
8.3
4.9
9.4
Cyclopteridae
0.0
0.0
0.0
0.0
0.0
0.0
0.7
0.5
1.4
1.2
0.7
6.0
Pleuronectidae
0.0
0.0
0.0
2.0
1.5
6.0
0.2
0.2
0.9
2.4
1.4
4.7
lOOf-
90
c
o
Q.
u
03
E
o
Q.
E
LU
80
70
60
50
40
30 -
20
10 -
Predator
—
fork
length
— 30-49 cm
—
— 50-69 cm
98
-- >70cm
—
\
\
\
\
227
^114
^.-^59
—
^^^'"'
—
344^^'
^^^
-'ioB
28'-'''
,
1 1
Spring
Summer
Autumn
Figure 5.— Percent of empty stomachs by season and Greenland halibut size. (Numbers are
sample sizes.)
stomach content weight of the fish 50-69 cm was
also fairly constant although much less than in the
larger fish. Although a midday drop in stomach con-
tent weight occurred in the two larger size groups,
stomach content weight for the smallest size group
appeared to steadily increase from 0100 h to 1200
h. The stomach content weight was highest but most
variable in size group 30-49 cm during the
1500-1800 h time period.
Daily Ration
The mean stomach content weight and the daily
ration varied among three size groups of Greenland
halibut (Table 3). By percent body weight, the mean
stomach content weight for 30-49 cm and 50-69 cm
size groups were 1.4% and 1.3% respectively. Fish
^70 cm had the greatest mean stomach content
weight (2.4% of body weight). Since gastric evacu-
686
YAN(; anri LIVINGSTON: FOOD HABITS OF GREENLAND HALIBIT
Table 3— Daily ration (D R ) and mean stomach content weight
(%BW ± SE) of Greenland halibut during each 3-h period. BW
= body weight; N = No. of stomachs (including empty stomachs).
Size class
Time
Mean stomach content
(Temperature)
period
N
weight (%BW ± SE)
30-49 cm
2400-0300
64
0.4700 ±0.1 772
(2.64°C)
0300-0600
21
0.9537 + 0.4470
fish as prey
0600-0900
44
1.1542 + 0.2926
a = 0.0143
0900-1200
59
1.7173 + 0.3097
R = 0.0194
1200-1500
27
1.8937 + 0.5661
DR. = 0.6644
1500-1800
52
2.5844+1.4731
1800-2100
49
0.7300 + 0.2425
squid as prey
2100-2400
34
1.9124 + 0.3710
a = 0.0079
R = 0.0107
S = 1 .4270
D.R. = 0.3665
50-69 cm
2400-0300
49
0.9141 +0.1306
(2.93°C)
0300-0600
78
1.5944 + 0.3292
+-<
fish as prey
0600-0900
108
1.6596 ±0.3011
sz
a = 0.0143
0900-1200
70
1.0256 + 0.2566
CT)
R = 0.0200
1200-1500
70
1.6406 + 0.2736
>•
■o
DR. = 0.6380
1500-1800
87
0.8967 + 0.2340
1800-2100
97
1.5372 + 0.3077
squid as prey
2100-2400
111
1.3652 ±0.2646
o
a = 0.0079
JD
R = 0.0111
S= 1.3292
C
D.R = 0.3541
u
>70 cm
2400-0300
25
2.8888 ±0.6365
(2.96°C)
0300-0600
22
2,4635 ±0.6325
CO
fish as prey
0600-0900
28
3.2101+0.6714
C
+->
a = 0.0143
0900-1200
21
1.0156 + 0.3308
R = 0.0201
1200-1500
40
2.8228 ±0.5439
c
o
DR. = 1.1712
1500-1800
40
1.7294 ±0.4454
o
1800-2100
41
2.5454 ±0.6559
^
2100-2400
33
2.7468 ±0.4782
S = 2.4272
o
CD
E
o
•1—'
ation rate (/?) is affected by prey type, the constant
a, used to calculate R for each size group, was deter-
mined by the percentage by weight of the main prey
items (fish or squid). In size group 30-49 cm, where
fish constituted 72% and squid constituted 25% by
weight of the diet, both a = 0.0143 (fish as prey)
and a = 0.0079 (squid as prey) were used for cal-
culating R. For size group 50-69 cm, fish constituted
47% of the diet and squid 52%, so gastric evacua-
tion rates were also calculated by using both a =
0.0143 and a = 0.0079. For size group ^70 cm, prey
fish comprised 94% of the diet, and a = 0.0143 was
used for calculating the gastric evacuation rate.
Large Greenland halibut (>70 cm) had the highest
daily ration value (1.17%), measured as percentage
of body weight per day (B WD), followed by the size
group 30-49 cm (0.66% BWD). The 50-59 cm size
group had the lowest daily ration value, 0.64% BWD
c/^
5r
0
>70cm
28
25
22
40
41 33
J- 40
21 t
}
J I I L
3r
2 -
1 -
0
50-69 cm
78 108
97 111
49
I
n|j.,{{
J I I L
30-49 cm
52
27
64
I
J L
44
5
59
34
I
■L 49
i
J I ■ ' I
0
12 18
Time (hours)
24
Figure 6.— Die) changes in mean stomach content weight (%BW
± SE) in the stomachs of three different size groups of Greenland
hahbut (the number above each bar was the sample size).
687
FISHERY BULLETIN: VOL. 86, NO. 4
for fish prey and 0.35% BWD for squid prey. Since,
fish and squid each constituted about one half of the
stomach contents by weight in this size group, the
actual daily ration lies within the range of the two
values, 0.64% and 0.35%.
DISCUSSION
This study demonstrates size-dependent prey pref-
erence by Greenland halibut; halibut <20 cm fed
primarily on euphausiids whereas those >20 cm
were largely fish and squid eaters. It is not surpris-
ing to find that walleye pollock was the dominant
prey (Table 1) of Greenland halibut because the esti-
mated biomass of walleye pollock in the Bering Sea
area is about 10,000,000 metric tons (Bakkala and
Wespestad 1983). Livingston et al. (1985^, 1986)
noted that walleye pollock is a major food source
not only for marine birds, marine mammals, and
man, but also serves as a major food source for domi-
nant components of the eastern Bering Sea ground-
fish complex. Other studies have also shown the
importance of walleye pollock as food of Greenland
halibut (Moiseev 1953: Mito 1974; Smith et al. 1978)
in the eastern Bering Sea.
In this study, large Greenland halibut (>70 cm)
ate fish almost exclusively. Bowering and Lilly
(1985) found that 65-69 cm was the length at which
Greenland halibut in the northwestern Atlantic
began to switch from smaller pelagic fish {Mallotus
villosus) to larger groundfish {Gadus m.orhua, Sebas-
tes sp., Anarhichadidae, Pleuronectidae, Zoarcidae)
as food. Mikawa (1963) noted increased piscivory
with size in Greenland halibut sampled in several
areas of the North Pacific. Mito (1974) also showed
the same trend; he found that 65-90 cm Greenland
halibut ate 20-40 cm long walleye pollock. These
observations suggest that large Greenland halibut
(^70 cm) feed on larger sized groundfish which may
be lower in the water column whereas the smaller
Greenland halibut (<70 cm) feed on smaller sized
pelagic fish in the upper water column.
Shuntov (1970) and Mikawa (1963) noted seasonal
depth migrations for Greenland halibut and inter-
preted the summer movement into shallower waters
as a feeding migration related to migrations of
walleye pollock. Based on the size distribution of
^Livingston, P. A., M. S. Yang, and D. Wencker. 1985. The
importance of juvenile pollock in the diet of key fish species in the
eastern Bering Sea. Unpubl. manuscr., 19 p. Presented as the
workshop on comparative biology, assessment, and management
of gadoids from the North Pacific and Atlantic Oceans, 24-28 June
1985. Northwest and Alaska Fisheries Center, National Marine
Fisheries Service, NOAA, Seattle, WA 98115.
walleye pollock in midwater trawl catches near the
Pribilof Islands and westward over the Aleutian
Basin, Livingston and Dwyer (1986) found that small
(age 0) pollock occurred in near-slope and shelf
areas, medium (age 1) pollock in shelf areas, while
larger (>1 year old) pollock occurred in all areas dur-
ing summer. Therefore, it can be concluded that in
the slope area, where juvenile walleye pollock (age
0 and 1) of the appropriate size for smaller Green-
land halibut were not available, the smaller sized
(30-69 cm) Greenland halibut ate the available prey,
cephalopods and deep-water fishes, while the larger
ones (^70 cm) consumed mostly larger walleye
pollock (>30 cm) and other fish regardless of depth
or season.
No clear diel feeding trends were found. The lack
of trends may be related to the large variations of
the time of sunrise and sunset in different seasons
in the Bering Sea. Other studies have varied find-
ings. Mito (1974) reported that Greenland halibut
fed primarily from sunset to midnight based on
hmited sample sizes (six specimens in some time
periods). Shuntov (1970) showed that this species
fed continuously in the Okhotsk Sea, although feed-
ing was somewhat higher during the night. By com-
paring day and night catch rates, Chumakov (1969)
concluded that Greenland halibut (in the Iceland
area) made daily vertical migrations (staying close
to the bottom during the day and moving up in the
water column at night). However, he did not cor-
relate this behavior with diel feeding. Thus, the
literature and this study show no definite diel feed-
ing trend in Greenland halibut.
Daily Ration
Daily ration calculations were based on the evacu-
ation rate of one prey item (pollock or squid) using
Elliott and Persson's (1978) model. Other authors
(Durbin et al. 1983; Dwyer 1984) have calculated
total daily ration by adding up the separate daily
rations of the different prey items. Persson (1984)
demonstrated that the evacuation of a specific food
item can be dependent on the ingestion of other food
items. Therefore, it may be erroneous to apply the
food consumption model to estimate the consump-
tion of individual prey types separately. Persson
(1984) suggested that the only practical solution to
calculate the daily rations of different prey items
is to calculate the mean weight of each food item
remaining in the digestive tract over 24 hours and
multiply the fraction it constitutes of the total mean
content with the total daily ration. This is necessary
688
VANC and LIVINGSTON: FOOD HABITS OF GREENLAND HALIBIT
since the complexity of feeding and evacuation pat-
terns in field populations of fish makes it impossi-
ble to estimate the consumption and evacuation of
different food items ingested at different times.
Bowering and Lilly (1985) estimated the consump-
tion rate of capelin, Mallotiis vUlosus, by Greenland
halibut in the northwestern Atlantic, using esti-
mates of gastric evacuation rate for Atlantic cod,
Gadus morhua, on capelin from Minet and Pero-
dou's (1978) study. Assuming a linear gastric evacu-
ation model, they found the time for Greenland
halibut to complete digestion of capelin at 2°-3°C
was 3-5 days. By using the gastric evacuation rate
calculated from this study, the time needed by large
Greenland halibut (^70 cm) to evacuate 99% of a
pollock meal at 3°C was 4.2 days, a value very
similar to Bowering and Lilly's (1985) estimate.
Livingston and Dwyer (1986'^) calculated an aver-
age daily ration for arrowtooth flounder of 0.62%
of body weight per day. This value is close to the
daily ration values calculated from this study. Since
Greenland halibut and arrowtooth flounder are eco-
logically and morphologically similar species, it is
not surprising to find their daily ration needs are
similar.
Huebner and Langton (1982) performed a gastric
evacuation study on winter flounder, Pseudopleuro-
nectes americanus. They used squid as food for fish
10-40 cm at 5.5°-7.0°C to get a gastric evacuation
rate of 0.079/h and calculated daily ration in the
range of 1.8-2.4% BWD. Compared to these values,
the daily ration of Greenland halibut calculated in
this study is low, possibly due to the lower temper-
atures in this study.
Greenland halibut ^70 cm apparently ate a higher
daily ration (1.17% of body weight per day) than did
those <70 cm (0.66% and 0.64% BWD for size group
30-49 cm and 50-69 cm, respectively). Flowerdew
and Grove (1979) studied the effects of body weight
and meal size on gastric emptying time in the turbot,
Scophthalmus maximus. Their results showed that
large fish emptied a meal of a given size from the
stomach at a faster rate than small fish, and large
meals in a given fish were processed at faster rate
than small meals. Dwyer (1984) also found a higher
daily ration value for larger walleye pollock in the
•'Livingston, P. A., and D. A. Dwyer. 1986. Food web inter-
actions of key predatory fish with northern fur seal, CallorhinMs
ursinus. in the eastern Bering Sea during summer 1985. In T. R.
Loughlin and P. A. Livingston (editors), Summan,' of joint research
on the diets of northern fur seals and fish in the Bering Sea dur-
ing 1985, p. 57-92. INTWAFC 86-19. Northwest and Alaska Fish-
eries Center, National Marine Fisheries Service, NOAA, 7600 Sand
Point Way N.E., Seattle, WA 98115.
eastern Bering Sea. However, Windell (1978) stated
that small fish generally consume proportionately
more food per unit weight, and some studies showed
this trend; Daan (1973) used a prey size dependent
evacuation model and found that ration decreased
with increasing fish size for North Sea Atlantic cod.
Huebner and Langton (1982) calculated daily ration
of winter flounder and found the largest fish (>300
g) had the smallest ration. Other studies (Elliott
1972; Hofer et al. 1982) showed that predator size
and meal size have little or no effects on gastric
evacuation rate. Durbin and Durbin (1980) concluded
from an extensive review of daily ration studies that
particle size and meal size relationships on gastric
evacuation rates deserve further study. Since the
>70 cm Greenland halibut in this study, which had
the highest daily ration, also consumed much larger
walleye pollock than the other predator size groups,
a particle size interaction with gastric evacuation
rate seems a likely avenue for further research.
ACKNOWLEDGMENTS
We are grateful to Sandra Noel, Karen Conlan,
and Wendy Carlson for their excellent work on the
figures. Thanks also go to Doug Milward and Geoff
Lang for their assistance on the stomach content
analyses. We also want to thank the two reviewers
for their comments on the earlier manuscript.
LITERATURE CITED
Bakkala, R. G., and V. G. Wespestad.
1983. Walleye pollock. In R. G. Bakkala and L. L. Low
(editors), Condition of groundfish resources of the eastern
Bering Sea and Aleutian Islands region in 1982, p. 1-27.
U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/N\VC-42.
Bowering, W. R., and G. R. Lilly.
1985. Diet of Greenland halibut off southern Labrador and
northeastern Newfoundland (Div. 2J-I-3K) in autumn of
1981-82, emphasizing predation on capelin. Northwest Atl.
Fish. Org. SCR Doc. 85/109, Ser. No. N1085, 16 p.
Chumakov, a. K.
1969. The Greenland halibut Reinhardtius hippoglossoides
(Walbaimi) in the Iceland area— the halibut fisheries and tag-
ging. J. Ichthyol. 9:909-912. (Engl. Transl. Vopr. Ikhtiol.)
Daan, N.
1973. A quantitative analysis of the food intake of North Sea
cod, Gadus morhua. Neth. J. Sea Res. 6:479-517.
Durbin, E. G.. and A. G. Durbin.
1980. Some factors affecting gastric evacuation rates in
fishes. Int. Counc. Explor. Sea CM. 1980/L:59.
Durbin. E. G., A. G. Durbin, R. W. Langton. and R. E.
Bowman.
1983. Stomach contents of silver hake, Merltuccius bilinearis,
and Atlantic cod, Gadtis morhua, and estimation of their
daily rations. Fish. Bull., U.S. 81:437-454.
689
FISHERY BfM.ETIN: VOL. m. NO. 4
DWYER, D. A.
1984. Feeding habits and daily ration of walleye pollock
{Theragra chalcogramma) in the eastern Bering Sea. M.S.
Thesis. Univ. Washington, Seattle, WA, 102 p.
DwYER, D. A., K. M. Bailey, and P. A. Livingston.
1987. Feeding habits and daily ration of walleye pollock
{Theragra chalcogramma) in the eastern Bering Sea, with
special reference to cannibalism. Can. J. Fish. Aquat. Sci.
44:1972-1984.
Elliott, J. M.
1972. Rates of gastric evacuation in brown trout, Salmo trut-
ta L. Freshwater Biol. 2:1-18.
Elliott, J. M., and L. Persson.
1978. The estimation of daily rates of food consumption for
fish. J. Anim. Ecol. 47:977-991.
Flowerdew, M. W., and D. J. Grove.
1979. Some observations on the effects of body weight,
temperature, meal size and quality on gastric emptying time
in the turbot, Sc&phthalmus maximus (L.) using radiography.
J. Fish Biol. 14:229-238.
Halliday, K., and Y. Umeda.
1986. Data Report: 1984 bottom trawl survey of the eastern
Bering Sea continental shelf. U.S. Dep. Commer., NOAA
Tech. Memo. NMFS F/NWC-108, 203 p.
Hart, J. L.
1973. Pacific fishes of Canada. Bull. Fish. Res. Board Can.
180, 740 p.
Haug, T., and B. Gulliksen.
1982. Size, age, occurrence, growth, and food of Greenland
halibut, Reinhardtius hippoglossoides (Walbaum) in coastal
waters of western Spitzbergen. Sarsia 68:293-297.
Hofer, R., H. Forstner, and R. Rettenwander.
1982. Duration of gut passage and its dependence on tem-
perature in roach, Rutilus rutilus L: Laboratory and field
experiments. J. Fish Biol. 20:289-299.
HuBBS, C. L., and N. J. Wilimovsky.
1964. Distribution and synonymy in the Pacific Ocean, and
variation, of the Greenland halibut, Reinhardtius hippo-
glossoides (Walbaum). J. Fish. Res. Board Can. 21:1129-
11.54.
Huebner, J. D., and R. W. Langton.
1982. Rate of gastric evacuation for winter flounder, Pseudo-
pleuronectes americanus. Can. J. Fish. Aquat. Sci. 39:
356-360.
Livingston, P. A., D. A. Dwyer, D. L. Wencker, M. S. Yang,
and G. M. Lang.
1986. Trophic interactions of key fish species in the eastern
Bering Sea. Int. North Pac. Fish. Comm. Bull. 47:49-65.
Mikawa, M.
1963. Ecology of the lesser halibut, Reinhardtius hippo-
glossoides Matsuurae Jordan and Snyder. Bull. Tohoku
Reg. Fish. Lab. 29:1-41.
MiNET, J. P., AND J. B. PeRODOU.
1978. Predation of cod, Gadus morhua, on capelin, Mallotus
villosus, off eastern Newfoundland and in the Gulf of St.
Lawrence. Int. Comm. Northwest Atl. Fish. Res. Bull.
13:11-20.
MiTO, K.
1974. Food relationships among benthic fish populations in
the Bering Sea. M.S. Thesis, Hokkaido Univ., Hokkaido,
Jpn.
Moiseev, p. a.
1953. Treska i Kambaly dalnevestochnykh morei (Cod and
flounders of Far-Eastern seas). [In Russ.) Izv. Tik-
hookean. Nauchno-Issled. Inst. Rybn. Khoz. Okeanogr.
40:1-287. (Transl. by Transl. Bur. Can. Dep. Seer. State,
576 p., available as Fish. Res. Board Can. Transl. Ser. 1 19.)
NlSKA, E. L., AND A. R. Magill.
1967. Occurrence of (Greenland halibut and Asiatic flounder
off Oregon. Fish. Comm. Oreg., Res. Brief 13:123-124.
Persson, L.
1984. Food evacuation and models for multiple meals in
fishes. Environ. Biol. Fishes 10:305-309.
Schmidt, P. J.
1934. On the zoogeographical distribution of the chief marine
food fishes in the western part of the North Pacific. Proc.
Pac. Sci. Congr. 5:3795-3799.
Shmidt, P. Y.
1950. Fishes of the Sea of Okhotsk. Acad. Sci. U.S.S.R.
Trans. Pac. Comm., Vol. VI. (Transl. from Russian by
Isarael Program for Sci. Transl., Jerusalem, 1965.)
Shuntov, V. P.
1970. Sezonnoe respredelenie chernogo i strlozubykh
paltusov v Beringovum more (Seasonal distribution of black
and arrow-tooth halibuts in the Bering Sea). [In Russ.] Tr.
Vses. Nauchno-Issled. Inst. Morsk. Rybn. Khoz, Okeanogr.
70 (Izv. Tikhookean. Nauchno-Issled. Inst. Rybn. Khoz.
Okeanogr. 72):391-401. [Transl. by Isr. Program Sci.
Transl., 1972, hi P. A. Moiseev (editor), Soviet fisheries in-
vestigations in the northeastern Pacific, Part 5, p. 397-408.
Available Natl. Tech. Inf. Serv., Springfield, VA, as TT
71-501271.]
Smith, R. L., A. C. Paulson, and J. R. Rose.
1978. Food and feeding relationships in the benthic and
demersal fishes of the Gulf of Alaska and Bering Sea. Final
Report to NOAA RU284 OCSEAP. Inst. Mar. Sci., Univ.
Alaska, p. 33-107. Vol. 1 June 1978.
VoDOPOViCH, D. S., and J. J. Hoover.
1981. A computer program for integrated feeding ecology
analyses. Bull. Mar. Sci. 31:922-925.
Westrkeim, S. J., AND F. T. Fletcher.
1966. First record of the twoline eelpout Bothrocara brun-
neum, Greenland halibut, Reinhardtius hippoglossoides, and
shortbelly rockfish, Sebastodes jordani, in British Columbia
waters. J. Fish. Res. Board Can. 23:309-312.
Windell, J. T.
1978. Digestion and the daily ration of fishes. In S. D.
Gerking (editor). Ecology of freshwater fish production, p.
159-183. John Wiley and Sons (Halsted Press), N.Y.
690
ASPECTS OF THE BIOLOGY OF TWO SCYLIORHINID SHARKS,
APRISTURUS BRUNNEUS AND PARMATURUS XANIURUS, FROM
THE UPPER CONTINENTAL SLOPE OFF SOUTHERN CALIFORNIA
Jeffrey N. Cross'
ABSTRACT
The distribution, abundance, reproductive cycle, and food habits of two scyliorhinid sharks are discussed.
Catsharks occurred on 87% of 71 longline sets and in 6% of 48 trawls. Longline catches were stratified
by habitat into banks (hard substrate) and mud (soft substrate). Apristurus brunneus occurred more
frequently on mud sets than on bank sets, but its abundance was similar in both habitats. Parmaturus
xaniuTus occurred equally frequently on mud and bank sets, but it was more abundant on bank sets.
Catches of both species consisted of adults and adolescents; juveniles were rare or absent. Historical
collections suggest that juveniles are mesopelagic.
Male P. xaniurus matured at a smaller size than msXeA. brunneus. Females of both species matured
at about the same size and fecundity increased with female size. The proportion of body weight devoted
to gonads and maximum oocyte size were greater among P. xaniurus, but fecundity and the proportion
of females carrying egg cases were greater among A. brunneus. Seasonal changes in gonadal develop-
ment were not well defined for either species. Members of both populations may have been reproduc-
tively active throughout the year.
The diets of both species comprised, in order of importance, crustaceans, teleosts, and squids. Most
prey consumed were pelagic; however, it is not known where in the water column the catsharks obtained
their prey.
The Scyliorhinidae is the largest family of living
sharks with about 94 valid species (Nelson 1984).
Commonly known as catsharks, they occur world-
wide from tropical to cold-temperate and arctic
waters from the intertidal to depths greater than
2,000 m. Little is known about the biology of most
scyliorhinid sharks despite their abundance and
widespread distribution (Springer 1979; Compagno
1984).
Apristurtis hrunneus Gilbert, the brown catshark,
occurs in the eastern Pacific Ocean from northern
British Columbia, Canada, to northern Baja Califor-
nia, Mexico, and perhaps south to Central America
and Peru. It is common on the outer continental
shelf and upper slope off British Columbia, Wash-
ington, and northern California (Springer 1979) but
is considered uncommon off central and southern
California (Miller and Lea 1972). DeLacy and Chap-
man (1935) and Cox (1963) described its egg case.
Jones and Geen (1977) made observations on its
distribution, reproduction, and food habits in British
Columbia waters.
^Southern California Coastal Water Research Project, 646 W.
Pacific Coast Highway, Long Beach, CA 90806.
Parmaturus xaniurus Gilbert, the filetail cat-
shark, occurs in the eastern Pacific Ocean from cen-
tral California to the Gulf of California, Mexico. It
is fairly common on the outer continental shelf and
upper slope (Compagno 1984). Cox (1963) described
its egg case. Lee (1969) reported that juveniles were
captured by midwater nets in the Santa Barbara
basin off southern California. Springer (1979) re-
ported that P. xaniurus were observed eating mori-
bund lanternfishes (Myctophidae) at the bottom of
the oxygen-poor Santa Barbara Basin.
The objective of this study was to increase the
knowledge of the life histories of A. brunneus and
P. xaniurus by analyzing data on the distribution,
abundance, reproduction, and food habits of these
species collected during a survey of the fishes of the
upper continental slope off southern California
(Cross 1987).
MATERIALS AND METHODS
Fishes occurring on or near the bottom between
290 and 625 m were collected by otter trawl and
longline. Forty-eight trawls were made between
November 1981 and August 1983 (Fig. 1). A single
warp semiballoon trawl with 7.6 m headrope, 8.8 m
Manuscript accepted June 1988.
FISHERY BULLETIN: VOL. 86, NO. 4, 1988.
691
FISHERY BULLETIN: VOL. 86, NO. 4
PT. DUME
1 18'00'
LOS ANGELES
34'00'
0 5 2°^^ ^°
KM
• OTTER TRAWL
LONGLINE
NEWPORT BEACH
DANA PT.
Figure 1.— Map of the study area.
footrope, 4.1 cm (stretched) body mesh, and 1.3 cm
(stretched) cod end liner was towed along an isobath
at approximately 2.5 knots for 10 minutes.
Seventy-one trips were made with commercial
longline fishermen between June 1983 and Novem-
ber 1984 (Fig. 1). A unit (tub) of longline gear con-
sisted of about 650 m of groundline bearing approx-
imately 600 hooks (4/0 and 5/0 standard rockcod) on
short leaders. Salted pieces of Engraulis mordax
and, to a lesser extent, Scomber japonictis were used
as bait. Three to six tubs tied together formed a set.
Usually, lines were set between the hours of 1000
and 1400 and retrieved the following day between
0600 and 1000.
Sinking and floating longlines were set. On sink-
ing sets, weights (bricks) were tied to either end of
the groundline and at intervals along the line. On
floating sets, weights and floats (soda bottles) were
tied alternately to groundline; the distance between
two weights encompassed 50-60 hooks. Anchors
and buoy lines were attached to each end of the
groundline. Floating and sinking lines were set on
mud, but only floating lines were set on banks. Sets
on the mud ranged from 400 to 600 m deep; sets
on the banks ranged from 350 to 550 m deep.
Scyliorhinid sharks were taken to the laboratory
where they were measured to the nearest 1 mm total
length (TL) and weighed to the nearest 0.1 g. The
left clasper of males was measured to the nearest
0.5 mm. The gonads were removed and weighed to
the nearest 0.1 g. All eggs larger than 4 mm in diam-
eter were separated from the ovary and measured
to the nearest 1 mm. Stomachs were removed and
placed in 10% buffered formalin; the contents were
692
CROSS: BIOLOGY OF SCYLIORHINID SHARKS
washed in water, sorted, and identified to tlie lowest
taxon practicable; pieces of bait were ignored.
Size at sexual maturity was estimated for males
by change in relative size of the claspers and testes,
and for females by change in relative size of the
ovaries and presence of egg cases in the oviducts
(Pratt 1979). Polynomial regressions were fit to the
data for males (SAS 1982). The reproductive season
was determined by enlarged testes in the males and
the presence of full-sized eggs in the ovaries of the
females. The gonadosomatic index (GSI) was cal-
culated for the sexes of both species from
GSI = (gonad weight/body weight) x 100.
Analyses of stomach data were summarized by the
index of relative importance (IRI) modified from
Pinkas et al. (1971):
IRI = PO{PN + PW)
where PO is percent occurrence, PN is percent
numbers, and PW is percent weight calculated for
each prey category. The length of intact prey was
measured.
Catch/effort data (where a unit of effort was one
tub of line) were transformed to logjo and analyzed
for habitat, season, and depth differences by analysis
of covariance (ANCOVA) for unbalanced designs
(SAS 1982) with habitat and season as the main ef-
fects and depth as the covariate. Catch per tub for
each positive set was determined by averaging the
catches of the constituent tubs.
Fish size data were analyzed for habitat, season,
and depth differences by ANCOVA for unbalanced
designs. The data were not transformed because size
was approximately normally distributed.
Geometric mean weight-length regressions were
calculated from the logarithmic transformation of
W = aL^
where W is weight in grams, L is total length in
millimeters, and a and b are fitted constants (Ricker
1973). The regression coefficients (b) were compared
by the method of Clarke (1980).
The sediments of the upper continenal slope off
Newport Beach are predominantly green silty clays.
Sand content is fairly constant down slope (mean
= 12% by dry weight); areas around the offshore
banks and the shoulders of the submarine canyons
are sandier (25-50% by dry weight). Organic con-
tent increases from 5 to 7% (as total volatile solids)
at 290 m to 11-14% at 625 m (SCCWRP 19832).
Between 600 and 700 m, the slope gives way to the
low-oxygen San Pedro Basin to the northwest and
to the deeper San Diego Trough to the southeast
(Fig. i).
Longline fishermen recognize two habitats on the
slope: hard substrate banks and soft, relatively
featureless (on a fathometer) mud bottom. Surface
sediments on the banks are a mixture of coarse sand,
shell hash, and occasional rocks. As used herein,
banks include submerged mountains, shoulders of
submarine canyons, and isolated mounds as small
as a few hundred meters across and 20-30 m high.
The mud bottom is green silty clay and is the pre-
dominant habitat on the slope.
Oceanographic measurements in the water column
off Newport Beach showed weak and decreasing
gradients with increasing depth. The mean annual
temperature was 8.3°C (SD = 0.3, N = 64, min =
7.5, max = 9.1) at 300 m and 6.5°C (SD = 0.2, N
= 25, min = 6.0, max = 6.9) at 500 m. Mean an-
nual dissolved oxygen was 1.21 ppm (SD = 0.26, A''
= 54, min = 0.76, max = 1.94) at 300 m and 0.48
ppm (SD = 0.10, N = 20, min = 0.31, max = 0.72)
at 500 m. Some of the variation at 300 m was the
result of seasonal changes related to upwelling. In
the spring, temperature and dissolved oxygen
decreased, and salinity and density increased
(SCCWRP fn. 2).
RESULTS
Distribution and Abundance
The occurrence of scyliorhinid sharks in trawl
catches was markedly different from longline
catches. Catsharks were caught in 3 (6%) of the 48
otter trawls. The six individuals collected accounted
for <0.1% of all fish caught in trawls. Catsharks
were caught on 62 (87%) of the 71 longline sets (212
tubs of gear examined). The 698 individuals collected
accounted for 5.8% of the fish caught (2.8% of catch
weight) on longlines.
Apristurus brunneus were caught on 50 (70%)
sets; 475 individuals were collected. Parmaturus
xaniurus were caught on 53 (75%) sets; 223 in-
dividuals were collected. The two species occurred
independently on the tubs of longline gear (x^ =
0.39, P > 0.05). They were equally abundant on bank
^SCCWRP. 1983. A survey of the slope off Orange County,
California. Report to Countj' Sanitation Districts of Orange Coun-
ty. Long Beach: Southern California Coastal Water Research
Project, 208 p.
693
FISHERY BULLETIN: VOL. 86, NO. 4
sets {t = 0.56, P > 0.05), but A. brunneus were more
abundant than P. xaniurus on mud sets (t = 3.50,
P < 0.01) (Table 1).
Apristurus brunneus occurred on 61% of the bank
sets and 79% of the mud sets. Catches were not
significantly different among habitats, seasons, or
depths (ANCOVA, P > 0.05). Mean catch per tub
in the mud habitat was not significantly different
between floating and sinking sets {t = 1.35, P =
0.19).
Parmaturus xaniurus occurred on 73% of the
bank sets and 76% of the mud sets. Catches were
significantly higher on bank sets than on mud sets
(ANCOVA, P < 0.05). Catches were nearly twice as
high on bank sets in the winter compared with bank
sets in the summer. Catch increased with increas-
ing depth on banks, but it decreased with increas-
ing depth on mud. Catches on the mud were not
significantly different between floating and sinking
sets {t = 0.76, P = 0.45.).
Size
There were no significant differences in the
regression coefficients of the weight-length relation-
ships between males and females of either A. brun-
neus or P. xaniurus over the range of sizes ex-
amined (Table 2; Fig. 2). There were no significant
differences in the size of A. brunnetis or P. xaniurus
among habitats, seasons, or depths (ANCOVA, P
> 0.05).
Table 1 .—Catch statistics for Apristurus brunneus and Parmaturus
xaniurus from longline collections on banks and mud. No^ =
number, wt = weight in kg, N = number of positive sets, X =
mean, SD = one standard deviation.
Banks
Mud
Species
N
X
SD
N
X
SD
A. brunneus
No./tub
20
2.3
2.1
30
3.7
4.6
Wt/tub
20
0.9
0.8
30
1.5
1.8
P. xaniurus
No./tub
24
2.0
2.0
29
1.1
1.1
Wt/tub
24
0.6
0.6
29
0.4
0.4
Reproduction
Based on relative change in clasper length and
gonad weight, A. brunneus males reached sexual
maturity between 450 and 500 mm TL; P. xaniurus
males reached sexual maturity between 375 and 425
mm TL (Fig. 3). Females of both species reached
sexual maturity between 425 and 475 mm TL (Fig.
4). Only the right ovary was functional in both
species.
Seasonal changes in gonadal development were
not pronounced among A. brunneus. Male and
female GSIs were highest in winter and lowest in
summer (Fig. 5). Large oocytes were present in the
right ovary of females throughout the year (Fig. 6).
700-
/
MALES y/
^-. 600-
FEMALES /
c/)
y /
2
y /
O 500-
y /
^-^
y yy^
^ 400-
y yy
y yy
O
y yy
y y
UJ 300-
y
^
y ^
y y^
200-
y ^y^^
y ^'^^
y ^
100-
350 400 450 500 550 600 650
800-
— 1 1
/
/
700-
/
/
2 600-
O
^ 500-
O 400-
LiJ
$ 300-
y^
^
^
/y
. y
/y
/y
200-
100-
ji ,
_
300 350 400 450 500 550 600
TOTAL LENGTH (MM)
Figure 2.— Weight-length relationships oi A-pristurus brunneus
(above) and Parmaturus xaniurus (below).
Table 2.— Geometric mean weight-length regressions for Apristurus brunneus and
Parmaturus xaniurus. N - sample size, min = minimum total length (TL) in mm,
max = maximum TL, W = body weight in g, /. = TL, LI = lower 95% confidence
interval for b, L2 = upper 95% confidence interval, r = correlation coefficient.
Species Sex N min max
W = aL^
LI
L2
A. brunneus
P. xaniurus
F
M
F
M
149
90
76
89
369
389
307
325
556
625
574
516
W = 2.379 X
W = 3.577 X
W = 9.377 X
W = 3.163 X
10"® ; 3 059
2.839
2.809
3.045
3.166
3.279
3.134
3.439
3.688
0.899
0.966
0.965
0.934
694
CROSS: BIOLOGY OF SCYLIORHINID SHARKS
§ 50-
f 40
O
y 30
20
10
a:
Q.
•
• •
|l|_-^
•
MW^*
\
yr^
••
9/
;>•
•
A_v*«
^-^i
^ 4.
CO
o
3-
2-
••Vt. ••
^^-•1 •*
•
1-
n-
i
400 450 500 550 600 650
TOTAL LENGTH (MM)
350 400 450 500
TOTAL LENGTH (MM)
Figure 3.— Left clasper length and gonadosomatic index (GSI) versus total length of male Apristurus brunneus
(left) and Parmaturus xaniurus (right).
.%?
. .s
-••-• — •\
20-
iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii .
<
Q
O
O
UJ
15-
•••?«?**• ••
10-
5-
n-
•
•
•
/
J,*
25-
20-
is-
le-
Illllllllllllllllllllllllllll.
•
s'
n-
3S0 400 4S0 SOO 550
TOTAL LENGTH (MM)
325 375 425 475 525
TOTAL LENGTH (MM)
575
Figure 4.— Number of oocytes (>4 mm) and maximum oocyte diameter versus total length of female Apristurus
b7~unneus (left) and Parmaturus xaniurus (right). Cross-hatched bar indicates size of fish carrying egg cases.
695
FISHERY BULLETIN; VOL. 86, NO. 4
NDJFMAMJJASON
MONTH
Figure 5.— Mean gonadosomatic index (GSI) otApristurus bruyi-
neus males (above) and females (below) from November 1983
through November 1984. Cross-hatched bar indicates months when
females were carrying egg cases. Vertical bars are 95% confidence
intervals.
The number of oocytes larger than 4 mm was cor-
related with size among mature females (r = 0.425,
P < 0.001); the maximum number was 29. Atretic
oocytes were observed from March through Novem-
ber and spent fish (no oocytes larger than 2-3 mm;
N = 2) were collected in December and July.
Egg cases were found in the oviducts of 48 A.
brunneus females larger than 450 mm (37% of
mature females); the egg cases were not complete-
ly formed in only one individual. Females with egg
cases were collected every month except February
and August (Fig. 5). More mature females carried
egg cases from December through May (42%) than
from June through November (29%), although the
difference was not significant {)c = 2.24, P > 0.1).
Egg case length (measured between the tendrils)
was correlated with female total length (r = 0.386,
P < 0.02). The mean ratio of egg case length to
female total length was 0.114 (SD = 0.019).
Seasonal changes in gonadal development were
not pronounced among P. xaniurus. Male GSI was
high throughout the year; female GSI was highest
in winter and lowest in summer (Fig. 7). Large
5 10 15 20
OOCYTE DIA (MM)
Figure 6.— Mean monthly oocyte (>4 mm) size-
frequency distribution of Apristuriis brunneiis
from November 1983 (bottom) to November
1984 (top). Sample size in parentheses.
696
CROSS: BIOLOGY OP^ SCYLIORHINID SHARKS
^ 2-
CO
o
1
0
<
^4-H.^,
^o
NOV
(20)
^
CD
U1
7-
.
6-
•
5-
4-
<
/
\
•
3-
2-
t
/
\
/
\
/
<
/' :
>
1-
<
►^^
/
/
^^<
>
nJ
iriiiiiiMiM
III
III
iiiiiii
DJFMAMJJASON
MONTH
Figure 7.— Mean gonadosomatic index (GSI) of Parmaturus
xaniurus males (above) and females (below) from December 1983
through November 1984. Cross-hatched bar indicates months when
females were carrying egg cases. Vertical bars are 95% confidence
intervals.
oocytes were present in the right ovary of females
every month except July (Fig. 8). The number of
oocytes larger than 4 mm was correlated with size
among mature females (r = 0.597, P < 0.001); the
maximum number was 23. Atretic oocytes were ob-
served in May, August, and November and spent
fish (no oocytes larger than 3-4 mm; A^ = 3) were
collected in January and March.
Egg cases were present in the oviducts of five P.
xaniurus females larger than 470 mm (10% of
mature females); the egg cases were not complete-
ly formed in one individual. One 450 mm TL female
had recently released two egg cases as evidenced
by distended oviducts. Egg cases were present in
females collected in December, January, August,
and November (Fig. 7). The mean ratio of egg case
length to female total length was 0.156 (SD =
0.023).
Food Habits
Stomachs from 211 A. brunneus were examined;
138 (65%) were empty or contained only traces of
OOCYTE DIA (MM)
Figure 8.— Mean monthly oocyte (>4
mm) size-frequency distribution oi Par-
maturus xaniurus from December
1983 (bottom) to November 1984 (top).
Sample size in parentheses.
well-digested prey. There was no seasonal trend in
the proportion of empty stomachs. Crustaceans,
teleosts, and molluscs constituted the diets of the
remaining individuals (Table 3). Natantian decapods
697
FISHERY BULLETIN: VOL. 86, NO. 4
Table 3.— Stomach contents of 73 Apristurus brunneus (mean size = 496
mm TL, SD = 40, min = 396, max = 619). PO = percent occurrence, PN
= percent number, PW = percent weight, IRI = index of relative importance.
PO
PN
PW
IRI
Crustacea
97.3
74.2
67.3
13,768
Isopoda
9.6
3.0
0.8
36
Epicaridea
9.6
3.0
0.8
36
Mysidacea
1.4
0.5
0.4
1
Euphausiacea
6.8
4.5
4.5
61
Decapoda
71.2
61.8
58.4
8,558
Natantia
61.6
52.7
48.9
6,259
Penaeidea
13.7
7.5
6.6
193
Sergestidae
12.3
5.6
4.7
127
Sergestes similis
11.0
4.9
3.9
97
Petalidium suspiriosum
1.4
0.7
0.8
2
Penaeidae
2.7
1.9
1.8
10
Bentheogennema sp.
2.7
1.9
1.8
10
Caridea
45.2
23.7
28.8
2,373
Pasiphaeidae
45.2
23.7
28.8
2,373
Pasiphaea sp.
27.4
14.0
16.3
830
Pasiphaea pacifica
13.7
5.1
8.3
184
Pasiphaea emarginata
1.4
1.4
1.4
4
Reptantia
15.1
6.6
7.8
217
Anomura
15.1
6.6
7.8
217
Galatheidae
15.1
6.6
7.8
217
Pleuroncodes planipes
5.5
1.9
2.0
21
Mollusca
20.5
8.4
11.4
406
Pelecypoda
1.4
0.3
0.1
<1
Cephalopoda
19.2
8.0
11,2
369
Coleoidea
19.2
8.0
11.2
369
Decapoda
19.2
8.0
11.2
369
Teuthoidea
19.2
8.0
11.2
369
Gonatus sp.
4.1
2.5
3.3
24
Abraliopsis felis
2.7
0.8
1.0
5
Loligo opalescens
1.4
0.5
0.8
2
Osteichthyes
42.5
16.6
21.0
1,598
tVlyctophidae
6.8
3.8
4.4
56
(40% of total IRI), especially carideans of the family
Pasiphaeidae (15%), dominated the contents. Tele-
osts (10%), including myctophids, were also impor-
tant. Most fish remains in the stomachs were
digested beyond recognition. Squids made up the re-
mainder of the diets (3%). There was no evidence
for a size-related change in diet; the dominant prey
occurred in all sizes of fish examined.
The stomachs of 73 A. brunneus contained an
average of 2.7 prey items (SD = 1.7, max = 10) and
the contents averaged 0.7% of body weight (SD =
0.4, max = 1.99). Content weight was not correlated
with body weight (Spearman r^ = 0.227, P > 0.2),
and carapace length of the most frequently occur-
ring prey, Pasiphaea spp., was not correlated with
fish size (r, = -0.168, N = 3S, P > 0.2). Relative
content weight was not significantly different
among months (Kruskal-WaUis test, H = 14.89, P
= 0.19).
Stomachs from 155 P. xaniurus were examined;
85 (55%) were empty or contained only traces of
well-digested prey. There was no seasonal trend in
the proportion of empty stomachs. Crustaceans,
teleosts, and molluscs constituted the diets of the
remaining individuals (Table 4). Reptantian deca-
pods (36% of total IRI), particularly the galatheid
Pleuroncodes planipes (12%), dominated the con-
tents. Natantian decapods (3%) and teleosts (7%),
including myctophids, were also important. Most
fish remains in the stomach were digested beyond
recognition. Squids made up a small part of the diets
(<1%). There was no evidence for a size-related
change in diet; the dominant prey occurred in all
sizes of fish examined.
The stomachs of 70 Parmaturus xaniurus con-
tained an average of 2.4 prey items (SD = 1.3, max
= 5) and the contents averaged 1.2% of body weight
(SD = 0.9, max = 4.7). Content weight was corre-
lated with body weight (Spearman Vg = 0.372, P <
0.002). Relative content weight was not significantly
different among months (Kruskal-Wallis test, H =
13.26, P = 0.35).
698
CROSS; BIOLOGY OV SCYLIORHINID SHARKS
Table 4.— Stomach contents of 70 Parmaturus xaniurus (mean size = 438
mm TL, SD = 47, min = 341, max = 547). PO = percent occurrence, PN
= percent number, PW = percent weight, IRI = Index of relative importance.
PO
PN
PW
IRI
Crustacea
97.1
75.5
72.1
14,332
Isopoda
4.3
1.6
0.5
9
Epicaridea
4.3
1.6
0.5
9
Bopyridae
2.9
1.0
0.3
4
Munidon parvum
2.9
1.0
0.3
4
Euphausiacea
24.3
8.6
5.4
340
Decapoda
81.4
56.5
59.2
9,421
Natantia
21.4
10.8
11.9
486
Penaeidea
11.4
4.6
4.0
98
Sergestidae
11.4
4.6
4.0
98
Sergestes similis
11.4
4.6
4.0
98
Caridea
12.9
5.0
7.0
155
Pasiphaeidae
12.9
5.0
7.0
155
Paslphaea pacifies
10.0
4.0
5.8
98
Reptantia
67.1
38.5
43.4
5,495
Anomura
65.7
38.2
43.0
5,335
Galatheidae
65.7
38.2
43.0
5,335
Pleuroncodes planipes
40.0
21.9
26.1
1,920
Brachyura
1.4
0.4
0.4
1
Mollusca
5.7
1.4
1.8
18
Cephalopoda
5.7
1.4
1.3
18
Coleoidea
5.7
1.4
1.8
18
Decapoda
5.7
1.4
1.8
18
Teuthoidea
5.7
1.4
1.8
18
Osteichthyes
48.6
20.8
25.3
1,036
Myctophidae
10.0
4.8
6.1
109
DISCUSSION
Distribution and Abundance
Apristurus hrunneus and Parmaturus xaniurus
were a common, though unwanted, part of the long-
line catch on the upper continental slope off south-
ern California. The abundance of both species was
underestimated by trawl. Among the 29 species
(12,074 individuals) caught on longlines, A. brunneus
ranked 7th in abundance and P. xaniurus ranked
10th. They ranked 29th and 34th, respectively,
among the 42 species (7,264 individuals) taken in
trawls (Cross 1987). The bias of small trawls against
large demersal fishes is well known (Day and Pearcy
1968; Haedrich et al. 1975). Most previous fish col-
lections on the slope off southern California were
taken with small trawls which explains why these
sharks are not considered common.
Catches of A. hrunneus and P. xaniurus were
similar on bank sets, but A. hrunneus was more
abundant on mud sets. The two species occurred in-
dependently at the scale of one tub of longline gear
(about 650 m). Apristurus brunneus was equally
abundant on mud and bank sets, and seasonal dif-
ferences in distribution and catch were not appar-
ent. Parmaturus xaniurus was more abundant on
bank sets than on mud sets, suggesting some habitat
selection. Catches of P. xaniurus were highest on
banks in the winter; the reason for this is not known.
Juvenile A. brunneus and P. xaniurus were con-
spicuously absent from longline and trawl collections
on the slope. Catsharks are generally regarded as
demersal fishes (Compagno 1984), but A. brunneus
and P. xaniurus have been captured in the water
column. An undisclosed number of A. brunneus
larger than 260 mm (TL assumed) were collected
up to 172 m above the bottom in 373 m of water off
British Columbia, Canada (Jones and Geen 1977).
Sixty-nine P. xaniurus (99-320 mm) were collected
in 43 midwater trawls from 9 to 490 m above the
bottom in 527-582 m of water in the Santa Barbara
Basin, CA (Lee 1969). The livers of P. xaniurus con-
tain a high proportion of squalene, a low specific
gravity oil that aids in hydrostatic balance (Springer
1979).
Juveniles and adolescents of both species were
taken in midwater trawls in the Santa Barbara
Basin (bottom depths between 490 and 576 m)
(UCSB3). Catsharks occurred in 31 (41%) of 75 mid-
^UCSB. Collections taken by the University of California, Santa
Barbara, with an opening and closing net between 1965 and 1967
and deposited in the Los Angeles County Museum of Natural
History.
699
FISHERY BULLETIN: VOL. 86, NO. 4
water trawls; 83% of the individuals were collected
within 250 m of the bottom. Fifty-four A. brunneus
(99-380 mm TL, median = 177) were collected in
22 trawls and 23 P. xaniurm (1 10-229 mm TL, me-
dian = 175) were collected in 17 trawls.
The occurrence of juveniles and adolescents of
both species in midwater collections suggests that
the water column, especially within 200-300 m of
the bottom, is their nursery area, and would explain
their conspicuous absence in benthic collections dur-
ing the present study. Caillief also concluded that
P. xaniums are mesopelagic as juveniles and demer-
sal as adults.
Interestingly, 119 midwater trawls in the Santa
Cruz Basin, CA and near Rodriguez Seamount (most
bottom depths between 1,200 and 2,200 m) (UCSB
fn 3) captured no scyliorhinid sharks (the trawls
were taken concomitantly with trawls in the near-
by Santa Barbara Basin). Bottom depths >1,000 m
may be beyond the range of both species. The deep-
est recorded collection of A. brunneus off southern
California was 933 m (Roedel 1951); the deepest
recorded collection of P. xaniums was 687 m
(Springer 1979). The absence of both species in mid-
water trawls in the Santa Cruz Basin and near
Rodriguez Seamount would not be surprising if
juveniles do not travel far horizontally and adults
do not occur below 1,000 m.
Size
The largest A. brunneus collected during the pres-
ent study (625 mm TL male) was less than the max-
imum recorded size (680 mm TL; Compagno 1984).
The largest P. xaniurus (574 mm TL female) was
greater than the maximum recorded size (550 mm
TL; Compagno 1984). Weight-length relationships
of males and females of both species were similar.
It is not known if males and females that were the
same size were the same age. Attempts at deter-
mining the ages of P. xaniurus were unsuccessful
(Cailliet 1986^).
Reproduction
Like many scyliorhinid sharks, A. brunneus and
P. xaniurus exhibit single oviparity: one fertilized
Cailliet, G. M. 1981. Ontogenetic changes in the depth distri-
bution and feeding habits of two deep-dwelling demersal fishes off
California: sablefish and filetail cat sharks. [Abstr.] Am. See.
Ichthyol. Herpetol. Sixty-first Annu. Meeting, Corvallis, OR.
''G. M. Cailliet, Moss Landing Marine Laboratories, Moss Land-
ing, CA, pers. commun. July 1986.
egg enters each oviduct and, after a short period,
is deposited in a tough egg case on the substrate
where it is anchored by tendrils (Nakaya 1975). Em-
bryonic development takes place largely outside the
mother and may require a year to produce a hatch-
ling (Compagno 1984).
Size at sexual maturity estimated during the pres-
ent study agrees with published observations for
both species (Jones and Geen 1977; Compagno
1984). Male P. xaniurus matured at a smaller size
than male A. brunneus. Females of both species
matured at about the same size and fecundity in-
creased with female size. Fecundity was greater
among A. brunneus as was the proportion of mature
females carrying egg cases, but the proportion of
body weight devoted to gonads and maximum
oocyte size were greater among P. xaniurus.
Seasonal changes in gonadal development were
not well defined for either species; individuals in
both populations may be sexually active at any time
of the year. Several observations suggest that
oocyte production was seasonal: the highest propor-
tion of adult females with oocytes <10 mm in diam-
eter and the highest frequency of atretic oocytes
were observed from summer through fall when GSI
was lowest.
There are few published observations on reproduc-
tive cycles of scyliorhinid sharks. Scyliorhinus
canicula, an abundant shelf and upper slope cat-
shark in the northeastern Atlantic Ocean, lays eggs
throughout the year. Seasonal maxima in egg cap-
sule production are apparent, but timing varies with
latitude. Size at sexual maturity also varies with
latitude; fish mature at a larger size at higher
latitudes (Ford 1921; Capape 1977). The fecundity
of female S. canicula (46-50 cm TL) ranges from
23 to 34 eggs; fecundity increases with fish size.
Eggs are 16 mm in diameter at ovulation. Annual
fecundity is about 96-115 eggs. In the Mediterra-
nean Sea, egg capsule incubation times range from
180 days for eggs deposited during the summer to
285 days for eggs deposited during the winter
(Capape 1977).
Food Habits
Apristurus brunneus and Parmaturus xaniurus
consumed, in order of importance, crustaceans, tele-
osts, and squids. Similar diets were reported for
both species by Jones and Geen (1977), Cailliet
(1981, see fn. 4), and Compagno (1984). The diets
of the two catsharks were broadly similar except for
the occurrence of crustaceans. Reptantian decapods
700
CROSS: BIOLOGY OF SCYLIORHINID SHARKS
dominated the diet of A. brunnetis (36% of total IRI)
while natantian decapods dominated the diet of P.
xaniurus (40% of total IRI). This may be a result
of habitat differences between the catsharks rather
than prey selection. The epicaridean isopods in the
stomach contents are crustacean parasites, and
were probably ingested with their hosts.
A very high percentage of A. brunneus and P.
xaniurus stomachs was empty. This may not be
representative of their respective populations.
Sharks caught by baited hooks have a higher pro-
portion of empty stomachs, and lower stomach
content weight, than sharks caught by gill net
(Medved et al. 1985).
Most of the prey consumed by A. brunneus and
P. xaniurus are pelagic. Demersal adult P. xaniurus
eat mostly crustaceans (including pelagic pasiphaeid
and sergestid shrimps), fishes (primarily mycto-
phids), and cephalopods. Mesopelagic juveniles eat
more and smaller pelagic pasiphaeid, euphausiid,
and sergestid shrimps and cephalopods, and fewer
fishes (Cailliet fn. 4). Similar pelagic prey also
dominate the diets of Galeus melastomus and
Apristurus spp., common outer shelf-upper slope
scyliorhinid sharks of the northeastern Atlantic
Ocean (Orsi and Wurtz 1977; Mattson 1981; Mauch-
line and Gordon 1983). Ontogenetic changes in diet
were reported for Scyliorhinus canicula: the impor-
tance of crustaceans declines and the importance of
teleosts increases with increasing fish size (Capape
1974).
Deep-sea benthic fishes can obtain pelagic prey by
1) feeding on prey whose vertical distribution ex-
tends close to the bottom [P. xaniurus were ob-
served eating moribund myctophids in the bottom
of the Santa Barbara Basin (Springer 1979)]; 2)
migrating into the water column to feed [A. brun-
neus and P. xaniurus are captured in midwater
(Jones and Geen 1967; Lee 1969; UCSB fn. 3)];
3) feeding on carcasses that sink to the bottom (mud
and terrestrial plant debris were found in several
catshark stomachs during the present study, and
both species took dead bait); or 4) feeding in the net
(not a factor in this study) (Pearcy and Ambler 1974;
Pearcy 1976; Sedberry and Musick 1978).
ACKNOWLEDGMENTS
This study was supported in part by a contract
from the County Sanitation Districts of Orange
County. Special thanks are extended to the dory
fishermen of Newport Beach for taking me on their
boats. H. Stubbs, M. Moore, T. Pesitch, and the
crews of the RV Westwind and RV Vantuna made
the trawl collections possible. D. Tsukada and J.
Laughlin diligently identified the prey organisms.
G. Cailliet, P. Klimley, J. Seigel, and S. Springer
provided helpful comments on earlier versions of the
manuscript.
LITERATURE CITED
CAFAPfi, C.
1974. Contribution a la biologic des Scyliorhinidae des cotes
tunisiennes. ll-Scyliorhimis canicula Linne, 1758: Regime
alimentaire. Ann. Inst. Michel Pacha 7:13-29.
1977. Contribution a la biologic des Scyliorhinidae des cotes
tunisiennes. l-Scyliorhinus canicula (Linne, 1758): Repar-
tition geographique et bathymetrique, sexualite, reproduc-
tion, fecondite. Buil. Off. natn. Pech. Tunisie 1:83-101.
Clarke, M. R. B.
1980. The reduced major axis of a bivariate sample. Bio-
metrika 67:441-446.
COMPAGNO, L. J. V.
1984. In FAO species catalogue. Sharks of the world. An
annotated and illustrated catalogue of shark species known
to date. Part 2. Carcharhinilbrmes. Vol. 4, Pt. 2, p. 251-
655. FAO Fish. Biol. Synop., No. 125.
Cox, K. W.
1963. Egg-cases of some elasmobranchs and a cyclostome
from California waters. Calif. Fish Game 49:271-289.
Cross, J. N.
1987. Demersal fishes of the upper continental slope off
southern California. Calif. Coop. Oceanic Fish. Invest. Rep.
28:155-167.
Day, D. S., and W. G. Pearcy.
1968. Species associations of benthic fishes on the continen-
tal shelf and slope off Oregon. J. Fish. Res. Board Can.
25:2665-2675.
DeLacy, a. C, and W. M. Chapman.
1935. Notes on some elasmobranchs of Puget Sound, with
descriptions of their egg cases. Copeia 1935:63-67.
Ford, E.
1921. A contribution to our knowledge of the life-histories
of the dogfishes landed at Plymouth. J. Mar. Biol. Assoc.
U.K. 12:468-505.
Haedrich, R. L., G. T. Rowe, and P. T. Pollonl
1975. Zonation and faunal composition of epibenthic popula-
tions on the continental slope south of New England. J.
Mar. Res. 33:191-212.
Jones, B. C, and G. H. Geen.
1977. Observations on the brown cat shark. ApiHsturus brun-
neus (Gilbert), in British Columbia coastal waters. Syesis
10:169-170.
Lee. R. S.
1969. The filetail catshark, Parmaturus xaniurus, in mid-
water in the Santa Barbara Basin off California. Calif. Fish
Game 55:88-90.
Mattson, S.
1981. The food of Galeus melastomus, Gadiculus argenteus
thori, Trisopterus esmarkii, Rhinonemus cimbrius, and
Glyptocephalus cynoglossu^ (Pisces) caught during the day
with shrimp trawl in a west-Norwegian fjord. Sarsia 66:
109-127.
Mauchline, J., and J. D. M. Gordon.
1983. Diets of the sharks and chimaeroids of the Rockwall
701
FISHERY BULLETIN: VOL. 86, NO. 4
Trough, northeastern Atlantic Ocean. Mar. Biol. (Berl.)
75:269-278.
Medved, R. J., C. E. Stillwell, and J. J. Casey.
1985. Stomach contents of young sandbar sharks, Car-
charhinus plumbeus, in Chincoteague Bay, Virginia. Fish.
Bull, U.S. 83:395-402.
Miller, D. J., and R. N. Lea.
1972. Guide to the coastal marine fishes of California. Calif.
Dep. Fish Game. Fish Bull. No. 157, 249 p.
Nakaya, K.
1975. Taxonomy, comparative anatomy and phylogeny of
Japanese catsharks, Scyliorhinidae. Mem. Fac. Fish. Hok-
kaido Univ. 23:1-94.
Nelson, J. S.
1984. Fishes of the world. 2ded. John Wiley & Sons, N.Y.,
523 p.
Orsi, L. R., and W. Wurtz.
1977. Patterns and overlap in the feeding of two selachians
of bathyal fishing grounds in the Ligurian sea. Rapp.
Comm. int. Mer Medit. 24:89-94.
Pearcy, W. G.
1976. Pelagic capture of abyssobenthic macrourid fish.
Deep-Sea Res. 23:1065-1066.
Pearcy, W. G., and J. W. Ambler.
1974. Food habits of deep-sea macrourid fishes off the Oregon
coast. Deep-Sea Res. 21:745-759.
PiNKAS, L., M. S. Oliphant, and I. L. K. Iverson.
1971. Food habits of albacore, bluefm tuna, and bonito in
California waters. Calif. Dep. Fish Game, Fish Bull. No.
152, 105 p.
Pratt, H. L.
1979. Reproduction in the blue shark, Prionace glauca. Fish.
Bull., U.S. 77:445-470.
Ricker, W. E.
1973. Linear regressions in fishery research. J. Fish. Res.
Board Can. 30:409-434.
Roedel, p. M.
1951 . The brown catshark, Apristurvs hrunneus, in Califor-
nia. Calif. Fish Game 37:61-63.
SAS Institute Inc.
1982. SAS User's Guide: Statistics. 1982 ed. SAS Institute
Inc., Gary, NC, 923 p.
Sedberry, G. R., and J. A. Musick.
1 978. Feeding strategies of some demersal fishes of the con-
tinental slope and rise off the mid- Atlantic coast of the USA.
Mar. Biol. (Berl.) 44:357-375.
Springer, S.
1979. A revision of the catsharks, Family Scyliorhinidae.
NOAA Tech. Rep. NMFS Circ. No. 422, 152. p.
702
FURTHER SUPPORT FOR THE HYPOTHESIS THAT
INTERNAL WAVES CAN CAUSE SHOREWARD TRANSPORT OF
LARVAL INVERTEBRATES AND FISH
Alan L. Shanks^
ABSTRACT
In areas of mesotides (tidal range 2 to 4 m) and narrow continental shelves (<30 km) internal waves
can transport (i.e., convey from one place to another) the larvae of coastal organisms shoreward. Research
reported here was in an area of microtides (tidal range <2 m) and a wide continental shelf (>80 km),
the South Atlantic Bight. Half of the sampled sets of internal waves were aligned parallel to shore and
probably originated at the shelf break. The higher densities of larvae and flotsam in the slicks over these
internal waves (convergence zones) than in the rippled water between shcks (divergence zones) indicates
that these waves were transporting larvae and flotsam shoreward. All nontransporting internal waves
were aligned at a sharp angle to shore and may have formed over shoals oriented perpendicular to shore.
To further test the hypothesis that internal waves can transport larvae, surface plankton were col-
lected from the waters over, in front, and behind a set of internal waves. The density of Portunus spp.
megalopae was significantly higher in waters in front of the set than behind. The average densities of
a variety of larval fish and invertebrates were significantly higher over the internal waves than in front
of the set of waves. These data indicate that internal waves can cause shoreward transport of larvae
and flotsam. Precompetent larval fish were not carried shoreward by this set of waves while competent
stages (i.e., juvenile through postflexion) were transported shoreward.
Recent papers have suggested that the planktonic
larvae of some coastal invertebrates and fish
(Shanks 1983, 1985, 1986; Jillett and Zeldis 1985;
Kingsford and Cheat 1986; Shanks and Wright
1987) as well as flotsam (e.g., an oil spill, Shanks
1987) can be transported (i.e., conveyed from one
place to another) by internal waves. As the tide ebbs
off the continental shelf or across some other sharp
change in the bottom relief (i.e., a reef or bank) a
lee wave is formed (Lee and Beardsley 1974;
Gargett 1976; Maxworthy 1979). When the tide
changes to flood this lee wave is "released" and
propagates away from its point of origin (see for ex-
ample Chereskin 1983). Most of the waves formed
at the continental shelf break propagate shoreward.
As the original internal wave moves onshore it
evolves from a solitary wave into a set of waves
(Osborne and Burch 1980). Stripes of glassy water,
slicks, are surface manifestations of currents over
the internal waves, and they delineate zones of con-
verging and downwelling currents situated between
the crest and the trough of an internal wave (Ewing
1950; LaFond 1959). The currents over the inter-
nal waves generate slicks by both perturbing small
^University of North Carolina at Chapel Hill, Institute of Marine
Sciences, 3407 Arendell Street, Morehead City, NC 28557.
surface waves and concentrating the organic sur-
face film (Ewing 1950; Gargett 1976). Buoyant flot-
sam will be carried into the convergence zone by the
surface currents, but because of the particle's buoy-
ancy they will not follow the water as it is down-
welled; the flotsam will be trapped at the surface
in the convergence and as the convergence zone
moves onshore so will the flotsam. The proposed
mechanism of larval transport suggests that any
organism which can remain at the surface in the con-
vergence zone either by swimming or other
behaviors (Shanks 1985) will, like flotsam, remain
in the slick and be transported onshore (Shanks
1983).
The conditions necessary for the production of
tidally generated internal waves (tides, sharp bot-
tom relief, and some water column density stratifica-
tion) are present in the waters adjacent to nearly
all land masses, and the surface manifestations of
internal waves have been observed from both ships
and satellites in numerous locations (Apel et al.
1975; Fu and Holt 1982; Sawyer 1983). Evidence
that internal waves may be capable of transporting
planktonic larvae onshore has been collected in the
Pacific Ocean off Southern California (Shanks 1983,
1985, 1986), in the San Juan Archipelago (Shanks
and Wright 1987), and in the waters off the North
Manuscript accepted July 1988.
FISHERY BULLETIN: VOL. 86, NO. 4, 1988.
703
FISHERY BULLETIN: VOL. 86. NO. 4
and South Islands of New Zealand (Zeldis and Jillett
1982; Jillett and Zeldis 1985; Kingsford and Choat
1986). These areas are characterized by mesotides,
i.e., tidal range 2 to 4 m (Davis 1964). The first pur-
pose of the research reported in this paper was to
test if transport occurred in the Atlantic Ocean and
in an area of microtides (tidal range <2 m, Davis
1964), the South Atlantic Bight.
To date researchers have compared the density
(no./m-) of larvae in the convergence zones (slicks)
over internal waves to the density present in the
divergence zone (ripples). Cases where the densities
of larvae were significantly higher in slicks than the
ripples were used as evidence for the proposed
hypothesis. Higher density in the convergence im-
plies that larvae tend to spend more time in the slick
than in the water between slicks. Because of the
speed at which internal waves are propagating, even
a brief residence in an internal-wave-slick could
cause appreciable shoreward transport. For exam-
ple, if a larvae spent an hour in a slick it could be
carried about 2 km (assuming an average internal
wave speed of 56 cm/s. Sawyer 1983).
Higher larval densities in the convergence zone
is only one of several predictions which should be
true if internal waves are transporting larvae shore-
ward. The second purpose of this research was to
test several additional predictions. As an internal
wave moves shoreward, larvae from in front of the
wave will be swept by the currents associated with
the wave into the convergence zone. If these larvae
can remain at the surface in the convergence, they
will be carried shoreward. Predictions include 1)
larval density in the waters in front of the set of in-
ternal waves should be significantly higher than in
the waters behind the set; 2) because larvae will
accumulate in the convergence, the observed den-
sity of larvae over an internal wave will be signifi-
cantly higher than the density in the waters in front
of the set of internal waves; and 3) a rare larval
type may be carried into an area from a distant
source in which case these particular larvae may
only be present in the waters over the internal wave
and will be concentrated in the slicks. To test these
predictions, replicate neuston net samples were col-
lected in the convergence and divergence zones over
a set of internal waves, and in the waters imme-
diately in front and behind this set of internal waves.
METHODS
The study was confined to the ocean waters at the
northern end of Onslow Bay within about 20 km of
Beaufort Inlet, NC, U.S.A. (long. 34°40'W, lat.
76°40'N; Fig. 1). Surface plankton tows from the
waters over internal waves were collected during
the summer of 1985 on 4 and 24 June, 14 and 20
July, and 21 August. On 24 June 1985 surface
plankton tows were also collected from immediate-
ly in front and behind a set of internal waves. On
14 June 1985 and 9 and 19 June 1986 samples were
collected at the surface and in the water column to
determine which taxa were exclusively neustonic in
distribution.
Winds stronger than 10 to 15 knots obliterate the
slicks, which delineate the convergence zones over
internal waves. Searches for internal-wave-slicks
and sampling of the associated plankton were
limited to periods with winds less than a moderate
breeze (Beaufort scale 4). The procedure for locating
internal- wave-slicks was to proceed along shore until
the estuarine front associated with Beaufort Inlet
was crossed and thence perpendicular to shore un-
til we found a set consisting of at least three large
(at least 30 m wide by about 500 m long) linear slicks
separated by one to several hundred meters of
rippled water. Sets of large slicks separated by rip-
pled water are a unique surface signature of large,
usually tidally generated internal waves (Ewing
1950; LaFond 1959; Apel et al. 1975; Gargett 1976;
Fu and Holt 1982; Chereskin 1983; Sawyer 1983).
To test if a set of internal waves was capable of
transporting flotsam, surface drifters (weighted
Styrofoam cups) were released in a line perpen-
dicular to and in front of the set of internal-wave-
slicks (Shanks 1983). Prior to the release of the sur-
face drifters and immediately after all sampling was
completed the position of the first slick in the set
was determined by either compass bearings on land-
marks or with loran. From these measurements we
were able to determine the distance that the set of
waves propagated during the period of observation.
While the currents acted on these surface drifters,
3 or 4 replicate 5- to 10-min surface (<20 cm depth)
plankton tows were made in the slicks and rippled
water between slicks. These 1985 samples were col-
lected on 4 and 24 June, 14 and 20 July, and 21
August 1985. In addition, on 24 June 1985 replicate
surface plankton samples were also collected from
the water immediately in front and behind the set
of internal waves (within 200 m of the internal wave
set). Plankton samples were collected using a manta
net (Brown and Cheng 1981) with a mouth opening
of 0.95 X 0.26 m and a net mesh of 0.333 mm. A
flow meter mounted in the mouth of the net mea-
sured the volume of water filtered. Between tows
704
SHANKS: SHOREWARD LARVAL TRANSPORT
35°-
34°
78
L
76°
ATLANTIC OCEAN
Figure L— Map of the study area. The dashed hne running along shore is the 20 m contour and the dashed
line running from just to the left of Beaufort Inlet to the tip of Cape Lookout Shoals encloses the study area.
in the different habitats the manta net was cleaned
by towing for 5 to 10 minutes with the cod end re-
moved. Most of the plankton was washed from the
net by this procedure. For convenience and in order
to more accurately sample the relatively small en-
vironments studied these plankton tows were made
from a 6 m motor boat. Plankton samples were
preserved in 5% formalin.
To assess the vertical distribution of larval types,
surface and oblique plankton tows were made on 14
June 1985 and 9 June 1986. On 19 June 1986 rep-
licate [n = 3) surface, oblique, and bottom tows were
made. The dimensions of the oblique and bottom sled
nets were 0.55 x 0.55 m and 0.80 x 0.45 m respec-
tively. Both nets had mesh of 0.333 mm, and flow
meters on the nets were used to measure the volume
of water filtered. The procedure for the oblique tows
was as follows: with the boat moving slowly forward
the net was lowered rapidly to near the bottom
(about 10 to 20 m) and then, maintaining a wire
angle of about 60°, the net was hauled slowly to the
surface. The bottom tows were made by lowering
the sled straight to the bottom with the boat station-
ary; as the boat moved slowly forward, the line was
paid out until a scope of 1 to 3 was achieved; tows
lasted 5 to 10 minutes; at the end of the tow the boat
was backed down onto the net; and the net was
pulled vertically back to the surface. Both the
oblique and bottom nets were open when they
passed through the surface and were thus contam-
inated by some surface plankton.
Organisms were sorted and identified with the aid
of a dissecting microscope. A variety of sources were
used to identify larval fish (Fritzsche 1978; Hardy
1978; Johnson 1978; Martin and Drewry 1978;
Fahay 1983). Larval fish were also sorted by devel-
opmental stage which were defined following the
terminology and descriptions in Ahlstrom et al.
705
FISHERY BULLETIN: VOL. 86, NO. 4
(1976) and Moser and Ahlstrom (1970). Megalopae
were identified from references sited in Williams
(1984). Portunus spp. and Callinectes spp. mega-
lopae were separated using the characters suggested
by Smyth (1980). Penaeids were identified using
Cook (1966). No attempt was made to precisely iden-
tify the other groups of organisms counted.
Statistical comparisons of the density of larvae
and flotsam in the different sampled habitats were
made using Wilcoxon's two-sample test (Sokal and
Rohlf 1969). Densities were considered to be sig-
nificantly different when P < 0.05.
RESULTS
During the summer of 1985 searches for internal-
wave-slicks were made on eight occasions. On three
days, despite favorable wind and sea conditions, no
internal-wave-slicks were observed. On 4 and 24
June, the observed sets of slicks were oriented
parallel to shore and the bottom topography while
on 14 and 20 July sets were oriented nearly perpen-
dicular to shore. On 21 August a set of slicks aligned
nearly perpendicular to shore was observed about
5 km offshore, and at about 9 km offshore a second
set of slicks was found oriented parallel to the bot-
tom topography. Both types of slicks, perpendicular
and parallel orientation, possessed all of the usual
characteristics of internal waves. Slicks were about
30 to 50 m wide, they were separated by one to
several hundred meters of rippled water, and the
sets moved (the parallel-to-shore slicks moved on-
shore while the perpendicular-to-shore slicks moved
north).
The surface drifters released in front of sets of
slicks oriented parallel to shore on 4 June and 21
August could not be located following the plankton
tows. On 24 June, all of the drifters located at the
end of the plankton tows (48 of 50 released) were
found in the first two slicks of the set of internal
waves. The drifters, which had been released in a
250 m long line in front of the set of internal waves,
had been caught, concentrated, and carried about
4 km shoreward by the internal waves. Clearly this
set of internal waves was capable of carrying buoy-
ant flotsam shoreward.
Sargassum floats were abundant in the plankton
tows. The floats are buoyant, do not extend above
the water (i.e., are not blown directly by the wind),
and, hence, the floats should act much like the re-
leased surface drifters. On all three dates when the
internal- wave-slicks were oriented parallel to shore,
the density of Sargassum floats was significantly
higher (14- to 300-fold higher; Tables 1, 2) in the con-
vergence than divergence zones. These data indicate
that not only was the set of internal waves sampled
on 24 June capable of carrying flotsam shoreward,
but the sets oriented parallel to shore on 4 June and
21 August were also capable of carrying buoyant
flotsam shoreward.
In contrast, on those days when the slicks were
oriented roughly perpendicular to shore (14 July, 20
July, and 21 August, Table 1) released surface
drifters were about equally distributed between slick
and rippled water and, despite the fact that the
sHcks moved northward during the observation
period, the drifters were not carried along with the
waves. The density of Sargassum floats was not
significantly higher in the slicks than the rippled
water between the slicks. These data suggest that
these internal waves were not transporting flotsam.
Before larval densities can be interpreted, it is
first necessary to determine which larval types are
found exclusively in the neuston. An increase in the
neustonic density of organisms, which inhabit both
the neuston and the water column, could be due to
forces concentrating just the neustonic portion of
the population, or it could be due to animals from
the water column augmenting the population in the
neuston. Given the sampling regime of this study
it was impossible to differentiate between these two
possibilities. Because of this limitation a series of
tows were made in the neuston, the water column,
and bottom water to determine which organisms ex-
clusively inhabited the neuston.
There was a distinct assemblage of megalopae and
larval fish that were caught in the oblique water
column and bottom plankton tows but were nearly
absent from the neuston tows (Table 3). Unfor-
tunately, organisms that were common in the
neuston tows made over or around internal waves
were abundant on only one of the three dates when
the vertical distribution samples were collected. On
this one date (19 June 1986, Table 3) there was a
group of larval fish and crabs that were only pres-
ent in the neuston tows. This latter group included
the megalopae of Portunus spp. and Callinectes spp.,
juvenile Monacanthus hispidus, and juvenile and lar-
val Hyporhamphus unifasciatus and Sphoeroides
maculatus. Previous research also suggests that
these larvae and postlarvae as well as others are in-
habitants of the surface waters. Both the behavior
(Sulkin andVan Heukelem 1981) and the vertical
distribution (Smyth 1980; Johnson 1985a) of Calli-
nectes spp. megalopae suggest that they are usually
neustonic. The megalopae of Portunus spp. and
706
SHANKS: SHOREWARD LARVAL TRANSPORT
Table 1 — The density of larval fish and invertebrates (mean no. /1 00 m + SE) in the slicks (convergence zones) and ripples (divergence
zones) over internal waves and the relative concentration of these organisms in the slicks (i.e., the ratio of slicks/ripples) over internal
waves oriented parallel (data in upper section of table) and perpendicular (data in lower section of table) to shore.
Dates when slicks were
oriented parallel to shore
1
4 June 1985
21
August 198£
)
Slicks
Ripples
Slicks/
Slicks
Ripples
Slicks/
n = 3
n = 3
ripples^
n = 3
n = 3
ripples^
Total fish
12 + 2.9
4.2 + 0.9
3*
1.7±0.9
0.8 ±0.3
2
Hypsoblennius hentzi
6.0±2.1
3.1 +0.2
2
0.1±0.1
0.4 + 0.1
0.3
Mem bras martinica
1.7+1.1
0.3 + 0.1
5
0
0
—
Hyporhamphus unlfasciatus
0.7 + 0.7
0.6 + 0.4
1
0
0
—
Brevootia tyrannus
1.6-0.8
0
—
0
0
—
Monacanthus hispidus
0.9 + 0.6
0
—
0.8 + 0.6
0
—
Chellopogen heterurus
0
0
—
0.2±0.1
0
—
Miscellaneous
1.1±0.3
0.3 ±0.2
4
0.4 + 0.4
0.2 + 0.1
2
Total Brachyura
41 + 12
1.7 + 0.6
24*
12±3.9
0.5 ±0.4
22*
Callinectes spp.
IVIegalopa
0.2 + 0.2
0
—
3.9+1.7
0.2 + 0.1
19*
First crab
1.5 ±0.6
0
*
0.2±0.2
0
—
Portunus spp.
Megalopa
36+11
1.6±0.8
23*
7.5±21
0.3 + 0.3
24*
First crab
1.2 ±0.6
0
—
0
0
—
Miscellaneous
2.0+1.1
0.1+0.1
15*
0.3 + 0.2
0
—
Amphipods
40+16
1.2 + 0.1
33*
2.2+1.5
0
—
Polychaete larvae
5.2+1.7
0.7 + 0.5
8*
0
0
—
Stomatopod larvae
0
0
—
32 + 8.8
8.1 ±2.6
4*
Cnidaria
0
0
—
9.9 + 2.9
18±6.2
1
Salpa
0
0
—
15±2.6
60 + 2.9
0.3*
Sargassum floats
11 ±6.8
0.8 ±0.6
14*
8.8 ±1.4
0.4 ±0.3
21*
Dates when slicks were oriented perpendicular to shore
14 July 1985
20 July 1985
21
August 1985
Slicks
Ripples
Slicks
Slicks/
Ripples
Slicks/
Slicks
Ripples
Slicks/
n = 3
n = 3
ripples^
n = 4
n = 4
ripples^
n = 4
n = A
ripples^
Total fish
2.6 + 0.4
0.9 + 0.5
3*
1.2 ±0.3
1.0 ±0.2
1
1.1 ±0.1
1.8±1.0
1
Hypsoblennius hentzi
0
0
—
0.6 ±0.2
0.8±0.1
1
0.3 ±0.2
1.5 ±0.9
0.2
Hyporhamphus unlfasciatus
0.5 + 0.3
0
—
0.4 + 0.2
0
—
0
0
—
Monacanthus hispidus
0
0.2 ±0.2
0
0
0
—
0.3±0.2
0.2 ±0.2
2
Chellopogen heterurus
1.0 ±0.4
0
—
0
0
—
0.2 + 0.2
0
—
Miscellaneous
1.2 + 0.4
0.4 + 0.4
2
0.1 +0.1
0.2 + 0.8
1
0.4 + 0.3
0.2 ±0.2
3
Total Brachyura
3.6±1.3
5.2±0.9
1
0.9±0.3
0.3±0.3
3
1.6±0.5
0.6 ±0.4
3
Callinectes spp.
Megalopa
0.4±0.3
0.6 ±0.6
1
0.4±0.2
0
—
0.2±0.2
0
—
Portunus spp.
Megalopa
0.9±0.6
2.2 + 0.6
0.4
0.2 + 0.2
0.2 + 0.2
3
0.1 +0.1
0.3 + 0.3
0.5
Miscellaneous
1.8±0.7
2.2±0.1
1
0.1±0.1
0.2 ±0.2
1
1.4 ±0.5
0.3 ±0.2
5
Amphipods
5.7±3.9
2.9±0.6
2
4.4 + 0.7
1.7 + 0.8
3
0
0
—
Polychaete larvae
18 + 7.3
8.1+5.2
2
0
0
—
2.6+1.3
0.3 + 0.2
8
Stomatopod larvae
0
0
—
0
0
—
9.9-1.5
9.1 ±2.6
1
Cnidaria
83+15
126 + 73
1
0
0
—
7.8+1.2
4.9+1.7
2
Salpa
91 ±36
103±78
1
0
0
—
239 ±35
130±38
2
Sargassum floats
22 ±16
3.6±1.7
6
0.4±0.1
0.1 ±0.1
5
5.2 ±2.4
1.1±0.3
5
'Also included in this category are the data from 24 June (Tables 2 and 4).
^he ratio of the abundance in the slick divided by the abundance in the rippled water.
P < 0.5 Wilcoxon's two-sample test (Sokal and Rohlf 1969).
Ocypoda spp. are abundant in neuston net tows
(Smyth 1980; Johnson 1985a). Juvenile filefish,
Monacanthus hispidus, are abundant in neuston
tows (Fahay 1975; Eldridge et al. 1977) and are com-
monly found associated with floating seaweeds
(Dooley 1972). Larval Hyporhamphus unifasciatus
and Hypsoblennius hentz are abundant in surface
plankton tows (Fahay 1975; Eldridge et al. 1977;
Fritzsche 1978). Both their behavior (Breder and
Clark 1947) and abundance in neuston tows (Eld-
ridge et al. 1977) suggest that larval Sphoeroides
maculatus are also residents of the surface waters.
These results suggest that there is an assemblage
of larvae unique to the neuston and that tests of the
707
internal-wave-mediated larval transport hypothesis
outlined in the introduction can be made.
The distribution of organisms followed a pattern
similar to that of Sargassum floats (Tables 1, 2, 4).
In the tows from internal waves oriented perpen-
FISHKRY HI.'LLETIN: VOL. 8t;, NO. 4
dicular to shore there was only one instance in which
the density of an organism was significantly higher
in the convergence than the divergence zone (total
larval fish, 14 July 1985, Table 1) indicating that
these internal waves were not transporting larvae.
Table 2. — The density (mean no. /1 00 m + SE) of invertebrates and Sargassum floats in front of tfie set of inter-
nal waves, tfie weighited average over thie internal waves, in tfie internal wave slick (convergence zone), tfie rippled
water between slicks (divergence zone), and befiind tfie internal waves. Significance compares tfie density between
tfiese samples using a Wilcoxon's two-sample test.
Density
Significance^
Mean no
./1 00 m^ + SE
;, f7 = 3
In front
Slicks
In front
vs.
Internal
vs.
vs.
internal
Species
In front
wave^
Slicks
Ripples
Befiind
befiind
ripples
wave
Total Bracfiyura
1.1+0.2
6.5+1.3
21+5.7
2.5 + 0.2
0.2 + 0.2
*
•
•
Portunus spp.
Megalopa
0.3 + 0
3.8 + 0.9
14 + 4.2
0.8 + 0.2
0
*
*
*
First crab
0
1.2 + 0.7
4.7+2.5
0.1 +0.1
0
*
*
Callinectes spp.
Megalopa
0.2 + 0.2
0.2 + 0.1
0.9 + 0.2
0
0.1 +0.1
*
Miscellaneous
Megalopa
0.6 + 0.3
1.7 + 0.3
1.8 + 0.7
1.7 + 0.2
0.1 +0.1
Amphiipods
30 + 8.1
176+112
754 + 546
0.3 + 0.3
20 + 6.5
*
Polycfiaete larvae
2.9+1.1
16 + 3.0
62 + 8.6
2.5+1.3
1.8 + 0.6
*
*
Penaeus spp.
Postlarvae
1.8+1.3
34 + 6.5
94 + 6.5
16 + 4.4
0.4 + 0.4
*
*
Sargassum floats
0.9±0.1
42 ±29
182±122
0.6±0.3
0
*
*
*
'Wilcoxon's two-sample test (Sokal and Rohlf 1969). • = P < 0.05.
^Weighted average density over an internal wave assuming a 30 m wide slick and a 100 m wide ripples. Internal wave
no./m^ X 30 m^) + (ripples no./m^ x 100 m2)/130 m^) x 100.
[(slick
Table 3. — Comparison of tfie density of various types of megalopae and larval fisfi caugfit in neuston, oblique, and
bottom plankton tows.
Density, no.
/1 00 m^
19 June 1986
14 June 1985
9 June
1986
Mean + se, n =
= 3
Species
Neuston
Oblique
Neuston
Oblique
Neuston
Oblique
Bottom
Crab megalopa
Portunus spp.
0
0
0
0
5.3 + 4.5
0
0
Callinectes spp.
0
0
0
0
3.1+2.4
0
0
Ocypoda quadrats
0
9.0
0
0
1.5 + 1.4
0
0
Pinnotfieridae
0.4
9.0
0
9.0
0.1 +0.1
310+160
18.7 + 9.2
Xantfiidae
0
88.8
0
9.0
0
690 + 250
17.0 + 9.6
Cancer spp.
0
0
0
0
0
66.7 + 40.6
1.4 + 0.8
Unknown and misc.
2.0
239
0.2
90.0
1.3 ±0.8
194 ±89.2
11.4 + 3.6
Fisfi larvae
Monacanthus hispidus^
0
0
1.1
0
0.1 +0.1
0
0
Hyporhamphus unifasciatus^
0
0
0
0
0.3 + 0.2
0
0
Sphoeroides maculatus
0
0
0
0
0.3 + 0.1
0
0
Membras martlnica
0.4
0
0
0
0
0
0
Hypsoblennius hentzi
0
0
0
0
0.3 + 0.3
6.0 + 6.0
0
Symphurus plagiusa
0
44.4
0
27.0
0
190 + 60
0.1+0.1
Prionotus evolans
0
18.0
0
136
0
510 + 210
15.1+6.0
Seriola spp.
0
0
0
27.0
0
46.0+17.8
0.3 + 0.3
Engraulidae
0
18.0
0
217
0
280 + 99
16.2 + 8.9
Unknown and misc.
0
9.0
2.2
63.0
0
150±110
3.9 + 3.0
'Juvenile or late postflexion stages only.
708
SHANKS: SHOREWARD LARVAL TRANSPORT
Table 4— The density (mean no. /1 00 m^ + SE) of larval fish in front of the set of internal waves, the weighted
average over the internal waves, in the internal wave slick (convergence zone), the rippled water between slicks
(divergence zone), and behind the internal waves. Significance compares the density between these samples using
a Wilcoxon's two-sample test. J = Juvenile, LPF = Late Post Flexion, ERF = Early Post Flexion, F = Flexion,
PF = Preflexion, and Total = Sum of all stages.
Mean no./
Density
MOO m^ ±
SE, n = 3
Significance'
In front
In front Slicks vs.
Internal
vs. vs. internal
Species
In front
wave
Slicks
Ripples
Behind
behind ripples wave
Monacanthus hispidus
1.1 +0.1
2.0 + 0.4
8.1+9.0
0.2 ±0.2
0.1 +0.1
* *
Hyporhamphus unifasciatus
J
0
0.1 +0.1
0.2±0.9
0
0
LPF
0.1+0.1
0.9±0.5
4.2±2.3
0
0
*
ERF
0
0.5±0.2
1.8±0.6
0
0
* *
F
0
0.1 ±0.1
0.2 + 0.2
0
0
*
PF
0
—
0
0
0
Total
0.1 ±0.1
1.5±0.6
6.5 ±2.6
0
0
* *
Membras martinica
J
0
0.1±0.1
0.4 + 0.4
0
0
LPF
0
—
0
0
0
ERF
0
—
0
0
0
F
0
0.1 +0.1
0.5 ±0.3
0
0
RF
0.2±0.1
0.2±0.1
14±4.2
0.4 ±0.4
1.0 + 0.6
* *
Total
0.2±0.1
0.2±0.1
15±44
0.4 ±0.4
1.0±0.6
* •
Hypsoblennius hentzi
J
0
0.2 + 0.2
0.8±0.8
0
0
LPF
0
0.2±0.1
0.5±0.3
0
0
*
ERF
0
0.2 + 0.0
0.7±0.1
0
0
* *
F
0.3 + 0.9
0.2 + 0.1
0.9 + 0.4
0
0
*
PF
14 + 2.4
4.5 ±1.8
16±6.0
0.9±0.5
2.6±0.2
* *
Total
14 + 2.0
5.2±1.6
19±5.5
0.9±0.5
2.6 + 0.2
* *
Brevootia tyrannus
F
0.2 + 0.1
0.1 +0.1
0.5 + 0.3
0
0.2 ±0.2
PF
0.7 + 0.2
0.2±0.1
0.6 ±0.3
0
0.1 ±0.1
Total
0.9 ±0.3
0.2±0.0
1.1 ±0.1
0
0.4 ±0.4
Miscellaneous
J
0
0.5 + 0.2
1.8±0.9
0
0
* *
LPF
0
0.2±0.1
0.8±0.9
0
0.1 ±0.1
* *
EPF
0
—
0
0
0
F
0.1 ±0.1
0.1 +0.1
0.4 + 0.4
0
0
RF
0
0.1 ±0.1
0.4±0.4
0
0
Total
0.1 ±0.1
0.8 ±0.2
3.4 ±1.0
0
0.1 ±0.1
* *
Grand total
16±1.6
13±1.0
53 ±3.6
1.6 ±0.2
4.2 ±0.4
* *
'Wilcoxon's two-sample test (Sokal and Rohlf 1969). * = P < 0.05.
^Weighted average density over an internal wave assuming a 30 m wide slick and a 100 m wide ripples. Internal wave = [(slick
no./m2 X 30 m^) + (ripples no./m^ x 100 m2)/130 m^] x 100.
In contrast, in tows from internal waves oriented
parallel to shore there were numerous instances in
which larval densities were significantly higher in
the slicks (Tables 1, 2, 4); the densities observed in
the slicks were 4- to >50-fold higher than those in
the rippled waters. These data demonstrate that a
variety of larval and postlarval invertebrates and
fish were transported shoreward by these internal
waves.
On 24 June 1985, surface plankton samples were
collected in the first two slicks over a set of inter-
nal waves oriented parallel to the shore, in the rip-
pled water between these slicks, and in front and
behind the entire set of internal waves. Ninety-six
percent of the surface drifters were caught by the
first two convergence zones and carried shoreward
about 4 km. The density of Sargassum floats was
highest in the slicks (>100-fold, Table 2) and they
were at significantly lower densities behind the set
of internal waves than in front. There were 0.9
Sargassum floats/100 m^ in the waters in front of
the set of internal waves and none in the waters
709
FISHERY BULLETIN: VOL. 8(;, NO. 4
behind the waves (Table 2). The distribution of tar
balls (spilled asphalt) over and around the internal
waves was similar to the distribution of Sargassum
floats (Shanks 1987). As these internal waves prop-
agated shoreward, the currents over the waves
swept buoyant flotsam from the waters in front of
the set of waves into the convergence zone where
the flotsam was caught and carried shoreward.
Calculation of the density of larvae or flotsam over
the internal waves (i.e., density in the slicks plus the
ripples) requires knowledge of the width of the slick
and rippled waters over the internal waves. Unfor-
tunately, in this initial study these measurements
were not made, necessitating that these widths be
approximated using values from the literature.
Slicks were assumed to be 30 m wide and the rip-
pled waters separating the slicks were assumed to
be 100 m wide. These values are consistent with my
experience and with published values (LaFond 1959;
Sawyer 1983). The observed density over one inter-
nal wave was calculated as (slick no./m^ x 30 m) -i-
(ripples no./m^ x 100 m)/130 m-. The density over
an internal wave was compared with the density in
an equal area of water in front of the set of inter-
nal waves. Making these calculations for Sargassum
floats gives an observed density of 42 floats/ 100 m^
over the internal wave vs. 0.9 floats/100 m^ in front
of the set; the density over the internal wave is
significantly larger (46-fold greater) than the den-
sity in front of the set (Table 2). Again the data
indicate that Sargassum floats were carried shore-
ward by the internal waves.
None of the types of larval fish characteristic of
water column samples (Table 3) were caught in any
neuston tow. Present in the neuston tows were only
those types of larval and juvenile fish which my
samples and the descriptions in the literature sug-
gest are characteristically neustonic.
The densities of larval and juvenile fish frequent-
ly were significantly higher in the slick samples than
the samples from the rippled waters between slicks
(Table 4). On 24 June 1985, most larval fish, espe-
cially juvenile and postflexion stage larvae, were
rare in both the waters in front and behind the set
of internal waves. On this date in 6 tows, 3 in front
and 3 behind the set of internal waves, only 4 juven-
ile and postflexion larvae were caught as compared
to 185 in the 3 tows made in the slicks. Probably
because of the rarity of larval fish in both in front
and behind tows there were only three cases, total
and preflexion Hypsoblennius hentz and total lar-
val fish, in which the density of fish in front of the
set of internal waves was significantly higher than
behind the set (Table 4). There are seven instances,
however, in which the density over the internal wave
of a larval or juvenile fish was significantly and at
least 10-fold higher than the density in front of the
set of internal waves (Table 4). In six of these in-
stances the fishes were at the juvenile, late post-
flexion, or early postflexion stages of development.
Larval and juvenile fish were grouped by stage
of development and the densities over and in front
of the internal waves were calculated for each
developmental stage (Fig. 2). The densities over the
internal waves of the juvenile, late postflexion, and
early postflexion developmental stages were signif-
icantly higher than the densities of these stages in
front of the set. There was not a significant differ-
ence between the densities over and in front of the
set of internal waves of flexion stage larvae and
preflexion larvae were significantly more abundant
in front of the set. These data suggest that internal
1 1 1 — I I I I I I 1 1 1 — I I M I I
0.1 1.0 10.0
Density, No./IOOm^
Figure 2.— Densities of larval fish by stage of development caught
on 24 June 1985 in the waters in front of the set of internal waves
(open circles) and over the internal waves (closed circles). Data are
presented as the mean + 95% confidence interval with the points
above or below this line being the actual observations. Asterisks
indicate cases where the in front density was significantly different
(Wilcoxon's two-sample test, P < 0.05) from the density over the
internal waves. The method of calculating the densities is described
in the text.
710
SHANKS: SHOREWARD LARVAL TRANSPORT
waves are capable of carrying fish larvae shoreward,
but shoreward transport seems to be confined to
postflexion stage larval or juvenile fish.
In the samples from 24 June 1985 there are a
number of cases in which a fish species and/or stage
of development was common in the tows from the
convergence zones but was absent or very rare in
the divergence zones, in front, or behind the set of
waves (i.e., zero or one caught in the nine tows from
these three habitats. Table 4). Either these larval
fish are extremely rare in the waters surrounding
the internal waves in which case the convergence
zone must have accumulated larvae from a large
volume of water or the internal waves transported
the larvae into the study area from a distant source.
On 24 June 1985, the density of total Brachyura
was significantly higher in the slicks over the inter-
nal waves than in any other area sampled (Table 2).
Portunus spp., a group which previous data had
demonstrated inhabited the neuston, made up the
bulk of the Brachyura caught. The densities of Por-
tuniis were significantly higher in the slicks than the
rippled water between slicks. The density of Por-
tunus spp. megalopae differed significantly from
0.3/100 m^ in front of the set of internal waves to
0/100 m- behind (Table 2). Portunus spp. first crabs
were absent from both the waters in front and
behind the set of internal waves; they were abun-
dant in the waters over the internal wave (Table 2).
The densities over the internal waves of Portunus
spp. megalopae and first crabs were significantly
higher than their densities in front of the set of
waves (Table 2). Callinectes spp. were uncommon
in the samples though their density was significantly
higher in the slick than in the rippled waters. In-
cluded under the category of miscellaneous mega-
lopae were the species Uca spp., Sesarma spp., and
majid crabs, all forms which were found to be more
abundant in the water column than in the neuston
(Table 2). There was not a significant difference
between their density in the slicks vs. the rippled
waters, suggesting that these megalopae were not
carried shoreward by the sampled set of internal
waves. The data indicate that definitely Portunus
spp. and possibly Callinectes spp. were carried on-
shore by the sampled set of internal waves.
Other organisms counted in the samples were
adult amphipods, polychaete larvae, and Penaeus
spp. postlarvae. Densities of these types of or-
ganisms were significantly higher in the slicks than
the rippled waters (Table 2). While in each instance
densities were lower behind the set of internal waves
than in front, the differences were not statistically
significant. The density of polychaete larvae and
Penaeus spp. postlarvae over the internal waves
were significantly higher than the density in front
of the set. These data suggest that these inverte-
brate larvae were also transported onshore by the
set of internal waves.
DISCUSSION
Tidally generated internal waves have been ob-
served in many areas of the world (Apel et al. 1975;
Fu and Holt 1982; Sawyer 1983). Larvae may util-
ize internal waves as a mechanism of onshore migra-
tion along many coastlines. Testing this hypothesis
would require the impossible task of making obser-
vations along all coastlines. An alternate technique
is to test for internal wave transport in areas with
different combinations of tidal range and shelf
width. Previous work has been done in areas of
mesotides (tidal range 2 to 4 m, Davis 1964) and
either narrow (<6 km. Shanks 1983) or moderate
(about 30 km, Kingsford and Choat 1986) shelf
widths. The Atlantic adjacent to the Beaufort Inlet
is characterized by microtides (tidal range <2 m,
Davis 1964) and a wide shelf (about 80 km). The first
purpose of this research was to test if onshore trans-
port of larvae could occur in an area with these
characteristics. On the three dates in which the
internal-wave-slicks were oriented roughly parallel
to shore the data suggest that internal waves were
transporting larvae and flotsam onshore. If the
waters around the Beaufort Inlet are representative
of localities with microtides and wide shelves, then
onshore migration of larvae via internal waves may
occur in other similar areas of the world.
On three dates the internal-wave-slicks were
oriented nearly perpendicular to shore and were pro-
pagating roughly northward. Over the continental
shelf, internal waves oriented roughly perpendicular
to shore have been observed in satellite photo-
graphs. These internal waves are invariably associ-
ated with submarine canyons (Apel et al. 1976).
There are no submarine canyons in Onslow Bay. The
internal waves causing these slicks may have been
formed over Frying Pan Shoals (Fig. 1). These
shoals extend about 50 km across the shelf from the
end of Cape Fear and form the southern boundary
of Onslow Bay. Just north of the study area, Cape
Lookout and the Cape Lookout Shoals also extend
out across the shelf (Fig. 1). On a flight over Cape
Lookout numerous slicks oriented parallel to Cape
Lookout Shoals were observed propagating north-
ward into Rayleigh Bay (pers. obs.). The geography
711
FISHERY BULLETIN: VOL. 86. NO. 4
and oceanography of the Cape Lookout and Frying
Pan Shoals are so similar that by analogy the inter-
nal waves oriented perpendicular to shore observed
during this study may have been formed over Fry-
ing Pan Shoals and propagated northward into
Onslow Bay. The mechanism for the formation of
these internal waves is unknown.
As was observed in a previous study (Shanks
1983), only some sets of internal waves transported
larvae or flotsam. In this study only those slicks
aligned parallel to shore caused transport. It is not
clear why some sets of waves cause transport while
others do not. The physical characteristics of tidal-
ly generated internal waves are quite variable. The
amplitude (Cairns 1968) and decay distance of in-
ternal waves varies over the fortnightly tidal cycle
(Brink 1988). Further, the depth, wave length, and
shape of the internal waves is dependent on the
relative depths of the thermocline and the bottom
and wave amplitude (LaFond 1959; Lee 1961; Cairns
1967, 1968). What is needed is simultaneous mea-
surements of the physical characteristics of a set of
internal waves much like those made by LaFond
(1959) with measurements of the transport of flot-
sam or larvae.
The second purpose of this research was to test
several new predictions derived from the hypothesis
that internal waves can transport larvae. If trans-
port was occurring, then one would predict that
1) due to the accumulation of larvae in the con-
vergence zones as internal waves propagate shore-
ward, the density of larval types transported by the
internal waves should be significantly lower behind
than in front of the set, 2) the observed density of
larvae over an internal wave should be significant-
ly higher than in the waters in front of the set of
waves, and 3) there may be types of larvae which
are only present in the slicks, suggesting that they
had been carried into the area from a distant source.
The appropriate samples to test these three predic-
tions were collected on 24 June 1985. The densities
of several larval types were significantly higher in
the waters in front of the internal wave set than
behind. There were many instances in which the
observed density over the internal waves of a type
of larvae was significantly higher than in the waters
in front of the set of internal waves. Lastly, there
were a number of organisms that were only caught
in the convergence zones over the internal waves.
In this set of observations the three predictions were
confirmed indicating that this set of internal waves
was carrying larvae and flotsam shoreward.
The significant differences in larval densities
observed on 24 June (i.e., in front vs. behind and
over the internal waves vs. in the front of the set
of waves) may have been due to fortuitous cross-
shelf patchiness in larval density. Because conditions
allowed only one opportunity to sample in front and
behind a set of internal waves, this alternate ex-
planation can not be rejected. Cross-shelf larval
patchiness is, however, probably not an adequate
explanation because of the very short distance over
which large differences in larval abundances were
observed. For example, the tows in front of the set
and the tows in the slick and rippled water over the
set were separated by at most 200 m, yet there were
many cases (14, Tables 2, 4) where larval abun-
dances were different by at least a factor of 10. Dif-
ferences in plankton abundance of this magnitude
and over this small a distance are almost always
associated with oceanographic features (e.g., fronts:
Boden 1952; Pingree et al. 1974; Owen 1981; Fogg
et al. 1985). The only apparent oceanographic
feature in the study area was the internal-wave-
slicks. The observed differences in larval density
were probably caused by the internal waves.
The data in Figure 2, the density of fish by stage
of development over the internal waves vs. in front
of the set of waves, suggest that only the later
developmental stages of fish (juvenile through early
postflexion) were transported onshore by the set of
internal waves sampled on 24 June 1985. All five
of the abundant fish species (Table 4) inhabit near
shore or estuarine habitats as adults (Fritzsche 1978;
Hardy 1978; Johnson 1978; Martin and Drewry
1978). Flexion and preflexion larval fish are clearly
not competent to adopt the adult or nursery habitat,
and the data suggest that they were not carried on-
shore by the internal waves. There may be adaptive
advantages to planktonic larvae avoiding estuarine
waters during their development (Strathmann
1982). Juvenile fish and, perhaps, also postflexion
larvae, are competent to recruit into the adult or
nursery habitat. Transport onshore by internal
waves may, therefore, be adaptively advantageous
for those fish whose adult or nursery habitat is
coastal or estuarine.
The larval development of the blue crab, Calli-
nectes sapidus, occurs at sea (Smyth 1980;
McConaugha et al. 1983; Johnson 1985b). The lar-
vae are present in the waters over the continental
shelf and out to the Gulf Stream (Smyth 1980). At
the end of the larval period the megalopae must
return to an estuarine habitat to continue its adult
existence. How the megalopae make this migration
is an open question (Johnson et al. 1984; Johnson
712
SHANKS: SHOREWARD LARVAL TRANSPORT
1985b). The behavior of the megalopae (Sulkin and
Van Heukelem 1981) and their distribution in the
plankton (Smyth 1980; Johnson 1985a) both suggest
that these larvae are inhabitants of the neuston. Lar-
vae that inhabit the neuston can be transported on-
shore by internal waves (Shanks 1985). On 4 and 24
June and 21 August the densities of Callinectes spp.
megalopae and first crabs were significantly higher
in the convergence zones over the internal waves
than in the divergences, suggesting that Callinectes
spp. were being transported onshore by the sampled
internal waves. On 24 June, when Callinectes spp.
were uncommon, there was not a significant differ-
ence in the density of Callinectes spp. in front vs.
behind the internal wave nor was there a significant
difference in the density of blue crabs over vs. in
front of the set of waves. These results are mixed,
but they do suggest that the megalopae of Calli-
nectes spp. may be transported shoreward in the con-
vergence zones over internal waves.
The megalopae of a variety of crab species were
only caught in the water column (Table 3 and see
Johnson 1985a). Some of these megalopae as adults
occupy near shore and even estuarine habitats. If
these megalopae migrate onshore they must be
utilizing some mechanism of onshore transport other
than slicks over internal waves.
In conclusion, the data presented in this paper in-
dicate that in an area where the tides are of small
amplitude and the continental shelf is wide, inter-
nal waves are nevertheless capable of transporting
larval invertebrates and fish shoreward. The sam-
ples collected on 24 June 1985 more critically test
the hypothesis that internal waves cause cross-shelf
transport and the results support the hypothesis.
ACKNOWLEDGMENTS
Assistance with the field work was enthusiastical-
ly provided by 0. McMillan, G. Safrit, J. Purifoy,
and W. Graham. G. Safrit, in addition, heroically and
with really very little complaint, sorted most of the
plankton samples. Comments by M. Kingsford and
R. Forward improved the manuscript. Figures were
prepared by V. and H. Page and F. Schwartz helped
to identify larval fish.
LITERATURE CITED
Ahlstrom E. H., J. L. Butler, and B. Y. Sumida.
1976. Pelagic stromiatoid fishes (Pisces, Perciforms) of the
eastern Pacific: kinds, distributions, early life histories and
observations on five of these species from the northwest
Atlantic. Bull. Mar. Sci. 26:285-402.
Apel, J. R., H. M. Byrne, J. R. Proni, and R. L. Charnell.
1975. Observations of oceanic internal and surface waves
from the Earth Resources Technology Satellite. J.
Geophys. Res. 90:865-881.
Boden, B. p.
1952. Natural conservation of insular plankton. Nature 169:
697-699.
Breder, C. M., Jr., and E. Clark.
1947. A contribution to the visceral anatomy, development,
and relationships of the Plectognathi. Bull. Am. Mus. Nat.
Hist. 88:287-320.
Brown, D. M., and L. Cheng.
1981. New net for sampling the ocean surface. Mar. Ecol.
Prog. Ser. 5:225-227.
Brink, K. H.
1988. On the effect of bottom friction on internal waves.
Cont. Shelf Res. 8:397-403.
Cairns, J. L.
1967. Asymmetry of internal tidal waves in shallow waters.
J. Geophys. Res. 72:3563-3565.
1968. Thermocline strength fluctuations in coastal waters.
J. Geophys. Res. 73:2591-2595.
Chereskin, T. K.
1983. Generation of internal waves in Massachusetts Bay.
J. Geophys. Res. 88:2649-2661.
Cook, H. L.
1966. A generic key to the protozoean, mysis, and postlar-
val stages of the littoral Penaeidae of the northwestern Gulf
of Mexico. Fish. Bull., U.S. 65:437-447.
Davis, J. L.
1964. A morphogenic approach to world shorelines. Z.
Geomorphol. 8:127-142.
Dooley, W. L.
1972. Fishes associated with the pelagic Sargassum com-
munity. Contrib. Mar. Sci. 16:1-32.
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. SC Mar.
Resour. Cent. Tech. Rep. No. 18, 22 p.
Ewing, G.
1950. Slicks, surface films and internal waves. J. Mar. Res.
9:161-187.
Fahay, M. P.
1975. An annotated list of larval and juvenile fishes captured
with surface-towed meter net in the South Atlantic Bight
during four RV Dolphin cruises between May 1967 and
February 1968. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS SSRF-685, 39 p.
1983. Guide to the early stages of marine fishes occurring
in western Atlantic Ocean, Cape Hatteras to the southern
Scotian shelf. J. Northwest Atl. Fish. Sci. 4:1-423.
Fogg, G. E., B. Egan, G. D. Floodgate, D. A. Jones, J. Y.
Kassab, K. Lochte, E. I. S. Rees, S. Scrope-Howe, and
C. M. TURLEY.
1985. Biological studies in the vicinity of a shallow-sea tidal
mixing front VII. The frontal ecosystems. Philos. Trans.
R. Soc. Lond. B310:555-571.
Fritzsche, R. a.
1978. Chaetodontidae through Ophidiidae. In Development
of fishes of the Mid- Atlantic Bight: an atlas of egg, larval
and juvenile stages, p. 1-340. U.S. Dep. Inter., Fish. Wildl.
Serv., Biol. Serv. Program Vol. 5.
Fu, L.-L., and B. Holt.
1982. Seasat views oceans and sea ice with synthetic-apera-
ture radar. Jet Propul. Lab. Puhl. 81-120.
713
FISHERY BULLETIN: VOL. 86, NO. 4
Gargett, a. E.
1976. Generation of internal waves in the Strait of Georgia,
British Columbia. Deep-Sea Res. 23:17-32.
Hardy, J. D., Jr.
1978. Anguillidae through Syngnathidae. In Development
of fishes of the Mid-Atlantic Bight: an atlas of egg, larval
and juvenile stages, p. 1-458. U.S. Dep. Inter., Fish. Wildl.
Serv., Biol. Serv. Program Vol. 2.
JiLLETT, J. B., AND J. R. ZELDIS.
1985. Aerial observations of surface patchiness of a plank-
tonic crustacean. Bull. Mar. Sci. 37:609-619.
Johnson, D. F.
1985a. The distribution of Brachyuran crustacean megalopae
in the waters of the York River, lower Chesapeake Bay and
adjacent shelf: implications for recruitment. Estuarine
Coastal Shelf Sci. 20:693-705.
1985b. Wind-forced dispersion of blue crab larvae in the Mid-
dle Atlantic Bight. Cont. Shelf. Res. 4:733-745.
Johnson, D. F., B. S. Hester, J. R. McConaugha.
1984. Studies of a wind mechanism influencing the recruit-
ment of blue crabs in the Middle Atlantic Bight. Cont.
Shelf. Res. 3:425-437.
Johnson, G. D.
1978. Carangiadae through Ephippidae. In Development of
fishes of the Mid-Atlantic Bight: an atlas of egg, larval and
juvenile stages, p. 1-324. U.S. Dep. Inter., Fish. Wildl.
Serv., Biol. Serv. Program Vol. 4.
Kingsford, M. J., and J. H. Choat.
1986. Influence of surface slicks on the distribution and on-
shore movements of small fish. Mar. Biol. (Berl.) 91:161-
171.
LaFond, E. C.
1959. Slicks and temperature structure in the sea. Navel
Electron. Lab., San Diego, CA Rep. 937:1-27.
Lee, C. Y., and R. C. Beardsley.
1974. The generation of long nonlinear internal waves in a
weakly stratified shear flow. J. Geophys. Res. 79:453-462.
Lee, 0. S.
1961. Observations on internal waves in shallow water.
Limnol. Oceanogr. 6:312-321.
Martin, F. D., and G. E. Drewry.
1978. Stromateidae through Ogcocephalidae. In Develop-
ment of fishes of the Mid-Atlantic Bight: an atlas of egg, lar-
val and juvenile stages, p. 1-416. U.S. Dep. Inter., Fish.
Wildl. Serv., Biol. Serv. Program Vol. 6.
Maxworthy, T.
1979. A note on the internal solitary waves produced by tidal
flow over a three-dimensional ridge. J. Geophys. Res.
84:338-345.
McConaugha, J. R., D. R. Johnson, A. J. Provenzano, and
R. C. Maris.
1983. Seasonal distribution of larvae of Callinectes sapidus
(Crustacea:Decapoda) in the waters adjacent to Chesapeake
Bay. J. Crust. Biol. 3:582-591.
MosER, H. G., and E. H. Ahlstrom.
1970. Development of lantern fishes (family Mcytophidae) in
the California Current. Part 1. Species with narrow-eyed
larvae. Bull. Los Angeles City Mus. 7:1-145.
Osborne, A. R., and T. L. Burch.
1980. Internal solitons in the Andaman Sea. Science 208:
451-460.
Owen, R. W.
1981 . Fronts and eddies in the sea: Mechanisms, interactions
and biological effects. In A. R. Longhurst (editor). Analysis
of marine ecosystems, p. 197-234. Acad. Press, N.Y.
Pingree, R. D., G. R. Forster, and G. K. Morrison.
1974. Turbulent convergent tidal fronts. J. Mar. Biol. Assoc.
U.K. 54:469-479.
Sawyer, C.
1983. A satellite study of ocean internal waves. U.S. Dep.
Commer., NOAA Tech. Memo. ERL PMEL-46.
Shanks, A. L.
1983. Surface slicks associated with tidally forced internal
waves may transport pelagic larvae of benthic invertebrates
and fishes shoreward. Mar. Ecol. Prog. Ser. 13:311-315.
1985. Behavioral basis of internal-wave-induced shoreward
transport of megalopae of Pachygrapmis crassipes. Mar.
Ecol. Prog. Ser. 24:289-295.
1986. Tidal periodicity in the daily settlement of intertidal
barnacle larvae and an hypothesized mechanism for the
cross-shelf transport of cyprids. Biol. Bull. (Woods Hole)
170:429-440.
1987. The onshore transport of an oil spil by internal waves.
Science 235:1198-1200.
Shanks, A. L., and W. G. Wright.
1987. Internal-wave-mediated shoreward transport of
cyprids, megalopae, and gammarids and correlated long-
shore differences in the settling rate of intertidal barnacles.
J. Exp. Mar. Biol. Ecol. 114:1-13.
Smyth, P. 0.
1980. Callinectes (Decapoda:Portunidae) larvae in the Mid-
dle Atlantic Bight, 1975-77. Fish. Bull., U.S. 78:251-265.
SOKAL, R. R., AND F. J. ROHLF.
1969. Biometry. Freeman, San Francisco, p. 776.
Strathmann, R. R.
1982. Selection for retention or export of larvae in estuaries.
In V. Kennedy (editor), Estuarine comparisons, p. 521-536.
Acad. Press, N.Y.
Sulkin, S. D., AND W. Van Heukelem.
1981. Larval recruitment in the crab Callinectes sapidus
Rathbun: an amendment to the concept of larval retention
in estuaries. In V. Kennedy (editor), Estuarine com-
parisons, p. 459-475. Acad. Press, N.Y.
Williams, A. B.
1984. Shrimp, lobsters, and crabs of the Atlantic coast of the
eastern United States, Maine to Florida. Smithson. Inst.
Press, Wash., D.C., p. 550.
ZELDIS, J. R., AND J. B. JiLLETT.
1982. Aggregation of pelagic Munida gregaria (Fabricius)
(Decapoda, Anomura) by coastal fronts and internal waves.
J. Plankton Res. 4:839-857.
714
TROPHIC RELATIONS OF THE BLUE ROCKFISH,
SEBASTES MYSTINUS, IN A COASTAL UPWELLING SYSTEM
OFF NORTHERN CALIFORNIA
Edmund S. Hobson and James R. Chess^
ABSTRACT
The planktivorous Sebastes mystinus in nearshore habitats off northern California feeds primarily on
relatively large, gelatinous zooplankters that originate offshore, including thaliaceans, ctenophores, and
pelagic hydrozoans. These prey organisms increase in number during spring and summer when surface
waters driven seaward by northerly winds carry upwelled nutrients to diatoms that nourish offshore
zooplankton populations. But the resulting increases in zooplankton during this upwelling season become
available to S. mystinus in the nearshore habitats only when the surface flow turns shoreward during
intermittent episodes of downwelling. Although some of this shoreward flow is driven by southerly winds,
much of it occurs during calms, or under northerlies lacking the velocities needed to drive surface waters
seaward. There is increasing shoreward transport during fall and winter, when downwelling episodes
are more frequent, but progressively fewer zooplankters are carried into the nearshore habitats. This
is because as less nutrients come into the system with the reduced upwelling, and as available sunlight
declines, the offshore zooplankton populations suffer from shortages of diatoms. Although S. mystinus
compensates for decreased numbers of zooplankters during most of the year vnth increased consump-
tion of specific plant materials, i.e., Nereocystis sori, or the monostromatic epiphytes Porphyra nereocystis
and Smithora naidum (depending on the season), these too are in short supply during winter. In winter,
therefore, S. mystinus experiences its poorest feeding conditions. Thus, S. mystinus is adapted to feeding
opportunities created by alternating episodes of strong upwelling and strong downwelling, and is most
abundant within its range along the west coast of North America where both conditions are well developed.
Coastal marine fishes in temperate latitudes experi-
ence major seasonal changes in their environment.
In study off northern California's Mendocino coast
(lat. 39°13'N, long, 123°14'W), we studied the ef-
fects of seasonal change on the trophic relations of
the blue rockfish, Sebastes mystinus (Fig. 1).
Sebastes mystinus, a major species in the recrea-
tional fishery off northern and central California
(Frey 1971), is a planktivore that feeds on scypho-
zoans, ctenophores, copepods, amphipods, thalia-
ceans, fishes, and algae (Gotshall et al. 1965; Love
and Ebeling 1978; Hallacher and Roberts 1985).
Although its diet is known to vary with the season,
relationships involving specific environmental
features, and the availability of prey, remain
unclear.
The marine environment off California is pro-
foundly affected by seasonal variations in wind-
driven movement of the surface water (Reid et al.
1958; Bolin and Abbott 1962; Bakun and Parrish
^Southwest Fisheries Center Tiburon Laboratory, National
Marine Fisheries Service, NOAA, 3150 Paradise Drive, Tiburon,
CA 94920.
1980). This force, known as Ekman transport
(Ekman 1905), is strongest off the Mendocino coast
(Bakun 1973), where it must be a major influence
on the trophic relations of S. mystinus. Because
Ekman transport is the basis of coastal upwelling
(e.g.. Smith 1968), it is of great importance to
biological productivity (e.g., Boje and Tomczak 1978)
and, therefore, to the availability of food.
Upwelling develops along the coast of northern
California when surface waters driven seaward by
northerly winds, such as those characteristic of
spring and summer, are replaced by colder subsur-
face waters from offshore that flow up into the near-
shore habitat (Smith 1968; Bakun 1973). On the
other hand, an opposing condition develops when
surface waters driven shoreward by southerly
winds, such as those characteristic of winter storms,
flow over the colder nearshore waters to produce
the condition sometimes called downwelling (Bakun
1973). But despite the strong seasonality evident in
both upwelling and downwelling, short-term rever-
sals lasting just a few days occur throughout the
year (see Bakun [1973] and Mason and Bakun
[1986]).
Manuscript accepted July 1988.
FISHERY BULLETIN: VOL, 86, NO. 4, 1988.
715
FISIIKKV HILLKTIN: VOL. 8(;. NO. 4
Figure \.— Sebast.es mystinus next to canopy of bull kelp, Nereocystis leutkeana, off Mendocino.
In this paper we consider how the trophic rela-
tions of 5. mystinus respond as seaward and shore-
ward movements of the surface water produce
alternating episodes of upwelling and downwelling.
Emphasis is on how the resulting environmental
changes alter the relative availability of food.
Among studies of marine fishes, this is, to our
knowledge, the most comprehensive attempt yet
made to integrate data on food, potential food, and
environmental variables— all key elements in trophic
relations.
METHODS
The study lasted from the winter of 1976-77 to
the summer of 1981, with the first 15 months in-
volving exploratory work along about 15 km of the
Mendocino coast south of Point Cabrillo. A study
site was then established off Salmon Point (Fig. 2)
during the spring of 1978, and from that time sam-
pling followed a set regime.
Study Site
The study site (Fig. 3) was in 10-15 m of water,
about 300 m from shore. Rocks the size of houses
jutted 10-15 m above the water at the seaward
perimeter of the site, but despite the shelter offered
by these rocks, most of the area was regularly swept
by wind and sea. Except for isolated pockets of sand,
the site was floored by rock pavement and boulders
(some 5-15 m in diameter), largely swept clean by
the turbulence and surge that prevailed most of the
time.
Environmental Variables
During each sampling session, we noted the gen-
716
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
42°N
41°N
40°N
39°N
38°N
37"N
125°W 124°W 123°W 122°W 121°W
Figure 2.— The coast of northern California.
eral state of the weather, including wind direction
and estimated velocity. In this paper (except in
Figures 4 and 9, as noted below) we consider winds
from between northwest and northeast to be
"northerly", and from between SW and SE to be
"southerly". We also noted sea conditions and
recorded sea-surface temperatures. More precise
wind data became available after May 1980, when
a NOAA weather station was established at Mendo-
cino. During the same month, we placed recording
thermographs at depths of 6 and 20 m in the study
area, and although the deeper site was abandoned
after 1981 in favor of replicating the record with
two instruments at the shallower site, these have
given us a continuous record of water temperatures
from then until the present (Spring of 1988). Based
on these observations and in situ visual assessments
of the plankton and water characteristics, usually
we could tell whether upwelling or downwelling
predominated during a sampling session even
though most of the time conditions were to some
extent mixed.
Diet and Occurrences of Food
To relate the diet of S. mystinus to foods present
at the time of feeding, we took concurrent samples
of gut contents and plankton, as well as some selec-
tive samples of the benthos. Complications from
normal diel variability were reduced by taking all
samples between the hours of 1100 and 1300. The
sampling schedule was strongly influenced by the
weather, as much of the time work was prevented
by high seas and/or long-period swells and result-
717
FISHERY BULLETIN: VOL. 86, NO. 4
Figure 3.— Study site off Salmon Point.
ing turbulence. Although all seasons were sampled,
most collections and observations were made when
the sea was relatively calm, a condition that gen-
erally lasted no more than a few days at a time.
Nevertheless, because often the weather turned
while observations were under way, turbulent con-
ditions were well represented.
Fish
The 5. mystinus studied at Mendocino were spa-
tially segregated by size and age. This paper con-
siders only adult fish of more than 200 mm SL,
which forage by day in aggregations of up to several
hundred individuals in the upper levels of the water
column. The analysis of gut-contents involved 247
individuals, 200-350 mm SL, taken from these ag-
gregations using handheld spears. Juveniles and
subadults are excluded because they forage at lower
levels of the water column and consume organisms
not taken by adults.
Plankton
Most sampling sessions included two plankton
collections: one at the sea surface, and the other
between 1 and 2 m above the sea floor. Only the 27
surface collections are considered here, however,
because the others sampled at levels of the water
column below where adults usually feed. For each
collection we used scuba to push the net (0.333 mm
mesh in a 78 cm x 78 cm frame) through the water
for 5 minutes. Occasionally the net broke the sur-
face, but usually was kept underwater to avoid foul-
ing the sample with floating debris. (See Hobson and
Chess 1976 for additional information on this pro-
718
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
cedure, and how the samples were processed and
analyzed.) We visually assessed the larger zooplank-
ters in the water column during each collection, as
well as at other times during each sampling session,
and these observations greatly enhanced our abil-
ity to interpret the samples.
Benthos
Although S. mystinus is a planktivore, it is impor-
tant to include the benthos when documenting
environmental changes that affect its feeding. Many
of the actual or potential foods of 5. mystinus
originate on the underlying substrata, and their oc-
currences in the water column relate to changes in
the benthos. Thus, assessments of the benthos done
during other studies concurrently under way in the
study area, including visual counts and airlift sam-
ples, produced data that are incorporated into this
report where pertinent.
Ranking Prey Taxa
To estimate the relative importance of the various
prey organisms in the diet of S. mystinus, we
grouped related forms in multispecies categories
that were then ranked. The ranking was based on
an index calculated as: relative frequency of occur-
rence in diet x mean number of individuals con-
sumed X percent of total diet volume that was
represented by that category. This index is similar
to the widely used Index of Relative Importance
(IRI) of Pinkas et al. (1971), but puts more weight
on numbers consumed. It is important to emphasize
numbers consumed in quantifying the feeding ac-
tivity of S. mystinus (and most other planktivores)
because the many small prey are ingested indi\adual-
ly, making the capture of each a discrete act.
RESULTS
During the four years of this study, spring and
(to a lesser extent) summer constituted an upwell-
ing season, while fall and (to a greater extent) winter
constituted a downwelling season. The spring tran-
sition between downwelling and upwelling seasons
occurred over just a few days between late March
and early April, whereas the less distinct fall tran-
sition between upwelling and downwelling seasons
was extended over a month or more between mid-
August and late October.
Despite this seasonal pattern, however, there
were short-term reversals of just a few days that
had profound effects. Reactions of the environment
to major wind changes were virtually immediate.
Sharply reduced water temperatures signifying the
intrusion of upwelled water often followed within
hours of increased northerly winds, while warmer
water rich in such readily visible zooplankters as
ctenophores, hydrozoans, pteropods, thaliaceans,
and larvaceans, often flowed into the nearshore
habitats within a day after the onset of southerly
winds. Conditions during each sampling session are
listed in Table 1.
Although downwelling conditions were most in-
tense under southerly winds, they also developed
during calms and with weak northerlies. We found
that generally it took northerlies of 10 knots (K) or
more to produce upwelling conditions, although as
little as 5 K were effective if held steady for several
days. In the absence of sufficient force, however,
upwelling ceased and warmer water rich in offshore
zooplankters entered the nearshore habitat, gen-
erally moving in a southerly direction along the
coast.
The close relation between shifts in prevailing
wind and alternations between upwelling and down-
welling is illustrated by comparing the wind direc-
tion and velocity measured by the nearby NOAA
weather station to the sea temperatures recorded
by our thermographs. Although these data became
available only during the last year of the study,
subsequent years have produced similar profiles (ex-
cept 1982-83, when there w^as a strong El Nino).
Thus, the pattern of sea temperatures and wind dur-
ing the upwelling season of 1981 and during the
downwelling season of 1980-81, recounted in detail
below, are representative.
The Upwelling Season
The upwelling season began in late March or early
April with a precipitous drop in sea temperatures.
It was a drop of about 3 °C— typically from
11°-12°C to 8°-9°C-that coincided with the onset
of strong, persistent northerly winds characteristic
of this time of year. The pattern of sea temperatures
and wind during the 1981 upwelling season (Fig. 4)
illustrates how upwelling and downwelling related
to prevailing winds during that period.
Habitat Conditions
At the start of the upwelling season, the nearshore
habitats appeared barren. Storm seas during the
previous winter had carried away most of the bull
719
FISHERY HILI.ETIN; VOL. 86. NO. 4
25
>-
20
_l
a:
UJ
15
1-
on
o
10
z
c
Jd
5
o.
0
>-
_l
a:
5
LU
X
3
10
O
CO
15
EVENT
o
a:
r)
!<
oc
UJ
Q.
:s
UJ
12 -1
11
10
9
8
7
APRIL
MAY
JUNE
JULY
Figure 4.— Sea temperatures and wind off Mendocino during the 1981 upwelling season. Plotted sea tem.pefatures are estimated
daily means from record of continuously recording thermograph. Plotted wind velocities and directions are averages of 2-5
daily readings. Southerly values represent winds from south of east-west axis, northerly values represent winds from north
of this axis. The highly infrequent winds from due west or due east were entered as zero. In calculating values for days of
both northerlies and southerlies, the former were considered positive, the latter negative. The following accounts of events
identified in the figure cite wind velocities that exceed plotted values, which are averages.
Event 1— Seven days of southerly winds, 19-25 March, with sea temperatures rising to 11.1°C, represented the last down-
welling episode of the 1980-81 downwelling season.
Event 2— The 1981 upwelling season began on 26 March, as 2 weeks of strong northerlies (to 4 April) resulted in an abrupt
drop of more than 3°C in sea temperature.
Event 3— Water temperatures fell to the lowest point of the year, 7.3°C, on 13 May following over 2 weeks of 10-30 K northerlies.
Event 4— The first of a series of downwelling episodes during this upwelling season developed as 8-12 K southerlies on 17
and 18 May resulted in a sharp rise in sea temperature to 8.5°C.
Event 5— Temperatures dipped when the wind shifted back to the north on 19 May. Over the next month variable northerlies
blew at less than 10 K, and although sea temperatures did not vary more than a few tenths of a degree, they were consistently
about 1°C warmer than before the 17-18 May downwelling episode.
Event 6— On 28 June, the first of 6 consecutive days of 4-10 K southerlies, sea temperatures began a steep climb to above
10°C that marked the upwelling season's second major downwelling episode.
Event 7— Northerlies returned on 3 July, and sea temperatures immediately dropped below 10°C again. Despite northerlies
of 5-7 K over the next 2 weeks, sea temperatures remained at about 9.5°C, again about 1 °C warmer than before the 28 June-2
July downwelling episode.
Event 8— With southerlies of 5-10 K on 7 of the 9 days between 12 and 20 July, sea temperatures once again rose above
10°C to mark the season's third major downwelling episode.
720
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
Table i .
-Conditions during sampling sessions off Salmon Point, Mendocino
County, 1978-81.
Sea
Wind
Plankton
Fish sampled
Size
temp.
dir. : vel.
condi-
range
Mean
Date
CO
(knots)
tion^
No.
(mm)
(mm)
1. Upwelling
season.
upwelling conditions
5/10/78
8.5
NW: 10
U/M
7
270-312
289.7
8/08/78
10.5
NW: 10
U
6
285-309
2930
6/12/79
9.0
NW: 15
U/M
7
241-312
239.3
6/24/80
10.0
None
U
9
224-308
250.0
4/09/81
8.1
NW: 15
U
8
235-300
265.6
5/19/81
9.5
WNW: 10
U/M
1
260
260.0
6/17/81
9.0
NNW: 10
U
13
260-336
301.0
II. Upwelling
season
downwelling
conditions
6/21/78
9.0
None
D/M
3
228-308
270.3
6/29/78
10.0
SW: 2
D/M
6
240-315
254.7
8/23/78
11.0
None
D
6
230-280
259.0
9/08/78
14.0
None
D
4
240-295
277.5
4/24/79
10.0
None
M
9
225-313
277.7
7/24/79
11.5
None
M
11
208-302
239.5
8/27/79
12.8
NW: 5
M
12
204-335
264.6
5/29/80
9.5
None
D/M
20
213-330
276.4
7/17/80
11.0
NW: 5
D/M
5
240-304
270.8
8/21/80
11.0
S: 5
D
9
230-305
270.6
III. Downwell
ng season, downwell
ng conditions.
fall
10/04/78
11.0
WNW: 10
D +
3
270-310
286.7
10/18/78
12.0
S: 5
D +
9
220-350
296.2
12/08/78
9.0
S: 8
D +
17
227-331
281.2
10/16/79
14.0
NW: 8
D
13
218-311
280.1
11/04/80
12.0
N: 6
D
17
230-335
296.6
1 2/1 7/80
12.0
NNW: 5
D
14
225-331
276.3
IV. Downwei
ing season, downwell
ing conditions
winter
2/04/79
10
S: 12
D
12
231-320
273.3
3/06/79
11
W: 8
D
2
224-260
242.0
1/31/80
11
None
M
16
200-345
275.3
1/13/81
13
SW: 5
D
8
275-305
286.3
'Plankton condition. D = Downwelling condition: zooplankters considered to be off-
shore species, especially relatively large gelatinous forms, numerous. U = Upwelling con-
dition: offshore zooplankters absent. M = Mixed condition: some offshore zooplankters
present.
kelp, Nereocystis leutkeana, which is the major com-
ponent of the kelp forests. These same seas, often
heavily laden with sediment, had also scoured the
seafloor of much of the algal understory, as well as
many of the sedentary invertebrates. Once the up-
welling season had been established, however, the
habitats changed rapidly.
Often during early-season upwelling the water
was relatively transparent, although after rains visi-
bility often was limited by suspended sediment dis-
charged into the nearshore habitat from coastal
streams. By mid-May, however, the numbers of
planktonic diatoms (primarily Chaetoceros sp. and
Nitzchia sp.) had greatly increased to depths of
Event 9— In the last major upwelling of this upwelling season, sea temperatures fell to 9.4°C on 2 September after 10 con-
secutive days of 5-8 K northerlies.
Event 10— Sea temperatures rose steadily to 11.4°C on 17 September after 1-3 day periods of 4-8 K northerlies had alter-
nated with 1-4 day periods of 4-8 K southerlies through the first half of that month.
Event 11— Sea temperatures fell to 10.5°C on 24 September after 7 consecutive days of 5-15 K northerlies. A shift to southerlies
on 25 September started another climb, and the transition to the downwelling season clearly was under way. This transition,
however, was not nearly so well defined as that which had introduced the upwelling season 6 months before.
721
FISHERY BULLETIN: VOL. 86, NO. 4
10-20 m, which often limited visibility to a few
meters. These diatom blooms were most evident
during the intermittent episodes of downwelling.
The benthos similarly proliferated. Within a few
weeks after the initial upwelling, a growth of ben-
thic diatoms (primarily Isthmia nervosa and Tricer-
tium americana) appeared on the previously baron
rocks, and even on sand sheltered from wave surge.
Then larger elements of the biota began to increase
in size and number. Particularly evident were cal-
careous sponges (especially Leucosolenia spp.),
hydroids, and certain bryozoans. Benthic algae,
predominantly Desmarestia ligulata, rapidly over-
grew much of the rock substrata, while young A^ereo-
cystis leutkeana, which first appeared on the sea-
floor in May, had grown to the water's surface by
mid-June. Swarms of mysids were increasingly
numerous near the seafloor in April (they had been
prominent for a month or more), and caprellid and
gammaridean amphipods on rocky substrata began
a sharp increase in numbers.
The bull kelp had formed a dense canopy by mid-
July, and at about the same time large numbers of
sori (reproductive structures) began falling from the
kelp's fronds. (A frond that had recently lost a sorus
appears at the left side of Figure 1.) Planktonic
diatoms continued to proliferate during periodic
blooms, but by this time the benthic diatoms that
had carpeted much of the seafloor during the spring
were mostly gone. Similarly, the mysid swarms,
which had peaked in May and June, usually began
to decline by mid-July. Other elements of the biota,
Table 2.— Food of adult Sebastes mystinus relative to near-surface planl<ton during upwelling episodes of the upwelling
season, n = 7.
Food organism
In diet
In planl<ton
Size
(mm)
%
occur.
X %
vol.
Size
(mm)
%
occur.
X %
vol.
Rank
Taxa (ranl< index)
X no.
X no.'
1
PLANTS (3751 .20)
NR2
72
NR
52.1
NR
NR
NR
NR
Nereocystis sori
15-20
15
1.10
10.6
NR
NR
NR
NR
Porphyra sp.
NR
15
NR
9.9
NR
NR
NR
NR
Smithora naidum
NR
21
NR
17.0
NR
NR
NR
NR
Others
NR
21
NR
14.6
NR
NR
NR
NR
2
PELAGIC HYDROZOA (1373.30)
Hydromedusae
10-30
21
4.51
14.5
<1-4
86
120.08
2.3
Eutonina indlcans
—
0
0.00
0.0
—
0
0.00
0.0
Others
—
0
0.00
0.0
<1-2
86
68.65
0.6
Syphonophora
Muggiaea atlantica
—
0
0.00
0.0
—
0
0.00
0.0
Stephanomia bijuga
—
0
0.00
0.0
—
0
0.00
0.0
Others
—
0
0.00
0.0
3-4
14
51.43
1.7
Chondrophora
Velella velella
10-30
21
4.51
14.5
NA^
NA
NA
NA
3
MYSIDACEA (203.15)
5-7
3
10.26
6.6
—
0
0.00
0.0
Acanthomysis spp."*
5-7
3
10.26
6.6
—
0
0.00
0.0
Others
—
0
0.00
0.0
—
0
0.00
0.0
4
SCYPHOZOA (61.60)^
NR
8
NR
7.7
NR
NR
NR
NR
Fragments
NR
8
NR
7.7
NR
NR
NR
NR
5
EUPHAUSIACEA (35.10)
5-8
15
1.95
1.2
<1-4
57
363.60
3.1
Larvae
—
0
0.00
0.0
<1-4
57
363.60
3.1
Thysanoessa spp.
5-8
15
1.95
1.2
—
0
0.00
0.0
Others
—
0
0.00
0.0
—
0
0.00
0.0
6
GAMMARIDEA (1.80)
1-8
8
0.75
0.3
1-8
86
10.31
5.3
Atylus tridens
—
0
0.00
0.0
2-8
29
4.63
5.0
Ishyrocerus n. sp.
4
3
0.31
0.1
—
0
0.00
0.0
Jassa falcata^
1-3
8
0.28
0.1
<1-3
14
0.78
NR
Polycheria osboumi
—
0
0.00
0.0
—
0
0.00
0.0
Others
2-8
5
0.16
0.1
1-2
57
4.90
0.3
7
CALANOIDA (0.90)
5-6
8
0.16
0.7
<1-7
100
2454.09
38.6
Nauplii
—
0
0.00
0.0
<1
14
72.00
0.1
Acartia spp.
—
0
0.00
0.0
1-2
52
224.23
4.4
Calanus pacificus
—
0
0.00
0.0
2-3
57
24.68
5.3
Eucalanus californlcus
6
8
0.13
0.6
4-7
71
35.48
1.4
Rhincalanus nasutus
—
0
0.00
0.0
—
0
0.00
0.0
Others'
5
3
0.03
0.1
<1-3
1
2097.70
27.4
722
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
however, reached maximum development during the
summer. Prominent among them were benthic
algae, e.g., Desmarestia ligulata and Laminaria set-
chelli, as well as certain sedentary animals including
various sponges, e.g., Leucilla nuttingi; ascidians,
e.g., Trididemnum opacum; hydroids, e.g., Obelia
spp.; and bryozoans, e.g., Bugula spp. The benthic
caprellids and gammarideans attained peak numbers
during early summer, when they literally carpeted
some areas of the seafloor. Samples taken with an
airlift during July 1978 measured densities of over
10,000 caprellids (mostly Metacaprella kennerleyi),
and 108,000 gammarids (mostly Jassa spp.) in m-
quadrats. Their numbers declined sharply during
August, however, and by September they occurred
only in scattered patches.
Feeding Conditions
The diet of S. mystinus relative to foods present
during the upwelling season was assessed with
samples of gut contents and near-surface plankton
taken during 7 upwelling episodes (Table 2) and 10
downwelling episodes (Table 3).
More prey were consumed during the downwell-
ing episodes (e.g., x no. prey taken = 110.8, vs.
20.1 during upwelling episodes). Thaliacians (Fig.
6) were the primary food during the upwelling
season, but all were taken during downwelling
episodes— the only times when the guts were packed
with food. These relatively large, gelatinous zoo-
plankters did not occur either in the plankton or in
the diet of S. mystinus during upwelling episodes.
Table 2— Continued.
In diet
In plankton
Food organism
^JTO
%
occur.
xVo
vol.
Size
(mm)
%
occur.
x%
vol.
Rank Taxa {ranl< index)
(mm)
X no.
X no.^
8 CAPRELLIDEA (0.54)
10
5
1.08
0.1
6
14
2.25
0.4
Metacaprella kenneriyi
10
5
1.08
0.1
—
0
0.00
0.0
Others
—
0
0.00
0.0
6
14
2.25
0.4
9 CHAETOGNATHA (0.09)
23
3
0.03
1.0
3-12
57
20.80
2.1
Undetermined species
23
3
0.03
1.0
3-12
57
20.80
2.1
10 FISHES (0.02)
12
3
0.05
0.1
2-14
57
14.92
1.1
Larvae
12
3
0.05
0.1
2-14
57
14.92
1.1
OTHER CATEGORIES
Polychaeta
—
0
0.00
0.0
<1-12
86
74.83
1.4
Molluscan larvae
—
0
0.00
0.0
<1-1
57
149.92
1.7
Pelagic gastropoda
—
0
0.00
0.0
1-5
29
2.32
8.6
Cladocera
—
0
0.00
0.0
1
29
64.28
0.7
Harpacticoida
—
0
0.00
0.0
1-2
1
54.77
0.7
Cyclopoidea
—
0
0.00
0.0
1
71
75.08
0.9
Cirripedean larvae
—
0
0.00
0.0
<1-1
1
1850.92
15.0
Reptantian larvae
—
0
0.00
0.0
<1-2
86
134.48
1.9
Natantian larvae
—
0
0.00
0.0
2-3
43
72.00
1.4
Larvacea
—
0
0.00
0.0
3-8
43
53.48
1.1
Eggs, undetermined
—
0
0.00
0.0
<1
43
3024.00
9.7
Eggs, fish
—
0
0.00
0.0
<1-2
71
312.68
1.7
UNIDENTIFIABLE MATERIAL^
—
—
—
15.9
—
—
—
—
No. fish examined: 51
No.
plankton
collections: 7
224-336, X = 284.3 mm SL
X no. zooplankters: 4940.57
No. empty = 13
X no. prey: individuals = 20.13
taxa = 2.03
'Value is estimated mean number per 100 m^ of water, based on water filtered (54.8 m^) during the 5-min collection.
2NR = not recorded. The enumeration was either omitted or unfeasible,
^Velella velella floats on the water's surface, where it was not effectively sampled by our net.
"Most mysids sampled were Acanthomysis sculpta.
^Adult S mystinus often were seen feeding on large individuals of Cyanea capillata (Fig. 5), which were avoided by us during plankton
collections because they would have made collections unmanageable.
^According to Kathleen Conlan (National IVIuseum of Canada, P.O. Box 3443, Station D, Ottawa, Canada KIP 6P4, pers. commun.
26 f^ay 1 987), Jassa falcata does not occur in California, and forms along the coast considered to be this species (including the form(s)
referred to here) are undescribed.
't^any of the calanoids from the plankton included in this category were juveniles and other undetermined stages of the species
distinguished above. Most were at the lower end of the size range indicated.
^Foods digested beyond recognition.
723
FISHERY BULLETIN: VOL. 86. NO. 4
«IN
•
.'^g'Vrf.iM:^w.\v'^.']'{-;i^>^^V*?"''^.
»'^L.^yi-.f:-iK'i;.-x • wf.
Figure 5.— Adult Sebastes mystinus feeding on a scyphozoan, Cyanea capillata.
Table 3.— Food of adult Sebastes mystinus relative to near-surface plankton during downwelling episodes of the
upwelling season, n = 10.
Food organism
In
diet
In plankton
- Size
(mm)
%
occur
X no.
X %
vol.
Size
(mm)
%
occur.
X no.'
X %
vol.
Rank
Taxa (rank index)
1
2
THALIACEA (124990.86)
Undetermined species
PLANTS (2030.40)
Nereocystis sori
Porphyra sp.
Smithora naidum
Others
1-10
1-10
NR2
NR
NR
NR
NR
47
47
54
18
22
15
28
76.20
76.20
NR
NR
NR
NR
NR
34.9
34.9
37.6
9.7
11.5
5.5
10.9
3-45
3-45
NR
NR
NR
NR
NR
30
30
NR
NR
NR
NR
NR
130.70
130.70
NR
NR
NR
NR
NR
8.9
8.9
NR
NR
NR
NR
NR
724
HOBSON and CHESS; TROPHIC RELATIONS OF THE BLUE ROCKFISH
Table 3.— Continued.
Food organism
In diet
In plankton
Size
%
x%
Size
%
x%
Rank
Taxa (rank index)
(mm)
occur.
X no.
vol.
(mm)
occur.
X no.^
vol.
3
CALANOIDA (778.78)
3-6
20
16.93
2.3
<1-8
100
5557.10
31.9
Nauplii
—
0
0.00
0.0
<1-1
50
594.00
1.8
Acartia spp.
—
0
0.00
0.0
1-2
1
1121.00
4.6
Calanus pacificus
—
0
0.00
0.0
2-5
70
105.84
2.4
Eucalanus californicus
4-6
20
16.92
2.3
1-8
70
796.68
14.6
Rhincalanus nasutus
—
0
0.00
0.0
5-6
10
92.88
2.7
Others^
3
1
0.01
<0.1
<1-3
1
2846.70
5.8
4
HYPERIIDEA (405.13)
2-13
49
6.36
1.3
1-7
60
17.10
0.6
Hyperoche medusarum
2-6
5
0.38
0.1
1-7
20
9.72
0.3
Vibilia spp.
5
24
4.35
0.7
—
0
0
0
Others
2-13
23
1.63
0.5
1-3
40
7.38
0.3
5
PELAGIC GASTROPODA (334.66)
Heteropoda
1-20
16
5.81
3.6
1-18
20
37.44
1.1
Caranaria japonica
10-20
8
2.35
2.0
—
0
0.00
0.0
Pteropoda
Corolla spectabilis
10-15
11
3.45
1.6
15-18
10
4.32
0.8
Limacina helicina
3
1
0.01
<0.1
1-2
20
33.12
0.3
Others
—
0
0.00
0.0
—
0
0.00
0.0
PELAGIC HYDROZOA (318.24)
Hydromedusae
Eutonina indicans
Others
Syphonophora
Muggiaea atlantica
Stephanomla bijuga
Others
Chondrophora
Velella velella"
LARVACEA (1.12)
Undetermined
POLYCHAETA (0.87)
Larvae
Postlarvae
FISHES (0.73)
Larvae
10
SCYPHOZOA (0.02)
Fragments
OTHER CATEGORIES
Molluscan larvae
Cladocera
Harpacticolda
Cirripedean larvae
Gammaridea
Euphausiacea
Reptantian larvae
Natantian larvae
Chaetognatha
Eggs, undetermined
Eggs, fish
UNIDENTIFIABLE MATERIAL^
No. fish examined: 85
204-335, X = 266.2 mm SL
No. empty = 8
X no. prey: individuals = 110.84
taxa = 3.35
2-35
8-9
NR
2-16
25-35
2-7
2-7
4-25
4-25
8-45
8-45
NR
NR
NR
NR
1-8
2-12
NR
30
0
0
5
5
14
7
8
8
19
0
19
4
4
3
3
1
0
3
0
12
9
0
0
0
12
0
1.56 6.8
0.00
0.00
0.27
0.31
0.51
0.47
0.70
0.70
0.46
0.00
0.46
0.14
0.14
0.01
0.01
0.01
0.00
0.03
0.00
0.32
0.11
0.00
0.00
0.00
0.74
0.00
0.0
0.0
0.2
0.2
2.7
3.7
0.2
0.2
0.1
0.0
0.1
1.3
1.3
0.5
0.5
<0.1
0.0
NR
0.0
0.5
0.7
0.0
0.0
0.0
0.3
0.0
1-20
10-20
1-5
6-9
2-6
NA"*
2-6
2-6
<1-7
<1-4
7
2-11
2-11
NR
NR
«1-<1
«1-<1
<1-2
<1-2
<1-3
<1-11
<1-3
1-6
6-18
«1-<1
<1-2
70
10
50
10
0
20
NA
70
70
70
70
10
50
50
NR
NR
70
40
70
100
90
60
70
60
50
50
60
240.66 8.8
0.54
187.56
2.88
0.00
49.68
NA
142.74
142.74
103.50
90.36
13.14
4.50
4.50
NR
NR
631.08
115.92
55.25
1372.44
12.24
229.86
119.52
43.20
18.54
2921.40
48.96
— — — 11.0
No. plankton collections: 10
X no. zooplankters: 6622.50
0.3
5.8
0.6
0.0
2.1
NA
2.2
2.2
0.7
0.3
0.4
0.4
0.4
NR
NR
4.9
1.7
1.1
11.8
0.6
1.0
2.1
0.7
1.2
5.0
0.9
'Value is estimated mean number per 100 m^ of water, based on water filtered (54.8 m^) during the 5-min collection.
2NR = not recorded. Ttie enumeration was eithier omitted or unfeasible.
^IVIany of the calanoids from tfie plankton included in this category were juveniles and other undetermined stages of the species
distinguished above. Most were at the lower end of the size range indicated
^Velella velella floats on the water's surface, where it was not effectively sampled by our net.
^Foods digested beyond recognition.
725
FISHERY Hl'LLKTlN: VOL. 86. NO, 4
Figure 6.— Thaliacean, Cyclosalpa bakeri, off Mendocino. Solitary individual (upper) and aggregate of individuals (below).
Scale indicator = 1 cm.
726
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
Plant material ranked as the top food-category
during upwelling episodes (Table 2) and as the
second-ranked food during downwelling episodes
(Table 3). In both cases, certain algae dominated the
diet on days when offshore zooplankters were in
short supply during the latter part of the upwelling
season (upwelling episodes on 8 August 1978, 12
June 1979, and 17 June 1981; downwelling episodes
on 24 July 1979 and 27 August 1979; see Table 1).
Algae were not taken earlier in the season, however.
Thus, while the six individuals collected during a
plankton-poor upwelling episode of 8 August 1978
(Table 1) were full of algae (90% of gut contents)
and epibenthic crustaceans (10%), all eight collected
during a plankton-poor upwelling episode of 9 April
1981 (Table 1) were empty. Similarly, the occurrence
of algae in the diet during downwelling episodes
(Table 3) is based mostly on 20 adults collected dur-
ing two days (24 and 27 August 1979) when en-
vironmental conditions indicated downwelling but
an absence of offshore zooplankters was noted
(Table 1). There was no indication of plants being
ingested for epiphytic animals.
Of the wide variety of plant materials in the diet,
only the three most frequently ingested forms ap-
peared to some extent digested. These were the sori
of N. leutkeana, Porphyra nereocystis (an epiphyte
on A^. leutkeana), and Smithora naidum (an epiphyte
on certain seagrasses). The sori of A'^. leutkeana
developed as variably sized areas (typically about
50-150 mm long) in fronds near the water's surface.
They dropped from the fronds when mature, and
we saw several ingested by 5. mystinus as they
drifted toward the bottom. Sori recovered from the
intestines of S. mystinus were more translucent
than sori from the stomach, but comparison under
magnification of sectioned material from both
regions of the gut (Fig. 7) indicated that only zoo-
spores were digested. Ingested fragments of P.
nereocystis and S. naidum, both species having
monostromatic thalli, ranged from intact, but flaccid
and blanched, to disintegrating. Ingested S. naidum
often were attached to pieces oiZostera marina or
Phyllospadix torreyi, but these seagrasses never
evidenced digestion. Nor did other plant forms pres-
ent among the gut contents appear to be digested,
including D. ligulata and A^. leutkeana (vegetative
tissue), along with various unidentified phaeophytes
and rhodophytes. Thus, plants contributed less food
than is indicated by the tables and histograms,
where their rank is inflated by undigested materials.
Zooplankters taken during upwelling episodes dif-
fered from early to late in the season, whereas those
taken during downwelling episodes remained much
the same throughout (as listed in Table 3). Thus, of
the food categories listed for the upwelling condi-
tion (Table 2), pelagic hydrozoans, mysids, and
scyphozoans were taken only during the spring,
whereas euphausiids and caprellids were taken only
in summer. The only notable departure from the
downwelling condition depicted in Table 3 occurred
on 24 April 1979 (Table 1), when conditions were
at the time judged to be mixed. Upwelling had been
unusually weak during the first month of that up-
welling season, and there was no wind at the time
of sampling. The sea was calm and, at 10°C, warm
for April. Although these conditions usually indicate
downwelling (which is why the data are assigned to
that category), we neither collected nor saw organ-
isms typical of downwelHng conditions. Further-
more, of the nine adult S. mystinus (224-313, x
= 277.7 mm SL) collected, eight (89%) were empty.
The ninth contained one sorus from A'^. leutkeana
(an unusually large number of A'', leutkeana had per-
sisted through the previous winter, which had been
exceptionally mild) and also organisms not seen or
collected by us in the environment at the time: two
Corolla spectabilis (a pelagic gastropod typical of
downwelling conditions) and six Velella velella (a
pelagic hydrozoan typical of upwelling conditions).
The latter float on the water's surface (Fig. 8), and
often we saw adult S. mystinus break the surface
to feed on them.
The Downw^elllng Season
Between late August and mid-September it
became evident that transition to the downwelling
season was under way. Winds had become light and
variable (but generally remained either northerly or
southerly), and for a growing number of days at a
time, the water was notably blue and transparent.
Records of sea temperatures and wind for the
1980-81 downwelling season illustrate how occur-
rences of downwelling and upwelling related to
prevailing wind during that period (Fig. 9).
Habitat Conditions
The downwelling season developed with less wind
from the north and more wind from the south, but
either way with winds that tended to be light, so
that relatively tranquil conditions prevailed. Flow-
ing into the nearshore habitat with offshore surface
waters was a rich supply of relatively large, mostly
gelatinous zooplankters that were major prey of 5.
727
FISHERY BULLETIN: VOL. 86, NO. 4
Figure 7.— Sections through surface region of sori (stained with hematoxylin) from gut of Sebastes mystinus. Upper section is from
the stomach, lower is from the intestine. Granular objects in upper section but largely absent in lower section are zoospores in zoosporangia.
mystinus. But, whereas the predominant of such
forms during spring and summer usually were thali-
aceans, during the fall they were usually cteno-
phores (Fig. 10), pelagic hydrozoans— mostly sipho-
nophores (Fig. 11), and hydromedusae. The pelagic
hydrozoan Velella velella, which had been prominent
during the upwelling season, was not seen. Pelagic
gastropods (Fig. 12), pteropods at least, also were
more abundant during fall downwelling. On the
other hand, these waters were poor in the pelagic
diatoms that had bloomed periodically during spring
and summer. Furthermore, the nearshore habitat
had by this time lost much of its benthos. Many of
the more insecurely anchored Nereocystis plants, for
example, had been carried away by strong wave
surge that frequently swept through the nearshore
habitat even during relatively tranquil periods. Al-
though the loss of these plants greatly increased
interplant distances, Nereocystis beds remained
dominant features. This was because the plants still
in place continued to grow, to produce sori, and to
thicken the surface canopy. Probably at least partly
because the canopy's increased thickness blocked
sunlight from the seafloor below, the algal under-
story, Desmarestia lingulata in particular, was
greatly reduced. Similarly, many benthic animals
that proliferated during the previous upwelling
season had become scarce, including the amphipods
and mysids noted above to be declining during late
summer.
As the downwelling season progressed through
fall toward winter, there were major transforma-
tions of the nearshore habitat that, to at least some
extent, resulted from southwesterly storms. The
728
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
Figure 8.—Velella velella being driven shoreward by a northwest wind off the coast of northern California.
major effects of storms on the nearshore habitats
came from 1) physical force of waves and surge,
and 2) suspended sediments carried by coastal
runoff into the nearshore habitats after heavy rains.
The first of these was the more apparent, at least
in exposed locations where often many of the ben-
thic algae and sessile invertebrates were swept
away. This force, intensified when the water car-
ried abrasive sediments, swept away most of the
Nereocystis, although some plants survived the
winter in sheltered places where they continued to
produce sori. Suspended sediments had their most
obvious effects in locations sheltered from water
movement, where they frequently settled to blanket
the benthos. Probably a more profound effect of
materials in suspension, however, was reduced
transparency of the water that limited the amount
of light (already at low levels owing to the shorter
days and low sun angle) reaching phototrophic
organisms.
As a general result of these, and probably other
forces in combination, the nearshore habitat ap-
peared relatively barren during the latter stages of
the downwelling season. That storms were a major
factor in reducing the vitality of this habitat was ap-
parent during the relatively mild winter of 1978-79,
when the effects described above were reduced.
That was the only year, for example, when mysids
were noted to be conspicuous in the nearshore
habitat throughout the winter.
Feeding Conditions
The diet of S. mystinus relative to foods available
during the downwelling season was assessed dur-
ing 10 days of sampling under downwelling condi-
729
FISHERY BULLETIN: VOL. 86, NO. 4
ui
X
a:
o
20
15
10
c
Z
5
0
5
a:
UJ
o
(/5
EVENT
o
UJ
UJ
Q.
UJ
10
15
20
13 -1
12 -
11 -
10 -
9 -
8
i kkkk A A
10 U12.13.14 15 16
T fTTT TT
A,
A
17
T
AUG
SEPT
OCT
NOV
DEC
JAN
FEB
26
MAR
Figure 9.— Sea temperatures and wind off Mendocino during the 1980-81 downwelling season,
methods used in obtaining and plotting the data.
See capture of Figure 4 and text for
Event 1— Sea temperature fell to 8.5°C on 16 August after 9 consecutive days of 7-10 K northerlies, producing the last episode of
strong upwelling during the 1980 upwelling season.
Event 2— Sea temperature rose to 10.4°C on 26 August after southerlies of 5-12 K on 9 of the preceding 11 days. Then began a 6-wk
transition to the downwelling season, during which 1-4 day periods of 2-10 K northerlies, with slightly lowered sea temperatures, alter-
nated with 1-4 day periods of southerlies, with slightly elevated temperatures.
Event 3— The last bit of upwelling before the season's first major downwelling episode began on 1 4 October with 7 consecutive days
of northerlies— 20 K during the first 2, 4-8 K during the next 5.
Event 4— Sea temperatures rose to 11.2°C on 7 November after 16 straight days of 2-12 K southerlies or weak (2-5 K) northerlies.
Event 5— Sea temperatures fell to 9.7°C during a major cooling that coincided with 8 days of 3-9 K northerlies (with gusts to 20 K)
from 8 to 15 November.
Event 6— On 21 November the wind shifted to the south, with gusts to 20K, and sea temperatures began rising.
Event 7— A 3-day storm (1-3 December) with 20-30 K southerlies (and rain) briefly accelerated the rise in temperature.
Event 8— The warming trend was briefly interrupted by several 1-3 day periods of 4-10 K northerlies.
Event 9— By 22 December sea temperatures had risen to levels above 12°C that characterized the 1980-81 dowmwelling season. Although
southerlies predominated over the next 3 weeks, on some days reaching 25 K (e.g., on 25 December), there were no further large in-
creases in temperature.
Event 10— Sea temperatures anomalously fell during the first 2 weeks of January, even though winds during that period were mostly
southerlies of up to 10 K.
Event 11— An 8-day storm (15-22 January) with 12-20 K southerlies (and rain) resulted in sea temperatures rising to 12.7°C— the warmest
of the year (equaled 3 weeks later).
730
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
Figure 10.— Ctenophores off Mendocino. Beroe forskali (left), missing piece probably taken by planktivore, and Pleurobrachia hachei
(right). Scale indicator = 1 cm.
Event 12— The process abruptly reversed with a shift to northerlies that gusted to 30 K for 2 days, and sea temperature fell to 11.5°C
on 25 January.
Event 13— In another abrupt reversal, 3 days (27-29 Januarj') of 15-25 K southerlies (with rain) resulted in sea temperature rising to 12°C.
Event 14— Reversing again, northerlies of 12-20 K on 30-31 January drove sea temperatures down to 10.5°C, at which point (1 February)
the thermograph malfunctioned. Southerlies predominated while the thermograph was inoperative, and when it was reinstalled on 11
February, rising sea temperature 2 days later (13 February) equaled the season high of 12.5°C. But when northerlies returned the next
day the downwelling season began its decline.
Event 15— Two days of 20-25 K northerlies (19-20 February) resulted in the season's largest 24-h change in sea temperatures.
Event 16— Sea temperature rose to 11.5°C on 3 March following 8 consecutive days of 8-15 K southerlies (with rain), or weak north-
erlies (to 6 K). An abrupt shift followed, as 2 days of 15-25 K northerlies (4-5 March) resulted in the season's second largest 24-h change
in sea temperature— the first time since November that sea temperature had fallen below 10°C.
Event 17— After a week of light variable winds, during which sea temperatures remained essentially unchanged at about 9.8°C, southerlies
increased to 15-20 K (on 15 March) and sea temperatures began rising to begin the final downwelling episode of the 1980-81 down-
welling season (which was followed by the abrupt reversal that marked the beginning of the 1981 upwelling season, see Figure 4).
731
FISHERY BULLETIN: VOL. 86, NO. 4
Figure 11.— Siphonophores off Mendocino. The physonect Stephanomia bijuga (left), and the calycophore Praya dubia (right).
indicator = 1 cm.
Scale
tions, but because fall and winter were so different,
we consider them separately. We did not observe
upwelling conditions during this period.
The major foods during six days of fall downwell-
ing, based on the ranking indices (Table 4), were
pelagic hydrozoans, specifically siphonophores, but
vegetation comprised a larger part of the total diet
volume. Virtually all plant materials taken at these
times, however, were sori of A'', leutkeana. As was
true during the upwelling season, there tended to
be more vegetation in the diet when there were
fewer of the larger gelatinous zooplankters in the
water column. For example, during the sampling
session of 18 October 1978 (Table 1), when the sur-
face plankton-collections took 400 siphonophores
and ctenophores, only 1 of the 9 fish collected had
consumed plant material (one sorus ofN. leutkeana).
On the other hand, during the sampling session of
16 October 1979 (Table 1), when the surface plank-
ton-collection took only 30 siphonophores and cteno-
phores, 10 of 13 fish collected had consumed vege-
tation (x diet volume = 80%, virtually all of it
sori of A^. leutkeana; number taken = 1-21, x =
9.0), and two others were empty.
Our assessment of the diet and concurrent com-
position of the plankton during winter downwelling
is limited to four days of sampling (Table 5). Data
from these collections are combined for consistency
Figure 12.— Pelagic gastropods off Mendocino. The heteropod
Caranaria japonica (upper), and the pteropod Corolla spectabilix
(lower). Scale indicator = 1 cm. Often in areas where blue rockfish
are feeding many of the C. spectabilis present have lost the bulbous
central part of their body (pseudoconch and viscera).
732
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
733
FISIIKKY Hl'LLKTIN: VOL. 86, NO. 4
Table 4. — Food of adult Sebastes mystinus relative to near-surface plankton during downwelling episodes of fall.
n = 6.
Food organism
In diet
In plankton
Size
(mm)
O/n
y o/n
J^i7P
o/n
xo/o
vol.
Rank
Taxa (rank index)
occur.
X no.
vol.
(mm)
/u
occur.
X no.^
1
PELAGIC HYDROZOA (10455.98)
Hydromedusae
2-15
46
18.04
12.5
1-25
83
1747.12
26.5
Eutonina indicans
18
1
0.03
<0.1
18-25
33
1516.81
14.7
Others
15
4
0.02
<0.1
<1-20
33
26.10
0.2
Sypfionophiora
Muggiaea atlantica
6-15
10
0.34
0.7
5-12
67
15.83
2.0
Stephanomia bijuga
10
17
1.33
0.8
10
17
2.39
0.3
Otfiers
2-15
31
16.32
10.9
1-8
4
185.99
9.3
Cfiondropfiora
Velella velella^
—
0
0.00
0.0
NA^
NA
NA
NR
2
PLANTS (4123.00)
NR^
54
2.06
37.1
NR
NR
NR
NR
Nereocystis sori
NR
51
1.94
35.6
NR
NR
NR
NR
Porphyra sp.
NR
2
NR
0.2
NR
NR
NR
NR
Smithora naidum
NR
6
0.06
<0.1
NR
NR
NR
NR
Others
NR
6
0.06
1.3
NR
NR
NR
NR
3
CTENOPHORA (1742.66)
3-15
33
5.74
9.2
<1-10
0.67
295.79
5.4
Pleurobrachia bachei
3-15
31
5.73
9.2
<1-10
50
288.59
5.2
Others
12
1
0.01
<0.1
3
17
7.20
0.2
4
PELAGIC GASTROPODA (554.00)
Heteropoda
2-160
25
3.20
6.8
<1-45
100
187.82
8.4
Caranaria japonica
12-160
15
0.33
2.1
30-45
17
1.21
0.7
Pteropoda
Corolla spectabilis
5-22
16
2.69
4.7
2-20
50
72.90
7.5
LImacina helicina
2
1
0.18
<0.1
1-3
17
109.80
0.2
Others
—
0
0.00
0.0
<1-3
17
3.91
<0.1
5
SCYPHOZOA (89.96)
10-35
18
0.51
9.8
NA
NA
NA
NA
Fragments
10-35
18
0.51
9.8
NA
NA
NA
NA
6
THALIACEA (31.97)
5-29
16
0.54
3.7
2-5
17
3.01
NR
Undetermined species
5-29
16
0.54
3.7
2-5
17
3.01
NR
7
EUPHAUSIACEA (22.93)
4-10
21
0.78
1.4
<1-9
67
162.90
2.2
Larvae
4
1
0.01
<0.1
4
50
129.01
1.2
Thysanoessa spp.
4-10
19
0.74
1.4
5-9
33
33.89
1.0
Others
8
1
0.03
<0.1
—
0
0.00
0.0
with the other tables and figures, but the combina-
tions obscure the great variation in feeding condi-
tions at this time of year. For example, taxa of the
major food categories listed in Table 5 were numer-
ous in the diet and plankton only during the two
sampling sessions in February and March of 1979.
Of the 14 fish collected at those times, all but one
was well fed (x no. prey = 26.7), with the excep-
tion being a pregnant female whose gut was empty.
Thaliaceans dominated on these occasions, both in
the diet and in the plankton, and hyperiid amphi-
pods, Vihilia spp. (which are parasites of thaliaceans
(Laval 1980)), were similarly abundant. In contrast,
the collecting session of January 1980 indicated
there were more zooplankters in the water column,
but that they were exceptionally small. The plank-
ton collection took 1,488 zooplankters (compared
with 109 and 715 in the two 1979 collections), but
only 2% were of species that occurred as large as
2 mm (compared with 86% in 1979). That these small
zooplankters were unsuitable as prey of adult S.
mystinus is implicit in the fact that of 16 fish col-
lected, only 3 contained food— all of it the alga Por-
phyra sp. (The other 13 represented 76% of all fish
with empty guts in the winter collections.) Signif-
icantly, of the taxa identified as food of adult
S. mystinus during the winter (Table 5), only one,
the calanoid Calanus pacificus, was represented in
the January 1980 plankton collection. Conditions
were intermediate during the sampling session of
January 1981, when some of the Usted food taxa
occurred in both diet and plankton (though in
reduced numbers) and five of eight fish sampled con-
tained food.
734
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
Table a.— Continued.
In diet
t
In plankton
Food organism
— Qi70
%
occur.
x%
vol.
Size
(mm)
%
occur.
X %
vol.
Ranl< Taxa (rank index)
(mm)
X no.
X no.^
8 CALANOIDA (14.32)
2-7
22
0.93
0.7
<1-7
100
9528.11
28.3
Nauplii
—
0
0.00
0.0
<1-1
33
68.99
<0.1
Acartia spp.
—
0
0.00
0.0
<1-1
83
1751.42
4.3
Calanus pacificus
3
6
0.13
<0.1
<1-4
83
4609.21
15.8
Eucalanus californlcus
4-7
15
0.67
0.7
2-7
100
255.39
1.2
Rhincalanus nasutus
4
4
0.06
<0.1
—
0
0.00
0.0
Others"
2-3
3
0.07
<0.1
<1-3
100
2843.10
7.0
9 HYPERIIDEA (8.28)
1-8
24
1.15
0.4
1-7
83
292.19
0.4
Hyperoche medusarum
2
3
0.16
0.1
1-7
50
273.01
0.2
Vibilia spp.
—
0
0.00
0.0
—
0
0.00
0.0
Others
1-8
21
0.99
0.3
1-3
33
19.18
0.2
10 POLYCHAETA (3.20)
4-40
16
0.40
0.5
<1-7
83
118.21
1.7
Larvae
—
0
0.00
0.0
<1-4
33
103.81
1.5
Postlarvae
4-40
16
0.40
0.5
3-7
50
14.40
0.2
OTHER CATEGORIES
Molluscan lan/ae
—
0
0.00
0.0
«1-<1
50
189.59
0.7
Cladocera
—
0
0.00
0.0
<1-1
33
142.20
0.8
Cyclopoida
NR
1
0.01
NR
<1-1
67
648.00
4.0
Clrripedean larvae
—
0
0.00
0.0
«1-<1
67
128.70
0.2
Reptantian larvae
—
0
0.00
0.0
<1-6
83
52.20
0.3
Natantian larvae
—
0
0.00
0.0
1-5
33
237.60
NR
Bryozoan larvae
—
0
0.00
0.0
<1
33
244.49
0.2
Larvacea
2
3
0.12
0.1
<1-4
67
844.20
2.8
Chaetognatha
10-11
3
0.06
0.2
5-20
83
225.59
5.0
Eggs, undetermined
—
0
0.00
0.0
<1
17
1535.99
4.3
Eggs, fish
3
1
0.01
<0.1
<1-2
33
40.19
4.4
Fishes
6-24
4
0.04
0.2
4-20
33
5.40
0.7
UNIDENTIFIABLE MATERIAL^
—
—
—
17.4
—
—
—
—
No. fish examined: 73
No. plankton colled
:ions: 6
218-350, X = 290.3 mm SL
X
no. zooplankters:
9663.0
No. empty = 7
X no. prey: individuals = 32.4
taxa = 3.6
'Value Is estimated mean number per 100 m^ of water, based on water filtered (54.8 m^) during the 5-min collection.
^Velella velella floats on the water's surface, where it was not effectively sampled by our net.
^NR = not recorded. The enumeration was either omitted or unfeasible.
■■Many of the calanoids from the plankton included in this category were juveniles and other undetermined stages of the species
distinguished above Most were at the lower end of the size range indicated.
^Digested beyond recognition.
DISCUSSION
It is clear that wind-driven movement of the sur-
face water profoundly influences feeding by Sebastes
mystinus off northern California. Water set in
motion by the wind can be tens of meters deep
(Bakun 1973; Barber and Smith 1981), and so carries
most of the foods of nearshore planktivores. The
movement is seaward (with upwelling) under north-
erly winds and shoreward (with downwelling) under
southerly winds. Thus, with winds along the Men-
docino coast being northerly or southerly about
80% of the time (based on records of the NOAA
weather station there), adult S. mystinus in that
area alternate between periods when planktonic
foods are being carried into their habitat and pe-
riods when these foods are being carried away.
This perception of alternations between upwell-
ing and downwelling is simplified, perhaps overly
so, to emphasize features we consider essential to
the feeding of S. mystinus, and also because details
of what clearly is a complex oceanographic system
remain unclear. In particular, we stress the impor-
tance of shoreward surface transport in carrying
prey to S. mystinus in nearshore habitats. The major
prey of adult 5. m^s^iwws— thaliaceans, pelagic
hydrozoans, and other relatively large, gelatinous
zooplankters— tend to be concentrated in areas of
oceanic convergence, and dispersed in areas of
oceanic divergence (e.g., Bakun and Parrish 1980).
Thus, when offshore surface waters converge on the
coast, the planktonic foods of S. mystinus become
concentrated near shore.
735
FISllKKV HILLKTIN; VOL. 86, NO. 4
Table 5.— Food of adult Sebastes mystinus relative to near-surface plankton during downwelling episodes of winter.
n = 4.
Food organism
In
diet
In plan
ikton
*^i7P
O/n
X %
'^i7P
O/n
k- 0/n
Rank
Taxa (rank index)
(mm)
/u
occur.
X no.
vol.
(mm)
/u
occur.
X no.'
A /U
vol.
1
THALIACEA (29034.03)
7-20
71
12.90
31.7
5-45
50
9.90
34.8
Undetermined species
7-20
71
12.90
31.7
5-45
50
9.90
348
2
HYPERIIDEA (3278.81)
1-5
62
23.40
2.2
1-6
75
2.25
0.8
Hyperoche medusarum
—
0
0.00
0.0
4
25
0.45
NR3
Vibilia spp.
1-5
57
23.06
2.1
5-6
25
0.90
0.8
Others
6
5
0.34
0.1
1
25
0.90
NR
3
EUPHAUSIACEA (1567.61)
6-12
43
11.76
3.1
10
75
2.25
12.5
Larvae
—
0
0.00
0.0
—
0
0.00
0.0
Thysanoessa spp.
6-12
38
11.67
3.1
10
75
2.25
12.5
Others
6
5
0.09
<0.1
—
0
0.00
0.0
4
SCYPHOZOA (1054.20)
NR
38
1.43
19.4
NR
NR
NR
NR
Fragments^
NR
38
1.43
19.4
NR
NR
NR
NR
5
PELAGIC GASTROPODA (726.34)
Heteropoda
6-30
48
2.91
5.2
1-20
25
8.10
0.8
Caranaha japonica
15-30
10
0.24
0.5
—
0
0.00
0.0
Pteropoda
Corolla spectabilis
6-20
48
2.67
4.7
20
25
0.45
0.8
Limacina helicina
—
0
0.00
0.0
1-2
25
7.65
NR
Others
—
0
0.00
0.0
—
0
0.00
0.0
6
PELAGIC HYDROZOA (437.91)
Hydromedusae
3-13
43
1.52
6.7
1-6
75
14.85
1.3
Eutonina indicans
—
0
0.00
0.0
—
0
0.00
0.0
Others
10-22
40
0.33
3.5
1-4
50
12.60
0.5
Syphonophora
Muggiaea atlantica
8-13
38
0.86
2.1
6
50
1.35
0.8
Stephanomia bijuga
—
0
0.00
0.0
—
0
0.00
0.0
Others
3-20
19
0.33
1.1
3
25
0.90
NR
Chondrophora
Velella velella
—
0
0.00
0.0
NA'
NA
NA
NA
7
PLANTS (380.00)
NR
38
NR
10.0
NR
NR
NR
NR
Nereocystis sori
NR
5
0 10
0.1
NR
NR
NR
NR
Porphyra sp.
NR
19
NR
5.2
NR
NR
NR
NR
Smithora naidum
—
0
0.00
0.0
NR
NR
NR
NR
Others
NR
19
NR
4.7
NR
NR
NR
NR
Shoreward transport can be either wind-driven
(Ekman transport), or result simply from relaxation
of the forces that drive up welling. But in either case
our observations indicate that shoreward flowing
surface waters override the colder waters near
shore, a process we refer to as downwelling. Usually
the term downwelling is limited to conditions that
result from shoreward Ekman transport (e.g.. Gross
1977), but we have found that relaxation of upwell-
ing has essentially the same effect on the nearshore
ecosystem, the difference being simply in degree of
effect.
Some studies have concluded that warming of the
nearshore surface waters during relaxation of up-
welling results from alongshore advection (e.g.,
Send et al. 1987), but even though zooplankters
entering our study area during downwelling gen-
erally moved southward along the coast, the char-
acteristic presence of such forms as thaliaceans,
ctenophores, and pteropods indicate that the advec-
tion is from offshore. So despite the complexities
of circulation and mixing that occur in the coastal
waters off northern California (e.g., Winant et al.
1987), the net result affecting the trophic relations
of S. mystinus are alternations between seaward
and shoreward transport.
These water movements follow a strong seasonal
pattern that is evident in upwelling indices for lat.
39°N (which crosses Mendocino) produced by the
Pacific Fisheries Environmental Group of the South-
west Fisheries Center, NMFS, NOAA. In addition
to the seasonal trend, short-term episodes of sea-
ward and shoreward transport produce day-to-day,
even hour-to-hour, changes in the foods available to
736
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
Table 5.— Continued.
In diet
In plankton
Food organism
Qi70
%
x%
Size
%
X %
— oize
Rank Taxa (rank index)
(mm)
occur.
X no.
VOL
(mm)
occur.
X no.'
vol.
8 CTENOPHORA (4.73)
11-20
10
0.43
1.1
0
0.00
0.0
Plerobrachia bachei
11
5
0.09
0.1
—
0
0.00
0.0
Others
20
5
0.34
1.0
—
0
0.00
0.0
9 MYSIDACEA (4.47)
6-20
14
0.29
1.2
3-8
50
1.80
0.5
Acanthomysis sculpts
6
5
0.19
1.0
—
0
0.00
0.0
Others
18-20
10
0.10
0.2
3-8
50
1.80
0.5
10 CALANOIDA (1.18)
2-5
14
0.28
0.3
1-5
100
684.00
25.6
Nauplii
—
0
0.00
0.0
—
0
0.00
0.0
Acartia spp.
—
0
0.00
0.0
<1-1
25
229.50
7.5
Calanus pacificus
2-3
14
0.14
0.1
1-3
100
110.70
3.8
Eucalanus callfornicus
5
10
0.14
0.2
3-5
50
12.60
0.8
Rhincalanus nasutus
—
0
0.00
0.0
—
0
0.00
0.0
Others
—
0
0.00
0.0
1-2
100
331.20
13.5
OTHER CATEGORIES
Molluscan larvae
—
0
0.00
0.0
«1-<1
50
38.70
2.5
Cyclopoida
—
0
0.00
0.0
1
75
61.20
1.5
Cirripedean larvae
—
0
0.00
0.0
1-2
75
24.75
0.5
Reptantian larvae
4
<1
0.05
<0.1
<1-3
100
240.30
9.3
Larvacea
—
0
0.00
0.0
3-6
50
6.30
1.3
Chaetognatha
—
0
0.00
0.0
5-35
75
14.85
5.0
Fishes
8-10
<1
0.10
NR
3-7
75
7.65
2.8
UNIDENTIFIABLE MATERIAL^
—
—
—
19.1
—
—
—
—
No. fish examined: 38
No.
plankton collections: 4
200-345, X = 265.3 mm SL
X no. zooplankters: 634.3
No. empty: 17
X no. prey: individuals = 45.2
taxa = 4.8
'Value is estimated mean number per 100 m^ of water, based on water filtered (54.8 m^) during the 5-min collection.
^Adult S. mystmus often were seen feeding on large individuals of Cyanea capillata, wfiicfi were avoided by us during plankton col-
lections because they would have made collections umanageable.
3NR = not recorded. The enumeration was either omitted or unfeasible.
'Velella velella floats on the water's surface, where it was not effectively sampled by our net.
^Digested beyond recognition.
planktivorous fishes. The following discussion con-
siders how the diet of S. mystinus, summarized in
Figure 13, is influenced by these alternations in sur-
face transport during distinct upwelling and down-
welling seasons.
The Upwelling Season
The spring-summer upwelling season produces op-
timal feeding conditions for S. mystinus. During this
period the combined effects of increased nutrients
(from strong upwelling) and increased daylight (from
longer days, higher sun-angle, and less storm-pro-
duced sediments in suspension) result in growth of
diatom populations that constitute the food-base of
the zooplankton community.
Seaward Ekman transport in response to the
season's persistent northerly winds carries the up-
welled nutrients and increasing number of diatoms
offshore, where the response of zooplankters can be
spectacular. Consider, for example, thaliaceans,
which are a major prey of adult S. mystinus. Re-
cent study has shown that populations of Salpa
fusiformis (a common thaliacean in the California
Current) normally are food-limited, but can grow
rapidly when diatoms are abundant (Silver 1975).
In response to a diatom bloom, Thalia democratica
(another salp common in the California Current), can
increase in size by up to 10%/hour and in numbers
up to 2.5 times/day, the highest rate recorded for
a metazoan animal (Heron 1972a, b).
Zooplankters thus increased in size and number
are then carried to S. mystinus near shore by the
shoreward flow that develops with relaxation of up-
welling, or, more forcefully, with shoreward Ekman
transport under southerly winds. It remains uncer-
tain, however, whether the numbers of zooplankters
entering the Mendocino nearshore habitats are in
fact related to the productivity of local upwelling.
Wickett (1967) concluded that zooplankton abun-
737
%
100
75
50
a
FISHERY BULLETIN: VOL. 86, NO. 4
%
lOOr
75-
diet
volume
%
100
50
50
25
_^
r7777l
IX X I
._VZ1
0
o if
Percent
stomachs
empty
.^ E
75
50
25
diet
volume
100
75
50
25
Rn:^^
^
Meon
stomach
fullness
CO
.yo
O. CO
o
u c
.■5 m
Q. 3
%
TOO
75-
50
25
L_^_
0
Percent
stomochs
empty
o
.^E
Figure 13.— Seasonal variations in the diet of Sebastes mystiniLs:
a. Diet during upwelling episodes of the upwelling season, n = 51.
b. Diet during downwelling episodes of the upwelling season, n = 85.
738
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
75
diet
volume
7.
lOQ-
75-
50-
25-
%
100
75
50
25
r^^r^
r^^
^^ V7\
o tl
-5
Percent
o o
stomachs
i^"*-
empty
0, D
.■PE
%
lOOr
75
50
oLE^a_
%
100
75
50
.^YZl
25-
0
t »
o t^
o □
0) o
Percent
stomachs
empty
Figure 13.— Coniiwzterf— Seasonal variations in the diet of Sebastes mystinus :
c. Diet during fall of the downwelling season, n = 73.
d. Diet during winter of the downwelling season, n = 38.
739
FISHERY BULLETIN: VOL. 86. NO. 4
dances off California depend on nutrients from the
Gulf of Alaska, and Chelton et al. (1982) concluded
not only that phytoplankton off California depend
on nutrients from higher latitudes but also that
wind-forced coastal upwelling is relatively unim-
portant in supplying these nutrients. Despite this
uncertainty, growth in thaliaceans is so rapid that
it would seem at least many of those off the Men-
docino coast could result from local upwelling.
Regardless of what determines the growth of tha-
liaceans in this region, their appearance in large
numbers next to the beach is evidence that shore-
ward transport has developed. These animals are
readily detected by in situ observations, and also by
their occurrences in samples of both plankton and
gut contents (Table 3).
Intermittent shoreward transport during the up-
welling season seems especially strong off northern
California. This is evident in the upwelling indices
of Bakun (1973), and also in that fewer thaliaceans
get inshore at this time of year, both to the north,
off Oregon (Hubbard and Pearcy 1971), and to the
south, off central and southern California (Black-
burn 1975). Thus, it is clear that the major foods of
5. mystinus along the Mendocino coast are most
available during downwelling episodes of the upwell-
ing season.
Despite the increased productivity of the upwell-
ing season, S. mystinus experiences relatively poor
feeding conditions during that season's upwelling
episodes. Not only are fewer prey taken during
upwelling than during downwelling, a higher pro-
portion of the fish have empty stomachs (Fig. 13).
This is because the shoreward flow that transports
zooplankters from offshore during downwelling is
replaced by the seaward flow that is part of the
upwelling condition.
There is, however, one relatively large gelatinous
zooplankter from offshore that is most available as
prey during upwelling. This is the chondrophore
Velella velella, a pelagic hydrozoan known as "by-
the- wind-sailor". It is, in fact, entirely because of
this animal that the food category "Pelagic Hydro-
zoa" ranked second as food during upwelling epi-
sodes (Table 2). Because V. velella floats on the
water's surface and is equipped with a sail-like struc-
ture (Fig. 8), its movements are determined more
by wind than by current. The species includes two
forms distinguished by whether their sails are ori-
ented to the left of or to the right of the main body
axis. This orientation determines their direction in
sailing before the wind— left-handed individuals sail
to the left of the wind direction and right-handed
individuals sail to the right (Bieri 1959). Although
it has been reported that the right-handed form
predominates off California (Morris et al. 1980), all
those we examined from off Mendocino were left-
handed, and so would have been driven shoreward
by the northerly winds that generated upwelling.
In the absence of favored open-water zooplank-
ters, 5. mystinus increased consumption of near-
shore hyperbenthic zooplankton, e.g., mysidaceans
and gammarideans (Table 2). But fewer of these
organisms were taken than might be expected,
based on their great abundance during much of the
upwelling season. Acanthomysis sculpta, the mysid
most often taken as prey, typically aggregates in
large swarms within 2 m of the seafloor. To prey
on them, adult 5. mystinus must leave the upper
levels of the water column in a departure from their
usual feeding mode that may reduce feeding effec-
tiveness. In addition, most hyperbenthic zooplank-
ters probably are too small to be ready prey of these
fish. Although 5-7 mm mysids (Table 2) should be
large enough, most other taxa are less than 2 mm.
Organisms as small as 1 mm occur in the diet, in-
cluding some thought to be strictly benthonic, e.g.,
smaller of the gammaridean Jassa sp. (which also
occurred in plankton collections; Table 2). But such
forms may be ingested (and taken by plankton
nets) while attached to drifting plant fragments.
Although 5. mystinus has a smaller mouth than
most of its congeners, presumably as an adaptation
to planktivory (Hallacher and Roberts 1985), the
adults appear unable to consume the larvae of neritic
species, e.g., cirripedeans, that, with maximum
dimensions of 1 mm or less, often are the most
numerous of the zooplankton (Table 2). These lar-
vae are major prey of juvenile 5. mystinus (unpubl.
data; Gaines and Roughgarden 1987), which further
suggests it is their small size that precludes them
as prey of the adult.
Foods most often consumed in the absence of pre-
ferred zooplankters, however, were plant materials.
In fact, during upwelling episodes more plants were
consumed than anything else (Table 2), and even
during downwelling plants were the second-ranked
food category (Table 3). Although these rankings are
inflated by undigested plant tissues, certain algal
materials appear to be important foods. The avail-
ability of plant foods to supplement prey shortages
was strongly seasonal, however. Thus, the sharply
reduced availability of plants during winter and
early spring undoubtedly contributed to the preva-
lence of empty stomachs among fish collected at
those times.
740
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
Although a wide variety of plants were ingested,
S. mystinus seemed able to utilize only certain algal
tissues. These included zoospores of Nereoeystis leut-
keana and the two epiphytes, Porphyra nereocystis
and Smithora naidum. Probably this is because as
members of a carnivorous family they have only
limited abilities to digest plant material.
The zoospores of A'', leutkeana are especially vul-
nerable to S. mystinus when the sori have dropped
from the plants' fronds, because at this time the sori
have lost their epidermis (an adaptation of the
mature sori that facilitates release of the zoospores
(David C. Walker^)). Zoospores may be appropriate
food for this largely carnivorous fish because they
are, as their name implies, animal-like: they have
cell membranes but not the cellulosic cell walls
(Wilson 1952) that preclude plants as food for many
fishes (e.g., Lobel 1981). The other algae apparent-
ly utilized— Porp%ra nereocystis and Smithora
naidum— may be appropriate forage for a fish with
limited herbivorous abilities because they are mono-
stromatic plants only 25-60 fi thick (Abbott and
Hollenberg 1976). So while Gotshall et al. (1965)
reported that algae in the guts of S. mystinus gen-
erally are undigested, some forms appear to be im-
portant foods.
It is possible that plant materials are among items
of little or no food value that are ingested during
food shortages simply because at such times the
adult 5. mystinus become less discriminating in their
choice of drifting objects. On the other hand, it is
also possible that these fish have unusual herbiv-
orous abilities as a result of adaptations to a diet
rich in thaliaceans, which are among the few animals
with cellulosic tissues (Berrill 1961). This second
possibility is weakened, however, by the fact that
the cellulosic tunics of the thaliaceans appear to pass
through the gut undigested (Gotshall et al. 1965; our
observations).
The Downwelling Season
As the downwelling season developed in the fall,
S. mystinus experienced progressively poorer feed-
ing conditions. Offshore water flowing into the near-
shore habitats at these times tended to be poor in
phytoplankton (hence its transparency and blueness,
in contrast to the turbid greenness of a few months
before), and so generally lacked the herbivorous
^David C. Walker, Department of Botany, University of British
Columbia, Vancouver, British Columbia V6T 1W5, Canada, pers.
commun. 10 May 1977.
thaliaceans that were major prey during periods of
peak feeding. Relatively large gelatinous zooplank-
ters continued on occasion to be numerous in the
shoreward flow, but the species at this time tended
to be carnivores rather than herbivores, and S.
mystinus did not feed on some of them. For exam-
ple, while siphonophores made "Pelagic Hydrozoa"
the top-ranked food category (Table 4), the most
numerous pelagic hydrozoan in the water column,
Eutonina indicans, went virtually untaken. (The
small medusae visible throughout Figure 1, especial-
ly against the dark kelp, are of this species.) On the
other hand, another relatively large, gelatinous
zooplankter, the ctenophore Pleurobrachia bachei
(Fig. 10), was prominent both as prey and in the
plankton. Probably the presence of P. bachei ac-
counted for the gut-content occurrences of the
hyperiid Hyperoche medusarum, which is a parasite
of this ctenophore (Brusca 1970).
As was true during late spring and summer, cer-
tain plant materials became major foods during the
fall when favored zooplankters were in short supply.
But unlike late spring and summer, when many
different plant forms were taken, virtually the only
plant materials consumed in the fall were the sori
of N. leutkeana. This reflected a decreasing abun-
dance of other plants in the en\dronment at the time.
Nevertheless, while sporophytes of A'', leutkeana
were fewer in fall than in summer, those present
were larger, more mature, and produced more sori.
Although winter produced the poorest feeding
conditions of the year, with by far the highest in-
cidence of empty stomachs among the fish exam-
ined (Figure 13; also noted by Gotshall et al. 1965),
occasionally the waters flowing into the nearshore
habitats were rich in offshore zooplankters, in-
cluding thaliaceans, and at these times the fish fed
well. Our sampling of the highly varied winter con-
ditions was not frequent enough to recognize a pat-
tern, but they differed from fall conditions. In
general, the distinctive fall and winter conditions
off Mendocino matched the "oceanic" and "David-
son Current" oceanographic seasons defined by
Skogsberg (1936) and Bolin and Abbott (1962) for
Monterey Bay.
Ekman Transport and the Distribution
of Sebastes mystinus
Ekman transport may be important to the distri-
bution of S. mystinus. This is implicit in our finding
that alternations between seaward and shoreward
surface transport produce feeding opportunities for
741
FISHERY BULLETIN: VOL. 86, NO. 4
which the species is particularly well adapted. Along
the west coast of North America, these alternations
are best developed off northern and central Califor-
nia (Bakun 1973; Bakun et al. 1974; Mason and
Bakun 1985). And not only is S. mystinus most
numerous off this same section of the coast, it is
perhaps the dominant fish in the nearshore habitat
there (Hallacher and Roberts 1985; Bodkin 1984; our
observations). Although the species is reported from
northern Baja California, Mexico, to the Bering Sea
(Miller and Lea 1972), its numbers are sharply
reduced northward from northern California (Alver-
son et al. 1964; Frey 1971; Hart 1973) and south-
ward from central California (Hubbs 1948; Lim-
baugh 1955; Quast 1968). Although undoubtedly
other factors are involved, we suggest that occur-
rences of S. mystinus in these northern and south-
ern regions are limited by less favorable feeding
conditions.
ACKNOWLEDGMENTS
Daniel Howard assisted in the field work and
performed various aspects of data processing.
Michael Bowers prepared the histological sections
represented by Figure 7, and Louis Rouleau, NOAA
weather observer at Mendocino, provided wind data
used in Figures 4 and 9. For constructive criticism
of the manuscript we thank Andrew Bakun, John
Hunter, William Pearcy, and Richard Rosenblatt.
Rahel Fischer typed the manuscript.
LITERATURE CITED
Abbott, I. A., and C. J. Hollenberg.
1976. Marine algae of California. Stanford Univ. Press,
Stanford, 827 p.
Alverson, D. L., a. T. Pruter, and L. L. Ronholt.
1964. A study of demersal fishes and fisheries of the north-
eastern Pacific Ocean. H. R. MacMillan Lectures in Fish-
eries. Inst. Fish. Univ. British Columbia, 190 p.
Bakun, A.
1973. Coastal upwelling indices, west coast of North America,
1946-71. U.S. Dep. Commer., NOAA Tech. Rep. NMFS
SSRF-671, 103 p.
1975. Daily and weekly upwelling indices, west coast of North
America, 1967-73. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS SSRF-693, 114 p.
Bakun, A., D. R. McLain, and F. V. Mayo.
1974. The mean annual cycle of coastal upwelling off western
North America as observed from surface measurements.
Fish. Bull., U.S., 72:843-844.
Bakun, A., and R. H. Parrish.
1980. Environmental inputs to fishery population models for
eastern boundary current regions. In G. A. Sharp (editor).
Workshop on the effects of environmental variation on lar-
val pelagic fishes, p. 67-104. Workshop Rep. 28, Intergov.
Ocean. Comm., UNESCO. Paris.
Barber, R. T., and R. C. Smith.
1981. Coastal upwelling systems. In A. R. Longhurst
(editor). Analysis of marine ecosystems, p. 31-68. Acad.
Press, N.Y.
Berrill, N. J.
1961. Salpa. Sci. Am. 204:150-160.
BlERl, R.
1959. Dimorphism and size distribution in Velella and
Physalia. Nature 184:1333.
BoJE, R., and M. Tomczak (editors).
1978. Upwelling systems. Springer-Verlag, N.Y., 303 p.
Blackburn, M.
1975. Thaliacea of the California Current region: rela-
tions to temperature, chlorophyll, currents and upwelling.
CALCOFI Rep. 20:183-214.
Bodkin, J. L.
1986. Fish assemblages in Macrocystis and Nereocystis kelp
forests off central California. Fish. Bull., U.S. 84:799-808.
Bolin, R. L., and D. p. Abbott.
1962. Studies on the marine climate and phytoplankton of the
central coastal area of California, 1954-1960. CALCOFI
Rep. 9:23-45.
Chelton, D. B., p. a. Bernal, and J. A. McGowan.
1982. Large scale interannual physical and biological inter-
actions in the California Current. J. Mar. Res. 40:1095-
1125.
Ekman, V. W.
1905. On the influence of the earth's rotation on ocean cur-
rents. Ark. Mat. Astron. Fys. 2,11:1-52.
Frey, H. W. (editor).
1971. California's living marine resources and their utiliza-
tion. Calif. Dep. Fish Game, 148 p.
Gaines, S. D., and J. Roughgarden.
1987. Fish in offshore kelp forests affect recruitment to inter-
tidal barnacle populations. Science 235:479-481.
Gotshall, D. W., J. G. Smith, and A. Holbert.
1965. Food of the blue rockfishSe6astodes7nt/s<twws. Calif.
Dep. Fish Game 51:147-162.
Gross, M. G.
1977. Oceanography. 2d ed. Prentice-Hall, N.J. 497 p.
Hallacher, L. E., and D. A. Roberts.
1985. Differential utilization of space and food by the inshore
rockfishes (Scorpaenidae: Sebastes) of Carmel Bay. Environ.
Biol. Fish. 12:91-110.
Hart, J. L.
1973. Pacific fishes of Canada. Fish. Res. Board Can.,
Ottawa, 740 p.
Heron, A. C.
1972a. Population ecology of a colonizing species: the pelagic
tunicate Thalia democratica. I Individual growth and
generation time. Oecologia (Berl.) 10:269-293.
1972b. Population ecology of a colonizing species: the pelagic
tunicate Thalia democratica. II population growth
rate. Oecologia (Berl.) 10:294-312.
HoBSON, E. S., AND J. R. Chess.
1976. Trophic interactions among fishes and zooplankters
near shore at Santa Catalina Island, California. Fish. Bull.,
U.S., 74:567-598.
Hubbard, L. T., and W. G. Pearcy.
1971. Geographic distribution and relative abundance of
Salpidae off the Oregon Coast. J. Fish. Res. Board Can.
28:1831-1836.
Hubbs, C. L.
1948. Changes in the fish fauna of western North America
742
HOBSON and CHESS: TROPHIC RELATIONS OF THE BLUE ROCKFISH
correlated with changes in ocean temperature. J. Mar. Res.
7:459-482.
Laval, P.
1980. Hyperiid amphipods as crustacean parasitoids associ-
ated with gelatinous zooplankton. In M. Barnes (editor),
Oceanogfraphy and marine biology, annual review 18, p.
11-56. Aberdeen Univ. Press, Aberdeen.
LiMBAUGH, C.
1955. Fish life in the kelp beds and the effects of kelp
har\'esting. Univ. Calif. Inst. Mar. Resour., IMR Ref. 55-9.
LOBEL, P. S.
1981. Trophic biology of herbivorous reeffishes: Alimentary
pH and digestive capabilities. J. Fish Biol. 19:365-397.
Love, M. S., and A. W. Ebeling.
1978. Food and habits of three switch feeding fishes in the
kelp forests off Santa Barbara, California. Fish. Bull., U.S.
76:257-271.
Mason, J. E., and A. Bakun.
1986. Upwelling index update, U.S. west coast, 33N-48N
latitude. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-
SWFC-67, 81 p.
Miller, D. J., and R. N. Lea.
1972. Guide to the coastal marine fishes of California. Calif.
Dep. Fish Game, Fish Bull. 157, 235 p.
Morris, R. H., D. P. Abbott, and E. C. Haderlie.
1980. Intertidal invertebrates of California. Stanford Univ.
Press, Stanford, 690 p.
PlNKAS, C. M., M. S. OLIPHANT, and I. L. K. IVERSON.
1971. Food habits of albacore, bluefin tuna and bonito in
California Waters. Calif. Dep. Fish Game, Fish Bull. 152,
105 p.
Reid, J. L., G. I. RoDEN, and J. G. Wyllie.
1958. Studies of the California Current system. CalCOFI
Prog. Rep. 1/7/56 to 1/1/58, p. 27-90.
Send, V., R. C. Beardsley, and C. D. Winant.
1987. Relaxation from upwelling in the Coastal Ocean
Dynamics Experiment. J. Geophys. Res. 92:1683-1698.
Silver, M. C.
1975. The habits of Salpafusiformis in the California Current
as defined by indicator assemblages. Limnol. Oceanogr.
20:230-237.
Skogsberg, T.
1936. Hydrography of Monterey Bay, California. Ther-
mal conditions, 1929-1933. Trans. Am. Phil. Soc. 29,
152 p.
Smith, R. L.
1968. Upwelling. In H. Barnes (editor), Oceanography and
marine biology, annual review, 6, p. 11-45. George Allen
and Unwin Ltd., Lond.
QUAST, J. C.
1968. Observations on the food of the kelp-bed fishes. In
W. J. North and C. L. Hubbs (editors), Utilization of kelp-
bed resources in southern California, p. 109-142. Calif.
Dep. Fish Game, Fish Bull. 139.
WiCKETT, P.
1967. Ekman transport and zooplankton concentration in
the North Pacific Ocean. J. Fish. Res. Board Can. 24:581-
594.
Wilson, C. L.
1952. Botany. Dryden Press, N.Y., 483 p.
Winant, C. D., R. C. Beardsley, and R. E. Davis.
1987. Moored wind, temperature, and current observations
made during Coastal Ocean Dynamics Experiments 1 and
2 over the northern California continental shelf and upper
slope. J. Geophys. Res. 92:1569-1604.
743
THE REPRODUCTIVE BIOLOGY OF TILEFISH,
LOPHOLATILUS CHAMAELEONTICEPS GOODE AND BEAN, FROM
THE UNITED STATES MID-ATLANTIC BIGHT, AND
THE EFFECTS OF FISHING ON THE BREEDING SYSTEM
Churchill B. Grimes/ Charles F. Idelberger,^ Kenneth W. Able,^
AND Stephen C. Turner*
ABSTRACT
To investigate the reproductive biology of tilefish, LopholatAlus chamaeleonticeps Goode and Bean, we
sampled the commercial longline fishery from 1978 to 1982. Results suggested that tilefish are frac-
tional spawners from March to November with a spawning peak from May to September. Estimates
of fecundity ranged from about 195 x 10^ to 10 x 10*^ (for 53 cm FL, 2.1 kg and 91 cm FL, 13 kg
females), but only 60-85% of the ovarian eggs appeared to have been released by the end of the spawn-
ing season.
Tilefish are apparently gonochoristic however, some adult males had slight amounts of ovarian tissue
within the testicular mass. Sex ratios were skewed in favor of males at larger sizes; however, both sexes
were present at most ages. All juveniles and unsexed fish (<400 mm FL) appeared to be female (i.e.,
gonad cell structure consisted of oogonia and previtellogenic oocytes), suggesting that some tilefish may
undergo prematurational sex reversal, or that early gonad development involves an all female appear-
ing stage.
Tilefish may have a complex breeding system that is behaviorally mediated. Both sexes are func-
tionally mature at ca. 50 cm FL and 5 years, but some males appear to delay participation in spawning
for 2-3 years and 10-15 cm in additional length. Tilefish are sexually dimorphic, with males attaining
larger sizes and developing conspicuously enlarged predordal adipose flaps (an apparent indicator of male
breeding status) at 65-70 cm FL (6-7 years), when males begin to participate in spawning (i.e., develop
large testes), not when they are functionally mature (i.e., producing sperm).
The rapidly expanding fishery from 1978 to 1982, which reduced population density by one-half to
two-thirds, may have altered the breeding system by causing males to spawn at smaller sizes (10 cm)
and younger ages (2-2.5 years) in 1982 than in 1978.
Tilefish, Lopholatilus chamaeleonticeps (Branchio-
stegidae), is a demersal gonochoristic species found
along the outer continental shelf from Nova Scotia
south to Surinam (Dooley 1978; Markle et al. 1980).
Within the Mid-Atlantic Bight (continental shelf
between Cape Cod, MA and Cape Hatteras, NC),
they inhabit a narrow zone of relatively warm tem-
peratures (9°-14°C) in about 80-240 m depths. Fol-
lowing a brief period as pelagic larvae (Fahay and
Berrien 1981; Berrien 1982), juveniles settle to the
bottom. Adults are sexually dimorphic, males hav-
ing larger adipose flaps ( = predorsal crest of Dooley
'Southeast Fisheries Center Panama City Laboratory, National
Marine Fisheries Service, NOAA, 3500 Delwood Beach Road,
Panama City, FL 32407.
^Florida Marine Research Institute, 100 8th Ave. S.E., St.
Petersburg, FL 33701.
^Rutgers University, Marine Field Station, Tuckerton, NJ 08087.
■•Southeast Fisheries Center Miami Laboratory, National Marine
Fisheries Service, NOAA, 75 Virginia Beach Drive, Miami, FL
33149.
1978) than females (Katz et al. 1983). Growth in both
sexes is about 10 cm per year for the first four years,
then it slows, but more so in females than males
(Turner et al. 1983). Maximum size of females is
about 100 cm FL and 112 cm FL in males, and max-
imum ages are 35 and 26 years, respectively (Turner
et al. 1983).
Both juveniles and adults select sedimentary sub-
strata, and seek shelter in a variety of habitats.
Grossman et al. (1985) found a strong correlation
between sediment composition and tilefish occur-
rence. Juveniles and adults occupy highly conta-
geously distributed vertical burrows, the primary
habitat, in Pleistocene clay substrata (Able et al.
1982; Grimes et al. 1986). Adults also inhabit hori-
zontal excavations in more vertically oriented clay
sediments of submarine canyon walls called "pueblo
habitats" (Cooper and Uzmann 1977; Warme et al.
1977; Grimes et al. 1986), and scour depressions
under and around glacial erratic boulders (Valen-
Manuscript accepted April 1988.
fishery BULLETIN: VOL. 86, NO. 4, 1988.
745
FISHERY BULLETIN: VOL. 86. NO. 4
tine et al. 1980; Grimes et al. 1986). Several crusta-
ceans and fishes are concentrated in and near tilefish
excavations, which are formed by the combined
activity of tilefish and their associates (Able et al.
1982; Grimes et al. 1986). Several associates are im-
portant components of tilefish diets, suggesting
close trophic linkages within the burrow community
(Turner and Freeman MS^; Grimes et al. 1986).
Limited mark-recapture data suggests that tilefish
are sedentary (Grimes et al. 1983). Tilefish and their
associates have definite temporal activity patterns,
and male-female pairing may occur (Grimes et al.
1986). Habitat preference, contagious distribution
within suitable clay substratum, definitive activity
patterns, limited movement, and pair bonding pro-
vide opportunities for social interaction and organi-
zation within populations.
Commercial landings of the Mid-Atlantic-southern
New England tilefish stock (Katz et al. 1983) have
varied widely since 1916 when 4,500 metric tons (t)
were landed in 10 months (Freeman and Turner
1977). In the early 1970s an important longline fish-
ery developed, centered in New York-New Jersey,
and landings increased. This fishery became one of
the most valuable finfisheries in both states during
most years since 1978 (Grimes et al. 1980; U.S.
Department of Commerce 1980a-c; D. J. Christen-
sen pers. commun.^). From 1977 to 1980, annual
landings were 2,061, 3,412, 3,840, and 3,575 t, but
catches declined to 3,200 and 1,900 t in 1981 and
1982 (U.S. Department of Commerce 1980a-c; D. J.
Christensen fn. 6). Effects of fishing on the stock
have been drastic, reducing stock size by one-half
to two-thirds from 1978 to 1982 (Turner 1986).
The purpose of this paper is to describe the repro-
ductive biology of tilefish in the northern stock, and
present evidence that participation in spawning by
males is socially mediated. Furthermore, we show
that the mating system has been modified by com-
mercial fishing.
MATERIALS AND METHODS
Samples of tilefish from the United States Mid-
Atlantic-southern New England area were obtained
from domestic longline and New Jersey recreational
(headboat^) fisheries. Information gathered for each
^Turner, S. C, and B. L. Freeman. Food habits of tilefish,
Lopholatilus chamaeleonticeps. Unpubl. manuscr. Southeast
Fish. Cent., Natl. Mar. Fish. Serv., NOAA, Miami, FL 07732.
•^D. J. Christensen, Northeast Fisheries Center Sandy Hook
Laboratory, National Marine Fisheries Service, NOAA, Highlands,
NJ 07732, pers. commun. 1983.
sample included the statistical nature of the sam-
ple (random or nonrandom), capture method, loca-
tion (LORAN C), date, and length (FL, cm), sex,
weight (whole and eviscerated), and height (mm) of
the adipose flap of each animal. When possible,
gonads were excised, weighed (0.1 g) and preserved
in 10% formalin. We assigned each fish to one of
three classifications (male, female, or unknown) and
one of six stages of sexual maturity following visual
examination (Nikolsky 1963). Some fish used for
reproductive studies were aged using thin sections
of their saggital otoliths (Turner et al. 1983).
Routine histological sectioning and staining (hae-
motoxylin and eosin) techniques and light micro-
scropy (450 X and 1000 x) were used to examine
gonad structure of six small fish (<50 cm), and assess
the state of sexual maturity. Ovarian development
corresponded to that described by Moe (1969) for
red grouper, Epinephelus morio, and Yamamoto
(1956) for the flounder Liopsetta obscura. Sperma-
togenic development was identical to that of Tilapia
leucosticta (Hyder 1969) and toadfish, Opsanus tau,
(Hoffman 1963). Females were assessed as imma-
ture when ovaries contained only previtellogenic and
early vitellogenic oocytes (Yamamoto 1956; Moe
1969; Waltz et al. 1982; Ross 1978; Erickson and
Grossman 1986). Males were considered mature
when active spermatogenesis was occurring and
spermatozoa were present in spermatogenic tubules
(Ross 1978). Limited spermatogenesis occurred in
the testes of some immature males (Erickson and
Grossman 1986).
A gonosomatic index (GSI) was calculated for
GW
females according to the formula GSI = ^fr^ x 100,
where GW = fresh gonad weight (g) and BW =
fresh eviscerated body weight (g) (Nikolsky 1963),
for describing spawning seasonality. The seasonal
progression of mean ovum diameters was also used
to establish the reproductive seasonality, and the
ovum-diameter frequency distribution for ripe
females was used to indicate spawning frequency
(isochronal vs. heterochronal) (Hickling and Ruten-
berg 1936). Separated ova from each fish were
placed in liquid and stirred; a random sample for
measuring was then obtained by extracting three
aliquots with a large syringe. Diameters of 500-
1,000 ova from each female were measured (near-
est ocular micrometer unit) using a binocular dis-
secting scope and a filar micrometer. We assumed
''Headboats are vessels which charge anglers for fishing on an
individual, thus "per head", basis.
746
GRIMES ET AL.: REPRODUCTIVE BIOLOGY OF TILEFISH
uniform ova size (developmental stage) distribution
among anterior to posterior ovarian lobe locations,
as has been demonstrated for Lopholatilus chamae-
leonticeps from the South Atlantic Bight (Erickson
et al. 1985).
To determine if the liver was being utilized to store
energy in the form of fat reserves to be used in
gonad maturation, we calculated a hepatosomatic
index (HSI) = liver weight (g)/gutted body weight
(g) X 100 (Htun-Han 1978).
Ovaries used to estimate fecundity ( = ovarian egg
count of Gale and Deutsch (1985)) were preserved
in modified Gilson's fixative (Bagenal and Braum
1978). The ovarian tunic was removed and washed
free of adhering ova. Developing ova were separated
from folicular material and most oogonia by wash-
ings under a stream of water. Based upon initiation
of yolk accumulation, all oocytes with diameters
>0.15 mm were included in ovarian egg counts.
Each sample of developing ova was diluted in water
and stirred, then at least two subsamples pipetted;
each subsample was placed in a 6 x 6 cm gridded
Petri dish for counting. Ova were counted in six ran-
domly selected grid squares, and the average num-
ber of ova in the six squares was then adjusted to
the total subsample count by multiplying the aver-
age by the total number of grid squares in the dish.
The sample and subsamples were oven dried at 40°C
for at least 24 hours and weighed to the nearest
0.001 g on a Mettler^ balance. Fecundity (total
ovarian egg count) was estimated as the number in
the subsample multiplied by sample weight divided
by subsample weight.
Predictive equations of fecundity from length and
weight were fit using least squares regression and
converted to functional regressions (Ricker 1973).
Fecundity was separately regressed on FL and
gutted weight using all possible combinations of un-
transformed log and semi-log models. We inspected
residuals, plots and coefficients of determination to
evaluate fits.
RESULTS
Gonad Structure and Sex Determination
Males smaller than 600 mm FL were difficult to
sex by gross gonadal structure because testes were
small and undeveloped. However, in females larger
than about 400 mm FL ovaries were sufficiently
developed to visually determine sex easily. There-
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
fore, we histologically examined gonadal tissues of
155 fish (50 males, 545-814 mm FL; 52 females,
241-678 mm FL; and 53 juveniles, 146-400 mm FL).
Based upon gonad microstructure, seven fish be-
tween 467 and 592 mm FL that were macroscop-
ically unsexable were males. All juveniles and un-
sexed fish <400 mm FL (79 fish 121-400 mm FL)
appeared to be females, i.e., had cell structure con-
sisting of only oogonia. These cells had slightly aci-
dophilic cytoplasm and distinct, but irregular, cell
membranes. The nuclear membrane was less ob-
vious because it was masked by basophylic nuclear
material located peripherally.
The paired ovaries of adult L. chamaeleonticeps
are suspended below the swimbladder from the dor-
sal body wall in the extreme posterior of the body
cavity. Ovarian microstructure and development are
identical to Caulolatilus microps (Ross and Merriner
1983) and L. chamaeleonticeps (Erickson et al. 1985)
from the South Atlantic Bight and are described in
detail by Idelberger (1985).
The testes of tilefish are solid, smooth textured,
and more elongate than ovaries. In males smaller
than about 65 cm FL, testes were not obviously
lobed and were pinkish in color. Only in large adult
males were testes creamy white and heavily lobed.
Microstructure of the testes was typical of teleosts
(Hoffman 1963; Smith 1965; Hyder 1969) and was
identical to C. microps (Ross and Merriner 1983),
and, like the ovary structure, was described in detail
by Idelberger (1985). Two males (604 and 609 mm
FL) had slight amounts of ovarian tissue within the
testicular mass (Fig. 1). That ovarian tissue con-
sisted of several resting or previtellogenic (peri-
nucleolar) oocytes located in testicular mass with
definite spermatogenic tubules and crypt structure.
Sexual Dimorphism
The predorsal adipose flap is sexually dimorphic
in adult tilefish and can easily be used to determine
sex in animals larger than about 70 cm FL. The size
of the predorsal flap was significantly larger in
males than in females older than age 5 years and
larger than 65 cm FL (Table 1). It was not possible
to compare predorsal flap sizes of the largest males
to those of females, because females do not grow
as large as males (Turner et al. 1983; Harris and
Grossman 1985).
Sexual Maturity
We determined the general pattern of age and size
of sexual maturity using both visual staging and
747
FISHERY BILLETIN: VOL. 86. NO. 4
B
Figure 1.— Photomicrographs of histological sections of tilefish testes from a 60.4 cm FL male at x400 (A)
and a 60.9 cm FL male at x 100 (B) showing residual oocytes (RO) and spermatogenic cripts (SC).
748
GRIMES ET AL.: REPRODUCTIVE BIOLOGY OF TILEFISH
Table 1— Mean (X) and sample size (A/) for predorsal adiposal
flap height at size and age for male and female tilefish, and
students Mests of flap height differences between sexes.
Males Females
X(mm) N X(mm) N
(-value
Proba-
bility
Size
(FL, mm)
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85-89
90-94
95-99
100-104
Age
(years)
3
4
5
6
7
8
9
10
11
12
14
15
20.1
23.1
25.7
30.6
37.1
49.8
61.8
72.3
82.4
68.0
63.0
98.0
20.5
36.2
33.2
44.0
51.4
57.4
65.0
112
13
29
36
40
73
67
30
30
12
1
1
8
0
0
6
23
41
40
17
8
4
0
0
1
17.1
23.6
25.9
27.6
31.6
31.7
37.6
37.7
28.8
38.5
5.0
8.1
20.4
21.6
27.9
30.1
30.4
32.9
31.9
30.9
36.5
36.8
19
49
62
105
101
51
25
11
8
2
0
0
6
7
16
32
51
42
40
35
16
9
2
14
1.02
0.33
0.04
0.91
2.63
3.58
4.05
1.90
1.84
Not testable
Not testable
Not testable
Not testable
Not testable
0.018
6.63
3.42
5.64
5.97
5.45
5.59
Not testable
Not testable
Not testable
N.S.
N.S.
N.S.
N.S.
<0.01
<0.01
<0.01
<0.05
<0.05
N.S.
<0.001
<0.01
<0.001
<0.001
<0.001
<0.001
histological data pooled over all years of the study.
The two methods gave substantially different results
for males; however, both visual and histological data
suggested that all females matured at 60-65 cm FL
and 8-9 years of age (Figs. 2, 3). Visual staging in-
dicated that virtually all males were mature at 80-85
cm FL and 10-11 years. In contrast, histological
results indicated almost all males were mature at
65-70 cm FL (estimated age 7-8 years, Turner et
al. 1983) (Figs. 2, 3). That is, visual staging sug-
gested that females matured about 20 cm smaller
and 2-3 years younger than males; however, accord-
ing to histological analysis, both males and females
were mature and producing gametes at a similar size
(60-70 cm FL). Thus, these results show that al-
though males were producing sperm at 65-70 cm
FL (estimated 7-8 years, Turner et al. 1983), they
were not developing a large testicular mass until
80-85 cm FL and age 10-11 years.
We reasoned that predorsal adipose flap size
might be a sign of maturity or reproductive status
of males, because the height of the flap was larger
in males than females above 65 cm and age 5 years
(see previous section on Sexual Dimorphism). Addi-
tionally, the development of enlarged adipose flaps
in males coincided with the size of 50% maturity
(60-65 cm FL) as judged by development of enlarged
testes (i.e., visual method), not at the smaller size
of 50% maturity (50-55 cm FL) determined his-
tologically (Fig. 2). Therefore, if predorsal adipose
flap height were a sign of reproductive status in
males, then mature males should have larger flaps
than immature males; and this was so (Fig. 4).
Furthermore, ANCOVA (with FL as the covariate)
indicated that flap height was significantly larger
in mature males than immature males (Table 2).
Table 2. — Mean and ANCOVA (with FL as the covariate) of predor-
sal adipose flap height in sexually mature and immature male
tilefish. Flap height data set tested was restricted to a size range
containing both mature and immature males (50-75 cm FL). df
= degrees of freedom, ss = sum of squares, W = sample size, and
X = mean.
MEAN
N
X flap height (mm)
X FL (cm)
Immature
Mature
94
87
28.7
42.7
59.4
68.6
ANCOVA
Source
df
ss F
P> F
Maturity
FL
FL X maturity
Error
Total
1
1
1
177
180
504.4
7,423.4
605.1
13,895.7
30,365.8
6.42
94.56
7.71
0.0121
0.0001
0.0061
Because we had studied reproduction during a
period of rapid expansion of the fishery, i.e., the
commercial longline fishery had decreased the
tilefish population by one-half to two-thirds from
1978 to 1982 (Turner 1986), we decided to examine
the data to determine if the fishery had affected the
size at which tilefish were developing enlarged
testes (i.e., attaining maturity as assessed by the
visual method). A comparison of visually assessed
sexual maturity at size and age for 1978 and 1982
suggested that females <50 cm FL were maturing
at a smaller size in 1978 than in 1982, but that
females >50 cm FL matured at very slightly larger
sizes in 1978 than in 1982 (Table 3). Log-likelihood
contingency tests of these data for females indicated
that maturity-at-length was significantly different
in 1978 and 1982, but the difference was due to the
numbers of mature and immature fish between 41
and 55 cm (Table 3). The data for males <75 cm FL
indicated that males matured at smaller size in 1982
than in 1978 (Table 4). A log-likelihood contingency
749
FISHERY BULLETIN: VOL. 86, NO. 4
100—,
LU
IX
ID
I-
<
50-
4^ — o — o — o — o
" 22 5 2 1
o o VISUAL
• • HISTOLOGICAL
~i 1 1 1 1 ' i 1 1 1 — r
Figure 2.— The relationship of sexual maturity (by
both visual staging and histological methods) and size
(5 cm FL intervals) for female and male tilefish. Sam-
ple size is shown next to each point. Curves are fitted
by eye.
100—1
z
LU
O
50-
^ ^X^'os
73 20 8 10 4 1
O O O O O
9 31
o o VISUAL
. , HISTOLOGICAL
T 1 u I u- I 1— •— I 1 1 1 1 1 1 1 1 1 1 1 r-
40 60 80 100
LENGTH (cm)
100-1
lU
<
50-
z 100
LU
o
GC
LU
Q.
50 —
"T I I I I I 1 1 1 1 1 1 1 1 I r
2 4 6 8 10 12 14 >15
Figure 3.— The relationship of sexual maturity (by visual
staging) and age for female and male tilefish. Sample size
is shown next to each point. Curves are fitted by eye.
AGE (years)
750
GRIMES ET AL.: REPRODUCTIVE BIOLOGY OF TILEFISH
Table 3— Female sexual maturity at length (visually determined)
and log-likelihood contingency test that maturity at fork length (FL)
in female tilefish was not different for 1 978 and 1 982. Critical chi-
.2
square values are x oos
(11 df) = 19.7andx 0 01 (■'■' df) = 24.7.
Table 4.— tvlale sexual maturity at length (visually determined) and
log-likelihood contingency tests that maturity at fork length (FL) in
male tilefish was not different for 1978 and 1982. Critical chi-
square values are x^oos C^ "^0 = "'^-^l and x^o.oi CO dO =
23.2. NT = not testable.
N 1 = noi
lesiat
3ie.
1978
1982
1978
1982
.
FL
Maturity
Sample
(Maturity
Sample
G-
FL
(cm)
Maiuiiiy
No. %
Sample
size
No.
%
Sample
size
6-
statistic
(cm)
No.
%
size
No.
%
size
statistic
31-35
0
0
4
N
0 d
a
t a
NT
26-30
N
0 d
a t a
0
0
1
NT
36-40
0
0
15
N
0 d
a
t a
NT
31-35
0
0
7
0
0
3
0
41-45
5
28
18
11
52
21
4.4
36-40
3
21
14
0
0
17
0
46-50
10
24
42
50
79
63
59.9
41-45
11
49
23
4
7
60
30.4
51-55
13
12
105
16
43
37
47.3
46-50
34
81
42
11
15
72
89.6
56-60
0
0
15
19
73
26
0
51-55
54
59
92
41
89
46
56.5
61-65
8
44
18
12
71
17
5.2
56-60
40
80
50
35
100
35
0
66-70
15
60
25
16
89
18
13.7
61-65
63
98
64
49
100
49
0
71-75
31
79
39
15
94
16
8.9
66-70
57
100
57
27
100
27
0
76-80
47
96
49
20
91
22
1.8
71-75
12
100
12
29
100
29
0
81-85
28
100
28
14
100
14
0
76-80
12
92
13
14
93
15
0.2
86-90
10
100
10
5
100
5
0
81-85
10
100
10
4
100
4
0
91-95
3
100
3
2
100
2
0
86-90
6
100
6
N
0 d
a t a
NT
96-100
8
100
8
N
0 d
a
t a
NT
91-95
3
100
3
1
100
1
0
101-105
3
100
3
N
0 d
a
t a
NT
96-100
N
0 d
a t a
1
100
1
NT
106-110
1
100
1
N
0 d
a
t a
NT
Total
393
360
176.7
111-115
Total
1
100
1
384
N
0 d
a
t a
241
NT
141.2
_ 80-
E
o
g 60
I
CL
<
i 40H
LU
CO
O
Q.
Q
<
20-
O IMMATURE n = 38
MATURE n = 92
1
O.
O O20
5
o.
o.
"^ \ r
40
-^ \ ' r
60
n r
80
LENGTH
(cm)
Figure 4.-Mean predorsal adipose flap height at length (5 cm FL intervals) for sexually mature
and immature male tilefish. Maturity was assessed by visual staging. Sample size is indicated for
each interval.
751
FISHERY HI'LLKTIN: VOL. 8fi. NO. 4
test indicated that maturity-at-age in males for the
two years was highly significantly different (Table
4).
Analysis of visual maturity at age data gave less
ambiguous results. Percent of females mature at age
in 1978 and 1982 was not significantly different
(Table 5). However, males matured at younger ages
in 1982 than 1978, and the differences between
years were highly statistically significant (Table
6).
Table 5.— Female sexual maturity at age (visually determined) and
log-likelihood contingency tests that maturity at age in female
tilefish was not different for 1978 and 1982. Critical chi-square
values are x^o.os (9 df) = 16.9 and x^o oi (9 df) = 21 .7. NT = not
testable.
1978
1982
Age
Maturity
Sample
Maturity
Sample
G-
(yr)
No.
%
size
No.
%
size
statistic
3
0
0
1
0
0
3
0
4
1
8
13
1
5
19
0.3
5
35
43
82
5
33
15
3.1
6
19
83
23
3
75
4
0.6
7
47
90
52
8
100
8
0
8
68
97
70
20
95
21
0.7
9
14
100
14
14
100
14
0
10
2
100
2
12
100
12
0
11
N
0 d
a t a
7
100
7
NT
12
1
100
1
8
100
8
0
13
N
0 d
a t a
4
100
4
NT
14
N
0 d
a t a
2
100
2
NT
>15
9
100
9
4
100
4
0
Total
267
121
4.7
Sex Ratio
To minimize the chances of obtaining biased
results caused by the effects of fishing we estimated
sex ratio at size and age using only data collected
in 1978 when the tilefish population was lightly ex-
ploited. Proportions of males and females were
similar in the 46-50 and 51-55 cm FL intervals, al-
though the ratio in the 51-55 cm FL interval tested
significantly different at P < 0.05 (Table 7). Males
were significantly predominant (70-80%) between
71 and 90 cm FL. At the 91-95 cm FL size females
were predominant once again (>50%), although not
significantly so (Table 7). Above 100 cm FL only
males were collected.
The general pattern of sex ratio at age in 1978
seems to have been equal proportions of the sexes
through about age 20 years, with only females pres-
ent from ages 29 to 36 years. Log-likelihood tests
of sex ratio at age showed no significant differences
Table 6.— Male sexual maturity at age (visually determined) and
log-likelihood contingency tests that maturity at age in male tilefish
was not different for 1978 and 1982. Critical chi-square values
are x^o.os (^ df) = 15.5 and x^o.oi (^ ^^ = 20.1 . NT = not testable.
1978
1982
Age
(yr)
Maturity
No. %
Sample
size
Maturity
No. %
Sample
size
G-
statistic
4
1
6
16
3
75
4
34.5
5
10
10
96
7
41
17
44.8
6
3
19
16
8
80
10
27.6
7
6
38
16
6
75
8
22.3
8
41
77
53
15
75
20
0.6
9
23
85
27
19
100
19
0
10
5
100
5
6
86
7
0
11
2
100
2
1
100
1
0
12
1
100
1
2
100
2
0
>15
11
100
11
N
0 d
a t a
NT
Total
246
88
129.8
Table 7.— Sex ratio at length and log-likelihood tests that
sex ratio was not different from 1:1 at 5 cm FL intervals.
All G scores were calculated using Yates correction for
small sample sizes. Critical chi-square values are x^oos (^
df) = 3.84 and x^ooi (^ df) = 6.64. NT = not testable.
Fork length
Number
Number
G-
(cm)
of females
Percent
of males
statistic
46-50
46
46
54
0.6
51-55
95
41
134
6.7
56-60
56
78
16
23.5
61-65
66
78
19
27.5
66-70
74
69
33
16.1
71-75
18
30
43
10.5
76-80
14
21
53
25.8
81-85
13
30
31
7.6
86-90
7
37
12
1.3
91-95
5
63
3
0.4
96-100
0
0
9
NT
101-105
0
0
4
NT
106-110
0
0
2
NT
111-115
0
0
1
NT
from equality except for ages 7 and 8 years; sex ratio
at age was not testable for ages 29-36 years because
only females were present (Table 8).
Although sex ratio was skewed with age and size,
the estimated sex ratio for the entire population was
different from 1:1 in 1978 (Table 9). We calculated
the population to be 46.2% males, and could not re-
ject the null hypothesis that sex ratio is not differ-
ent from 1:1 (x^ = 0.15, df = 1, x"o.o5 = 3.84) in the
population >50 cm FL.
Spawning
Several lines of evidence suggest that tilefish in
the Mid- Atlantic-southern New England area are
752
GRIMES ET AL.: REPRODUCTIVE BIOLOGY OF TILEFISH
Table 8.— Sex ratio at age and log-likelihood tests that sex
ratio at age was not different from 1:1. All G scores were
calculated using Yates correction for small sample sizes.
Critical chi-square values are x^qos (^ <^^ = 3.84 and x^o.oi
(1 df) = 6.64. NT = not testable.
Age
Number
Number
G-
(yr)
of females
Percent
of males
statistic
4
1
100
0
NT
5
43
46
50
0.7
6
10
67
5
2.5
7
40
78
11
18.8
8
69
62
42
7.1
9
20
49
21
1.0
10
2
29
5
2.4
11
0
0
2
NT
16
2
40
3
0.8
19
2
50
2
Equality
29
1
100
0
NT
30
1
100
0
NT
31
1
100
0
NT
32
2
100
0
NT
33
1
100
0
NT
34
1
100
0
NT
35
1
100
0
NT
36
1
100
0
NT
Table 9.— The proportion of male tilefish in the 1978
n
population >50 cm FL, P = X L, M, where L and M =
the proportion of all fish and males, respectively, in the /'^
5 cm FL interval, and n = the number of size intervals. L,
was determined from the 1978 commercial longline catch
of tilefish.
Proportion
Proportion
Fork length
of
Proportion
of males in
(cm)
all fish
of males
population
/
L
M
P
<50
0.026
0
0
51-55
0.058
0.540
0.032
56-60
0.158
0.253
0.026
61-65
0.073
0.167
0.012
66-70
0.162
0.267
0.044
71-75
0.191
0.646
0.127
76-80
0.110
0.792
0.089
81-85
0.087
0.744
0.066
86-90
0.682
0.611
0.043
91-95
0.021
0.400
0.008
96-100
0.011
0.875
0.010
>101
0.036
1.000
0.004
= 0.462
fractional spawners from about March through
November, although most of the reproduction
evidently occurs from May to September. Some
females with free ova in the ovarian lumen (running
ripe) v^ere present in March through August and in
October and November. From May through August,
89-98% of the females were ripe or running ripe
(Fig. 5). Running ripe or ripe males were not as
frequently observed as were females in a similar
reproductive state. In fact, only very large males
(75-80 cm FL) were observed with a large creamy
white swollen testicular mass. Ripe males were
found in January, March, May through August,
October, and November, but the highest proportions
(23-46%) were present in May through August (Fig.
5).
GSI data for females indicated a similar seasonal
spawning pattern (Fig. 6). Highest GSI values con-
sistently occurred from May through August, when
ovaries accounted for 3.5-8.3% of gutted body
weight.
Analysis of ovum-diameter data also suggested
that spawning occurred mostly from May to Sep-
tember, and indicated that spawning was fractional,
i.e., ova were released in batches. Highest monthly
mean developing ovum diameters (0.30-0.42 mm)
occurred in May through August (Fig. 7). During
these months mean ovum diameter was usually
>0.35 mm (Fig. 7). During other months, mean
ovum diameter was always <0.30 mm. The size-
frequency distribution of ova from running ripe
females was polymodal (Fig. 8), suggesting multi-
ple spawnings by individual females during the
reproductive season.
Tilefish may not utilize the liver and soma to store
energy as fat for mobilization to the gonads in prep-
aration for spawning as many species are thought
to do (Hoar 1957). HSI for both males and females
showed a distinct pattern of seasonal variation, but
highest values occurred during summer (spawning
season) and lowest in winter (Fig. 9). Somatic con-
dition factor (eviscerated weight, g/FL^) showed no
discernable seasonal pattern in males or females.
Fecundity
Because tilefish are fractional spawners, fecun-
dity (ovarian egg count) was estimated from females
collected early in the spawning season (i.e., May to
early June) to minimize the chance of using a par-
tially spawned ovarj' and underestimating egg num-
ber. Estimates of fecundity ranged from approx-
imately 195,000 for a 53 cm FL (2.1 kg) female to
10 miUion for a 91 cm FL (13 kg) female, with a
mean egg count of 2.28 million (r? = 49, SD = 1.02).
The 91 cm FL female with 10 million eggs was ex-
ceptional; all other estimates were <4.1 million, even
for other large females 80 and 86 cm FL. Therefore,
we judged that the two largest fish, 91 and 86 cm
FL, were outliers and developed predictive
equations for egg count without using outlier data.
753
FISHERY BULLETIN: VOL. 86, NO. 4
STAGE
4 Running
fmamj jasond
MALE
100
N=45
LLI
O
IT
lU
Q.
STAGE
"1
3 + 4 Rip
2 Dev.
V
5 + 6
Spent/
Resting
Figure 5.— Reproductive seasonality of female and male tilefish as indicated by the percent of various
visual maturity stages collected by month from 1978 to 1982.
Log-transformed models produced slightly super-
ior fits, with length proving a slightly better pre-
dictor of ovarian egg count than weight (log^ Y =
4.75 log, FL - 5.2, r^ = 0.62, n = 48; log, Y =
1.48 log, W + 2.48, r^ = 0.59, n = 48). Based upon
our estimates, a first spawning female would pro-
duce <500,000 eggs.
DISCUSSION
Seasonality and Spawning
It seems clear that the northern stock of Lopho-
latilus chamaeleonticeps consists of fractional
spawners over an 8 or 9 month season, with peak
spawning from May to September. Our findings
agree with the limited information previously re-
ported. Collins (1884) reported ripe fish in July;
Bigelow and Schroeder (1953) in August; Dooley
(1978) in February, March, June, and July; Morse
(MS)^ March through August; and Freeman and
Turner (1977) from mid-March to mid-September.
Other members of the Branchiostegidae (Caulola-
tilus microps, C. chrysops, C. princeps, C. affinis,
'Morse, W. W. Length, weight, spawning and fecundity of the
tilefish, Lopholatilus chamaeleonticeps, from New Jersey waters.
Unpubl. manuscr. Northeast Fish. Cent. Sandy Hook Lab., Natl.
Mar. Fish. Serv., NOAA, Highlands, NJ 07732.
754
GRIMES ET AL.: REPRODUCTIVE BIOLOGY OF TILEFISH
10
8 -
X
m
o
6-
Q
<
Z
O
O 4,
z
<
2-
jIfImIaImIjIjIaIsIoInId
1979
jIfImIaImIjIjIaIsIoInIdIjIfImIaImIjIjIaI
1980 I 1981
SAMPLING DATE
Figure 6.— Reproductive seasonality of female tilefish as indicated by monthly mean gonad index. Sample
size is indicated for each month.
60-1
50'
E
E
tr 40-
<
Q 30-
>-
o
o
o
<
m
20-
10-
j'f'm'a'm'jIjIa's'o'n'dIj'f'm'a'm'j'j'a
1980
1981
SAMPLING DATE
Figure 7.— Reproductive seasonality of female tilefish as indicated by monthly mean oocyte diameters. Sample size is indicated
for each month.
755
FISHERY BULLETIN: VOL. 86, NO. 4
n = 521
-J
.5 .6 7 8 ,9 1.0 1.1
OOCYTE DIAMETER (05 mm intervals)
r
1.2
T
1.4
Figure 8.— Frequency distribution of oocyte diameters from a spawning 66 cm FL female tilefish (running ripe) collected
in June 1979.
Figure 9.— Monthly mean heptosomatic index
for male (upper) and female (lower) tilefish.
Sample size is shown next to each mean.
30 -\
X 20-
LU
a
z
tr
> 10
30
20 -
lC
m
>
3 101
a|sJo|n|d j| f|m|a|m|j|j|a|s|o|n|d
1979 1980
SAMPLING DATE
18
1611 . 20
J|F|M|A|VI|J|J|A|
1981
A I S| 0| N I D
1979
J|F|M|A|M|J|J|A|S|0|N|D
1980
SAMPLING DATE
J|F|M|A|M|J|J|A|
1981
756
GKIMES ET AL.: REPRODUCTIVE BIOLOGY OP' TILEFISH
Branchiostegus japonicus) all exhibit extended re-
productive seasons centered around summer (Haya-
shi 1977, 1979; Dooley 1978; Ross and Merriner
1983). However, Erickson et al. (1985) reported that
L. chamaeleonticeps spawned from March through
June off Georgia in the South Atlantic Bight. The
reason for the shorter reported spawning season for
L. chamaeleonticeps in the South Atlantic Bight is
uncertain.
Fractional spawning by individual females was in-
dicated by the polymodal frequency distribution of
ovum diameters. This asynchronous follicular devel-
opment is typical of fractional spawning (deVlam-
ing 1983), and has been reported for L. chamaele-
onticeps in the South Atlantic Bight (Erickson et al.
1985) and Mid-Atlantic Bight (Morse fn. 9). Frac-
tional spawning based upon polymodal ovum-diam-
eter distributions has also been reported for the con-
familials C. microps and B. wardi (Dooley 1978;
Ross and Merriner 1983).
Our estimates of fecundity in tilefish are consis-
tent with those in the literature. We agree with Gale
and Deutsch (1985) that the term fecundity is incor-
rectly applied to many fishes, especially fractional
spawners, because in most cases there is no reason-
able means to determine which or how many oocytes
or developing ova will be released, or how many
ovarian ova will be resorbed after spawning and
never released. Therefore, we recognize that our
ovarian egg count data provides only a rough esti-
mate of actual fecundity. Erickson and Grossman
(1986) found the relationship between fecundity and
weight to be best described by a log transformation
(log, F = 1.497 log, W + 12.59, r- = 0.93); while
Morse (fn. 9) found the relationship best fitted to
a linear form (F = -966,471 + 887 W, r"~ = 0.61).
To compare our estimates further we calculated
relative fecundity using extreme point estimates of
fecundity reported by other authors (Erickson and
Grossman 1986, 414 -g"^ and 950 -g'' for 2.0 and
8.0 kg fish; Morse fn. 9, 543 -g-^ and 867 -g-^ for
3.5 and 9.0 kg fish) and this study (119 -g"' and
769 -g"' for 2.1 and 13.0 kg fish), and using predic-
tive equations (Table 10). Comparing the range of
point estimates, the three studies are similar, how-
ever our findings agree more closely with Erickson
and Grossman (1986) when the comparison is based
upon predictive equations. The inconsistency in the
comparison is because Morse (fn. 9) used a linear
equation, thus assuming that egg production per
gram of body weight was constant for all body
weights, and Erickson and Grossman (1986) and this
study chose curvilinear equations. It is possible to
conclude that small tilefish, like some other fishes
(Grimes 1987), produce fewer eggs per unit body
weight than larger fish.
We compared numbers of ova actually found in
ripe ovaries collected on 1 July (n = 3) and 28
August {n = 6) to ovarian egg numbers predicted
for the same size females to estimate what propor-
tion of ova had been spawned in the mid- and late-
spawning season. The analysis suggested that ap-
proximately 25% of ovarian eggs were spawned by
July and 50% by the end of August. The ovaries of
two postspawning females collected on 10 October
contained considerable quantities of ovarian eggs
>0.15 mm in diameter. These ova accounted for
15-20% of the predicted ovarian egg number. By
December 11 (n = 3) the number of ovarian eggs
>0.15 mm in diameter had decreased to about 5%
of the predicted maximum ovarian egg number.
Since atretic ova were observed in histological
preparations of ovaries in spent and resting stages,
and resorption is well documented in other teleosts
(Hoar 1957; Smith 1965; Combs 1969; Foucher and
Beamish 1977; LaRoche and Richardson 1980;
Waltz et al. 1982), it seems reasonable to conclude
that at least 15-20% of the maximum ovarian egg
number are never spawned and are resorbed dur-
ing the winter.
Sexuality
Our results agree with Erickson and Grossman
(1986) that tilefish are gonochoristic and that sec-
ondary gonochorism is a possibility. Gonad micro-
structure and development of adult (>50 cm FL)
tilefish were typical of most male and female ovi-
parous teleosts, and identical to that described for
Table 10. — Comparison of ovarian egg number and
relative ovarian egg number for small (2.0 kg) and large
(9.0 kg) female tilefish calculated following Erickson
and Grossman (1986), Morse,' and this study.
Body
weight
Egg
number
Relative
egg number
Erickson and
Grossman
2.0
9.0
828,723
7,875,326
414
875
Morse
2.0
9.0
1,773,003
7,982,003
887
887
This study
2.0
9.0
917,434
8,498,245
459
944
'Morse, W. W. Length, weight, spawning and fecundity of
the tilefish, Lopholatilus chamaeleonticeps. from New Jersey
waters, Unpubl. manuscr Northeast Fish. Cent. Sandy Hook
Lab , Natl Mar. Fish. Serv., NOAA, Highlands, NJ 07732.
757
FISHERY KILLKTIN: VOL. 8«. NO. 4
male and female Caulolatilus microps (Ross and
Merriner 1983) and female Lopholatilus chamaele-
onticeps (Erickson et al. 1985). Like Ross and Mer-
riner (1983) and Erickson and Grossman (1986), we
also found a few adult males (2 of 50) with ovigerous
tissue (previtellogenic residual oocytes) in the testic-
ular mass. Histological sections of testes revealed
no gross structural features that indicated prior
functional female status (e.g., remnants of an
ovarian lumen (Sadovy and Shapiro 1987)). Further-
more, no ovary was ever observed in transition to
a testis (transitional ovotestes), however transition
can occur within a matter of weeks. Similar to other
branchiostegids (C microps, Ross and Merriner
1983; Brayichiostegus wardi and B. seTratus, Dooley
1978; L. chamaeleonticeps , Erickson and Grossman
1986), sex ratios were skewed in favor of males at
large sizes. However, both sexes were present at
most sizes and ages (45-95 cm FL and 5-10, 16, and
19 years) and only females were present at ages
29-36 years, ruling out protogyny. Disparate sex
ratios at size are apparently due to differential
growth and mortality rates between sexes (Turner
et al. 1983; Harris and Grossman 1985).
Our histological examination of juvenile gonads
suggest that some L. chamaeleonticeps may undergo
prematurational sex reversal. All juvenile gonads
examined (63 fish, 146-400 mm FL) appeared to be
female, based upon gonad cell structure (i.e., pres-
ence of only oogonia and previtellogenic oocytes).
We are very tentative about the determination that
all of these small fish (<400 mm FL) were females,
because undifferentiated gonia, oogonia, and sper-
matogonia are very similar in appearance (Yama-
moto 1956; Hoffman 1963; Hyder 1969; Ross 1978).
Once a gonad has developed gross structure such
as spermatogenic tubules and crypts, or an ovarian
lumen, determining sex is straightforward. We
found residual oogonia in 2 of 50 histologically ex-
amined testes. We found no juveniles with truly
intersexual gonads, nor were we able to observe a
lumen. Ross and Merriner (1983) suggested that the
confamilial C. microps underwent prematurational
sex reversal, and that gonochorism in C. microps
might be a regression from monandric protogyny.
Their conclusions were based upon findings from
four juveniles (one specimen with a totally ovarian
gonad and the remaining three with gonads that
contained substantial amounts of testicular tissue),
and adult males (8 of 41 examined) with residual
oocytes in the testicular mass.
That prematurational sex reversal has been ob-
served among several families and species of fish
suggests that either prematurational sex reversal
is more common among fishes than suspected, or
that early gonad development in fishes involves an
all-female or female-appearing stage. For example,
the salmonid Salmo gairdneri (Mrsic 1923); the
cyprinidsBrac^danio rerio (Takahashi 1977), Bnr-
bus tetrazona (Takahashi unpubl. data cited in Taka-
hashi 1977), Rodeus ocellatus (Shimizu 1979),
Cyprinus carpio (Davis and Takashima 1980), and
Carassius auratus (Stromsten 1931; Takahashi and
Takano 1971); and the anabantids Mac ropoc/us eon-
color and M. opercularis (Schwier 1939) are hatched
as all females having ovaries with no testicular char-
acteristics. About one-half continue normal devel-
opment to mature females and about one-half under-
go a transitory intersexual stage before becoming
adult males. The female juvenile cyprinids Brachy-
danio rerio, Barbus tetrazona, Rhodeus ocellatus,
and Carassius auratus all develop an ovarian lumen,
as well as oogonia and previtellogenic oocytes. Pre-
maturational sex change is reportedly common
among the hermaphroditic Sparismatinae (Scaridae)
that spend their entire life as males (Robertson and
Warner 1978).
Social Control
We interpret the data on sexual maturity and sex-
ual dimorphism to suggest that some sexually
mature males delay participation in spawning for
up to 3 years and 10-15 cm, and that this mating
system is socially mediated. Histological assess-
ments of sexual maturity revealed that both males
and females produced mature gametes by about 50
cm FL and 5-6 years. However, visual inspection
of gonads to determine maturity gave the same
result as the histological evidence in females, but
indicated that males were not mature until attain-
ing 65-70 cm FL and 7-8 years. Ross and Merriner
(1983) also reported that some Caulolatilus microps
males that were visually assessed as immature were
later shown by histological methods to be produc-
ing mature gametes, and Erickson and Grossman
(1986) found L. chamaeleonticeps off Georgia that
showed active, yet incomplete spermatogenesis.
Sexual dimorphism in tilefish is manifested in the
size of the adipose flap which is conspicuously larger
in males than females and larger ultimate body size
in males; size of the adipose flap may be an indica-
tion of male breeding status. The adipose flap of
males became larger at 65-70 cm FL (7-8 years)
when males were judged sexually mature by visual
inspection (i.e., had developed large testes), not
758
CRIMES ET AL.: REPRODUCTIVE BIOLOGY OF TILEFISH
when they were functionally mature at 50 cm FL
as revealed by histological methods.
A nonrandom mating system of pairing, involv-
ing mate selection by females, is consistent with the
both sexual dimorphisms observed in tilefish.
Female mate selection can convey a reproductive
advantage to large males (Ghiselen 1969; Howard
1979), but in randomly mating species, females are
usually larger than males (Ghiselen 1969). Female
ability to discriminate males is required to support
female mate selection (Howard 1979), which often
leads to development of specialized structures and
colorations by males for display (Krebs 1972;
Warner and Robertson 1978). The enlarged adipose
flap in tilefish certainly represents a conspicuous,
highly visible feature in some adult males that could
serve as a visual cue to signal male breeding status.
The evolution of a female mate selection system
requires that a male have the ability to control
resources important to the female (Howard 1979;
Krebs and Davies 1984), and several lines of evi-
dence suggest that male tilefish may be territorial.
Burrowing is apparently the rule in the family
(Able et al. 1987). Direct observation from sub-
mersibles and mark-recapture data indicated that
tilefish orient to particular burrows, and may be
long-term residents of their habitats (Grimes et al.
1983, 1986). Furthermore, time-lapse photography
showed the same male-female pair of tilefish utiliz-
ing the same burrow over a 26-h period (Grimes et
al. 1986), and pair formation has been observed in
the branchiostegids Malacanthus plumieri (Clark
and Ben-Tuvia 1973), Hoplolatilus sp. (Thresher
1984), and H. starcki and H. cuniculus (Randall and
Dooley 1974).
We believe that our data indicate that tilefish have
a mating system consisting of two classes of sex-
ually mature males, a category actively engaged in
spawning and a category of satellite males that do
not spawn. Similar breeding systems have been
described for several species of hermaphroditic reef
fish (Popper and Fishelson 1973; Fishelson 1975;
Warner and Robertson 1978; Robertson and Warner
1978; Warner and Hoffman 1980; Shapiro 1984).
For example, there are scarid and labrid populations
with large territorial terminal phase males that have
preferred mating status, and nonterritorial initial
phase males that have nonpreferred breeding status
(Warner and Robertson 1978; Robertson and
Warner 1978). Hermaphroditic populations of the
serranid Anthias squamipinnis in the Gulf of Eilat
have two behaviorally distinct types of males, a
dominant territorial male that actively courts, inter-
acts, and spawns with females, and smaller males
that do not interact or spawn with females in the
social group. The latter male category have filamen-
tous degenerative gonads (Popper and Fishelson
1973; Fishelson 1975). These mating systems are
characterized by strong sexual selection and main-
tenance of reproductive territories by males, and by
being reef systems in which fish are habitat limited.
Tilefish are also severely habitat limited, i.e., to bur-
rowable clay substrate generally (Able et al. 1982;
Grossman et al. 1985; Grimes et al. 1986).
During the period we studied reproduction, the
fishery for tilefish was rapidly expanding, and one
effect of fishing seems to have been to alter the
structure of the mating system. Based upon both
age-structured and non-age-structured population
modeling, tilefish population density was reduced
by about one-half to two-thirds from 1978 to 1982,
apparently due to the rapid expansion of the com-
mercial longline fishery (Turner 1986). Female size,
and particularly age, at maturity do not seem to
have been altered in any consistent fashion by the
population reduction from fishing. Males, on the
other hand, appear to have experienced profound
changes in visually assessed maturity; they clearly
were mature at smaller size (10 cm) and younger
age (2-2.5 years) in 1982 than in 1978.
We interpret the decrease in size/age of maturity
in males to be the effect of fishing. Fishing lowered
population density, and in so doing may have made
mating territories available to smaller and younger
males. This interpretation is supported by the find-
ings of Warner and Robertson (1978) and Robert-
son and Warner (1978) that the ratio of the two
categories of sexually mature males (initial and ter-
minal phase) in western Caribbean scarid and labrid
populations was density dependent; i.e., relatively
more initial phase males were found in dense popu-
lations.
ACKNOWLEDGMENTS
Our tilefish research was initiated at the sugges-
tion of the late Lionel A. Walford. We tender
grateful appreciation to the following individuals
and institutions for their assistance in this research.
Joseph Desfosse, Susan Shepherd, Gary Shepherd,
and Stuart Katz helped extract data from samples.
Richard Trout and Bruce Babiarz provided statis-
tical and histological advice, respectively. Fran
Puskus, Louis Puskus, John Larsen, and the fisher-
men of Barnegat Light and Sea Isle City, NJ coop-
erated in obtaining tilefish samples. Daniel Erickson
759
FISHERY BULLETIN: VOL. 86, NO. 4
and an anonymous reviewer provided useful
editorial comments. Support for this research was
provided by New Jersey Sea Grant (RF-2), the New
Jersey Agricultural Experiment Station (AES
12409), and the Center for Coastal and Environmen-
tal Studies, Rutgers University.
LITERATURE CITED
Able, K. W., C. B. Grimes, R. A. Cooper, and J. R. Uzmann.
1982. Burrow construction and behavior of tilefish,
Lopholatilus chamaeleo'nticeps, in Hudson Submarine
Canyon. Environ. Biol. Fishes 7:199-205.
Able, K. W., D. C. Twichell, C. B. Grimes, and R. S. Jones.
1987. Tilefishes of the genus Caulolatilus construct burrows
in the sea floor. Bull. Mar. Sci. 40:1-10.
Bagenal, T. B., and E. Braum.
1978. Eggs and early life history. In T. B. Bagenal (editor),
Methods for assessment of fish production in fresh water,
p. 165-201. Blackwell Sci. Publ., Oxford.
Berrien, P.
1982. Larval fish distribution in the Middle Atlantic Bight.
In M. D. Grosslein and T. R. Azarovitz (editors), Fish dis-
tribution, p. 23-44. MESA New York Bight Atlas, Moongr.
15. New York Sea Grant Institute, Albany, NY.
Bigelow, H. B., and W. C. Schroeder.
1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv.
Fish. Bull. 53:1-577.
Clark, E., and A. Ben-Tuvia.
1973. Red Sea fishes of the family Branchiostegidae with a
description of a new genus and species Assynoetrwrus oreni.
Bull. Sea Fish. Res. Stn., Haifa 60:63-74.
Collins, J. W.
1884. History of the tilefish. Rep. U.S. Fish. Comm. 10:
237-294. (Doc. 81.)
Combs, R. M.
1969. Embryogenesis, histology and organology of the ovary
of Brevoortia patronus. Gulf Res. Rep. 2:333-434.
Cooper, R. A., and J. R. Uzmann.
1977. Ecology of juvenile and adult clawed lobsters, Homarus
americanus, H. gammarus and Nephrops norvegicus. In
B. F. Phillips and J. S. Cobb (editors). Workshop on lobster
and rock lobster ecology and physiology, p. 187-208.
Commonw. Sci. Ind. Res. Organ., Div. Fish. Oceanogr. Circ.
No. 7.
Davis, P. R., and F. Takashima.
1980. Sex differentiation in common carp, Cyprinus carpio.
J. Tokyo Univ. Fish. 66:191-199.
deVlamming, V. L.
1983. Oocyte development patterns and hormonal in-
volvements amoung teleosts. In E. R. Duggan (editor). Con-
trol processes in fish physiology, p. 176-199. Croom-Helm,
Lond.
DOOLEY, J. K.
1978. Systematics and biology of the tilefishes (Perciformes:
Branchiostegidae and Malacanthidae), with descriptions of
two new species. U.S. Dep. Commer., NOAA Tech. Rep.
NMFS Circ. 411, 78 p.
Erickson, D. L., and G. D. Grossman.
1986. Reproductive demography of tilefish from the South
Atlantic Bight with a test for the presence of protogynous
hermaphroditism. Trans. Am. Fish. Soc. 115:279-285.
Erickson, D. L., M. J. Harris, and (J. 1). (Jrossman.
1985. Ovarian cycling of tilefish, Lopholatilus chamaele-
nnticeps Goodc and Bean, from the South Atlantic Bight,
U.S.A. J. Fish Biol. 27:131-146.
Fahay, M. p., and p. Berrien.
1981. Preliminary description of larval tilefish Lojihulatilus
chamaeleonticepn. Rapp. I'.-v. Reun. Cons. int. P^xplor. Mer
178:600-602.
FiSHELSON, L.
1975. Ecology and physiology of sex reversal in Anthiaa
squamipinnis (Peters), (Teleostei: Anthiidae). ht R. Rein-
both (editor), Intersexuality in the animal kingdom, p. 284-
294. Springer-Verlag, N.Y.
FOUCHER, R. P., AND R. J. BEAMISH.
1977. A review of the oocyte development in fishes with
special reference to Pacific hake {Merlucciiis productus).
Can. Fish. Mar. Serv. Tech. Rep. 755, 16 p.
Freeman, B. L., and S. C. Turner.
1977. Biological and fisheries data on tilefish, Lopholatilua
chamaeleonticeps Goode and Bean. NOAA, Natl. Mar. Fish.
Serv., Northeast Fish. Cent., Sandy Hook Lab. Tech. Ser.
Rep. No. 5, 41 p.
Gale, W. F., and W. G. Deutsch.
1985. Fecundity and spawning frequency of captive tessel-
lated darters - fractional spawners. Trans. Am. Fish. Soc.
114:220-229.
Ghiselen, M. T.
1969. The evolution of hermaphroditism among animals. Q.
Rev. Biol. 44:189-208.
Grimes, C. B.
1987. Reproductive biology of the Lutjanidae: A review. In
J. J. Polovina and S. Ralston (editors), Tropical snappers and
groupers, biology and fishery management, p. 239-294.
Westview Press, Boulder, CO.
Grimes, C. B., K. W. Able, and R. S. Jones.
1986. Tilefish, LopholatilTis chamaeleonticeps, habitat, be-
havior and community structure in Mid-Atlantic and south-
ern New England waters. Environ. Biol. Fishes 15:273-
292.
Grimes, C. B. K. W. Able, and S. C. Turner.
1980. A preliminary analysis of the tilefish, Lopholatilus
chamaeleonticeps, fishery in the Mid-Atlantic Bight. Mar.
Fish. Rev. 42(11):13-18.
Grimes., C. B., S. C. Turner, and K. W. Able.
1983. A technique for tagging deepwater fish. Fish. Bull.,
U.S. 81:663-666.
Grossman, G. D., M. J. Harris, and J. E. Hightower.
1985. The relationship between tilefish (Lopholatilus cha-
maeleonticeps) abundance and sediment composition off
Georgia. Fish. Bull., U.S. 83:443-447.
Harris, M. J., and G. D. Grossman.
1985. Growth, mortality, and age composition of a lightly ex-
ploited tilefish substock off Georgia. Trans. Am. Fish. Soc.
114:837-846.
Hayashi, Y.
1977. Studies on the maturity and the spawning of the red
tilefish in the East China Sea - L Estimation of the spawn-
ing season from the monthly changes of gonad index. Bull.
Jpn. Soc. Sci. Fish. 43:1273-1277.
1979. Studies of the maturity and the spawning of the red
tilefish in the East China Sea - II. Spawning pattern esti-
mated from the monthly changes of ovarian egg diameters.
Bull. Jpn. Soc. Sci. Fish. 45(12):1475-1480.
Hickling, C. F., and E. Rutenberg.
1936. The ovary as an indicator of the spawning period in
760
GRIMES ET AL.; REPRODUCTIVE BIOLOGY OF TILEFISH
fishes. J. Mar. Biol. Assoc. U.K. 21:311-316.
Hoar, W. S.
1957. The gonads and reproduction. In M. E. Brown
(editor), The physiology of fishes, p. 287-321. Acad. Press,
N.Y.
Hoffman, R. A.
1963. Gonads, spermatic ducts, and spermatogenesis in the
reproductive system of male toadfish, Opsanus tau. Chesa-
peake Sci. 4:21-29.
Howaku, R.
1979. Estimating reproductive success in natural populations.
Am. Nat. 114(2):221-231.
Htu.n-Han, M.
1978. The reproductive biology of the dab, Limanda liman-
da (L), in the North Sea: gonosomatic index, hepatosomatic
index condition factor. J. Fish. Biol. 13:369-378.
Hyder, M.
1969. Histological studies on the testis of Tilapia leucosticta
and other species of the genus Tilapia (Pisces: Teleostei).
Trans. Am. Microsc. Soc. 88:211-231.
iDELBERGER, C. F.
1985. Reproductive biology of the tilefish, Lopholatilus
ckamaeleonticeps, in the Middle Atlantic Bight. M.S.
Thesis, Rutgers Univ., New Brunswick, NJ, 51 p.
Katz, S. J., C. B. Grimes, and K. W. Able.
1983. Delineation of tilefish. Lopholatilus ckamaeleonticeps,
stocks along the United States east coast and in the Gulf
of Mexico. Fish. Bull., U.S. 81:41-50.
Krebs, C. J.
1972. Ecology: the experimental analysis of distribution and
abundance. Harper and Row, N.Y., 694 p.
Krebs, C. J., and N. B. Davies.
1984. Behavioral ecology: an evolutionary approach. 2nd ed.
Sinauer Assoc, Sunderland, MA, 500 p.
La Roche, J. L., and S. L. Richardson.
1980. Reproduction of northern anchovy, Engrauiis mordax,
off Oregon and Washington. Fish. Bull., U.S. 78:603-618.
Markle, D. F., W. B. Scott, and A. C. Kohler.
1980. New and rare records of Canadian fishes and the
influence of hydrography on resident and non-resident Sco-
tian Shelf ichthyofauna. Can. J. Fish. Aquat. Sci. 37:49-65.
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. No. 10, 95 p.
Mrsic, W.
1923. Die Spatbefruchtung und deren Einfluss auf Entwick-
lung und Geschlechtsbildung, experimentell nachgepruft an
der Regenbogenforell. Arch. Mikrosk. Anat. Entwicklung-
smech. 98:129-209.
NiKOLSKY, G. V.
1963. The ecology of fishes. Acad. Press, N.Y., 352 p.
Popper, D.. and L. Fishelson.
1973. Ecology and behavior oiAnthias squamipinnis (Peters,
1855) (Anthiidae, Teleostei) in the coral habitat of Eilat (Red
Sea). J. Exp. Zool. 184:409-424.
Randall, J. E., and J. K. Dooley.
1974. Revision of the Indo-Pacific branchiostegid fish genus
Hoplolatiius, with descriptions of two new species. Copeia
1974:457-471.
RiCKER, W. E.
1973. Linear regressions in fishery research. J. Fish. Res.
Board Can. 30:409-434.
Robertson, D. R., and R. Warner.
1978. Sexual patterns in the labroid fishes of the western
Caribbean, II: the parrotfishes (Scaridae). Smithson. Con-
trib. Zool. No. 255, 26 p.
Ross, J. L.
1978. Life history aspects of the gray tilefish, Caulolatilus
microps (Goode and Bean, 1978). M.S. Thesis, College of
William and Mary, Williamsburg, VA, 120 p.
Ross, J. L., and J. V. Merriner.
1983. Reproductive biology of the blueline tilefish, Caulola-
tilus microps, off North Carolina and South Carolina. Fish.
Bull, U.S. 81:553-568.
Sadovy, v., and D. Y. Shapiro.
1987. Criteria for the diagnosis of hermaphroditism in fishes.
Copeia 1987:136-156.
Schwier, H.
1939. Geschlechtsbestimmung und - differenzierung bei
Macropodus opercularis, concolor, chinensis und deren Ar-
bastarden. Z. Indukt. Abstammungs-Vererbungsl. 77:291-
335.
Shapiro, D. Y.
1984. Sex reversal and sociodemographic processes in coral
reef fishes. In G. W. Potts and R. J. Wootten (editors). Fish
reproduction: strategies and tactics, p. 103-118. Acad.
Press, N.Y.
Shimizu, M.
1979. Studies on morphogenesis and sex differentiation of the
gonad in several teleost fishes. M.S. Thesis, Hokkaido
Univ., Hokodate, Japan, 74 p.
Smith, C. L.
1965. The patterns of sexuality and the classification of ser-
ranid fishes. Am. Mus. Novit. 2207:1-20.
Stromsten, F. a.
1931. The development of the gonads in the goldfish,
Carassius auratus (L). Iowa Stud. Nat. Hist. 13:3-45.
Takahashi, H.
1977. Juvenile hermaphroditism in the zebrafish, Brachy-
danio rerio. Bull. Fac. Fish. Hokkaido Univ. 28:57-
65.
Takahashi, H., and K. Takano.
1971. Sex hormone-induced precocious hj^jertrophy and cilia-
tion of epithelial cells in the ovarian lumen of the goldfish.
Annot. Zool. Jpn. 44:32-41.
Thresher, R. E.
1984. Reproduction in reef fishes. TFH Publ., Inc., Neptune
City, NJ, 399 p.
Turner, S. C.
1986. Population dynamics of, and impact of fishing on,
tilefish, Lopholatilus chamaeleonticeps, in the Middle
Atlantic-Southern New England region during the 1970's
and early 1980's. Ph.D. Thesis, Rutgers Univ., New Buns-
wick, NJ, 289 p.
Turner, S. C, C. B. Grimes, and K. W. Able.
1983. Growth, mortality and age/size structure of the fish-
eries for tilefish, Lopholatilus chamaeleonticeps, in the Mid-
dle Atlantic-Southern New England region. Fish. Bull.,
U.S. 81:751-763.
U.S. Department of Commerce.
1980a. New Jersey landings. December 1979. U.S. Dep.
Commer., NOAA, NMFS, Current Fish. Stat. 7973.
1980b. New York landings, annual summary. U.S. Dep.
Commer., NOAA, NMFS, Current Fish. Stat. 7979.
1980c. Rhode Island landings, December 1979. U.S. Dep.
Commer., NOAA, NMFS, Current Fish. Stat. 7975.
Valentine, P.C, J. R. Uzmann, and R. A. Cooper.
1980. Geology and biology of Oceanographer Submarine
Canyon. Mar. Geol. 38:283-312.
761
FISHERY Hl'LLKTIN: VOL. 86, NO. 4
Waltz, C. W., W. A. Roumillat, and C. A. Wenner.
1982. Biology of the whitebone porgy, Calamus leucosteus,
in the South Atlantic Bight. Fish. Bull., U.S. 80:863-874.
Warme, J. E., R. A. Slater, and R. A. Cooper.
1977. Bioerosion in submarine canyons, /w D. J. Stanley and
G. Kelling (editors). Submarine canyon, fans and trench
sedimentation, p. 65-70. Hutchinson and Ross Publ.,
Dowden.
Warner, R. R., and S. G. Hoffman.
1980. Local population size as a determinant of mating
system and sexual composition in two tropical marine fishes
{Thalassoma spp.). Evolution 34:508-518.
Warner, R. R., and D. R. Robertson.
1978. Sexual patterns in the labroid fishes of the western
Caribbean, I: the wrasses (Labridae). Smithson. Contrib.
Zool. No. 254, 27 p.
Yamamoto, K.
1956. Studies on the formation of fish eggs I. Annual cycle
in the development of ovarian eggs in the flounder, Linp-
setta obscura. J. Fac. Sci. Hokkaido Univ., Ser. VI, Zool.
12:362-373.
762
A FIELD METHOD FOR DETERMINING PREY PREFERENCE OF
PREDATORS
N. B. Hargreaves^
ABSTRACT
A new field method for determining prey preferences of fish that feed on juvenile salmon is described.
The basic elements of this method consist of capturing, tagging or marking, and releasing prey with
known characteristics, and comparing these characteristics with those of tagged prey subsequently
recovered from the stomachs of predators. The feasibility of this approach is illustrated by two experiments
conducted in 1985, designed to assess prey size preferences of predators feeding on juvenile pink salmon
during the early sea-life period. The results indicate that yearling coho salmon, Oncorhynchus kisutch,
were size selective when feeding on juvenile pink salmon, 0. gorbuscha, preferring the smaller prey.
A major advantage of this method is that it eliminates the need to determine the abundances of various
prey t\T)es in the field. It also allows the investigator to control or precisely measure many of the variables
that are known to affect the availability of prey to predators.
Predation plays an important role in shaping the
ecological structure of many biological systems. One
aspect of predation that has attracted considerable
interest is the observation that, when offered a
choice of prey types, predators typically show a pref-
erence for one of them. The result is that more of
the preferred prey are consumed than would be ex-
pected, based on the relative abundances of the
various prey types.
There have been many attempts to quantify the
food or prey preferences of predators (e.g., Hess and
Swartz 1940; Ivlev 1961; Schneider 1981) using a
wide variety of mathematical indices of preference
(reviewed by Cock 1978; Pearre 1982). In some
situations, however, this approach clearly is not
suitable. For example, in many systems the relative
abundances and species composition of the prey can
vary substantially over the normal feeding range of
the predators. This is particularly evident in fish-
eries, where piscivorous predators and their prey
are often very mobile, and can travel considerable
distances during even a single feeding period. In
such cases, the proportions of the various prey found
in the stomachs of predators may result from varia-
tions in the relative concentrations or availability
of prey over the extensive area searched by the
predator, rather than from any prey preference. It
is typically very difficult to determine the concen-
^Department of Fisheries and Oceans, Fisheries Research
Branch, Pacific Biological Station, Nanaimo, B.C. V9R 5K6,
Canada.
trations and species composition of prey over such
large areas, so indices of preference may not pro-
vide much insight into the predation process.
The purpose of this paper is to describe a new
method for determing the prey preference of pred-
ators in the wild. This method consists of a field
experiment in which a number of potential prey with
known characteristics are released. These prey are
tagged prior to release to allow positive identifica-
tion even if they mix with other prey of the same
type after they are released. The prey preference
of the predators is assessed directly by comparing
the characteristics of the tagged prey found in the
stomachs of predators with those of the prey that
were released. The major advantage of this ap-
proach is that the relative abundances and other
characteristics (species ratios, size composition, etc.)
of the various prey types are known in advance and
additional field measurements are not required. In
addition, if only tagged prey are compared, there
is much less ambiguity in assessing the preference
of predators for each type of prey because the major
alternative explanations are eliminated. Although
applicable to a wide variety of predator-prey inter-
actions, the details and potential utility of this
method are illustrated by two experiments con-
ducted in Masset Inlet and Masset Sound, B.C.,
Canada. In both cases the goal was to test the
hypothesis that natural fish predators were size
selective when feeding on juvenile pink salmon,
Oncorhynchus gorbuscha, during the early sea-life
period.
Manuscript accepted July 1988.
FISHERY BULLETIN: VOL. 86, NO. 4, 1988.
763
FISIIKKV HlLl.KTIN: VOL. 86, NO. 4
METHODS
The first experiment was conducted on 26 April
and the second experiment during 6 and 7 June
1985. Pink salmon fry were captured in the Yakoun
River during 10-18 April using two inclined plane
traps with mouth openings of 1.1 m'^. The traps
were emptied each morning and the pink salmon
transferred by truck to several 1 m^ net pens at-
tached to the research vessel Velella, anchored at
Marinelli Point (Fig. 1).
All pink salmon were tagged and marked to allow
positive identification after they were released. Dur-
ing 15-18 April 6,000 pink salmon were anesthe-
tized with tricaine-methanesulfonate (MS-222) and
tagged with half-length binary-coded wire tags by
an experienced tagging crew. Proper placement of
each tag was confirmed by passing all fish through
"% ^''
^ :'''' '
K ''t;^
i
\l ~^^^*>'»s»^
1.
■^^^>=^?si^^S^
MASSET SOUND AND INLET vS 54°oo'-
-\^'£j/^5''\ 0 5nm g i i I'i:;
W^^^ British \ 'f\'(v'
>> ^>*^ Columbia \ /i ^ "^\
Pacfic ^^^ \ ^^^ %'^
- T^^^ ^°
>5^!fe Y /
Ocean ^LAiV
/ ir
\ '^ U S A
/ ^/
K
f tii^T ^^°^°'~
- 130° 1251 45°-
1 1
/v^^
:J ir f.
:f^/f |,
Y^^^^-^i^^ i
_^X^ - ^v-^'fet
u-»*^=''^^^^^'^*^'~x_.,^ji?**'^^ '. ''' _ii "^r
r|;v , ■■'J'^0^^ * '^'tP^'''^'^''^^'^^^ \ 1 53°40'-
- *^ * '
^ly
I32»30' I32»20' I32°I0'
1 1 1
Figure 1.— Chart of Masset Inlet showing locations of first (Marinelli Point) (A) and second
(B) releases of tagged fish and saltwater enclosure (C) used to hold tagged fish until required
for the second experiment.
764
HARGREAVES: PREY PREFERENCE OF PREDATORS
a 2.54 cm (1 inch) diameter Quality Control Device
(QCD), manulactured by Northwest Marine Technol-
ogy^ (Shaw Island, WA 98286). Samples of tagged
fish were also dissected and examined visually and
microscopically to verify tag implantation. Three
thousand of these pink salmon were also marked
externally by amputating either the left or right pec-
toral fin. No other fins were amputated. The pur-
pose of the fin clipping was to determine which
method (nose tag or fin clip) was more effective for
identifying fish recovered from the stomachs of
predators.
After tagging and marking, all pink salmon were
transferred back into the small, saltwater enclo-
sures. On 20 April the three thousand tagged but
unmarked pink salmon were transferred to a larger
(51 m'^) enclosure anchored near the southwest end
of Masset Inlet, and held until required for the sec-
ond experiment. This enclosure was shallowed by
hand once each week to check the condition of the
fish and cleaned with a high-pressure hose every
10-14 days to remove algae growing in the meshes.
The food supply of the pink salmon held in this
enclosure was not controlled and consisted of what-
ever came through the meshes. No supplementary
food was added. Under this regime the pinks re-
mained very active and appeared healthy.
The first experiment was initiated at 1600 on 26
April, by releasing 3,000 tagged and fin-clipped pink
salmon at Marinelli Point. A sample of 99 fish was
removed, anesthetized with 2-phenoxyethanol, and
the live fork length of each fish measured to deter-
mine the size distribution of fish at the time of
release. A single beach seine set was made three
hours later, in the immediate vicinity of the release
site, to collect a sample of potential predators.
The second experiment was initiated by releasing
1.800 tagged pink salmon at 1130 on 6 June, into
the boat harbor at Masset (Fig. 1). None of these
fish were fin-clipped. A sample of 100 pink salmon
was removed prior to the release, and each fish
measured to determine the size distribution of fish
at the time of release. Two beach seine sets were
made to collect potential predators prior to the time
of release, the first at 0930 and the second at 1015.
A total of sixteen additional sets were made after
the release, seven between 1245 and 1826 on 6 June,
and nine more between 0900 and 1400 the follow-
ing day.
The beach seine used to capture potential pred-
ators was 46 m long and constructed of 6.4 mm
stretched nylon mesh. All potential predators were
examined immediately after capture. Each mea-
sured fish was anesthetized in 2-phenoxyethanol,
and the stomach contents obtained by either
hydraulic flushing or dissection. All fish remains in
the stomach contents were examined visually or
microscopically to identify prey to the species level,
using Hart (1973) as a general reference and Phillips
(1977) to identify juvenile salmon. All juvenile
salmon found in the stomachs of predators and any
live juvenile salmon captured along with the pred-
ators were measured (if possible), examined for
missing fins, and passed through the QCD several
times to determine if they were tagged.
Yearling coho salmon, Oncorhynchus kisutch, cap-
tured on the first day of the second experiment were
used to obtain an estimate of the total population
of coho salmon in the vicinity of the release site. All
coho salmon captured on 6 June were retained,
anesthetized with 2-phenoxyethanol, and marked
using a hot-wire branding device. At 2000 on 6 June,
these coho were sorted to remove any that did not
appear healthy and fully recovered, and the remain-
ing 170 were released. All coho captured the follow-
ing day were inspected to determine if they were
marked. The numbers of marked and unmarked
coho were used to derive a simple (single census)
Petersen estimate of the total population of year-
ling coho salmon in the vicinity of the release site
using the following equation (Ricker 1975):
A^
{M + 1) (C + 1)
(R + 1)
(1)
^Reference to trade name does not imply endorsement by the
National Marine Fisheries Service, NOAA.
where A^ = total number of coho salmon at time
of marking
M = number of coho salmon marked and
released on 6 June
C = total number of coho salmon captured
on 7 June
R = number of marked coho salmon recap-
tured on 7 June.
The 95% confidence interval for the population
estimate was obtained by substituting into this equa-
tion the fiducial limits of the number of recaptured
coho salmon, from tables of the Poisson distribution
(Ricker 1975). These figures were used to estimate
the total number of tagged pink salmon eaten by
coho salmon in the second experiment, using the
equation
765
FISHERY HTLLKTIN: VOL. 8«. NO. 4
T = (P * N)/E
(2)
where T = total number of pink salmon eaten
P = total number of pink salmon in all
stomachs examines
N = estimated total population of coho
salmon, from Equation (1)
E = total number of stomachs examined.
RESULTS
The 3,000 tagged and marked pink salmon re-
leased at Marinelli Point on 26 April ranged in size
from 34 mm to 41 mm, with an average of 38 mm
(Table 1). A total of 57 juvenile sockeye salmon {On-
corhynchus nerka), 5 pink salmon, 2 chum salmon
(0. keta), 4 starry flounder (Platichthys stellatus),
1 sturgeon poacher {Agonies acipenserinus), and 210
yearling coho salmon were captured in the single
beach seine set done after the tagged fish were
released. All five pink salmon had been fin-clipped
and tagged. Three of these pinks had fork lengths
of 37 mm and the other two were both 38 mm long.
The two chum salmon had fork lengths of 41 mm
and 40 mm. The four starry flounders and single
sturgeon poacher were all small and were imme-
diately released. All 210 coho salmon were measured
and their stomach contents examined for evidence
of predation on juvenile pink salmon. A total of 90
coho salmon had remains of fish in their stomachs,
including 51 pink salmon, 1 chum salmon, and 14
juvenile salmon that could be identified only as
either pink or chum salmon, due to extensive
digestion (Table 2). One additional coho salmon had
eaten two Pacific sandlance, Anmiodytes hexapterus,
but all other fish remains were too digested to
positively identify. Out of the 23 tagged pink salmon
found in the stomachs of these coho salmon, only
18 were sufficiently intact to permit measurements
of their fork length (Table 2). The average length
of these pink salmon was significantly less {t = 8.02;
P < 0.001) than the average length of the pink
salmon that were released (Fig. 2). Twelve of the
eighteen tagged and measurable pink salmon found
in the stomachs of the coho were clearly missing a
pectoral fin. The pectoral fins of all the other fish
were too digested to be certain whether or not they
had been fin-clipped. The average length of the 12
Table 1 .—Characteristics of tagged juvenile pink salmon released in each experiment and those subsequently recovered by
beach seining. Nr = no. released; Ns = no. fish in prerelease sample; Min = minimum fork length; Max = maximum fork
length; X = mean fork length; C.I. = 95% confidence interval forX; Nc = no. of tagged salmon recaptured by beach seining;
H = no. hours after release of tagged fish.
Tagged fish released
Tagged fish
recaptured
Date
Species
Nr Ns Min Max X
C.I.
Species
Nc
Min
Max
X
C.I.
H
4/26/85
pink
3,000 99 34 41 38.1
37.9-38.4
pink
5
37
38
37.4
36.7-38.1
1.5
6/6/85
pink
1,800 100 36 51 44.4
43.8-44.9
pink
3
42
46
44.3
39.2-49.5
1-3
pink
12
40
50
44.7
42.7-46.6
3-5
pink
4
40
47
42.8
38.0-47.5
5-7
pink
0
—
—
—
—
21-26
Table 2.— Number and size of juvenile salmon found in stomachs of predators in each experiment. Sets = no.
beach seine samples; S = no. of stomachs examined; N = total no. juvenile salmon in stomachs; Nc = no. of
tagged fish in stomachs; Nm = no. of tagged fish whose fork length could be measured; Min =_minimum fork length;
Max = maximum fork length; X = average fork length; C.I. = 95% confidence interval forX; JS = juvenile pink
or chum salmon.
Date
No.
sets
Predator
S
Juvenile salmon found in
predator stomachs
Experi-
ment
All salmon
Species N
Tagged salmon only
Nc
Nm
Min
Max
X
C.I.
1
2
4/26/85
6/6/85
1
16
coho
coho
staghorn
Dolly Varden
210
374
24
15
pink
chum
JS
pink
pink
pink
51
1
14
17
0
0
23
0
0
15
18
15
34
36
38
45
35.6
40.7
35.0-36.2
39.3-42.1
766
HARGREAVES: PREY PREFERENCE OF PREDATORS
26-
19
cc
5-
b
34 35 36 37 38 39 40 41 42
FORK LENGTH (mm)
Figure 2.— Size-frequency distribution of sample of tagged pink
salmon released in the first experiment (A) and all tagged pink
salmon subsequently recovered from stomachs of predators (B).
fin-clipped fish was 35.4 mm, which is not signifi-
cantly different {t = 0.47; P > 0.50) from the aver-
age length of the 18 tagged and measurable pink
salmon.
Pink salmon released in the second experiment
ranged in size from 36 to 51 mm fork length, with
an average of 44.4 mm (Table 1). A total of 1 juvenile
sockeye salmon, 1 juvenile chum salmon, 28 year-
ling coho salmon, 6 Pacific staghorn sculpin {Lep-
tocottus armatus), 10 Pacific herring (Clupea
harengus pallasi), approximately 500 larval walleye
pollack (Theragra chalcogramma), 20 Pacific snake
prickleback (Lumpenus sagitta), 1 red Irish lord
{Hemilepidotus hemilepidotus), 6 Dolly Varden
(Salvelinus malma), and 10 starry flounder were
captured in the 2 beach seine sets done before the
release. Examination of the stomach contents of all
the coho salmon, five staghorn sculpins, the red Irish
lord, and all the Dolly Varden provided no indica-
tion that any of these fish had recently eaten juvenile
pink or chum salmon. The seven additional beach
seine sets done on 6 June after the release of the
tagged pink salmon captured a total of 271 coho
salmon, 33 Dolly Varden, 46 staghorn sculpin, 2
coastal cutthroat trout {Salmo clarki clarki), 55 pink
salmon, and 3 chum salmon. Nineteen pink salmon
had been tagged (Table 1). All the untagged pinks
were larger.
The stomach contents of 246 coho salmon, 14
Dolly Varden, and 16 sculpins were examined. There
was no evidence that any of the Dolly Varden or
sculpins had eaten any juvenile pink or chum salmon,
although one sculpin had eaten a juvenile coho
salmon. However, 13 of the coho salmon had eaten
a total of 17 juvenile pink salmon, of which 15 had
been tagged. The head was missing from one of the
two untagged pink salmon so it could not be mea-
sured, but the fork length of the other was 51 mm.
The tagged pink salmon found in the stomachs of
the coho salmon were typically the smaller ones (Fig.
3), with an average length significantly less {t =
5.06; P < 0.001) than the average length of all pink
salmon that were released in this experiment (Table
2).
No juvenile salmon were found in the stomachs
of predators captured on 7 June. A total of 1 juvenile
pink salmon, 141 yearling coho salmon, 1 Dolly
Varden, and 115 staghorn sculpins were captured
in 9 beach seine sets. The pink salmon was not
tagged and none of these predators had recently
eaten any juvenile pink or chum salmon. Five of the
coho salmon had been branded, so the estimated
total population of coho salmon in the vicinity of the
release site was 4,047, with a 95% confidence inter-
14-
10-
cr
LJ
m 6
-z.
2-1
36 38 40 42 44 46 48
FORK LENGTH (mm)
50 52
Figure 3.— Size-frequency distribution of sample of tagged pink
salmon released in the second experiment (A) and all tagged pink
salmon subsequently recovered from stomachs of predators (B).
767
FISHERY BULLETIN: VOL. 86, NO. 4
val of 1,912 to 9,339. Based on these figures, it is
estimated that coho salmon consumed a total of 162
(9%) of the tagged pink salmon released on 6 June,
with a 95% confidence interval ranging from 76 (4%)
to 375 (21%).
DISCUSSION
Determining the prey preference of predators in
the wild is an important but difficult problem. The
two experiments reported here illustrate a new ap-
proach to determining the selectivity of predators
in the wild. Aside from logistic problems, the suc-
cess of this method depends on the validity of four
main assumptions: 1) predators that are captured
and examined, and their stomach contents, are truly
representative of the total predator population of
interest; 2) tagging or marking the prey does not
result in abnormal behavior of either the prey or
predators; 3) ingestion and partial digestion of the
prey by the predators does not significantly alter
the characteristics of the prey that are of primary
interest; 4) all of the tagged prey remain equally
"available" to the predators for the duration of the
experiment.
The first assumption should be valid if the sam-
pling program is appropriately designed, consider-
ing the statistical tests that will be used to analyze
the data. This is a complex topic and an in-depth
discussion is beyond the scope of this paper. How-
ever, many extensive references are available (e.g.,
Anderson and McLean 1974; Cochran 1977; Mont-
gomery 1976).
The two experiments reported here were designed
only to demonstrate the utility of this approach and
do not clearly show how generally applicable the
results are. Only a small number of tagged fish were
released and samples of predators were collected
with a beach seine, which undoubtedly is biased to
some degree in terms of the species and sizes of fish
that were captured. In addition, the samples were
collected at only one location in the first experiment
and over a relatively small area in the second experi-
ment. However, extensive sampling and examina-
tion of the stomach contents of fish predators over
a 3-yr period indicates that yearling coho salmon are
the major predators of juvenile pink salmon through-
out Masset Inlet and Masset Sound (Hargreaves in
press). The results of these two experiments are also
consistent with those obtained from enclosure ex-
periments, which indicate that yearling coho salmon
are size selective when feeding on juvenile pink or
chum salmon (Parker 1971; Hargreaves and LeBras-
seur 1986). Thus, although quite limited in scope,
these two experiments provided results that are con-
sistent with those obtained by two other indepen-
dent, but considerably more expensive and labori-
ous, methods.
The assumption that tagging and marking the
prey does not affect the behavior of predators or
prey can be assessed either by direct observation
or by conducting additional experiments. In some
cases it may be possible to design the experiment
to allow observation of both the predators and prey
throughout the experiment and directly observe any
unusual behavior. However, in many cases, addi-
tional experiments will probably be required. For
the two experiments reported here, the pink salmon
used in the first experiment were tagged and fin-
clipped; the fish released in the second experiment
were tagged but were not fin-clipped. The results
of the first experiment indicate that tagging was
more effective than fin-clipping for recognizing fish
recovered from the stomachs of predators. In terms
of behavioral changes, previous work indicated that
the mortality of tagged and untagged juvenile
salmon was not significantly different when exposed
to predators and that tagging juvenile salmon had
no noticeable affect on the behavior of either the
predators or prey (Hargreaves and LeBrasseur
1986). However, the tagged fish used in these en-
closure experiments were not fin-clipped.
Earlier studies have indicated that amputation of
fins from juvenile salmon typically results in lower
survival rates (Ricker 1949). Marked pink salmon
fry also suffer higher mortality than unmarked fry,
possibly due to a bias on the part of predators for
marked prey (Parker et al. 1963). This is not a major
concern in the two experiments reported here be-
cause the intent was to determine the size selectivity
of predators, rather than any selectivity for marked
or unmarked prey. In addition, fin-clipped pink
salmon were used only in the first experiment but
predators consumed significantly more of the
smaller prey in both experiments. This supports the
assertion that fin-clipping the pink salmon did not
substantially affect the prey size selectivity of the
predators. In general, the possibility that the results
of these types of experiments may not apply to un-
tagged or unmarked prey can be eliminated by using
only tags or marks that are known to have negli-
gible effects on the behavior of both the prey and
predators.
The third assumption, that ingestion and partial
digestion of the prey by the predators does not sig-
nificantly alter the important characteristics of the
768
HARGREAVES: PREY PREFERENCE OF PREDATORS
prey, can often be directly verified. For the two ex-
periments reported here, prey size (fork length) was
the characteristic that was most important. The
observed difference between the average length of
the tagged prey that were released and the tagged
prey subsequently recovered from the stomachs of
predators was 2.5 mm (6.6%) in the first experiment,
and 3.7 mm (8.3%) in the second experiment. These
are small differences, and the possibility that they
might be due to experimental error rather than
predator selectivity must be considered. All length
measurements were made to the nearest millimeter
and numerous remeasurements indicated that mea-
surement errors were negligible at this level of ac-
curacy. To eliminate the possibility that the length
of the tagged fish might decrease if they were pre-
served (Parker 1963), live fish were used to deter-
mine the length-frequency distribution of the prey
prior to release and all prey recovered from the
stomachs of predators were immediately measured.
Burgner (1962) reported that the length of sockeye
salmon smolts decreased by 2-3% because of rigor
mortis alone. However, experiments conducted in
Masset Inlet in 1984 indicated much smaller changes
occur after death in juvenile pink salmon. At tem-
peratures of 9°-10°C, the average length of 26
juvenile pink salmon of known length, fed to and
subsequently recovered from the stomachs of 19
yearling coho salmon, decreased less than 1% for
periods of up to four hours after ingestion (Har-
greaves unpubl. data). In the two experiments re-
ported here, numerous beach seine sets were made
to capture potential predators, but all of the tagged
pink salmon found in their stomachs were recovered
within four hours of the releases of tagged prey.
Shrinkage of the prey after ingestion therefore can
account for only a small portion of the observed
differences in size between the prey that were
released and those that were found in the stomachs
of predators.
The fourth assumption, that all tagged or marked
prey remain equally "available" to predators
throughout the experiment, will usually prove to be
the most difficult to assess and verify. The avail-
ability of prey to predators frequently depends on
characteristics of the predators (hunger level, visual
acuity, mobility, body or gape size, individual or
group behavior, etc.), the prey (abundance, colora-
tion, size, speed, endurance, behavior, etc.), and the
environment (habitat complexity, light conditions,
etc.). These parameters can interact in a complex
manner, so that it is typically only in the simplest
situations that all factors that affect the availabil-
ity of prey to a predator can be thoroughly inves-
tigated and understood (Curio 1976; Zaret 1980).
Predators consumed significantly more of the
smaller prey in both experiments reported here,
despite substantial differences in the physical char-
acteristics of the two release sites, time of year, and
various characteristics of the predators (abundance,
species composition, size, feeding history, etc.). This
suggests that the availability of prey to the preda-
tors was not substantially affected by variations in
the characteristics of either the predators or the en-
vironment. It also appears reasonable to assume
that prey of all sizes remained equally available to
predators during both experiments. All of the
tagged prey were one species and received identical
treatment prior to release. There is no reason to
think there were any substantial differences in the
physical characteristics among the prey at the time
of release, aside from the desired variation in size.
It is conceivable, however, that differences in prey
behavior or size might have indirectly influenced the
availability of prey to the predators. For example,
extensive sampling of juvenile salmon in Masset
Inlet has indicated a tendency for larger pink salmon
to be concentrated further offshore than smaller
pink salmon during the early sea-life period (Har-
greaves et al. 1987a, b). Swimming speeds of salmon
are also known to increase rapidly with increasing
body size (Brett 1965). Thus, if there was any
tendency for tagged salmon to rapidly swim away
from the release sites, larger fish may have left
quicker than smaller fish. The result could be a
decrease in the average size of tagged salmon found
in the immediate vicinity of the release site and the
incorrect conclusion that predators were selective-
ly feeding on the smaller prey.
In fact, however, this possibility appears unlike-
ly. In both experiments the size of the live, tagged
fish recovered along with the predators was not
significantly different than the size of the fish that
had been released as much as nine hours earlier
(Table 1). There is also no indication that the mean
size of these fish changed in a consistent manner
over the course of the second experiment. These
results suggest that, if there was any segregation
of tagged prey after release, it was probably minor
and did not appreciably affect the availability of prey
to the predators.
In general, complications arising from variations
in the availability of prey to the predators may be
reduced or eliminated by limiting the duration of the
experiment. If all of the prey are released at one
time and location, it is reasonable, and in most cases
769
FISHERY BULLETIN: VOL. 86, NO. 4
probably valid, to assume that all prey are equally
available to any predators captured in the immediate
vicinity a short time later. The amount of time dur-
ing which the prey subsequently remain equally
available to predators will likely vary from one situa-
tion to the next. If the prey are very mobile, prob-
ably some will eventually become less or more acces-
sible to predators than others. This possibility can
be eliminated or at least minimized by keeping the
experiment short enough to ensure that the prey do
not have sufficient time to segregate or move away
from the release site. The magnitude of this prob-
lem and thus the appropriate duration for each ex-
periment may be assessed by recapturing some of
the tagged prey after the release. The experiment
should be terminated when the characteristics of the
recaptured prey begin to diverge significantly from
those of the original prey population.
Determining the prey preference of predators in
the wild is a concern to many biologists. All methods
of determining the prey selectivity of predators in
the wild are, and will likely continue to be, hampered
by the complexity of the related problem of deter-
mining the relative "availability" of prey to pred-
ators. The advantage of the method proposed here
is that it allows the investigator to control some of
the major variables that are known to affect the
availability of prey. The most important character-
istics of the prey (species ratios, abundance, size
ranges, etc.) can be determined before any preda-
tion occurs and in many cases can also be precisely
controlled. The predators remain free to feed on all
types of prey in the study area, but for the purposes
of the investigator, the choice of prey can effectively
be reduced to those with known characteristics and
origin. This is a major advantage when compared
with the more traditional approach of calculating
selectivity indices, as it eliminates the need to deter-
mine the relative abundances of prey in the field.
It also substantially reduces the ambiguity associ-
ated with interpreting selectivity indices for highly
mobile predators, where typically there is little or
no information available concerning the area
traveled by the predator during the feeding period
and thus what prey were actually available to the
predator.
The specific goal of the two experiments reported
here was to determine if predators were size selec-
tive when preying on juvenile pink salmon during
the early sea-life period. The results indicate that
yearling coho salmon were the dominant predator
of juvenile pink salmon at two locations, one in
Masset Inlet and the other in Masset Sound, and
that the average size of juvenile pink salmon con-
sumed by these predators was significantly less than
the average size of pink salmon that were released.
These results are consistent with those obtained
from two other independent approaches and suggest
this method may be a viable and cost-effective alter-
native for determining the prey preferences of pred-
ators in the wild. It may be particularly useful for
assessing prey preferences of predators feeding on
juvenile salmon near hatchery facilities in Canada
and the United States, where millions of juvenile
salmon are currently tagged and released each year.
ACKNOWLEDGMENTS
Suggestions and comments from an anonymous
reviewer are greatly appreciated. Robin LeBrasseur
and Owen Kennedy assisted with the supervision of
the field experiments and analyses of the stomach
samples. Bruce Patten, Lui Marinelli, Tom Poole,
Josette Weir, Rick Hobbs, Ted Carter, and Bob
Hungar helped to capture, tag, and sort the juvenile
salmon and assisted in the beach seining and pro-
cessing of samples collected in the various experi-
ments. Trans-Provincial Airlines generously per-
mitted unlimited use of their seaplane wharf and
other facilities at Masset, B.C. to conduct the sec-
ond experiment during 6 and 7 June 1985.
LITERATURE CITED
Anderson, V. L., and R. A. McLean.
1974. Design of experiments: A realistic approach. Marcel
Dekker Inc., N.Y.
Brett, J. R.
1965. The relation of size to rate of oxygen consumption and
sustained swimming speed of sockeye salmon (Oncorhynchus
nerka). J. Fish. Res. Board Can. 22:1491-1501.
BURGNER, R. L.
1962. Studies of red salmon smolts from the Wood River
Lakes, Alaska. In T. S. Y. Koo (editor), Studies of Alaska
red salmon. Univ. Wash. Publ. Fish. N.W. I(6)-251-316.
Cock, M. J. W.
1978. The assessment of preference. J. Anim. Ecol. 47:805-
816.
Cochran, W. G.
1977. Sampling techniques. 3d ed. John Wiley and Sons,
N.Y., 428 p.
Curio, E.
1976. The ethology of predators. Springer- Verlag, N.Y., 250
P-
Hargreaves, N. B.
In press. Predation of juvenile pink (Oticorhynchus gorbuscha)
and chum (0. keta) salmon in Masset Inlet, B.C. Can. J.
Fish. Aquat. Sci.
Hargreaves, N.B., E. W. Carter, and R. J. LeBrasseur.
1987a. Beach seine catches of juvenile salmon and other fish
770
HARGREAVES: PREY PREFERENCE OF PREDATORS
in Masset Inlet and Masset Sound, B.C., in 1984. Can. Data
Rep. Fish. Aquat. Sci. 640, 79 p.
HARGREAVES, N. B., AND R. J. LeBRASSEUR.
1986. Size selectivity of echo salmon (Oncorhynchus kisutch)
salmon preying on juvenile chum salmon (0. keta). Can. J.
Fish. Aquat. Sci. 43:581-586.
HARGREAVES, N. B., B. A. PaTTEN, AND R. J. LeBRASSEUR.
1987b. Beach seine catches of juvenile salmon and other fish
in Masset Inlet and Masset Sound, B.C. in 1985. Can. Data
Rep. Fish. Aquat. Sci. 632, 256 p.
Hart, J. L.
1973. Pacific fishes of Canada. Fish. Res. Board Can. Bull.
180, 74 p.
Hess, A. D., and A. Swartz.
1940. The forage ratio and its use in determining the food
grade in streams. Trans. 5th North Am. Wildl. Conf., p.
162-164.
Ivlev, V. S.
1961. Experimental ecology of the feeding of fishes. Yale
Univ. Press, New Haven.
Montgomery, D. C.
1976. Design and analysis of experiments. John Wiley and
Sons, N.Y., 418 p.
Parker, R. R.
1963. Effectsof formalin on length and weight of fishes. J.
Fish. Res. Board Can. 20:1441-1455.
1971. Size selective predation among juvenile salmonid fishes
in a British Columbia inlet. J. Fish. Res. Board Can. 28:
1503-1510.
Parker, R. R., E. C. Black, and P. A. Larkin.
1963. Some aspects offish marking mortality. Int. Comm.
Northwest Atl. Fish. Spec. Publ. No. 4, p. 117-122.
Pearre. S., Jr.
1982. Estimating prey preference by predators: uses of
various indices and a proposal of another based on x^- Can.
J. Fish. Aquat. Sci. 39:914-923.
Phillips, A. C.
1977. Key field characteristics identifying young marine
Pacific salmon. Fish. Mar. Serv. Tech. Rep. 746, 13 p.
RiCKER, W. E.
1949. Effect of removal of fins upon the growth and survival
of spiny-rayed fishes. J. Wildl. Manage. 13:29-40.
1975. Computation and interpretation of biological statistics
offish populations. Bull. Fish. Res. Board Can. 191, 382 p.
Schneider, D. C.
1981. Size-selective predation of mysids by birds. Mar. Ecol.
Prog. Serv. 5:223-224.
Zaret, T. M.
1980. Predation and freshwater communities. Yale Univ.
Press, New Haven, CT, 187 p.
771
EXPERIMENTAL MANIPULATION OF POPULATION DENSITY AND
ITS EFFECTS ON GROWTH AND MORTALITY OF JUVENILE
WESTERN ROCK LOBSTERS, PANULIRUS CYGNUS GEORGE
Richard F. Ford/ Bruce F. Phillips, ^ and Lindsay M. Joll^
ABSTRACT
A density manipulation experiment was conducted at Seven Mile Beach, Western Australia, to compare
growth and mortality for different density groups of juvenile western rock (spiny) lobsters, Panulirus
cygniis, inhabiting limestone patch-reefs. Juveniles on a control reef were left at their natural, high den-
sities while those on a treatment reef were reduced to approximately 25% of the original, natural density
by trapping and reintroduction, which maintained the original size-frequency distribution. Mark-recapture
studies were conducted on each reef at three monthly intervals for a year to estimate size-specific growth
rates, population densities, and mortality rates. Direct counts of individuals were made by divers to
estimate total numbers of juveniles on each reef. There were no statistically significant differences in
growth rates for any age category between the control and treatment reefs, but there were significant
differences in size-specific mortality rates between the treatment and control groups, with much lower
mortality on the treatment reef. Our results suggest that markedly reduced densities of juveniles on
a reef may lead to a corresponding reduction in mortality, but no effect on growth was evident. However,
part of the apparently higher mortality on the control reefs may instead have been due to emigration
of tagged individuals to other reefs. The difficulties of conducting manipulation experiments in the field
on a highly mobile species are discussed.
The western rock (spiny) lobster, Panulirus cygnus,
occurs on the coast of Western Australia from North
West Cape (lat. 22 °S) to Cape Naturaliste (lat.
34° S). Juveniles (2-5 years old) inhabit the coastal
limestone reefs, primarily at depths of 1-10 m (Chit-
tleborough and Phillips 1975). They remain on the
reefs for several years, apparently with little move-
ment from one area to another (Chittleborough
1974a). Joll and Phillips (1984) found that the juve-
niles feed on a variety of animals and plants asso-
ciated with the seagrass beds surrounding the reefs.
Following the spring molt, larger animals (ages 4-6)
move to the adult habitats in depths of 30-150 m
(Morgan et al. 1982).
Chittleborough (1970) observed that, near the
center of the geographical range, recruitment of
small juveniles appeared to exceed the holding capa-
city of the reef system. He concluded that density-
dependent mortality of the juveniles at such sites
limits their recruitment to the adult stock. He also
observed reduced growth rates of animals at these
'Department of Biology, San Diego State University, San Diego,
CA 92182.
^Division of Fisheries Research, CSIRO Marine Laboratories,
P.O. Box 20, North Beach, 6020, Australia.
^Fisheries Department, W. A. Marine Research Laboratories,
P.O. Box 20. North Beach, 6020, Australia.
sites and considered that the available food re-
sources may be inadequate for the maintenance of
optimum growth at such high densities.
Although there have been many field studies on
the ecology of spiny lobsters (see review by Kanciruk
1980), none have attempted experimental manipu-
lation to elucidate the effects of population density
on growth and survival. This paper considers the
growth and survival of juvenile P. cygnus inhabiting
experimental and control patch-reefs at Seven Mile
Beach, following a manipulation designed to reduce
the density of juveniles of the experimental reef. The
hypothesis to be tested was that high population
densities of juvenile lobsters limit the growth and
survival of the western rock lobsters. Despite its in-
herent practical problems, a manipulation approach
was adopted as the one most likely to yield direct
evidence to evaluate the hypothesis.
Manipulation experiments are best done using rep-
licated experimental areas and both increases and
decreases in the density of the species under con-
sideration (Connell 1974; Underwood 1979). How-
ever, where species (such as lobsters) cannot be
transplanted or enclosed effectively, the only prac-
tical option is to simply reduce densities (Connell
1983). In the case of the study described here, there
are also practical limitations in finding sufficiently
Manuscript accepted June 1988.
FISHERY BULLETIN: VOL. 86, NO. 4, 1988.
773
FISHERY BULLETIN: VOL. 86, NO. 4
similar experimental areas to employ as replicates
and in selecting control reefs which are equivalent
to the premanipulation condition of experimental
reefs. In this experiment juveniles on one reef,
selected as the control site, were left at natural, high
densities, while those on another, selected as the
treatment site, were reduced to approximately 25%
of their original, natural density. A third reef, also
left at its natural high density, was monitored at a
lower level to examine the degree of representative-
ness of the control.
METHODS
Reef Study Sites
Three limestone patch-reefs at Seven Mile Beach,
Western Australia (lat. 29°08'S; long. 114°54'E),
designated as reefs I, III, and V (Fig. 1), were used
as the study sites. These three test reefs are typical
of those at Seven Mile Beach in terms of both their
structure and biota. Observations by divers also in-
dicated that the size structure of P. cygnus on each
reef was similar. The patch-reefs occupy a lagoon
environment between the beach and a limestone bar-
rier reef approximately 400 m offshore. Each patch-
reef is surrounded by a calcareous, sandy substrate
and areas of limestone, which both support exten-
sive seagrass beds of mixed species composition, the
primary feeding areas for juvenile P. cygnus (Cobb
1981; Joll and Phillips 1984). Seagrass species of the
genera Amphibolis, Heterozostera, and Halophila
dominate in these beds. The reefs themselves are
covered by Amphibolis spp. and by a variety of algal
species.
The approximate area of reef III, the treatment
reef, is 0.104 ha, and that of reef V, the main con-
trol reef, is 0.103 ha. Reef I, the secondary control
reef, has an area of approximately 0.071 ha. Reef
I is located approximately 750 m south of reef V,
while reef III is located approximately 60 m direct-
ly west and offshore from reef V (Fig. 1). Adequate
separation of the treatment and control reefs from
each other for the purposes of the experiment was
assumed, based on maximum foraging ranges of up
to 50 m for P. cygnus in the Seven Mile Beach area
reported by Chittleborough (1974a). Water depths
are 2-3 m around reefs I and V and 3-5 m around
reef III. The tops of the reefs are nearly exposed
at low tide.
Single Molt Increments, Annual
and Seasonal Growth
The numbers of juvenile P. cygnus from each reef
which were sexed, measured, and marked or re-
FlGURE 1.— Map of study area at Seven Mile Beach, Western Australia, showing the locations of reefs, I, III, and V (dark shaded) in
relation to other reefs (light shaded).
774
FORD ET AL.: POPULATION OF WESTERN ROCK LOBSTERS
moved in January 1981 at Seven Mile Beach are
given in Table 1 . Animals with carapace lengths (CL)
^40 mm were marked with individually numbered
western rock lobster tags (Chittleborough 1974b).
Animals with CL <40 mm were marked with in-
dividually numbered Floy^ No. FD-68B spaghetti
tags (Davis 1978). Growth data were obtained from
tagged individuals recaptured on the three test reefs
during resampling in February, March, May, Aug-
ust, and September 1981, and January and February
1982. Growth of tagged P. cygnus between recap-
tures over the period January to May was used to
provide data on single molt increments.
Size and Age Structure
Age classes were identified from length-frequency
distributions, as described by Chittleborough (1970)
and Chittleborough and Phillips (1975). From anal-
ysis of the size structure present in January 1981,
juveniles up to 38.0 mm CL were judged to be 2
years of age at that time; 38.1-55.0 mm, 3 years
of age; 55.1-68.0 mm, 4 years of age; and those
>68.1 to be 5 years of age or older. Similarly for
■•Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
January 1982, animals up to 42.5 mm CL were con-
sidered to be 2 years of age at that time; 42.6
mm-55.0 mm, 3 years of age; those 55.1-68.0 mm
to be 4 years of age; and those >68.1 to be 5 years
of age or older.
Population Size, Density, and
Mortality Rates
Estimates of the population size, density, and mor-
tality rates of P. cygnus juveniles on reef I have been
made at Seven Mile Beach since 1970, using recap-
tures from baited traps (Chittleborough 1970; Chit-
tleborough and Phillips 1975). These estimates for
reef I were continued during the period of this study
on reef I and, in addition, estimates were also under-
taken for reefs III and V, using the same traps and
mark-recapture methods. In the present study,
12-13 traps were set around the perimeter of each
test reef to ensure catches directly from the area
of the reef. At each sampling time an array of 12-13
traps was set simultaneously around reefs III and
V in an attempt to reduce attraction by traps of P.
cygnus from one reef to another.
During the initial tagging in January 1981, trap-
ping was conducted on each reef for four consecu-
tive days, while during each subsequent recapture
Table 1 .—Numbers of juvenile Panulirus cygnus examined on three reefs at Seven Mile Beach.
Western Australia.
Sampli
ng dates
Reef
1981
1981
1981
1981
1981
1981
1982
1982
9-10
6-7
26-27
21-22
26-27
26-27
Jan.
Feb.
Mar.
May
Aug.
Sept.
Jan.
Feb.
Reef 1
No. caught
463
447
—
—
362
300
498
419
No. tagged
463
—
—
—
—
—
—
—
No. recaptured
—
103
—
—
98
72
88
^5
11-15
4-7
10-11
5
24-25
19-20
24-25
23-25
Jan.
Feb.
Mar.
May
Aug.
Sept.
Jan.
Feb.
Reef III
No. caught
1,202
785
475
233
221
359
791
862
No. tagged
304
—
—
—
—
—
—
—
No. removed
from reef
898
306
150
—
—
—
—
—
No. recaptured
—
' +
+
54
52
55
204
'13
11-15
4-7
10-11
5
24-25
19-20
24-25
23-25
Jan.
Feb.
Mar.
May
Aug.
Sept.
Jan.
Feb.
Reef V
No. caught
141
682
596
424
599
631
703
696
No. tagged
1,141
—
—
—
—
—
—
—
No. recaptured
—
278
214
154
212
285
140
28
'Additional to those recaptured in January 1982.
2+ = not scored.
775
FISHERY Kl'LLETIN: VOL. 8H, NO. 4
period trapping was conducted for two consecu-
tive days. The single mark-recapture estimate of
Bailey (1951) was used to calculate population
sizes.
Visual Estimates of Population
Size
Direct estimates of the total number of juveniles
present on each test reef were made by two divers
during the day. Initially the surface of each test reef
was mapped, and the map was transferred to ace-
tate writing sheets for use underwater, so that
major features such as crevices, holes, and sections
of ledges could be recognized and searched in a
uniform manner during each census. The two divers
moved slowly around and over the reef, counting
and recording juvenile P. cygnus. Underwater lights
were used to aid in this process. Repeated counts
were often necessary to obtain consistent results for
sections of the reef with large aggregations of juve-
niles. One diver followed approximately 2-3 m
behind the other, and after each section of the reef
was censused, the numbers of juvenile P. cygnus
recorded by the two divers were compared. Only
three observers conducted all of the visual censuses
and, after experience was gained initially, differ-
ences between total counts by any two divers on a
reef usually were less than 5%. Counts by the two
divers were compared for 12 of these censuses,
employing separate Wilcoxon signed-rank tests in
which the data recorded by each diver for a given
section of the reef were paired. The results for all
12 censuses indicated no significant differences in
counts between divers {P > 0.05).
There are two primary sources of error in this
method. One is that few individuals <40 mm CL can
be seen on the surface of the reef or in holes or crev-
ices. The other results from reduced visibility caused
by turbidity and water turbulence. This second prob-
lem was largely avoided by only doing counts when
conditions of turbidity and water turbulence were
favorable.
During January-June 1981 and October 1981-
February 1982, visual density estimates were ob-
tained monthly or bimonthly in reefs III and V.
However, storm conditions and poor visibility
during the remainder of 1981 precluded observa-
tions. Visual density estimates were conducted on
reef I during January 1981 and in January and
February 1982, while in January and February
1983 estimates were conducted on all three test
reefs.
Density Manipulation Experiment
All of the juvenile P. cygnus caught on reefs I and
V (the control reefs) in January 1981 were tagged,
measured, and released. On reef III (the treatment
reef) 1,202 P. cygnus were caught during four con-
secutive days in January 1981 and graded into size
categories (5 mm CL size intervals). This was done
by measuring the animals and holding them in water
in mesh bags suspended from the side of the boat
during the 2-3 hours required for processing.
To reduce the population of lobsters on reef III
by approximately 75%, three out of each four ani-
mals in each size group were removed from the reef
and translocated to another locality out of the Seven
Mile Beach area. Selection was done by removing
the appropriate number of individuals from each size
category blindly to avoid bias. This helped to assure
that the groups of juveniles returned to reef III had
a size frequency and sex ratio similar to those of the
original population. The remaining 304 juveniles
caught from reef III were remeasured, tagged, and
released on that reef. During the next two sampling
periods in February and March 1981, any untagged
P. cygnus caught on reef III were removed (Table
1) to aid in maintaining the density at approximately
25% of its natural level.
Analysis of the Growth Data
Two types of growth data were examined: 1) the
single molt increments of animals with <4 legs miss-
ing (based on the growth of animals recaptured
within four months of a previous capture) (Chittle-
borough 1976), and 2) the average relative growth
rate (Sandland and McGilchrist 1979), which has
been shown to be appropriate for analyses of P.
cygnus growth (Phillips et al. 1983). The data were
classified by reef, sex, age class, duration, and time
of year at liberty and by the number of legs miss-
ing at the time of tagging.
The average relative growth rate data of P. cygnus
were condensed into a three-factor nonorthogonal
experimental design with missing cells. The data
were analyzed by examining differences in growth
between test reefs for each age and recapture inter-
val, using the Wilcoxon rank sum test, and by ex-
amining age, recapture interval effects, and their
interaction in a two-factor nonorthogonal analysis
of variance.
In addition to P. cygnus caught on the reef of
original tagging, 75 individuals (12%) were recap-
tured on both reefs III and V. The data for the 49
776
FORD ET AL.: POPULATION OF WESTERN ROCK LOBSTERS
of these which showed growth between tagging and
recapture were initially analyzed separately.
Chittleborough (1970) found that loss of more
than three legs depressed growth. This was con-
firmed for our data by an analysis comparing the
average relative growth rates of P. cygnus in the
two leg-loss classes (<3 legs missing and >i legs
missing), using a f-test. There was a significant dif-
ference in growth rates between the two leg-loss
classes {P > 0.05). Although this difference is con-
founded with differences in the other factors, the
result was considered as sufficient evidence, when
combined with Chittleborough's findings, to exclude
from further analysis data for P. cygnus with >4 legs
missing.
A further analysis was performed to determine
if the growth rates of males and females differed.
These analyses were done separately for each
reef, age class, and recapture interval, using the
Wilcoxon rank sum test. The results revealed no
significant differences in average relative growth
rates between the sexes (P > 0.05), so the data for
males and females of the same age group were
pooled.
RESULTS
Of the 304 tagged P. cygnus on reef III, 186
(61.2%; 87 males and 99 females) were never recap-
tured, while of the 1,141 animals tagged on reef V,
636 (55.7%; 324 males and 312 females) were never
recaptured. Similarly, of the 463 P. cygnus tagged
in January 1981 on reef I, 307 (66.3%; 162 males
and 145 females) were never recaptured. Compari-
son of the size ranges and size-frequency distribu-
tions of P. cygnus tagged on the three reefs in
January 1981 with those not recaptured (Fig. 2)
indicated that the "losses" were distributed equal-
ly over the full size range and therefore may be
assumed to be random. Ratios of males and fe-
males in these "losses" did not differ significantly
from the malerfemale ratios in the original tagged
population (chi-square test of independence, P >
0.05).
None of the tagged P. cygnus on reef I were recap-
tured on reef III or V. However, 75 (12%) of the
tagged P. cygnus either on reef III or V were subse-
quently recaptured on the other reef, and some were
caught several times on reefs III and V. However,
only four of the animals originally tagged on reef
III were ever recaptured on reef V, suggesting a
general movement of P. cygnus from reef V to reef
III, i.e., inshore to offshore.
Size and Age Structure
The size-frequency distributions of P. q/g^ntts juve-
niles on reef I in January 1981 and January 1982
and that of the juveniles recaptured in January 1982
are shown in Figure 3. Similarly, the size-frequency
distributions of juveniles on reefs III and V in Jan-
uary, May, and September 1981 and in January 1982
are shown in Figure 4. There was change in the com-
position of the population on reef I between Janu-
ary 1981 and 1982, indicating immigration of 2- and
3-yr-old animals into the population of this reef.
Between January and September 1981 there was
an indication of immigration of untagged animals
to both reefs III and V, principally of animals ^3
years of age, while in January 1982 there was also
an obvious immigration of 2-yr-old animals to both
reefs.
Population Size, Density, and
Mortality Rates
The estimates of population size, density, and mor-
tality rates for P. cygnus juveniles >3 years of age
on reefs I, HI, and V are given in Table 2. The
estimate made for the population on reef III in Feb-
ruary 1981 assumes that the 304 tagged P. cygnus
released in January were still within the population
in February.
The population densities of >3-yr-old juveniles on
reef I estimated each January from 1970 to 1982
are shown in Figure 5. The density in January 1981,
at the time the study began, was the highest ever
recorded. The density in January 1982 was also very
high and only exceeded by the levels in January 1974
and January 1981.
The annual mortality coefficient for juveniles aged
Table 2— Population size, population density and mortality-rate
estimates for juvenile Panulirus cygnus >3 years of age on three
test reefs at Seven Mile Beach, Western Australia.
'Reef 1
^Reef III
^Reef V
(0.071 ha)
(0.104 ha)
(0.103 ha)
Date
Population SD
Population SD
Population SD
Jan, 1981
1,990 170
2,192 194
2,644 120
Feb. 1981
— —
1,879 195
— —
Aug. 1981
625 38
725 54
1 ,042 40
Jan. 1982
1 ,273 93
1,951 92
1,841 101
'Annual instantaneous mortality coefficient for those >3 years in 1981
1.655
^Annual Instantaneous mortality coefficient for tfiose >3 years in 1981
1.302.
^Annual instantaneous mortality coefficient for tfiose >3 years in 1981
1.825.
777
FISHERY BULLETIN: VOL, 86, NO. 4
60-,
50-
40-
30-
20-
10-
REEF I
TOTAL NUMBER 463
ALL TAGGED
NOT RECAPTURED
_■ __ ^*- _; ,^ ., . : «i -^ //-, ,.r^ , — : IM m^ rf\ fr\ I \
O'^CMtDCOOCM^CDcnOCNJ
■* CD CO O
h- r^ r^ oD
CO
{I.
LU
h-
<S)
m
O
O
cc
LU
CD
REEF III
TOTAL 1 202
DTAGGED 304
REMOVED
NOT RECAPTURED
u^Lnir)inincr>LninuitnuiLnLnij^ir)inu^ir)inir)iouiu7ini/^inir>Lnin
■^(£)C0OCNJ'^^C0OC\J'<S-U3C0
O 04 t to 00 O
Ln LO If) in lO (O
(O CO
(O
O CM rj- U3 OO O
t-^ r^ r^ r^ r^ co
REEF V
TOTAL 1141
ALL TAGGED
NOT RECAPTURED
Lnuiu^LOuiLnLOir)ur>LnLOuii-nLr)LOLr)u')LnLOLOLOir)Lnu')Lr)LOuo
Tt(bo6c)c\ittcbcdoc\i'<s-cbc6oc>j ■^cbc6oc\J'^uDopoc\J'^u3
CVJCNJCNJCOCOnCDCO-^-^'S-'^'*" ' '
ininLnLni/5tDcDCDcDCDf^r~r^i^
CARAPACE LENGTH (mm)
Figure 2.— Size-frequency distributions of juvenile Panulirus cygnus in January 1981 on three test reefs
(reefs I, III, and V) at Seven Mile Beach, Western Australia and size-frequency distributions of animals
removed or tagged. Size-frequency distributions of tagged animals not subsequently recaptured also are
indicated.
>3 years on reef III was slightly lower than that for
juveniles on reefs I and V. Comparison of the esti-
mates of population size on which these mortality
coefficients were based showed that annual survival
of animals ^3 years old was significantly higher on
reef III than on reef V (chi-square test of indepen-
dence, P < 0.005).
The numbers of juveniles tagged and released on
each test reef in January 1981 are shown in Table
1, while the numbers and percentages of those same
individuals recaptured on the same reef in January
or February 1982 are summarized in Table 3 for
each age group. These data were used, in part, to
provide more specific estimates of age-specific sur-
778
FORD ET AL.: POPI'LATION OF WESTERN ROCK LOBSTERS
vivorship over the 1-}t period, based on assumptions
considered later in this paper. A striking feature of
the data for reef III is the very high percentage of
recaptures, ranging from 97.4 to 100% for 3-, 4-,
and ^5-yr-old individuals, and with a somewhat
lower value of 58.1% for 3-yr-old animals. Percent-
age recaptures for all age groups combined were
71.4% on reef III, as compared with much lower
REEF
JANUARY 1981
ALL TAGGED 463
11
in
t/J
C/D
9
<
8
_J
O
7
LU
6
N
5
OD
4
Z
3
C/)
?
_l
<
1
^
-z.
<
11
II
O
10
^
9
vp
o^
R
^ — '
>
7
o
R
2
IXI
b
Z)
4
O
3
LU
OC
2
REEF I
JANUARY 1982
TOTAL CAUGHT 503
TAG RECAPTURES 93
LnunLnunLnLnLnirjLninirjLOLnLnLnioun
O^ OOOJUD O'^OOCXJtDO'^OOCNJ (X)0 ^
CNjojcMnco^TrM-LriLncDCDUDr^i-^cooo
CARAPACE LENGTH (mm)
Figure 3.— Size-frequency distributions of juvenile Panulirus
cygnus on Reef I at Seven Mile Beach. Western Australia, in
January 1981 and 1983.
values on reef V (14.7%) and reef I (20.1%). Sep-
arate comparisons employing chi-square tests of in-
dependence indicated that the number of juveniles
recaptured were significantly higher on reef III than
on reef V for each of the four age groups {P < 0.005).
Visual Estimates of Population
Size
Estimates of the mean total numbers of P. cygnus
juveniles on the three test reefs, based on visual
sampling by two divers, are summarized in Table 4.
The data for reef I are incomplete but show a
dramatic decrease from 265 on 10 January 1981 to
66 on 19 January 1981. There was also an overall
reduction from 265 in January 1981 to 181 in Janu-
ary 1982 and 173 in January 1983.
The data for reef III (Table 4) indicate that re-
moval of P. cygnus by trapping on 11-15 January
1981 reduced the number of juveniles from 705 pres-
ent on 10 January to about half of that number, with
395 (56%) observed on 19 January and 346 (49%)
on 3 February. Further removal by trapping on 4
and 5 February 1981 reduced the number of juve-
niles to approximately 29% of the original 10 Janu-
ary level (205 observed on 18 February).
From March through June 1981 the mean esti-
mated numbers of juveniles on reef III varied from
189 to 260, representing approximately 27-48 (S
= 36%) of the original, natural population level ob-
served on 10 January 1981 (Table 4). By January
1982, the number of juveniles on reef III had in-
creased to 430, or approximately 61% of the level
observed in January 1981 and by January 1983 the
number had increased to 590 or approximately 84%
of the natural level two years before.
The numbers of juveniles on reef V were rather
more variable, ranging from a low of 72 (25 March
1981) to a high of 780 (October 1981). Of special im-
Table 3— Number and percentage of juvenile Panulirus cygnus tagged in each age class on three reefs at Seven Mile
Beach, Western Australia, in January 1981, and the numbers present and recaptured in January and February 1982.
Reef 1
Reef III
Reef V
Age
1981
1982
1981
1981
1982
1981
1981
1982
1981
class
No.
Total no.
Recaptures
No.
Total no.
Recaptures
No.
Total no.
Recaptures
(Yr)
tagged
caught
No.
%
tagged
caught
No.
%
tagged
caught
No.
12
o/o
2
117
260
12
10.3
38
227
37
97.4
192
338
6.3
3
245
196
62
24.4
203
375
118
58.1
777
267
102
13.1
4
77
36
14
18.2
57
109
56
98.2
149
112
45
30.2
>5
15
6
5
33.3
6
92
6
100.0
23
14
9
39.1
Total
463
503
93
20.1
304
804
217
71.4
1,141
731
168
14.7
'In January 1981
779
FISHERY KULLKTIN: VOL. 8fi, NO. 4
REEF
JANUARY 1981
D TOTAL CAUGHT 1202
■ NUMBER TAGGED 304
CD ■^ c6c\jcbo'^o6c\i(£>0'^ooc\icbo
CD
C/D
O
LU
N
CO
C/5
_l
<
REEF III
MAY 1981
TOTAL CAUGHT 233
TAG RECAPTURES 54
— I — I — — I — I — ^ — ^ — ^ — I — '
TtcpcviiboTtcgcNJcd
O ■^ OO CNJ (D o
C\J CM C\J 00 CO Tj-
o
REEF III
SEPTEMBER 1981
TOTAL CAUGHT 359
TAG RECAPTURES 55
IT) in If) in IT) in
Tf OOCVJtOOTt OOCvJtOO'^ OOCNJCOO'*
c\ic\jcocoT}-TtTtinincotD(Ot-^r---oooo
Figure 4.— Size-frequency distributions of juvenile
Panulirus cygnus on reefs III and V at Seven Mile
REEF III
JANUARY 1982
TOTAL CAUGHT 804
TAG RECAPTURES 217
ininininm ininininmininminin
o •^ oo <M CO
C\J CM CVJ CO CD
o ■*
00C\ICDO'«J-000gC0O'*
tinmuDtDcDr-^t^oooo
CARAPACE LENGTH (mm)
780
FORD ET AL.: POPLLATION OF WESTERN ROCK LOBSTERS
REEF V
JANUARY 1981
ALL TAGGED 1141
•*o6c\iu3C3'^a)c\icbo'*o6oJcbd
■^ m HI (£)(£) io r-~ r^ao
c\j c\j n n Tj-
5
en
en
<
o
LU
N
C/5
<
REEF V
MAY 1981
TOTAL CAUGHT 424
TAG RECAPTURES 154
-■"Tooc^iOQTfobrvjuDCJ
s
in ld in LO
00 oj lb o
CD r-- r~. 00
o
-^
>
o
z
LU
Z)
o
LU
DC
LL
10
9
8
7
6
5-1
y^^
X
REEF V
SEPTEMBER 1981
TOTAL CAUGHT 631
I TAG RECAPTURES 235
i^-J^^
irjiriifjinioLnunLni-nLOLnLnLnLnuiLniD
OT3-o6cNitbO'<4-COC\IUDO'<TCOC\Jcbod
CNJCMOJCOCO'^TtTtLniOCDCOCDr-^I^OOCD
REEF V
JANUARY 1982
TOTAL CAUGHT 731
TAG RECAPTURES 168
Figure 4— Continued— Beach. Western Australia,
during the period January 1981 through January
1982.
inLniouniDLOLnLniounLnLOiDLnLn
pTrgDc\i(£JOTtcoc\i<£)C3Tj-odc\jcD
CMCMCNjcDcortTj-Ti-LnuncDtDcDr-^r-
CARAPACE LENGTH (mm)
781
FISHERY BULLKTIN: VOL. 86. NO. 4
portance was the 34% decline in numbers between
10 January and 18 February on reef V. By mid-
February 1981 the estimated population on reef V
was below that on reef III. The mean number of
animals for reef V in February 1982 (401) was 71%
of that observed in January 1981 (563), although in
October, November, and December 1981 it exceeded
the January 1981 level. In January 1983, the number
of juveniles (537) was essentially the same as that
observed two years before (Table 4).
Figure 5.— Densities of 3-yr-old juvenile
Panulirus cygnus in January on reef I at
Seven Mile Beach, Western Australia, for the
period 1970 through 1982. Mean and 95%
confidence limits are shown. Data for
1970-80 from Morgan et al. (1982).
Table 4. — Mean total numbers of juvenile
Panulirus cygnus on three test reefs at Seven
Mile Beach, Western Australia, based on visual
estimates made by two observers. — = no
visual estimate made.
Date
Reef 1
Reef III
Reef V
10 January 1981
265
705
563
19 January 1981
66
395
286
3 February 1981
—
346
235
18 February 1981
—
205
191
9, 13 March 1981
—
335
227
25 March 1981
—
189
72
23 April 1981
—
233
220
15, 17 June 1981
—
260
451
15 October 1981
—
374
780
19 November 1981
—
424
680
10 December 1981
—
321
647
29 January 1982
181
430
—
23 February 1982
42
474
401
11 January 1983
173
590
537
16 February 1983
121
521
501
29r
28-
27-
26-
CO
JC
25-
n
z
24-
CO
23-
ULI
22-
1-
?1 -
U)
CO
?n-
o
_l
iy-
^
18-
n
O
1/ -
cc
16-
II
O
1b-
^
14-
CO
13-
z
1?-
LL!
Q
11 ■
10-
9-
1
\ h
■f
1970 71 72 73 74 75 76 77 78 79 80 81 82
YEAR
Growth
Single Molt Increments
Mean single molt increments of male and female
P. cygnus on reefs III and V, the treatment and con-
trol reefs, (Table 5) showed no significant difference
within either the 2-, 3-, or 4-yr-old age-classes (^test,
P > 0.05). Only four animals estimated to be ^5 years
of age were recaptured and these were all from reef
V. After pooling the molt increment data for the two
sexes there were no significant differences between
the mean single molt increments of 2-, 3-, or 4-yr-
old animals on reefs III and V (^test, P > 0.05).
Table 5.— Mean molt increments in carapace length (mm) for sex age-class groups
of juvenile Panulirus cygnus from three test reefs at Seven Mile Beach, Western
Australia during January-May 1981. Data are mean BE (n).
Age
class
Reef!
Reef III
Reef V
Males
Females
Males
Females
Males
Females
2
—
—
2.58(0.40)
(5)
1.84(0.11)
(5)
1 .75(0.28)
(4)
2.41(0.20)
(10)
3
2.85(0.56)
(6)
1.74(0.21)
(9)
2.01(0.33)
(16)
1.76(0.15)
(16)
2.27(0.14)
(76)
2.17(0.26)
(60)
4
—
—
2.75(0.61)
(4)
1.94(0.26)
(5)
3.17(0.26)
(16)
2.88(0.19)
(18)
782
FORD ET AL.: POPULATION OF WESTERN ROCK LOBSTERS
The data from reef I, the secondary control reef,
were only sufficient for an examination of the single
molt increments of 3-yr-olds. Within this age class
there was no significant difference in mean molt in-
crement between the two sexes and the data for the
whole age class were pooled. There were no signif-
icant differences between the mean single molt in-
crement for 3-yr-olds from reef I and the mean
single molt increment for this age class on either
reefs III or V (t-test, P > 0.05).
Annual and Seasonal Growth
The mark-recapture data from reefs I, III, and V
are considered within four time periods: From Janu-
ary 1981 to May 1981 (the "summer-autumn"
period), from May 1981 to September 1981 (the
"winter" period), from September 1981 to Febru-
ary 1982 (the "spring-summer" period) and from
January 1981 to January-February 1982 (the "an-
nual" growth period). As described in Methods sec-
tion above, average relative growth rates (ARGRs)
of males and females were not significantly differ-
ent and therefore the data were pooled. Because of
the small number of ^5-yr-old lobsters, their growth
data were combined with the data for the 4-yr-old
individuals. The ARGRs of the different age groups
from all three reefs over each of the four time
periods are given in Figure 6.
Comparison of growth data for the test reefs for
each age group and recapture interval were made
using Wilcoxon rank sum tests. The primary pur-
pose was to evaluate whether growth data for the
three reefs could be pooled for later analysis. The
comparisons are of interest in their own right, but
caution is required in interpreting some of the dif-
ferences established because of small sample sizes
in some cells.
The results showed no significant differences in
ARGRs of any age group between reefs III and V
(P > 0.05), the treatment and the principal control
reefs respectively in the density manipulation ex-
periment. There were significant differences in
ARGRs between reefs I and III for individuals 3
years of age in January-May 1981, May-September
1981, and January 1981 -January 1982 and for in-
dividuals >4 years of age in January-May 1981.
There also were significant differences in ARGRs
between reefs I and V for individuals 3 years of age
in January-May 1981 and May-September 1981,
and for individuals >A years of age in January-May
1981 (P < 0.05). The significant differences all
showed consistently higher ARGRs on reef I than
on reefs III and V.
>
2
0)
k-
(U
Ol
(0
4)
>
10
c
15
4)
JAN -MAY 81
Reef I III V
0.20 r
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
_Ll
I I I I I I I I I I I
MAY -SEPT. 81
I III V
i.
1 1 1 1 1 1 1
SEPT-JAN. 82 JAN. 81 -JAN. 82
I III V I III V
I I I ] I
I I I I I I I
2 3?4 2 3?4 2 3*4 2 3*4 2 3?4 2 3*4 23*4 2 3*4 2 3*4 2 3*4 2 3*4 2 3*4
Age ( Years )
Figure 6.— Average relative growth rates (ARGRs) of juvenile Panulirus cygmis
on test reefs at Seven Mile Beach, Western Australia.
783
FISHERY BULLKTIN: VOL. 86, NO, 4
The data for both reefs III and V were then
pooled. The data for 49 P. cygnus which were recap-
tured on reefs III and V were included in this set
(Fig. 7). Comparisons of these combined data were
made with the data for reef I, and the results in-
dicate that all of the significant differences in
growth described above were maintained except
that for 3-yr-old individuals in January 1981 -Janu-
ary 1982. The lack of a consistent significant differ-
ence between reefs at all ages and recapture inter-
vals suggests that the effect is not a simple response
to a superior environment, but rather that interac-
tions between reef and age and between reef and
recapture interval are present.
DISCUSSION
The density manipulation described in this paper
is an attempt to use an ecological field experiment
in a nonbenign, sub tidal habitat to study the popu-
lation processes of a spiny lobster. Practical limita-
tions within this environment related to wave effects
on drifting plant material and the need to enclose
enough area to adequately provide for the foraging
range of juvenile P. cygnus precluded the use of
large enclosures which would have effectively pre-
vented migration. Also, because of the potential
mobility of P. cygnus, it was not possible to increase
and maintain the density of juveniles above natural
levels on a reef without such an enclosure. There-
fore, only an experimental reduction in density was
attempted:
1. Data from the visual estimates show that we
were successful in reducing the number of juve-
nile P. cygnus on reef III to approximately 30%
of the original, natural level by removing animals
in January, February, and March 1981. Follow-
ing the last removal, the estimated numbers of
animals varied, but showed a slow increase over
the 1-yr period of the manipulation experiment,
with a mean of 36% of the original January level
during March through June 1981 and a mean of
55% of the original level during October 1981
through January 1982. However, the number of
juveniles on both reefs I and V, the control reefs,
also declined during this period, and the popula-
tion on reef V was estimated to be below that
of reef III on 18 February 1981. Therefore, it is
difficult to separate the effects of the systema-
tic removals from a general decline in numbers
indicated by what was observed on reefs I and V.
The size and age structures and the sex ratios
of P. cygnus juveniles on reefs I, III, and V in
January 1981 were very similar. Size and age
structures and the sex ratios of these juveniles
0.14 r-
JAN - MAY '81 MAY - SEPT '81 SEPT - JAN '82 JAN '81 - JAN '82
2
0.12
w
£
**
3
o
0.10
S3)
V
>
♦-
n
0.08
V
Ol
(0
0)
0.06
>
re
c
to
V
0.04
s
0.02 -
I
I
I
I
I
I
I
3 *4 2 3 *4
Age ( years )
3 ?4
Figure 7.— Average relative growth rates (ARGRs) of juvenile Panulinis
cygnus on test reefs III and V (combined) at Seven Mile Beach, Western
Australia.
784
FORD ET AL.: POPl'LATION OF WESTERN ROCK LOBSTERS
also were very similar among the reefs at all
subsequent sampling times, suggesting that the
drastic reduction in the density of P. cygnus on
reef III, and the method of reconstituting the
population, did not have an evident effect on
these characteristics over the 1-yr period of the
experiment.
2. The population sizes of ^3-yr-old P. cygnus juve-
niles, estimated from catches with baited traps,
were significantly lower on reef III than on reef
V in January 1981 and August 1981, but were
similar (i.e., not significantly different) in Janu-
ary 1982. However, these data, and the estimates
of mortality derived from them, are subject to
the usual problems associated with mark-recap-
ture techniques. Phillips (unpubl. data) has found
in more recent studies that the baited traps are
capable of attracting juveniles from over a wide
area and hence the population size and density
estimates applied to individual reefs probably are
inaccurate. However, the trends in the popula-
tion estimates from the mark-recapture data are
supported by the visual estimates of population
size, indicating that they reflect what was actual-
ly happening on these two reefs.
3. Clearly, the visual estimates provide a more spe-
cific set of information about population levels
of juveniles living on the reef. The method also
allows direct estimates of numbers over short
time intervals and with minimal disturbance of
the western rock lobsters. However, although
there is no doubt about the drastic decline in the
numbers of tagged lobsters on reefs I and V after
the period of initial tagging, it is not possible to
determine if this was as a result of tagging mor-
tality, an emigration as a response to handling
and tagging or part of the normal behavior
pattern.
Despite the attempts to select directly comparable
reefs for the experiment, it is possible that reefs III
and V do provide different environments for the
resident P. cygnus. The movement of 71 individuals
from reef V to reef III may be part of a typical move-
ment from shallower to deeper reefs. It seems
unlikely that water depth, per se, is the primary fac-
tor involved, because the difference in depth be-
tween the two reefs is no more than 2 m. The reefs
used by Chittleborough (1970) at Garden Island, on
which he found P. cygnus juveniles remained for
several years with little movement even from one
part of the reef to another, were similar in depth
to reef III.
Observations during sampling indicated that on
reefs I and V, which are both located close to the
beach, relatively large amounts of plant detritus
build up around the base of the reef and under
ledges and that turbidity of the water is sometimes
quite high. These effects also occur on reef III, but
are less pronounced. This suggests that less favor-
able conditions on the shallower, inshore patch-
reefs may cause some P. cygnus to seek reefs
slightly farther offshore which have more suitable
conditions. It also may help to explain the higher
variability in numbers of juveniles from the visual
estimates made on reef V compared with reef III.
Some individuals may temporarily emigrate from
areas such as reef V during periods of adverse
conditions.
Survival and mortality data, which were obtained
from both the mark-recapture estimates of popula-
tion size and the numbers of tagged animals recap-
tured on the same reef, indicate that survival of P.
cygnus juveniles was significantly higher on reef III
than on reef V over the 1-yr period of the experi-
ment. This was evident for individuals of all four age
groups. It suggests that the experimental reduction
in numbers of juveniles on reef III, resulting in
relatively low population densities (29-61% of the
original, natural number) during the 1-yr experi-
mental period, led to significantly higher survival
than on reef V, where P. cygnus juvenOes were pres-
ent at natural density levels. One explanation for
this is that reduced densities of juveniles on a reef
may lead to a corresponding reduction in predation
mortality and adverse interspecific effects of crowd-
ing. An alternative explanation for the very high
survival of P. cygnus juwenWes estimated on reef III,
relative to reefs I and V, is that the tagged juve-
niles on reef III remained for the entire year while
many of those tagged on reef I and V emigrated to
other reefs.
These mortality data are subject to several sources
of error, including the basic problems associated
with mark-recapture sampling to obtain population
estimates (Bailey 1951). The age-specific estimates
of survival obtained from comparisons of numbers
of tagged individuals released on each reef in Janu-
ary 1981 and recaptured there in January or Feb-
ruary 1982 requires a major assumption. It is that
all of the tagged individuals not recaptured on a
particular reef after one year have died. This un-
doubtedly is not the case because some of those in-
dividuals probably moved to other reefs at Seven
Mile Beach after their release. To the extent that
this occurred, the survivorship estimates are low.
785
FISHERY BULLRTIN: VOL. 86. NO. 4
No quantitative information is available to correct
the estimates for this effect.
The analyses of the growth data, including both
the single molt increments and the ARGRs, clearly
indicated high variability in growth rates between
age groups, and that site (reef) and season varia-
tions (recapture interval) were compounding factors.
Estimates of growth made during the 1-yr period
of the study were similar to those obtained on reef
I in previous studies at Seven Mile Beach (Chittle-
borough 1970, 1975, 1976; Chittleborough and
Phillips 1975; Joll and Phillips 1984). However,
despite the fact that the population densities in 1981
were the highest ever recorded, the molt increments
of the 3-yr-old P. cygnus were significantly higher
than at the low densities in 1971-74.
Comparisons of the growth data for P. cygnus
juveniles showed no significant differences in
growth within any age group between reefs I, III,
and V. The reduced densities of juveniles on reef
III had no apparent effect on either their overall
growth rates or their molt increments. This suggests
that the food resources on the surrounding seagrass
beds may not limit growth within the range of P.
cygnus densities present on these reefs during the
field experiments. However, other factors may be
involved. The foraging ranges of juveniles on reef
III may overlap those animals from reef V and other
nearby patch-reefs. More recent acoustic tracking
studies by Jernakoff (unpubl. data) suggest that this
is probably the case. If so, then reducing the den-
sity of P. cygnus on reef III might not produce a
significant increase in their growth rates, because
they could still be sharing their food resources with
animals from nearby reefs.
The best tests of hypotheses about the effects of
limited resources are those where the densities of
P. cygnus are experimentally manipulated in repli-
cated experimental areas and which incorporate ap-
propriate controls (Connell 1974; Underwood 1979).
This attempt has highlighted a number of problems.
Nevertheless, it may be useful to conduct modified
manipulation experiments of this kind in the futiire.
Obviously, one of the problems with the present ex-
periment was the lack of replication, and replication
should be incorporated in the design of any future
experiments. This was not possible in the present
study because of high time and manpower require-
ments associated with the use of trapping and mark-
recapture techniques. Furthermore, our observa-
tions indicate that it will be extremely difficult to
find a series of patch-reefs similar enough in size,
structure, and other features to serve as true rep-
licates. As has been shown in this study, even small
differences in water depth, or as yet unidentified
characteristics, make the selection of reefs as equi-
valents very difficult. Selection of such reefs also
will not be easy because often several reefs are
within the known foraging range of the juveniles.
Evaluation and refinement of the visual estima-
tion technique also will be necessary before further
manipulation experiments are undertaken, as the
usefulness of mark-recapture techniques is dubious.
Without the development of such a refined method
which would permit rapid and frequent estimates
of population size, the effect of subsequent changes
in population levels cannot be properly monitored.
Without such a method it also would not be possi-
ble to determine the extent to which migrations of
P. cygnus juveniles to and from the reefs are induced
by the use of the baited traps or by handling and
other disturbances during the mark-recapture
process.
ACKNOWLEDGMENTS
We thank David Wright, David Evans, Simon
Braine, and Leo Olsen of the CSIRO Division of
Fisheries Research, Marmion, Western Australia,
for their assistance with field and laboratory work.
We also thank Frank W. Reneke of San Diego State
University and R. Sandland of the CSIRO Division
of Mathematics and Statistics for their assistance
with computer data summaries and statistical anal-
yses. This study was sponsored by the National Sci-
ence Foundation U.S. -Australia Cooperative Science
Program through NSF grant INT 7927203 to
Richard F. Ford and by the CSIRO Division of
Fisheries Research.
LITERATURE CITED
Bailey, N. T. J.
1951. On estimating the size of mobile populations from
recapture data. Biometrika. 38:293-306.
Chittleborough, R. G.
1970. Studies on recruitment in the Western Australian rock
lobster Panulirus longipes cygnus George: density and
natural mortality of juveniles. Aust. J. Mar. Freshwater
Res. 21:131-148.
1974a. Home range, homing and dominance in juvenile west-
ern rock lobsters. Aust. J. Mar. Freshwater Res. 25:227-
234.
1974b. Development of a tag for the western rock lobster.
Rep. Div. Fish. Oceanogr. CSIRO Aust. No. 56, 19 p.
1975. Environmental factors affecting growth and survival
of juvenile western rock lobsters Panulirus longipes (Milne-
Edwards). Aust. J. Mar. Freshwater Res. 26:117-196.
1976. Growth of juvenile Panulirus longipes cygnus George
786
FORD ET AL.: POPULATION OF WESTERN ROCK LOBSTERS
on coastal reefs compared with those reared under optimal
environmental conditions. Aust. J. Mar. Freshwater Res.
27:279-295.
Chittleborough, R. G., and B. F. Phillips.
1975. Fluctuations of year-class strength and recruitment in
the western rock lobster Panuiirus long-ipes (Milne-Edwards.
Aust. J. Mar. Freshwater Res. 26:317-328.
Cobb, J. S.
1981. Behavior of the Western Australian spiny lobster,
Panulirus cygnus George, in the field and the laboratory.
Aust. J. Mar. Freshwater Res. 32:399-409.
CONNELL, J. H.
1974. Field experiments in marine ecology. In R. Mariscal
(editor), Experimental marine biology, p. 21-54. Acad.
Press, N.Y.
1983. On the prevalence and relative importance of inter-
specific competition: evidence from field experiments. Am.
Nat. 122:661-696.
Davis, G. E.
1978. Field evaluation of a tag for juvenile spiny lobsters,
Panulirus argus. Trans. Am. Fish. Soc. 107:100-103.
JoLL, L. M., AND B. F. Phillips.
1984. Natural diet and growth of juvenile western rock
lobsters Panulirus cygnus George. J. Exp. Mar. Biol. Ecol.
75:145-169.
Kanciruk, p.
1980. Ecology of juvenile and adult Palinuridae (spiny lob-
sters). In J. S. Cobb and B. F. Phillips (editors), The biology
and management of lobsters, Vol. II, p. 59-92. Acad. Press,
N.Y.
Morgan, G. R., B. F. Phillips, and L. M. Joll.
1982. Stock recruitment relationships in Panulirus cygnus,
the commercial rock (spiny) lobster of Western Australia.
Fish. Bull., U.S. 80:475-486.
Phillips, B. F., L. M. Joll, R. L. Sandland, and D. Wright.
1983. Longevity, reproductive condition and growth of the
western rock lobster, Panulirus cygnus George, in aquaria.
Aust. J. Mar. Freshwater Res. 34:419-429.
Sandland, R. L., and C. A. McGilchrist.
1979. Stochastic growth curve analysis. Biometrics 35:255-
271.
Underwood, A. J.
1979. The ecology of intertidal gastropods. Adv. Mar. Biol.
16:111-210.
787
A LONG-TERM STUDY ON THE BEHAVIOR AND SURVIVAL OF
EARLY JUVENILE AMERICAN LOBSTER, HOMARUS AMERICANUS, IN
THREE NATURALISTIC SUBSTRATES: EELGRASS, MUD, AND ROCKS
Diana E. Barshawi and Donald R. Bryant-Rich^
ABSTRACT
An 8-month study on the behavior, growth, and survival of early juvenile American lobsters, HoTnarus
americanus, was conducted in three different naturalistic habitats of mud, rocks with algae, and eelgrass.
Fifteen narrow aquaria (10 cm wide) allowed visual observations of American lobster's activities in five
replicates of each of the three habitats. After a 3-month acclimation period to establish "natural" ben-
thic communities which entered through the water supply, three stage IV American lobsters were intro-
duced into each aquarium. Observations were made on the settling, burrowing, activity, and feeding
behavior of these lobsters.
American lobsters in eelgrass and rock habitats settled into the substrate more quickly, had burrows
a greater percent of the time, and spent less time repairing their burrows than lobsters in mud habitats.
The lobsters in eelgrass had a lower mortality rate than lobsters in either rocks or mud. None of the
lobsters in any substrate were observed foraging for food outside of their burrows. However, the behavior
of these American lobsters indicated that they were able to capture plankton drawn into their burrows
by pleopod fanning. Six lobsters molted during the coldest part of the year when the water temperature
was approximately 1° to 2°C.
Stage IV of the American lobster, Homarus ameri-
canus, is best described as transitional between
larval and juvenile (Phillips et al. 1980). During this
stage major behavioral changes take place, which
coincide with the morphological changes occurring
in the molt. These behavioral and morphological
changes cause the stage IV lobsters to descend from
the upper layers of the water column to the bottom
where they build a burrow (Botero and Atema 1982;
Ennis 1975).
Knowledge of the American lobster's behavior
from the onset of settlement until they reach a size
of approximately 20 mm in carapace length (CL)
remains Hmited because juveniles of this size
range have been found in the field only sporadi-
cally.
Several laboratory experiments sought to deter-
mine the substrate preferences of stage IV Ameri-
can lobsters. Howard and Bennett (1979) found that
lobsters [H. gammarus) generally choose the largest
size of gravel provided (approximately 20 mm in
diameter), because larger rocks have more available
'Boston University Marine Program, Marine Biological Labora-
tory, Woods Hole, MA 02543; present address: Marine Field Sta-
tion, Rutgers University, Great Bay Blvd., Tuckerton, NJ 08087.
^Boston University Marine Program, Marine Biological Labora-
tory, Woods Hole, MA 02543.
space between them for burrows. If given a choice
between a gravel substrate or a silt/clay substrate,
American lobsters prefer the gravel (Pottle and
Elner 1982). In choice tests, stage IV American
lobsters preferred rocks with macroalgae, followed
by, in order of decreasing preference, mud, rocks
on sand, and sand. If not afforded a choice, the
lobsters settled most quickly on the rocks with
macroalgae, followed by rocks on sand, mud, and
sand (Botero and Atema 1982).
MacKay (1926) recorded observations on the
lobsters' ability to burrow in mud. Subsequently
Cobb (1971), Berrill and Stewart (1973), and Botero
and Atema (1982) have described the methods
by which juvenile American lobsters make burrows
in both mud and rocky substrates. No observa-
tions have been made on American lobsters burrow-
ing into other substrates, such as eelgrass or
peat.
Cobb et al. (1983) followed stage IV H. ameri-
canus for short periods of time following their re-
lease into the field. They observed behavior which
may indicate that American lobsters test different
substrates and continue moving if they are on un-
satisfactory substrates such as sand or mud; how-
ever, only two lobsters were actually seen reject-
ing a substrate.
Manuscript accepted July 1988.
FISHERY BULLETIN: VOL. 86, NO. 4, 1988.
789
FISHERY BULLETIN: VOL. 86. NO. 4
None of the previous studies lasted for more than
a few days, and long-term behavioral observations
have never been recorded on early juvenile Ameri-
can lobsters. There are virtually no field data on sim-
ple life history parameters such as their preferred
substrate, growth rate, diet, and behavior. It is
unwise to proceed with experimental laboratory
studies on an organism without having a descrip-
tive life history background to provide context. Both
Cobb (1987) and Fogarty (1987) recognized the need
for more studies on the behavior and ecology of the
postsettled prerecruits used in this study. At pres-
ent it is difficult to gather such information in the
field. However, this study was designed in order to
provide such a background by carefully creating
naturalistic habitats in the laboratory. We present
quantified behavioral observations, survival, and
growth of early juvenile American lobsters in three
different substrates: mud, rocks, and eelgrass, over
an 8-mo period.
MATERIALS AND METHODS
Fifteen "ant farm" aquaria (45 cm deep x 30 cm
long X 10 cm wide) were constructed to optimize
our ability to observe the American lobsters inside
their burrows. Five of the aquaria were two-thirds
filled with cohesive mud (particle size <0.06 mm)
collected from mud flats in the Woods Hole, MA
area. Five of the aquaria were two-thirds filled with
rocks, collected from subtidal areas in such a way
that a representative distribution of rock sizes
was obtained (0.1-20 mm diameter). Some of the
rocks in each tank had macroalgae Coleus (sp.)
and Fucus (sp.), or both growing on them. Eelgrass
collected from local eelgrass beds was placed in the
last five aquaria; less substrate was used in these
latter tanks so that the eelgrass leaves had room
to grow.
The aquaria were randomly distributed in a sys-
tem which provided running, unfiltered seawater at
ambient temperatures seasonally ranging from
23° to 0°C. Plankton were always visible; also the
three habitats occasionally had plankton blooms,
during which algae and zooplankton were plentiful.
The tanks were maintained on an ambient light/dark
regime with a light intense enough to keep the
eelgrass alive. Removable, opaque, black plastic was
placed around each aquarium to the level of the
substrate to ensure that the lobster burrows were
dark. To establish "natural" benthic communities,
the tanks acclimated from 20 July until 16 October
1982, before beginning the experiment.
Stage IV American lobster siblings from the
hatchery at St. Andrews, New Brunswick, Canada
were introduced, one per day into each aquarium
for 3 consecutive days. Thus, the total number of
lobsters at the start of the experiment was 45, 15
per treatment. Observations were recorded continu-
ously for the first half hour after each introduction
and then for the following 1.5 hours; observations
were recorded by scanning (taking an instantanious
reading of the lobsters' behavior) every 10 minutes.
Observations were made of the following: 1) loca-
tion of the animal in the aquarium, 2) motion (walk-
ing, swimming, or resting), 3) burrowing activity
(pleopod fanning, bulldozing, or digging), and 4)
shape, size, and location of final burrow. This pro-
cedure was similar to that used in the substrate
choice tests done by Atema et al. (1982).
During the first introduction of American lobsters
into several of the tanks, mud crabs, Neopanope
sayi, immediately consumed them. The mud crabs
were subsequently removed and new American lobs-
ters were placed into these tanks.
After the American lobsters had been introduced
into each of the 15 tanks, long-term observations
began of each lobster in each tank at intervals rang-
ing from daily to twice per week. The observation
periods were at different times during the day with
5.1% during the dark period, although it was dif-
ficult to see the lobsters in low light because of the
cryptic nature of some burrows. There was a total
of 195 observations periods. Each lobster that was
visible was watched for at least one minute; if the
lobster was active, observations lasted until the ac-
tivity ended. However, for the quantitative analysis
of lobster activity only the first minute of observa-
tions were used. A total of 495 hours of observa-
tions were made averaging 11 hours per individual
lobster. For each lobster we recorded 1) the loca-
tion of the lobster in relation to its burrow, 2)
whether the lobster had molted, 3) the lobster's ac-
tivity, and 4) the shape of the burrow (with a quick
sketch). The activities observed are described in
Table 1.
The experiment lasted approximately eight
months, from 21 October 1983 to 1 July 1984. The
lobsters were not fed during that time; we assumed
they would find food from the communities in which
they lived. At the end of the experiment, the sur-
viving lobsters were weighed and their carapace
length was measured. Additionally, the sediment in
each tank was sieved through a 1 mm screen, and
all organisms were collected, weighed, and identified
to the genus or species level.
790
BARSHAW and BRYANT-RICH: EARLY JUVENILE AMERICAN LOBSTER
Table 1 .—Description of the different activities observed through-
out the experiment.
Activity
Description
Rest No movement for at least 30 seconds. Groom-
ing was not considered movement, and was not
recorded separately from rest.
Pleopod fan Movement of the pleopods; if the fanning was
(PPF) being used to repair the burrows, i.e., sediment
was being moved, then the activity was recorded
as burrow repair.
Burrow repair Any activity which caused sediment to be
(BR) moved, including bulldozing (pushing sediment
toward with the claws spread apart), pleopod
fanning, and digging (loosening sediment by
pushing claws into it).
Investigate Standing at the entrance of the burrow with
(INVEST) antennules out and antennae flicking.
Feed Eating anything larger than 1 mm. Activity that
looked like filter feeding was not included in this
category (it was part of the pleopod fan). It is
discussed in the text.
Walk Walking on the sediment. Does not include
"walking" in the burrow.
Swim Swimming in the water column.
RESULTS
Burrowing
The American lobsters in the eelgrass and rock
substrates started burrow construction more quickly
than the ones in the mud substrate (1 way ANOVA,
Newman-Keuls test, P < 0.05). There was no sig-
nificant difference in the time to initial burrowing
between lobsters in eelgrass and lobsters in rock
substrates (Table 2A).
American lobsters used the same methods to make
burrows in eelgrass as in mud and rocks. They
typically started at the base of an eelgrass plant and
then established a burrow under the rhizomes by
pleopod fanning and bulldozing. The burrows usually
had two openings although burrows were seen with
from one to six openings. These openings were
smaller and more difficult to see than similar open-
ings in mud or rock substrates. Although lobsters
in all substrates had burrows for the majority of the
observations, because their burrow had collapsed,
the lobsters in the mud substrate were without a
burrow for a greater percent of the observations
than the lobsters in the eelgrass or rock substrates
(arcsine transformation, 1 way ANOVA, Newman-
Keuls test, P < 0.05, Table 2B). For this analysis
the lobster had to be visible; if neither the lobster
nor its burrow were visible during a given observa-
tion period, that observation was excluded from the
analysis.
Activity
American lobsters were not seen to forage out-
side of their burrows. If a lobster had a burrow, it
was never seen outside of that burrow in any of the
treatments during the entire experiment. During
the day periods, these lobsters were seen in their
burrows 1,503 times, and outside of their burrows
0 times. Therefore, by using sampling theory, one
can calculate that the lobsters were spending at least
99.8% of their time during light periods in their bur-
rows (binomial distribution, P = 0.05). During the
night periods lobsters were seen in their burrows
103 times, outside of their burrows 0 times. There-
fore, the lobsters were spending at least 97.0% of
the time in their burrows during the dark (binomial
distribution, P = 0.05). The difference between the
night and day percentages is a function only of the
greater number of observations made during the
day.
The cumulative times that the American lobsters
spent at various activities were influenced by sub-
Table 2. — (A) The average time in minutes that it took each lobster in the
eelgrass, rock, and mud treatments to start construction of their burrow
(eelgrass vs. mud and rocks vs. mud, P < 0.05). (B) The percent of observa-
tions throughout the experiment during which the lobsters in each substrate
did not have a burrow. N varied from 160 to 68, depending on how many
lobsters were visible (eelgrass vs. mud and rocks vs. mud, P < 0.05). (C)
The average weight, in grams, and the carapace length (CL), in mm, of the
lobsters in each treatment at the end of the experiment.
A
Time to burrow
B
No burrow
C
Weight and CL
Eelgrass
Rock
Mud
7.92 + 2.02
11.92 -1- 3.22
49.17 + 14.08
4.70 + 1.7
3.82 + 1.22
12.2 + 3.7
4.95 ± 0.95 18.15 ± 1.25
3.18 ± 0.16 15.22 ± 1.34
3.22 ± 1.23 15.85 ± 1.92
791
FISHERY BULLETIN: VOL. 86, NO. 4
CO
z
g
<
>
LU
CO
ffl
O
u.
O
UJ
o
MX
a.
60
50
40
30
20
10
0
I
■:i
\
0 !
REST
B EG
0 ROCK
H MUD
X
rp^/i
INVEST
WALK
FEED
ACTIVITIES
Figure 1 . — The percent of observations in which the American lobsters were engaged in each of the listed activities. BR: eelgrass
vs. mud, P < 0.001. Invest: eelgrass vs. mud and rock vs. mud, P < 0.05. A'^ varied from 160 to 68.
strate (Fig. 1). The lobsters in the mud substrate
spent a significantly greater percent of their time
repairing their burrows than the lobsters in either
eelgrass or rocks (arcsine transformation, 1 way
ANOVA, Newman-Keuls test, P < 0.001). Based on
percent of observations, the lobsters in mud spent
significantly less time investigating than did the
lobsters in rocks (tests as above, P < 0.05). There
was no significant difference, however, between the
percent of observations spent investigating in the
mud vs. the eelgrass, or in the percent of observa-
tions spent investigating in the eelgrass vs. the
rocks.
Based on the percent of observations, the time
the lobsters spent resting and pleopod-fanning
was considerable (18-45%) in all substrates and did
not differ between them. Walking was only observed
when a lobster did not have a burrow. Feeding
occurred only on the few occasions when some
edible object landed close enough to the burrow so
that the lobster could reach it without entirely
leaving its burrow. Twice lobsters were seen catch-
ing swimming amphipods at the entrance to their
burrow.
American lobsters were observed creating a cur-
rent by pleopod-fanning, which was seen to draw
plankton through their burrows. During these
periods the lobster stood with its clawed limbs held
up and apart. The mouth parts, particularly the sec-
ond and third maxillipeds moved rapidly, and the
first pair of walking legs were often brought up to
the mouth. Occasionally the lobster would jerk for-
ward and snap its claws. All of the above-mentioned
appendages are covered with various types of setae
(Factor 1978), which could help the lobsters to catch
the plankton both by "filtering" with their maxilli-
peds, claws, and first walking legs, and by seizing
the plankton with their claws. These observations,
supported by Lavalli and Barshaw (1986) and Bar-
shaw (in press) show that American lobsters are
able to catch plankton while remaining in their
burrow.
792
BARSHAW and BRYANT-RICH: EARLY JUVENILE AMERICAN LOBSTER
Mortality
There was an initial mortality of the American
lobsters in all habitats followed by no deaths in the
winter and another die-off in the spring (Fig. 2). The
mortality rate for American lobsters in eelgrass was
significantly lower than those in the mud or in the
rocks (arcsine transformation, linear, least-square
regression, r^ = 0.75 eelgrass, 0.86 rock, 0.94
mud, comparison of slopes, P < 0.001).
Molting and Size
At the end of the experiment there was no signif-
icant difference in the size of the lobsters between
habitats, although the American lobsters in eelgrass
tended to be larger (Table 2C). There was also no
significant difference in the number of observed
molts between treatments. We observed molting by
six lobsters during the coldest part of the year when
the water temperature was between 1° and 2°C.
Possible Prey
The biomass of American lobsters (>1 mm) was
not significantly different among treatments, but
the biomass of American lobsters in eelgrass tended
to be higher (ANOVA, Newman-Keuls test, P <
0.01, Table 3). There were fewer different genera
residing in the mud habitats, with the greatest diver-
sity in rock.
DISCUSSION
Initial observations showed that stage IV Ameri-
can lobsters started to burrow more quickly in eel-
grass and rock habitats. While several investigators
have shown that American lobsters choose rocks
over mud (Howard and Bennett 1979; Pottle and
Elner 1982; Botero andAtema 1982), no choice ex-
periments have used eelgrass as a substrate. Like-
wise, in this experiment we have not directly shown
that the lobsters prefer the eelgrass substrate
because they were not offered a choice. Speed of
settling would be indicative of a preference, how-
ever, if the method that lobsters use to choose a
substrate is to keep swimming if the habitat is un-
suitable, but settle if it is suitable. Such behavior
was observed in laboratory experiments by Botero
and Atema (1982). Indications that lobsters keep
swimming over unsuitable substrates was also ob-
served by Cobb et al. (1983) in the field.
Q
<
LU
Q
H
Z
LU
O
QC
LU
Q.
NOV
DEC
JAN
FEB
MAR
APR
MAY
Figure 2.— The cumulative percent of American lobsters that died in the three treatments, each
point represents a day when a lobster died. Fifteen lobsters per treatment were present at the begin-
ning of the experiment. Eelgrass vs. mud and eelgrass vs. rock, P < 0.001.
793
FISHERY BULLETIN: VOL, 86. NO. 4
Table 3. — Average biomass (wet weight) and species list of organisms larger than 1 mm liv-
ing in the sediment of each treatment at the end of the expehment.
Treatment
Bivalve
Polycheate
Other
Eeigrass: Mean biomass per tank
Mercenaha mercenaria
Yoldia (spp.)
Rock: Mean biomass per tank =
Andora oval is
Andora transversa
Mercenaria mercenaria
Mucoma (spp.)
Yoldia (spp.)
8.2 ± 2.7 g
Capitella (spp.)
Clemenalla (spp.)
Glycera (spp.)
Lumbrineris (spp.)
Neris (spp.)
Spiondae (gen.)
Terebellida (gen.)
3.28 ± 1.07 g
Chrysopetalidae
{Dysponetus pygmaeus)
Nephytidae (spp.)
Nereis (spp.)
Phyllodocdae (gen.)
Sapella (gen.)
Terebellida (gen.)
Sipunculoidea
{Golfingia gouldi)
Sipuculoidea
{Golfingia gouldi)
Crustacea
(Neopanope sayi)
Mud: Mean biomass per tank = 4.4 + 1.07 g
Geukendia demessus
Solemya velum
Yoldia (spp.)
Glycera (spp.)
Nephtys (spp.)
Nereis (spp.)
Orbiniidae (spp.)
Terebelleda (gen.)
The mud substrate appeared to be the least
suitable of the three tested since American lobsters
in eeigrass and rocks were without burrows for less
time than those in mud. However, even in the mud
habitat the lobsters had burrows for an average of
87.8% of the observations (Table 2B). The three
substrates used in this study were chosen partly
because of their differences, yet the lobsters man-
aged to build and maintain burrows in all three
substrates for eight months. This result clearly
shows that early juveniles have flexible behavior and
modify it to adapt to different substrates.
The way that the American lobsters partitioned
the amount of time they spent on different activities
was also affected by the substrate in which they
lived. While lobsters in eeigrass and rock habitats
spent little time repairing their burrows, lobsters
in mud spent considerable time on repair (Fig. 1).
This result is consistent with the characteristics
of the three substrates. Eeigrass stabilizes the
underlying sediment (decreases erosion) by baffling
the water currents with its leaves and binding the
sediment with its roots (Scoffin 1970); rocks, al-
though usually found in areas of stronger currents,
provide a ready made solid roof; mud, however, is
more easily disturbed (Rhoads and Young 1970). If
one lobster did not have a burrow in the mud tanks,
its walking often destroyed the other lobsters'
burrows.
The extra time that the American lobsters in mud
spent repairing their burrows was subtracted mainly
from investigation time (Fig. 1), perhaps because
there was not as much prey in the mud for the
lobsters to detect, so this activity was the most ex-
pendable. No significant differences were found
between the time budgets of the lobsters in eeigrass
and the lobsters in rocks; however, the lobsters in
eeigrass spent more time resting than investigating,
while the opposite was true of the lobsters in rock
substrate.
The American lobsters in the eeigrass had a lower
mortality rate than those in either mud or rocks.
This result could have been due to the greater bio-
mass of possible prey animals living in the eeigrass
habitat, and/or the greater complexity of the eel-
grass habitat, which in essence separated the lob-
sters and ameliorated the effects of high density.
Seagrass beds in nature have also been shown to
have a greater biomass of species living in them than
the biomass of species living in less complex sub-
strates such as mud or sand (Orth 1973; Thayer et
al. 1984).
The lower mortality rate of lobsters in eeigrass
led to a greater number of lobsters per tank in this
treatment. Higher concentrations of lobsters have
been shown to cause slower rates of growth in
lobster living in fairly unnatural substrates (Cobb
and Tamm 1974). In this study, the American
794
BARSHAW and BRYANT-RICH: EARLY JUVENILE AMERICAN LOBSTER
lobsters in eelgrass were not smaller perhaps also
because the eelgrass substrate ameliorated the
effects of higher density. We did not observe any
differences in the activity budgets of the lobsters
owing to higher density (Fig. 1).
This study shows that early juvenile American
lobsters differ dramatically in their behavior from
older lobsters. They seldom, if ever, forage for food
outside of their burrows, but instead remain inside
of them. This was true even though there were no
predators present other than other juveniles.
The lobsters' main activities within their bur-
rows reflected their needs. Because they did not for-
age outside of their burrows, all of the early juve-
niles' nourishment must have been found inside
of their burrow, or within reach of the entrance.
Lobsters could forage on polycheates, meiofauna,
and on any other organisms residing inside their bur-
rows or draw plankton in by pleopod-fanning. In this
study lobsters were seen to catch swimming am-
phipods at the entrance to their burrow, and Ber-
rill (1974) observed similar behavior. Besides rest-
ing, the lobsters mainly "investigated" for anything
edible in the entrance of their burrow and pleopod-
fanned to draw in plankton. If they were forced to
burrow in a relatively unstable substrate, such as
mud, they spent a significant amount of time main-
taining that burrow.
The claws of early juvenile American lobsters are
smaller and weigh less relative to the abdomen than
those of older lobsters, and, by external appearance
the two claws are not differentiated from each other.
Furthermore, the speed of the tail flip reflex is faster
at sizes smaller than 20 mm carapace length (Lang
et al. 1977). These morphological characteristics
along with the behavioral results from this study,
and field observations that juveniles become easier
to find at a carapace length of 20-40 mm (Cooper
and Uzmann 1980; Able et al. in press), indicate that
the juvenile stage of the American lobster can ac-
tually be divided into two substages: 1) the early
juvenile stage, spanning settlement to the time un-
til claws begin to differentiate, during which period
the lobsters seldom, if ever, leave their burrow; and
2) the late juvenile stage, starting when the claws
are differentiated and become larger in relation to
the abdomen and ending with sexual maturity. At
this stage, the lobsters start to forage for food
outside of their burrows, and behave more similar-
ly to adults (Cooper and Uzmann 1980; Able et al.
1988).
We suggest the following scenario for the life
history of early juvenile American lobsters. After
settling onto a suitable substrate the lobsters build
a burrow where they remain for the duration of the
"early juvenile" substage. By catching food, both
in the substrate around their burrow entrance and
by drawing plankton into their burrow by pleopod-
fanning, the early juvenile lobsters manage to sur-
vive without foraging outside their burrow.
ACKNOWLEDGMENTS
We thank Diane Cowan and Ruth lannazzi for help
with observations and maintenance of the system.
We also thank Mike Eagles then of the St. Andrews
Marine Biological Field Station, New Brunswick,
Canada for supplying us with stage IV American
lobsters. We also thank Kenneth W. Able for review-
ing the manuscript and Judith Capuzzo and Stanley
Cobb for reviewing an earlier draft. This research
was supported in part by a National Wildlife Feder-
ation - American Petroleum Institute, Environmen-
tal Conservation Fellowship to Diana Barshaw.
LITERATURE CITED
Able. W. K., K. L. Heck, Jr., M. P. Fahay, and C. T. Roman.
1988. Habitat utilization of juvenile lobsters in a Cape Cod
estuary: evidence for the importance of salt marshes. Estu-
aries 11:83-86.
Atema, J., D. F. Leavitt, D. E. Barshaw, and M. C. Cuomo.
1982. Effects of drilling muds on the behavior of the Ameri-
can lobster, Homarus americaniis, in water column and
substrate exposures. Can. J. Fish. Aquat. Sci. 39:675-690.
Barshaw, D. E.
In press. Growth and survival of early juvenile lobsters,
Homarus americanus, on a diet of plankton. Fish. Bull.
Berrill, M.
1974. The burrowing behavior of newly-settled lobsters.
Homarus vulgaris (Crustacea-Decapoda). J. Mar. Biol.
Assoc. U.K. 54:797-801.
Berrill, M., and R. Stewart.
1973. Tunnel-digging in mud by newly-settled lobsters,
Homarus americanus. J. Fish. Res. Board Can. 30:285-
287.
Botero, L., and J. Atema.
1982. Settlement and substrate selection during larval set-
tling in the lobster, Homarus americanus. J. Crust. Biol.
2:59-69.
Cobb, J. S.
1971. The shelter-related behavior of the lobster. Homarus
americanus. Ecology 52:108-115.
1987. Summary of session 6: Ecology of population structure.
Can. J. Fish. Aquat. Sci. 43:2389-2391.
Cobb, J. S., T. Gulbansen, B. F. Phillips, D. Wang, and M.
Syslo.
1983. Behavior and distribution of larval and early juvenile
Homarus americanus. Can. J. Fish. Aquat. Sci. 40:2184-
2188.
Cobb, J. S., and G. R. Tamm.
1974. Social conditions increase intermolt period in juvenile
795
FISHERY BULLETIN: VOL. 86. NO. 4
lobster Homaras americanus. J. Fish. Res. Board Can.
32:1941-1943.
Cooper, R. A., and J. R. Uzmann.
1980. Ecology of juvenile and adult Hcymarun. In J. S. Cobb,
and B. F. Phillips (editors). The biology and management
of lobsters, Vol. I, p. 97-142. Acad. Press, N.Y.
Ennis, G. p.
1975. Behavioral responses to changes in hydrostatic pres-
sure and light during larval development of the lobster
Homarus americanus. J. Fish. Res. Board Can. 32:271-
281.
Factor, J. R.
1978. Morphology of the mouthparts of larval lobsters,
Homarus americanus, with special emphasis on their setae.
Biol. Bull. (Woods Hole) 154:383-408.
Fogarty, M. J.
1987. Summary of session 8: Models of stock-recruitment
relationships. Can. J. Fish. Aquat. Sci. 43:2392-2393.
Howard, A. E., and D. B. Bennett.
1979. The substrate preference and burrowing behavior of
juvenile lobsters, Homanis gammarus. J. Nat. Hist. 13:
433-438.
Lang, F., C. K. Govind, W. J. Costello, and S. I. Green.
1977. Developmental neuro- ethology: changes in escape and
defense behavior during growth of the lobster. Science
197:682-685.
Lavalli, K. L., and D. E. Barshaw.
1986. The ability of post-larval American lobsters to re-
move plankton from the water column. Am. Zool. 26:
90 A.
MacKay, D. a.
1926. Post-larval lobsters. Science 64:530.
Orth, R. J.
1973. Benthic infauna of eelgrass, Zostera marina, beds.
Chesapeake Sci. 14:258-269.
Phillips, B. F., J. S. Cobb, and R. W. George.
1980. General biology. In J. S. Cobb, and B. F. Phillips
(editors), The biology and management of lobsters, Vol. I,
p. 1-81. Acad. Press, N.Y.
Pottle, R. A., and R. W. Elner.
1982. Substrate preference behavior on juvenile American
lobsters, Homarus americanus, in gravel and silt-clay sedi-
ments. Can. J. Fish. Aquat. Sci. 39:928-932.
Rhoads, D. C, and D. K. Young.
1972. The influence of deposit-feeding organisms on sediment
stability and community trophic structure. J. Mar. Sci. 28:
150-178.
SCOFFIN, T. P.
1970. The trapping and binding of subtidal carbonate sedi-
ments by marine vegetaion in Bimi Lagoon, Bahamas. J.
Sediment Petrol. 40:249-273.
Thayer, G. W., W. J. Kenworthy, and M. S. Fonseca.
1984. The ecology of eelgrass meadows of the Atlantic coast:
a community profile. U.S. Wildl. Serv. FW5/0B5-84/02.
796
DISTRIBUTION AND ABUNDANCE OF THE BOTTLENOSE DOLPHIN,
TURSIOPS TRUNCATUS (MONTAGU, 1821), IN VIRGINIA^
Robert A. Blaylock^
ABSTRACT
The distribution and abundance of the bottlenose dolphin, Tursiops truncatus, was examined by conduct-
ing aerial surveys of the Chesapeake Bay mouth and nearshore coastal waters of Virginia in 1980 and
1981. Bottlenose dolphin density was estimated using line transect methods and a 4-term Hermite
polynomial was chosen to model the detection function. Six surveys in the Chesapeake Bay mouth resulted
in an average density estimate of 0.159 dolphins/km". Ten surveys along the southern Virginia coast
produced an average density estimate of 3.446 bottlenose dolphins/km' within 2 km of shore. Average
bottlenose dolphin abundance in the Chesapeake Bay mouth and along the southern Virginia coast was
estimated at 340 dolphins (±104, 95% C.I.). An estimate of 0.208 bottlenose dolphins/km" along the
northern Virginia coast is tenuous because only one survey was conducted there. Dolphin sightings were
distributed uniformly along the southern Virginia coast with the exception of some clustering of herd
sightings at the capes bordering the Chesapeake Bay mouth. The percentage of calves per herd aver-
aged 7.5% in Chesapeake Bay mouth, 4.3% in the southern coastal area, 9.0% in the northern coastal
area, and peaked in June. Five of seven bottlenose dolphins identified by unique dorsal fin shapes in
1980 were resighted in 1981, suggesting seasonal residency of individuals.
Of the 23 cetacean species occurring along the
Virginia coast (Leatherwood et al. 1976; Blaylock
1985) the bottlenose dolphin, Tursiops truncatus
(Montagu, 1821), is the only cetacean found near
shore regularly and in large numbers. However,
there are few quantitative data available to assess
the abundance, distribution, and seasonal occur-
rence of Tursiops truncatus (hereafter referred as
Tursiops) in Virginia coastal waters.
Those bottlenose dolphins occurring seasonally in
Virginia are believed to form part of a population
distributed from northern North Carolina to New
Jersey during the summer. This population was the
focus of a sporadic fishery along Hatteras Island,
NC from circa 1797 to 1929, the primary products
of the fishery being hides and oil (True 1891; Town-
send 1914; Mead 1975). From cumulative catch
records, Mitchell (1975) estimated a historical popu-
lation size of 13,748-17,000 dolphins and inferred
annual migration from biannual peaks in catches
during the fall and spring. True (1891) earlier sug-
gested a north-south migration, reporting on fish-
ermen's observations that bottlenose dolphins were
usually seen traveling south in the fall and north in
the spring, with only a few remaining near Hatteras
during the summer.
Analysis of large-scale aerial surveys along the
northern and mid-Atlantic U.S. coast revealed a
bimodal longitudinal Tursiops distribution, inter-
preted as separate nearshore and offshore areas of
abundance (CETAP 1982). These areas represent
the habitats of two distinct morphological types of
T. truncatus. The offshore type is slightly larger at
the onset of physical and sexual maturity than the
nearshore types and ultimately attains a greater
size^.
An important finding of the CETAP surveys was
the presence of multiple latitudinal peaks in coastal
sightings indicating discontinuities in the north-
south distribution of nearshore Tursiops (CETAP
1982). These observations indicate either an uneven
distribution of nearshore Tursiops or the presence
of multiple coastal populations or subpopulations.
However, a recent epidemic suggests that the
U.S. east coast Tursiops may represent a single
stock.
Tursiops mortalities south of North Carolina
during autumn of 1987 increased sharply with the
apparent emigration of Tursiops from Virginia
'Contribution No. 1464 from the Virginia Institute of Marine
Science.
^Virginia Institute of Marine Science, Gloucester Point, VA
23062.
'J. G. Mead, Division of Mammals, Smithsonian Institution,
Washington, DC, pers. commun. June 1978.
Manuscript accepted June 1988.
FISHERY BULLETIN; VOL. 86, NO. 4, 1988.
797
FISHERY BULLETIN: VOL. 8«. NO. 4
waters.^ If the high level of Tursiops mortalities ex-
perienced in the mid-Atlantic coast during the sum-
mer of 1987 was because of an infectious agent, then
its spread to conspecifics in more southerly regions
may have been caused by contact between in-
dividuals from different areas and more extensive
migration than has been previously suggested.
In the present study I used aerial surveys to esti-
mate the abundance and examine the distribution
of T. truncatus in Virginia coastal waters, including
the Chesapeake Bay mouth. I also investigated
natality periods by monthly comparison of the aver-
age percentage of calves present and residency pat-
terns using photographic records of identifiable
individuals.
METHODS
Aerial surveys were conducted during July-
October 1980, and May-June 1981, from a high-
winged, single-engine aircraft (U6A DeHavilland
Beaver^) at an altitude of 152 m and at an air-
speed of 147 km/h. Observers sitting in the two
passenger seats searched each side of the transect
for bottlenose dolphins. A recorder/navigator sitting
forward of the observers and next to the pilot helped
to maintain predetermined transect lines and
recorded sightings which were communicated via
intercom.
Upon sighting a bottlenose dolphin herd, the per-
pendicular distance from the flight path to the herd
center was determined from calibrated, taped mark-
ings on the wing struts with the aircraft in level
flight or a hand-held inclinometer. The transect was
then temporarily halted and the herd circled at a
lower altitude to count individuals. The herd loca-
tion, direction of travel, behavior, and the number
of calves were also noted. Transect lengths and the
survey area were measured with a digital planimeter
from NOS/NOAA navigation charts.
Depending upon the area surveyed (Fig. 1), two
types of survey schemes were used. Systematic, lati-
tudinally oriented transects were used in the Chesa-
peake Bay mouth (CBM) during 1980. The northern
starting point for each survey was randomized, and
each transect was located 7.4 km south of the pre-
vious transect. Two exceptions to this regime oc-
■•D. M. Burn, Southeast Fisheries Center, National Marine Fish-
eries Service, NOAA, 75 Virginia Beach Drive, Miami, FL 33149,
pers. commun. June 1988.
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
curred, but in neither case was the distance between
transects less than 3.7 km. Three or four transects
were flown during each CBM survey and each
survey covered approximately 30% of the total
survey area. CBM surveys were not conducted in
1981.
Longshore surveys were flown from north to
south in 1980 and 1981, parallel to the coast and 1
km offshore from Cape Charles to False Cape (32.3
km). Those conducted in 1980 were flown imme-
diately upon completing the CBM surveys so that
there was no possibility of counting herds in the
longshore area that were counted during CBM
surveys, except perhaps during transit between
Cape Charles and Cape Henry, which was flown
over open water on the shortest line between the
two points. One additional survey was flown along
the northern Virginia coast.
After subtracting the minimum distance from the
transect that could be observed because of limited
visibility directly beneath the aircraft, the perpen-
dicular sighting distance data were truncated at 1
km. Data from all three study sites were then pooled
for the calculation ofg{x), the detection function for
line transect, and [/'(O)], the probability density func-
tion of perpendicular sighting distances evaluated
at the transect. In line transect the detection func-
tion gix)is the conditional probability of observing
an object at perpendicular distance x from the tran-
sect line and /(a:;) is ^'(a;) scaled to integrate to one
(Burnham et al. 1980). Each survey was treated as
a replicate to determine the analytical variance of
/(O). Herd density was then calculated separately for
each of the survey areas using /(O) estimated from
the pooled sightings.
Several estimates of /(O) and its analytical vari-
ance were calculated by fitting parametric and
nonparametric models to the distribution of perpen-
dicular sighting distances using the Fortran
programs TRANSECT (Laake et al. 1979) and
HAZARD and HERMITE (Buckland 1985). Maxi-
mum likelihood estimates and large-sample vari-
ances were found using the procedure of Burnham
et al. (1980: 135-136). The Fortran program SIZE-
TRAN (Drummer and McDonald 1987) was used to
test the hypothesis of independence between herd
size and perpendicular sighting distance using a
likelihood ratio test and thus determine if the detec-
tion function was biased by herd size.
Herd density was estimated as (Burnham et al.
1980, p. 18, eq. 1.3):
D = nf{0)/2L
798
BLAYLOCK: DISTRIBUTION OF THE BOTTLENOSE DOLPHIN
76°|00'
76° 00
Figure 1.— Coastal survey areas and bottlenose dolphin sightings (dosed circles)
and Chesapeake Bay mouth survey area (enclosed by dark lines) with dolphin sight-
ings (open circles). Dashed line represents transit during coastal surveys. Sightings
near Fishermen's Island and Cape Henry which occurred during coastal surveys
were included only in the coastal survey analyses.
where n is the number of herds detected and L is
the transect length in kilometers. The variance of
D was estimated as (Burnham et al. 1980, p. 51, eq.
1.17):
SHD) = Dnicvin)f + (cv(f(0))2].
Herd size was not significantly different between
study areas (Kruskal-Wallis test (K-W test), x" =
0.9953, df = 2, P = 0.61, Sokal and Rohlf 1981) and
sightings were pooled to determine the overall mean
herd size. Herd sizes were not normally distributed
(Fig. 2) and therefore were normalized by log trans-
formation to calculate the geometric mean (Sokal
and Rohlf 1981) and its variance. Bottlenose dolphin
density (P) is the_product of herd density 0) and
mean herd size (H). The variance of P, following
Goodman (1960, p. 710, eq. 7) is
,2^ _ H's^jD) DH^{H) sHD)sHH)
S {r) — , + — — , —
niD)
n{H)
n{Dyn{H)
with 710) equal to the number of herd sightings in
the survey area and n{H) equal to the number of
herds used in the estimation of H. This assumes
799
FISHERY BULLETIN: VOL. 86, NO. 4
C/)
Q
OC
lU
I
DC
LU
CD
25
20-
15
10-
5-
22
r~i . r
1 1
o
o
O
O
in
o o o o o
ix) r^ 00 en o
o
o
o
ro
O
2
O
in
C\j
io
5^
in ID r^ i£> ^
HERD SIZE
o
^
ro
§
Figure 2.— Distribution of bottlenose dolphin herd sizes. Numbers above
bars denote the number of herds in that size class.
independence between herd size and perpendicular
sighting distance.
Bottlenose dolphin abundance is the product of P
and the area surveyed. In the CBM area this may
be extrapolated to the total area if the transects are
distributed randomly with respect to dolphin sight-
ings.
The recorder did not distinguish between observ-
er's sightings when recording them, thus observer
bias was not investigated. The effects of sea state
and sun glare on detectability were not investigated.
Surveys were not conducted when sea states were
above two on the Beaufort scale, and it is unlikely
that sea state influenced the results. However, the
effect of glare reduced the observers' field of view,
which decreased the number of animals detected and
resulted in an underestimation of P.
I conducted photographic surveys from a 7 m boat
on five occasions in 1980 and six in 1981 for the
purpose of identifying individual bottlenose dolphins
by the shape of, or markings on, their dorsal fins.
Contact prints of the 35 mm photographs were
examined under a dissecting microscope at 40 x
magnification.
RESULTS
Six aerial surveys in the CBM averaged 119.4 km
per survey, covered an area of 762 km^, and re-
sulted in five herd sightings of bottlenose dolphins.
Ten surveys along the southern Virginia coast re-
sulted in 49 herd sightings. Each coastal survey was
32.3 km in length and covered an area of 65 km^.
An additional survey along the northern Virginia
coast was 108 km in length, covering an area of 216
km^, and resulted in two herd sightings.
In line transect the distance at which a bottlenose
dolphin herd is sighted is assumed to be indepen-
dent of its size (Burnham et al. 1980; Seber 1986).
Although it seems reasonable that larger herds
would be detected at greater distances, analysis of
herd size and sighting distance using the method of
Drummer and McDonald (1987) showed no signifi-
cant size-bias (P > 0.05). As a check, I also regressed
herd size against perpendicular sighting distance.
There was no apparent association between herd
size and distance from the transect (r^ = 0.001)
(Fig. 3). The geometric mean herd size was 14.4
bottlenose dolphins/herd (SE = 4.0, n = 56).
Truncation of bottlenose dolphin sightings at 1 km
resulted in the discarding of one herd sighting in
the southern coastal area, none in the CBM, and one
in the northern coastal area. The truncated sighting
in the northern coastal area was at approximately
1,200 m from the transect and the herd was ap-
parently feeding in the wake of a trawler, thus the
sighting was atypical of other sightings during this
study and probably influenced by the presence of
the trawler (see Leatherwood 1975). Both sightings
in this area occurred farther offshore than sightings
in the other study areas.
Several parametric and nonparametric models
were investigated for fit to the pooled perpendicular
sighting distances (Table 1). None of the models dif-
fered significantly from the observed distance dis-
tributions (chi-square test, P > 0.05). The coefficient
800
BLAYLOCK: DISTRIBUTION OF THE BOTTLENOSE DOLPHIN
1.4
I ,2
O
LU
C/5
z
<
tr
O
DC
U.
LU
O
CO
Q
10
0-
r^ = OOOI
, N= 56
•
•
••
• •
•
•
•
• •
•
•
# . •
1 1
••
• •
•
• • •
1 r
•
1 1
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160
HERD SIZE
Figure 3.— Scatterplot of sighting distance (in kilometers) versus herd
size of bottlenose dolphins. (Note that some of the points in the 1-10 size
class and the interval 0-0.2 km represent more than one herd.)
Table 1 .—Models used in calculation of the detection function g{x)
for bottlenose dolphin. N is the number of terms used in the model
and f(0) is f(x) evaluated at the transect. SE[f(0)] was calculated
assuming asymptotic normality with Z = 1.96.
Model
N
^"(0)
SE(f"(0)]
Reference
Hermite polynomial
3
2.849
0.441
Buckland 1985
Hermite polynomial
4
3.104
0.522
Buckland 1985
Hazard rate
na
3.004
1.345
Buckland 1985
Fourier series
4
3.323
0.551
Burnham et al.
1980
Negative exponential
na
3.216
0.936
Burnham et al.
1980
Exp. power series
2
3.649
2.198
Burnham et al.
1980
Exp. polynomial
2
3.001
0.866
Burnham et al.
1980
Half normal
na
1.950
0.398
Burnham et al.
1980
of variation of the 3-term Hermite polynomial was
slightly less than that of the 4-term (0.154 vs. 0.168),
but the 4-term Hermite polynomial model provided
a better fit to the observed perpendicular sighting
distance distribution than either the 3-term Hermite
polynomial or the 4-term Fourier series models
(Table 2).
The appropriate model for the observed perpen-
dicular sighting distances should fit the data most
closely near the centerline of the transect (Schweder
1977). The 4-term Hermite polynomial model closely
approximates the observed sighting distances in the
interval 0-200 m and also in the subsequent inter-
vals (Fig. 4).
The nonparametric 4-term Hermite polynomial
model yielded an estimate of /(O) = 3.104 (SE =
0.522). Dolphin density in the southern Virginia
coastal area (3.446 dolphins/km-) was much greater
than that in the CBM area (0.159 bottlenose dol-
phins/km^); however, the abundance estimates are
of similar magnitude (219 vs. 121 bottlenose dol-
phins, respectively) due to the greater area sampled
Table 2.— Observed and expected distribution of bottlenose dolphin herd sightings by distance in-
tervals perpendicular to the transect (PSD in meters) with chi-square values. Figures in parentheses
are degrees of freedom. Expected values are rounded to one decimal place for clarity of presen-
tation.
4-term Fourier
PSD
(m)
Observed
4-ferm Hermite
3-term Hermite
series
Expected
x'
Expected
x'
Expected
x'
0-200
27
27,2
0.120E-2
26.1
0.320E-1
27.4
0.698E-2
200-400
10
9.8
0.441 E-2
11.1
0.117
7.4
0.913
400-600
9
9.4
0.205E-1
8.2
0.803E-1
11.6
0.593
600-700
5
4.4
0.819E-1
4.5
0.423E-1
4.9
0.273E-2
700-800
2
2.3
0.338E-1
3.3
0.819
1.9
0.197E-2
800-1,000
1
1.0
0.400E-3
0.9
0.169E-1
0.7
0.114E-2
Cumulative x^
0.137(1)
1.108 (2)
1.519(1)
801
FISHERY BULLETIN: VOL. 8fi. NO. 4
3.104
>
M
Z
111
o
>■
<
a
o
DISTANCE FROM TRANSECT (km)
Figure 4.— Four-term Hermite probability density function fit to histogram of bottle-
nose dolphin sighting frequency and perpendicular distance (rescaled to account for
the "blind spot" beneath the aircraft). Numbers above bars denote number of sight-
ings in interval.
in the CBM (Table 3). CBM and coastal surveys were
flown sequentially with no delay between them and
the relatively high speed of the aircraft prevented
counting of the same herd twice. Therefore, the
abundance estimates in the CBM and the southern
coastal survey areas may be considered additive and
totaled 340 bottlenose dolphins (±104, 95% C.I.)-
The survey altitude limited observations of herd
composition to the percentage of calves in each herd.
Bottlenose dolphins much smaller than the others
and accompanied by a larger bottlenose dolphin
were considered to be calves. The mean percentage
Table 3. — Summary of aerial survey results of bottlenose dolphins.
N is the number of surveys; L is the total length of transects at each
location in km; and n is the number of herd sightings within one
km of the transect. D is estimated herd density (herds/km^); P is
estimated dolphin density; and A is estimated dolphin abundance
(standard errors in parentheses).
Survey
location
N
L
n
6
P
A
+ 95%
C.I.
Chesapeake Bay
mouth
6
717
5
0.011
(0.003)
0.159
(0.017)
121
(13)
33
Southern Virginia
coast
10
323
48
0.239
(0.063)
3.446
(0.193)
219
(54)
122
Eastern Shore
1
108
1
0.014
(0.004)
0.208
(0.058)
45
(13)
—
of calves in all herds from all study areas peaked
in June at 9.5% (SE = 0.8, n = 3) and declined
thereafter until September (Fig. 5); however, there
were no significant differences between months
(K-W test, r = 9.1930, df = 5, P > 0.10). Con-
sidering the total study period, the mean percentage
of calves in herds in the CBM area was 7.5% (SE
= 3.2%, n = 5); in the southern coastal area, 4.3%
(SE = 1.0%, n = 49); and in the northern coastal
area, 9.0% (SE = 1.8%, n = 2). The mean percent-
age of calves in herds did not differ significantly
among areas (K-W test, x" = 2.8196, df = 2, P =
0.24).
Bottlenose dolphins were never sighted more than
1.6 km from shore during CBM surveys nor during
subsequent surveys up to 8 km offshore of the
coastal study area. Also, bottlenose dolphins were
not found in depths greater than 10 m except in the
Eastern Shore area. Plotting of bottlenose dolphin
sightings (Fig. 1) shows a uniform distribution along
the southern coastal area with some clusters of
sightings at Cape Henry and at Cape Charles.
During 1980, seven bottlenose dolphins which
were recognizable by the shape of the trailing edge
of their dorsal fins were identified and photo-
graphed. In 1981, 17 individuals identified from
dorsal fin photographs included 5 which had been
photographed in 1980. Thus, of the 19 recognizable
802
BLAYLOCK: DISTRIBUTION OF THE BOTTLENOSE DOLPHIN
individuals present during either of those two years,
at least 26% were present at some time during both
summers.
14
o
cc
lU
X
z
CO
lU
>
<
u
GC
UJ
a.
12-
10-
2-
16
13
20
MAY
JUN
JUL AUG
MONTH
SEP
OCT
Figure 5.— Percentage of bottlenose dolphin calves in herds by
month. Bars represent the standard error of the mean (horizontal
line within bars) and vertical lines, 95% confidence intervals. Num-
bers above bars denote the number of herds sighted per month.
DISCUSSION
The choice of a model for g{x), the probability of
detecting an object at a distance x from the tran-
sect, is the primary analytical consideration in a line
transect estimate of density (Burnham et al. 1980;
Seber 1982, 1986). Burnharn et al. (1980) thoroughly
review the subject of density estimation from line
transect surveys and recommended the Fourier
series as a general model for g(x). However, addi-
tional models which meet their criteria have since
been proposed (Bnckland 1985; Seber 1986). Buck-
land (1985) suggested the use of a model where the
cosine terms in the Fourier series equation are re-
placed by Hermite polynomials.
Buckland (1985) warned that if the model requires
four or more terms to fit the distributional data, one
or more of the assumptions of line transect theory
may be violated. I suggest that, in aerial surveys
of cetaceans, the primary assumption that all objects
on the transect are observed with a probability of
one [giO) = 1] is routinely violated. The diving be-
havior of cetaceans during different activities may
vary widely, thus the probability of the animals be-
ing at the surface when the observers pass may also
vary. Also, active dolphins may be more readily
detected than resting dolphins. In spite of this, this
assumption is somewhat less restrictive than the
primary assumption of strip census which assumes
that all objects within the strip are detected. If the
other assumptions are met, the major consequence
of failure to meet the assumption of ^'(0) = 1 is that
density will be underestimated.
A further assumption is that perpendicular dis-
tances are measured without error. Even using an
inclinometer, vertical motion of the aircraft and
inaccuracy of the altimeter introduce error into
distance measurements. Grouping distance mea-
surements into discrete intervals is a logical way in
which to compensate if the model used is robust to
grouping.
The assumption of random location of transects
with respect to bottlenose dolphin distribution was
met by randomization of the starting point of each
survey in the CBM. It is obvious from the cluster
of sightings at Cape Henry and Fisherman Island
that bottlenose dolphins were not distributed ran-
domly in the coastal study area (Fig. 1). This could
occur if bottlenose dolphins were counted more than
once; however, their movement was slow compared
with that of the observers and, because longshore
surveys were flown immediately upon completion
of CBM surveys, it is unlikely that dolphins were
counted more than once. It is more likely that the
cluster of sightings was because of an environmen-
tal factor, such as the attraction of dolphins to con-
centrations of prey in fronts between estuary and
ocean waters.
According to Essapian (1963), mating by the
bottlenose dolphin occurs in the spring and birth
occurs about one year later (McBride and Kritzler
1951; Tavolga and Essapian 1957). Mead (1975),
citing True (1891), stated that "Information received
from the fishermen at the Hatteras fishery indicated
that fetuses were generally small in September, in-
creasing in size as the season progressed." This im-
plies that natality occurs primarily in the spring.
Townsend's (1914) data (also cited in Mead 1975)
suggest an additional autumn peak in natality. The
June peak in the percentage of calves agrees with
those observations suggesting a spring natality
peak; however, because of the slight increase in the
803
FISHERY Bl'LLKTIN: VOL. 86. NO. 4
percentage of calves in September, a second autumn
peak cannot be ruled out.
Resighting of 26% of the identifiable bottlenose
dolphins in 2 successive years is evidence that some
of these return to the same area. Although the prob-
ability of resighting individuals twice in 2 successive
years is low if the individuals are transient, knowl-
edge of the length of stay is required to infer
seasonal residency. A study similar to that con-
ducted on bottlenose dolphins in Argentina (Wiir-
sig and Wiirsig 1978) could provide information on
the length of individual residency and should be con-
sidered. This would facilitate interpretation of the
data presented here, as well as that gathered from
currently ongoing surveys.
Because of the violation of several important
assumptions, the accuracy of the density and abun-
dance estimates reported here is difficult to assess.
The CETAP (1982) summer average density esti-
mate of nearshore Tursiops in the mid-Atlantic
region was 0.0093 dolphins/km^. This is much
lower than my estimate of 0.159 dolphins/km- in
the Chesapeake Bay mouth and 3.446 dolphins/km-
in the southern coastal region. Besides differences
in survey altitude and airspeed, one possible reason
for this discrepancy is the larger area surveyed dur-
ing the CETAP program. If the coastal Tursiops
are generally found close to shore (within 2 km) and
the area surveyed extends far beyond this distance,
then the density of coastal Tursiops in its typical
habitat will be underestimated. Alternatively, a
heterogeneous coastal distribution could account for
this discrepancy.
The importance of an average bottlenose dolphin
density estimate which may be used as an index of
abundance has recently been emphasized by an inci-
dence of disease which resulted in the deaths of over
200 Tursiops along the Virginia coast and over 400
along the Mid- Atlantic Bight during the summer of
1987.^ The rather large gap in the coastal Tursiops
abundance data base renders assessment of the im-
pact of the 1987 mortalities on local Tursiops stocks
problematic. Future monitoring of the coastal Tur-
siops may provide answers as to the rate of recovery
and allow assessment of the impact of future catas-
trophic events. A coordinated, long-term program
to monitor coastal Tursiops abundance would per-
mit temporal comparisons of abundance indices and
provide a greater understanding of natural popula-
tion fluctuations. Because the coastal Tursiops
^J. G. Mead, Division of Mammals, Smithsonian Institution,
Washington, DC, pers. commun. June 1988.
inhabit an area where human activity is rapidly in-
creasing, such a monitoring program should receive
high priority.
ACKNOWLEDGMENTS
I greatly appreciate the participation of R. Bow-
man, S. White, and especially R. A. Byles in the
aerial surveys. J. G. Mead provided welcome advice
and imparted to me much of his personal knowledge
of the bottlenose dolphin, for which I am grateful.
S. T. Buckland and T. D. Drummer graciously pro-
vided me with Fortran programs for line transect
analysis. My thanks to J. A. Musick, R. A. Byles,
G. P. Scott, and anonymous Fishery Bulletin re-
viewers who made especially helpful comments and
suggestions on earlier drafts of the manuscript. This
study was conducted under contract No. NA-80-
FA-D-0008 to H. M. Austin from the National
Marine Fisheries Service, NOAA, and an assistant-
ship to the author from the Virginia Institute of
Marine Science.
LITERATURE CITED
Blaylock, R. a.
1985. The marine mammals of Virginia. Va. Inst. Mar. Sci.
Sea Grant Ed. Ser. No. 35, 37 p.
Buckland, S. T.
1985. Perpendicular distance models for line transect sam-
pling. Biometrics 41:177-195.
BuRNHAM, K. P., D. R. Anderson, and J. L. Laake.
1980. Estimation of density from line transect sampling of
biological populations. Wildl. Monogr. No. 72, 202 p.
CETAP.
1983. A characterization of marine mammals and turtles in
the mid- and north-Atlantic areas of the U.S. outer continen-
tal shelf. Cetacean and Turtle Assessment Program, Univ.
Rhode Island. Bur. Land Manage. Contract Rep. NTIS
PB83-2158.55, 570 p.
Drummer, T. D., and L. L. McDonald.
1987. Size bias in line transect sampling. Biometrics 43:13-
21.
Essapian, F. S.
1963. Observations on abnormalities of parturition in captive
bottlenosed dolphins, TursiOTps truncatus, and concurrent
behavior of other porpoises. J. Mammal. 44:405-414.
Goodman, L. A.
1960. On the exact variance of products. J. Am. Stat. Assoc.
55:709-713.
Laake, J. L., K. P. Burnham, and D. R. Anderson.
1979. User's manual for program transect. Utah State Univ.
Press, Logan.
Leatherwood, S.
1975. Some observations of feeding behavior of bottle-nosed
dolphins (Tursiops truncatus) in the northern Gulf of Mex-
ico and (Tursiops cf T. gilli) off southern California, Baja
California, and Nayarit, Mexico. Mar. Fish. Rev. 37(9):
10-16.
804
BLAYLOCK: DISTRIBUTION OF THE BOTTLENOSE DOLPHIN
Leatherwood, S., D. K. Caldwell, and H. E. Winn.
1976. WTiales, dolphins, and porpoises of the western North
Atlantic. A guide to their identification. U.S. Dep. Com-
mer., NOAA Tech. Rep. NMFS Circ. 396, 176 p.
McBride, a. F., and H. Kritzler.
1951. Obser\'ations on pregnancy, parturition, and postnatal
behavior of the bottlenose dolphin. J. Mammal. 32:251-266.
Mead, J. G.
1975. Preliminary report on the former net fisheries for Tur-
siops truncatvs in the western North Atlantic. J. Fish. Res.
Board Can. 32:1155-1162.
Mitchell, E.
1975. Porpoise, dolphin, and small whale fisheries of the
world. Status and problems. lUCN Monogr. No. 3, 129 p.
Int. Union Conserv. Nat. Nat. Res., Morges, Switzerland.
Schweder, T.
1977. Point process models for line transect experiments.
In J. R. Barba, F. Brodeau, G. Romier, and B. Van Cutsem
(editors), Recent developments in statistics, p. 221-242.
North-Holland Publ. Co., N.Y.
Seber, G. a. F.
1982. The estimation of animal abundance and related param-
eters. 2d ed. Charles Griffin & Company, Ltd., London
and High Wycombe.
1986. A review of estimating animal abundance. Biometrics
42:267-292.
SOKAL, R. R., AND F. J. ROHLF.
1981. Biometry. 2d ed. W. H. Freeman and Co., San
Franc.
Tavolga, M. M., and F. Essapian.
1957. The behavior of the bottlenose dolphin {Tursiops trun-
catvs): Mating, pregnancy, parturition, and mother-infant
behavior. Zoologica 42:11-31.
TOWNSEND, C. H.
1914. The porpoise in captivity. Zoologica 1:289-299.
True, F. W.
1891. Observations on the life history of the bottlenose por-
poise. Proc. U.S. Nat. Mus. Vol XHI, No. 812. (1890), p.
197-203.
WCrsig, B., and M. WOrsig.
1978. Occurrence and group organization of Atlantic bottle-
nose porpoises (Tursiops truncatus) in an Argentine
bay. Biol. Bull. (Woods Hole) 154:348-359.
805
NOTES
THE FEEDING HABITS OF TWO DEEP
SLOPE SNAPPERS, PRISTIPOMOIDES ZONATUS
AND P. AURICULA, AT PATHFINDER REEF,
MARIANA ARCHIPELAGO
The lutjanid snappers belonging to the genus
Pristipomoides are among the most prized and
valuable commercial fish resources in tropical and
subtropical regions of the Pacific Ocean (Polovina
and Ralston 1987). These fishes normally inhabit
escarpments with high vertical relief. During an in-
tensive bottom fish survey conducted at Pathfinder
Reef in the Mariana Archipelago, two species of
snappers, P. zonatus and P. auricilla, comprised
more than 68% of the total catch (Polovina 1985).
Depth of capture data on these two species demon-
strated overlap in their bathymetric distribution
(Polovina et al. 1985; Ralston and Williams 1988).
Numerous feeding studies have been conducted
on snappers that inhabit shallow (<100 m) water;
however, published information on the diets of deep
slope snapper species in the tropical Pacific is nearly
nonexistent (see review by Parrish 1987). Kami
(1973) noted prey items for four species of Pristi-
pomoides in Guam with total sample sizes ranging
from one to six individuals, and Kluegel (1921)
presented information on the diet oiP.filaTnentosus
in Hawaii based upon four fish. The present paper
examines how two coexisting species, P. zonatus and
P. auricilla, partition food resources. With recent
efforts to expand and develop commercial fisheries
for tropical snappers as well as other deep dwell-
ing bottom fishes, there is an increasing need to
recognize the resources that support these fishes.
The results presented here will therefore be useful
for developing fishery management strategies and
will lead to a better understanding of the ecology
of tropical demersal communities.
Methods
Stomach and spew samples from 106 P. zonattis
and 72 P. auricilla were collected at Pathfinder
Reef during an intensive fishing experiment on
10-19 April and 5-7 May 1984. Located in the Mari-
ana Archipelago at lat. 16°30'N, long. 143°05'E,
Pathfinder Reef is a circular, volcanic pinnacle ris-
FISHERY BULLETIN: VOL. 86, NO. 4, 1988.
ing to about 16 m beneath the surface. The fishes,
ranging from 30.2 to 44.4 cm fork length (x =
38.4 cm, standard deviation = 2.81), were caught
with hook and line on hydraulic-powered gurdies.
The terminal rig and gurdy specifications are
described in Uchida and Uchiyama (1986). While
fishing, the vessel was usually allowed to drift over
the banks and steep slopes, targeting species in the
150-275 m depth range. Our study species were cap-
tured mainly in depths between 180 and 220 m. Fish-
ing was conducted around the entire perimeter of
the bank during daylight hours (Polovina 1986).
Typical of feeding studies conducted on deepwater
species with swimbladders, regurgitation of stomach
contents posed serious problems (Bowman 1986).
The rapid ascent to the surface forced most of the
stomachs to evert and lose an unknown quantity of
the contents. Many food items, however, were
caught in the throat or gill rakers, picked out, and
saved. These items, referred to as spews, and full
stomachs were frozen and returned to the labora-
tory for analysis.
Laboratory procedures for examination of the
samples were similar to those employed by Harrison
et al. (1983). Briefly, food samples were sorted,
counted, and identified to the lowest practical taxon.
The volume of prey items was measured by water
displacement. Fishes were predominantly identified
by osteological and external anatomical characters
and morphometries. Invertebrates were identified
by undigested hard parts and external morpho-
logical features including shells for gastropods and
exoskeletons for crustaceans.
To analyze the data, we used percent frequency
of occurrence to provide a qualitative picture of the
food spectrum and percent volume of prey to de-
scribe relative biomass of prey items (Hyslop 1980).
Because a high percentage of the diets was com-
posed of urochordates, we did not employ any
numerical analysis of the prey items. These pelag-
ic, colonial tunicates (all Pyrosoma spp.) did not
occur in discrete units and were thus difficult to
count.
An interspecific comparison of the two snapper
species requires some measure of diet overlap and
niche breadth. Diet overlap (Q) was computed by
using the formula developed by Morisita (1959) and
807
modified by Horn (1966). Q is formulated as
C, =
2 ^ Pa ■ Pjh
j s
J. p,ir + 2! pj,
h = l
h = l
where p,;, and Pjj, are the biomass proportions of a
prey item h in the diets of species i and j, respec-
tively, and s is the total number of major prey taxa
in the food spectrum. C^ varies from zero, when
there is no overlap between the diets of species i
andy, to one, when all prey items are in equal pro-
portions. Niche breadth (B) for each species was
computed by using the formula developed by Levins
(1968):
B; =
1
crustaceans, must have been captured at the sub-
stratum. Conversely, the 72 food samples from P.
auricilla were composed of 88.2% invertebrates and
11.8% fishes by volume; all were predominantly
pelagic forms. Major invertebrate prey included the
pelagic heteropod Atlanta spp., the pelagic tunicate
Pyrosoma spp., and numerous species of cavolinid
pteropods. Few fishes, most of them small and
unidentifiable, were eaten by this snapper species.
The dietary overlap value (Q) was calculated to
be 0.04. Based on the Langton (1982) convention of
0.00-0.29 as the low overlap, 0.30-0.60 as the
medium overlap, and X).60 as the high overlap, very
little overlap occurred between the diets of the two
snappers at Pathfinder Reef.
The niche breadth values (B) calculated for the
two species indicated P. zonatus had a higher food
breadth (3.82, .s = 10) than P. auricilla (2.05, s =
6), suggesting more food specialization among the
latter. Evenness in the proportion of the diet attrib-
uted to each prey type, however, was similar for the
two species (5„ = 0.38 in P. zonatus; B„ = 0.34 in
P. auricilla).
where again, p^)^ is the biomass proportion of a
prey item h in the diet of species i and s is the total
number of major prey taxa. The breadth values
range from one, when prey items consist only of one
category, to s, when all food items are in equal pro-
portion in the same diet. These breadth values were
then normalized as B„ = Bis, which ranges from a
value of zero, representing the most uneven distribu-
tion of prey composition in the diet, to one, repre-
senting a totally even distribution. For the calcula-
tion of both breadth and overlap indexes, only prey
items identified at least to the family level were
used.
Results
The diets of P. zonatus and P. auricilla differed
considerably from each other (Table 1). By volume,
the 106 food samples from P. zonatus collected for
this study were composed of 59.0% invertebrates
and 41.0% fishes. The dominant invertebrate prey
included the pelagic tunicate Pyrosoma spp. and
galatheid crabs, nearly all Munida japonica. The
most frequently occurring fishes in the diet were the
ophichthid eel, Shultzidia johnstonensis, and the
symphysanodontids Symphysanodon maunaloe and
iS. typv^. In general, P. zonatus preyed heavily upon
benthic organisms. Many of the prey items, such as
echinoderms, octopods, and the various benthic
Discussion
As previously mentioned, a few fragmentary
reports (cf. Parrish 1987) from various localities
have addressed the trophic relationships of any
tropical snapper inhabiting waters deeper than 100
m. He further attributes this lack of information to
the normally remote fishing localities and, most of
all, to the loss of stomach contents through regur-
gitation during capture.
The problem of regurgitation has plagued trophic
studies of demersal fishes, regardless of whether
fishing involved trawls (Mauchline and Gordon 1984;
Clark 1985; Bowman 1986) or hook and line (Kluegel
1921; Forster etal. 1970; Seki 1984). Likewise, most
food samples for our study were salvaged, regurgi-
tated spe wings retained in the mouth, throat, or gill
rakers of the fishes. It is possible that the material
occurring as spews may consist predominantly of
certain prey items more likely retained because of
size or some morphological structure, thereby result-
ing in a biased interpretation of the diet. We never-
theless employ the assumption that what is recov-
ered is representative of the diets at the time of
capture.
Based upon depth of capture information, P.
zonatus and P. auricilla occupy nearly the same
habitat, and considerable spatial overlap occurs in
their foraging zones (Polovina 1986; Ralston and
808
Williams 1988). During submersible dives at John-
ston Atoll, Ralston et al. (1986) verified this cohabi-
tation with visual observations of the two species.
PHstipomoides zonatus were observed between 215
and 250 m (median, 230 m) and P. auricilla between
205 and 295 m (median, 240 m). Our dietary anal-
ysis suggests that these two demersal species par-
tition food resources by selecting prey from differ-
ent microhabitats. Pristipomoides zonatus is best
described as a demersal carnivore, with its diet con-
taining benthic and demersal invertebrates together
with tunicates and small fishes. In contrast, P.
auricilla feeds primarily on large pelagic plankton,
as evidenced by the abundance of heteropods,
pteropods, and tunicates among the prey items.
Kami (1973) also found tunicates {Pyrosoma spp.)
in four of the five P. auricilla food samples in Guam.
The differences in diet composition are consistent
with the suggestion, based upon catch rates and the
taking of a baited hook, that P. zonatus is the more
aggressive predator of the two species (Polovina
1986).
Although our study revealed little overlap be-
tween the diets of P. zonatus and P. auricilla, the
common occurrence of Pyrosoma spp. in the diets
of both species seems significant. Similarly, these
tunicates were also found among the stomach con-
tents of three other congeneric species (Kami 1973;
Parrish 1987). Kashkina (1987) reported intensive
predation on pyrosomes, as well as salps, among
Table 1.— Diet composition of Pristipomoides zonatus and P. auricilla at Pathfinder Reef,
f^/lariana Archiipelago.
Pristipomoides zonatus
Pristipomoides auricilla
(N =
= 106)
{N
= 72)
Species
% volume
% frequency
% volume
% frequency
Invertebrates
Hydrozoa
Siphonophora
—
—
3.2
20.8
Ctenophora
0.3
0.9
—
—
Polychaeta
0.3
0.9
5.4
15.3
Gastropoda
0.1
0.9
0.3
4.2
Atlantidae
—
—
0.9
12.5
Cavolinidae
0.2
4.7
38.1
72.2
Cephalopoda
0.3
0.9
7.4
2.8
Teuthoidea
—
—
0.2
1.4
Octopoda
0.1
0.9
—
—
Crustacea
2.6
10.4
5.5
44.4
Stomatopoda
0.3
0.9
0.3
1.4
Euphausiacea
Euphausiidae
—
—
0.7
4.2
Decapoda
Caridea
0.5
2.8
0.3
2.8
Pandalidae
1.2
2.8
—
—
Palinura
—
—
0.7
1.4
Anomura
Galatheidae
19.6
28.3
—
—
Brachyura
17.1
13.2
0.2
1.4
Echlnodermata
Ophiuroidea
0.1
0.9
—
Tunicata
Pyrosomatidae
16.3
34.9
25.0
37.5
Fishes
Osteichthyes
(Unid. fishes)
16.3
22.6
8.7
16.7
Anguilliformes
8.2
7.6
—
Ophichthidae
8.6
10.4
—
—
Myctophiformes
—
—
2.5
1.4
Ophidiiformes
Ophidiidae
0.1
0.9
—
—
Perciformes
Serranidae
4.0
1.9
Symphysanodontidae
3.1
5.7
—
—
Chaetodontidae
0.1
0.9
—
Gempylidae
—
—
0.6
2.8
Tetraodontiformes
Balistidae
0.6
0.9
—
—
809
numerous pelagic and demersal fish species. Origi-
nally thought to be of little nutritional value, these
tunicates have been found to contain filtered con-
centrations of phytoplankton and microzooplankton,
thereby elevating the preys' food value. Regional
and highly localized oceanographic processes will
affect the distribution of such potential planktonic
prey, and exploitation of these resources may influ-
ence the local distribution of predator species (Brock
and Chamberlain 1968; Bray 1981). Deepwater
snappers are most abundant on slopes of upcurrent
exposure and near underwater headlands at John-
ston Atoll (Ralston et al. 1986). These abundance
patterns were attributed to planktonic concentra-
tions created by mesoscale oceanographic processes
as noted with other fishes on bank or slope habi-
tats (Isaacs and Schwartzlose 1965; Pereyra et al.
1969).
Polovina (1986) suggested that fishing may selec-
tively deplete one species, such as P. zonatus,
preferentially over another (P. auricilla) and there-
by alter the species composition in a given locality.
With the small degree of dietary overlap between
the two species in this study, such selective removal
of P. zonatus will decrease predation pressure on
the demersal prey resources of this species and, as
evidenced in Larson (1980), may ultimately allow
greater niche breadth for coexisting predator
species.
In conclusion, this study has provided qualitative
insight into dietary habits of two sympatric, deep-
water species and has permitted some inferences
regarding their ecology. Much more study is needed
to comprehend fully the role of these predators in
the ecosystem. Logistical constraints, including the
problem of regurgitation, will continue to make
quantitative assessments of diet a difficult task un-
til new capture methodologies are developed. How-
ever, the rewards in improved understanding of
deepwater ecology and increased ability to manage
these valuable stocks suggest that the efforts will
be justified.
Acknowledgments
We thank G. W. Boehlert, R. E. Bowman, C. B.
Grimes, J. D. Parrish, J. J. Polovina, S. Ralston, and
the two anonymous reviewers for their helpful com-
ments on various drafts of the manuscript. This
paper is a result of the Resource Assessment Inves-
tigation of the Mariana Archipelago at the South-
west Fisheries Center Honolulu Laboratory, Nation-
al Marine Fisheries Service, NOAA.
Literature Cited
Bowman, R. E.
1986. pjffect of reffurptation on stomach content data of
marine fishes. Environ. Biol. Fishes 10:171-181.
Bray, R. N.
1981. Influence of water currents and zooplankton densities
on daily foraging movements of blacksmith, Chnmiia punrti-
pinniH, a planktivorous reef fish. Fish. Bull., U.S. 78:
829-841.
Brock, V. E., and T. C. Chamberlain.
1968. A geological and ecological reconnaissance off western
Oahu, Hawaii, principally by means of the research sub-
marine "Asherah." Pac. Sci. 22:373-394.
Clark, M. R.
1985. The food and feeding of seven fish species from the
Campbell Plateau, New Zealand. N.Z. J. Mar. Freshwater
Res. 19:339-363.
FoRSTER, G. R., J. R. Badcock, M. R. Longbottom, N. R.
Merrett, and K. S. Thomson.
1970. Results of the Royal Society Indian Ocean Deep Slope
Fishing Expedition, 1969. Proc. R. Soc. Lend., B Biol. Sci.
175:367-404.
Harrison, C. S., T. S. Hida, and M. P. Seki.
1983. Hawaiian seabird feeding ecology. Wildl. Monogr. 85,
71 p.
Horn, H. S.
1966. Measurement of "overlap" in comparative ecological
studies. Am. Nat. 100:419-424.
Hyslop, E. J.
1980. Stomach contents analysis— a review of methods and
their application. J. Fish Biol. 17:411-429.
Isaacs, J. D., and R. A. Schwartzlose.
1965. Migrant sound scatterers: Interaction with the sea
floor. Science 150:1810-1813.
Kami, H. T.
1973. The Pristipomoides (Pisces: Lutjanidae) of Guam with
notes on their biology. Micronesica 9:97-117.
Kashkina, a. a.
1987. Feeding of fishes on salps (Tunicata, Thaliacea). J.
Ichthyol. 26:57-64.
Kluegel, E.
1921. The food of Hawaiian food fishes. M.S. Thesis, Univ.
Hawaii, Honolulu, HI, 18 p.
Langton, R. W.
1982. Diet overlap between Atlantic cod, Gadiis morhua,
silver hake, Merluccius bilinearis, and fifteen other north-
west Atlantic finfish. Fish. Bull., U.S. 80:745-759.
Larson, R. J.
1980. Competition, habitat selection, and the bathymetric
segregation of two rockfish (Sebastes) species. Ecol.
Monogr. 50:221-239.
Levins, R.
1968. Evolution in changing environments. Princeton Univ.
Press, Princeton, 120 p.
Mauchline, J., AND J. D. M. Gordon.
1984. Feeding and bathymetric distribution of the gadoid and
morid fish of the Rockall Trough. J. Mar. Biol. Assoc, U.K.
64:657-665.
MORISITA, M.
1959. Measuring of interspecific association and similarity
between communities. Mem. Fac. Sci., Kyushu Univ., Ser.
E Biol. 3:65-80.
Parrish, J. D.
1987. The trophic biology of snappers and groupers, /w J. J.
810
Polovina and S. Ralston (editors). Tropical snappers and
groupers: Biologj' and fisheries management, p. 405-463.
Westview Press, Boulder and Lond.
Pereyra, W. T., W. G. Pearcy, and F. E. Carvey, Jr.
1969. Sebastodesflavidjis, a shelf rockfish feeding on meso-
pelagic fauna, with consideration of the ecological implica-
tions. J. Fish. Res. Board Can. 26:2211-2215.
Polovina, J. J.
1985. Variation in catch rates and species composition in
handline catches of deepwater snappers and groupers in the
Mariana Archipelago. Proc. Fifth Int. Coral Reef Congr.,
Tahiti 5:515-520.
1986. A variable catchabOity version of the Leslie model with
application to an intensive fishing experiment on a multi-
species stock. Fish. Bull., U.S. 84:423-428.
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.
Polovina, J. J., and S. Ralston (editors).
1987. Tropical snappers and groupers: Biology and fisheries
management. Westview Press, Boulder and Lond., 659 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. 84:141-154.
Ralston, S., and H. A. Willums.
1988. Depth distributions, growth, and mortality of deep
slope fishes from the Mariana Archipelago. U.S. Dep.
Commer., NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-
SWFC-113, 47 p.
Seki, M. p.
1984. The food and feeding habits of the grouper,
Epinephelus quemvs Seale 1901, in the Northwestern
Hawaiian Islands. In R. W. Grigg and K. Y. Tanoue
(editors). Proceedings of the Second Sj-mposium on Resource
Investigations in the Northwestern Hawaiian Islands, vol.
2, May 25-27, 1983, University of Hawaii, Honolulu, Hawaii,
p. 179-191. LTS;IHI-SEAGRANT-MR-84-01.
Uchida, R. N., and J. H. Uchiyama (editors).
1986. Fisher>' atlas of the Northwestern Hawaiian Islands.
U.S. Dep. Commer., NOAA Tech. Rep. NMFS 38, 142 p.
Michael P. Seki
Southwest Fisheries Center Honolulu
Laboratory
National Marine Fisheries Service, NOAA
2570 Dole St.
Honolulu, HI 96822-2396
Michael W. Callahan
Hawaii Cooperative Fishery Research Unit
University of Hawaii
Honolulu, HI 96822
SEASONALITY AND DEPTH DISTRIBUTION
OF LARVAL FISHES IN
THE NORTHERN GULF OF MEXICO
ABOVE LATITUDE 26°00 N
Justification and Methods
Information on early life stages of fishes is impor-
tant for a better understanding of recruitment pro-
cesses and for the proper management of fisheries.
Knowledge of seasonal occurrence and depth dis-
tribution of larval fishes is essential for planning and
scheduling ichthyoplankton cruises (Colton et al.
1979) and juvenile surveys, so that sampling for
target species can be concentrated during periods
and at depths where effort will be most effective
(Saville 1964). In addition, knowledge of the seasonal
occurrence of early life stages is an important aid
in identifying larvae. Because eggs and yolk-sac lar-
vae are planktonic for only a relatively few days
after being spawned, the presence and distribution
of early life stages also suggests proximity of adult
spawning concentrations (Houde 1974), aiding the
definition of spawning areas and seasonal spawn-
ing migrations of adults. Since some commercial and
recreational fisheries (e.g., red drum, Sciaenops
ocellatus, and black drum, Pogonias cromis) exploit
spawning aggregations, encroachment on these
aggregations could have an adverse impact on the
fishery.
Colton et al. (1979) summarized larval seasonality
data and spawning areas for marine continental
shelf fishes between Nova Scotia and North Caro-
lina; whereas, Herrema et al. (1985) inferred spawn-
ing seasons of coastal fishes off eastern Florida
based on examination of enlarged gonads. The sea-
sonal occurrence of larvae of many species from the
northern Gulf of Mexico (GOMEX), however, is not
well documented. The northern GOMEX is herein
defined as waters north of lat. 26°00'N; this area
approximates the U.S. Fishery Conservation (i.e..
Exclusive Economic) Zone. For discussion, the study
area was subdivided into three regions (Fig. 1) based
on longitude as follows: eastern GOMEX (waters
east of long. 86°00'W), central GOMEX (those
between 86°00'W and 94°00'W), and western
GOMEX (waters west of 94°00'W). Seasonality data
are scattered throughout the grey literature, and
many studies have focused on either select taxa or
are limited in spatial or temporal coverage. The most
comprehensive studies of the larval ichthyofauna
community in the northern GOMEX were those of
Houde et al. (1979) from continental shelf waters
FISHERY BULLETIN: VOL. 86, NO. 4. 1988.
811
Figure 1.— Location of the northern Gulf of Mexico (GOMEX) and its subregions. Studies conducted in waters east of long. 86°00'W
are considered northeastern GOMEX, those between 86°00'W and 94°00'W are north-central GOMEX, and west of 94°00'W are
northwestern GOMEX.
off west Florida, and Finucane et al. (1977, 1979b)
from the south Texas outer continental shelf. Ditty
(1986) presented data on seasonality of larval fishes
from neritic continental shelf waters off Louisiana
that included a discussion of peak seasonal occur-
rence of abundant taxa; however, those data were
limited in areal coverage, and there were significant
gaps in the temporal occurrence of many taxa.
Therefore, data on larvae of approximately 200
coastal and oceanic fishes from 61 families were
compiled from unpublished plankton surveys\ as
well as other published studies from throughout the
northern GOMEX, to further clarify the seasonal-
ity and peak seasonal occurrence (Table 1), and
depth distribution (Table 2) of larval fishes from this
area. Since the occurrence of early developmental
stages of fishes suggests recent spawning (Colton
et al. 1979; Ruple 1984), only studies that primari-
ly collected larvae <10 mm SL were used. Because
the taxonomy of larvae of many northern GOMEX
fishes (e.g., most exocoetids, blennies, and gobies)
are poorly understood, no attempt was made to com-
pile seasonality or depth distribution data for these
taxa or for anguilliform, myctophiform, or salmoni-
form fishes; other poorly understood taxa (e.g., en-
^Louisiana Department of Wildlife and Fisheries. P.O. Box
98000, Baton Rouge, LA 70898-9000.
graulids and cynoglossids) were assigned only to
genus.
Discussion
Although not all northern GOMEX ichthyoplank-
ton studies were readily comparable because of
differences in gear and tow type, plankton net mesh
size, and seasonal and areal coverage (Table 3), few
major discrepancies in either seasonality or peak
seasonal occurrence of larvae resulted from such dif-
ferences. For example, Pearson (1929) suggested a
secondary spawning from late July to November for
black drum off Texas, but these dates have not been
reported from elsewhere in the literature. Likewise,
Finucane (1976) reported round scad, Decapterus
punctatus, <6 mm SL in December-January, but
based on sampling at the same location in subse-
quent years (Finucane et al. 1977, 1979b) and the
seasonal occurrence of this species as suggested by
others (Table 1), these may have been larvae of the
rough scad, Trachurus lathami, rather than those
of round scad. Houde et al. (1979) collected larvae
of Spanish sardine, Sardinella aurita, during all
months of their Florida continental shelf survey;
however, their winter occurrences of Spanish sar-
dine were only from the southern part of the survey
area (i.e., <26°00'N). This may account for the dis-
812
Table 1.— Seasonality (X) and peak seasonal occurrence (*) of larval fishes (<10 mm SL) in the northern Gulf of IVIexico (GOfVlEX) above
lat. 26°00 N. Sources 1-31 and 75 are studies conducted in the northeastern GOMEX (waters east of long. 86°00 W); 33-59 and 74,
north-central GOMEX (those between 86°00 W and 94°00 W); 61-73, northwestern GOIVIEX (waters west of 94°00 W); and 23, 25, 32,
48, 60, and 76-80 are Gulf-wide studies. Assignment of taxa to Families follows Robins et al. (1980) except Epigonus sp. and Sym-
physanodon typus which follows Johnson (1984). Numbers in Source, see end of table.
Family
Taxa
J
F
M
A
M
J
J
A
s
0
N
D
Source
Elopidae
Elops saurus
X
X
X
X
X
X
X
X
X
1,2,41,54,61,63,66
Albulidae
Megalops atlanticus
Albula vulpes
X
X
X
X
X
2,6
54,66
Clupeidae
Brevoortia spp.
X
X
X
X
X
X
X
X
1-3,61-63
B. patronus
X
X
X
X
X
33,40-42,46,47,49,53,55,
57-59,61-63,66,69,74
B. smithi
X
X
X
X
5,55
Etrumeus teres
*
X
X
X
X
X
2,5,12,29,40,41,43,50-52
59,61,63,66,68
Harengula jaguana
X
X
*
•
*
X
X
X
1-3,5,30,40,46,47,50-53,
56-59,61-63,66,68,69
Opisthonema oglinum
X
X
•
*
*
*
X
X
X
2,5,31,40,46,50-52,
56-59,61-63,68,69
Sardinella auhta
X
X
X
X
X
X
X
X
X
2,22,50-52,56,59,61,
63,68
Engraulidae
Anchoa spp.
X
X
*
*
*
*
X
X
X
1-3,23,25,40,41,46,47,
53,56-59,61-64,66,68,69
Anchoviella 1 Engraulis
X
X
X
X
X
X
X
X
X
X
X
X
40,41,46,61-63,68
Gobiesocidae
Gobiesox strumosus
X
X
X
X
X
X
X
X
X
1,3,40,41,46,47,56,58,
59,66,68
Ogcocephalidae
Ogcocephalus sp.
X
X
X
68
Bregmacerotidae
Bregmaceros atlanticus
X
X
X
X
X
X
X
X
X
X
X
X
2,4,40,59,61-64,68
B. canton
X
X
X
X
X
X
X
*
*
*
*
X
2,4,40,59.64
B. houdei
X
X
X
X
X
*
*
*
*
2,4,64
B. macclellandi
X
X
X
X
X
X
X
2,4
Gadidae
Uroptiycis spp.
U. cirrata
U. florldanus
X
X
X
X
X
X
X
X
2,40,41,57.59,61-63,68
64
63
Ophidiidae
Brotula barbata
X
X
X
40,59,64
Lepophidium spp.
X
X
X
X
X
X
X
X
X
X
X
X
18,40,59,63,68
Ophidian spp.
X
X
X
X
X
X
X
X
X
40,59,61-63,68
O. selenops
X
X
X
X
X
X
X
X
X
X
18,40
Ophidian Type 1
X
X
X
X
X
X
X
X
X
X
18
Ophidian Type 2
X
X
X
X
X
X
X
X
X
X
18
Otaphidium omostigmum
X
X
X
X
X
X
X
X
X
X
18,68
Carapidae
Carapus bermudensis
X
X
X
X
X
X
X
X
X
2,63
Echiadan sp.
X
X
X
X
X
X
X
X
X
X
2,64
Exocoetidae
Hemiramphus balaa
H. brasiliensis
X
X
X
X
X
X
2
2
Hyparhamphus unifasciatus
X
X
X
X
X
X
X
X
2,40,69
Belonidae
Ablennes hians
X
X
62
Atherinidae
Membras martinica
X
X
X
X
2,40,41,46,47,58,69
Holocentridae
Adioryx vexillarius
Holocentrus sp.
X
X
2
40
Caproidae
Antigania spp.
A. capros
X
X
X
X
X
2
2
Trachipteridae
Trachipterus sp.
X
59
Fistulariidae
Fistulaha sp.
X
2
Centriscidae
Macrorhamphosus scalapax
X
X
2,40,41,59,63,64
Serranidae
Anthias spp.
X
X
X
X
X
X
X
X
X
X
X
X
2,41,61,63,64
Centrapristis spp.
X
X
X
X
X
X
X
X
X
X
41,53,59,61-64,68
C. striata
X
X
X
X
X
X
X
2,7
Diplectrum spp.
X
X
X
X
*
*
*
*
X
X
X
X
53,59,61-64,68
D. farmasum
X
X
X
♦
*
X
X
•
*
X
X
2,7
Epinephelus spp.
X
X
X
X
X
X
X
X
X
2,64,68
Goniaplectrus hispanus
X
64
Hemanthias aureorubens
X
X
X
X
X
X
X
2
H. leptus
X
2,40
H. vivanus
X
X
X
X
X
X
X
X
X
X
X
X
2,61,63,64
Holanthias martinicensis
X
X
64
Lioprapoma spp.
X
X
X
X
X
X
X
X
X
2,61,63,64
Myctoperca spp.
X
64
Plectranthias garrupellus
X
2
Serraniculus pumilio
X
*
•
*
*
X
X
2,7,40,61-63
Serranus spp.
X
X
X
X
X
X
X
X
X
X
X
2,63,64,68
813
Table ^.— Continued.
Family
Taxa
J
F
M
A
M
J
J
A
s
0
N
D Source
Grammistidae
Pseudogramma gregoryi
X
X
X
2
Rypticus spp.
X
X
X
X
•
*
X
X
2,7,40,53,61-63,68
R. saponaceus
X
X
X
X
X
61-63,68
Priacanthidae
Phacanthus spp.
P. arenatus
P. cruentatus
Pristigenys alta
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
2,32,64
64
64
2,32
Apogonidae
Apogon sp.
X
41
Epigonidae
Epigonus sp.
X
64
Malacanthidae
Caulolatilus spp.
C. cyanops
X
X
X
X
X
X
X
X
X
X
X
X
X
X 2,61,63,64
2
Pomatomidae
Pomatomus saltatrix
X
X
X
X 40,59,61,63,64,67
Rachycentridae
Rachycentron canadum
X
X
X
X
X
X
2,37,62,63
Carangidae
Alectis crinitus
X
X
X
X
48,62
Caranx sp.
X
X
X
X
X
X
X
X
X
X
X
X 2,8,61,63,68,69
C. crysos
X
X
X
*
*
•
X
X
X
2,8,24,41,50-52,61
C. hippos/latus
X
X
X
X
X
32,48,60,69
Chloroscombrus clirysurus
X
X
*
X
1-3,8,40,41,46,47,50-53,
56-59
Decapterus punctatus
X
*
*
*
*
*
X
X
2,8,10,24,40,50-53,59,
61-63,68
Elagatis bipinnulata
X
X
X
X
X
X
2,56,59
Hemicaranx amblyrhynchus
X
X
X
X
X
X
48,66
Oligoplites saurus
X
X
*
*
X
X
X
1-3,8,10,32,40,41,46,48,
53,56,59-61,68,69
Selar crumenopthalmus
X
X
X
X
X
X
X
32,40,48,53,59,60
Selene spp.
X
X
X
X
2,8,10,32,40,41,48,53,
59-61,63,68
Seriola spp.
X
X
X
X
X
X
X
X
X
X
X
X 2,8,32,40,48,56,59,60,
63,68
Trachinotus spp.
X
X
X
X
X
X
2,8,32,47,48,61,63
T. carolinus
X
X
X
X
32,41,48,60,66
T. falcatus/goodei
X
X
X
48,60
Trachurus lathami
*
*
X
X
X
X
X 2,8,23,40,41,50-52,59,
61-63,68
Coryphaenidae
Coryphaena spp.
X
X
X
X
X
X
X
2,41,61,64
C. equisetis
X
X
X
X
X
X
X
2,17,23,25,40
C. hippurus
X
X
X
X
X
X
X
X
X
X
2,17,23,25
Lutjanidae
Etelis oculatus
X
X
X
X
X
62,63
Lutjanus spp.
X
X
*
*
*
X
X
X
2,40,53,59,61,63,64
L. campechanus
X
X
X
X
X
X
X
62,64,73
Pristipomoides aquilonahs
X
X
X
X
X
X
X
X
2,64
Rhomboplltes aurorubens
X
X
X
X
X
X
X
X
2,64
Acropomatidae
Symphysanodon typus
X
X
X
2
Gerreidae
Eucinostomus spp.
X
X
X
X
X
X
X
X
40,47,53,59,61-63,66
Haemulidae
Haemulon spp.
X
X
X
X
61,63
Orthopristis chrysoptera
X
X
*
X
X
2,40,46,47,59,66,69
Sparidae
Archosargus probatocephalus
X
*
*
X
1,2,40,47,48,53,59,60,
69,70
DIplodus holbrookl
X
X
X
X
X
2,68
Lagodon rhomboides
X
X
X
X
* 1-3,11,40,41,46-48,53,
56,57,59,60,62,63,66,68,
69
2,68
Pagrus pagrus
X
X
X
Stenotomus caprlnus
X
X 48,61
Sciaenidae
Bairdiella ctirysoura
X
*
*
X
X
1-3,32,39-41,46,47,53,
57-59,62,63,66,69,71
Cynoscion arenarius
X
*
X
X
X
X
1-3,32,34,35,39,40,46,
53,56-59,61,63,68,69,71
C. nebulosus
X
X
*
*
*
*
X
X
1-3,32,35,39-41,46,47,
53,56-59,68-72
C. nothius
X
X
X
X
*
X
2,32,39-41,46,53,56,59,
61-63,68
Larimus fasciatus
X
X
X
X
X
X
X
34,39-41,46,53,56,58,59,
61-63,69
Leiostomus xanthurus
X
X
X
X
X
• 1-3,34,39-41,44,46,47,
53,56-59,61-63,66,
68-71,74
814
Table 1 . — Continued.
Family
Taxa
J
F
M
A
M
J
J
A
s
o
N
D
Source
Menticirrhus spp.
X
X
X
X
X
X
X
X
X
X
X
1-3,32,34,39-41,46,47,
53,56-59,61-63,66,68,69
Mlcropogonias undulatus
*
X
X
X
X
*
*
*
1-3,34,39-41,44,46,47,
53.56-59,61-63,66,
68-71,74
Pogonias cromis
X
*
X
X
1-3,34,39,40,46,47,56,
58,59,61,63,66,68-70,72
Sciaenops ocellatus
X
*
*
X
1,28,39,40,46,47,53,56,
58,59,61,63,66,69,70,72
Stellifer lanceolatus
X
X
X
X
X
X
X
32,39-41,46,53,56-59,
66,69
Mullidae
Mullus auratus
X
2
Kyphosidae
Kyphosus spp.
X
X
X
X
X
X
X
2,61,63
Ephippidae
Chaetodipterus faber
X
X
X
40,41,46,47,53,56,57,59,
61-64,66,68
Pomacentridae
Abudefduf saxatills
X
63
Mugilidae
Mugil spp.
X
X
X
X
X
X
X
X
X
X
X
X
2,61,63,68
M. cephalus
X
X
X
X
40,41,44,47,53,57,59,
61-64,69,71
M. curema
X
*
*
*
X
40.46,47,57,59,63,64,68
Sphyraenidae
Sphyraena spp.
X
X
*
*
•
X
X
X
40,46,53,57,59,61-64,66,
68
2,64
S. barracuda
X
X
X
X
X
S. borealis
X
X
X
X
2,61,68
S. guachancho
X
X
X
X
X
2,61,63
Polynemidae
Polydactylus octonemus
X
64
Opistognathidae
Lonchopisthus macrognathus
Gnathagnus egregius
X
X
X
68
63
Callionymidae
Callionymus spp.
C. bairdi
C. pauciradiatus
X
X
X
X
X
X
61,68
59
40,59
Microdesmidae
Microdesmus spp.
X
X
X
X
X
X
X
X
X
2,40,46,57,59,61-64,68
Acanthuridae
Acanthurus spp.
X
X
X
X
X
X
X
2,64
Gempylidae
Gempylus serpens
Nealotus tripes
Nesiarchus nasutus
X
X
X
X
X
X
X
2,61
2,41
2,59
Diplospinous multistriatus
X
X
X
X
X
X
X
X
X
X
2,40,41,59,61,63
Trichiuridae
Lepidopus sp.
X
61
Trichiurus lepturus
X
X
X
X
X
X
X
X
X
X
X
X
40,41,56,59,61,63,68
Scombridae
Acanthocybium solanderi
X
X
X
X
X
X
2,25
Auxis spp.
X
X
X
X
*
*
*
*
*
X
X
2,21,23,25,32,40,59,61,
63,64,68,79
Euthynnus alletteratus
X
*
*
*
*
*
X
X
2,21,23,25,32,40,53,57,
59,61-64,68,79
E. pelamis
X
X
X
X
X
X
X
2,21,23,25,32,40,61,63,
64,79
Sarda sarda
X
53
Scomber japonicus
X
X
X
X
X
X
X
X
X
2,32,40,61,63,68
Scomberomorous cavalla
X
X
X
*
X
X
2,21,23,25,27,32,40,
50-52,59,61-65,68
S. maculatus
X
X
X
X
*
*
X
2,23,25,27,32,40,50-53,
56,57,59,61-63,65,66,68
T. albacares
X
*
*
*
X
X
40,63
T. atlanticus
X
•
•
X
X
X
X
2,21,23,25,32,40,61,63,
64,79
T. thynnus
X
X
X
2,21,23,25,32,40,61,64,
68,76-78,80
Xiphiidae
Xiphias gladlus
X
X
X
*
*
•
X
X
X
X
X
X
19,23,25,64,79
Istiophoridae
Istiophorus sp.
X
X
X
X
X
2,16,23,25,61-64
Stromateidae
Ariomma spp.
X
X
X
X
X
X
X
X
X
X
2,40,41,57,59
Cublceps pauciradiatus
X
X
X
X
X
X
X
X
X
X
X
X
2,40,41,59,61,63,64
Hyperoglyphe byttiites
X
X
X
X
36
Nomeus gronovii
X
X
40,41
Peprilus paru
X
X
*
*
*
X
X
X
2,32,38,40,41,53,56,57,
59,61-63,66
P. burti
*
*
*
X
X
X
X
X
X
X
•
•
2,32,38,40,41,53,57,59,
61-64,66,68
Psenes spp.
X
X
X
2
P. cyanoptirys
X
2
815
Table ^ .—Continued.
Family
Taxa
J
F
M
A
M
J
J
A
S
O
N
D
Source
P. pellucidus
X
X
X
X
X
2
Tetragonurus atlanticus
X
59
Scorpaenidae
Scorpaena spp.
X
X
X
X
X
X
X
X
X
X
X
X
2,40,41,53,56,57,61-63,
66,68
Triglidae
Peristedion spp.
X
2
Prionotus spp.
X
X
X
X
X
X
X
X
X
X
X
X
1-3,40,41,46,53,56-59,
61-64,66,68,69
Dactylopteridae
Dactylopterus volitans
X
2,40
Bothidae
Bothus spp.
X
X
X
X
X
X
X
X
X
X
32,40,41,45,59,61-64,68
8. ocellatus
X
X
*
*
*
X
X
X
X
X
X
X
61-63,68
B. roblnsi
X
X
X
•
*
•
*
*
*
*
2,9
Citharichthys spp.
X
X
X
X
X
X
X
X
X
X
X
X
32,40,46,56-59,63,66,68,
69
2,9,32,40
C. cornutus
X
X
X
•
*
•
•
X
X
X
C. gymnorhinus
X
X
X
X
X
X
X
X
X
X
2,9,32,40,59
C. macrops
X
X
X
X
*
X
X
X
X
X
X
2,9,26
C. spilopterus
X
X
X
X
X
X
X
X
X
X
X
2,26,40,41,45,53,59
Cyclopsetta spp.
X
X
X
X
X
*
*
*
X
X
32,40,41,45,61-63,68
C. fimbriata
X
X
X
*
X
X
*
X
X
2,9,62,63
Engyophrys senta
X
X
X
X
X
X
2,20,32,40,41,45,53,59
Etropus crossotus
X
X
X
X
X
X
X
X
26,32,40,41,45,46,56,57,
59
2,9,26
E. rimosus
*
*
*
*
X
X
X
X
X
X
•
•
Monolene sessilicauda
X
X
X
X
X
X
X
X
2,9,13,32,40,61,62,68
Paralichthys spp.
*
X
X
X
X
X
1,2,40,41,45-47, 53,57,
59,61-63,66,68-70
Syacium spp.
X
X
*
*
*
X
X
X
X
32,40,41,45,53,59,61-64,
68
32,40,57,61-63,66
S. gunteri
X
*
*
X
X
X
X
S. papillosum
X
*
*
*
*
X
X
2,9,14,32,40
Trichopsetta ventralis
X
X
X
X
X
X
2,9,15,32,40,45
Soleidae
Achirus lineatus
Gymnachirus sp.
G. melas
X
X
X
X
X
X
*
X
*
*
X
X
*
X
X
X
X
2,3,40,41,45-47,53,56,
57,59,75
61,63
2
Trinectes maculatus
X
X
*
X
X
X
1,40,46,53,58,66
Cynoglossidae
Symphurus spp.
X
X
X
X
X
X
X
X
X
X
X
X
2,3,40,45-47,53,56-59,
61-64,66,68,69
Balistidae
Balistes sp.
Monacanthus sp.
M. hispidus
M. setifer
X
X
X
X
X
X
X
X
X
X
X
61
61,62,68
46,53,69
40
Ostraciidae
Lactophrys sp.
X
X
X
66,68
Tetraodontidae
Lagocephalus laevigatus
X
X
68
Sphoeroides spp.
X
X
X
X
X
X
X
X
X
X
X
X
1,2,40,41,46,53,56,57,
59,61-64,66,68,69
Molidae
Ranzania laevis
X
64
'Blanchet 1979.
2' Juarez 1976.
"'Ditty and Truesdale
1984.
siRinucane et al. 1977.
2Houde et al. 1979.
22Houde 1976.
^Fore 1970.
"Finucane et al. 1979a.
^Phillips et al. 1977.
23Kelley et al. 1986.
"Fore 1971.
"Finucane et al. 1979b.
"Houde 1981.
s-iMontolio 1976.
"Fruge 1977.
e^McGowan 1985.
sHoude and Fore 1973. ^sRjchards et al. 1984.
«Kuhr
1 1979
1.
"McEachran et al. 1980.
«Smith 1980.
2«Tucker 1978.
«Rupli
e 1984.
««Hoese 1965.
'Houde 1982.
2'Dwinell and Futch
1973,
"'Sabins 1973.
«'Barger et al. 1980.
sLeak 1981.
28peters and McMlchael 1987.
"OSEAMAP 1985.
68Finucane 1976.
9Dowd 1978.
23Houde 1977a.
i^Shaw et al
. 1985.
e^Allshouse 1983.
'"Aprieto 1974.
30Houde 1977b.
50Shaw and Drull
inger
1985
'"King 1971.
"Caldwell 1957.
3'Houde 1977c.
s'Shaw and Drull
inger
1986
"Guillen and Landry 1981.
'2Houde 1973.
32SEAMAP 1983.
"Shaw et al
. 1987.
"Pearson 1929.
'3Futch 1971.
33Christmas and Waller 1975.
53Stuck and
Perry 1982.
"Collins et al 1980
'"Futch and Hoff 1971. 34Cowan 1985.
"Thompson
and
Deegan 1982.
"Sogard et al. 1987.
'SFutch 1977.
35Daniels 1977.
ssTurner 1969.
"Houde et al (1970, cited in
'«Gehringer 1957.
36Dawson 1971a.
"Vecchione
et al
. 1982.
Phillips et al. 1977).
''Gibbs and Collette 1959. ^'Dawson 1971b.
s'Walker 1978
'^Richards 1976.
isGordon 1982.
38Ditty 1981.
sswilliams 1983.
"Richards 1977.
'^Grall et al. 1983.
39Ditty 1984.
59LDWF 1983 uni
3Ubl.
data.
'SRichards and Potthoff 1980a.
20Hensley 1977.
""Ditty 1986.
60SEAMAP 1984.
"Richards and Potthotf 1980b.
soMcGowan and Richards 1986.
816
Table 2.— Primary depth distribution of larvae (<10 mm SL) of some abundant taxa of fishes from the
northern Gulf of Mexico above lat. 26°00'N. Depths are those reported In the literature at which >75%
of larvae were collected. Asterisk (') indicates larvae are estuarine-dependent.
Depth (m)
Taxa
<25
<50
<100 50-200
>150 Source
Chloroscombrus chrysurus
X
1,2,3,4,5,6,8,10
Orthopristis chrysoptera
X
1
Cynoscion nebulosus '
X
1,9,10
C. arenarius
X
1,2,9,10,11
Pogonias cromis '
X
11
Archosargus probatocephalus '
X
1
Chaetodipterus faber
X
2
Peprilus paru
X
1,25,26
Anchoa spp.
X
1
Harengula jaguana
X
1,4,5,6,13
Opisthonema oglinum
X
1,4,5,6,10,29
Brevoortia patronus '
X
14,15,16,17,32
Sardinella aurita
X
1,4,6,18
Diplectrum formosum
X
1
Serranlculus pumilio
X
1
Centropristis striata
X
1
Lagodon rhomboides '
X
1
Leiostomus xanthurus '
X
1,11,19,32
Micropogonias undulatus *
X
1,11,19,20,32
Scomberomorus maculatus
X
1,4,21,22,23,24
Decapterus punctatus
X
1,3,4,6
Peprilus burti
X
25,26
Etrumeus teres
X
1,5,6,11,12,27,28,30
Caranx crysos
X
1,3,4,6
Trachurus lathami
X
1,2,3,5,30
Hemanthias vivanus
X
1
Auxis sp.
X
1,21,22,23
Euthynnus alletteratus
X
1,21,22,30
Scomberomorus cavalla
X
20,23,24
Lutjanus campechanus
X
31
Xiphias gladius
X 7,21,??
Istiophorus spp.
X 21,22
Euthynnus pelamis
X 1,21,22,23
'Houde et al. 1979.
'2Houde 1977b.
23Dwinell and Futch 1973.
2Ditty and Truesdale 1984.
'3Houde 1977c.
2"McEachran et al. 1980
3Leak 1981.
'"Shaw et al. 1985.
25Ditty 1981.
"Shaw et al 1987.
'spore 1970.
26SEAMAP 1983.
5Shaw and Drulllnger 1986.
'«Turner 1969.
"Fore 1971.
^Shaw and Drulllnger 1985.
"Chnstmas and Waller 1975.
28Houde 1973.
'Grail et al 1983.
'SHoude 1976.
29Houde 1977a
sMontollo 1976.
'^Fruge
1977.
^epinucane et al 1979b,
'Daniels 1978.
^oplnucane et al
. 1977.
3'Collins et al. 1980.
'"Walker 1978
2'Richards et al.
1984.
32Sogard et al 1987.
"Cowan 1985.
22Kelley
et al. 1986.
33Richards and Potthoff 1980b.
crepancy in Spanish sardine seasonal occurrence
reported by Houde and Fore (1973) (September-
February) and that found by Shaw and DrulHnger
(1985, 1986), Shaw et al. (1987), and this review
(March-November). Finally, seasonality data re-
ported in Ditty (1986) for larvae of finescale men-
haden, Brevoortia gunteri, was based on specimens
subsequently believed to be a morph of gulf menha-
den {B. patronus).
In general, the seasonal occurrence of many taxa
of larval fishes (i.e., scaled sardine, Harengula
jaguana, and Atlantic thread herring, Opisthonema
oglinum) began both earlier and extended later in
studies conducted on the Florida continental shelf
and southern parts of the study area than those in
the northern-most GOMEX. Likewise, in studies in-
volving bays (e.g., Phillips et al. 1977; Blanchet
1979; Williams 1983), tidal passes (e.g.. King 1971;
Sabins 1973; Allshouse 1983), and barrier islands
(e.g., Guillen and Landry 1981; Ruple 1984), peak
seasonal occurrence of larvae of some fishes was
usually later than in studies conducted offshore,
primarily because of the time delay necessary for
shoreward migration of estuarine-dependent larvae
such as Atlantic croaker (Micropogonias undulatus),
striped mullet {Mugil cephalus), pinfish {Lagodon
817
Table 3.— Location and sampling characteristics for northern Gulf of Mexico (GOMEX) ichthyoplankton studies. For months sampled:
Q = quarterly, M = monthly, S = synoptic, B = biweekly. Numbers in columns with asterisk (*) refer to footnotes. Numbers separated
by semicolons refer to specific gear types.
Gear
Mesh
Station
Months/freq.
Study type/
Study
type*
size 1
(mm)
depths (m)
sampled*
taxa studied
Tow type*
Study location
Phillips et al. 1977
2
0.363
<10
10, B
Survey
13
Tampa/Hillsborough Bay, FL
Blanchet 1979
2
0.505
2-6
11, M
Survey
14
Lower Apalachicola Bay, FL
Houde et al. 1979
1,
2
0.505
10-200
9, Q
Survey
13
Florida continental shelf
Turner 1969
4
0.571
<58
9, Oct. -Apr.
Brevoortia
15,
19
Eastern and central GOMEX
Williams 1983
5
0.505
<10
11, M
Survey
14,
17
Lower Mobile Bay, AL
Struck and Perry 1982
2
1.050
10
9, M
Survey
14,
17
Off Mississippi Sound
Ruple 1984
4;
6
0.505,
0.571
4-7; 0.5
12, M
Survey
14,
19; 21
Horn Island, MS
Juarez 1976
2
0.505
>183
Q, Apr.-Nov.
Scombrids
13
Central GOMEX
Montolio 1976
2
0.505
>183
Q, Apr.-Nov.
Carangids
13
Central and western GOMEX
Fore 1971
—
—
27-108
M, Dec. -Apr.
Etrumeus
—
Central GOMEX
Sabins 1973
8
0.480
<1.5
11, M
Survey
19,
21
Caminada Pass, LA
Fruge 1977
1,
2
0.505
<91
Nov.
Micropogonias
13
Mississippi River Delta to
Atchafalaya Bay, LA
Kuhn 1979
1
0.505
10-91
Nov., Jan., July
Bothids
13
Mississippi River Delta to
Atchafalaya Bay, LA
Ditty and
1
0.505
10-91
Jan., July
Survey
13
Mississippi River Delta to
Truesdale 1984
Atchafalaya Bay, LA
Walker 1978
2
0.505
10-45
July, Aug., Dec.
Survey
15,
19
Mississippi River Delta to
Timbalier Bay, LA
Ditty 1986
1;
2
0.363
10; 30
11, M; 9, 0
Survey
14,
14
14,
16, 17;
Off Caminada Pass, LA
Vecchione et al. 1982
1
0.333,
0.505
10
11, M
Survey
16, 17
Southwest Louisiana off
Calcasieu Lake
Cowan 1985
1
0.335
6-130
M, Dec. -Apr.
Sciaenids
18
Southwest Louisiana
Shaw et al. 1985
1
0.335
6-130
M, Dec. -Apr.
Brevoortia
18
Southwest Louisiana
Guillen and
8
1.000
<1.5
M, Jan. -Aug.
Survey
19,
21
Galveston Island, TX
Landry 1981
McGowan 1985
1
0.333
Over reef
9, Q
Survey
13,
20
Flower Gardens Reef, TX
King 1971
4
1.000
<5
9, M
Survey
14,
16, 17
Cedar Bayou Pass, TX
Hoese 1965
2
0.086
6-50
9, M
Survey
14
Off Port Aransas, TX
Allshouse 1983
4
0.505
3-12
11, M
Survey
13
Aransas Pass, TX
Finucane 1976
2
0.250
18-135
Dec, Apr., Aug.
Survey
13
South Texas continental
shelf
Finucane et al. 1977,
1
0.333,
0.505
18-182
9, M
Survey
13
South Texas continental
1979b
shelf
Finucane et al. 1979a
1
0.333,
0.505
17
July, Oct., Feb.
Survey
13
Off Galveston, TX
Fore 1970
7
—
7-75
M, Sept.-Apr.
Brevoortia
18
Central and western GOMEX
Christmas and
2;
4
0.505;
1.000
2-110
9, Oct. -Apr.
Brevoortia
13;
14
Gulf-wide
Waller 1975
Richards and
1;
3
0.505;
0.946
>183
9, Apr.-May
Scombrids
13;
14
Gulf-wide
Potthoff 1980a, b
SEAMAP 1983-1985
1;
3
0.333;
0.946
>10
9, S
Survey
13;
14
Gulf-wide
'60 cm bongo.
21 m met.
^1 X 2 m neuston.
■»0.5 m net.
=1 X 0.5 net.
«2 X 0.5 m.
'Gulf-V sampler
BRenfro beam trawl.
sMultiyear.
'015 months.
"12 months.
'214 months.
'^Oblique.
'••Surface.
'^Subsurface.
'^Middepth.
"Bottom,
'^Stepwise oblique
"Horizontal
2ooblique at select depth intervals.
2'Surf zone.
rhomboides), black drum, red drum, and gulf men-
haden.
The seasonal occurrence of several taxa are known
only from a few scattered specimens (e.g., cornet-
fish, Fistularia sp.; ribbonfish, Trachipterus sp.;
cowfish, Lactophrys sp.; slender mola, Ranzania
laevis; threadfish, Polydactylus sp.; and cobia,
Rachycentron canadum) or a single monthly record
(Table 1). Larvae of other taxa (e.g., searobins, Pri-
onotus spp.; anchovies, Anchoa spp.; tonguefishes,
Symphums spp.; and puff erf ishes, Sphoeroides spp.)
were collected during all months (Table 1), reflect-
ing the many species that comprise each genus. In
general, larvae of most scombrids were collected
beyond the 50 m depth contour, except Spanish
mackerel, Seomberomorus maculatus, which oc-
curred primarily within 50 m. Finucane et al. (1979b)
collected 50% of their king mackerel, 5. cavalla, lar-
vae off Texas during September and found this
species relatively more abundant and over greater
water depths (35-183 m) than Spanish mackerel
(<35 m). In the north-central GOMEX, larvae of
818
Spanish mackerel were more abundant than those
of king mackerel, and spawning of this latter species
probably occurs over shallower depths than in the
northwestern GOMEX (Shaw et al. 1987). Only six
larvae of king mackerel were collected by Houde et
al. (1979), suggesting that this species does not use
the northeastern GOMEX as a major spawning area.
Larvae of cero, 5. regalis, have not been reported
from the study area. Larvae of sciaenids and en-
graulids occurred primarily within the 50 m depth
contour, with several species of sciaenids primarily
collected inside the 25 m contour (Table 2). Both
sciaenids and engraulids are relatively more abun-
dant in the north-central and northwestern than in
the northeastern GOMEX (Finucane et al. 1977,
1979b; Houde et al. 1979; Richards et al. 1984;
Kelley et al. 1986) (Table 4). Most larvae of clupeids
occur in shelf waters of <50 m depth, except those
of round herring, Etrumeus teres, which primarily
occur beyond the 50 m contour (Table 2). Larvae of
gulf menhaden are more abundant in the north-
central than either the northeastern or northwest-
ern GOMEX and occur primarily around the
Mississippi River Delta (Fore 1970; Christmas and
Waller 1975; Sogard et al. 1987; and others). Scaled
sardine and Atlantic thread herring larvae are
abundant in all three subregions of the northern
GOMEX; Spanish sardine are rare in the north-
central but relatively abundant in the northeastern
and northwestern GOMEX (Shaw and Drullinger
1985, 1986; Shaw et al. 1987). Larvae of gulf butter-
Table 4— Comparison of 10 most abundant families of larval fishes
collected during major ichthyoplankton surveys of thie Gulf of Mex-
ico. Rank is based on number of individuals collected. Numbers
are % of total collection.
Finucane
Finucane
Houde
Kelley
Finucane
et al.
et al.
et al.
et al.
Taxa
1976
1977
1979b
1979
1986
Engraulidae
17.4
13.5
6.2
14.6
Gobiidae
16.6
20.6
15.8
15.1
6.2
Bregmacerotidae
14.5
12.5
7.3
2.7
5.4
Clupeidae
8.5
5.0
8.1
20.5
9.6
Sciaenidae
5.3
2.1
3.1
Carangidae
4.8
3.1
3.7
3.9
4.1
Bothidae
4.6
8.4
6.1
6.4
4.4
Synodontidae
4.5
3.6
10.9
3.0
Myctopfiidae
3.0
5.0
4.9
5.1
11.2
Serranidae
2.0
2.2
4.9
Cynoglossidae
3.4
Scombridae
1.9
Ophidiidae
2.9
Labridae
2.3
Gonostomatidae
3.3
Mugilidae
2.5
Totals
81.2
75.7
68.6
66.8
64.4
fish, Peprilus burti, and harvestfish, P. paru,
are most common in the north-central GOMEX
(SE AMAP 1983), with only a few of their larvae (8
and 25, respectively) having been collected by Houde
et al. (1979) in the northeastern GOMEX. Although
gulf butterfish larvae have been collected during
every month (Table 1), larvae are most common
from November to March with very limited spawn-
ing during the summer (Ditty 1981; SE AMAP 1983).
Houde et al. (1979) also collected >90% of both
sparid and haemulid larvae inside the 50 m contour
during their Florida continental shelf survey and
found that although lutjanids occurred at all depths,
they were most abundant from 30 to 100 m. Unlike
larvae of other speciose families which primarily oc-
curred either within (e.g., clupeids, engraulids, and
sciaenids) or beyond (e.g., scombrids) the 50 m depth
contour, those of bothids, carangids, and serranids
were widely distributed and occurred at all depths.
Larvae of swordfish, Xiphias gladius, and sailfish,
Istiophorus sp., are oceanic and occurred primarily
outside the 200 m depth contour; leptocephali of
tarpon, Megalops atlanticus, and bonefish, Albula
vulpes, were seldom collected in the study area. Of
those taxa whose larvae were most abundant within
the 25 m contour (Table 2), several were most com-
monly collected at considerably shallower depths
(e.g., pigfish, Orthopristis chrysoptera <20 m; black
drum <18 m; and spotted seatrout, Cynoscion nebu-
losus <15 m). Larvae of other taxa which occurred
primarily within the 25 m contour includes leather-
jacket, Oligoplites saurus (<20 m; Houde et al.
1979); kingfish, Menticirrhus spp. (<20 m; Walker
1978; Houde et al. 1979; Cowan 1985); and spottail
pinfish, Diplodus holbrooki (<15 m; Houde et al.
1979). Larvae of hogchoker, Trinectes maculatus,
and silver perch, Bairdiella chrysoura, were occa-
sionally collected in neritic offshore studies, but
were most abundant in pass/estuarine studies (e.g.,
Sabins 1973; Allshouse 1983).
In conclusion, these data represent the current
knowledge of the seasonality, peak occurrence, and
primary depth distribution of larval fishes in the
northern (rOMEX. This information provides a foun-
dation upon which sound management decisions con-
cerning both the commercial and recreational ex-
ploitation of spawning aggregations of fishes and
the potentially adverse impact on these fisheries
resulting from such exploitation can be based.
Acknowledgments
We would like to thank the Louisiana Department
819
of Wildlife and Fisheries and the Louisiana Offshore
Oil Port (LOOP, Inc.) for providing unpublished
seasonality data from 1983 plankton surveys; the
Southeast Area Monitoring and Assessment Pro-
gram (SEAMAP) for providing data on selected
species of clupeids, priacanthids, carangids, scom-
brids, and bothids collected during 1982 surveys of
northern GOMEX waters; and to William J. Rich-
ards for sharing information on Peprilus burti and
P. paru collected during 1982 SEAMAP cruises.
This paper is dedicated to Robert and Katherine
Ditty whose love and support through the years are
an inspiration. Louisiana State University Coastal
Fisheries Institute Contribution No. LSU-CFI-87-
25.
Literature Cited
Allshouse, W. C.
1983. The distribution of immigrating larval and post-larval
fishes into the Aransas-Corpus Christi Bay complex. M.S.
Thesis, Corpus Christi State Univ., Corpus Christ, TX, 118 p.
Aprieto, V. L.
1974. Early development of five carangid fishes of the Gulf
of Mexico and the South Atlantic coast of the United States.
Fish. Bull., U.S. 72:415-443.
Barger, L. E., L. a. Collins, and J. H. Finucane.
1978. First record of bluefish larvae, Pomatomics saltatrix,
in the Gulf of Mexico. Northeast Gulf Sci. 2:145-148.
Blanchet, R. H.
1979. The distribution and abundance of ichthyoplankton in
the Apalachicola Bay, Florida area. M.S. Thesis, Florida
State Univ., Tallahassee, 143 p.
Caldwell, D. K.
1957. The biology and systematics of the pinfish, Lagodon
rhomboides (Linnaeus). Bull. Fla. State Mus., Biol. Ser.
2:77-173.
Christmas, J. Y., and R. S. Waller.
1975. Location and time of menhaden spawning in the Gulf
of Mexico. Gulf Res. Rep., 20 p.
Collins, L. A., J. H. Finucane, and L. E. Barger.
1980. Description of larval and juvenile red snapper, Lutja-
nus campechanus. Fish. Bull, U.S. 77:965-974.
CoLTON, J. B., Jr., W. G. Smith, A. W. Kendall, Jr., P. L.
Berrien, and M. L. Fahay.
1979. Principal spawning areas and times of marine fishes.
Cape Sable to Cape Hatteras. Fish. Bull., U.S. 76:911-
915.
Cowan, J. H., Jr.
1985. The distribution, transport and age structure of drums
(Family Sciaenidae) spawned in the winter and early spring
in the continental shelf waters off western Louisiana. Ph.D.
Thesis, Louisiana State Univ., Baton Rouge, 182 p.
Daniels, K. L.
1977. Description, comparison, and distribution of larvae of
Cynoscion nebulosus and Cynoscion arenarius from the
northern Gulf of Mexico. M.S. Thesis, Louisiana State
Univ., Baton Rouge, 47 p.
Dawson, C. E.
1971a. Notes on juvenile black driftfish, Hyperoglyphe byth-
ites, from the northern Gulf of Mexico. Copeia 1971:732-
735.
1971b. Occurrence and description of prejuvenile and early
juvenile Gulf of Mexico cobia, Rachycentron canadum.
Copeia 1971:65-71.
Ditty, J. G.
1981. Comparative morphological development of Pepriliis
burti, P. triacanthus. and P. paru from the western North
Atlantic. M.S. Thesis, Louisiana State Univ., Baton Rouge,
48 p.
1984. Seasonality of sciaenids in the northern Gulf of Mex-
ico. Assoc. Southeastern Biol. Bull. 31(2):55.
1986. Ichthyoplankton in neritic waters of the northern Gulf
of Mexico off Louisiana: composition, relative abundance,
and seasonality. Fish. Bull.. U.S. 84:935-946.
Ditty, J. G., and F. M. Truesdale.
1984. Ichthyoplankton surveys of nearshore Gulf waters off
Louisiana: January-February and July, 1976. Assoc.
Southeastern Biol. Bull. 31(2):55-56.
DowD, C. E.
1978. Abundance and distribution of Bothidae (Pisces, Pleuro-
nectiformes) larvae in the eastern Gulf of Mexico, 1971-1972
and 1973. M.S. Thesis, Univ. Miami, Miami, 106 p.
Dwinell, S. E., and C. R. Futch.
1973. Spanish and king mackerel larvae and juveniles in the
northeastern Gulf of Mexico, June through October 1969.
Fla. Dep. Nat. Res., Mar. Resour. Lab. Leafl. Ser. Vol. IV,
Part 1(24):1-14.
Finucane, J. H.
1976. Baseline survey of ichthyoplankton. In W. B. Jackson
(editor). Environmental studies of the south Texas outer con-
tinental shelf 1975. Vol. 1: Plankton and fisheries, p. 20-42
and 163-305. NOAA Final Report to BLM, Contract No.
08550-1A5-19, 425 p.
Finucane, J. H., L. A. Collins, and L. E. Barger.
1979a. Determine the effects of discharges on seasonal abun-
dance, distribution, and composition of ichthyoplankton in
the oil field. In W. B. Jackson (editor), Environmental
assessment of an active oil field in the northwestern Gulf
of Mexico, 1977-1978. Vol. 2 - Data management and bio-
logical investigation, p. 2.3.6.-1 thru 2.3.6.-157. NOAA
Final Report to EPA, Contract No. EPA-IAG-D5-E693-EO.
Finucane, J. H., L. A. Collins, L. E. Barger, and J. B.
McEachran.
1979b. Ichthyoplankton/mackerel eggs and larvae. In W. B.
Jackson (editor). Environmental studies of the south Texas
outer continental shelf 1977. NOAA Final Report to BLM,
Contract No. AA550-1A7-21, 504 p.
Finucane, J. H., L. A. Collins, and J. D. McEachran.
1977. Ichthyoplankton/mackerel eggs and larvae. /wW. B.
Jackson (editor), Environmental studies of the south Texas
outer continental shelf 1976. NOAA Final Report to BLM,
Contract No. AA550-TA7-3, 484 p.
Fore, P. L.
1970. Oceanic distribution of eggs and larvae of the Gulf
menhaden. In Report of the Bureau of Commercial Fish-
eries Biological Laboratory, Beaufort, N.C., for the fiscal
year ending June 30, 1968, p. 11-13. U.S. Fish. Wildl. Serv.
Circ. 341.
1971. The distribution of the eggs and larvae of the round
herring, Etrumeu^ teres, in the northern Gulf of Mexico.
Assoc. Southeastern Biol. Bull. 18(1):34.
Fruge, D. J.
1977. Larval development and distribution of Micropogon un-
dulatus and Leiostomus xanthurus and larval distribution
820
oiMugil cephalus and Bregrrmceros atlanticus of the south-
eastern Louisiana coast. M.S. Thesis, Louisiana State
Univ., Baton Rouge, 75 p.
FUTCH, C. R.
197L Larvae of Monolene sessilicauda Goode, 1880
(Bothidae). Fla. Board Conserv., Mar. Lab. Leafl. Ser. IV,
Part l(21):l-n.
1977. Larvae of Trichopsetta ventralis (Pisces: Bothidae),
with comments on intergeneric relationships within the
Bothidae. Bull. Mar. Sci. 27:740-757.
FUTCH, C. R., AND F. H. HOFF, Jr.
1971. Larval development of Syacium papillosum (Bothidae)
with notes on adult morphology. Fla. Board Conserv., Mar.
Lab. Leafl. Ser. IV, Part l(20):l-22.
Gehringer, J. W.
1957. Observations on the development of the Atlantic sail-
fish, IstiophoTTus americanus (Cuvier), with notes on an
unidentified species of istiophorid. Fish. Bull., U.S. 57:
139-171.
GiBBS, R. H., Jr., and B. B. Collette.
1959. On the identification, distribution, and biology of the
dolphins, Coryphaemi hippurus and C. equiselis. Bull. Mar.
Sci. Gulf Caribb. 9(2):117-152.
Gordon, D. J.
1982. Systematics and distribution of larval fishes of the
subfamily Ophidiinae (Pisces, Ophididae) in the eastern
Gulf of Me.xico. M.S. Thesis, Univ. Miami, Coral Gables, 121
P-
Grall, C, D. p. de Sylva, and E. D. Houde.
1983. Distribution, relative abundance, and seasonality of
swordfish larvae. Trans. Am. Fish. Soc. 112:235-246.
Guillen, G. J., and A. M. Landry.
1981. Species composition and abundance of ichthyoplankton
at beachfront and saltmarsh environments. Proc. Annu.
Conf. Southeast Assoc. Game Fish Agencies 34:388-403.
Hensley, D. a.
1977. Larval development of Engyophrys senta (Bothidae),
with comments on intermuscular bones in flatfishes. Bull.
Mar. Sci. 27:681-703.
Herrema, D. J., B. D. Peery, N. Williams-Walls, and J. R.
Wilcox.
1985. Spawning periods of common inshore fishes on the
Florida east coast. Northeast Gulf Sci. 7:153-155.
HOESE, H. D.
1965. Spawning of marine fishes in the Port Aransas, Texas
area as determined by the distribution of young and larvae.
Ph.D. Thesis, Univ. Texas, Austin, 144 p.
Houde, E. D.
1973. Estimating abundance of sardine-like fishes from egg
and larval surveys. Eastern Gulf of Mexico: preliminary
report. Proc. Gulf Caribb. Fish. Inst. 25:68-78.
1974. Research on eggs and larvae of fishes in the eastern
Gulf of Mexico. In R. E . Smith (editor). Proceedings of the
marine environmental implications of offshore drilling in the
eastern Gulf of Mexico, p. 187-204. State Univ. Syst. Fla.,
Inst. Oceanogr., St. Petersburg.
1976. Abundance and potential for fisheries development of
some sardine-like fishes in the Eastern Gulf of Mexico.
Proc. Gulf Caribb. Fish. Inst. 28:73-82.
1977a. Abundance and potential j-ield of the Atlantic thread
herring, Opisthonema oglinum. and aspects of its early life
history in the eastern Gulf of Mexico. Fish. Bull., U.S.
75:493-512.
1977b. Abundance and potential yield of the round herring,
Etrumeus teres, and aspects of its early life history in the
eastern Gulf of Mexico. Fish. Bull., U.S. 75:61-89.
1977c. Abundance and potential yield of the scaled sardine,
Harengula jaguana, and aspects of its early life history in
the eastern Gulf of Mexico. Fish. Bull., U.S. 75:613-628.
1981. Distribution and abundance of four types of codlet
(Pisces: Bregmacerotidae) larvae from the eastern Gulf of
Mexico. Biol. Oceanogr. 1(1):81-104.
1982. Kinds, distributions and abundances of sea bass larvae
(Pisces: Serranidae) from the eastern Gulf of Mexico. Bull.
Mar. Sci. 32:511-522.
Houde, E. D., and P. L. Fore.
1973. Guide to identity of eggs and larvae of some Gulf of
Mexico clupeid fishes. Fla. Dep. Nat. Resour., Mar. Res.
Lab. Leafl. Ser. 4, Part 1, No. 23:1-14.
Houde, E. D., C. R. Futch, and R. DETvri-LER.
1970. Development of the lined-sole, Achirus lineatus, de-
scribed from laboratory- reared and Tampa Bay specimens.
Fla. Dep. Nat. Resour., Mar. Res. Lab. Tech. Ser. 62, 43 p.
Houde, E. D., J. C. Leak, C. E. Dowd, S. A. Berkeley, and
W. J. Richards.
1979. Ichthyoplankton abundance and diversity in the east-
ern Gulf of Mexico. Report to BLM, Contract No. 11550-
CT7-28, 546 p.
Johnson, G. D.
1984. Percoidei: Development and relationships. In H. G.
Moser et al. (editors). Ontogeny and systematics of fishes,
p. 464-498. Am. Soc. Ichthyol. Herpetol. Spec. Publ. No. 1.
Juarez, M.
1976. Distribucion de las formas larvarias de algunas especies
de la familia Scombridae en aguas del Golfo de Mexico.
Rev. Invest., INP. 2(l):33-65.
Kelley, S., T. Potthoff, W. J. Richards, L. Ejsy'mont, and
J. V. Gartner.
1986. SEAMAP 1983- Ichthyoplankton: Larval distribution
and abundance of Engraulidae, Carangidae, Clupeidae, Lut-
janidae, Serranidae, Sciaenidae, Coryphaenidae, Istiophori-
dae, Xiphiidae, and Scombridae in the Gulf of Mexico. U.S.
Dep. Commer., NOAA Tech. Memo., NMFS-SEFC-167.
King, B. D., III.
1971. Study of migratory patterns of fish and shellfish
through a natural pass. Texas Parks Wildl. Dep. Tech. Ser.
No. 9, 54 p.
KUHN, N. A.
1979. Occurrence and distribution of larval flatfish (Pleuro-
nectiformes) off the southeastern Louisiana coast during four
cruises including brief descriptions of the early larval stages
of Citharichthys spilopterus and Etropics crossotus. M.S.
Thesis, Louisiana State Univ., Baton Rouge, 79 p.
Leak, J. C.
1981. Distribution and abundance of carangid fish larvae in
the eastern Gulf of Mexico, 1971-1974. Biol. Oceanogr.
l(l):l-28.
McEachran, J. D., J. H. FiNUCANE, AND L. S. Hall.
1980. Distribution, seasonality and abundance of king and
Spanish mackerel larvae in the northwestern Gulf of Mex-
ico (Pisces: Scombridae). Northeast Gulf Sci. 4:1-16.
McGowAN, M. F.
1985. Ichthyoplankton of the Flower Garden Banks, North-
west Gulf of Mexico. Ph.D. Thesis, Univ. of Miami, Miami,
376 p.
McGOWAN, M. F., AND W. J. RICHARDS.
1986. Distribution and abundance of bluefm tuna (Thunnus
thynniis) larvae in the Gulf of Mexico in 1982 and 1983 with
estimates of the biomass and population size of the spawn-
ing stock for 1977, 1978 and 1981-1983. Coll. Vol. Sci.
821
Pap., Int. Comm. Conserv. Atl. Tunas 24:182-195.
MONTOLIO, M. A.
1976. Estudio taxonomico y morfometrico de los estadios lar-
vales de dos especies de Carangidae, Decapteriis punctatus
(Agassiz, I82d) y Carayixcrysos (MitcheW, 1815), y sudistri-
bucion en el Golfo de Mexico. Rev. Invest., INP. 2(2):
85-125.
Pearson, J. C.
1929. Natural history and conservation of redfish and other
commercial sciaenids of the Texas coast. Bull. U.S. Bur.
Fish. 44:129-214.
Peters, K. M., and R. H. McMicheal, Jr.
1987. Early life history of the red drum, Sciaenops ocellattts
(Pisces: Sciaenidae), in Tampa Bay, Florida. Estuaries 10:
92-107.
Phillips, T. D., J. M. Lyons, J. M. Daily, and M. Sigurdson.
1977. A study of ichthyoplankton seasonality and entrain-
ment by the Big Bend Power Plant, Tampa Bay, Florida.
In R. D. Garrity, W. J. Tiffany, and S. Mahadevan (editors).
Ecological studies at Big Bend Steam Electric Station: An
analysis and summary of studies on the effects of the cool-
ing water system on aquatic fauna. Vol. Ill, Ch. 5, p. 1-128.
Richards, W. J.
1976. Spawning of bluefin tuna (Thunnus thynnus) in the
Atlantic Ocean and adjacent seas. Coll. Vol. Sci. Pap., Int.
Comm. Conserv. Atl. Tuna. 5:267-278.
1977. A further note on Atlantic bluefin tuna spawning.
Coll. Vol. Sci. Pap., Int. Comm. Conserv. Atl. Tuna. 6:335-
336.
Richards, W. J., and T. Potthoff.
1980a. Distribution and abundance of bluefin tuna larvae in
the Gulf of Mexico in 1977 and 1978. Coll. Vol. Sci. Pap.,
Int. Comm. Conserv. Atl. Tunas. 9:433-441.
1980b. Larval distributions of scombrids (other than bluefin
tuna) and swordfish in the Gulf of Mexico in the spring of
1977 and 1978. Coll. Vol. Sci. Pap., Int. Comm. Conserv.
Atl. Tunas. 9:680-694.
Richards, W. J., T. Potthoff, S. Kelley, M. F. McGowan,
L. Ejsymont, J. H. Power, and R. M. Olvera L.
1984. SEAMAP 1982- Ichthyoplankton larval distribution
and abundance of Engraulidae, Carangidae, Clupeidae, Lut-
janidae, Serranidae, Coryphaenidae, Istiophoridae, Xiphii-
dae, and Scombridae in the Gulf of Mexico. U.S. Dep.
Commer., NOAA Tech. Memo., NMFS-SEFC-144, 4 p.
Robins, C. R., R. E. Bailey, C. E. Bond, J. E. Brooker, E. A.
Lachner, R. N. Lea, and W. B. Scott.
1980. A list of common and scientific names of fishes from
the United States and Canada. Am. Fish. Soc. Spec. Publ.
No. 12 (4th ed.), 174 p.
Ruple, D. L.
1984. Occurrence of larval fishes in the surf zone of a north-
ern Gulf of Mexico barrier island. Estuarine Coastal Shelf
Sci. 18:191-208.
Sabins, D. S.
1973. Diel studies of larval and juvenile fishes of the Cami-
nada Pass area, Louisiana. M.S. Thesis, Louisiana State
Univ., Baton Rouge, 163 p.
Saville, a.
1964. Estimation of the abundance of fish stocks from egg
and larval surveys. Rapp. P.-v. Reun. Cons. int. Explor.
Mer 153:164-170.
SEAMAP.
1983. (plankton). ASCII characters. Data for 1982. Fish-
eries-independent survey datae/National Marine Fisheries
Service, Southeast Fisheries Center: Gulf States Marine
Fisheries Commission (producer).
1984. (plankton). ASCII characters. Data for 1983. Fish-
eries-independent survey datae/National Marine Fisheries
Service; Southeast Fisheries Center: Gulf States Marine
Fisheries Commission (producer).
1985. (plankton). ASCII characters. Data for 1984. Fish-
eries-independent survey datae/National Marine Fisheries
Service; Southeast Fisheries Center: Gulf States Marine
Fisheries Commission (producer).
Shaw, R. F., J. H. Cowan, Jr., and T. L. Tillman.
1985. Distribution and density oi Brevoortia patronus (gulf
menhaden) eggs and larvae in the continental shelf waters
of western Louisiana. Bull. Mar. Sci. 36:96-103.
Shaw, R. F., J. G. Ditty, and J. Lyszkowski-Shultz.
1987. Fisheries-independent data on coastal herrings and
associated species (including mackerels) from the northern
Gulf of Mexico. Final Report to the MARFIN Program
(NOAA Contract No. NA 86-WC-H-06117) for FY 1986-87,
104 p. NMFS Southeast Regional Office, 9750 Roger Blvd.,
St. Petersberg, FL, 33702.
Shaw, R. F., and D. L. Drullinger.
1985. The early life history of coastal pelagic finfish in Loui-
siana. La. Board Regents Res. Dev. Prog., Baton Rouge,
115 p.
1986. Early life history of coastal pelagic finfish in Louisiana.
La. Board Regents Res. Dev. Prog. Final Rep., Baton Rouge,
272 p.
Smith, D. G.
1980. Early larvae of the tarpon, Megalopa atlantim Valen-
ciennes (Pisces: Elopidae), with notes on spawning in the
Gulf of Mexico and the Yucatan Channel. Bull. Mar. Sci.
30:136-141.
Sogard, S. M., D. E. Hoss, and J. J. Govoni.
1987. Density and depth distribution of larval gulf menhaden,
Brevoortia patronus, Atlantic croaker, Micropogonias un-
dulatus, and spot, Leiostomiis xanthurus, in the northern
Gulf of Mexico. Fish. Bull., U.S. 85:601-609.
Stuck, K. C, and H. M. Perry.
1982. Ichthyoplankton community structure in Mississippi
coastal waters. In Fishery monitoring and assessment com-
pletion report, 1 January 1977 to 31 December 1981, p. VI-
I-l thru VI— 1-53. Gulf Coast Res. Lab. (Ocean Springs,
MS), Proj. No. 2-296-R.
Thompson, B. A., and L. A. Deegan.
1982. Distribution of ladyfish (Elops saurus) and bonefish
(Albula vulpes) leptocephali in Louisiana. Bull. Mar. Sci.
32:936-939.
Tucker, J. W., Jr.
1978. Larval development of four species of bothid flatfish
in the Citharichthys-Etropus complex: C. cornutus, C. gym-
norhinus, C. spilopterus, and Etropus crossotvs. Ph.D.
Thesis, North Carolina State Univ., Raleigh, 213 p.
Turner, W. R.
1969. Life history of menhadens in the eastern Gulf of Mex-
ico. Trans. Am. Fish. Soc. 98:216-224.
Vecchione, M., C. E. Meyer, and C. L. Stubblefield.
1982. Zooplankton. In West Hackberry brine disposal
project pre-discharge characterization, Ch. 8, p. 8-1 thru
8-69. Dep. Energy Strategic Pet. Reserve Proj., D.O.E.
Contract No. DE-AC96-80P010228, 729 p.
Walker, H. J., Jr.
1978. Ichthyoplankton survey of nearshore Gulf waters
between Barataria Bay and Timbalier Bay, Louisiana, dur-
ing July, August, and December, 1973. M.S. Thesis, Loui-
siana State Univ., Baton Rouge, 59 p.
822
Williams, L. W.
1983. Larval fish assemblages of lower Mobile Bay. M.S.
Thesis, Univ. South Alabama, Mobile, 55 p.
James G. Ditty
Louisiana Department of Wildlife and Fisheries
Seafood Division
P.O. Box 98000. Baton Rouge, LA 70898-9000
Present address:
Coastal Fisheries Institute
Center for Wetland Resources
Louisiana State University
Baton Rouge. LA 70803-7507
Glen G. Zieske
Louisiana Department of Wildlife and Fisheries
Seafood Division
P.O. Box 98000, Baton Rouge. LA 70898-9000
Richard F. Shaw
Coastal Fisheries Institute
Center for Wetland Resources
Louisiana State University
Baton Rouge. LA 70803-7507
UTILIZATION OF A WASHINGTON
ESTUARY BY JUVENILE
ENGLISH SOLE, PAROPHRYS VETULUS
The use of west coast estuaries and protected bays
as nursery grounds by English sole, Parophrys
vetulus Girard, a significant component of Pacific
coast groundfish landings, has been well docu-
mented (Westerheim 1955; Kendall 1966; Smith and
Nitsos 1969; Misitano 1970). From data collected off
Oregon, Laroche and Holton (1979) showed that
English sole also utilize nearshore areas along the
open coast as nursery grounds. Krygier and Pearcy
(1986) determined that estuarine dependence for
juvenile English sole was indeed significant relative
to the open coastal area off Oregon, although their
survey design made it difficult to compare absolute
abundance in these areas. In addition, the estuaries
studied by Krygier and Pearcy were much smaller
than the Washington estuaries of Grays Harbor and
Willapa Bay, making it difficult to extrapolate their
results.
In the present study our objectives were to 1)
compare relative density and estimates of abun-
dance of 0-age English sole between a Washington
estuary. Grays Harbor, and the adjacent area along
the open coast; 2) compare fish density between
several subareas (strata) of each system; and 3) note
timing of immigration to and emigration from the
estuary. Specific gear was developed to efficiently
sample small benthic organisms and was used in
both the estuary and open coast survey areas, elim-
inating the need for gear selectivity intercalibration.
In addition, the statistical design of the survey en-
abled population estimates with confidence intervals
to be made for each area.
Methods and Materials
Survey Design
For this study, we specifically developed a plumb
staff beam trawl with an effective width of 2.3 m.
We designed it for a quantitative assessment of
juvenile fishes and crustaceans closely associated
with the bottom. Its fine mesh (4 mm) cod end liner
retained newly settled flatfish (15-25 mm total
length). A complete account of its construction,
method of deployment, and field testing was given
by Gunderson and Elhs (1986).
We selected two separate survey areas for the
study, the Grays Harbor estuary and the adjacent
nearshore area along the open coast. The estuarine
survey was based on a stratified random statistical
design and the open coast survey on a systematic
trackline. Both areas were surveyed in 1983 and
1984.
The estuary was stratified into four geographic
areas (Fig. 1). Each stratum was divided into 1 x
1 km grids (1 km intervals in the case of narrow
channels), and several stations were then random-
ly selected with the constraint that no two be adja-
cent. Additional stations were added in both STR
(stratum) 1 and 2 for the 1984 survey.
For the open coastal survey, three tracklines
oriented perpendicular to the bathymetry were
located off Copalis Head, Westport, and Willapa Bay
(Cape Shoalwater) (Fig. 2). We established a sys-
tematic series of stations along each trackline at 9
m depth intervals from 9 to 64 m. Whenever wave
conditions permitted, we sampled an additional sta-
tion at 5.5 m. In 1984, the 64 m stations were
dropped on each trackline because of consistent gear
damage in 1983. Also in 1984, replicate tows were
made at the 27 and 37 m stations.
Sampling Schedule
We sampled the estuary twice monthly from April
through September 1983 and 1984, and a single trip
was made in January 1984 for continuity. The two
fishery BULLETIN: VOL. 86, NO. 4, 1988.
823
i _j|j«*^Wiiifir'^ -rt^ ^x, 0 2 A 6
>
^ ,y^ ,- — « ^ ^"^^Vi /'^ Kilometers
I
1
R 1' — . ° '• ^fi»s^^<«;iaj^si Ji?>>^A>t
4-' a» ' ^ f ""^i "* ^^^-«aL. rK^f^ r\\'
% \.' : / ° ■ ^^^ '%
^
^
»' ^'Z / • y*"!*^^^
j^J^^(T) °'''"°' '"''^ ° °_->rr?^^^^^^^
^^^"^ " ^T^-'^'- O STtATUM NUMBER
T-T-^^ © ^^^^s^^Wtv STRATUM BOUNDARY
Pt ChehalisV. fy* ° 'Z^*' "'^"H^
V ii''^' T • '^^ Stratum Area (Ha) Sites
\ ¥^\° >T 'y^^^^^^t^^ 1 3651 5
V \* • 4&^r^^ 2 2516 6
\\ ^%0' ^: 3 1548 4
U '4 /«.r^^\%'- * ^^ ^
tl 'fe*' "^sT^T^^
Figure 1.— Chart of Grays Harbor, WA, showing the four strata and random trawl sites selected from grid lines.
trips per month coincided with low tides when
navigation and location of stations within the nar-
row channels characteristic of the North Bay and
Inner Harbor would be easiest. Stations situated in
the shallow, unmarked channels were sampled at
low water, while stations in more navigable areas
were sampled during the following periods of high
water. All hauls were confined to slack tide periods
so that strong tidal currents would not interfere
with the operation of the beam trawl. We sampled
the offshore study area once a month from April
through September.
Field Techniques
Within the estuary, the plumb staff beam trawl
was operated from a 6 m outboard boat, while off-
shore, either a 17 m or 20 m vessel was chartered.
Mechanical or hydraulic winches allowed the re-
trieval of the beam trawl while underway, reducing
escapement of captured organisms from the net.
At all stations, the beam trawl was hauled for a
distance dependent on the amount of epibenthic
material expected. Hauls of <10 minutes duration
(200-300 m) were necessary within the estuary
because the small mesh cod end liner would clog with
mud and organic debris, making the net difficult to
retrieve. Although the same problem existed in the
open coast survey area, hydraulic lifting gear en-
abled hauls of up to 20-30 min duration (800-1,000
m) at most stations. We attempted to tow at 3 km/h
with a scope of about 5:1; at depths <5 m, the scope
was increased to 10:1.
Within the estuary, marker buoys were deployed
where we estimated the trawl started and stopped
fishing. The distance fished was determined by
measuring between buoys with an optical range-
finder. Along the open coast, Loran C readings were
recorded at points we estimated the trawl first con-
tacted and left bottom and were later converted to
distance fished.
We sorted most of the catch by species (some
fishes to family only), and information was recorded
for total weight and number per haul. Random sub-
sampling of the catch (never <20% by weight) was
performed when necessary to speed processing.
Length frequencies and individual lengths and
weights were recorded for selected species.
824
Figure 2.— Chart of the open coast adjacent to Grays Harbor, WA, showing the isobaths,
tracklines, and systematic trawl stations.
Data Analyses
Nautical charts of Grays Harbor and the adjacent
open coast were digitized and strata areas were
calculated (Figs. 1, 2), using computerized algor-
ithms available in the software library of the
NWAFC Resource Assessment and Conservation
Engineering Division. Strata were defined for the
offshore area by first extending the north and
south boundaries 5 nmi from the Copalis and Willapa
transects, respectively, and then determining the
area between the 5, 15, 40, and 75 m depth con-
tours.
Estimates of mean density of young English sole
by stratum by month, population by stratum, and
total population for each survey area were calcu-
lated, using slight modifications of the methods of
Pereyra et al. (1976). In calculating density esti-
mates (number/hectare) from catch per kilometer,
the efficiency coefficient of the gear was assumed
to be unity (i.e., all fishes in the path would be cap-
tured). Length frequencies for 5 mm (TL) size inter-
vals were used to calculate size composition. Age
classes were then determined by visual inspection
of the resultant length-frequency histograms. After-
wards, the proportion of individuals within the size
range of the 0-age group in a particular month was
multiplied by the total population for that month to
provide an estimate of the abundance of 0-age
recruits.
Results
Over the course of the 2-yr study, we completed
a total of 592 trawls, 349 within Grays Harbor and
243 along the open coast (Table 1). Hauls within the
825
Table 1.— Allocation of sampling effort indicating the number of
successful trawls completed for each trip.
Survey
1983
1984
area
Month
Trips
Trawls
Month
Trips
Trawls
Estuary
Apr.
2
9,13
Jan.
1
13
May
2
13,16
Apr.
1
5
June
2
16,14
May
2
15,16
July
2
16,16
June
2
18,18
Aug.
2
14,16
July
2
18,17
Sept.
1
16
Aug.
2
19,17
Oct.
1
16
Sept.
1
18
Total
175
174
Open coast
May
1
17
Apr.
19
July
1
20
May
23
Aug.
1
25
June
26
Sept.
1
48
Aug.
Sept.
32
33
Total
110
133
estuary averaged 0.25 km at 2.52 km/h, while in the
coastal study area they averaged 0.91 km at 3.32
km/h.
Distribution and Abundance
We captured a total of 13 species of flatfishes in
the two study areas during the 2-yr survey (Table
2). Juvenile sanddab, sand sole, and English sole
were found to be abundant both in Grays Harbor
estuary and along the open coast. Two species in
particular, English sole and starry flounder, were
found to have much higher densities in the estuary
than along the open coast (Table 2). English sole
were found to be the most abundant flatfish in the
estuary for both years and along the open coast in
1984.
For both years within Grays Harbor, densities
of English sole generally were highest in STR 1, 2,
and 4 and lowest densities in STR 3 (Table 3).
Apart from changes due to recruitment, fluctuation
in abundance in the four areas of the estuary was
likely affected by movement of juveniles between
strata and between the estuary and open coast.
Along the open coast, greatest densities were
observed nearshore (Table 4) in depths <40 m.
In 1984, English sole were most abundant at 5-15
m owing to the presence of high numbers of 0-age
fish.
Mean density (all months combined) within the
estuary was over 20 times greater than the open
coast in 1983. Settlement of fish <25 mm TL
was much greater in 1984 than 1983 in both
areas, but estuarine density was still higher than
that of the open coast. A t-test was performed on
log-transformed CPUE (Zar 1984) because catch
data are typically nonnormally distributed. Re-
sults showed a significantly greater density (P <
0.0001) of juvenile English sole in the estuary
than along the open coast for both 1983 (562 vs.
23 mean no./ha) and 1984 (1,149 vs. 178 mean
no. /ha).
Population Estimates
In making population estimates, mean densities
of fish less than age V by month were multiplied by
strata areas. Comparison of length frequencies of
fish we collected with published age-at-length rela-
tionships (Van Cleve and El Sayed 1969) indicated
that fish beyond age IV were rarely captured. Older
fish may not have been present in the study areas
Table 2.— Stratified mean densities and 2 SE (in parentheses) for juvenile flatfish captured in the two survey
areas for 1983-84.
Species
English sole, Parophrys vetulus
Pac. Sanddab, Citharichthys sordidus
Butter sole, Isopesetta isolepsis
Sand sole, Psettichthys melanosticus
Dover sole, Microstomus pacificus
Rex sole, Glyptocephalus zachirus
Slender sole, Lyopsetta exills
Starry flounder, Platichthys stellatus
Petrale sole, Eopsetta jordani
Rock sole, Lepidopsetta bilineata
C-0 sole, Pleuronichthys coenosus
Curlfin sole, Pleuronichthys coenosus
Cal. tonguefish,^ Symphums atricauda
1983 densities (no./ha)
1984 densities (no./ha)
Estuary
Open coast
Estuary
Open coast
562(132)
23(10)
1,149(237)
178(101)
147(30)
94(24)
94(33)
65(22)
<1
37(10)
<1
13(4)
27(11)
10(6)
36(14)
23(11)
—
11(5)
—
8(4)
—
10(5)
—
2(2)
—
2
—
1
25(16)
<1
10(7)
1
—
<1
—
<1
—
<1
—
<1
<1
<1
—
—
<1
<1
—
<1
<1
<1
—
—
'Occurrence considered anomolous (Dinnel and Rogers 1986).
826
Table 3— Population estimates for juvenile English sole in the estuary survey area by date
(month/year) and by stratum (1-4). The 95% confidence interval for the total population is ob-
tained by adding or substracting the value m parentheses.
Date
Stra-
tum
Density
(no./ha)
Population
(millions)
Date
Stra-
tum
Density
(no./ha)
Population
(millions)
4/83
1
2
3
4
82
127
312
383
0.301
0.318
0.483
0.318
1/84
1
2
3
4
66
100
207
158
0.239
0.251
0.320
0.131
Total
1.421 (0.851)
Total
0.942 (0.472)
5/83
1
2
3
4
1,138
297
84
1,825
4.154
0.747
0.130
1.515
5/84
1
2
3
4
2,191
4,858
575
2,045
7.999
12.223
0.890
1.697
Total
6.545 (5.045)
Total
22.808 (8.686)
6/83
1
2
3
4
681
961
71
801
2.488
2.417
0.110
0.665
6/84
1
2
3
4
1,092
1,049
160
835
3.986
2.639
0.248
0.693
Total
5.679 (3.089)
Total
7.565 (4.195)
7/83
1
2
3
4
717
775
222
740
2.616
1.950
0.344
0.614
7/84
1
2
3
4
1.360
2,470
212
1,087
4.966
6.214
0.328
0.902
Total
5.524 (3.836)
Total
12.411 (5.569)
8/83
1
2
3
4
648
662
390
468
2.365
1.666
0.604
0.388
8/84
1
2
3
4
1,285
939
417
402
4.691
2.362
0.645
0.333
Total
5.023 (2.898)
Total
8.031 (3.557)
9/83
1
2
3
4
765
291
195
524
2.793
0.733
0.301
0.435
9/84
1
2
3
4
124
558
755
262
0.451
1.403
1.169
0.217
Total
4.262 (2.772)
Total
3.241 (1.745)
10/83
1
2
3
4
Total
125
598
564
115
0.457
1.505
0.873
0.095
2.931 (1.251)
or, more likely, were better able to avoid the nar-
row beam trawl.
Within the estuary, English sole were most nu-
merous in May of both years, but the peak of 22.8
million in 1984 was more than 3 times higher than
the peak of 6.5 million in 1983 (Table 3). Although
the distribution of the English sole population within
the estuary was highly variable, the bulk of the
population was in STR 1 or STR 2 for most months.
Along the open coast, the English sole population
peaked at the same time and at generally the same
levels as the estuary for both years, about 5.9 and
23.2 million for 1983 and 1984, respectively (Table
4). Comparable populations of young fish occurred
in both areas despite the 18 times greater geograph-
ic extent of the offshore survey area.
Recruitment
Relative recruitment to the two survey areas was
measured in terms of populations of 0-age fish.
Within the estuarine study area, recently trans-
formed (<25 mm TL; Laroche et al. 1982) English
sole were observed from April to July in 1983, and
in January and from May to August in 1984. Peak
abundance of this size range was observed in May
of both years, but it is highly likely that early spring
peaks were missed since our study lacked adequate
coverage of fall, winter, and early spring months.
Duration of settlement along the Oregon coast is
known to be much longer than observed in our study
(Laroche and Richardson 1979; Boehlert and Mundy
1987). Along the open coast, settlement seemed to
827
Table 4.— Population estimates for juvenile English sole in the open coast survey area
by date (month/year) and by stratum (1-4). The 95% confidence interval for total popula-
tion is obtained by adding or substracting the value in parentheses.
Date
Depth
(m)
Density
(no./ha)
Population
(millions)
Depth Density
Date (m) (no./ha)
Population
(millions)
5/83
7/83
8/83
9/83
5-15
15-40
40-75
Total
5-15
15-40
40-75
Total
5-15
15-40
40-75
Total
5-15
15-40
40-75
Total
6
100
11
2
20
8
54
35
12
12
32
3
0.115
4.881
0.887
5.883 (4.666)
0.044
0.983
0.680
1.707 (1.840)
1.028
1.709
0.991
3.728 (2.390)
0.230
1.547
0.269
2.045 (0.902)
4/84
5/84
6/84
8/84
9/84
5-15
165
3.124
15-40
47
2.318
40-75
55
4.525
Total
9.967 (7.071)
5-15
190
3.598
15-40
80
3.911
40-75
191
15.667
Total
23.176 (19.225)
5-15
226
4.282
15-40
48
2.356
40-75
9
0.732
Total
7.370 (3.110)
5-15
903
17.133
15-40
14
0.670
40-75
21
1.724
Total
19.527 (18.981)
5-15
97
1.836
15-40
322
15.761
40-75
5
0.402
Total
18.000 (11.494)
be of shorter duration, though again, the survey
lacked sampling effort from October through March.
Recently transformed juveniles were captured only
during May in 1983 and were found from April to
September in 1984. Peak abundance was observed
in May as in the estuary.
Populations of 0-age fish were determined using
estimates of total juvenile population (Tables 3, 4)
and size-frequency data (Table 5). These estimates
show that English sole had higher recruitment to
the estuary than to the open coast for early spring
of both years. The estuarine population of young-
of-the-year exceeded that of the open coast by over
four times (6.4 vs. 1.5 million) in May 1983. The dif-
ference was less pronounced in May 1984, but the
estuarine population was again higher (22.8 vs. 19.0
million). Later in summer in both years, the estu-
arine population of 0-age fish declined to a greater
extent than that along the open coast, but some of
the relative change is likely due to emigration from
the estuary.
Densities were plotted by 5 mm length interval
for the estuarine and open coast study areas for May
each year, a period of high settlement (Fig. 3). Den-
sities of juveniles <25 mm were more than 10 times
greater in the estuary, indicating disporportionately
higher direct settlement and/or higher mortality of
newly settled juveniles in the open coast area. When
Table 5.— Length ranges (mm TL) of 0-age English sole deter-
mined from visual inspection of length-frequency distributions by
study area and by date (month/year). Modal and mean lengths, and
the proportion of total population comprised by the 0-age group
are also indicated.
Range
Mode
Mean
Propor-
Study area
Date
(mm)
(mm)
(mm)
tion
Estuary
4/83
19- 84
65
57.6
0.85
5/83
19- 99
30
50.5
0.97
6/83
18-114
40
62.5
0.98
7/83
26-119
55
73.8
0.99
8/83
43-124
70
79.1
0.99
9/83
49-145
80
87.0
1.00
10/83
62-130
85
92.8
1.00
1/84
25- 54
35
34.9
0.58
5/84
15- 84
20
26.9
1.00
6/84
20- 99
30
46.4
0.96
7/84
25-114
85
66.9
0.99
8/84
30-124
90
84.8
0.99
9/84
50-135
90
92.0
1.00
Open coast
5/83
21- 60
35
33.3
0.25
7/83
99-115
100
104.0
0.07
8/83
43-165
120
115.4
0.80
9/83
53-200
110
134.7
0.89
4/84
15- 35
20
19.5
0.90
5/84
15- 90
20
26.5
0.82
6/84
15-120
30
48.6
0.87
8/84
20-160
95
91.0
0.98
9/84
20-180
130
119.0
0.89
828
Figure 3.— Density of juvenile English sole
by length interval for May and September
of 1983 and 1984. Note the higher density
of 0-age fish in the estuary each spring.
r
\
o
z.
c
Q
10'
10'
10'
Mag 1983
10" -
10"
10*
Estuary
Open Coast
o
o upe
50 100
September 1983
150
eoo
1)
X
\
o
z:
IT
«->
U)
C
10' -
10"
10"
lO"
O Estuary
« Open Coast
■r- '^ ~ ' I ' '
150
200
Mag 1984
I
\
o
z:
in
c
0;
lO' '
10* -
10' -
10' -
10"' -
10
-?
o Estuarg
® Open Coast
f f f V
50 100
September 1984
150
200
I
\
o
z:
c
C3
10*
10'
lO" -
10
-1
10"
Length (mm)
829
densities were plotted by length interval the fol-
lowing September of each year, it was evident
that fish over 140 mm were not available in the
estuary and had probably emigrated to the open
coast.
Discussion
If there is an adaptive advantage in utilizing estu-
arine nursery grounds rather than the open coast,
there must exist a mechanism for 0-age fish to enter
estuarine systems. Although English sole larvae are
abundant in coastal waters (Richardson and Pearcy
1977), early stages have not been prevalent in estu-
arine larval surveys (Pearcy and Meyers 1974;
Misitano 1977). Large transforming larvae (18-23
mm) have been collected in Humboldt Bay and
Columbia River estuary (Misitano 1976, 1977), and
in Yaquina Bay (Boehlert and Mundy 1987). Immi-
gration of 0-age English sole to the Grays Harbor
estuary may be accomplished by direct settlement
of transforming larvae after simple advection by
ocean water into the bay, or by movements of new-
ly settled benthic juveniles. Such movement could
be accomplished either actively or by selective tidal
transport as noted for juvenile flatfishes in the
North Sea (DeVeen 1978).
During the period of this study, newly trans-
formed English sole were found both within the
Grays Harbor estuary and along the adjacent open
coast. English sole have also been shown to enter
Yaquina Bay after settlement (Boehlert and Mundy
1987) so it is likely that both transforming larvae
and settled juvenile English sole may enter Grays
Harbor. Krygier and Pearcy (1986) found newly
transformed English sole to be more abundant in
open coastal areas and presumed movement into
Oregon estuaries to occur predominantly after
transformation. The occurrence of recently trans-
formed benthic juveniles in such high numbers
throughout Grays Harbor suggests direct settlement
of late stage larvae after advection into the estuary
may also be an important mode of entry.
Emigration of the largest fish to the open coast
took place during late summer, and all fish larger
than 140 mm were found exclusively in the open
coast area by September. Studies of other estuarine
nursery areas have indicated that the emigration
process involves the larger size classes of 0-age fish
(Herke 1971; Weinstein 1983). Emigration from Ya-
quina Bay of the larger 0-age English sole has been
noted in the fall (Westrheim 1955; Olsen and Pratt
1973; Bayer 1981). Angell et al. (1975) observed a
similar phenomenon for young English sole in a
Puget Sound nursery area.
The departure of larger juveniles later in summer
may be in response to changing environmental con-
ditions and may be indicative of the limits of the
carrying capacity of estuaries being exceeded for
populations of juvenile fish (Krygier and Pearcy
1986). Alternatively, a change in dietary preferences
of larger 0-age fish may cause them to leave estu-
aries in search of prey items (Toole 1980), thus
reducing intraspecific competition. The advantage
in utilization of estuarine nurseries then, may be
more for protection of vulnerable sizes rather than
for accelerated growth (Rosenberg 1982).
Even though our study found a great deal of inter-
annual variability, the Grays Harbor estuary and
other nearby estuaries (Shi 1987) clearly are impor-
tant nursery grounds for juvenile English sole,
which had similar size populations in the estuarine
and offshore study areas despite the greater geo-
graphic extent of the latter. Peak population esti-
mates for 0-age English sole for the month of May
show that 81% and 54% of 0-age fish in the Grays
Harbor area were found in the estuary in 1983 and
1984, respectively. This is probably an underesti-
mate of the degree of estuarine dependence, how-
ever, because some juveniles may move into the
estuary later in the summer. Nevertheless, our
results show that at least half of the 0-age English
sole in the Grays Harbor nearshore area make use
of an estuary during the first year of life. This kind
of information will prove useful in assessing the
economic impact on commercial fisheries from
navigation and industrial development projects,
which may contribute to habitat degradation in
Grays Harbor.
Acknowledgments
This note represents part of a Masters Thesis sub-
mitted to the University of Washington School of
Fisheries by C. W. Rogers. Work was supported
primarily by the Washington Sea Grant Program
(NOAA grant NA81AA-D00030, R/F-49). Data pro-
cessing assistance from the Northwest and Alaska
Fisheries Center is gratefully acknowledged. Logis-
tic support was provided by the U.S. Coast Guard
and the Washington Department of Fisheries. The
senior author also acknowledges the sponsors of the
Melvin G. Anderson Memorial Scholarship and the
Graduate School of the University of Washington
for support of portions of this work. We are grate-
fully indebted to all who assisted in field collections,
830
especially D. Samuelson, K. Carrasco, A. R. Black,
and B. Gutermuth.
Literature Cited
Angell, C. S., B. S. Miller, and S. R. Wellings.
1975. Epizootiology of tumors in a population of juvenile
English sole (Parophrys vetulus) from Puget Sound, Wash-
ington. J. Fish. Res. Board Can. 32:1723-1732.
Bayer, R. D.
1981. Shallow-water intertidal ichthyofauna of the Yaquina
estuary, Oregon. Northwest Sci. 55:182-193.
BOEHLERT, G. W., AND B. C. MUNDY.
1987. Recruitment dynamics of metamorphosing English
sole, Parophrys vetulus, to Yaquina Bay, Oregon. Estua-
rine Coastal Shelf Sci. 255:261-281.
DeVeen, J. F.
1978. On selective tidal transport in the migration of North
Sea plaice (Pleuronectes platessa) and other flatfish species.
Neth. J. Sea. Res. 12(2): 115-147.
DiNNEL, P. A., AND C. W. ROGERS.
1986. Northern range extension for California tonguefish,
Symphurus atricauda, to Washington State. Calif. Fish
Game 72:119-221.
GUNDERSON, D. R.. AND I. E. ELLIS.
1986. Development of a plumb staff beam trawl for sampling
demersal fauna. Fish. Res. 4:35-41.
Hercke, W. H.
1971. Use of natural and semi-impounded Louisiana tidal
marshes as nurseries for fishes and crustaceans. Ph.D.
Thesis, Louisiana State Univ., Baton Rouge, 264 p.
Kendall, A. W., Jr.
1966. Sampling juvenile fishes on some sandy beaches of
Puget Sound, Washington. M.S. Thesis, Univ. Washington,
Seattle, 77 p.
Krygier, E. E., AND W. G. Pearcy.
1986. The role of estuarine and offshore nursery areas for
young English sole, Parophrys vetulus Girard, of Oregon.
Fish. Bull, U.S. 84:119-132.
Laroche, J. L., AND S. L. Richardson.
1979. Winter-spring abundance of larval English sole, Paro-
phrys vetulus, between the Columbia River and Cape Blanco,
Oregon during 1972-1975 with notes on occurrences of three
other pleuronectids. Estuarine Coastal Mar. Sci. 8:455-
476.
Laroche, J. L., S. L. Richardson, and 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.
Laroche, W. A., and R. L. Holton.
1979. Occurrence of 0-age English sole, Parophrys vetulus,
along the Oregon coast: An open coast nursery area?
Northwest Sci. 53:94-96.
MiSITANO, D. A.
1970. Aspects of the early life history of English sole (Paro-
phrys vetulus) in Humboldt Bay, California. M.S. Thesis,
Humboldt State College, Areata, CA, 57 p.
1976. Size and stage of development of larval English sole,
Parophrys vetulus, at time of entry into Humboldt Bay.
Calif. Fish Game 62:93-98.
1977. Species composition and relative abundance of larval
and post-larval fishes in the Columbia River estuary, 1973.
Fish. Bull., U.S. 75:218-222.
Olsen, R. E., and 1. Pratt.
1973. Parasites as indicators of English sole {Parophrys
vetulus) nursery grounds. Trans. Am. Fish. Soc. 102:405-
411.
Pearcy, W. G., and S. S. Meyers.
1974. Larval fishes of Yaquina Bay, Oregon: a nursery
ground for marine fishes? Fish. Bull., U.S. 72:201-213.
Pereyra, W. T., J. E. Reeves, and R. G. Bakkala.
1976. Demersal fish and shellfish resources of the Eastern
Bering Sea in the baseline year 1975. U.S. Dep. Commer.,
Natl. Mar. Fish. Serv., NWAFC Proc. Rep., 619 p.
Richardson, S. L., and W. G. Pearcy.
1977. Coastal and oceanic fish larvae in an area of upwelling
off Yaquina Bay, Oregon. Fish. Bull., U.S. 75:125-145.
Rosenberg, A. A.
1982. Growth of juvenile E nglish sole, Parophrys vetulus, in
estuarine and open coastal nursery grounds. Fish. Bull.,
U.S. 80:245-252.
Shi, Y.
1987. Population dynamics of juvenile English sole, Paro-
phrys vetulus, in the estuarine and adjacent nearshore areas
of Washington. M.S. Thesis, Univ. Washington, Seattle, 86
P-
Smith, J. G., and R. J. Nitsos.
1969. Age and grovrth studies of English sole, Parophrys
vetulus, in Monterey Bay, California. Pac. Mar. Fish.
Comm. Bull. 7:73-79.
Toole, C. L.
1980. Intertidal recruitment and feeding in relation to optimal
utilization of nursery areas by juvenile English sole (Paro-
phrys vetulus: Pleuronectidae). Environ. Biol. Fishes 5:
383-390.
Van Cleve, R., and S. Z. El-Sayed.
1969. Age, growth, and productivity of an English sole (Paro-
phrys vetulus) population in Puget Sound, Washington.
Pac. Mar. Fish. Comm. Bull. 7:51-71.
Weinstein, M. p.
1983. Population dynamics of an estuarine-dependent fish,
the spot (Leiostomus xanthurus), along a tidal creek-seagrass
meadow coenocline. Can. J. Fish. Aquat. Sci. 40:1633-
1638.
Westrheim, S. J.
1955. Size composition, growth and seasonal abundance of
juvenile English sole (Parophrys vetulus) in Yaquina Bay.
Fish. Comm. Oreg. Res. Briefs 6:4-9.
Zar, J. H.
1974. Biostatistical analysis. Prentice-Hall, Inc.. Englewood
Cliffs, NJ, 718 p.
Christopher W. Rogers
School of Fisheries
University of Washington
Seattle, WA 98195
present address:
College of Marine Studies
Robinson Hall, Room 305
University of Delaware
Newark, DE 19716
School of Fisheries
University of Washington
Seattle, WA 98195
Donald R. Gunderson
David A. Armstrong
831
LENGTH-WEIGHT RELATIONSHIPS FOR GULF
FLOUNDER, PARALICHTHYS ALBIGUTTA,
FROM NORTH CAROLINA
Ginsburg (1952) resolved that the gulf flounder,
Paralichthys albigutta, ranging from North Carolina
to Laguna Madre, TX (Topp and Hoff 1972; Hoese
and Moore 1977; Robins and Ray 1986), was aPara-
lichthys. Topp and Hoff (1972) summarized the
many distributional records known throughout its
range. Other than keys to the species of Paralich-
thys (Gutherz 1967), much of the biology of the gulf
flounder remains unknown even though it abounds
east of the Mobile Bay system (Joseph and Yerger
1956; Topp and Hoff, 1972; Shipp 1986).
Some researchers cite a 390 mm (total length, TL)
Cedar Key, FL specimen (Jordan and Swain 1885)
as the largest size attained by the gulf flounder
(Hoese and Moore 1977; Robins et al. 1986). Vick
(1964) noted a 710 mm, 5 kg, specimen in the sport
fishery off Panama City, FL but did not furnish data
on specimens larger than 380 mm TL.
We present length-weight regression data for
North Carolina gulf flounder from 263 to 673 mm
TL and 318 to 3,706 g.
Methods
Since 1975, 75 gulf flounder were speared while
scuba diving along the Cape Lookout rock jetty (13
km east) and the artificial fishing reef (3 km SE) off
Morehead City, NC. Most dives occurred in Novem-
ber and December, when P. albigutta and other
paralichthids congregated in nearshore ocean
waters off Carteret County, NC prior to their off-
shore spawning migration. Specimens were weighed
to the nearest gram on beam balances and meas-
ured (total length) in millimeters within hours of
capture.
Observations
While gulf flounder are not abundant in North
Carolina (Schwartz et al. 1979, 1982), they are cap-
tured by hook and line or spear fishermen when the
fish frequent high saline nearshore ocean waters or
inlets (Schwartz 1979, 1982). Species of Paralich-
thys can usually be separated from each other by
the number of gill rakers on the lower first gill arch,
fin ray count, spotting, body width, salinity prefer-
ence, and depth distribution preference (Gutherz
1967). Gulf flounder possess 9-12 (usually 10-11) gill
rakers on the lower first arch, 53-63 anal rays, and
three prominent ocellated spots arranged in a tri-
angular pattern. North Carolina gulf flounder had
9-13 gill rakers on the lower first arch (65 speci-
mens) and 54-67 anal fin rays. Complete anal rays
were not counted in 17 specimens.
Gulf flounders caught consisted of 13 males and
62 females. Males ranged from 310 to 426 mm TL
260
300
340 380 420 460 500 540
Total length (TL) mm
580
620
660
700
Figure 1.— Length-weight relationship for gulf flounders, Paralichythys albigutta, from North Carolina. Asterisk represents Vick's
specimen.
832
FISHERY BULLETIN: VOL. 86, NO. 4, 1988.
and weighed 318 to 949 g, while females ranged
from 263 to 673 mm TL and 408 to 3,706 g (Fig.
1). The length-weight relationship for North Caro-
lina gulf flounder can best be expressed as log w =
-5.24 + 3.134 log / for 75 specimens (sexes com-
bined), r = 0.957 (Fig. 1). Little change occurred
when the male data was removed because the female
length-weight relationship was virtually the same:
log w = -5.018 + 3.053 log I (N = 62), r = 0.955.
Vick's (1964) large specimen, 710 mm TL, 5,000 g,
fits right on our regression curve. Thus, there is
little doubt that his specimen was P. albigutta be-
cause P. dentatus does not occur in the Gulf of Mex-
ico (Robins and Ray 1986), and P. squamilentus or
P. lethostigma possess other distinguishing meristic,
morphometric, and ecological requirements (Vick
1964; Gutherz 1967).
The maximum known upper size and weight can
now be raised to at least 673 mm and 3,706 g in
North Carolina.
Acknowledgments
Thanks are extended to the North Carolina
Marine Reef program personnel, which included
G.W.S., for assisting in collecting flounders in 1975.
Val and Henry Page produced Figure 1; Charleen
Miller, Texas A&M Research Foundation, was in-
strumental in locating Vick's 1964 report; and
Brenda Bright typed the manuscript.
Literature Cited
GiNSBURG, I.
1952. Flounders of the genus Paralichthys and related genera
in American waters. U.S. Fish. Wildl. Serv., Fish. Bull.
52:267-351.
Gutherz. E. J.
1967. Field guide to the flatfishes of the family Bothidae in
the western North Atlantic. U.S. Fish Wildl. Serv., Circ.
263, 47 p.
HoESE, H. D., AND R. H. Moore.
1977. Fishes of the Gulf of Mexico, Texas, Louisiana, and ad-
jacent waters. Texas A&M Univ. Press, College Station,
327 p.
Jordan, D. S., and J. Swain.
1885. Notes on fishes collected by David S. Jordan at Cedar
Keys, Florida. Proc. U.S. Natl. Mus. 7 (1884):230-234.
Joseph, E. B., and R. W. Yerger.
1956. The fishes of Alligator Harbor, Florida, with notes on
their natural history. Fla. State Univ. Stud. No. 22, Pap.
Oceanogr. Inst. No. 2, p. 111-156.
Robins, C. R., and G. C. Ray.
1986. A field guide to Atlantic coast fishes of North America.
Houghton Mifflin Co., Boston, 354 p.
Schwartz, F. J., W. T. Hogarth, and M. P. Weinstein.
1982. Marine and freshwater fishes of the Cape Fear estuary,
North Carolina, and their distribution in relation to environ-
mental factors. Brimleyana No. 7, p. 17-37.
Schwartz, F. J., P. Perschbacker, L. Davidson, K. Sandoy,
J. Tate, M. McAdams, C. Simpson, J. Duncan, and D. Mason.
1979. An ecological study of fishes and invertebrate macro-
fauna utilizing and Cape Fear River estuary, Carolina Beach
Inlet, and adjacent Atlantic Ocean. Inst. Mar. Sci. Publ.,
Morehead City, NC, 326 p.
Shipp, R. L.
1986. Dr. Bob Shipp's guide to the fishes of the Gulf of Mex-
ico. Dauphin Island Sea Lab Press, Dauphin Island, AL,
256 p.
Topp, R. W., and F. H. Hoff, Jr.
1972. Flatfishes (Pleuronectes), Vol. 4, Pt. II. Hourglass
cruises. Fla. Dep. Nat. Resour., St. Petersburg, Fla., 135 p.
Vick, N. G.
1964. The marine ichthyofauna of St. Andrew Bay, Florida,
and nearshore habitats of the northeastern Gulf of Mexico.
Texas A&M Univ. Res. Found. Proj. 286-D, 77 p.
Glen W. Safrit
Frank J. Schwartz
Institute of Marine Sciences
University of North Carolina
Morehead City, NC 28557
FUNCTIONAL REGRESSION EQUATIONS FOR
ZOOPLANKTON DISPLACEMENT VOLUME,
WET WEIGHT, DRY WEIGHT, AND CARBON:
A CORRECTION
The objective of this note is to point out the fact that
the first nine equations published by Wiebe et al.
(1975, table 2) were those appropriate for the Y on
X regressions; they were not the functional regres-
sion equations as originally claimed. This mistake
was discovered as a result of correspondence with
F. A. Ascioti (Dip. di Biologia Animale Ed Ecologia
Marina; Universita di Messina; Via Dei Verdi 75;
98100 Messina ITALY). This table 2 should have had
the following equations in it:
Equation
Regression equation
N
1
LOG(DV)
■■ -143Jf + 0.820
LOG(C)
87
2
LOG(WW) =
■■ -1.537 + 0.852
LOG(C)
70
3
LOG(DW)
0.i99 + 0.991
LOG(C)
195
h
LOG(DV)
■■ -1.842 + 0.865
LOG(DW)
163
5
LOG(WW) =
■■ -2.002 + 0.950
LOG(DW)
95
6
LOG(DV)
0.139 + 1.003
LOG(WW)
77
7
LOG(BWW) =
■■ -1.9J^7 + 1.050
LOG(BDW)
U21
8
LOG(BDV) =
■■ -1.887 + 1.007
LOG(BDW)
uou
9
LOG(BDV) =
0.005 + 0.981
LOG(BWW)
J,03
FISHERY BULLETIN: VOL. 86, NO. 4, 1988.
833
10 LOG(PDW =
11 LOG(DV)
12 LOG(WW) =
13 LOG(DV)
0.853 + 1.02J, LOG(PC) 45
1.037 + 0.839 LOG(DW) 111
0.975 + 0.9A6 LOG(DW) 9It
0.107 + 1.082 LOG(WW) 76
To prepare these GM functional regression equa-
tions (Ricker 1973:412), the original data were
checked and log transformed, and the regressions
recomputed. The variances and correlation coeffi-
cients remain essentially unchanged. Note that in
Wiebe et al. (1975) and above, equations 1-10 are
based on biomass standardized to per cubic meter
while the remainder are not. Equations 6 and 10 in
table 2 of Wiebe et al. (1975) also contained errors;
the intercept of equation 6 was incorrectly printed
as 0.670 rather than 0.067; the intercept of equa-
tion 10, which was 0.558, has been corrected to
0.853.
The equations listed above, when compared with
those originally presented, provide estimates of x
given yory given x, which generally differ by less
than 6% in the central part of the data set and by
less than 15% in the tails of the data set. Samples
similar in biomass and taxonomic composition to
those used in this study, which we have analyzed
for size of individuals as a function of taxonomic unit
(Davis and Wiebe 1985), contain animals which
typically range in size from 0.35 mm to 100 mm.
For a number of samples for which we reported
carbon values, nitrogen values were also obtained
from the CHN analyzers we used (Table 1 , this note).
Although carbon to nitrogen conversion factors exist
in the literature, few are based on data ranging over
as many hydrographic regimes as does data pre-
sented in Wiebe et al. (1975). Since a growing num-
ber of mathematical models use nitrogen rather than
carbon as the basic currency, we take this oppor-
tunity to present these data (Fig. 1) and the result-
ing carbon/nitrogen ratio.
Functional regression of carbon versus nitrogen
yielded the following significant {P < 0.01; r' =
0.99) relationship:
10'
CO
c
CD
CD
O
10^
10'
„ 1,,
1
)
< - A n - 48
□ -An-71
X - BBAY
-
- 1
1
• - GOS- 166
O - GOS - SW
1
1(D^
10^
10^
Carbon (mg/m^)
Figure 1.— Plot of data used in calculating the regression relating carbon and nitrogen in zooplankton collections from areas specified
in Table 1.
834
Table 1.— Number of zooplankton samples for each cruise from
which both carbon and nitrogen were measured. The general loca-
tion of the stations for these samples are given in Wiebe et al.
(1975).
volume, wet weight, dry weight, and carbon. Fish. Bull.,
U.S. 73:777-786.
Peter H. Wiebe
Cruise or area
Date
No. of
measure-
ments
Diameter
of net
(mesh size)
Buzzards Bay
Jan. -June
1972
16
70 cm (240 nm)
Slope Water
(RV Gosnold)
June-Aug.
1972
12
100 cm (333 ^^m)
Atlantis II 48
(Gulf of Mexico)
Nov. 1968
19
70 cm (240 ^m)
Gosnold 166
(New York Bight)
June 1970
33
70 cm (240 /jm)
Atlantis II 71
(Sargasso Sea)
Sept. 1972
39
100 cm (333 ^im)
Nitrogen (mg/m^) = -0.0247
+ 0.2324 carbon (mg/m^)
Essentially the carbon/nitrogen ratio of the bulk zoo-
plankton we collected is a constant (4.30) over a
broad range of values and oceanographic habitats.
As an atomic ratio, 5.02, this value is lower than that
predicted by the Redfield ratio, 6.63 (Redfield et al.
1963), an indication that zooplankton are nitrogen
rich relative to their phytoplankton counterparts.
Acknowledgments
We would like to express our appreciation to F. A.
Ascioti for his correspondence and attention to
detail which enabled us to correct the mistakes in
our previous publication. This research was sup-
ported by NSF Grant OCE-8709962 and is Contri-
bution No. 6839 from the Woods Hole Oceanograph-
ic Institution.
Literature Cited
Davis, C. S., and P. H. Wiebe.
1985. Macrozooplankton biomass in a Warm-Core Gulf
Stream Ring: Time series changes in size structure, tax-
onomic composition and vertical distribution. J. Geophys.
Res. 90:8871-8884.
Redfield, A. C, B. H. Ketchum, and F. A. Richards.
1963. The influence of organisms on the composition of sea-
water. In M. N. Hill (editor). The Sea, Vol. 2, The composi-
tion of sea-water, p. 26-77. Intersci. Publ., John Wiley and
Sons, N.Y.
Richer, W. E.
1973. Linear regressions in fishery research. J. Fish. Res.
Board Can. .30:409-434.
Wiebe, P. H., S. H. Boyd, and J. L. Cox.
1975. Relationships between zooplankton displacement
Woods Hole Oceanographic Institute
Woods Hole, MA 0251,3
ELECTROPHORETIC IDENTIFICATION OF
EARLY JUVENILE YELLOWFIN TUNA,
THUNNUS ALBACARES
Early juveniles, 13 mm standard length (SL) or
larger, of yellowfin tuna, Thunnus albacares, and
bigeye tuna, T. obesus, cannot be distinguished on
the basis of meristic, morphological, or pigmenta-
tion characters (Matsumoto et al. 1972). Collette et
al. (1984) reported that most species of the genus
Thunnus can be distinguished at the larval stage by
melanophore patterns. Matsumoto et al. (1972) and
Nishikawa and Rimmer (1987) suggested that T.
albacares and T. obesus larvae can be separated by
the respective absence or presence of postanal ven-
tral melanophores. Confirmation of the identifi-
cation of T. albacares larvae has been obtained
through laboratory rearing studies (Harada et al.
1971; Mori et al. 1971). However, the use of post-
anal ventral pigmentation patterns as reliable char-
acters to distinguish yellowfin and bigeye tuna
larvae has been questioned by Richards and Pothoff
(1974). Nishikawa and Rimmer (1987) stated that
it is virtually impossible to identify to species the
early juvenile stages, 15 to 60 mm SL, of Thunnus
because larval pigmentation patterns become
obscured and are no longer diagnostic. Further-
more, Pothoff (1974) was unable to separate T.
albacares and T. obesus as early juveniles, 8 to 100
mm SL, on the basis of osteological characters.
Electrophoresis of water soluble proteins has been
used to distinguish morphologically similar larval
and early juvenile marine fishes (Morgan 1975;
Smith and Crossland 1977; Sidell et al. 1978; Smith
and Benson 1980). Sharp and Pirages (1978)
presented starch gel electrophoretic patterns for
several loci of adults of many scombrid species, in-
cluding most members of the genus Thunnus. Al-
though electrophoretically very similar, adults of
yellowfin and bigeye tuna can be unambiguously
distinguished by the electrophoretic pattern of the
muscle isozyme of glycerol-3-phosphate dehydrog-
FISHERY BULLETIN: VOL. 86. NO. 4, 1988.
835
enase (alpha glycerol phosphate). This locus has been
used to identify adults of the genus Thunnus in the
absence of complete morphological data (Dotson and
Graves 1984). This paper describes an application
of the above techniques, modified for work with
small tissue samples, to the identification of an op-
portunistic collection of early juvenile tuna, in ex-
cellent morphological condition, which were frozen
shortly after capture. The early juveniles in this col-
lection encompassed the pigmentation patterns
reported for both T. albacares and T. obesus.
Specimens were collected aboard the MV Royal
Polaris, a San Diego-based sportfishing boat, about
1 km off Clipperton Island in the eastern Pacific (lat.
10°23'N, long. 109°15'W), 8 May 1986, from hours
of 2100 to 2400. The early juveniles were caught
underneath floodlights at a depth of about 1 m with
a fine (1 mm) mesh, long-handled dip net. They were
not present at the surface. After each pass, the dip
net contents were sorted for scombrid larvae. Ap-
proximately 100 specimens were collected, most of
which were quickly frozen in seawater.
Adult specimens of T. albacares and T. obesus
were collected by hook and line off the Pacific coast
of southern and northern Baja California, Mexico,
respectively. White muscle tissue samples were
removed from freshly caught specimens and quick-
ly frozen.
The early juveniles collected off Clipperton Island
were thawed in the laboratory and examined under
a dissecting microscope. Those juveniles positively
identified to the morphologically indistinguishable
T. albacares/T. obesus complex were measured for
total length (TL) to the nearest millimeter and ex-
amined for postanal ventral pigmentation pattern.
Heads were removed and placed in 95% ethanol for
otolith studies. The remaining trunk and tail mus-
culature was placed in a small (1.0 mL) microfuge
tube, and 60 /iL of cold grinding buffer (0.1 M Tris,
pH 7.5) was quickly added. Tissues were homog-
enized with a cold ground-glass rod contoured to fit
snugly within the microfuge tube. Approximately
10 seconds of rod rotation were required to com-
pletely disrupt the tissues. The homogenate was cen-
trifuged for 2 minutes in a microfuge and stored on
ice until electrophoresis.
Two grams of adult tissue were disrupted for ap-
proximately 20 seconds in 4 volumes of cold grind-
ing buffer in a motor-driven, ground-glass tissue
homogenizer. The homogenate was centrifuged at
5,000 g, 4°C, for 10 minutes. The supernatant was
removed, diluted 10:1 with cold grinding buffer, and
stored on ice until electrophoresis.
Horizontal starch gel electrophoresis was per-
formed on 12% (W/V) gels run in the Tris/Citrate
II system of Selander et al. (1971). Gels were run
at 45 to 50 mA for 3.5 hours. Glycerol-3-phosphate
dehydrogenase was stained using the protocol of
Shaw and Prasad (1970). Three sets of standards
composed of the supernatants of muscle tissue
homogenates of adult yellowfin and bigeye tuna
were placed in each gel to score the early juveniles.
A total of 77 early juveniles, ranging in length
from 10 to 21 mm TL, were processed. Glycerol-3-
phosphate dehydrogenase activity was scored for 68
individuals. All early juveniles displayed a muscle-
tjrpe glycerol-3-phosphate dehydrogenase band of
low anodal mobility, identical to that of the yellow-
fin tuna adults (Fig. 1). No individuals with the
faster migrating T. obesus glycerol-3-phosphate
dehydrogenase band were detected.
The lack of bigeye tuna juveniles in this study
could be the result of two possibilities: either the
early juveniles were all yellowfin tuna or both
yellowfin and bigeye tuna early juveniles share a
muscle-type glycerol-3-phosphate dehydrogenase
isozyme of similar electrophoretic mobility. How-
ever, differential ontogenetic expression of electro-
phoretically distinct isozymes has not been reported
for fishes in studies that have used adult allozymes
to identify larvae or early juveniles (Morgan 1975;
Smith and Crossland 1977; Sidell et al. 1978; Smith
and Benson 1980) or in investigations of ontogene-
tic expression of electrophoretic loci (Shaklee et
al. 1974; Siebenaller 1984). Thus, the electropho-
retic similarity of the glyceraldehyde-3-phosphate
alleles of the early juveniles investigated in this
study most likely indicates that they were all yellow-
fin tuna.
On the same trip during which the early juveniles
were collected at Clipperton Island, about 300 adult
yellowfin tuna were caught on hook and line but no
bigeye tuna were taken. Histological examination
of ovarian tissue from several of the adult yellow-
fin tuna revealed postovulatory follicles, indicating
that spawning was taking place (Anonymous 1987).
On the basis of this information, it is not unexpected
that all the early juveniles identified electrophoret-
ically in this study proved to be T. albacares.
A wide range of postanal ventral pigmentation
patterns (red and black) was displayed by the early
juveniles. Since these specimens were collected at
night, the pigment cells, when present, were distinct
and brightly colored as reported by Matsumoto et
al. (1972). While some individuals had no melano-
phores in this region, others had from one to eight.
836
•
1
8
10 11 12 13 14
Figure 1.— Photograph (A) and line drawing (B) of a gel demonstrating electrophoretic mobilities of muscle-tj-pe glycerol-3-phosphate
dehydrogenase alleles of yellowfin and bigeye tuna. Bigeye tuna display an allele with a greater anodal mobility. The gel includes
adult bigeye (lanes 1 and 13) and yellowfin (lanes 2 and 14) tuna standards and 7 early juveniles, all identified as yellowfin tuna.
Note that three individuals (lanes 4, 9, and 11) did not have sufficient activity to stain.
According to Matsumoto et al. (1972) and Nishikawa
and Rimmer (1987), yellowfin larvae less than about
12 mm SL have no black pigment spots in the ven-
tral tail region. However, Mori et al. (1971) reported
and illustrated that there is black pigmentation on
the ventral edges of the tail in laboratory-reared
yellowfin larvae at 7.8 mm TL. Twenty-one in-
dividuals in the 10 to 12 mm TL size range were elec-
trophoretically typed as yellowfin tuna in this study.
Six of these early juveniles had black postanal ven-
tral pigmentation (characteristic of bigeye tuna),
while 15 had no black postanal ventral pigmenta-
tion (characteristic of yellowfin tuna).
Variability in larval and early juvenile pigmenta-
tion within species, including large changes in pig-
mentation over small size ranges, is found within
many marine fishes (Powles and Markle 1984).
Richards and Pothoff (1974) have suggested that the
variability of postanal ventral pigmentation is not
consistent with specific differentiation within the T.
albacares/T. obesus complex. This study supports
their claim.
Early life history studies are necessary for an
understanding of recruitment within each species
of tuna. Due to the morphological similarity of
yellowfin and bigeye tuna larvae and early juveniles,
specific separation has not been possible. This study
provides a simple method for identifying yellowfin
and bigeye tuna larvae and early juveniles. With this
technique and additional material, it may be possi-
ble to find a morphological character that will allow
rapid identification of these two species.
837
Acknowledgments
We would like to thank the owner, Frank
LoPreste, and skipper, Steve Loomis, as well as the
crew of the MV Royal Polaris for providing the op-
portunity to collect specimens. William H. Bayliff
and Witold L. Klawe reviewed the manuscript.
Literature Cited
Anonymous.
1987. Annual report of the Inter-American Tropical Tuna
Commission 1986. Inter-Am. Trop. Tuna Comm., 264 p.
COLLETTE, B. B., T. POTTHOFF, W. J. RICHARDS, S. UEYANAGI,
J. L. RUSSO, AND Y. NiSHIKAWA.
1984. Scombroidei: development and relationships. In H. G.
Moser et al. (editors), Ontogeny and systematics of fishes,
p. 591-620. Am. Soc. Ichthyol. Herpetol., Spec. Publ. No. 1.
DOTSON, R. C, AND J. E. Graves.
1984. Biochemical identification of a bluefin tuna establishes
a new California size record. Calif. Fish Game 70:58-64.
Harada, T., K. Mizuno, 0. Muriata, S. Miyashita, and H.
hurutani.
1971. On the artificial fertilization and rearing of larvae
in yellowfin tuna. Mem. Fac. Agric. Kinki Univ. 4:145-
151.
Matsumoto, W. M., E. H. Ahlstrom, S. Jones, W. L. Klawe,
W. J. Richards, and S. Ueyanagi.
1972. On the clarification of larval tuna identification par-
ticularly in the genus Thunnus. Fish. Bull., U.S. 70:1-12.
Morgan, R. P.
1975. Distinguishing larval white perch and striped bass by
electrophoresis. Chesapeake Sci. 16:68-70.
Mori, K. S., S. Ueyanagi, and Y. Nishikawa.
1971. The development of artifically fertilized and reared lar-
vae of the yellowfin tuna, Thunnus albacares. Bull. Far
Seas Fish. Res. Lab. (Shimizu) 5:219-232.
Nishikawa, Y., and D. W. Rimmer.
1987. Identification of larval tunas, billfishes and other scom-
broid fishes (suborder Scombroidei): an illustrated guide.
Aust. CSIRO Mar. Lab. Rep. 186:1-20.
Potthoff, T.
1974. Osteological development and variation in young tunas,
genus Thunnus (Pisces, Scombridae), from the Atlantic
Ocean. Fish. Bull., U.S. 72:563-588.
Powles, H., and D. E. Markle.
1984. Identification of larvae. In H. G. Moser et al. (editors),
Ontogeny and systematics of fishes, p. 31-33. Am. Soc.
Ichthyol. Herpetol., Spec. Publ. No. 1.
Richards, W. J., and T. Potthoff.
1974. Analysis of the taxonomic characters of young scom-
brid fishes, genus Thunnus. In J. H. S. Blaxter (editor). The
early life history of fish, p. 623-648. Springer-Verlag,
Berlin.
Selander, R. K., M. H. Smith, S. Y. Yang, W. E. Johnson, and
J. B. Gentry.
1971. Biochemical polymorphism and systematics in the ge-
nus Peromyscus. I. Variation in the old-field mouse (Pero-
ynyscus polionotus). Univ. Texas Publ. 7103:49-90.
Shaklee, J. B., M. J. Champion, and G. S. Whitt.
1974. Developmental genetics of teleosts: biochemical anal-
ysis of lake chubsucker ontogeny. Dev. Biol. 38:356-382.
Sharp, G. D., and S. W. Pirages.
1978. The distribution of red and white swimming muscles,
their biochemistry, and the biochemical phylogeny of selected
scombrid fishes. In G. D. Sharp and A. E. Dizon (editors).
The physiological ecology of tunas, p. 41-78. Academic
Press, N.Y.
Shaw, C. R., and R. Prasad.
1970. Starch gel electrophoresis of enzymes— a compilation
of recipes. Biochem. Genet. 4:297-320.
SiDELL, B. D., R. G. Otto, and D. A. Powers.
1978. A biochemical method for distinction of striped bass
and white perch larvae. Copeia 1978:340-343.
Siebenaller, J. F.
1984. Analysis of the biochemical consequences on on-
togenetic vertical migration in a deep-living teleost fish.
Physiol. Zool. 57:598-608.
Smith. P. J., and P. G. Benson.
1980. Electrophoretic identification of larval and 0-group
flounders {Rhombosolea spp.) from Wellington Harbour, N.Z.
J. Mar. Freshwater Res. 14:401-404.
Smith, P. J., and J. Crossland.
1977. Identification of larvae of snapper, Chrysophrys
auratus Forster, by electrophoretic separation of tissue en-
zymes. N.Z. J. Mar. Freshwater Res. 11:795-798.
John E. Graves
Marie A. Simovich
Department of Biology.
University of San Diego
Alcala Park.
San Diego, CA 92110
Kurt M. Schaefer
Inter- American Tropical Tuna Commission
do Scripps Institution of Oceanography
La Jolla, CA 92093
A COMPARISON OF DEMERSAL ZOOPLANKTON
COLLECTED AT ALLIGATOR REEF, FLORIDA,
USING EMERGENCE AND REENTRY TRAPS
Demersal zooplankton have been shown to be im-
portant components of a number of marine com-
munities, including coral reefs (Porter and Porter
1977; Alldredge and King 1977), kelp beds (Ham-
mer 1981), and other habitats (Thomas and Jelley
1972). They probably play an important role in the
flux of particulate material through benthic com-
munities (Porter and Porter 1977). Demersal zoo-
plankton can also be important prey for fish and
other consumers (Alldredge and King 1977, 1980).
Demersal zooplankton are usually sampled by
techniques that take advantage of their migration
into or out of the plankton. Studies by Alldredge and
King (1980, 1985) and Youngbluth (1982) suggest
838
FISHERY BULLETIN: VOL. 86, NO. 4. 1988.
that sampling methods used to collect demersal
zooplankton can have a significant effect on the
numbers and kinds of animals collected. The most
widely used method to date is emergence trapping,
in which zooplankton are collected as they migrate
out of the substrate, e.g.. Porter and Porter (1977).
Another method, reentry trapping, captures zoo-
plankton as they enter the substrate, e.g., Alldredge
and King (1980). A primary purpose of this study
was to compare results from simultaneous emer-
gence and reentry trapping, and verify Alldredge
and King's (1980) finding that reentry traps capture
significantly more zooplankton than emergence
traps.
Horizontal currents are likely to have major but
variable effects on the transport of demersal organ-
isms in the plankton. Organisms migrating high into
the water column are likely to be transported lateral-
ly to a greater extent than those that remain in the
near-bottom layer, as Alldredge and King (1985)
have suggested. Thus, the implicit assumption of
some other investigators that demersal zooplankton
are characteristic residents of a habitat needs
qualification. Horizontal transport of migrating
demersal zooplankton is probably an important
recruitment mechanism and a means by which
planktivorous fish associated with reefs and other
structures are supported. Demersal zooplankton
have been shown to be associated with sand bottoms
by Alldredge and King (1977, 1980), so sand bottoms
might be an important source of demersal zooplank-
ton for reef communities. Therefore, our purpose
in this study was also to assess the abundance of
demersal zooplankton on sand bottom habitat. The
study addressed the questions: how abundant are
different taxa of demersal zooplankton associated
with sand bottom habitat? Can different trapping
techniques distinguish zooplankters with different
propensities for horizontal transport by currents?
How do the abundance and taxonomic composition
of demersal zooplankton on open carbonate sand
compare with the finding of others studying similar
sand bottoms?
Materials and Methods
Study sites were located approximately 1.8 km
east of Alligator Reef Buoy, ca. 7.5 km southeast
of Upper Matecumbe Key in the Florida Keys (lat.
24°41.26'N, long. 80°35.68'W). The sites were char-
acterized by a sub tidal, carbonate sand bottom that
sloped gently from depths of 2-4 m to depths of over
40 m. Macrophages occurred sparsely on the bottom
(principally Thalassia sp.), with the closest coral
structure and associated fish populations several
hundred meters downstream. Currents were mini-
mal (<0.05 m s^') and flowed to the northeast,
parallel to the axis of the Gulf Stream just offshore.
Sampling at the site was conducted with the aid
of scuba, supported by the NOAA-National Under-
sea Research Program's RV Seahawk, which op-
erated in the Florida Keys region during March
1985.
Demersal zooplankton were sampled simulta-
neously by emergence trapping and reentry trap-
ping. The emergence traps were a simplified ver-
sion of those used by Hobson and Chess (1979),
consisting of a single cone of 95 f^m mesh netting
1 m high and 21 cm diameter at the mouth (Fig. 1).
The mouth of the net was weighted with lead
weights sewn into the mouth collar to aid placement
on the bottom. AIL polyethylene bottle with a
polyethylene funnel in the throat was attached to
the upper end of the net; the bottles' natural buoy-
ancy extended the traps. Reentry traps were square
polyethylene pans, 21 cm per side and 5 cm deep,
with removable snap-top lids, approximately one-
third filled with defaunated local sand.
Divers deployed six of each type of trap in adja-
cent groups just after dusk and retrieved them just
after dawn on two successive nights in March 1985.
Reentry traps were closed and trap bottles from
emergence traps capped by divers before return to
the surface. The moon was nearly new and rose late
each night, thus providing little illumination. The
first night's collections were made for 12.75 hours
at a depth of 7 m and the second night's for 11.5
hours at a depth of 20 m.
Animals in the collection bottles of the emergence
traps were concentrated on a 95 ^m mesh sieve,
washed into sample bottles, and fixed in a buffered
formalin-Rose Bengal solution (ca. 5% formalin final
strength). Animals that had entered the reentry
traps were washed from the sand with fresh water,
causing them to release their grasp on sand grains
(R. Higgins^). The contents of each trap were
washed at least 5 times, and until no further animals
could be seen in the wash water. All washings were
sieved through a 95 /im mesh sieve. The retained
material was then poured into sample bottles and
fixed as above.
The stained, fixed animals in each sample were
examined, identified, and counted with a stereo-
microscope. Identifications were made to genus and
'R. Higgins, Smithsonian Institution, Washington, DC 20560.
839
Emergence Trap
95 urn
mesh
sail cloth
20.5cm
Figure 1.— Diagram of emergence trap used in this study.
species when possible; numbers of major taxa and
groups are presented here for simpHcity and com-
parative purposes.
Results
The zooplankton caught in the traps at Alhgator
Reef were dominated by copepods. Other taxa and
categories that were collected included chaeto-
gnaths, nematodes, ostracods, polychaetes, cuma-
ceans, isopods, amphipods, and mysids, as well as
numerous others that were less common or fre-
quently caught (Table 1). Many of these groups, such
as the harpacticoid copepods, are well known as ben-
thic or near-bottom forms. Others, such as the
calanoid copepods, are known as primarily plank-
tonic animals that are sometimes observed close to
the bottom or other substrate. Many larval forms
of benthic macroinvertebrates were captured, espe-
cially settling stages, such as barnacle cyprids and
megalopae.
Comparisons of the total mean numbers of animals
caught by each trap type and at each location show
that approximately 4 times as many animals were
caught by reentry traps as by emergence traps and
that about twice as many animals were captured at
the 7 m site as at the 20 m site (Table 1). Twelve
common groups together accounted for 97.9% of
all the animals caught. The results of two-way
ANOVAs showed that reentry traps captured sig-
nificantly more harpacticoid and cyclopoid copepods,
ostracods, mysids, polychaetes, and nematodes than
emergence traps, while the emergence traps cap-
tured significantly more calanoid copepods and
isopods (Table 2). Significantly more cumaceans,
gammarid amphipods, polychaetes, and nematodes
were captured at the 20 m site than at the 7 m site,
while significantly more calanoid copepods and chae-
tognaths were captured at the shallower location
(Table 2). The two-way ANOVAs showed significant
interaction effects between the trap types and sam-
pling locations on the numbers of harpacticoid and
840
Table 1 .—Demersal zooplankton captured in reentry and emergence traps at sand bottom sites
(7 m and 20 m) near Alligator Reef, Florida in March 1985. Data are mean numbers of animals
m
(SE).
Emergence
Reentrv
r
Group
7i
m
20 m
7 m
20nfi
Harpacticoids^
754 (188)
3,040 (726)
16,108 (2,289)
8,866 (845)
Cyclopoids'
2,039
(508)
740 (165)
5,943 (1,071)
3,713 (478)
Nematodes'
11
(11)
13(9)
8,840 (1,241)
823 (121)
Calanoids'
2,023
(421)
359 (128)
231 (80)
574 (209)
Copepod nauplii'
830
(248)
338 (99)
130 (53)
515 (121)
Ostracods'
51
(14)
35 (15)
498 (93)
360 (80)
Chaetognaths'
289
(47)
17(8)
357 (70)
126 (65)
Halacarids
4
(4)
0
538 (160
0
Polychaetes'
29
(27)
69 (42)
94 (39)
314 (56)
Caprellid amphipods
0
1 1 7 (38)
4(4)
170 (18)
Gammarid amphipods'
11
(5)
100 (49)
54 (19)
119 (32)
Cumaceans'
14
(5)
74 (51)
58 (23)
101 (18)
Mysids'
7
(5)
39 (8)
72 (27)
43 (18)
Larvaceans
112 (87)
0
29 (29)
11 (11)
Isopods'
62
(15)
35(11)
7(7)
33 (9)
Lancelets
0
0
14(7)
69 (9)
Pycnogonids
4
(4)
39 (34)
0
0
Tanaids
0
36 (11)
0
0
Cyphonautes
0
0
0
29 (13)
Penaeids
0
9(5)
0
18(7)
Chiton larvae
0
0
11 (7)
7(5)
Pagurid crabs
0
4(4)
7(5)
4(4)
Cyprids
0
0
11 (7)
0
Hippid crabs
0
0
0
7(5)
Magelona polychaetes
0
4(4)
0
0
Fish larvae
0
4(4)
0
0
Brachyuran crabs
0
4(4)
0
0
Totals
6,240
5,076
33,006
15,902
Trap totals
11,316
48,908
Site totals: 7 m:
39,250
20 m:
20,978
Grand total
60,228
"common" groups caught by both traps at both sites.
Table 2.— Results of 2-way ANOVAs comparing effects of sampling location and trap type on numbers of animals
caught from 12 common groups. Data are mean/m^ and F value for associated comparisons for each category;
significantly higher means are underlined. For all comparisons df = 1, 19.
Location
Trap
Interaction
Group
7 m
20 m
F
Emergence
Reentry
F
F
Harpacticoids
8,431
6,218
2.81
1,793
12,487
65.57**
14.33**
Cyclopoids
3,991
2,361
4.14
1,449
4,828
17.83**
1.17
Nematodes
4,426
454
36.80**
12
4,842
54.20**
41.70**
Calanoids
1,127
476
6.45**
1,267
402
11.39**
14.53**
Copepod nauplii
480
427
3.12
606
305
3.33
4.00
Ostracods
274
207
1.00
43
424
36.30**
1.75
Chaetognaths
323
77
19.51**
165
242
1.88
1.00
Polychaetes
61
203
12.07**
47
204
14.82**
3.69
Gammarid amphipods
32
110
7.26*
51
87
1.00
1.00
Cumaceans
36
89
4.41*
41
79
2.32
2.97
Mysids
40
41
1.00
22
58
4.92*
3.45
Isopods
34
33
1.00
49
20
6.94*
5.41*
= P < 0.05,
P < 0 01
841
calanoid copepods, isopods, and nematodes caught
(Table 2), indicating that the capture rates of the
two trap types varied between locations. Differences
in the types and numbers of animals caught by each
trap are more important, however.
Reentry traps were much more effective than
emergence traps at capturing a greater variety of
animals. The most striking differences are the much
larger numbers of harpacticoid and cyclopoid cope-
pods captured in reentry traps. In addition to quan-
titative differences the reentry traps also caught in-
dividuals of six groups that were not found in the
emergence traps (Table 1). The six groups included
the lancelets and five types of demersal larvae.
Conversely, emergence traps were more effective
at capturing calanoid copepods and isopods. Speci-
mens of five other taxa were captured only in
emergence traps (Table 1).
Analysis of the capture rate of each common
group by the two trap types shows significantly dif-
ferent assemblages (x" = 881068, df = 11, P «
0.001) (Fig. 2a). Furthermore, each trap type caught
some relatively rare groups, meaning that the lower
capture rate of the emergence traps did not prevent
them from capturing groups that did not appear in
the reentry traps. Analysis of the capture frequen-
cies of these rarer groups shows that the two trap
types capture different assemblages of organisms
(x^ = 25806, df = 9, P « 0.001) (Fig. 2b). There-
fore, the reentry and emergence traps sampled dif-
ferent fauna or sampled the same fauna differently.
Discussion
Varying migration patterns and swimming be-
haviors by the various taxa and groups can explain
the differences between the assemblages caught by
the two trap types. Ascending animals would have
to move 1 m off the bottom in order to be captured
by the emergence traps. Descending animals would
not have been captured at all by the emergence
traps, but would only have needed to be a few cm
off the bottom to enter the reentry traps. Thus,
reentry traps are more likely to capture demersal
organisms during their migration than emergence
traps if many of these organisms never move very
far up into the water column, as Alldredge and King
(1985) have shown. Reentry traps also captured set-
tling larvae, which presumably are migrating in only
one direction prior to establishing a sessile mode of
life. Such larval forms were a small fraction of the
total numbers of animals caught, but could be a sig-
nificant portion of the reentering fauna at times.
Both trap types may have also captured some ani-
mals that are holoplanktonic as noted by Robichaux
et al. (1981), despite our efforts to prevent this
during deployment and recovery of the traps. Some
animals may have entered the traps by crawling
rather than from the plankton, as Scheibel (1974)
observed. Finally, placement of the traps after dusk
may have missed animals migrating at or before
dusk, but the errors caused by this artifact, as well
as errors due to incomplete recovery of animals, are
not likely to alter our results significantly.
Another possible explanation for at least some of
the differences between capture rates of the two
trap types is differential avoidance of one trap type,
in this case the emergence traps. Given that emer-
gence traps consist of materials quite unlike those
that demersal zooplankton would normally en-
counter it should not be surprising that they might
seek to avoid contact with them. The narrow fun-
nel placed in the mouth of the collection bottles,
while necessary to retain animals that have entered
the bottle, may exclude others altogether. Some
demersal zooplankton, such as calanoid copepods,
are well known to exhibit an escape response when
placed in contact with surfaces. Reentry traps, on
the other hand, work partly by replicating natural
sand substrate, reducing the potential for avoidance.
The results show clearly that different sampling
techniques yield variable numbers of animals, even
within the same taxon, and collect different groups
of animals. Thus, evaluation of the demersal zoo-
plankton depends strongly on sampling techniques.
Adoption of a single standard sampling technique
might appear to be a resolution of the problem, but
a standard approach should sample all the organ-
isms that exhibit demersal behavior in a given area,
and neither emergence trapping nor reentry trap-
ing does. Furthermore, Stretch (1985) has observed
that not all members of a demersal population
migrate each night, so trapping techniques that
depend on animal migration must consistently
underestimate the actual abundance of demersal
organisms in association with a given substrate.
Tendency to migrate may vary among species,
within the life cycle of a given species and from day
to day, making accurate sampling of the demersal
zooplankton by trapping a logistical impossibility.
Collection techniques that directly sample demer-
sal organisms in or on the substrate, such as airlift
sampling (Stretch 1985) or sediment coring tech-
niques commonly used to sample meiofauna, should
give more accurate abundance estimates, but must
be used in conjunction with one or more trapping
842
O)
«
E
to
<
■o
0)
Q.
(0
O
50
40
30
20
10
a
d
.z%^
<^%^'
EL
*^^ ^c>'' <^* ^^* ^^ ^-
^ r^ri.
j:^
-*
M
O
Q.
E
o
O
• - none captured
^.-H.
n
• •— i-L
x:k
•„ •
.^^'
^>s
.*
,.»* .s^^
T T ••
.v<*^
<-' -<> \* x^-' rA"- ><~ <>'
O^ <i'0^
^<^
.«^'
<ir
Taxon/Group
Figure 2.— Percentage of the total animals caught by each trap type by taxon or group, in rank order of total abun-
dance, pooled over sites; a) common groups, b) rare groups.
843
techniques to distinguish migratory from nonmigra-
tory populations.
Comparisons of demersal zooplankton abundances
among studies are also made difficult by variation
among the trap types and approaches used. All-
dredge and King (1980) compared reentry and emer-
gence traps, showing as we have that reentry traps
captured very much larger numbers and different
proportions of demersal organisms. Aside from
studies by Stretch (1985) and ourselves (unpubl.
data), we are unaware of any effort to use a direct
sampling technique to calibrate a trapping tech-
nique. Thus, published abundance estimates for
demersal zooplankton abundance are probably low
and biased, reflecting the preponderant use of
emergence trapping.
Robichaux et al. (1981) pointed out that animals
entering traps by crawling can be a significant
artifact. Such contamination would probably be a
greater problem for reentry traps than for emer-
gence traps. Our reentry traps captured large
numbers of nematodes and harpacticoid copepods,
which can enter by crawling, but even when they
are eliminated altogether from the trap totals, reen-
try traps still caught twice as many animals as
emergence traps (Table 1). Furthermore, at least
some nematodes and harpacticoids do swim freely,
even if they do not move very far up into the water
column, as Alldredge and King (1985) have shown.
Thus, we think that reentry trapping reliably yields
higher estimates of demersal zooplankton abun-
dance that are more realistic than results from
emergence trapping but probably not truly accurate.
Robichaux et al. (1981) also argued that con-
tamination of demersal zooplankton traps by holo-
planktonic and crawling organisms causes an over-
estimate of the actual importance of demersal
zooplankton in benthic food webs. We dispute this
view on several grounds. First, the emergence trap-
ping technique used by Robichaux et al. (1981), as
is the case with others' use of emergence trapping,
probably yielded significant underestimates of the
actual abundance of demersal zooplankton, as we
discuss above. Second, we suspect that all trapping-
techniques are likely to miss animals that are not
migrating actively or that avoid traps, causing fur-
ther population underestimates. Finally, estimates
of demersal zooplankton populations resident within
a given habitat may fail to reflect the actual avail-
ability of these animals as consumers or prey via
transport.
Sand bottom habitats may be important sources
of demersal zooplankton for consumers in other
habitats. Currents can carry demersal organisms
passively to other habitats. Animals that migrate
high into the water column, such as the groups cap-
tured especially well by emergence traps, may be
carried relatively great distances compared with
those that crawl or stay within the near-bottom
boundary layer. Furthermore, off-reef foraging by
reef dwellers may allow exploitation of demersal
organisms on sandy bottoms in the absence of advec-
tion. If so, estimates of demersal zooplankton abun-
dance derived from reentry trapping will again be
realistic, if not accurate, from the standpoint of com-
munity ecology. Therefore, demersal zooplankton
are potentially quite important to marine benthic
communities, even if the techniques used to sample
them are imperfect.
Acknowledgments
This research was made possible with support pro-
vided by the NOAA-National Undersea Research
Program at the University of North Carolina,
(UNC), Wilmington, to Lawrence B. Cahoon and G.
Simmons, who graciously shared shiptime. Support
was also provided by UNC Sea Grant (R/MG 84-07
and R/MG 85-01), the National Science Foundation
(RII 8311 486), the American Philosophical Society,
the Lerner-Grey Fund for Marine Research and the
North Carolina Collegiate Academy of Sciences.
Literature Cited
Alldredge, A. L., and J. King.
1977. Distribution, abundance, and substrate preference of
demersal reef zooplankton at Lizard Island Lagoon, Great
Barrier Reef. Mar. Biol. (Berl.) 41:317-333.
1980. Effects of moonlight on the vertical migration patterns
of demersal zooplankton. J. exp. mar. Biol. Ecol. 44:133-
156.
1985. The distance demersal zooplankton migrate above the
benthos: implications for predation. Mar. Biol. (Berl.) 84:
253-260.
Hammer, R. M.
1981. Day-night differences in the emergence of demersal
zooplankton from a sand substrate in a kelp forest. Mar.
Biol. (Berl.) 62:275-280.
HoBSON, E. S., and J. R. Chess.
1979. Zooplankters that emerge from the lagoon floor at Kure
and Midway Atolls, Hawaii. Fish. Bull., U.S. 77:275-280.
Porter, J. W., and K. G. Porter.
1977. Quantitative sampling of demersal zooplankton migrat-
ing from different reef substrates. Limnol. Oceanogr. 22:
553-556.
Robichaux, D. M., A. C. Cohen, M. L. Reaka, and D. Allen.
1981. Experiments with zooplankton on coral reefs, or will
the real demersal zooplankton please come up? P.S.Z.N.L
Mar. Ecol. 2:77-94.
844
SCHEIBEL, W.
1974. Submarine experiments on benthic recolonization of
sediments in the western Baltic Sea. II. Meiofauna. Mar.
Biol. (Bed.) 28:165-168.
Stretch, J. J.
1985. Quantitative sampling of demersal zooplankton: re-
entry and airlift dredge sample comparisons. J. exp. mar.
Biol." Ecol. 91:125-136.
Thomas, M. L., and E. Jelley.
1972. Benthos trapped leaving the bottom in Bideford River,
Prince Edward Island. J. Fish. Res. Board Can. 29:1234-
1237.
YOUNGBLUTH, M. J.
1982. Sampling demersal zooplankton: a comparison of field
collections using three different emergence traps. J. exp.
mar. Biol. Ecol. 61:111-124.
Lawrence B. Gaboon
Craig R. Tronzo
Department of Biological Sciences
University of North Carolina at Wilmington
Wilmington. NC 28403
845
INDEX
Fishery Bulletin: Vol. 86, No. 1-4
ABLE, KENNETH W.-see GRIMES et al.
"Accumulation of age pigments (lipofuscin) in two cold-water
fishes," by Maria Vernet, John R. Hunter, and Russell D.
Vetter, 401
"Age and growth of larval gulf menhaden, Brevoortia patronus,
in the northern Gulf of Mexico," by Stanley M. Warlen, 77
"Age, morphology, developmental biology, and biochemical genetic
variation of Yukon River fall chum salmon, Oncorhynchus keta,
and comparisons with British Columbia populations," by Terry D.
Beacham, Clyde B. Murray, and Ruth E. Withler, 663
"Age-specific vulnerability of Pacific sardine, Sardinops sagax, lar-
vae to predation by northern anchovy, Engraulis mordax" by John
L. Butler and Darlene Pickett, 163
Ageing studies
lipofuscin in larval fish, 401
menhaden, gulf, 77
see also Otoliths
Algae, benthic
recruitment of spiny lobsters, 331
Allocyttus sp.— see Oreo, black
Alosa sapidissima—see Shad, American
Alvinocaris markensis
new decapod crustacean species, 263
Alvinocaris muricola
new decapod crustacean species, 263
Alvinocaris spp.
key to species, 263
Alvinocaris stactophila
new decapod crustacean species, 263
AL-YAMANI, J.-see DAGG et al.
Amphipods
shrimp diets, 543
"Analyses of the relationship between the distribution of searching
effort, tuna catches, and dolphin sightings within individual purse
seine cruises," by Tom Polacheck, 351
Anchovy, northern
predation on sardine larvae, 163
Anoplopoma fimbria—see Sablefish
"Appendage injury in Dungeness crabs. Cancer magister, in south-
eastern Alaska," by Susan M. Shirley and Thomas C. Shirley, 156
Apristurus hrunneus—see Shark, scyliorhinid
Armorhead, pelagic
biogeography, 453
ARMSTRONG, DAVID A.-see ROGERS et al.
ARNOLD, CONNIE R.-see BROWN-PETERSON et al.
"Aspects of the biology of two scyliorhinid sharks, Apristurus
brunneus and Parmaturus xaniurus, from the upper continental
slope off southern California," by Jeffrey N. Cross, 691
Atheresthes evermanni—see Flounder, kamchatka
Atheresthes stomias—see Flounder, arrowtooth
Atlantic croaker
distribution and abundance, 129
BAGLIVO, JENNY A.-see BROUSSEAU and BAGLIVO
Banded drum
distribution and abundance, 129
BARLOW, JAY, "Harbor porpoise, Phocoena phocoena, abun-
dance estimation for California, Oregon, and Washington: I. Ship
surveys," 417
BARLOW, JAY, CHARLES W. OLIVER, TERRY D. JACKSON,
and BARBARA L. TAYLOR, "Harbor porpoise, Phocoena pho-
coena, abundance estimation for California, Oregon, and
Washington: II. Aerial surveys," 433
BARSHAW, DLANA E., and DONALD R. BRYANT-RICH, "A
long-term study on the behavior and survival of early juvenile
American lobster, Homarus americanus, in three naturalistic sub-
strates: eelgrass, mud, and rocks." 789
Bass, striped
stomach contents, 397
BAYLIFF, WILLIAM H., "Integrity of schools of skipjack tuna,
Katsuwonus pelamis, in the eastern Pacific Ocean, as determined
from tagging data," 631
BEACHAM, TERRY D., CLYDE B. MURRAY, and RUTH E.
WITHLER, "Age, morphology, developmental biology, and
biochemical genetic variation of Yukon River fall chum salmon,
Oncorhynchus keta, and comparisons with British Columbia
populations," 663
BECKER, D. SCOTT, "Relationship between sediment character
and sex segregation in English sole, Parophrys vetulus," 517
"Behavior of southern right whales, EuJbalaena australis, feeding
847
feeding on the Antarctic krili, Euphaiisia superba," by William
M. Hamner, Gregory S. Stone, and Bryan S. Obst, 143
Behavior studies
English sole, 830
lobster, American, 791
plankton, 842
Biochemical genetics— see Genetic studies
Black drum
distribution and abundance, 129
trotline fishery, 109
BLAYLOCK, ROBERT A., "Distribution and abundance of
bottlenose dolphin, Tursiops truncatus (Montagu, 1821), in
Virginia," 797
Bluefish
spawning season, 237
BOEHLERT, GEORGE W., and TAKASHI SASAKI, "Pelagic
biogeography of the armorhead, Pseudopentaceros wheeleri, and
recruitment to isolated seamounts in the North Pacific Ocean,"
453
BOLZ, GEORGE R., and R. GREGORY LOUGH, "Growth
through the first six months of Atlantic cod, Gadus morhua, and
haddock, Melanogrammus aeglefinus, based on daily otolith
increments," 223
Boonea impressa—see Snail, pyramidellid
Bootstrap method
sablefish abundance, 445
BOURKE, ROBERT E.-see WATSON et al.
Breeding habits— see Reproductive studies
Brevoortia patronus—see Menhaden, gulf
BRILL, RICHARD W.-see WATSON et al.
BROOKS, E. R.-see MULLIN and BROOKS
BROUSSEAU, DIANE J., and JENNY A. BAGLIVO, "Life
tables for two field populations of soft-shell clam, Mya arenaria,
(Mollusca: Pelecypoda) from Long Island Sound," 567
BROWN-PETERSON, NANCY, PETER THOMAS, and CONNIE
R. ARNOLD, "Reproductive biology of the spotted seatrout,
Cynoscion nebulosus, in South Texas," 373
BRUCE, B. D., "Larval development of blue grenadier,
Macruromis navaezelandiae (Hector), in Tasmanian waters," 119
BRYANT-RICH, DONALD R.-see BARSHAW and BRYANT-
RICH
BUTLER, JOHN L.-see WATANABE et al.
BUTLER, JOHN L., and DARLENE PICKETT, "Age-specific
vulnerability of Pacific sardine, Sardinops sagax, larvae to preda-
tion by northern anchovy, Engraulis mordax," 163
BUTLER. MARK J., IV-see HERRNKIND et al.
Bythugraea mesatlantica
new decapod species, 263
CAHOON, LAWRENCE B., and CRAIG R. TRONZO, "A com-
parison of demersal zooplankton collected at Alligator Reef,
Florida, using emergence and reentry traps," 838
CALLAHAN, MICHAEL W.-see SEKI and CALLAHAN
Cancer magistei — see Crab, Dungeness
Cancer oregoneTisis—see Crab, Cancer
Carbon isotopes
shrimp diets, 543
Caridea: Bresilidae
key to species, 263
CASTRO, MARGARIDA, and KARIM ERZINI, "Comparison of
two length-frequency based packages for estimating growth and
mortality parameters using simulated samples with varying
recruitment patterns," 645
Catsharks— see Shark, scyliorhinid
Chacellus filiformis
bathymetric and geographic distribution, 67
Char, Arctic
predation on salmon, 611
CHEN, CHE-TSUNG, TZYH-CHANG LEU, and SHOOU-JENG
JOUNG, "Notes on reproduction in the scalloped hammerhead,
Sphyma lewini, in northeastern Taiwan waters," 389
CHESS, JAMES R.-see HOBSON and CHESS
Clam, soft-shell
life tables, 567
Cod, Atlantic
growth, 223
"Cojoined twin adult shrimp (Decapoda: Penaeidae)," by Austin
B. Williams, 595
COLLINS, MARK R., and CHARLES A. WENNER, "Occurrence
of young-of-the-year king, Scomheromorus cavalla, and Spanish,
S. maculatus, mackerels in commercial-type shrimp trawls along
the Atlantic coast of the southeast United States," 394
Cololabis saira—see Saury, Pacific
"A comparison of demersal zooplankton collected at Alligator Reef,
Florida, using emergence and reentry traps," by Lawrence B.
Cahoon and Craig R. Tronzo, 838
"Comparison of two length-frequency based packages for esti-
mating growth and mortality parameters using simulated samples
with varying recruitment patterns," by Margarida Castro and
Karim Erzini, 645
848
"A comprehensive theory on the etiology of burnt tuna," by Cheryl
Watson, Robert E. Bourke, and Richard W. Brill, 367
CONOVER, DAVID O.-see NYMAN and CONOVER
Copepod nauplii
distribution and abundance, 319
COSTON-CLEMENTS, LINDA-see HOSS et al.
COUTU, J.-M.-see DUTIL and COUTU
COWAN, JAMES H., JR., and RICHARD F. SHAW, "The
distribution, abundance, and transport of larval sciaenids collected
during winter and early spring from the continental shelf waters
off west Louisiana," 129
Crab. Cancer
occurrence off British Columbia, 525
Crab, Dungeness
abundance, 603
appendage injury, 156
dissolved oxygen levels, 604
megalopae associations, 299
Crab, Gulf stone
megalopa stage, 289
CRANE, S. A.-see DAVIES et al.
Crassostrea gigas—see Oyster, Japanese
Crassostrea virginica—see Oyster, American
CRECCO, VICTOR A.-see SAVOY and CRECCO
Croaker, Atlantic
metabolic responses to temperatures, 483
Croaker, white
food-seeking, 251
CROSS, JEFFREY N., "Aspects of the biology of two scylio-
rhinid sharks, Apristurus hrunneus and Pannatu7~iis xaniurus,
from the upper continental slope off southern California," 691
Crustacean, decapod
bathymetric and geographic distribution, 67
CURRENS, KENNETH P., CARL B. SCHRECK, and HIRAM
M. LI, "Reexamination of the use of otolith nuclear dimensions
to identify juvenile anadromous and nonanadromous rainbow trout,
Salmo gairdneri," 160
Cynoscion arenarius—see Seatrout, sand
Cynoscion nebulosu^see Seatrout, spotted
Cynoscion regalis—see Weakfish
DAGG, M. J., P. B. ORTNER, and J. AL-YAMANI, "Winter-time
distribution and abundance of copepod nauplii in the northern Gulf
of Mexico," 319
DAVIES, N. M., R. W. GAULDIE, S. A. CRANE, and R. K.
THOMPSON, "Otolith ultrastructure of smooth oreo, Psevdocyt-
tus macuiatus, and black oreo, Allocyttus sp., species," 499
DEW, C. BRAXTON, "Stomach contents of commercially caught
Hudson River striped bass, Morone saxatilis, 1973-1975," 397
Diet— see Food habits
Disease studies
oysters, American, 553
Dissodactylus juveniles
bathymetric and geographic distribution, 67
Dissolved oxygen levels
prawn mortality, 601
"The distribution, abundance, and transport of larval sciaenids col-
lected during winter and early spring from the continental shelf
waters off west Louisiana," by James H. Cowan, Jr. and Richard
F. Shaw, 129
"Distribution and abundance of bottlenose dolphin, Tursiops trun-
catus (Montagu, 1821), in Virginia," by Robert A. Blaylock, 797
DITTY, JAMES G., GLEN G. ZIESKE, and RICHARD F.
SHAW, "Seasonality and depth distribution of larval fishes in the
northern Gulf of Mexico above latitude 26°00'N," 811
DOERZBACHER, JEFF F.-see McEACHRON et al.
Dolly Varden
salmon predation, 613
Dolphin, bottlenose
population studies, 797
Dolphin, spotted
sightings within tuna, purse seine cruises, 351
DUTIL, J-D., and J.-M. Coutu, "Early marine life of Atlantic
salmon, Salmo salar, postsmolts in the northern Gulf of St.
Lawrence," 197
Early marine life of Atlantic salmon, Salmo salar. postsmolts in
the northern Gulf of St. Lawrence," by J.-D. Dutil and J.-M.
Coutu, 197
"An econometric analysis of net investment in Gulf shrimp fishing
vessels," by John B. Penson, Jr., Ernest 0. Tetty, and Wade L.
Griffin, 151
Economic studies
fishery capital stock and investment approach, 339
shrimp fishing vessels, 151
tuna, "burnt", 367
"The effect of the ectoparasitic pyramidellid snail, Boonea im-
pressa, on the growth and health of oysters, Crassostrea virginica,
under field conditions," by Elizabeth A. Wilson, Eric N. Powell,
and Sammy M. Ray, 553
"The effects of siltation on recruitment of spiny lobsters,
Panulirus argus," by William F. Herrnkind, Mark J. Butler IV,
and Richard A. Tankersley, 331
849
El Nino
ichthyoplankton off Chile, 1
"Electrophoretic identification of early juvenile yellowfin tuna,
Thunnus albacares," by John E. Graves, Marie A. Simovich, and
Kurt M. Schaefer, 835
ELSTON, RALPH A.-see FARLEY et al.
EMLEN, JOHN M.-see McINTYRE et al.
English sole
population studies, 826
Engraulis mordax—see Anchovy, northern
Environmental studies
ichthyoplankton off Chile, 1
rockfish, blue, 715
see also Feeding habitats; Habitat studies
Enzymes
tuna, "burnt", 367
EPIFANIO, CHARLES E., DAVID GOSHORN, and TIMOTHY
E. TARGETT, "Induction of spawning in the weakfish, Cynos-
cion regalis," 168
ERZINI, KARIM-see CASTRO and ERZINI
"Estimation of natural mortality in fish stocks: A review," by E. F.
Vetter, 25
Estuarine studies
shrimp, penaeid, 543
Eubalaena avstralis—see Whales, southern right
Euchirogra'psus americanus
bathymetric and geographic distribution, 67
Ewphausia superba—see Krill, Antarctic
"Evaluation of variability in sablefish, Anoplopoma fimbria, abun-
dance indices in the Gulf of Alaska using the bootstrap method,"
by Michael F. Sigler and Jeffrey T. Fujioka, 445
"Experimental manipulation of population density and its effects
on growth and mortality of juvenile western rock lobsters,
Panulirus cygnus George," by Richard F. Ford, Bruce F. Phillips,
and Lindsay M. Joil, 773
"Extractable lipofuscin in larval marine fish," by M. M. Mullin and
E. R. Brooks, 407
FARLEY, C. AUSTIN, PETER H. WOLF, and RALPH A.
ELSTON, "A long-term study of 'microcell' disease in oysters
with a description of a new genus, Mikrocytos (g. n.), and two new
species, Mikrocytos mackini (sp. n.) and Mikrocytos roughleyi (sp.
n.)," 581
"The feeding habits of two deep slope snappers, Pristipomoides
zonatus and P. auricilla, at Pathfinder Reef, Mariana Archi-
pelago," by Michael P. Seki and Michael W. Callahan, 807
Feed habitats
mesotide areas, 703
see also Food habits
"A field method for determining prey preferences of predators,"
by N. B. Hargreaves, 763
FELDER, DARRYL L.-see MARTIN et al.
FINN, JAMES E.-see McINTYRE et al.
Fish habitats— see Habitat studies
FISHER, JOSEPH P.-see PEARCY and FISHER
Fishery, commercial
capital stock and investment approach, 339
flounder, yellowtail, 91
Hawaiian handline, 367
reducing bycatch using trotline, 109
tuna, 351, 367
Fishes
associations with crab megalopae, 299
Flatfish
population studies, 826
Flounder, arrowtooth
morphological comparison, 608
Flounder, gulf
growth studies, 832
Flounder, kamchatka
morphological comparison, 608
Flounder, yellowtail
variation in CPUE, 91
Food habits
bass, striped, 397
croaker, white, 251
rockfish, blue, 717
salmon. Pacific, 213
salmon postsmolts, 197
salmon predation, 613
shark, scyliorhinid, 697
shrimp, penaeid, 543
snapper, 807
whales, southern right, 143
"Food habits and daily ration of Greenland halibut, Reinhardtius
hippoglossoides, in the eastern Bering Sea," by M. S. Yang and
P. A. Livingston, 675
"Food pathways associated with penaeid shrimps in a mangrove-
fringed estuary," by Allan W. Stoner and Roger J. Zimmerman,
543
FORD, RICHARD F., BRUCE F. PHILLIPS, and LINDSAY M.
JOLL, "Experimental manipulation of population density and its
effects on growth and mortality of juvenile western rock lobsters,
Panulirus cygnus George," 773
FUJIOKA, JEFFREY T.-see SIGLER and FUJIOKA
850
"Functional regression equations for zooplankton displacement
volume, wet weight, dry weight, and carbon: A correction," by
Peter H. Wiebe, 833
"Further support for the hypothesis that internal waves can cause
shoreward transport of larval invertebrates and fish," by Alan L.
Shanks, 703
Gadus morhua—see Cod, Atlantic
GAULDIE, R. W.-see DA VIES et al.
Gear, trawls
crab sampling, 529
Genetic studies
salmon, chum, 665
Genyonemus lineatus—see Croaker, white
Glycogen, muscle
spawning of sea scallop, 597
GOSHORN, DAVID-see EPIFANIO et al.
GOULD. EDITH, DIANE RUSANOWSKY, and DONNA A.
LUEDKE, "Note on muscle glycogen as an indicator of spawn-
ing potential in the sea scallop, Placopecten magellanicus," 597
GRAVES, JOHN E., MARIE A. SIMOVICH, and KURT M.
SCHAEFER, "Electrophoretic identification of early juvenile
yellowfin tuna, Thunnus albacares," 835
GREEN, ALBERT W.-see McEACHRON et al.
Grenadier, blue
larval development, 119
GRIFFIN, WADE L.-see PENSON et al.
GRIMES, CHURCHILL B., CHARLES F. IDELBERGER, KEN-
NETH W. ABLE, and STEPHEN C. TURNER, "The reproduc-
tive biology of tilefish, Lopholatilus chamaeleontice-ps Goode and
Bean, from the United States Mid-Atlantic Bight, and the effects
of fishing on the breeding system," 745
"Growth of Pacific saury, Cololabis saira, in the northeastern and
northwestern Pacific Ocean," by Yoshiro Watanabe, John L.
Butler, and Tsukasa Mori, 489
Growth studies
cod, Atlantic, 223
flounder, gulf, 832
haddock, 223
lobster, American, 789
lobster, spiny, 775
menhaden, gulf, 77
saury. Pacific, 489
tropical fish, 645
"Growth through the first six months of Atlantic cod, Gadus
morhua, and haddock, Melanogrammus a^glefinus, based on daily
otolith increments," by George R. Bolz and R. Gregory Lough,
223
GUNDERSON, DONALD R.-see ROGERS et al.
Habitat studies
fishing efffects, 746
mesotide areas, 703
lobster, American, 791
rockfish, blue, 719
Haddock
growth, 223
Halibut, Greenland
food habits, 675
HAMNER, WILLIAM M., GREGORY S. STONE, and BRYAN
S. OBST, "Behavior of southern right whales, Eubalaena
aiistralis, feeding on the Antarctic krill, Euphausia superba,"
143
Haplosporidium nelsoni
parasitic effects on oysters.
553
"Harbor porpoise, Phocoena phocoena, abundance estimation for
California, Oregon, and Washington: I. Ship surveys," by Jay
Barlow, 417
"Harbor porpoise, Phocoena phocoena, abundance estimation for
California, Oregon, and Washington: II. Aerial sun^eys," by Jay
Barlow, Charles W. Oliver, Terry D. Jackson, and Barbara L.
Taylor, 433
HARGREAVES, N. B., "A field method for determining prey
preferences of predators," 763
Hatchery studies
salmon, coho, 173
smolt release, 655
Hedonic approach
fishery capital stock and investment.
339
HERRNKIND, WILLIAM F., MARK J. BUTLER IV, and
RICHARD T. TANKERSLEY, "The effects of siltation on
recruitment of spiny lobsters, Panulirus argus," 331
HOBSON, EDMUND S., and JAMES R. CHESS, "Trophic rela-
tions of the blue rockfish, Sebastes mystinus, in a coastal upwell-
ing system off northern California," 715
Homarus americanus—see Lobster, American
HOSS, DONALD E., LINDA COSTON-CLEMENTS, DAVID S.
PETERS, and PATRICIA A. TESTER, "Metabolic responses of
spot, Leiostomus xanthwnis, and Atlantic croaker, Micropogonias
undulatus, larvae to cold temperatures encountered follownng
recruitment to estuaries," 483
Humboldt Current area
ichthyoplankton off Chile, 1
HUNTER, JOHN R.-see VERNET et al.
Hydrothermal discharge
Alvinocaris spp., 263
851
Ichthyoplankton— see Plankton studies
IDELBERGER, CHARLES F.-see GRIMES et al.
Identification methods
tuna, yellowfin, 835
"Induction of spawning in the weakfish, Cynoscion regalis," by
Charles E. Epifanio, David Goshorn, and Timothy E. Targett, 168
"Integrity of schools of skipjack tuna, Katsuwonus pelamis, in the
eastern Pacific Ocean, as determined from tagging data," by
William H. Bayliff, 631
"Interannual variation of ichthyoplankton composition and abun-
dance relations off northern Chile, 1964-85," by Valerie J. Loeb
and Omar Rojas, 1
JAHN, A. E., D. M. GADOMSKI, and M. L. SOWBY, "On the
role of food-seeking in the suprabenthic habit of larval white
croaker, Genyonemus lineatus (Pisces: Sciaenidae)," 251
JAMIESON, GLEN S., and ANTAN C. PHILLIPS, "Occurrence
of Cancer crab (C. magister and C. oregonensis) magalopae off the
west coast of Vancouver Island, British Columbia," 525
JAMIESON, GLEN S., and ELLEN K, PIKITCH, "Vertical
distribution and mass mortality of prawns, Pandalus platyceros,
in Saanich Inlet, British Columbia," 601
JOLL, LINDSAY M.-see FORD et al.
JOUNG, SHOOU-JENG-see CHEN et al.
Kingfish, southern
distribution and abundance, 129
KIRKLEY, JAMES E., and DALE E. SQUIRES, "A hmited in-
formation approach for determining capital stock and investment
in a fishery," 339
KOSKI, K V.-see MURPHY et al.
Krill, Antarctic
feeding of whales, 143
Larimtcs fasciatus—see Banded drum
"Larval development of blue grenadier, Macruronus novaeze-
landiae (Hector), in Tasmanian waters," by B. D. Bruce, 119
Larval studies
cod, Atlantic, 223
continental shelf, 703
copepod nauplii, 319
crab, cancer, 525
crab, Dungeness, 299
crab. Gulf stone, 289
croaker, Atlantic, 483
croaker, white, 251
fishes, 299
grenadier, blue, 119
852
haddock, 223
ichthyoplankton off Chile, 1
lipofucin for ageing. 407
menhaden, gulf, 77
mesotide areas, 703
northern Gulf of Mexico, 811
sardine. Pacific, 163
sciaenids, 129
spot, 483
Laurencia spp.— see Algae, benthic
LeiostoTmis xanthurus—see Spot
"Length- weight relationships for gulf flounder, Paralirhthys
albigutta, from North Carolina," by Glen W. Safrit and Frank J.
Schwartz, 832
LEU, TZYH-CHANG-see CHEN et al.
LI, HIRAM M.-see CURRENS et al.
Life history studies
clams, soft shell, 567
juvenile Pacific salmon, 213
shark, scyliorhinid, 691
"Life tables for two field populations of soft-shell clam, Mya
arenaria, (Mollusca: Pelecypoda) from Long Island Sound," by
Diane J. Brousseau and Jenny A. Baglivo, 567
"A limited information approach for determining capital stock and
investment in a fishery," by James E. Kirkley and Dale E.
Squires, 339
LIN, BIING-HWAN, and NANCY A. WILLIAMS, "Specifying
a functional form for the influence of hatchery smolt release on
adult salmon production," 655
Line transect methods
porpoise, harbor, 429
Lipofuscin
ageing methods, 401
larval marine fish, 407
Lithodes maja
bathymetric and geographic distribution, 67
LIVINGSTON, P. A.-see YANG and LIVINGSTON
Lobster, American
behavior and survival, 789
Lobster, spiny
population studies, 773
siltation and recruitment, 331
LOEB, VALERIE J., and OMAR ROJAS, "Interannual varia-
tion of ichthyoplankton composition and abundance relations off
northern Chile, 1964-85," 1
"A long-term study of 'microcell' disease in oysters with a descrip-
tion of a new genus, Mikrocytos (g. n.), and two new species,
Mikrocytos mackini (sp. n.) and Mikrocytos roughleyi (sp. n.)," by
C. Austin Farley, Peter H. Wolf, and Ralph A. Elston, 581
"A long-term study on the behavior and survival of early juvenile
American lobster, Homarus ainericanus, in three naturalistic
substrates: eelgrass, mud, and rocks," by Diana E. Barshaw and
Donald R. Bryant-Rich, 789
Lopholatilus chamaeleonticeps—see Tilefishes
LOUGH, R. GREGORY-see BOLZ and LOUGH
LUEDKE, DONNA A. -see GOULD et al.
Mackerel, king
occurrence of in shrimp trawls, 394
Mackerel, Spanish
occurrence of in shrimp trawls, 394
Macruronus novaezelandiae—see Grenadier, blue
MARTIN, JOEL W., FRANK M. TRUESDALE, and DARRYL
L. FELDER, "The megalopa stage of the Gulf stone crab,
Menippe adina Williams and Felder, 1986, with comparison of
megalopae in the genus Menippe," 289
Mathematical methods
abundance survey, 448
chi-square tests, 634
CPUE for yellowtail flounder, 91
investment in shrimp fishing vessels, 151
length-frequency distribution, 645
mortality in fish stocks, 25
population estimates, 617
regression equations, 833
salmon survival, 655
MATLOCK, GARY C.-see McEACHRON et al.
MAYO, RALPH K.-see O'Brien and MAYO
McEACHRON, LAWRENCE W., JEFF F. DOERZBACHER,
GARY C. MATLOCK, ALBERT W. GREEN, and GARY E.
SAUL, "Reducing the bycatch in a commercial trotline fishery,"
109
McINTYRE, JOHN D., REGINALD R. REISENBICHLER,
JOHN M. EMLEN, RICHARD L. WILMOT, and JAMES E.
FINN, "Predation of Karluk River sockeye salmon by coho
salmon and char," 611
"The megalopa stage of the Gulf stone crab, Menippe adina
Williams and Felder, 1986, with comparison of megalopae in the
genus Menippe," by Joel W. Martin, Frank M. Truesdale, and
Darryl L. Felder, 289
Melanogrammus aeglefinus—see Haddock
MENDELSSOHN, ROY, "Some problems in estimating popula-
tion sizes from catch-at-age data," 617
Menhaden, gulf
age and growth, 77
Menippe adina— see Crab, Gulf stone
Menticirrhus americanussee Kingfish, southern
"Metabolic responses of spot, Leiostomus xanthurus, and Atlan-
tic croaker, MicropogonioH undulatus, larvae to cold temperatures
encountered following recruitment to estuaries," by Donald E.
Hoss, Linda Coston-Clements, David S. Peters, and Patricia A.
Tester, 483
"Microcell" disease
oysters, 581
Micropogonias undulatiis—see Croaker, Atlantic
Microstomias pacificics—see Sole, Dover
Migration studies
salmon, Atlantic, 197
salmon, coho, 173
salmon. Pacific, 213
"Migrations of coho salmon, Oncorhynchus kisutch, during their
first summer in the ocean," by William G. Pearcy and Joseph P.
Fisher, 173
Mikrocytos mackini
"microcell" disease in oysters, 581
Mikrocytos roughleyi
"microcell" disease in oysters, 581
Mobula japanica
key to species, 56
Mobula munkiana
key to species, 60
Mobula, spp.— see Rays, mobulid
Mobula tarapacana
key to species, 62
Mobula thurstoni
key to species, 49
Monte Carlo method
abundance survey method, 448
MORI, TSUKASA-see WATANABE et al.
Morone axatilis—see Bass, striped
"Morphological differences between two congeneric species of
pleuronectid flatfishes: Arrowtooth flounder, Atheresthes stomias,
and Kamchatka flounder, A. evermanni," by Mei-Sun Yang, 608
Mortality
estimates, 671
length-frequency analysis, 645
lobster, spiny, 777
methods of fish stock estimation, 25
MULLIN, M. M., and E. R. BROOKS, "Extractable lipofuscin
in larval marine fish," 407
Munida forceps
bathymetric and geographic distribution, 67
853
Munida longipes
bathymetric and geographic distribution, 67
Munidopsis alvisca
new decapod species, 263
Munids
abundance, 602
dissolved oxygen levels, 604
MURPHY, MICHAEL L., JOHN F. THEDINGA, and K V.
KOSKI, "Size and diet of juvenile Pacific salmon during seaward
migration through a small estuary in southeastern Alaska," 213
MURRAY, CLYDE B.-see BEACHAM et al.
Mya arenaria—see Clam, soft-shell
"Natural history of the rays on the genus Mobula in the Gulf of
California," by Guiseppe Notarbartolo-di-Sciara, 45
"New marine decapod crustaceans from waters influenced by
hydrothermal discharge, brine, and hydrocarbon seepage," by
Austin B. Williams, 263
NOTARBARTOLO-DI-SCIARA, GUISEPPE, "Natural history of
the rays on the genus Mobula in the Gulf of California," 45
"Note on muscle glycogen as an indicator of spawning potential
in the sea scallop, Placopecten magellanicus," by Edith Gould,
Diane Rusanowsky, and Donna A. Luedke, 597
"Notes on decapod and euphausiid crustaceans, continental
margin, western Atlantic, Georges Bank to western Florida,
USA," by Austin B. Williams, 67
"Notes on reproduction in the scalloped hammerhead, Sphyrjia
leuiini, in northeastern Taiwan waters," by Che-Tsung Chen, Tzyh-
Chang Leu, and Shoou-Jeng Joung, 389
NYMAN, ROBERT M., AND DAVID 0. CONOVER, "The rela-
tion between spawning season and the recruitment of young-of-
the-year bluefish, Pomatomus saltatrix, to New York," 237
O'BRIEN, LORETTA, and RALPH K. MAYO, "Source of varia-
tion in catch per unit effort of yellowtail flounder, Limanda fer-
ruginea (Storer), harvested off the coast of New England," 91
OBST, BRYAN S.-see HAMNER et al.
"Occurrence of young-of-the-year king, Scomberomorus cavalla,
and Spanish, S. maculatus, mackerels in commercial-type shrimp
trawls along the Atlantic coast of the southeast United States,"
by Mark R. Collins and Charles A. Wenner, 394
"Occurrence of Cancer crab (C. magister and C. oregonensis)
megalopae off the west coast of Vancouver Island, British Colum-
bia," by Glen S. Jamieson and Antan C. Phillips, 525
"Oceanographic associations of neustonic larval and juvenile fishes
and Dungeness crab megalopae off Oregon," by Jonathan M.
Shenker, 299
854
"On the role of food-seeking in the suprabenthic habit of larval
white croaker, Genyonemus lineatics (Pisces: Sciaenidae)," by A. E.
Jahn, n. M. Gadomski, and M. L. Sowby, 251
Onchorhynchus gorhuscha—see Salmon, pink
Oncorhynchus keta—see Salmon, chum
Oncorhynchus kisutch—see Salmon, coho
Oncorhynchus nerka—see Salmon, sockeye
Oncorhynchus spp.— see Salmon, Pacific
Oreo, black
otolith ultrastructure, 499
Oreo, smooth
otolith ultrastructure, 499
ORTNER, P. B.-see DAGG et al.
Ostrea edulis—see Oyster, flat
Ostrea lusida—see Oyster, Olympia
"Otolith ultrastructure of smooth oreo, Pseudocyttus maculatus,
and black oreo, Allocyttus sp., species," by N. M. Davies, R. W.
Gauldie, S. A. Crane, and R. K, Thompson, 499
Otoliths
bluefish, 237
cod, Atlantic, 223
haddock, 223
menhaden, gulf, 77
oreo, black, 499
oreo, smooth, 499
saury. Pacific, 489
trout, rainbow, 160
Oyster, American
growth and health, 553
Oyster, flat
"microcell" disease, 581
Oyster, Japanese
"microcell" disease, 581
Oyster, Olympia
"microcell" disease, 581
Pandalus jordanisee Shrimp, pink
Pandalus platyceros—see Prawn, spot
Panulirus argussee Lobster, spiny
Panulirus cygnus—see Lobster, spiny
Parasite studies
oyster growth, 553
oysters, 581
Parmaturus xaniurus—see Shark, scyliorhinid
Parophrys vetuliissee English sole
Pathology
oysters, 581
PEARCY, WILLIAM G., and JOSEPH P. FISHER. "Migrations
of coho salmon, Oncorhynchus kisutch, during their first summer
in the ocean," 173
"Pelagic biogeography of the armorhead, Pseudopentaceros ivheel-
eri, and recruitment to isolated seamounts in the North Pacific
Ocean," by George W. Boehlert and Takashi Sasaki, 453
Penaeus brasiliensis—see Shrimps, penaeid
Penaeus notialis—see Shrimps, penaeid
Penaeus subtilis—see Shrimps, penaeid
PENSON, JOHN B., JR., ERNEST 0. TETTY, and WADE L.
GRIFFIN, "An econometric analysis of net investment in Gulf
shrimp fishing vessels," 151
Perkinsus marinus
parasitic effects on oysters, 553
PETERS, DAVID S.-see HOSS et al.
PHILLIPS, ANTAN C.-see JAMIESON and PHILLIPS
PHILLIPS, BRUCE F.-see FORD et al.
Pkocoena phocoena—see Porpoise, harbor
PICKETT, DARLENE-see BUTLER and PICKETT
PIKITCH, ELLEN K.-see JAMIESON and PIKITCH
Placopecten magellanicus—see Scallop, sea
Plankton studies, 1, 129, 299, 704, 723, 811, 833, 838
Pogonias cromissee Black drum
POLACHECK, TOM, "Analyses of the relationship between the
distribution of searching effort, tuna catches, and dolphin sightings
within individual purse seine cruises," 351
Polychaetes, capitellid
shrimp diets, 543
Pomatomus saltatrix—see Bluefish
Population studies
dolphin, bottlenose, 797
English sole, 823
larval fishes. 811
lobster, spiny, 775
mackerel, Pacific, 622
methods, 617
porpoise, harbor, 417, 433
sablefish, 445
salmon, chum, 663
shark, scyliorhinid, 694
Porpoise, harbor
abundance estimation, 417, 433
Postsmolts
salmon, Atlantic, 197
POWELL, ERIC N.-see WILSON et al.
Prawn, spot
distribution and mortality, 601
Predation
salmon, 763
sardine larvae, 163
"Predation of Karluk River sockeye salmon by coho salmon and
char," by John D. Mclntyre, Reginald R. Reisenbichler, John M.
Emlen, Richard L. Wilmot. and James E. Finn, 611
Pristipomoides auricilla—see Snapper
Pristipomoides zonatus—see Snapper
Pseudocyttus maculatus—see Oreo, smooth
Pseudopentaceros wheeleri—see Armorhead, pelagic
Purse seine
salmon, coho, 173
searching effort, tuna catches, and dolphin sightings, 351
RAY, SAMMY M.-see WILSON et al.
Rays, mobulid
natural history, 45
key to species, 48
Red drum
reducing bycatch, 109
"Reducing the bycatch in a commercial trotline fishery," by
Lawrence W. McEachron, Jeff F. Doerzbacher, Gary C. Matlock,
Albert W. Green, and Gary E. Saul, 109
"Reexamination of the use of otolith nuclear dimensions to iden-
tify juvenile anadromous and nonanadromous rainbow trout, Salmo
gairdneri," by Kenneth P. Currens, Carl B. Schreck, and Hiram
M. Li, 160
Reinhardtius hippoglossoides—see Halibut, Greenland
REISENBICHLER, REGINALD R.-see McINTYRE et al.
"The relation between spawning season and the recruitment of
young-of-the-year bluefish, Pomatomus saltatrix, to New York,"
by Robert M. Nyman and David 0. Conover, 237
"Relationship between sediment character and sex segregation in
English sole, Parophrys vetulv^," by D. Scott Becker, 517
"Reproductive biology of the spotted seatrout, Cynoscion nebu-
losus, in South Texas," by Nancy Brown-Peterson, Peter Thomas,
and Connie R. Arnold, 373
855
"The reproductive biology of tilefish, Lopholatilus chamaeleon-
ticeps Goode and Bean, from the United States Mid-Atlantic Bight,
and the effects of fishing on the breeding system," by Churchill
B. Grimes, Charles F. Idelberger, Kenneth W. Able, and Stephen
C. Turner, 745
Reproductive studies
bluefish, 237
larval fishes, 811
oyster, American, 553
scallop, sea, 597
seatrout spawning, 129
seatrout, spotted, 373
shark, hammerhead, 389
tilefishes, 745, 752
weakfish, 168
Rockfish, blue
habitat studies, 715
ROGERS, CHRISTOPHER W., DONALD R. GUNDERSON, and
DAVID A. ARMSTRONG, "Utilization of a Washington estuary
by juvenile English sole, Parophrys vetulus," 823
ROJAS, OMAR-see LOEB and ROJAS
RUSANOWSKY, DIANE-see GOULD et al.
Sablefish
abundance indices, 445
Saccostrea commercialis—see Oyster, flat
SAFRIT, GLEN W., and FRANK J. SCHWARTZ, "Length-
weight relationships for gulf flounder, Paralichthys albigutta, from
North Carolina," 832
Salmo gairdneri—see Trout, rainbow
Salmo salar—see Salmon, Atlantic
Salmon, Atlantic
postsmolts, 197
Salmon, chum
population spawning, 663
size and diet of juveniles, 213
smolts and adult production, 655
Salmon, coho
ocean migrations, 173
predation, 611
predators, 763
size and diet of juveniles, 213
smolts and adult production, 655
Salmon, Pacific
size and diet of juveniles, 213
Salmon, pink
predators, 763
size and diet of juveniles, 213
Salmon, sockeye
predation, 611
856
Salvelinus alpinus—see Char, Arctic
Salvelinus malma—see Dolly Varden
Sampling techniques
plankton, 838
sole, English, 517
Sardine, Pacific
vulnerability to predation, 163
Sardinops sagax—see Sardine, Pacific
SASAKI, TAKASHI-see BOEHLERT and SASAKI
SAUL, GARY E.-see McEACHRON et al.
Saury, Pacific
growth, 489
SAVOY, THOMAS F., and VICTOR A. CRECCO, "The timing
and significance of density-dependent and density-independent
mortality of American shad, Alosa sapidissima," 467
Scallop, sea
spawning potential, 597
SCHAEFER, KURT M.-see GRAVES et al.
Schooling
tuna, skipjack, 631
SCHRECK, CARL B.-see CURRENS et al.
Sciaenops ocellatus—see Red drum
Scomberomorus cavalla—see Mackerel, king
Scomberomorus maculatus—see Mackerel, Spanish
"Seasonality and depth distribution of larval fishes in the northern
Gulf of Mexico above latitude 26°00'N," by James G. Ditty, Glen
G. Zieske, and Richard F. Shaw, 811
Seatrout, sand
larval spawning, 129
Seatrout, spotted, 109
reproductive biology, 373
Sebastes mystinus—see Rockfish, blue
Sediment character
sex segregation in English sole, 517
SEKI, MICHAEL P., and MICHAEL W. CALLAHAN, "The
feeding habits of two deep slope snappers, Pristipomoides zonatus
and P. auricilla, at Pathfinder Reef, Mariana Archipelago," 807
Sex segregation
sole, EngHsh, 517
Shad, American
mortality rates, 467
SHANKS, ALAN L., "Further support for the hypothesis that
internal waves can cause shoreward transport of larval in-
vertebrates and fish," 703
Shark, scyliorhinid
biology, 691
Sharks, hammerhead
reproduction, 389
SHAW, RICHARD F.-see DITTY et al.
SHAW, RICHARD F.-see COWAN AND SHAW
SHENKER, JONATHAN M., "Oceanographic associations of
neustonic larval and juvenile fishes and Dungeness crab megalopae
off Oregon," 299
SHIRLEY, SUSAN M., and THOMAS C. SHIRLEY, "Appendage
injury in Dungeness crabs, Cancer magister, in southeastern
Alaska," 156
SHIRLEY, THOMAS C.-see SHIRLEY and SHIRLEY
Shrimp, gulf
investment in fishing vessels, 151
Shrimp, pink
abundance, 603
dissolved oxygen levels, 604
Shrimp, roughback
cojoined (two headed) specimen, 595
Shrimps, penaeid
food pathways, 543
SIGLER, MICHAEL F., and JEFFREY T. FUJIOKA, "Evalua-
tion of variability in sablefish, Anoplopoma fimbria, abundance in-
dices in the Gulf of Alaska using the bootstrap method," 445
SIMOVICH, MARIE A.-see GRAVES et al.
"Size and diet of juvenile Pacific salmon during seaward migra-
tion through a small estuary in southeastern Alaska," by Michael
L. Murphy, John F. Thedinga, and K V. Koski, 213
Snail, pyramidellid
parasitic effects on oysters, 553
Snapper
food habits, 807
Sole, Dover
lipofuscin for ageing, 401
Sole, English
sex segregation, 517
"Some problems in estimating population sizes from catch-at-age
data," by Roy Mendelssohn, 617
"Sources of variation in catch per unit effort of yellowtail flounder,
Limanda ferruginea (Storer), harvested off the coast of New
England," by Loretta O'Brien and Ralph K. Mayo, 91
Spawning season
bluefish, 237
Spawning studies— see Reproductive studies
"Specifying a functional form for the influence of hatchery smelt
release on adult salmon production," by Biing-Hwan Lin and Nancy
A. Williams, 655
Sphyma lewini—see Sharks, hammerhead
Sphyma mokarran—see Sharks, hammerhead
Sphyma zygaena—see Sharks, hammerhead
Spot
distribution and abundance, 129
metabolic responses to temperatures, 483
SQUIRES, DALE E.-see KIRKLEY and SQUIRES
Stenella attenuata—see Dolphin, spotted
Stomach contents
halibut, Greenland, 676
"Stomach contents of commercially caught Hudson River striped
bass, Morone saxatilis, 1973-1975," by C. Braxton Dew, 397
STONE, GREGORY S.-see HAMNER et al.
STONER, ALLAN W., and ROGER J. ZIMMERMAN, "Food
pathways associated wdth penaeid shrimps in a mangrove-fringed
estuary," 543
Submersibles
decapod crustaceans, 67
prawn distribution, 602
Survey, Pacific
growth, 489
Surveys
aerial, 433
bootstrap, 445
ichthyoplankton off Chile, 1
lobster, spiny, 331
porpoise, harbor, 417
ship, 417
sole, English, 517
Tagging studies
tuna, skipjack, 631
TANKERSLEY, RICHARD T.-see HERRNKIND et al.
TARGETT, TIMOTHY E.-see EPIFANIO et al.
Temperature studies
spot. 483
croaker, Atlantic, 483
TESTER, PATRICIA A.-see ROSS et al.
TETTY, ERNEST O.-see PENSON et al.
857
THEDINGA, JOHN F.-see MURPHY et al.
THOMAS, PETER-see BROWN-PETERSON et al.
THOMPSON, R. K.-see DAVIES et al.
Thunnus albacares—see Tuna, yellowfin
Thunnus obesus—see Tuna, bigeye
Tide studies, 703
Tilefishes
reproductive biology, 745
"The timing and significance of density-dependent and density-
independent mortality of American shad, Alosa sapidissima," by
Thomas F. Savoy and Victor A. Crecco, 467
Trachypenaeus similis—see Shrimp, roughback
Trawls, shrimp
occurrence of mackerel, 394
TRONZO, CRAIG R.-see CAHOON and TRONZO
"Trophic relations of the blue rockfish, Sebastes mystinus, in a
coastal upwelling system off northern California," by Edmund S.
Hobson and James R. Chess, 715
Trotlines, 109
Trout, rainbow
lipofuscin for ageing, 401
otolith identification, 160
TRUESDALE, FRANK M.-see MARTIN et al.
Tuna
distribution within purse seine cruises, 351
Tuna, bigeye
burnt tuna 367
Tuna, "burnt"
etiology, 367
Tuna, skipjack
schooling, 631
tag studies, 631
Tuna, yellowfin
burnt tuna, 367
identification of juveniles, 835
Turbot— see Halibut, Greenland
TURNER, STEPHEN C.-see GRIMES et al.
Tursiops tumcatiLs—see Dolphin, bottlenose
Upwelling and downwelling
northern California coast, 715
South Atlantic Bight, 703
"Utilization of a Washington estuary by juvenile English sole,
Paropkryts vetulus," by Christopher W. Rogers, Donald R. Gunder-
son, and David A. Armstrong, 823
VERNET, MARIA, JOHN R. HUNTER, and RUSSELL D.
VETTER, "Accumulation of age pigments (lipofuscin) in two
cold-water fishes," 401
"Vertical distribution and mass mortality of prawns, Pandaltis
platyceros, in Saanich Inlet, British Columbia," by Glen S.
Jamieson and Ellen K. Pikitch, 601
VETTER, E. F., "Estimation of natural mortality in fish
stocks: a review," 25
VETTER, RUSSELL D.-see VERNET et al.
WARLEN, STANLEY M., "Age and growth of larva! gulf
menhaden, Brevoortia patronus, in the northern Gulf of
Mexico," 77
WATANABE, YOSHIRO, JOHN L. BUTLER, and TSUKASA
MORI, "Growth of Pacific saury, Cololabis saira, in the north-
eastern and northwestern Pacific Ocean," 489
WATSON, CHERYL, ROBERT E. BOURKE, and RICHARD W.,
BRILL, "A comprehensive theory on the etiology of burnt
tuna," 367
Wave slicks, 703
Weakfish
induction of spawning, 168
WENNER, CHARLES A.-see COLLINS and WENNER
Western rock lobster— see Lobster, spiny
Whales, southern right
behavior, 143
feeding, 143
WIEBE, PETER H., "Functional regression equations for
zooplankton displacement volume, wet weight, dry weight, and
carbon: a correction," 833
WILLIAMS, AUSTIN B., "Cojoined twin adult shrimp
(Decapoda: Penaeidae)," 595
WILLIAMS, AUSTIN B., "New marine decapod crustaceans
from waters influenced by hydrothermal discharge, brine, and
hydrocarbon seepage," 263
WILLIAMS, AUSTIN B., "Notes on decapod and euphausiid
crustaceans, continental margin, western Atlantic, Georges Bank
to western Florida, USA," 67
WILLIAMS, NANCY A.-see LIN and WILLIAMS
WILMOT, RICHARD L.-see McINTYRE et al.
WILSON, ELIZABETH A., ERIC N. POWELL, and SAMMY M.
RAY, "The effect of the ectoparasitic pyramidellid snail, Boonea
858
impressa, on the growth and health of oysters, Crassostrea
virginica, under field conditions," 553
"Winter-time distribution and abundance of copepod nauplii in the
northern Gulf of Mexico," by M. J. Dagg, P. B. Ortner, and J.
Al-Yamani, 319
WITHLER, RUTH E.-see BEACHAM et al.
WOLf^ PETER H.-see FARLEY et al.
YANG, M. S., and P. A. LIVINGSTON, "Food habits and daily
ration of Greenland halibut, Reinhardtius hijypoglossoides, in the
eastern Bering Sea," 675
YANG, MEI-SUN, "Morphological differences between two con-
generic species of pleuronectid flatfishes: Arrowtooth flounder,
Atheresthes stomias, and Kamchatka flounder, A. evermanni,"
608
Yukon River salmon, 663
ZIESKE, GLEN G.-see DITTY et al.
ZIMMERMAN, ROGER J.-see STONER and ZIMMERMAN
Zooplankton— see Plankton studies
859
NOTICE
NOAA Technical Reports NMFS published during first 6 months of 1988.
63. Stock assessment of the Atlantic menhaden, Brevortia
tyrannus fishery. By Douglas S. Vaughan and Joseph
W. Smith. January 1988, iii + 18 p., 13 tables, 17 figs.
64. Illustrated key to penaeoid shrimps of commerce in the
Americas. By Isabel Perez Farfante. April 1988, iv +
32 p., 49 figs.
65. History of whaling in and near North Carolina. By Ran-
dall R. Reeves and Edward Michell. March 1988, iii +
28 p., 5 tables, 10 figs.
66. Atlas and zoogeography of common fishes in the Bering
Sea and northeastern Pacific. By M. James Allen and
Gary B. Smith. April 1988, iii + 151 p., 8 tables, 4 figs.
Some NOAA publications are available by the purchase from the Superinten-
dent of Documents, U.S. Government Printing Office, Washington, D.C.
20402.
860
ERRATA
Fishery Bulletin Vol. 86, No. 3
Hoss, Donald E., Linda Coston-CIements, David S. Peters, and Patricia A. Tester, "Metabolic responses
of spot, Leiostomus xanthurus, and Atlantic croaker, Micropogonias undulatus, larvae to cold temper-
atures encountered following recruitment to estuaries," pages 483-488.
Bottom of page 484 should read as follows:
Figure 1.— Representative water-temperatures, salinities, and developmental stages of larval spot from the spawning area
to Beaufort Inlet, North Carolina. Drawings from Powell and Gordy (1980) and Lippson and Moran (1974).
484
U S Poitsl S«rw>ce
STATEMENT OF OWNERSHIP, MANAGEMENT AND CIRCULATION
lA.TITLE OF PUBLICATION
Fishery Bulletin
IB PUBLICATION NO
3 6 6-370
2 DATE OF FILING
28 Sept. 1988
3. FREQUENCY OF ISSUE
Quarterly
3A. NO OF ISSUES PUBLISHED
ANNUALLY
4
3B ANNUAL SUBSCRIPTION
PRICE
$16.00
4. COMPLETE MAILING ADDRESS OF KNOWN OFFICE OF PUBLICATION ISlrcrl. On. Count): Slair and ZIP*4 Codel (Not frlnten)
Scientific Pu):ilications Office, MtlFS
7600 Sand Point Way NE, BIN C15700, Seattle, WA 98115
5- COMPLETE MAILING ADDRESS OF THE HEADQUARTERS OF GENERAL BUSINESS OFFICES OF THE PUBLISHER (Not prinirr)
National terine Fisheries Service, troAA/DOC
Universal Building, South, 1825 Conn. Ave. , mv, Washington, D.C. 20235
6. FULL NAMES AND COMPLETE MAILING ADDRESS OF PUBLISHER, EDITOR, AND MANAGING EDITOR IThis item MUST NOT be blank/
PUBLISHER /Name and Complete Mailing Address]
Jack ffcCormick, Scientific Publications Office, MIFS
7600 Sand Point Way tlE, BIN C15700, Seattle, WA 9B115
EDITOR (Name and Complete Mailing Address)
l^ri' Fukuyama, Scientific Publications Office, NMFS
7600 Sand Point Way tE, BIN C15700, Seattle, WA 9B115
MANAGING EDITOR /Name and Complete Mailing Address)
same as above.
OWNER (1/ owned by a corporation, its name and address must be staled and also immediately thereunder Die names and addresses of slocliliolders
owning or holding I percent or more of total amount of stock If not owned by a corporation, the names and addresses of the individual owners must
be given If owned by a partnership or other unincorporated firm, its name and address, as well as that of each individual must be given If the publica-
tion IS published by a nonprofit organitalion. its name and address must be slated ) (Item must be completed.)
FULL NAME
COMPLETE MAILING ADDRESS
ri..S. n^rvirtmt^'nt of foirmf^rce
14t-h .St., NW, W,iBhingfr.n, D.C. 20230
KNOWN BONDHOLDERS. MORTGAGEES, AND OTHER SECURITY HOLDERS OWNING OR HOLDING 1 PERCENT OR MORE OF TOTAL
AMOUNT OF BONDS, MORTGAGES OR OTHER SECURITIES (If there are none, so state)
FULL NAME
COMPLETE MAILING ADDRESS
None
9 FOR COMPLETION BY NONPROFIT ORGANIZATIONS AUTHORIZED TO MAIL AT SPECIAL RATES (Section 423 1 2 DMM only)
The purpose, function, and nonprofn stalus of this organization and the exempt status for Federal rncome tax purposes (Check one)
Q
HAS NOT CHANGED DURING
PRECEDING 1 2 MONTHS
D
HAS CHANGED DURING
PRECEDING 12 MONTHS
(If changed, publisher must submit explanario
change with this statement )
EXTENT AND NATURE OF CIRCULATION
(See instructions on reverse side)
AVERAGE NO COPIES EACH
ISSUE DURING PRECEDING
12 MONTHS
ACTUAL NO, COPIES OF SINGLE
ISSUE PUBLISHED NEAREST TO
FILING DATE
A. TOTAL NO. COPIES (Net Press Run)
2094
2111
B. PAID AND/OR REQUESTED CIRCULATION (liandled bv U.S
1 Sales through dealers and earner?, streei vendors and counter sales, i
GPO
n.r. ?fi?4n)
2 Mall Subscription
(Paid and/or requested)
C TOTAL PAID AND/OR REQUESTED CIRCULATION
(Sum of tOBl and I0B2)
600
600
D- FREE DISTRIBUTION BY MAIL, CARRIER OR OTHER MEANS
SAMPLES, COMPLIMENTARY, AND OTHER FREE COPIES
1485
1494
E. TOTAL DISTRIBUTION ^Sumo/Ca/idD;
2085
2094
F. COPIES NOT DISTRIBUTED
1 Office use. left over, unaccounted, spoiled after printing
17
2, Return from News Agents
G. TOTAL (Sum of E. Fl and 2 -should equal net press run show
2094
2111
I certify that the statements made by
me above are correct and complete
ER, BUSlfJESS MANAGER, OR OWNER
PS Form 3526, Dec. 1985
30:
30
INFORMATION FOR CONTRIBUTORS TO THE FISHERY BULLETIN
Manuscripts submitted to the Fishery Bulletin will reach print faster if they conform to the following instructions.
These are not absolute requirements, of course, but desiderata.
CONTENT OF MANUSCRIPT
The title page should give only the title of the paper,
the author's name, his/her affiliation, and mailing address,
including ZIP code.
The abstract should not exceed one double-spaced page.
In the text, Fishery Bulletin style, for the most part,
follows that of the U.S. Government Printing Office Style
Manual. Fish names follow the style of the American
Fisheries Society Special Publication No. 12, A List of Com-
mon and Scientific Names of Fishes from the United States
and Canada, Fourth Edition, 1980.
sibility of the author. Abbreviations of names of periodicals
and serials should conform to Serial Sources for the BIOSIS
Data Base. {Chemical Abstracts also uses this system,
which was developed by the American Standards
Association.)
Common abbreviations and symbols, such as mm, m,
g, L, mL, mg, °C (for Celsius), %, '1m, and so forth,
should be used. Abbreviate units of measure only when
used with numerals. Periods are only rarely used with
abbreviations.
We prefer that measurements be given in metric units;
other equivalent units may be given in parentheses.
Text footnotes should be typed separately from the text.
Figures and tables, with their legends and headings,
should be self-explanatory, not requiring reference to the
text. Their placement should be indicated in the right-hand
margin of the manuscript.
Preferably figures should be reduced by photography
to 5% inches (for single-column figures, allowing for 50%
reduction in printing), or to 12 inches (for double-column
figures). The maximum height, for either width, is 14
inches. Photographs should be printed on high quahty or
glossy paper.
Do not send original drawings to the Scientific Editor;
if they, rather than the photographic reductions, are need-
ed by the printer, the Scientific Publications Office will
request them.
Each table should start on a separate page. Consistency
in headings and format is desirable. Vertical rules should
be avoided, as they make the tables more expensive to
print. Footnotes in tables should be numbered sequentially
in arable numerals. To avoid confusion with powers, they
should be placed to the left of numerals.
Acknowledgments, if included, are placed at the end
of the text.
Literature is cited in the text as: Lynn and Reid (1968)
or (Lynn and Reid 1968). All papers referred to in the text
should be listed alphabetically by the senior author's sur-
name under the heading "Literature Cited." Only the
author's surname and initials are required in the literature
cited. The accuracy of the literature cited is the respon-
FORM OF THE MANUSCRIPT
The original of the manuscript should be typed, double-
spaced, on white bond paper. Please triple space above
headings. We would rather receive good duplicated copies
of manuscripts than carbon copies. The sequence of the
material should be:
TITLE PAGE
ABSTRACT
TEXT
LITERATURE CITED
TEXT FOOTNOTES
APPENDIX
TABLES (Each table should be numbered with an
arabic numeral and heading provided).
LIST OF FIGURES (Entire figure legends)
FIGURES (Each figure should be numbered with an
arabic numeral; legends are desired)
ADDITIONAL INFORMATION
Send the ribbon copy and two duplicated or carbon copies
of the manuscript to:
Dr. Andrew E. Dizon, Scientific Editor
Fishery Bulletin
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 271
La Jolla, CA 92038
Fifty separates will be supplied to an author free of
charge and 50 supplied to his organization. No covers will
be supplied.
(Contents— Continv£d)
BLAYLOCK, ROBERT A. Distribution and abundance of the bottlenose dolphin, Tur-
siops truncatus (Montagu, 1821), in Virginia 797
Notes
SEKI, MICHAEL R, and MICHAEL W. CALLAHAN. The feeding habits of two deep
slope snappers, PHstipomoides zonatus and P. auricilla, at Pathfinder Reef, Mariana
Archipelago 807
DITTY, JAMES G, GLEN G. ZIESKE, and RICHARD F. SHAW. Seasonality and depth
distribution of larval fishes in the northern Gulf of Mexico above latitude 26°00'N ... 811
ROGERS, CHRISTOPHER W., DONALD R. GUNDERSON, and DAVID A. ARMSTRONG.
Utilization of a Washington estuary by juvenile English sole, Parophrys vetulus. . . 823
SAFRIT, GLEN W, and FRANK J. SCHWARTZ. Length-weight relationships for gulf
flounder, Paralichthys albigutta, from North Carolina 832
WIEBE, PETER H. Functional regression equations for zooplankton displacement
volume, wet weight, dry weight, and carbon: a correction 833
GRAVES, JOHN E., MARIE A. SIMOVICH, and KURT M. SCHAEFER. Electro-
phoretic identification of early juvenile yellowfin tuna, Thunnus albacares 835
CAHOON, LAWRENCE B., and CRAIG R. TRONZO. A comparison of demersal zoo-
plankton collected at Alligator Reef, Florida, using emergence and reentry traps . . . 838
Index 847
Notice 860
1
• GPO 791-008
Mm. WHUI LIBRARY
UH nuG