ca
a.
o
(ShJ313l^) HldJO
9i
E gz
327
FISHERY BULLETIN: VOL. 76, NO. 2
Average Vertical Structure
Comparing Ring-D biomass partitioned accord-
ing to depth, the upper 200 m in the ring con-
tained, on the average, less biomass during both
sampling periods than did the Sargasso Sea (Table
2, top). This was true both day and night during
the August and November cruises. In contrast,
ring biomass between 200 and 800 m was higher
both day and night (Figures 1, 2). The range of
200-800 m biomass values in the ring and in the
Sargasso Sea does not even overlap. The combina-
tion of lower average surface biomass and higher
average subsurface biomass in the ring is highly
significant (Sign test, P<0.01, computed using
sums of 0-200 m and 200-800 m cm3/l,000 m^ de-
rived from Figures 1 and 2). The regional weighted
averages of percent 0-800 m biomass present in
the upper 200 m in August were Sl'/f , 347c , and
2T7c in the Sargasso Sea, slope water, and ring,
respectively. In November these averages were
457c, 327c, and 257^ (Table 4). Although very dif-
ferent sampling systems and tow strategies were
employed, data from Atlantis II cruise 71 corrobo-
rate the direction of difference of these observa-
tions in that the percentages of 0-800 m biomass
found at night in the upper 300 m were 64% and
52% for the Sargasso Sea and ring, respectively
(Table 5). In addition, the 300-800 m biomass was
1.73 times larger in this latter ring than in the
surrounding Sargasso Sea.
Diel Migration
Complicating these general observations and
contributing to sample variability are day/night
differences in biomass distributions (Table 4). In
Table 5.-
Area
-Ring and Sargasso Sea zooplankton biomass-
Atlantis II cruise 71 (mg/m^).
0-300 m
0-800 m
0-300
0-800
100
■Ring
Sargasso Sea
954
1,648
963
2,080
930
1,704
828
1,728
858
1,344
798
1,072
765
1,640
921
1,304
52%
64%
all day/night sample pairs the fraction of 0-800 m
biomass present in the 0-200 m interval is larger
in the night sample (Sign test, P <0.01). This re-
sults from either diel migration or day/night dif-
ferences in avoidance within the comparatively
well-illuminated surface layers. Avoidance does
not appear to be an important factor because at
some stations the day 0-800 m biomass exceeds the
night 0-800 m biomass. This is true in all Sargasso
Sea 0-800 m sample pairs and at one slope water
station (Figures 1, 2). Furthermore, some species
of zooplankton taxa already enumerated, e.g.,
euphausiids and pteropods, exhibit strong diel
migration patterns in all three areas.
Since we believe diel migration to be the ap-
propriate explanation, the data further suggest
that while essentially the same percentage of
0-800 m biomass was migrating into the surface
layers of the Sargasso Sea (24-30% during both
sampling periods), there was a reduced percentage
migrating in the ring in November ( 2 1% in August
versus 9% in November — Table 4). Although a
smaller proportion of the biomass may have been
migrating in the ring relative to the Sargasso Sea,
there was a significantly greater (Mann- Whitney
t/-test, P<0.05) day/night biomass ratio in the
Table 4. — Percent of 0-800 m slope water, ring, and Sargasso Sea zooplankton biomass in the upper 200 m
(800 m tows only). D = Day; N = Night.
August
: 1975
November 1975
Percentages of
Percentages of
Region
individual tows
D
N
N-D
(D + N)/2
individual tows
D
N
N-D
(D + N)/2
Sargasso Sea
Di = 32
N, = 57
D2 = 41
N2 = 69
39
63
24
51
z p
II II
30
60
30
45
Ring fringe
D, = 46
46
Cold core ring
D, = 16
N, = 32
N2 = 42
16
37
21
27
N, = 29
Di = 21
D2 = 19
20
29
9
25
Slope water
'D, = 3
'N, = 93
D2 = 7
N2 = 61
7
61
54
34
Di = 32
Ni = 55
D2 = 13
N2 = 27
23
41
18
32
Warm core ring
Di = 30
N, = 46
30
46
16
38
'On this tow series, MOC 18 and 19, salps were extremely dominant. These tows are excluded from averages.
328
ORTNER ET AL.: SARGASSO SEA ZOOPLANKTON BIOMASS DISTRIBUTION
upper 200 m in the ring (Table 6). This apparent
contradiction results from the fact, already noted,
that the percentage of 0-800 m biomass present in
the upper 0-200 m was very much greater in the
Sargasso Sea. Day/night ratio of biomass in the
upper water column is often used to measure in-
tensity of diel migration; clearly the meaning of
this ratio is highly dependent upon average verti-
cal biomass distribution.
Slope water day/night sample pairs may be in-
terpreted as documenting diel migration, but the
data are extremely variable both within and be-
tween cruises (Table 4). There may have been a
less well-developed migration in the fall, but the
generality of this is questionable.
Table 6. — Day night differences in slope water, ring, and Sar-
gasso Sea zooplankton biomass in the upper 200 m.
Ratio night day
Region
Augi
ist 1975
November
1975
Sargasso Sea
'1.78
'1.84
21.37
20.99
'1.37
21.86
Cold core ring
'2.01
23 10
'2 48
21 86
Slope water
'406
2.368 33
2.31482
'2 10
2227
21 70
'Based on 0-200 m tows
2Based on 0-800 m tows
3Ratio affected by extreme salp dominance.
Shallow Biomass Structure
In the 0-200 m biomass profiles, an intermediate
biomass peak occurred between 50 and 100 m
depth at nearly every station in August 1975 (Fig-
ure 1: MOC 1, 3, 6, 7, 10, 16, 17). At all butone of
the Sargasso Sea and ring stations this inter-
mediate peak is the highest observed value in the
0-200 m tows. At slope water stations of the same
cruise this intermediate peak is the second highest
observed value. If we rank each interval in a
profile in order of zooplankton abundance, we can
test the significance of this observation. For in-
stance, the individual summer tows in the ring
and the Sargasso Sea exhibit significant concor-
dance as to which depth intervals have the larger
and which the smaller zooplankton biomass
(Friedman 2-way analysis of variance on ranks,
P<0.005). Given this result, the best estimate of
the differences between intervals is the order of
their summed ranks (i.e., 50-75 m>75-100
m>100-125 m>25-50 m>0-25 m>150-175
m>125-150 m>175-200 m). Applying a procedure
for testing differences between individual depth
intervals (Nemenyi 1963), we see that concor-
dance results from the fact that the 50-75 m
biomass is significantly greater than the biomass
in the intervals 125-150, 150-175, and 175-200 m,
and the 75-100 m biomass is greater than the
175-200 m biomass (P<0.05). An intermediate
peak is not a notable feature of any of the 0-200 m
profiles taken on the fall cruise with the exception
of the Sargasso Sea sample pair (Figure 2: MOC
23, 26).
DISCUSSION
Wiebe, Hulburt, Carpenter, Jahn, Knapp, Boyd,
Ortner, and Cox ( 1976) have discussed the forma-
tion and decay of an idealized cold core ring. Ini-
tially conditions inside a ring core are identical to
those in the slope water just northward of the Gulf
Stream at the time of ring formation. Through
time the ring decays; the isotherms deepen, the
water becomes more saline, the O2 minimum
deepens, and the constituent flora and fauna
either die off or become diluted by populations
from the surrounding Sargasso Sea. Because zoo-
plankton populations are generally suited to the
environmental conditions they encounter within
their normal range, this decay process may be
viewed as the gradual imposition of a complex
environmental stress upon an entire community.
Wiebe, Hulbert, Carpenter, Jahn, Knapp, Boyd,
Ortner, and Cox ( 1976) have documented some of
the intermediate stages in this idealized process.
In fact, this process can be aborted when a ring is
reabsorbed by the Gulf Stream (Fuglister 1972;
Richardson et al. 1977). All biological and physical
properties are not equally conservative so their
decay rates would not be the same.
Regional Contribution of Cold Core Rings
PRIMARY PRODUCTIVITY.— It is well
known that slope water is more productive than
the Sargasso Sea. Ryther (1963) estimated that
slope water is about twice as productive on an
annual basis ( 120 g C/m^ per yr versus 60 g C/m^
per yr). Although our own data are scanty, rings
on the average are intermediate between slope
water and the Sargasso Sea (Table 7). A few
simplifying assumptions permit budgetary com-
putations to be made regarding the overall effect
of rings on the carbon budget of the northern Sar-
gasso Sea. Let us assume an average ring life of 1
yr and a linear rate of decay of productivity (i.e.,
that annual ring production is the arithmetic
mean of annual Sargasso Sea and slope water pro-
duction). Allowing 6 to 13% as the areal contribu-
329
FISHERY BULLETIN: VOL. 76, NO 2
Table 7. — Summary of slope water, ring, and Sargasso Sea primary productivity (mg C/m^ per day), phytoplankton carbon' (mg/m^),
and chlorophyll a (mg/m^) measurements.
March 1974
August
1975
Phytoplankton
carbon
November 1975
Phytoplankton
Region
Productivity
Chlorophyll
Productivity
Chlorophyll
Productivity
Chlorophyll
carbon
Sargasso Sea
2285
46.4
^207
100
133
334
86.5
2522
12,0
21
Cold core ring
440 1
333.1
73.0
83
106
17.3
45
4832
155-5
1865
103
282
Slope water
1.025.5
70.4
175
50.5
1,302.2
368 4
270
280
824.0
376,2
363.7
39
287
'Based on counts of cells larger than 4-5 txm
^The high value in average mg C m^ per day observed at this station is a consequence of one unusually high surface value.
tion of rings to the northern Sargasso Sea as
explained earlier, and Ryther's estimate of a
twofold difference in annual production, the net
annual production of the geographic northern
Sargasso Sea is then 3 to 7% higher than if it
contained no rings (i.e., 6 x 1.5 = 9, 9 + 94 = 103
and 13 X 1.5 = 20, 20 + 87 = 107). Our assumption
of linear decay is most certainly an oversimplifica-
tion. In November 1975, the ring water column,
like the slope water, began its winter overturn
before the surrounding Sargasso Sea. Mixing
eroded the seasonal thermocline that had been
observed in Ring-D in August 1975. The decay we
have assumed was reversed, and ring productivity
was enhanced (Table 7).
ZOOPLANKTON STANDING CROP.— Simi-
lar calculations can be made regarding the rela-
tive contribution of rings to the mean zooplankton
biomass of the geographic northern Sargasso Sea.
Neglecting one station which had anomalously
high values due to extreme salp dominance, the
average of slope water biomass values is 3.5 times
the observed Sargasso Sea biomass (Table 3).
Given this ratio and the same linear [i.e., (3.5 + 1)
-i- 2 = 2.25] and areal assumptions made earlier,
rings may augment the zooplankton standing crop
of the geographic northern Sargasso Sea by 8 to
16% (i.e., 6 X 2.25 = 14, 14 + 94 = 108 and 13 x
2.25 = 29, 29 + 87-1 16). Our ratio of slope water
to Sargasso Sea biomass may be compared with
that of Grice and Hart (1962), who reported the
slope water standing crop as three to four times
that of the Sargasso Sea. They also excluded ex-
tremely salp-rich samples in making this com-
parison. Our assumption of 2.25 as an annual
mean ring/Sargasso biomass ratio (i.e., linear de-
cay) may be an overestimate considering the aver-
age biomass ratio obtained on all cruises to date
and the average ring age sampled (Table 3). On
the other hand some rings do last longer than a
year and the lowest ring:Sargasso Sea ratios that
we have observed are approximately 1.3 (i.e.,
>1.0).
We have noted a highly significant decline in
0-800 m biomass from August to November in
slope water, ring, and Sargasso Sea both in
data presented here and in data more recently
collected.'^ This observation is consistent with
those of Grice and Hart ( 1962) with respect to the
slope water. They noted, however, no such decline
in the Sargasso Sea. Neither is there a summer-
to-fall decline in the Sargasso Sea data of Deevey
(1971). The Sargasso Sea and slope water data of
Fish (1954) exhibit irregular fluctuations in
biomass throughout the summer and fall. Moore
( 1949) presented some Sargasso Sea data indicat-
ing a progressive decline of biomass from a spring
maximum to a fall minimum. Their data substan-
tiate that interseasonal fluctuations in the Sar-
gasso Sea are less marked than in the slope water.
Vertical Structure
We have pointed out that, compared with either
the Sargasso Sea or slope water, an unusually
small percentage of 0-800 m biomass is present in
the upper 200 m of a ring. We found a relatively
large fraction of the 0-800 m zooplankton biomass
above 200 m in the northern Sargasso Sea. The
netting employed by Leavitt ( 1935, 1938) was rel-
atively coarse (1.0 mm) so it is difficult to compare
our results with his. Nonetheless, at his two Sar-
gasso Sea stations (2462, 2463) the percentages of
^Some of this data is presented in figure 4 of Richardson, P, L,,
J, Schmitz, and P. H, Wiebe, 1977, Gulf Stream ring experi-
ment, Polymode News 25:3. Unpubl. manuscr.
330
ORTNER ET AL.: SARGASSO SEA ZOOPLANKTON BIOMASS DISTRIBUTION
0-800 m biomass present in the upper 200 m were
42 and 49Vf which corresponds closely to our val-
ues (Table 3). Both of our results are virtually
identical with those obtained by Menzel and
Ryther ( 1961). From their table 1 we can calculate
the percentages of 0-500 m biomass and 0-1,000 m
biomass present above 200 m. Averaging the re-
sults, we find 449c of the 750-m biomass was pres-
ent during the day above 200 m. lashnov (1961)
presented data for the Sargasso Sea in which 90%
of the 0-1,000 m plankton was present above 200
m, but he used a relatively fine mesh net (0.180
mm). Unfortunately, Deevey and Brooks (1971)
characterized 500-m depth intervals with horizon-
tal tows at the midpoint of each interval to 2,000
m, while Grice and Hart (1962) sampled only the
upper 100-200 m.
Several authors suggest that a vertical biomass
structure similar to our slope water and Sargasso
Sea observations is to be expected in temperate or
subtropical oceanic environments relatively free
of advective inputs. Vinogradov (1968: figure 47
and stations 3206 and 3829 in table 18) gave
examples of oceanic regions with such a distribu-
tion. Zenkevich and Birstein (1956) agreed that
zooplankton biomass in the North Pacific rather
steadily decreases from the surface downwards,
although the most marked reduction they discuss
might be below our lowest standard sampling
depth. The one very deep tow series we obtained in
a ring, however, gave no indication of such a re-
duction (Figure 2, MOC 31).
Zooplankton biomass profiles obtained by
Murano et al. (1976) in the northwest Pacific
above the Sagami Trough exhibit the expected
decrease with depth. Reanalyzed in our manner,
the data of Marlowe and Miller ( 1975) for Station
P in the North Pacific support the above generali-
zation; the percentage of their 0-500 m biomass
found at night in the upper 200 m was 579^ . If one
extrapolates their 500-m values as approximately
applicable to the 500-800 m interval — a conserva-
tive approach for this argument — the resulting
percentage becomes 49*7^ (N). This is not unlike
our average slope water percentage of 5 1% ( N ) and
quite distinct from the average ring percentage of
33% (N) (Table 4). Station P is very different from
Ring-D in respect to its vertical biomass distribu-
tion.
In slope water, the intermediate biomass peak
in the upper 200 m approximately coincides with
the depth of a nitrite maximum of the type discuss-
ed by Vaccaro and Ryther ( 1960). Our results and
those of Marlowe and Miller (1975) appear to dif-
fer: they felt that the shallow nitrite peak of Sta-
tion P was avoided by zooplankton. Since the
levels of nitrite we have observed at the maximum
are only slightly lower than those reported by
Marlowe and Miller (0.2-0.5 fxg A-N-NOg/l versus
0.64 /Ltg A-N-NO2/I), our findings cast doubt on
their speculation that nitrite toxicity might have
been involved in the maintenance of the biomass
minima they observed.
Explanations for Ring Biomass Structure
Given the relatively high zooplankton biomass
of the slope water, it is clear why cold core rings
have a higher average zooplankton biomass than
the Sargasso Sea. Further, their higher average
primary productivity appears responsible for this
differential persisting 10-12 mo after ring forma-
tion. Our data suggest the decline in ring biomass
takes place rather slowly; the oldest rings sampled
(10-12 mo) had ring/Sargasso biomass ratios only
20% smaller than the same ratios in the newest
rings sampled (3.0 and 3.5 mo, Table 3). Although
physically and chemically intermediate between
slope water and Sargasso Sea, rings appear to be
unique in their vertical distribution of biomass.
We offer two logically distinct explanations for
the small fraction of the 0-800 m biomass found
within the upper 200 m of a ring. They are not
mutually exclusive and the relative importance of
these explanations is species dependent. The sim-
pler argument stresses the importance of a physi-
cal factor — temperature. If a slope water animal
were physiologically restricted to a particular
temperature range, its habitat would descend as
the ring decayed and isotherms sank. To the ex-
tent that the zooplankton population in the slope
water exhibited this behavior, ring biomass dis-
tributions would deepen. This could apply only to a
species which in its home range — the slope wa-
ter— remains beneath the seasonal thermocline
(i.e., moderately deep-living and exhibiting li-
mited diel migration). Such a species would most
likely have to be either carnivorous or omnivor-
ous. Wiebe and Boyd ( 1978) have documented such
a phenomenon for the slope water euphausid
species, Nematoscelis megalops.
A more complex explanation stresses the impor-
tance of a biological factor — food resources. The
kinds of changes that accompany ring decay must
have a substantial effect upon zooplankton-
phytoplankton interactions. Using unpublished
331
FISHERY BULLETIN: VOL. 76, NO 2
data obtained in August 1975 from 5 nine bottle
hydrocasts, the number of phytoplankton cells per
liter averaged 10,000 in the slope water, 2,500 in
the ring, and 2,000 in the Sargasso Sea. Cells
smaller than 4-5 /xm were not enumerated and
were, therefore, excluded from these computa-
tions. Values were integrated from 0 to 200 m — a
conservative procedure tending to reduce slope
water versus ring or Sargasso Sea differences. The
species composition of the ring, while distinct, was
more like that of the Sargasso Sea than that of the
slope water. Again, considering the 0-200 m depth
interval, the number of different phytoplankton
species an animal would have encountered in a
liter of water would, on the average, have been 6.0
(slope water), 9.6 (ring), and 10.4 (Sargasso Sea).
Converting the mean cell volume of each species to
carbon (Strathmann 1967) and multiplying by the
number of individuals present, yielded values of
average phytoplankton carbon of 1,400, 200, and
140 ng C/1. Thus, to acquire the same ration of
food, a herbivore would have had to filter more
than five times more water in the ring than in the
slope water, and even more in the Sargasso Sea. In
addition, the evenness of species' carbon equiva-
lence was 0.46, 0.75, and 0.76. That is, the total
carbon per liter was more evenly distributed
among different species in the ring and the Sar-
gasso Sea than in the slope water. (Evenness
equals HIH„^^^ (Pielou 1966) where H is the
Shannon-Weaver diversity index computed upon
species carbon equivalence rather than abun-
dance and //max ^ ^ogj, S where S = number of
species.) This last result implies that a herbivore
capable of selecting by carbon content (i.e., parti-
cle size) would have found it less advantageous to
concentrate on a particular species in the Sargasso
Sea and the ring than in the slope water.
These properties of the phytoplankton popula-
tion, i.e., species composition, carbon concentra-
tion, cell concentration, and cell carbon distri-
bution, have profound effects on a filter-feeding
herbivore's harvesting ability. We believe that
early in ring evolution herbivorous slope water
species are deleteriously affected and, therefore,
may be replaced by Sargasso Sea forms more
quickly than deeper living carnivorous or om-
nivorous slope water species. If we are correct, ring
biomass distribution may deepen in part because a
ring's 0-200 m biomass declines more rapidly than
does its 200-800 m biomass.
Identification of some of the taxa in August 1975
samples, although limited, support the argument
that in Ring-D epizooplanktonic herbivores were
replaced before epizooplanktonic carnivores or
omnivores. The species list of Ring-D thecosoma-
tous pteropods, a largely herbivorous group, was
quite similar to that of the surrounding Sargasso
Sea.^ Grice and Hart (1962) found that chaetog-
naths, a purely carnivorous group, were consider-
ably more abundant in the Sargasso Sea than they
are in slope water. In 6 nine-net fine-mesh tow
series (12.5 cm diameter, Clarke-Bumpus nets
with 67 ^tm mesh) taken in August, chaetognaths
were five to ten times more abundant in the sur-
rounding Sargasso Sea than they were in Ring-D.
Other epizooplanktonic carnivores, e.g., Stylo-
cheiron suhmii and S. abbreviatum, which are
routinely found in the Sargasso Sea were not
found in Ring-D August MOCNESS tows.
Organic Flux to Deep Sea
Rings may contribute a disproportionate frac-
tion of the utilizable organic material available to
the northern Sargasso deep sea. We feel this is
likely both because of their generally higher pro-
ductivity and because of their unique zooplankton
biomass distribution and the factors that have re-
sulted in that distribution. Ring zooplankton
biomass below 200 m, in that it exceeds Sargasso
Sea biomass and ultimately declines to a similar
level, contributes to this augmentation. Differen-
tial seasonal mixing processes could also increase
downward particulate flux. For example, in
November 1975 we observed that winter mixing
had proceeded further in Ring-D than in the sur-
rounding Sargasso Sea water column. Herbivor-
ous ring zooplankton (i.e., Sargasso forms) may
have been unable to fully capitalize upon the sud-
den opportunity afforded by the increased primary
production that accompanied the mixing (Table 7).
If so, a larger fraction of this enhanced phyto-
plankton production would sink into the aphotic
depths. Physical evidence obtained on two cruises
undertaken to study rings during the summer has
suggested to us that the seasonal thermocline may
often be less stable in rings than in the Sargasso
Sea.
Finally, there is a possibility of enhanced con-
tribution of organic matter into the deep sea due to
a lower overall trophic efficiency within the upper
200 m of rings (and slope water). If we divide
^John Wormuth, unpubl. data; cited with permission.
332
ORTNER ET AL.: SARGASSO SEA ZOOPLANKTON BIOMASS DISTRIBUTION
average 0-200 m zooplankton biomass (milligrams
of carbon per square meter calculated using equa-
tion 4, table 2 in Wiebe et al. 1975) by 0-200 m
phytoplankton carbon (milligrams of carbon per
square meter from Table 5), excluding salp-rich
MOC 18 and 19, we obtain the following ratios:
Aug. 1975
Nov. 1975
Sargasso Slope
Sea Ring water
253 138 84
332 131 28
Ratios in the ring are low, as are those in the slope
water. Lower ratios suggest to us lower overall
trophic efficiency within the upper 200 m. Al-
though biased in that many cells are quite small,
particularly in the Sargasso Sea, phytoplankton
carbon of cells >5 ixm is probably a reasonable
estimate of the food available at the time of sam-
pling to many of the herbivorous animals caught
by our 0.333-mm mesh nets. The direction of dif-
ference noted above conforms with ideas expressed
by Menzel and Ryther ( 1961 ), Heinrich ( 1962), and
others who argued that especially close
phytoplankton-zooplankton coupling may charac-
terize oceanic tropical-subtropical waters.
The biomass data presented here illustrate the
fact that geographic demarcation of oceanic faunal
provinces is not sufficient. Hydrographic as well as
faunal mapping is essential in explaining that
portion of station-to-staion variability associated
with mesoscale hydrographic variability resulting
from phenomena like Gulf Stream cold core rings.
ACKNOWLEDGMENTS
We express our grateful appreciation to Alfred
Morton for his assistance at sea; to Michael Stro-
man who measured many of the displacement vol-
umes; and to our typist Jane Peterson, for her
remarkable tolerance and good humor. James Cox
and Margo Haygood critically read the manu-
script. This study was supported by ONR
NOOO14-66-C-0240, NOOO14-24-C-0262 NR
083-004, NSF DES 74-02783A1, the Woods Hole
Oceanographic Institution Graduate Education
Program, and the Tai Ping Foundation.
LITERATURE CITED
AHLSTROM, E. H., AND J. R. THR.'MLKILL.
1963. Plankton volume loss with time of preserva-
tion. Calif. Coop. Oceanic Fish. Invest. Rep. 9:57-73.
Be, a. W. H., J. M. FORNS, AND O. A. ROELS.
1971. Plankton abundance in the North Atlantic Ocean.
In J. D. Costlow, Jr. (editor), Fertility of the sea, Vol. 1, p.
17-50. Gordon and Breach Sci. Publ., N.Y.
Deevey, G. B.
1971. The annual cycle in quantity and composition of the
zooplankton of the Sargasso Sea off Bermuda. I. The upper
500 m. Limnol. Oceanogr. 16:219-240.
Deevey, g. b., and a. L. Brooks.
1971. The annual cycle in quantity and composition of the
zooplankton of the Sargasso Sea off Bermuda. II. The
surface to 2000 m. Limnol. Oceanogr. 16:927-943.
Fish, C. J.
1954. Preliminary observations on the biology of boreo-
arctic and subtropical oceanic zooplankton populations. In
Symposium on marine and fresh-water plankton in the
Indo-Pacific, Bangkok, p. 3-9. Diocesan Press, Madras.
FUGLISTER, F. C.
1972. Cyclonic rings formed by the Gulf Stream 1965-
66. In A. L. Gordon (editor). Studies in physical oceanog-
raphy. A tribute to George Wiist on his 80th birthday, Vol.
1, p. 137-168. Gordon and Breach, N.Y.
Grice, G. D., and a. D. Hart.
1962. The abundance, seasonal occurrence, and distribu-
tion of the epizooplankton between New York and Ber-
muda. Ecol. Monogr. 32:287-309.
Heinrich, a. k.
1962. The life histories of plankton animals and seasonal
cycles of plankton communities in the oceans. J. Cons.
27:15-24.
lASHNOV, V. A.
1961. Vertikal'noe respredelenie massy zooplanktona
tropicheskoi oblasti Atlanticheskogo Okeana (Vertical
distribution of the mass of zooplankton throughout the
tropical zone of the Atlantic). Dokl. Akad. Nauk. SSSR
136:705-708.
JAHN, a. E.
1976. On the midwater fish faunas of Gulf Stream rings,
with respect to habitat differences between Slope water
and Northern Sargasso Sea. Ph.D. Thesis, Woods Hole
Oceanogr. Inst., Woods Hole, 172 p.
Lai, D. Y., and p. L. Richardson.
1977. Distribution and movement of Gulf Stream rings.
J. Phys. Oceanogr. 7:670-683.
LEAVITT, B. B.
1935. A quantitative study of the vertical distribution of
the larger zooplankton in deep water. Biol. Bull. (Woods
Hole) 68:115-130.
1938. The quantitative vertical distribution of macrozoo-
plankton in the Atlantic Ocean basin. Biol. Bull. (Woods
Hole) 74:376-394.
Marlowe, C. J., and C. B. Miller.
1975. Patterns of vertical distribution and migration of
zooplankton at Ocean Station "P". Limnol. Oceanogr.
20:824-844.
MCGOWAN, J. A., AND D. M. BROWN.
1966. A new opening-closing paired zooplankton net.
Scripps Inst. Oceanogr. Ref 66-23:1-56.
MENZEL, D. W., AND J. H. RYTHER.
1961 . Zooplankton in the Sargasso Sea off Bermuda and its
relation to organic production. J. Cons. 26:250-258.
MOORE, H. B.
1949. The zooplankton of the upper waters of the Bermuda
area of the North Atlantic. Bull. Bingham Oceanogr.
Collect., Yale Univ. 12(2), 97 p.
333
FISHERY BULLETIN: VOL. 76, NO. 2
MURANO M., R. MARUMO, T. NEMOTO, AND Y. AlZAWA.
1976. Vertical distribution of biomass of plankton and
micronekton in the Kuroshio water off Central Ja-
pan. Bull. Plankton. Soc. Jap. 23:1-12.
Nemenyi, p. B.
1963. Distribution-free multiple comparisons. Ph.D.
Thesis, Princeton Univ., Princeton, 127 p.
Parker, C. E.
1971. Gulf Stream rings in the Sargasso Sea. Deep-Sea
Res. 18:981-993.
PlELOU, E. C.
1966. The measurement of diversity in different types of
biological collections. J. Theor. Biol. 13:131-144.
Richardson, P.
1976. Gulf Stream rings. Oceanus 19(3):65-68.
RICHARDSON, P. L., R. E. CHENEY, AND L. A. MANTINI.
1977. Tracking a Gulf Stream ring with a free drifting
surface buoy. J. Phys. Oceanogr. 7:580-590.
RYTHER, J. H.
1963. Geographic variations in productivity. In M. N.
Hill (editor). The sea. Ideas and observations on progress
in the study of the seas, Vol. 2, p. 347-380. Interscience
Publ., Lond. >
Saunders, P. M.
1971. Anticyclonic eddies formed from shoreward mean-
ders of the Gulf Stream. Deep-Sea Res. 18:1207-1219.
STRATHMANN, R. R.
1967. Estimating the organic carbon content of phyto-
plankton from cell volume or plasma volume. Limnol.
Oceanogr. 12:411-418.
Vaccaro, R. F., and J. H. RYTHER.
1960. Marine phytoplankton and the distribution of ni-
trite in the sea. J. Cons. 25:260-271.
Vinogradov, M. E.
1968. Vertical distribution of the oceanic zooplankton. [In
Russ.] Isdatel'stvo Nauka, Moskva, 319 p. (Isr. Prog. Sci.
Transl., 1970, 339 p.)
Wiebe, p. H., and S. H. Boyd.
1978. Limits of Nematoscelis megalops in the Northwest-
em Atlantic in relation to Gulf Stream cold core rings.
Part I. Horizontal and vertical distributions. J. Mar.
Res. 36:119-142.
Wiebe, P. H., S. Boyd, and J. L. Cox.
1975. Relationships between zooplankton displacement
volume, wet weight, dry weight, and carbon. Fish. Bull.,
U.S. 73:777-786.
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., E. M. Hulburt, E. J. Carpenter, A. E. Jahn, G.
P. Knapp, III, S. H. Boyd, P. B. Ortner, and J. L. Cox.
1976. Gulf Stream cold core rings: large-scale interaction
sites for open ocean plankton communities. Deep-Sea
Res. 23:695-710.
Zenkevich, L. a., and J. A. Birstein.
1956. Studies of the deep water fauna and related prob-
lems. Deep-Sea Res. 4:54-64.
334
RELATIVE CONTRIBUTION OF HUDSON, CHESAPEAKE, AND
ROANOKE STRIPED BASS, MORONE SAXATILIS,
STOCKS TO THE ATLANTIC COAST FISHERY
Thomas J. Berggren' and Joel T. Lieberman^
ABSTRACT
Morphological characters were used in discriminant analysis to quantitatively estimate the relative
contribution of striped bass, Morone saxatilis, stocks from various estuaries to the striped bass fishery
along the Atlantic coast. Representative samples of the spawning stocks of the Hudson River,
Chesapeake Bay system, and Roanoke River were collected and counts and measurements were taken
on each specimen. Discriminant functions based on five morphological characters correctly classified
approximately TS^c of the specimens. The effectiveness of three types of estimates based on these
functions in accurately estimating stock proportions was investigated in a simulation study. Results of
the simulation study indicated which type of estimate was least biased. A sampling design using
geographical and temporal strata was then employed to sample the Atlantic coastal fishery from Cape
Hatteras, N.C., to Maine. Observations for the morphological characters were taken on collected fish
and the resulting data entered into discriminant functions obtained from spawning-stock collections.
The specimens were classified by area of origin and the three types of estimates of relative contribution
of the Hudson, Chesapeake, and Roanoke stocks were obtained. Results indicated that the Chesapeake
stock was the major contributor to the Atlantic coastal striped bass fishery and the Hudson and
Roanoke stocks were minor contributors.
The striped bass, Morone saxatilis, is an important
sport and commercial fish in the estuaries and
coastal waters of the Atlantic seaboard from
Maine to North Carolina (Koo 1970). Recruitment
to the striped bass fishery is from various stocks of
striped bass spawned and developed in rivers and
estuaries along the Atlantic coast. Recapture loca-
tions of tagged striped bass indicate that individu-
als from all spawning areas north of Cape Hat-
teras, N.C., utilize much of the Atlantic coast
north of their respective spawning areas during a
northward migration in the spring and a south-
ward migration in the fall (Merriman 1941; Raney
et al. 1954; Alperin 1966; Schaefer 1968; Flor-
ence^; Texas Instruments^). The major spawning
areas which potentially contribute individuals to
the fisheries operating during the northward and
southward migrations are the tributaries of
'Texas Instruments Inc., Buchanan, N.Y.; present address:
Biometrics Unit, 337 Warren Hall, Cornell University, Ithaca,
NY 14853.
^Texas Instruments Inc., P.O. Box 237, Buchanan, NY 1051 1.
^Florence, B. 1974. Tag returns from 1375 large striped bass
tagged in two Maryland spawning rivers. Outdoor Message.
Organized Sportsmen of Mass. Oct. 1974.
■•Texas Instruments Inc. 1976. Report on relative contribution
of Hudson River striped bass to the Atlantic coastal fishery.
Prepared for Consolidated Edison Company of New York, Inc.,
101 p.
Chesapeake Bay and the Roanoke and Hudson
Rivers.
Although tagging data have not led to quantita-
tive estimates of relative contribution, they have
led to conflicting ideas as to which major stock of
striped bass predominates in the fishery: the Hud-
son stock or the Chesapeake stock. Most published
works have generally concluded that the striped
bass stock from the Chesapeake Bay system is the
major contributor to the fisheries north of
Chesapeake Bay (Merriman 1941; Vladykov and
Wallace 1952; Alperin 1966; Schaefer 1968; Porter
and Saila^; Raney^). However, Clark'' and
Goodyear^ concluded that the striped bass stock
Manuscript accepted August 1977.
FISHERY BULLETIN: VOL. 76, NO. 2. 1978.
^Porter, J., and S. B. Saila. 1969. Final report for the coopera-
tive striped bass migration study. U.S. Fish. Wildl. Serv. Con-
tract no. 14-16-005, 33 p.
"Raney, E. C. 1972. The striped bass, Morone saxatilis, of the
Atlantic coast of the United States with particular reference to
the population found in the Hudson River. Testimony before
USAEC Safety and Licensing Board for Indian Point, Unit no. 2.
Docket no. 50-247, Oct. 30, 105 p.
'Clark, J. 1972. Effects of Indian Point Units 1 and 2 on the
Hudson River aquatic life. Testimony before USAEC Safety and
Licensing Board for Indian Point, Unit no. 2. Docket no. 50-247,
Oct. 30, 63 p.
^Goodyear, C. P. 1974. Origin of the striped bass of the middle
Atlantic coast. Testimony presented to the Committee on Mer-
chant Marine and Fisheries of the U.S. House of Representa-
tives. Feb. 19, 40 p.
335
FISHERY BULLETIN: VOL. 76, NO. 2
from the Hudson River is the major contributor to
the coastal fishery from New Jersey to Mas-
sachusetts because the number of striped bass
tagged in Chesapeake Bay and recaptured outside
the Bay was too low to indicate a large contribu-
tion of Chesapeake stock to that fishery.
Because of the controversy of which stock pre-
dominates, we conducted a study to obtain quan-
titative estimates of relative percentage of the
major stocks in the coastal fishery. A previous
study (Grove et al. 1976) demonstrated the feasi-
bility of using discriminant analysis on mor-
phological characters (counts and morphometric
ratios) to distinguish among Hudson, Chesapeake,
and Roanoke spawning stocks of striped bass. That
study showed that adequate segregation of spawn-
ing stocks within the Chesapeake Bay system was
not possible. Quantitative estimates of stock com-
position based on morphological characters and
discriminant analysis have been obtained for
sockeye salmon (Fukuhara et al. 1962; Anas and
Murai 1969), pink salmon (Amos et al. 1963), and
Atlantic herring (Messieh 1975). The present
study establishes discriminant functions based on
collections of spawning-stock specimens to classify
striped bass collected in the Atlantic coastal
fishery from southern Maine to Cape Hatteras.
The percentage of specimens collected that were
classified into each stock was used to estimate
the relative contribution of that stock to the
fishery.
METHODS AND MATERIALS
Collection of Spawning-Stock Specimens
During the spawning season of 1975, mature
striped bass were collected from the natal rivers of
major stocks along the Atlantic coast. These fish
were assumed to have originated from the rivers
(i.e., that striped bass, like salmon and other
anadromous fishes, home to their natal stream to
spawn). This assumption was supported by tag-
ging studies in which striped bass tagged on
spawTiing grounds were recaptured on the same
spawning grounds in successive years (Mansueti
1961; Nichols and Miller 1967). Collections were
composed of 232 mature striped bass from the
Chesapeake Bay tributaries (70 from the Rap-
pahannock River, 53 from the Potomac River, 52
from the Choptank River, and 57 from the Elk
River and Chesapeake and Delaware Canal), 168
from the Hudson River, and 99 from the Roanoke
River. Only 19 sexually ripe striped bass were
collected from the Delaware River above the en-
trance to the Chesapeake and Delaware Canal,
which confirms findings by Chittenden ( 1971) that
spawning in the Delaware River is not substan-
tial. Therefore specimens from the Delaware
River were omitted from subsequent analyses.
Collections were made primarily during April in
the Chesapeake Bay tributaries, Delaware and
Roanoke Rivers, and during May in the Hudson
River. Most specimens were obtained fresh from
commercial fishermen using pound nets, haul
seines, and gill nets. Some were netted by study
personnel.
To assure an adequate representation of the
sexes and multiple year classes in spawning-stock
collections, sampling was designed to obtain
nearly equal numbers of male and female striped
bass and a minimum of 10 individuals in each of
the following length categories: ^399, 400-549,
550-699, 700-849, and &850 mm. Discriminant
functions based on male and female specimens
from multiple year classes are needed to analyze
an oceanic population which consists of a different
sex ratio and broader age structure than that of
the spawning stocks.
Processing of Spawning-Stock Specimens
Scale samples, counts, measurements, sex, and
state of maturity were obtained from each speci-
men while in fresh condition. Scale samples from
above the lateral line between the first and second
dorsal fins were pressed on acetate cards. Ages
were determined by the scale annulus method
(Mansueti 1961). Measurements from the focus to
the first and second annuli were made on mag-
nified scale images. The following counts and
measurements were taken: number of lateral line
scales, left pectoral rays, right pectoral rays, sec-
ond dorsal rays, anal rays, upper-arm gill rakers,
fork length, snout length, head length, and inter-
nostril width. Methods used were those discussed
by Hubbs and Lagler (1958) and Grove et al.
(1976).
Counts, measurements, and age determinations
were replicated by a second observer and a set of
tolerances was established to reduce observation
error. When differences between replicated obser-
vations exceeded tolerances, the observations
were retaken. Means of the replicated counts and
means of ratios of the replicated measurements
were used in subsequent analysis.
336
BERGGREN and LIEBERMAN: RELATIVE CONTRIBUTION OF STRIPED BASS
Analysis of Spawning-Stock Specimens
Choice of morphological characters for segrega-
tion of Hudson, Chesapeake, and Roanoke spawn-
ing stocks followed three stages of statistical
analysis: correlation analysis between each
character and fork length (FL), analysis of the
effects of sex and age on each character, and dis-
criminant analysis. Analysis involved only speci-
mens with observations on all counts and mor-
phometric and scale-annulus measurements.
Since spawning stocks do not include immature
specimens which occur in the coastal waters, we
chose only those characters that were independent
(i.e., not highly correlated) offish size and could
therefore be used to segregate specimens from the
entire stock. Characters were considered to be in-
dependent of length when variations (r^) attribut-
able to length in any stock were ssO.lO. Characters
not independent of length were used in further
analysis when the distribution of character values
had small overlap among spawning stocks since
such characters help identify stock origin.
Multivariate statistical tests were made to de-
termine the effect of sex and age on the characters
used to determine the discriminant functions,
since one assumption of discriminant analysis was
that each stock was homogeneous. Differences in
character values among ages for males or females
and between sexes within each stock were tested
with a procedure that combined tests of equality of
means and equality of covariance matrices (An-
derson 1958). Assuming equal covariance ma-
trices, rejection of the null hypotheses of equal
distributions indicated that one or more of the
character means differed among ages or between
sexes.
Multivariate discriminant analysis was used to
gain maximum separation among stocks. Linear
and quadratic discriminant functions (Anderson
1958; Kendall and Stuart 1968) for each spawning
stock were determined from character values ob-
tained from collections of that stock. A stepwise
procedure on the linear function was used to indi-
cate the subset of characters which best separated
the stocks. The quadratic function based on this
subset was formed if the assumption of a common
covariance matrix among spawning stocks needed
for the linear function was not met. The assump-
tion in discriminant analysis that characters had
a multivariate normal distribution was investi-
gated with histograms.
Ability of the discriminant functions to separate
stocks and accurately estimate stock proportions
was assessed using functions based on total
spawning-stock collections and functions obtained
from a cross-validation procedure (Mosteller and
Tukey 1968). In this procedure collections were
randomly divided in half and discriminant func-
tions were determined from one-half and applied
to each half. Percentages of correct classification
and estimates of stock proportion were obtained
for each subset and compared with those from the
total sample. Comparisons were also made be-
tween estimated and known spawning-stock per-
centages.
Although these estimates of stock percentages
may accurately approximate true percentages in
spawning-stock collections, they may deviate sub-
stantially from stock percentages in oceanic col-
lections. Fukuhara et al. (1962) stated that the
bias in these estimates increased as stock percen-
tages became more disproportionate. Since stock
percentages in oceanic collections may be more
disproportionate than stock percentages in
spawning-stock collections (i.e., 347^ Hudson, 469c
Chesapeake, and 20*^ Roanoke stocks), less biased
estimates of stock percentages may be needed.
Adjusting Estimates of Stock Percentages
Two procedures were developed to obtain esti-
mates of stock percentages that were less biased
than the as-classified (i.e., classifications obtained
directly from discriminant functions) estimates.
The first procedure adjusted estimates using a
technique described by Worlund and Fredin
(1962) which generalized to the three population
case methodology developed in Fukuhara et al.
(1962). This procedure used percentages of speci-
mens from each spawning stock that were mis-
classified into other stocks to correct as-classified
estimates for bias due to misclassifications. When
adjusted estimates were negative, as-classified es-
timates were modified by methodology developed
by Schuermann and Curry. ^
The second procedure iteratively reclassified
specimens based on updated prior probabilities
that specimens originated from each of the spawn-
ing stocks. The first stage of the procedure is the
same as the as-classfied procedure; therefore as-
^Schuermann, A. C, and G. L. Curry. 1973. Notes on paramet-
ric programming. Unpubl. manuscr. Dep. Ind. Eng. Texas A&M
Univ., College Station.
337
FISHERY BULLETIN: VOL. 76, NO. 2
classified estimates of stock contribution are ob-
tained at the end of this stage. However, these
estimates are then used in the second stage as
prior probabilities that specimens come from the
three stocks. For example, the as-classified esti-
mate of Hudson stock contribution obtained at the
end of the first stage was used at the beginning of
the second stage as our best guess of the proportion
of specimens in the sample that originate from the
Hudson. These prior probabilities are then used to
weight the decision to classify each specimen into
one of the stocks. Similarly, the proportion of
specimens classified into each stock in the second
stage were used as priors in the third stage. The
procedure was carried out for nine stages.
The effectiveness of adjusted and iterative esti-
mates in reducing bias in the as-classified esti-
mate due to misclassification was investigated in a
simulation study. Discriminant functions from
the cross-validation study were used to classify a
subset of specimens from the independent half of
the spawning-stock collections, and each of the
three types of estimates of relative percentage
were obtained and compared with the known stock
percentage. For percentages of Hudson stock rang-
ing from 0 to 907f , the difference between each
estimate of Hudson percentage and the known
percentage of Hudson specimens in the subsample
was obtained as a measure of bias in the estimate.
Collection, Processing, and Analysis of
Atlantic and Hudson River Specimens
Assessment of the relative contribution of vari-
ous stocks of striped bass to the Atlantic coastal
fishery required a stratified sampling design that
provided samples from the entire coastal fishery
and considered the migratory nature of striped
bass; therefore a geographically and temporally
stratified sampling design was used. The geo-
graphical stratification consisted of 10 strata from
southern Maine to Cape Hatteras, with 2 to 4
substrata within each stratum to compensate for
variations in stock composition within the
stratum ( Figure 1 ). The Rhode Island stratum was
not subdivided because of its small size. Tempo-
rally, the year was divided into six 2-mo periods to
obtain estimates of stock composition by stratum
throughout the year.
Collections of striped bass from the coastal
fishery were obtained primarily from sport and
commercial fishermen; however, in areas where
adequate sport and commercial fisheries did not
exist, study personnel used haul seines and gill
nets to collect specimens. Collections were limited
to striped bass caught during the same day (i.e.,
within 24 h) to assure freshness. In many in-
stances the entire catch was used, but due to the
size of some catches, a random sample propor-
tional to the number of small ( <550 mm), medium
(550-850), and large (>850) striped bass caught
was obtained.
Oceanic and overwintering specimens were pro-
cessed in the same manner as spawning-stock
specimens. Two replicates of 10 counts and mea-
surements were taken from each specimen, and
scale samples were obtained for subsequent age
and growth rate determinations in the laboratory.
A total of 2,737 oceanic specimens with a complete
set of meristic, morphometric, and scale charac-
ters were processed (Table 1). Additionally, 79
striped bass overwintering in Croton Bay on the
Table l . — Number of striped bass with complete character sets' collected by spatial stratum and period from Atlantic coastal
fishery in 1975.
Spatial
stratum
8
9
10
Total
Locality
Legal/
subiegai^
S Maine-N Mass
S. Mass.
Rhode Is.
E. Long Is. Sound
W Long Is Sound
E Long Is S. Shore
W Long Is S Shore
N.J,
Del.-Md -N Va.
S Va.-N C.
Jan. -Feb. Mar.-Apr. May-June July-Aug. Sept -Oct. Nov. -Dec. Total
Legal
82
58
74
214
Legal
91
. 90
82
263
Legal
60
43
56
159
Legal
96
140
99
335
Sublegal
5
1
6
Legal
1
38
14
15
89
157
Sublegal
2
42
85
10
139
Legal
1
89
102
86
106
384
Sublegal
8
17
19
44
Legal
30
58
93
120
124
425
Sublegal
4
11
15
Legal
34
113
28
73
117
365
Legal
71
3
6
100
180
Legal
27
24
51
28
180
672
531
755
571
2,737
'Measurements and counts taken on all variables used in the character set
^Sublegal-sized striped bass (< 406.5 mm FL) from New York waters (strata 4 to 7) were analyzed separately.
338
BERGGREN and LIEBERMAN: RELATIVE CONTRIBUTION OF STRIPED BASS
ATLANTIC OCEAN
ST CROIX R>vEft(
Figure l.— Collection regions
for the Atlantic coastal fishery
for striped bass showing geo-
graphical stratification and
substratification; collection
sites for spawning-stock speci-
mens indicated by dots on source
rivers.
CEAN
10-3
ICAPE HATTERAS
WILMINGTON, SC
CAPE FEAR
339
FISHERY BULLETIN: VOL. 76, NO. 2
Hudson River from 6 December 1974 through 20
March 1975 were processed.
Three estimates of stock contribution, i.e., "as-
classified," "adjusted," and "iterative" estimates,
were calculated for collections of legal-sized,
sublegal-sized, and overwintering striped bass by
geographical and temporal strata. Sublegal-sized
<406.5 mm or 16 in FL) and overwintering striped
bass collected in New York waters were not con-
sidered to be a part of the coastal fishery and were
analyzed separately. In each stratum, the
percentage of striped bass allocated to a stock pro-
vided an estimate of that stock's relative contribu-
tion. Mean 1975 estimates of stock contribution of
legal-sized striped bass were calculated by averag-
ing strata estimates within periods then averag-
ing across the six periods. Relative contribution
estimates by age were also obtained.
The influence of the Hudson stock in coastal
strata adjacent to the Hudson River was investi-
gated by comparing the relative contribution of
Hudson, Chesapeake, and Roanoke stocks within
"inner" and "outer" zones designed by the U.S.
Nuclear Regulatory Commission.'*^ The inner
zone encompassed western Long Island Sound
(stratum 5), the New York Bight (stratum 7), and
northern New Jersey (stratum 8-1), whereas the
outer zone encompassed the remaining waters
from Cape May, N.J., to Maine (strata 1 to 4, 6, 8-2,
8-3). Estimates of relative contribution for inner
and outer zones were calculated for each period by
summing the number of Hudson-, Chesapeake-,
and Roanoke-classified fish within appropriate
strata. Mean estimates of contribution within
each zone were calculated for the year by averag-
ing across temporal strata.
RESULTS AND DISCUSSION
Establishment of Discriminant Functions
Five characters were established as the charac-
ter set best able to discriminate among Hudson,
Chesapeake, and Roanoke stocks. They are, in
order of importance (as established by stepwise
linear discriminant analysis): 1) the ratio of snout
length/internostril width, 2) the scale ratio of first
to second annulus/focus to first annulus measure,
'"U.S. Nuclear Regulatory Commission. 1975. Final environ-
mental statement related to operation of Indian Point Nuclear
Generating Plant Unit no. 3 Consolidated Edison Company of
New York, Inc. Office of Nuclear Reactor Regulation. Docket no.
50-286, Vol. 1:V-166-V-178.
3) a character index (Raney and deSylva 1953), 4)
the upper-arm gill raker count (which includes
rudimentary rakers), and 5) the lateral line scale
count. The character index, i.e., the sum of left and
right pectoral, second dorsal, and anal fin rays,
was used since Grove et al. (1976) demonstrated
that individual fin ray characters did not add sig-
nificant discriminatory ability.
The five characters satisfied the criterion for
independence with fish length in each stock with
only one exception. The snout length/internostril
width ratio for the Roanoke stock has a coefficient
of determination of nearly 0.20 but was retained
because its distribution had the least overlap
among spawning stocks of all characters, thus
making it a potentially good discriminator.
Results of the test of homogeneity indicated that
only the Hudson stock was homogeneous among
ages and between males and females. Significant
differences (a = 0.05) were found among ages and
between sexes in the Chesapeake spawning stock
and among ages in the Roanoke spawning stock.
Differences found in the Chesapeake spawning
stock may have resulted from pooling collections
from its four major tributaries.
Quadratic functions (Table 2) were used to dis-
criminate among stocks as a result of the investi-
gation of underlying assumptions of discriminant
analysis. Significant differences (a = 0.05) were
found among covariance matrices of Hudson,
Chesapeake, and Roanoke spawning stocks which
suggested that quadratic functions would better
discriminate among these stocks than linear func-
tions. Histograms suggested that no radical de-
parture of multivariate normality was evident,
although normality of individual characters does
not assure multivariate normality of the character
set. Therefore multivariate normality of the
character sets was assumed.
Percentage of spawning-stock specimens cor-
rectly classified by the quadratic functions and
estimated stock percentages resulting from the
use of these functions closely agreed with results
obtained by the cross-validation procedure (Table
3). For the total set of collections, 76. 8^^ of Hudson
specimens, 67.T7( of Chesapeake specimens, and
85. 9^^ of Roanoke specimens were correctly clas-
sified, resulting in an overall correct classification
of 74.4%. This was similar to overall percentages
of 73.2 and 77.1 obtained for the cross-validation
subsets. Estimated relative percentages for each
stock varied <3 percentage points among the total
set and cross-validated subsets, whereas varia-
340
BERGGREN and LIEBERMAN: RELATIVE CONTRIBUTION OF STRIPED BASS
Table 2. — Quadratic discriminant functions' based on Hudson, Chesapeake, and Roanoke spawning-stock specimens of striped bass
and used to classify spawning-stock, oceanic, and overwintering specimens.^
Hudson;
''hud =^ 1.489.070559 -
+ 0 090968 UZ +
+ 24 321052 X r
Chesapeake
''CHES = - 1.368 946420
+ 0.321075 C/Z -
+ 25 294896 X +
Roanoke:
'^ROAN = - 1.650 902863
+ 0.228873 UZ
+ 25 512087 X -
(0.077516 U^ i 0.256954 W^ >
0,047441 WX + 0.023246 WY +
7 985031 Y > 381 695141 Z
- (0.089560 U^* 0 242459 W?
0.092151 WX 0,000861 WY
7.014936 Y » 323 469441 Z
1 171065X2 ^
0 164200 WZ
2 536320 Y^
0.457365 Xy
123 907000 Z2
2 799760 XZ
- 0.019058 UW
2.861250 YZ) +
<- 0.015160 UX - 0.007057 UY
8.776221 U + 28,127772 W
1 122690X2 + 2 155850/2 + 117 554000 Z^- 0.007099 UW + 0.005302 UX
2 363980 WZ + 0 381082 XY + 3.623860 XZ - 1 590090 VZ) + 1 1 316822 U
- (0.107062 U2
0.293615 WX
22.351388 Y
+ 0.316254 W2 t
+ 0 129292 WY -
469 422957 Z
2 063540X2
1 009790 WZ
0.842590 Y^ + 139.577500 Z2
0 106776 Xy - 0 606466 XZ +
0 062826 UW + 0.015703 UX
4 416000 YZ) + 10 320202 U ^
0.015500 uy
21 749040 W
0 043640 UY
27 000888 W
'Except for an additive constant ( -2.5 In 2tt) common to each function.
2F = discriminant score. U = lateral line scale count, W = character index, X = upper-arm gill raker count, Y
measurement ratio, and Z = snout length mternostril width ratio.
first to second annulus/focus to first annulus
Table 3. — Comparison of correct-classification percentages and estimated and known stock percentages
among the total set of spawning-stock specimens of striped bass and cross-validation subsets.
Correctly
Spawning classified
stock (%)
Random set'
Known
stock
(%)
Estimated
stock
(%)
Correctly
classified
(%)
Independent set2
Known
stock
(%)
Estimatec
stock
(%)
Correctly
classified
(%)
T^tal^t^
Known
stock
(%)
'Randomly sampled half of total spawmng-stock collections used to determine quadratic functions for cross-validation.
^Remaining half of spawmng-stock specimens classified by quadratic functions based on the random set.
^AII specimens from spawmng-stock collections classified by quadratic functions based on the total set.
Estimated
stock
(%)
Hudson
81 0
337
36.5
72.6
33.6
35.2
76.8
337
36.9
Chesapeake
69 8
466
40.2
68 1
464
424
67.7
465
40.3
Roanoke
878
19.7
233
86.0
200
224
859
198
22.9
Overall
77 1
732
74.4
tions between estimated and known stock
percentages within sets was as much as 9 percent-
age points. The quadratic functions thus provided
slightly biased estimates of stock percentages
when applied to collections composed of 34^^ Hud-
son, 46*^ Chesapeake, and 20'7( Roanoke stocks.
Best Estimator of Relative Contribution
The best estimate of the percentage of Hudson
River specimens in subsamples from the simula-
tion studies was the estimate from the third itera-
tion of the reclassification procedure (Table 4). On
the average, this iterative estimate was less
biased than estimates from other iterations, the
as-classified estimated (i.e. , estimate from the first
iteration), and the adjusted estimate for most per-
centages of Hudson stock considered. In addition,
the variance of the bias of the iterative estimate
was often less than that of the other estimates. For
percentages of Hudson stock 'e 50*^ , the iterative
and adjusted estimates closely agreed and the bias
in each estimate was small ( 'SS percentage
points). The iterative estimate will, therefore, be
used to estimate Hudson stock contribution in
oceanic collections, and the adjusted estimate will
be used to substantiate estimates of Hudson con-
TABLE 4. — Mean and standard deviation of absolute bias' of
estimated relative percentages of Hudson River stock of striped
bass in replicated random samples from spawning-stock collec-
tions.^
Known percent
of Hudson
Estimates of absolute bias
As-classified
Mean SD
Iterat
ive
Adju
Mean
sted
River stock
Mean
SD
SD
^90
23.0
2.40
4.3
2.97
14.4
5 73
80
20.2
5.14
7.4
882
14.3
7.78
75
17.6
3.47
8.4
382
128
5.00
70
13.0
3.32
4.7
452
7.3
4.80
65
10.8
3.11
33
298
7.5
4.07
60
9.0
3.21
53
284
68
5.00
55
7.5
2.95
4.8
3 52
74
4.02
50
5.5
1.99
4.2
266
4.7
3.44
45
2.2
2.17
3.5
1 85
4.7
2 14
40
1.2
1.04
3.2
3.44
43
421
35
2.2
1.45
"4.2
"2 09
44
2.87
30
5.9
3.18
3.3
243
3.4
1.79
25
7.8
3.07
4.0
225
4.2
3.27
20
9.5
2.26
2.8
1.72
1,8
099
15
12.1
4.19
3.5
3.24
33
262
10
15.1
3.80
4.5
3.36
50
3.61
5
17.4
4.03
4.5
3.72
43
385
0
18.1
2.53
2.5
1.73
1.5
1.69
Overall mean
11.0
4.4
6.2
'Absolute value of the difference between the true relative percentage of
Hudson River stock in the subsample and the estimated relative percentage
based on nine replicates of varying Chesapeake and Roanoke proportions in
the subsamples
2Estimates were based on random samples from one-half of spawmng-stock
collections which were classified as to area of origin by quadratic functions
obtained from the other half of the collections.
^Based on two replicates.
■■Based on eight replicates.
tribution « 50^^. The iterative estimate will also
be used to estimate Chesapeake and Roanoke
stock contributions.
341
FISHERY BULLETIN: VOL. 76, NO. 2
Estimates of Stock Contribution for
Oceanic and Overwintering Collections
Iterative estimates of relative contribution of
Hudson, Chesapeake, and Roanoke stocks indi-
cated that the Chesapeake stock was the major
contributor to the striped bass fishery along the
Atlantic coast while the Hudson and Roanoke
stocks were minor contributors (Table 5). The
Chesapeake stock predominated in 34 of 35 geo-
gi-aphical and temporal strata while the Hudson
stock predominated in the remaining stratum.
Iterative estimates of Chesapeake contribution to
the fishery exceeded SO'^ in all strata not adjacent
to the Hudson River. Iterative estimates of the
Hudson stock were largest in western Long Island
Sound and the New York Bight with values ex-
ceeding 20*7^ during some periods. Although itera-
tive estimates of Roanoke stock contribution
never exceeded 20^7^, they were highest in North
Carolina waters (stratum 10) and in strata from
Massachusetts to Maine (strata 1, 2).
The Hudson stock contribution in strata from
Massachusetts north to Maine and from New Jer-
sey south to North Carolina (strata 8 to 10) should
be low as indicated by iterative estimates (Table 5)
and results of tagging studies. Zero estimates in
northern waters do not necessarily indicate an
absence of Hudson River striped bass since the
simulation study has shown that such estimates
may be obtained in situations where true con-
tribution is low. In fact, data on adult striped bass
tagged in the Hudson River during spawning sea-
son and recaptured in waters as far north as Bos-
ton Harbor, Mass., have indicated a northern mi-
gration of a portion of the Hudson stock (Texas
Instruments see footnote 4). However, these data
support near-zero estimates of Hudson contribu-
tion in southern waters since tagged striped bass
were not recaptured south of northern New Jer-
sey. Data (Chapoton and Sykes 1961) on adult
striped bass tagged along the outer coast of North
Carolina and recaptured on the spawning grounds
of Chesapeake Bay and Albemarle Sound
Table 5. — Estimates of relative contribution of Hudson, Chesapeake, and Roanoke stocks of legal-sized
striped bass' to 1975 oceanic collections by period and spatial strata. As-cl. = As-classified, Iter. =
Iterative, and Adj. = Adjusted estimates.
stratum
Sample
size^
Hudson
Chesapeake
Roanoke
Period
As-cl.
Iter.
Adj.
As-cl.
iter.
Adj.
As-cl.
Iter.
Adj.
Jan -Feb.
10
27
25.9
37
6.7
63.0
926
90.7
11.1
3.7
2.6
Mar. -Apr.
5
38
52.6
57.9
54.2
42.1
42.1
458
5.3
0.0
0.0
7
30
23.3
3.3
0.0
73-3
96.7
100,0
33
0.0
00
8
34
235
8.8
08
676
882
992
8.8
29
0,0
9
71
85
0.0
0.0
77,5
97.2
98.9
14.1
28
11
May-June
1
82
110
0.0
00
683
902
885
20.7
98
11,5
2
91
14.3
0-0
00
71 4
95,6
964
14.3
4.4
36
3
60
30.0
3.3
139
600
967
84 5
10,0
0.0
1,6
4
96
21 9
10
0,0
698
990
1000
83
0-0
0.0
5
14
357
286
23.0
57 1
71.4
770
7.1
0,0
0.0
6
89
258
56
5.4
652
933
94.6
9.0
1,1
0.0
7
58
41 4
259
33.7
51 7
707
663
6.9
3,4
0.0
8
113
23.9
0,0
1,5
673
100.0
985
8.8
0.0
0.0
July- Aug
1
58
19.0
00
0,0
67,2
948
95.4
13.8
52
4.6
2
90
7.8
00
0,0
722
96 7
908
20 0
3.3
9.2
3
43
302
2.3
10,3
65 1
977
897
4.7
0.0
0.0
5
15
26 7
0,0
5 1
66 7
100.0
94 9
6.7
0.0
0,0
6
102
225
7,8
1,6
63,7
88.2
927
13.7
3.9
5.7
7
93
333
15,1
13,4
65,6
84.9
86.6
1.1
0,0
0.0
8
28
21 4
0.0
0,0
71 4
100.0
100.0
7 1
0,0
0.0
Sept -Oct
1
74
13.5
0,0
0-0
77,0
98.6
1000
9.5
1,4
00
2
82
12.2
0,0
0,0
58,5
85.4
760
29.3
146
24.0
3
56
250
3,6
7,5
58,9
94.6
833
16.1
1,8
9.2
4
140
16.4
0,7
00
64.3
94.3
886
19.3
5,0
11.4
5
89
41,6
40,4
37,2
46.1
57.3
565
12.4
2.2
6.4
6
86
15.1
0,0
0,0
73.3
96.5
99 8
11.6
3.5
0.2
7
120
233
1,7
29
63.3
950
91.6
13.3
33
5.2
8
73
16.4
0,0
0,0
76.7
986
100.0
68
14
0.0
9
6
16.7
0,0
00
66.7
100,0
922
16.7
00
78
Nov -Dec.
4
99
21.2
00
00
667
980
969
12,1
2.0
3,1
6
106
16.0
00
0,0
698
99 1
959
14,2
0,9
4 1
7
124
21.0
4,8
0,0
76.6
95,2
100 0
24
0,0
00
8
117
21,4
0,0
0.0
72.6
100,0
1000
6.0
0,0
0.0
9
100
8.0
0,0
0.0
80.0
990
100.0
12.0
1,0
0.0
10
24
12.5
0.0
0.0
62.5
83.3
82.0
25.0
16.7
18.0
Overall mean
23.0
6.5
6.6
66.0
90.8
90.2
11.0
2.7
3.2
'Not included
are striped bass <406.5 mm
FL from New York waters.
^Sample sizes
of five specimens or less in
any stratum
are not
included.
342
BERGGREN and LIEBERMAN: RELATIVE CONTRIBUTION OF STRIPED BASS
tributaries also support near-zero estimates of
Hudson River contribution in waters off North
Carolina.
Comparison between iterative and adjusted es-
timates indicated close agreement for each stock
within the 35 strata. The largest difference be-
tween estimates was 12.2 percentage points, but
differences of <5 percentage points occurred in
809c of the strata for the Hudson stock, 71% of the
strata for the Chesapeake stock, and 869r of the
strata for the Roanoke stock. The adjusted esti-
mates therefore substantiate low iterative esti-
mates of contribution of Hudson and Roanoke
stocks.
Comparison of mean iterative and adjusted es-
timates of relative contribution indicated that the
two estimates differed by <1 percentage point for
each stock. Mean iterative and adjusted estimates
were, respectively, 6.5 and 6.6^^ Hudson, 90.8 and
90.2*7^ Chesapeake, and 2.7 and 3.2% Roanoke
contribution.
The contribution of the Hudson stock to the
coastal fishery was greater in strata adjacent to
the Hudson River than in the remaining strata.
Mean iterative estimates of relative contribution
of the Hudson River stock to inner and outer zones
were 16.0% (15.0% adjusted) and 2.8% (0.0% ad-
TabLE 6. — Mean estimates' of relative contribution of Hudson,
Chesapeake, and Roanoke stocks of legal-sized striped bass^ to
1975 oceanic collections within- USNRC zones.^
Inner zone
Outer zone
Estimate
Hudson
Chesa-
peake
Roa-
noke
Hudson
Chesa-
peake
Roa-
noke
As-classified
Iterative
Adjusted
31 7
160
15.0
629
83,1
84.2
55
09
0.8
19.2
2.8
0.0
68.0
94.2
96.4
12.8
3.0
3.6
'Average of five temporal strata since only one striped bass collected in inner
zone during period 1 (Jan. -Feb. ).
^Not included are striped bass <406.5 mm FL from New York waters.
^U.S. Nuclear Regulatory Commission inner zone corresponds to study
strata 5,7, and 8- 1 : the outer zone corresponds to study strata 1 to 4, 6. 8-2, and
8-3.
justed), respectively, for the year (Table 6). Al-
though the Chesapeake stock was the predomi-
nant contributor to both inner and outer zones, the
contribution of the Hudson stock exceeded that of
the Roanoke stock in the inner zone but was less in
the outer zone.
The Hudson stock predominated in collections of
sublegal-sized striped bass in western Long Island
Sound, the New York Bight, and in collections of
specimens overwintering in Croton Bay on the
Hudson River (Table 7). Iterative (and adjusted)
estimates of the percentage of sublegal-sized fish
classified into the Hudson stock in western Long
Island Sound (primarily in Little Neck Bay) and
the New York Bight were at least 80% , but were
less than 40% along the southeastern shore of
Long Island (stratum 6) from May through Oc-
tober. The iterative (and adjusted) estimated of
contribution of the Hudson stock to the overwin-
tering population in the Hudson River was greater
than 95% .
This study has provided additional information
in the importance of dominant year classes of
striped bass. Approximately 52% of the specimens
collected from the coastal fishery in 1975 were
from the 1970 year class, and 77% of them were
classified as Chesapeake fish. Schaefer (1972)
stated that production of young-of-the-year
striped bass in Chesapeake Bay during 1970 was
the largest ever recorded and that this year class
should provide excellent fishing in New York wa-
ters for 6 to 8 yr after recruitment. The presence of
this dominant year class of Chesapeake fish
confirms the rationale used by Merriman (1941)
and Schaefer (1968) to conclude that the
Chesapeake stock predominates in the coastal
fishery. A summary of the occurrence of dominant
year classes in the Atlantic coastal fishery has
been given by Schaefer (1968).
Table 7.— Estimates of relative contribution of Hudson, Chesapeake, and Roanoke stocks of sublegal-sized striped bass' to
New York waters by period and spatial stratum and of legal-sized striped bass to the overwintering population in the Hudson
River. As-cl. = as-classified, Inter, = iterative, and Adj. = adjusted estimates.
Period
Stratum
Sample
size^
Hudson
Chesapeake
Roanoke
Population
As-cl.
Iter.
Adj.
As-cl.
23.7
Iter
2.6
Adj.
4.3
As-cl.
0.0
Iter.
0.0
Adj
Ovenwintering
76
76.3
97.4
95.7
0.0
Sublegal
IVIay-June
5
42
92.9
100.0
100.0
7,1
0.0
0.0
0.0
0.0
0.0
6
8
12.5
0.0
00
500
62 5
64.3
37.5
37,5
3b /
7
11
81 8
81.8
1000
18,2
182
00
0.0
0.0
0,0
July-Aug.
5
85
88.2
100.0
100.0
11 8
00
00
0.0
0.0
0,0
6
17
41.2
35.3
392
41.2
588
47.4
17.6
5.9
13,4
Sept.-Oct.
5
10
80.0
80.0
100.0
20.0
20.0
0,0
0.0
0.0
0,0
6
19
26.3
15.8
20.8
36.8
47.4
41.9
36.8
36.8
3/.3
'Striped bass <406.5 mm FL from New York waters.
^Sample sizes of five specimens or less in any stratum are not included. Three sublegal-sized specimens collected ovenwintering in the Hudson River
were classified as Hudson fish.
343
FISHERY BULLETIN: VOL, 76. NO. 2
SUMMARY AND CONCLUSIONS
A study was conducted to identify the origin of
striped bass collected in the Atlantic coastal
fishery and estimate the relative contribution of
major stocks to the fishery. Quadratic discrimi-
nant analysis was applied to values of five mor-
phological characters obtained from Hudson,
Chesapeake, and Roanoke spawning-stock speci-
mens to determine functions which best separated
the stocks. Correct-classification percentages of
76.8, 67.7, 85.9'7f were obtained for the Hudson,
Chesapeake, and Roanoke spawning stocks, re-
spectively, resulting in an overall correct clas-
sification of 74. 47c of the specimens.
A simulation study was conducted to investi-
gate the bias in as-classified, iterative, and ad-
justed estimates of relative contribution due to
misclassification error inherent in the discrimi-
nant functions. Results indicated that iterative
estimates may best approximate the true con-
tribution of the Hudson stock in oceanic collec-
tions.
A stratified sampling design was used during
six 2-mo periods in 1975 to collect representative
samples of striped bass in the Atlantic coastal
fishery from southern Maine to Cape Hatteras.
This provided estimates of stock composition by
stratum throughout the year.
Oceanic samples were classified by discriminant
functions and as-classified, iterative, and revised
estimates of relative contribution of the major
stocks were obtained. Mean iterative estimates of
relative contribution for 1975 are 6.59c Hudson,
90.87c Chesapeake, and 2.7% Roanoke stocks.
Iterative estimates of Hudson contribution for
legal-sized striped bass exceeded 207c only in
western Long Island Sound and the New York
Bight during certain months. In collections from
Western Long Island Sound and the New York
Bight, iterative estimates of the percentage of
sublegal-sized fish classified into the Hudson stock
were at least 807c during the May through October
periods. For Hudson River collections of overwin-
tering striped bass, an iterative estimate of 97.4%
Hudson stock was obtained.
The occurrence of a dominant year class was
noted. Approximately 52% of the legal-sized
specimens collected in the 1975 oceanic sampling
program were from the 1970 year class, and 77% of
these were classified as Chesapeake in origin.
Major conclusions drawn from the study are: 1)
the Chesapeake stock is the major contributor to
the Atlantic coastal striped bass fishery from
southern Maine to Cape Hatteras; 2) the
Chesapeake stock is also the major contributor of
legal-sized striped bass in the vicinity of the Hud-
son River (western Long Island Sound and the
New York Bight); 3) sublegal-sized striped bass
collected in the vicinity of western Long Island
Sound and the New York Bight are predominantly
of Hudson origin; and 4l striped bass overwinter-
ing in the Hudson River are predominantly of
Hudson origin.
ACKNOWLEDGMENTS
We acknowledge Thurman L. Grove who in-
itiated the study and was instrumental in its suc-
cessful completion, and George A. Roth who
helped gather and process the data and aided in
report writing. We also thank Eddie Baldocchi,
Dana Grass, Michael Locke, Edwin Manter,
Ronald McGratten, Thomas Orvosh, Martin Ot-
ter, and Donald Strout for their help in gathering
and processing the data; John Bennett and
Leanna Pristash for their work in computer pro-
gramming; and Dennis DuBose for his help with
the Schuermann and Curry methodology. This
study was carried out under contract to Consoli-
dated Edison Company of New York, Inc., as part
of the Hudson River Ecological Survey.
LITERATURE CITED
ALPERIN, I. M.
1966. Dispersal, migration and origins of striped bass from
Great South Bay, Long Island. N.Y. Fish Game J.
13:79-112.
AMOS. M. H., R. E. ANAS. AND R. E. PEARSON
1963. Use of a discriminant function in the morphological
separation of Asian and North American races of pink
salmon, Oncorhynchus gorbuscha (Walbaum). Int.
North Pac. Fish. Comm., Bull. 11:73-100.
ANAS. R. E., AND S. MURAL
1969. Use of scale characters and a discriminant function
for classifying sockey salmon (Oncorhynchus nerka) by
continent of origin. Int. North Pac. Fish. Comm., Bull.
26:157-192.
ANDERSON, T. W.
1958. An introduction to multivariate statistical anal-
ysis. John Wiley & Sons, N.Y., 374 p.
CHAPOTON. R. B., AND J. E. SYKES
1961. Atlantic coast migration of large striped bass as
evidenced by fisheries and tagging. Trans. Am. Fish. Soc.
90:13-20.
CHITTENDEN, M. E., jR.
1971. Status of the striped bass, Morone saxatilis, in the
Delaware River. Chesapeake Sci. 12:131-136.
344
BERGGREN and LIEBERMAN: RELATIVE CONTRIBUTION OF STRIPED BASS
FUKUHARA, F. M., S. MURAI. J. J. LALANNE,
AND A. SRIBHIBHADH
1962. Continental origin of red salmon as determined from
morphological characters. Int. North Pac. Fish. Comm.,
Bull. 8:15-109.
Grove, T. L., T. J. Berggren, and D. A. Powers
1976. The use of innate tags to segregate spawning stocks
of striped bass, Morone saxatilis. In M. Wiley (editor),
Estuarine processes. Vol. 1, p. 166-176. Academic Press,
N.Y.
HUBBS, C. L., AND K. F. LAGLER.
1958. Fishes of the Great Lakes region. Revised ed.
Cranbrook Inst. Sci., Bull. 26, 213 p.
KENDALL, M. G., AND A. STUART.
1968. The advanced theory of statistics. Vol. 3. Hafner
Publ. Co., N.Y., 557 p.
KOO, T. S. Y.
1970. The striped bass fishery in the Atlantic States.
Chesapeake Sci. 11:73-93.
MANSUETI, R.
1961. Age, growth, and movements of the striped bass,
Roccus saxatilis, taken in size selective fishing gear in
Maryland. Chesapeake Sci. 2:9-36.
MERRIMAN, D.
1941. Studies on the striped bass, Roccus saxatilis, of the
Atlantic coast. U.S. Fish Wildl. Serv., Fish. Bull. 50:1-
77.
Messieh, S. N.
1975. Delineating spring and autumn herring populations
in the southern Gulf of St. Lawrence by discriminant
function analysis. J. Fish. Res. Board Can. 32:471-477.
MOSTELLER, F., AND J. W. TUKEY.
1968. Data analysis, including statistics. In G. Lindzey
and E. Aronson (editors). The handbook of social psychol-
ogy, 2d ed., p. 80-203. Addison- Wesley, Reading, Mass.
NICHOLS, P. R., AND R. V. MILLER.
1967. Seasonal movements of striped bass, Roccus
saxatilis (Walbaum), tagged and released in the Potomac
River, Maryland, 1959-61. Chesapeake Sci. 8:102-124.
Raney, E. C, and D. p. DE Sylva
1953. Racial investigations of the striped bass, Roccus
saxatilis (Walbaum). J.Wildl. Manage. 17:495-509.
Raney, E. C, W. S. Woolcott, and a. G. Mehring.
1954. Migratory pattern and racial structure of Atlantic
coast striped bass. Trans. 19th North Am. Wildl. Nat.
Resour. Conf , p. 376-396.
SCHAEFER, R. H.
1968. Size, age composition emd migration of striped bass
from the surf waters of Long Island. N.Y. Fish Game J.
15:1-51. 1972. Striped bass. Conservationist 27:27,
46.
Vladykov, V. D., AND D. H. Wallace.
1952. Studies of the striped bass, Roccus saxatilis (Wal-
baum), with special reference to the Chesapeake Bay re-
gion during 1936-1938. Bull. Bingham Oceanogr. Col-
lect 14:132-177.
WORLUND, D. D., AND R. A. FREDIN.
1962. Differentiation of stocks. In Symposium on pink
salmon, p. 143-153. H. R. MacMillan Lectures in
Fisheries, Univ. B.C., Vancouver, Can.
345
COPPER SENSITIVITY OF PACIFIC HERRING,
CLUPEA HARENGUS PALLASl,
DURING ITS EARLY LIFE HISTORY
D. W. Rice, Jr and F. L. Harrison'
-i
ABSTRACT
Embryos and larvae of the Pacific herring, Clupea harengus pallasi, were exposed to copper, using a
flow-through bioassay system. Herring embryos were exposed continuously from 12 h after fertiliza-
tion until hatching, and larvae were exposed from the time of hatching until yolk sac absorption.
Embryos were also exposed to 36-h duration pulses of copper in order to evaluate the sensitivy of
different developmental stages of herring embryos to copper. Pulsed exposures started at 62, 98, or 136
h after fertilization. The following measurements were taken as indices of the toxic effects of copper:
cumulative mortality, percent hatching, and larval length upon hatching.
The onset of mortality of herring embryos continuously exposed to copper began 90 h after fertiliza-
tion, with deaths occuring over a short interval thereafter (response period). Significant embryo
mortalities occurred at a copper concentration as low as 35 /xg/1. Herring larvae continuously exposed
to copper showed significant mortality at 300 ju.g/1 copper, with no delay in the onset of mortality.
Embryos exposed to 36-h pulses of copper during different developmental stages showed reduced
sensitivity when exposed after the response period. Larvae that hatched from eggs exposed to a 36-h
pulse of copper before the response period grew significantly less than those hatched from eggs exposed
during later developmental stages.
Numerous studies have shown that many aquatic
animals are adversely affected by increased levels
of copper in water; most of the work on fishes has
been restricted to freshwater species (Becker and
Thatcher 1973; Brungs et al. 1976). Since 90^f of
the world's marine fish are taken from the conti-
nental shelf and nearshore upwelling areas (Wal-
dichuk 1974), increases in copper pollution in
coastal aquatic ecosystems are of particular con-
cern.
The concentration of copper in unpolluted near-
shore waters ranges from 0.3 to 3.8 /Lig/1 (Chester
and Stoner 1974). Increased concentrations of cop-
per in coastal waters have resulted from the re-
lease of municipal waste waters (Mytelka et al.
1973; Mitchell and McDermott 1975) and of
effluents from power plants ( Hoss et al. 1975; Mar-
tin et al. 1977). In polluted waters, concentrations
as high as 13,900 /xg/l copper have been reported
(Mitchell and McDermott 1975).
Examination of the toxic effects of copper on
coastal marine fisheries is important for the estab-
lishment of water quality standards that will pro-
tect fishery resources of coastal zones. Eggs and
'Environmental Sciences Division, Lawrence Livermore
Laboratory, University of California, Livermore, CA 94550.
Manuscript accepted September 1977.
FISHERY BULLETIN; VOL. 76, NO. 2, 1978.
larval stages of fish are reported to be the life
history stages that are most sensitive to a variety
of pollutants (Skidmore 1965; Pickering and Gast
1972; Struhsaker et al. 1974; Christensen 1975).
The necessity of conducting toxicity tests during
the most susceptible stage in the life history of an
organism has been emphasized by Hynes (1970),
and the sensitivity of vertebrate embryos to heavy
metals has been suggested as a criterion for water
quality by Birge and Just (1973).
While some work has been done to assess the
toxicity of copper to the early life history stages of
freshwater fishes (Mount 1968; Hazel and Meith
1970; McKim and Benoit 1971; O'Rear 1972;
Gardner and La Roche 1973; Benoit 1975), little
assessment has been made of toxic effects of copper
on marine fishes. Such studies should be very im-
portant since mortalities that occur during the
early life history stages of marine fish strongly
influence the strength of a given year class offish
(May 1974; Bannister et al. 1974; Postuma and
Zijlstra 1974; Gushing 1975; Vaughan and Saila
1976). Toxic effects that have an impact upon sur-
vival during early developmental stages would
also act to reduce the strength of a given year class
offish. The embryos and larvae of the Pacific her-
ring, Clupea harengus pallasi, represent a useful
347
FISHERY BULLETIN: VOL. 76. NO 2
test organism for evaluating the toxic effect of
copper upon the early life history stages of marine
fish. The Pacific herring is a commercially impor-
tant fish that spawns along both eastern and w^est-
ern Pacific coasts (Eldridge and Kail 1973; Hart
1973). Herring spawn great numbers of demersal,
adhesive eggs on shallow intertidal substrates.
The egg is relatively large, 1.3 to 1.6 mm in diame-
ter, and is covered by a thick, three-layered,
opaque chorion (Blaxter and Holliday 1963). De-
velopment of the embryo is comparatively slow,
taking 7 to 9 days at 14 °C (Alderdice and Velsen
1971). The tough chorion permits easy collection
and handling, and the slow development of the
embryo allows observation time not available in
more rapidly developing species.
Three bioassays were conducted to evaluate the
sensitivity of herring embryos and larvae to cop-
per. The first two assays were designed to evaluate
the sensitivity of embryos and larvae to continu-
ous copper exposure, while the third examined the
sensitivity of embryos to brief copper exposures.
Since the form of copper to which the herring em-
bryos and larvae were exposed may play a sig-
nificant role in the toxic response ( Pagenkopf et al.
1974), the partitioning of copper among the com-
ponents of the water in the bioassay system was
also determined.
MATERIALS AND METHODS
Collection and Handling of Test Organisms
Intertidal collections of Pacific herring eggs
were made along the shore of Belvedere Island and
the Tiburon Peninsula, San Francisco Bay, Calif.
The eggs were collected directly into a 15-gal, in-
sulated ice chest containing aerated seawater
from the egg collection site and were transported
to the laboratory within 2 h after collection. The
water temperature at the collection site was
11.0°-11.5°C and upon arrival at the laboratory
the temperature of the water increased to no more
than 13.5°C. Only eggs deposited in single layers
onFucus si).,Laminaria sp., or Gracilaria sp. were
chosen for testing. Before placing the eggs into
exposure chambers, they were removed from the
seaweed by bending the frond and then gently
brushing them with a finger into a sorting dish
containing seawater kept at 12°C. The eggs were
examined with a microscope at 20 x; only viable
embryos at the same stage of development were
chosen. No more than 51 embryos were placed in
any exposure chamber. All transfers of embryos or
larvae were carried out with a large-bore, polished
glass pipette.
Embryos at two different stages of development
were used for the tests. The age of the earlier stage
embryos, collected 7 February 1975, was esti-
mated to be 12 h after fertilization since they were
undergoing epiboly (Ahlstrom's stage IV
( Ahlstrom 1943)). These embryos were exposed to
copper continuously, each of seven groups being
exposed to a different copper concentration. The
age of the later stage embryos, collected 26 Feb-
ruary 1975, was estimated to be 48 to 50 h after
fertilization (Ahlstrom's stage IX). These embryos
were divided into four groups. Three groups were
exposed for 36 h to the same copper concentration
but during different developmental stages. The
fourth group was maintained in flowing seawater
from the time of collection to within 1 h after
hatching, and then continuously exposed to differ-
ent copper concentrations as larvae.
Bioassay System
The organisms were exposed to copper in 5-1
clear plastic bowls (Figure 1). The exposure solu-
tion was introduced into each chamber by gravity
flow from a mixing chamber into which seawater,
at a rate of 11 ml/min, and copper chloride solu-
tion, pH 3, at a rate of 1 ml/h, were pumped con-
tinuously. Approximately 17 h was required for
replacement of 90'7f of the water in the chamber.
The height of water in the exposure chambers was
maintained by a constant-level out-flow siphon.
The diameter of the mouth of the out-flow siphon
Seawate
Copper
Plastic
mixing
chamber
Outflow siphon
Clear plastic cover
Clear plastic o o
^exposure chamber °^
Figure 1 . — Diagram of the exposure chamber and flow- through
dehvery system used to expose Pacific herring embryos and
larvae to copper.
348
RICE and HARRISON: COPPER SENSITIVITY OF PACIFIC HERRING
was greater than that of the tubing to reduce the
flow velocity at the mouth of the siphon. The
mouth of the siphon was covered with nylon net-
ting (505-)U,m pore size) to prevent the loss of or-
ganisms from the chamber. A gentle stream of
bubbles delivered to the bottom of the chamber
provided aeration and mixing. All exposure
chambers were immersed in a water bath whose
temperature was monitored. Illumination was
provided by the fluorescent lighting in the
laboratory and followed the regular ambient
photoperiod.
The exposure period, the nominal copper con-
centration, and the total number of embryos or
larvae exposed during a typical experiment are
given in Table 1 . Each experiment was repeated at
least once. All exposures were initiated by the
addition of appropriate amounts of copper chloride
to the chambers. The continuous exposure of the
test organisms continued until all animals died or,
in the case of embryos, until hatching occurred or,
in the case of the larvae, until yolk sac absorption
occurred. The pulsed exposures were terminated
by transferring exposed embryos to an exposure
chamber containing control seawater.
The following measurements were taken as in-
dices of the toxic effect of copper: cumulative mor-
tality with time, percent hatching, and larval
length at hatching. The embryos or larvae were
examined within the exposure chambers at each
observation period with a 7 x beam dissection mi-
croscope with a 21-cm depth of field. The criterion
for embryo death was the lack of heart beat or body
movement. Since the embryos were attached in
clusters, dead embryos were not removed until the
termination of a test. Larvae that hatched from
pulse exposed embryos were collected, anes-
thesized with a 1% quinaldine solution, and pre-
served in 5% Formalin^ in seawater. Measure-
ments of the hatched larvae were made with an
ocular micrometer and all obvious deformities
noted. The criterion for larval death was a failure
to respond to a gentle prod with a polished glass
rod. Dead larvae were removed during each obser-
vation period and preserved in 5% Formalin in
seawater.
Total copper concentrations were measured two
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
Table l. — Experimental conditions and median lethal times for bioassays determining the sensitivity of Pacific herring embryos
and larvae to copper.
Exposure
Nominal copper
Number of
Mean total copper
Number of
Time to median lethal
period
concentration
organisms
concentrations
water
level (LTso ± 95% confi-
Experiment
(h)
(mq/i)
exposed
(fxgn ± SD)
samples
dence interval) (h)
Embryos, continuous
i 180
Control
150
4.3 ± 1.9
3
(')
exposure
(12 h atter •
25
48
27 9 ± 7.4
3
V)
fertilization
35
49
38.1 ± 9.4
3
2144.9 ± 8.3
through
45
50
44.1 ± 11.6
3
2134.4 ± 5.2
hatching)
55
49
51.2 ± 9.7
3
2134.8 ± 3.0
100
53
127,9 ± 34.6
2
2115.4 ± 2.8
200
49
235.5 ± 47,2
2
298.7 ± 2.1
Larvae, continuous
300
Control
100
2.5 ± 0.8
3
D
exposure
(Hatching
300
49
274.0 ± 24.1
4
(1,3)
through
600
49
572.1 ± 31.5
4
(1,2)
yolk sac
1,400
50
1,349.0 ± 247.0
4
241.7 ± 7.3
absorption)
2,000
51
1,969 0 ± 148.0
2
223.8 ± 1.5
2,500
49
2,425.4 ± 89.0
2
220.9 ± 2.4
3,500
51
3,430.5 ± 710.6
2
"15.6
Time to median lethality
following termination of
pulse ± 95% confidence
interval (h)
Embryos, pulsed
exposures;
Pulse 1
36
Control
49
3.0 ± 0.08
2
(')
(62 through 98
100
44
93,8 ± 6.3
3
238.4 ± 2.0
h after
fertilization)
Pulse II
38
Control
93
3.9 ± 1.6
3
D
(98 through 136
100
94
111.9 ± 13.7
4
253.9 ± 6.1
h after
fertilization)
Pulse III
36
Control
46
6,1 ± 0.7
3
0
(136 through
100
48
101.6 ±29.3
3
(')
172 h after
fertilization)
'50% mortality not achieved at this concentration.
2Slope significantly different from control slope (P<0.01) (Snedecor and Cochran 1967).
^Slope significantly different from control slope (P<0.05) (Snedecor and Cochran 1967).
•"Determined according to the method of Litchfield and Wilcoxon (1949).
349
FISHERY BULLETIN: VOL. 76. NO. 2
to four times during each bioassay to determine
the actual concentrations to which the organisms
were exposed. Water samples were collected in
acid-washed polyethylene jars and acidified to pH
2 with concentrated HCl. Total copper was
analysed by the APDC-DDDC-MIBK extraction
method described by Kinrade and VanLoon
(1974). The copper concentration in extracted
MIBK solutions was determined with a model 303
Perkin Elmer atomic absorption spectrophotom-
eter, using an HGA-2100 graphite furnace with a
deuterium background corrector.
Since the chemistry of copper in seawater is
complex, more than one form of copper may be
present in the bioassay water. To examine the
form of the copper in the bioassay system water,
out-flow samples were collected from the bioassay
system before organisms were introduced to de-
termine the particulate-bound fractions
(>0.45^im), ionic fraction (bound by Chelex-100
resin (Riley and Taylor 1968)), and complexed
fraction (not bound by Chelex-100 resin). The
analysis scheme is summarized in Figure 2. To
monitor the partitioning of copper into each of
these fractions, copper-64 was equilibrated with
water samples after they were withdrawn from
the bioassay system. The partitioning of stable
copper in the seawater of the bioassay system was
indicated by the percentage of the initial activity
recovered in each of the described fractions.
Statistical Analysis
The measures of toxicity determined in this
study were the time to 50^f mortality at each con-'^
centration of copper tested (median lethal time,
LT50) and the concentration of copper resulting in
50% mortality over a given time (median lethal
concentration, LC50). These toxicity measures
were deterimed by performing a weighted linear
regression analysis on the sets of cumulative mor-
tality data using the logistic function. The
straight line transform of the logistic function is:
logit P = In PIQ =a + (3x, so that if logit P is plot-
ted against X, the points will fall on a straight line
with a as the intercept and /3 as the slope (Berkson
1953). In our calculations of LT50, x represented
the time from the onset of the reaction period in
the case of continuous embryo exposures, from
hatching in the case of continuous larval expo-
sures, and from the termination of a given pulse in
the case of pulsed embryo exposures. In our calcu-
lations of LC50, X represented concentration,
350
Seawater
APDC-DDDC-MIBK
extraction
0.45 u filter
Total
fraction
Chelex 100
ion exchange
resin
APDC-DDDC-MIBK
extraction
on waste
water
Particulate
bound
fraction
Ionic
fraction
_ Complexed
fraction
Figure 2. — Analysis scheme for the separation of copper frac-
tions recovered from the bioEissay system used to expose Pacific
herring embryos and larvae to copper.
and our method followed that outlined by the
American Public Health Association (1976) with
logit analysis used in place of probit analysis.
A computer was used to calculate the LC50 and
LT50 values, and for each fitted line the program
determined: the LT50 or LC50, the 959^ confidence
limits associated with the LT50 or LC50, Pearson's
rho (p), the slope (/3) and the intercept (a), and the
mean square error (EMS); no assumptions of
homogeneity were made and the EMS was calcu-
lated in every case, rather than assuming an EMS
of 1 for homogeneous data (Finney 1964).
In the case of embryos that were exposed con-
tinuously or exposed to pulses of copper, deaths
prior to the delayed reaction period or the onset of
the pulsed exposure, respectively, were not used in
the data analysis. In no case were mortalities dur-
ing these periods greater than 6%.
The relationship between time to 50% mortality
during continuous exposure of both embryos and
larvae and concentration was determined follow-
ing the method outlined in the American Public
Health Association (1976). The resulting toxicity
curve was used to estimate the lethal threshold
concentration (incipient LCgo) (Sprague 1969).
RESULTS
Physical Parameters of the Bioassay System
Mean copper concentrations measured during
each test are reported in Table 1. The partitioning
RICE and HARRISON. COPPER SENSITIVITY OF PACIFIC HERRING
of copper-64 among particulate-bound, ionic, and
complexed fractions of copper recovered from the
bioassay water indicates that the copper was
primarily in the ionic form {Table 2). The mean pH
of the water in exposure chambers in all tests was
8.08 (SD = ±0.024). The mean temperature for all
tests was 13.3°C (SD = ±0.8°C).
Table 2. — Percentages of copper-64 in fractions of sea water
recovered from the bioassay system used to expose Pacific her-
ring embryos and larvae to copper.
Nominal copper
concentration
(^g/i)
Particulaie
bound
Ionic
Complexed
Total'
10
5.2
83.6
4.3
93.7
50
4.2
88.3
3.7
96.5
100
1.1
89.9
2.1
940
500
0.7
938
1,7
970
1.000
0.6
91,8
1.1
969
2,000
1.0
96.4
1.0
99.9
'Total Includes copper-64 remaining in Chelex-100 resin after elution.
Continuous Exposures to Copper
Survival of embryos continuously exposed to
copper was high at all concentrations of copper
tested until 90 h of exposure, at which time dose-
related mortalities occurred (Figure 3). The period
during continuous exposure when embryo deaths
begin is termed the reaction period. Mortalities of
developing embryos at copper concentrations of 35
/ug/l and higher were significantly different from
controls (P<0.01). Virtually no hatching occurred
at concentrations above 45/Ltg/l copper. Develop-
mental features observed in the control embryos
during the onset of the reaction period included
the appearance of eye pigmentation, the onset of
coordinated body movements, and the initiation of
heart beat. Embryos continuously exposed to cop-
per concentrations of 100 /i.g/1 copper and higher
developed an opaque cast to the chorion, which
was followed later by whitish discoloration of the
body. Embryos continuously exposed to 200 /u.g/1
copper developed an opaque change in the chorion
at 60-72 h from fertilization, with body discolora-
tion, spasmodic contractions, quiverings, and re-
duced fin fold development occurring at 84-96 h
from fertilization.
Herring larvae continuously exposed to copper
were many times less sensitive to copper than
herring embryos. Larval mortalities differed sig-
nificantly from controls at concentrations of 300
/Ltg/l copper and higher (P<0.05) (Figure 4). Prior
to death the larvae sank to the bottom of the expo-
sure chamber and patches of whitish discoloration
were observed over the bodies. Spasmodic quiver-
ing and whole body contractions were observed in
larvae at concentrations of 1,400 /xg/1 copper and
above.
The toxicity curves for continuously exposed
herring embryos and larvae are shown in Figure 5.
Median lethal times for each copper concentration
tested and 95% confidence limits are detailed in
Table 1.
100
Figure 3. — Percent cumulative mor-
tality of Pacific herring embryos con-
tinuously exposed to copper (micro-
grams per liter). Mortality curves
shown are the fitted logit curves used to
establish median lethal times.
+->
C
0)
u
s.
4J
o
>
Z3
-H Hatching
0 20 40 60 80 100 120 140 160 180
Hours from fertilization
351
FISHERY BULLETIN: VOL. 76, NO. 2
FIGURE 4. — Percent cumulative mor-
tality of newly hatched Pacific herring
larvae continuously exposed to copper
(micrograms per liter). Mortality curves
shown are the fitted logit curves used to
establish median lethal times.
4->
c
Q)
O
S-
0)
Q.
■4->
S-
o
>
+->
o
^■a n I n I D u °i — \j^
<0^^=^^
Control
J \ \ I
0 30 60 90 120 150 180 210 240 270 300
_J Yolk sac L
absorbtion ^
Hours from hatching
o
-M
S-
o
o
o
E
240
192
144k
96
24
10
T — r
^--1- Hatching completed -^-
■1-
Initiation of
yolksac
absorption
EMBRYOS
A=LC5o
A = LT50 ± 95% confidence limits
LARVAE
D = LC50
■ = LT50 ^ 95% confidence limits
J L
10
20 30
50
1000
100 300
Copper - pg/£
Figure 5. — Toxicity curves for Pacific herring embryos and larvae continuously exposed to copper.
3000
352
RICE and HARRISON: COPPER SENSITIVITY OF PACIFIC HERRING
The toxicity curve for herring embryos is pre-
sented for the purpose of discussion only, since the
90 h delay until the onset of mortality, regardless
of concentration, biases the toxicity curve for com-
parison with other organisms without a reaction
period. Sprague (1969) recommended that a con-
centration that killed SO^r of the population dur-
ing an exposure sufficiently long that acute lethal
action has ceased (incipient LC50) be used as the
single most useful criterion for toxicity. The inci-
pient LC50 is not influenced by the bias introduced
by the reaction period. The estimated incipient
lethal level for herring embryos was found to be 33
/xg/1 copper.
Only larval deaths earlier than 100 h after
hatching were considered in the construction of
the larval toxicity curve since larvae surviving
beyond approximately 200 h after hatching have
begun yolk sac absorption, and the apparently
synergistic effects of copper stress and starvation
can be observed in the larval time vs. percent
mortality curves (Figure 4). The estimated incip-
ient lethal level for herring larvae was found to
be 900 /xg/1 copper.
Thirty-six Hour Pulsed Embryo Exposures
Pulses of copper exposure for 36 h showed that
the sensitivity of herring embryos to copper
100
changed as the embryos developed (Figure 6,
Table 1). A 36-h pulse of 100 /xg/l copper delivered
during the reaction period (Pulse I) had the
greatest effect upon hatching and the length of
larvae at hatching (Table 3). A 36-h pulse of 100
fxg/\ copper delivered just before hatching (Pulse
III) had a signficant effect on larval length at
hatching, but the percentage of embryos hatching
was actually greater than controls.
Table 3. — Percent Pacific herring embryos hatching and mean
larval length at hatching for three groups of Pacific herring
embryos exposed to 36-h pulses of 100 /ng/1 copper. Each group
received a pulse at a different time during development.
Item
Mean larval length
(mm ± SD)
Percent hatching
Controls
6.10 ± 0.47
92
Pulse 1
3.77 ± 0.2'
6
Pulse 2
4.23 ± 0.31*
47
Pulse 3
5.75 ± 0.62-
98
•Significantly different from controls (P<0.01) (Snedecor and Cochran
1967).
DISCUSSION
Several features of the toxic response of herring
at various stages of their early life history are of
interest. Previous tests examining the sensitivity
of other fish embryos and larvae to copper have
found that the larval stage is the more sensitive
stage (Hazel and Meith 1970; McKim and Benoit
1971, O'Rear 1972; Gardner and La Roche 1973;
c
a;
o
0)
Q.
80 -
> 60 -
(0
•M
L.
o
E
>
JO
Z3
E
D
o
40
20 -
0
1
1 1 /I
A>r*
//
Continuous—^ f
W^- Pulse! J
1 1 /■
/ / ■/
/ A /
■^M
/ / Pulse 11-^
1 / ^Pulse III/
—
/ \ yf- Control
/ / x^ i n ^-t-r"^
/ / ^^'''^v Uj-*"^"^'^^
1
1 y ^^..^m^S-^^^m^
40
80 120 160
Hours from fertilization
200
Figure 6. — Percent cumulative mor-
tality of three groups of Pacific herring
embryos exposed to 36-h pulses of 100
/j.g/1 copper. Each group received a pulse
at a different time during development.
The cumulative mortality observed for
Pacific herring embryos continuously
exposed to 100 /xg/l copper (See Figure
3) is shown for comparison.
Pulse
I
Pulse
Pulse
353
FISHERY BULLETIN: VOL. 76, NO. 2
Benoit 1975). Contrary to these findings, we found
that embryos of the Pacific herring appear to be
the stage that is more sensitive to copper. It should
be noted that the fishes examined in previous
studies spawn in fresh or brackish waters and
cannot be considered true marine species as is the
herring.
Another interesting feature of the toxic re-
sponse of the herring embryos and larvae was that
the behavior prior to death was similar to that of
adult fish exposed to copper. Jerky, uncoordinated,
and spontaneous movements were noted by Baker
(19691 in the winter founder, Pseudopleuronectes
americanus, acutely exposed to 3,200 and 1,000
juig/1 copper. Bluegill, Lepom/s macrochirus,
chronically exposed to 162 /u,g/l copper showed
periodic involuntary spasms several weeks prior
to death (Benoit 1975). The spasmodic contrac-
tions and quiverings noted in herring embryos and
larvae prior to death might be of a similar nature.
Baker noted that these symptoms are similar to
those of Wilson's disease which also manifests
spasmodic muscle contractions and quiverings in
mammals. Wilson's disease is the result of an in-
born error of metabolism that results in an excess
of unbound copper in the blood stream (Adelstein
and Vallee 1962). Goldfish, Carassius auratus,
subjected to doses of 1,000 /xg/1 copper exhibit se-
vere neurotoxic symptoms and accumulate copper
in nervous tissues at levels similar to those seen in
Wilson's disease (Vogel 1959).
Some of the toxic effects observed in herring
embryos and larvae were similar to those reported
for other heavy metals. Striped bass, Morone
saxatilis, embryos exposed to copper or zinc
(O'Rear 1972) and Baltic needlefish, Belone bel-
one, exposed to cadmium (Dethlefsen et al. 1975)
developed opaque discoloration of the chorion dur-
ing exposure. In the present study, the chorion of
the herring embryos became increasingly opaque
as exposure to copper continued. Wedemyer (1968)
found that in coho salmon, Oncorhynchus kisutch,
70% of the total zinc-65 uptake during exposure
was firmly bound to the chorion, 26% was bound in
the perivitelline space, and only 2% reached the
yolk and 1% reached the embryo. Wedemyer
(1968) also demonstrated that copper is bound by
the salmon embryo's chorion. The opaque discol-
oration noted in herring embryos with continued
exposure to copper may well be a reaction result-
ing from copper uptake by the chorion.
The observation of a reaction period during
bioassays with herring embryos has been noted
previously. A reaction period for herring embryos
continuously exposed to cadmium (Rosenthal and
Sperling 1974) and high temperatures and
salinities (Alderdice and Velsen 1971) occurred at
about the time of the onset of heart beat. The
sensitivity of this developmental period in the
herring was further borne out by our findings in
which 36-h pulses of 100 /xg/1 copper during the
reaction period caused higher mortalities than
36-h pulses during later developmental periods.
Pacific herring embryos may be vulnerable to
toxic effects from effluents now being discharged
into coastal environments. A survey of 108 muni-
cipal waste effluents on the Atlantic coast showed
that 50% of the waste effluents contained >100
H.gl\ copper; some discharges were as high as 5,900
/Ltg/1 copper (Mytelka et al. 1973). A survey of six
municipal waste discharges along the southern
California coast revealed concentrations ranging
from 74 to 13,900 ^xgl\ copper with an average
annual mass emission rate of 532 t of copper dur-
ing 1971-74 (Mitchell and McDermott 1975).
While the amount of copper discharged in the ionic
form was not reported, the potential for environ-
mental exposure levels approaching the incipient
LC50 of 33 /x.g/1 copper found for herring embryos in
the present study should be considered in estab-
lishing water pollution control standards.
Frequently authors conducting bioassays using
copper or other heavy metals have not examined
the chemical state of the metal in their bioassay
system. Such characterizations are important
since different chemical forms of metals may have
different toxic effects (Lee 1973). The method out-
lined in this work for examining the particulate
bound fraction, the ionic fraction, and the com-
plexed fraction of metals in seawater provides a
means of examining the important chemical forms
of copper in aquatic bioassay systems. With the
use of appropriate isotopes this method could eas-
ily be applied to other metals. In the case of the
system used to expose Pacific herring embryos and
larvae it appears that the ionic form of copper
predominated. In freshwater the ionic form of cop-
per seems to be the most toxic (Pagenkopf et al.
1974). This is probably also the case for Pacific
herring embryos and larvae exposed to copper in
seawater.
ACKNOWLEDGMENTS
We express our appreciation to Richard E. Tul-
lis, California State University, Hayward, for re-
354
RICE and HARRISON: COPPER SENSITIVITY OF PACIFIC HERRING
viewing the manuscript and for his helpful criti-
cisms and suggestions throughout this study. The
statistical program used was prepared by John
Koshiver, University of California, Lawrence
Livermore Laboratory, and we thank him. We also
acknowledge the helpful comments regarding the
handling of herring embryos and larvae from
Maxwell E. Eldridge, Southwest Fisheries Center
Tiburon Laboratory, National Marine Fisheries
Service, NOAA. We especially thank John
Kriegsman, Larry Schramm, and Bob Fountain of
the California Department of Fish and Game for
their timely information on the location of herring
spawn and for their aid in the collection of these
eggs. This work was performed under the auspices
of the U.S. Energy Research and Development
Administration Contract No. W-7405-ENG-48.
LITERATURE CITED
ADELSTEIN, S. J., AND B. L. VALLEE.
1962. Copper. In C. L. Comar and F. Bronner (editors).
Mineral metabolism, an advanced treatise, Vol. II (B), p.
371-401. Academic Press, N.Y.
AHLSTROM, E. H.
1943. Appendix. In Studies on the Pacific pilchard or
sardine (Sardinops caerulea). 4. — Influence of tempera-
ture on the rate of development of pilchard eggs in nature,
p. 9-12. U.S. Fish Wildl. Serv., Spec. Sci. Rep. 23.
ALDERDICE, D. F., AND F. P. J. VELSEN.
1971. Some effects of salinity and temperature on early
development of Pacific herring ( C lupea pallasi). J. Fish.
Res. Board Can. 28:1545-1562.
American Public Health association.
1976. Standard methods for the examination of water and
wastewater. 14th ed. Am. Public Health Assoc, Am.
Water Works Assoc., Water PoUut. Control Fed., Wash.,
D.C., 1193 p.
Baker, J. T. P.
1969. Histological and electron microscopical observa-
tions on copper poisoning in the winter flounder (Pseudo-
pleuronectes americanus) . J. Fish. Res. Board Can.
26:2785-2793.
Bannister, R. C. a., D. Harding, and S. J. Lockwood.
1974. Larval mortality and subsequent year-class
strength in the plaice ( Pleuronectes platessa L.). In J. H.
S. Blaxter (editor), The early life history offish, p. 21-39.
Springer- Verlag, N.Y.
Becker, C. D., and T. O. Thatcher.
1973. Toxicity of power plants chemicals to aquatic
life. Publ. WASH-1249, U.S. At. Energy Comm., 222 p.
benoit, d. a.
1975. Chronic effects of copper on survival, growth, and
reproduction of the bluegill (Lepomis macro-
chirus). Trans. Am. Fish. Soc. 104:353-358.
Berkson, J.
1953. A statistically precise emd relatively simple method
of estimating the bio-assay with quemtal response, based
on the logistic fimction. J. Am. Stat. Assoc. 48:565-599.
Birge, w. J., AND J. J. Just.
1 973 . Sensitivity of vertebrate embryos to heavy metals as
a criterion of water quality. Ky . Water Resour. Res. Rep.
61, 20 p.
Blaxter, J. H. S., and F. G. T. Holliday.
1963. The behaviour and physiology of herring and other
clupeids. Adv. Mar. Biol. 1:261-393.
Brungs, W. a., J. R. Geckler, and M. Gast
1976. Acute and chronic toxicity of copper to the fathead
minnow in a surface water of variable quality. Water
Res. 10:37-43.
Chester, R., and J. H. Stoner.
1974. The distribution of zinc, nickel, manganese, cad-
mium, copper, and iron in some surface waters from the
world ocean. Mar. Chem. 2:17-32.
Christensen, G. M.
1975. Biochemical effects of methylmercuric chloride,
cadmium chloride, and lead nitrate on embryos and ale-
vins of the brook trout, Salvelinus fontinalis. Toxicol.
Appl. Pharmacol. 32:191-197.
Gushing, D. H.
1975. Maine ecology and fisheries. Camb. Univ. Press,
Camb., Engl., 278 p.
Dethlefsen, v., h. von westernhagen, and H. Rosen-
thal.
1975. Cadmium uptake by marine fish larvae. Helgol.
wiss. Meeresunters. 27:396-407.
Eldridge, M. B., and W. M. Kaill.
1973. San Francisco Bay area's herring resource — a color-
ful past and a controversial future. Mar. Fish. Rev.
35(11):25-31.
Finney, D. J.
1964. Statistical method in biological assay. 2d ed. Hafner,
N.Y., 668 p.
Gardner, G. R., and G. LaRoche.
1973. Copper induced lesions in estuarine teleosts. J.
Fish. Res. Board Can. 30:363-368.
Hart, J. L.
1973. Pacific fishes of Canada. Fish. Res. Board Can.,
Bull. 180, 740 p.
Hazel, C. R., and S. J. meith
1970. Bioassay of king salmon eggs and sac fry in copper
solutions. Calif Fish Game 56:121-124.
Hoss, D. E., L. C. Coston, J. P. Baptist, and D. W. engel.
1975. Effects of temperature, copper and chlorine on fish
during simulated entrainment in power-plant condenser
cooling systems. In Environmental effects of cooling sys-
tems at nuclear power plants, p. 519-527. Int. At.
Energy Agency, Vienna.
HYNES, H. B. N.
1970. The biology of polluted waters. Univ. Toronto
Press, Toronto, Can., 502 p.
Kinrade, J. D., AND J. C. Van Loon.
1974. Solvent extraction for use with flame atomic absorp-
tion spectrometry. Anal. Chem. 46:1894-1898.
Lee, G. F.
1973. Review paper: Chemical aspects of bioassay tech-
niques for establishing water quality criteria. Water
Res. 7:1525-1546.
Litchfield, J. T., Jr , and F. Wilcoxon.
1949. A simplified method of evaluating dose-efTect exper-
iments, J. Pharmacol. Exp. Ther. 96:99-113.
Martin, M., M. D. Stephenson, and J. H. Martin.
1977. Copper toxicity experiments in relation to abalone
deaths observed in a power plant's cooling waters. Calif.
Fish Game 63:95-100.
355
FISHERY BULLETIN: VOL. 76, NO. 2
MAY. R. C.
1974. Larval mortality in marine fishes and the critical
period concept. In J. H. S. Blaxter (editor), The early life
history offish, p. 3-19. Springer- Verlag, N.Y.
MCKIM, J. M., AND D. A, BENOIT.
1 97 1 . Effects of long-term exposures to copper on survival,
growth, and reproduction of brook trout (Salvelinus fon-
tinalis). J. Fish. Res. Board Can. 28:655-662.
MITCHELL, F. K., AND D. J. MCDERMOTT.
1975. Characteristics of municipal wastewater dis-
charges, 1974. South. Calif Coastal Water Res. Proj.,
Annu. Rep. 1975, p. 163-165.
Mount, D. I.
1968. Chronic toxicity of copper to fathead minnows
(Pimephales promelas, Rafinesque). Water Res. 2:215-
223.
MYTELKA, A. I., J. S. CZACHOR, W. B. GUGGINO, AND H. GOLUB.
1973. Heavy metals in wastewater and treatment plant
effluents. J. Water Pollut. Control Fed. 45:1859-1864.
O'REAR, C. W., Jr
1972. The toxicity of zinc and copper to striped bass eggs
and fry with methods for providing confidence limits.
Southeast. Assoc. Game Fish Comm., 26th Annu. Conf ,p.
484.
PAGENKOPF, G. K., R. C. RUSSO, AND R. V. THURSTON.
1974. Effect of complexation on toxicity of copper to
fishes. J. Fish. Res. Board Can. 31:462-465.
PICKERING, Q. H., AND M. H. GAST
1972. Acute and chronic toxicity of cadmium to the fat-
head minnow (Pimephales promelas). J. Fish. Res.
Board Can. 29:1099-1106.
POSTUMA, K. H., AND J. J. ZIJLSTRA.
1974. Larval abundance in relation to stock size, spawning
potential and recruitment in North Sea herring. In J. H.
S. Blaxter (editor), The early life history offish, p. 113-
128. Springer- Verlag, N.Y.
RILEY, J. P., AND D. Taylor.
1968. Chelating resins for the concentration of trace ele-
ments from sea water amd their analytical use in conjunc-
tion with atomic absorption spectrophotometry. Anal.
Chim. Acta. 40:479-485.
Rosenthal, H., and K.-R. Sperling.
1974. Effects of cadmium on development and survival of
herring eggs. In J. H. S. Blaxter (editor). The early hfe
history of fish, p. 383-396. Springer- Verlag, N.Y.
SKIDMORE, J. F.
1965. Resistance to zinc sulphate of the zebrafish
(Brachydanio rerio) at different phases of its life his-
tory. Ann. Appl. Biol. 56:47-53.
Snedecor, G. W., and W. G. Cochran.
1967. Statistical methods. 6th ed. Iowa State Univ. Press,
Ames, 593 p.
SpragUE, J. B.
1969. Measurement of pollutant toxicity to fish. L Bioas-
say methods for acute toxicity. Water Res. 3:793-821.
Struhsaker, J. W., M. B. Eldridge, and T. Echeverria.
1974. Effects of benzene (a water-soluble component of
crude oil) on eggs and larvae of Pacific herring and north-
em anchovy. In F. J. Vemberg and W, B. Vemberg
(editors). Pollution and physiology of marine organisms,
p. 253-284. Academic Press, N.Y.
Vaughan, D. S., and S. B. Saila.
1976. A method for determining mortality rates using the
leslie matrix. Trans. Am. Fish. Soc. 105:380-383.
VOGEL, F. S.
1959. The deposition of exogenous copper under experi-
mental conditions with observations on its neurotoxic and
nephrotoxic properties in relation to Wilson's diseeise. J.
Exp. Med. 110:801-810.
WALDICHUK, M.
1974. Coastal marine pollution and fish. Ocean Manage.
2:1-60.
WEDEMEYER, G.
1968. Uptake and distribution of Zn*^ in the coho salmon
egg iOncorhynchus kisutch). Comp. Biochem. Physiol.
26:271-279.
\
356
ESTIMATED ZOOPLANKTON PRODUCTION AND THEIR
AMMONIA EXCRETION IN THE KUROSHIO AND ADJACENT SEAS
TsuTOMU Ikeda' and Sigeru Motoda^
ABSTRACT
Production and ammonia excretion of zooplankton in the Kuroshio and adjacent seas were estimated
from field data of biomass, size distribution, and habitat temperature of zooplankton, and from
experimental data of respiration and ammonia excretion rates as functions of body size and tempera-
ture. Winberg's basic balanced equations were applied to calculate production from respiration data.
Further, mortality related to the lifespan and the ratio of herbivores to carnivores in the zooplankton
community were estimated from theoretical assumptions.
In this study, 18-72% of primary production was grazed by herbivorous zooplankton, and production
of herbivorous zooplankton ( = secondary production) was 10-60 mg C/m^ per day. The ecological
efficiency between primary and secondary production was 5-22%. Ammonia-nitrogen excreted by
zooplankton was 4-24 mg N/m^ per day, which can support 11-44% of the nitrogen requirements of
primary production.
In marine ecosystems solar energy photosynthet-
ically fixed as oi'ganic matter by phytoplankton is
channelled through zooplankton to nektonic fishes
and crustaceans at higher trophic levels. Impor-
tant features of the roles of zooplankton in this
scheme are their extremely high conversion ef-
ficiency of phytoplankton organic matter (in con-
trast with terrestrial ecosystems, see Wiegert and
Owen 1971; Steele 1974) and the simultaneous
regeneration of nutrients through their excretory
activities. The latter role is considered an impor-
tant mechanism in maintaining constant primary
production levels in the seas, especially in oligo-
trophic areas (Ketchum 1962; Corner and Da vies
1971).
These dynamic functions of zooplankton have
seldom been quantitatively evaluated in the field.
One difficulty lies in the fact that the zooplankton
community includes animals belonging to a vari-
ety of phyla and a number of species which differ
geographically. Information from detailed studies
on one or a few species is not adequate for this
purpose, and collection of all necessary data on
each component species in the community is not
practical. Therefore, the development of some al-
ternative approach is needed to overcome this
problem.
^Australian Institute of Marine Science, P.M.B 3, Townsville,
MSO, Queensland, Australia.
^Marine Science and Technology, Tokai University, Shimizu,
Japan.
Manuscript accepted November 1977.
FISHERY BULLETIN; VOL. 76, NO. 2, 1978.
METHODS
In this study, we treat the zooplankton commun-
ity as an assemblage of different sizes of animals
and use body size-related constant functions for
respiration and ammonia excretion from labora-
tory experiments to estimate feeding, production,
and ammonia regeneration in the Kuroshio and
adjacent seas. A systematic survey of the study
area had been carried out by Japanese parti-
cipants in the CSK (Co-operative Study of the
Kuroshio and adjacent region) organized by
UNESCO during 1965-67 (Motoda et al. 1970; Irie
and Yamazi 1972).
Biomass, Habitat Temperature, and Size
( = Weight) Distribution of Zooplankton
Zooplankton were sampled vertically from 150
m with a NORPAC standard net (mesh aperture,
0.35 mm) in summer (June-October 1965 and
1966) (Figure lA) and winter seasons
(December- April 1965, 1966, and 1967) (Figure
2A). From the average biomass of zooplankton
summarized by Yamazi (1971) for 0-150 m, the
present study area was divided into four density
classes (<10, 10-50, 50-100, and >100 mg wet
weight/m^). The isopleth for 100 mg wet weight/m^
shifted northward in the cold season and south-
ward in the warm season, especially in the east
China Sea (Motoda et al. 1970; Irie and Yamazi
1972). Seasonal difference in the composition of
357
FISHERY BULLETIN: VOL. 76, NO. 2
Figure l. — A. Sampling stations, zooplankton biomass, and isotherms (100-m depth, continuous hnes; 50-m depth, broken lines)
during the warm season (June-October) in the Kuroshio and adjacent seas. B. Distribution of estimated secondary production.
358
IKEDA and MOTODA: ZOOPLANKTON PRODUCTION AND AMMONIA KXCRKTION
,0
ZOOPLANKTON BIOMASS
(mg wet wt/m-*)
>100
50-100
10-50
<10
2S\
0\
SECONDARY
PRODUCTION
(mg C/m^/day)
&
31-60
11-30
\ ca.10
_L
Figure 2. — A. Sampling stations, zooplankton biomass, and isotherms (100-m depth, continuous lines; 50-m depth, broken lines)
during the cold season (December- April) in the Kuroshio and adjacent seas. B. Distribution of estimated secondary production.
359
FISHERY BULLETIN: VOL. 76, NO. 2
zooplankton taxonomic groups among stations
was less pronounced, with copepods dominant
(56-657r of total individual number), followed by
Noctiluca (S-lS'/f ), appendicularians (G-T^'r), and
chaetognaths (4-5'7f) (Yamazi et al. 1972).
Biomass expressed per cubic meter was converted
to per square meter by multiplying by depth of
sampling.
The habitat temperature of zooplankton from
0-150 m was represented by that at 100 m
(Japanese Oceanographic Data Center 1967,
1969). In the east China Sea, which is shallower
than 150 m, the temperature at 50 m was taken as
the habitat temperature (Figures lA, 2A).
From data summarized by Yamazi (1971), the
biomass of zooplankton per haul was divided by
total number of individuals per haul to obtain
average body weight. Values thus obtained at all
sampling stations were grouped into warm or cold
season, and assumed as a general size distribution
in each season (Figure 3). The highest frequency
was observed in the range 0.1-0.2 mg wet weight/
animal in both seasons. Faunal differences south
50
40
30
^20
z 10-
UJ
o
u
a
li. 0
«
I
•7.
30-
hm
ft^
-15 -0 5 0 5
LOG »V BODV WT
s~^ \-
•.'.0
"=30
20-
10-
I I""t
N:U7
•/,
20-
(0-3'/.)
m
flv
•> n n
-15 -05 05
LOG AV BODY WT
-~Ti~>-^
(2 6'/.)
T 1 1 1 1 1 1 1 1
0 0 0501 0 3 0 5 0 7 0 9 11 13 15 17 19
AVERAGE BODY WEIGHT OF A ZOOPLANTER (mg wet wt)
Figure 3. — Relative frequency of average size of zooplankton
(biomass/number of zooplankton at each sampling station) in
warm (June-October) (upper figure) and cold (December- April)
(lower figure) seasons in the Kuroshio and adjacent seas. A
normalized frequency distribution fitted by logarithmic trans-
formation of body weights is superimposed on the right side of
each figure. N is number of sampling stations.
and north of the subarctic boundary (ca. lat. 40°N)
reported by Motoda and Marumo (1965) were ig-
nored here, because no systematic difference was
found in average body size of zooplankton between
these areas. The skewed size distribution was con-
verted to a normal distribution curve by
logarithmic transformation (base 10). Fitness to
the curve was tested primarily by the normal
probability paper (Harding 1949) and finally
confirmed by chi-square test (warm season: x^ =
17.85,df = 6, P<0.01; cold season: x^ = 7.24, df =
6, 0.25
Mean % ovigerous ?
Mean min size ovigerous 9
% 9 >13.8 mm carapace length
7.5 (8,956)
9.7 (6,347
19.4(6,347)
13 8(6,347)
21.1 (6,347)
6.0 (6,951)
7.5 (7,730)
1.1 (7,730)
13.8 (7,730)
2.7 (7,730)
DISCUSSION
Methods for Measuring Growth
The growth rate of crustaceans in nature,
though of considerable research interest, has been
difficult to measure for several reasons. Primarily,
all of the hard parts of the animal are cast off with
the molt, making the marking of them all but
impossible until recent years. Wenner et al. ( 1974)
discussed the problems associated with measuring
growth for crustaceans, and their table 1 sum-
marized possible patterns of growth for the Crus-
tacea. That table stressed the relative contribu-
tion of two factors in crustacean growth: molt
increment and molt frequency. Both of these may
be responsive to different environmental parame-
ters, altering the growth pattern of a species.
The standard methods for measuring field
growth rate for crustaceans (caging, mark and
recapture, and analysis of modal size classes with-
out corroborative data) are unsatisfactory to com-
pare field growth rates for different populations of
E. analoga. Since none of these methods alone
suffices for this kind of comparative measurement
with E. analoga, the instantaneous growth rate
approach was used in this comparative analysis.
The method has qualities common to some of the
other methods mentioned, but avoids some of the
inherent problems of those methods. This
technique allows direct observation of size-specific
molt frequency and molt increment, while
minimizing the handling effects normally as-
sociated with laboratory impoundment. It is likely
that molt frequency estimates by this method are
more accurate for the larger crabs, for which the
5-day holding period is a relatively smaller pro-
portion of the intermolt period. The method has
allowed comparison of growth factors (molt incre-
ment and frequency) in detail {orE. analoga and
gave repeatable data such as that found for Goleta
Bay in 1974 and 1975. Thus a technique for mea-
surement of crustacean growth has been de-
veloped here which may hold promise for such
comparative studies as this, where field caging is
impractical.
Growth of Emerita analoga
The large difference in the growth rate of E.
analoga between beaches of Goleta Bay and a
Santa Cruz Island bay is remarkable in view of
their proximity (about 42 km apart) but not in
view of the different environmental conditions
found at each beach. The combination of colder
water and reduced filterable material in suspen-
sion in the water appears to have slowed the
growth of .E. analoga on Santa Cruz Island. This
difference is evidence of the sensitivity of these
two factors of sand crab growth to variation in
environmental qualities.
It is tempting to construct growth curves from
such data on molt increment and molt frequency,
having arrived at estimates for these. Both of
these factors, however, have been shown to be
highly responsive to environmental conditions. In
fact, they vary widely in time and space with no
clear pattern emerging as yet. A growth curve
constructed from these data would apply only
under a specific set of environmental conditions.
Certainly these large differences in growth rate in
nearby beaches precludes the use of modal size
classes from several beaches for the determination
of growth for£^. analoga, as Efford ( 1967) did ear-
lier.
373
FISHERY BULLETIN: VOL. 76, NO. 2
Efford (1967:84, figure 3) presented a growth
curve forE. analoga, constructed from data taken
from 22 beaches on the Pacific coast between En-
senada, Mexico, and Tofino, Canada (a 2,400 km
distance). Three-fourths of the data presented
were gathered over a period of only 2 mo (between
17 June and 17 August 1961). The remaining data
were collected in 1959 and 1963. Where size-
frequency data were bimodal, the author assumed
that two year classes were present. To construct a
growth curve from his data, Efford also had to
assume that: 1) growth rate was the same year to
year (temporally stable, at least during the grow-
ing season); and 2) longshore migration did not
take place for E. analoga.
Dillery and Knapp (1970) demonstrated that an
average E. analoga individual of about 26 mm
carapace length travels about 15 m/day
alongshore in an easterly direction on local
beaches in Santa Barbara. This implies that the
individuals in inhabiting a particular location
may change from day to day. Barnes and Wenner
(1968) suggested that the interpretation of size-
frequency data is considerably simplified if sex
reversal is assumed for this species, and some di-
rect evidence (Wenner 1972) supports this as-
sumption. However, recent laboratory data (Fu-
saro 1977) suggest that a differential growth rate
for males and females between 9 and 14 mm
carapace length may account for the observed
size-frequency distributions and sex ratio pat-
terns, rather than protrandry.
Combining data from different beaches, as Ef-
ford ( 1967) did, also carries with it the assumption
that the growth rate is relatively the same for the
various parts of the range o^E. analoga (spatially
stable). However, in an analysis of E. talpoida
data presented by Wharton (1942), Efford
suggested that the growth pattern of this latter
species may differ in the southern part of its range.
It is likely that temperature was responsible for
the difference, as it is likely that temperature has
an effect on the growth of £. analoga in different
parts of its Pacific coast range.
Wenner et al. ( 1974) presented data (their figure
5) which suggested that for E. analoga, even dif-
ferent local populations may display different
growth patterns, at least as indicated by size at
sexual maturity. Data presented in this report
imply that molt increment and molt frequency are
indeed different in different environmental re-
gimes in nature. Growth curves constructed for
such different areas would likely differ. To com-
bine these sets of data would be to obscure the real
differences in growth rate observable in such local,
proximate populations.
The instantaneous growth rate estimate,
though, may be used as an index of how well a
population fares under a given set of environmen-
tal conditions. Consider this instance. It has been
shown that a population oiE. analoga on a beach
at Santa Cruz Island grew about one-third as fast
as a population in Groleta Bay (molt increment
depressed by one-third and molt frequency de-
pressed by one-half). Assuming a fixed number of
molts to maturity (e.g., Wenner et al. 1974, table
1), the island population would reach maturity at
a smaller size and in about twice as long a time. In
fact, population structure data (Table 3) shows
that sand crabs from the island population
reached maturity at about the same size as those
from Goleta Bay. If a fixed size at maturity is
assumed, the island population sampled would
take about three times as long to reach that fixed
size.
The third possible assumption, that there is a
fixed length of time to maturity, is argued against
by all available data. In any of these cases, how-
ever, the population of sand crabs inhabiting the
beach of the Santa Cruz Island bay location was at
a distinct disadvantage in terms of reproductive
success when compared with the population oiE.
analoga inhabiting the Goleta Bay beach. This
reproductive disadvantage was brought about at
least in part by the large observed differences in
molt frequency and molt increment at the two
locations. A much smaller percentage of females
were of reproductive size, probably due to the dif-
ferential growth rate (see Table 3).
Cox and Dudley ( 1968) also reported large vari-
ations in the size of the smallest egg bearing
female £^. analoga found in their collections. Data
presented here may account for such previously
problematical observations, in that differences in
growth rate may affect the size distribution and
abundance of mature females.
As these data suggest, then, the growth rate of a
crustacean population plays an important role in
the life history of that species in its particular
environmental situation. Of course, when dealing
with a species which has pelagic larval stages, it
becomes difficult to study local populations under
the assumption that they are genetically different.
Recruitment patterns are not generally well
known for species with pelagic larvae (see Thorson
1950; Efford 1970; Mileikovsky 1971; Strathmann
374
FUSARO: GROWTH RATE OF THE SAND CRAB
1974), thus confounding the issue of reproductive
success for a population in a particular habitat.
Thus "relative reproductive success" may not be as
good a criterion between populations such as these
as it is between species. Measurement of differ-
ences in such life history factors as growth rate
may, therefore, take on added significance in the
comparison of two populations or species in differ-
ing environments, inasmuch as they do not depend
on the assumption of genetic isolation but concern
themselves more with the relationship of the
population to its particular environmental cir-
cumstances.
ACKNOWLEDGMENTS
I thank A. M. Wenner for consultation through-
out the study and for his critical review of the
manuscript, and J. Childress, A. Oaten, and J.
King for their critical review of the manuscript. I
thank C. Stanton of Santa Cruz Island Company
for use of the island site studied, and L. Laughrin
of the University of California Santa Cruz Island
Field Station for the support facilities under his
charge. I also thank P. Kearney for assistance in
typing the manuscript.
LITERATURE CITED
Barnes, N. B., and a. M. Wenner
1968. Seasonal variation in the sand crahEmerita analoga
(Decapoda, Hippidae) in the Santa Barbara area of
CaUfomia. Limnol. Oceanogr. 13:465-475.
Cox, G. W., AND G. H. Dudley
1968. Seasonal pattern of reproduction of the sand crab,
Emerita analoga, in southern California. Ecology
49:746-751.
DILLERY, D. G., AND L. V. KNAPP.
1970. Longshore movements of the sand crab, Emerita
analoga (Decapoda, Hippidae). Crustaceana 18:233-
240.
Drach. p.
1939. Mue et cycle d'intennue chez les crustaces deca-
podes. Ann. Inst. Oceanogr. Monaco 19:103-391.
EKFORD. I. E.
1966. Feeding in the sand crab, Emerita analoga (Stimp-
son) (Decapoda, Anomura). Crustaceana 10:167-182.
1967. Neoteny in sand crabs of the genus Emerita (Ano-
mura, Hippidae). Crustaceana 13:81-93.
1970. Recruitment to sedentary marine populations as
exemplified by the sand crab, Emerita analoga ( Decapoda,
Hippidae). Crustaceana 18:293-308.
FUSARO, C.
1977. Population structure, growth rate and egg produc-
tion of the sand crah, Emerita analoga (Hippidae): a com-
parative analysis. Ph.D. Thesis, Univ. California, Santa
Barbara, 182 p.
Green, J. P., and M. R. Neff
1972. A survey of the fine structure of the integument of
the fiddler crab. Tissue Cell. 4:137-171.
MILEIKOVSKY, S. A.
1971. Tjqies of larval development in marine bottom in-
vertebrates, their distribution and ecological significance:
a re-evaluation. Mar. Biol. (Berl.) 10:193-213.
Neushul, M., W. D. Clarke, and D. W. Brown
1967. Subtidal plant and animal communities of the
Southern California Islands. In R. N. Philbrick (editor),
Proceeding of the Symposium on the biology of the
California Islands, p. 37-57. Santa Barbara Botanical
Garden, Santa Barbara.
Strathmann, R.
1974. The spread of sibling larvae of sedentary marine
invertebrates. Am. Nat. 108:29-44.
THORSON, G.
1950. Reproductive and larval ecology of marine bottom
invertebrates. Biol. Rev. (Camb.) 25:1-45.
WENNER, A. M.
1972. Sex ratio as a function of size in mjirine Crustacea.
Am. Nat. 106:321-350.
Wenner, A. M., C. Fusaro, and A. Oaten.
1974. Size at onset of sexual maturity and growfth rate in
crustacean populations. Can. J. Zool. 52:1095-1106.
Wharton, G. W.
1942. A typical sand beach animal, the mole crah, Emerita
talpoida (Say). In A. S. Pearse, H. T. Humm, and G. W.
Wharton. Ecology of sand beaches at Beaufort, N. C, p.
157-164. Ecol. Monogr. 12.
375
TIME-DEPENDENT SOLUTIONS AND EFFICIENT PARAMETERS
FOR STOCK-PRODUCTION MODELS
R. Ian Fletcher'
ABSTRACT
The time-dependent formulations of the Graham-Schaefer and Pella-Tomlinson systems are re-
structured so as to accommodate directly the critical-point parameters of their respective governing
graphs; the resulting parametric system accounts for the behavior of either model wholly in terms of its
management components. The indeterminate exponent and the coefficients of the Pella-Tomlinson
equations are uncoupled and the dual formulations associated with the conventional casting of the
system are eliminated; the governing equations and corresponding solutions are cast into composite
forms and the sign changes of coefficients become automatic. The previously obscure relationships
between management parameters and variable graph curvature in the Pella-Tomlinson model are
expressly formulated; maximum sustainable yield is shown to be independent of the indeterminacy of
the system. Time-delay estimators for both systems are formulated.
We analyze here, in a deterministic setting, cer-
tain of the transient, nonlinear mechanisms
employed in the modelling of stock and yield dur-
ing periods of imbalance between fishing removals
and stock productivity. The general method of
analysis, which appeals primarily to the direct
parameterization of critical points, will apply to
any nonlinear scheme of exploitation and gross
production, but it applies in particular to the
Graham-Schaefer hypothesis (Graham 1935;
Schaefer 1954) and to the "generalized" model of
Pella and Tomlinson ( 1969). Since control of either
system rests ultimately with the control of critical
points, we restructure the parametric definitions
accordingly and the governing equations for both
systems are then controlled directly by parame-
ters of management significance.
Typically, either system reflects the determinis-
tic premise that a stock of fishes, otherwise held by
exploitation at levels below a prior abundance,
will constantly strive to recover its numbers in
accord with some innate, self-regulating, and re-
peatable mechanism of restoration. Any such res-
toration must accrue from the productivity of the
stock, and by Graham's hypothesis, the inherent
or latent capacity for productivity in a stock of
fishes depends jointly on the current size of the
stock (in numbers or biomass) and the difference
between the current and potentially maximum
'Center for Quantitative Science in Forestry, Fisheries, and
Wildlife, University of Washington, Seattle, WA 98195.
sizes. Whence, in terms of time-dependent
biomass B, and with the proportionality
coefficient defined as the ratio of "intrinsic"
growth rate k and 6^ , Graham's formula for latent
productivity P takes on the familiar form
P(B) = kB
B
'-B\
(1)
Manuscript accepted September 1977.
FISHERY BULLETIN: VOL. 76. NO. 2. 1978.
Of the two expanded terms, the first governs the
intrinsic, exponential capacity for growth of the
population's biomass, while the negative, non-
linear term provides the damping that ultimately
slows growth as B(t) approaches its asymptotic
maximum B^. The two terms, in their algebraic
sum, govern the latent productivity of the stock at
any stock size between zero and JS^^- Parameter k,
as we shall see, is coupled analytically and
phenomenologically to parameter By:, but the de-
pendence of k on root Bx in Equation ( 1) can be
supressed in favor of the direct parameterization
of maximum productivity (which, in the complete
exploitation model, we identify with maximum
yield rate I.
In the Pella-Tomlinson model, the parametric
controls for latent productivity exceed by one the
total number of such parameters in Graham's
formulation, an increase in freedom that comes at
considerable cost to tractability, both analytical
and statistical. The differential equation that gov-
erns latent productivity in the Pella-Tomlinson
system has the indeterminate form
377
PiB) = CfB + C2 5",
(2)
with exponent n the additional parameter, but
with the signs of the coefficients now dependent on
the range of definition of n. As before, the com-
bined terms describe, at any stock size B, the
stock's latent capacity for productivity. With n
undetermined (its determination being a part of
the empirical demonstration), solutions of Equa-
tion ( 2 ) constitute infinitely many growth laws. By
setting n = 2, and with Cj >0, C2<0, Equation (2)
reduces to the Graham equation (Equation (1)).
PellaandTomlinson (1969) attribute Equation (2)
to Richards ( 1959). For a detailed analysis of (2) as
a general growth form, see Fletcher (1975); the
anticedents of this analysis appear there.
In either of the two systems, exploitation enters
the formulation for productivity by the direct dif-
• • •
ference P - Y, with Y signifying the rate of
biomass removal owed to exploitation and P the
latent productivity of the stock. Wherefore, in
writing
B(B) = P(B) - Y(B),
(3)
we interpret B(B) as being the resultant produc-
tivity, at stock size fi, that nets to the stock for its
growth. The net may be positive, negative, or zero
accordingly as P and Y vary with B. That is
P > Y implies B > 0: the stock's latent
productivity exceeds the rate of exploitation; a
positive net productivity remains to the stock
and the stock so tends to a higher level of
biomass.
• • •
P < Y implies B < 0: the rate of biomass re-
moval exceeds the stock's capacity for growth;
the stock adjusts to the deficit in net productiv-
ity by tending to a lower level of total biomass.
• • •
P = Y implies B = 0: the exploitation rate just
balances latent productivity, and biomass
trajectory B(t) exhibits an extremum. Should B
= 0 over finite time, stock biomass remains
stationary and the state called "equilibrium"
prevails.
Although the detailed time course of any real
stock biomass is actually determined by varia-
tions in renewal, survival, member growth, and
the age- or size-dependent probabilities of capture,
such effects are not usually separated in the mod-
els of interest here, and yield rate Y customarily
takes the form
378
FISHERY BULLETIN: VOL. 76, NO. 2.
Y(t) ^F(t)'B(t), (4)
with the implication that all fish of the fishable
stock are presumed to share, in equal measure, the
force of fishing mortality F, irrespective of age or
size. By admitting Equation (4) into Equation (3),
our general form for net productivity becomes
B =P - F'B,
(5)
where the time variation of F is usually prescribed
by average effort f on the assumption that F =
qfiT, quantity q being the individual probability of
capture per unit of effort and r the averaging in-
terval measured in fractions of the dimensional
time unit of F.
ANALYSIS OF
THE GRAHAM SYSTEM
Figure 1 illustrates the phase-plane graph of
Equation (1), the latent productivity of a Graham
stock. Maximum productivity m occurs at stock
sizep. And regardless of the conventions employed
in the formulation of Equation (2), essential
parametric control in the equation resides spe-
cifically with its nonzero root By~ and with coordi-
nate m of the critical point (p, m). Parameter m
and Bx constitute a complete, minimum set of
analytically independent parameters for latent
^cxmtrcHJUxbLe
parcuneter-
m = Pm^
Figure l. — Latent productivity P as a function of stock size B,
the Graham model. See Equations (1) and (la).
FLETCHER: TIME-DEPENDENT SOLUTIONS AND EFFICIENT PARAMETERS
productivity in the Graham system, and they rep-
resent the whole extent of available control over
the graph of Equation (1). Coordinate p of the
critical point has the fixed value B^/2, and the
graph of Equation (1) has a fixed curvature of
second degi'ee.
Wherefore, productivity Equation (1), cast di-
rectly in terms of analytical parameters m and
B^, takes on the form
p=4m[|-]-4m[|-]^ ,la,
and intrinsic rate k, as it turns out, bears a propor-
tionality dependence on maximum productivity
and maximum biomass in the relationship
k =
i>oo L ^maxj
And with the substitution of Equation (la) into
Equation (5), the formula for the net productivity
of a Graham stock becomes
B = 4m
["LJ-^^-ll:]
FB.
(6)
In the integrated, equilibrium versions of the
Graham system, maximum latent productivity m
becomes maximum sustainable yield (MSY),
hence parameter m may be directly interpreted as
MSY in any optimization procedure on Equation
(6).
If we restrict the time-dependence of F to abrupt
changes so that any solution of Equation (6) cor-
responds on its interval of validity, however brief,
to some constant value of F, then the time-
dependence of B in Equation (6) becomes
productivity (Equation (6)) and the biomass solu-
tion (Equation (7)) for cases where
F <
4m
Boo
As indicated in the figure, root fi* becomes the
adjustment level to which biomass trajectory B(t)
will trend when F is less than critical quantity
AmlBy^ (and obviously, Bit) trends to By- in Equa-
tion ( 7 ) when F is zero) . The system is governed by
the positive branch of Equation (6) when Y
0), and by the negative branch of
• • •
Equation (6) when Y > P (in which case, B < 0).
But this partitioning of F into subranges for nega-
tive or positive B is a density-dependent process.
Although we must have F < 4m /B^ foJ* positive
B*, the values of F on that range that drive the
stock either up or down will depend on initial stock
size B,,. To insure, for arbitrary B^, that F < P in
Equation (6), mortality F must have a value such
that
0 < F <
4m
Bno
[ - tl
in which case B(t) increases from initial value B^
towards a higher adjustment level B.;,. But for any
value of F on the interval
4m
Boo
Br
Be
< F <
4m
Boo
then Y > P and B(t) decreases from B„ towards a
lower adjustment level B.: .
Figure 3 illustrates the relationship between
net productivity (Equation (6)) and the biomass
solution (Equation (7)) when
Bit) =
B.
l + Cne-<*'"/^^-^^>'
B.
r FBool
(7)
Be
and with initial time tQ set arbitrarily to zero, the
integration constant in Equation (7) becomes
Co =
B, -B(
b7~
Figure 2 illustrates the relationship between net
F >
4m
Boo
in which case the adjustment level of biomass cor-
responds to the zero root of Equation (6). As indi-
cated by the figure, any mortality F so great as to
equal or exceed the quantity AmlBy., if main-
tained, will fish a Graham stock to extinction.
Since Equation (6) governs the relationship be-
tween transient biomass and nonequilibrium re-
moval, we look to its solution (Equation (7)) for
time delays between equilibria. But the asympto-
tic behavior of Equation (7) is a minor analytical
annoyance to be circumvented here. Let us
379
FISHERY BULLETIN: VOL 76, NO. 2.
6
\
5<0
5
6,
M^c)/
e>>o
e>(t)
A B
Figure 2.— A. Typical phase-plane graph of net productivity B = P - Y, Equation (6). the Graham system, with mortality F
constrained to the interval 0 0 and the positive branch applies. B. Typical solution
graphs of stock biomass B( t) , Equation ( 7 ) . When Y >P, biomass declines from initial value B^ towards adjustment level S* . When Y <
P, biomass increases from initial value S^ towards adjustment level B^.
e>(t)
r>p
A B
Figures.— A. Typical graph of net productivity B =P -Y, Equation(6), the Graham system, with mortality F ^AmlBy,. Foranysuch
value of F, the zero root of Equation (6) applies, removal rate Y exceeds latent productivity P, and B <0. B. Typical solution trajectory
B(t) of Equation (7) when F & Am!B^. Biomass declines from initial value B,, towards extinction level B = 0.
380
FLETCHER: TIME-DEPENDENT SOLUTIONS AND EFFICIENT PARAMETERS
presume that no practical technique of estimation
will have a precision of resolution better than
some assignable percentage of true stock size, and
let us reflect that practical uncertainty in our
analysis by expanding the asymptotic bound of
Equation (7) to a region of radius €'B::, around the
analytical value of the bound ( e being the measure
of the uncertainty). Whence, with B,j and Sj now
signifying initial and adjustment levels, and by
supposing thatF changes abruptly at time /„ from
value Fq to some new value F^, Equation (7) be-
comes
(1± 6) =
l + Cne~^^'"/^«-'''i^'
the plus sign applying when Fj >F„ and the minus
sign whenFj 0.
F = 2m/Bx; stock size B(t) implies p (p being
the biomass level Bx/2 where maximum latent
productivity occurs; Figure 1), which implies
that Y-^'m. Accordingly, we may identify
parameter m, in any of the rate equations here,
with MSY (which, we should remember, is it-
self a yield rate).
Since, by Equation (4), instantaneous removal
varies in time as Y(t) = Fit)B(t), then over the
course of the adjustment interval that follows an
abrupt change in F, yield from a Graham stock
will accumulate as
381
FISHERY BULLETIN: VOL. 76. NO. 2.
7(0 = (B^-B^) In
B* =
B
i4mlB^-F
..-1)
Boo
FB^
Am
(9)
m
^ 6(1 -n) U1i/(i-">
the plus sign applying to Equation (10) and the
minus sign to Equation (11).
ANALYSIS OF
THE PELLA-TOMLINSON SYSTEM
As noted in the foregoing section, the maximum
latent productivity m of a Graham stock always
occurs at a biomass value exactly one-half the
unexploited maximum 5^. In turn, MSY of the
equilibrium model must also occur at the stock
level B^I2. So as to gain control over the locations
of those extrema, Pella and Tomlinson ( 1969) mod-
ify the Graham system by writing^the differential
equation for latent productivity P essentially in
the form of Equation (2), which, by the customary
treatment, has a troublesome, dual formulation
owing to the sign changes at /? =1 of coefficients
c, ,c.^. On the interval Q 1
latent productivity takes on the basic form
P = bB
aB"
(11)
(where c\ = b, c.2 = -a, with a and b positive). In
either case, the bound B-^, the maximum produc-
tivity m, and the ordinate p (which governs the
biomass level where m occurs), all depend on the
numerical value assigned to exponent n. That is,
root By is given by
Boo ~
l/(l-n)
the ordinate p is determined by
P =
an
b
i/(i-M)
while maximum productivity m, by the conven-
tional casting of the model, must be determined
from the formula
n > 1
6«/^ n= 1
P(5)
m
Figure 4.— Typical graph of Equation (12), latent productivity
P as a function of stock size B, the Pella-Tomlinson system.
Coordinate p. in its location with respect to root
By., directly reflects the value assigned to expo-
nent n, as indicated by Figure 4. When n takes any
value between zero and unity, coordinate p falls on
the range between zero and B-^/e ( ~ 0.3679 B^), in
which case Equation (10) applies. When n takes
any value greater than unity, coordinate p falls on
the range between Byle and By, in which case
Equation (11) applies. But the coordinate m has no
essential dependence on exponent n, and its ap-
parent coupling with n ( as indicated by the formu-
lation above) is merely an inconvenient artifact of
the conventional analysis. With parameters m
and n uncoupled (see Fletcher 1975), the Equa-
tions (10) and (11) that govern latent productivity
in the Pella-Tomlinson system can be consolidated
into the single governing equation
P = ym
[£] --[£]"■ <-
with y a purely numerical factor wholly prescribed
by n as
382
FLETCHER: TIME-DEPENDENT SOLUTIONS AND EFFICIENT PARAMETERS
7 =
n
n-1
(13)
Be
With the coefficients so cast, the sign reversals at
turning point n ^ 1 become automatic. In con-
sequence, the consolidated interval of definition
for n becomes 0 < n < x (the point n ^ 1 being a
removable singularity). With parameter m thus
separated from n in Equation (12), the undeter-
mined exponent n can be defined solely by the
fraction pBy. in the relationship
Br
= n"^'
■n)
(14)
Consolidated Equation ( 12) now takes on the role
in the Pella-Tomlinson system that Equation ( la)
takes on in the Graham system. In fact, when n =
2, Equation (12) reducesto Equation (la), in which
case y — A and p/B^ = V2. As an interesting aside
here, we note that Equation (12), at the turning
point n = I, takes on the form
P =
e m
[Boo] [Poo_
(e being Napier's constant), while ratio ( 14), in the
limit as n-^l, has the value
In fact. Fox (1970) constructed a stock-production
model around this special case, but since the ratio
p/Bx has the fixed value 1/e, Fox's model "has as
rigid a form as the Graham model" (Ricker 1975:
331).
Quantities m, p, and B^ constitute a complete,
minimum set of independent parameters for la-
tent productivity in the Pella-Tomlinson system.
Collectively they control the behavior of govern-
ing Equation (12), but the influence of any one
parameter remains independent of the remaining
two. Figure 5 illustrates their separate effects on
the graph of Equation (12).
By appealing to the same piecewise constraints
that enter the Graham productivity equations, we
substitute Equation (12) into the general produc-
tivity formula (Equation (5)) and net productivity
in the Pella-Tomlinson system becomes
B = ym
[boo]
ym
B
Be
FB. (15)
And over any time interval, however brief, that
mortality F might be presumed to have a fixed
value, biomass variable B in Equation ( 15) has the
general time-dependent solution
B^: unexploited stock
level [the nonzero
root of Equation ( 12)].
p: biomass level for
maximum productivity
[the coordinate of S in Equa-
tion (12) where m occurs].
m: maximum productivity
[the extremum coordinate
fmax i"^ Equation (12)].
Figure 5. — The graph of Equation ( 12), latent productivity in the Pella-Tomlinson system, as controlled by independent parameters
m, p. and By..
383
FISHERY BULLETIN: VOL. 76. NO. 2,
B{t)
5,^-"+Cexp ((ymlBoo-F)il-n)?j
lia-n)
(16)
B* =
[ym-FB^j
l/d-M)
B.
By setting initial time ^^ arbitrarily at zero, the
integration constant C in Equation (16) becomes
C = B,
l-n
B.
Biomass Equation (16) will apply immediately
upon a change in F and remain valid thereafter for
the time that F remains constant. Over such time,
population biomass will trend up or down in accord
with Equation (16) from initial size B^ towards
adjustment level B^. Should nonzero root B^. be
negative (which is possible only when n > 1), then
the adjustment level corresponds to the zero root of
Equation (15) and the population tends to extinc-
tion by Equation (16).
The critical relationships between fishing mor-
tality, productivity, and time-dependent yield rate
in the Pella-Tomlinson system are considerably
more complex than the relationships between F,
P, and Y in the Graham system. Figure 6 illus-
trates the behavior of P - Y when n < 1, and
Figures 7 and 8 illustrate P -Y when n > 1. The
ratio ymlB-j- becomes the critical quantity in the
Pella-Tomlinson system [AmlB-j. being its coun-
terpart in the Graham system).
As indicated by Figure 6, the biomass level p
where maximum productivity occurs must lie on
the range 0 P. And
those values of F, for which Sff,* either increases or
decreases to B.,.. , depend on the critical ratio ym/B^c
and initial biomass value B^. To insure, for arbi-
b<0
5
Tttnzshoid &co/&
.T"^ —
/ fZorujc of p
/ when
1 &
5>0
e>it)
F >
Ym
1-
itnplies Y>Py E>o^£>ifc
ddfustmant
n—i
levcJ
1-
ao"-'
B, "-'
'«»
impUcs Y0^
B
Figure 6.— a. Typical phase-plane graph of net productivity Equation ( 15), the Pella-Tomlinson system, for values ofn where 0 0 and the
positive branch applies. B. Typical solution graphs of stock biomass Br ^J, Equation (16), when 0 < n < 1. Should Y >P, biomass
declines from initial value B^ towards adjustment level B+
384
But when Y < P, biomass increases from B^ towards B^, .
FLETCHER: TIME-DEPENDENT SOLUTIONS AND EFFICIENT PARAMETERS
trary fig, that Y 0 and the positive branch of
Equation (15) applies. Trajectory B(t) then in-
creases, in accord with Equation ( 16), from initial
value 5p towards a higher adjustment level B...
But for any value of F such that
F >
ym
Ban
B,
n-l -I
Be
n-l
then y > P and the negative branch of Equation
(15) applies; trajectory B(t) decreases from B^ to-
wards a lower adjustment level B^.
Although the sign of B and the consequential
course of B(t) is a density-dependent process for
given F, we should note here that when
n
1
ym
Ban
(17)
then B(t)->'Y and -► m, irrespective of initial con-
ditions. Accordingly, we may identify parameter
m with MSY in any of the (reformulated) rate
equations of the system.
As indicated by Figures 7 and 8, the biomass
level p where m occurs must lie on the range By^/e
1. And with n so prescribed,
root B.., of Equation (15) may have positive or
negative values accordingly as F has a value less
or greater than the critical ratio ymlB-x_. Figure 7
illustrates the behavior of Equations (15) and ( 16)
for the constraints
n > 1
0 < F <
ym
Boo
0 < B* < Boo,
in which case, root Bi.. of Equation (15) becomes
the adjustment level such that B(t}-^B:i, by
Equation (16). But whether Bf^ trends up or down
to B:^ depends on the further partitioning of F with
respect to initial biomass value B^. To insure, for
arbitrary B„, that Y
P, B < 0, and the negative branch of
Equation (15) applies; trajectory B(t) decreases
from By towards a lower adjustment level B^. as
indicated by the upper curve of Figure 7b.
Should mortality F equal or exceed the critical
ratio ym/By. in a Pella-Tomlinson system where n
exceeds unity, the corresponding stock, over
sufficient time, will trend to extinction. Figure 8
illustrates the behavior of Equations ( 15) and (16)
for the constraints
n > 1
ym
F >
Br
B. < 0,
in which case the zero root of Equation (15)
applies, and we have B < 0 and B(t) -►O, irrespec-
tive of initial conditions.
By expanding the asymptotic bound of Equation
(16) to a region of radius e'Bn:, and by appealing to
arguments similar to those that led to the delay
estimate (Equation (8)) of the Graham system, we
calculate from Equation ( 16) the transition times
for a Pella-Tomlinson stock as being
Be
4ag
(l-n)(7m-FiBoo)
In
"l-(l±6)^-" "I
_l-(Bo/Bi)i-"J
(18)
where e represents the imprecision of stock-
abundance estimates, and where B^, andBj signify
initial and adjustment levels as they correspond to
mortality values F„ and Fj . Again we suppose that
F changes abruptly at zero reference time from
value F„ to the new value Fj, the plus sign of
Equation (18) applying when Fj > F^ and the
minus sign when Fj < F^.
By Equation (4) and the assumption that F var-
ies in time by taking on fixed values of finite dura-
tion, we can write the transient yield rate for the
Pella-Tomlinson system in the consolidated form
385
FISHERY BULLETIN: VOL. 76, NO. 2.
5
\ n>i
I
I
Thmshxyldipooje.
5
5(t;
to
Ym
&,
-/-
an— I
o
rt—t
< F <
"67
implies V >P, £>o^&:ic
implLes Y\ and F0. Should Y>P, the
negative branch of Equation ( 15) applies; should Y\ and FP, biomass trajectory B(t) declines from initial value Bq toward adjustment level B*. Should
,Yi
I 1
/
b(t)
Ex-iinction
A B
Figures.— A. Phase-plane graph of net productivity Equation (15) when « > 1 andF ^ymlBy-. Forany such combination of n andF,
S* < 0 and the zero root of Equation ( 15) applies. B. Typical solution trajectory. Equation ( 16), when « > 1 and F > ymlB^, in which
case the stock declines from initial value B„ towards extinction.
386
FLETCHER: TIME-DEPENDENT SOLUTIONS AND EFFICIENT PARAMETERS
.1-
Y{t) = FB,
1 -(l-(Bo/B*)'"7 exp ((7m/Boo-F)(l-n)/)y^^''"^
(19)
which is valid for all values of n save n ^ 1. Owing
to the range of definition on exponent 1/1 -n, I
have not found a closed form for the general time
integral of Equation (19) (although existence is
fairly easy to show for n positive and either less or
greater than unity). But the usefulness of the
analysis does not suffer too greatly for that omis-
sion, since one may accommodate Equation ( 19) to
a numerical equation solver for finite measures of
yield 8Y on associated intervals 8t.
When F changes abruptly (as we have assumed
throughout), yield rate Y changes abruptly, but
the ensuing trends of adjustment are governed, in
the Pella-Tomlinson system, by the following rela-
tionships:
0 1:
F < ym/By,: stock size B(t)->B^ (Figure 7),
which implies that Y-^FB^..
F^ymlBy.; stock s\zeB(t) -►O (Figure 8), which
implies that Y-^O.
n > 0 (both ranges):
F = (\-lln)ym!B~f.\ stock size Bit) ->p, which
implies that Y -^■m (and we may identify
maximum latent productivity m with
maximum yield rate in any of the time-
dependent formulations of the analysis).
The quantity ym/By-, which plays such a promi-
nent role in the analysis, can be identified as the
"intrinsic growth rate" of the stock whenever ex-
ponent n > 1, in direct analogy to quantity k of the
Graham system (and, in fact, with n = 2, then y =
4 and Am/By. = k). But as a consequence of the
indeterminate power form of the Pella-Tomlinson
system and the switching of coefficient signs in the
governing equations, the intrinsic growth rate
turns out to be density-dependent when n takes on
values between zero and unity. That is, by Equa-
tion ( 12), the intrinsic rate (if we may call it so) has
the form
_ JUL 5"-i
Boo
when n falls on the interval 0 < n < 1 (in which
case, 7 < 0).
DISCUSSION
Any nonlinear stock-production system may be
restructured along the lines of the critical-point
analysis described in the foregoing sections; such a
treatment will generate parametric variables
most likely to be those essential to management
analysis. A synopsis of the parameters that appear
in the restructured Graham and Pella-Tomlinson
systems is given by Table 1.
Table l. — Parameters of the restructured Graham and Pella-
Tomlinson systems as they apply to management components.
Control
parameters
Graham
Pella-Tomlinson
Management components
system
system
Maximum stock size
s^
Boo
Maximum productivity
(corresponds to MSY)
m
m
Stock size for maximum
productivity (the optimum"
stock size)
e^/2 (fixed)
P
Ratio p/e^
V2(fixed)
n1/(1-n)
Fistiing mortality
F
F
General ad|ustment level
(consult text for mortality
conditions)
e.. or 0
S., or 0
Fisfiing mortality for
adjustment level p (the
"optimum" F)
2/77/e^
(1 - 1/n) ymiBy,
Graph curvature
fixed
n
For optimization procedures on the Graham sys-
tem, the essential parameters are {F, m, By-}
augmented by the auxiliary parameters B^ and
B;;:. For the Pella-Tomlinson system we may
choose the combination {F, m.p.B,^} or the combi-
nation {F, m, n,Bx}, either of which constitutes an
essential set of mutually independent parameters.
In the first set, p and By. determine n; in the
second, n and S^ determine p. The relationships
in either case are governed by Equation (14).
Although the parametric influence of n is
wholly prescribed by the ratio p/Byr, exponent n
also determines the curvature of all graphs of the
Pella-Tomlinson system. Therefore, when the par-
ticularization of the system depends primarily on
general curve fitting, the likelihood always exists
that ill-determination of parameters will follow,
387
owing to stochastic displacement of datum points
at biomass levels remote from locations p and B-^.
As revealed by Equation (14), exponent n is quite
unstable to small perturbations in the ratio p/B-^.
The variational response in n exceeds the pertur-
bation in p/B^ by an order of magnitude near n =
1, and the instability increases as p/fi^-^1. But the
location of p with respect to fix is far more critical
to management analysis than graph curvature
and its associated "good fit," since, to the left of p,
the stock produces biomass at a positively acceler-
ated rate, while to the right of p productivity de-
celerates.
The trait of degeneracy in the system has been
noted by Pella and Tomlinson (1969) and by Fox
(1971, 1975), but the exact relationships between
exponent n and the quantities m, p, and S^ have
been obscured heretofore by the conventional
castings of the system. With the restructured gov-
erning equations and the explicit formulations of
critical parameters, much of the statistical degen-
eracy associated with previous routines can be
constrained. And since the management parame-
ters appear directly in the equations of the system,
their variances can be calculated directly in the
estimation procedure and appeals to indirect
methods are avoided.
FISHERY BULLETIN: VOL. 76, NO. 2.
LITERATURE CITED
Fletcher, R. I.
1975. A general solution for the complete Richards func-
tion. Math. Biosci. 27:349-360.
FOX, W. W., JR.
1970. An exponential surplus-yield model for optimizing
exploited fish populations. Trans. Am. Fish. Soc. 99:80-
88.
1971. Random variability and parameter estimation for
the generalized production model. Fish. Bull., U.S.
69:569-580.
1975. Fitting the generalized stock production model by
least-squares and equilibrium approximation. Fish.
Bull., U.S. 73:23-37.
Graham, M.
1935. Modern theory of exploiting a fishery, and applica-
tion to North Sea trawling. J. Cons. 10:264-274.
PELLA, J. J., AND P. K. TOMLINSON.
1969. A generalized stock production model. Inter-Am.
Trop. Tuna Comm., Bull. 13:419-496.
RICHARDS, F. J.
1959. A flexible growth function for empirical use. J.
Exp. Bot. 10:290-300.
RICKER, W. E.
. 1975. Computation and interpretation of biological statis-
tics offish populations. Fish. Res. Board Can., Bull. 191,
382 p.
SCHAEFER, M. B.
1954. Some aspects of the dynamics of populations impor-
tant to the management of the commercial marine
fisheries. Inter-Am. Trop. Tuna Comm., Bull. 1:25-56.
388
TROPHIC RELATIONSHIPS IN JUVENILES OF THREE SPECIES OF
SPARID FISHES IN THE SOUTH AFRICAN MARINE LITTORAL
M. S. Christensen*
ABSTRACT
The feeding habits of three sparids, Diplodus sargus, D. cervinus, and Sarpa salpa, were studied.
Juveniles of these fishes occur commonly in the intertidal and immediately subtidal regions of
southeast Africa, while adults were only observed in these zones at high tide. Small juvenile D. sargus
feed largely on harpacticoid copepods, amphipods, and, in spring and early summer, cirripede nauplii,
chironomid larvae, and an unidentified trochophore larva. Larger individuals mainly take amphipods
and green algae. Successive size classes of D. cervinus feed mainly on harpacticoid copepods and
chironomid larvae, the shrimp Palaemon pacificus, amphipods, and then polychaetes. Sarpa salpa
ingest harpacticoids when small, diatoms and red algae as a large juvenile, and red and green algae as
an adult. Corresponding changes in gut length and dentition are reported for S. salpa.
Marked ecological separation of the three species was observed. Small juveniles appear at different
times of the year and feed on different foods (dietary and temporal separation). Larger juveniles and
subadults have different diets or feed in separate parts of the littoral zone (behavioral, dietary, and
spatial separation).
A brief review of methods of analyzing stomach contents is included and it is suggested that a
combination of points and ranking indices would be the most valuable. The method, here termed the
comparative feeding index, is described.
The food and feeding relationships of fishes in the
intertidal zone of South Africa are poorly
documented and the results are largely qualita-
tive. The most important of these studies deal with
one or two species of the families Gobiidae (Pitt-
Kennedy^), Sparidae (Hutchings^), Cheilodac-
tylidae (Butler^), and Gobiesocidae (Stobbs^).
Three species of sparids were investigated in the
present survey, Sarpa salpa (Linnaeus 1758), Di-
plodus sargus (Linnaeus 1758), and D. cervinus
(Lowe 1838). Barnard (1927), Smith (1965), and
Hutchings (see footnote 3) described S. salpa
(strepie) as being primarily a herbivore, whereas
Talbot ( 1954) found the fish to be omnivorous. The
'J. L. B. Smith Institute of Ichthyology, Rhodes University,
P.O. Box 94, Grahamstown, 6140, South Africa.
*Pitt-Kennedy, S. 1968. A preliminary investigation of feeding
in two gobies Coryphopterous caffer (Giinther) and C. nudiceps
(C. and V.) with not«s on their sexual maturity. Unpubl. honors
proj., 39 p. Zool. Dep., Univ. Cape Town, S. Afr.
^Hutchings, L. 1968. A preliminary investigation into the
diets of two littoral teleosts, Sarpa salpa (Linnaeus) and
Pachymetopon blochii (Valenciennes), with notes on their biol-
ogy. Unpubl. honors proj., 25 p. Zool. Dep., Univ. Cape Tovm, S.
Afr.
■•Butler, G. S. 1975. An investigation into the biology of two
inter and infratidal species of Cheilodactylidae (Pisces: Teleo-
stei). Unpubl. honors proj., 29 p. Zool. Dep., Rhodes Univ.,
Graham-stown, S. Afr.
'Stobbs, R. E. In preparation. Preliminary investigations into
the feeding behaviour and food preferences of Chorisochismus
dentex (Pisces: Gobiesocidae).
Manuscript accepted August 1977.
FISHERY BULLETIN: VOL. 76, NO 2, 1978.
latter study was made in an estuary, however,
where algae are generally less abundant than in
the intertidal region. Diplodus sargus (blacktail)
is described as being an omnivore (Biden 1954;
Talbot 1954; Smith 1965), as isD. cervinus (zebra).
Little, however, has been published on the food
and feeding habits of the juveniles of these three
species, although they are abundant in the inter-
tidal and immediately subtidal regions of this
coast.
The objectives of this study have, therefore,
been to determine: 1) the diet of juveniles of these
three species; 2) how feeding changes with age and
season; 3) the degree of overlap between species,
possibly resulting in competition; 4) recruitment
times and approximate growth rates of the fish;
and 5) the relationship between dentition, gross
gut morphology, and diet.
During the course of this study, existing
methods of fish feeding analysis were found to be
inadequate and a new technique is described
which overcomes some of the problems.
MATERIALS AND METHODS
Fish were collected from February to December
1975 at 2-wk intervals during spring tide, in spite
of the possibility of introducing biases, as diving
389
FISHERY BULLETIN: VOL. 76, NO. 2
conditions were most suitable at this time. Hand
nets were used in the intertidal pools and multi-
prong spear guns in the subtidal area. Hook and
line, poison, traps, and gill nets were not used as
further biases may be induced to the feeding data
(Randall 1967). Fish collected with hand nets were
immediately placed in a 10% Formalin^ solution,
whereas this procedure was delayed for up to iy2 h
in the case of those taken by spear. It was con-
cluded that death stops or greatly slows digestion,
as the stomach contents were found to be in an
equally digested state in both groups on later
analysis. This has also been observed by Hobson
(1974).
The fish were left in Formalin for 10 to 14 days.
This time period was maintained throughout to
standardize any length and weight changes in-
duced by the fixative (up to 5%, Royce 1972).
About 10 scales were removed from under the pec-
toral fin and cleaned with a camel hair brush after
having been soaked overnight in water with a
trace of carbolic acid (Pinkas 1966). They were
mounted dry and examined over a white
background using a low-power binocular micro-
scope. Standard lengths to the nearest millimeter
were taken and the stomach removed and placed
in 45% n-propyl alcohol.
The stomach is here defined as that part of the
gut between the last gill arch and the gut caecae.
The intestines were not examined as some food
items are more resistant to digestion than others,
with resultant biases as one moves along the gut
(Randall 1967; Kionka and Windell 1972; Gannon
1976). Food items were identified to species where
possible.
Numerous methods have been employed in
analyzing the food habits of fishes and volumetric
and gravimetric techniques are being more widely
used today with the current trend towards greater
accuracy (Windell 1971). Both suffer from the
same limitation in that digestion of the food both
reduces its volume and weight. This has resulted
in the use of reconstructed weights and volumes
where the live weight and/or volume is back-
calculated from a measureable parameter, e.g.,
carapace length (M. Bruton, pers. commun.). In
this particular study some of the fish had fed on
diatoms and it is not feasible to determine the
volume of such small items (Windell 1971). Simi-
larly, the reconstructed weight could not be de-
termined as a sample of monospecific, uncontami-
nated diatoms is impracticable to obtain and
contains an indeterminate number of dead frus-
tules which varies from sample to sample (Round
1971).
In such cases, the points (Swynnerton and
Worthington 1940) and ranking index methods
(Hobson and Chess 1973) would appear to be more
satisfactory and were initially used in the present
study. The points system was modified by Frost
(1943) and subsequently by Hynes (1950) to take
into account gut fullness, 30 points being allotted
when the stomach was distended, 20 when full, 10
when half full, and so on. One, two, four, eight, or
sixteen points were assigned to each food item
rather than fractions of the total allotted to each
stomach in proportion to their volumes. This is an
artificial situation and the method was revised as
described below.
After removal, the stomach is allotted between 0
and 30 points in proportion to its fullness. This is
very subjective but overcome to some extent when
large numbers of guts are handled. The contents
are then sorted, identified, and the percentage
volume estimated for each food item with the aid of
Data Sheet No. 6 of Geotimes.'' All estimations are
made with the organisms spread out to an even
depth throughout the microscope field or equiva-
lent surface. The total number of points allocated
to that stomach is then subdivided amongst the
food items in proportion to their percentage vol-
umes. The points gained by each food item are
summed for the total sample offish and the mean
calculated. The values are then scaled down to a
percentage to give the dietary composition of the
fish examined. In the case of the ranking index
(RI), the volume is estimated as above and the
mean calculated for each food item per fish. The
mean volume is then multiplied by the ratio of the
number of fish containing that item to the total
sampled.
The points method, however, places too much
weight on single food items that have been fed on
to distension by a few fish, whereas the RI method
fails to consider stomach fullness. It is therefore
suggested that an alternative, here termed the
comparative feeding index (CFI), would be more
suitable as it takes into account all three factors,
i.e., the volume, fullness, and frequency of occur-
rence of each food item. The method involves the
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
390
'Available from the American Geological Institute, 2101 Con-
stitution Avenue, N.W., Washing^n, D.C.
CHRISTENSEN; TROPHIC RELATIONSHIPS OF SPARID FISHES
allotment of points to each food organism as de-
scribed above and the mean value per fish is then
multiplied by the percentage of the total sample of
fish that contain that item. As can be seen, the CFI
combines the properties of both points and RI
methods and thus reduces to some extent the effect
of the problems discussed.
The diet of these fish was also determined by the
occurrence method (Hynes 1950) as this indicates
the feeding preferences rather than the food's vol-
umetric value. This is determined as the percent-
age of fish in the sample analysed which contain
that particular food item.
The dietary composition of D. sargus was
analysed for the winter (February- July) and
summer (August-December) periods, as it was
found to be seasonal. This was not done for the
other two species, as feeding seasonality is
synonymous in these fish with the change in diet
with age, as they exhibit discontinuous recruit-
ment.
All skeletal material was cleared and stained
using the trypsin maceration, alizarin stain
method of Taylor ( 1967). The gut was dissected out
in the same specimens, drawn and measured, and
the gut length to standard length ratio (G/S) was
calculated as in Weatherly (1972).
STUDY AREA
The study area is situated about 3.2 km north of
Kleinemonde in the eastern Cape and is known
locally as Clayton's Rocks (Figure 1). The
shoreline consists of a gently shelving, sandy
beach with broken rocky areas of varying extent.
33* 15'
'
1 CvT 1
0
Srahomstown
\Great Fis^ River
33"X'
V\
KleinemondevV'-i;^
Port Alfredjk'-'''''"''^
Clayton's Rocks
t SOUTH Z'
, ,„^
r^ *
10 km
Chart Area
X?\<
J
26 30
2645
27 00
27 15
Figure l. — The study area in the Eastern Cape Province, South
Africa. Adapted from Topographical Chart 3326, Grahamstown.
The smaller rocky outcrops are continually cov-
ered and uncovered by sand, as the beach is un-
stable and backed by large, shifting sand dunes
which move at right angles along the coast. The
rocky area under study is made up of sandstone
which strikes east-west and dips steeply south-
wards. This has resulted in the development of
gullies and pools partially sheltered from wave
action by ridges of resistant rock (Figure 2). The
maximum collection depth at the seaward edge of
the gullies was 3 m at low tide.
Environmental conditions vary greatly and
salinities may fall to 25%o at low tide, which is
caused by freshwater seepage into the pools from
springs in the beach. During the day, at low tide,
surface water temperatures have been recorded
ranging from 26° (summer) to 15°C (winter) in the
intertidal and from 22° (summer) to 14°C (winter)
in the open sea.
The other major fish species coexisting in the
study area are listed below with their general
biological characteristics, where known.
ARIIDAE
T achy sums feliceps: occurs singly, as a juvenile,
in crevices.
SPARIDAE
Lithognathus lithognathus: occurs in small
groups of juveniles.
Rhabdosargus holubi: juveniles, in small
groups.
Sparodon durbanensis: juveniles, observed
from October to March, either singly or in
small groups.
CHEILODACTYLIDAE
Chirodactylus brachydactylus: as juveniles and
subadults, singly, mainly from June to
November.
MUGILIDAE
Unidentified species: occur all year round as
juveniles.
CLINIDAE
Clinus cottoides: a purely intertidal species,
lives in weed, juveniles appeared about
June/July.
Clinus superciliosus: lives in weed, juveniles
only observed in November/December.
GOBIIDAE
Caffrogobius caffer: intertidal species, juveniles
seen from June to November.
TETRAODONTIDAE
Amblyrhyncotes honckenii: singly or in small
groups.
391
FISHERY BULLETIN: VOL. 76, NO. 2
^.■^-'^.
Figure 2.— The study site at Clayton's Rocks.
RECRUITMENT AND GROWTH
Di plod us sargus
The monthly and total length-frequency dis-
tribution is given in Figure 3. The lumped sample
show^s a mode in the 10- to 20-mm size class, indi-
cating that larger fish tend to emigrate to deeper
water. The juveniles appear in the littoral zone
when between 9 and 10 mm standard length (SL),
and leave when about 90 mm long. It appears that
large fish utilize the intertidal area at high tide as
two fish of 107 and 108 mm SL were collected in
the intertidal area some 2 h after low tide and
another 164 mm long was collected 3 h after low
tide.
Visibility was <15 cm in the pools of the re-
search area during September, October, and De-
cember due to flooding of the Fish River. No dives
could be made during this period in the subtidal
area with the result that fish >40 mm were not
collected.
Recruitment of the juveniles into the littoral
appeared to be relatively constant as no monthly
peaks of abundance were found during the survey.
This tends to confirm Biden's (1954) suggestion
that females of this species spawn throughout the
year, though mainly in summer.
No monthly modes could be followed over the
period of study as it is a continuously recruiting
Standard Length (mm)
Figure 3. — Monthly and total length- frequency distribution of
Diplodus sargus, showing three age-classes (O-t-, 1-I-, and 2 + ).
392
CHRISTENSEN: TROPHIC RELATIONSHIPS OF SPARID FISHES
population and so growth rates were not esti-
mated. The age of three immature specimens was
determined as being 1+ years old (107 and 108
mm SL) and 2+ years old (164 mm SL).
Diplodus cervinus
The length-frequency distribution is given in
Figure 4. The total pooled sample shows that this
species first appears in the littoral zone when
about 8 mm long and leaves again when about 140
mm long. Large fish were observed to move into
the intertidal area after low tide, as was the case
with D. Sargus, but no specimens were obtained.
Monthly modes were observed, which are age-
classes 0-^, 1 + , and 2+ (Figure 4). Recruitment of
juveniles into the tide pools is discontinuous, oc-
curring between August and November, with a
peak in October. Two fish sampled in August ( 132
and 129 mm SL) had just formed the second ring,
giving the approximate time of scale-ring forma-
tion. The estimated average growth rate as deter-
mined for mode 1+ is 45 mm from February to
December, which is about 54 mm/yr.
10
5
0
3
5
3
Total
n=67
2* ^n fi
y^^
Mar- .n
n=17
A^
Apr.
n = 5
May
n=1
^'*»^'''^^^^
Jun.
n = 3
Jul.
Jl=ji
.-^ iTs^A A
Aug
n = 8
Sep
n=0
Oct. „
n=7
Oh
^/ilA
Nov.
n=12
,^> <^^.
Dec.
n = 5
50 100
Standard Length (mm)
150
Sarpa salpa
The length-frequency distribution shows that
the majority of the population is from 9 to 45 mm
SL (Figure 5). The juveniles appear in the intertid-
al when 2=9 mm, and fish >100 mm were never
observed in the littoral at low tide; two specimens
of age-class 2 -I- were collected 4 h after low tide.
Three age-classes were observed, labelled as 0 + ,
1 + , and 2 + . Recruitment of juveniles into the tide
pools is discontinuous, occurring between June
and September. Age-classes H- and 2+ were ap-
proximately % and iy2 yr old, respectively, when
sampled. The time of scale-ring formation is then
likely to have been about June. The average
growth rate is estimated to be 45 mm in 5 mo, or
108 mm/yr for the age-class 0-I-. Two fish were
obtained in April 1976 with lengths of 68 and 76
mm, which indicates that the estimate may be
slightly high, the predicted length being 81 mm.
DIETARY COMPOSITION
The composition of the diet is illustrated by oc-
currence and CFI values scaled down to percent-
ages.
30"
20
10
0'
3
3
■Si 3
IT
•b 3
10
10
5
^^^^ i>
Total
2* n=116
Feb.
Mar.
n=2
Apr
May
j}=L
JiJi
n=1
Z^
Jul.
J1=S
^>N.
Aug
il=33
Sep.
^A.
ii=i
Oct.
_Q=]a
^\r^
Nov.
n=11
Dec
n=A
50 100
Standard Length (mm)
150
Figure 4. — Monthly and total length -frequency distribution of
Diplodus cervinus , showing three age-classes ( 0 -(- , 1 -f , and 2 + ).
FIGURE 5. — Monthly and total length-frequency distribution of
Sarpa salpa, showing three age-classes (0-I-, l-i-, and 2 + ).
393
FISHERY BULLETIN: VOL. 76. NO. 2
Diplodus sargus
Winter Feeding
The diet is composed mainly of harpacticoid
copepods, amphipods, algae, isopods, polychaetes,
and ostracods (Figure 6; Table I, n = 88).
The diet of the smallest size class (5-15 mm) is
composed almost equally of harpacticoid copepods
and amphipods, but the percentage consumed of
the former increases in the next size class whereas
that of the latter decreases. The diet remained
similar in the following two size classes. In the 35 -
to 50-mm size class, the fish fed little on harpac-
ticoid copepods, the diet being largely composed of
amphipods. The situation was similar in the
largest size class, although algae and polychaetes
were increasingly taken.
Summer Feeding
8
&
100
80
60
40
20-
Length (mm) 5
No. of Fish 9
Figure 6. — Changes in diet with length of Diplodus sargus,
collected between February and July 1975, as shown by the
comparative feeding index. Food items included in Others are:
brachyurans, crab zoaea, diatoms, echinoderms, hydrozoans,
leptostracans, molluscs, mysidaceans, Palaemon pacificus.
rhydophytan algae, sand, and unidentifiable animal fragments.
Although similar to the winter diet, chironomid
larvae, diatoms, crab zoaea, and leptostracans are
more significant. Cirripede nauplii and an uniden-
tified trochophore larva were also commonly
taken, and these were not found in winter speci-
mens (Figure 7; Table 2, n = 149).
The diet of the smallest size class (5-15 mm SL)
is composed mainly of harpacticoid copepods. The
next size class (15-20 mm SL) fed on a similar diet,
although the percentage of harpacticoids taken
decreased and that of polychaetes and cirripede
nauplii increased. These changes were further
magnified in the 20- to 25-mm size class. In the
next size classes (25-50 mm), there was a change
and the green alga, Ulva sp., contributed sig-
nificantly to the diet.
Poor diving conditions in September, October,
and December reduced the sample size and only
nine fish in the 50- to 165-mm size range were
analysed. In general, these fish showed an increas-
ing tendency to take more amphipods, and less
Table 1. — Changes in the percentage composition of the food of Diplodus sargus with length during the period
February to July 1975, as assessed by the comparative feeding index (CFI) and occurrence (Occ.) methods. In the
case of the former, all values exceeding SC/f have been italicized in order to emphasize those food items which
contribute maximally to the diet ( — = absent).
Size classes (mm)
S^Ts 15-20 20-25 25-35 35-50 50-165
Taxon CFI Occ. CFI Occ. CFI Occ. CFI Occ. CFI Occ, CFI Occ.
Chlorophyla ______ 1.3 33.3 1.0 25.0 12.8 41.7
Rhodophyta — — — — — — — — 0.2 8.3 1.6 16.7
Chrysophyta — — 02 118 — — — — — — 0.3 2.8
Polychaefa 2.4 33,3 11.8 76.5 16.7 75.0 3.6 50,0 14 33.3 12,8 47,2
Crustacea
Amphipoda 47.4 77,8 5,8 29,4 5,0 25,0 7.8 66,7 80,7 66,6 58,7 72.2
Ostracoda 0.5 22.2 4.2 58.8 1.2 75,0 0.2 16.7 0.1 8.3 — —
Harpacticoid copepoda 45.9 88.9 74 2 82,4 69,3 1000 80.9 66.7 1.3 25,0 0,4 11.1
Isopoda 1.2 33.3 1,0 35.3 7,4 75,0 5,0 50,0 3,3 50,0 4,2 33.3
Brachyura (zoaea) — — — — 0,1 12,5 — — — — — —
Tanaidacea _ _ _ _ _ _ 0 1 16,7 0,1 8,3 — —
Macrura ______ 0,1 16.7 0,2 16,7 — —
Mysidacea _ — — — — — 0,2 16.7 — — 0,1 2,8
Insecta _ — — — 0,2 12 5 — — — — 0 1 2 8
Mollusca ________ 01 83 4,1 22,2
Echinodermata — — — — — — — — — — 0.1 2.8
Unidentifiable fragments 2,6 44.4 3.0 52.9 0.1 12.5 0 8 50.0 12.2 66.7 5.4 52.8
No of fish examined 9 17 8 6 12 36
Average no. of points
allotted per stomach 18.8 13.4 7.3 20.3 17.2 13.2
394
CHRISTENSEN: TROPHIC RELATIONSHIPS OF SPARID FISHES
lOOi
80
60
40
20'
Ottiers
■,rww~--| Till IN, irn— --i^ - - ..T^
Length (mm) 5 15 20 25 35 50 165
No of Fish 72 36 K 12 6 9
Figure 7. — Changes in diet with length of Diplodus sargus,
collected between August and December 1975, as shown by the
comparative feeding index. Food items included in Others are
the same as for Figure 8, in addition to the unidentified
trochophore larva.
algae, diatoms, chironomid larvae, and hydrozoa
as they grow larger.
Identity of Food
Diatoms: two species of the genus Licmophora,
predominantly L. pfannkucheae , as well as L.
ehrenbergii.
Other algae: the chlorophytan algae of the
genus Ulua most frequent, although some of the
larger size classes also fed on Caulerpa filiformis
(50-165 mm: 7.5% in summer and 2.9% in winter),
Bryopsis sp., Enteromorpha sp., and Valonia sp.
Some rhodophytans taken, including Ceramium
sp., Hypnea spicivera, Polysiphonia sp., and
Tayloriella spp.
Harpacticoid copepods: 12 species, only 4 com-
mon, identification was not possible.
Amphipods: 28 species — 7 caprellid species in-
cluding Caprella danilevskii, C. penantis, C.
scaura, and Caprellina longicollis; Cerapus
tubularis; Corophium? acherusicum, and C?
triaenunyx; Cymadusa sp.; two Gammaropsis
species including G. holmesi; Jassa spp.;
Lysianassa ceratina; and L. variegata; two Maera
species; Paramoera capensis; Parelasmopus
suluensis; two Photis species; Temnophlias sp.;
Urothoe sp.; and three unidentified species.
Isopods: nine species — Cymodocella pustulata,
C. siiblevis, Dynamenella huttoni, D. mac-
rocephala. Exosphaeroma antikraussi, Gnathia
sp., Janiropsis sp., Panathui-a sp., and a Stene-
trium species.
Polychaetes: Dodecaceria pulchra. E alalia
triliueata, two Nereis species, an Onuphis sp., and
terebellid tentacles most commonly found in the
gut contents, as well as Pista sp.. Pumatoleois
kraussi, and Serpula vermicularis.
Ostracods were not identified.
Table 2. — Changes in percentage composition of the food of Diplodus sargus with length during the period August
to December 1975, as assessed by the comparative feeding index (CFI ) and occurrence (Occ.) methods. In the case of
the former, all values greater than 30% have been italicized in order to emphasize those food items which
contribute maximally to the diet ( — = absent).
size classes (mm)
5^15 15-20 20^55 25-35 35-50 50-165
Taxon CFI Occ. CFI Occ. CFI Occ. CFI Occ. CFI Occ. CFI Occ.
Chlorophyta 02 42 06 83 2,9 28 6 33 4 50 0 135 33,3 143 55.6
Rhodophyta ______ 0,9 33,3 0,8 33,3 — —
Chrysophyta ______ 0,7 25,0 87 33,3 7,4 11,1
Hydrozoa — — — — — — — — — — 11 111
Polychaeta 11 20,8 10,4 47.2 4,0 35 7 14,7 33,3 09 16 7 4.0 22.2
Crustacea
Amphipoda 0.3 6 9 17 27 8 14 1 57 1 2 4 25 0 6 5 50 0 62 5 100.0
Ostracoda 6.0 37 5 5.5 55 6 6 4 64 3 0 8 16 7 0 4 16 7 0 1 111
Harpacticoid copepoda 67 8 86.1 58.? 94.4 47.6 100.0 29 6 83.3 176 100.0 — —
Isopoda 1.0 16 7 2 1 25 0 3 4 42 9 3 0 25 0 6 9 50 0 9 8 44 4
Brachyura 0.1 5.6 - - 0 2 14.3 0 2 8 3 — — 0 2 111
Cirnpedia 8.8 36 1 139 66 7 215 42 9 13 83 — — - -
Leptostraca — — 03 56 — — —
Tanaldacea — — — — 0 2 14 3 — —
Macrura — — — — 0 7 71 —
Insecta 8 8 36 1 4 8 27 8 0 5 214 3 8 33 3 37 2 50 0 — —
Moliusca — — — — — — — — — — 0311,1
Trochophore larvae 4.6 30.6 0.3 5.6 16 14 3 12 83 — — — —
Unidentifiable fragments 1.3 20.8 2.3 30.6 2.9 28.6 8.1 41.7 7.5 50.0 0.2 11.1
No. of fish examined 72 36 14 12 6 9
Average no. of points
allotted per stomach 13.7 17^ 204 12^0 14^8 182
395
FISHERY BULLETIN: VOL. 76, NO. 2
Insects: only the larva of the chironomid Tel-
matogeton minor.
Brachyurans: Rhyncoplax bovis and the gut of
an unidentified crab (in a single case).
Molluscs: Gibbula rosea, Helcion pruinosus,
Philine aperta, and a rhaciglossid (no shell, so not
possible to identify further).
Hydrozoans: Symplectoscyphus sp. and
Thecocarpus formosus were ingested by one fish in
the 50- to 70-mm size class.
Echinoderms: Parechinus sp.
Tanaidaceans: Leptochelia barnardi most com-
monly found, also an Apseudes sp.
Mysidaceans: only one species, Mysidops
similis, could be identified with any certainty.
Di plod us cervintis
The diet of Diplodus cervinus is illustrated in
Figure 8 and Table 3 (n = 67). The juveniles
(10-20 mm) fed mainly on harpacticoid copepods
and chironomid larvae. In the next size class
(20-35 mm), juveniles of the sand shrimp,
Palaemon pacificus, were taken instead of
chironomid larvae. This trend continues in the 35-
to 50-mm size class, the diet being composed
largely of P. pacificus as well as harpacticoid
copepods. Polychaetes are more important in this
size group, a trend which is maintained in all
larger size classes. In the larger fish, there was
again a changeover, the percentage of amphipods
taken being 65.7% (50-75 mm) and 27.6% (75-100
mm). Unidentifiable crustacean fragments in
100'
80'
60
40
20
Length (nun) 10
No.of Fish
135
Figure 8. — Changes in diet with length in Diplodus cervinus , as
shown by the comparative feeding index. Food items included in
Others are: chlorophytan algae, cirripede nauplii, coralline al-
gae, molluscs, mysidaceans, ostracods, tanaidaceans, and the
unidentified trochophore larva.
these size classes composed 12.6% and 47.9%, re-
spectively, which is partly explained by the fact
that 6 of the 40 fish were taken at night and their
stomach contents were largely digested and thus
indistinguishable. The diet of the largest size class
( 100-135 mm) was made up mainly of polychaetes.
Identity of Food
The diet was composed of almost all the food
species listed for D. sargus, although in differing
proportions, as well as the following: Cymodocella
eutylos (isopod), Littorina knysnaensis (mollusc),
Table 3. — Changes in the percentage composition of the food of Diplodus cervinus with length, as assessed by the
comparative feeding index (CFI) and occurrence (Occ.) methods. In the case of the latter, all values exceeding 30%
have been italicized in order to emphasize those food items which contribute maximally to the diet ( — = absent).
Size classes (mm)
10-20 20-35 35-50 50-75 75-100 100-135
Taxon CFI Occ. CFI Occ. CFI Occ. CFI Occ. CFI Occ. CFI Occ.
Chlorophyta — — — — 4.6 12.5 — — — — — —
Rhodophyta __________1.9 16.7
Polychaeta 0.2 25 0 3 8 33.3 17 6 50 0 17 2 72 4 11.4 18.0 65.4 82.3
Crustacea
Amphipoda 4.1 25 0 10 33.3 0 4 25 0 65 7 74 9 27 6 45 0 5.8 33.3
Ostracoda — — 02 22 2 — — — — 0.6 9.0 — —
Harpacticoid copepoda 56.5 100.0 47,6 77 8 16.5 87.5 2.2 51.7 0.1 9.0 — —
Isopoda — — 10.0 44.4 4.2 25.0 1.5 20.7 11.0 27.0 1.7 16.7
Cirripedia (nauplii) — — — — — — 0.5 6,9 — — — —
Macrura — — 4 7.2 44.4 53.2 50,0 — — 1.0 9.0 — —
Tanaidacea — — — — — — 0.2 6.8 — — — —
Mysidacea — — — — — — 0.1 3.4 — — — —
Insecta 39 0 50.0 2.2 55.6 0 4 12.5 — — — — — —
Mollusca — — — — — — — — 0,4 9 0 1.9 16.7
Trochophore larvae 0.2 25.0 — — ___ — ___ —
Unidentifiable fragments — — — — 3.1 50.0 12.6 79.3 47.9 36.0 23.3 33.3
No. of fish examined 4 9 8 29 11 6
Average no of points
allotted per stomach 8.3 19.3 8.0 10.0 6.4 11.1
396
CHRISTENSEN: TROPHIC RELATIONSHIPS OF SPARID FISHES
Phoxostoma sp. (amphipod), and Corallina sp.
(rhodophytan alga).
Sarpa salpa
The percentage composition of the diet of Sarpa
salpa is illustrated in Figure 9 and Table 4. The
food habits of this species changed from being
primarily a carnivore as a juvenile to a herbivore
as a subadult.
Juveniles in the 10- to 25-mm size classes fed
mainly on harpacticoid copepods. In the next size
100-
Figure 9. — Changes in diet with length of Sarpa salpa, as
shown by the comparative feeding index. Food items included in
Others are: bryozoans, cirripede nauplii , crab zoaea, fish muscle,
insects, isopods, leptostracans, mysidaceans, ostracods,
pMjlychaetes, rhaciglossid molluscs, and tanaidaceans.
class (25-35 mm), however, the total animal con-
tribution is only 15.1%, diatoms and rhodophytan
algae being most important. Diatoms are taken in
decreasing amounts from then on and those fish
>75 mm SL fed predominantly on chlorophytan
and rhodophytan algae.
Identity of Food
Chlorophytan algae: eight species — Bryopsis
sp., Caulerpa filiformis, Chaemaedoris delphini,
Cladophora spip.,Enteromorpha sp.,Rhizoclonium
sp., and Ulua sp.
Rhodophytan algae: Ceramium sp., Champia
compressa, Hypnea spicifera, Polysiphonia sp.,
and Tayloriella sp. commonly taken as well as Ac-
rosorium sp., Arthrocardia sp., Centroceras sp.,
Corallina spp., Polyzonia elegans, and Ptero-
siphonia cloiophylla.
Chrysophytan algae: three species of
diatoms — Isthmia enervis, Licmophora ehrenber-
gii, and L. pfannkucheae .
Hydrozoans: Gattya humilis commonly ingested
by the 75- to 100-mm size class, whereas Ser-
tularella sp. and Thecocarpus formosus were un-
common.
Polychaetes: only two species of Nereis.
Isopods: uncommon, but six species were
found — Dynamenella huttoni, D. macrocephala,
Gnathia sp., Janiropsis sp., Panathura sp., and
Stenetrium sp.
Table 4. — Changes in the percentage composition of the food of Sarpa salpa with length, as assessed by the comparative
feeding index (CFI) and occurrence (Occ.) methods. In the case of the latter, all values exceeding 30% have been italicized in
order to emphasize those food items which contribute maximally to the diet ( — = absent).
Size classes (mm)
10-15 15-25 25-35 35-50 50-75 75-100 125-150
Taxon CFI Oca ^R Oca "CFJ Oca CFI Occ. CFI Occ. CFI Occ. CFI Occ.
Chlorophyta 0.3 8 7 — — 0 9 313 37.2 92 0 34.0 917 30 6 100.0 48.3 100.0
Rhodophyla — — 0.4 9.1 21.3 62.5 323 76.0 28.6 66.7 45.9 83.3 50.7 100,0
Chrysophyta 0.2 87 9.3 24,2 62,7 1000 32 4 68,0 33.4 66.7 _ _ _ —
Hydrozoa _______ — 0,1 16.7 20.5 50.0 0.8 50.0
Polychaeta 5.1 17.4 2.6 30,3 _ _ _ _ 0.4 8,3 0,1 167 — —
Crustacea
Isopoda 17 17.4 0 9 30.3 0,1 6 3 — — — — — — — —
Amphipoda 73 30,4 17,8 63,6 4,2 50 0 0.9 16.0 0.1 16.7 0.3 33.3 0.8 50 0
Ostracoda 67 26 1 14 36 4 01 63 01 12 0 ______
Harpacticoid copepoda 77.0 87,0 64 5 72 7 0,5 18 8 1,4 24 0 ______
Cirnpedia (nauplii) 5,5 22,0 2 1 18,2 _____ — _ — — —
Brachyura (zoaea) — — 0,1 3,0 0,1 12.5 ________
Leptostraca 0 1 4 4 — — — — — — — — — — — —
Insecta 0,1 4 4 0,2 3.0 0.1 12.5 — — — — — — — —
Bryozoa ________ — — 01 16 7 — —
Mollusca __________ 0.1 16,7 — —
Pisces ____ 0,2 6,3 — — — — — — — —
Unidentified fragments
and sand 2.0 26.1 0.7 24.2 9.8 31.3 1.7 32.0 3.4 41.7 2.4 50.0 — —
No, of fisfi examined 23 33 16 25 12 6 2
Average no of points
allotted per stomach 7.8 9.5 13.8 15.7 16.5 15.5 28.0
397
FISHERY BULLETIN: VOL. 76. NO. 2
Amphipods: also uncommon except in juveniles
which took the following 15 species: Caprella cicur
and two other species of^ Caprella, Cerapus sp.,
Corophium sp., Gammaropsis sp., Lysianassa
ceratina, L. uariegata, Jassa sp., Maera sp.,
Paramoera capensis, Parelasmopus suluensis,
Photis sp., and two unidentified species.
Harpacticoid copepods: eight species.
Insects: larvae of the chironomid Telmatogeton
minor.
Tanaidaceans: Leptochelia barnardi, uncom-
mon.
Molluscs: rhaciglossid.
DENTITION AND GUT MORPHOLOGY
Diplodus ceruinus — there are six upper and four
lower incisors which are narrower than those of Z).
sargus, and there are fewer molars, the number
increasing with age (Figure lOA). This would in-
dicate that adult D. cervinus feed on softer foods
than D. sargus , which is borne out as the diet of the
former consists primarily of polychaetes, whereas
the latter took amphipods and molluscs as well.
The gut of D. ceruinus is short with a G/S ratio of
0.7 in a 16.5-mm fish and 0.95 in one 74 mm long.
Diplodus sargus — there are four stout incisors
and three to four rows of fairly large molars in
each jaw (Figure lOB), the latter increasing in size
and number with age. The teeth are those of a
typical omnivore (Weatherly 1972). The G/S ratio
was 0.76 in a 16.5-mm fish and this is within the
range of omnivores as defined by Nikolsky ( 1963).
Sarpa salpa — this species shows a change in
dentition correlated with age and diet. The young
fish are carnivorous and have short, pointed coni-
FlGURE 10.— Dentition. A. Medial view of
the left upper and lower jaws oi Diplodus
cervinus (MSC 75-36, 94 mm SL). B. Me-
dial view of the left upper and lower jaws
ofZ). sargus (MSC 75-34, 107 mm SL). C.
Lateral view of the upper and lower jaws
of a juvenile Sarpa salpa (MSC 75-39, 20
mm SL). D. Lateral view of a single tooth
of a subadult S. salpa (MSC 75-37, 39 mm
SL). E. Lateral view of the upper and
lower jaws of an adult S. salpa (RUSI 74-
323, 99 mm SL).
398
CHRISTENSEN: TROPHIC RELATIONSHIPS OF SPARID FISHES
cal teeth to grasp prey (Figure IOC). Multicusped,
incisiform teeth begin to break through when the
fish are about 20 mm long (Figure lOD). The
pointed teeth are completely replaced by the time
the fish are 35 mm long, after which they feed
predominantly on algae ( approximately 60% CFI).
The multiple cusps wear away and the teeth are
bicuspid incisiform by the time the fish are 65 to 75
mm long (Figure lOE). These can nip at algae and
the diet is composed of 65 to 117c plant matter at
this stage.
The gut shows a corresponding change in length
with diet, a long gut being characteristic of a her-
bivore. The G/S ratio increases from 0.86 in a
20-mm fish (Figure llC, typically omnivorous) to
1.36 in a 39-mm fish (Figure IIB), and 2.66 in a
99-mm individual (Figure llA). This latter value
is typical of a herbivore (Nikolsky 1963), although
not as pronounced as in some other herbivorous
fish species.
DISCUSSION
A number of fish species occur as juveniles or
spend their entire life cycle in the eastern Cape
intertidal (see description under Study Area). The
family Sparidae includes the largest number and
the present investigation of the trophic relation-
ships of three of these was initiated as competitive
interaction is often most vigorous in closely re-
lated fish ( Fryer and lies 1972). There is an intense
dietary overlap in some cases and the available
resources are subdivided in two main ways. Re-
cruitment of juveniles of the three species takes
Figure ll. — Lateral views ofSarpa salpa with the gut unravel-
led and displayed to illustrate the increase in gut length with
size. A. 99 mm SL IRUSI 74-323). B. 39 mm SL (MSC 75-37). C.
20 mm SL (MSC 75-39).
place at different times of the year and this re-
duces competition between those size groups in
which the greatest feeding overlap was observed.
The remaining size classes were separated as their
diets were different.
Small juveniles of the three species have the
most similar diets of all size classes studied. The
resulting competition is reduced by two
mechanisms. Firstly, juveniles of Sarpa salpa
occur in the tide pools primarily from July to Sep-
tember (Figure 5) whereas those of Diplodus cer-
vinus were found during October and November
(Figure 4) and D. sargus was present throughout
the year (Figure 3). Secondly, at the time of maxi-
mal competition (July-November), the diet of
small D. sargus includes food items not taken at
other times of the year, e.g., chironomid larvae
and cirripede nauplii (Figure 7). This may be due
to either the presence of these prey items only at
that time of the year and/or to the effects of com-
petition forcing D. sargus to include them in its
diet. A combination of both factors would appear to
be operative in the case of the chironomids as the
larvae were obtained in bottom samples taken in
October-November and not in March. No data are
available for cirripede nauplii, crab zoaea, and the
unidentified trochophore larva as plankton sam-
ples were not taken.
Competition for food is greatly reduced by the
time the three sparids are about 25 to 30 mm long.
At this stage, S. salpa feeds mainly on diatoms and
red algae (Figure 9); D. cervinus ingests Palaemon
pacificus, harpacticoid copepods, and isopods (Fig-
ure 8); and D. sargus takes green algae, harpac-
ticoids, chironomid larvae, and cirripede nauplii
(Figures 6, 7). The separation is equally distinct in
subadult fish as S. salpa is then a herbivore, D.
cervinus takes polychaetes, some amphipods, and
isopods, while D. sargus feeds on amphipods and
399
FISHERY BULLETIN: VOL. 76, NO. 2
green algae. The overlap on amphipods by the
latter two species may be partially compensated
for by behavioral separation. Diplodus cervinus is
a secretive substrate feeder whereas Z). sargus is a
more open water fish tending to feed on vertical
rock surfaces away from the bottom. The fact that
neither species was very common intertidally in
these size classes may also contribute towards a
reduction in competition.
The diet of large juvenile S. salpa is unusual in
that it consists mainly of diatoms and epiphytic
rhodophytan algae which occur commonly on
corallines, Hypnea spicifera and Tayloriella spp.
(M. H. Giffen, pers. commun.). The fish must,
therefore, selectively separate these food items as
few fragments of the algae on which they grow
were found in the stomach contents. This is in
contrast to Rhabdosargus holubi, also a sparid,
which ingests algae for their epiphytic diatoms
rather than separating them, even though the
algae are not digested (Blaber 1974). The situation
may be similar to this in larger S. salpa as the
rectal contents appeared to be relatively undi-
gested and fewer diatoms were observed on the
algae (Figure 12).
Temporal separation of juveniles to reduce com-
petition has not been reported previously for tide
pool fish species as far as I am aware, although it
has been observed in two pelagic plankton feeders
from the Adriatic, the anchovy and sardine ( Vuce-
tic 1975). Large dietary overlaps have been noted
in several intertidal fish, including blennies,
clinids, gobies, and labrids (Gibson 1968, 1972).
These fed predominantly on crustaceans and it is
possible that similar mechanisms reduce competi-
tive pressure amongst them, although this was not
determined as samples were only taken for 2 mo.
The data presented indicates that/), cervinus is
a carnivore, D. sargus is an omnivore, and S. salpa
an omnivore when juvenile and a herbivore when
adult. Similar feeding habits were found for adults
of the same three species in the Klein River es-
tuary (Talbot 1954). The dentition and gross gut
morphology changed with size and this was most
marked in S. salpa, corresponding with the ob-
served diet. Comparable transformations have
been reported for other fish species (Nikolsky
1963).
Two other sparids, Sparodon durbanensis and
R. holubi, cohabit with those studied in the littoral
zone. The few specimens examined had fed mainly
on harpacticoid copepods as small juveniles, with a
resultant overlap with D. sargus as all three oc-
curred in the research area in November-
December. Large specimens were not examined,
but R. holubi appears to be an omnivore as a
juvenile in estuaries and a carnivore feeding on
molluscs as an adult (Talbot 1954; Blaber 1974).
Adult S. durbanensis are carnivores feeding on
small fish, molluscs, and crustaceans (Biden
Figure 12.— Scanning electron micrograph of the surface of a chlorophytan alga, Ulva sp., removed from the gut ofSarpa salpa (147
mm SL) to show the disappearance of diatoms. A. Oesophageal sample. B. Rectal sample.
400
CHRISTENSEN: TROPHIC RELATIONSHIPS OF SPARID FISHES
1954). Competitive pressure is, therefore,
minimized between the five sparid fish species
commonly occurring in the littoral zone.
ACKNOWLEDGMENTS
This study was submitted in partial fulfillment
of the requirements for an M.S. degree in the J. L.
B. Smith Institute of Ichthyology, Rhodes Univer-
sity, Grahamstown. I thank G. S. Butler, B. J. Hill,
P. B. N. Jackson, M. van Harten, and R. Winter-
bottom for assistance in the field and/or valuable
criticisms. I am also grateful to the following, who
identified many of the food items: C. Griffiths (am-
phipods) and B. F. Kensley (isopods and
brachyurans). University of Cape Town; M. H.
Giffen (diatoms). Fort Hare University, Alice; and
S. C. Seagrief (algae), Rhodes University,
Grahamstown. R. Winterbottom provided useful
suggestions and criticisms as my supervisor, and
carefully reviewed and commented on the manu-
script, as did R. N. Gibson, Dunstaffnage Marine
Research Laboratory, Argyll, Scotland, and E. S.
Hobson, National Marine Fisheries Service,
NCAA, Tiburon, Calif. Their assistance was
greatly appreciated as was that of J. Pote who
typed the manuscript. Finally, I thank the follow-
ing for financial assistance: J. L. B. Smith Insti-
tute of Ichthyology (Hugh Le May scholarship),
CSIR (research bursary), and Rhodes University
(Research Council Grant No. 2857, R. Winterbot-
tom principal investigator).
LITERATURE CITED
Barnard, K. H.
1927. A monograph of the marine fishes of South Africa.
Ann. S. Afr. Mus. 21(2):419-1065.
BIDEN, C. L.
1954. Sea-anghng fishes of the Cape: a natural history of
some of the principal fishes caught by sea anglers and
professional fishermen in Cape waters. 2d ed. Juta and
Co. Ltd., Capetown and Johannesb., 304 p.
BLABER, S. J. M.
1974. Field studies of the diet of Rhabdosargus holubi
(Pisces: Teleostei: Sparidae). J. Zool. (Lond.) 173:407-
417.
Frost, W. E.
1943. The natural history of the minnow, Phoxinus phox-
inus. J. Anim. Ecol. 12:139-162.
Fryer, G., and T. D. Iles.
1972. The cichlid fishes of the great lakes of Africa: their
biology and evolution. Oliver & Boyd, Edinb., 641 p.
Gannon, J. E.
1976. The effects of differential digestion rates of zoo-
plankton by alewife, Alosa pseudoharengus , on determi-
nations of selective feeding. Trans. Am. Fish. Soc.
105:89-95.
Gibson, R. N.
1968. The food and feeding relationships of littoral fish in
the Banyuls region. Vie Milieu, Ser. A, 19:447-456.
1972. The vertical distribution and feeding relationships
of intertidal fish on the Atlantic coast of France. J.
Anim. Ecol. 41:189-207.
HoBSON, E. S.
1974. Feeding relationships of teleostean fishes on coral
reefs on Kona, Hawaii. Fish. Bull., U.S. 72:915-1031.
HoBsoN, E. S., AND J. R. Chess.
1973. Feeding oriented movements of the atherinid fish
Pranesus pinguis at Majuro Atoll, Marshall Islands. Fish.
Bull., U.S. 71:777-786.
hynes, h. b. n.
1950. The food of fresh-water sticklebacks (Gasterosteus
aculeatus and Pygosteus pungitius), with a review of
methods used in studies of the food of fishes. J. Anim.
Ecol. 19:36-58.
KlONKA, B. C, AND J. T. WINDELL.
1972. Differential movement ofdigestible and indigestible
food fractions in rainbow trout, Salmo gairdneri . Trans.
Am. Fish. Soc. 101:112-115.
NIKOLSKY, G. V.
1963. The ecology of fishes. Academic Press, N. Y., 352 p.
PINKAS, L.
1966. A management study of the California barracuda
Sphyraena argentea Girard. Calif Dep. Fish Game,
Fish. Bull. 134, 58 p.
RANDALL, J. E.
1967. Food habits of reef fishes of the West Indies. Stud.
Trop. Oceanogr. (Miami) 5:665-847.
ROUND, F. E.
1971. Benthic marine diatoms. Oceanogr. Mar. Biol.
Annu. Rev. 9:83-139.
ROYCE, W. F.
1972. Introduction to the fishery sciences. Academic
Press, N.Y., 351 p.
SMITH, J. L. B.
1965. The sea fishes of southern Africa. Cape and Trans-
vaal Printers Ltd., Cape Town, 580 p.
SWYNNERTON, G. H., AND E. B. WORTHINGTON.
1940. Note on the food of fish in Haweswater (Westmor-
land). J. Anim. Ecol. 9:183-187.
Talbot, F. H.
1955. Notes on the biology of the white stumpnose, Rhab-
dosargus globiceps (Cuvier), and on the fish fauna of the
Klein river estuary. Trans. R. Soc. S. Afr. 34:387-407.
Taylor, W. R.
1967. An enzyme method of clearing and staining small
vertebrates. Proc. U.S. Natl. Mus. 122(3596), 17 p.
VUCETIC, T.
1973. Synchronism of the spawning season of some pelagic
fishes (sardine, anchovy) and the timing of the maximal
food (zooplankton) production in the Central Adria-
tic. Pubbl. Staz. Zool. Napoli, 39 Suppl. 1:347-365.
Weatherly, a. H.
1972. Growth and ecology of fish populations. Academic
Press, Lond., 293 p.
WINDELL, J. T.
1971. Food analysis and rate of digestion, /n W. E. Ricker
( editor) , Methods for assessment offish production in fresh
waters, p. 215-226. Blackwell Scientific Publications, Oxf.
and Edinb.
401
EFFECT OF STARVATION ON THE HISTOLOGICAL AND
MORPHOLOGICAL CHARACTERISTICS OF JACK MACKEREL,
TRACHURUS SYMMETRICUS, LARVAE
Gail H. Theilacker^
ABSTRACT
Histological and morphological criteria were developed to assess the nutritional condition of
laboratory-reEired jack mackerel, Trachurus symmetricus, larvae. A comparison of the histological
features of fed and starved larvae revealed that the digestive tract and its associated glands were the
first tissues to be affected by starvation. The extent of cellular deterioration increased with time of
starvation. To classify larval condition, histological characteristics of the pancreas and gut were given
numerical grades. The histological technique correctly classified 839c of the feeding and starving
larvae.
The morphometric analysis relied upon a stepwise discriminant analysis that used a combination of
five measurements (standard length, head length, eye diameter, body depth at the pectoral, and body
depth at the anus) to estimate individual larval condition. The morphometric method was as sensitive
as the histological examination in determining whether or not a larva was fed or starved. Ultimately,
these histological and morphological criteria may be useful for estimating larval survival in the field by
assessing the condition of sea-caught larvae.
Fishery scientists generally agree that observed
fluctuations in recruitment of young fish to a fish
stock may be the consequence of mortality during
the larval stage. Because starvation is probably
one of the principal causes of mortality (Hunter
1976a), a need exists to develop criteria for detect-
ing the incidence of starvation in sea-caught
specimens. Several scientists have suggested that
the differences in body form between feeding and
starving larvae could be used to identify the nutri-
tional status of larvae caught in sea surveys. For
example, Shelbourne ( 1957) based his assessment
of the condition of ocean-caught plaice,
Pleuronectes platessa, larvae on their external ap-
pearance. Certain morphometric measurements
also can be indicative of starvation. A decrease in
thickness of the larval fish body has been corre-
lated with starvation for several marine and
freshwater fish larvae [herring, Clupea harengus,
and plaice (Ehrlich et al. 1976); northern anchovy,
Engraulis mordax (Arthur 1976); anchovy, E.
japonica (Honjo et al. 1959; Nakai et al. 1969);
pike, Esox lucius, and carp, Cyprinus carpio (Kos-
tomarova 1962)]. Other morphological features
(Ehrlich et al. 1976) considered to be indicative of
'Southwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, La Jolla, CA 92038.
Manuscript accepted August 1977.
FISHERY BULLETIN: VOL. 76, NO. 2, 1978.
starvation in herring and plaice were a decrease in
the angle of the pectoral girdle, a change in the
ratio of the head to eye height (herring only), and a
decrease in the relative condition factor. Coinci-
dent with morphometric differences caused by
starvation, Ehrlich et al. (1976) described his-
tological changes in the gut and liver. The his-
tological approach was used to classify yellowtail,
Seriola quinqueradiata, larvae into "feeding,"
"semi-feeding," and "starving" groups by Umeda
and Ochiai (1975). This technique was also effec-
tive for diagnosing starvation in northern an-
chovy larvae (O'Connell 1976). In both species,
degeneration of cells of digestive organs was the
best indicator for identification of starvation. Sev-
eral other studies also have correlated starvation
in fish larvae with degeneration of the digestive
organs, mainly the gut. Kostomarova (1962) de-
scribed a retardation in development of the gut in
larvae of starved carp and pike and a reduction in
the depth of the epithelial cells lining the gut.
Reduced gut cell height was also reported for the
larvae of starved yellowtail (Umeda and Ochiai
1975), herring, and plaice (Ehrlich et al. 1976).
Morphological criteria are preferable to his-
tological ones because they take much less time to
determine and require no special preservation
techniques. However, histological criteria may be
more accurate for classifying individual larva.
403
FISHERY BULLETIN: VOL. 76, NO. 2
The purpose of this study was to develop mor-
phological and histological criteria for assessing
the nutritional condition of jack mackerel larvae
and to evaluate these criteria by comparing their
success in identifying fed and starved larvae
reared in the laboratory. Ultimately, criteria
based on these results may be useful for estimat-
ing larval survival in the field by assessing the
condition of sea-caught larvae.
MATERIALS AND METHODS
Jack mackerel eggs were collected by towing a
1-m (mouth diameter, 0.505-mm mesh) plankton
net just below the sea surface at various locations
between 20 and 200 mi (32 and 320 km) off the
coast of southern California in June and July 1975
and in May 1976. The eggs were separated from
most of the plankton at sea and then sorted by
developmental stage at the Southwest Fisheries
Center, La Jolla, Calif. Temperature was main-
tained at 15°C during sorting and in the larval
rearing containers. The light cycle was 12 h light
and 12 h dark. Five hundred normally developing
eggs from a single day's spawning were transfer-
red into 100 1 black Kydex^ circular rearing tanks
containing filtered seawater {5fxm, Cuno filtered).
There were three experiments and two treatments
in each experiment; larvae in one tank were of-
fered food while those in the other were not. The
fed larvae were given a diet of a naked dinoflagel-
late, Gymnodinium splendens (50/ml), a rotifer,
Brachionus plicatilis (30-40/ml), and a copepod,
Tisbe sp. (1 or 2/ml). This feeding method has been
described (Lasker et al. 1970; Theilacker and
McMaster 1971; Hunter 1976b).
Histological criteria were developed in the first
two experiments. The sampling procedure and the
number of larvae sampled differed depending on
the requirements of the analysis. Collectively, a
total of 152 larvae were examined. In the third
experiment, usually 15 larvae were sampled daily
for 5 days from the "fed" tank (n =69) and 3 days
in the "starved" tank (n = 48). All larvae were
examined both histologically and morphologi-
cally. No dead larvae were sampled because the
postmortem change which takes place in tissues of
fish larvae, due to digestion by their own enzymes
(autolysis), resembled antemortem destruction
caused by starvation. Standard length of each
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
404
larva was measured on a slide; then seawater was
removed and replaced by Bouin's fixative. Pre-
serving individual larvae in this manner assured
that each would be straight and flat, facilitating
subsequent morphometric measurements. Five
measurements were taken after preservation to
monitor daily changes in larval body form and
determine effects of starvation: standard length
(SL, tip of upper jaw to tip of notochord), head
length (HL, tip of upper jaw to cleithrum), eye
diameter (ed), body depth at the pectoral (bd-1),
and body depth at the anus (bd-2). Standard length
shrinkage in Bouin's fixative was 11.5%. Next,
measured larvae were prepared for histological
examination using standard techniques. Larvae
were transferred from Bouin's to 70% ethyl al-
cohol after 24 h, dehydrated with an ethyl n-butyl
alcohol series in a Fisher Tissuemation, and em-
bedded in Paraplast-plus. The Paraplast paraffin
blocks were frozen with fluorocarbon spray
(Cryokwick) just before the larvae were serially
sectioned at 5 fxm in a sagittal plane. The mounted
sections were stained with Harris' hematoxylin
and eosin and mounted in synthetic resin.
Histological Grading System
Recently O'Connell (1976) developed a numeri-
cal, histological grading system to characterize
the nutritional condition of individual northern
anchovy larvae. He examined tissues of the larvae
microscopically to determine the features of tissue
microstructure that were affected by starvation. A
grade was assigned to each feature based on the
degree of similarity or dissimilarity of the his-
tological microstructure between the starved lar-
vae and fed larvae. I followed this method and
modified it as necessary for the tissues of jack
mackerel.
Each of the histological characteristics used to
assess starvation in jack mackerel larvae was
evaluated and assigned a grade. A grade of "3" was
given to a characteristic which resembled that in
"normal or healthy" larvae, a grade of "2" was
given to an intermediate condition, a grade of "1"
to the starved condition. Since these criteria were
established by comparing actively feeding, seem-
ingly healthy larvae with moribund larvae, it was
assumed that an average of the 12 graded features
examined for each larva classified the larva into
the correct nutritional group: "healthy group,"
average grade = 2.34 to 3.00; "intermediate
group" = 1.67 to 2.33; and "starved" = 1.00 to
THEILACKER: EFFECT OF STARVATION ON JACK MACKEREL
1.66 (the break points establish three equal
groups).
Data Analysis
In the main, conclusions are based on the results
of a stepwise discriminant analysis (SWDA). A
discriminant analysis allows one to distinguish
between two or more groups, given a set of vari-
ables that describe the characteristics in which
the individuals in each group are expected to dif-
fer. In the stepwise discriminant analysis, all the
variables are introduced into a SWDA computer
program and the best set of variables, based on the
generalized Mahalanobis distance (Rao 1952), is
selected. The first variable chosen will usually be
the one which gives the best score when classify-
ing the individuals into their predetermined
groups. The score is equal to the number correctly
classified. The selection of each succeeding vari-
able improves the score until a subset is chosen
which is as good as the full set of variables for
discriminating the groups. All variables not in-
cluded in the final subset are considered superflu-
ous, or not necessary for classification.
RESULTS
between the round cells. In starved larvae, mitotic
activity was arrested and many of the cells were
shrunken, which caused large clear areas to ap-
pear between densely stained, atrophied cells.
Liver (Figures 5-8)
In the normal larval jack mackerel liver the
hepatic cords were two cells thick. Within each
hepatocyte, the nucleus (3) was regular in shape
and distinct. The cytoplasm (4) was well dispersed
with intracellular spaces, probably an area where
glycogen and lipid are stored. Sinusoid areas,
where metabolic exchanges take place between
the hepatic cords, contained blood cells. After
starvation for a few days, the liver atrophied, the
cytoplasm condensed, stained darkly, and in-
tracellular spaces had disappeared. There were
focal degenerative or necrotic areas and accumu-
lations of eosinophilic granules and masses. Nuc-
lei were often irregular in shape and granules
appeared around their periphery; these darkened
areas are presumed to be condensed, inactive
chromatin (Stein et al. 1975). The gallbladder (5)
in starved larvae was always enlarged; normally
it discharges its contents under the stimulus of
food (Love 1970).
Histological
The histological condition of yolk-sac and ac-
tively feeding larvae ("normal") was compared
with that of 3-day starved larvae and many differ-
ences were noted in the cells, tissues, and organs.
The degree of apparent histological deterioration
of 1- and 2-day starved larvae was intermediate
between the "normal" and moribund status. This
condition was termed "semi-starved" by Umeda
and Ochiai (1975) and "intermediate" by O'Con-
nell(1976).
The following section describes the normal his-
tology of actively feedingjack mackerel larvae and
that of starving larvae. Twelve histological
criteria, which appeared to be indicators of starva-
tion, were identified; they are numbered in the test
below and are referenced in the photomicrographs
(Figures 1-12).
Brain (Figures 3, 4)
The primitive brain cells exhibited a high inci-
dence of mitotic activity (1) in normal larvae and
there was relatively little intercellular space (2)
Pancreas (Figures 5-8)
Cells of the exocrine pancreas were arranged in
series, in a circular fashion, as a secretory unit
called an acinus. The nucleus (6), clear and dis-
tinct, was located in the basal portion of the
pyramidal cells. The acinar arrangement of the
pancreatic cells (7) was found to be very sensitive
to deprivation of food. A breakdown in the sym-
metry in the acinus was slight but usually detect-
able after 1 day of starvation. Tissue degeneration
after 3 days of starvation was extreme. The nu-
cleus was irregular and uniformly stained and
there was no detectable acinar arrangement. The
presence of zymogen, digestive proenzyme se-
creted by the acinus and stored as granules at its
central apex, was usually associated with starva-
tion.
Digestive Tract (Figures 7-12)
The columnar epithelial cells of the midgut were
closely united (8) in a single layer. Microvilli were
visible along the border of the lumen giving a
brush effect. In starved larvae, the midgut cells
405
FISHERY BULLETIN: VOL. 76, NO. 2
Pancreas Swim bladder
Muscle
Hindgut
Notochord
Midgut
Brain
Liver
Heart
Pancreas
Notochord
Swim bladder
Hindgut
Midgut
Liver
Gall bladder
m^-^
N
A RBO
:^. itz ^a
foodi'
i^ BC--
M6
FG .
M K
W ^
Figure l. — Trachums symmetricus larva, day 8, fed for 3 days. All 12 histological criteria graded as "healthy." 32 x.
Figure 2. — Trachums symmetricus larva, day 8, starved for 3 days. All 12 histological criteria graded as "starved." 32 x.
Figure 3. — Head of fed larva, graded "healthy." Mitotic activity (1 h) is indicated. Note close proximity of primitive brain cells to
each other. 200 x. h = histological grade = healthy.
Figure 4. — Head of starved larva, graded "starved." Atrophied and darkly stained primitive brain cells (1 s); large intercellular
spaces (2 s). 200 x. s = histological grade = starved.
Figure 5. — Day 8 fed larva. Histological features graded "healthy." 200 x . See enlargement. Figure 7. B = brain, BC = blood cells
(white or immature red), FG = foregut.I = Islet of Langerhans (endocrine pancreas), K = kidney, L = liver, MG = midgut, M = mus-
cle, N = notochord, P = exocrine pancreas, RBC = red blood cells, SB = swim bladder.
Figure 6. — Day 8, starved larva. Histological features graded "starved." 200 x. See enlarge, Figure 8. GB = gallbladder, O = oil,
Y = yolk, see Figure legend 5 for rest of symbols.
Figure 7. — Day 8 fed Trachurus symmetricus larva. Enlargement of Figure 5. Midgut cells in close union (8 h); prominent nuclei (3
h) and large intracellular spaces (4 h) in the liver; pancreatic nuclei distinct (6 h) and cells arranged in a circular unit (acinus) (7 h);
gallbladder (GB) "normal." 480 x. h = histological grade = "healthy," L = liver, MG = midgut, P = exocrine pancreas.
406
THEILACKER: EFFECT OF STARVATION ON JACK MACKEREL
* etr
^''Wj
MG.Sh »^
. L
;^5k
7
M,llh
MG
EC^
Hi M
10
•'^— ^
II
Figure 8. — Day 8, starved Trachurus symmetricus larva. Enlargement of Figure 6. Loss of integrity of midgut cells (8 s); atrophied
liver with dark staining and irregular nuclei (3 s); no acinar cellular arrangement in pancreas (7 s); separated muscle fibers (11 s);
swollen kidney (K); distended gallbladder (GB); note presence of yolk and oil (Y, O), and eosinophilic mass (EM) in liver. 480 x.
s = histological grade = "starved," M = muscle, MG = midgut, P = exocrine pancreas.
Figure 9. — Day 8 fed larva. Large eosinophilic inclusions in hindgut (10 h); muscle fibers closely packed (11 h); thin, epithelial
integumental cells (EC) are prominent below gut and above trunk musculature. 200 x. h = histological grade ="healthy,"
HG = hindgut, M = muscle, MG = midgut.
Figure id. — Day 8, starved larva. Loss of cellular structure in hindgut; enlarged epithelial integument cells (EC); separated muscle
fibers (11 s). 200 x. s = histological grade = "starved," HG = hindgut, M = muscle, N = notochord, SP = spinal cord.
Figure ll. — Day 7 larva, starved 2 days and histologically graded "intermediate." No inclusions in hindgut; cellular separation in
midgut (MG) and hindgut (HG); midgut cells sloughing (9 s); muscle fibers beginning to separate (11 i). 200 x. i = histological
grade = intermediate, s = histological grade = "starved," M = muscle, N = notochord, SP = spinal cord.
Figure 12. — Day 7 larva, starved 2 days, graded "intermediate." Abundant intermuscular tissue (12 h); no muscle (M) fiber
separation; pancreatic nuclei (6 i) not distinct and cellular acinar arrangement lacking (7 i); midgut cells separating (8 i) and sloughing
(9 s). 480 X. i = histological grade = intermediate, s = histological grade = "starved," MG = midgut, P = exocrine pancreas.
407
FISHERY BULLETIN: VOL. 76, NO. 2
began to separate from each other. It appeared
that the midgut was extremely vulnerable to a
deficiency of food and usually after 1 day of starva-
tion, single mucosal cells could be seen sloughed
(9) into the lumen. The margin of the lumen con-
tinued to lose its integrity as starvation advanced.
Cells of the hindgut of feeding larvae exhibited an
accumulation of eosin staining inclusions. The in-
clusion bodies, which may be the sites of intracel-
lular digestion, have been observed in other
marine and freshwater fish (Kostomarova 1962;
Iwai and Tanaka 1968a, b; Iwai 1968; Umeda and
Ochiai 1975; O'Connell 1976). The amount, size,
and intensity of the staining (10) of these inclu-
sions varied in feeding larvae. They were not pres-
ent in starving larvae.
Musculature (Figures 9, 10, 12)
In feeding jack mackerel larvae, individual
muscle fibers were close together (11); they were
composed of closely packed, striated, and parallel
myofibrils. After a period of starvation, the fibers
separated, the fibrils were not distinct, and occa-
sionally they lost their parallel structure. Be-
tween some fibers there was a granular,
basophilic, nucleated substance called "intermus-
cular tissue" (12) by O'Connell (1976). In starved
larvae, this tissue was usually absent.
General Histological Characteristics
After jack mackerel larvae had starved for 3
days, signs of depletion were widespread. In addi-
tion to changes in major tissues and organs there
was a general atrophy and disintegration of all
cells and tissues including those of cartilage, kid-
ney, endocrine pancreas, and swim bladder. The
number of pyknotic nuclei (i.e., darkening and
shrinking nuclei, which give the first indication
that a cell is dying) increased in all tissues (see eye
and brain, Figure 4). Epithelial cells of the in-
tegument were hypertrophic, twice as large as
normal in 3-day starved larvae (Figure 10), and
kidney tubules were swollen (Figure 6). There was
always a larger yolk reserve retained by starving
larvae (Figure 6). A decrease in yolk absorption in
starving larvae was also reported by Kostomarova
(1962) for pike and carp and by Umeda and Ochiai
(1975) for yellowtail.
Histological Grading
To determine whether the classification of a jack
mackerel larva required the grading of all his-
tological features or a lesser number, a group of 27
408
larvae, 14 feeding and 13 starving, was examined
and the resulting grades for each criterion were
submitted to a SWDA. The experimental treat-
ment (fed or starved) was unknown until after all
larvae were microscopically examined. The larvae
were 7 days old and had been feeding or starving
for 2 days. The grading system classified all fed
larvae (n = 14) into the healthy group (individual
average grade of the 12 histological features
ranged between 2.42 and 2.92). The average
grades for the 2-day starved larvae were more
variable. The larvae were classified, about
equally, into each of the three nutritional groups:
four had a grade range between 2.35 and 2.54,
ranking in the healthy group; four were classified
as intermediate, grade range 2.08 to 2.31; and five
larvae were ranked as starved with the average
grades ranging between 1.15 and 1.54.
Results of the SWDA on the above data disclosed
that grading only two histological characteristics,
the arrangement of the cells in the pancreas (vari-
able 7 ) and the sloughing of mucosal cells from the
midgut (variable 9), gave the same conclusions as
using all 12 features. Therefore, in all subsequent
histological assessments, the average grade of
these two criteria, variables 7 and 9, was used as
the index of larval condition.
Morphological
The jack mackerel larvae were 2.45 mm SL (pre-
served) at hatching and initiated feeding at 3.35
mm, 5 or 6 days after hatching (hatching = day 0,
Figure 13A). Atthetimeof first feeding, some yolk
and oil were present but the yolk sac was not
discernible. The relationship between the five
morphological characteristics, measured to de-
termine the effects of starvation, and days of star-
vation is illustrated in Figure 13B-F. Since no
data have been published on the daily growth rate
of field-caught jack mackerel larvae, I used length
as an estimate of age. When the morphometric
measurements were plotted against length, no
single measurement was a reliable index of star-
vation, as illustrated by pectoral body depth plot-
ted for fed and starved larvae (Figure 14). How-
ever, some limits can be set from this graph: 1) all
larvae <3.30 mm SL that do not have a yolk sac
probably are starving (feeding is initiated at 3.35
mm); and 2) larvae with a body depth >0.47 mm
are feeding. This leaves the size class between 3.30
and 3.55 mm where the cases cannot be separated.
Most individuals in this class (29 fed and 24
THEILACKER; EFFECT OF STARVATION ON JACK MACKEREL
4.2^ A
• FED
A STARVED
0.60
E
E 0.70
X
Q- 0.601-
UJ
o
Q 0.50-
o
iij
a.
0.40-
0.30
6 7 8 9 10
h-^-^
6 7 8 9
DAYS
10
'g 0.35
E
ui
ijj
>-
UJ
0.30-
<
Q 0.25
0.20
8
10
Figure 13. — A. Growth ofTrachurus symmetricus larvae, with means and standard deviations. Sample size on any day was usually 15.
B-F. Relationship between five morphological characteristics of T. symmetricus larvae and days of starvation, with means and standard
deviations. Sample size was usually 15.
starved) have been feeding or starving for 1 or 2
days.
To determine whether a set of several mor-
phometric variables could predict the condition of
larvae in the 3.30 to 3.55 mm size class, a SWDA
was run. Eleven morphometric variables, in which
the two predetermined groups (fed and starved)
were expected to differ, were entered into the
SWDA: 1) HL, 2) ed, 3) bd-1, 4) bd-2, 5) HL/SL, 6)
ed/SL, 7) bd-l/SL, 8) bd-2/SL, 9) ed/HL, 10) bd-1/
409
FISHERY BULLETIN: VOL. 76, NO. 2
Figure 14. — Relationship between
pectoral body depth and standard
length of Trachurus symmetricus lar-
vae which were fed (open circles),
starved 1 and 2 days (dots), and starved
3 days (x's).
6
E
bJ
Q
>-
Q
O
CD
<
cr
o
I-
UJ
Q.
U./'S
070
-
0 0
o
065
—
0
0
060
-
0
00 0 OCD
O 0 0 0
0
055
-
0 OCD 0
0
00
050
-
X
0 CDOOA O
0
0
Oflftfto 0 CDOOO ^
0 aco
045
X
X
X
X
X
>«-perplanei de-
rived by describing these regions as multivariate
normal density distributions with common
variance-covariance matrices i Welch 1939 ».
Quadratic discriminant functions have been de-
veloped ' Smith 1947 '. The resulting decision sur-
faces are nonlinear. The quadratic discriminant
function does not require common variance-
covariance matrices. Anas and Murai > 1969 > com-
pared the classificatory abilities of the linear and
quadratic discriminant functions. They found ' in
agreement with Isaacson 1954' that even if the
assumption that the distributions have common
variance-covariance matrices is violated, the
linear discriminamt function would still give good
results for large sample sizes. But the quadratic
function gave slightly bener results.
All investigators utilizing discriminant
analyses to separate races of Pacific salmon have
assumed that the density distributions of mea-
surements from a particular class of salmon were
multivariate normal. The frequency distributions
of scale characters in Major et al. ' 1975 > show that
multimodal and skewed distributions occilt for
chinook salmon scale characters even in the uni-
variate case. In many other cases, the underh-ing
distribution functions may be non-Gaussian. Dis-
criminant functions based upon non-Gaussian dis-
tributions or obtained by distribution-free
methods are preferable to those based upon an
unrealized assumption of normality.
Nearly all of the discriminant function analyses
used in the investigations of Pacific salmon have
been two-class analyses designed to determine the
continent of origin of salmon taken on the high
seas. For the two-class situation only one discrim-
inant function need be calculated. These two-class
problems are a special case of the many-class prob-
lems in which a separate discriminant function is
calculated for each class. Bilton and Messinger
<1975> calculated discriminant functions for each
of several runs in a classification study on sockeye
salmon. If several stocks of salmon intermingle
and are to be classified, analyses of this t>-pe are
needed.
Spechts 1966' polynomial discriminant
method does not require that the underhing den-
sity distributions be multivariate normal nor that
they have common variance-covariance matrices.
Since this method is nonparametric. various scale
characteristics may be used for discrimination
with no particular regard to the underlying dis-
tributions. Thus, the method is flexible and practi-
cal.
Specht 1966' uses an estimated probability
density function of the form described by Parzen
1 1962 1 and extended by Murthy > 1966 • to the mul-
tivariate case. The underh"ing multivariate den-
sity for each class is modeled by a sum of functions
that are multivariate Gaussian in form, one such
function for each fish in the learning sample for
that class. This set of functions is complete. There-
fore, for each class the underhnng continuous
probability density. Gaussian or not, may be ap-
proximated arbitrarily closely by such a sum. A
power series expansion of this estimated denisty
then results in a pohTiomial term in the density
function, the coefficients of which are functions of
the obser\"ations i fish i in the learning sample. One
such set of coefficients is computed for each class to
be considered. These polynomials determine the
nonlinear decision surfaces and are the basis for
discrimination.
The indi\'idual multivariate Gaussian functions
I which when summed model the underlying mid-
tivariate distribution for that class' contain a
"'smoothing parameter." cr. which appears in the
place of a standard error. This parameter is then
incorporated in the estimates of the pol\*nomial
coefficients. The reader is referred to Specht ' 1966 1
for a discussion of the effect of this smoothing
parameter and for the algorithm for the calcula-
tion of the sets of pohTiomial coefficients { -Dfe ■
k- ... fefj ( • The f>olynomial discriminant func-
tion is:
P(X) = Do ^ Z)iXi - D0X2 -
DpX,
'11^1
D^^X■^~ ' . . . + Df^^f^^Xf^^Xf^^
D^pX\ ^ 1^11 A'
+ . . . + Dfe^;;2fe3^fei^fe2'^fe3
"^ DpppX p
Dk.
kh ^fei
-^fei
416
COOK and LORD: IDENTIFICATION OF STOCKS OF SOCKE'i'E SALMON
where p = dimension of the vector X (set of scale
characters )
1 h,P' (X)
for all s =f r
where d(X)
p'iX)
/2; =
the decision on an unknown X
the classes (origins)
the polynomial value for X
calculated using the discrimi-
nant function for class B;
the a priori probability, the
uses of which will be de-
scribed later.
APPLICATION OF THE METHOD
Three scale sample sets are required to imple-
ment the polynomial discriminant method: learn-
ing samples, test samples, and onknowTi samples.
The learning and testing samples are collected
from each subpopulation when they are segre-
gated (i.e.. in the rivers of origin i. Scale characters
to be measured in the unknown sample for the
required discrimination are determined by
evaluating characters measured in the learning
samples. The learning samples and the characters
selected are used to calculate the coefficients in the
pohTiomial discriminant functions. To calculate
these coefficients, the value for the smoothing
parameter and the point at which the discrimin-
ant function should be truncated must be deter-
mined. Various circumstances will dictate differ-
ent choices. When a smoothing parameter of 1.5
was chosen, all terms in the discriminant function
greater than the fourth order contributed negligi-
bly to polynomial values and so were truncated in
our applications. Often, polynomial discriminant
functions of lower order yield adequate results.^
*A poKiiomial discriminant function with six variables and of
the fourth order will contain 210 terms. Since our calculations
were performed by computer, we chose not to delete the third or
fourth degree terms. However, if more than six vso-iables are
used, it would be wise to truncate further in order to keep the
number of terms down.
The fish comprising the test samples are classified
to test the effectiveness of the polynomial discrim-
inant method and to determine the a priori prob-
abilities. 'Each test sample consists offish from
one class. I Finally, fish collected from the zone of
intermingling are classified to determine the de-
gree of intermingling in the area of interest.
Appraisal of the method using scale samples of
sockeye salmon collected from the 1967 escape-
ment in five Bristol Bay rivers showed large per-
centages of fish comprising the test samples were
correcth- classified. However, misclassified fish in
the test group • set of test samples from all rivers
being considered i were not assigned to the rivers
in proportion to the known relative test sample
sizes. To balance these misclassifications. wher-
ever a greater number of fish comprising the test
group was assigned to a particular river than
should have been i according to the relative test
sample sizes ). the a priori probability for that river
w-as lowered. Corresponding increases were made
for those classes with insufficient assignment. By
alternatively using the decision procedure of the
polynomial discriminant method and adjusting
the a priori probabilities, we obtained solutions so
that the number offish belonging to a certain river
that were misassigned to all other rivers approxi-
mately equaled the number of fish misassigned to
that certain river from all other rivers. Thus the a
priori probabilities were not used in the manner
their name suggests, but a priori knowledge may
dictate test sample sizes. The relative test sample
sizes in the test group may be in the relative pro-
portions to be expected in the unknown sample
( i.e.. historical relative run sizes >. The adjustment
procedure, then, shifts the nonlinear decision sur-
faces between the probability- densities so that the
incorrectly identified samples are assigned to the
various rivers in the proportions dictated by the
test sample sizes in the test group. However, the
primary purpose of the adjustment procedtire is
not to balance the misclassifications but to
maximize the number of correct classifications. As
the misclassifications are balanced, the number of
correct classifications generally increases. At this
point the result is a classification method that
maximizes the total number of correct classifica-
tions and balances misclassification rates for a test
group in which the test sample sizes are in particu-
lar proportions.
However, it is obvious that the proportions of
fish from the various classes in the test group
would rareh- be identical to those proportions in
417
FISHERY BULLETIN: VOL. 76, NO. 2
the unknown sample. Thus, imbalance among the
misclassified fish will recur, unless the expected
accuracy of classification is very good (near 100%).
We have devised a method to correct for this.
Based upon the results of classification of the
known test group, the classification matrix, C, is
estimated:
C =
C\\ Ci2
21 •'22
- C„i C„2
In
2"
c
where c,j is an estimate of the fraction of fish
allocated to class i belonging to class j, such
that 2 Cjj = 1.0, Vy. (Note that for each 7 the
i = i
c,j 's are a set of estimated multinomial prob-
abilities and that each test sample size should be
adequate.) If the discrimination is error-free, C
would be an identity matrix. The adjustment of a
priori probabilities causes the initially estimated
classification matrix to evolve to the point where
CT =Rf such thatT ^R,.
The ith component of the vector T is the fraction of
fish in the test group from test sample i (class /),
and the ith component of the vector J?, is the
fraction offish in the test group allocated to class i
by the adjusted polynomial discriminant method.
The test samples comprising T are not indepen-
dent of the classification scheme since they are
used to determine the a priori probabilities used in
the decision rule. Hence, the estimated prob-
abilities in the classification matrix may not be
unbiased. However, we did chi-square tests that
show elements of the classification matrix are not
significantly different when estimated with either
the test samples used to determine the a priori
probabilities or a second independent test group.
Thus, we prefer to use only one test group to de-
termine the a priori probabilities and to estimate
the elements of the classification matrix because
the test sample sizes will be larger (and the var-
iance of the Cy's smaller) if we do not subdivide
the fish available.
Now, let u , be the fraction of fish in a sampled
group that belong to the ith class. The vector U is
then unknown except for the obvious side condi-
tion 2 u,- = 1. The classification matrix now
; = 1
operates on U to give:
CU =i?/
where the ith component of/?,, is the fraction of
fish in the unknown sample allocated to the ith
class. Since C is estimated, R ,, is known and since
C is usually nonsingular, we can estimate U by
U =C ' R^.
Each point estimate iu,) obtained will have some
variance. This variability will depend upon the
accuracy with which fish from class i are classified,
the accuracy with which the elements of C are
estimated, and variance due to sampling error en-
countered when obtaining the unknown sample.
Thus, if any u , is small, then its estimate ( u , ) may
be negative. Such solutions are meaningless. In
such cases the classes with negative solutions
should be dropped (assume such u, ~ 0) and the
analyses repeated.
We did simulation work to evaluate the classi-
fication matrix correction procedure for the two-
and three-class situations. Five hundred simu-
lated experiments were done for each situation.
For the two-class case the average error of the
classification results was 0.100 while that of the
corrected estimates was 0.055. In 84% of the exper-
iments the corrected estimate was closer to the
true value than classification result. For the
three-class case the average error of the classifica-
tion results was 0.127 while that of the corrected
estimates was 0.054. In 89% of the experiments
the corrected estimate was closest to the true
value. The results of these simulations show that
the classification correction procedure improves
estimates of the true proportion of a class present.
This classification matrix correction procedure
will reduce to the correction procedure developed
for the two-class case by Worlund'' in the following
manner:
*A similar relationship and a least squares solution technique
is given by Worlund and Fredin (1962).
■'Worlund, D. D. 1960. A method for computing the variance of
an estimate of the rate of intermingling of two salmon popula-
tions. Unpubl. manuscr., 13 p. Bur. Commer. Fish., Biol. Lab.,
Seattle, Wash.
418
COOK and LORD: IDENTIFICATION OF STOCKS OF SOCKEYE SALMON
-1
U = C" R
or
Generally
Since
substitution yields u^ =
■"l"
'C22''l -
Cl2'-2
C\\Ci2 -
^21^12
"2
Cll'-2 -
C2l'"l
/11C22 ~
C21C12
=
'i^ii -
^un
"«
CiiCjj -
CjiCij '
0
= I-'-/,
Cji
= 1 - Cii ,
Cjj
= 1 - Cij ,
H<; ;/ .
n - Cij
Cii
which is the correction formula of Worlund and
Fredin (1962) (except for differences in notation
and terminology) that has been used in many
two-class Pacific salmon stock identification
studies.
Application to Sockeye Salmon Samples
Taken in High Seas Sampling
A problem of interest to the nations bordering
the North Pacific Ocean is the origin of sockeye
salmon taken on the high seas. The rivers of origin
of sockeye salmon south of the central Aleutian
Islands in summer are of particular interest to the
United States since an index of their overall rela-
tive abundance is used to forecast the numbers of
mature fish returning to Bristol Bay in the follow-
ing year (Rogers 1975). These fish are primarily of
Bristol Bay origin (Hartt 1962, 1966; Hartt et al.
1975). Knowledge of the relative abundance of the
various runs of the Bristol Bay stock south of the
central Aleutians would be useful for forecast pur-
poses and might provide insight into the high seas
life history of the various runs.
In order to recognize age 2.2 immature sockeye
salmon on the high seas in 1976, the freshwater
growth patterns of scales from three of the major
rivers in Bristol Bay were examined.® Scales from
the smolt outmigrations of 1974 for the Kvichak
and Naknek Rivers were used as learning and
*Age designation indicates fish which migrated to sea after
two winters in freshwater and have spent two winters at sea.
They are expected to return from the ocean primarily at age 2.3,
or after sf)ending three winters at sea.
testing samples. For the Egegik River scales from
age 2.2 adult fish returning to spawn in 1976 were
used as learning and testing samples because
smolt scales were unavailable. The freshwater
scale patterns offish from these runs were used to
classify the sockeye salmon captured south of
Adak Island during summer 1976 after having
spent two winters in the ocean.
The scale patterns were examined under a mi-
croprojector of the type described by Dahlberg and
Phinney (1968). The widths of the summer,
winter, and plus growth zones were measured in
terms of circuli counts and distance. The width of
the widest circulus was also measured. Each scale
character was then ranked over all classes (rivers)
and the Kruskal-Wallis statistic (Kruskal and
Wallis 1952) calculated. The difference between
the average sum of ranks for each pairwise class
combination was also calculated. On the basis of
these statistics the scale characters providing the
best univariate separation were selected for use in
the polynomial discriminant method. Highly de-
pendent scale characters were not used.
By examining the learning samples, six scale
characteristics were chosen for use in the polyno-
mial discriminant method: 1) The number of the
circuli in the first winter growth zone, 2) the
number of circuli in the second summer growth
zone, 3) the number of circuli in the plus growth
zone, 4) the width of the first summer growth zone,
5) the width of the second winter growth zone, and
6) the width of the widest circulus.® Learning
sample sizes of 25, 25, and 24 for the Egegik,
Kvichak, and Naknek River classes, respectively,
were used to calculate the coefficients in the
polynomial function for each class. The classi-
ficatory ability of these functions was then tested.
The relative test sample sizes for each class were
determined by examining run size data. According
to the average run sizes of age 2.3 salmon for the
last 8 yr approximately equal numbers offish from
each class were expected to occur in the unknown
sample. However, since the Kvichak River test
sample size was twice that of the Egegik or Nak-
nek River sample size, the fish in the latter test
samples were given a weight of 2 when the a priori
'It should be mentioned that all data points were "nor-
malized." That is, the mean and standard deviation for each scale
character were calculated from the learning samples (all
categories combined). All data points were then transformed by
subtracting off" the mean and dividing by the standard deviation
for the appropriate scale character. This is done for numerical
purposes.
419
FISHERY BULLETIN: VOL. 76, NO. 2
probabilities were adjusted. After adjusting the a
priori probabilities, we obtained the results given
in Table 1. The classification matrix was then es-
timated:
C =
0.800 0.040 0.167
0.080 0.740 0.208
0.120 0.220 0.625
where the subscripts of the matrix elements ( c,^'s)
were 1, 2, and 3 for the Egegik, Kvichak, and
Naknek River classes, respectively. Seventy-two
percent of the fish in the test group were correctly
classified. The fish in the high seas sample were
then classified with the adjusted polynomial dis-
criminant method.
Of the 101 sockeye salmon, 25 were classified as
Egegik River fish, 22 as Kvichak River fish, and 54
as Naknek River fish. The resultant vector was:
Ru =
0.267
0.222
0.511
The estimated unknown vector was thus:
C~' R.. =
1.300 0.037
-0.360
0.267
-0.078 1.498
-0.478
0.222
-0.222 -0.534
1.837
0.511
0.17f
0.067
= U.
0.761
Based upon preliminary data for the 1977 Bristol
Bay sockeye salmon run from the Alaska Depart-
ment of Fish and Game, the actual unknown vec-
tor was:
U
0.325
0.061
0.614
The classification matrix correction procedure
gave a slightly better estimate than the direct
results of the polynomial discriminant method.
The differences between the u, 's and the u , 's were
due to bias and variability. (We are presently
examining methods to reduce the variability of
our i/,'s.)
A problem with the high seas sample is that
some of these sockeye salmon originate in rivers
other than those considered. Although the three
Table l. — Results of the polynomial discriminant method on a
known test group of Bristol Bay sockeye salmon. The a priori
probabilities were 0.340, 0.332, and 0.328 for the Egegik,
Kvichak, and Naknek River classes, respectively.
Calculated
Correct decisions
Total (all calcu-
decisions
Egegik
Kvichak
Naknek
lated decisions)
Egegik
Kvichak
Naknek
Total (all
correct
decisions)
40
4
6
50
2
37
11
50
8
10
30
48
50
51
47
148
classes considered will account for nearly all of the
age 2.2 sockeye salmon bound for Bristol Bay,
some may be non-Bristol Bay fish. When the Bris-
tol Bay runs are at a low point in their cycle, up to
20% of the high seas sockeye salmon at Adak Is-
land may be non-Bristol Bay fish (Hartt et al.
1975). The possible bias from classifying the non-
Bristol Bay fish into the classes established should
be considered since 1977 is a low year in the sock-
eye salmon run cycle.
In conclusion, the polynomial discriminant
method can be used to identify certain runs of
sockeye salmon on the high seas by differences in
freshwater scale growth patterns. Possibly the
relative proportions of sockeye salmon that will be
returning to inshore areas can be predicted. Even-
tually the method will be used to predict one year
in advance the relative run sizes to the major Bris-
tol Bay rivers by sampling these sockeye on the
high seas.
Application to
Inshore Fishery Stock Separation
A problem of interest to the Alaska Department
of Fish and Game is the separation of stocks in
commercial catches in inshore areas, particularly
the separation of Kvichak, Naknek, and Egegik
River sockeye salmon. The Division of Commer-
cial Fisheries is collecting data on scale measure-
ments for growth studies. They are interested in
how well these data and the polynomial discri-
minant method can separate Bristol Bay sockeye
salmon stocks.
Scale data from samples of the 1973 spawning
escapement were examined. Each of two age-
classes was examined separately. Distance and
circuli counts to both the freshwater and saltwater
annuli were examined for use in the polynomial
discriminant method with the Kruskal-Wallis and
multiple comparison procedures. The accuracy of
classification for age 1.2 and age 2.2 sockeye salm-
420
COOK and LORD: IDENTIFICATION OF STOCKS OF SOCKEYE SALMON
on was examined for each age-group with known
test groups.
The degree of separation for age 1.2 sockeye
salmon is shown in Table 2. (Egegik River fish are
historically insignificant in this age-class.) The
scale characters providing this separation were: 1)
the circuli count to the first annulus, 2) the dis-
tance to the first annulus, 3) the distance from the
first to the second annulus, 4) the distance from
the second to the third annulus, 5) the circuli count
from the third annulus to the edge of the scale, and
6) the distance from the third annulus to the edge
of the scale. Ninety-five percent of the fish in the
test group were correctly classified.
The degree of separation for age 2.2 sockeye
salmon is shown in Table 3. The scale characters
providing this separation were: 1) the circuli count
to the first annulus, 2) the distance to the first
annulus, 3) the circuli count from the first to the
second annulus, 4) the distance from the second to
the third annulus, 5) the distance from the third to
the fourth annulus, and 6) the circuli count from
the fourth annulus to the edge of the scale.
Seventy-seven percent of the fish in the test group
were correctly classified.
Thus, the polynomial discriminant method can
provide adequate separation with a given data
base. The data collected for growth studies provide
good separation in some cases. Sockeye salmon
from the Egegik, Kvichak, and Naknek Rivers are
distinguishable in terms of these scale measure-
ments and it should be possible to estimate their
relative proportions in catch samples.
Table 2 . — Results of the polynomial discriminant method on 1 . 2
age Bristol Bay sockeye salmon from 1973. The a priori prob-
abilities were 0.52 and 0.48 for the Kvichak and Naknek River
classes, respectively.
Calculated
Correct decisi
ons
Total (all calcu-
decisions
Kvichak
Naknek
lated decisions)
Kvichak
Naknek
Total (all correct
decisions)
18
2
20
0
19
19
18
21
39
T.\BLE 3. — Results of the polynomial discriminant method on 2.2
age Bristol Bay sockeye salmon from 1973. The a priori prob-
abilities were 0.342, 0.330, and 0.328 for the Egegik, Kvichak,
and Naknek River classes, respectively.
Calculated
Correct decisions
Total (all calcu-
decisions
Egegik
Kvichak
Naknek
lated decisions)
Egegik
Kvichak
Naknek
Total (all
correct
decisions)
20
1
5
26
3
22
1
26
3
4
14
21
26
27
20
73
COMMENTS AND CONCLUSIONS
The key to successful implementation of the
polynomial discriminant method is the choice of
scale characters that reflect differences between
the subpopulations of concern. The scale charac-
ters that are most likely different are those that
are formed when the populations are geographi-
cally separated. Genetic and environmental
influences on scale formation probably interact to
create these differences. Although it is likely that
no single characteristic will provide the required
separation, a group of characteristics analyzed
with multivariate techniques ( e.g. , the polynomial
discriminant method) will often provide this re-
quired separation. The polynomial discriminant
function technique requires no consideration of
the underlying probability density functions for
these scale characters because these density func-
tions are estimated nonparametrically. Once the
characters that provide the best separation are
determined (by rank order comparison procedures
in this paper) the discriminant function analysis
may be implemented.
A learning sample is needed to calculate the
discriminant function for each subpopulation.
These fish comprising these samples must be col-
lected before or after the populations intermingle
(either as smolts or returning adults in the respec-
tive rivers). Learning samples must be taken from
the same year class and freshwater age-group as
the unknown (mixed) population if the scale
characters are known or thought to vary from year
to year. Using Specht's (1966) algorithm and the
data from these learning samples, the coefficients
in the discriminant functions are calculated. The
next step is to appraise the effectiveness of these
polynomial discriminant functions.
By classifying a group of test samples the pro-
portion of correctly identified fish and the clas-
sification error rates can be determined. The pro-
portion of correctly identified fish will likely be low
until a good set of a priori probabilities is deter-
mined. As the a priori probabilities are adjusted to
balance the classification error rates, the propor-
tion of correctly identified fish will generally in-
crease. The proportion of correctly identified fish,
when the classification error rates are satisfactor-
ily balanced, gives an indicator of the effectiveness
of the polynomial discriminant method. The clas-
sification error rates specific to these final a priori
probabilities are now estimated so that they may
be corrected for when the polynomial discriminant
421
method is applied to the unknown mixed sample.
This is done with the classification matrix correc-
tion procedure.
First, the fish in the unknown mixed sample are
classified with the polynomial discriminant
method (using the adjusted a priori probabilities).
The proportions resulting for each subpopulation
and the decision matrix allow simple algebraic
solution for the estimated true proportions of the
various subpopulations in the zones of interming-
ling.
Estimates of this type are often needed in par-
ticular management situations involving Pacific
salmon. By using scale samples and the polyno-
mial discriminant method, the proportions of the
major classes present in areas where the subpopu-
lations mix can be estimated. We have considered
only two possible applications in this paper: high
seas monitoring for predictive purposes and the
analysis of catch samples. Many other possibilities
exist for other situations and other salmon species:
the timing of inshore runs could be examined in
estuarine areas or in river systems, the continent
of origin of salmon on the high seas could be
examined (for those species or areas not already
analyzed), or the intermingling of hatchery and
native populations could be analyzed for certain
fisheries. Since scale samples are relatively easy
to collect and exchange and since computers are
readily available to do the necessary calculations,
the polynomial discriminant method is a flexible
and practical tool for the racial analysis of Pacific
salmon, particularly sockeye salmon.
ACKNOWLEDGMENTS
Many thanks are due to Colin Harris, Allan C.
Hartt, and Robert L. Burgner for their editorial
advice and guidance. We also wish to thank James
B. Scott for his meticulous work with the scale
data. Many others providing services to the deep
sea tagging project at the Fisheries Research In-
stitute deserve thanks. The scale samples and
data from the Bristol Bay river systems were col-
lected by the Alaska Department of Fish and
Game. We are grateful to Paul Krasnowski and
other Alaska Department of Fish and Game per-
sonnel for providing these vital samples. This re-
search was primarily supported by NOAA, Na-
tional Marine Fisheries Service, under Contract
No. 03-6-208-35470. The inshore stock separation
project was supported by the Alaska Department
of Fish and Game.
FISHERY BULLETIN: VOL. 76, NO. 2
LITERATURE CITED
Amos, M. H., R. E. Anas, and R. E. Pearson.
1963. Use of discriminant function in the morphological
separation of Asian and North American races of pink
salmon, Oncorhynchus gorbuscha (Walbaum). Int.
North Pac. Fish. Comm., Bull. 11:73-100.
Anas, R. E.
1964, Sockeye salmon scale studies. Int. North Pac. Fish.
Comm., Annu. Rep. 1963:158-162.
Anas, R. E., and S. Mural
1969. Use of scale characters and a discriminant function
for classifying sockeye salmon (.Oncorhynchus nerka) by
continent of origin. Int. North Pac. Fish. Comm., Bull.
26:157-192.
BILTON, H. T., AND H. B. MESSINGER.
1975. Identification of major British Columbia and Alaska
runs of age 1.2 and 1.3 sockeye from their scale characters
Int. North Pac. Fish. Comm., Bull. 32:109-129.
DAHLBERG, M. L., AND D. E. PHINNEY.
1968. A microprojector for use in scale studies. Prog.
Fish-Cult. 30:118-120.
Dark, T. A., and B. J. Landrum.
1964. Analysis of 1961 red salmon morphological da-
ta. Int. North Pac. Fish. Comm., Annu. Rep. 1962:110-
115.
Das Gupta, S.
1973. Theories and methods in classification: A review.
In T. Cacoullos (editor), Discriminant analysis and appli-
cations, p. 77-137. Academic Press, N.Y.
Fisher, R. A.
1936. The use of miltiple measurements in taxonomic
problems. Ann. Eugen. 7:179-188.
Fukuhara, F. M., S. Murai, J. J. LaLanne, and a.
Sribhibhadh.
1962. Continental origin of red salmon as determined from
morphological characters. Int. North Pac. Fish. Comm.,
Bull. 8:15-109.
Hartt, A. C.
1962. Movement of salmon in the North Pacific Ocean and
Bering Sea as determined by tagging, 1956-1958. Int.
North Pac. Fish. Comm., Bull. 6, 157 p.
1 966. Migrations of salmon in the North Pacific Ocean and
Bering Sea as determined by seining and tagging, 1959-
1960. Int. North Pac. Fish. Comm., Bull. 19, 141 p.
Hartt, A. C, G. E. Lord, and D. E. Rogers.
1975. Monitoring migrations and abundance of salmon at
sea. Int. North Pac. Fish. Comm., Annu. Rep.
1973:73-79.
ISSACSON, S. L.
1954. Problems in classifying populations. In O.
Kempthome, T. A. Bancroft, J. W. Gowen, and J. L. Lush
(editors), Statistics and mathematics in biology, p. 107-
117. Iowa State Coll. Press, Ames.
KONOVALOV, S. M.
1971. Differentiation of local populations of sockeye salm-
on Oncorhynchus nerka (Walbaum). (Translated by L. V.
Sagen, 1975. Univ. Wash. Pub. Fish., New Ser., 6, 290 p.)
KRUSKAL, W. H., AND W. A. WALLIS.
1952. Use of ranks in one-criterion variance analysis. J.
Am. Stat. Assoc. 47:583-621.
Major, R. L., S. Murai, and J. Lyons.
1 975. Scale studies to identify Asian and western Alaskan
422
COOK and LORD: IDENTIFICATION OF STOCKS OF SOCKEYE SALMON
Chinook salmon. Int. North Pac. Fish. Comm., Annu.
Rep. 1973:80-91.
Mason, J. E.
1966. Sockeye salmon scale studies. Int. North Pac. Fish.
Comm., Annu. Rep. 1964:117-118.
Ml'RTHY, V. K.
1966. Nonparametric estimation of multivariate densities
with applications. In P. R. Krishnaiah (editor), Mul-
tivariate analysis, p. 43-56. Academic Press, N.Y.
Parzen, E.
1962. On estimation of a probability density function and
mode. Ann. Math. Stat. 33:1065-1076.
Patrick, E. a.
1972. Fundamentals of pattern recognition. Prentice-
Hall, Englewood Cliffs, N.J., 504 p.
ROGERS, D. E.
1975. Forecast of the sockeye salmon run to Bristol Bay in
1976. Univ. Wash., Fish. Res. Inst. Circ. 76-1, 49 p.
Smith, C. A. B.
1947. Some examples of discrimination. Ann. Eugen.
13:272-282.
SPECHT, D. F.
1966. Generation of polynomial discriminant functions for
pattern recognition. Stanford Univ., Tech. Rep. 6764-5,
127 p.
Welch, B. L.
1939. Note on discriminant functions. Biometrika
31:218-220.
Worlund, d. D., and r. a. FREDIN.
1962. Differentiation of stocks. In Symposium on pink
salmon, p. 143-153. H. R. MacMillan Lectures in
Fisheries, Univ. B.C., Vancouver, Can.
423
EFFECTIVENESS OF ESCAPE VENT SHAPE IN TRAPS FOR
CATCHING LEGAL-SIZED LOBSTER, HOMARUS AMERICANUS,
AND HARVESTABLE-SIZED CRABS, CANCER BOREALIS
AND CANCER IRRORATUS^
Jay S. Krouse^
ABSTRACT
During 1976 a study was conducted to find an escape vent that would select similar sized lobsters as the
rectangular vent, yet retain Cancer crabs s90 mm carapace wddth. Analysis of the size composition of
research and commercial catches from experimental traps revealed that circular (58 mm in diameter)
and rectangular (44.5 x 152.4 mm) vents release shorts and retain legal lobsters ( 3^81 mm carapace
length) equally well, and decidedly more marketable-sized crabs were captured in traps with circular
vents. Length-width relationship shows that crabs 3=90 mm carapace width have lengths 3^58 mm,
thus precluding the possibility of marketable-sized crabs exiting through an opening 58 mm in
diameter. Escapement studies for lobsters confirm that with the present minimum legal size of 3^/i6 in,
a 58-mm diameter vent vnll select legals and allow most of the sublegals to escape.
Accordingly, the Maine Department of Marine Resources recommends that either circular ( 358 mm
in diameter) or oblong (3=44.5 x 152.4 mm) escape vents be incorporated in all crab and lobster traps
along the Maine coast.
Although rectangular escape vents are a very
beneficial type of savings gear for the lobster
fishery (Templeman 1939; Wilder 1945, 1948,
1954; Krouse and Thomas 1975; Krouse 1976),
this vent does not retain marketable-sized rock
crab, Cancer irroratus, and Jonah crab, C.
borealis. Since these commercially important crab
species are often caught incidental to lobsters, I
undertook the present study to find an escape
opening that would retain harvestable-sized crabs
and have similar fishing selectivities for the lob-
ster, Homarus americanus, as the rectangular
vent.
In designing a trap to catch crabs and exclude
lobsters, Stasko (1975) observed in laboratory
tests that circular holes retained commercial-
sized crabs yet allowed small lobsters to escape;
however, the effectiveness of escape holes was not
tested in the field. Jow (1961) demonstrated the
advantages of circular escape openings in the trap
fishery for Dungeness crab, C. magister.
In this paper I evaluate the relative efficiency of
'This study was conducted in cooperation with the U.S. De-
partment of Commerce, National Marine Fisheries Service,
under Public Law 88-309, as amended, Commercial Fisheries
Research and Development Act, Project 3-153-R.
^Maine Department of Marine Resources, West Boothbay
Harbor, ME 04575.
circular and rectangular vents by examining data
from: 1) commercial and research catches com-
piled from vented and nonvented traps; 2) studies
of escapement from traps; and 3) certain mor-
phometric relationships of crabs and lobsters.
METHODS
From November 1976 through March 1977 a
commercial fisherman recorded and provided me
with catch data from traps with circular vents
(58-mm diameter) fished alongside traps without
vents. This experimental gear was arranged into
two groups with four trawls [series of six traps
spaced about 6 fathoms ( 11.0 m) apart with a sur-
face buoy at either end] per group. In each group
half the traps in a trawl had no vents, while the
remainder had either single (end of trap) or paired
(side of trap) vents depending upon the group (Fig-
ure IB, D). Every time the fisherman hauled
these traps he recorded the following information:
1 ) number of days traps were set between hauls;
and 2) number of lobsters &81 mm carapace
length, CL (keepers), and <81 mm CL (shorts)
caught in the vented and nonvented traps for each
trawl string.
From July through November 1976, project per-
sonnel fished commercial lobster traps near
Manuscript accepted July 1977
FISHERY BULLETIN: VOL. 76, NO. 2, 1978.
425
FISHERY BULLETIN: VOL. 76, NO. 2
7>\
^
D
Figure l. — Lobster traps having a rectangular vent positioned vertically (A) and horizontally (C) and single (B) and paired (D) circulai
vents.
Boothbay Harbor, Maine, with: 1) circular [58-mm
(2.3 in) and 61-mm (2.4 in) diameter] vents; 2)
vertical and horizontal rectangular [44.5 mm ( 1.8
in) X 152.4 mm (6.0 in)] vents; and 3) traps with-
out vents. Carapace length of lobsters was mea-
sured from posterodorsal edge of eye socket to
posterior margin of carapace and carapace width
(CW) of crabs, distance between the two m^ost pos-
terior notches on the anterolateral border of the
carapace, to the nearest millimeter.
Trap escapement was studied by placing
lobsters of known sizes in traps with circular open-
ings of 58, 60, and 61 mm in diameter. Side en-
trances of each trap were closed so escapement had
to be via the vents. Traps were secured to the
laboratory dock and usually checked daily for es-
capement for about a week.
To determine whether or not a crab or lobster
could pass through a round opening of a given size,
we correlated carapace length of lobsters with
carapace height (CH), and the carapace width of
crabs with carapace length . For 2 1 7 lobsters ( sexes
combined), ranging from 70 to 98 mm CL,
carapace height was determined by positioning
the lobster's ventral surface on a flat board and
then measuring the greatest perpendicular dis-
tance from the board to the top of the carapace.
Carapace length of crabs was measured from the
anterior margin of the frontal region to the poste-
rior border of the intestinal region. Measurements
for the two Cancer species were treated separately
due to the species disparities in body shapes. We
recorded carapace length for 103 male rock crabs
(females were excluded due to commercial unim-
portance) ranging from 90 to 122 mm CW, and 96
Jonah crabs (sexes combined) ranging from 96 to
132 mm CW.
RESULTS AND DISCUSSION
Lobsters in Research Gear
There are marked differences in size composi-
tion and number of lobsters caught in nonvented
426
KROUSE: EFFECTIVENESS OF ESCAPE VENT SHAPE
Table l. — Lobsters caught with nonvented and various types of vented traps from July through
November 1976.
Catch
Catch effort
Mean
Vent type
Total
no.
Sublegals;
legals
carapace
length
(mm)
Standard
error
Legals per
trap haul
No of
trap hauls
Months
fished
Nonvented
749
4.3:1
76.2
±0.28
0.53
265
July-Nov.
Horizontal
198
0.6:1
839
±062
0.54
229
July-Nov.
Vertical
107
0.5:1
84.3
±1.05
0.56
129
July- Sept.
Circular:
58 mm
25
0.6:1
828
±1.35
0.30
53
Oct. -Nov.
61 mm
42
0.4:1
85.9
±1.35
0.47
66
Sept.
and vented traps (Table 1). Vented traps caught
fewer sublegal lobsters per trap-haul than non-
vented traps (t-test, P<0.01).
The ratio of sublegal to legal lobsters did not
differ among the four types of vents (^-test,
P>0.1), with the exception of 61-mm circular
vents which caught fewer sublegals than horizon-
tal vents (P<0.01). As will be discussed later, the
61-mm hole is slightly oversize for a minimum size
of 81 mm CL, thus some smaller legal lobsters and
most shorts escape. Nevertheless this information
suggests that circular openings are as effective as
the rectangular vent ( Krouse and Thomas 1975) in
permitting escapement of short lobsters.
To further assess the relative efficiencies of the
various vents, catch-effort values (numbers of
lobsters per trap haul set over day, CPUE) were
calculated and plotted for legal-sized and all-sized
lobsters combined for each vent type (Figure 2).
For this figure, 58- and 61-mm circular vent data
were pooled because of the small sample size and
similar catch values. Figure 2 graphically shows
that the CPUE for legal-sized lobsters was similar
for all vent types; however, for combined catches of
legals and sublegals, the CPUE for nonvented
traps was several fold greater. Thus, this indicates
that all traps tested were about equally efficient in
capturing legal lobsters; but, as to be expected,
nonvented traps caught substantial numbers of
short lobsters which probably would have escaped
from vented traps. Most importantly, these data
support an earlier conclusion that circular vents
select about the same size lobsters as do rectangu-
lar vents.
Lobsters in Commercial Gear
Catch data provided by a local lobsterman were
compiled according to the following categories of
gear: 1) end vented traps with a single circular
hole of 58 mm diameter (Figure IB); 2) side vented
traps with paired round openings of 58 mm diame-
4.2 r
3.9
3.6
3.3
_l
=) 3.0-
<
^ 2.71-
a.
< 2.4
CE
>- 2-1
S 1.8
a.
1.5
)
S 1.2
Z .6
.3
1.2
.9
.6
.3
LEGALS and SUBLEGALS
-I- HORIZONTAL VENT
-O— NO VENT
O CIRCULAR VENT
A VERTICAL VENT
2 3
SET OVER DAYS
Figure 2. — Comparison of the number of lobsters (legals only;
sublegal and legals combined ) per trap haul set over day for lobster
traps with rectangular (horizontal and vertical) and circular vents
(58 and 61 mm combined) and traps without vents.
ter (Figure ID); and 3) two groups of nonvented
traps (one for nonvented traps fished in the same
trawl string with end vented traps and the second
for traps paired with side vented traps). Compari-
sons of the CPUE and the ratios of sublegals to
legals indicated that vented traps caught fewer
sublegal-sized lobsters than the corresponding
groups of nonvented traps (^-test, P<0.01) (Table
2). Higher CPUE values for vented traps show
that circular vents are at least as efficient if not
more effective in catching legal-sized lobsters
than nonvented traps (f-test, P<0.01). In an ear-
lier study Krouse and Thomas (1975) reported
that traps with 44.5 x 152.4 mm rectangular
vents were more successful in catching legal
lobsters than traps with smaller vents or no vents.
427
FISHERY BULLETIN: VOL 76, NO. 2
Table 2. — Comparison of commercial catches of sublegal and
legal-sized lobsters caught in traps with 58-mm circular vents and
traps without vents.
Trap type
Side vent:
double opening
No vent
End vent:
single opening
No vent
No. of
sublegals;
trap haul
0 98
233
1.80
2.76
No. of
legals/
trap haul
0 56
0 49
0 72
057
No of
trap
hauls
144
132
144
144
Sublegals:
legals
1.8:1
47:1
25 1
4.9:1
HORIZONTAL VENT
n =124
X = 91.2
SE = 0 89
60 70 80 90 100 IIC 120
Possible explanations for these disparities in ef-
ficiency may be that: 1) larger lobsters are less
likely to enter traps containing several other
lobsters, and/or 2) after legal lobsters are caught,
their attempts to escape might be intensified as
the density of lobsters increases w^ithin the trap.
Aside from the previously mentioned differ-
ences in the number of shorts caught per trap haul
for vented and nonvented traps, end vents (single
hole) captured 1.80 shorts/trap haul, whereas side
vents (double hole) caught only 0.98 shorts/trap
haul (Mest, P<0.01). Apparently, the additional
vent will insure greater escapement.
Crabs in Research Gear
Since male C. irroratus attain larger sizes than
females (Krouse 1972), commercial catches of this
species are comprised almost entirely of males, so
in the following analyses only catches of male
crabs are considered. Variations in size composi-
tion of catches with different vents as manifested
by width-frequency histograms (Figure 3) and
mean carapace widths which are statistically dif-
ferent ( Duncan's new multiple range test, P <0.01 )
indicate that: 1) fewer large crabs ( ^90 mm CW)
were captured in traps with horizontal vents
(mean 91.2 mm CW); and 2) as many, if not more,
larger crabs were collected with circular (mean
96.5 mm CW) than nonvented traps (mean 93.8
mm CW). According to this data, the 58-mm circu-
lar vent is at least as efficient in retaining
marketable-sized crabs as the nonvented trap and
certainly much more efficient than the horizontal
vent. Escapement of subcommercial-sized crabs
through circular openings has long been recog-
nized by west coast States with Dungeness crab
fisheries (Miller 1976). These states require crab
traps to have one or two escape rings with diame-
ters &4 in.
This situation was further evaluated by compar-
ing the numbers ( crabs ^90 mm CW) per trap haul
428
35
30
25
UJ
D
O 20
Ld
(T 15
Ll-
10
t-
Z 5
UJ
O 0
cr
35
30
25
20
15
10
5
0
60 70
80
n
— 1
NO
VENT
n
= 265
ii
= 93.8
— 1
SE
1
= 0.55
90
100
no
120
5R mm CIRCULAR VENT
n =93
7 " 96.5
SE = 0.66
60 70 80 90 100 110 120
CARAPACE WIDTH (mm)
Figure 3. — Width-frequency distributions for male rock crabs
caught with nonvented traps, traps with 58-mm circular vents,
and traps with horizontal vents fished near Boothbay Harbor,
Maine.
set over day for each of the different vents (Figure
4). CPUE values were highest for circular vents,
lowest for horizontal vents, and intermediate for
vertical and nonvented traps. Thus circular vents
relative to the other vents were most effective in
retaining crabs 3^90 mm CW and based on the
following ratios of nonkeepers ( <90 mm CW) to
keepers (^90 mm CW), selectively fished for
larger crabs:
Vent type
Circular Nonvented Horizontal Vertical
Nonkeepersikeepers 1.5:1 2.5:1 4.1:1 5.7:1
Even though smaller crabs can egress quite read-
ily from traps with horizontal and vertical vents,
the above values at first glance appear to reflect
the converse, i.e., more nonkeepers are caught in
KROUSE: EFFECTIVENESS OF ESCAPE VENT SHAPE
45r
_l 42
< 3 9
Q. 36
<
a: 3 3
H
5 24
UJ 2 1
m
S 1.8
2^ 15
1.2
9
6
3
K —
HORIZONTAL VENT
o —
NO VENT
0
CIRCULAR VENT
A-
VERTICAL VENT
12 3 4 5
SET OVER DAYS
Figure 4. — Comparison of the number of rock crabs ( 390 mm
carapace width i per trap haul set over day captured with non-
vented and circular and rectangular (horizontal and vertical) ven-
ted lobster traps.
over days in Figure 4, particularly for circular and
nonvented traps where escapement could only re-
sult via the entrances, vividly demonstrate the
crab's ability to escape as trap soak time is in-
creased. Evidently, after the voracious crabs be-
come satiated by eating the trap's bait, which fre-
quently occurs in 1 or 2 days during the summer,
the trap loses its attractiveness and crabs try to
escape. Therefore, crab fishermen can maximize
their catches by hauling their traps daily, particu-
larly during periods of high catches. Contrasted to
declining crab catches with greater soak times are
lobster CPUE values which increase until 4 or 5
set over days, after which catches begin to di-
minish (Figure 2). Similar trends in CPUE data
for commercial catches have been reported by
Thomas (1973). Thus it appears that crabs are
more adept at escaping from traps than lobsters.
Escapement and Morphometric Studies
Lobsters
horizontal and vertical vented traps than in the
circular and nonvented traps. Actually, horizontal
and vertical vents, unlike circular and nonvented
traps, also permit harvestable crabs to escape, re-
sulting in reduced catches of keepers. Con-
sequently, the proportion of nonkeepers to keepers
is markedly greater for horizontal and vertical
holes.
As evidenced by the aforementioned catch data,
selectivity features of horizontal and vertical rec-
tangular vents are similar; however, compared to
circular vents they are unsatisfactory for catching
large crabs. Prior to field testing, guided by opin-
ions of some fishermen and our own thoughts, it
seemed plausible that a vertically positioned rec-
tangular opening (Figure lA) might inhibit es-
capement of those crabs with carapace length ex-
ceeding the vent's width (smallest dimension). Of
course, this was predicated on the assumption that
when a crab encounters such a narrow upright
opening it will only attempt to egress in a horizon-
tal plane and will not tilt the body diagonally.
However, laboratory observations and size com-
position of catches in traps with vertical vents
indicate crabs will readily turn on end or side to
exit.
Prior to this, only escapement through the vent
itself has been discussed; this certainly does not
preclude escapement through entrance heads.
Diminishing CPUE values plotted against set
Passage of lobsters through a round hole is re-
lated to the lobster's carapace height (greatest
cross-sectional dimension) relative to hole diame-
ter. Figure 5 shows that: 1) most legal-sized
95%
Prediction
Intervals
85 90
LENGTH (mm)
Figure 5. — Carapace length-carapace height relationship for
lobsters with 95% confidence and prediction intervals.
429
FISHERY BULLETIN; VOL. 76, NO. 2
lobsters with 81 mm CL had <58 mm CH; 2) about
half those lobsters with 84 mm CL had <58 mm
CH; and 3) lobsters >90 mm CL had ^58 mm CH.
Based on this relationship alone, it appears that
many lobsters ranging from 8 1 to 89 mm CL would
be able to squeeze through a 58-mm diameter hole;
however, this is refuted by the previous sections on
the commercial and research catches of lobsters
with circular vented traps and the following dis-
cussion of escapement studies. Lobster escape-
ment through a round opening cannot be accu-
rately determined by carapace height alone since
this measurement excludes the walking legs
which contribute to the lobster's overall height or
depth. Whether or not a lobster is successful in
passing through a round hole will be determined
not only by the lobster's greatest transverse di-
mension (carapace height plus protruding legs)
but also by the lobster's ability to maneuver
through a tight opening.
Obvious limitations with the aforementioned
morphometric relationship caused me to seek an
alternate approach to assess escapement. Thus, I
decided to determine the largest size lobster that
could be manually passed through a 58-mm
diameter hole. Lobsters 81 mm CL passed through
the hole rather easily following careful manipula-
tion of the walking legs and 82-mm CL lobsters
required considerable force, often causing bodily
harm, while larger lobsters ( >82 mm CL) could
not pass through the opening.
Patterns of escapement for lobsters ranging
from 78 to 84 mm CL from traps with 58-, 60-, and
61-mm diameter vents varied decidedly as de-
picted by retention curves in Figure 6. Only the
58-mm vent retained all legal-sized lobsters and
still had reasonably high escapement of sublegals;
whereas, the other vents which were merely 2 or 3
mm larger allowed legal-sized lobsters to escape.
These data emphasize the importance of accu-
rately producing the 58-mm opening, else the
vent's desired effect will be lost.
Crabs
Carapace width-length relationships for C.
borealis and C. irroratus graphically show that
crabs >90 mm CW (commercially harvested size)
have carapace lengths (dimension limiting es-
capement) which exceed 58 mm (Figures 7, 8).
Accordingly, commercial-sized crabs of either
species cannot egress through a circular opening
58 mm in diameter. In fact, if the vent diameter
77 78 79 80 81
CARAPACE LENGTH (mm)
83 84
Figure 6. — Retention curves for lobsters placed in lobster traps
with circular vents of 58, 60, and 61 mm in diameter.
were increased to as large as 65 mm (certainly, an
over estimate) to accommodate an upward shift in
the lobster minimum size (Maine Department of
Marine Resources recommends an increase from
3^/i6 to 3y2 in CL by Vi6-in increments annually
over a 5-year period) this would have little or more
likely no effect on catches of marketable crabs.
RECOMMENDATIONS
In view of the findings of this study and past
investigations (Krouse and Thomas 1975; Krouse
1976), all lobster and crab traps fished in Maine
waters should have a rectangular escape vent not
less than 1.75 in (44.5 mm) by 6 in ( 152.4 mm) or at
least two circular escape vents not less than 2.28
in (58 mm) in diameter. To insure maximum es-
capement of sublegal lobsters, vents should be in-
stalled next to the sill on the side or end of the
trap's parlor section.
Although fishermen should certainly have the
option to fabricate their own vents, provided that
the prescribed dimensions are adhered to, the use
of synthetic, prefabricated vents is highly recom-
mended (Krouse and Thomas 1975). Recently, a
plastics manufacturer assured me that vents could
be produced and retailed for about 200 each. At
this low price and with today's high price of laths
(about 50 each), if a synthetic vent replaces two
laths every 3 yr, then after 6 yr the original cost of
430
KROUSE: EFFECTIVENESS OF ESCAPE VENT SHAPE
90r
95 /oPredtction
Intervals
95 ''o Confidence
Intervals
110 120
CARAPACE WIDTH (MM)
30
Figure 7. — Carapace width-carapace length relationship for male rock crabs with 95^^ confidence
and prediction intervals.
90-
5
X
O
§ 80
LU
U
<
<
<
70
60f-
90
--♦a
95%
Prediction
Intervals
100
110 120
CARAPACE WIDTH(MM)
I3O
Figure 8. — Carapace width-carapace length relationship for Jonah crabs with 'dWc confidence and
prediction intervals.
431
the vent will be defrayed by the replacement cost
of the laths, resulting in a cost savings.
Therefore, those fishermen interested in captur-
ing only lobsters and, perhaps, minimizing their
crab catches, would be encouraged to use rectan-
gular vents, while fishermen interested in both
lobsters and crabs or solely the latter should
employ circular vents.
ACKNOWLEDGMENTS
I thank David A. Libby for his assistance in field
collections, data compilation, and figure prepara-
tion; and Charles Begin, a commercial fisherman,
who often at an inconvenience to himself, fur-
nished me with invaluable catch information.
LITERATURE CITED
Jow, T.
1961. Crab trap escape-opening studies. Pac. Mar. Fish.
Comm. 5:49-71.
KROUSE. J. S.
1972. Some life history aspects of the rock crab, Cancer
irroratus, in the Gulf of Maine. J. Fish. Res. Board Can.
29:1479-1482.
1973. Maturity, sex ratio, and size composition of the
natural population of American lobsters, Homarus
FISHERY BULLETIN: VOL. 76. NO, 2
americanus, along the Maine coast. Fish. Bull., U.S.
71:165-173.
1977. Incidence of cull lobsters, Homarus americanus, in
commercial and research catches off the Maine
coast. Fish. Bull, U.S. 74:719-724.
KROUSE, J. S., AND J. C. THOMAS.
1975. Effects of trap selectivity and some population
parameters on size composition of the American lobster,
Homarus americanus, catch along the Maine
coast. Fish. Bull., U.S. 73:862-871.
Miller, R. J.
1976. North American crab fisheries: Regulations and
their rationales. Fish. Bull., U.S. 74:623-633.
Stasko, a. B.
1975. Modified lobster traps for catching crabs and keep-
ing lobsters out. J. Fish. Res. Board Can. 32:2515-2520.
TEMPLEMAN, W.
1939. Investigations into the life history of the lobster
(Homarus americanus) on the west coast of Newfound-
land, 1938. Newfoundland Dep. Nat. Resour., Res. Bull.
(Fish.) 7, 52 p.
Thomas, J. C.
. 1973. An analysis of the commercial lobster (Homarus
americanus) fishery along the coast of Maine, August
1966 through December 1970. U.S. Dep. Commer.,
NOAA Tech. Rep. NMFS SSRF-667, 57 p.
Wilder, D. G.
1945. Wider lath spaces protect short lobsters. Fish. Res.
Board Can., Atl. Biol. Stn. Circ. G-4, 1 p.
1948. The protection of short lobsters in market lobster
areas. Fish. Res. Board Can., Atl. Biol. Stn. Circ. G-11, 1 p.
1954. The lobster fishery of the southern Gulf of St. Law-
rence. Fish. Res. Board Can., Gen. Ser. Circ. 24, 16 p.
432
SCHOOL STRUCTURE OF THE SQUID LOLIGO OPALESCENS
Ann C. Hurley'
ABSTRACT
The squid Loligo opalescens forms schools which are similar in many respects to those of obligate
schooling fishes. These schools are marked by parallel orientation of individuals and strong cohesive-
ness. Laboratory experiments indicate that the main sensory modality regulating schooling is vision.
Squid on opposite sides of a clear rigid Plexiglas barrier readily schooled. The structure of schools of six
squid depended on size of individuals and was modified by environmental disturbance. Parallel
orientation was weaker in schools of smaller squid (ca. 7 cm dorsal mantle length) than it was in larger
ones (ca. 12 cm). In the field, L. opalescens schools are composed of uniformly sized individuals.
Laboratory experiments designed to determine whether this was due to actual size selection were
inconclusive, but they did suggest mechanisms which might be important in determining squid
position in the school.
Considerable effort has been spent in understand-
ing the schooling behavior of fish in terms of
physiological mechanisms and possible survival
value and ecological consequences. (See reviews
by Radakov 1973 and Shaw 1970, 1978.) Virtually
no work has been done on schooling behavior of
invertebrates which occur in the same environ-
ments as schooling fish. The most evident school-
ing invertebrates in the pelagic environment are
the squid. Squid and fish play very similar ecologi-
cal roles and the two groups of organisms possess a
large number of similarities. (See Packard 1972,
for a discussion of convergent evolution.)
Loligo opalescens is common off the west coast of
North America with a reported range from Baja
California to lat. 55°N (Fields 1965; Bernard
1970). Relatively little is known of the behavior or
general ecology of L. opalescens in spite of the fact
that there is a fishery for this species in California.
The fishery is based primarily upon the tendency
of squid to spawn in large aggregations in shallow
water (McGowan 1954; Fields 1965). Very little is
known about the distribution or location of newly
hatched squid as well as squid in later stages of
life. Attempts to catch the juveniles have often
been unsuccessful (Okutani and McGowan 1969)
and only recently have attempts been made to
catch nonspawning adults. Even though field
data are difficult to obtain, it is possible to keep
both juvenile (Hurley 1976) and adult L. opales-
cens alive in the laboratory. Schooling in the
laboratory was examined to provide insights
about the function of schooling in squid.
METHODS
The squid used in the behavioral studies were
obtained either by dipnetting them after they had
been attracted to an underwater light or by pur-
chasing them from a local bait dealer. In the
laboratory, the squid were maintained in a 3-m
diameter circular tank with rapidly circulating
seawater. They were fed irregularly on small fish
(either mosquitofish, Gambusia affinus, or
goldfish, Carassius auratus). Mosquitofish were
taken much more readily than were the goldfish.
Occasionally, the squid could be trained to take
dead food. This was accomplished by first getting
them to accept live fish and by then throwing dead
fish in along with the live ones. In this manner, the
squid could also be coaxed to accept pieces of frozen
northern anchovy, Engraulis mordax. If the squid
were undamaged when they arrived at the
laboratory and there was an abundant supply of
small fish available, it was relatively easy to keep
them for over a month.
Experiments designed to examine various as-
pects of schooling behavior were run in a 2 x 3 m
rectangular Plexiglas^ tank which was filled to a
depth of 0.4 m. The tank was painted flat white
and the primary source of lighting (in addition to
general room illumination) was provided by
'Moss Landing Marine Laboratories, P.O. Box 223, Moss
Landing. CA 95039.
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
Manuscript accepted November 1977.
FISHERY BULLETIN: VOL. 76. NO. 2. 1978.
433
FISHERY BULLETIN: VOL. 76, NO. 2
fluorescent lights placed around the perimeter of
the tank which shone through the walls. This pro-
vided even, diffuse light in the tank. The water in
the tank was noncirculating.
Schooling behavior was recorded on Tri-X film
using a 35 mm camera with motor drive and a
variable setting automatic timer. A mirror was
placed above the tank and pictures were taken of
the squid by photographing the surface image
reflected in the mirror. A black plastic barrier
surrounded the experimental tank. A small hole
in the barrier allowed observations to be made of
the squid without disturbing them. Exact
methods, timing of pictures, etc. varied with the
experiment and will be described in the appro-
priate section.
After the films were developed, they were
analyzed using a Scientific Data Products data
tablet (Graf-Pen) coupled with a PDP 11-45 com-
puter. The data tablet is a set of microphones
placed at right angles which record the sound pro-
duced by an electrical spark made by a special
marking pen. The x and y coordinates of a point
were relayed to the computer by pressing the pen
down on the tablet at that point. This device al-
lowed the recording of large amounts of squid posi-
tion data. In each frame, the tip of the tail, the tip
of the outstretched arms, and a point midway be-
tween the two eyes were recorded for each squid.
Other information, such as the position of
barriers, was recorded in the same way. Mea-
surements taken from the photographs were sub-
sequently converted to real distances by multipli-
cation by appropriate scale factors.
Students of schooling have examined school
geometry both as a two-dimensional system on a
horizontal plane (Breder 1959; Williams 1964;
John 1964; Hunter 1966, 1968; Van 01st and
Hunter 1970) and as a three-dimensional struc-
ture (Cullen et al 1965; Symons 1971a, b; Pitcher
1973). Since squid schools do have a three-
dimensional structure in nature, a three-
dimensional analysis will eventually be necessary
to determine all of the structural details of the
school. A three-dimensional analysis, however, is
much more difficult than a two-dimensional
analysis. It was felt that a two-dimensional
analysis would suffice to examine certain aspects
of squid schooling behavior. In these experiments,
the squid were very nearly confined to a two-
dimensional plane by the shallow water depth in
the experimental tank. Observations of small
schools (up to six squid) in a deeper tank (1 m
depth) indicated that the two-dimensional struc-
ture observed in the experimental tank was not
uncommon.
Three indices were chosen to quantify the angu-
lar orientation of individuals in a school, the over-
all shape of the school, and the distance between
neighboring individuals in a school. These indices
were proposed by Hunter ( 1966) and he includes a
detailed discussion of their properties. The three
indices are:
1. Mean separation distance: An average of the
horizontal distances separating each squid from
every other squid in the school. It is influenced by
school shape, distance between neighboring squid,
and number of squid in the school. Distances be-
tween all possible pairs of squid are measured and
these values are averaged. Distance is measured
between the two closest points on the midline of
the bodies ( including outstretched arms) of the two
squid.
2. Mean distance to nearest neighbor: An aver-
age of distances from each squid in the school to its
nearest neighbor. The measurement is made be-
tween the two closest points on the midline of the
bodies (including outstretched arms) of the two
squid. The same measurement is used twice if two
squid are closer to each other than to any other
squid.
3. Mean angular deviation: This is a measure of
the differences in orientation among squid within
a school. The heading of each squid is determined
and the resultant direction of the school is com-
puted by assigning each squid a value of one and
adding the headings vectorally. The mean number
of degrees individual squid deviated from this re-
sultant direction of the school was calculated as
the index of orientation.
One difference between squid and most of the
schooling fish which have been studied is that
squid readily swim both forward and backward.
Thus, a squid with an orientation that was 180°
out of phase with the rest of the school might still
be swimming with the rest of the school. For this
reason, one orientation measure was calculated
which regarded the squid as a line segment rather
than as a directed vector and measured the small-
est angular deviation between line segments.
Such measurements were rarely different from
measurements made considering the orientation
of the squid and therefore will not be considered
further in this paper.
434
HURLEY: SCHOOL STRUCTURE OF LOLIGO OPALESCENS
Where measurements of squid length are given,
they are of dorsal mantle length from tip of the pen
to the tip of the tail. The total length of the squid
(including arms but excluding tentacles) is about
1.3-1.5 times the dorsal mantle length (Fields
1965).
RESULTS
Response to Disturbance
One set of factors that caused changes in school-
ing can be grouped under "external disturbances."
These included introducing objects (such as a net)
into the water near the squid or tapping on the side
of the tank. The typical response was for the squid
to group more tightly and. in cases where it was
not already marked, to increase the degree of
parallel orientation. The amount of change in
schooling behavior and the temporal characteris-
tics of this change depended upon the nature and
intensity of the disturbance and upon its duration.
One attempt at quantifying the stimulus in-
volved placing an aquarium air stone in the tank.
Pressurized air delivered to this air stone in differ-
ing amounts and duration produced a stream of
bubbles which could be used as a disturbance
stimulus of varied intensity and duration. A small
stream of bubbles produced little squid reaction,
while vigorous water action due to the bubble
stream produced marked changes in behavior.
Figure 1 shows the changes in three of the school-
ing indices in response to a moderate disturbance
caused by turning on the air bubble stimulus. The
degree of parallel orientation, which was already
pronounced, did not change appreciably. But the
squid did draw noticeably closer together.
Schooling Structure as
a Function of Squid Size
Six squid of nearly equal size were haphazardly
taken from the holding tank and placed in the
experimental tank. The squid swam in this tank
for an hour before measurements were made. With
the exception of one experiment, a picture was
taken of the squid every minute for approximately
1 h. During this other experiment, a picture was
taken every 10 s for 10 min. This set of experi-
ments was conducted during the daylight hours of
two different days. All of the squid used in this set
of experiments had been captured on the same
night.
There was a decrease in the mean angular de-
viation as the size of the squid increased (Table 1).
Since small values of the mean angular deviation
index are associated with increased parallel orien-
tation, the degree of parallel orientation is
greatest in schools composed of large individuals.
Even in the case of the small individuals, however,
the value of the index does not approach what
would be expected if the squid were each orienting
in a random direction. In a simulation of 1 million
values for six randomly oriented fish. Hunter
(1966) found that the mode of the frequency dis-
tribution was 69°.
Although the average values for mean angular
deviation do give a measure of average departure
from parallel orientation for a whole experiment,
they do not give an indication of how variable a
particular group of squid is in its orientation over
time. For example, an experiment of 30 pictures
and an average value of the mean angular devia-
tion index of 20° could have had all of the 30 values
close to 20°. This would indicate a consistent mod-
erate degree of parallel orientation over time. On
the other hand, such an average value could also
come from a situation where the squid had strong
parallel orientation part of the time and were
much more loosely oriented the rest of the time
(e.g., the index value could have been 10° on 15
frames and 30° on 15). This kind of difference can
be detected if a measure of the variability of the
mean angular deviation index for each experi-
ment is calculated. The variability (standard de-
viation, SD, Table 1) increased with decreasing
squid size, indicating that not only do the smaller
squid not orient on the average in as parallel a
manner as larger squid, but they are also more
temporally variable in their orientation. This dif-
ference can also be seen if individual experiments
are examined. Figure 2 shows the values for mean
separation distance and mean angular deviation
Table l . — Relationship between average size of Loligo opales-
cens and parallel orientation and separation of individuals in the
six-squid experiments. Each index was calculated for each
frame.
Group
number
Mean
mantle
length
(cm)
No.
frames
examined
Mean angular
deviation in-
dex (degrees)
Mean separation
distance Index (cm)
X
SD
X
SD
1
7.5
44
32.0
18.4
32.3
26.5
2
7.6
58
290
15.3
25.6
8.0
3
7.7
60
18.1
10.7
162
4.5
4
9.7
19
18.5
5.7
14.0
4.2
5
97
62
16.2
62
18.7
5.2
6
102
69
17.2
6.6
20.9
5.0
7
11.9
65
11.1
56
18.5
4.0
8
120
55
9.6
28
15.3
2.7
9
12.0
46
9.1
4.2
13.8
3.0
435
FISHERY BULLETIN: VOL. 76. NO 2
30
20
r¥m^.
20
CD C/1
FiGL'RE 1. — Values of schooling indices
for a school of six Loligo opalescens be-
fore and after disturbance (turning on
bubbler). Dashed line indicates when
air was turned on. Pictures were taken
every minute for 64 min.
30
--^ 20
LU •—<
s:° 10
for two (Groups 3 and 7) of the experimental runs
summarized in Table 1. The parallel orientation is
stronger and the variability less in the larger
squid (Figure 2C, D). The mean separation dis-
tance index is not as clear a function of size (Table
1; Figure 2A, B).
Schools in the Ocean
Very little is known of the natural behavior of
Loligo opalescens when it is not in large mating
schools. In many areas, there often is a large con-
centration of squid in the vicinity of the deep-
scattering layer (C. Recksiek, Moss Landing
Marine Laboratories, Moss Landing, CA 95039,
436
TIME
pers. commun., October 1976) and large layered
concentrations of L. opalescens have been reported
by those involved in submersible exploration (A.
Flechsig, Sea Grant Marine Advisory Service,
University of California at San Diego, La Jolla,
CA 92093, 1973). There is evidence to indicate,
however, that L. opalescens is often found in much
smaller schools and that these schools contain a
narrow size range of individuals. Fields (1965)
presents data on the uniformity of size of young
squid taken from the same fish catch (presumably
the same squid school ) and speculates that the size
ranges in the schools he observed represent ap-
proximately one-half or less than one-half of a
year's growth.
HURLEY: SCHOOL STRUCTURE OF LOLIGO OPALESCENS
30
A
20 .
2 10
V
/ V"
.' S-.
\/\
\y
\/
30 .
20
10 -
^/\/\ A.; WW n
10
~T —
20
30
MINUTES
~T —
40
— I-
50
60
•a
MINUTES
Figure 2. — Values of mean separation distance and mean angular deviation for two (Groups 3 and 7) of the
experiments presented in Table 1. Mean size ofLoligo opalescens in the experiment represented in A and C was 7.7 cm
mantle length. Mean size of squid in the experiment represented in B and D was 11.9 cm mantle length.
437
FISHERY BULLETIN: VOL. 76, NO. 2
I also obtained data on the uniformity of size in
individuals of the same school. The squid were
caught during a 1-wk period in August in locations
ranging from San Diego to Santa Catalina Island,
Calif. A night-light was placed off the stern of the
ship in the center of an L-shaped 3-m long mesh
net. Squid were attracted to the light and would
rush into the net. The net was then raised and the
squid could be removed with dip nets. The
"schools" were all of the squid which swam into the
net at the same time. Squid caught during this
period ranged from 5.8 to 17.3 cm dorsal mantle
length. But for a given school, they were much
more uniform in length. The average size range for
29 schools of 2 to 32 individuals was 2.5 cm.
Maintenance of
School Structure and Orientation
Experiments in the laboratory have indicated
that vision is sufficient sensory input to mediate
schooling behavior. Squid on different sides of a
clear, rigid Plexiglas barrier will readily school
with each other and they appear to maintain the
same type of parallel orientation that is present in
normal schooling behavior. Preliminary experi-
ments using such Plexiglas barriers were run to
try to elucidate the mechanisms by which spacing
is maintained.
Two-Squid Experiments
Experiments were run to determine whether
squid would school in the same manner with or
without a clear Plexiglas barrier in place. Mea-
surements were obtained for squid swimming to-
gether and for the same squid swimming on oppo-
site sides of a Plexiglas barrier which divided the
tank into two compartments. The order of the
treatment was randomized for each pair of squid.
Squid ranged in size from 7 to 13 cm mantle
length. For a given experiment, the two squid
were of similar length. Pictures were taken of the
squid in each treatment every 10 s for 3 min after
they first came together and again every 10 s for 3
min after the squid had been left undisturbed for
15 min. If the squid did not come together to within
at least 0.5 m within 1 min, the experiment was
terminated.
Table 2 shows the results of five such experi-
ments. The first 3-min periods have been compared
with each other, as have the later runs. This was to
see if the pattern of schooling changed after the
Table 2. — Median nearest distances and median separation
angles for two-squid (Loligo opalescens) experiments.
With
Without
Item
barrier
barrier
Difference'
Nearest dis-
20.6
16.2
P = 0 05 barrier greater
tance (cm)
113
6.6
P<0.01 barrier greater
first 3 mm
38.6
11.65
P<0 01 barrier greater
14.9
7.9
P<0.01 barrier greater
13.8
6.4
P<0.01 barrier greater
Nearest dis-
24.2
15.8
P<0.01 barrier greater
tance (cm)
18.4
6.3
P<0 01 barrier greater
second 3
23.9
18.35
P<0.05 barrier greater
mm
14.35
13.1
P<0.05 barrier greater
12.5
8.2
P<0.01 barrier greater
Separation
16.4
52.7
P<0.01 barrier less
angles (de-
24.3
11.2
P<0 05 barrier greater
grees) first
75.1
21.0
P<0 01 barrier greater
3 mm
21.2
12.2
NS^
17.0
30.9
P<0.05 barrier less
Separation
28.2
15.4
P<0.05 barrier greater
angles (de-
15.4
13.6
NS
grees) sec-
19.1
18.0
NS
ond 3 mm
25.4
18.3
NS
11.1
24.9
P <0.05 barrier less
'Significance of difference in medians from IVlann- Whitney U-test.
^NS = no significant difference. P 0.05
squid became more adapted to the experimental
regime. This table presents results for the median
nearest distance between the two squid for each
run and for the median separation angle for these
same runs. Separation angle for each frame is
simply a measurement of the angle between the
two squid and is a measure of orientation (0° sep-
aration angle indicating parallel alignment facing
the same direction).
The barrier has an effect upon the separation
distance between the two squid. In all cases, there
was a significant difference between the distance
between squid with and without the barrier. When
the Plexiglas barrier was present, the squid
tended to space themselves farther apart. There is
not a clear relationship between angular separa-
tion and the presence of the barrier. Of the six runs
showing significant differences, three had greater
median separation angles with the barrier in place
and three had greater median separation angles
when the barrier was not present.
Three-Squid Experiments
The experimental tank was divided crosswise
into three equal compartments (1 x 2 m each) by
clear Plexiglas partitions. A squid was chosen
from the holding tank and was placed in the cen-
tral compartment. Then a squid for each of the
outer compartments was selected. These squid
were assigned at random to each of the outer com-
partments. The squid were allowed to adapt to the
experimental situation for 15 min and then were
filmed for 5 min (one picture every 10 s). The two
438
HURLEY. SCHOOL STRUCTURE OF LOLIGO OI'ALESCENS
outer squid were then switched from one outer
compartment to the other and the squid were
again allowed to adapt for 15 min. They were then
filmed for 5 min (once every 10 s). Squid in these
experiments ranged from 9.2 to 15.3 cm mantle
length.
It was hoped that this experimental design
would indicate whether the center squid, if given a
choice, would choose to school with a larger or
smaller squid or one closer to its own size. One way
to determine whether such a choice is being made
would be to determine whether the center squid
spends more of its time closer to one outer squid
than to the other. Each 5-min run was considered
as a unit and each frame was scored according to
which outer squid the center squid was nearest.
For each run, the data were compared with a
binomial distribution which assumed that the
center squid had an equal probability of being
closest to either outer squid. Of the 17 runs, 16
showed a significant deviation from the expected
binomial distribution (Ps:0.05 for 1; P^O.Ol for
15). These 16 runs were now grouped according to
whether the center squid was closest to the larger
or smaller outer squid. In 8 of the 16 cases, the
center squid was nearest the larger outer squid,
while in the other 8 cases, it was nearest the small-
er squid. There is no evident preference for large
versus small squid. The data can also be arranged
to determine whether the center squid spent most
of its time near the squid closer to its own size.
There were 14 runs for which it was possible to say
that the center squid was closer in size to one of the
outer squid. Of these 14 runs, the center squid was
significantly nearer to the squid closer to its own
size 9 times and nearer to the squid farther from
its own size 5 times.
These experiments may be viewed in another
way by looking at the absolute position of the
squid in the tank. The nearest distances of the
squid to the Plexiglas barriers were calculated for
each frame. These data are summarized in Figure
3 for the 17 runs. The side squid usually are very
near the barrier which separates them from the
center compartment, while the center squid varies
his position within the center compartment, but
approaches the Plexiglas barriers much less often.
DISCUSSION
Pelagic fish and squid represent a striking case
of convergent evolution, not only morphologically
(Packard 1972), but behaviorally as well. One as-
pect of behavior where this is particularly appa-
rent is schooling. Since many of the same ecologi-
cal pressures exist for both pelagic groups, it is not
surprising that some sort of schooling behavior
would have developed in both fish and squid. What
is surprising, given the very different physiology
and mode of locomotion, is that so many aspects of
this behavior are the same.
Loligo opalescens fits Breder's (1967) definition
of obligate schoolers. Single L. opalescens are
rarely caught in the field, and they immediately
come together when placed in a tank in the
laboratory. As has been reported for many species
offish (Radakov 1973), L. opalescens schools con-
sist of individuals of approximately the same size.
It has been suggested that the reason that fish
school in such groups has to do with swimming
speed. Small and large individuals would not
swim at the same speed and thus would not nor-
mally stay together. This is possibly also true for
squid, but data on the swimming speed of large
and small L. opalescens are not available to sub-
stantiate the argument. For several reasons, the
swimming speed hypothesis seems less plausible
for squid than for fish. In schools offish which show
parallel orientation, the fish continually maintain
forward motion and thus swimming speed is likely
to be an important factor. But field and laboratory
observations have indicated that individuals in
squid schools spend much of their time hovering in
the same position in the water column with only
16
14
12
10
8
6 .
0
2 .
0
m-
Figure 3. — Histograms of mean
nearest distance between Loligo opales-
cens and barrier in the 17 three-squid
experiments. Distances are broken up
into 10-cm intervals. From left to right:
left outer squid to left barrier, center
squid to left barrier, center squid to
right barrier, right outer squid to right
barrier.
439
FISHERY BULLETIN: VOL 76, NO. 2
slight backward and forward motion caused by jets
of water from the siphon. Even when disturbed,
the squid do not make long extended swims which
would tend to sort out those of differing swimming
ability. In the field, the most common response of a
squid school to a disturbance (the presence of a
scuba diver or a shark) is to clump closer together
and move off a slight distance. On one occasion
when I was diving in a large spawning school
(several thousand individuals), the squid executed
the same type of maneuver that has been reported
for fish schools. Instead of moving off, the school
completely enclosed me, leaving a spherical space
of approximately 3 m radius around the "pred-
ator."
One other piece of evidence suggests that it is
not differences in swimming speed alone which
cause the squid to school according to size. While
diving in the Bahamas in the Hydrolab underwa-
ter habitat, we observed a school of squid which
routinely visited the habitat. This school was
composed of Doryteuthisplei, a species which quite
closely resembles L. opalescens and presumably
has similar swimming ability. This school con-
sisted of seven squid and, in this case, was not
composed of individuals of the same size, the
largest individual being at least two times the
length of the smallest individual. We chased this
school several times but were never able to force
them to separate. The smallest squid maintained
the same swimming speed as the largest squid.
It is possible that squid maintain schools of in-
dividuals of a fairly narrow size range because of
social factors. Generally, workers studying school-
ing have assumed that all of the fish in a school
may be treated as equivalent individuals in the
production of the behavior and that there is no
social structure in the schools. In fact, some work-
ers have suggested that schooling is really just a
modified form of individual cover-seeking be-
havior (Williams 1964; Hamilton 1971). This as-
sumption of equality of individuals may be an
untenable one for squid schools. In the field,
Hochberg and Couch ( 1971 ) observed signaling by
some members of a school of Sepioteuthis sepioidea
which they felt prevented other squid from joining
the school. Furthermore, in the laboratory, I have
observed complicated agressive interactions in L.
opalescens which certainly demonstrate that all
squid cannot be considered behaviorally equiva-
lent individuals at all times (Hurley 1977).
One aspect of schooling in fish which has been
emphasized by many workers is that the structure
of schools may change as a function of the age or
the physiological condition of the fish. Van 01st
and Hunter ( 1970), for instance, found that in five
species of marine fishes, schools of young fish were
less compact and showed greater differences in
angular headings than did schools of adult fish. In
addition. Hunter ( 1966) showed that distances be-
tween jack mackerel tended to increase with food
deprivation, while Keenleyside (1955) noticed
that sticklebacks were more densely packed in a
school when well fed than when starved.
I attempted to determine whether similar
phenomena were observable in squid schools.
Schools of small squid ( 7-9 cm mantle length) gave
the impression of being less cohesive than schools
of larger squid ( 13-15 cm mantle length). This was
supported to some extent by the quantitative mea-
surements, particularly those of angular orienta-
tion. The variability was also higher for all of the
indices for the smaller squid. It has been suggested
for fish (Van 01st and Hunter 1970) that the ob-
served change with size could have been an adap-
tation to the higher food requirements of the
juvenile fish. This speculation is supported by ob-
servations that a number of species school less
cohesively under conditions of food deprivation.
The same explanation may also hold for squid, but
my existing data do not support it. I ran two exper-
iments in which schools of six squid were filmed
before and after feeding. In one experiment, there
were no significant differences in the schooling
indices before and after feeding, while in the other,
there was significantly less school cohesion and
parallel orientation after feeding. In any event, it
is not possible to guess which factors are instru-
mental in this increased cohesiveness and consis-
tent geometry.
As is the case for fish, vision seems to be the
primary sensory system used in squid schooling.
Squid will readily school across a clear, rigid
Plexiglas barrier, although they tend to stay
somewhat farther apart than they normally
would. Investigators dealing with fish also found
that the presence of a clear, rigid barrier caused
abnormalities in the spacing between individuals,
in some cases increasing the fish-to-fish distance
( Cahn 1972) and in some cases decreasing it (Shaw
1969). These workers speculated that this change
was due to lack of lateral line input and a resultant
loss of information concerning the position of the
adjacent fish. Squid do not have a similar exten-
sive vibration-sensitive system, although they
may be able to detect vibrations with their stato-
440
HURLEY; SCHOOL STRUCTURE OF LOLIGO OPALESCENS
cysts. In the case of L. opalescens, the most likely
explanation for this change in spacing is that the
presence of the barrier physically limits the extent
to which each squid can compensate for the other
individual's movements. In the experimental
tank, squid seemed to differ in their motivation to
school. When the barrier was not present, a squid
with a strong tendency to school could always
maintain proximity to another squid. But if the
barrier were present, that squid could only follow
another squid as far as the barrier and had to
remain there until the other squid returned.
I had hoped that the experiments with the three
squid separated by Plexiglas barriers would give
some clue as to whether the squid actively chose to
school with individuals of the same size, but the
results were inconclusive. The results did indi-
cate, however, a possible mechanism for mainte-
nance of spacing within a school. The center squid
tended to stay toward the middle of the compart-
ment, while the side squid maintained positions
very near the Plexiglas barrier. It is possible that
the center squid was attempting to equalize the
visual angle subtended by the squid on each side,
while the outer squid were attempting to get into
positions with squid on each side. The measure-
ments of visual angle which I can get from my
photographs are not accurate enough to determine
whether this is happening. If outer squid are con-
tinually trying to achieve a position where they
have squid on either side of them, individuals in a
school should be continually shifting positions.
Casual observations have indicated that this does
happen some of the time; but at other times, the
individuals maintain the same positions relative
to one another.
An area where a comparison of squid and fish
schooling may be useful is in the speculation con-
cerning the evolution of the schooling behavior
and its possible advantages. Many recent papers
have concentrated on the hydrodynamic advan-
tages offish schooling (e.g., Breder 1976) and base
their explanations of many of the details of school
structure on the fish mode of tail-flip locomotion
and the vortices which are subsequently created.
Van 01st and Hunter ( 1970) suggest that the typi-
cal nearest neighbor distance in fish schools is
about one-half a body length and that this distance
may be explained by considering the amplitude of
the tail beat in swimming. It is interesting that in
squid, with their very different mode of locomotion
(jet propulsion as opposed to tail flips), the spacing
between nearest neighbors is still maintained be-
tween one-half and one body length in undis-
turbed squid.
Other investigators have speculated that a
primary function of schooling is as a defense
against predation. ( See reviews by Shaw 1970 and
Radakov 1973.) Squid have many of the same
reactions to disturbance that fish do. They both
clump more closely together as a result of distur-
bance and both have been seen to surround their
predators. Further evidence which suggests that
predator defense may be an important function of
squid schooling comes from the development of the
behavior in juvenile squid. In the course of rearing
L. opalescens (Hurley 1976), I made observations
on schooling behavior. The newly hatched squid
appeared to have no attraction to each other, but
after 6 or 7 wk schooling was occasionally ob-
served. This schooling was only evident in re-
sponse to disturbance (tapping on the tank or put-
ting a net into the water). When the squid were
feeding undisturbed, there was no obvious school-
ing behavior.
ACKNOWLEDGMENTS
I would like to thank P. Hartline, J. Hunter, and
G. D. Lange for their assistance during this study;
E. Shaw, G. Cailliet, and J. Nybakken for valuable
comments and suggestions; and Rosemary
Keegan for typing the manuscript. This work was
supported by NIH grant NS-09342 and NSF grant
GH-41809 to the laboratory of G. D. Lange, Uni-
versity of California, San Diego, and the South-
west Fisheries Center, La Jolla Laboratory, Na-
tional Marine Fisheries Service, NOAA, while I
held a NOAA associateship at that laboratory.
LITERATURE CITED
BERNARD, F. R.
1970. A distributional checklist of the marine molluscs of
British Columbia, based on faunistic surveys since
1950. Syesis 3:75-94.
BREDER, C. M., JR.
1959. Studies on social groupings in fishes. Bull. Am.
Mus. Nat. Hist. 117:393-482.
1967. On the survival value of fish schools. Zoologica
(N.Y.) 52:25-40.
1976. Fish schools as operational structures. Fish. Bull.,
U.S. 74:471-502.
CAHN, p. H.
1972. Sensory factors in the side-to-side spacing and posi-
tional orientation of the tuna, Euthynnus affinis, during
schooling. Fish. Bull., U.S. 70:197-204,
441
FISHERY BULLETIN: VOL. 76, NO 2
CULLEN, J. M., E. Shaw, and H, A. Baldwin.
1965. Methods for measuring the three-dimensional struc-
ture of fish schools. Anim. Behav. 13:534-543.
FIELDS, W. G.
1965. The structure, development, food relations, repro-
duction, and life history of the squid Loligo opalescens
Berry. Calif Dep. Fish Game, Fish Bull. 131:1-108.
Hamilton, W. D.
1971. Geometry for the selfish herd. J. Theor. Biol.
31:295-311.
HOCHBERG, F. G., AND J. A. COUCH.
1971. Biology of cephalopods. In J. W. Miller, J. G. Van-
Derwalker, and R. A. Waller (editors), Tektite 2. Scien-
tists in the sea, p. VI-221-VI-228. U.S. Dep. Inter., Wash.,
D.C.
Hunter, J. R.
1966. Procedure for analysis of schooling behavior. J.
Fish. Res. Board Can. 23:547-562.
1968. Effects of light on schooling and feeding of jack mac-
kerel, Trachurus symmetricus . J. Fish. Res. Board Can.
25:393-407.
Hurley, a. C.
1976. Feeding behavior, food consumption, growth, and
respiration of the squid Loligo opalescens raised in the
laboratory. Fish. Bull., U.S. 74:176-182.
1977. Mating behavior of the squid Lo/tgoopa/escens. Mar.
Behav. Physiol. 4:195-203.
JOHN, K. R.
1964. Illumination, vision, and schooling of Astyanax
mexicanus (Fillipi). J. Fish. Res. Board Can. 21:1453-
1473.
Keenleyside, M. H. a.
1955. Some aspects of the schooling behavior of fish. Be-
havior 8:183-248.
McGowan, J. A.
1954. Observations on the sexual behavior and spawning
of the squid, Loligo opalescens, at La Jolla, California.
Calif Fish Game 40:47-54.
Okutani, t., and J. A. McGowan.
1969. Systematics, distribution, and abundance of the
epiplanktonic squid (Cephalopoda, Decapoda) larvae of
the California Current, April, 1954 - March, 1957. Bull.
Scripps Inst. Oceanogr. 14:1-90.
Packard, A.
1972. Cephalopods and fish: the limits of convergence.
Biol. Rev. Cambr. Philos. Soc. 47:241-307.
PITCHER, T. J.
1973. The three-dimensional structure of schools in the
minnow, Phoxinus phoxinus (L.). Anim. Behav.
21:673-686.
RADAKOV, D. V.
1973. Schooling in the ecology of fish. John Wiley and
Sons, N.Y., 173 p.
Shaw, E.
1969. The duration of schooling among fish separated and
those not separated by barriers. Am. Mus. Novit. 2373,
13 p.
1970. Schooling in fishes: critique and review. In L. R.
Aronson, E. Tobach, D. S. Lehrman, and J. S. Rosenblatt
(editors). Development and evolution of behavior, p. 452-
480. W. H. Freeman and Co., San Franc.
1978. Schooling fishes. Am. Sci. 66:166-175.
SYMONS, P. E. K.
1971a. Spacing and density in schooling threespine
sticklebacks (Gasterosteus aculeatus and mummichog
{Fundulus heteroclitus). J. Fish. Res. Board Can.
28:999-1004.
1971b. Estimating distances between fish schooling in an
aquarium. J. Fish. Res. Board Can. 28:1805-1806.
VAN OLST, J. C, AND J. R. HUNTER.
1970. Some aspects of the organization of fish schools. J.
Fish. Res. Board Can. 27:1225-1238.
Williams, G. C.
1964. Measurement of consociation among fishes and
comments on the evolution of schooling. Publ. Mich.
State Univ. Mus. 2(7):351-383.
442
NORTHERN ANCHOVY SCHOOL SHAPES AS RELATED TO
PROBLEMS IN SCHOOL SIZE ESTIMATION
James L. Squire, Jr.'
ABSTRACT
Horizontal fish school profiles of the northern anchovy, Engraulis mordax, taken from day aerial
photographs and video tapes of school bioluminescence at night were examined to determine the
percentage of school area within a circular field of view and the school length and width ratios. Schools
observed during the day had an average length to width ratio of 2.09:1, at night the ratio was 2.53:1.
The percent coverage of the school's area in relation to a circle drawn tangent about the school averaged
42. 19c during the day and 29. 2^^ during the night. The effect of school shape on estimation of individual
school area as observed with a side-looking sonar was determined. School width measurements, similar
to that obtained by the sonar, were used to determine school area and indicated a possible average
overestimate of the actual school area of 1.72:1. The relation of school length and width to the error was
determined, indicating the gfreater the length to width ratio the greater the error.
Profiles offish schools as viewed and photographed
in the horizontal plane from an airborne platform
have been published by numerous authors.
Radakov (1972), in his review offish schooling,
described the characteristic horizontal shapes of
fish schools in nature as being very diverse and
extremely changeable. He stated that a spherical
shape of a school is the rarest of all and also that a
school's shape, size, or density is a result of the
interaction between the fish and the physical and
biological environment.
School shape and behavior in nature have been
studied with such techniques as aerial observa-
tion, hydroacoustic measurements, and underwa-
ter observation. Each of these methods has limita-
tions. Underwater visual observations are subject
to restrictions due to illumination and restricted
visibility. Aerial observation is limited in the day
by water transparency, illumination, and reflec-
tance from the water surface resulting from wind
and wave action. Visual observation of school
shape at night, as outlined by bioluminescent or-
ganisms, is limited to the moon's dark cycle or to
periods of no moon, and is affected by water trans-
parency and the density of bioluminescent or-
ganisms present in the water. Both day and night
observations are limited by the school's proximity
to the surface in relation to the factors affecting
water visibility.
Hydroacoustic observations using lower fre-
'Southwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038.
quency sounders of the type used commonly in
sonar fish surveys give imprecise images of fish
schools in the form of echograms that must be
interpreted. Greater resolution of fish school
shapes, but with limited range, can be obtained
with ultrasonic scanning equipment (Voglis and
Cook 1966).
All of these observation techniques may alter
the environment and in many cases may result in
modification of fish school behavior. Fish school
behavior is affected when in close proximity to
ships, submersibles, and divers, and aircraft
(noise, shadow) could possibly modify the school,
though this is not documented.
Surveys and research studies using variations
of these three observation techniques are cur-
rently in use for direct biomass estimation offish
populations by observation of individual schools,
school groups, and the internal structure of the
school.
Hydroacoustic research on schooling fish is cur-
rently being conducted by the Southwest Fisheries
Center (Smith 1970; Hewitt et al. 1976). Coastal,
hydroacoustic surveys are conducted by the State
of California (Mais 1974) to determine a relative
abundance estimate of the northern anchovy, En-
graulis mordax. These surveys are conducted dur-
ing the daylight hours, as comparative tests indi-
cate an increased probability of detection during
this period (Smith 1970).
Aerial observations by commercial fish spotters,
in the form of school counts and estimates of total
tonnage, are being used by the Southwest
Manuscript accepted August 1977.
FISHERY BULLETIN: VOL. 76. NO. 2. 1978.
443
FISHERY BULLETIN: VOL. 76, NO. 2
Fisheries Center to calculate indices of apparent
abundance for several coastal species, including
the northern anchovy (Squire 1972). To aid in the
detection and quantification of pilchard, Sar-
dinops ocellata, shoal occurrence off South Africa,
Cram (1974) used an airborne, low-light-level,
electron image intensifier to view the ocean's sur-
face, detecting the bioluminescence offish schools.
During these night aerial surveys the school's
horizontal surface area was interposed on the in-
strument's circular field of view, and running es-
timates of the percentage of coverage were made.
These percentage estimates were then used in the
computation of biomass estimates.
The intensifier used by the Sea Fisheries of
South Africa has been used by the author off the
southern California coast on an experimental
basis. Due to the highly variable school shapes
encountered, making estimates of the percentage
of school coverage in the circular field of view are
difficult. Experience indicated that examination
of aerial color photographs and night low-level
video tapes of anchovy school shapes for determi-
nation of the percentage of school coverage within
a circle would be useful, particularly if in the fu-
ture, surveys were to be conducted at night using
this method for the development of biomass esti-
mates for the northern anchovy and other near-
surface schooling pelagic species.
In addition, an analysis of anchovy horizontal
school shapes may assist hydroacoustic research-
ers in determining error parameters for computa-
tion of sonar biomass estimates. Hydroacoustic
surveys currently conducted for the northern an-
chovy use both side- and vertical-looking sonar to
detect and measure fish schools and school groups
during the day along a predetermined survey
track line. The acoustic "beam" used in these sur-
veys varies according to the unit and is of ±5° to
10°. When detecting the school, the side-looking
sonar measures the maximum dimension in one
aspect of the school, either normal to or parallel to
the ship's track. For the purpose of calculating
horizontal area, in contrast to the aircraft's verti-
cal view of the actual horizontal school area, the
echogram school width is assumed to be elliptical
(Smith 1970). Preliminary attempts at biomass
estimation from sonar surveys have used the sim-
ple assumption that a series of estimates of the
width of an elliptical school from random aspects
will result in an unbiased estimate of school hori-
zontal area. In a side-looking sonar the school
width is measured and provides two points of ref-
erence with the orientation of these points about
the school's profile being unknown. If an ellipse is
fitted randomly between these two points, the re-
sulting average area will equal a circle, a condi-
tion that was not observed in aerial photographs of
anchovy schools.
METHODS
To examine the shape of northern anchovy
schools as observed during day and night and to
determine what percentage the school occupies of
a circle tangent to two points along the school's
edge and containing the school inside the circle,
a circle was drawn about school profiles ob-
tained from a series of 20 day oblique aerial color
photographs (from the photographic files of the
Southwest Fisheries Center) and of 20 night
photographs of fish school bioluminescence. The
bioluminescent anchovy school shapes were
photogi'aphed from a television monitor as it pro-
jected video tapes recorded from an airborne low-
light-level television camera used during anchovy
resource surveys off northwestern Mexico. The
night photographs were made available through
the courtesy of Zapata, Inc.^ (Zapata Fisheries),
Houston, Tex. The night surveys using low-light-
level television were conducted at elevations of up
to 1,828 m (6,000 ft) and this survey technique is
effective because the northern anchovy commonly
migrates to the near-surface area during hours of
darkness (Squire 1972).
The actual area of the schools observed in the
photographs are unknown due to lack of data on
the aircraft's altitude, camera angle, and camera
geometry; however, all were taken from angles
approaching vertical. However, all area calcula-
tions are expressed in percentages of a circle
drawn tangent about the school's edge.
The day school profiles were further analyzed to
determine what the school area would be if the
width measurement were considered to be equal to
the school's diameter and what the area would be
if viewed systematically from six points 30° arc)
about an arc of 180° around the school (based on
school width or diameter as determined similar to
the measurements made from a hydroacoustic
sounder). These area data calculated from the six
points of observation to determine school width
were then compared to the actual school area.
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
444
SQUIRE: ANCHOVY SCHOOL SHAPES
School length to width ratios were determined
for both day and night schools.
The school area, as expressed in terms of
percentage of the circle drawn about it, was de-
termined by tracing the school profile upon paper,
cutting the school out of the circle and weighing
both the school profile and the nonschool portion of
the circle on a sensitive laboratory balance. School
areas for the six points of school width (diameter)
measurement about the 180° arc were computed
statistically.
RESULTS
Night Observations
Figure 1 illustrates the profiles of schools re-
corded by the airborne low-light-level television
system. On each figure are given the area percent-
age of a tangent circle that the fish school occupies
and the length to width ratio.
The average percent coverage of a night an-
chovy school, in relation to the circle tangent to
the school, is 29. 27^ The average ratio of school
length to width is 2.53:1.
Day Observations
Figure 2 illustrates profiles of schools observed
during the day. On each figure are given the per-
centage of a circle tangent to the school that is
occupied by the fish school, the length to width
ratio, and the 30° arc points that were used to
determine the school's estimated diameter and
area, simulated as randomly viewed by a side-
looking sonar. The ratio of the actual school area
to the calculated area of the school's average, high
and low estimate, as viewed every 30° of a 180° arc
based on simulated sonar measurements of width,
is given in Table 1.
The average percent coverage of a day anchovy
school to the tangent circle about the school is
\2.V7c. The average length to width ratio for all
day schools is 2.09:1. The ratio of estimate of the
area of all day schools, as calculated from mea-
surements from 30° arc points about the school, to
the actual area of the school is 1.72:1. The ratios,
length to width, and estimate of school area to
actual school area were compared and Figure 3
graphs the relationship. The graph displays the
variables plotted on log-log paper showing two
main points: One, that the variance is changing
proportionally to the mean. This is expected as
there should be more variation as the school
Table l. — Ratio of the actual anchovy school area to the aver-
age area based on six points of observation as viewed every 30° of
a 180° area, and the high and low ratio.
School
Average
High
Low
ID
2.61
1:3 95
1:0 83
2D
96
1:2 91
090
3D
22
1:1 56
0 85
4D
2
12
1:3,71
063
5D
13
1:1.41
080
6D
2
73
1:4.53
1.18
7D
50
1 2.35
060
8D
32
1 2 00
054
9D
49
1:1 88
1.19
10D
31
1:1 60
1 04
11D
2
65
1:4 82
062
12D
33
1:2.02
0.85
13D
47
1:2
35
0.77
14D
68
1:2
54
049
15D
97
1.3
30
065
16D
60
1:2
69
0.44
17D
38
1:2
10
0.84
180
38
1:2
00
068
19D
99
1:3
52
0.42
20D
77
1:2.90
0.66
length to width increases. Two, the plotted regres-
sion line indicates that more bias (higher esti-
mated actual school ratio) is introduced as the
school length to width ratio increases. The line is
significant at the 95'7f confidence interval as
proven by the ^test ( 1.98 <2.298). The confidence
limits are from 0.0545 to 0.734.
SUMMARY AND COMMENTS
The data on day/night school length to width
ratios support what is commonly known about the
schooling shapes of the northern anchovy. They
are more common in the near-surface area at
night, generally in large elongate thin surface
schools. These elongate schools tend to group to-
gether in the early morning hours and descend to
depth to form more compact schools during the day
(Mais 1974). Studies by Squire ( 1972) of aerial fish
spotter data show anchovy schools to be more fre-
quently observed, and observed in larger quan-
tities at night, when compared with day observa-
tions.
The schools percent area coverage of the tangent
circle at night is 12.9*^ less than its coverage dur-
ing the day and the length to width ratio is greater
by 0.44. In addition, analysis of school length to
width ratios compared to the ratio of estimated
school size to actual size (Figure 3) shows that as
the length to width ratio increases a greater error
in school area estimate will occur. Schools with a
length to width ratio of 2:1 have an estimated to
actual error of about 1.5:1 while more elongate
schools of a ratio of 3:1 have an estimated error of
about 1.75:1.
445
FISHERY BULLETIN: VOL. 76, NO. 2
NIGHT I NIGHT 2 NIGHT 3 ^'^^T "^
1195 14 98% 1420 1840% 1138 2967% '338 12 12%
NIGHTS ^'GHT6 NIGHT 7^ ^ NIGHTS^ ^
e (1) e €)
1456 1284% 1263 2159% 12 28 3425% 1135 4326%
NIGHT 9 NIGHT 10 NIGHT II NIGHT 12
(/) (J) e e
1381 2125% 1321 2868% 1297 2180% ||73 4050%
NIGHT 13 NIGHT 14 NIGHT 15 NIGHT ^6^
1300 2160% 1191 42 35% 12 51 2787% ' ^ ^^ 31 96 /«
NIGHT 17 NIGHT 18 NIGHT 19 NIGHT 20
1153 25 38% 12 00 2739% 1207 4075% 1126 6654%
Figure l. — Profiles of anchovy schools observed at night off southern California, indicating the width to length ratio
and the percentage of a tangent circle about the school.
446
SQUIRE: ANCHOVY SCHOOL SHAPES
DAY I DAY 2
I 2.05
DAY 5
3021%
DAY 6
DAY 3
I 139
DAY 7
6671%
I i 39
6978%
DAY 4
12.20
3152%
12.29
'K)84%
I 1.92
4918%
DAY 10
DAY II
I 127
57.59%
I 1.28
6240%
DAYJ2
I 3.54
2074%
1:1.59
50 36%
DAY 13
DAY 14
I 1.78
41.92%
DAY 15
DAY 16
1252
12.70
28.80%
h2.76
3721%
DAY 17
DAY 18
I 1,64
4760%
DAY 19
DAY 20
1285
2801%
1236
3392%
FIGURE 2. — Profiles of anchovy schools observed during the day off southern California, indicating the width to length ratio and
the percentage of a tangent circle about the school. Measurements of school width were taken at the six points (long arrow shaft)
indicated about a 180° arc.
447
FISHERY BULLETIN: VOL. 76, NO. 2
1.0
1.5 2.0
RATIO- SCHOOL
2.5 3.0 4.0 5.0
LENGTH TO WIDTH
Figure 3. — Regression plot for the ratio estimated anchovy
school size to actual size compared with the school length to
width ratio.
For simulated sonar observations of school
widths used in the calculation of school area, the
preliminary examination of these data indicates a
possible 1.72:1 average overestimate ofarea due to
school shape deviations from a circle or ellipse.
Fish schools, being highly variable in horizontal
profile, are probably equally complex in vertical
structure; the relationship of horizontal complex-
ity to vertical complexity is not known. Also un-
known is the question of whether the individual
school's axis is oriented in the same general direc-
tion within a group of schools, a possible factor
which, if it occurs, could provide a source of sub-
stantially higher or lower school area error esti-
mates from sonar track line surveys.
The problem of accurately estimating the per-
centage of school area within view of a low-light-
level viewer is difficult, as the examples of school
shapes within the target circle would indicate.
Parameters of human viewing error could be es-
tablished for this survey technique. However, the
conduct of surveys using a low-light-level televi-
sion system where the video signal can be recorded
and later electronically analyzed with the aid of an
image analyzer, should result in a higher degree of
survey accuracy.
School shapes were taken from photographs
randomly selected from an aerial photo file. Many
of the photos were taken in the nearshore areas.
There is the possibility that schools may be
slightly more elliptical in shape over deep water
than in the nearshore areas, but this is not
documented. If this were true the error estimate
would be reduced. This and other aspects of school
profile and orientation should be investigated
further and estimates of length to width ratios
from aerial surveys, done in conjunction with each
acoustic survey, may be useful for determination
of a correction factor for the acoustic data.
ACKNOWLEDGMENTS
The suggestions of Reuben Lasker and Paul
Smith and the assistance of Jim Zweifel in the
calculation of the weighted linear regression are
appreciated.
LITERATURE CITED
Cram, D. L.
1974. Rapid stock assessment of pilchard populations by
airrraft-bome remote sensors. Proc. 9th Int. Symp. on
Remote Sensing. Ann Arbor, 15-19 April, p. 1043-1050.
Hewitt, R. P., P. E. Smith, and J. C. Brown
1976. Development and use of sonar mapping for pelagic
stock assessment in the California Current area. Fish
Bull, U.S. 74:281-300.
Mais, K. F.
1974. Pelagic fish surveys in the California Cur-
rent. Calif Dep. Fish Game, Fish Bull. 162, 79 p.
RADAKOV, D. V.
1972. Schooling in the ecology offish. [In Russ.] Izdatel.
"Nauka," Moscow. (Engl, transl., 1973. 173 p. Isr. Pro-
gram Sci. Transl. Publ, John Wiley and Sons, N.Y.)
Smith, P, E.
1970. The horizontal dimensions and abundance of fish
schools in the upper mixed layer are measured by so-
nar. In G. B. Farquhar (editor), Proc. International
Symposium on Biological Sound Scattering in the Ocean,
p. 563-591. Maury Cent. Ocean Sci., Dep. Navy, Wash.,
D.C.
Squire, J, L., jr.
1972. Apparent abundance of some pelagic marine fishes
off the southern and central California coast as surveyed
by an airborne monitoring program. Fish. Bull., U.S.
70:1005-1019.
VOGLIS, G. M., AND J. C. COOK
1966. Underwater applications of an advanced acoustic
scanning equipment. Ultrasonics 4:1-9.
448
SYNERGISTIC EFFECTS OF ENVIRONMENTAL VARIABLES ON
THE METABOLISM OF THE COPEPOD EUTERPINA ACUTIFRONS
FROM TWO DIFFERENT AREAS OFF THE COAST OF THE
STATE OF SAO PAULO, BRAZIL^
G. Scares Moreira^ and W. B. Vernberg^
ABSTRACT
The combined efFects of temperature and salinity on the respiratory rate of two populations of the
copepod Euterpina acutifrons have been determined. One population was taken from a nonpolluted
area, Sao Sebastiao Channel, and the other from a polluted area, Santos Bay, both off the coast of the
State of Sao Paulo, Brazil. Four groups of copepods were used in the experiments: 1) Sao Sebastiao
animals kept in Sao Sebastiao water (35%« salinity); 2 1 Santos animals kept in Santos water (28%o
salinity); 3) Sao Sebastiao animals kept in Santos water; and 4) Sao Sebastiao animals kept in diluted
Sao Sebastiao water (28%o salinity). Results showed that Sao Sebastiao copepods in either full strength
seawater (35%«) or lower salinity seawater (28%o) could metabolically regulate over a wider range of
salinities than could Santos copepods in Santos water or Sao Sebastiao copepods maintained in Santos
water. It was concluded that the water quality of the Santos Bay was responsible for changes in the
metabolic regulatory capacity of the copepods exposed to Santos water.
The planktonic harpacticoid Euterpina acutifrons
(Dana) is distributed in the warm waters of the
world between lat. 66°N and 40°S (Haq 1972). It is
a euryhaline species and has been reported in
salinities ranging from 8%o (Cananeia Estuary,
southern Brazil, Tundisi 1972) to 39%o (Mediter-
ranean Sea, El-Maghraby 1965). Laboratory
studies have shown that reproduction can occur
over a salinity range of 15 to 45%o (Moreira and
Yamashita 1975). Euterpina acutifrons is an im-
portant link in the marine trophic web serving as
food source for both adult and larval fishes
(Pouchet and de Guerne 1887; Lebour 1918; Blin
1923; Carvalho 1945; Marques 1951; Thayer et al.
1974).
In an earlier paper, Moreira (1975) reported
that salinity tolerances for Brazilian populations
of £. acutifrons from Santos were very different
from those of populations of this species from Sao
Sebastiao. This in itself is not surprising since the
salinity regimes of the two areas are different. The
salinity in the Santos Estuary varies widely from
17 to 30%o depending on the tide and season of the
'Contribution no. 214 of the Belle W. Baruch Institute for
Marine Biology and Coastal Research.
^Physiology Department, Institute of Biosciences, and Insti-
tute of Marine Biology, University of Sao Paulo, Brazil.
^School of Pubhc Health, Belle W. Baruch Institute for Marine
Biology and Coastal Research and Department of Biology, Uni-
versity of South Carolina, Columbia, SC 29208.
Manuscript accepted September 1977.
FISHERY BULLETIN: VOL. 76, NO. 2, 1978.
year, while in Sao Sebastiao Channel the salinity
is approximately 35%o throughout the year. Water
temperatures in both areas are essentially the
same, ranging from 19° to 30°C depending upon
season. It was not determined, however, if the
observed differences in salinity tolerances of the
two populations were genetically or environmen-
tally induced. Subsequently, a study was initiated
to resolve this question by measuring metabolic
response patterns of specimens from both popula-
tions to different thermal-salinity regimes. It soon
became apparent that environmental parameters
other than temperature and salinity were factors
in determining the metabolic response patterns of
these copepods.
A detailed chemical analysis of the water in
Santos Bay is not available, but great numbers of
tankers and other vessels continuously operate
near shore, discharging ballast water and con-
taminating seawater and adjacent regions with
petroleum. In addition, there are a large number
of industries that discharge wastes directly into
the water. One sample analysis of Santos Bay
seawater was found to contain 270 ppb lead and
200 ppb nickel (unpublished data). Furthermore,
to minimize the effects of human waste or degra-
dation products, approximately 400 tons of
chlorine are added monthly near shore. The data
presented in this paper demonstrate that
449
FISHERY BULLETIN: VOL 76, NO. 2
metabolic response patterns of £. acutifrons to the
normal fluctuations found in estuarine systems
were significantly altered by the water quality of
Santos Bay water.
MATERIALS AND METHODS
The copepods were collected in two fixed loca-
tions off the coast of the State of Sao Paulo, one in
Santos Bay (lat. 23°59'S; long. 46"19'W), the
other in Sao Sebastiao Channel (lat. 23°50'S;
long. 45°25'S). Collections were made with a
nylon plankton net (20 m/x) during the winter
when water temperatures averaged approxi-
mately 21°C. All samples were brought im-
mediately into the laboratory whereupon E.
acutifrons were sorted from the plankton using a
mouth pipette under a binocular microscope. The
copepods were placed in 2-1 crystallizing dishes, 20
cm in diameter and 15 cm high. In the first two
series of experiments, the copepods were placed in
the water obtained from the collection points, i.e.,
copepods from Sao Sebastiao Channel were placed
in Sao Sebastiao water (35%o) and copepods from
Santos were placed in Santos water (28%o). In the
last two series of experiments, copepods from Sao
Sebastiao were placed either in Santos water or in
Sao Sebastiao water diluted to 28%o. The copepods
were maintained under temperature and photo-
period regimes approximating field conditions:
19°-24°C and 11 L:13 D. The copepods were fed
with Phaedactylum and Platymonas daily and
kept in the laboratory at least 1 wk before being
used in the respiration experiments.
Oxygen uptake was determined using Carte-
sian diver respirometers (Holter 1941), which
have a total volume of 8-13 />tl. Only nongravid
females were used. Two or three copepods were
placed in each diver, depending upon the salinity/
temperature regime of the experiment. The oxy-
gen uptake was determined during a 2-h interval.
The first 30-min reading was discarded; after this
initial reading, uptake rates remained constant.
Oxygen uptake rates were measured under the
following environmental conditions: Sao Sebas-
tiao animals maintained in Sao Sebastiao water,
Santos animals in Santos water, and Sao Sebas-
tiao animals in Santos water, 15°, 20°, 25°, 30°,
and 32°C at 15, 25, 35, 45, and 55%o salinities. The
oxygen uptake of Sao Sebastiao animals main-
tained in Sao Sebastiao water diluted to 28%o was
determined at 15°, 25°, and 30°C over the same
salinity ranges used in the other experiments. Ten
determinations were made under each set of en-
vironmental conditions. Distilled water or freeze-
concentrated brine was added to filtered seawater
to attain the desired salinities. Salinities were
determined by titrating against silver nitrate
(Harvey 1955).
Dry weights for the copepods were obtained
using a TorbaP torsion balance, 0.01 mg sensitiv-
ity. The copepods were rinsed with distilled water
and dried at 70 °C for 24 h before they were
weighed. Three replicates of 200 nongravid fe-
males from each area of collection were used. Re-
sults were expressed as microliters of oxygen per
milligram per hour. Significant difference of
means was calculated by the method of Simpson et
al. (1960) for small samples.
The metabolic data obtained in the first three
series of experiments were analyzed statistically
using multiple regression techniques. The basic
experimental design used in this study is usually
referred as a factorial design. Specifically, the plan
was a 5 X 5 factorial using five levels of tempera-
ture and five levels of salinity, making in all 25
combinations of experimental conditions. Since 10
determinations of oxygen uptake were made in
each combination, a total of 250 observations were
made in each series. Thus, the 250 observations
may reasonably be considered as continuous re-
sponses of a function of the two factors and interac-
tions.
Oxygen uptake data were analyzed as percent-
age of oxygen consumption relative to that at
25°C and 35%o salinity for Sao Sebastiao animals
in Sao Sebastiao water and at 25°C and 25%o salin-
ity for Santos animals in Santos water and Sao
Sebastiao animals in Santos water, i.e., the rate
under these "standard" conditions was assumed to
be lOO'/f , and rates obtained under other regimes
were calculated as the percent deviation from that
rate. Since the observations are treated as percent-
age measurements generated by data from bino-
mial populations, the transformation Y = arc sin
v'jc, where x is observed percent respiration, is
appropriate to stabilize variances (Mendenhall
1968). Analysis of variance for this data indicated
which of the factors (temperature, salinity, or
temperature-salinity interactions) had significant
effects on the metabolism of the copepods. The
program was run on an IBM 360 computer.
''Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
450
MOREIRA and VERNBERG: SYNERGISTIC EFFECTS OF ENVIRONMENTAL VARIABLES
RESULTS
Animals From Sao Sebastiao
in Sao Sebastiao Water (35%o)
The metabolic rate ofEuterpina acutifrons from
Sao Sebastiao which were maintained in water
from Sao Sebastiao Channel was less influenced
by changes in salinity than were the other groups
o{ E. acutifrons (Figure 1). Rate did, however, in-
crease with increasing temperature up to 30°C; at
32 °C, rates either leveled offer decreased. Great-
est increases in respiration rates were observed
between 15° and 20°C. These increases are
reflected in the relatively high metabolic rates
within this thermal range obtained from this
group of animals.
At 20°C metabolic rates were not significantly
different over the entire salinity range of 15-55%o,
and at 15°C the copepods were able to regulate
their metabolism over a range of 25-55%o. At
higher temperatures (25°, 30°, and 32°C) rates
generally were lower at the salinity extremes (15,
45,55%o) and highest at 35%o. Rates varied from a
minimum of 4.87 to a maximum of 22.36 /xl/mg h ^^
dry weight (Figure 1).
Statistical analysis indicated that 66% of the
observed variability in the rates could be
explained by the temperature-salinity combina-
tions (Table 1), although the linear effect of tem-
perature was the single significant factor (1%
level). The linear effect of salinity, the quadratic
effects of the temperature and salinity, and the
temperature-salinity interaction did not contri-
bute significantly to the observed changes in res-
piration rates. Figure 2A shows the response sur-
face contours fitted over the experimental design.
Table l. — Analysis of variance of data for
Euterpina acutifrons (Sao Sebastiao animals
in Sao Sebastiao water, 35%o). T = tempera-
ture, S = salinity.
Variable
r2
Significance level
T
054953
1%
72
0,65769
Not significant
S
0.65796
Not significant
S2
0.65885
Not significant
Tx S
065933
Not significant
Animals From Santos in
Santos Water (28%o)
In the Santos animals maintained in Santos wa-
ter, the rate of oxygen uptake also increased over
the temperature range to 30°C for the entire salin-
ity range. In most of the salinities, the largest
metabolic increase occurred between 25° and
30°C. This contrasts with Sao Sebastiao copepods
which exhibited the largest increase between 15°
and 20°C. The Santos animals did not show the
metabolic regulation observed in the Sao Sebas-
tiao animals maintained in Sao Sebastiao water,
and tended to have low metabolic rates at the
salinity extremes, i.e., 15, 45, and 55%o. Highest
rates occurred at salinities of 25-35%o. The rates
varied from a minimum of 7.97 to a maximum of
28.20 Atl/mg h^Mry weight (Figure 1).
Statistical analysis indicated that only 46% of
the observed variability in the respiration rates of
these animals could be explained by the
temperature-salinity combinations (Table 2). The
significant factors were the quadratic effects of
temperature and salinity (0.05% level) and the
linear effect of salinity (0.05% level). The
temperature-salinity interaction was not a sig-
nificant factor. Figure 2B shows the response sur-
face contours fitted over the experimental design.
Table 2. — Analysis of variance of data for
Euterpina acutifrons (Santos animals in San-
tos water 28%o). T = temperature, S = salin-
ity.
Variable
Significance level
T2
S^
S
T X S
024350
0,28955
046193
046198
0.05%
0.05%
0.05%
Not significant
Animals From Sao Sebastiao in
Santos Water
Transfer of Sao Sebastiao copepods into water
from Santos markedly altered their metabolic re-
sponses, especially their response to salinity. Res-
piration rates increased with temperature up to
25°C at the extreme salinities (15, 45, 55%o) and
up to 30°C at salinities of 25 and 35%o, before
leveling off or decreasing. The copepods which
were transferred to Santos water did not regulate
metabolically at any temperature at the salinity
extremes. Lowest rates were obtained at salinities
of 15 and 55%o, and the highest rates were ob-
served at 25%o at 15° and 25°C. At 30° and 32°C,
peak metabolic rates occurred at 35%o (Figure 1).
Rates varied from a minimum of 6.80 to a
maximum of 37.23 /xl/mg h^ dry weight (Figure
1).
451
FISHERY BULLETIN; VOL. 76. NO. 2
15
1 10
> 9
■c 8
r 7
y-
y /
,1
,) ..- ■.. 1
1
■c 6
4
,
15 25 35 45 55
Salinity %o
20*'C
- 30
,
^ ^
5 25
-5 20
. ,'-'' ^%
en
E
' M.
/ V
j: 15
• / JJ
3.
l_/ ■
"""^ ^1^^ ""h
10
■ / ' ^^
9
15 25 35 45
Salinity %o
55
35
30
^ 25
?15
20 ■
10
9
8
25''C
15
25
35
Salinity %o
45
55
40
35
D
§
30
>
■c
25
ai
P
10
^
CN
O
lb
3.
o\j \^
r
/'-^
/ ^
' K
^\y-^^>^'
\
•I \
/ yy^
1 ■■■.\
K
\
/y^
X
■ L>
M
■■■^r~~"vi
*f 1 -.j:J|i
15 25 35 45 55
Salinity %o
40
35
25 ■
I 30
■a 20
E
■^ 15
O
10
9
8
32*0
15 25 35 45
Salinity %o
55
Figure l. — Euterpina acutifrons: metabolic rate of animals from
Sao Sebastiao maintained in Sao Sebastiao water (solid line), ani-
mals from Santos maintained in Santos water (dashed line), ani-
mals from Sao Sebastiao maintained in Santos water (dot-dash
line), and animals from Sao Sebastiao maintained in Sao Sebastiao
diluted water (dotted line), in different temperature and salinity
regimes.
452
MOREIRA and VERNBERG: SYNERGISTIC EFFECTS OF ENVIRONMENTAL VARIABLES
15
20 25
Temperature °C
30
15
20 25
Temperature °C
15
20 25
Temperature "C
Figure 2. — Response-surface estimation of percent metabolic
rate of Euterpina acutifrons maintained at 25 different tempera-
ture and salinity combinations. A. Copepods from Sao Sebastiao
maintained in Sao Sebastiao water. B. Copepods from Santos
maintained in Santos water. C. Copepods from Sao Sebastiao
maintained in Santos water.
Table 3. — Analysis of variance of data for
Euterpina acutifrons (Sao Sebastiao animals
in Santos water, 28%o). T = temperature, S =
salinity.
Variable
/•2
Significance level
T
0-17164
0.05%
S2
0-32500
0.05%
8
046808
0.05%
T2
0 59715
0.05%
T ■ S
062298
0.05%
All of the analyzed factors, i.e., linear tempera-
ture and salinity, as well as the quadratic effect of
these factors, and the temperature-salinity in-
teraction, contributed significantly (0.057f level)
to the observed variability in respiration rates
(Table 3). A total of 62% of the variability could be
explained by these various factors. Figure 2C
shows the response surface contours fitted over the
experimental design.
Animals From Sao Sebastiao in
Sao Sebastiao Diluted Water
In this series the respiration experiments were
run at three temperatures to test whether or not
the results obtained for Santos animals were the
result of acclimation to a lower salinity. The respi-
ration rates at 28%o were essentially the same as
those for animals maintained in undiluted Sao
453
FISHERY BULLETIN: VOL. 76, NO. 2
Sebastiao water. At 15°, 25°, and 30°C, metabolic
rates were not significantly different over the
range of 15-45%o. The rates varied from a
minimum of 4.32 to a maximum of 19.17 ix\/mg h '
dry weight (Figure 1).
DISCUSSION
In Brazilian waters, populations of £■. acutifrons
thrive over a wide range of salinities and variable
salinity alone does not seem to be a limiting factor
in their distributional patterns (Tundisi 1972;
Moreira and Yamashita 1975). Indeed, of the vari-
ous environmental variables tested, temperature
alone significantly affected the metabolic rates of
these copepods. The present data demonstrate
that specimens of copepods from the unpolluted
Sao Sebastiao Channel have the capability of
metabolic regulation over a wide range of
salinities when tested using Sao Sebastiao water.
On the other hand, marked diminution in the
capability to regulate metabolically at salinity ex-
tremes was noted in E. acutifrons from the Santos
population and specimens from Sao Sebastiao
maintained in Santos water. For both groups of
animals salinity, as well as temperature, proved to
exert a statistically significant effect at the 5%
level (or less) on their oxygen uptake rates. These
marked changes in metabolic control in the
copepods taken from or exposed to Santos water
compared with that of copepods from Sao Sebas-
tiao are depicted in Figure 2.
While we did not measure population densities
of the E. acutifrons in our two study areas (Santos
Bay and Sao Sebastiao Channel), there is some
indication in the literature that population size is
sensitive to polluted waters. Gabriel et al. (1975)
reported a decrease in abundance of this species
in the Milford Haven Estuary following its de-
velopment into the largest oil port in the United
Kingdom in the 1960's, and there are several
examples that indicate that pollutants can affect
the survival of copepods and planktonic larvae.
Barnes and Stanbury ( 1948) have studied the toxic
action of copper and mercury salts on the copepod
Nitocra spinipes and verified that mercuric
chloride is a very effective poison; in contrast,
these animals are very resistant to copper.
D'Agostino and Finney (1974) have found that
copper and cadmium inhibit growth and develop-
ment of the copepod Tigriopus Japonicus at 0.064
mg/1 and 0.044 mg/1, respectively. Heinle (1969)
suggested that the high mortality rate ofAcartia
tonsa in a power plant effluent was due to the
chlorination of the cooling water, correlating the
apparent periodicity in the mortality rate with the
chlorination schedule. Latimer et al. (1975)
studied the toxicity of 30-min exposures of re-
sidual chlorine to two species of copepods, Lim-
nocalanus macrurus and Cyclops bicuspidatus
thomasi. The predicted "safe" concentrations were
0.9 mg/1 for L. macrurus and 0.5 mg/1 for C. b.
thomasi. Roberts et al. (1975) studied the acute
toxicity of chlorine to some estuarine species, in-
cluding molluscan larvae, copepods, shrimps, and
fishes. They found that molluscan larvae and
Acartia tonsa were the most sensitive species
tested, with 48-h TLg,, values at chlorine levels
<0.005 ppm. Gray (1974) demonstrated that lead
(Pb(N03)2) at 0.3 ppm reduced the growth rate of
the marine ciliate protozoan Cristigera by 11.7%
and at 0.15 ppm by 8.46%. Mercury was found to
have an effect on survival, metabolism, and be-
havior of the planktonic larvae of Uca pugilator
(DeCoursey and Vernberg 1972; Vernberg et al.
1973). Generally, larvae are much more sensitive
to toxicants than are adults and very low concen-
trations of a toxicant can interact with environ-
mental factors to cause increased mortalities
among larvae (Vernberg 1975).
Detailed chemical analyses of Santos water ob-
viously are needed, but the very high concentra-
tion of lead and nickel which were found in one
sample, plus the oil and other industrial effluents
that are being discharged, leave little doubt that
the Santos Estuary is highly polluted. Data pre-
sented in this paper strongly suggest that speci-
mens living in the Santos Estuary do so at a high
cost energetically. This high metabolic cost for
survival following exposure to salinity extremes
would almost certainly be a factor limiting the
distribution of £. acutifrons in polluted estuaries,
since fluctuating salinity regimes are characteris-
tic of this environment. Results obtained in this
study highlight the fact that the physiological re-
sponses of marine organisms may be markedly
modified if test animals are taken from or exposed
to polluted waters.
LITERATURE CITED
Barnes, H., and F. A. Stanbury.
1948. The toxic action of copper and mercury salts both
separately and when mixed on the harpacticid copepod,
Nitocra spinipes (Boeck). J. Exp. Biol. 25:270-275.
454
MOREIRA and VERNBERG: SYNERGISTIC EFFECTS OF ENVIRONMENTAL VARIABLES
BLIN. [F.)
1923. Note sur ralimentation de la Sardine. Euterpes et
Sardines. Bull. Soc. Zool. Fr. 48:99-105.
CARVALHO, J. DE PAIVA.
1945. Copepodos de Caioba e Taia de Guaratuba. Arq.
Mus. Parana. 4:83-116.
D'AGOSTINO, A., AND C. FiNNEY.
1974. The effect of copper and cadmium on the develop-
ment ofTigriopusjaponicus. In F. J. Vernberg and W. B.
Vemberg (editors). Pollution and physiology of marine
organisms, p. 445-463. Academic Press, N.Y.
DECOURSEY, p. J., AND W. B. VERNBERG.
1972. Effect of mercury on survival, metabolism and be-
haviour of larval Uca pugilator (Brachyura). Oikos
23:241-247.
El-Maghraby, a. M.
1965. The seasonal variations in length of some marine
planktonic copepods from the eastern Mediterranean at
Alexandria. Crustaceana 8:37-47.
Gabriel, P. L., N. S. Bias, and a. nelson-Smith.
1975. Temporal changes in the plankton of an indus-
trialized estuary. Estuarine Coastal Mar. Sci. 3:145-
151.
Gray, J. S.
1974. Synergistic effects of three heavy metals on growth
rates of a marine ciliate protozoan. In F. J. Vernberg and
W. B. Vemberg (editors). Pollution and physiology of
marine organisms, p. 465-485. Academic Press, N.Y.
Haq, S. M.
1972. Breeding of Euterpina acutifrons, a harpacticid
copepod, with special reference to dimorphic males. Mar.
Biol. (Berl.) 15:221-235.
Harvey, H. W.
1955. The chemistry and fertility of sea waters. Camb.
Univ. Press, N.Y., 224 p.
Heinle, D. R.
1969. Temperature and zooplankton. Chesapeake Sci.
10:186-209.
Holter, H.
1941. Technique of the Cartesian diver. C. R. Trav. Lab.
Carlsberg (Chim.) 24:399-478.
Latimer, D. L., A. S. Brooks, and A. M. Beeton.
1975. Toxicity of 30-minute exposures of residual chlorine
to the copepods Limnocalanus macrurus and Cyclops
bicuspidatus thomasi. J. Fish. Res. Board Can.
32:2495-2501.
LEBOUR, M. V.
1918. The food of post-larval fish. J. Mar. Biol. Assoc.
U.K. 11:433-469.
Marques, E.
1951. Copepodes encontrados no conteudo gastrico de
algun clupeideos da Guine Portuguesa. Anais Junta In-
vest. Colon. 6:11-18.
MENDENHALL, W.
1968. Introduction to linear models and the design and
analysis of experiments. Wadsworth, Belmont, Calif.,
465 p.
Moreira, G. S.
1975. Studies on the salinity resistance of the copepod
Euterpina acutifrons (Dana). In F. J. Vemberg (editor),
Physiological ecology of estuarine organisms, p. 73-
79. Belle W. Baruch Libr. Mar. Sci. 3, Univ. S.C. Press,
Columbia.
Moreira, G. S., and C. Yamashita.
1975. Influencia de la salinidad en la reproduction y desar-
rollo de Euterpina acutifrons (Dana). In Memorias I
Simposion Latino-americano sobre Oceanografia
Biologica (Mexico), p. 236-245.
POUCHET, G., AND J. DEGUERNE.
1877. Sur la nourriture de la Sardine. C. R. Acad. Sci.,
Paris 104:712-715.
ROBERTS, M. H., Jr., R. J. DIAZ, M. E. BENDER, AND R. J.
HUGGETT.
1975. Acute toxicity of chlorine to selected estuarine
species. J. Fish. Res. Board Can. 32:2525-2528.
Simpson, G. G., a. Roe, and R. C. Lewontin.
I960. Quantitative zoology. Revised ed. Harcourt, Brace
and Co., N.Y., 450 p.
THAYER, G. W., D. E. HOSS, M. A. KJELSON, W. F. HETTLER,
JR., AND M. W. LaCROIX.
1974. Biomass of zooplankton in the New^port River es-
tuary and the influence of postlarval fishes. Chesapeake
Sci. 15:9-16.
TUNDISI, T. M.
1972. Aspectos ecologicos do zooplankton da jegiao la-
gunas de Cannaneia com especial referenda aos Copepoda
(Crustacea). Ph.D. Thesis, Sao Paulo Univ., Sao Paulo.
Vernberg, W. B.
1975. Multiple factor effects on animals. In F. J. Vem-
berg (editor), Physiological adaptation to the environ-
ment, p. 521-537. Intext Educational Publishers, N.Y.
Vernberg, W. B., p. J. deCoursey, and W. J. Padgett.
1973. Synergistic effects of environmental variables on
larvae of Uca pugilator. Mar. Biol. (Berl.) 22:307-312.
455
DESCRIPTION OF LARVAE OF A HIPPOLYTID SHRIMP, LEBBEUS
GROENLANDICUS, REARED IN SITU IN KACHEMAK BAY, ALASKA
Evan Haynes^
ABSTRACT
Larvae of Lebbeus groenlandicus , a hippolytid shrimp, were reared in situ in Kachemak Bay, Alaska,
from the first zoea (Stage I) through themegalopa (Stage III). Each of the three stages is described and
illustrated, and then compared with descriptions of larvae of Lebbeus spp. given by other authors.
Information on the larval stages of the genus Leb-
beus is meager. Pike and Williamson (1961), in
their summary of the generic characteristics of
Spirontocaris and related genera, note that the
only larva o^ Lebbeus known for certain is a larva
of L. polaris dissected from a well-developed egg.
During studies on rearing larvae of pandalid
shrimp for descriptive purposes (Haynes 1976,
1978), I succeeded in rearing larvae of L. groen-
landicus to the megalopa stage. This report de-
scribes and illustrates each of the two zoeal stages
and megalopa of L. groenlandicus , and compares
the stages obtained from rearing in situ and from
plankton in Kachemak Bay with provisionally
identified larvae of L. groenlandicus reported by
other authors.
METHODS
A complete discussion of rearing technique,
methods of measurement, techniques of illustra-
tion, and nomenclature of gills and appendages is
given by Haynes (1976). Briefly, the rearing
technique consists of obtaining Stage I zoeae from
known parentage in the laboratory and then rear-
ing the zoeae to postlarvae in 500-ml flasks sus-
pended upright beneath the surface of the sea.
Cast skins and larvae removed from the flasks
were examined in the laboratory to determine
sequence and morphology of each stage. Larval
stage was also verified using larvae from plankton
reared in the same manner as larvae obtained in
the laboratory.
In the illustrations (Figures 1-3), for clarity,
setules on setae are usually omitted but spinulose
'Northwest and Alaska Fisheries Center Auke Bay Labora-
tory, National Marine Fisheries Service, NOAA, P.O. Box 155,
Auke Bay, AK 9982L
Manuscript accepted November 1977.
FISHERY BULLETIN: VOL. 76, NO. 2,
setae are shown. The terms are defined as follows:
setose - set with bristles (setae)
spinose - bearing many spines
spinous - spinelike
spinulose - set with little spines.
The figures are in part schematic and represent
typical setal counts.
STAGE I ZOEA
Total length of Stage I (Figure lA) 6.9 mm
(range 6.4-7.4 mm; 10 specimens). Live specimens
characterized by bright orange color extending
along ventral surface of body from antennules to
fourth abdominal segment, orange gut, small
orange chromatophore at anus, and greenish in-
ternal thoracic organs; remainder of zoea trans-
lucent. Rostrum slightly sinuate, without teeth,
about two-thirds length of carapace. Carapace
with dorsal rounded prominence at base of ros-
trum and near posterior edge; no supraorbital
spines. Usually at least two minute spinules occur
along ventral margin of carapace immediately
posterior to pterygostomian spine.
ANTENNULE (FIGURE IB).— First antenna, or
antennule, consists of an unsegmented cylindrical
basal portion and two distal conical projections;
largest conical projection bears four aesthetascs of
various lengths; smallest conical projection bears
a single heavily plumose seta.
ANTENNA (FIGURE IC).— Consists of inner
flagellum (endopodite) and outer antennal scale
(exopodite). Flagellum two-segmented, about
twice length of scale; distal segment styliform and
terminating in narrow projection. Two simple
457
1978.
FISHERY BULLETIN: VOL. 76, NO. 2
J
L
J
0. 5 mm 0.25 mm
Figure l. — stage I zoea ofLebbeus groenlandicus: A, whole animal; B, antennule; C, antenna; D, mandibles (right and left).
458
HAYNES: LEBBEUS GROEXLANDICUS LARVAE
1 . 0 mm
Figure l. — Stage I zoea of Lebbeusgroenlandicus: E, maxillule; F, scaphognathite of maxilla; G, first maxilliped; H, second maxilliped.
setae occur at joint. Antennal scale distally di-
vided into four segments (proximal joint often in-
complete) and fringed with 11 heavily plumose
setae along terminal and inner margins. A
small seta often occurs proximally near lateral
margin. Protopodite bears two simple spines ven-
trally, one at base of flagellum and one at base of
scale.
MANDIBLES (FIGURE ID).— Without palps;
459
FISHERY BULLETIN: VOL 76, NO. 2
0.5 mm
1 . 0 mm
Figure L — Stage I zoea of Lebbeus groenlandicus: I, third maxilliped; J, first pereopod; K, third pereopod, L, first pleopod; M, second
pleopod; N, telson.
well developed. Incisor process of left mandible
bears five teeth, one of them located near movable
premolar denticle (lacinia mobilis), in contrast to
triserrate incisor process of right mandible. Both
mandibles bear well-developed denticles along
terminal margin. Truncated end of molar process
of right mandible formed into curved lip. Only left
mandible bears a subterminal process.
460
HAYNES: LEBBEUS GROENLANDICUS LARVAE
MAXILLULE (FIGURE IE).— First maxilla, or
maxillule, bears coxal and basial endites and an
endopodite. Proximal lobe (coxopodite) bears 15
setae, most of them spinulose. Median lobe
(basipodite) bears 24 spines terminally, 9 of them
spinulose; and 2 spines subterminally, 1 of them
plumose and the other simple. A series of fine hairs
occurs in vicinity of the simple spine. Endopodite
originates from lateral margin of basipodite and
bears three terminal and two subterminal
spinulose setae. No evidence of outer seta on
maxillule.
MAXILLA (FIGURE IF).— Bears platelike
exopodite (scaphognathite) with 33 long, plumose
setae along outer margin, and a longer, thick seta
at the proximal end. Endopodite not segmented;
setae spinous, setation formula 2, 2, 1, 2, 3. Both
basipodite and coxopodite bilobed. Basipodite
bears 29 setae, 14 on distal lobe and 15 on proxi-
mal lobe. Coxopodite bears 23 setae, 5 on distal
lobe and 18 on proximal lobe.
FIRST MAXILLIPED (FIGURE IG).—
Protopodite segmented; bears 27 setae on distal
segment and 8 on proximal segment, most of them
spinulose. Endopodite four-segmented; setation
formula 4, 3, 3, 7. Basal segment of endopodite
bears conspicuous setulose spine. Exopodite seg-
mented at base; bears four natatory setae. Epipo-
dite distinctly bilobed.
SECOND MAXILLIPED (FIGURE IH).—
Protopodite not segmented; bears nine setae, five
of them spinulose. Endopodite five-segmented;
fourth segment expanded somewhat laterally;
terminal segment tipped by five setae and bears
single seta subterminally; basal segment bears
conspicuous setulose spine like that on basal seg-
ment of endopodite of first maxilliped; setation
formula 6, 4, 2, 3, 4. Exopodite about three times
longer than endopodite; bears five natatory setae.
Epipodite present but not bilobed.
THIRD MAXILLIPED (FIGURE II).—
Protopodite not segmented; bears three setae. En-
dopodite five-segmented; as long as exopodite;
number of setae somewhat variable. Exopodite
bears five natatory setae. No epipodite.
FIRST PEREOPOD (FIGURE IJ).— Endopodite
relatively short, wide, and partially segmented;
chela partially formed; dactylopodite bears three
simple spines. Exopodite a small lobe.
SECOND PEREOPOD.— Similar in shape to first
pereopod except narrower, exopodite smaller, and
chela more deeply cleft.
THIRD (FIGURE IK) TO FIFTH PEREO-
PODS. — Each pair essentially identical except
that they decrease slightly in size from third to
fifth. No exopodites.
PLEOPODS.— First pleopod (Figure IL) slightly
cleft, without joints or setae. Second pleopod (Fig-
ure IM) bilobed; outer lamella segmented; inner
lamella usually only partially segmented but
bears bud of appendix interna. Third to fifth
pleopods essentially identical to second pleopod
except both lamellae distinctly segmented.
ABDOMEN AND TELSON (FIGURES lA,
IN). — Abdomen consists of five segments and tel-
son (somite six is fused with telson in Stage I).
Fourth and fifth abdominal segments each with
pair of posterolateral spines nearly as long as
segments themselves. Telson slightly emar-
ginated distally; bears 19-21 densely plumose
setae; small spinules occur between bases of all
setae except two outermost pairs. Enclosed
uropods visible. Anal spine present.
STAGE II ZOEA
Total length of Stage II 8.3 mm (range 8.1-8.7
mm; 8 specimens). Color similar to Stage I zoea but
more diffuse. Rostrum (Figure 2 A) arched upward;
slightly blunter than in Stage I; without teeth.
Carapace bears supraorbital, antennal, and
pterygostomian spines in addition to several
spinules along anteroventral margin.
ANTENNULE (FIGURE 2B).— Shows considera-
ble change from Stage I. Largest conical projection
segmented at tip; terminal segment bears three
setae of different lengths; proximal segment bears
six groups of five aesthetascs each in addition to
row of four aesthetascs laterally and single seta
distally. Smallest conical projection bears three
nonplumose setae, one long and two short. Pedun-
cle of antennule rounded laterally, not segmented,
and bears five plumose setae that originate ven-
trally.
ANTENNA (FIGURE 2C).— Flagellum of an-
tenna still two-segmented, but slightly stouter
and projection at tip smaller than in Stage I; a few
461
FISHERY BULLETIN: VOL 76, NO 2
0. 5 mm
0. 5 mm
Figure 2.— Stage II zoea of Lebbeus groenlandicus: A, carapace; B, antennule; C, antenna; D, first pereopod; E, second pereopod; F,
second pleopod.
462
HAYNES: LEBBEUS GROENLANDICUS LARVAE
small setae occur along lateral margin. Antennal
scale distally divided into two segments and
fringed with 29 or 30 thin, plumose setae along
terminal and inner margins; distal outer projec-
tion a stout spine. Protopodite bears two stout
spines, one at base of flagellum and other at base of
scale.
MANDIBLES, MAXILLULE, AND MAX-
ILLA.— Essentially identical to Stage I except
scaphognathite of maxilla usually bears 35 setae
along outer margin, in addition to the longer and
thicker seta at proximal end, and proximal cleft
slightly deeper.
MAXILLIPEDS.— Essentially identical to Stage I
except exopodites of first, second, and third maxil-
lipeds bear 5, 16, and 16 natatory setae, respec-
tively.
FIRST PEREOPOD (FIGURE 2D).— Segmented;
without exopodite; chela functional.
SECOND PEREOPOD (FIGURE 2E).— Adult in
shape; chela functional; ischiopodite articulates
somewhat laterally with basipodite. No exopodite.
THIRD TO FIFTH PEREOPODS.— Similar to
Stage I except ischiopodite articulates somewhat
laterally with meropodite and basipodite.
PLEOPODS.— First pleopod slightly more de-
veloped than in Stage I but still only about one-
third length of second pleopod and without appen-
dix interna. Second pleopod ( Figure 2F) larger and
narrower than in Stage I; outer lamella about
one-fourth longer than inner lamella; both lamel-
lae and appendix interna fully segmented at their
bases. Third to fifth pleopods essentally identical
to second pleopods.
ABDOMEN AND TELSON.— Posterolateral
spines on fourth and fifth abdominal somites still
present, those on fourth somite being only slightly
shorter in relation to length of somite than in
Stage I. Telson essentially identical to Stage I
except segmented from sixth abdominal segment
and bears 20 or 21 densely plumose setae. Uropods
still enclosed.
STAGE III (MEGALOPA)
Total length of Stage III 7.5 mm (range 7.4-7.6
mm; two specimens). Antennal spine of carapace
larger and pterygostomian spine smaller than in
Stage II; no evidence of minute spinules along
anteroventral margin. Rostrum (Figure 3 A) short;
bears single tooth at base in addition to dorsal
protuberance. Antennules similar in shape to
adult; outer flagellum six-segmented; inner flagel-
lum five-segmented; peduncle three-segmented,
Figure 3. — Stage III (megalopa) of Lebbeus groenlan-
dicus: A, carapace; B, telson.
0.5 mm
463
FISHERY BULLETIN: VOL 76, NO. 2
lateral spine of proximal segment well developed.
Antennal flagellum with at least 30 segments;
about four times length of scale. Mandibles with
unsegmented palps bearing four or five short
teeth. Endopodite of maxillule reduced. Maxil-
lipeds shaped as in adult, exopodites reduced. Dac-
tylopodites of first and second pereopods well
developed; carpopodite of second pereopod six-
(sometimes seven-) segmented. Lateral margins of
pleopods fringed with setae; appendix internae
with minute cincinnuli. Posterolateral spines on
abdominal segments four and five remnant or
lacking. Telson (Figure 3B) rectangular in shape;
bears two pairs of spines terminally and one pair
laterally (one or two additional spines may occur
centrally on terminal margin). Uropods exposed;
fully developed except transverse hinge not com-
plete.
COMPARISON OF LARVAL STAGES
WITH DESCRIPTIONS BY
OTHER AUTHORS
Under the name "Spirontocaris-larva No. lA,"
Stephensen (1935) included four specimens that
were morphologically identical to zoeae provi-
sionally identified by him as Stage I S. polaris
(= Lebbeus polaris (Sabine)) except that they dif-
fered by lacking spines on abdominal segments
four and five, exopodites on any pereopods, or free
uropods. He regarded these four zoeae as belong-
ing to either Spirontocaris groenlandica ( = L.
groenlandicus), S. gaimardii (= Euahis gaimar-
dii (H. Milne Edwards)), or S. spinas (Sowerby).
Pike and Williamson (1961) have shown that the
absence of spines on abdominal segments four and
five eliminates the zoeae from being either E.
gaimardii or S. spinus. They agree with Stephen-
sen that his specimens oC'Spirontocaris-larwa No.
lA" are closely allied to zoeae he tentatively de-
scribed earlier (Stephensen 1917, 1935) as S.
polaris (= L. polaris). They suggest, therefore,
that Stephensen's "Spirontocaris -larva No. lA"
probably belongs to the genus Lebbeus and spe-
cifically to L. groenlandicus.
Comparison of my zoeae of L. groenlandicus
with the descriptions given by Stephensen for
"Spirontocaris-larva No. lA" shows that
"Spirontocaris -larva No. lA" are not zoeae of L.
groenlandicus. My Stage I zoeae bear remnant
exopodites on the first and second pereopods and
lateral spines on abdominal segments four and
five, but Stephensen's Stage I zoeae bear neither
the exopodites nor the spines. My Stage I zoeae do
not bear supraorbital spines, the peduncle of the
antennule is without joints or a ventral spine, and
there is no indication of the carpopodite of the
second pereopod being jointed; Stephensen's Stage
I zoeae bear supraorbital spines, the peduncle of
the antennule is three-jointed and bears a distinct
ventral spine, and the carpopodite of the second
pereopod is partially jointed. In addition, the
chelae of the first and second pereopods are not as
well formed in my Stage I zoeae as they are in
Stephensen's Stage I zoeae.
Several of the morphological characteristics de-
scribed by Stephensen as pertaining to "Spiron-
tocaris-larva No. lA" are typical of later stage
zoeae, a fact already noted by Pike and Williamson
( 1961) in their discussion of the morphology of the
zoeae of L. polaris and which prompted them to
suggest that Stephensen's zoeae were actually in
the second, or penultimate, zoeal stage. Even if
Stephensen was mistaken in identifying his zoeae
as Stage I rather than Stage 11, the morphological
differences between my zoeae and his are too great
to consider them identical species. My Stage II
zoeae bear spines on abdominal somites four and
five and the telson is segmented from the sixth
abdominal somite, whereas Stephensen's zoeae do
not bear spines on abdominal somites four and five
and the telson is not segmented from the sixth
abdominal somite. Also, in my Stage II zoeae the
peduncle of the antennule does not bear a ventral
spine and is unsegmented but in Stephensen's
zoeae the peduncle bears a ventral spine and is
segmented.
I have no further evidence on the identity of
Stephensen's "SpjVontocans-larva No. lA." Of the
three members of the genus recorded from Green-
land waters, L. polaris, L. groenlandicus, and L.
microceros (cf. Holthuis 1947; Squires 1966), L.
microceros was not recorded by Stephensen. Ap-
parently it is rare and its larvae have not been
described. Also, the advanced development of
Stephensen's "Spirontocaris-larva No. lA" makes
it unlikely that it belongs to another genus of the
spirontocarid group (cf. Pike and Williamson
1961). Apparently Stephensen's "Spirontocaris-
larva No. lA" is either the zoea of L. microceros or
that of another species of Lebbeus not yet recorded
from Greenland waters.
On the basis of descriptions of "Spirontocaris-
larva No. lA" by Stephensen (1935) and a late
stage embryo of Hippolyte polaris ( = L. polaris ) by
Kr^yer (1842), Pike and Williamson (1961)
464
H'WNES LEBBEUS GROEM.ANDICUS LAKVAE
characterized larvae of the genus Lehbeus as hav-
ing two (or three) zoeal stages, five-segmented
pereopods, and a small rostrum in Stage I, and
pereopods without exopodites in the last zoeal
stage. My description of larvae oiL.groenlandicus
confirms the generic characteristics for Lebbeus
larvae as given by Pike and Williamson. As noted
by Pike and Williamson, however, larvae are de-
scribed for only a few species of hippolytids, in-
cluding the genus Lebbeus, and further confirma-
tion of the generic characteristics of the larvae is
desirable.
LITERATURE CITED
Haynes, E.
1976. Description of zoeae of coonstripe shrimp, Pandalus
hypsinotus, reared in the laboratory. Fish. Bull., U.S.
74:323-342.
1978. Description of larvae of the humpy shrimp, Pan-
dalus goniurus, reared in situ in Kachemak Bay, Alas-
ka. Fish. Bull., U.S. 76:235-248.
HOLTHIUS, 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. 140, Monogr. 39a», 100 p.
KR0YER.
1842. Monografisk Fremstilling af Slaegten Hippolyte's
nordiske Arter. Med Bidrag til Decapodernes Udvik-
lingshistorie. K. Dan. Vidensk. Selsk. Naturv. Math.
Afh. Kbh. 9:209-360. (This work has not been seen by the
author.)
Pike, R. b., and D. I. Williamson.
1961. The larvae of Spirontocaris and related genera ( De-
capoda, Hippolytidae). Crustaceana 2:187-208.
Squires, H. J.
1966. Distribution of decapod Crustacea in the northwest
Atlantic. Ser. Atlas Mar. Environ., Am. Geogr. Soc.
Folio 12.
STEPHENSEN, K.
1917. Zoogeographical investigation of certain fjords in
Southern Greenland, with special reference to the Crus-
tacea, Pycnogonida and Echinodermata, including a list of
Alcyonaria and Pisces. Medd. Gr0nl. 53:229-378.
1935. Crustacea Decapoda. The Godthaab Expedition,
1928. Medd. Gr0nl. 80:1-94.
465
PREDICTING ABUNDANCE OF STRIPED BASS, MORONE SAXATILIS,
IN NEW YORK WATERS FROM MODAL LENGTHS *
Herbert M. Austin^ and Clarence R. Hickey, Jr.^
ABSTRACT
The abundance of cohorts for any given year class of striped bass, Morone saxatilis, prior to their
leaving Chesapeake Bay is inversely related to the modal length offish in that year class 2 yr later in
New York waters. The modal length of bass in their third year migrating into the New York area is a
reliable index of the abundance of that year class. When back extrapolated modal lengths at the end of
the second year of life are considered for the dominant year classes in the New York fishery (ages
ni-VI), a high degree of inverse correlation is found between age II and modal length and reported
landings suggesting that this is an effective method of predicting the abundance of the stock for the
fishery.
In discussing natural fluctuations in fish popula-
tions, Royce (1972) posed the question, "... can
we forecast their occurrence to take maximum
advantage of periods of high abundance and pro-
tect populations during periods of scarcity?" This
question is pertinent to the striped bass, Morone
saxatilis, stocks of the Atlantic coast of the United
States. The Atlantic coast commercial catch of this
species, while following a pattern of fluctuations,
has been in an upward trend in recent years, ap-
parently as a result of an increasing abundance of
fish (Koo 1970; McHugh 1972). This increasing
abundance has been reflected by an increased
commercial harvest in the State of New York.
Concurrently, although not as well documented, is
an increase in the number of recreational fisher-
men utilizing the resource. Both phenomena
necessitate the gathering of management infor-
mation while the resource is still in good condi-
tion.
Most ( >807c ) of the New York commercial har-
vest of striped bass occurs in the waters of eastern
Suffolk County (Figure 1) where the major
fisheries are primarily with haul seine and pound
net. Fish taken in this region are predominantly of
Chesapeake Bay origin (Neville et al. 1939; Alper-
in 1966; Schaefer 1968, 1972; Koo 1970; Austin
CONNECTICUT
ATLANTIC OCEAN
Figure l. — Location of Long Island, N.Y., and the southeastern
tip near Montauk Point where striped bass were collected during
1972 and 1974.
and Custer 1977; Austin and Hickey;'* Texas In-
struments, Inc.^).
This study was designed as one phase of a pro-
gram to tag and monitor "short" or prerecruit
striped bass (less than the legal, 406-mm New
York State limit). As stated by Talbot ( 1966), little
is known of these fish outside of their nursery
areas. Monitoring of these fish, then, permits
study of the next year's catch, a segment of the
striped bass population often overlooked in fishery
investigations.
Prerecruit striped bass in New York waters of
eastern Long Island are predominantly 2- and
■New York Ocean Science Laboratory, Contribution No. 82.
Funded by a grant to the New York Ocean Science Laboratory
from the State of New York, Project No. BR74-17F.
^Division of Fisheries Sciences and Services, Virginia Insti-
tute of Marine Science, Gloucester Point, VA 23062.
^Environmental Specialists Branch, U.S. Nuclear Regulatory
Commission, Washington, DC 20555.
Manuscript accepted September 1977.
FISHERY BULLETIN: VOL. 76, NO. 2, 1978.
■» Austin, H. M., and C. R. Hickey, Jr. 1974, Migration and
mortality of striped bass tagged in eastern Long Island, p. 11-
16. Proc, Am. Littoral Soc./N.Y. Ocean Sci. Lab. Fish Tag
Seminar, Dec. 1974, Montauk, N.Y.
^Texas Instruments, Inc. 1976. Report on relative con-
tribution of Hudson River striped bass to the Atlantic coastal
fishery. Unpubl. rep., 110 p. Texas Instruments, Inc., Dallas,
Tex.
467
FISHERY BULLETIN: VOL. 76, NO. 2
3-yr-old fish which are making their first annual
migration from their Chesapeake Bay nursery
areas to the northern summer feeding ground
(Austin and Hickey see footnote 4). The concen-
trated study of one age-group of fish permits
monitoring of the cohorts for successive years
starting with first departure from their home
grounds and, thereby, permits a description of dif-
ferences or variations in migration and abundance
on an annual basis, as well as an accurate evalua-
tion of year class mortality in successive years.
METHODS AND MATERIALS
Prerecruit striped bass were randomly removed
from the catches of commercial haul seine and
pound net fishermen in the waters of East
Hampton on the southeastern end of Long Island,
N.Y. (Figure 1). Samples were collected during
May and June 1972 and April-June 1974, thus the
age II fish were of the 1970 and 1972 year classes,
respectively. Fork lengths were measured in the
field to the nearest millimeter and scale samples
were removed for age determination. The fish
were then tagged (Floy*^ FD-69B anchor tags) and
released. The initial purpose of the study was
tagging of prerecruit fish to monitor the seasonal
migration and mortality of cohorts as they reached
legal size in the different states. The feasibility
study was focused on the 1970 and 1972 year
classes. Large differences in the modal size of the
fish in their third year (11 + ) existed between the
two year classes ( Figure 2 ) . The smaller sized 1970
year class of fish were from the most abundant
Chesapeake Bay year class on record (Schaefer
1972). Examination of the literature shows that
the length of cohorts may be inversely propor-
tional to the abundance or density of the fish (Ste-
vens 1977; Texas Instruments'' ), suggesting to us
that the length of the striped bass, when they first
appear in New York waters, could be an indicator
of year class strength and subsequently a means of
predicting stock abundance in local waters. Con-
sequently the focus of the study was redirected
towards examination of these differences.
Schafer (1968, 1972) stated that most commer-
cially harvested striped bass in New York are of
four age-classes, III- VI. Based on this, Schaefer
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
■'Texas In.struments, Inc. 1975. First annual report for the
multiplant impact study of the Hudson River estuary. Unpubl.
rep., vol. 1, p. VIII-8-VIII-12. Texas Instruments, Inc., Dallas,
Tex.
r-l
YC 1970
(in 1972)
I
YC 1972
(in 1974)
I I
i_J
350
FORK LENGTH IN MILLIMETERS
Figure 2. — Length-frequency distribution of age II striped bass
captured by commercial fishing gear near eastern Long Island,
N.Y., during 1972 and 1974.
(1972) related the New York harvest to a 4-yr
mean brood production (year class strength; ex-
pressed as annual mean number of juveniles per
standard seine haul in Chesapeake Bay, Md.) 3 to
6 yr prior to the harvest. He concluded that ap-
proximately 70% of the variability in annual New
York landings could be explained by annual fluc-
tuations in year class strength in Maryland wa-
ters of Chesapeake Bay.
We hypothesized that the growth rate of striped
bass, and, therefore, the body length at the end of
the 2-yr residence time in the Chesapeake Bay
nursery grounds, is a density dependent function
with the length inversely proportional to the year
class abundance (number offish). This hypothesis
was tested via a correlation analysis using modal
lengths at age 11+ (our data combined with pub-
lished data of Alperin 1966 and Schaefer 1968)
and year class abundance indices supplied by the
Maryland Department of Natural Resources. The
analyses were performed using a Hewlett-
Packard Model 9100B programmable calculator
with an X-Y plotter, which provided both a regres-
sion line and a correlation analysis and coefficient.
The relationship resulting from the above
analysis suggested that the density dependent
hypothesis is true. Since Schaefer ( 1972) described
a relationship between the annual New York har-
vest (reported commercial landings) of striped
bass and the Chesapeake Bay year class abun-
dance, and since we have described a probable
relationship between year class abundance and
modal length at age 11+ in New York waters, it
seemed reasonable to test the correlation between
468
AUSTIN and HICKEY: PREDICTING ABUNDANCE OF STRIPED BASS
the New York harvest and the modal length at age
11+ via Model II correlation analysis.
These analyses were performed in an effort to
describe a method for predicting the commercial
harvest (and therefore the apparent abundance) of
striped bass in New York waters. As each of the
several steps in the analyses were dependent on
the results of those previously calculated, they are
discussed in more detail along with the results
below.
The reliability of the suggested technique for
predicting the abundance of striped bass in New
York waters is dependent on several assumptions:
1 ) The Chesapeake Bay stock of fish is the major
contributor of striped bass to the New York
commercial fishery, as suggested by the several
authors noted above;
2 ) The annual relative contribution of the several
Atlantic coastal breeding stocks to the coastal
stock of fish and, therefore, to the New York
fishery remains constant or that it fluctuates or
cycles in a consistent manner;
3) The commercial fishery for striped bass in New
York effectively collects representative "sam-
ples" of the Chesapeake Bay stock of fish; this
assumption appears to be valid based upon the
relationships described by Schaefer (1972),
Texas Instruments, Inc. (see footnote 5) and
those described herein, and based upon our ob-
servations and those of Schaefer (1972) that
many size classes of fish are present in the
commercial catch — small age II prerecruits to
large mature fish >16 kg total weight;
4) The forecast of commercial striped bass land-
ings is based upon past historical landings in
relation to past life history events of the species
(year class abundance and length at age 11+)
and does not reflect changes in commercial
fishing effort or any changes in the contribu-
tions to the reported landings by recreational
fishermen; we have assumed a constant fishing
effort, as did Schaefer (1972), and thus com-
pared our results with his; while we recognize
the weakness in this assumption, there is no
alternative as there is no estimate of effort.
RESULTS AND DISCUSSION
Year Class Strength and Modal Size
The lengths of striped bass at age 11+ near Long
Island are probably related to ecological cir-
cumstances encountered by the fish during their
first 2 yr of residence in the rivers of Chesapeake
Bay (density, competition, amount of available
food). Similarly, Cushing ( 1968) found a close rela-
tionship between the mean length of age III Atlan-
tic herring, Clupea harengus, and the density of
their food source in the sea, and Clark (1967) de-
scribed reduced growth rates for sunfish, Lepomis,
due to overcrowding, excessive competition, and
reduced food supply. It has also been demonstrated
by Anthony ( 1 97 1 ) that the growth of young ( age I
and II) Atlantic herring is inversely related to
their abundance, and Wagner (1969) has stated
that in most fishes the growth rate per individual
is inversely related to their density.
If an inverse relationship exists between the
abundance of a year class of striped bass and the
cohort length at age II + , a similar relationship
should exist between the commercial harvest (as
an index of abundance) and the length at age II + ,
assuming that fishing effort remains approxi-
mately constant. To test this hypothesis, age 11 +
modal length data for year classes 1970 and 1972
(Figure 2) were combined with other published
modal length data for year classes at age 11+ in
New York waters from Alperin (1966) and
Schaefer ( 1968 ) (Table 1 ), providing a total of eight
annual data points. A correlation analysis was
performed between these eight annual modal
lengths of age 11+ fish and their respective
Chesapeake Bay year class strengths 2 yr earlier
(Figure 3). The year class strength data (supplied
by the Maryland Department of Natural Re-
sources) are expressed as the annual mean
Table l. — Comparison of observed and computed modal fork
lengths for age II striped bass in New York waters.
Observed modal
Computed modal
Year
Year class
length at age II
length at age 11^
class
strength'
(mm)
(mm)
1954
5.2
^31 3
318
1958
18 1
"285
278
1959
1,3
3335
330
1960
6.4
3310
314
1961
14.4
3290
289
1962
12.2
3300
296
1970
26.8
=245
251
1972
8.5
5295
307
Mean
297.9 ±53.0
297.9 ±50.4
Standard deviation
26.5
25.2
tr = 6.99
n =8
P<0.001
'Courtesy Joseph Boone. Maryland Department of Natural Resources, An-
napolis, Md,, data expressed as annual mean number of ageO+ juveniles per
standard seme haul.
^Based on the relationship y = 333 - 3X, where Vis the modal length of age
11+ fish and X is the strength of the year class (Figure 3).
^Extrapolated from Schaefer (1968),
■•Extrapolated from Alperin (1966).
^Data from the present investigation.
469
FISHERY BULLETIN: VOL. 76, NO. 2
400-^
^1959
i"
-I — I — I — I — r-
^1962
*I972
~~~~-~_J96l
.__^58
1970 a
10
15
20
25 3
STRENGTH OF YEAR CLASS (NO /SEINE HAUL 1
Figure 3.— Modal size (mm fork length) of age II striped bass
from Long Island waters as a function of Chesapeake Bay year
class strength. Year classes are indicated.
number of age 0+ juveniles per standard seine
haul near the Maryland shores of Chesapeake
Bay. These are the same data used by Schaefer
(1972). The relationship (Y = 333 - 3X) yielded a
correlation coefficient of -0.95 ir^ = 0.90),
suggesting that 909c of the annual variation in
modal length at age 11+ for striped bass in New
York waters can be explained by annual fluctua-
tions in year class abundance in the waters of
Chesapeake Bay.
Modal Size and
the New York Commercial Harvest
The equation described above (Y =
333 - 3X) was used to calculate (and thus to esti-
mate) modal lengths of age II -I- fish for those 8 yr
for which actual modal lengths exist. A ^test com-
parison between the observed age II modal sizes
and those computed using the correlation formula
above showed no significant difference at the 0.001
probability level (Table 1). Since no significant
difference existed between the observed and calcu-
lated modal lengths, the assumption was made
that reliable modal lengths could be calculated for
years in which no actual measurements exist. The
equation described above was, therefore, used to
estimate modal lengths of age 11+ striped bass for
all years between 1954 and 1972, using the corre-
sponding year class abundance data. A correlation
analysis was then performed (similar to that done
by Schaefer 1972) between the New York land-
ings of striped bass ( Y) and a 4-yr mean of the
computed modal lengths of age 11+ fish 1 to 4
yr prior to harvest (X). The relationship
(Y = 15,205,309 - 46,859X) (Figure 4A) yielded
a correlation coefficient of -0.86 (r^ = 0.74) sig-
nificant at the 0.001 probability level it, = 6.06,
n = 13). This expression permits the hindcasting
of New York landings as well as a forecast 1 yr in
advance, with 95'7r confidence limits. The hind-
casts and 1-yr forecasts (for 1975) are superim-
posed on the actual New York landings in Figure
5A.
As stated by Schaefer (1968, 1972), the New
York harvest is predominantly fish of ages III-VI.
Close examination of his catch data for 1962, how-
ever, revealed that age VII fish, although <27c of
the catch in number, could constitute a significant
proportion of the catch by weight. Schaefer's
(1968) age-frequency distribution shows that in
1962 the age III fish outnumbered the age VII by
about 10:1. Using the mean age- weight relation-
ships of Mansueti (1961) as 1.8 lb at age III and
12.5 lb at age VII, the age III fish in Schaefer's
( 1968) 1962 catch thus outweighed the age VII fish
by less than 1.5:1. Similarly, the age VI fish ( mean
295 300 305
MEAN MODfiL Size I - 5 YEARS PRIOR TO HARVEST
Figure 4.— Relationship of New York commercial landings of
striped bass to the mean modal size at age II; A) 1 to 4 yr prior to
harvest; B) 1 to 5 yr prior to harvest.
470
AUSTIN and HICKEY PREDICTING ABUNDANCE OF STRIPED BASS
2000-
flCTUflL LANDINGS
CALCULATED LANDINGS
^
1500-
^
f\"
\ \
\ \
^ \ '
r
h \
1
V
1000-
\
7
/X /
1
500-
/
_-'
400-
^'\/\y
300-
A
ACTUAL
LANDINGS
CALCULATED LANDINGS
1500-
f< * v'
/ A
\( \
I
f '^- \
//
u
1000-
!••
k
/
/i
\
I \
1
1 \
'/\ ^
t
%
/y
u\
500-
/U/
400-
^^
/\
/
300-
B
Figure 5. — Actual New York commercial landings of striped
bass from 1954 through 1974 with calculated landings through
1975 sujjerimposed using: A) 4-yr mean modal sizes of age II fish;
B) 5-yr mean modal sizes of age II fish.
weight 8.1 lb) outnumbered the age VII fish by 2:1,
but outweighed them by only 1.3:1.
It was apparent that during some years the New
York harvest of striped bass may be dominated by
five age-groups rather than four, as suggested by
Schaefer (1972). Another correlation analysis
was, therefore, performed between the New York
landings (Y) and a 5-yr mean of the computed
modal sizes of age II + fish 1 to 5 yr prior to harvest
iX). This 5-yr function was expressed as a linear
relationship (7 = 17,315,491 - 53,810X) (Figure
4B) with a correlation coefficient of -0.83
(r^ = 0.69), significant at the 0.001 probability
level it, =5.05, n =12). Although this
coefficient was reduced slightly from that of the
4-yr function above (r = -0.86), the fit of
estimated-to-actual landings (with 95^f confi-
dence limits) was better for many years (Figure
5B) and was closer to the actual landings than
the calculated predictions of Schaefer (1972)
(Table 2).
Size, Age, and Migration
As stated, age II + modal sizes may be computed.
Another method of size determination at age II + is
by back calculation of scale radii from larger, older
fish. Although no age II modal sizes determined by
this method were used in the predictive models,
our attempts to do so produced some interesting
information. Mansueti (1961) described the body
length-scale length relationship of striped bass as
an allometric linear function, permitting the back
calculation of size at each year of age using the
scale radii method. Scales from 142 age III striped
bass captured in eastern Long Island waters dur-
ing 1973 (year class 1970) were made available to
the authors by the New York State Department of
Environmental Conservation. The ages were re-
checked and the fork lengths at age II determined
by back calculation from body length:scale radii
ratios. The length-frequency distribution of back-
calculated data was bimodal, with equal peaks at
205 mm and 235 mm. The second peak was 4.1%
lower than the observed unimodal size of 245 mm.
Although the back calculated values were slightly
lower than the observed, the fit suggests that back
calculations may be used for obtaining age II sizes
of striped bass during years when these data are
lacking.
Unpublished length-frequency data for age II
striped bass of the year classes 1968, 1969, and
1971 taken in the Virginia rivers of the
Chesapeake Bay System were made available to
the authors by John V. Merriner of the Virginia
Institute of Marine Science. These data showed
bimodal distributions similar to that of the back-
calculated age II lengths above (Merriner pers.
commun.). Merriner suggested that multimodal
frequencies occurred because the fish were from
different river systems. Merriner's data were from
Virginia rivers while our data (Austin and Hickey
Table 2. — Comparison of actual and calculated commercial
landings of striped bass in the State of New York 1972-75.
Calculated
Calculated
Calculated
Year
Actual
landings'
landings
4-yr function^
landings
5-yr function^
landings by
Schaefer (1972)
1972
818.150
926,903
852,860
908.000
1973
1,673,984
1,447,975
1,496,965
1.455,000
1974
1,378,529
1,592,301
1,477.594
1 .607,000
1975
1,137,074
1,639.160
1,500.732
—
'Courtesy Fred Blossom, National Marine Fisheries Service, NOAA, Patch-
ogue. NY
^Forecasts using the linear regression formulae discussed in the text.
471
FISHERY BULLETIN; VOL 76, NO. 2
see footnote 4) suggest that the bass we examined
for back calculation of length were from both
Maryland and Virginia rivers, which could ex-
plain the differences in the results of the back calcu-
lations and the observed lengths. These data
suggest that the size-frequency distribution of age
II striped bass on Long Island could be bimodal
rather than unimodal. The fact that they were not
may be due to striped bass migrating by size
rather than by age. Two observations of prerecruit
striped bass near Long Island lend support to this
theory: 1) lOO'/r of the small 454 sublegal fish
tagged in 1972 were age-group II, and 2 1 only 28'7(
of the 696 sublegal fish tagged in 1974 were age-
group II (year class 1972), the remaining 72*7^
were age-groups III (659<^ ) and IV {T7( ). Those fish
probably were the larger 1972 and the smaller
1971 and 1970 fish. This large overlap in length
ranges permitted an intermingling of the age-
classes during the migration of 1974.
Management Implications
The size increments between different year
classes at the same age, and the size differences of
individuals within the same year class have sev-
eral implications:
1) Faster growing large individuals of any given
year class or a less abundant year class of
larger individuals are subject to earlier exploi-
tation in Chesapeake Bay and along the entire
Atlantic seaboard;
2) Slower growing individuals or small individu-
als of a large year class may be recruited sev-
eral months later than normal in Chesapeake
Bay, but perhaps not until a full year later
among the northern Atlantic States; a late re-
cruitment in the Chesapeake area might result
in more available fish to the fisheries in the
other coastal states when the fish migrate out
of the bay;
3) Projecting sizes offish on the basis of age or vice
versa may be invalid, e.g., age II fish in 1972
compared with age II fish of Merriman ( 1941 ) or
Mansueti (1961).
The use of a mean 4- or 5-yr modal function for
prediction of landings treats all year classes
equally. A weighted mean providing greater rep-
resentation to more abundant year classes might
result in more accurate predictions of landings.
Such a method could be used by any State simply
472
by monitoring the spring catches of age 11+ pre-
recruit fish taken by commercial fishermen. This
would require, in New York for example, annual
monitoring of the spring run with measurements
of sublegal fish. The use of observed modes rather
than computed modes for prediction of landings
will probably result in more accurate estimates, as
suggested in Table 3.
Table 3. — Comparison of actual New York commercial landings
of striped bass with those calculated using computed and observed
age II modal values, for years in which sufficient empirical data
exist.'
Item
1964
1965
Actual N Y landings
925.500
702,935
4-yr function
Landings calculated using:
Computed modes
1,021,090
807,881
Empirical modes
913,314
618,102
5-yr function
Landings calculated using.
_
Computed modes
799.588
1,098,771
Empirical modes
—
849,631
'Computed and observed age II modal values are those on Table 1.
The eastern New York commercial harvest of
striped bass is primarily dependent upon the year
class abundance of the Chesapeake Bay stock. The
harvest is influenced not only by the larger and
older individuals, but also by the annual recruit-
ment of age III fish, especially when dominant
year classes are present.
Knowledge of Chesapeake Bay year class
strength or age 11+ modal sizes in New York wa-
ters offers a means of forecasting the New York
commercial harvest, and thus the apparent abun-
dance of striped bass in New York waters.
If, as suggested, the level of the New York har-
vest is primarily related to the Chesapeake Bay
stock of fish, than the former can be used as a
qualitative measure of the latter.
Such predictive tools as those discussed should
be flexible to allow for the occurrence of more than
four age-groups of fish in the catch. This may be
especially important when dominant year classes
are present for several years. Necessary, then, is
the annual monitoring of the prerecruit fish in the
commercial catch by age or year class, length, and
weight. Age- weight data are especially important
as commercial landings are recorded by weight of
catch and not by numbers offish. Differences noted
between calculated and observed landings may be
due to environmental variability, changes in
fishing effort, the dominance of a particular year
class in the fishery, and the fluctuation in the
relative contributions offish from the several At-
AUSTIN and HICKEY: PREDICTING ABUNDANCE OF STRIPED BASS
lantic coastal breeding grounds. Future research
and managment efforts should take these into
consideration.
ACKNOWLEDGMENTS
We acknowledge the help of the following:
Joseph Boone, Maryland Department of Natural
Resources, Annapolis, Md., for unpublished Mary-
land year class data; John V. Merriner, Virginia
Institute of Marine Science, Gloucester Point, Va.,
for his unpublished Virginia data; Bryon H.
Young, New York State Department of Environ-
mental Conservation, Stony Brook, N.Y., for the
use of unpublished data; and Fred Blossom, Na-
tional Marine Fisheries Service, Patchogue, N.Y.,
for data on New York landings. J. L. McHugh,
State University of New York, and Jack P. Wise
and Richard H. Schaefer, National Marine
Fisheries Service, reviewed the manuscript. Their
comments are appreciated.
LITERATURE CITED
ALPERIN, I. M.
1966. Dispersal, migration and origins of striped bass from
Great South Bay, Long Island. N.Y. Fish Game J.
13:79-112.
AUSTIN, H. M., AND O. CUSTER.
1977. Seasonal migration of striped bass in Long Island
Sound. N.Y. Fish Game J. 24:53-68.
ANTHONY, V. C.
1971. The density dependence of growth of the Atlantic
herring in Maine. Rapp. P.-V. Reun Cons. Int. Explor.
Mer 160:197-205.
Clark, G. L.
1967. Elements of ecology. John Wiley and Sons, N.Y. ,
560 p.
CU,SHINC., D. H.
1968. Fisheries biology. Univ. Wis. Press, Madison, 200 p.
KOO, T. S. Y.
1970. The striped bass fishery in the Atlantic States.
Chesapeake Sci. 11:73-93.
MANSUETI, R. J.
1961. Age, growth and movements of the striped bass,
Roccus saxatilis, taken in size selective fishing gear in
Maryland. Chesapeake Sci. 2:9-36.
MCHUGH, J, L.
1972. Marine fisheries of New York State. Fish. Bull.,
U.S. 70:585-610.
Merriman, D.
1941. Studies on the striped bass {Roccus saxatilis) of the
Atlantic Coast. U.S. Fish Wildl. Serv., Fish. Bull. 50: 1-
77.
Neville, W. C, C. L. Dickinson, and J. R. Westman.
1939. Striped bass (Roccus saxatilis). In A biological
survey of the salt waters of Long Island, 1938. Part I, p.
107-113. N.Y. State Conserv. Dep., Suppl. 28th Annu.
Rep., 1938, No. 14.
ROYCE, W. F.
1972. Introduction to the fishery sciences. Academic
Press, N.Y., 351 p.
Schaefer, R. H.
1968. Size, age composition and migration of striped bass
from the surf waters of Long Island. N.Y. Fish Game J.
15:1-51.
1972. A short-range forecast function for predicting the
relative abundance of striped bass in Long Island waters.
N.Y. Fish Game J. 19:178-181.
Stevens, D. E.
1977. Striped bass (Morone saxatilis) year class strength
in relation to river flow in the Sacramento-San Joaquin
estuary, California. Trans. Am. Fish. Soc. 106:34-42.
Talbot, G. B.
1966. Estuarine environmental requirements and limit-
ing factors for striped bass. In A symposium on es-
tuarine fisheries, p. 37-49. Am. Fish. Soc. Spec. Publ. 3.
Wagner, F. H.
1969. Ecosystem concepts in fish and game manage-
ment. In G. M. Van Dyne (editor). The ecosystem con-
cept in natural resource management, p. 259-307.
Academic Press, N.Y.
473
NOTES
OBSERVATIONS ON A WHITE-SIDED
DOLPHIN, LAGEMORHYNCHUS ACUTUS,
PROBABLY KILLED IN GILL NETS IN
THE GULF OF MAINE
On 20 July 1976, a white-sided dolphin, Lageno-
rhynchus acutus, was observed floating with its
beak out of the water on Jeffreys Ledge, Maine
(lat. 43°09'N, long. 70°04'W). The 201-cm long
female weighed 113.2 kg and was freshly dead,
still bleeding freely from symmetrical injuries to
the left and right sides of both the upper and lower
jaws and the flippers. The lungs contained foamy
materials and were mottled white, indicating
drowning as the immediate cause of death. Many
gill nets were present in the area, and the sym-
metrical nature of the injuries indicated that the
animal had become entangled in the mesh,
drowned, and perhaps been freed or discarded dur-
ing hauling of the net. A humpback whale,
Megaptera novaengliae, was entangled in a gill net
for 2 h before freeing itself on the same day in the
same general area.
Gross autopsy revealed several cysts in the ab-
dominal muscles of the lower left side and a 5 cm x
7.5 cm yellow, pussy abscess 15 cm anterior and
dorsal to the right mammary gland, perhaps
caused by a bladderworm stage (plerocercoid) of
Monorygma grimaldi (Geraci et al.M. No other
parasites were found, although all major organs
except the brain were inspected. Tissue and organ
weights are shown in Table 1. The length and
'Geraci, G., S. A. Testaverde, D. J. St. Aubin, and T. H. Loop.
1976. A mass stranding of the Atlantic white-sided dolphin,
Lagenorhynchus acutus: a study into pathobiology and life his-
tory. Unpubl. manuscr., 166 p. submitted to Marine Mammal
Commission by New England Aquarium, Boston.
Table l. — Tissue and organ weights of a Lagenorhynchus
acutus from Jeffreys Ledge, Maine. Weights not corrected for
blood loss.
Weight
Weight
Tissue or organ
(kg)
Tissue or organ
(kg)
Muscle
69.0
Lett kidney
0.473
Blubber and tins
245
Right adrenal
0.011
Gastrointestinal tract
6.8
Left adrenal
0.010
Liver
3.2
Spleen
0.084
Heart
0.959
Right ovary
0.0051
Right lung
1.117
Left ovary
0.0047
Left lung
1.149
Bones'
5.4
Right kidney
0.476
Total
113.2
weight of this animal fit well on a regression line
developed for this species (Geraci et al. see foot-
note 1). From a length-age relationship (Geraci et
al. see footnote 1), it is likely that this female was
between 2y2 and 3 yr old, and was immature.
The stomach contained 980 g of food, including
four 25- to 30-cm herring, Clupea harengus, three
partly digested (total weight 340 g) and one skele-
ton (15 g); and one partly digested short-finned
squid, Illex illecebrocus (anterior mantle length
17.5 cm, weight 90 g), with remains of 10 other
squid of this species (represented by 5 complete
pairs of beaks plus 5 single anterior beaks). Mean
length of anterior beaks (± SD) was 1.42 ± 0.03
cm, corresponding to mantle lengths from 17 to 19
cm (Testaverde^). Also 10 left and 11 right otoliths
from silver hake, Merluccius hilinearis, (mean size
± SD = 1.15 ± 0.06 cm) indicated consumption of
at least 11 fish of 22-26 cm fork length (Nichy
1969).
From these data and the literature it appears
that C harengus and/, illecebrocus are staples in
the summer diet of white-sided dolphins. A 161-kg
female collected on 14 September 1954, off Cape
Cod contained 12 fresh herring, digested fish (ap-
parently herring), and squid (Schevill 1956). A
180-cm long male driven ashore with pothead
whales, Globicephala melaena, in Newfoundland
on 30 July 1954, contained herring and short-
finned squid (Sergeant and Fisher 1957). Short-
finned squid was the most common food in the
stomachs of white-sided dolphins which mass-
stranded at Lingley Cove, Maine, on 6 September
1974; however, no herring were found despite the
fact that "brit" herring were present in the cove
(Geraci et al. see footnote 1). Smelt, Osmerus mor-
dax, remains, were found in five individuals; silver
hake had been eaten by one individual; and un-
identified crustacean remains were found in
another stomach.
Schools of white-sided dolphins were unusually
common in the Gulf of Maine in 1976, perhaps
because squid were abundant, possibly as a result
of this year's unusually high sea temperatures
'Bones were bleached and dried in the laboratory.
^Testaverde, S.A. 1975. An informal discussion concerning the
cestode Phyllobothrium sp. in squid, Illex illecebrocus illece-
brocus and its possible relationship to marine mammals. Un-
publ. manuscr., 20 p.
475
(Anonymous 1976; Prescott and Moore 1976).
Silver hake, normally of variable abundance here
(Bigelow and Schroeder 1953) was also abundant
during 1976. On several different occasions,
groups of 6-30 white-sided dolphins were seen by
one of us (SKK) swimming close to pods of either
finback whale, Balaenoptera physalus, or
humpback whale, Megaptera novaeangliae , and
apparently feeding with them.
Acknowledgment
We are grateful to the crew of the MV Exxon Bay
State for procuring this specimen; to John E. Fitch,
California Department of Fish and Game, for
otolith identification; and to Fred Nichy, North-
east Fisheries Center, National Marine Fisheries
Service, NOAA, Woods Hole, Mass., for providing
data on otolith growth.
Literature Cited
ANONYMOUS.
1976. Rise in squid Down East traced to warmer waters.
Ellsworth (Maine) American, August 26, 1976, p. 1.
BIGELOW, H. B., AND W. C. SCHROEDER.
1953. Fishes of the Gulf of Maine. U.S. Fish Wildi. Serv.,
Fish Bull. 53, 577 p.
NiCHY, F.
1969. Growth patterns on otoliths from young silver hake,
Merluccius bilinearis (Mitch.). Int. Comm. Northwest
Atl. Fish. Res. Bull. 6:107-117.
PRESCOTT, R., AND M. MOORE.
1976. Cape Cod squid influx and pelagic bird gathering.
Event 93-76, The Center for Short-Lived Phenomena, Re-
view of Annual Events. Camb., Mass., Nov. 19, 1976.
SCHEVILL, W. E.
1957. Lagenorhynchus acutus o^ Cape CoA. J. Mammal.
37:128-129.
Sergeant, D. E., and H. D. Fisher.
1957. The smaller Cetacea of eastern Canadian waters.
J. Fish. Res. Board Can. 14:83-115.
College of the Atlantic
Bar Harbor, ME 04609
10 Riggs Street
Gloucester, MA 01930
103 Riverdale Drive
Orono, ME 04473
Steven K. Katona
Salvatore a. Testa VERDE
Bradley Barr
reciprocal hybridization between
the california and gulf of california
grunions, leuresthes tenuis and
leuresthes sardina (atherinidae)
The California grunion, Leuresthes tenuis, and the
Gulf of California grunion, L. sardina, are the only
fishes that temporarily leave the water during
spring high tides to deposit their eggs in beach
sand (Walker 1952). The eggs develop in the
nearly dry sand and hatch when uncovered and
agitated by the surf of the next series of high tides.
The grunions have an allopatric distribution.
The California grunion ranges from Monterey
Bay, Calif., to Bahia Magdalena, Baja California
Sur. The Gulf grunion is endemic to the Gulf of
California, ranging from Bahia Concepcion, Baja
California Sur, and Guaymas, Sonora, Mexico to
the mouth of the Rio Colorado (Moffatt and Thom-
son 1975).
Recent comparisons show that morphological,
physiological, and behavioral differences exist be-
tween the grunions. Morphologically very similar,
the most diagnostic characterictics distinguishing
them are lateral scale row counts; the mean
number in L. tenuis is 75 and in L. sardina is 55.
Gulf grunion adults are also significantly longer,
more slender, have a smaller eye diameter, and
are more lightly pigmented than those of the
California grunion (Moffatt 1974; Moffatt and
Thomson 1975). Gulf of California grunion have
wider embryonic and larval thermal tolerances, a
higher larval preferred temperature, and wider
larval salinity tolerances (Reynolds and Thomson
1974a, b, c; Reynolds et al. 1976, 1977; Moffatt
1977).
Light response remains positive in Gulf grunion
through adulthood, whereas the response shifts
from positive in the larvae to negative in the
adults of the California grunion (Walker 1952;
Reynolds and Thomson 1974c; Reynolds et al.
1977). In response to the shorter wave period in
the northern Gulf of California, the duration of the
spawning act of the Gulf grunion females is much
briefer than that of the California grunion females
(Thomson and Muench 1976; Muench 1977).
Only recently has the congeneric status of the
grunions been recognized (Moffatt 1974; Moffatt
and Thomson 1975). Evidence to date indicates
that the California grunion, the less primitive of
the two species, has adapted to the less fluctuating
tidal and thermal regimes of the California coast,
following isolation from an ancestral type by the
476
Baja California peninsula (Moffatt and Thomson
1975; Moffatt 1977).
Hybridization and hybrid survival experiments
have been widely used as indices of divergence and
have made valuable contributions as a tool in the
definition of phylogenetic relationships (Hubbs
1967, 1970). In an attempt to further illuminate
the relationship between the grunions, we made
artificial and reciprocal crosses and we report on
the first successful reciprocal hybridization of
Leuresthes tenuis and L. sardina.
Materials and Methods
Adult grunions, although easily obtained in
large numbers, are difficult to maintain and
transport alive. On 18 March 1976 (2330 PST),
milt from six California grunion males was col-
lected at Scripps Beach, La Jolla Calif., mixed in
the beaten yolks of two hen eggs (Bratanov and
Dikov 1961), and transported to El Golfo de Santa
Clara, Sonora, Mexico. The milt-yolk mixture,
maintained between 16° and 20 °C, was used to
fertilize the eggs from 8 to 10 Gulf grunion females
obtained at El Golfo on the following day ( 19
March) during a spawning run which began about
1700 MST. During this same run Gulf grunion
milt was collected from 6 to 7 males, transported in
the same manner and used to fertilize California
grunion eggs from about 10 females obtained dur-
ing a run that night at La Jolla at 0115 PST (20
March). One prior and four subsequent attempts
to hybridize the grunions were made during the
1975, 1976, and 1977 spawning seasons, but these
were unsuccessful because one or both grunions
failed to spawn.
The female grunions were rinsed thoroughly in
clean seawater before their eggs were stripped
directly into the milt-yolk mixture. The mixture
was diluted slightly with fresh seawater to in-
crease sperm motility, gently agitated, and kept
cool until the end of the spawning run. The eggs
were then strained, rinsed with seawater, and
placed in plastic refrigerator containers between
moist paper towels for transport and incubation.
Conspecific control embryos of each species were
obtained by mixing eggs and milt in a bucket of
seawater, one-third full, and did not involve the
transportation or preservation of milt in hen yolk.
When spawning individuals were plentiful, as at
the Gulf grunion run on 19 March, six to nine
males were stripped per one female in order to
achieve maximum fertilization levels (Moffatt
1977).
Both sets of hybrid fertilized eggs and the con-
specific controls of L. tenuis (18 March) and L.
sardina (19 March) were transported from San
Diego, Calif., to the University of Arizona at Tuc-
son aboard commercial airlines. Upon arrival (22
h postfertilization in L. sardina x L. tenuis and L.
sardina controls; 13 h in L. tenuis x L. sardina;
and 32 h inL. tenuis controls) each set of eggs was
inspected. Their development was monitored daily
thereafter.
Both California grunion spawning runs were
sparse at La Jolla. Therefore, the greatest portion
of eggs and sperm available were devoted to the
hybridization experiments and a low conspecific L.
tenuis sample size resulted. Consequently, the de-
velopmental and hatching data reported herein
for these embryos are a compilation of these few
controls and egg sets obtained on other occasions,
incubated at 20°C from 12 h postfertilization (Mof-
fatt 1977).
Yolk-sac larvae of the two hybrids and the con-
specific controls were placed in separate tanks
containing artificial seawater and raised on newly
hatched Artemia, freeze-dried marine zooplank-
ton, commercial staple food, and frozen Artemia
nauplii. Larvae of the hybrids and controls were
maintained for nearly 5 mo although initial mor-
tality rates (first 2 mo) in all groups were high
( >90% ). On 19 August, 141 days posthatching, the
aquaria air lines were fouled by compressor oil and
the few remaining hybrids and controls died. Only
two L. tenuis x L. sardina and nine L. sardina x
L. tenuis individuals survived to a size ( >12 mm)
at which the scale rows could be counted. This is
not to imply that scales might not have been pre-
sent prior to this time, merely that no attempt was
made to count them.
Results
At 22 h postfertilization, cleavage had pro-
gressed to the gastrula stage in L. sardina x L.
tenuis embryos as it had in theL. sardina controls.
The L. tenuis x L. sardina hybrids had reached a
32-cell blastodic stage at 13 h postfertilization as
do L. tenuis embryos.
Artificial fertilization levels in the conspecific
controls fell between 85 and 99'^ during the peaks
of their spawning seasons when male to female
ratios of 6 or 9:1 were available. The fertilization
477
levels of both hybrids ranged from 60 to 70%.
These diminished levels in the hybrids may have
resulted from a combination of several factors
such as: the low male to female ratios used ( <1:1);
decreased sperm motility in the viscous hen-yolk
medium; high sperm mortality due to time, star-
vation, t^imperature shock, handling, etc. or par-
tial reproductive isolation between the species in
the form of mild fertilization block to non-
conspecific spermatozoa.
The grunions, L. tenuis and L. sardina, showed
similar developmental rates (Moffatt 1977). De-
velopment proceeded normally in the hybrids and
at about the same rate as the controls. No unusual
embryonic mortality was observed in the hybrids,
evidence that these embryos were not gynogenetic
hybrids (Moore 1955).
Preliminary trials showed that hen's yolk and
seawater alone will not initiate cleavage in Gulf
grunion eggs. Precautions were taken to prevent
conspecific milt contamination. Preliminary
examination of cellular nuclei smears of develop-
ing embryos immersed in colchichine revealed
somatic chromosome numbers of about 2n = 40 in
all four sets of embryos (controls and hybrids),
further evidence that these embryos were true
diploid hybrids.
Grunion embryos will hatch after vigorous agi-
tation in seawater. On 31 March at 284 h (11.8
days) postfertilization, 65.6% of the L. sardina x
L. tenuis embryos hatched and 66.5% of the L.
tenuis x L. sardina embryos hatched at 272 h (1 1.4
days). These hatch times are similar to those of the
controls. Hatching can be induced in both grun-
ions at 10.2 days postfertilization when embryos
are incubated at 20° C (Moffatt 1977).
Newly hatched L. tenuis larvae are typically
more darkly pigmented; they have a larger eye
diameter; they are stronger swimmers; and they
are more capable of escaping net capture than
newly hatched L. sardina larvae (Moffatt 1977).
Leuresthes tenuis larvae are 10% longer (mean
total length = 7.70 mm) than those of L. sardina
( mean total length = 6.93 mm). The greater length
of the California grunion yolk-sac larvae occurs in
the postanal region as in the adults. California
grunion larvae are also 52%. heavier (mean dry
weight = 0.340 mg) whereas, the mean dry weight
ofL. sardina equals 0.223 mg (Moffatt 1977). The
greater length and weight of the California grun-
ion at hatching may be attributable to the 4.10
times greater ovum volume (Moffatt 1977; Moffatt
and Thomson in press).
These differences which distinguish the prolar-
vae of L. tenuis and L. sardina were also observed
in the hybrids. In most characteristics the L.
tenuis x L. sardina larvae were not visibly distin-
guishable from the maternal controls (L. tenuis),
e.g., size, pigmentation, and swimming ability.
However, the L. sardina x L. tenuis larvae ap-
peared to be somewhat intermediate to the con-
trols in extent of pigmentation and swimming
ability. At 2 wk after hatching the length and
pigment differences between the larvae were more
pronounced. Premaxillary teeth were visible in
the L. sardina x L. tenuis larvae but not in the
reciprocal hybrids. Again, hybrids closely resem-
bled the maternal controls. Gulf grunion adults
typically have much stronger dentition than do
the adults of the California grunion (Moffatt and
Thomson 1975).
As previously mentioned, the most diagnostic
differences between the adult grunions are the
lateral scale row counts. Scale counts of the 141-
day-old controls were essentially the same as
those of the adults (Table 1). The counts of the
hybrids were intermediate and significantly dif-
ferent from each other. Those shown by both hy-
brids were significantly different from those of both
parental species. The lateral scale rows of the hy-
brids were closer in number to those of the mater-
nal controls. Mean counts of L. sardina x L. tenuis
were 32% closer to those of L. sardina; and L.
tenuis x L. sardina were 20% closer to L. tenuis
than to those of the paternal parents, L. tenuis and
L. sardina. respectively.^ The intermediate counts
indicate paternal genome influence and that these
are indeed diploid hybrids.
'A mean hybrid count greater or less than 65 (the midvalue
between the parental species) indicated the affinity to one parent
or the other. The numerical affinities (percentages) were calcu-
lated as the ratio of the differences between 65 and the hybrid
count and between 65 and the adult counts.
Table l. — Means, ranges, n, and P values of lateral scale row
counts observed in 141-day-old hybrids and controls and adults
of the grunions, Leuresthes tenuis and L. sardina.
Parents
:
L tenuis
.
L sardina
Juveniles
Adults
P
Juveniles
Adults
P
L tenuis
L. sardina
X = 75 5
(75-76)
n = 2
X = 67.0*
(66-68)
n =2
P--0.01
X = 74 6
(69-80)
n = 143
•0 6
X = 61.8
(61-63)
n = 9
X = 55.1
(54-57)
n = 9
P< 0.001
X = 55 3
(51-60)
n = 177
>0.7
'Student's f-test comparison of the lateral scale rovi( counts between the two
hybrids P- 0 001.
478
Discussion
Natural hybridizations are reportedly more
common among freshwater fishes than among
marine fishes (Hubbs 1955). Hubbs (1970) stated
that, "teleost hybrids are relatively easily pro-
duced and if the parental morphology is similar
the hybrids are easily reared." The results of
natural and artificial amphibian and teleost
crosses have been widely employed for estimating
degrees of phylogenetic divergence, revealing sys-
tematic patterns, and explaining mechanisms
controlling development and differentiation.
Davidson (1968) reports that the closer the
phylogenetic relationship between the species
hybridized, the less likely the hybrid genome con-
trol will be displayed early in development. This is
because the mechanical aspects of early develop-
ment tend to be similar in closely related species
and may be primarily under the control of mater-
nal RNA accumulated in the egg prior to fertiliza-
tion. Davidson believes this, at least in part, ac-
counts for the commonly observed resemblance of
hybrids in early developmental stages to the mat-
ernal parent. The genetic influence of the paternal
genes in the hybrid genome may not be apparent
phenotypically until long after the onset of dif-
ferentiation (Davidson 1968).
It is possible that such mechanisms account for
the maternal resemblance pattern observed in the
grunion hybrids as well. The hybrids resembled
the maternal parents in overall size and body
proportions, coloration, swimming ability, net-
escape capability, and dentition until long after
hatching. Only when the lateral scale rows were
counted at 141 days after hatching did the
influence of paternal genes become visibly and
quantitatively apparent.
The numerous artificial and two natural hy-
bridizations (interspecific and intergeneric) re-
ported among the Atherinidae are reviewed by
Hubbs and Drewry (1959), Rubinoff (1961), and
Hubbs (1970). Natural hybrids reported between
Menidia menidia and M. beryllina along the At-
lantic coast of Florida (Gosline 1948) exhibit in-
termediate counts (i.e., scales and fin rays). Most
of the experimental crosses between these
atherinids resulted in low developmental success
and low survival rates except those of M. beryllina
9 X M. menidia 6 (Rubinoff 1961). Rubinoff did
not report whether any intermediate characteris-
tics existed in these hybrids nor was the reciprocal
cross attempted.
Geographically isolated species forms adapt to
their respective environments by the evolution of
appropriate gene complexes. Then, if sympatry
reccurs and hybridization takes place, hybrid in-
dividuals will usually be selected against (Mayr
1963; Ford 1964). Hybrids not selected against
will usually be successful over only a narrow geo-
graphical range, since in animals, natural hybrid-
ization is commonly associated with environ-
mental perturbation (Mayr 1963; Manwell and
Baker 1970).
The Menidia species are sympatric and hybrid-
ization does occur in northern Florida, a very nar-
row portion of the overlap in their ranges (Gosline
1948). Like these species, grunions are marine
fishes with similar, but not identical, ecological
preferences. However, the grunions are allopatric
and natural hybridization is not possible.
According to Mayr (1963), some investigators
argue that renewed sympatry with hybridization
is required as a process of speciation in order to
"perfect isolating mechanisms," and, therefore,
unlike the Menidia species, the heterospecific
status of the grunions may be questioned, espe-
cially in light of the hybridization success reported
herein. We conclude that, despite our success at
hybridizing L. tenuis and L. sardina, the mor-
phological, physiological, and behavioral distinc-
tions between them warrant their continued rec-
ognition as separate species.
Acknowledgments
We thank O. M. Moffatt, V. J. Moffatt, and D.
Dutcher for accompanying and assisting the
senior author on the collecting excursions. We
especially thank O. M. Moffatt for saving the pre-
served milt of the Gulf grunion from being swept
away by an unusually high wave at La Jolla. We
acknowledge the time and efforts of P. C. Cook in
preparing the preliminary somatic-cell smears,
and we appreciate the recommendations concern-
ing the preparation of this manuscript given by W.
W. Reynolds, C. D. Ziebell, E. A. Stull, and J. S.
Frost.
Literature Cited
BRATANOV, C, and v. DIKOV.
1961. Sur certaines particularites du sjserme chez les pois-
sons. Proc. Int. Congr. Anim. Reprod. 4:895-897.
Davidson, E. H.
1968. Gene activity in early development. Academic
Press, N.Y., 375 p.
479
FORD, E. B.
1964. Ecological genetics. John Wiley & Sons, Inc., N.Y.,
335 p.
GOSLINE, W. A.
1948. Speciation in fishes of the genus Menidia. Evolu-
tion 2:306-313.
HUBBS, C. L.
1955. Hybridization between fish species in nature. Syst.
Zool. 4:1-20.
HUBBS, C.
1967. Analysis of phylogenetic relationship using hy-
bridization techniques. Bull. Nat. Inst. Sci. India 34:48-
59.
1970. Teleost hydridization studies. Proc. Calif Acad.
Sci. 38:289-298.
HUBBS, C, AND G. E. DREWRY.
1959. Artificial production of an intergeneric atherinid
fish hybrid. Copeia 1959:80-81.
Manwell, C, and C. M. a. Baker.
1970. Molecular biology and the origin of species,
heterosis, protein polymorphism and animal breed-
ing. Univ. Wash. Press, Seattle, 394 p.
MAYR, E.
1963. Animal species and evolution. Belknap Press of
Harvard Univ. Press, Cambr., 797 p.
MOFFATT, N. M.
1974. A morphometric and meristic comparison of the Gulf
grunion, Leuresthes sardina (Jenkins and Evermann),
and the California grunion, Leuresthes tenuis ( Ayres). MS
Thesis, Univ. Arizona, Tucson, 36 p.
1977. Thermal effects on the survival and development of
embryonic grunions, Leuresthes sardina and L.
tenuis. Ph.D. Thesis, Univ. Arizona, Tucson, 88 p.
MOFFATT, N. M., AND D. A. THOMSON.
1975. Taxonomic status of the Gulf grunion (Leuresthes
sardina) and its relationship to the California grunion (L.
tenuis). Trans. San Diego Soc. Nat. Hist. 18:75-84.
In press. Tidal influence on the evolution of egg size in the
grunions (Leuresthes). Environ. Biol. Fishes.
MOORE, J. A.
1955. Abnormal combinations of nuclear and cytoplasmic
systems in frogs and toads. Adv. Gen. 7:139-182.
MUENCH, K. A.
1977. Behavioral ecology and spawning periodicity of the
Gulf of California grunion, Leuresthes sardina. Ph.D.
Thesis, Univ. Arizona, Tucson, 92 p.
REYNOLDS, W. W., AND D. A. THOMSON.
1974a. Tem|)erature and salinity tolerances of young Gulf
of California grunion, Leuresthes sardina (Atherini-
formes: Atherinidae). J. Mar, Res. 32:37-45.
1974b. Ontogenetic change in the response of the Gulf of
California grunion, Leuresthes sardina (Jenkins & Ever-
mann), to a salinity gradient. J. Exp. Mar. Biol. Ecol.
14:211-216.
1974c. Responses of young Gulf grunion, Leuresthes sar-
dina, to gradients of temperature, light, turbulence and
oxygen. Copeia 1974:747-758.
REYNOLDS, W. W., D. A. THOMSON, AND M. E. CASTERLIN.
1976. Temperature and salinity tolerances of larval
California grunion, Leuresthes tenuis (Ayres): a com-
parison with Gulf grunion, L. sardina (Jen-
kins & Evermann). J. Exp. Mar. Biol. Ecol. 24:73-82.
1977. Responses of young California grunion, Leuresthes
tenuis, to gradients of temperature and light. Copeia
1977:144-149.
480
RUBINOFF, I.
1961. Artificial hybridization of some atherinid fishes.
Copeia 1961:242-244.
THOMSON, D. A., AND K. A. MUENCH.
1976. Influence of tides and waves on the spawning be-
havior of the Gulf of California grunion, Leuresthes sar-
dina (Jenkins and Evermann). Bull. South. Calif Acad.
Sci. 75:198-203.
WALKER, B. W.
1952. A guide to the grunion. Calif Fish Game 38:409-
420.
Nancy M. Moffatt
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 271, La Jolla, CA 92038
DONALD A. Thomson
Department of Ecology and Evolutionary Biology
University of Arizona
Tucson, AZ 85721
TYCHOPLANKTONIC BLOODWORM,
GLYCERA DIBRANCHIATA, IN
SULLIVAN HARBOR, MAINE
The bloodworm, Glycera dibranchiata, is distri-
buted from the Gulf of St. Lawrence to the Gulf of
Mexico and from central California to lower
California and Mexico. It occurs from intertidal
water to 402 m depth (Pettibone 1963), but it is
more abundant in shallow coastal water. In Maine
and Nova Scotia the worms are dug commercially
along the coast from the upper layers of the inter-
tidal sand-silt-clay strata (Dow and Creaser 1970;
Anonymous 1974; Glidden^).
Spawning bloodworms are briefly pelagic occur-
ring in large numbers as they swarm in the after-
noon. Creaser ( 1973) observed swarming in Maine
during June. Simpson (1962) reported swarming
both in June and November-December, suggest-
ing a biannual spawning in Maryland. Klawe and
Dickie (1957) did not observe swarming by blood-
worms in Nova Scotia, although other evidence
indicated that the worms spawned in mid-May.
They suggested that the worms had a short noc-
turnal swarming period making them difficult to
observe. Simpson (1962) checked this possibility
•Glidden, P. E. 1951. Three commercially important poly-
chaete marine worms from Maine: Nereis (Neanthes) virens,
Glycera dibranchiata, Glycera americana. Rep. to Maine Dep.
Sea Shore Fish., Augusta, Maine.
in Maryland by making 40 observations with a
night-light between June and November. No
worms appeared at the surface under the light.
Individual bloodworms occasionally are pelagic
when not spawning. Pettibone (1963), when not-
ing the sightings of others, reported a bloodworm
swimming at the surface of Eel pond. Woods Hole,
Mass., on the evening of 17 August 1943; another
at the surface perhaps at the same pond on 28
January 1876; and another in Delaware Bay on 29
January 1957. No time was given for the two
January sightings. On 2 October 1969, E. P.
Greaser, Jr. sighted a bloodworm at the surface
near a dock on McKown Point, Boothbay Harbor,
Maine. The large nonspawner was observed at
noon swimming during a flood tide. We have found
that nonspawning bloodworms may also occur as
fairly abundant members of the tychoplankton —
bottom dwellers that are either swept upward
with tidal currents or migrate upward at night.
This study was originally designed to sample lar-
val Atlantic herring, Clupea harengus harengus
Linnaeus, and these results will be presented la-
ter. The implications of a large incidental catch of
bloodworms prompted our writing this note.
Materials and Methods
The site of this investigation, Sullivan Harbor,
is an embayment along the eastern coast of Maine.
It is divided into northern and southern sectors by
a constriction formed by an island, point of land,
and ledges (Figure 1). The southern sector opens
onto Frenchman Bay, which in turn opens onto the
Gulf of Maine. At its upper end, the northern sec-
tor constricts into a tidal falls. A narrow channel
extends north of the falls eventually bifurcating
into broad extensive shallows. Only small streams
enter these shallows about 5 km north of the
highway bridge. Sullivan Harbor is thus rela-
tively saline (31-32%o).
Six sampling stations were located within the
northern sector of the harbor; two in the landward
end of the channel (No. 3, 4), two in the seaward
end (No. 1,2), and one at each seaward entrance to
the subtidal flats (No. 5, 6). At each station within
the channel, four lines of buoyed and anchored
nets were set (Graham and Venno 1968). On each
line one net fished near the surface and a second at
3 m just above the edge (4 m) of the subtidal chan-
nel (Figure 1). A third net fished below the edge at
10 m and a fourth near the bottom (12-20 m). At
the entrance to the subtidal flats, one net was
suspended near the surface and another at 3 m just
above the bottom.
The nets were set at each station at dusk and
retrieved at dawn, fishing approximately one tidal
cycle. Calibrated meters centered within the nets
determined the amount of water strained. The
contents of the nets were preserved in the field
using a 5% Formalin^ solution. The sexes of the
worms were determined at a later date by inspec-
tion of the coelomic contents. Since variable
shrinkage of the worms made length measure-
ments unreliable, dry weight was obtained for
each worm.
Results
The nets strained 72 bloodworms from tidal cur-
rents during 6 of 10 cruises in autumn and winter
1974-75. During 1974, the nets captured 2 worms
on 14 October, 7 on 1 1 November, 2 on 5 December,
51 on 10 December, and 1 on 19 December. During
1975, the nets captured nine worms on 2 De-
cember. Only five worms were immature; their
weights varied from 0.02 to 0. 1 1 g. Mature females
outnumbered mature males about two to one
(41:24). The mean weights of the two sexes were
similar, 0.57 g, and their range varied from 0. 1 1 to
1.47 g.
Bloodworms were dispersed throughout the
water column and over both the channel and sub-
tidal flats. Nets at all stations and depths captured
worms. The average number netted was three and
ranged from one to seven. Of the 72 worms from all
cruises, nets set in the channel contained 58
worms and those over the subtidal flats held 14.
Their numbers decreased vertically: 33 near the
surface, 17 at 3 m, 15 at 10 m, and 7 near the
bottom.
An exceptionally large catch per unit effort was
obtained on 10 December. During the 10 cruises
the nets strained approximately 8,000 to 20,000
m^ of tidal water per cruise. Five of the sets yielded
catch rates varying from 0. 1 to 0.7 worm/1,000 m^.
A sixth set (10 December) yielded 3.38 worms/
1,000 m^. This catch rate was sufficiently large to
permit comparison of synoptic catch rates with
location and depth. The four lines of nets in the
channel strained 39 worms from 10,194 m^, yield-
ing a catch rate of 3.8 worms/1,000 m^ Those nets
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
481
68
15' Longitude W
43 32
43 30-
FlGURE 1— Sampling stations (1-6) for
larval Atlantic herring with sets of
buoyed and anchored nets in Sullivan
Harbor, Maine.
43 28-
-43 32
-43 30
43 26
in the flats strained 12 worms from 4,889 m^,
yielding 2.4 worms/1,000 m^. Shallow nets, above
the channel edge and those near the surface of the
flats, captured 37 worms by straining 9,614 m^ for
a catch rate of 3.8 worms/1,000 m^. Deep nets,
below the channel edge and those near the bottom
of the flats, captured 14 worms by straining 5,469
m^ for a catch rate of 2.6 worms/1.000 m^.
The numbers of worms captured in the nets were
few when compared with the numbers of smaller
tychoplankters, such as amphipods. Each indi-
vidual weight, however, was relatively large com-
pared with those individuals of more numerous
taxa and suggested that a large biomass of blood-
worms sometimes enters the water column of the
harbor.
Discussion
The mature bloodworms captured during winter
in buoyed nets at Sullivan Harbor were not free-
swimming spawners. Creaser (1973) sampled a
small worm flat at Wiscasset, Maine, from No-
vember 1967 to August 1969. During that time,
among the many worms dug, only three spawners
occurred during winter. Analysis of his collections
showed that egg diameters increased somewhat
during December and January but ceased growth
during the colder months of February and March.
Spawning was triggered in June by formation of
the epitoke, the growth of eggs to the spawning
"range" and a water temperature of at least 13°C.
These conditions were not found in the present
482
study. Also, we did not detect any morphological
changes that accompany formation of the epitokes
as described by Simpson (1962).
Swimming bloodworms at night have also been
reported for two other Maine inshore waters.
Dean^ saw 22 bloodworms during observations
made between 24 January and 29 March 1977 on
33 nights. The worms were present during five
nights in March and 15 were collected under a
night-light in the Damariscotta River, Maine — 8
on 11 March and 7 on 12 March. The gametes of
the worms were not sexually mature and the pre-
sence of the worms near the surface at night was
not related to spawning. Dean also reported that
buoyed and anchored nets set in Montsweag Bay
and the Sheepscot estuary between 1970 and pre-
sent captured 22 glycerids, some of which were G.
dibranchiata. In contrast, the senior author of this
paper did not capture bloodworms in buoyed and
anchored nets set in the Sheepscot estuary over
the same time period and in the same vicinity.
Possibly, the swimming of bloodworms at night is
sporadic.
A recent study of residual currents in Sullivan
Harbor suggested that the relatively shallow nets
above the edge of the channel (Figure 1) and at the
surface over the tidal flats strained a residual
seaward flow transporting tychoplankters and the
relatively deep nets strained a residual landward
flow. Distribution of bloodworms throughout the
water column would, therefore, insure their wide
dispersal by horizontal tidal currents, and it is
unlikely that after a tidal cycle they would regain
the location of their original burrows.
We hope to study further the bloodworms of
Sullivan Harbor and do not wish to speculate on
their origin or fate at this time. Rather, it is our
purpose to suggest that researchers investigating
bloodworms within their bottom habitat should
also examine their possible role as tychoplankters
for two reasons: populations of this important
commercial species in separate flats may become
intermixed, introducing problems in their man-
agement; and the reestablishment of worm popu-
lations previously destroyed by pollution or other
environmental catastrophe might proceed more
rapidly in those areas where there is winter trans-
port of mature worms, as well as the "normal"
dispersion of late spring larvae.
Acknowledgements
We thank C. Adams and D. Clifford of the Maine
Department of Marine Resources for collecting the
worms, sometimes under severe winter condi-
tions, and for processing the worms in the labora-
tory. We thank D. Dean for permitting us to cite
his unpublished manuscript.
Literature Cited
ANONYMOUS.
1974. Environmental inventory, benthic invertebrates. A
socio-economic and environmental inventory of the North
Atlantic Region, Vol. 1, p. 71-73. Res. Inst. Gulf Maine.
Greaser, E. p., Jr.
1973. Reproduction of the bloodworm (Glycera dibran-
chiata) in the Sheepscot Estuary, Maine. J. Fish. Res.
Board Gan. 30:161-166.
Dow, R. L., AND E. P. Greaser, Jr.
1970. Marine bait worms, a valuable inshore resource.
Atl. States Mar. Fish. Gomm. Leafl. 12, 4 p.
Graham, J. J., and P. M. W. Venno.
1968. Sampling larval herring from tidewaters with
buoyed and anchored nets. J. Fish Res. Board Can.
25:1169-1179.
KLAWE, W. L., AND L. M. DICKIE.
1957. Biology of the bloodworm, Glycera dibranchiata
Ehlers, and its relation to the bloodworm fishery of the
Maritime Provinces. Fish. Res. Board Gan., Bull. 115,
37 p.
PETTIBONE, M. H.
1963. Marine polychaete worms of the New England re-
gion. U.S. Natl. Mus., Bull. 227, 356 p.
Simpson, M.
1962. Reproduction of the polychaete Glycera dibran-
chiata at Solomons, Maryland. Biol. Bull. (Woods Hole)
123:396-411.
JOSEPH J. Graham
EDWIN P. Greaser, Jr.
Maine Department of Marine Resources
Research Laboratory
West Boothbay Harbor, ME 04575
'Dean, D. The swimming of bloodworms (Glycera spp.) at
night. Unpubl. manuscr.
SIMULATED FOOD PATCHES AND
SURVIVAL OF LARVAL BAY ANCHOVY,
ANCHOA MITCHILLI. AND SEA BREAM,
ARCHOSARGUS RHOMBOIDALIS
Survival rates of laboratory-reared marine fish
larvae often are directly related to prey concentra-
tion. Best survival usually has been reported
when prey are available at concentrations
> 1,000/1 ( O'Connell and Raymond 1970; Laurence
483
1974, 1977). Houde (in press) recently demon-
strated that survival of three species of marine fish
larvae from hatching to metamorphosis was 109c
or higher when mean prey concentrations were
only 34-130/1. But, he also found enhanced surviv-
al when food concentrations were increased. For
significant numbers of larvae to survive the tran-
sition stage from yolk nutrition to active feeding,
some researchers believe that dense patches of
prey must occur in the sea (O'Connell and
Raymond 1970; Hunter 1972). Such patches might
occur at densities of 10 to 1,000 times above the
mean prey density. Lasker (1975) has discussed
the dense patches of the dinoflagellate Gym-
nodinium splendens, which serves as prey for lar-
val northern anchovy, Engraulis mordax, in the
California Current and their possible relationship
to larval survival. Hunter and Thomas (1974) de-
monstrated that larval northern anchovies were
able to remain in patches of G. splendens that were
artificially created in laboratory experiments.
In two series of laboratory experiments we have
examined the effect of two simulated patches of
prey on survival in the bay anchovy, Anchoa
mitchilli, and the sea bream, Archosargus rhom-
boidalis. Patches were simulated during the first 6
days after hatching, when these larvae are most
susceptible to starvation mortality. The purpose of
the experiments was to determine if prey at high
density that were offered for more than some
minimum period would result in survival rates of
larvae that approached those obtained at a high,
constant prey concentration. This would indicate
that the larvae were able to obtain a daily ration
suitable for maintenance and growth by increas-
ing their feeding rate during the period of expo-
sure to the patch concentration of prey. At the low
prey concentrations usually found in the sea, a
relatively great expenditure of energy would be
required by larvae to obtain the minimum daily
ration for maintenance and growth. Such larvae
might weaken or fail to grow and thus be more
susceptible to starvation or predation.
Methods
Larvae were hatched from fertilized eggs that
were collected in plankton nets from Biscayne
Bay, Fla. In each experimental trial 140 sea bream
eggs were stocked (2.0/1) and 280 bay anchovy eggs
were stocked (4.0/1) in a 76-1 glass aquarium. Lar-
vae were reared for 10 days at 26±1°C. Salinities
ranged from 30.0 to 32.5%o for bay anchovy and
484
33.0 to 33.5%o for sea bream. Lighting was pro-
vided at 2500-2800 Ix by 40-W, cool-white fluores-
cent tubes. A 13 h light-11 h dark schedule was
maintained. Tanks were isolated in a black plastic
enclosure and all light was extinguished during
the dark periods. Sea bream and bay anchovy lar-
vae do not feed in the dark. At the end of experi-
ments, survivors were preserved in 5% Formalin^
and measured using an ocular micrometer.
Prey were the nauplii and copepodid stages of
copepods, approximately 50-100 /xm in diameter,
that were collected in 53-/Ltm mesh plankton nets.
Prey concentrations were determined by counting
organisms in 100- to 200-cm'^ aliquots from the
rearing tank (Houde 1975, 1977) several
times per day during the 13-h feeding period.
Background (i.e., nonpatch) prey levels were set at
25-50/1; this concentration was maintained when
patch concentrations were not offered from 2-6
days after hatching and continuously from 7-10
days after hatching. The patch concentration was
500 prey/1. Patches were provided for periods rang-
ing from 1.5 to 11 h (Tables 1,2). Both Oh, at which
no patches were provided, and 13h, at which a
constant 500/1 prey concentration was main-
tained, also were included in the series of experi-
ments for each species. The patch schedules were
maintained for only the first 5 days of active feed-
ing because larvae that survived that period had
greatly increased their searching ability and were
less dependent on high prey concentrations for
successful feeding.
Patches were created by adding prey to obtain
the 500/1 concentration. After larvae had fed at the
patch concentration for the desired period, prey
were reduced to 25-50/1 by siphoning them out of
the system through a 280-/>tm mesh screen and
replacing the siphoned water with 26°C filtered
seawater from a 150-1 header tank. Sea bream
larvae had no difficulty avoiding the siphon and its
screen during water exchanges, but precautions
were necessary for bay anchovy larvae. A 280-)U,m
mesh partition was used to "herd" anchovy larvae
toward one end of the tank prior to each siphoning
procedure. Siphoning procedures and water ex-
changes also were carried out in the 0-h and 13-h
patch period experiments to insure that those lar-
vae were exposed to the same procedural distur-
bances as larvae in experiments where prey con-
centration was being varied.
'Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
Larvae were exposed to the patch concentration
twice during each 13-h feeding period to obtain the
total desired time at the patch level. For example,
for a 6-h patch exposure, the prey concentration
was adjusted to 500/1 from 0800 to 1100; it was
then quickly reduced to 25-50/1, where it was
maintained until 1800, when the prey concentra-
tion was readjusted to 500/1 for the remaining 3 h
of the light cycle.
Results
Table 2. — Survival and standard lengths of Ardiosargus rham-
boidalis larvae at 10 days after hatching, based on 140 eggs and
variable patch exposure time. A patch is a prey concentration of
500/1. Nonpatch levels were 25-50/1. Patch conditions were pre-
sented to larvae on days 2-6 after hatching.
Patch exposure
time (h)
Survival
Standard l(
sngth (mm)
Percent
No.
Mean
SD
'0.0
3.57
5
4,21
0.44
1.5
22.00
31
404
0.56
3.0
32 14
45
3 77
0.56
6.0
59.29
83
387
0.40
9.0
41.43
58
3.31
0.41
11.0
6643
93
4.18
028
2130
42.00
59
4.21
0.31
Anchoa mi t chilli
'Food concentration was tneld constant at 25-50,'l during days 2-6.
^Food concentration was held constant at 500/1 dunng days 2-6, then re-
duced to 25-50/1 during days 7-10
Percent survival ranged from 0.36% at 0-h
patch exposure to 22.86% at 13 h (Table 1). The
steady increase in survival as patch exposure time
was increased was described by an exponential
function, Y = 0.3038 e034i9x^ where Y = percent
survival and X = hours at 500/1 prey concentration
(coefficient of determination, r^ = 0.98). For the
500/1 patch concentration, there was no minimum
time of exposure above which larval anchovy sur-
vival increased sharply or equalled the survival
obtained when larvae were exposed throughout
the day to the 500/1 prey concentration.
Surviving bay anchovies at 10 days after hatch-
ing differed significantly in mean standard
lengths (Table 1) among patch exposure times
(analysis of variance, P<0. 001). Mean lengths at
3-, 6-, and 9-h patch exposure times were sig-
nificantly greater than those at 11 and 13 h
(Student-Newman-Keuls test, P <0.05).
Table l. — Survival and standard lengths of Anchoa mitchiUi lar-
vae at 10 days after hatching based on 280 eggs and variable
patch exposure times. A patch is a prey concentration of 500/1.
Nonpatch levels were 25-50/1. Patch conditions were presented to
larvae on days 2-6 after hatching.
Patch exposure
Survival
Standard
length (mm)
time (h)
Percent
No
Mean
SD
'0.0
0.36
1
6.75
—
3.0
0.70
2
7.79
0.17
6.0
1 79
5
7.50
0.68
9.0
929
26
7.18
064
11.0
13.93
39
6.41
0.76
213.0
2286
64
6.58
0.58
'Food concentration was held constant at 25-50/1 during days 2-6.
^Food concentration was held constant at 500/1 during days 2-6. then re-
duced to 25-50/1 durino davs 7-10.
Archosargus rhoviboidalis
Survival ranged from 3.57 to 66.43% for sea
bream larvae over the range of patch exposure
times (Table 2). The relationship between percent
survival and patch exposure time was described by
a power function, 7 = 25. 07395s: 0 2878^ where Y =
percent survival and X = hours at 500/1 prey con-
centration. Although the power function described
the relationship reasonably well (coefficient of de-
termination, r^ = 0.94), an asymptotic regression
might be better to describe the relationship be-
cause sea bream larvae exposed to a 500/1 patch
density for between 3 and 6 h daily apparently
survived as well as when the 500/1 prey concentra-
tion was offered throughout the day. The power
function is retained here because fits to the data by
asymptotic regressions gave lower coefficients of
determination, due to the relatively high variabil-
ity in observed survival as patch exposure times
increased.
Mean lengths of survivors at 10 days (Table 2)
differed significantly among patch exposure times
(analysis of variance, P <0.001 ), but there was no
clear relationship between the mean lengths that
differed significantly (Student-Newman-Keuls
test, P<0.05) and the time of exposure.
Discussion
There was a marked difference in response of
bay anchovy and sea bream larvae to the simu-
lated patch conditions. Sea bream survival im-
proved greatly when larvae were presented with
prey at 500/1 for more than 3 h/day, the observed
survival then equaling that when they were of-
fered a constant 500/1 prey concentration. Bay an-
chovies were less successful in using the patch
conditions to improve their survival, although in-
creased survival rates did occur when larvae were
exposed for more than 6 h to the patch concentra-
tion. Results imply that first feeding bay anchovy
may require a high and stable prey density to
attain best survival in the sea, but that sea bream
485
are better adapted to survive under fluctuating
food conditions.
Survival observed in these experiments can be
compared with that reported previously (Houde in
press), when survival was related to prey den-
sities that were held constant from day 2 to day 16.
Predicted survivals at constant prey densities of
25-50/1 and 500/1 were 0.72-3.86% and 29.31%,
respectively, for bay anchovy larvae; and 5.94-
16.61% and 70.45% for sea bream larvae. Ob-
served survivals at 0-h and 13-h patch exposures
(Tables 1, 2), which correspond to the 25-50/1 and
500/1 constant prey concentrations, were only
slightly lower than those reported in the constant
prey level experiments (Houde in press). The
small differences probably were caused by the
siphoning and water exchange procedures which
did subject larvae to some stress. The similarity of
results in the two reports indicates that the patch
simulation procedure was effective in demonstrat-
ing the impact of patches on larval survival.
Growth results were inconclusive. Significant
differences in mean lengths were observed among
patch exposure times for both species (Tables 1,2).
In sea bream there was no clear relationship be-
tween mean lengths and patch exposure times,
but, unexpectedly, bay anchovy mean lengths
were smallest at the longest patch exposure times.
Presumably only the hardiest larvae survived
when patches were presented for only a short time,
and these larvae also may have had a relatively
great potential for growth. At the long exposures
to patch densities, survival was better, but no im-
provement in growth was noted, possibly because
some larvae with relatively poor growth potential
survived, or because of density-dependent effects
on growth that have been previously observed
(Houde 1975, 1977). Another compensating factor
was that patches were only presented on day 2 to
day 6 of the experiments, the prey concentrations
in all experiments being held constant at 25-50/1
from day 7 to day 10.
Only one possible patch regime was used in
these experiments. It is possible that other patch
densities or exposure schedules might alter re-
sults or conclusions. An infinite number of possi-
ble patch conditions could be simulated but future
experiments should be delayed until the temporal
and spatial scales of patchiness of organisms con-
sumed by marine fish larvae are better known.
Conditions that were simulated in these experi-
ments do not discount the possible ability of larvae
in the sea to maintain themselves within prey
486
patches that retain their integrity for days or
weeks. Hunter and Thomas (1974) demonstrated
that northern anchovy larvae could maintain
themselves within small patches of Gymnodinium
splendens in laboratory tanks. Lasker (1975)
found that feeding northern anchovy larvae were
relatively more abundant in the chlorophyll
maximum layer of the Los Angeles Bight, where
G. splendens was abundant, than in surface wa-
ters, and he suggested that larvae might be able to
maintain themselves in this rich source of food.
Bay anchovy larvae in our experiments derived
small benefits from the patch regime that we pro-
vided, but there may be stable patch conditions in
the sea which could greatly increase their poten-
tial for survival.
Acknowledgments
This study was supported by the National
Science Foundation, Biological Oceanography
Program, Grant OCE 74-18141. John Hunter re-
viewed and criticized an early draft of the manu-
script. Scott Siddall and A. Keith Taniguchi as-
sisted with the experiments.
Literature Cited
Houde, e. d.
1975. Effects of stocking density and food density on sur-
vival, growth and yield of laboratory-reared larvae of sea
bream Archosargus rhomboidalis (L.) (Sparidae). J.
Fish Biol. 7:115-127.
1977. Food concentration and stocking density effect on
survival and growth of laboratory-reared larvae of bay
anchovy Anchoa mitchiUi and lined sole Achirus
lineatus. Mar. Biol. (Berl.) 43:333-341.
In press. Critical food concentrations for larvae of three
species of subtropical marine fishes. Bull. Mar. Sci.
HUNTER, J. R.
1972. Swimming and feeding behavior of larval anchovy,
EngrauUs mordax. Fish. Bull., U.S. 70:821-838.
HUNTER, J. R., AND G. L. THOMAS.
1974. Effect of prey distribution and density on the search-
ing and feeding behaviour of larval anchovy EngrauUs
mordax Girard. In J. H. S. Blaxter (editor). The early
life history offish, p. 559-574. Springer- Verlag, N.Y.
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.
Laurence, G. C.
1974. Growth and survival of haddock [Melanogrammus
aeglefinus) larvae in relation to planktonic prey concen-
tration. J. Fish. Res. Board Can. 31:1415-1419.
1977. A bioenergetic model for the analysis of feeding and
survival potential of winter flounder, Pseudopleuronectes
americanus, larvae during the period from hatching to
metamorphosis. Fish. Bull, U.S. 75:529-546.
O'CONNELL, C. P., ANn L. P. RAYMOND.
1970. The effect of food density on survival and growth of
early post yolk-sac larvae of the northern anchovy iEn-
graulis mordax Girard) in the laboratory. J. Exp. Mar.
Biol. Ecol. 5:187-197.
Edward D. Houde
Richard C. Schekter
Division of Biology and Living Resources
Rosenstiel School of Marine and Atmospheric Science
University of Miami
4600 Rickenbacker Causeway, Miami FL 33149
DISCOVERY OF JUVENILE PACIFIC SALMON
(COHO) IN A SMALL COASTAL STREAM
OF NEW BRUNSWICK
Three juvenile Pacific salmon (Figure 1) were dis-
covered in a small coastal stream in southern New
Brunswick (Figure 2) in October 1976 while young
Atlantic salmon, Salmo salar, were being col-
lected for laboratory experiments. The Pacific
salmon were not recognized by the electrofishing
team, and their presence among the Atlantic salm-
on was not realized until the fish were sorted in
the laboratory some days or weeks later. Iden-
tification as either coho salmon, Oncorhynchus
kisutch, or chinook salmon, O. tshawytscha, was
later confirmed by W. B. Scott of Huntsman
Marine Laboratory, St. Andrews, N.B. Positive
identification to species of these juvenile fish was
not possible, but they were almost certainly coho
salmon because of recent introductions of this
species to the Atlantic coast.
Coho salmon are not native to the Atlantic, and
no populations reproducing in natural streams of
the Atlantic coast are known. Two aquaculture
operations using coho salmon are under way in
Maine, and coho salmon smolts have been released
in streams in New Hampshire and Massachusetts
since 1969 and 1971, respectively (Figure 2, inset).
Presumably, the parents were from one or more of
these four operations. No adults have been re-
ported from New Brunswick streams.
When the coho calmon were recognized, further
trips were made to obtain an estimate of their
numbers in the stream, their size, and habitat
preference in comparison with Atlantic salmon
and brook trout, Salvelinus fontinalis , which were
also present.
The stream, known locally as Frost Fish Creek,
drains into the estuary of the Digdeguash River
about 250 m from the Digdeguash Falls. It is a
small stream approximately 3 m wide in the lower
kilometer where all fishing took place. Its drain-
age area is approximately 570 ha. Discharge dur-
ing low summer fiow reaches as little as 80 1/s
(Symons and Harding 1974). The lowermost 0.25
km is steep with cascades and pools. The stream
here is either open to the sky or overhung with
alders. Most of the Atlantic salmon yearlings
occur in this portion of the stream. Through the
next 0.25 km upstream the gradient decreases;
occasional riffles are separated by pools and slow-
flowing water. Bankside cover consists of conifer-
ous softwoods partially clearcut. Atlantic salmon
yearlings and underyearlings occur in the riffles of
this section while the pools and quieter water are
inhabited by brook trout. Above this section the
Figure l.— Underyearling coho salmon captured on 28 October 1976 in Frost Fish Creek, N.B.
487
Figure 2. — Streams of southern New Brunswick and their access. Dots, sites of spot checks; star, site where coho salmon were
captured. Inset, province of New Brunswick, Canada, and northeastern United States showing location of aquaculture opera-
tions (Maine) or release sites (N. H. and Mass.) of coho salmon (squares) with respect to location of underyearling coho salmon
discovered in New Brunswick (star).
stream gradient becomes lower, the surrounding
area is swampy, the stream is choked with alders
and inhabited almost exclusively by brook trout.
Coho salmon occurred through the middle riffle-
pool section and extended in diminishing numbers
into the swamp area upstream.
To estimate the numbers of young coho in the
creek, two equal-effort electrofishings were per-
formed through the riffle-pool section and approx-
imately 50 m into the swampy section. The lower
fast section was fished separately during the first
trip (28 October), but since it contained no coho
salmon it was omitted on the second 20 days later.
Although some coho salmon might have moved
downstream in the period between the two
fishings, coho salmon were scarce in most up-
stream areas on both occasions, suggesting there
were few above the point where fishing ceased.
Twelve coho salmon were caught on the first trip
and five on the second. The total population esti-
mated by the depletion method ( Seber and Le Cren
1967) was 21. Three coho salmon had been cap-
488
tured during the collection trip on 9 October, so
that the total estimated population of coho salmon
in the stream was 24.
During electrofishing, particular note was
taken of the kind of habitat in which coho, Atlantic
salmon, and brook trout were captured. Coho salm-
on were found where there was immediate or
nearby overhead cover in the form of overhanging
banks, tree roots, or fallen trees or brush, and
where the water current was slow ( <30 cm/s). This
kind of habitat was also frequently occupied by
brook trout. On at least one occasion, a brook trout
and coho salmon were captured together. Atlantic
salmon were scarce above the lowermost steep sec-
tion of the stream. However, in October and
November three or four Atlantic salmon were cap-
tured in slow water where they had never been
seen in summer (Symons and Harding 1974).
These observations suggested that summer
habitat requirements of coho salmon were more
similar to those of brook trout than of Atlantic
salmon, although the latter may utilize brook
trout-coho salmon habitat in winter.
All captured coho salmon were retained and
taken to the laboratory for measuring and weigh-
ing. The average fork length of all coho salmon
captured was 89 mm, ranging from 75 to 100 mm.
There was no statistical difference between aver-
age lengths in October (89 mm) and in November
(91 mm). Examination of scales revealed that
these coho salmon were underyearlings. They
were considerably larger than underyearling At-
lantic salmon (60-70 mm fork length) and under-
yearling brook trout (40-60 mm) captured at the
same time. The coho salmon were retained for use
in laboratory experiments through the winter,
and the 1 0- 1 5 that survived were returned to Frost
Fish Creek the following April.
To investigate whether coho salmon might be
present elsewhere, spot checks were made in 17
nearby locations (Figure 2) between 28 October
and 17 November. Spot checks consisted of 10-35
min of electrofishing with most effort being ex-
pended in parts of streams having habitat similar
to that in which coho salmon were caught in Frost
Fish Creek. No coho salmon were found at any of
these sites. Brook trout were caught in all
streams, and Atlantic salmon were caught in
streams where they were known to occur. Brown
trout, Salnjo trutta, were caught in Frost Fish
Creek (2 individuals), Burns Brook ( 1 ), and Sorrel
Ridge Brook (1), all tributaries to the Digdeguash
River. Brown trout were introduced to the Dig-
deguash as early as 1921 (MacCrimmon and Mar-
shall 1968), and they continue to exist there in
small numbers.
In sum, an estimated population of 24 under-
yearling coho salmon was found in Frost Fish
Creek in fall 1976. No coho were discovered in
neighboring streams during a cursory search. Al-
though adult coho salmon are known to spawn in
small, gravelly coastal streams (Scott and
Crossman 1973), spawning may not have occurred
in the creek. Atlantic salmon apparently do not
spawn there despite the presence of young which
are thought to arrive from the main Digdeguash
River, having descended the falls into the estuary
and then reentering the nearest available fresh-
water. The young coho salmon may have arrived
by the same route. Regardless of the exact location
in which coho salmon spawned, should they estab-
lish a run in the river system, it would probably be
revealed by continued sampling of fish in the
creek.
Acknowledgments
Constructive criticisms of an earlier draft of the
manuscript were made by J. W. Saunders and R. L.
Saunders, whom we thank.
Literature Cited
MacCrimmon, H. R., and T. C. Marshall.
1968. World distribution of brown trout, Sa/mo^rw«a. J.
Fish. Res. Board Can. 25:2527-2548.
Scott, W. B., and E. J. Crossman.
1973. Freshwater fishes of Canada. Fish. Res. Board
Can., Bull. 184, 966 p.
SEBER, G. A. F., AND E. D. LE CREN.
1967. Estimating population parameters from catches
large relative to the population. J. Anim. Ecol. 36:631-
643.
Symons, p. E. K., and G. D. Harding.
1974. Biomass changes of stream fishes after forest spray-
ing with the insecticide fenitrothion. Fish. Res. Board
Can., Tech. Rep. 432, 47 p. + append.
PHILIP E. K. Symons
Department of Fisheries and Environment
Fisheries and Marine Service
Biological Station. St. Andrews, N.B.
Present address: Pacific Biological Station
Nanaimo, B.C. V9R 5K6
James D. Martin
Department of Fisheries and Environment
Fisheries and Marine Service
Biological Station. St. Andrews, N.B. EOG 2X0
489
SUBSAMPLER FOR ESTIMATING THE NUMBER
AND LENGTH FREQUENCY OF SMALL,
PRESERVED NEKTONIC ORGANISMS'
When many samples, containing large numbers of
organisms, must be processed it is often necessary
to take subsamples and assume that they are rep-
resentative of the total sample. Frequently sub-
samples are taken in some arbitrary fashion
which is described in such terms as "100 fish were
randomly selected." However, it is doubtful
whether any selection can be adequately random.
Therefore, numerous devices have been designed
in attempts to secure more representative sub-
samples and to increase the speed and efficiency of
subsampling.
Most subsamplers have been designed for use
with plankton, small benthos, and invertebrate
drift samples and are generally unsuitable for
larger organisms. However, Lewis and Garriott
( 1971) modified a Folsom plankton splitter for use
on meter net samples containing larval fish up to
19 mm long, and Hightower et al. ( 1976) described
a subsampler specifically designed for use with
nektonic organisms.
In the present paper I describe the design, oper-
ation, and efficiency of a subsampler originally
built for research on estuarine nekton (Herke
1971). The subsampler proved to be useful for es-
timating the number and length frequencies of
small nektonic organisms such as the bay an-
chovy, Anchoa mitchilli, tidewater silverside,
Menidia heryllina, and brown shrimp, Penaeus az-
tecus, as well as young of larger species such as
gulf menhaden, jBreyoor^/apafronus, and Atlantic
croaker, Micropogon undulatus. Although differ-
ent from most subsamplers, the design is fairly
similar to that described by Hightower et al.
(1976); it bears some similarities to those de-
scribed by Hewitt and Burrows ( 1948) for subsam-
pling live hatchery fish, by Gushing (1961) for
plankton, and by Sodergren (1974) and Hickley
(1975) for benthos.
My sampler differs from that of Hightower et al.
(1976) in at least four respects: 1) it has fewer
moving parts; 2) fewer water jets are required to
achieve through mixing of the sample; 3) a cen-
tral pillar or cylinder prevents organisms from
clumping in the center; and 4) the total sample is
'Contribution no. 24 of the Louisiana Cooperative Fishery
Research Unit: Louisiana State University, Louisiana Wildhfe
and Fisheries Commission, and U.S. Fish and Wildhfe Service
cooperating.
subdivided by raising vanes through the mixed
sample, rather than allowing the sample to settle
into baskets. Also, spin-dry weighing is required,
but it takes <1 min to complete the subsampling
process (after the organisms are placed in the sub-
sampler), rather than several minutes as required
for the subsampler described by Hightower et al. I
have made no comparative tests between the two
subsampler designs, however; individual cir-
cumstances may determine which would be most
practical in any given situation.
Subsampler Construction
My subsampler can be constructed of various
materials, and the same general design can be
used for large and small models. A small Plexi-
glas^ version ( Figure 1 ) has an outside diameter of
305 mm, and Herke (1971) also illustrated one
with a 580-mm outside diameter that utilized part
of a 208-1 steel drum for the outer cylinder, a 19-1
bucket for the inner cylinder, and plywood for the
false floor.
The subsampler in Figure 1 was constructed
primarily of Plexiglas about 6 mm thick. Plexiglas
joints were bonded with solvent (methylene
chloride and trichlorethylene). The major parts
and their functions are as follows (numbers refer
to the parts labeled in Figure 1):
1. Base.
2. Brass hinge for attaching base to edge of table
top.
3. Outer cylinder bonded to base; in addition to
solvent, a suitable cement may be required to
ensure a watertight seal to the base.
4. Gentral pillar of Plexiglas tube bonded to base
at exact center of circle formed by the outer
cylinder (3).
5. Rubber stopper in (4) to prevent material from
falling inside the piller.
6. Inner cylinder, which slides smoothly up and
down over (4).
7. Locking pin for holding (6) in the raised posi-
tion. Rubber bands around (6) and over a peg
through the shaft of (7) hold the pin in place
(these are omitted from the diagram to avoid
cluttering).
8. Vane bonded to (6); the outer edge almost
touches the outer cylinder. In the raised posi-
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
490
Figure l. — Basic design of the nektomc subsampler; see text for explanation.
tion shown, the three vanes subdivide the
sample into portions approximating 0.2, 0.2,
and 0.6 of the total.
9. False floor consisting of three sections bonded
to the inside of the outer cylinder (3) at a
height so that the upper surface is exactly
even with the upper edges of the vanes when
(6) is lowered to the base. The vanes move up
and down through the slits left between the
sections of the false floor. Enough space must
be left between the inner edges of the false
floor and the inner cylinder (6) and its at-
tached vanes to allow the inner cylinder and
vanes to move freely up and down. Con-
versely, the clearance must be small enough
to prevent organisms from falling into the
space below the false floor. Omitted from the
diagram are braces extending from the base to
near the inner edges of the two smaller sec-
tions of false floor.
10. Hinged door for removing subsampled or-
ganisms.
11. Latch holding the other door closed.
12. One side of a spout into which the water and
organisms pour when doors are opened; the
organisms are collected in a sieve below the
spout.
13. Rubber tubes to carry water; the middle one
enters the cylinder (3) beneath the false floor
(9).
491
14. Copper tube with outlets for each rubber tube
( 13). Water to operate the subsampler comes
through a large-diameter garden hose and
pistol-type hose nozzle (not shown) attached to
this tube.
15. Overflow tube attached to the outside of the
cylinder (3). The cut edges of a longitudinal
section of Plexiglas tubing are bonded to the
cylinder from the overflow intake to the bot-
tom of the base (1). Below the base the tube is
not sectioned (i.e., left intact) so a drain hose
can be attached to it.
16. Aluminum window screen covering overflow
intake; bottom of intake opening is level with
the top of the vanes (8) when they are raised.
17. Rubber stopper in drain hole below spout.
Also omitted from the diagram are "stops" on
the bottom edges of the vanes (which prevent the
vanes from pulling through the false floor) and
spongy, foam gaskets attached to the doors with
rubber cement.
Subsampling Procedure
In subsampling, one pushes the inner cylinder
(6) with the attached vanes down until it rests on
the base; in this position the tops of the vanes are
even with the top of the false floor so that the vanes
and floor form a single flat surface. The entire
sample is then placed on the false floor. The hose
nozzle trigger is squeezed fully open, squirting
water rapidly through the rubber tubing. (Nor-
mally, the space below the false floor is still filled
with water from previous use.) Some of the water
rises through the three vane slits in the false floor,
thereby inhibiting downward passage of the
smaller specimens; most of the water squirts out of
the upper tubes, causing the water above the false
floor to swirl rapidly. Turbulence thoroughly
mixes the sample as both sample and water re-
volve. When the water almost reaches the bottom
of the overflow intake, the inner cylinder (6) and
attached vanes are quickly raised as far as possi-
ble so that the locking pin (7) slides farther
through its hole in (6) and over the top edge of (4);
simultaneously, the hose nozzle trigger is re-
leased. The sample has now been divided into
three parts equal to about 0.2, 0.2, and 0.6 of the
whole.
The entire subsampler is next tilted on its
hinges (2) in preparation for emptying. If a 0.2
subsample is desired, only one door is opened and
492
the contents of that compartment flow through the
spout (12) into a sieve. (To avoid bias, the user
should always open the same door first. Occasion-
ally fish balance on top of the vanes; the user can
avoid personal bias by always pushing the fish so it
falls headfirst.) Opening both doors produces a 0.4
subsample and the remainder of the material in
the subsampler constitutes a 0.6 subsample. The
0.6 subsample is removed by first taking out the
0.4 subsample and then lowering the vanes as far
as they will go. The 0.6 subsample may then be
washed into a sieve below the spout. ( When remov-
ing any subsample, it is easier to wash the or-
ganisms out of the subsampler than to pick or push
them out.) A wide variety of subsample ratios can
be obtained by sequentially subsampling subsam-
ples (e.g., 0.8 x 0.2 x 0.2 = 0.032).
Small organisms do occasionally fall through
the vane slits into the space between the base and
the false floor. Such losses are normally insig-
nificant compared with the total number being
subsampled, but they are noticeable through the
Plexiglas. These organisms may be recovered by
washing them out through the drain hole plugged
by the rubber stopper (17).
No special leveling of the subsampler is re-
quired for proper operation; it may be mounted on
any reasonably level surface such as a table top or
laboratory bench.
Discussion
The subsampler is useful for estimating both
total numbers in a sample and the total length-
frequency distribution. If the total sample is not
first separated by species, one should at least make
a thorough scan of the sample, before subsam-
pling, to remove any unusually large or odd spec-
imens. As stated by Hightower et al. ( 1976), these
can later be added to the total estimate, which is
derived by extrapolating the subsample results.
However, subsampling can give erratic results for
inconspicuous species present in small numbers.
Therefore, I think it usually is best to first sepa-
rate the total sample into individual species, and
subsample only the abundant ones. For each of
these species, a subsample is first taken, and its
weight and that of the remainder are obtained by
the spin-dry method described by Herke (1973).
(In contrast to plankton, preserved fishes and
many crustaceans can be easily and precisely
weighed without damage by using the spin-dry
method.) All organisms in the subsample are then
counted and the number in the total sample is
estimated on the basis of the weights of the sub-
sample and total sample. Since the estimate is
based on weight rather than volume, the three
vanes need not divide the subsampler into exactly
0.2, 0.2, and 0.6 segments.
If a length-frequency estimate is desired, the
subsample can be further subsampled. Since the
number in the first subsample is now knovra, any
desired number for the length-frequency subsam-
ple can be closely approximated by selecting the
proper sequence of subsamples. For instance, sup-
pose the first subsample contains 3,371 anchovies
and a length frequency is desired from approxi-
mately 100 fish; 3,371 x 0.2 x 0.4 x 0.4 = 108.
Therefore, subsamples taken in this sequence
should produce the desired number for measuring.
The consistency with which the desired number
is obtained may be judged (Table 1) by comparing
the "theoretical" and "actual" numbers obtained
in 20 successive trials. The two subsamplers used
in these trials had a tendency to slightly exceed
the desired number; one or both of the smaller
compartments in each subsampler probably con-
tained a bit more than 0.2 of the whole. However,
the increased subsample size actually improves
the probability of obtaining an accurate length-
frequency estimate. Also, with use, one soon
learns whether the tendency is to obtain more or
fewer than the theoretical number and can select
the subsampling sequence accordingly.
How well the length-frequency estimates de-
rived from subsampling groups of anchovies and
menhaden represented the true length frequen-
cies of the groups was examined by using the
Kolmogorov-Smirnov one-sample, two-tailed test,
which is a test of goodness of fit. The test involves
comparing the observed cumulative frequency
distribution from a subsample with the cumula-
tive frequency distribution of the total sample. It
is sensitive to any kind of difference between the
two distributions — differences in location (central
tendency), in dispersion, in skewness, etc. Accord-
ing to Siegel ( 1956) the Kolmogorov-Smirnov test
is definitely more powerful than the chi-square
test when samples are small, and may be more
powerful in all cases.
The cumulative length-frequency distribution
for only one subsample was significantly different
(oc = 0.05) from its corresponding total sample
(Table 1 ). In the other 19 tests, the probability was
greater than 0.20 that a divergence of the observed
magnitude would occur if the observations were
really a random subsample from the total sample
(0.20 is the highest probability listed in Siegel's
table).
Table L— Results of 20 tests to determine the correspondence between: 1) the theoretical and actual number of bay
anchovies or gulf menhaden in the subsample, and 2) the cumulative length frequency distribution of fish in the
subsample and in the corresponding total sample. Subsamples were returned to the total sample after each trial. The
cumulative distribution shown in italics (in the same row with the number in the total sample) was the true
distribution obtained by measuring every fish in the sample.
Number in
total sample
Subsample
sequence
Final subsample no.
Standard length
in millimeters'
Theoretical
Actual
15
20
25
30
35
40
45
50
55
60
3,371
0.539
.481
0 772
.797
0B2^
835
0.867
880
0914
.947
0.957
.970
0.984
.977
0.995
992
7 000
anchovies
(0.2) (0.4) (0.4)
108
133
1,000
124
.524
.758
838
863
.911
960
.976
1 000
134
2.418
.739
.791
858
932
962
1.000
141
.489
.709
.773
822
894
.950
.979
993
1.000
146
.479
.740
781
.856
925
938
986
1.000
1.505
. 0.367
.347
553
.551
.643
.571
.774
673
%46
734
908
.795
964
.917
.997
.958
998
999
.999
anchovies
(0.2) (0.2)
60
49
59
.322
.576
.661
729
814
848
.933
967
1.001
77
.338
520
559
.676
793
832
949
988
1 001
70
.357
528
628
728
799
885
.956
985
999
71
.389
.570
.598
667
.778
.875
.958
1 000
Same
(0.4) (0.2)
120
128
.328
586
672
766
836
.914
992
1.000
1,505
125
.272
.544
600
.776
864
.920
.976
1 000
anchovies
133
.353
.556
.654
789
842
.880
.940
985
.993
1 001
134
.306
.507
589
.768
.843
.903
970
1.000
152
.283
.526
.598
710
.780
881
934
,987
1 000
1.221
.020
000
273
.278
756
.656
.980
1.000
.998
7.000
menhaden
(0,4) (0,2)
98
90
116
.026
.198
.733
1.000
115
.017
252
.765
991
1.000
128
.031
.242
664
969
984
1.000
113
.027
.345
.796
1.000
'Measured in 5-mm increments: i.e.. 15 = 15.0-19.9, 20 = 20.0-24.9, etc.
^The probability of a divergence this large in a random subsample from the total sample was between 0.05 and 0.01 The probability for the 1 9
other subsamples was >0.20
493
Number in subsample :
133
124
134
141
146
3,371 (total)
FIGURE 2.— Length-frequency distri-
bution of a total sample of 3,371 bay
anchovies, and of each of five subsam-
ples taken from the total. (From Herke
1971.)
If)
o
(M
ID
CM
in
o
in
O
in
o
in
ro
^
'J
in
in
ID
s
Stondard length m millimeters
It is difficult to visualize, from inspection of the
cumulative length-frequency distributions, how
well the percentage of fish in each subsample
length group represents the percentage in the cor-
responding length group in the total sample.
Therefore, this comparison is shown graphically
(Figure 2) for the first five subsamples listed in
Table 1.
Literature Cited
Gushing, C. E., Jr.
1961. A plankton sub-sampler. Limnol. Oceanogr.
6:489-490.
Herke, 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.
University Microfilms, Ann Arbor, Mich. (Diss. Abstr.
32:2654-B.)
1973. Spin-drying of preserved fishes and macroinverte-
brates. Trans. Am. Fish. Soc. 102:643-645.
HEWITT, G. S., AND R. E. Burrows.
1948. Improved method for enumerating hatchery fish
populations. Prog. Fish-Cult. 10:23-27.
HICKLEY, P.
1975. An apparatus for subdividing benthos samples.
Oikos 26:92-96.
HIGHTOWER, G. M., K. T. KIMBALL, AND C. A. BEDINGER, JR.
1976. A water-powered mechanical device for accurately
subsampling large numbers of nektonic organisms.
Trans. Am. Fish. Soc. 105:509-513.
LEWIS, S. A., AND D. D. GARRIOTT.
1971. A modified Folsom plankton splitter for analysis of
meter net samples. Proc. Annu. Conf. Southeast. Assoc.
Game Fish Comm. 24:332-337.
SIEGEL, S.
1956. Nonparametric statistics for the behavioral sci-
ences. McGraw-Hill, N.Y., 312 p.
SODERGREN, S.
1974. A simple subsampler for stream-bottom-fauna sam-
ples. Arch. Hydrobiol. 73:549-551.
WILLIAM H. HERKE
Louisiana Cooperative Fishery Research Unit
Louisiana State University
Baton Rouge, LA 70803
494
ERRATUM
Fishery Bulletin, Vol. 76, No. 1
Berrien, Peter L., "Eggs and larvae o( Scomber scombrus and Scomber japonicus in continental shelf
waters between Massachusetts and Florida," p. 95-115.
1) Page 99, left column, line 9, correct line to read:
or seven pterygiophores in the 2d through 5th (deleting the zero)
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 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 Publi-
cation No. 6, A List of Common and Scientific
Names of Fishes from the United States and
Canada, Third Edition, 1970.
Text footnotes should be typed separately
from the text.
Figures and tables, with their legends and
headings, should be self-explanatory, not requir-
ing reference to the text. Their placement should
be indicated in the right-hand margin of the
manuscript.
Preferably figures should be reduced by photog-
raphy 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. Photo-
graphs should be printed on glossy paper.
Do not send original drawings to the Scientific
Editor; if they, rather than the photographic re-
ductions, are needed 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 arabic
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 re-
ferred 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 literature cited.
The accuracy of the literature cited is the re-
sponsibility of the author. Abbreviations of names
of periodicals and serials should conform to Bio-
logical Abstracts List of Serials with Title Abbrevi-
ations. (Chemical Abstracts also uses this system,
which was developed by the American Standards
Association.)
Common abbreviations and symbols, such as
mm, m, g, ml, mg, °C (for Celsius), %, °/oo and so
forth, should be used. Abbreviate units of mea-
sure 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.
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 car-
bon copies. The sequence of the material should
be:
TITLE PAGE
ABSTRACT
TEXT
LITERATURE CITED
APPENDIX
TEXT FOOTNOTES
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. Jay C. Quast, Scientific Editor
Fishery Bulletin
Northwest and Alaska Fisheries Center
Auke Bay Laboratory,
National Marine Fisheries Service, NOAA
P.O. Box 155, Auke Bay, AK 99821
Fifty separates will be supplied to an author
free of charge and 100 supplied to his organiza-
tion. No covers will be supplied.
Contents-continued
HAYNES, EVAN. Description of larvae of a hippolytid shrimp, Lebheus groenlan-
dicus, reared in situ in Kachemak Bay, Alaska 457
AUSTIN, HERBERT M., and CLARENCE R. HICKEY, JR. Predicting abundance
of striped bass, Morone saxatilis, in New York waters from modal lengths 467
Notes
KATONA, STEVEN K., SALVATORE A. TESTAVERDE, and BRADLEY
BARR. Observations on a white-sided dolphin, Lagenorhynchus acutus, probably
killed in gill nets in the Gulf of Maine 475
MOFFATT, NANCY M., and DONALD A. THOMSON. Reciprocal hybridization
between the California and Gulf of California grunions, Leuresthes tenuis and
Leiiresthes sardina ( Atherinidael 476
GRAHAM, JOSEPH J., and EDWIN P. CREASER, JR. Tychoplanktonic blood-
worm, Glycera dibranchiato, in Sullivan Harbor, Maine 480
HOUDE, EDWARD D., and RICHARD C. SCHEKTER. Simulated food patches and
survival of larval bay anchovy, Anchoa mitchilli, and sea bream, Archosargus
rhomboidalis 483
SYMONS, PHILIP E. K., and JAMES D. MARTIN. Discovery of juvenile Pacific
salmon (coho) in a small coastal stream of New Brunswick 487
HERKE, WILLIAM H. Subsampler for estimating the number and length frequency
of mall preserved nektonic organisms 490
•i: GPO 796-049
,*
.<< °>^.
•at
^o
Fishery Bulletin
National Oceanic and Ati noi
^^ATES O^ ^
^M^&^dr^lAMiS^^ iMMlMke Fisheries Service
LIBRARY
NOV 0 0 1978
WHHfJSf HRI§: M;!c:c;
Vol. 76, No. 3
July 1978
CLARKE, THOMAS A. Diel feeding patterns of 16 species of mesopelagic fishes
from Hawaiian waters 495
FLETCHER, R. IAN. On the restructuring of the Pella-Tomlinson system 515
RIVARD, D., and L. J. BLEDSOE. Parameter estimation for the Pella-Tomlinson
stock production model under nonequilibrium conditions 523
LEIS, JEFFREY M. Systematics and zoogeography of the porcupinefishes iDi-
odon, Diodontidae, Tetraodontiformes), with comments on egg and larval develop-
ment 535
HEARD, WILLIAM R. Probable case of streambed overseeding — 1967 pink salmon,
Oncorhynchus gorbuscha. spawners and survival of their progeny in Sashin Creek,
southeastern Alaska 569
YOUNG, RICHARD EDWARD. Vertical distribution and photosensitive vesicles of
pelagic cephalopods from Hawaiian waters 583
VENRICK, E. L. Systematic sampling in a planktonic ecosystem 617
PEARCY, WILLIAM G. Distribution and abundance of small flatfishes and other
demersal fishes in a region of diverse sediments and bathymetry off Oregon . . . 629
PEARCY, WILLIAM G., and DANIL HANCOCK. Feeding habits of Dover sole.
Microstomas pacificus; rex sole, Glyptocephalus zachirus; slender sole, Lyopsetta
exilis; and Pacific sanddah, Citharichthys sordidus, in a region of diverse sediments
and bathymetry off Oregon 641
BARKLEY, RICHARD A., WILLIAM H. NEILL, and REGINALD M. GOODING.
Skipjack tuna, Katsuwoniis pelamis, habitat based on temperature and oxygen
requirements 653
QUINN, WILLIAM H., DAVID O. ZOPF, KENT S. SHORT, and RICHART T. W.
KUO YANG. Historical trends and statistics of the Southern Oscillation, El
Nino, and Indonesian droughts 663
FIEDLER, PAUL C. The precision of simulated transect surveys of northern an-
chovy, Engraulis mordax, school groups 679
Notes
PEREZ FARFANTE, ISABEL. Intersex anomalies in shrimp of the genus Penaeop-
sis (Crustacea: Penaeidae) 687
BONE, QUENTIN, JOE KICENIUK, and DAVID R. JONES. On the role of the
different fibre types in fish myotomes at intermediate swimming speeds 691
(Continued on back cover)
Seattle, Washington
0
U.S. DEPARTMENT OF COMMERCE
Juanita M. Kreps, Secretary
NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION
Richard A. Frank, Administrator
Terry L. Leitzell, Assistant Administrator for Fistieries
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.
n
EDITOR
Dr. Jay C. Quast
Scientific Editor, Fishery Bulletin
Northwest and Alaska Fisheries Center
Auke Bay Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 155, Auke Bay, AK 99821
Editorial Committee
Dr. Elbert H. Ahlstrom
National Marine Fisheries Service
Dr. Bruce B. Collette
National Marine Fisheries Service
Dr. Edward D. Houde
University of Miami
Dr. Merton C. Ingham
National Marine Fisheries Service
Dr. Reuben Lasker
National Marine Fisheries Service
Dr. Sally L. Richardson
Oregon State University
Kiyoshi G. Fukano, Managing Editor
The Fishery Bulletin is published quarterly by Scientific Publications Otflce. National Marine Fisheries Service. NOAA, Room 450,
1107 NE 45th Street, Seattle. WA 98105. Controlled circulation postage paid at Tacoma, Wash.
Although the contents have not been copyrighted and may be reprinted freely, 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 through 31 December 1978.
I
Fishery Bulletin
CONTENTS
Vol. 76, No. 3 July 1978
CLARKE. THOMAS A. Diel feeding patterns of 16 species of mesopelagic fishes
from Hawaiian waters 495
FLETCHER, R. IAN. On the restructuring of the Pella-Tomlinson system 515
RIVARD. D., and L. J. BLEDSOE. Parameter estimation for the Pella-Tomlinson
stock production model under nonequilibrium conditions 523
LEIS, JEFFREY M. Systematics and zoogeography of the porcupinefishes (Di-
odon, Diodontidae. Tetraodontiformes), with comments on egg and larval develop-
ment 535
HEARD, WILLIAM R. Probable case of streambed overseeding — 1967 pink salmon,
Oncorhynchus gorbuscha, spawners and survival of their progeny in Sashin Creek,
southeastern Alaska 569
YOUNG, RICHARD EDWARD. Vertical distribution and photosensitive vesicles of
pelagic cephalopods from Hawaiian waters 583
VENRICK, E. L. Systematic sampling in a planktonic ecosystem 617
PEARCY, WILLIAM G. Distribution and abundance of small flatfishes and other
demersal fishes in a region of diverse sediments and bathymetry off Oregon . . . 629
PEARCY, WILLIAM G., and DANIL HANCOCK. Feeding habits of Dover sole,
Microstomus pacificus: rex sole, Glyptocephalus zachirus; slender sole, Lyopsetta
exilis; and Pacific sanddab, Citharichthys sordidus. in a region of diverse sediments
and bathymetry off Oregon 641
BARKLEY, RICHARD A.. WILLIAM H. NEILL, and REGINALD M. GOODING.
Skipjack tuna, Katsuwonus pelaf7}is. habitat based on temperature and oxygen
requirements 653
QUINN, WILLIAM H., DAVID O. ZOPF, KENT S. SHORT, and RICHART T. W.
KUO YANG. Historical trends and statistics of the Southern Oscillation, El
Nino, and Indonesian droughts 663
FIEDLER, PAUL C. The precision of simulated transect surveys of northern an-
chovy, Engraulis mordax, school groups 679
Notes
PEREZ FARFANTE, ISABEL. Intersex anomalies in shrimp of the genus Penaeop-
sis (Crustacea: Penaeidae) 687
BONE, QUENTIN, JOE KICENIUK, and DAVID R. JONES. On the role of the
different fibre types in fish myotomes at intermediate swimming speeds 691
(Continued on next page)
Seattle, Washington
For sale by the Superintendent of Documents, U.S. Government Prmting Office. Washington,
DC 20402 — Subscription price per year; $12.00 domestic and $15.00 foreign. Cost per single
issue: $3.00 domestic and $.3.75 foreign.
Contents-continued
JEWETT, STEPHEN C. Summer food of the Pacific cod, Gadus macrocephalus, near
Kodiak Island, Alaska 700
LEMING. THOMAS D., and HILLMAN J. HOLLEY. A computer software system
for optimizing survey cruise tracks 706
Notices
NOAA Technical Reports NMFS published during the first 6 mo of 1978 715
Data on fisheries subjects accessioned through NMFS by NODC during the first 6 mo
of 1978 715
Vol. 76, No. 2 was published on 29 June 1978.
The National Marine Fisheries Service (NMFS) does not approve, rec-
ommend or endorse 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 pro-
motion 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.
DIEL FEEDING PATTERNS OF 16 SPECIES OF
MESOPELAGIC FISHES FROM HAWAIIAN WATERS
Thomas A. Clarke^
ABSTRACT
Diel patterns of stomach fullness, as percent of dry weight, were determined for 16 species of
mesopelagic fishes. Nine species of myctophids and one melamphaid, all vertical migrators, appeared
to feed solely or principally at night in the upper layers. These species encountered higher tempera-
tures and prey concentrations at night. Four species of stomiatoid fishes appeared to feed during the
day regardless of the extent of their migration or the absence thereof Prey concentrations encountered
by the stomiatoids during the daytime appeared to be higher than or similar to those encountered at
night. One myctophid and one gonostomatid showed no diel pattern; diel changes in the environmental
factors considered were relatively small in spite of the fact that both species undertook limited vertical
migrations.
Crude estimates of instantaneous evacuation rate and daily ration were made from data for four
species. These indicated that evacuation rate was increased at night in the upper layers and that daily
rations of species which migrated into the upper layers were similar to values for shallow-living
zooplanktonivores, while rations of deeper living species were lower. Thus while the adaptive value of
upward migration in the species which feed at night is obviously related to feeding activity, the upward
ascent by the daytime feeders may allow processing of larger daily rations than if they remained at low
temperature all day.
The extensive diel vertical migrations of certain
mesopelagic fishes have been well documented in
a vanety of oceanographic situations. While a
number of theories have been proposed to explain
the adaptive value of the behavior — in both fishes
and migrating invertebrates as well — data to
support any of them are few. One of the most
frequently proposed hypotheses (e.g., Marshall
1960) is that the organisms ascend at night to feed
in the upper layers where food is presumably at
higher concentrations and descend during the day
to avoid predation while the upper layers are well
lighted. Several studies of mesopelagic fishes (to
be cited below) have considered the relationship
between feeding chronology and vertical distribu-
tion in an effort to support at least one-half of the
hypothesis, but the results have for the most part
been rather equivocal. Apparent diel trends in
stomach fullness or details thereof are often ques-
tionable owing to low numbers of specimens
examined, insensitive methodology, or incomplete
diel coverage. Furthermore, all such studies on
mesopelagic fishes, with the exception of Merrett
and Roe (1974), have been conducted in high
'University of Hawaii, Department of Oceanography and
Hawaii Institute of Marine Biology, P.O. Box 1346, Kaneohe, HI
96744.
Manuscript accepted March 1978.
FISHERY BULLETIN: VOL. 76, NO. 3, 1978.
latitude or neritic situations and have dealt with
only one or, at most, three species.
This study considered the feeding chronology of
16 species from 5 families of mesopelagic fishes
from the north central Pacific Ocean. Vertical dis-
tribution and certain other apsects of the ecology
of these fishes are covered in Clarke (1973, 1974)
and Clarke and Wagner (1976); results from re-
lated investigations in the same study area are
summarized in Maynard et al. ( 1975). Comparison
of diel patterns of stomach fullness and diel
changes in temperature and prey concentration
allow consideration of adaptive value of the verti-
cal migrations undertaken by most of these
species. In four species, rough calculations of daily
ration are possible using equations similar to
those presented by Eggers (1977).
MATERIALS AND METHODS
Field Sampling
Specimens for this study were all collected with
a 3-m Isaacs-Kidd midv/ater trawl ca. 20 km west
of the island of Oahu, Hawaii (ca. lat. 2r20-30'N;
long. 158"20-30'W) in waters 2,000-3,000 m deep.
In order to reduce the concentration of zoo-
plankton in the cod end of the net and thus
495
FISHERY BULLETIN: VOL. 76, NO. 3
minimize bias due to fishes' feeding after capture,
the net terminated in a 1-m diameter cone of ca.
3-mm (Vs-in) knitted nylon mesh instead of the
"normal" plankton netting.
Specimens were taken in oblique tows which
sampled vertically migrating species at nine dif-
ferent periods of the 24-h cycle. At night, cable was
paid out in increments over a period of 1.5 h such
that the trawl fished roughly equal amounts of
time at all depths between the surface and ca. 350
m. The trawl was retrieved immediately after-
wards for a total towing time of about 2 h. Four
such tows were made between last light at dusk
and first light at dawn. During the day, 1,200 m of
wire were paid out initially. This placed the trawl
at ca. 350-400 m. Subsequently, cable was paid out
in increments such that the trawl fished between
this depth and ca. 1,100-1,200 m over a period of
1.5 h and then retrieved for a total fishing time of
ca. 2.5 h below 350-400 m. Three such tows were
made during the day. At dusk, 1,500 m of cable
were paid out initially, placing the trawl at ca. 500
m. Cable was then retrieved in increments such
that the trawl fished between 500 m and the sur-
face over 1 .5 h. The trawl reached maximum depth
just before sunset and was on deck shortly after
last light. At dawn, the process was reversed, and
the trawl shot before first light, and fished from
the surface to ca. 500 m over 1.5 h such that it
reached maximum depth ca. 1 h after sunrise. Ship
speed was ca. 1 m/s (2 kn) for cable retrieval and
ca. 2 m/s (4 kn) for all other phases.
In order to collect sufficient numbers of speci-
mens for as many species as possible, three 24-h
series of nine tows each (dusk, four at night, dawn,
and three during the day) were made 27-30 Au-
gust 1973. These dates were chosen to bracket new
moon (August 28) and minimize avoidance of the
trawl at night (Clarke 1973). One day tow of this
series was fouled and could not be repeated until
13 September 1973. The total range of time fished
by equivalent tows of each series (Table 1)
overlapped — considerably so for the night series
due to one night's fishing proceeding ahead of
schedule. The overlap was effectively less than
shown in Table 1 since most of the fishes analyzed
were probably taken below 50 m (based on previ-
ously cited studies of vertical distribution) and not
during the first 15 min or the last 5 min of each tow
when the trawl was shallower than 50 m. Con-
sequently, equivalent tows from each 24-h series
were considered replicates and specimens were
combined for data from each period. The nine
496
sampling periods will subsequently be designated
as follows: SS for sunset and SR for sunrise; Nl,
N2, N3, N4 for the four night periods in chronolog-
ical sequence; and Dl, D2, D3 for the three day-
time periods in sequence.
Danaphos oculatus, a nonmigrating species,
was not taken in the shallow night tows described
above. Nighttime data for this species were based
on specimens from three night series of three tows
each taken 30 August-1 September and 13-14 Sep-
tember 1973 (Table 1) using the same towing
schedule described above for daytime (ca. 400-
1,000 m). Thus only eight periods of the diel cycle
were considered. The three nighttime periods for
D. oculatus were designated dNl, dN2, and dN3.
In order to obtain more specimens of three
species of stomiatoids, I utilized specimens taken
24-25 May 1974 in seven tows at the same location
with the same net, and with the same procedure
and timings as Nl-4 and Dl-3. The numbers of
specimens used from this series will be noted in
the results. All specimens of the other species
came from 1973 collections.
The catch was immediately preserved in 4-5%
formaldehyde in seawater. The specimens re-
mained in this solution for up to 2 yr before pro-
cessing, but since all specimens of a given species
were processed within a period of 2-3 wk, any
between-sample differences in weight loss due to
leaching can be considered negligible.
Laboratory Analyses
The ratio of the dry weight of the stomach con-
tents to that of the fish as percent was used as an
index of stomach fullness. Where sufficient speci-
mens of a given species were available, 20 from
each of the nine sampling periods were examined.
If possible, the least damaged specimens (or
perhaps more appropriately — equivalently dam-
aged specimens) were selected from a narrow size
range. For many species, however, it was neces-
sary to use specimens damaged to various degrees
and of all sizes between recently (but fully)
metamorphosed juveniles and mature adults. In
cases where a specimen was damaged beyond loss
of scales or fin rays, i.e., where tissue was missing,
I used the median dry weight of other specimens of
the same standard length.
Each fish was briefly rinsed with tapwater and
gently blotted; standard length was measured to
the nearest millimeter. The stomach (anterior end
of the esophagus to the pyloric valve) was removed
CLARKE: DIEL FEEDING PATTERNS OF MESOPELAGIC FISHES
Table L— Towing times (Hawaiian Standard Time) for three 24-h series of oblique tows to sample vertically migrating mesopelagic
fishes and three all night series for deep-living nonmigratory fishes. Times given for dusk (SS), dawn (SR), and shallow night (Nl-4) are
for the entire tow; those for day tows (Dl-3) and deep night tows (dNl-3) are for the time the trawl fished below ca. 350-400 m.
Period
27-28 Aug.
28-29 Aug.
29-30 Aug.
Midpoint
Period
30-31 Aug.
31 Aug.-l Sept.
13-14 Sept.
Midpoint
SS
1754-1956
1815-1955
1820-2023
1910
dNI
2040-2300
2005-2233
2000-2235
2130
N1
2045-2240
2001-2155
2040-2233
2120
dN2
2352-0215
2325-0154
2330-0155
0050
N2
2315-0110
2207-0005
2255-0045
2340
dN3
0308-0540
0250-0515
0255-0525
0415
N3
0120-0320
0015-0210
0113-0310
0150
N4
0330-0520
0220-0420
0318-0515
0350
SR
0535-0742
0515-0725
0533-0745
0630
D1
0822-1047
0807-1034
0820-1045
0930
D2
1143-1425
1125-1350
1135-1410
1300
03
1510-1740'
1448-1710
1515-1740
1615
'The D3 tow for 28 August was fouled; time given is for tow made on 13 September.
and its contents, if any, placed on a clean glass
slide. The fish including the empty stomach was
placed in a preweighed aluminum pan. After
examination, the stomach contents were rinsed
into a second preweighed pan using distilled wa-
ter.
The stomach contents were examined only
casually. A rough estimate of fullness was made
and degree of digestion noted. Prosome length
(PL) of copepods and total length (TL) of other prey
were recorded from intact items. Intact prey items
could usually be identified to genus, but no serious
attempt was made to determine composition of the
diet from these samples. The remarks below on
types of prey include only the most frequently
encountered items and are not meant to be taken
as detailed analyses of diets.
Both fish and stomach contents were dried at
60°C for 24 h ( somewhat longer for a few large fish)
and allowed to cool under partial vacuum before
weighing. The pans with stomach contents were
weighed to the nearest 0.01 mg on a microbalance,
and the content weight determined by subtrac-
tion. Both control pans and reweighing of several
pans with dried stomach contents after a second
period in the drying oven or desiccator indicated
that the weighing and handling error was of the
order of ±0.02 mg. There was no indication that
error was proportional to the amount of material
in the pan. Pans with fish were weighed on a
semimicro balance; the reading was recorded to
0.01 mg on small fish and to 0.1 mg on those over
ca. 100 mg. Based on changes in weight of control
pans and reweighing offish after a second period of
desiccation, the error was <1% of the fish weight.
While the weighing and handling error was
such that estimates of stomach fullness were af-
fected only to the fourth or possibly third decimal
place, other errors or biases inherent in the mate-
rial should be mentioned. As noted above, an un-
known fraction of the material was lost due to
leaching. Damage to the fish positively biased the
ratios since there was some loss of skin, scales, or
fin rays in almost all specimens. Such errors were
unrelated to the time of collection and were more
likely to increase variability and thus to obscure
rather than cause diel trends in the data. The
intestinal contents, which were dried and weighed
with the fish, may have varied with time and thus
introduced a systematic error in fish weights.
Based on visual examination, however, largest
amounts of materials in the intestine were almost
certainly <1% of the total fish weight, and, con-
sequently, affected the stomach fullness index by
<0.1%.
The 2-3 h durations of the tows were a possible
source of bias and high variability. Bias in
stomach fullness could result from evacuation of
stomach contents between capture and death
(Eggers 1977). It is likely that this was negligible
since the fishes considered here were probably
dead soon after capture by the net. The 2-3 h possi-
ble differences in capture time for fishes from the
"same" period of the diel cycle almost certainly
contributed to the variability in stomach
fullness — particularly during periods when the
latter was changing rapidly.
Stomach fullness could possibly be biased nega-
tively by regurgitation after capture or positively
by feeding in the net. (Either type of bias would
tend to obscure rather than cause diel differences
in stomach fullness.) Regurgitation apparently
occurred infrequently in all species considered ex-
cept Lampanyctus nobilis. Except for the latter
(see below), specimens with partially digested food
remains in the mouth or everted stomachs were
not used. Hopkins and Baird (1975) showed that
feeding in the net is an unimportant source of
error even when a fine-mesh cod end is used, and
there was little indication of net feeding in the
present study. Zooplankton in good condition,
usually crustaceans with appendages erect and
extended, were infrequently found in the mouth.
These were assumed to have lodged there during
497
FISHERY BULLETIN: VOL. 76, NO. 3
capture and were not counted, but the fish and any
other contents were used. Items part way down the
esophagus with appendages flattened against the
body or the body folded were assumed to have been
eaten before capture and were included. I consi-
dered such "esophagus" items unlikely to have
been eaten after capture because concurrent
analyses of diet (Clarke in prep.) on the same
species collected by the same net indicate that
there is no difference in species composition be-
tween such items and items clearly in the stomach
and partially digested.
Stomach fullness values for a single species and
single time period were rarely distributed nor-
mally. Usually the values were skewed to the left,
but variably so — the mean being sometimes close
to the median and sometimes close to the 75th
percentile. Consequently, the entire set of
stomach fullness values for each species were
ranked and tested for between-period differences
by the Kruskal-Wallis nonparametric equivalent
of analysis of variance (//-test). The test is mainly
sensitive to differences in position (Tate and Clel-
land 1957), and significance implies differences
among the medians for the separate time periods
but does not single out which sets of data are
different. Each adjacent (in time) pair of data sets
was tested for differences in the median with the
Mann-Whitney or Rank sum test (Tate and Clel-
land 1957); however, because of multiple testing
on the same data, the significance levels from this
cannot be taken rigorously.
Neither test used is sensitive to differences in
variability, and no separate testing was done.
Some idea of differences in frequency distribution
can be gleaned from relative position of the mean
and median. Other gross differences, e.g., bi-
modality vs. unimodality, will be pointed out in
the results. Likewise, I did not test for possible
correlations between sex or size of the fish and
stomach fullness. The data from each period were,
however, ranked and compared (by inspection)
with sex and rank in length; no obvious correla-
tions were found.
RESULTS
A total of 15 vertically migrating species (10
myctophids, 4 stomiatoids, and 1 melamphaid)
and 1 nonmigrating stomiatoid were investigated.
These included species for which 20 individuals
were collected at most of the nine periods sampled
plus a few, less frequently taken species selected to
498
give broader coverage with respect to systematic
position or vertical distil jution pattern. In addi-
tion to graphical presentations (cited specifically
below), ancillary data for all species are sum-
marized in Table 2. In the subsequent presenta-
tion, stomachs were considered "empty" if
stomach fullness was <0.1'7r. This included both
visually empty stomachs and those with only a
trace of digested remains in the pyloric end of the
stomach. Types of prey organisms, state of diges-
tion, and other aspects not obvious from the
figures or Table 2 are considered in individual
species accounts below.
Comments on vertical distribution of prey items
are based on preliminary analyses of opening-
closing plankton tows taken in the study area and
their general agreement with data in the litera-
ture for the same or closely related species in other
central water mass localities. The plankton
tows — 16 taken in September 1973 and 20 in
November 1974 — covered the depth ranges of the
fishes considered both day and night. Euphausiids
from all samples have been counted and identified,
and copepods either counted (shallow night sam-
ples) or sufficiently examined to at least roughly
determine the depth ranges of the important prey
species. The apparent depth ranges agree gener-
ally with those given by Brinton (1967) and Roe
(1972). These two important types of prey can,
with a high degree of certainty, be classified as
shallow nonmigrators (above 200-300 m both day
and night), vertical migrators (above 200-300 m at
night and below this depth by day), and deep living
(below 300 m day and night). Similar statements
cannot be made for ostracods, the other important
crustacean group, nor for other taxa of zooplank-
ton.
Myctophidae
Bentbosema suhorhitale (Figure 1)
The //-test indicated highly significant
(P <0.005) differences in stomach fullness over the
diel cycle. The data from SS and Nl were charac-
terized by low averages, narrow percentile limits,
and high proportions of empty stomachs. Sub-
sequently stomach fullness generally increased
until SR and decreased throughout the day. The
most frequent prey items were copepods of the
genera Pleuromamma, Candacia, and Paracan-
dacia. Euphausia spp. and occasionally small de-
capods contributed significantly to the weight of
CLARKE: DIEL FEEDING PATTERNS OF MESOPELAGIC FISHES
Table 2. — Summary of data for each of the 16 species of fishes examined from each of the periods of the diel cycle. In each species/time
block, the first line gives the number of specimens examined and the number with stomach fullness <0.1% of fish dry weight in
parentheses; the second line, the size range of the specimens in millimeters standard length; and the third, the range of stomach fullness
values in percentage offish dry weight. All values of stomach fullness are rounded to the nearest 0.1%.
Species
SB
N1
N2
N3
N4
SR
D1
D2
D3
Benthosema suborbitale
20(12)
24-30
0-09
20( 8)
15-32
0-2.8
20( 3)
18-31
0-60
20( 1)
19-29
0.1-2.7
20( 5)
22-31
0 1-28
19( 2)
17-27
0-2.6
20( 2)
16-30
0.1-3.6
20( 3)
18-32
0-3.8
20( 2)
17-30
0-2.2
Bolinichthys longipes
20( 7)
24-46
0-1.4
20( 2)
21-46
0.1-2.0
20( 2)
20-51
0.1-2.5
20( 0)
17-49
02-15
20( 1)
18-43
0.1-2.4
20( 0)
21-32
0.3-4.6
20( 0)
16-50
0.2-48
20( 0)
17-46
0 1-15
20( 0)
19-48
0 1-11
Ceratoscopelus warmingi
20( 9)
22-53
0-28
20( 1)
18-45
0 1-98
20( 2)
24-45
0-81
20( 2)
23-50
0-50
20( 0)
25-47
0.1-7.6
5( 0)
18-23
0.9-4.9
20( 1)
19-48
0.1-2.4
20( 0)
19-59
0.1-4.5
20( 2)
18-52
0.1-7.4
Diaphus schmidti
20( 2)
25-39
0 1-23
20( 2)
20-41
0.1-1.7
20( 0)
19-41
0.3-1,4
20( 0)
19-40
02-60
20( 0)
17-38
03-38
20( 0)
20-38
0.1-22
20( 0)
14-37
0.2-2.3
20( 0)
15-38
0.2-2.7
20( 2)
19-38
0-2.3
Hygophum proximum
20(14)
26-33
0-1.0
20( 0)
19-32
0.2-67
20( 2)
18-38
0-5.8
20( 1)
18-42
0 1-3.6
20( 0)
18-37
0.2-7.0
20(14)
19-43
0-0 5
20(16)
19-42
0-0.4
20(13)
19-40
0-13
18(11)
18-46
0-1.3
Lampanyctus niger
20(14)
65-84
0-1.9
20(11)
65-85
0-1.0
20( 7)
63-83
0-1.7
20( 8)
64-85
0-13
20( 4)
51-85
0-1.5
0
20(10)
52-87
0-4.0
20(11)
68-85
0-3.7
20(10)
64-85
0-3.9
Lampanyctus nobilis
9( 4)
24-84
0-62
20( 3)
29-81
0-3.3
20( 2)
27-88
0-7.5
20( 4)
24-80
0-8.5
20( 1)
26-98
0-5.5
0
14( 2)
25-94
0-3.6
14( 4)
25-90
0-103
18( 6)
25-94
0-20
Lampanyctus steinbecki
20( 2)
25-39
0-27
20( 4)
22-48
0-3,7
20( 3)
22-41
0-4.4
20( 1)
21-39
0-35
20( 0)
22-42
0.2-4.0
20( 2)
18-36
0 1-9.8
20( 0)
22-42
0.1-4.5
20( 4)
23-44
0-2.1
20( 2)
27-42
0-5.5
Notolychnus valdiviae
20( 0)
19-24
0.2-3.4
20( 1)
20-23
01-2.9
20( 3)
19-23
0-3.0
20( 0)
19-23
0.3-4.7
20( 0)
18-23
0.3-3.6
20( 0)
19-24
0.3-35
20( 2)
20-23
0-1.7
20( 1)
17-22
0-25
20( 0)
19-24
0.1-3.5
Triphoturus nigrescens
20( 3)
18-34
0-5.9
20( 6)
17-31
0-36
16( 3)
15-34
0-9.0
15( 3)
15-35
0-6.6
19( 0)
16-33
1.0-14.3
20( 0)
15-36
07-9.9
20( 4)
16-34
0-16,0
20( 3)
18-33
0-6,7
20( 4)
18-34
0-4.5
Melamphaes danae
3( 0)
16-20
04-0.6
12( 3)
16-21
0-32
14( 2)
16-21
0-30
12( 0)
16-21
0.4-3.5
9( 0)
16-22
05-4.0
0
3( 0)
19-21
0.3-1 5
2( 0)
16
1.1-1.5
14( 1)
16-22
0-1.6
Gonostoma atlanticum
4( 0)
48-58
11-3.3
20( 2)
36-62
0-6.3
20( 1)
32-65
0-45
20( 2)
31-64
0.1-4.3
17( 4)
26-62
0-3.5
7( 2)
48-65
0.1-1.6
9( 0)
22-64
0.3-2.4
20( 1)
50-68
0.1-4.6
8( 0)
48-57
0.4-11.2
Gonostoma elongatum
20( 2)
26-112
0.1-2.6
20( 2)
31-126
0-13.1
20( 1)
30-135
01-59
20( 9)
30-132
0-5.0
20( 3)
32-79
0-2.0
8( 0)
30-150
02-80
20( 4)
24-125
0-166
14( 3)
30-149
0-5.2
10( 1)
29-87
0.1-7.3
Vinciguerria nimbaria
20( 0)
21-34
0.7-8.9
20( 0)
24-35
0.2-8.9
20( 3)
24-36
0-7.9
20( 3)
25-32
0-3.1
20(13)
23-34
0-0.6
20(15)
24-33
0-1.8
20( 9)
20-35
0-12.7
20( 1)
20-33
0.1-6.3
20( 2)
20-30
01-14.2
Danaphos oculatus
15( 0)
27-40
0.5-25
9( 0) 11( 0) 11( 0)
27-39 29-40 29-36
0.6-2.3 0.3-10 0.3-18
15( 2)
28-42
0-0.8
20( 2)
31-40
0-1.0
4( 0)
28-38
0.6-1.7
10( 0)
27-39
0.2-2.3
Valenciennellus
tripunctulatus
6( 0)
25-30
1.2-4.5
20( 0)
22-32
1.2-3.6
12( 0)
23-32
0.9-2.2
9( 2)
21-32
0-2.2
10( 1)
21-32
0.1-1.0
3( 2)
21-31
0-0.2
7( 2)
25-30
0-0.6
5( 0)
21-30
0.9-2.5
10( 0)
22-33
1.0-3.5
food. The food from specimens taken by day sam-
ples was generally well digested; except for the
thoracic spots or "buttons" from Pleuromamma,
prey was rarely recognizable beyond general
category.
Bolinichthys longipes (Figure 1)
Analyses of B. longipes were complicated by the
frequent presence in the stomachs of digenetic
trematodes. These were 1-10 mm long (most were
1-5 mm) and occurred in 419f of the stomachs
examined. They were mingled with the food and
appeared to have been fixed while wrapping
around or holding to items. As a probable conse-
quence, whole prey were rarely found inB. longipes'
stomachs. The parasites were, however, easily
separated from the food; they were not included
with either the fish or stomach content weight.
The number of trematodes was roughly a func-
tion of size of the fish. Fish < ca. 30 mm SL usually
had 0-2 individuals while several > 40 mm con-
tained 10-20. Since there was little between-
period difference in size composition of the fish
examined, there was no apparent correlation of
trematode number with time of day. Also there
499
FISHERY BULLETIN: VOL. 76. NO. 3
i
!.U
3. suborbitole
-
o *
r -1
r
' ' ^
r
"
-
-|
r
1
.0
: ^
:
T
o
— (
• (
I 1
1
i t
_
X
(
1
'-
J
L T
-1
L
—
X 1 J
L
-J
L
l/l
0
i *
'-
UO
1 1
r
2. Or
1.0-
B. longipes
<
> i
it o
O -1
r *
1
1
1 1 1 1
I
V
1
1' i
1
1 1
1 T
1 1
1
1 ■
1 ;
1
(
1 " '
1 i
I
D3 SS N) N2 N3 N4 SR Dl D2 D3
2 0 f—
"" C. warmingi
»«
3.05
20r
1.0
~1 1 — \ 1 1 1 r~
D3 SS Nl N2 N3 N4 SR Dl D2 D3
I I \ 1 1 1 1
1600 2000 0000 0400 0800 1200 1600
TIME OF DAY
D3 SS Nl N2 N3N4 SR Dl
D. schmidti
D2 D3
h
*
0
' 1
o
1 o
1 .
, 1 1
1 i.
1 . 1
( 1
< 1
f
— I 1 1 — I — I — I 1 1 1 1 —
D3 SS Nl N2 N3 N4 SR Dl D2 D3
I 1 1 1 1 1 1
1600 2000 0000 0400 0800 1200 1600
TIME OF DAY
Figure l. — Medians (dots), means ( x's), and ranges between 25th and 75th percentiles (solid vertical lines) of stomach fullness as
percentage of body weight throughout the diel cycle for four species of myctophids: Benthosema suborbitale, Bolinichthys longipes,
Ceratoscopelus warmingi, and Diaphus schmidti. Values are positioned at the midpoint of each sampling period (Hawaiian Standard
Time). Dashed vertical lines indicate significant differences between adjacent pairs (circle — 0.052'^ ). Because
of the latter, ranges and percentile limits were
broad, and means were much higher than me-
dians.
Ceratoscopelus warmingi fed on a wider variety
of taxa and sizes of prey than did the other species
covered here. The most frequent items were
copepods, ostracods, and small euphausiids, but
heteropods, siphonophores, and other zooplankton
also occurred. Intact items of such relatively small
prey were recorded mostly from specimens col-
lected at night; remains from day-collected speci-
mens were usually well digested. Ceratoscopelus
warmingi also took items up to 10'7( of bodily
weight; squid, other fishes, and large euphausiids
or decapods occurred in specimens >35-40 mm.
Such single large items accounted for nearly all
the fish with high values of stomach fullness, and
intact prey of this size occurred at all times of the
day. Most such items were vertically migrating
species that could have been taken at night, but
remains of nonmigrating Cyclothone spp., which
could have only been encountered between dawn
and dusk, were found in 1 1 specimens. Thus, while
the overall trend of the data indicates that C.
warmingi feeds principally on small zooplankton
in the upper layers at night, it probably takes
large prey whenever encountered.
Diaphus schmidti (Figure 1)
Diel differences in stomach fullness for D.
schmidti were highly significant (P<0.005), and
the trend was similar to that of the preceding
myctophids except for timing; the maximum value
occurred at Dl instead of SR. Empty stomachs
occurred only in a few specimens from D3, SS, and
Nl. Diaphus schmidti took a large variety of prey
items; the dominant taxa were small crustaceans
(ca. 0.5-3.0 mm PL or TL): ostracods, copepods, and
larval and juvenile malacostracans. Heteropods,
pteropods, polychaetes, and chaetognaths were
also noted. Excepting chaetognaths, few items
were >4-5 mm. Frequency of intact items was
highest at SR, and lowest at D3 and SS.
Hygophum proximum (Figure 2)
Diel differences in stomach fullness for//, prox-
imum were highly significant (P<0.005), and the
trend quite different from those of the other
species examined here. Most stomachs were
empty, and even 75th percentile values were zero
or nearly so between SR and SS; the peak value
occurred at N2. Hygophum proximum fed princi-
pally on medium-sized copepods (1-3 mm PL) and
occasionally other crustaceans. Less than 109f of
the stomachs were empty for any of the night
periods, but intact items were found frequently
only in stomachs from Nl. By N2 most of the prey
were unrecognizable, and only six items were rec-
ognizable to even general category in all the other
samples.
Larnpanyctus niger (Figure 2)
This species, one of three forms of the L. niger-
complex which occur near Hawaii, has minute
pectoral fins and lower AO counts than the others;
it was designated as "Form B" in Clarke (1973).
Zahuranec^ has recently identified the form as L.
niger (sensu stricto). There was evidence from
deep night tows taken during the same sampling
period that a fraction of the population ofL. niger
did not vertically migrate; consequently, some of
the day-caught specimens may not have ascended
to the upper layers the previous night. (Such
"non-migration" was also recorded in previous
studies, see Clarke 1973.)
The //-test indicated no significant diel differ-
ences in stomach fullness (P>0.10), and none of
the adjacent pairs differed significantly. The me-
dians from nighttime show a trend similar to that
of other myctophids, but the means were highest
during the day. No specimens were available from
SR. Values of stomach fullness were overall much
lower than observed in other species. Stomach
fullness exceeded 19c in only 21 of the 160 speci-
mens, and over 509c of the stomachs were empty at
all periods except N2, N3, and N4.
The most frequent food items were large
copepods of the familes Metridiidae, Euchaetidae,
and Aetideidae and small (<10-15 mm TL)
euphausiids. Occasionally small fishes were
found. Intact prey items were found in stomachs
from all periods. Deep-living copepods such as
Metridia and Pseudochirella were noted in day-
caught specimens indicating that at least some
feeding occurs during the day.
^B. J. Zahuranec. Oceanic Biology Program, Office of Naval
Research, Arlington, VA 22217. Personal communications, June
1977.
501
FISHERY BULLETIN: VOL. 76. NO. 3
A Or- !i proximum
3 0
2.0
1.0
:: II
1 1 1 r
1.0
1 1
o
1 •— « 1
111 1
-h-^ — f-+
D3 SS N1 N2 N3 N4 SR Dl D2 D3
3.U
L. nobilis
-1
r
■
2 0
*
T
r
o
1.0
0
1 i
~ X 1
" I I' J
1
1
1
1
1 1
1
1
1
1
1
1
1/1
Z D3 SS Nl N2 N3N4 SR Dl D2 D3
^30r N.valdiviae
U-
2.0
iin
-1 1 1 1 — I — I 1 1 I r~
D3 SS Nl N2 N3 N4 SR Dl D2 D3
D3 SS Nl N2 N3 N4 SR Dl D2 03
I I I \ \ 1 1
1600 2000 0000 0400 0800 1200 1600
TIME OF DAY
l.U
L. niger
J
"■
:
~ X
X
0.5
-
r 1
r
.
y- ., ::
X ■■ "
T T
T "■
1 1
I 1
1 ^
0
~
1 .
<
1 -L
1 1
>
1
.
D3 SS Nl N2 N3 N4 SR Dl D2 D3
3. Op L. steinbecki
2.0
10
60 mm. The small specimens had eaten mostly
copepods and amphipods 1-3 mm and Euphausia
spp. <10 mm, while the large ones had taken large
copepods ( >3 mm PL) and euphausiids, mysids,
sergestiids, and fishes 10-30 mm long. Intact prey
were found frequently in night specimens and oc-
casionally in those caught by day. The latter were,
with the exception of a single Lophothrix humili-
frons (apparently a deep-living copepod), migrat-
ing species that could have been taken at night.
One specimen from N4 contained; among the re-
mains of a euphausiid, crab megalopa, and
copepods; a partially digested insect (probably a
hymenopteran).
Lampanyctus steinbecki (Figure 2)
Stomach fullness values for L. steinbecki dif-
fered significantly (P<0.005) over the diel cycle.
The medians generally increased from SS to SR
and thereafter stayed at about 1% until a sharp
decrease between D3 and Nl. The percentage of
fish with empty stomachs was low for all periods.
The principal prey ofL. steinbecki were copepods >
ca. 2 mm PL — mostly aetideids, Pleuromamma,
and Candacia — and euphausiids. A few intact
items were found in specimens from Dl and D2 but
all were shallow-living or migrating species that
could have been taken the previous night. With
the exception of a s\r\g\ePareuchaeta sp. (probably
a deep-living nonmigrator), the prey from D3 and
SS were all well digested.
Notolychnus valdiviae (Figure 2)
The //-test indicated highly significant
(P<0.005) diel differences in stomach fullness for
N . valdiviae. Median values were low early in the
night and increased to a peak at N4. The minimum
value at Dl was slightly below the early night
values. Stomach fullness increased slightly until
SS and then decreased at SS-N 1 . The percentage of
fish with empty stomachs was low at all periods.
The positions of the 75th percentiles indicated
higher percentages of fish with relatively full
stomachs at N3, N4, SS, and D3.
Notolychnus valdiviae had taken a wide variety
of sizes (ca. 0.5-4.0 mm PL) and species of
copepods, but the bulk of the food in terms of
weight was made up by large (relative to the
weight of N. valdiviae) items such as
Pleuromamma xiphias, Candacia longimana, and
2-4 mm aetideids. Intact prey were more fre-
quently noted in specimens from N3 and N4 than
in those from the apparent "secondary peak" in
stomach fullness at D3 and SS. Considering only
those specimens with stomach fullness >2%
(whose numbers distinguish the peak periods from
others), only three of the nine from D3 and SS
contained intact or partially intact Pleuromam-
ma. The other six contained remains that were
either unrecognizable or barely so. In contrast, of
the 15 specimens from N3 and N4, 12 contained
1-3 intact items, while only 3 contained unrecog-
nizable remains. This plus the absence of any ap-
parent significant differences associated with the
D3/SS peak indicate that the latter was due to a
chance collection of a few more specimens that had
taken large meals the previous night rather than
to extensive daytime feeding.
Triphoturus nigrescens (Figure 2)
Overall diel differences in stomach fullness
were highly significant (P<0.005). Both medians
and means rose from low values at SS and Nl to a
peak at N4 and then, except for a slight increase at
D3, declined until SS. Due to the broad overlap in
503
ranges and percentiles for most pairs, only the
large increase between Nl and N2 was even mar-
ginally significant. The percentage of empty
stomachs was highest at Nl and zero at and just
after the peak at N4.
Triphoturus nigrescens fed principally on
Pleuromamma and Euphausia spp. Intact prey
were recorded more frequently during N2-N4 than
in other periods. As in the case of N. valdiviae, the
apparent peak at D3 was due to a few fishes' con-
taining large amounts of well-digested material
rather than freshly taken items.
Melamphaidae
Melamphaes danae (Figure 2)
Few M. danae were taken at any period. None
were taken at SR, and only two or three were
taken at SS, Dl, and D2. The data indicate a diel
trend similar to that of several myctophids, but
the//-test indicated that diel differences were only
marginally significant (P = ca. 0.10). If the data
from SS, Dl, and D2 were not included, the //-test
indicated significance atP = ca. 0.05 and the N4
and D3 values differed atP<0.05. This latter, and
statistically dubious, manipulation indicates that
the apparent trend in the data is real, but that
more specimens would be needed to confirm it
properly.
Melamphaes danae fed on a wide variety of zoo-
plankton including polychaetes and chaetognaths
as well as crustaceans — mostly small copepods
and ostracods. The copepods identified were all
either vertical migrators or shallow-living, non-
migrating species. Intact items were present in
nighttime specimens; those from daytime con-
tained remains barely identifiable to general tax-
on.
Gonostomatidae
Gonostoma atlanticum (Figure 3)
Relatively few G. atlanticum were available
from four periods even though 23 additional
specimens from the May 1974 collections were in-
cluded. Still there were significant (P<0.05) dif-
ferences in stomach fullness over the diel cycle.
Median values rose steadily from SR to D3, re-
mained at ca. 2% between D3 and N2, and then
dropped sharply between N2 and N3. Though the
median for N4 was slightly higher than that for
504
FISHERY BULLETIN; VOL. 76, NO. 3
either N3 or SR, the percentage of empty stomachs
was highest at N4 and SR, indicating an overall
trend for decrease during the late night.
Gonostoma atlanticum fed on large
copepods — mostly Pleuromamma xiphias, Can-
dacia longimana, and aetideids and scolecithricids
of several genera — and small (<10-15 mm)
euphausiids. Intact prey were found in stomachs
from all periods, but were mostly from the period
between Dl and N2. The majority of the contents
from N3 and N4 were well digested.
Gonostoma elongatum (Figure 3)
Relatively few G. elongatum were available
from three periods and the size range of individu-
als used was extremely broad (26-150 mm). There
is evidence from past studies that fractions of the
population occasionally do not migrate (Clarke
1974), but catches from deep night tows taken
during the same sampling period did not clearly
indicate whether or not this occurred during this
study.
The //-test indicated marginally significant dif-
ferences (0.05
«
1.0-
ihi
1
{
D3 SS Nl N2 N3 N4 SR Dl D2 D3
V. tripunctulotus
n
~~l \ 1 1 1 T T
D3 SS Nl N2 N3 N4 SR Dl D2 D3
D3 SS dNl dN2 dN3 SR Dl D2 D3
I 1 1 1 1 I I
1600 2000 0000 0400 0800 1200 1600
TIME OF DAY
70r
60-
5.0-
4.0-
1.0-
-
G. e
ongatom
~
-1
*
1
-I
r
- -
1
r
1
■»
:
1
-r
1
■\
"
1
r T •
1
J
L , 1
:, " '• I '
-'- 11
1 1
r
r
r
3.0- T
2.0-
i.\J
D. oculatus
o
T
:
1 -1
1
., 1
-^ 1' , ■
O o
1
1.0
1—
1 1 J
1
^ -1
[ '■! '
1 .
1
L J
0
III 1
i
1 ( 1
1
1
1 1
D3 SS Nl N2 N3N4 SR Dl D2 D3
6.U
V. nimbaria
5.0
♦ o
4.0
-
1 1
■ 1 ♦
3.0
-
-
' 0 o
*
(
' ' 1 1
2 0
-
' 1 1
' 1 1
1.0
0
1
1
1
1 (
1 1
1 1
1 1
. !'
1
II II
I 1
D3 SS Nl N2 N3 N4 SR Dl D2 D3
I 1 1 1 1 I I
1600 2000 0000 0400 0800 1200 1600
TIME OF DAY
Figure 3. — Stomach fullness throughout the diel cycle for five species of stomiatoids: Gonostoma atlanticum, Gonostoma elongatum,
Valenciennellus tripunctulatus , Danaphos oculatus, and Vinciguerria nimbaria. Symbols and format as in Figure 1.
rather than the medians, because there were
marked differences in frequency distribution for
these periods — differences to which the median is
not sensitive. At D2 the data were skewed to the
left with most values <1% and very few full
stomachs. At D3 the data were bimodal; 7 values
were higher than the mean of ca. 2.757f and 11
<1%. By SS, the data were again unimodal and
skewed slightly to the right with 16 values >2%
and only 2 <1%. Thus the trend between D2 and
SS was one of a gradual change in percentages of
the fish with very full stomachs, and the abrupt
increase in median values between D3 and SS
occurred as the high values became the dominant
mode. The percentages of empty stomachs showed
a trend opposite to that of the average values, i.e.,
an increase during the night and a decrease be-
tween SR and D2.
505
Vinciguerria nimbaria fed upon a wide variety
of sizes and taxa of prey. Small ( < ca. 2 mm PL)
copepods and ostracods were most frequent, but
larger copepods and small euphausiids occurred
regularly. Both the number of prey items and ab-
solute values of stomach fullness for the peak
period were higher than for most of the the other
species examined here; in several instances the
remains of 20-40 prey items were found in a single
stomach. Intact items were most frequent at SS
and common in specimens from day samples. Some
intact items were noted from Nl and a few from
N2, but stomachs from N3, N4, and SR contained
practically nothing but well-digested remains.
Sternoptychidae
Datictphos oculatus (Figure 3)
Few D. oculatus were available for any period
except Dl, and numbers were particularly low for
D2. Nine of the specimens used came from the May
1974 series. In spite of this, there was an evident
and highly significant (P<0.005) diel trend in
stomach fullness. Median values rose steadily
from a minimum at SR to a maximum at SS and
declined nearly constantly throughout the night.
There were a few empty stomachs at SR and Dl
and none at other periods. Danaphos oculatus fed
almost exclusively on Pleuromamma xiphias,
Euchaeta media, and similar-sized juveniles and
adults of several aetideid species. Intact items
were most frequently noted in D3 and SS speci-
mens; some were found in those from Dl and D2.
Almost none of the night specimens contained any
but well-digested remains.
Valenciennellus tripumtulatus (Figure 3)
Few V. tripunctulatus were available from any
period except Nl; 31 of the total examined came
from the May 1974 collections. Still, like D.
oculatus, V. tripunctulatus showed a clear and
highly significant iP <0.005) diel trend in stomach
fullness. Medians rose from zero at SR to a
maximum at SS and declined throughout the
night. The principal prey items were P. xiphias, P.
abdominalis, E. media, and similar-sized
aetideids. The stomachs from D2 to SS were nearly
uniformly packed with intact prey while those
from late night and SR were either empty or con-
tained only traces of well-digested remains.
FISHERY BULLETIN: VOL. 76, NO. 3
DISCUSSION
Feeding Chronology
Interpretation of data on stomach fullness is
limited because observed fullness is a function of
two rate processes — feeding rate and stomach
evacuation rate. Diel changes in stomach fullness
indicate that one or both rates vary over the diel
cycle, but without independent estimates of one or
the other, the only certain statements that can be
made are that feeding exceeds evacuation during
periods when fullness increases, the opposite
when fullness decreases, and that both rates are
zero when the stomach is empty. Notes on state of
digestion of stomach contents are helpful, but
must be interpreted with caution. Absence of in-
tact items indicates that feeding rate is zero, but
presence of intact items does not necessarily mean
feeding rate was positive during a given period
since some items may remain intact for an un-
known time after feeding ceases. Still, it is possi-
ble within these limits to qualitatively consider
changes in the two rates and to relate them to
environmental changes which the fishes en-
counter over the diel cycle.
The species considered here undergo diel
changes in numerous environmental factors, some
of which are likely to affect either feeding or
stomach evacuation rate in a qualitatively pre-
dictable manner. The migrating species encounter
higher temperatures at night. Diel temperature
changes for each species (Table 3) were deter-
mined using temperature-depth profiles from the
study area (Maynard et al. 1975 give profiles from
several seasons of three different years) and depth
ranges of the fishes (Clarke 1973, 1974; Clarke
and Wagner 1976). Because all species considered
occur below the steepest part of the thermocline
during the day, the magnitude of the diel tempera-
ture change is mostly a function of nighttime
depth range and not day depth or absolute range of
migration. For the same reason, juveniles, which
occur shallower than adults in most species
(Clarke 1973), incur greater temperature change
than adults of the same species. The migrating
species also encounter lower pressures and higher
oxygen concentrations at night (oxygen-depth
profiles for a site near the study area are given in
Gordon 1970).
Unless the fishes are able to regulate metab-
olism over the range of diel changes, the day-night
506
CLARKE: DIEL FEEDING PATTERNS OF MESOPELAGIC FISHES
Table 3. — Depth ranges, estimated diel changes in tempera-
ture, and probable day-night differences in prey concentration
for the 16 species of fishes considered. (See text for sources of
estimates.)
Depth range (m)
Temperature
Night
change ("C)
Prey density
Species
Day
Night-day
Night vs. day
Benthosema
0-100
18-19
N =D
suborbitale
500-600
Bolinichthys
50-150
16-19
N >D
longipes
500-700
Ceratoscopelus
0-150
16-20
N > -D
warmingi
600-1,000
Diaphus
0-75
17-19
N>D
schmidti
500-600
Hygophum
0-150
15-19
N>D
proximum
500-700
Lampanyctus
75-200
10-17
N>>D
steinbecki
600-1,000
Lampanyctus
50-150
16-20
N>>D
nobilis
600-1,200
Notolychnus
80-150
15-16
N>D
valdiviae
500-650
Triphoturus
25-100
17-19
N >D
nigrescens
550-750
Melamphaes
75-200
10-18
N>>D
danae
750-1,200
Danaphos
450-650
0
N<D
niger
650-900
differences in temperature and oxygen concentra-
tion both predict lower rates of metabolic proces-
ses in general and in particular lower feeding or
stomach evacuation rate during the day. Childress
(1975) and Childress and Nygaard (19731 indi-
cated that mesopelagic organisms can regulate
over a wider range of oxygen partial pressures
than these fishes encounter off Hawaii. Thus
temperature changes are more likely to affect rate
processes. Teal ( 1971 1 showed that increased pres-
sure can stimulate metabolic rates and thus
mediate or cancel effects of temperature; however,
it seems likely that temperature effects are pre-
dominant for the species considered here since
these fishes migrate through a much stronger
thermocline than did the shrimps studied by Teal.
As a consequence of vertical migration — by the
fishes and by many of their prey — the fishes en-
counter diel differences in prey concentration,
with which feeding rate is likely to be positively
correlated. As noted above, the depth distributions
of all prey species in the study area are not known
in detail; however, general, qualitative features
were evident from the available plankton samples
(see above). Most of the important prey species
were either shallow-living nonmigrators that oc-
curred above ca. 200 m day and night or were
vertical migrators with maximal concentrations
at ca. 300-450 m by day. Some important genera,
e.g. Euphausia, Pleuromamma, and Euchaeta, oc-
curred as deep as 600 m during the day but not at
high densities. At night, most copepods and many
of the euphausiids occurred at highest densities
above ca. 150-200 m. Many prey species occurred
between 200 and 300 m at night, but except for a
few euphausiid species, concentrations were much
lower than in the upper 200 m. Below ca. 600 m by
day and below ca. 300 m at night, total zoo-
plankton concentration was low and that of impor-
tant prey species nearly zero. Based on the above
features and the fishes' depth ranges, qualitative
estimates of day-night differences in prey concen-
tration were made for each species (Table 3).
Nine species of myctophids and probably
Melamphaes danae had similar diel patterns in
that median values of stomach fullness were min-
imal at or near dusk and increased only at night,
but details of the patterns were variable. Six
species, Benthosema suborbitale, Bolinichthys
longipes, Ceratoscopelus warmingi, Diaphus
schmidti, Lampanyctus steinbecki, and L. nobilis
(Figures 1, 2), had two periods of increasing
stomach fullness during the night separated by a
decline. Maximum stomach fullness occurred at or
near dawn, and the fish reached day depth with
relatively full stomachs. Stomach fullness ap-
peared to decrease during the day in some species
and showed no clear trend in others, but in most
there was a significant decrease at or near dusk. In
Notolychnus valdiviae, Triphoturus nigrescens,
and possibly Melamphaes danae (Figure 2), me-
dian stomach fullness appeared to increase stead-
ily throughout the night to a peak value just before
dawn. In the first two of these species, stomachs
were partially evacuated by the time they reached
day depth. In Hygophum proximum (Figure 2)
median fullness reached a peak value early in the
night, and stomachs were completely evacuated
by dawn.
For most of the above species there was no evi-
dence of significant feeding at depth during the
day. Intact items were more frequent at night, and
stomach contents of day-caught fish were usually
507
FISHERY BULLETIN: VOL 76, NO. 3
well digested. Lampanyctus nobilis and L. stein-
becki occasionally take deep-living copepods dur-
ing the day, and C. warmingi apparently takes
large items whenever it encounters them. Still,
the instances of definite day feeding were so few in
even the latter three species that the medians and,
therefore the diel patterns, were only marginally
affected.
All of these myctophids undergo diel changes in
temperature and prey concentration (Table 3) that
correlate with the observed pattern of feeding sole-
ly or mostly at night while in the upper 200 m. All
are at much higher temperatures at night. Al-
though some species occur as shallow as ca. 500-
600 m during the day and thus partially overlap
the daytime depth ranges of certain of their prey,
all occur below daytime maxima of prey concen-
trations and almost certainly encounter higher
concentrations at night. Certain details of the pat-
terns of stomach fullness indirectly indicate that
stomach evacuation rate may be lower during the
day as predicted by temperature differences. In
many species, stomach fullness did not clearly de-
crease during the day; since feeding rate was ap-
parently zero then, the evacuation rates must
have been low or zero. The sharpest declines in
stomach fullness occurred at or near dusk in most
species, near dawn in N. valdiviae and T. nigres-
cens, and during the night in H. proximum — not
during periods when the fishes remained within
their day depth ranges. In all cases except H. prox-
imum, however, something related to vertical
migration itself, e.g., activity, could be responsible
for the apparent increases in evacuation rates.
Four species of stomiatoids, Gonostoma atlan-
ticum, Danaphos oculatus, Valenciennellus
tripunctulatus, and Vinciguerria nimharia, fed
only during the day. The last three species occur
somewhat shallower by day then do the myc-
tophids and are consequently at or near depths of
maximum concentration of their prey then. The
upward migration of V. nimbaria is similar in
extent to that of its prey. Thus this species en-
counters little or no diel change. Danophos
oculatus does not migrate, and Valenciennellus
tripunctulatus migrates less than do its prey. Con-
sequently, both species occur below high concen-
trations of prey at night. The adults of G. atlan-
ticum (as were most specimens used here) occur
near the lower depth limits of most prey species
both day and night, and the day-night difference is
probably minor. Thus in these species, the day
depth ranges, rather than the occurrence or up-
508
ward extent of migration, seem more related to
observed feeding pattern.
All four species feed at nearly the same, low
temperature. Diel temperature change is zero for
D. oculatus, and relatively small for V.
tripunctulatus and large G. atlanticum because
they penetrate only part way through the ther-
mocline. Vinciguerria nimbaria undergoes a
change similar to that of the myctophids. The
temperature changes or lack thereof obviously
have no effect on feeding periodicity; however, the
steepness of the nighttime decline in stomach
fullness seems roughly correlated with nighttime
temperature indicating an effect on stomach
evacuation rates. This trend is considered in more
detail below.
Lampanyctus niger and G. elongatum, the two
species which showed no diel pattern in stomach
fullness, do not undergo large diel changes in
either temperature or prey concentration in spite
of the fact that they migrate. The large individuals
of both species I as were all theL. niger and mostG.
elongatum) undergo a relatively small tempera-
ture change. Likewise, only the smallest juveniles
of either species encounter markedly higher prey
concentrations at night. The relatively low values
of stomach fullness in both species and the pres-
ence of deep-living, nonmigrating zooplankton in
L. niger indicate that these two species feed at a
low rate whenever and wherever they encounter
prey.
Relationship to Previous Studies
Comparison of the present results with those of
previous studies is restricted because method-
ology in all cases was different from that of the
present study and in many cases equivocal or
probably not sensitive enough to discern diel
trends or lack thereof. With the exception of the
study by DeWitt and Cailliet (1972), appropriate
statistical testing was not done, and it is impossi-
ble to do so from the published data.
The most directly comparable study is that by
Holton (1967) on Lampanyctus (= Triphoturus)
mexicanus. Using 10 fish from each of eight
periods of the day, he determined dry weights, but
for some unknown reason weighed the entire
alimentary canal with the food. The minimal val-
ues observed, presumably from empty stomachs,
indicate that his "% nutrition" values should be
decreased by about 2.5-3 to make them roughly
comparable to those of the present study. Though
CLARKE: DIEL FEEDING PATTERNS OF MESOPELAGIC FISHES
the ranges and standard deviations of the data are
broad relative to the differences in means, the diel
trend in the latter is similar to that observed here
for H. proximum, i.e., peak value was reached
early in the evening and then dropped to low val-
ues and probably zero before the dawn descent.
Most previous studies have used visual esti-
mates of fullness with a scale of 3-5 ranks. Because
of the lack of "intercalibration" between inves-
tigators, only the rank for "empty" can be com-
pared unequivocally, and it is not certain in what
manner the ranks might correlate with percent-
ages of the fishes' dry bodily weight. Finally the
validity or absence of trends and details thereof
are questionable because scales of only 0-3 or 0-4
are rather insensitive. (Had only visual estimates
of fullness been used for the present study, only in
a few cases, e.g.,//. proximum or Valenciennellus
tripunctulatus, would the diel trends have been
obvious.)
Anderson's ( 1967) data on T. mexicanus indicate
a peak in stomach fullness just before sunrise, but
his data on degree of digestion indicate that fresh
food items were most frequent between sunset and
midnight. His data for Bathylagus stilbius (car-
diac portion of the stomach only) indicate two
separate periods of increasing fullness at night
and the sharpest decrease prior to ascent at dusk.
This pattern correlates with frequency of less-
digested prey items and is very similar to that
observed for several myctophids in this study.
Similar indices were used in the studies of four
species of high latitude myctophids: Benthosema
glaciale (Gjosaeter 1973) and Stenobrachius
leucopsarus, Diaphus theta, and Tarletonbeania
crenularis (Tyler and Pearcy 1975). Both studies
examined large numbers of specimens from each
of a few, very broad time periods. Their data indi-
cated highest percentages of full or nearly full
stomachs at night and highest percentages of low
values during the day. The occurrence of some full
stomachs during the day led Gjosaeter to conclude
that diel variation in feeding was not great and
Tyler and Pearcy to conclude that there was no
evidence against diurnal feeding. Both studies
noted a higher degree of digestion during the day.
These results are, however, consistent with the
possibility that like many of the myctophids in the
present study, their species descended at dawn
with full stomachs and did not evacuate them
completely until the dusk ascent. The latter may
well have not been detected in these studies due to
the broad time periods used.
Data on myctophids from recent studies by Mer-
rett and Roe (1974) and Baird et al. (1975) are
consistent with nocturnal feeding but are
equivocal to varying degrees due to low numbers
of specimens, incomplete diel coverage, or
methodology. Both studies based stomach fullness
estimates on counts of identifiable prey items. Ap-
parently, the presence of a single resistant part,
e.g., a Pleuromamma button, was counted the
same as an intact, whole individual of the same
taxon. Because of this and the likelihood that some
prey taxa or parts of prey are digested — and con-
comitantly rendered unrecognizable — at different
rates (e.g., Pandian 1967; Gannon 1976), such
counts seem to be insensitive or possibly biased
estimates of gut fullness — especially so when the
counts are used to back-calculate dry weight as
done by Baird et al. Furthermore, neither study
corrected the fullness estimate for fish weight,
which (using standard length ranges given by
these authors and assuming that weight is
roughly porportional to the cube of the length)
varied by factors of ca. 7-15 in the myctophids
covered by Merrett and Roe and ca. 2.75 in D.
taaningi, the species studied by Baird et al.
Merrett and Roe's data for L. cuprarious indi-
cated peak fullness in the middle of the night and a
decrease before the dawn descent — a pattern simi-
lar to that of//, proximum. Their data (or Lobian-
chia dolfleini and A^. valdiviae include no samples
between dusk and near dawn, but show fuller
stomachs at dawn. Data of Baird et al. for D.
taaningi are also similar to that for H . proximuni .
The rise in fullness from empty or nearly empty
stomachs in the afternoon to fairly high values in
early evening is evident and based on 39 and 9
specimens, respectively; however, the subsequent
decline is based on a single specimen from late
night and 4 from just after dawn ( 1 which con-
tained a fair amount of food).
Fewer stomiatoids have been examined
elsewhere, but much of the data available is con-
sistent with diurnal feeding. Perhaps the most
convincing data (because of good diel coverage and
numerous specimens) presented by Merrett and
Roe (1974) is that for Valenciennellus
tripunctulatus, which does not migrate in their
study area. The pattern is clearly similar to that
observed for the Hawaiian specimens. Hopkins
and Baird ( 1977) cited their own unpublished data
also indicating diurnal feeding for the same
species. Merrett and Roe (1974) interpreted dusk
peaks of numbers of items/nonempty stomach as
509
FISHERY BULLETIN: VOL. 76, NO. 3
an indication of dusk feeding activity in two
species of A rgy rope lee us; however, the data for A.
hemigymnus seem to me more consistent with in-
creasing stomach fullness throughout the day and
a nighttime decline. Except for high dawn values
(based on only three specimens from two tows), A.
aculeatus shows a similar trend.
DeWitt and Cailliet ( 1972) found no diel trend in
feeding ofCyclothone signata, but, based on fewer
empty stomachs in fish caught in the upper part of
the depth range, proposed that this species, al-
though it does not undertake diel vertical migra-
tions, may ascend irregularly to levels of higher
prey concentration to feed. Their data also indi-
cated that a deeper living species C. acclinidens,
had a higher percentage of empty stomachs by
day; as noted by the authors, the latter seems to
defy any reasonable explanation.
Legand et al. ( 1972) considered feeding chronol-
ogy of 14 species of mesopelagic fishes from the
South Pacific. Though trends in stomach fullness
of some species are similar to those noted here,
e.g., that for Triphoturus microchir (which almost
certainly = T. nigrescens) is very similar to that
for T. nigrescens near Hawaii, a number of species
show patterns quite different from those reported
by either the present or other studies. Interpreta-
tion of the validity of such "exceptions" is difficult
owing to the sparse presentation of Legand et al.
Though total numbers of specimens are fairly
high, it is not clear that they were equitably dis-
tributed among diel periods, from the same area,
or from the same season, etc. The percent fullness
values are obviously based on wet weights — an
imprecise measurement, particularly for stomach
contents — and it is not clear whether all fish and
stomach contents were weighed or some sort of
averaging or regression procedure was employed.
The feeding patterns shown by previous studies
cannot be compared in detail with those presented
here; however, there is general agreement in data
on the two dominant groups of mesopelagic fishes.
Myctophids feed mostly at night, while stom-
iatoids tend to feed by day. My interpretations
indicate that near Hawaii, the differences are at
least partially related to different diel relation-
ships of the fishes to vertical distributions of their
prey. Other interpretations are obviously possible,
e.g., the feeding patterns may prove to be charac-
teristic of the two taxa regardless of relationship
to prey distribution. It would be of particular in-
terest to investigate myctophids with vertical dis-
tribution patterns similar to those of the
stomiatoids, i.e., with shallow day depth ranges at
or near high daytime concentrations of zoo-
plankton. (Certain Myctophym and Diciphus spp.
from Hawaii meet this criterion [Clarke 1973], but
were not captured in sufficient numbers to be in-
cluded in this study.)
The diel feeding patterns of mesopelagic fishes
could well be related to light rather than (or in
addition to) temperature and prey concentration.
No data on diel light changes near Hawaii are
available; however, data of Kampa ( 1970) from a
similar area of clear oceanic water in the North
Atlantic show that during full moon the diel
change in depths of relevant isolumes is of the
order of 300-350 m. Even allowing for considerable
differences in extinction coefficients between
Hawaii and Kampa's study area, the diel change
in isolumes at new moon (when the present sam-
ples were taken) off Hawaii is probably at least
300-350 m and could be as great as 500 m. The
absolute diel change in depth for most of the myc-
tophids is over 500 m while that for the 4 day-
feeding stomiatoids is ca. 400 m or less (Table 3).
Thus it is possible that feeding in both groups
occurs when higher light levels are encoun-
tered— at night for the myctophids and by day for
the stomiatoids.
Estimation of Rates
As mentioned previously, neither feeding rate
nor stomach evacuation rate can be considered
quantitatively without an independent estimate
of the other. Because of the difficulty in keeping
mesopelagic fishes alive for grazing or evacuation
experiments, it will likely be a long time before
independent estimates are available. For a few
species considered here it is, however, possible to
derive "quasi-independent" estimates of evacua-
tion rate given certain plausible assumptions.
These allow, with further assumptions, rough es-
timates of feeding rate and daily ration.
For any period where feeding rate is zero,
changes in stomach fullness are due to evacuation
alone, and, if temperature, pressure, etc., remain
essentially constant during that period, the rate of
evacuation can be assumed to be proportional to
the amount of food in the stomach (Kjelson and
Johnson 1976; Eggers 1977). The change in
stomach fullness would then be described by:
dSldt = -kS orS, = So^*'
(1)
510
CLARKE: DIEL FEEDING PATTERNS OF MESOPELAGIC FISHES
where S is stomach fullness as percentage of fish
weight; So and S,, the values at the beginning and
end of a period oft hours; and/?, the instantaneous
evacuation rate in per hour.
For most of the species considered here, there is
no extended period of decline in stomach fullness
where the above assumptions are met, but a rough
estimate of /? is possible for//, prox^'mum and three
species of stomiatoids. Hygophum proximum ap-
parently ceases feeding early in the night, and
stomach fullness declines from N2 to SR under
essentially constant conditions, i.e., the fish re-
main in the upper layers. Stomach fullness de-
clines from SS to SR in Vinciguerria nimbaria,
Valenciennelliis tripunctulatus, and Danaphos
oculatus, and except for relatively brief periods of
migration in the first two species, they remain at
the same temperature, etc., for this period.
The values of ^ for these four species were calcu-
lated by simply using the integral form of Equa-
tion (1) and the median values of S for the begin-
ning and end of the periods mentioned above
(Table 4). (Other fitting procedures, such as least
square methods, require that a number of ques-
tionable statistical assumptions be made.) The
values of k are inversely correlated with night
depth and thus positively with temperature being
lowest for D. oculatus, highest for Vinciguerria
nimbaria and H. proximum, and intermediate for
Valencienellus tripunctulatus.
For each of the four species, prey concentration
and temperature, pressure, etc., were essentially
constant throughout the period when feeding oc-
curred (SS to N2 for H. proximum and SR to SS for
the stomiatoids). It is not unreasonable to assume,
as a first approximation, that feeding rate was
constant during the periods of increasing stomach
fullness. Changes in fullness would then be de-
scribed by:
dS/dt ^ F - k'S
(2)
where /j ' is the instantaneous evacuation rate dur-
ing the period of feeding, andF is the feeding rate
in percentage bodily weight per hour. Integrating
and rearranging gives an equation forF in terms
of ^', the duration of the feeding period /' in
hours, and median fullness at the beginning (Sq ' )
and end (S, ') of the feeding period:
F =
k' (S/ X S„'e-''-'
1
-*'t'
(3)
(In some cases, there were a few relatively high
values of stomach fullness among the data for a
given period; consequently, the feeding rate of
some individuals may have been lowered due to
satiation. Such values had little effect on the me-
dian, and thus the assumption of constant feeding
rate is probably not seriously violated as long as
medians are used in the calculations.)
Estimates of feeding rate and daily ration ( =
Ft') were calculated (Table 4) using median values
of stomach fullness at SR and SS as So' and S,',
respectively, for the stomiatoids and, similarly, SS
and N2 for H. proximum. Since both D. oculatus
and H. proximum feed at the same temperatures
as those under which the instantaneous evacua-
tion rates were estimated above, /? ' in Equation (3)
was assumed equal to/? calculated from Equation
( 1 ). The daytime or "feeding" temperatures of Vin-
ciguerria nimbaria and Valenciennellus tripunc-
tulatus are lower than those under which k was
estimated from Equation (1). During the day both
species occur at nearly the same temperature as
does D. oculatus both day and night. Con-
sequently, for each of the two migrating
stomiatoids, two values of feeding rate and daily
ration are given in Table 4 — one calculated using
Table 4. — Estimates of instantaneous stomach evacuation rates, feeding rates, and daily rations for four species of mesopelagic fishes
based on changes in median stomach fullness over the dial cycle. The first three columns give the sampling periods (Table 1 ) between
which feeding rate was assumed to be zero, the duration of this interval it), and the calculated instantaneous stomach evacuation rate
ik). The last five columns give the sampling periods between which feeding rate was assumed constant and positive, the duration of this
interval (f), the instantaneous stomach evacuation rate assumed for the feeding periods (k '), and calculated feeding rate (F in % of
bodily weight per hour) and daily ration (/? = Ft' in '7c of bodily weight per day). For both Valenciennellus tripunctulatus and
Vinciguerria nimbaria, two values of /e ', F, and R are given: the higher values under the assumption of constant stomach evacuation
rate night and day ik' = k), the lower under the assumption that stomach evacuation rate during the feeding period was lower and equal
to that estimated for the nonmigrating, deep-living, Danaphos oculatus. See text for formulae and further explanation.
Species
Nonfeeding period
t (h)
k (h-')
Feeding period
r (h)
k' (h-
F(%h)
,;d)
Hygophum proximum
N2-SR
6.8
-0 52
SS-N2
45
-0.52
1.26
5.7
Danaphos oculatus
SS-SR
11.3
-0.10
SR-SS
127
-0.10
015
1.9
Valenciennellus
SS-N4
8.7
-0.22
SR-SS
127
-0.22
0.57
7.3
tnpunctulatus
-0.10
0.34
4.3
Vinciguerria
SS-SR
11.3
-0.38
SR-SS
127
-0.38
1.42
18.1
nimbaria
-0.10
0.51
6.5
511
FISHERY BULLETIN: VOL. 76, NO. 3
k' = k from Equation (1) and the other using k ' =
0.10, the value for D. oculatus.
The estimated ration for Vinciguerria nimbaria
seems inordinately high (18%) if/?' is assumed
equal to k, the nighttime estimate of evacuation
rate. Such values have been estimated for very
young, rapidly growing zooplanktonivorous
fishes, e.g., A/osa aestivalis (Burbidge 1974) and
Oncorhynchus gorbuscha (Parsons and LeBras-
seur 1970). Data from Kjelson and Johnson's
1 1976) study of postlarval Lagodon rhomboides
and Leiostomus xanthurus feeding rates on zoo-
plankton yield estimates of daily ration of only 9.4
and 8.69c , respectively, in terms of wet weight ( my
calculations from their data). The estimated ra-
tion for V. nimbaria using the low value for /? ' and
that for//, proximum lie within the range of val-
ues observed for larger individuals in the first two
studies cited above and for Morone chrysops
juveniles feeding on zooplankton (Wissing 1974).
The daily ration of the California sardine, Sar-
dinops caerulea, which is a larger zooplanktoni-
vore, is apparently slightly lower; judged from
Lasker's (1970) estimates of metabolic and grovvd:h
requirements, the daily ration is probably about
3-4% ( in terms of calories) for the sizes considered.
The above comparisons are admittedly
stretched and ignore, among other things, possible
differences due to environmental temperature,
but the similarity of estimated daily rations of//.
proximum and V. nimbaria to those of shallow-
living planktonivores is not entirely unexpected.
Childress and Nygaard (1973) have shown that
the chemical composition of mesopelagic fishes
which migrate to the upper layers at night is more
similar to that of epipelagic species than to non-
migrating, deep-living forms.
Differences between estimates for the three
stomiatoids are correlated with the extent of verti-
cal migration. Nighttime stomach evacuation rate
is highest in V. nimbaria, lowest in D. oculatus,
and intermediate in Valenciennellus
tripunctulatus. Feeding rate and daily ration es-
timates show the same trend regardless of
whether or not daytime stomach evacuation rates
are assumed lower. The absolute values of
stomach fullness at the end of the feeding period
(Figure 3) are also highest in Vinciguerria nim-
baria and lowest in D. oculatus. These trends in-
dicate a possible adaptive value for the upward
migrations of some stomiatoids. The higher tem-
peratures encountered at night by migrators could
allow processing of larger meals and presumably
faster growth, turnover, etc., rates than for species
which remain at depth day and night.
ACKNOWLEDGMENTS
I thank the captain and crew of the RV Teritu
for their cooperation on their vessel's final cruise;
the many people who assisted in collection of the
samples; P. J. Wagner and G. L. Hoff for careful
and competent sorting of the fishes and dry weight
determinations, respectively; and K. Gopala-
krishnan for identification of euphausiids and de-
capods. This research was supported by NSF GA-
38423 and the State of Hawaii, Hawaii Institute of
Marine Biology.
LITERATURE CITED
Anderson, R.
1967. Feeding chronology in two deep-sea fishes off Cal-
ifornia. M.S. thesis, Univ. South. Calif., Los Ang., 11 p.
Baird, R. C, T. L. Hopkins, and D. F. Wilson.
1975. Diet and feeding chronology of Diaphus taaningi
(Myctophidae) in the Cariaco Trench. Copeia
1975:356-365.
BRINTON, E,
1967. Vertical migration and avoidance capability of
euphausiids in the California Current. Limnol.
Oceanogr. 12:451-483.
BURBIDGE, R. G.
1974. Distribution, growth, selective feeding, and energy
transformations of young-of-the-year blueback herring,
Alosa aestivalis (Mitchill), in the James River, Virgin-
ia. Trans. Am. Fish. Soc. 103:297-311.
CHILDRESS, J. J.
1975. The respiratory rates of midwater crustaceans as a
function of depth of occurrence and relation to the oxygen
minimum layer off Southern California. Comp.
Biochem. Physiol. 50A:787-799.
Childress, J. J., and M. H. Nygaard.
1973. The chemical composition of midwater fishes as a
function of depth of occurrence off southern
California. Deep-Sea Res. 20:1093-1109.
Clarke, T. A.
1973. Some aspects of the ecology of lantemfishes (Myc-
tophidae) in the Pacific Ocean near Hawaii. Fish. Bull.,
U.S. 71:401-434.
1974. Some aspects of the ecology of stomiatoid fishes in
the Pacific Ocean near Hawaii. Fish. Bull., U.S.
72:337-351.
Clarke, T. A., and P. J. Wagner.
1976. Vertical distribution and other aspects of the ecology
of certain mesopelagic fishes taken near Hawaii. Fish.
Bull., U.S. 74:635-645.
DeWITT, F. a., JR , AND G. M. CaILLIET.
1972. Feeding habits of two bristlemouth fishes, Cv-
clothone acduudens and C. signata (Gonostomatidae).
Copeia 1972:868-871.
EGGERS, D. M.
1977. Factors in interpreting data obtained by diel sam-
512
CLARKE: DIEL FEEDING PATTERNS OF MESOPELAGIC FISHES
pling of fish stomachs. J. Fish. Res. Board Can. 34:290-
294.
Gannon, J. E.
1976. The effects of differential digestion rates of zoo-
plankton by alewife, Alosa pseudoharengus. on determi-
nations of selective feeding. Trans. Am. Fish. Soc.
105:89-95.
GJOSAETER, J.
1973. The food of the myctophid fish, Benthosema glaciale
(Reinhardt), from western Norway. Sarsia 52:53-58.
Gordon, D. C.
1970. Chemical and biological observations at Sta. Gol-
lum, an oceanic station near Hawaii, January 1969 to
June 1970. Rep. Hawaii Inst. Geophys., Univ. Hawaii,
H.I.G.-70-22,44 p.
HOLTON, A. A.
1969. Feeding behavior of a vertically migrating lan-
temfish. Pac. Sci. 23:325-331.
HOPKINS, T. L., AND R. C. BAIRD.
1975. Net feeding in mesopelagic fishes. Fish. Bull., U.S.
73:908-914.
1977. Aspects of the feeding ecology of oceanic midwater
fishes. In N. R. Andersen and B. J. Zahuranec (editors).
Oceanic sound scattering prediction, p. 325-360. Plenum,
N.Y.
KAMPA, E. M.
1970. Underwater daylight and moonlight measurements
in the eastern North Atlantic. J. Mar. Biol. Assoc. U. K.
50:397-420.
KJELSON, M. a., and G. N. JOHNSON.
1976. Further observations of the feeding ecology of post-
larval pinfish, Lagodon rhomboides, and spot, Leiostomus
xanthurus. Fish. Bull., U.S. 74:423-432.
LASKER, R.
1970. Utilization of zooplankton energy by a Pacific sar-
dine population in the California current. In J. H. Steele
(editor). Marine food chains, p. 265-284. Univ. Calif.
Press, Berkeley.
Legand, M., p. Bourret, p. FOURMANOIR, R. GRANDPERRIN.
J. A. GUEREDRAT, A. MICHEL, P. RANCUREL, R. REPELIN,
AND C. ROGER.
1972. Relations trophiques et distributions verticales en
milieu pelagique dans I'Ocean Pacifique intertropi-
cal. Cah. O.R.S.T.O.M., ser. Oceanogr. 10:303-393
Marshall, N. B.
I960. Swimbladder structure of deep-sea fishes in relation
to their systematics and biology. Discovery Rep. 31:1-
122.
MAYNARD, S. D., F. V. RIGGS. AND J. F. WALTERS
1975. Mesopelagic micronekton in Hawaiian waters:
Faunal composition, standing stock, and diel vertical
migration. Fish. Bull., U.S. 73:726-736.
Merrett, N. R., AND H. S. J. Roe
1974. Patterns and selectivity in the feeding of certain
mesopelagic fishes. Mar. Biol. (Berl.) 28:115-126.
PANDIAN, T. J.
1967. Transformation of food in the fish Megalops cyp-
rinoides. I. Influence of the quality of food. Mar. Biol.
(Berl.) 1:60-64.
Parsons, T. R., and R. J. LeBrasseur.
1970. The availability of food to different trophic levels in
the marine food chain. In J. H. Steele (editorl. Marine
food chains, p. 325-343. Univ. Calif. Press, Berkeley.
Roe, H. S, J.
1972. The vertical distributions and diurnal migrations of
calanoid copepods collected on the SOND Cruise, 1965. 1.
The total population and general discussion. J. Mar.
Biol. Assoc. U.K. 52:277-314.
Tate, M. w., and R. C. Clelland.
1957. Nonparametric and shortcut statistics in the social,
biological, and medical sciences. Interstate Printers and
Publishers, Danville, 111., 171 p.
TEAL, J. M.
1971. Pressure effects on the respiration of vertically mi-
grating decapod Crustacea. Am. Zool. 11:571-576.
Tyler, H. R., Jr.. and W. G. Pearcy.
1975. The feeding habits of three species of lanternfishes
(family Myctophidae) off Oregon, USA. Mar. Biol.
(Berl.) 32:7-11.
WISSING, T. E.
1974. Energy transformations by young-of-the-year white
hass Moronechrysops (Rafinesque) in Lake Mendota, Wis-
consin. Trans. Am. Fish. Soc. 103:32-37.
513
ON THE RESTRUCTURING OF THE PELLA-TOMLINSON SYSTEM
R. Ian Fletcher*
ABSTRACT
The time-dependent analysis of an earlier work is extended to the equilibrium case of the Pella-
Tomlinson system, and the relationships between the equilibrium and nonequilibrium versions of the
restructured system are developed. The dual formulations of the conventional analysis are avoided and
maximum sustainable yield is separated from the indeterminacy of the system. All arbitrary
coefficients are eliminated and the management components incorporated directly into the system
equations. The source of the statistical degeneracy in the model is revealed and explicitly formulated,
and in the companion article by D. Rivard and L. J. Bledsoe (this issue of the Fishery Bulletin) the
restructured model is treated by a new statistical method that subdues the estimation failures
associated with past treatments of the Pella-Tomlinson system.
Because the equilibrium versions of all stock-
production models follow from steady-state inte-
grations, the strategy of fishery regulation be-
comes a strategy of accommodation, so to speak, as
determined by a pattern of balanced model states
where removals just equal the productivities
otherwise surplus to the maintenance needs of the
stock. Population status usually enters the process
in the simple, robust form of integrated numbers
or biomass, and the removals of fishing constitute
direct fractions of the whole fishable stock without
reference to age or weight distributions. Since the
appearance of Schaefer s work (Schaefer 1954) the
strategy has been applied to the management of
many fisheries. Schaefer devised a rational,
linearized method for estimating the parameters
of Graham's equilibrium model (Graham 1935)
from the actual nonequilibrium yields and effort
expenditures of a fishery, a contribution that is
often misunderstood. In applying Schaefer's
method or like schemes of synthesis, it is not so
much that one hopes to observe a pattern of
equilibrium levels in a fishery or even expects
them to come about, but rather, by knowing the
response history of a stock to various exploitation
pressures, one might then be guided by the model
in bringing a stock, through a sequence of man-
agement actions, into a state where some desired
level of sustainable yield most likely abides. The
philosophy is widely accepted in fisheries man-
agement but its application is often censured.
'Center for Quantitative Science in Forestry, Fisheries and
Wildlife, University of Washington, Seattle, WA 98195.
either on economic or biological grounds (see, for
example, Larkin 1977)
The exploitation model of Pella and Tomlinson
( 1969), as it is customarily thought of, extends the
more "basic" model of Graham from a system of
second degree in nonlinearity to a flexible or more
"general" system of indeterminate degree. The in-
creased flexibility comes into the Pella-Tomlinson
model through the addition of a single exponential
parameter, but the analytical peculiarities that
accompany the improvement often lead to
paradoxical ends since the equations of the system
then permit the simultaneous generation of good
data fits and poor parameter estimates (see the
commentary of Ricker 1975:323-326 and the
treatments of Fox 1971, 1975). This disturbing
trait of the statistical model arises from the
conflict between the variable (or parametric) cur-
vature of the analytical model and the coupling of
that curvature, in the conventional formulations,
with all the coefficients of the system. As shown in
a prior work (Fletcher 1978), those effects may be
separated in the time-dependent analysis by re-
structuring the system equations so as to accom-
modate directly the critical-point coordinates of
the system graphs. In this paper we extend the
analysis to the equilibrium version of the Pella-
Tomlinson system, and we show the relationships
between the equilibrium model and the (restruc-
tured) time-dependent equations.
For a stock of mixed age classes, the most
difficult problem in applying any equilibrium
model will lie, essentially, in the interpretation of
time-dependent transitions between idealized
states (however momentary, long-enduring, or
Manuscript accepted March 1978.
FISHERY BULLETIN: VOL. 76. NO 3. 1978.
515
FISHERY BULLETIN: VOL. 76, NO. 3
unobserved such states may be), since the stock
will include simultaneously the young and the old,
the older having accumulated a probabilistic his-
tory of mortality, fecundity, and growth which
may differ considerably from the current schedule
that affects both. Various tactics for adjusting the
parametric mechanics of stock-production models
to such long-term, delayed influences are given by
Gulland (1969), Fox (1975), Walter,^ and others,
but in the case of the Pella-Tomlinson system the
difficulties have been compounded by artifacts of
the conventional analysis and by an instability
inherent to the mathematical indeterminacy of
the system itself. With the critical-point analysis,
most of those impediments will convert to tracta-
ble relationships or vanish altogether. We can
suppress the troublesome dual formulations as-
sociated with the conventional casting of the sys-
tem, we can uncouple the indeterminate exponent
and the coefficients of the governing equations,
and we can make explicit the relationships be-
tween parametric graph curvature and the man-
agement components of the system.
THE REFORMULATED
GOVERNING EQUATIONS
Stock-production models, as they are usually
defined, arise from the common premise that a fish
stock, when reduced by exploitation to a level
below some prior abundance, will always strive to
recover its former size in accord with some latent,
self-regulating mechanism of restoration. Irre-
spective of the compensatory details, any such re-
covery must accrue directly from the productivity
of the stock, and in the conventional representa-
tion of the Pella-Tomlinson system, the latent
capacity for biomass production in a stock of fishes
is given the dual formulation
P(B) = ±aB" + b B.
la)
lb)
•
P(B) being the production rate of the stock at stock
sizeB. Equation ( la) applies when exponents falls
on the range 01. In either case, all the critical compo-
nents of the system — maximum stock size,
maximum productivity, the stock level where
maximum productivity occurs — depend in some
^Walter, G. G. 1975. Non-equilibrium regulation of fisheries.
Int. Comm. Northwest Atl. Fish. Res. Doc. 75/IX/131, 12 p.
516
way on the numerical value assigned to exponent
n. That is, root B^c is given by
lll—n
B
the critical ordinate p (which corresponds to the
stock level where maximum productivity occurs)
is determined by
P =
1/1—??
while extremum coordinate m (which corresponds
to productivity P ma.x ' must be determined from
the formula
m
n \b )
the plus sign applying to Equation (la) and the
minus sign to Equation (lb).
Although exponent n controls the graph curva-
tures of Equations ( la) and ( lb), the nonzero roots
and extrema are controlled hy By- and the coordi-
nate pair (p, m). As shown by Fletcher (1975),
coordinate m has no essential dependence on ex-
ponent n, and with the appropriate transforma-
tions the dual formulation (Equations ( la, b)) may
be suppressed. In consequence, either of the
parametric sets {m, p,Bx}or{m,/7.5x}will consti-
tute a complete set of independent governing
parameters for latent productivity in the Pella-
Tomlinson system, and the dual formulation
(Equations (la, b)) converts to the single differen-
tial equation for latent productivity
P = ym
it) -HI)"- <^»
with y a purely numerical factor wholly prescribed
by n as
,n/n — l
y
1 ■
(3)
With the coefficients so cast, the sign reversals at
turning point /; = 1 become automatic, and the
consolidated interval of definition for n becomes
00 and the stock increases. Should
• • •
Y = P, then B = Q and biomass trajectory Bit)
exhibits an extremum, which is the necessary
condition for equilibrium fishing. Yield rate Y cus-
tomarily takes the form
With initial time /„ set at zero, the integration
constant C in Equation (8) becomes
^0^
- B.
The quantity B ;, , when positive in Equation (8),
becomestheadjustment level such thatBf^^ -► B*
over time. When, for certain ranges of n and F,
quantity B*<0, then the zero root of Equation (7)
applies and B(t) -► 0. When mortality F takes
the value
MSY
■(^)
■ym
bZ
(9)
irrespective of the value of parameter n, then
B(t) -►p and Y ->m (which are the conditions, in
the equilibrium limit, for maximum sustainable
yield). In terms of the parameter set { m, p, B^^ },
Equation (9) becomes, simply,
F.
MSY
m
P
Figure 1 gives a summary of the general con-
straints on the time-dependent system; for a more
detailed treatment of system behavior, see Flet-
cher (1978).
THE RESTRUCTURED
EQUILIBRIUM SYSTEM
Yit) = F{t) ' B{t)
(6)
with the assumption that all fish of the fishable
stock share equal probabilities of capture. By ad-
mitting Equations (2) and (6) into Equation (5),
the differential equation that governs net produc-
tivity in the restructured system becomes
B = 7m
B
B
ym
©"
FB (!)
and over any time interval, however brief, that
mortality F might be presumed to have a fixed
value, biomass variable B in Equations (6) and (7)
has the general time-dependent solution
B(0
= (b,i-" + Cexp ((7m/B„
* [ym-FB^)
By Equations (2) and (5), the time-varying rate
of yield in the reformulated Pella-Tomlinson sys-
tem takes the form
y = ,„ (y _ ,„ (y _ s. ,10,
and when, for given F and n, governing Equation
(7) exhibits a positive root, then B(t) -► B^ and
B -► 0 in Equation (10), and yield rate Y, over
sufficient time, approaches a constant value. In
the steady-state tor "equilibrium") limit, yield
then accumulates as
F)(l
-n)t)y"'-''\
(8)
1 /(!-«)
B.
517
FISHERY BULLETIN: VOL. 76. NO. 3
00
Y — Fb^
A
6,
5-
n-i\
r>p
MSV
n >-/'
E>oo
Y/77
' MSV
Figure l. — Time-dependent response of the Pella-Tomlinson system to parametric variations of exponent n and mortality F. The
upper diagram summarizes system response when n falls on the range 0 < /! < 1 . The adjustment level of biomass is never zero for this
range of n however great the value ofF, and mortality F ms-, has no absolute constraints; such a stock cannot be fished to extinction.
The lower diagram summarizes system behavior when n falls on the range n > I. Mortality F ^sv 's then constrained to the interval
indicated by the diagram. WhenF exceeds the critical value ym/By^, then the stock, over sufficient time, trends to extinction.
f
(lY =
ym
and for any such equilibrium interval t, the inte-
grated yield rate iY^h) takes on the parametric
formulation
= 7m
(11)
with maximum latent productivity m of the time-
dependent system becoming the maximum sus-
tainable yield rate (the MSY) of the equilibrium
system. With B^ as the parametric variable in
Equation (11), a zero left endpoint exists for Y^.Ij
when /? > 1 and F = ymlB^. Should F exceed the
critical value ymlBy- when exponent n >1, no
equilibrium state exists; such conditions in the
518
time-dependent system correspond to extinction
trends. But when n has any value on the range
00 no matter how great the value
of F. That is, when 01,
however, the corresponding stock can have non-
zero adjustment levels B^ only whenF1 and when fishing mortality ex-
ceeds the critical value ym/Bx, the "adjustment"
level corresponds to extinction and Equation (12)
does not apply.
Upon the substitution of Equation (12) into
Equation (11), the direct relationship between
equilibrium yield and equilibrium fishing mortal-
ity becomes
1
T V ym J
B.
(13)
and the fishing mortality that maximizes Equa-
tion ( 13) is given by Equation (9). That is, with the
substitution of F^gy into Equation (13) then
Y*/t = m.
Under the equilibrium conditions, the conven-
tional quantity U (which signifies accumulated
catch per unit of fishing effort as a function of
fishing intensity f/r) can be cast into the restruc-
tured form
\ °° jm T I
(14)
which eliminates the explicit appearance of catch-
ability coefficient g, permitting instead the direct
quantification of maximum sustainable yield m.
Quantities U and J7x have the customary mean-
ings
Y
U = —f' ^Y* being the yield accumulated over
time interval r as a consequence of ef-
fort f).
U^ ^ qB^ (q being the individual probability of
capture per unit of fishing effort f).
Should the accumulation interval t be a year, the
variable U becomes annual CPUE (catch per unit
of effort) and the variable [It becomes effort per
annum. With exponents >1 in the Pella-Tomlinson
system, no steady-state CPUE exists for a fishing
intensity in excess of critical value ym /U^.. But if n
519
FISHERY BULLETIN: VOL. 76. NO. 3
has any value on the interval 0 Inionmcubion itouo
> PlOQiam contAol ilow
Vautjo. -input
Vzcla-ion aZgoiyCthm
Calculation algo-
lithm
an analytical form for the uncertainties in the
final values of the parameters (Bevington 1969).
By letting SiO) be the weighted residual sum of
squares for the final parameter estimates, how-
ever, the variance-covariance matrix of the esti-
mates (Bard 1974) can be approximated by
imation in the neighborhood of O is appropriate. A
necessary and sufficient condition for the
F-distribution to be appropriate here is that dif-
ferences in true and estimated parameter values
are independent and approximately normally dis-
tributed with zero mean and equal variance.
Ve =
j^ J
S(e)/(r-5).
(12)
Some idea of the joint variability of the parame-
ters can be obtained by evaluating the ellipsoidal
confidence region, based on the assumption that
the linearized form has validity around B ( Draper
and Smith 1966). The confidence region is then
given by
[e-B] J'^ J[B-B]'
F(5,r-5,l-a),
(13)
where F(5, r-5, l-«) is the standard tabulated
F-statistic. The ellipsoid is not a true confidence
region, of course, since the dependent variable, Y,
is a nonlinear function of B. The intervals ob-
tained are valid to the extent that a linear approx-
DETERMINATION OF
STARTING VALUES
In order to reduce the number of iterations re-
quired to minimize Equation (8), reasonably accu-
rate starting values should be employed. Starting
values can be calculated from a linearization and
simplification of the basic model.
STEP 1. By using y,,y2, .. . , Y, and /■,, /a, . . .
fr, find an estimate of g from the Delury technique.
Note that this procedure generally underesti-
mates (? (see Ricker 1975). Correction forq will be
provided in step 4.
STEP 2. Find estimates of B, + i from the equa-
tion
B
..1 = (^M.l^^..l,..2)/2'
(14)
526
RIVARD and BLEDSOE: PARAMETER ESTIMATION FOR PELLA-TOMLINSON MODEL
where (by assuming ft_t^\ constant over the inter-
val t,t + \)
mates of m and B x- Finally, Bo is approximated by
B
t.t+\ i.t+i'^ 't.t + \
(15)
Note that Y,j + y and /",, ^i correspond to y, and/', of
Equations (6) and (5).
STEP 3. Let /z= 2, as in the Graham-Schaefer
model, and estimate m andB-^ by fitting the linear
model
(16)
where y =
- Tjn—l
t B^dt
+ qf^, x^ = B';-\
Equation (16) is derived from Equations (1) and
(3). However, Equation (16) requires an estimate
of the relative growth rate dB,/Bidt, say i?,. As
suggested by Causton ( 1969), the mean value of 7?
between / and t+2 is given by
^M.2 = (lnB^^,-lnB^)/2.
(17)
For the purpose of fitting Equation ( 16), quantity
/?,,+2 n^^y t>6 considered an estimate ofRi + i,
which corresponds to B, + ^. Whence. Equation
(16) provides estimates of/?? andSx as
m =
h %"]
(l/l-n)
7
B =
ct.
d.
1/1— ^7
(18)
(19)
STEP 4. Steps 2 and 3 are repeated iteratively
for increasing values of q. The value of g which
provides the minimum residual sum of squares
[1 (Y, - y, )2] is accepted as the appropriate start-
ing value for q.
STEP 5. Step 3 is repeated iteratively for in-
creasing values of n, parameter q being kept con-
stant. The value of n which provides the minimum
residual sum of squares [1 (y, - Y,)^] is accepted
as the appropriate starting value for n . In the last
iteration, Equations (18) and (19) provide esti-
^0 = ^0,1
a
(20)
where B, andB,, , are estimated by Equations ( 14)
and (15), respectively.
Steps 1 through 5 provide a set of starting values
for the optimization algorithm (11). Usually the
starting values are near the solution and few iter-
ations will be needed. Of course, it would be possi-
ble to derive algorithms for more accurate starting
values, but our purpose here is to find a rough
estimate for each coefficient and to let the iterative
procedure (11) converge to the minimum. Some-
times, by experience or by prior information, it is
possible to provide starting values as satisfactory
as those provided by the algorithm given above.
MONTE-CARLO SIMULATIONS
The parameter values that we chose to generate
the data of Table 1 (deterministic model) were
recovered exactly by the estimation procedure.
Results of fitting 18 stochastic versions of the de-
terministic model are also included in Table 2.
Based on our simulation results, there do not ap-
pear to be any serious problems with bias of
parameter estimates. The bottom line of Table 2,
which gives the coefficients of variation of the
parameter estimates, reveals that estimates of the
three parameters of principal interest to the man-
ager have the smallest variability. Those
parameters are maximum sustainable yield (m,
C.V. = 147f ), optimal effort level(/MSY> C.V. = 67r ),
and yield per unit of effort at optimum effort
(t/ivjsY' C.V. = 9%). Our results confirm the ob-
servations of Fox (1971) and Pella and Tomlinson
(1969) on the robustness of m and /"msy with re-
spect to error in the measurement of the yield
data. From Table 2, we can also compare variance
estimates from Equation ( 12 » with variance of es-
timates for 10 replicates at o" = 0.200. Equation
(12) appears to give (approximately) unbiased es-
timates of the variance of the sampling distribu-
tion of G. Also, out of the 19 cases considered, the
true parameter value lay outside the arbitrary ±2
(SD) confidence interval twice form and only once
each forB^,/?, andB,,. Although we did not employ
an extensive Monte-Carlo simulation, our results
suggest that the normal approximation to the
sampling distribution of Q is an acceptable ap-
proximation, at least for management purposes.
527
FISHERY BULLETIN: VOL. 76, NO. 3
In a few additional simulations (replicates 13
and 15), parameters obtained by the five-para-
meter procedure were ill determined. A parameter
is considered ill determined if its estimated
value responds unreasonably to seemingly insig-
nificant variations in the data (Bard 1974). The
basic difficulty is that the model is extremely gen-
eral and capable of several types of behavior over
the space of 0. In the Pella-Tomlinson system,
ill determination often occurs whenever an itera-
tion of the algorithm (11) gives an estimate of O
such that the point (m , /msy* o^ ^he yield-effort
plane lies outside the concentration of data. In
such a circumstance the exponent n takes on small-
er and smaller values in the successive iterations
and the solution of system (1) and (3) degenerates
to an exponential form for which only four
parameters are required for uniqueness. That is,
as/? ->0. in Equation (1), then (B/By.)" -►I and y
-► - 1 . The five-parameter procedure then over-
prescribes the system, which in turn predisposes
the coefficient estimates to extremely large var-
iances. The ultimate irony here is the fact that
wholly unrealistic parameter estimates still gen-
erate good fits to the catch-effort history (i.e. small
residuals). For example, in Figure 2 the fitted
five-parameter curve predicts /"jv^gy near infinity
while in the true model /'msy actually corre-
sponds to 174,000 units of effort. However dif-
ferent the equilibrium curves are, the five-
parameter procedure still generates a good fit to
the catch history (S(0) = 1.10). Incompleteness of
information over a wide range of effort values, as
well as excessive noise in the catch-effort data,
will tend to bring about such pathological condi-
tions.
To overcome these difficulties, reformulation of
the estimation problem is necessary. By the fol-
lowing considerations, the five-dimensional
parameter space can be reduced to three dimen-
sions. First, we will approximate £„ by Equation
( 20). Furthermore, if the data contain information
on the yields under low exploitation, we may
define Sx as
B
= MAXiYJq f.)
r.
(21)
By using Equations (20) and ( 21 ),fi, I andfixcanbe
deleted from G, leaving only m, q, and n as the
independent parameters requiring estimation. It
is important to understand at this point that B^
and fix are not fixed; they are reevaluated by
Equations (20) and (21) at each iteration, along
with the parameters tu ,q, and n . In fact, the solu-
tion of Equations ( 1) and (3), as well as Equations
(20) and (21), specify a new model with unknowns
O =[m,q, nY. By this restructuring, much of the
degeneracy associated with the model can be
eliminated. As shown in Figure 2, this procedure
also provides a closer correspondence between the
"estimated" and the "true" equilibrium model.
Furthermore, the three-parameter procedure still
generates an adequate nonequilibrium catch his-
tory (S(G) = 1.40). In a Monte-Carlo simulation
study, parameter estimates obtained by using
these transformations fell within reasonable
FlOURE 2. — Comparison of the "true"
model with the models obtained by
using the estimation procedure on three
and five parameters respectively- Solid
lines show equilibrium yield curves;
data points show nonequilibrium simu-
lated (dots) yields and predicted (circles)
yield values from the three-parameter
approach. Dashed vertical lines indi-
cate the magnitude of residuals.
UJ
528
RIVARD and BLEDSOE: PARAMETER ESTIMATION FOR PELLA-TOMLINSON MODEL
hounds (Table 3). Out of the 20 cases considered,
the true parameter vahie lay outside the arbitrary
±2 (SDl confidence interval only once tor ru and/?.
Also, variance estimates were comparable with
the variance estimates of the five-parameter pro-
cedure (compare Tables 2 and 3).
Table 2. — Estimated parameters for the deterministic model and for 18 stochastic replicates. The
Levenberg-Marquardt algorithm is employed in a five-dimensiona! parameter space i m ,By.,n,q,Bg). For each
parameter and replicate, the parameter estimate ± its estimated standard deviation from Equation ( 12) are
tabulated. Replicates 13 and 15 have been e.xcluded due to degeneracy of the model, as discussed in the text.
Repli-
m
Sx
q
Bo
'msv
cate
(f
= S (6) (r-3).
^For 12 replicates with -r = 0 200
^Overall standard deviation of parameter estimates for 12 replicates with u = 0.200.
529
FISHERY BULLETIN: VOL. 76. NO. 3
CASE STUDIES
We applied the three-parameter method to the
catch-efibrt data of the yellowfin tuna fishery of
the eastern tropical Pacific, 1934 through 1967
[the same data that were analyzed by Pella and
Tomlinson ( 1969) and by Fox ( 1971)]. Table 4 gives
a comparison of results, and our final equilibrium
model is shown by Figure 3. As indicated by Table
4, the parameter estimates of the Levenberg-
Marquardt method are comparable with the esti-
mates that Fox obtained with his search al-
gorithm. Pella and Tomlinson also employed a
searching algorithm but their minimization
criterion was an unweighted least-squares func-
tion. Our standard deviation estimate is very
small for w( MSY) but relatively large f or B^,n,q,
and fi„, which is a consequence of insufficient in-
formation in the yellowfin tuna data on yield at
high fishing rates. With such limited information,
one can anticipate that neither the shape nor the
location of the descending portion of the equilib-
rium curve (dashed in Figure 3) could be deter-
mined with much accuracy, and the large variance
estimates on the system coefficients reflect this
situation. Of course, the variance estimates for
/^j^Y ^^^ ^MSY ^^^ always be calculated by the
delta method, and to avoid the complex deriva-
tions that accompany the presence of covariance
terms, an alternative would be to define a new
parameter space so as to estimate /"msy oi' ^^msy
directly. The variance-covariance matrix for the
coefficients would then provide the desired infor-
mation on the variability of those parameters.
Our final example is based on the data from the
Pacific halibut fishery in International Pacific
Halibut Commission Area 2, as given in Ricker
(1975, table 13.1). To analyze these data, Ricker
derived an estimate of c/ from the age composition
of the catch. Then he obtained parameter esti-
mates for a Graham-Schaefer model by regressing
Ye/B against B and Yg/f against / (Ricker 1975,
examples 13.5 and 13.6). In both cases, Ricker
employed GM and Nair-Bartlett regression. The
results Ricker obtained by fitting the Graham-
Schaefer model were compared with the results we
obtained from fitting the generalized stock pro-
duction model by our three-parameter version of
the Levenberg-Marquardt method (Table 5). The
latter provided estimates of m, q, and n with rela-
tively small variance estimates. Furthermore, the
estimate of n appears to be significantly different
from 2.00, which validates the use of the Pella-
Tomlinson model. Nevertheless, estimates of m
are not significantly affected by the choice of the
wrong model, while estimates of /msy are slightly
Table 4. — Comparison of parameter estimates obtained by Pella and Tomlinson (1969), by Fox ( 1971) and by the
Levenberg-Marquardt algorithm for the yellowfin tuna in eastern Pacific Ocean. Values that follow the ± signs
are the standard-deviation estimates for each parameter.
Bx
q
8o
m
Reference
P
(10«)
n
(10--)
(108)
(108)
'MSV
^MSV
Residuals
Pella and Tomlinson (1969,
table 5)
—
—
1 40
450
—
1 826
35.300
5,173
1 78 ■ 10^8
Fox (1971. table 4)
—
1 427
2 10
8 10
1 206
1 926
32.700
5.890
0,736
Levenberg-Marquardt
—
1 448
208
8 01
1 192
1.924
32.700
5.884
0735
algorithm
±0 890
±0 75
±4 9
±1 24
±090
Levenberg-Marquardt
algorithm
0 27
1 274
2 30
9 08
1 079
1 962
32.170
6.097
0641
(correlated error)
-025
*0 653
±0 55
±4 7
±0553
±0 106
Table 5.— Comparison between the estimates of Ricker ( 1975) for the Pacific halibut ( Interna-
tional Pacific Halibut Commission Area 2) and those obtained by the Levenberg-Marquardt
algorithm. Values that follow the ± signs are the standard-deviation estimates for each
parameter.
B ,
q
m
Reference
P
MO'')
n
(10-')
(10^)
'msv
^MSY
Ricker (1975. example 13 5)
GM regression
—
204
2.00
907
31 2
337
926
Nair-Bartlett regression
—
195
2 00
9 07
31 0
3 50
88 6
Ricker (1975. example 13 6)
GM regression
—
256
2 00
9 07
33 0
2 84
1162
Nair-Baniett regression
—
239
2 00
9 07
31 8
294
108 2
Levenberg-Marquardt algorithm
—
187
1 28
1445
31 6
2 83
1117
±18
±0 09
±1 36
±083
Levenberg-Marquardt algorithm
0 33
188
1 28
14 33
31 8
2 84
1120
(correlated error)
-0,16
±22
±0 12
±1 67
±10
530
RIVARn and RLEDSOE PARAMETER ESTIMATION FOR PELLA-TOMLINSON MODEL
2 5 0
2 0 0-
1 5 0
Q 10 0-
111
>-
5 0
Figure 3. — Equilibrium stock produc-
tion model for yellowfin tuna data, from
1934 through 1967, as determined by
the Levenberg-Marquardt algorithm.
EFFORT (ilO boil-liri)
>
overestimated in most applications of the
Graham-Schaefer model (note also that Ricker's
treatment assumes equilibrium). In contrast to
the yellowfin tuna data, analysis of the equilib-
rium model for halibut data indicates that fishing
effort has been concentrated slightly to the right of
/msy (compare Figures 3 and 4).
In the preceding case studies, the Levenberg-
Marquardt algorithm gave estimates with rela-
tively small coefficients of variation. In both cases,
Figure 4. — Equilibrium stock produc-
tion models for Pacific halibut (Interna-
tional Pacific Halibit Commission Area
2), from 1910 through 1957, as deter-
mined by Ricker (1975, examples 13.5
and 13.6, Nair-Bartlett regressions)
and by the Levenberg-Marquardt al-
gorithm (three-parameter version).
10 11
EFFORT (ilO ikitll)
531
FISHERY BULLETIN: VOL 76, NO 3
e.g.. the coefficients of variation for the estimates
of maximum sustainable yield {rh ) were below G^c .
It is questionable, however, whether the data can
justify such precision. Variability of the exploited
population due to migration, to changes in fishery
regulations over time, and to expansion of the
fishing areas, as well as variability of q due to
learning by fishermen and to technological de-
velopments, are important factors underlying the
complexity of events influencing the serial catch-
effort information. In future research, alternative
forms of the model in which g is a variable
parameterized with respect to time will be
explored. Furthermore, in a randomly fluctuating
environment, equilibrium population levels (and
MSY, by extension) are not constant and the
equilibrium points are instead described by a
probabilistic cloud representing the equilibrium
probability distribution (May 1974). The knowl-
edge of this equilibrium probability distribution
would give us some idea of the probability of
achieving the desired management goal ( MSY, for
instance).
W =
(r-iy
Y.Y^P
Y,Y,P y,yp
y.' : Kyy-''
YYp
1—1 f-<
(r-l)
Y.Y^p
(r-2)
Parameter p, constrained between 0 and 1, is a
measure of the importance of lags and can be esti-
mated along with the parameters of the differen-
tial equations (1) and (3). Itcanbeseen that Equa-
tion (9) is a particular case of Equation (22), where
the off-diagonal elements of W are null.
The Levenberg-Marquardt algorithm, as formu-
lated in Equation (11), is designed to minimize
directly a sum of squares of residuals as given by
Equation (9). In order to minimize Equation (22)
by using Equation (11), we must scale W by the
transformation
DISCUSSION ON
ERROR STRUCTURE
In the preceding examples, we found runs in the
time sequence plot of residuals. Those runs indi-
cate correlations among the residuals. Serial cor-
relation, as we usually find in applying production
models to catch data, indicates that the real sys-
tem is working differently than the presupposed
model and that some minor effects have been neg-
lected (such as age composition or environmental
factors). But as indicated by Draper and Smith
(1966), the effects of correlation can be ignored
when the ratio {r - p )/r tends to unity (p being the
number of estimated parameters). In certain situ-
ations, of course, this ratio is likely to become
small (tending to zero as r approaches p) and we
may want to consider weights (W,) which account
for both the inequality of variance and the correla-
tions. In our estimation procedure, the assumption
of uncorrelated error can be relaxed by writing
Equation (9) in the more general form (J. J. Pella,
pers. commun.)
X = D~^ W D
(23)
SiO,p) = [Y-Y] W-i [Y-Yl'
(22)
where Y is the row vector of observed yields, Y is
the row vector of predicted yields, and W is the
symmetric, positive definite matrix
532
where D is a diagonal matrix having elements D,
1 r), and write W as
y, (I
W = D U A U'T D.
(24)
where U.\U^ is the eigenvalue and eigenvector
decomposition ofX. Note thatX is actually the
correlation matrix of errors. Therefore Equation
(22) becomes
sie,p)
= [Y-Y] D-i U A~i U'^ D~^ [Y
Y]''. (25)
Then Equation (25) has the same form as Equation
(9), where the weights iW,) are the square roots of
the eigenvalues of X and where the residuals are
given by [Y— Y J D ' U. Such a procedure requires,
however, diagnoalization of an r by r matrix.
Moreover, diagonalization must be repeated at
least p times for each iteration. This procedure
produces a 10-fold increase in computing time.
Although an exhaustive study of all possible
stochastic effects on the model was not attempted,
some simulations were done to determine the
magnitude of error in parameter estimates due to
serial correlations of the e,. Results are given on
Tables 4 and 5. For the yellowfin tuna data, p =
RIVARD and BLEDSOE: PARAMETER ESTIMATION FOR PELLATOMLINSON MODEL
0.27. For the Pacific halibut data, p = 0.33. In
either case, p exhibits a relatively large coefficient
of variation when compared with the elements of
O. One could anticipate such results since p
reflects the "persistence" of fluctuations in popula-
tion size, and the estimation of p would therefore
require a longer catch history in order to achieve a
greater precision. But more importantly, the val-
ues of G and Var[OJ were not significantly altered
by the inclusion of the additional parameter. And
while the errors of any particular catch history
might indeed by correlated, the minimization
criterion (9) will provide satisfactory estimates of
B despite the fact that correlations do not enter
into its formulation. The limited results contained
herein suggest that serial correlation can be safely
ignored when the ratio (r - p)/r is near unity.
Under such a condition the estimation procedure
is robust with respect to the assumption of inde-
pendence of errors in actual data.
CONCLUSION
modifications of the model to incorporate such
hypothetical effects as migration or stock interac-
tions can be made easily. Of course, to the extent
that the esti mation procedure must rely strictly on
catch-effort data, it will be subject to the same
information uncertainties as any other method.
But within the basic estimation procedure, we can
combine the catch-effort data with prior informa-
tion and thereby reduce the uncertainties in our
estimates. The prior information can be any in-
formation on a state variable, s\xch.asB(t), or even
any prior knowledge of the coefficients as express-
ed by B ± Var(B). Suppose, for example, that we
have information from independent surveys on
stock density (acoustic surveys, indirect estima-
tion from knowledge of larval densities, or even
virtual population analysis from catch records).
Such surveys would then provide us with esti-
mates B^ each having a variance ViB, ), let us say,
at various times t. We can easily introduce such
information into the estimation procedure by
defining the new objective function
The purpose of this paper has been to examine a
version of the Levenberg-Marquardt algorithm as
an alternative method for estimating the
coefficients of the generalized stock production
model. The parameter values obtained by this pro-
cedure are close to those obtained by previous
studies on yellowfin tuna and Pacific halibut. Ob-
viously, data requirements are such that a full
range of effort values (ranging over low and high
exploitation rates) are necessary to insure con-
vergence in the estimation procedure and to pro-
duce estimates with small variability. Our simu-
lations reveal that with the Levenberg-Marquardt
method both the estimates of coefficients and the
estimates of variances remain approximately un-
biased when white noise is considered. If present,
such bias is sufficiently small as to be obscured in
the variability associated with catch error. The
simulations also showed the range of variability in
parameter estimates that might be expected for
given levels of normally distributed error in catch
data.
Because the parameters of interest appear
explicitly in the system equations, the estimation
procedure for the parameters also produces the
variance estimates directly. Moreover, the method
has a reliability and an efficiency of computation
somewhat greater than previous methods. And
since the estimation procedure relies on a numeri-
cally integrated system of differential equations.
s(B) = s w.iY.-yy + s
ill I j
V -l^D _DX2
{B.-B.y. (26)
Introduction of the second term in the objective
function constrains the optimization and thereby
improves convergence. If the prior information
has extremely large variance, then this informa-
tion is of no value; the second term of Equation (26)
will tend to zero and the objective function then
reduces to Equation (9). In general, the alteration
permits the simultaneous employment of the two
state variables. Therefore, the final coefficients
are no longer based solely on catch and effort data;
their determination includes our knowledge of
previous stock densities.
As observed here in a statistical setting, and
by Fletcher (1978a, b) in the exact analysis,
the Pella-Tomlinson system exhibits internal in-
stability in its parametric relationships. That
behavior arises from the variable nature of the
system's nonlinear! ty, which would not be particu-
larly detrimental if our problems were limited
strictly to the geometric syntheses of data by curve
fitting. But for the purposes of management and
preservation of stocks, the subject is elevated
partly at least to the status of parameter estima-
tion "where we look for procedures to obtain val-
ues of the parameters that not only fit the data
well, but also come on the average fairly close to
the true value" (Bard 1974). Although the Pella-
533
FISHERY BULLETIN; VOL. 76, NO. 3
Tomlinson system exhibits a convenient flexibil-
ity with a minimum number of coefficients, the
peculiar coupling of the coefficients to the non-
linearity of the system often provides more flexi-
bility than we care to have, and a conventional
least-squares statistic may not be sufficient to con-
trol the system in the estimation procedure. In
consequence, many constraints have to be imposed
on the system in order to obtain convergence in the
estimation procedure and to insure reliability in
the coefficient values thus estimated.
ACKNOWLEDGMENTS
We would like to thank R. I. Fletcher, J. J. Pella,
and C. G. Walters, each of whom reviewed the
manuscript and offered many helpful suggestions.
This research was supported by NORFISH, a
marine research project of the University of
W^ashington Sea Grant Office and the National
Marine Fisheries Service (Grant 04-7-158-44021,
Office of Sea Grant, National Oceanic and Atmos-
pheric Administration, U.S. Department of Com-
merce). Financial support was also provided by the
National Research Council of Canada and by the
Minister of Education of Quebec. Contribution
No. 492 of the College of Fisheries, University of
Washington.
LITERATURE CITED
B.AKD, Y.
1974. Nonlinear parameter e.stimation. Academic Press,
N.Y.. 341 p.
Bevington. p. R.
1969. Data reduction and error analysis for the physical
sciences. McGraw-Hill Book Co., N.Y., 336 p.
Brown, K. M., and J. E. Dennls
1972. Derivative free analogues of the Levenberg-
Marquardt and Gauss algorithms for nonlinear least
squares approximations. Numer. Math. 18:289-297.
C.M^STON, D. R.
1969. A computer program for fitting the Richards func-
tion. Biometrics 25:401-409.
CONWAY. G. R., N. R. GLA.S.S. AND J. C. WiLCOX.
1970. Fitting nonlinear models to biological data by Mar-
quardt's algorithm. Ecology 51:503-507.
Draper. N. R., and H. Smith
1966. Applied regression analysis. John Wiley & Sons
Inc., N.Y.,407 p.
Fletcher, R, I.
1975. A general solution for the complete Richards func-
tion. Math. Biosci. 27:349-360.
1978a. Time-dependent solutions and efficient parameters
for stock-production models. Fish Bull., U.S. 76:377-
388.
1978b. On the restructuring of the Pella-Tomlinson sys-
tem. Fish. Bull., U.S. 76:515-521.
FOX, W. W., JR
1971. Random variability and parameter estimation for
the generalized production model. Fish. Bull., U.S.
69:569-580.
1975. Fitting the generalized stock production model by
least-squares and equilibrium approximation. Fish.
Bull., U.S. 73:23-37.
LEVENBERG, K.
1944. A method for the solution of certain non-linear prob-
lems in least squares. Q. Appl. Math. 2:164-168.
May, R. M.
1974. Stability and complexity in model ecosystems. 2d
ed. Princeton Univ. Press, Princeton, N.J., 265 p.
MARQUARDT, D. W.
1963. An algorithm for least-squares estimation of non-
linear parameters. J. Soc. Ind. Appl. Math. 11:431-441.
Pella, J, J., and P, K. Tomlinson.
1969. A generalized stock production model. Inter-Am.
Trop. Tuna Comm., Bull. 13:419-496,
RICKER, W. E.
1975. Computation and interpretation of biological statis-
tics of fish populations. Fish. Res. Board Can., Bull. 191,
382 p.
SCHNUTE, J,
1977. Improved estimates from the Schaefer production
model: theoretical considerations. J. Fish. Res. Board
Can. 34:583-603.
WALTER. G. G.
1975. Graphical methods for estimating parameters in
simple models of fisheries. J. Fish. Res. Board Can.
32:2163-2168.
534
SYSTEMATICS AND ZOOGEOGRAPHY OF THE PORCUPINEFISHES
(DIODON, DIODONTIDAE, TETRAODONTIFORMES), WITH
COMMENTS ON EGG AND LARVAL DEVELOPMENT^
Jeffrey M. Leis^
ABSTRACT
The porcupinefish genu.s Diodon is composed of five species: D. hystrix Linnaeus and D. eydouxii
Brissout de Bameville are closely related species, each of which has a relatively elongate body, spines
on the caudal peduncle, and high dorsal and anal fin ray counts; D. holocanthus Linnaeus and D.
Uturosus Shaw form a second species pair, each of which has a round body, no caudal peduncle spines,
and moderate dorsal and anal fin ray counts; D. nicthemerus Cuvier is a round-bodied species but
differs from D. holocanthus and D. Uturosus in meristic characters and spination.
Diodon hystrix. D. holocanthus. and D. eydouxii are distributed circumtropically. The Atlantic
population of D. holocanthus has diverged from the Indo-Pacific (including eastern Pacific) popula-
tions. Diodon eydouxii is pelagic, and both D. hystrix and D. holocanthus have pelagic juvenile stages.
Diodon Uturosus is found in the Indo-West Pacific, and D. nicthemerus is limited to Tasmania and
southern Australia. It is not known whether the latter species have pelagic stages.
The egg and larval stages of D. hystrix and D. holocanthus (the latter identified by rearing) are
similar. The pelagic eggs are 1.6-2.1 mm in diameter and hatch in about 5 days at 25°C. The larvae
metamorphose to spiny juveniles at ca. 4 mm in about 3 wk. Both species have pelagic juvenile stages of
long duration: D. hystrix remains pelagic to ca. 200 mm standard length, thus providing ample time for
dispersal. Eggs and larvae of the other species are unknown.
The identities of the species of the genus Diodon
have been confused since the time of Linnaeus.
The most recent description of a valid "new"
species was in 1846, but, unfortunately, time has
done little to clarify the situation. Twenty-eight
nominal species attributable to Diodon have been
described since 1758, and most contemporary au-
thors recognize two or three species. However, Le
Danois (1959), in the only recent review of the
genus as a whole, recognized six species.
The present study grew out of attempts to iden-
tify juvenile Diodon that resulted from rearing of
pelagic eggs taken in Kaneohe Bay, Oahu, Hawaii
(Watson and Leis 1974). These juveniles could not
be identified using existing keys. While current
literature recognized only two species of Diodon in
Hawaiian waters, examination of museum speci-
mens revealed that three were present there. This
discovery, together with the encouragement of
J. E. Randall of the Bernice P. Bishop Museum, led
to the present study clarifying the identities of all
of the species of Diodon and the description of their
'Hawaii Institute of Marine Biology Contribution No. 548.
^Department of Oceanography, and Hawaii Institute of
Marine Biology, University of Hawaii, Kaneohe, Hawaii; pres-
ent address: Marine Ecological Consultants, 533 Stevens Av-
enue, Solana Beach, CA 92075.
Manuscript accepted December 1977.
FISHERY BULLETIN: VOL. 76, NO. 3, 1978.
development. An attempt was made to obtain in-
formation on existing type-specimens and this,
along with the examination of a large number of
specimens, has led to the conclusion that the genus
is composed of five species, three of which are dis-
tributed circumtropically. Further, it is shown
that the present taxonomic confusion is attributa-
ble to inadequate original descriptions, reliance
on poor characters for differentiation, the close
similarity of several of the species, and unusual
aspects of the life histories of the species of Diodon .
All of the nominal species could be distinguished
with some certainty with two exceptions: the type
of Diodon echinii.s Rafinesque 1810 could not be
located and the original description provides no
clue to its identification; the holotype of Trichocy-
clus ennaceus Giinther 1870 (BMHN 1976.2.23.1)
is a small fish in especially poor condition, giving
the appearance of having been obtained from a
stomach of some predator, and, while it is cer-
tainly a Diodon, more specific identification could
not be made. Diodon dussumieri Bibron (see Le
Danois 1959, 1961) is a nomen nudum, but exami-
nation of the "type" (MNHN 1306) by J. E. Randall
of the Bernice P. Bishop Museum indicates that Le
Danois was correct in placing D. dussumieri in
synonomy with D. holocanthus.
535
FISHERY BULLETIN; VOL. 76, NO. 3
Although basically shorefishes, the diodontids
(at least Dwdon and Chilomycterus) are strongly
tied to the pelagic environment through pelagic
eggs and well-developed pelagic juvenile stages.
In Diodon these juveniles remain pelagic for
weeks or months ( judging from size) and are often
found far from shore. In fact, juvenile Diodon spp.
are commonly encountered in the stomachs of such
pelagic predators as dolphins (Gibbs and Collette
1959), and one species, D. eydouxii, is apparently
pelagic throughout its life cycle.
The eggs of diodontids are poorly known.
Nichols and Breder (1926) described the unfer-
tilized eggs of Chilomycterus schoepfi from New
Jersey as demersal, nonadhesive, transparent,
and about 1.8 mm in diameter. However, Breder
and Clark (1947) suggested that the eggs of C.
schoepfi may be normally pelagic. The pelagic eggs
of D. holocanthus and D. hystrix from Hawaii were
briefly described as Diodon sp. and "diodontid II,"
respectively, by Wat,son and Leis (1974). Sanzo
( 1930) described the development of what are ap-
parently the pelagic eggs of D. hystrix (identified
as Crayracion sp.?) from the Red Sea. Wolfsheimer
(1957) reported an aquarium spawning of D.
holocanthus (identified by him as D. hystrix), but
provided little descriptive information on the
eggs. The eggs mentioned by Wolfscheimer sank,
but did not adhere, to the bottom of the aquarium.
They did not develop, .so it is likely that they were
not fertilized.
Larval and juvenile Diodon are no better known
than the eggs. Blanco and Villadolid (1951) illus-
trated a juvenile "Diodon bleekeri" but this fish is
clearly a juvenile tetraodontid. Many juvenile tet-
raodontids have prominent spines, particularly on
the ventral surfaces. Fowler (1928) illustrated a
juvenile Diodon, identified as D. hystrix, but the
figure does not show spines on the caudal peduncle
(see below), so this identification is apparently
incorrect (assuming the drawing is accurate). No
locality or other descriptive data are given by
Fowler, so a specific identification cannot be made.
Sanzo ( 1930) illustrated two larvae that resulted
from rearing of his D. hystrix eggs and a juvenile
Diodon captured in a plankton tow. The illustra-
tion of this latter fish shows no peduncle spines,
but in other respects it resembles D. hystrix. Mito
( 1966) illustrated a larval and a juvenile Diodon,
both identified as D. holocanthus. The pigmenta-
tion and the relatively small eye .shown in Mito's
illustrations more closely resemble the specimens
of D. hystrix studied here. At least four species of
536
Diodon occur in Japanese waters, and Mito's
specimens could be any of these, because he gives
no information as to how the identifications were
made. Nishimura (1960) reported on juvenile
Diodon cast ashore in the Sea of Japan, but did not
provide specific identifications.
MATERIALS AND METHODS
Measurements and counts are as defined by
Hubbs and Lagler (1958:19-28) unless otherwise
stated. Measurements routinely were made with
needle point dividers to the nearest 0.5 mm. Fish
<10 mm and all eggs were measured under a dis-
secting microscope to the nearest division of the
ocular micrometer ( ±0.02 mm at 50 x, the power
normally used). All measurements are from pre-
served specimens.
Unspecified lengths are in millimeters standard
length. Caudal peduncle length was measured
from the posterior base of the dorsal fin to the end
of the hypural plate. Head width was measured
immediately behind the eyes. Body width was
measured at the base of the pectoral fin. Width of
the eye was taken horizontally across the clear
cornea. Measurements are given as proportions of
standard length.
Dorsal and anal fin ray counts included all rays,
branched and unbranched. The last two rays were
counted separately because they have separate
bases. Pectoral fin ray counts excluded the upper
ray. This ray, although well developed in small
( <30 mm) juveniles, is a rudiment in adults and is
often not visible because it is embedded. In large
specimens, the fin bases are especially fleshy and
accurate fin ray counts are difficult to make with-
out dissection or radiography.
Body measurements are given as range, mean
(I), and standard deviation (SD). The sample size
for the measurements is given in parentheses at
the beginning of the description of each species.
Morphometries are included only from individuals
>50 mm. Fin ray counts are included for all
specimens on which counts could be made (Table
1). In most cases, rays in both pectoral fins were
counted. Fin rays were not counted on specimens
with fin damage or on specimens that had rays
obscured due to the thick bases of the dorsal and
anal fins. Radiography was tried unsuccessfully to
obtain vertebral counts: the dermal spines and
their bases obscured the vertebrae, and made ac-
curate counts impossible. The vertebral counts
LEIS: SYSTEMATICS AND ZOOGEOGRAPHY OK THE PORCUPINEFISHES
Table l. — Fin ray counts of Djodon species.
Dorsal fin rays
12
13
14
15
16
17
18
X
D eydouxii
Atlantic
3
1
17.25
Indo-Paciflc
4
25
6
17.06
D hystnx
Atlantic
2
7
1
15.90
Indo-Pacific
2
20
8
15.20
D holocanthus
Atlantic
22
4
14.15
Indo-Paclfic
9
39
7
13.96
0. liturosus
1
12
15
15.50
D nicthemerus
1
10
12.91
Anal fin rays
12
13
14
15
16
17
18
X
D eydouxii
Atlantic
3
1
17.25
Indo-Pacific
2
22
11
17.26
D. hystnx
Atlantic
3
7
15.70
Indo-Pacific
1
15
6
15.23
D holocanthus
Atlantic
9
18
13.67
Indo-Pacific
26
27
1
13.54
D liturosus
10
17
1
14.68
D nicthemerus
3
5
2
12.80
Pectoral fin rays
19
20
21
22
23
24
25
X
D eydouxii
Atlantic
5
3
21.38
Indo-Pacific
7
58
17
20.12
D. hystnx
Atlantic
1
10
10
1
23.50
Indo-Pacific
2
16
34
6
22.76
D holocanthus
Atlantic
7
31
20
22.22
Indo-Pacific
1
17
57
27
5
22.17
D liturosus
2
11
24
17
2
23.11
D. nicthemerus
2
9
9
20.35
given for D. holocanthus were made on cleared and
stained material.
The dermal spines require special terminology
and measurements, as given below. Measure-
ments, except for shaft length, were taken on dis-
sected spines (Figure 1).
The spine shaft is that portion bearing the
pointed tip, but excluding the shaft extension. The
length of the spine (= shaft length) was taken
from the lower portion of the lateral arm to the tip
of the shaft. The starting point for this measure-
ment can be found most easily by probing around
the base of the spine.
The shaft extension is the portion of the shaft
extending past the lateral arms of the base, and its
length was measured from the lower portion of the
lateral arm to the tip of the extension.
The lateral arms of the base are the subdermal
portions of the spine upon which the spine pivots
during erection. The length of the spine base was
the straight line distance from tip to tip of the
lateral arms.
The frontal spines are those of the anteriormost
row on the head between the eyes. The pectoral
axil spines are the spines immediately posterior to
the base of the pectoral fin.
Figure l. — Typical Diodon body spine: (A) spine (or shaft)
length, (B) length of the shaft extension, (C) length of the spine
base. The tip of the spine shaft points caudad.
The number of spines in a longitudinal row over
the dorsum from the snout to the dorsal fin base
(S-D spines) and the spines in a longitudinal row
over the ventrum from the lower jaw to the anus
(S-A spines) were counted. These rows of spines
are irregular and difficult to follow, so the counts
should be considered approximate. With practice,
repeated counts of ±1 can be achieved consis-
tently. The numbers of spines between pectoral
fins, both over the dorsum (P-D-P spines) and ven-
trum (P-V-P spines), were also counted, but these
counts are even less reproducible than the lon-
gitudinal counts.
Repeated reference is made to the spines on the
caudal peduncle. In some species the only spines in
the region of the caudal peduncle are some rather
large spines associated with the dorsal and anal
fin bases. Although these spines extend over the
peduncle, their subdermal bases (lateral arms and
shaft extension) are at least partially anterior to a
line between the base of the posteriormost rays of
the dorsal and anal fins, and they are considered
not to be on the peduncle. In other species, there
are relatively small spines which are wholly pos-
terior to the line defined above on the dorsal and
dorsolateral surfaces of the peduncle; these spines
are considered to be on the peduncle (Figure 2).
Larvae were obtained from plankton samples
(field specimens) and rearing experiments using
eggs from plankton tows (reared specimens). All
eggs and larvae were captured around the
Hawaiian island of Oahu. Rearing took place in
537
FISHERY BULLETIN: VOL. 76, NO. 3
B
Figure 2. — Semidiagrammatic lateral view of the caudal
peduncle and posteriormost spines of (A) a slender-bodied, long
peduncled species (Diodon eydouxii) and (Bl a round-bodied,
short peduncled species (D. holocanthus) .
the laboratory under ambient temperature (ca.
25°C) and a variety of conditions. Generally, the
eggs were hatched in unaerated 4-1 beakers filled
with seawater from the collection area. Hatched
larvae were transferred to 10-20 1 containers and
provided with overhead illumination. The con-
tainers were wrapped in black plastic. Wild zoo-
plankton (ca. 60-200 ixm) from a plankton pump
were added on alternate days; this was later
supplemented with Artemia nauplii. Water was
changed twice a week and .specimens were re-
moved periodically for preservation. Many rear-
ing attempts were made, but since fewer than 20
eggs usually were available per attempt, few of the
attempts were successful.
Some larvae were cleared and stained using the
KOH-alizarin red method of Hollister (1934).
Measurements and definitions of stages generally
follow those of Leis (1977), unless otherwise noted.
All drawings of eggs and larvae were made with
the aid of a camera lucida.
The institutions housing the examined speci-
mens are as follows: Academy of Natural Sci-
ences of Philadelphia (ANSP); Australian
Museum, Sydney (AMS); Bernice P. Bishop
Museum, Honolulu (BPBM); British Museum
(Natural History) (BMNH); California Academy
of Sciences (CAS); Gulf Coast Research Labora-
tory and Museum (GCRL); George Vanderbilt
Foundation (GVF), deposited in CAS; Hawaii In-
stitute of Marine Biology (HIMB); Los Angeles
County Mu.seum of Natural History (LACM);
Museum National d'Histoire Naturelle, Paris
(MNHN); National Marine Fisheries Service,
Honolulu, Hawaii (NMFS H), La Jolla, Calif.
(NMFS LJ), and Miami, Fla. (NMFS M);
Naturhistorisches Museum, Vienna (NMV); J. L.
B. Smith Institute of Ichthyology at Rhodes Uni-
versity, South Africa (RUSI); Scripps Institution
of Oceanography (SIO); Tulane University (TU);
National Museum of Natural History, Smithso-
nian Institution (USNM); University of Arizona
(UA). A catalog number is given when available;
many GVF specimens were uncataloged and
therefore the register or station number is given.
The synonymies include all known original
usage of names. In addition, references of sys-
tematic or zoogeographic interest are included. If
the identification of a nominal species is question-
able, it is preceded by a question mark (?). Pre-
Linnaean literature is cited in the text if appro-
priate, but is omitted from the synonymies.
GENUS DIODON LINNAEUS
Diodon Linnaeus 1758:334, after Artedi
1738. Type-species D. hystrix Linnaeus by
subsequent designation of International Com-
mission on Zoological Nomenclature, opinion
77.
Paradiodon Bleeker 1865:49. Type-species D.
hystrix Linnaeus by original designation.
Trichodiodon Bleeker 1865:49. Type-species D.
pilosus Mitchill by original designation.
Trichocyclus Giinther 1870:316. Type-species T.
erinaceus Giinther by monotypy.
Diagnosis. — Body rotund, width 0.25-0.54, depth
varies greatly depending on degree of inflation.
Eyes large, 0.05-0.17. Swim bladder bilobed.
Teeth in each jaw fused into a single beaklike
unit without a median suture dividing upper or
lower jaws into right and left halves. Gill opening
a short, vertical slit immediatelv anterior to the
538
LEIS: SYSTEMATICS AND ZOOGEOGRAPHY OF THE POKCT'PINEKISHKS
pectoral fin base. Approximately 20 vertebrae.
Dorsal and anal fins usually rounded, set far back
on body, with 12-18 rays. Caudal rounded, with 9
rays (there are no secondary rays). Pectoral fin
slightly emarginate, with 19-25 rays, the upper-
most ray (not counted) greatly reduced in adults.
No pelvic fins. Body covered with long spines, all
but a few (around the gill opening, dorsal fin base,
and caudal peduncle! of which are erectile. Erec-
tile spines consisting of a long pointed shaft, tw'o
subdermal lateral bases lying in nearly the same
plane as the shaft, and usually a shaft extension
which is shorter than the shaft. The shaft exten-
sion may be greatly reduced. Nasal organs consist-
ing of a short tentacle with a pair of lateral open-
ings near the tip. In larger individuals of some
species the tissue closing the end of the tentacle
may be absent, giving rise to a bifid nasal tentacle
without nostrils. Both species whose reproductive
habits are known (D. hystnx and D. holocanthus)
spawn pelagic spherical eggs of 1.6-2.1 mm in
diameter.
Ren2arks. — Only Bleeker's (1865) proposal of
Paradiodon for the species here considered to be-
long in Diodon (because of page priority, he be-
lieved Diodon should apply to those species usu-
ally referred to Chilomycterus) has disturbed the
stability of the usage of the name Diodon.
Trichodiodon and Trichocyclus are names applied
to juvenile stages oi Diodon.
Although subgeneric status seems unwar-
ranted, Diodon can be broken into two groups on
the basis of body width, caudal peduncle length,
and squamation. The species of the slender-bodied
group, D. eydouxii and D. hystrix, have a rather
narrow body (Figure 3, Table 2), long caudal
peduncle (Figure 3, Table 2), and several small
spines in the dorsal and dorsolateral surfaces of
the peduncle. The species of the round-bodied
group, D. holocanthus, D. liturosus, and D. nicth-
emerus, have a wider body, shorter caudal pedun-
cle (Figure 3), and lack spines on the caudal
peduncle (although there are strong spines, pro-
jecting over the peduncle, at the base of the dorsal
and anal fins). Upon inflation, the dorsal and anal
fins are engulfed by the expanding skin. In the
round-bodied group, the caudal peduncle and fin
are also largely obscured in inflated specimens
and the large spines mentioned above provide
added protection. In the slender-bodied group, the
peduncle remains largely uncovered and is pro-
tected only by the relatively small spines on its
upper surfaces. Diodon nivthemerus, although
clearly a member of the round-bodied group, ap-
pears to have undergone a reduction in spine
number and base size, and is thus separable from
D. holocanthus and D. liturosus.
70-1
E 60-
E
-LITUROSUS
-EYDOUXII
-HYSTRIX
-NICTHEMERUS
-HOLOCANTHUS
X
100
150
200
Figure 3. — Plotted regression lines of (top) caudal peduncle
length vs. standard length and (bottom) body width vs. stan-
dard length for the five species of Diodon. Lines plotted only-
over size range of specimens examined. The line with arrow-
head for D. hystrix extends to 571 mm SL. Regression data in
Table 2.
Table 2. — Regression equations for caudal peduncle length
(PL) and body width (BW) vs. standard length (SL) in the five
species of Diodon (see also Figure 3).
Species
Regression equation
r
'slope
df
D, hystrix
PL = 0.189 SL - 2.79
0.97
21.16
31
D eydouxii
PL = 0.226 SL - 4.79
0.95
17.44
33
D. liturosus
PL = 0.159 SL- 2.40
096
17.88
26
D. nicthemerus
PL = 0.151 SL- 1.38
0.94
785
8
D- holocanthus
PL = 0.152SL+ 4.69
0.90
1586
61
D. hystrix
BW = 0.338 SL+ 6.01
097
23.41
31
D. eydouxii
BW = 0.262 SL + 5 27
0.88
10.45
33
D. liturosus
BW = 0.333 SL + 10.29
0.90
10.58
26
D. nicthemerus
BW =0 313SL ^ 13.11
0.93
6.44
6
D holocanthus
BW = 0.368 SL ^ 6.29
0.96
2575
62
539
FISHERY BULLETIN; VOL. 76. NO. 3
KEY TO THE SPECIES OF
THE GENUS DIODON
la Two or more small spines wholly on the dorsal or dorsolateral surfaces of the caudal
peduncle (Figure 2A); color pattern of adults dominated by small (smaller than eye)
spots; at least D, P, and C fins of adults with dark spots 2
lb No spines w^holly on the caudal peduncle (Figure 2B); color pattern of adults dominated
by large dorsal and lateral bars or blotches; fins of adults without spots except in some
cases at base 3
2a P 19-22. both D and A 16-18; D and A of adults falcate; S-A spines^U; head width less
than 30'y SL D. eydouxii
(circumtropical, oceanic)
2b P 22-25 (rarely 21), D 14-17, A 14-16; D and A of adults rounded; S-A spines&14; head
width greater than 307^ SL D. hystrix
(circumtropical, shore fish but juveniles pelagic)
3a No small, fixed, tribase spine immediately above the gill opening; no small, flat spines on
the anterior border of the depression surrounding the gill opening (Figure 4); S-A spines
^11; adult color pattern dominated by four large lateral bars, dorsum uniformly
dark D. nicthemerus
(Australia)
Figure 4. — Head of Dwdon nicthemerus (AMS 1.16990-004) showing arrangement of spines in
the region of the gill opening. Note that spines anterior to gill opening are not flattened. Also note
tubular nostril.
3b One or two small, fixed, tribase spines above the gill opening; three or four small, flat
spines forming the anterior border of the depression surrounding the gill opening ( Figure
5); S-A spines S3l2; adult color pattern dominated by several dorsal blotches
540
LEIS: SYSTEMATICS AND ZOOGEOGRAPHY OF THE PORCUPINEFISHES
Figure 5.— Head of Dwdon liturosus (CAS
30967) showing arrangement of spines in the re-
gion of the gill opening. Note that spines on an-
terior border of opening are short and flattened.
Also note the small dow-nward-pointing spine
below the anterior border of the eve.
Longest Frontal Spine/ SL
— I © 1 D. liturosus
1 ® 1 D. hystrix
1 0 1 D. eydouxii
At! D. holocanthus ^ ^ — i
la-Pac. D. holocanthus 1 1© —
D nicthemerus
^ H
FlGURE 6. — Ratio of frontal spine length to standard length for
the five species of Diodon. Line indicates range, circle and bar
indicate mean, and vertical bars alone denote ±1 SD. Note dif-
ference of spine length between Atlantic and Indo-Pacific
specimens of Z). holocanthus . Number of specimens given in de-
scription for each species.
I 1 1 r
.04 .08
.12
-! 1 —
.16
.20
— I —
.24
4a Frontal spines 0.04-0.10 (Figure 6i, much shorter than pectoral axil spines; 17-22 S-A
spines; a small downward-pointing spine below the anterior margin of the eye present;
dorsal blotches with a distinct light border; a dark gular band from eye to eye under the
lower jaw D. liturosus
(Indo-Pacific)
4b Frontal spines 0.13-0.28 (Figure 6), slightly shorter to much longer than pectoral axil
spines; 12-15 S-A spines; a small downward-pointing spine below the anterior margin of
the eye absent (Indo-Pacific specimens) or present (most Atlantic specimens); dorsal
blotches without a distinct light border; no gular band D. holocanthus
(circumtropical)
541
FISHERY BULLETIN VOL 76. NO 3
DIODON EYDOUXII BRISSOUT DE
BARNEVILLE
Pelagic Porcupinefish (Figure 7)
Diodon cyduuxii Brissout de Barneville 1846:142
(eastern Pacific); Troschel 1847:364; Dumeril
1855:278.
Diodon melanopsis Kaup 1855:228 (no locality
given).
Diodon spinosissimu.s (not of Cuvier): Giinther
1870:307 (Cape of Good Hope, Siam).
Diagnosis. — A slender-bodied Diodon , head width
0.25-0.30, peduncle length 0.16-0.22. Caudal
peduncle armed dorsally with short spines. Body
spines long and slender, moderate in number, S-D
spines 13-17, S-A spines 10-14. Pectoral axil
spines 0. 11-0.16, usually longer than longest fron-
tal spines. A short, fixed tribase spine im-
mediately above gill opening. D 16-18, A 16-18, P
19-22. Nasal tentacle with a pair of lateral open-
ings. No barbels or fleshy tentacles. Dorsal and
anal fins falcate (rounded in juveniles). Color pat-
tern dominated by small (ca. = to pupil) dark
spots dorsally and laterally. These often as-
sociated with the spine axils. A dark gular band
starting from below the eyes and continuing under
the chin, usually with a branch extending dorsally
between eye and gill opening.
Descn'p^/o/;.— (35 specimens) D 16-18, A 16-18, the
first two or three rays unbranched; P 19-22. Head
width 0.25-0.30 (.v = 0.27; SD - 0.01), body width
0.25-0.35 (.V = 0.30; SD = 0.02), peduncle length
0.16-0.22 (.V = 0.19; SD = 0.02), eye 0.05-0.10
(X = 0.08; SD = 0.01). Dorsal and anal fins falcate,
not rounded. Nasal tentacles with a pair of lateral
openings.
S-D spines 13-17, S-A spines 10-14, about 12
spine rows over the dorsum between pectoral fin
bases, about 21 spine rows over the ventrum be-
tween pectoral fin bases. Four or five frontal
spines. Longest frontal spine 0.07-0.15 ix = 0.11;
SD =0.05), pectoral axil spines 0.11-0.16
ix = 0.14; SD = 0.01). Pectoral axil spines usually
the longest on the body, 0.61-1.03 ix = 0.78;
SD = 0. 1 1 ) in frontal spines. Spines long and slen-
der. Frontal, middorsal, and ventral spines of
about the same length. Pectoral axil spines and
those dorsolateral spines from over eye to over
pectoral fin among the longest on body (ca. 0.8 in
frontal spines). Spines on caudal peduncle short
(ca. 1.5 in frontal spines) and fixed due to a rather
long shaft extension (ca. 2 in shaft). Shaft exten-
sion on other spines reduced, never more than 159f
of the shaft length. Subdermal bases moderate in
extent, and, except for spines around fin bases and
caudal peduncle, always shorter than shaft. No
spines markedly reduced other than on caudal
peduncle; the latter spines generally arranged in
one or two bilateral pairs along the dorsolateral
edge of the peduncle. Approximately 40^^ (14 of
36) of the specimens examined also possess a
single dorsomedial spine on the caudal peduncle.
A short, fixed tribase spine immediately above the
gill opening and a second slightly posterior to it
above the pectoral base. Three short, flat spines
Figure l.— Diodon eydouxii, 128 mm SL, central Pacific (NMFS H CHG 55-71)
542
LEIS: SYSTEMATICS AND ZOOGEOGRAPHY OF THK PORCHPINEFISHES
with broad lateral bases form the anterior border
of the gill opening. No spines on the snout.
No barbels or fleshy tentacles.
Dorsally the ground color is light grey to brown
grading to white ventrally. Dorsal and lateral sur-
faces marked with dark ovoid spots ( 1.5 mm) and numerous oil droplets serves to
distinguish the eggs of D. holocanthus from
those of all other pelagic eggs except those of
other tetraodontiform species. The eggs of the
molid Ramania laevis have been described by
Leis ( 1977). Ramania laevis eggs may be distin-
guished from D. holocanthus eggs by the
former's smaller size (1.4-1.65 mm) and by the
extensive pigment which develops on the ven-
tral surface of the yolk sac of R. laevis in the
middle stage.
Hawaiian ostraciid eggs (Ostracion and Lac-
toria) may be distinguished by their slightly
oblong shape, fewer oil droplets ( <10), but most
reliably by a patch of bumps on the chorion
surrounding the micropyle. This "rough patch"
is easilv overlooked.
FISHERY BULLETIN; VOL. 76. NO, 3
Diocion hystrix eggs are the only other Diodon
eggs known ( see section on D. hystnx). They can
be distinguished from those of D. holocanthus by
their larger size ( >1.9 mm), greater number of
oil droplets ( >30), and the orange (rather than
red) pigment.
Larval Development: Fifteen reared and 12
field-collected larvae in good enough condition
for descriptive purposes were available. Mor-
phometric data are summarized in Table 4.
The newly hatched larva has well-developed,
apparently functional eyes, jaws, and gas blad-
der (Figure 22). The pectoral fins are quite
large, although no rays are formed. The larvae
are 1.9-2.1 mm SL at hatching and the body is
rotund. Development in reared larvae is slow.
Dorsal and anal fin anlagen form by day 10 (2.4
mm, Figure 22); the olfactory pit also forms by
this time and the eyes have become proportion-
ally larger. The oldest reared larva available
was 16 days old, but it was smaller than the
Table 4. — Morphometric and meristic data for larval and juvenile Diodon holocanthus
(measurements in mm). ? indicates individuals of unknown age, from plankton samples;
X indicates damaged.
Age (days)
Notochord
Snout
Fin ray count;:;
of reared
or standard
to anus
Width
Head
Head
Mouth
—
fish
length
length
of eye
length
width
width
D
A
P
Larvae
1
2.0
1-5
03
09
1,1
0,5
0
0
0
1
2.1
15
03
0,8
1,2
05
0
0
0
1
2.0
15
03
0,8
12
04
0
0
0
1
2.0
1.4
03
08
1 2
04
0
0
0
1
1.9
1,4
03
07
12
05
0
0
0
?
1.9
1,5
03
0,8
1,1
0,6
0
0
0
?
1.9
1 4
03
0,9
—
0,4
0
0
0
?
1.«
1 6
04
09
1,2
06
0
0
0
?
1.9
1 0
03
06
—
—
0
0
0
7
2.0
1,6
04
08
1,1
06
0
0
0
?
2.0
1 6
03
07
1 1
03
0
0
0
5
1.8
14
03
08
12
0,6
0
0
0
6
1.8
14
04
0,8
1 0
06
0
0
0
7
1.9
14
03
0.8
1,0
05
0
0
0
8
2.1
1 4
04
0,9
1 1
06
0
0
0
8
2.2
1 7
0 5
09
14
07
0
0
0
8
2.1
15
04
0,8
1,1
06
0
0
0
9
2.0
1 4
0,4
0,8
10
0,5
0
0
0
?
2.2
1 5
03
0.8
1,2
0,5
0
0
0
?
2.3
2,0
05
0,8
1,5
06
0
0
0
?
2.3
1,5
03
0,8
1,1
0,5
0
0
0
10
2.4
20
05
0,9
1 5
08
0
0
0
10
2.2
1 7
05
09
13
06
0
0
0
?
2.5
2 1
05
07
—
04
0
0
0
?
2.6
2,0
0 5
0,7
08
04
0
0
0
?
2.7
22
05
07
—
—
0
0
0
16
11.9 .
1,5
0,5
10
1,2
0,7
0
0
0
Juveniles:
?
3.8
34
08
1,9
2,6
1,0
22
25
4.8
40
1 0
23
35
1,9
14
14
23
?
5.5
4,8
1,1
2,8
3,3
16
—
—
—
?
6.0
5,3
12
3,3
3,5
18
—
—
—
33
6.7
5,2
1,4
29
43
1,7
15
.
21
?
7.2
5,5
1 5
3,4
37
1 7
14
14
23
ca 30
8.1
67
1 8
39
4,8
1,8
14
14
21
7
11.0
9,0
20
55
63
2,7
—
—
—
?
14.1
109
2,7
63
7,6
32
15
14
23
'Fish in emaciated condition.
560
LEIS: SYSTEMATICS AND ZOOGEOGRAPHY OF THE PORCrPlNEFISHES
Figure 22. — Reared larvae ofDiodon holocanthus: (top) newly
hatched larva 2.0 mm, (middle) 10-day-old larva 2.4 mm, and
(bottom) dorsal view of 10-day-old larva with pigment omitted.
day-10 larvae and appeared emaciated. There
are incipient fin rays and bases visible in the
fins of the 16-day-old larva, but it otherwise is
not obviously advanced over the 10-day-old
specimen. There is no sign of development of
the caudal fin complex. The largest larva avail-
able is a 2.7-mm field-collected specimen which
is no more advanced than the day- 16 larva. The
dermal sac is inflated in young larvae (Figure
22), but the subdermal space is virtually gone
by day 10 (Figure 22).
The larvae are more or less uniformly pig-
mented with scattered melanophores on the
dorsal surfaces at all stages. The pigment
spreads laterally, but there is little below the
level of the pectoral fin and the ventral surfaces
remain devoid of melanophores until metamor-
phosis. The newly hatched larvae have no
melanophores posterior to the anus (Figure 22),
but by day 10 postanal pigment has spread to
the middle of the dorsal fin anlage. In life, the
newly hatched larva is covered with widely
scattered red chromatophores on the dermal
sac and fins. The red pigment persists through
the larval stage and on about day 2 it is
supplemented by a yellow background pigment
covering all the body surfaces (not the dermal
sac), but being most obvious ventrally due to a
lack of melanophores there.
A 2.0-mm field-collected specimen was
cleared and stained. The only ossified struc-
tures were the cleithrum, coracoid, and six
branchiostegals.
Juvenile Development: Metamorphosis appar-
ently occurs at ca. 3 mm at an age of about 3
wk. The smallest juvenile available is 3.8 mm
and resembles Mito's (1966) illustration of a
3.7-mm juvenile except that Mito's fish had
smaller eyes. The caudal, dorsal, anal, and pec-
toral fins are all formed as are the teeth, and
the body is covered with small spines. The
spines do not appear to be erectile, but the fish
is capable of inflation. The spines are covered
with a sheathlike tissue. They elongate rapidly
with growth and by 4.8 mm SL (Figure 23) they
are obviously erectile. The nostrils are formed
in the 3.8-mm flsh, although the nasal tentacle
with two lateral openings is not formed until
4.8 mm SL, and in fish as large as 6.0 mm. it
may be open at the ends. The 4.8-mm fish is in
all respects a miniature adult with all external
structures formed and functional. External
changes to the adult stage involve only changes
in proportion; the spines in particular elongate,
the body becomes less rotund and the eye rela-
tively smaller. Morphometric and meristic data
are summarized in Table 3.
A 33-day-old juvenile of 6.7 mm was cleared
and stained. The vertebral column and skull are
incompletely ossified but all other structures
are ossified. The vertebral formula is
12 + 9 = 21 and the vertebral column is
strongly arched. There are 1 1 dorsal and 1 1 anal
pterygiophores which are associated with ver-
tebrae 12-16 and 13-17, respectively.
At metamorphosis, pigment changes radi-
cally. The background color in live material is
still predominantly yellow with scattered red
chromatophores but this does not persist. Dor-
sally, the melanophores are scattered fairly uni-
formly, with a concentration at the pectoral base
and very little pigment on the caudal peduncle.
561
FISHERY BULLETIN. VOL. 76, NO. 3
Ventrally, however, a number of distinct spots
have formed that cover the belly (Figure 23).
The spots (pelagic spotting) are at first close
together but become less numerous and propor-
tionately larger, aligning in rows with growth
(Figure 19). Dorsal spotting (always more dif-
fuse than ventral spotting) begins to form at
around 10 mm and the characteristic dorsal
blotch pattern is generally visible by 30 mm,
although in pelagic specimens the contrast with
the background color is not great. The pelagic
spotting is retained in all pelagic individuals
examined (to 86 mm) and in some specimens
collected inshore. The fins remain unpigmented
except for a few melanophores along the fin rays
of the dorsal fin.
Identification of Larvae and Juveniles: Diodon-
tid larvae are likely to be confused only with the
rotund, heavily pigmented, sac enclosed
ceratioid larvae and other tetraodontiform lar-
vae. Reference to Bertelsen's (1 95 1 ) work should
allow ceratioid larvae to be distinguished as
such. Rotund tetraodontiform larvae may be
distinguished from diodontid larvae as follows:
molids by their body spination and early form-
ing pectoral rays; ostraciids by their pigmenta-
tion and early forming pectoral rays; tetraodon-
tids by their relatively more elongate body
shape and early forming fin rays. Diodon larvae
are heavily pigmented only on dorsal surfaces,
do not develop fin rays until near or at
metamorphosis, have very wide heads and
bodies ( >body depth), and have very wide
mouths.
The larvae of D. hnlocanthus can be distin-
guished from the putative D. hystrix larvae, the
only other larval diodontid known, by the less
Figure 23.— Reared juvenile of
Diodon holocanthus, 4.8 mm SL, 25
days old. Note pelagic spotting.
Hawaiian material.
well-developed condition at hatching of the lat-
ter (see section on D. hystrix). In addition, D.
hystrix larvae are predominantly orange upon
hatching while those of D. holocanthus are yel-
low. Melanophores of D. holocanthus do not
extend onto the postanal myomeres past the
middle of the dorsal and anal fin anlagen; the
postanal myomeres of D. hystrix are moderately
pigmented. Lastly, the eyes of Z). hystrix larvae
are smaller than those of D. holocanthus larvae
(Tables 2,3).
Once the spines form, the lack of caudal
peduncle spination, fin ray counts and spine
placement serve to distinguish D. holocanthus
from all other Diodon species (see Key).
The duration of the pelagic stage is unknown,
but judging from reared specimens, metamor-
phosis occurs about 3 wk after hatching at about
4 mm SL. The largest individual captured pelag-
ically was 86 mm while the smallest captured
inshore was 60 mm. A certain amount of plastic-
ity in the duration of the pelagic stage is indi-
cated, but its length clearly must be measured
in terms of months. No special adaptations for
pelagic life are evident in these juvenile stages
except, perhaps, in color. In the tetraodontiform
fishes (except the molids) the larval stage is
short and relatively unspecialized, while a rela-
tively unmodified pelagic juvenile stage may be
quite long (see Remarks under D. eydouxii).
This strategy (for dispersal?) is in marked con-
trast to that in many advanced perciform
shorefishes (e.g., Acanthuridae, Chaetodon-
tidae) where bizarrely modified and long-lived
larval and pelagic prejuvenile stages are de-
veloped which subsequently undergo marked
(and rapid) metamorphosis upon becoming
benthic.
562
LEIS: SYSTEMATICS AND ZOOGEOGRAPHY OF THE PORCUPINEFISHES
Diodon holocanthus eggs and larvae have
been found in Hawaiian waters from February
through September, with an apparent peak in
abundance in May-June, although they are
never common. Larvae usually occurred singly
in plankton tows ( volume filtered 200- 1 ,000 m-'^).
Although as many as 30 eggs 1.000 m'^ have
been taken, 1-5 eggs/1,000 m^ were more usual,
and most tows contained none. Eggs were usu-
ally found close to shore, but larvae rarely were
found closer than 1 km from shore (pers. ob-
serv.).
Holotype. — No holotype or type-series is known to
exist. Linnaeus based his description on that of
Artedi (1738).
Distribution . — Diodon holocanthus is circumtrop-
ical in distribution, but is seemingly absent in the
southwest and central Pacific east of the andesite
line (the separation of continental from oceanic
rocks, Figure 14). However, it reappears in
Hawaii, Pitcairn, and Easter Islands. Cuvier's
holotype of D. quadrimaculatus was allegedly col-
lected by Peron in Tahiti (see Le Danois 1961).
Inasmuch as it is known that much of the locality
data accompanying Peron's specimens are incor-
rect (associated with a shipwreck, see Whitley
1931:25) this record is questionable. There is evi-
dence of divergence of the Atlantic population! s)
from those of the Indo-Pacific (see Remarks).
Remarks. — I follow the spelling holocanthus
(rather than holocanthus of many authors) which
was used consistently by both Linnaeus and Ar-
tedi (see also Bailey et al. 1970), and is thus not
considered to be a misprint as maintained by Jor-
dan and Evermann (1891). Linnaeus' description
is brief; the only useful information being the
statement that the spines are terete and ex-
tremely long on the head and nape. However, this
can apply only to D. nicthemerus or D. holocan-
thus. Assuming that "Habitat in India" means
India as understood today, and not the entire
Indo-Pacific, D. nicthemerus is eliminated. How-
ever, even if "Habitat in India" means the entire
Indo-Pacific, it is unlikely that specimens of D.
nicthemerus, a species apparently confined to
southern Australia, could have reached Artedi by
1738. In any case, subsequent usage and stability
demand that the name D. holocanthus apply to the
species described above.
Diodon pilosus is synonymized with D. holocan-
thus on the basis of Mitchill's observation that no
spines were present between the dorsal and caudal
fins of his small (ca. 38 mm) New York specimen.
Diodon holocanthus is the only Atlantic species
that lacks peduncle spines. Mitchill distinguished
D. pilosus on the basis of its flxible spines, but this
is the usual condition in small specimens. No
holotype is known to exist.
Cuvier's types are extant. Information and
photographs of these specimens (catalog numbers
and other information are given by Le Danois
1961) provided by M. L. Bauchot (pers. commun.,
MNHN, 20 May 1975) clearly establish D.
novemmaculatus, sex?naculatus . quadrimac-
ulatus, and multimaculatus (all of Cuvier) as
junior synonyms of D. holocanthus . Inasmuch as
Cuvier's (1818) descriptions are relatively clear,
only his D. novemmaculatus requires comment.
The holotype of D. novemmaculatus (MNHN
A. 9928, 107 mm) is D. holocanthus, apparently
from the Atlantic (no locality data are available
for this specimen). A spine is present below the
anterior margin of the eye and the eye bar is dis-
continuous over the interorbital. Unfortunately,
Cuvier's figure resembles D. liturosus as much as
D. holocanthus (the figure shows the frontal spines
shorter than they actually are). This probably led
Bleeker (1865) to apply the name D. novem-
maculatus to D. liturosus.
Diodon maculifer Kaup ( 1855) is included here
with some questions. Kaup's description is of little
help, and no type material can be found in the
British Museum where it would be expected to
reside. The holotype may have been part of Kaup's
lost personal collection (A. C. Wheeler, pers. com-
mun.). Examination of one of the South African
(Kaup's type-locality) specimens of "Diodon
maculifer" listed by Giinther (1870) (BMNH
1845.7.3.103, 100 mm, loaned by A. C. Wheeler)
reveals it to be an inflated, dried D. holocanthus.
In this specimen, inflation is so great ( an artifact of
stuffing and drying?) that the subdermal spine
bases project through the dried skin. Thus, the
base of the spines appear to be expanded and
transversely compressed. The only characteristic
feature of Kaup's description is the compressed
nature of the spines, and it seems likely that his
description was based on a dried, inflated D.
holocanthus.
Steindachner's Atopomycterus bocagei can be
placed in the synonomy of D. holocanthus on the
basis of information on the holotype (NMV 63848)
563
FISHERY BULLETIN; VOL. 76, NO. 3
provided by P. Kahsbauer (pers. commun., NMV,
1975). Steindachner's (1866) description is essen-
tially correct and unquestionably refers to D.
holocanthus. The placement of this specimen in
Atopumycterus was apparently based on the split
nasal tentacle (see section on D. nicthemerus). A
single split nasal tentacle was present on only 3 of
the more than 100 specimens of D. holocanthus
examined, so this condition is rare but not unpre-
cedented.
Both D. liturosus Shaw and D. maculatiis
Lacepede (the Latinized version of Le Diodon
Tachete) have been incorrectly applied to D.
holocanthus by various authors (see section on D.
liturosus).
For about the past 50 yr the chief sources of
confusion on the identity of D. holocanthus have
been confusion with D. histrix by some (mostly
American) authors and the lumping of D. liturosus
under D. holocanthus by nearly all authors. The
latter problem is discussed under D. liturosus.
The confusion between D. hystnx and D.
holocanthus stems primarily from three sources.
Many authors (e.g., Gosline and Brock 1960) have
conjectured that D. holocanthus is the young of D.
hystrix because the former does not reach a large
size, and few, if any, small specimens of the latter
were available. However, as discussed under D.
hystrix, this species is pelagic to ca. 200 mm and is
thus unavailable to inshore collecting. Inasmuch
as D. holocanthus does not commonly exceed 200
mm, the confusion was perhaps understandable.
Second, many early descriptions are poor and
keys often rely solely on the size of frontal spines
relative to the pectoral axil spines to distinguish
the two species. Especially in Atlantic specimens
of D. holocanthus, the frontal spines are likely to
be approximately the same size or even shorter
than the pectoral axil spines.
Finally, as noted by Clark andGohar( 1953) (see
alsoBagniset al. 1972:225), living/), hystrix ohen
display a dorsal blotch pattern not unlike that of
D. holocanthus. I have not observed this color pat-
tern in preserved D. hystrix.
The apparent divergence of the Atlantic and
Indo-Pacific populations of D. holocanthus men-
tioned above is of interest. At present, since D.
holocanthus is apparently absent from the Red Sea
and the Mediterranean, gene flow could occur only
around southern Africa. Evidence that this is ap-
parently not happening comes from the Indian
Ocean specimens which lack a snout spine and
have very long frontal spines in contrast to the
Atlantic specimens (Table 5). In addition, Poll's
( 1959) description (as D. hystrix ) of a west African
specimen is typical of the specimens from the
western Atlantic examined by me. The apparent
increase in frontal spine length from the Atlantic
to the Pacific to the Indian Oceans is curious.
Based on studies of other groups (Ekman 1967)
affinities might be expected between the Atlantic
and eastern Pacific populations, but no extension
to Hawaii and Easter and Pitcairn Islands would
be expected. The lack of the snout spine in all but
the Atlantic population and one Hawaiian speci-
men may indicate that the Atlantic population is
distinct. Fin ray counts are of little help in resolv-
ing this question. Because all the characters
which appear to differ between the Atlantic
specimens and those from other areas are rather
variable (although some are significantly differ-
ent in a statistical sense), I choose not to distin-
guish formally the populations nomenclaturally
at the subspecific level. If future study shows this
split to be desirable, the proper name for the At-
lantic specimens would be Diodon holocanthus
pilosus Mitchill.
Le Danois ( 1954 ) reported sexual dimorphism in
D. holocanthus, but her illustration of a female D.
holocanthus (p. 2355:fig. 3) appears to be D.
liturosus.
Material examined. — 141 specimens, 5-289 mm.
EASTERN PACIFIC: NMFS LJ (1;18.5) 18°56'N, 104 = 10 "W;
NMFS LJ D31-133.25 ( 1:64. 5l 26°04.5'N, 112°48.0'W; NMFS LJ
TO-5801 ( 1:85.5) 5°29.5'N, 77°57'W; NMFS LJ ( 1:73.5) "350 mi.
west of Costa Rica"; NMFS LJ B-5011 157.40 (2:41-41.5)
21°32.5'N, 111°14.5'W; UA 66-39-18 (1:242) San Agustin Bay,
Sonora, Mexico; UA 69-35-25 ( 1:245) Guaymas, Sonora, Mexico;
Table 5. — Comparison of selected characters of Diodon holocanthus from five regions (see also
Figure 6). n = number of individuals examined for snout spine.
No. with
Frontal
Fin
rays (X)
Area
n
snout spine
spine/SL
D
P
Interorbital bar
Atlantic
58
52
0.146
14 15
22 15
Usually discontinuous
E. Pacific
11
0
0.154
1380
22 10
Usually continuous
Hawaii. Pitcairn. and Easter Is.
24
1
0.174
14 44
22.57
Usually continuous
W Pacific
29
0
0.205
14.11
22.17
Usually continuous
Indian
6
0
0.200
13.80
21.92
Usually continuous
564
LEIS: SYSTEMATICS AND ZOUGECHIKAl'HV OF THE Pt)KtlPINEFlSHES
UA 7 1-63-8 (1 : 1 45 ) Puerto Vallarta, Jalisco, Mexico; UA 7 1-65-9
(1:126) IslaJaltemba, Jalisco, Mexico; SIO 59-373 (l:ca. 200) La
Jolla, Calif.; SIO 63-82 (l:ca. 90) Cape Marco, Colum-
bia. HAWAIIAN IS.: HIMB (3:135-289), HIMB 67-58 (1:67)
Kaneohe Bay, Oahu; HIMB (1:181) Punaluu, Oahu; BPBM
10635 ( 1:63), BPBM 6977 (1:167) Diamond Head, Oahu; BPBM
5124 (1:129) French Frigate Shoals; NMFS H TC32-6,9,11,14
(6:12.5-30) 21°22'N, 158°14'W; NMFS H TC32-23 (1:14)
2r00'N, 158°30'W; NMFS H TC32-73 (1:7.0) 19°31'N,
156°06'W. SOUTHEAST PACIFIC: (all BPBM) 16459 (2:144-
168), 13251 (1:135), 16455 (1:122) Pitcairn I.; 6797 (1:150.5),
6798 ) 1:185), 6799 (1:158), 6800 ( 1: 156.5) Easter I. WESTERN
PACIFIC: GVF stn HK91 (2:85-109) 19"38'N, 111°30'E; GVF
2269 ( 1:128) Gulf ofThailand; CAS 29126 (1:32) Ternate, Moluc-
cas; CAS 6987 (1:41) Misaki, Japan; CAS 6752 (3:100-114)
Wakanoura Kii, Japan; CAS 53402 (1:225) Hachijo I., Japan;
CAS 15849 (10:90-125) Taiwan Strait. AUSTRALIA: AMS
1.17228-001 (10:67-91) New South Wales. INDIAN OCEAN:
RUSI 2782 (1:47.5) Knysna, South Africa; RUSI 3709 (1:60.5)
East Cape, South Africa; RUSI 3710 ( 1:65) Inhaca, Mozambique:
BPBM 19022 (2:173-188) Negombo, Ceylon; BPBM 20255
( 1:195) Wolmar, Mauritius. WESTERN ATLANTIC OCEAN:
CAS 4761 (1:150) Jamaica; CAS 54039 (1:94) Havana, Cuba;
CAS 18182 (2:50.5-57.5) 29°14'N, 88°19'W; CAS 17184 (1:91)
Pine I., Fla.;GCRLVTS:11184( 1:113) San Bias, Panama; LACM
1463 ( 1:84.5) Key Biscayne, Fla.; LACM 6281, 6282, 6283, 6284,
5781, 5872 (23:64-159) southern Jamaica; NMFS LJ Gill 3-64
(1:59) 33°29'N, 76^40 'W; NMFS LJ Silver Bay 3458 (1:60)
29°03'N, 78°04'W; NMFS M Oregon 77-72-39-144 (1:12.5)
23°34'N, 82°22'W, 39-73 (1:13) 21°31'N, 86°14'W, 39-50 (1:24)
16°50'N, 80°13'W, 39-48 (1:24.5) 17°26'N, 79°26'W, 39-58 ( 1:30)
21°01'N, 80°14'W, 39-63 (2:10-40) 19°41'N, 84°13'W, 39-01
(1:23) 13°00'N, 60°00'W, 39-39 (1:45) 18°00'N, 73°00'W, 39-11
(1:56.5) 17°25'N, 63°00'W; NMFS M Bowers 75-126-8 (1:28)
26°00'N, 79°30'W; NMFS M Oregon 77-76-66-19786 (2:23-32)
18°18'N, 75°22'W, 66-19789 (2:20-30) 18°49'N, 74°44'W, 66-
19790 (6:27-34) 19°22 'N, 75°44'W, 66-19791 ( 18:19-33) 17°50'N,
74°47'W.
Note. — Since this paper was accepted for publica-
tion, NMFS H and most HIMB specimens were
transferrred to BPBM.
ACKNOWLEDGMENTS
I am grateful to the following individuals for
information on and loans of specimens: D. G.
Smith, Marine Biomedical Institute, University
of Texas; M. M. Smith and R. Winterbottom,
RUSI; T. Potthoff, NMFS M; S. J. Karnella,
USNM; M. L. Bauchot, NMHN; J. Moreland, Na-
tional Museum of New Zealand; E. H. Ahlstrom
and B. Y. Sumida, NMFS LJ; P. M. Sonoda and
W. N. Eschmeyer, CAS; D. A. Thomson, UA; P.
Kahsbauer, NMV; J. M. Dixon, National
Museum of Victoria; A. C. Wheeler, BMNH; C. E.
Dawson, GCRL; D. F. Hoese and J. R. Paxton,
AMS; R. J. Lavenberg, LACM; R. H. Rosenblatt,
J. Copp, and J. Pulsifer, SIO; W. F. Smith-Vaniz,
ANSP; and A. Suzumoto, BPBM. Special thanks
go to J. C. Tyler (NMFS M) and C. Baer (HIMB)
for critically reading the manuscript and W. I.
Follett and L. Dempster (CAS) for valuable ad-
vice on nomenclature. J. E. Randall (BPBM) pro-
vided the initial stimulus for this study, and his
help and encouragement are gratefully acknowl-
edged. L. Y. Maluf (UA) gave valuable assis-
tance in the final stages of manuscript prepara-
tion. W. Waston (HIMB) drew Figure 2L D.
Hashimoto (HIMB) conducted the rearing exper-
iments that produced two of the metamorphosed
juveniles of D. holocanthus.
LITERATURE CITED
Artedi, p.
1738. Ichthyologia (Lugduni Batavorum). Pars III Genera
Piscium. 1962 reprint, Historae Naturalis Classica 15,
J. Cramer, Weinheim, p. 203-294.
BAGNIS, R., p. MAZELLIER, J. BENNETT, AND E. CHRISTIAN.
1972. Fishes of Polynesia. European Book Co., San
Franc, 368 p.
Bailey. R. M., J. E. Fitch, E. S. Herald, E. A. Lachner,
C. c. LiNDSEY, C. R. Robins, and W. B. Scott
1970. A list of common and scientific names of fishes from
the United States and Canada. 3d ed. Am. Fish. Soc.
Publ.6, 150p.
Bertelsen,E.
1951. The ceratioid fishes. Ontogeny, taxonomy , distribu-
tion and biology. Dana-Rep. Carlsberg Found. 39, 276 p.
Blanco, G. J., and D. V. Villadolid.
1951. The young of some fishes of Luzon. Philipp. J.
Fish. 1:67-93.
BLEEKER, P.
1865. Atlas Ichthyologique des Indes Orientales Neer-
landaises, Vol. V. Baudroies. Ostracions, Gymnodontes,
Balistes. Amst.,152p.
BLOCH.M.E.
1785. Naturgeschichte der Auslandischen Fische. 2 vol.:
1:1-136,2:1-160. Berl.
Bloch, M. E., and J. G. Schneider.
1801. Systema ichthyologiae. Berl., 584 p.
Breder, C. M., Jr., and E. Clark.
1947. A contribution to the visceral anatomy, develop-
ment, and relationships of the Plectognathi. Bull. Am.
Mus. Nat. Hist. 88:287-319.
Breder, C. M., and D. E. Rosen.
1966. Modes of reproduction in fishes. Nat. Hist. Press,
Garden City, N.Y., 941 p.
BRIS.SOUT DE BARNEVILLE, M. L.
1846. Note sur les diodoniens. Rev. Mag. Zool. Pure
Appl. 9:136-143.
Castelnau,F.
1872. Contribution to the ichthyology of Australia.
Proc. Zool. Acclim. Soc. Victoria 1:29-247.
Clark, E., and H. A. F. Gohar.
1953. The fishes of the Red Sea: Order Plectog-
nathi. Mar. Biol. Stn. Al Ghardaga, Publ. 8, 80 p.
Coleman, N.
1974. Australian marine fishes in colour. A. H. and A.
W. Reed, Sydney, 108 p.
565
FISHERY BULLETIN: VOL. 76, NO. 3
CURTISS, A.
1938. A short zoology of Tahiti. Priv. publ., 193 p.
CUVIER, M. G.
1818. Surlesdiodons, vulgairementorbesepineux. Mem.
Paris Mus. Natl. Hist. Nat. 4:121-138.
Day, F.
1878. The fishes of India; being a natural history of the
fishes known to inhabit the seas and fresh waters of In-
dia, Burma and Ceylon. Priv. publ., Lond., 778 p.
De Beaufort, L. F.
1962. The fishes of the Indo- Australian archipelago, Vol.
XI. E.V. Brill, Leiden, Neth., 481 p.
DeKAY,J.E.
1842. Zoology of New York; or. The New York fauna. In
Natural History of New York Geological Survey, Vol. I,
Part 4, Fishes, 415 p.
DUMERIL, A.H. A.
1855. Note sur un travil inedit de Bibron relatif aux pois-
sons plectognathes Gymnedontes (Diodons et Tetro-
dons). Rev. Mag. Zool. Pure Appl., 2d Ser., 7:274-282.
EgeR,W. H.
1963. An exotoxin produced by the puffer Aro^/jro^ his-
pidus, with notes on the toxicity of other plectognath
fishes. M.S. Thesis, Univ. Hawaii, Honolulu, 88 p.
ElGENMANN, C. H.
1885. A review of the genera and species of Diodontidae
found in American seas. Ann. NY. Acad. Sci. 3:297-311.
EK\L-\N,S.
1967. Zoogeography of the sea. Sidgwick & Jackson,
Lond., 417 p.
EYDOUX, J. F. T., AND F. L. A. SOULEYET.
1841. Voyage de la Bonite, 1836-1837. Zool. Vol.
I. Poissons p. 157-215. Paris.
Fowler, H.W.
1928. The fishes of Oceania. Mem. Bemice P. Bishop
Mus. 10, 540 p.
Fraser-Brunner, a.
1943. Notes on the plectognath fishes. — VIII. The clas-
sification of the suborder Tetraodontoidea , with a snyop-
sis of the genera. Ann. Mag. Nat. Hist., Ser. 11, 10:1-18.
GIBBS, R. H., AND B. B. COLLETTE.
1959. On the identification, distribution, and biology of
the dolphins, Coryphaena hippurus and C.
equiselis. Bull. Mar. Sci. Gulf Caribb. 9:117-152.
GOSLINE, W. a., AND V. E. BROCK.
1960. Handbook of Hawaiian fishes. Univ. Hawaii
Press, Honolulu, 372 p.
Grant, EM.
1972. Guide to fishes. Dep. Primary Ind, Brisbane, 472
P-
GUNTHER, A. C. L. G.
1870. Catalogue of the fishes in the British Mu-
seum. Vol. 8, 549 p. Lond.
1910. Andrew Garrett's Fische der Siidsee, Vol. IX. J.
Mus. GodefTroy 17:389-515.
HALSTEAD, B. W.
1967. Poisonous and venomous marine animals of the
world. Vol. 2: Vertebrates. U.S. Gov. Print. Off., 1070 p.
HERRE,A. W. C.T.
1924. Poisonous and worthless fishes. An account of the
Philippine plectognaths. Philipp. J. Sci. 25:415-511.
HOBSON,E.S.
1974. Feeding relationships of teleostean fishes on coral
reefs in Kona, Hawaii. Fish. Bull., U.S. 72:915-1031.
566
Hollister,G.
1934. Clearing and dyeing fish for bone study. Zoologica
(N.Y.) 12:89-101.
HUBBS, C. L., AND K. F. LAGLER.
1958. Fishes of the Great Lakes region. (Revised ed.)
Cranbrook Inst. Sci., Bull. 26, 213 p.
JENKINS. O. P.
1904. Report on collections of fishes made in the Ha-
waiian Islands, with descriptions of new species.
Bull. U.S. Fish. Comm. 22:417-511.
JORDAN, D. S., AND B. W. EVERMANN.
1891. The fishes of North and Middle America. Bull.
U.S. Natl. Mus. 47(2):1241-2183.
Jordan, D. S., and C. H. Gilbert.
1883. List of fishes collected at Panama by Captain John
M. Dow, now in the United States National
Museum. Proc. U.S. Natl, Mus. 5:373-378.
Jordan, D. S., and J. O. Snyder.
1902. A review of the gymnodont fishes of Japan. Proc.
U.S. Natl. Mus. 24:229-264.
KAUP.J. J.
1855. Ubersicht iiber die Species einiger Familien der
Sclerodermen. Arch. Naturgesch. 21( l):215-233.
Klunzinger, C.B.
1871. Synopsis der Fische des Rothen Meeres, II
Theil. Verb. Zool. Bot. Ges. Wien 21:441-688.
Lacepede,B. G. E.
1798. Histoire naturelle des poissons. Vol. 2. Plassan,
Paris, 632 p.
LeDanois, Y.
1954. Sur le dimorphisme sexuel des poissons de la
famille des Diodontides. C. R. Hebd. Seances Acad. Sci.
(Paris) 238:2354-2356.
1959. Etude osteologique, myologique et systematique
des poissons du sous-ordre des Orbiculates. Ann. Inst.
Oceanogr. (Monaco), New Ser., 36:1-273.
1961. Catalogue des types de poissons orbiculates du
Museum National d'Histoire Naturalle. II. Families des
Tetraodontidae, Lagocephalidae, Colomesidae, Diodon-
tidae et Triodontidae. Bull. Mus. Natl. Hist. Nat., Ser.
2,33:462-478.
Leis, J. M.
1977. Development of the eggs and larvae of the slender
mola, Ranzania laevis (Pisces: Molidae). Bull. Mar. Sci.
27:448-466.
Linnaeus, C.
1758. Systema Naturae. 10th ed. Holmiae, Impensis
Direct Laurentii Salvi, Tomus 1, 824 p.
Marshall, T.C.
1965. Fishes of the Great Barrier Reef and coastal waters
of Queensland. Livingston Publ. Co., Narberth, Pa.,
566 p.
Masuda, H., D. araga, and T. Yoshino.
1975. Coastal fishes of southern Japan. Tokai Univ.
Press, Tokyo, 379 p.
Meek, S. E., and S. F. Hildebrand.
1928. The marine fishes of Panama. Field Mus. Nat.
Hist., Zool. Vol. 15, Part 3, 1045 p.
MITCHILL, S. L.
1815. The fishes of New York, described and
arranged. Trans. Lit. Philos. Soc. N.Y. 1:355-492.
MITO.S.
1966. Fish eggs and larvae. [In Jpn.] In S. Motoda
(editor), Illustrations of the marine plankton of Japan,
Vol . 7 , 74 p. Soy 6-Sha, Tokyo.
LEIS: SYSTKMATICS AM) /.OOCKOC.KAl'H Y OF IHK I'OKCCPIXKFISHES
NlCHOI-S. J. T., AND C. M. BHKDKH, JH.
1927. The marine fishes of New York and southern New
England. Zoologica (N.Y.) 9:1-192.
NlSHIMl'RA.S.
1960. Some aspects of the natural history of porcupine
puffers migrating to the Japanese waters. I. Spawning
and migration. [In Jpn., Engl, summ.l Jpn. Ecol.
10:6-11.
Orsi,J. J.
1974. A check list of the marine and freshwater fishes of
Vietnam. Publ. Seto Mar. Biol. Lab. 21:153-177.
PARRA.D. A.
1787. Descripcion de diferentes plezas de historia natural,
las mas del rama maritimo. representadas en setenta y
cinco laminas. Havana, 543 p.
POLL,M.
1959. Expedition oceanographique beige dans les eaux
cotieres Africaines de I'Atlantique Sud. ( 1948-1949). Vol.
IV. Fascicule 3B. Poissons V. Teleosteen acanthoptery-
giens (deuxieme partie). Bruxelles, 417 p.
RAFINESQUE.C.S.
1810. Indiced'ittiologiaSiciliana, Messina, 70 p.
RANDALL, J. E.
1967. Food habits of reef fishes of the West Indies. Stud.
Trop. Oceanogr. (Miami) 5:665-847.
1968. Caribbean reef fishes. T. F, H. Publ., Inc., Nep-
tune City, 318 p.
Sanzo, L.
1930. Ricerche biologiche su material! raccolti dal Prof. L.
Sanzo nella campagna idrografica nel Mar Rosso della R.
N. Ammiraglio Magnaghi 1923-1924. VII. Plectog-
nathi. Mem. Com. Talassogr. Ital. 167:1-111.
SCOTT, E. O. G.
1957. Observations on some Tasmanian fishes: Part VIII.
Pap. Proc. R. Soc. Tasmania 91:145-156.
Scott, T. D.
1962. The marine and freshwater fishes of South Austra-
lia. Gov. Print, Adelaide, 338 p.
SHAW,G.
1804. General zoology or systematic natural his-
tory. Pisces 5(2): 1-463. Lond.
STEINDACHNER, F.
1866. Zur Fischfauna von Port Jackson in Austra-
lien. Sitzungsber. Akad. Wiss. Wien 53:424-481.
TORCHlo,M.
1963. Accertata presenza di un rappresentante della
famiglia Diodontidae in Mediterraneo (Osteichthyes Tet-
raodontiformes). Atti Soc. Ital. Sci. Nat. Mus. Civ. Stor.
Nat. Milano 102(3):l-7.
TR0SCHEL,F.H.
1847. Bericht iiber die Leistungen in der Ichthyolgie
wahrend des jahres 1846. Arch. Naturgesch. 13:343-
367.
waite, E. R.
1923. The fishes of South Australia. Gov. Print.,
Adelaide, 243 p.
Watson, W., and J. M. Leis.
1974. Ichthyoplankton of Kaneohe Bay, Hawaii. A one-
year study offish eggs and larvae. Univ. Hawaii, Hon-
olulu. Sea Grant Program TR-75-1, 178 p.
Whitley, G. P.
1931. Studies in ichthyology. No. 4. Rec. Aust. Mus.
18:96-133.
1933. Studies in ichthyology. No. 7. Rec. Aust. Mus.
19:60-112.
WOLFSCHEIMER,G.
1957. A spawning of procupine puffers. Aquarium
(Phila. 126:288-290.
567
PROBABLE CASE OF STREAMBED OVERSEEDING— 1967 PINK SALMON,
ONCORHYNCHUS GORBUSCHA, SPAWNERS AND SURVIVAL OF
THEIR PROGENY IN SASHIN CREEK, SOUTHEASTERN ALASKA
William R. Heard'
ABSTRACT
The 1967 escapement of 38,067 pink salmon, Oncorhynchus gorhuscha. to Sashin Creek, southeastern
Alaska, was the largest since 1942. Studies on distribution and density of spawners and freshwater
survival of their progeny indicated that deposition of excessive numbers of eggs caused a severe
compensatory mortality of alevins during winter. Spa wner density was 1.7, 1.6, and 1.2 females/m^ in
upper, middle, and lower study areas respectively. The greater density of spawners in the upper area in
the odd-numbered years may be determined by genetic factors like timing of escapements and by
greater marine survival of fry from the upper area. Based on the previously consistent relation between
timing of adult entry and resulting freshwater survival, 1967 spawners should have produced 8 million
fry rather than the 3 million that were produced.
Mortality of eggs and alevins was high during spawning, low between spawning and hatching, and
high between hatching and emergence. Between 1 December 1967 and 25 March 1968, 11.1 million
eggs or alevins, 10.7 million of which were alive on 1 December, disappeared within the streambed.
Initial mortality of these progeny probably occurred in the early alevin stage from oxygen privation,
whereas disappearance was probably related to rapid decomposition and invertebrate scavenging. A
"snowball effect" is postulated whereby alevins that die shortly after hatching place increasing
demands on available oxygen, causing accelerated mortality. A review of historical patterns of fry
production in Sashin Creek indicates that streambed overseeding occurred in 1967.
Studies of pink salmon, Oncorhynchus gorbuscha,
in Sashin Creek, Baranof Island, southeastern
Alaska, have shown that certain factors markedly
affect freshwater survival. These factors include:
1) seasonal timing of spawning (Skud 1958); 2)
density and distribution of adults on the spawning
grounds relative to ecological characteristics of
the stream, especially gradient (Merrell 1962);
and 3) quality of the intragravel environment,
including oxygen content of intragravel water and
amount of silt and fine particulate material in
streambed gravels (McNeil 1966, 1968). Other fac-
tors of significance, but believed to be of less
influence on freshwater survival in Sashin Creek,
include predation on eggs and alevins (McLarney
1967), stream discharge during spawning (Ellis
1969), and incubation (McNeil 1968).
The spawning ground of Sashin Creek extends
from the head of tidewater to an impassable falls
1,200 m upstream and includes 13,629 m^ of
streambed. Ninety-six percent (13,084 m^) of this
ground comprises three distinct ecological areas
'Northwest and Alaska Fisheries Center Auke Bay Labora-
tory, National Marine Fisheries Service, NOAA, P.O. Box 155,
Auke Bay, AK 99821.
Manuscript accepted February 1978.
FISHERY BULLETIN: VOL. 76. NO. 3. 1978.
that differ in gradient and size of particles in the
substrate. McNeil (1966) called the areas upper,
middle, and lower and described them briefly as
follows: upper (2,945 m^) — relatively steep gra-
dient (0.79'f ) and coarse streambed gravel; middle
(4,067 m^) — intermediate gradient (0.3*^) and
medium-sized streambed gravel; and lower (6,072
m^) — low gradient (0.1%) and relatively fine
streambed gravel. The remaining 4% (545 m^) of
spawning ground is located in a short section of
stream between the counting weir and the lower
area and is not treated in this paper.
Pink salmon spawners entering Sashin Creek
(the escapement) have been counted at a weir at
the mouth of the creek since 1934, and the result-
ing numbers of fry from these escapements have
been determined since 1940. During this time, the
number of spawners varied from as few as 8 to
more than 90,000 and the number of fry produced
varied from 50 to almost 6 million. The percentage
of freshwater survival, based on the estimated po-
tential egg deposition, ranged from 0.06 to 21.75%
(Table 1).
The high escapement of 38,067 pink salmon
spawners in 1967, following a long series of rela-
tively low escapements, gave me an opportunity to
569
FISHERY BULLETIN: VOL 76, NO. 3
Table l. — Number of adult pink salmon, potential egg deposi-
tion, number of fry produced, and freshwater survival in Sashin
Creek, 1934-67. (Modified from McNeil 1968.1
Brood
year
Number
of
adults
Potential
egg
deposition'
Number
of fry
produced
Percentage
freshvi^ater
survival
1934
7.917
—-
—
—
1935
6.323
—
—
—
1936
5.364
—
—
—
1937
9,085
—
—
—
1938
6.467
—
—
—
1939
16,830
—
—
—
1940
53.594
52.858.000
3.399,900
8.43
1941
84,303
88.678.000
1.024.300
1.16
1942
92.085
78.894.000
674.000
0.85
1943
14,883
14.980,000
227.800
1.52
1944
4,050
3,904,000
105.600
2.71
1945
5,465
5,062,000
43.100
0.85
1946
933
736.000
1.200
0.16
1947
1,486
1.330.000
27.600
2.07
1948
597
516.000
9.100
1.76
1949
4,902
4,800,000
176.200
3.67
1950
112
86,000
50
0.06
1951
4,366
4,062,000
412,500
10.15
1952^
45
—
740
—
1953
1,164
1.284.000
95.400
7.43
1954
21
12.000
660
548
1955
9,267
10,286,000
266,200
12,31
1956
933
1.018.000
5,040
050
1957
2.834
2.588.000
562,900
21,75
1958
217
174.000
10.700
6 13
1959
35.391
40.379,000
5.332.400
13.21
19602
162
—
480
—
1961
28.759
29,425,000
5,940,300
20.19
1962^
8
8,000
100
1 20
1963
16,757
16,640,000
3.256.300
19-57
1964
^2.193
2,230,000
■"SIO.OOO
1391
1965
14,833
12,668,000
"2.235.000
17.92
1966
5,761
6.255.000
"744,000
11.99
1967
38.067
44.384.000
3,007,200
6.78
'Based on 2,000 eggs female except when actual fecundity was calculated
^An attempt was made to destroy the spawners or their progeny.
^Natural returning adults (327) were supplemented by the introduction of
1.866 adults taken from Bear Harbor. Kuiu Island
"Fry weir not operated: figures are estimates of live alevins in the gravel |ust
before start of emergence
study the effects of a large spawning population on
freshwater survival. For the 1967 escapement I
studied 1 ) timing of entry into the stream and the
distribution and density of pink salmon on the
spawning grounds, and 2) survival of progeny by
time periods in the three ecological areas of the
stream and the overwinter disappearance of eggs
and alevins from streambed gravels. In this paper
I present all the available data on escapament size
and production of fry in Sashin Creek and develop
the hypothesis that streambed overseeding occur-
red in 1967. As Ricker (1962:186) pointed out,
detailed knowledge on the effects of overseeding is
important in understanding why pink salmon
populations fluctuate. He stated, "Because it
[overseeding] happens rarely nowadays, no
chance should be lost to make such a study if one
occurs." Simply stated, overseeding can be defined
as an egg density in spawning bed gravels that
leads to a significantly greater freshwater mortal-
ity than a lesser density would cause. As discussed
more fully later, it is a complex and dynamic in-
teraction between egg density, streambed ecology,
and specific climatic conditions.
TIMING OF ENTRY AND
DISTRIBUTION AND DENSITY
OF SPAWNERS
The timing of stream entry was determined
from daily counts of the adults at the Sashin Creek
weir. This timing apparently influences the
freshwater survival of the progeny. An inverse
relation between time of stream entry of spawners
and survival of progeny is usual in Sashin Creek
(Skud 1958; Merrell 1962; McNeil 1968; Ellis
1969): high survival has been associated with
early spawning and low survival with late spawn-
ing. Merrell ( 1962) further pointed out that pink
salmon spawn in Sashin Creek an average of 12
days earlier during odd years than even years. ^ As
a result, freshwater survival is usually higher
among progeny of spawners from odd years than
among those from even years (Table ll.
In 1967, 507f of the spawners had entered
Sashin Creek by 20 August, the second earliest
date on record. The early entry indicated that sur-
vival of eggs and alevins would be high, but this
did not prove to be the case.
Throughout the run, random lots of females
were tagged at the weir, and the distribution and
density of spawners were determined from daily
counts of both tagged and untagged females on the
spawning grounds. This technique, described by
McNeil (1968) and used by Ellis (1969), provides
two methods of estimating the numbers of females
spawning in the upper, middle, and lower areas.
One method assumes that tagged females distrib-
ute themselves among the three sections the same
as untagged females. In the other method, the
summed daily count of all females in each area is
divided by the average longevity on the spawning
grounds. The results from the two methods were
generally in agreement, except for the upper area
where estimates based on distribution of tagged
females were considerably higher than those
based on total females. The difference may reflect
the difficulty in making accurate counts on spawn-
ing riffles where densities of spawners are high; in
such a situation an observer might count small
^The date when 50% of the escapement to Sashin Creek had
entered the stream has been commonly used as an index of time
of spawning.
570
HEARD: PROBABLE CASE OF STREAMBED OVERSEEDING
numbers of tagged females more accurately than
large numbers of untagged females. Because the
relative accuracy of the two methods is unknown, I
averaged them to arrive at mean estimates of den-
sities of females in the three areas (Table 2).
Spawner density in Sashin Creek is usually un-
equal in the three study areas, depending in part
on the total number of spawners. Densities in 1967
were the highest recorded for specific areas-'^ of the
stream (Table 3). Merrell (1962) noted that in
years when many spawners were present, they
utilized all of the available spawning grounds, and
in years when few were present, they spawned
mostly in the lower portion of the stream. When
the upper area was used, survival of eggs and
alevins in that area was higher and the number of
fry produced was proportionally much greater
than in the middle and lower areas (Merrell 1962).
In addition, the sediment content and water qual-
ity of the stream in the upper area were better
than in the other two areas (McNeil 1966, 1968).
Sashin Creek thus presents an apparent
paradox — the least favorable areas are used in
years of relatively few spawners, and the best
areas are used only during years of great numbers
of spawners.
Merrell ( 1962) thought that che greater use of
the upper area was related primarily to density-
dependent spawner interactions. In 1967, how-
ever, the heavy use of the upper area was appar-
ently not the result of high densities downstream
forcing spawners into upstream areas: spawner
^Although the total number of spawners entering the stream
has been recorded since 1934 (Table 1), detailed studies on the
distribution of spawners in the upper, middle, and lower areas of
the stream have been available only since 1961.
Table 2. — Estimated densities of female pink salmon spawning
in three areas of Sashin Creek, 1967.
Females per square meter
Area
Based on counts of
tagged females only
Based on counts of tagged
and untagged females
Mean
Upper
Middle
Lower
1.90
1.49
1.19
1.59
1.76
1.15
1.74
1.62
1.17
Tables.-
—Estimated densities of female pink salmon spawning
in three areas of Sashin Creek, 1961-67.
Females per square
meter
Area
1961'
19632
1 9643 1 9654
19665
1967
Upper
Middle
Lower
1.00
1.00
1.00
0.59
089
059
0-01 0 58
0.09 062
0.13 0.44
0.04
0.27
0.28
1.74
1.62
1.17
'Extrapolated from subjective estimate (McNeil et al. 1964).
^Adjusted from McNeil (1966).
^McNeil et al (1969).
"McNeil (1968).
^Ellis (1969).
densities in the upper area built up rapidly before
spawning reached significant levels in the middle
and lower areas. Although the upper area contains
only 22'7f of the combined spawning grounds of the
three areas, in 1967, 62^/ of the first group of
female pink salmon tagged at the weir spawned in
the upper area (Table 4). In general, the intensity
of spawning in 1967 progressed to downstream
areas from the upper area rather than the reverse.
McNeil (1966, 1968) noted similar downstream
shifts in spawning in Sashin Creek in 1963 and
1965. Although McNeil (1966) felt that the shift
occurred because of heavy rainfall during the
spawning period, he later noted ( McNeil 1968) the
same phenomenon during an unusually dry year.
It appears that the upper area in Sashin Creek is
not necessarily used because of spawner overflow
but because of more complex factors. Two interre-
lated factors could account for the spawner dis-
tributions observed in recent years: 1) migratory
behavior associated with timing of the escape-
ment, and 2) a genetic tendency for odd-year
spawners to use upstream areas. Odd-year spawn-
ers enter the stream earlier than even-year spawn-
ers. A characteristic of early stream entry in
anadromous fishes may be a tendency to migrate
farther upstream than spawners associated with
late stream entry (Briggs 1955). In addition to
early entry and use of the upper area, odd-year
spawners for the past 9 or 10 generations have
consistently had higher escapements and, except
for 1967, higher freshwater survival of progeny
than even-year spawners (Table 1). Natural selec-
tion may be operating, in recent odd-year genera-
tions, to encourage progeny produced in the upper
area to spawn in that area. Wells and McNeil
( 1970) showed that fry produced in the upper area
of Sashin Creek were larger and presumably of
better quality than those produced in the
downstream areas. Differential marine survival
Table 4. — Dates of tagging and percentage of total escapement
counted through weir, numbers of female pink salmon tagged,
and spawning distribution of tagged females in three areas of
Sashin Creek, 1967.
Percentage of
total escape
Tagged
females
Percentage of tagged
fisti accounted for
Date of
ment
cour
ited
Females
observed
Upper
Middle
Lower
tagging'
througfi weir
tagged
spawning
area
area
area
10 Aug.
3
49
40
62
25
12
12 Aug.
13
50
40
22
37
40
17 Aug
26
50
42
21
23
55
20 Aug.
54
50
50
22
42
36
5 Sept.
98
50
40
22
30
48
'Females tagged on each date received color-coded tags that differentiated
them from females tagged on other dates.
571
FISHERY BULLETIN; VOL. 76, NO. 3
that favored fry produced in the upper area over
those produced in the downstream areas could ac-
count for the greater escapements of odd-year
spawners in recent years.
SURVIVAL OF EGGS AND ALEVINS
Survival of eggs and alevins from the 1967 brood
year was estimated in Sashin Creek for four time
periods: 1) from stream entry to end of spawning,
2) from end of spawning to hatching, 3) from
hatching to shortly before fry emergence, and 4)
from shortly before emergence to emergence and
downstream migration of fry.
The estimates of survival were based on esti-
mates of the potential egg deposition of female
spawners and estimates of the surviving eggs and
alevins in the three study areas. Potential egg
deposition was estimated by multiplying the
number of females by average fecundity. Densities
of eggs and alevins were determined after spawn-
ing, during hatching, and before fry emergence by
sampling randomly selected 0.1-m- points in the
streambed with a hydraulic sampling technique
described by McNeil (1964a). The number of fry
migrating from the stream were estimated on the
basis of daily counts of fry migrating through a
weir at the stream mouth.
Numbers of females entering the stream and
average fecundity were derived from counts and
samples taken at the weir. Of the 38,067 pink
salmon spawners entering Sashin Creek in 1967,
19,639 (52'7f ) were females. Total counts of mature
eggs from each of 35 females selected at random
from the run ranged from 810 to 2,954 (average
2,260) eggs/female (907r confidence limit of mean
fecundity was ±115 eggs).
The percentage of eggs available for deposition
that are actually buried in the streambed is partly
dependent on the density of spawners. McNeil
( 1964b) discussed the role of redd superimposition
and showed that at spawner densities approaching
3 to 4 females/ m^ of spawning ground, an upper
asymptotic limit on the density of eggs in the
streambed is reached. Factors other than spawner
density that may influence egg deposition include
loss of adults in the stream before spawning and
retention of eggs in the female's body (Neave
1953), type and characteristics of the spawning
substrate (McNeil 1966), streamflow during
spawning (Ellis 1969), and loss of eggs to verte-
brate predators during the spawning process
(Moyle 1966; McLarney 1967; Reed 1967).
The efficiency of egg deposition of pink salmon
spawners in Sashin Creek is highly variable, from
37 to 829^^ of the potential egg deposition (Ellis
1969). In 1967 the number of pink salmon eggs
potentially available for deposition was 44.4 mil-
lion, with 19.9 million of these (459^ of the poten-
tial) estimated to be in the streambed after spawn-
ing. The efficiency of egg deposition was 47'7f in the
upper area, 50% in the middle area, and 387f in the
lower area.
Although spawner densities were high in 1967
(Table 3), the ability of pink salmon to void most of
their eggs during spawning did not seem to be
affected. Egg retention is characteristically low in
Sashin Creek, usually less than 57c of fecundity
( McNeil 1966; Ellis 1969). In 1967, 1 examined the
body cavities of 402 spent female pink salmon
(about 2'7c of the total) and found that average egg
retention was 1.59^ of average fecundity.
The proportion of eggs actually deposited that
were alive at the end of the spawning period in
1967 was highest (93%) in the upper area, inter-
mediate (83%) in the middle area, and lowest
(74% ) in the lower area (Table 5). This high survi-
val in the upper area is consistent with that of
previous years. The ratio of live to combined live
and dead eggs and alevins was usually higher in
the upper and middle areas than in the lower area
through hatching to the beginning of fry
emergence (Table 5).
Survival of eggs and alevins varied among the
three time periods (during spawning, between end
of spawning and hatching, and between end of
hatching and emergence). Survival within each
time period for each area was higher between
spawning and hatching than during spawning or
between hatching and emergence (Table 6). As
previously discussed, survival during spawning
was related primarily to the ability of females to
successfully deposit their eggs because a high per-
centage of the eggs buried were alive shortly after
spawning. Survival between spawning and hatch-
ing and between hatching and emergence pertains
to survival of eggs and alevins within the
streambed.
The densities of live preemerged fry in the
streambed of Sashin Creek in late March 1968
were 382, 260, and 108/m- in the upper, middle,
and lower areas, respectively. From these den-
sities I estimated a population of 2.9 million fry in
the entire stream. Operation of the fry weir began
just after the late March streambed sampling was
572
HEARD: PROBABLE CASE OF STREAMBED OVERSEEDING
Table 5. — Potential egg deposition, number of live and dead eggs and alevins, ratio of live to combined live and dead eggs and
alevins, and estimated survival of 1967 brood year pink salmon in three areas of Sashin Creek.
Area
Potential egg deposition
per square meter
Mean
3,947
3,672
2.644
90°o confidence
limits of mean
Period
beginning
10 Aug
and ending
Combined live and dead eggs
and alevins per square meter
QO^o confidence
Mean limits of mean
Percentage of live to combined
live and dead eggs and alevins
Mean
90°o confidence
limits of mean
Percentage
calculated
survival
Upper
Middle
Lovi/er
::;201
:187
:136
1 Oct.
1.863
±254
93
± 1
43
1 Dec.
1.714
±295
86
± 4
37
25 Mar.
647
±138
59
±13
10
1 Oct.
1,826
±218
83
±12
41
1 Dec.
1,591
±226
70
± 7
30
25 Mar.
702
±147
37
± 2
7
1 Oct.
1,015
±120
74
±17
28
1 Dec.
989
±116
72
± 2
27
25 Mar.
350
±70
31
±10
4
Table 6. — Percentage of estimated survival of 1967 brood year
pink salmon eggs and alevins for three time periods in three
areas of Sashin Creek and for the entire stream, 1967.
Percentage
survival
Between end of
Between end of
During
spawning and
fiatching and
Area
spawning
fiatcfiing
emergence
Total
Upper
43
85
26
10
Middle
41
73
23
7
Lower
28
95
15
4
Entire stream'
37
83
22
68
' Data weigfited and adjusted to include spawning grounds not included in tfie
three study areas
completed. Relatively few fry migrated
downstream through the weir until mid-April; the
daily fry migrations increased steadily through
late April, reached a peak in early May, then de-
clined rapidly, and were essentially completed by
early June (Figure 1). The total number of pink
salmon fry estimated to migrate from Sashin
Creek from the 1967 brood year spawners was 3
million. Similar close agreements between esti-
mates based on densities of preemerged fry and
those based on number of fry counted at the weir
have occurred in previous years (McNeil 1968).
The 3 million fry migrating from Sashin Creek
in the spring of 1968 represent a total freshwater
survival of 6.8'''!^ of the 44.4 million potential egg
deposition. This is the lowest freshwater survival
in the odd-year line of pink salmon spawners in
Sashin Creek since 1949 (Table 1). I will sub-
sequently attempt to show that this reduced sur-
vival was primarily due to excessive seeding of the
streambed during spawning.
DISAPPEARANCE OF EGGS
AND ALEVINS
To determine the number of eggs and alevins
that disappeared from the streambed, I compared
the potential egg deposition with the numbers of
live and dead eggs at the end of spawning and the
number of eggs and alevins at the time of hatching
and just before emergence. In 1967, 55% of the
potential egg deposition disappeared during
spawning. The fate of these eggs is unknown, but
they were probably removed from the stream dur-
ing the spawning period by predators, scavengers,
or turbulent streamflow. McLarney (1967) and
McNeil ( 1968) discussed the roles offish predators
(especially sculpins) and water turbulence in re-
moving eggs from Sashin Creek during spawning
and between spawning and hatching.
McNeil ( 1968) found that eggs and alevins of the
1963 and 1965 brood years disappeared at differ-
ent rates in the upper, middle, and lower study
areas of Sashin Creek. Most of the 1963 brood year
progeny disappeared during spawning, and most
of the 1965 brood year progeny disappeared be-
tween hatching and emergence (over the winter). I
will examine closely the possible fate of eggs and
alevins during this period (December to March)
because the factors that caused a reduced freshwa-
ter survival of 1967 brood year progeny prevailed
during this period.
The estimated percentages of the potential egg
deposition that disappeared in the upper, middle,
and lower areas of Sashin Creek were similar
within each of the three periods. This disappear-
ance varied greatly between periods: 55% of the
progeny (eggs or alevins) had disappeared by 1
October, 4% between October and December, and
25% between December and March (Table 7). The
disappearance between hatching and emergence
(December and March) appears more significant
when expressed in terms of numbers present in
December; 56-65% of the eggs and alevins in the
upper, middle, and lower areas of Sashin Creek on
1 December had disappeared by 25 March (Table
8).
573
FISHERY BULLETIN: VOL. 76. NO. 3
Figure l. — Daily number of 1967
brood year pink salmon fry counted
through Sashin Creek weir in spring
1968.
i
'
^ 364,000
325
-
300
-
275
-
250
-
Q 225
z
<
c/)
O 200
-
u.
—
NUMBER OF
tn O
-
100
-
75
-
50
-
25
-
{III,
1
lllli... ......
u^ I •■■|"'"|""|""|' i'|i
II 1 1
25 5 15 25
MARCH APRIL
15
MAY
25
5 15
JUNE
Mortality Patterns in the Streambed
Mortality of eggs and alevins within the
streambed at Sashin Creek is evident in two ways:
1 ) as a reduction in the total population of eggs and
alevins within the streambed, i.e., they disappear,
and 2) as an increase in the number of dead eggs
and alevins in the streambed and a decrease in the
number of live eggs and alevins. In the first in-
stance, some factors that can cause eggs and ale-
vins to disappear are turbulent streamflow,
streambed scouring, predation, and scavenging.
Excluding predation, these same factors can also
cause dead eggs and alevins that may have died for
other reasons to disappear from the streambed. In
the second instance, factors causing an increase in
the number of dead eggs and alevins within the
streambed are generally relatable to desiccation,
freezing, and the quality of intragravel water.
In addition, dead eggs and alevins may disap-
pear because of at least two factors that do not
affect live eggs and alevins: biochemical decom-
position and consumption by intragravel inver-
tebrate scavengers. Thus, factors that cause eggs
574
HEARD: PROBABLE CASE OF STREAMBED OVERSEEDING
Table 7. — Percentage of potential egg deposition of 1967 brood
year pink salmon that disappeared from three areas of Sashin
Creek and from the entire stream by 1 October 1967, 1 December
1967, and 25 March 1968.
Estimated percentage ot
potential egg deposition disappearing
Area
By 1 Oct.
1 Oct.
to 1 Dec. 1
Dec. to 25 Mar.
Upper
Middle
Lower
Entire stream'
53
50
62
55
4
6
1
4
27
24
24
25
' Data weighted and adjusted to include spawning grounds not included in the
three study areas.
T.\BLE 8. — Estimated densities of all eggs and alevins (live and
dead) in three areas of Sashin Creek on 1 December 1967 and 25
March 1968, and percentage that disappeared between the two
dates.
Number of eggs and alevins
per square meter on
Area
1 Dec.
25 Mar.
Percentage that
disappeared
between dates
Upper
Middle
Lower
1,714
1,591
989
647
702
350
62
56
65
or alevins to disappear from streambed gravels
may or may not have been the initial cause of
death.
It is unlikely that turbulent streamflow,
streambed shifting, or predation were the reasons
that 1967 brood year eggs or alevins disappeared
between early December and late March,
Streamflows were generally low, and although an
intermittent ice cover was present on Sashin
Creek during January, February, and March,
there was no indication of streambed shifting be-
cause of ice scouring. A series of short metal stakes
driven into the streambed throughout the stream
in November to mark coho salmon, O. kisutch,
redds was still in place in March, indicating that
no streambed shifting had occurred.
Most fish in Sashin Creek that could eat pink
salmon eggs and alevins (juvenile coho salmon;
rainbow trout, Salmo gairdneri; Dolly Varden,
Salvelinus malma\ and coastrange sculpin, Cottus
aleuticus) are essentially dormant during the
winter when water temperatures are low. Chap-
man and Bjornn ( 1969) have shown that resident
stream salmonids may disappear into the sub-
strate when water temperatures fall below 4.4 °-
5.5 °C. I have observed similar behavior in Sashin
Creek. Stream temperatures in Sashin Creek
were below 4.4 °C from 13 December 196-7 to 20
April 1968.
Water ouzels, Cinclus mexicanus, and mergan-
sers, Mergus merganser, are occasionally present
on Sashin Creek during the winter and could ac-
count for the disappearance of some eggs and ale-
vins during open water periods. In considering the
magnitude of the disappearance of 1967 brood
year eggs and alevins in Sashin Creek between
December and March, it is unlikely that the
maximum possible loss to these sources is sig-
nificant. This conclusion is based on the small
numbers of mergansers and ouzels present, the
amount of time the stream was covered with ice,
and the large number of eggs or alevins that dis-
appeared. Between two and five mergansers were
noted in the vicinity periodically. When present,
these birds spent much of their time in the inter-
tidal portion of Sashin Creek or in the adjacent
estuary. Only four of the smaller and territorial
ouzels normally occur along the upper, middle,
and lower areas of Sashin Creek in winter, and
during periods of ice cover these birds go
elsewhere. Based on periodic observations and
temperature records, I estimate the stream was
covered with ice approximately half the 1967-68
winter.
The estimated population of live and dead pink
salmon eggs and alevins in Sashin Creek was 18.3
million on 1 December 1967 and 7.2 million on 25
March 1968 (Table 9), making a loss of 11.1 mil-
lion eggs and alevins between the two dates. Be-
cause there is little evidence that the loss was
caused by external factors that physically re-
moved eggs or alevins from the streambed, the loss
was likely due to factors within the intragravel
environment.
The disappearance of 11.1 million eggs and ale-
vins from the streambed between hatching and
emergence led me to examine the relation between
live eggs and alevins and dead eggs and alevins in
the streambed. The densities of dead eggs and
alevins in the upper, middle, and lower areas (Ta-
ble 10) indicated that the numbers of dead eggs
and alevins remained relatively stable between
the time periods. This does not necessarily indi-
cate that dead eggs in the streambed during one
sampling period were still there during a later
sampling period. Dead eggs can disappear at any
time for many reasons, but can persist in a
T.ABLE 9. — Estimated population of live and dead pink salmon
eggs and alevins in Sashin Creek on 1 October 1967, 1 December
1967, and 25 March 1968,
Millii
ons
of
eggs
and alevins in streambed
Sample date
Live
Dead Total
1 Oct. 1967
1 Dec. 1967
25 Mar. 1968
16.5
13.7
3.0
34 19.9
46 18.3
4.2 7.2
575
FISHERY BULLETIN: VOL. 76, NO. 3
streambed for as long as 18 mo ( McNeil et al. 1964).
Once hatching is completed, no new dead eggs can
be added to the streambed. Because hatching of
live eggs was well underway on 1 December ( about
35% completed), many of the dead eggs present in
March had already died by 1 December (Table 10).
Most of the eggs and alevins that disappeared
over the winter (about 11 million; Tables 7, 8, 9)
were individuals that had been alive on 1 De-
cember because the number of dead eggs and ale-
vins was essentially unchanged from December
(4.6 million) to March (4.2 million) (Table 9). Mor-
tality (in the form of disappearance) of live eggs
and alevins in the streambed between 1 December
and 25 March was 74% in the upper area, 77% in
the middle area, and 85% in the lower area (Table
11). Of the 11.1 million pink salmon eggs and
alevins that disappeared within the streambed be-
tween 1 December and 25 March, 10.7 million
were alive on 1 December.
As previously mentioned, the cause of the dis-
appearance of dead eggs and alevins in the
streambed may differ from the cause of their
deaths. This apparently occurred with the 1967
brood year pink salmon progeny in Sashin Creek,
and I offer the following theoretical sequence to
explain the major overwinter disappearance of
eggs and alevins.
The greatest number of fry produced in Sashin
Creek since 1940 was 5.9 million (Table 1). On 1
December 1967, 13.7 million live pink salmon
eggs and alevins were in the Sashin Creek
streambed (Table 9), a number that appears to
Table 10. — Estimated densities of dead pink salmon eggs and
alevins in three areas of Sashin Creek on 1 October 1967, 1
December 1967, and 25 March 1968.
Dead eggs
and alevins per square meter
Upper area
Eggs Alevins
Middle area
Loviier
area
Date
Eggs Alevins
Eggs
Alevins
1 Oct, 1967
1 Dec. 1967
25 Mar. 1968
129 0
223 7
196 69
310 0
459 18
334 108
264
266
199
0
11
43
Table ll. — Estimated densities of live pink salmon eggs and
alevins in three areas of Sashin Creek on 1 December 1967 and
25 March 1968 and disappearance of live eggs or alevms between
the two dates.
Area
Upper
Middle
Lovi/er
Live eggs and
alevins per square meter
1 Dec. 1967 25 Mar 1968
Alevins Eggs Alevins
Eggs
899
769
463
Percentage of live
eggs or alevins that
disappeared between
dates
exceed the capacity of the streambed for pink
salmon fry production. I postulate that the high
initial density of eggs led to a severe mortality of
embryos in the early alevin stage, probably be-
cause of widespread oxygen privation or a combi-
nation of oxygen privation and a buildup of toxic
metabolites. The rate of oxygen consumption by
embryos increases steadily with development
(Wickett 1954, 1962) and coincides with the gen-
eral lowering of streamflows during the late fall,
followed by stabilization of streamflows at near
the normal winter levels.^ This combination of
conditions permitted the embryo population to
survive up to, but not much beyond, the hatching
period. These recently hatched dead alevins then
apparently disappeared rapidly within the
streambed through the combined action of
biochemical decomposition and intragravel inver-
tebrate scavenging. As I will show later, the rapid
disappearance of recently hatched dead alevins in
the streambed seems consistent with this
hypothesis.
Although no intragravel water quality data are
available from Sashin Creek during or shortly
after hatching to support the above theory, a com-
parison of the rates of oxygen consumption by pink
salmon embryos of various ages indicates that
oxygen requirements do steadily increase during
the hatching period. The rates of oxygen consump-
tion reported for early stage eggs (7-26 days old)
have ranged from 0.0003 mg O2 /egg per h ( Wickett
1954) to 0.0005 mg 02/egg per h (Brickell 1971).
Brickell found that the rate of oxygen consump-
tion by 35-day-old pink salmon eggs was 0.0018 mg
Oj/egg per h, almost four times the rate he mea-
sured for 7-day-old eggs. Faintly eyed 38-day-old
eggs had an oxygen consumption rate of 0.002 mg
Og/egg per h (Wickett 1962) while 7-day-old ale-
vins had a consumption rate of 0.01 mg 02/alevin
per h (Wickett 1954). x
575
345
249
382
260
108
74
77
85
""Seasonally, stream discharge in Sashin Creek is usually
highest in fall and lowest in summer. Discharge in winter
months may also be low, but is normally above summer levels.
Because unseasonably low winter discharge could reduce oxygen
delivery to embryos below the normal seasonal pattern, I com-
pared the low monthly discharge during December, January,
February, and March for 1967-68 with low discharge patterns in
the same months for the period 1951-52 to 1966-67. The low
mean monthly discharge from Sashin Creek during December,
January, February, and March ranged from 18 to 62 ft^/s and
averaged 33 ft^/s for the 16-yr period. The mean minimum
monthly discharge during these same 4 mo in 1967-68 was 30
ft^/s (U.S. Geological Survey 1969), suggesting that low
streamflow levels during these months in 1967-68 were near
normal.
576
HEARD: PROBABLE CASE OF STREAMBED OVERSEEDING
In addition to the increasing oxygen require-
ments due to growth and development of Hve em-
bryos, Brickell (1971) found that rates of oxygen
consumption by dead intact pink salmon eggs ex-
ceeded those of early stage live eggs fourfold:
0.0018 mg Og/whole dead egg per h versus
0.0004-0.0005 mg Oa/T-day-old live egg per h. He
noted even gi'eater oxygen consumption for dead
eggs when the chorion was pierced or slit or the
egg was fragmented: mean oxygen consumption of
fragmented dead eggs in constant-flow cylinders
was 0.017 mgOs/eggper h, which exceeds the rate
Wickett (1954) found for 7-day-old live alevins. It
follows that alevins that die shortly after hatch-
ing, because of their soft, exposed, and readily
oxidizable tissue, would have higher rates of oxy-
gen consumption than whole intact dead eggs, live
eggs, or early stage live alevins.
These increases in oxygen consumption upon
death of developing pink salmon embryos are the
rationale for suggesting a "snowball effect" —
rapidly increasing deaths of embryos once lethal
oxygen concentrations were approached. With
high densities of live embryos already placing ex-
cessive demands on the oxygen and each death
increasing the demand, the resulting heavy mor-
tality could have caused fry production to plunge
below that expected from lower initial egg
densities — an excellent example of Neave's ( 1953)
theory of compensatory mortality.
Disappearance of Dead Eggs Versus
Disappearance of Dead Alevins
To test the hypothesis that dead alevins disap-
pear within the streambed more rapidly than dead
eggs, I conducted a small study in Sashin Creek in
the winter of 1968-69 to consider the relative per-
sistence of dead eggs and alevins in the streambed.
A series of Vibert boxes (small plastic perforated
containers), each containing a mixture of
streambed gravel, 10 dead eggs, and 10 dead ale-
vins (all from 1968 brood year pink salmon) were
buried in Sashin Creek on 14 December 1968. The
boxes were buried about 20.3 cm deep across a
riffle in the middle study area. At irregular inter-
vals, pairs of the boxes were removed from the
streambed and the contents were preserved for
examination.
Alevins disappeared from the Vibert boxes at a
much faster rate than eggs (Table 12). Fewer than
half of the original number of alevins were still
recognizable at the end of 2 wk; after 37 days only
Table 12. — Contents of Vibert boxes with dead pink salmon
eggs and alevins buried in Sashin Creek streambed between 14
December 1968 and 14 April 1969.'
Eggs
recovered
Alevins
recovered
Invertebrates recovered
No of days
buried
Insect larvae^
Planarian
worms^
0
20
20
0
0
9
20
10
14
4
16
20
4
37
20
24
20
2
40
34
30
20
1
27
144
37
20
4
36
11
44
20
0
55
5
51
19
0
70
3
71
19
0
196
4
86
20
2
72
9
96
20
0
135
6
109
20
0
149
29
121
19
0
102
36
'Each box originally contained 10 dead eggs and 10 dead alevins. Two boxes
were removed on each sample date and the contents combined tor reporting.
^Of all insect larvae recovered, 80°o were Plecoptera. 16°o Diptera, 3%
Trichoptera. and 1°o Ephemeroptera,
^Tentatively identified as Polycelis borealis. a species that Kenk (1953)
commonly found in clear cold streams in southern parts of Alaska,
one box contained identifiable alevins. Although
the dead alevins disappeared rapidly, only a few of
the dead eggs disappeared. In a study to determine
whether certain stonefly numphs were predators
or scavengers on salmon eggs and alevins, Ellis
(1970) found in one experiment that dead pink
salmon alevins buried in Vibert boxes in a stream
essentially disappeared within a 2-wk period.
Concurrently with the rapid disappearance of
dead alevins from the buried boxes was a rapid
buildup of invertebrates in the boxes. Although
invertebrates are commonly found with salmon
embryos (Briggs 1953; McDonald 1960; Nicola
1968), it is frequently impossible to determine if
predation or scavenging is occurring. Although
some groups of stonefly nymphs are known to at-
tack live salmon embryos (Stuart 1953; Claire and
Phillips 1968), Ellis ( 1970) concluded that nymphs
of the carnivorous genus Alloper la were basically
scavengers rather than predators.
In addition to various insect larvae, a planarian
worm tentatively identified as Polycelis borealis
was commonly found in the boxes buried in Sashin
Creek (Table 12). Little is known on the biology or
life history of this planarian, but under favorable
conditions it appears to rapidly increase its num-
bers in the streambed, and thus may be particu-
larly important in removing dead alevins. I have
observed successive seasonal increases in the rela-
tive abundance of planarians in samples taken
from the Sashin Creek streambed with the hy-
draulic sampler in the fall, winter, and spring.
Total counts of planarians removed from the
streambed with the hydraulic sampler are not pos-
577
FISHERY BULLETIN: VOL. 76, NO. 3
sible,^ but partial counts indicated that by March
the densities of planarians in some parts of Sashin
Creek commonly reached several thousand per
square meter. A similar seasonal increase in
streambed populations of planarians concomitant
with the seasonal occurrence of sockeye salmon, O.
nerka, embryos has been noted elsewhere.^
In Sashin Creek there is little doubt that high
planarian populations are related to the presence
of salmon eggs and alevins, because planarians
are scarce in streambed gravels above the impass-
able falls where salmon do not spawn. However,
the precise role of these organisms in the ecology of
spawning beds is unknown. To learn something
about the role of planarians, I conducted tests with
various combinations of planarians and live and
dead salmon eggs and alevins in experimental
containers. In these tests planarians did not prey
on and were not toxic to live embryos, nor did they
feed on dead eggs unless the chorion was broken
and the egg contents exposed.
EVIDENCE OF OVERSEEDING
In assessing the probability of streambed over-
seeding in Sashin Creek in the 1967 brood year, it
is most useful to compare fry production in 1967-
68 with production in other years. Since 1940,
^When large numbers of planarians are excavated with the
hydraulic sampler, many elongate their bodies and pass through
the meshes of the collecting net.
6W. L. Hartman, W. R. Heard, and C. W. Strickland. 1962. Red
salmon studies at Brooks Lake biological field station, 1961.
Unpubl. manuscr. on file, NWAFC Auke Bay Lab. NMFS,
NOAA, P.O. Box 155, Auke Bay. Alaska.
production has varied from 50 fry to almost 6 mil-
lion fry; corresponding parent escapements varied
from 8 to 92 ,085 ( Table 1 ) . Only three escapements
exceeded that of 1967, and only one of these ( 1940)
produced more fry (about 0.4 million more) (Table
1). When the numbers of fry are plotted against
potential egg deposition, a dome-shaped fry pro-
duction curve is derived for Sashin Creek (McNeil
1969). The relative position of fry production for
the 1967 brood year falls near the descending limb
of the curve; fry production from the 1941 and
1942 brood years indicates a continuing decrease
in fry production as escapements increased (Fig-
ure 2).
Data collected since 1961 on the density of eggs
in the three study areas at the end of spawning
provide a means of more precisely defining the fry
production potential of the stream. Plotting fry
production as a function of actual egg deposition
for each area produces curves that suggest the
potential maximum fry production in the upper
and middle areas is around 500 fry/m^ and the
potential in the lower area is about half that
number (Figure 3). In 1967 the actual density of
eggs deposited considerably exceeded twice the
theoretical maximum fry production in all three
areas and the fry production was considerably
below the maximum; it appears that overseeding
occurred in 1967.
Until 1967 the timing of entry of adults into
Sashin Creek had usually been an accurate indi-
cator of the freshwater survival of progeny (Mer-
rell 1962; McNeil 1968; Ellis 1969). The presumed
biological basis for the correlation between early
spawning and high survival was that embryos de-
500
Figure 2. — Production of pink salmon in
Sashin Creek, 1940-67. The shaded area
in the lower left is shown on a larger scale
in the upper right corner. Nine genera-
tions of the even-year line 1946-62 were
excluded because fry productions were all
v
^14
_
\^^
■ —
Q
> \(.76)
>.
z
\
O 12
_
\
5
^
_l
<
\
^0
-
v. 31)
\
z
\( 42)
0.
\
uj 3
_
(.15)
\
>
\
-J
^\..,.^
\
6
-
— -U"
\( 26) \
4
-
(~07)^^'^;;;;--V^
2
-
^0-
0
Aiir.
'SFPT ' nn
' ' MnV ' HF
r ' lAM ' FFR ' MAD ' AP
D ' MAV '
Figure 5. — Number of live pink salmon in Sashin Creek at the
beginning and end of three periods in freshwater for 1963, 1965,
and 1967 brood years. Numbers in parentheses show instan-
taneous monthly mortality coefficients from the egg through
alevin stages. Mortality during fry migration (April and May)
for the 1963 and 1967 brood years was negligible when measured
as the difference between streambed and weir estimates of total
fry production. The dotted extension of the 1965 brood year
assumes no mortality during this period.
1963 and 1967 (0.76 and 0.71); the lower mortality
during spawning in 1965 (0.31) probably reflects
efficient spawning during the low streamflow con-
dition prevailing that year (McNeil 1968). Mortal-
ity from spawning to hatching was similar, but
mortality from hatching to emergence was strik-
ingly different in each of the 3 yr (Figure 5).
McNeil (1968) suggested the increase in over-
winter mortality of 1965 brood year progeny (0.26)
over 1963 brood year progeny (0.07) might have
been related to a delayed mortality from low con-
centrations of dissolved oxygen in early embryo
development during drought conditions in late
summer and early fall in 1965. The number of
spawners in 1967 was more than double the
number in 1963 and 1965 (Table 1). The over-
winter mortality of 1967 brood year progeny (0.42)
was considerably higher than the high mortality
of the 1965 brood year (0.26). The heavy over-
winter mortality experienced by the 1967 brood
year progeny may also have been caused by low
dissolved oxygen concentrations. However, be-
cause no drought conditions existed while the
progeny were in the gravel, these poor oxygen
conditions probably resulted from the high density
of eggs and alevins in the streambed.
SUMMARY
1. In 1967, 38,067 pink salmon spawned in
Sashin Creek on Baranof Island, Alaska. Fifty-
two percent of the spawners ( 19,639) were females;
mean fecundity was 2,260 eggs/female and the
potential number of eggs available for deposition
totaled 44.4 million.
2. Entry of spawners into the stream was the
second earliest on record; based on the previously
consistent relation between time of entry and
freshwater survival, the production of fry should
have been greater than any previously recorded,
but the 3 million fry produced were less than half
the predicted number.
3. Mean female densities on the spawning
grounds were 1.74/m2 in the upper area, 1.62/m2
in the middle area, and 1.17/m2 in the lower area.
Densities were higher in the upper area at the
beginning of spawning before significant levels of
spawning occurred in the middle or lower areas.
The tendency for spawners in the odd-year line to
utilize the upper area of Sashin Creek may be due
to genetic factors, including timing of escape-
ments, and possibly differential marine survival
favoring fry produced in the upper area.
580
HEARD: PROBABLE CASE OF STREAMBED OVERSEEDING
4. Survival of progeny of the 1967 spawners
was determined a) from stream entry to end of
spawning, b) from end of spawning to hatching, c)
from hatching to shortly before fry emergence, and
d) from shortly before emergence to emergence
and downstream migration of fry. In general, sur-
vival in each of these time periods was greatest in
the upper area, lowest in the lower area, and in-
termediate in the middle area, a pattern consis-
tent with previous survival studies at Sashin
Creek. Total freshwater survival from potential
egg deposition to preemerged fry was 107f , 79c , and
49c in the upper, middle, and lower areas, respec-
tively, and 6.8% for the entire stream. The total
number of migrating fry agreed closely with the
estimates of preemerged fry in the streambed in
late March.
5. Mortality of eggs and alevins was high dur-
ing spawning, low between spawning and hatch-
ing, and high between hatching and emergence.
Between 1 December 1967 and 25 March 1968,
11.1 million eggs or alevins disappeared within
Sashin Creek streambed; 10.7 million of these
were alive on 1 December. The high densities of
eggs and alevins in the streambed after spawning
and at hatching are believed to exceed the
streambed capacity for fry production. High over-
winter mortalities appear to have occurred shortly
after hatching, probably from critical levels of dis-
solved oxygen in intragravel water. Critical oxy-
gen levels apparently developed under average
winter streamflow conditions due to the high
biochemical oxygen demand placed on the
streambed by high egg and alevin densities.
6. Recently hatched dead alevins disappear
rapidly within the streambed because of biochem-
ical decomposition and invertebrate scavenging.
In comparison with dead alevins, dead eggs disap-
pear slowly. In Sashin Creek, insect larvae and a
planarian, probably Polycelis borealis, may be
particularly important in removing dead salmon
eggs and alevins from the streambed.
7. Several aspects of the historical patterns of
pink salmon fry production in Sashin Creek
suggest that streambed overseeding occurred in
1967. Fry production from the 1967 spawners falls
on the descending limb of the fry production
curves, both for the stream as a whole (since 1940)
and for the individual stream areas (since 1961).
From the historical pattern of time of adult entry
and resulting freshwater survival, freshwater
survival of 1967 brood year progeny should have
been around 18% (or a production of 8 million fry).
Survival of progeny during spawning and between
spawning and hatching was adequate to reach
these predicted levels. Overwinter mortalities (be-
tween hatching and emergence), however, were
higher than any previously recorded. Compensa-
tory losses during this period were probably due to
the presence of too many eggs and alevins in the
gravel for existing environmental conditions —
streambed overseeding.
8. Overseeding does not invariably occur at
some precise density of eggs, but rather is a
dynamic interaction between densities of eggs and
alevins in the gravel, certain ecological charac-
teristics that define the fry production capability
of the streambed, and the prevailing climatologi-
cal features during the 6- to 8-mo period eggs and
alevins reside in the streambed.
ACKNOWLEDGMENTS
The following people assisted with field studies
on the 1967 spawners and freshwater survival of
the progeny: David Brickell, Robert Coats,
Richard Crone, Calvin Fong, Henry Kopperman,
Derek Poon, and Roger Winchester.
LITERATURE CITED
Brickell, D. C.
1971. Oxygen consumption by dead pink salmon eggs in
salmon spawning beds. M.S. Thesis, Univ. Alaska, Col-
lege, 53 p.
BRIGGS, J. C.
1953. The behavior and reproduction of salmonid fishes in
a small coastal stream. Calif. Dep. Fish Game, Fish
Bull. 94, 62 p.
1955. Behavior pattern in migratory fishes. Science
(Wash., D.C.) 122:240.
Chapman, D. W., and T. C. Bjornn.
1969. Distribution of salmonids in streams, with special
reference to food and feeding. In T. G. Northcote ( editor),
Symposium on salmon and trout in streams, p. 153-176. H.
R. MacMillan Lect. Fish. Inst. Fish. Univ. B.C., Van-
couver.
CLAIRE. E. W., AND R. W. PHILLIPS.
1968. The stone^y Acroneuria pacifica as a potential pred-
ator on salmonid embryos. Trans. Am. Fish. Soc. 97:50-
52.
Ellis, R. J.
1969. Return and behavior of adults of the first filial gen-
eration of transplanted pink salmon, and survival of their
progeny, Sashin Creek, Baranof Island, Alaska. U.S.
Fish Wildl. Serv., Spec. Sci. Rep. Fish. 589, 13 p.
1970. Alloperla stonefly nymphs: Predators or scavengers
on salmon eggs and alevins? Trans. Am. Fish. Soc.
99:677-683.
581
FISHERY BULLETIN: VOL 76. NO. 3
KKNK, R.
1953. The fresh-water triclads (Turbellaria) of Alaska.
Proc. U.S. Natl. Mus. 103:163-186.
McDonald. J. G.
I960. A possible source of error in assessing the survival of
Pacific salmon eggs by redd sampling. Can. Fish Cult.
26:27-30.
McL.XKNEY. W. 0.
1967. Intra-stream movement, feeding habits, and popula-
tion of the coastrangesculpin,Co»(/sa/e!^?;cKS. in relation
to eggs of the pink salmon, Oncorhynchus gorbuscha. in
Alaska. Ph.D. Thesis, Univ. Michigan, Ann Arbor, 131
P-
McNeil, W. J.
1964a. A method of measuring mortality of pink salmon
eggs and larvae. U.S. Fish Wildl. Serv., Fish. Bull.
63:575-588.
1964b. Redd supenmposition and egg capacity of pmk
salmon spawning beds. J. Fish. Res. Board Can.
21:1385-1396.
1966. Distribution of spawning pink salmon in Sashin
Creek, southeastern Alaska, and survival of their prog-
eny. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 538, 12
P-
1968. Migration and distribution of pink salmon spawners
in Sashin Creek in 1965, and survival of their prog-
eny. U.S. Fish Wildl. Serv., Fish. Bull. 66:575-586.
1969. Survival of pink and chum salmon eggs and alevins.
In T. G. Northcote (editor). Symposium on salmon and
trout in streams, p. 101-117. H. R. MacMillan Lect.
Fish. Inst. Fish. Univ. B.C., Vancouver.
McNeil. W. J., S. C. Smedley. and R. J. Ellis.
1969. Transplanting adult pink salmon to Sashin Creek,
Baranof Island, Alaska, and survival of their prog-
eny. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 587, 9 p.
McNeil. W. J., R. A. Wells, and D. C. Brickell.
1964. Disappearance of dead pink salmon eggs and larvae
from Sashin Creek, Baranof Island, Alaska. U.S. Fish
Wildl. Serv., Spec. Sci. Rep. Fish. 485, 13 p.
MERRELL, T. R., Jr.
1962. Freshwater survival of pink salmon at Sashin
Creek, Alaska. In N. J. Wiiimovsky (editor). Symposium
on pink salmon, p. 59-72. H. R. MacMillan Lect.
Fish. Inst. Fish. Univ. B.C., Vancouver.
Moyle. p.
1966. Feeding behavior of the Glaucous-winged Gull on an
Alaskan salmon stream. Wilson Bull. 78:175-190.
NEAVE. F.
1953. Principles affecting the size of pink and chum salm-
on populations in British Columbia. J. Fish. Res. Board
Can. 9:450-491.
Nicola. S. J.
1968. Scavenging by Alloperla (Plecoptera: Chloroper-
lidae) nymphs on dead pink (Oncorhynchus gorbuscha I
and chum lO. ketat salmon embryos. Can. J. Zool.
46:787-796.
Reed. R. J.
1967. Observation of fishes associated with spawning
salmon. Trans. Am. Fish. Soc. 96:62-67.
RICKER. W. E.
1962. Regulation of the abundance of pink salmon popula-
tions. In N. J. Wiiimovsky (editor). Symposium on pink
salmon, p. 155-201. H. R. MacMillan Lect. Fish. Inst.
Fish. Univ. B.C., Vancouver.
SKUD, B. E.
1958. Relation of adult pink salmon size to time of migra-
tion and freshwater survival. Copeia 1958:170-176.
Stuart. T. A.
1953. Spawning migration, reproduction and young stages
of loch trout (Sa/wo truttaL.) . Scott. HomeDep., Fresh-
water Salmon Fish. Res. 5, 39 p.
U.S. GEOLOGICAL Survey
1969. Water resources data for Alaska 1968. Part 1. Sur-
face water records. U.S. Geol. Surv., Water Resour. Div.,
155 p.
Wells. R. a., and W. J. McNeil.
1970. Effect of quality of the spawning bed on growth and
development of pink salmon embryos and alevins. U.S.
Fish Wildl. Serv., Spec. Sci. Rep. Fish. 616, 6 p.
WICKETT, W. P.
1954. The oxygen supply to salmon eggs in spawning
beds. J. Fish. Res. Board Can. 11:933-953.
1962. Environmental variability and reproduction poten-
tials of pink salmon in British Columbia. In N. J.
Wiiimovsky (editor). Symposium on pink salmon, p.
73-86. H.R. MacMillan Lect. Fish. Inst. Fish. Univ. B.C.,
Vancouver.
582
VERTICAL DISTRIBUTION AND PHOTOSENSITIVE VESICLES OF PELAGIC
CEPHALOPODS FROM HAWAIIAN WATERS
Richard Edward Young *
ABSTRACT
Vertical distribution data were obtained for 47 species ofpelagiccephalopods off Oahu, Hawaii. Peaks
in species richness occurred at 500-800 m during the day and in the upper 300 m at night. Over SO^f of
the individuals occurred in the upper 250 m at night. Approximately 60% of the species underwent diel
vertical migration, and most of these migrated into the upper 250 m. In five of nine groups of closely
related species, clear differences in habitat were found.
Deepwater spawning appeared to occur in a variety of cephalopods. Two of the bathypelagic octopods
brooded their young at or above the upper limit of the remaining adult population. In doing so, the
extent of the upward migration of newly hatched individuals was reduced.
Photosensitive vesicles occurred in all species. These organs probably detect downwelling daylight
for regulating vertical migration and counterillumination. The vesicles also appeared to form an
elaborate system for monitoring bioluminescent light from the animal's own photophores, from within
the mantle cavity, and from other animals located outside the visual field.
Cephalopods must occupy a wide variety of ecolog-
ical roles in the pelagic realm of the open ocean:
the highest diversification of families and genera
is found in this environment. In order to under-
stand these roles, the vertical distribution of these
animals must be determined. A number of papers
have treated various aspects of the vertical dis-
tribution of oceanic cephalopods (e.g., Pearcy
1965; Clarke 1969; Roper 1969; Gibbs and Roper
1971; Clarke and Lu 1974, 1975; Lu and Clarke
1975a, b; Roper and Young 1975). Their vertical
habitats, however, remain poorly known.
Data on the vertical distribution of cephalopods
is difficult to obtain: many species are uncommon,
and some avoid small trawls. In this study an
opening-closing net (modified Tucker trawl) pro-
vided unambiguous depth data, and a slightly
larger open net (3-m Isaacs-Kidd midwater trawl)
added considerable additional data; nevertheless,
fast-swimming species were poorly sampled.
Extraocular photoreceptive organs, the photo-
sensitive vesicles, were examined in each species
for clues that would indicate the role of light in
regulating vertical distribution patterns. The or-
gans in squid, known as the parolfactory photo-
sensitive vesicles, lie near the brain within the
confines of the cephalic cartilage. In octopods the
organs, known as epistellar photosensitive vesi-
cles, lie within the mantle cavity adjacent to the
'Department of Oceanography, University of Hawaii, Hono-
lulu, HI 96822.
stellate ganglia. The photosensitive vesicles are
paired organs. Each organ, as the name implies, is
generally composed of a large number of small
vesicles. The individual vesicles contain photo-
sensitive cells similar to those of the retina, and
their photoreceptive nature has been well estab-
lished ( Nishioka et al. 1966; Mauro and Baumann
1968; Mauro 1977). The specific functions of the
photosensitive vesicles are unknown in both neri-
tic and oceanic cephalopods although many
suggestions have been made (see Packard 1972).
Several papers discussing the relationship of
vertical distribution to eye structure, biolumines-
cence and/or development of photosensitive vesi-
cles in selected species have already appeared
(Young 1972a, 1973, 1975a, c, d, 1977). Some data
on distribution taken during the initial phases of
this program have been published by Roper and
Young (1975). This paper examines the vertical
distribution of all pelagic cephalopods taken off
Hawaii and the morphology and orientation of
their photosensitive vesicles.
MATERIALS AND METHODS
Specimens were collected off the island of Oahu
in the Hawaiian archipelago at long. 158°20'W,
lat. 21 = 20 'N over depths between 1,500 and 4,000
m. Collections were made from September 1969 to
November 1974 primarily from the RV Teritu.
Over 3,300 specimens were taken in horizontal
tows during about 1,000 h of trawling time.
Manuscript accepted January 1978.
FISHERY BULLETIN; VOL" 76. NO. 3, 1978.
583''^
FISHERY BULLETIN: VOL. 76. NO. 3
Cephalopods were collected primarily with two
types of nets: a 3-m opening-closing modified
Tucker trawl and a 3-m Isaacs-Kidd midwatej-
trawl (IKMT). Details of the trawling with the
Tucker trawl are given by Walters (1976). When
the Tucker trawl failed to close or close com-
pletely, the trawl was considered an open tow.
Tows usually were made at 5 to 6 km/h for a period
of 3 h. Twilight periods were generally avoided.
Tows made with net closed indicated the catch
contained almost no contamination. Contamina-
tion from previous tows was minimized by care-
fully cleaning the net after each tow. The trawl
tended to wander vertically when open; this was
most severe in deep tows. During the latter part of
the program, acoustic depth telemeters allowed
trawl depths to be continuously adjusted and
greatly reduced wandering. The distribution
figures indicate the extent of this wandering.
Trawl depths usually were determined with a
time-depth recorder attached to the trawl.
Clarke ( 1973) discussed trawling methods with
the IKMT. The trawl was lowered quickly then
towed horizontally at 5 to 6 km/h for usually 2 h.
Retrieval was rapid with the ship moving slowly
ahead. Vertical wandering of the net was not as
serious as with the Tucker trawl. All specimens
captured with the IKMT were assumed to come
from the modal trawling depth of the net, or if no
clear mode was present, from the midpoint of the
effective vertical range of the tow. The occasional
capture of a specimen during setting or retrieval of
a net results in an anomalous depth record below
the animal's normal habitat. Contamination of
the catch by animals from previous tows occasion-
ally occurred with the IKMT. This contamination
is especially serious as the error may be impossible
to detect.
IKMT data for a few of the most abundant
species are presented both as catch per trawling
effort and as actual catch figures (Table 1). The
remaining distribution figures are designed to
show animal size vs. depth relationships and to
indicate the precision and reliability of the data
(e.g., fishing range of the tow, open or opening-
closing tow). As a result, corrections in the data for
unequal sampling at various depths could not be
made. This bias was especially critical at depths
<400 m during the day and at depths > 1,000 m
during the day and night where sampling was low.
The magnitude of this error can be determined
from Table 2, which lists sampling time in each
100-m depth interval.
Depth data for most species taken over the en-
tire trawling period have been combined. There-
fore, short-term variation in depth distributions
may be obscured. Where sufficient data exist to
determine general distribution patterns based on
Tucker trawls alone, these data are presented
separately. For species with insufficient data, data
from both trawls are combined in the figures. In
most cases larvae, which usually have a different
vertical distribution than adults, have been
excluded from the distribution figures and the
Table l. — Depth distribution, capture rates, and numbers of the most abundant cephalopod species captured by the Isaacs-Kidd
midwater trawl. Day captures for Pterygioteuthis giardi are included in Figure 4. R = capture rate in numbers per 1,000 m^ of water
sampled. N = actual number captured. ND = no data.
Day
Night
m
o
CO
X
CO
CO
CO
n
,co
X
CO
3
am
Oi
b
Q.CD
O .
Q)
e
3
PTi
1^
o
5)
^
O
o
0)
o
en
O CO
1^
I
0)
51
S
u.
i:
Cl>
5:
E
u.
5:
Depth (m)
R
N
R
N
R
N
R
N
R
N
R
N
R
N
0-50
ND
ND
ND
ND
ND
ND
5.4
14
15.2
39
1.5
4
5.8
15
50-100
ND
ND
ND
ND
ND
ND
17.9
52
62,1
180
6.5
19
7.9
23
100-150
ND
ND
ND
ND
ND
ND
3.2
8
23,3
57
7.3
18
2.0
5
1 50-200
0
0
0
0
0
0
5,2
14
14,1
38
15.9
43
2.2
6
200-250
0
0
0
0
0
0
1,0
1
0
0
3.1
3
0
0
250-300
ND
ND
ND
ND
ND
ND
1.0
1
1-0
1
6.5
6
4.3
4
300-400
3.0
4
2.3
3
6.9
9
0
0
0
0
0
0
0
0
400-500
10.9
14
42.3
54
14.1
18
49
3
4.9
3
4.9
3
0
0
500-600
7.4
9
72.2
87
20.7
25
3,4
2
6,9
4
5.1
3
0
0
600-700
22.8
26
8.7
10
14,9
17
0
0
1,4
1
0
0
0
0
700-800
1.0
2
5.3
10
3.2
6
28
3
0
0
0
0
0
0
800-900
3.6
3
8.5
7
1.2
1
0
0
0
0
0
0
0
0
900-1,000
2.1
2
1.0
1
2.1
2
0
0
0
0
0
0
0
0
1,000-1,100
1.5
1
20.1
13
1.5
1
0
0
0
0
5.4
1
0
0
584
YOUNG: VERTICAL DISTRIBUTION AND PHOTOSENSITIVE VESICLES
Table 2. — Trawling time in minutes. Since trawls of two different sizes were used, a correction
factor of 0.6 was applied to the trawling times of the Tucker trawl to compute the adjusted total
trawling time. This factor represents the approximate difference in effective mouth areas of the
two nets. Open tows = timeofeach tow was assigned to one depth. Opening-closing tows = time
of each tow was apportioned among depth zones traversed by the trawl. IKMT = Isaacs-Kidd
midwater trawl.
Tucker trawl
1
IKMT
Total adjusted
to IKMT
Open
Opening-closing
Day Night
Open
Depth (m)
Day
Night
Day
Night
Day
Night
0-50
1.091
939
2,562
3,780
50-100
180
1,112
614
2.897
108
3,932
100-150
781
27
591
2.442
16
3,265
1 50-200
144
530
84
870
130
2,689
267
3,529
200-250
536
186
320
177
944
289
1,458
250-300
146
31
871
911
106
1,434
300-350
180
203
514
452
605
682
913
350-400
180
180
460
407
839
552
1,223
904
400-500
360
502
1,529
1,413
1,276
605
2,409
1,754
500-600
376
1,748
927
1,204
577
2,253
1,359
600-700
133
1.683
1,420
1,139
714
2,149
1,646
700-800
313
220
1.638
838
1,862
1,052
3,033
1,687
800-900
1,244
519
820
179
1,566
490
900-1,000
180
30
709
184
917
179
1,450
307
1,000-1,100
182
464
64
646
182
924
330
1,100-1,200
195
230
156
180
195
435
289
1,200-1,300
200
300
380
180
528
180
1.300-1,400
10
256
234
98
146
212
1,400-1,500
67
106
40
64
1.500-1,600
30
18
1.600-1,700
8
38
5
23
1,700-1,800
104
267
62
160
1,800-1.900
84
228
50
137
1.900-2,000
48
28
29
17
2.000-2.100
15
27
9
16
2.100-2.200
72
29
43
17
2,200-2.300
43
33
26
20
2.300-2.400
8
5
text. Specimens captured during twilight periods,
with a few exceptions, have also been excluded
from the charts.
Species examined are listed in Table 3. Larvae
or juveniles of several additional species were cap-
tured but are not included in this study. These are:
Tremoctopus violaceus, Argonauta sp., Cranchia
scabra, Thysanoteuthis rhombus, Onykia sp. One
pelagic species reported from Hawaii by Berry
(1914), Iridoteuthis iris, was not taken. This
species belongs in the genus Nectoteuthis and
probably lives in association with the ocean floor. ^
Photosensitive vesicles of most species were
sectioned. Material was fixed either in
glutaraldehyde-osmium tetroxide or Bouin's so-
lution and was embedded in Epon 812^ or paraffin.
All vesicles sectioned contained cells with photo-
sensitive processes and, therefore, appeared to be
functional. In only a few cases did the general
histology of the organs add to our understanding of
^Roper, C, and R. Young. Review of the Heteroteuthinae.
Unpubl. manuscr.
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
their function. As a result, histological details are
not included for most species.
In order to quantify the size of vesicles, an at-
tempt was made to obtain dry weights. Many types
of vesicles proved difficult to remove and clean
completely. Vesicles from a series of similar-sized
Pyroteuthis addolux, which are easily removed,
were weighed and found to vary by a factor of 1.5.
Because of the large individual variations and in-
accuracies due to difficulties in isolating many
types of vesicles, this method of quantification was
abandoned. As a result, camera lucida drawings of
photosensitive vesicles provide the only measure
of organ size: their relative size can be approxi-
mately determined by comparison with the brain
size.
RESULTS
Pyroteuthis addolux Young 1972
Vertical Distribution
During the day, 39 specimens captured by the
Tucker trawl indicate a vertical range for this
585
FISHERY BULLETIN: VOL. 76, NO. 3
Table 3. — Species of cephalopods considered.
Order Teutholdea
Family Enoploteuthidae
Pyroteuthis addolux Young 1 972
Pterygioteuthis microlampas Berry 1913
Pterygioteuthis giardi Fischer 1895
Abralia tngonura Berry 1913
Abralia astrosticta Berry 1909
Abraliopsis sp. A. n.sp being described by L, Burgess
Abraliopsis sp. B, n.sp. being described by L, Burgess
Abraliopsis sp. C, n.sp. being described by L. Burgess
Enoploteuthis sp. A. n.sp, being described by L. Burgess
Enoploteuthis sp B, n.sp. being described by L. Burgess
Thelidioteuthis alessandnnii (Verany 1851)
Family Ommastrephidae
Symplectoteuthis oualaniensis (Lesson 1830)
Hyaloteuthis pelagicus (Bosc 1802)
Notolodarus hawaiiensis (Berry 1912)
Family Histioteuthidae
Histioteuthis dofleini (Pfeffer 1912)
Histioleuthis celetana pacifica (Voss 1962)
Histioteuthis sp, under study by N. Voss
Family Neoteuthidae
Neoteuthis sp
Family Bathyteuthidae
Bathyteuthis abyssicola Hoyle 1885
Family Ctenopterygiidae
Ctenopteryx siculus (Verany 1851)
Family Onychoteuthidae
Onychoteuthis compacta (Berry 1913)
Family Octopoteuthidae
Octopoteuthis nielseni Robson 1948
Family Cycloteuthidae
Cycloteuthis serventyi Joubin 1919
Discoteuthis laciniosa Young and Roper 1969
Family Brachioteuthidae
Brachioteuthis sp.
Family Chiroteuthidae
Chiroteuthis n.sp., being described by Roper and Young
Chiroteuthis picteti Joubin 1 894
Chiroteuthidae ngen , n.sp. being described by Roper and Young
Planktoteuthis lippula (Chun 1908)
Gnwatditeuthis bomplandi (Verany 1837)
Family Mastigoteuthidae
Mastigoteuthis famelica (Berry 1909)
Mastigoteuthis inermis Rancurel 1972
Family Joubiniteuthidae
Joubiniteuthis portieri (Joubin 1912)
Family Cranchiidae
Liocranchia valdiviae (Chun 1906)
Liocranchia reinhardti (Steenstrup 1856)
Leachia pacifica (Issel 1908)
Phasmatopsis fishen (Berry 1909)
Taonius pavo (LeSueur 1821)
Sandalops melancholicus Chun 1906
Helicocranchia beebei Robson 1948
Bathothauma lyromma Chun 1 906
Order Octopoda
Family Bolitaenidae
Eledonella pygmaea Verrill 1884
Japetella diaphana Hoyle 1885
Family Amphitretidae
Amphitretus pelagicus Hoyle 1885
Family Vitreledonnelidae
Vitreledonnella nchardi Joubin 1918
Order Vampyromorpha
Family Vampyroteuthidae
Vampyroteuthis inlernalis Chun 1903
Order Sepioidea
Family Sepiolidae
Heteroteuthis hawaiiensis (Berry 1909)
species of 375 to 510 m; most captures came from
450 to 500 m (Figure 1). IKMT data lumped into
100-m increments show most day captures be-
tween 400 and 700 m (Table 1 ). At night, 38 of the
41 specimens captured by the Tucker trawl indi-
cate a vertical range of about 110 to 225 m; most
specimens camp from 150 to 200 m. Three speci-
mens were captured during the night in opening-
closing tows near their day habitat at depths be-
tween 360 and 480 m. Each of these three speci-
mens was taken in a separate tow during a cruise
in November 1972 within a few days of new moon.
Although the upper 200 m was not sampled on this
cruise, these captures indicate that at least part of
"1 — I — I — I"
200
400
Q600
SOD
1000
I I I I r T r
X 1 A O ■ '■O O ji ■ A oo ' o ' o
,iO O 'OOO. >0O O O;; 0|0*
:t
t ft
o o
o 9 9 90 o
10
22
26 30 34
MANTLE LENGTH ,m m
38
42
50
Figure. L— Vertical distribution of Pyroteuthis addolux. Symbols for Figure 1 and subsequent figures: open circles represent day
captures; closed circles represent night captures. A bar with a circle indicates an opening-closing tow with the bar representing the
depth range of the tow and the circle the most likely depth of the capture (the modal depth, or if no clear mode is present, the midpoint of
the vertical range of the tow). A circle without a bar indicates a capture in an open tow, A bar without an associated symbol indicates an
open oblique tow. Such bars do not always intersect the zero depth 1 ine as some gradual oblique tows were made between specific depths.
Solid bars represent night captures. Dashed bars represent day captures, A small dot represents a presumed contaminant.
586
YOUNG; VKRTirAl. DISTRIBLTION AND PHOTOSKNSITIVK VESICLES
the population was not migrating during this
period. IKMT data lumped into 50-m increments
show that most night captures were made between
50 and 200 m with peak catches between 150 and
200 m.
Photosensitive Vesicles (Figure 2 A)
The organs are very similar to those described
by R. E. Young (1977) \n Pterygioteuthis micro-
lampas. Pyroteuthis addolux has two sets of or-
gans. The dorsal organs (the more dorsal set) lie
embedded in the posterodorsal wall of the cephalic
cartilage and adjacent to the optic lobes of the
brain. Each ventral organ lies deeply embedded in
the posteroventral surface of the cephalic carti-
lage. Except for a thin medial extension on each
ventral organ, all organs are thick, compact, and
approximately circular to square in outline. The
histological structure of the dorsal and ventral
organs is similar. The integument adjacent to both
the dorsal and ventral organs lacks pigment and
thereby forms distinctive "windows" for the pas-
sage of light.
Nerves from both dorsal and ventral organs
enter the peduncle complex of the brain and their
fibers disperse in the base of the peduncle lobe
near its broad junction with the olfactory lobe.
Pterygioteuthis microlampas Bcrr\ I91.'^
X'ertical Distribution (Figure 3)
The vertical distribution has been described by
R. E. Young ( 1977). During the day, 48 specimens
captured with the Tucker trawl indicate a depth
range of 450 to 575 m; 857r of the specimens were
taken between 450 and 500 m. IKMT data lumped
into 100-m increments (Table 1) show most day
captures between 400 and 600 m. At night 56
specimens taken by the Tucker trawl indicate a
depth range of 25 to 180 m; nearly 85*^ of the
captures were made between 50 and 105 m. IKMT
data lumped into 50-m increments indicate a
range of 0 to 200 m with a strong peak in the 50- to
100-m depth zone. The night distribution was not
affected by moonlight (R. E. Young 1977).
D PV
POST
PS MASS
FIGURE 2.— A. Photosensitive vesicles oiPyroteuthis addolux. This illustration and most subsequent drawings show a side view of the
bram. The optic stalk has been cut (as indicated by cross-hatching) and the optic lobe removed. The esophagus can be seen passmg
through the bram. Three major subdivisions of the bram are apparent (i.e., the supraesophageal mass, the posterior subesophageal
mass, and the middle subesophageal mass.) A large nerve tract which extends anteriorly to the anterior subesophageal mass was cut
(indicated by dotted line) and the latter portion of the brain is not shown. B. Photosensitive vesicles of A 6ra/iops;s sp. B. Abbreviations
for Figure 2 and subsequent figures of photosensitive vesicles: AV. PV.— Anteroventral photosensitive vesicles; C. PV.— central
photosensitive vesicles; DOR.— dorsal; D. PV.— dorsal photosensitive vesicles; ES.— esophagus; GILL— gill; H. RET, M,— head retrac-
tor muscles; INT,— intestine; M. S, MASS-middle subesophageal mass of the brain; MV, PV.— midventral photosensitive vesicles;
N.-nerve; OP. ST.-optic stalk; PED. L.-pedunclelobe; PIG. S.-pigment screen; POST.-posterior; P. PV. -posterior photosensitive
vesicles; P. S. MASS— posterior subesophageal mass of the brain; PV.— Photosensitive vesicles; V. PV.— ventral photosensitive
vesicles; S. MASS — supraesophageal mass of the brain; VENT. — ventral.
587
FISHERY BULLETIN: VOL 76. NO. 3
200-
400
E
r
a.
LU
Q600
800
1000
9?
12
"^I — ' — ^ — WMVrry y^r-- ^
' * Mt^ nm* TTTT* ITTT ^T TT* T
13
•lo 00900 ^oo^oj
' ' ' ■ " I' '• ■ 4 _,
looo
14 IS 16 17 18
MANTLE LENGTH. mf
i?
^^
J L-v-l-
20 21 24
Figure 3.— Vertical distribution of
Pterygioteuthis microlampas. From R. E.
Young (1977). Symbols as in Figure 1.
Photosensitive Vesicles
The organs have been described by R.E. Young
(1977). They are essentially the same as in
Pyroteuthis addolux.
Pterygioteuthis giardi Fischer 1895
Vertical Distribution (Figure 4)
During the day 30 specimens captured from
both trawls indicate a depth range from about 390
to 525 m; over 907^ of the captures were made
between 390 and 450 m. One IKMT tow captured
eight badly damaged specimens at 630 m, well
below their zone of maximum abundance. The
previous tow had captured six specimens at 390 m
that were in excellent condition; specimens from
the deeper tow probably are contaminants. The
depth distribution of this species may be biased by
the relatively low sampling effort between 350
and 400 m.
At night, 39 captures with the Tucker trawl
indicate a depth range of 15 to 180 m; over Ih'^c of
the specimens came from 15 to 50 m. IKMT data
lumped into 50-m increments (Table 1) show the
maximum abundance at depths of 0 to 100 m.
Photosensitive Vesicles
The organs are essentially the same as in
Pyroteuthis addolux.
Ahralia trigotiura Berry 1913
Vertical Distribution (Figure 5)
Fifty specimens were captured by both trawls.
Excluding presumed contaminants, the day cap-
tures were made between 390 and 650 m with
nearly 809f between 450 and 560 m. At night cap-
tures were made between 30 and 200 m with over
759^ between 50 and 100 m.
800 -
Figure 4. — Vertical distribution of Pterygioteuthis giardi . Sym-
bols as in Figure 1.
588
200
1000
T-' — r
T r-
'O o o o
_l I ■ . I
22 26 30 34
MANTLE LENGTH, mm
42
Figure 5. — Vertical distribution oiAbralia trigonura. Symbols
as in Figure 1.
YOUNG: VERTICAL DISTRIBUTION AND PHOTOSENSITIVE VESICLES
Photosensitive Vesicles (Figure 6)
The arrangement of organs is similar to Ab-
raliopsis sp. described by Young ( 1973). Four sets
of organs are present; all lie adjacent to the
cephalic cartilage. One set is located dorsally, one
posteriorly, and two ventrally. Each dorsal organ
is situated in a concavity of the cephalic cartilage
at the posterodorsal edge of the head. The organ is
compact, dorsoventrally flattened and circular to
triangular in outline. The posterior organs are
located on the posterior surfaces of the optic lobes.
Each is approximately elliptical and very flat. The
posterior organs have a strong yellow pigment
which does not fade after fixation. Other organs
contain an orange pigment that is lost after fixa-
tion. The posterior organs lie immediately an-
terior and lateral to the opaque liver and directly
anterior to the attachment zone of the transparent
head retractor muscles (Figure 6B). The ventral
organs on each side consist of two, narrowly joined,
flattened lobes. One of these, the anteroventral
lobe, is located somewhat anterior and medial to
the other. The anteroventral lobe has its medial
and anterior ends in a deep depression of the
cephalic cartilage. The anteromedial edge of each
lobe nearly makes contact with its counterpart of
p PV
Figure 6. — A. Photosensitive vesicles oiAbralia trigonura . B.
Ventral view of A. trigonura with portion of mantle removed.
This illustration shows the relationship between the posterior
photosensitive vesicles, the opaque liver (L), and the mantle
cavity. Abbreviations as in Figure 2.
the opposite side. The more posterior of the two
lobes, the midventral lobe, lies between the ven-
tral surface of the optic lobe and the cephalic carti-
lage.
Circular windows, similar to those described in
Abraliopsis sp. (Young 1973), and characterized
by a reduced number of chromatophores, are pres-
ent above each dorsal organ. A large ventral win-
dow, totally lacking chromatophores, lies on the
ventral surface of the head above the funnel and
below the ventral vesicles.
Ahralia astrosticta Berry 1909
Vertical Distribution
No specimens were captured during the present
program. However, one specimen was taken in a
gill net set overnight by T. Clarke on the bottom in
180 m. The National Marine Fisheries Service,
NOAA, has captured 44 juveniles (7-38 mm ML
(Mantle length)) in pelagic trawls between 10 and
130 m at night and 10 adults in a benthic shrimp
trawl at 110 m at night near Hawaii. The type was
captured in a bottom dredge between 354 and 650
m, presumably during the day. Roper and Young
(1975) indicated that this animal lives near the
ocean floor even when migrating.
Photosensitive Vesicles
The organs and associated windows are basi-
cally the same as in A. trigonura.
Abraliopsis sp. A
Vertical Distribution (Figure 7)
Sixty-seven specimens were captured. Exclud-
O 600 -
800 -
-r^
••
•*
••
•
•
•
••
•
•
•
' 1
•
[ ■
— r
• •
•
I
I
-
•
•
•
• •
••
•
-
■
-
-
o
■
0
o
_
o
o
o
0
0 o
o
o
_
■
o
o
0
o
o
o
0 "
-
-
-
i-
I
1 1
1
1
i
22 26 30 34
MAN TIE LENGTH, mm
Figure 7. — Vertical distribution of Abraliopsis sp. A. Symbols
as in Figure 1.
589
FISHERY BULLETIN: VOL. 76. NO. 3
ing presumed contaminants, specimens captured
during the day came from depths of about 475 to
700 m; 80% were taken between 550 and 700 m. At
night, captures were made between about 20 and
200 m; nearly 80% were taken in the upper 100 m.
Photosensitive Vesicles
The organs and associated windows have been
described in detail by Young ( 1973); they are simi-
lar to those of Abralia trigonura.
Ahraliopsis sp. B
Vertical Distribution (Figure 8)
During the day, 23 specimens taken by the
Tucker trawl indicate a depth range of 500 to 650
m; most captures came from 500 to 600 m. IKMT
data lumped into 100-m increments indicate most
specimens came from 400 to 700 m.
At night, 19 specimens from the Tucker trawl
probably came from depths between 50 and 100 m.
IKMT data lumped into 50-m increments show a
strong peak in the 50- to 100-m interval. A few
IKMT captures were made as shallow as 15 m.
Photosensitive Vesicles (Figure 2B)
The dorsal and anteroventral organs are similar
to other species of Abraliopsis and Abralia. The
posterior lobes, however, are absent, and the mid-
ventral lobes are enlarged and extended dorsally.
In addition, a thin string of vesicles extends from
each midventral lobe dorsally between the brain
and the optic lobe to join with the dorsal lobe. The
structure of this string is slightly variable, and in
some specimens the vesicles found about at mid-
T T '
,(
•
}
' ^\ 1.'.
111.
T
— r-
•
— 1 1
• •
200
-
-
400
-
_
600
o
66 o6
'■''0'' o
b
6
o
6
0
f
o
o
9
too
innn
. 1 1
1
'
1
1 1
■
1
1
brain level are slightly enlarged and elongate
(Figure 2B). The yellow pigment characteristic of
the posterior lobes in related species does not occur
in any of the lobes of this species.
Ahraliopsis sp. C
Vertical Distribution (Figure 9A)
Only 12 specimens were captured. During the
day, five specimens were taken between 500 and
600 m. At night, all of the captures were made in
the upper 100m.
o 600 -
*
1
••
' — r ' 1
t
\
-
0
D
-
-
O
-
-
> 1
1 . 1
-
20 30 to 50
MANTLE LENGTH, m m
m ANTIE It NGTH
MANTLE LENGTH, ««
Figure 8. — Vertical distribution oi^ Ahraliopsis sp. B, Symbols
as in Figure 1.
Figure 9. — A, Vertical distribution ofEnoploteuthis sp. A (cir-
cles ) and Enoploteu this sp.B( squares ) . B . Vertical distribution
of Ahraliopsis sp. C (circles) and Thelidioteuthis allesandrinii
(triangles). Symbols as in Figure 1.
Photosensitive Vesicles
The organs are similar to Abralia trigonura and
Abraliopsis sp. A except that the posterior lobe is
smaller, slightly more medially located, and con-
tinuous with the midventral organ. This latter
connection, however, does not have the yellow
pigment that the posterior organ possesses. Also, a
few scattered vesicles lie on the posterior margin
of the nerve from the dorsal organ.
Erioploteuthis sp. A
Vertical Distribution (Figure 9B)
During the day, two captures were made be-
tween 500 and 600 m; and at night, three captures
were made in the upper 100 m.
Photosensitive Vesicles (Figure lOA)
The organs are similar to Abralia trigonura
590
YOUNG: VERTICAL DISTRIBUTION AND PHOTOSENSITIVE VESICLES
^^ - — - D PV Photosensitive Vesicles (Figure I OB)
Three sets of organs, dorsal, posterior, and ven-
tral, consist primarily of a loose association of var-
iously shaped, mostly independent vesicles
(Young 1977). The organs are broad, flat struc-
tures, with the greatest concentration of vesicles
along the lateral margins of the organs. The or-
gans lack yellow pigment.
Family Ommastrephidae
Symplectotenthis oualaniensis
(Lesson 1830)
PPV
Figure lO.— A. Photosensitive vesicles oi Enoploteuthis sp.
A. B. Photosensitive vesicles of Thelidioteuthis alessandrinii .
Abbreviations as in Figure 2.
with the following exceptions. The midventral
organ has a more irregular shape, is less compact,
and has a narrow connection with the posterior
organ. The posterior organ is continuous with the
dorsal organ via a strand of vesicles that extends
over the optic lobe. Except for a short segment
adjacent to the dorsal organ, this strand contains
yellow pigment as does the posterior lobe.
Enoploteuthis sp. B
Vertical Distribution (Figure 9B)
One specimen was captured during the day at
515 m and three were taken at night between 50
and 150 m.
Photosensitive Vesicles
The vesicles are the same as those of Enop-
loteuthis sp. A except for some differences in the
posterior organ. The posterior organ in Enop-
loteuthis sp. B is more elongate, more medially
located on the optic lobe, and lacks yellow pig-
ment.
Thelidioteuthis alesandrinii (Verany 1851)
Vertical Distribution (Figure 9A)
During the day, one specimen was taken in an
opening-closing tow between 720 and 780 m. At
night, three specimens were taken in open tows
between 80 and 100 m.
Vertical Distribution
Except for larvae, only one specimen was cap-
tured in the midwater trawls. This specimen was
taken at night by the IKMT which fished at 100 m.
This fast-swimming squid normally avoids our
trawls. Members of this species are commonly
seen at the surface at night around the night-light
and a number have been dipnetted. Little is
known, however, of their day distribution, al-
though Young (1975b) had assembled evidence
which indicates that they live in the upper few
hundred meters but may descend on occasion to
great depths.
Photosensitive Vesicles (Figure IIA)
Three sets of organs are present: a dorsal, cen-
tral, and ventral set. The ventral organ lies within
the cephalic cartilage at the posterior end of the
head and immediately above the posterolateral
portion of the funnel. It consists of a series of flat,
D PV
Figure ll. — A. Photosensitive vesicles of Symplecoteuthis
oualaniensis. B. Photosensitive vesicles of //ya/o24 mm ML and
all specimens captured above 600 m during the
day. One hundred fourteen specimens are plotted.
Although only a few shallow day captures were
made, the vertical distribution pattern is clear:
animals between 5 and 15 mm ML predominate in
599
FISHERY BULLETIN: VOL. 76, NO. 3
'tttt ♦
? 9
,o 4
i o
•••o'w
P? • o "" . 9
o o otPO°
• 900 O
10 20 30 10 50 60
MANTLE LENGTH, mm
Figure 29, — Vertical distribution of Liocranchia valdiviae.
Symbols as in Figure 1.
the upper few hundred meters. Descent to adult
depths begins within the 5- to 15-mm ML size
range or occasionally larger. Most specimens 15 to
25 mm ML are captured between depths of 500 and
700 m, while most animals >25 mm ML are found
deeper than 700 m with progressively larger
specimens found at progressively greater depths.
Diel vertical migration does not occur. Five large
specimens captured at depths of 40 to 525 m at
night, however, indicate that some specimens oc-
casionally wander into the upper depths at night.
Mature specimens were not captured.
Photosensitive Vesicles (Figure 30A)
Liocranchia valdiviae has a single set of small
organs. Each organ is elongate and extends along
the posterior side of the optic stalk. Each organ
usually consists of three elongated vesicles. A
strip of dark brown screening pigment with ir-
regular margins extends along much of the an-
teromedial edge of the ventral half of the organ.
The broad dorsal vesicle either lacks screening
pigment or has only a trace of it. The slender
middle vesicle has a narrow, often discontinuous
strip of pigment which widens ventrally. The ven-
tral vesicle, which is the largest, has a broad, con-
tinuous layer of screening pigment.
The vesicles of L. valdiviae grow allometrically.
At 30 mm ML the vesicles are small, and screening
pigment consists of a single small patch on the
ventral vesicle. In the largest specimen ( 102 mm
ML), the pigment screen is very extensive and
covers much of the anterior surface of the dorsal as
well as ventral portions of each organ. The dorsal
and ventral vesicles in each organ are somewhat
broader in this specimen, making the organ more
dumbbell shaped.
Liocranchia reinhardti (Steenstrup 1856)
Vertical Distribution (Figure 31)
All 12 juvenile specimens were captured at
night. Ten of the 12 specimens were taken in the
upper 100 m; the other 2 came from 150 to 200 m. A
single mature specimen was captured at 775 m in
a Tucker trawl that failed to close on retrieval.
This specimen was a female that had recently
spawned: remnants of what appeared to be sperm
reservoirs were attached to the inner right wall of
the mantle near the base of the funnel; the nida-
mental glands were gelatinous and extremely
swollen; the ovary was depleted; and the muscular
tissue of the mantle, fins, head, and arms was
flaccid.
Unfortunately, there are no data on the day
distribution of this species in Hawaiian waters.
•
•
••
1 1 1 1
•• T
I 1
•
200
.
_
400
-
-
600
-
-
800
1
1 . 1 1
•
1 1
Figure 30. — A. Photosensitive vesicles of Liocranchia val-
diviae. B, Photosensitive vesicles of L. reinhardti. Abbrevia-
tions as in Figure 2,
90 130 170
MANTLE LENGTH, mm
Figure 31, — Vertical distribution of Liocranchia reinhardti.
Symbols as in Figure 1,
600
YOUNG: VERTICAL DISTRIBUTION AND PHOTOSENSITIVE VESICLES
However, in the tropical North Atlantic, two
specimens of L. reinhardti (44 and 48 mm ML)
were captured during the day in opening-closing
tows between 510 and 600 m, while a 75-mm ML
specimen was taken in an open tow that fished
between 390 and 800 m (C.C. Lu pers. commun.).
Also, M. Clarke ( 1969) reported specimens of 46
and 69 mm ML from depths between 450 and 810
m during the day in the Atlantic.
Photosensitive Vesicles (Figure 30B)
The organs of L. reinhardti have been described
by Messenger (1967a). This species has a single
large set of organs lying along the posteromedial
surface of the peduncle lobe. Each organ consists of
a linear array of 20 to 25 tightly packed vesicles.
The vesicles are elongated in a transverse direc-
tion except for those at the dorsal and ventral ends
which are nearly circular. The vesicles are sepa-
rated from one another by a heavy brown pigment
screen which also covers most of the convex an-
terior side of the organ. Most vesicles within each
organ thus form elongate cups which presumably
admit light only from one surface. The dorsal vesi-
cle, however, lacks screening pigment from the
posterior lateral and dorsolateral surfaces. The
ventral vesicle is larger, with photosensitive pro-
cesses twice as long as those of the dorsal vesicle. It
lacks pigment on its ventral and anterior surface.
The curvature of the organ and the arrangement
of screening pigment allows light to enter dif-
ferent vesicles from a wide range of angles.
Specimens ^47 mm ML have no screening pig-
ment on the vesicles while those 2*70 mm ML
exhibit pigment as described above. The vesicles
in the largest specimen (spent female) are slightly
larger (especially the ventral vesicle) relative to
the brain size than in smaller specimens.
Leachia pacifica (Issel 1908)
Vertical Distribution (Figure 32)
The vertical distribution of L. pacifica has been
described elsewhere (Young 1975a). This species
reaches about 809^ of its maximum length in
near-surface waters. Large specimens (45-60 mm
ML) are found throughout the water column be-
tween 30 and at least 1,800 m with those taken
from progressively deeper water exhibiting pro-
gressively greater sexual maturity. Gravid
females were taken at depths > 1,300 m.
200
400-
600
800
)000
6 )200
I
t—
a.
Q )400
16001-
1800
2000
22001-
2400,
■
■>■•' ' ' _
;
°8
0
•
•
o
•
0 "■
8
•
°0
o
0
0
o
•
1
•
o
1
0
m *
• —
1 —
1 1 1 1 1 1 1
1 1
1 1 1
10 20 30 40 50 60
MANTLE LENGTH, mm
Figure 32.— Vertical distribution of Leachia pacifica. From
Young ( 1975a). Symbols as in Figure 1.
Photosensitive Vesicles (Figure 33A)
Leachia pacifica has a single set of organs lo-
cated on the posteroventral surface of the peduncle
lobe. Each organ consists of 4 or 5 cup-shaped
vesicles that are closely packed into a small oval
organ. A dark brown screening pigment covers
Figure 33. — A. Photosensitive vesicles of Leachia pacifica . B.
Photosensitive vesicles of Sandalops melancholicus. Abbrevia-
tions as in Figure 2.
601
FISHERY BULLETIN: VOL. 76, NO. 3
much of the anterior and slightly dorsal surface of
each organ. This pigment is also found in the walls
between some vesicles, tending to isolate them
from one another. Although the vesicles are mi-
nute in the larva and very small in the adults, a
small positive allometric growth of the vesicles
seems to occur. Screening pigment first appears on
the vesicles between about 20 and 30 mm ML. In
the adult, the screening pigment is most extensive
and covers the entire anteromedial surface of the
organ.
Phasmatopsis fisheri (Berry 1909)
Vertical Distribution (Figure 34)
Over 300 specimens of P. fisheri were captured
but most were larvae. Metamorphosis occurs at a
size of 40 to 50 mm ML.
During the day, six larvae were captured be-
tween 150 and 250 m. Seventeen juvenile and
adult specimens were captured between about 625
and 800 m; most captures were made between 650
and 775 m.
At night, larvae «30 mm ML were taken
primarily in the upper 50 m; larvae 31 to 40 mm
ML were found throughout the upper 200 m. Four-
teen juveniles and adults were taken at night be-
tween 90 and 225 m; most captures were made
between 100 and 200 m.
Photosensitive Vesicles (Figure 35)
Phasmatopsis fisheri has a single set of large
organs. Each organ consists of a broad, elongate
vesicle that extends from the optic gland on the
dorsal surface of the optic stalk ventrally over the
posterior surface of the peduncle complex onto the
side of the ventral subesophageal mass, where it
A PIG
PIG.S
Figure 35. — Photosensitive vesicles oiPhasmatopsis fisheri. A.
Larva, 35 mm ML. B. Juvenile, 70 mm ML. C. Adult, 130 mm
ML. Abbreviations as in Figure 2.
bends slightly dorsally. Each organ is thick later-
ally and medially. Most of the anterior, medial,
and lateral surfaces of each organ are covered by
dark brown pigment screen. The dorsal tip of each
organ lacks screening pigment on its lateral por-
tion and has limited pigment screen on its medial
portion. The curvature of each organ allows light
to enter various parts from a wide range of angles.
The anterior wall of each organ consists of little
more than a membrane backed by dense pigment.
The posterior wall and the walls of the dorsal and
ventral ends of each organ contain 4 or 5 layers of
sensory-cell bodies. Sensory processes are longer
(215 /Ltm) and thinner (inner diameter 2 to4/>tm) in
the ventral parts of the organ than in the dorsal
200 -
4 00 -
Figure 34.— Vertical distribution of
Phasmatopsis fisheri. Symbols as in
Figure L
0 600 -
800
m^//^)9om'////A^^ »!■ — ' — I — ' — I — ' — I — ' — t — ' — I — ' — r^'-T-'^
, • •«( •
lOOo' I'll — ^-
J J L
10 20 30
40
50 60 70
MANTLE LENGTH, mm
80
90
I *f \ \ I L.
100 130 170
602
YOUNG: VERTICAL DISTRIBUTION AND PHOTOSENSITIVE VESICLES
parts (length 155 /xm, inner diameter 3 to 6 ju,m).
The processes are long and slender and organized
in a straight, parallel alignment.
Each organ is small in larvae and lacks screen-
ing pigment. At 35 mm ML, the vesicles form a
narrow strip along the posterior surface of the
peduncle lobe (Figure 35A). The largest larvae
have relatively small organs without screening
pigment; the youngest juveniles have large organs
that are heavily pigmented. In the juvenile and
adult stages, the organ exhibits positive allomet-
ric growth (compare Figure 35B and C). In the
adult stages, the organ exhibits positive allomet-
ric growth (compare Figure 35B and C). In the
brain. The ventral half of the organ is particularly
enlarged and the organs on each side of the brain
contact broadly (but do not fuse) below the ventral
midline of the brain.
Taonius pavo (LeSueur 1821)
Vertical Distribution (Figure 36)
The vertical distribution of T. pavo has been
described by Young ( 1975d). Larvae probably live
in the upper 400 m, although only one capture was
made. Juveniles were found primarily between
600 and 650 m, and adults were captured between
725 and 970 m. Diel vertical migration does not
occur.
u
-, ^
1 1 (—
' T '
T '
I
1 '
< I
200
-
•
"
eioo
-
• •
,
-
^=600
-
b
-
o
"
0
b
o
-
80C
\r\ne\
1
1 ■ , 1
1
1 1
1
1 i
0
o
-
40 60 80 100 IJO HO 160 180 200 220
WANTIE LENGTH, mm
Figure 36. — Vertical distribution ofTaoniuspavo. From Young
U975d.). Symbols as in Figure 1.
F1GL!RE 37. — Photosensitive vesicles of Taonius pavo . A. Lar-
va. B. Juvenile, 140 mm ML. C. Adult, 220 mm ML.
lumen is unoccupied. Sensory processes occupy
about V5 (i.e., about 230 ^im) of the lumen diame-
ter on the anterior, posterior, and dorsal sides and
slightly more (about 300^t.m) on the ventral sides.
The processes are loosely packed and intertwined
to a large extent dorsally and more tightly packed
ventrally. Inner diameters of the processes vary
greatly from about 3 to 30 ^im. The wall on the
dorsal half of the organ contains about two layers
of sensory-cell bodies compared with about three
layers ventrally. In a 140-mm ML juvenile, the
dissected vesicle was also hollow, with an even
thinner region of the lumen occupied by sensory
processes. The organs exhibit positive allometric
growth (Figure 37).
Sandalops melancholicus Chun 1906
Vertical Distribution (Figure 38)
The vertical distribution in this species has been
reported by Young ( 1975d). Larvae were found in
the upper 400 m. Juveniles were captured between
450 and 674 m, and two adults were captured near
800 and 1,075 m. Diel vertical migration does not
occur.
Photosensitive Vesicles (Figure 37)
Taonius pavo has a single set of organs located
on the posteroventral side of the peduncle com-
plex. Each organ consists of a single oval vesicle.
No screening pigment is present. The large size of
the vesicle in a 220-mm ML specimen belies its
internal structure. The large central region of the
Photosensitive Vesicles (Figure 33B)
Sandalops melancholicus has a set of organs
located along the ventral surface of the peduncle
complex. Each organ consists of a single bilobed
vesicle (R. E. Young 1977). Slight positive allo-
metric growth of the vesicles occurs between the
juvenile and adult stages.
603
FISHERY BULLETIN VOL 76, NO 3
200-
400-
-600-
800-
1000-
1200
-
i.
*
\ 6
'c
1
'III
i
0 OQO •
ooj 0
•
, 1 , 1
' 1 1 1
•
, 1 , 1
1
1
20 40 60 80
MANTLE LENGTH, mm
100
Figure 38. — Vertical distribution of Sandalops melancholicus.
From Young ( 1975d). Symbols as in Figure 1.
Helicocrarichia heehei Robson 1948
sively greater depths, although the relationship of
size to depth is not very precise. The deepest cap-
ture was probably at 1,200 m. Mature specimens
were not captured.
Photosensitive Vesicles (Figure 40A)
One set of organs is present. Each organ consists
of a single small oval vesicle located on the poste-
rior surface of the peduncle complex. No screening
pigment is present. Very slight, if any, positive
allometric increase in the size of the vesicles oc-
curs from juveniles to the largest specimens.
Bathothanma lyromma Chun 1906
Vertical Distribution (Figure 41)
Although only 12 specimens were captured, a
general pattern of ontogenetic descent is evident.
Vertical Distribution (Figure 39)
Including larvae, 47 specimens were captured.
Although day and night captures are not well in-
termingled in Figure 39 (due largely to sampling
inequities), the data indicate that this species does
not migrate. Rather, it seems to undergo on-
togenetic descent. The youngest specimens were
captured between 100 and 200 m. Progressively
larger specimens were generally taken at progres-
FlGURE 40. — A. Photosensitive vesicles of Helicocrarichia
beebei. B. Photosensitive vesicles of Bathothauma lyromma.
Abbreviations as in Figure 2.
200-
400
6 600
X
a.
S 800-
1000-
1200-
1400
"1 1 1-
I '^
—tfi* . • 4 .^i
« •
10 20 30 40 50 60
MANTLE LENGTH, mm
1200-
20
40 60 80 100
MANTLE LENGTH, mm
140
Figure 3.9. — Vertical distribution of Helicocrarichia heehei.
Symbols as in Figure 1.
Figure 41. — Vertical distribution of Bathothauma lyromma.
Symbols as in Figure 1.
604
YOUNG: VERTICAL DISTRIBUTION AND PHOTOSENSITIVK VESICLES
Day and night captures were in the same depth
range, indicating that diel vertical migration does
not occur. The three specimens captured at the
greatest depths were gravid females. The speci-
men captured at 910 m had sperm receptacles im-
bedded in the back of the head and in the an-
terodorsal surface of the mantle. The nidamental
and oviducal glands were greatly enlarged and the
entire visceropericardial coelom was packed with
large eggs. The muscular tissue was slightly
flabby. The specimen captured at 1,125 m exhi-
bited almost identical features. The specimen cap-
tured at about 1,100 m had similarly placed sperm
reservoirs, less extensively enlarged nidamental
and oviducal glands, and lacked eggs (apparently
due to damage during capture*. This specimen
exhibited no sign of muscular degeneration. The
mantle cavity of this specimen had two very long
arms from another specimen (presumably a male)
attached to the inner wall of the mantle. The
largest specimen was an immature female. Its size
was largely due to its fixation in a relaxed state. In
this species, the pen is extraorinarily delicate and
accurate measurements of contracted, crumpled
specimens are nearly impossible.
Photosensitive Vesicles (Figure 403)
Bathothauma lyrommQ has a single set of or-
gans. Each organ consists of a flat oval vesicle
located on the posteroventral surface of the pedun-
cle complex. No screening pigment is present.
Slight positive allometric growth of the vesicles
occurs from juveniles to adults.
Galiteuthis pacifica (Robson 1948)
Vertical Distribution (Figure 42)
The 27 specimens captured indicate a broad ver-
tical range for this species. Fourteen of the 19
captures of specimens >20 mm ML came from
depths of 700 m or more. The data indicate that
diel vertical migration does not occur.
Photosensitive Vesicles
The vesicles of this species are similar to those of
G. phyllura described by Young ( 1972a). A single
set of organs is present. Each organ consists of a
large oval vesicle attached to the posteroventral
surface of the peduncle complex. Considerable
positive allometric growth of the vesicles occurs.
0| i r
200
400-
E
x"
■600
800-
1000-
1200
1 — ' — r
6oo o
o 9 •■ o o
1350m
.J I . 1 I
10 20 30 40 50 260
MANTLE LENGTH, mm
FiGL'RE 42. — Vertical distribution of Galiteuthis pad fica . Sym-
bols as in Figure 1.
Order Octopoda
Family Bolitaenidae
Eleciouella pygmaea Verrill 1884
N'ertical Distribution (Figure 43)
Eighty specimens were captured. Day and night
captures were in the same depth range (except
above 300 m where day trawling was minimal),
indicating that diel vertical migration does not
occur. Most specimens between 5 and 15 mm ML
\^ o • •
L». ,,.,,, J' f»
-I 1 1 r
-| r
"nil
'9 J 0,
nt,
20 ;5 30
M AN T I E le NGTM,
Figure 43. — Vertical distribution oi Eledonella pygmaea. Cir-
cles with crosses represent brooding females. Double circle rep-
resents a gravid female. Otherwise symbols as in Figure 1.
605
FISHERY BULLETIN: VOL. 76, NO. 3
were captured either around 200 m or below 600
m. Apparently the size at which young begin their
descent to adult depths is rather variable. The
deep captures exhibit a clear pattern of ontogene-
tic descent. At 25 mm ML or larger all specimens
(excluding brooding females) were captured be-
tween depths of 975 and 1,425 m. Four females,
apparently brooding, were captured between
about 800 and 870 m.
The pigmentation of the female changes as she
becomes gravid: the chromatophores over the
mantle and especially over the aboral surface of
the arms and web become more numerous, and the
oral surfaces of the arms and web develop an even
denser pigmentation. Nearly all iridophores are
lost. At the same time the arms and the web be-
come thicker. The web between the dorsal six arms
becomes more extensive, and the web between the
two ventral arms is reduced. These dark octopods
spawn and apparently brood their young (Young
1972b). Five specimens taken from horizontal
tows exhibited this increased pigmentation. In
four cases, the ovary was depleted, and in the fifth,
captured at 1,400 m, the eggs were not fully ma-
ture, but were considerably larger than in an im-
mature female of approximately the same size. In
two cases egg strings with developing embryos
were found in the same trawl with dark and pre-
sumably brooding females.
No mature males were taken. However, judging
from the development of the hectocotylus, the
penis, and the spermatophore glands, two speci-
mens captured at 1,200 and 1,425 m were nearly
mature. Another slightly less mature specimen
was taken at 1,325 m. Three still less mature
specimens were taken between 1,175 and 1,200 m.
while a large male taken at 1,025 m was the least
developed of all.
Photosensitive Vesicles (Figure 44 A)
The photosensitive vesicles consist of a single
pair of organs; each organ is a spherical vesicle
attached to the posterior margin of the stellate
ganglion.
Japetella diaphana Hoyle 1885
Vertical Distribution (Figure 45)
Seventy-four specimens were captured. Diel
vertical migration does not occur. Specimens <20
mm ML were captured mostly in two regions, be-
tween 170 and 270 m and between 500 and 800 m.
Oj 1 r— — I— — r-
<^ •
-1 1— I 1 1 r
,t.
Wo
.w?
%
* o
9
II
I . I
0 ®
20 30 40 50 60 70
MANTLE LENGTH, m •
Figure 45. — Vertical distribution of Japetella diaphana. Cir-
cles with crosses represent brooding females. Double circles rep-
resent gravid females. Otherwise symbols as in Figure 1.
*"^.
\
STG
STG
Figure 44. — A. Section through the photosensi-
tive vesicle of adult Eledonella pygmaea. B.
Photosensitive vesicles ofAmphitretuspelagicus.
ST. G. — Stellate ganglion. Otherwise symbols as
in Figure 1.
B
606
YOUNG: VERTICAL DISTRIBUTION AND PHOTOSENSITIVE VESICLES
where they exhibited an ontogenetic descent. The
depth range for specimens 5^20 mm ML was 725 to
1,065 m; nearly 909^ of the animals occur between
700 and 950 m and nearly 60'7f between 750 and
850 m. Gravid and brooding females were found at
the extremes of this range. Two gravid females
were captured at 1,050 and 1,065 m while three
spent and presumably brooding females were
taken between 725 and 800 m. As in£'. pygmaea,
the gravid and spent females have a very heavy
pigmentation and lack most of the iridophores
present in younger specimens. Five such females
were captured in horizontal tows. One gravid
female with a sperm mass embedded in the
gelatinous tissue between the second and third
arms was taken at 1,050 m. Another taken at
1,065 m had been gutted in the trawl but had not
spawned: the musculature was firmer than in
spent females, and the catch contained a large
number of octopod eggs which undoubtedly came
from the ruptured ovary. Three specimens taken
between 725 and 800 m probably had spawned:
two had depleted ovaries and the third was gutted
but had deteriorated musculature. In the same
tow with the last specimen were four newly
hatched larvae, presumably from the brood of the
female. One large, heavily pigmented female
taken in an oblique tow had the remnants of an
egg string dangling from one of the large suckers
of the third arm. Two eggs were completely en-
gulfed by the sucker, while a third dangled from
the broken egg string extending from the sucker.
No mature males were taken.
0
•
1
.
1 1
300
-
I
-
400
-
-
•
-
600
" ° i
-
■
•
• 00
-
•
■
•
1000
, — 1 — , — 1
1
MANTiE LENGTH
0
> 1 ' I
' I
-
200
-
—
400
-
-
600
-
-
8 00
_
O
•
_
.
o
.
1000
•
o
-
1200
_ • ■
,i\ , 1
1
-
20 40 60
MANTLE \.l NGTH.m m
Figure 46. — A. Vertical distribution of Amphitretus pelagicus
(squares) and Vitreledonella richardi (circles). B. Vertical dis-
tribution ofVampyroteuthis infernalis . Half-closed circles repre-
sent a twilight capture. Otherwise symbols as in Figure 1.
Photosensitive Vesicles
The vesicles are as in E. pygmaea.
Family Amphitretidae
Amphitretus pelagicus Hoyle 1885
Vertical Distribution (Figure 46A)
Two specimens were taken at night in the upper
350 m.
Photosensitive Vesicles (Figure 44B)
Amphitretus pelagicus has one set of organs.
They lie on the stellate ganglia immediately an-
terior to the entry points of the pallial nerves.
Each organ consists of a large complex of a dozen
or more generally circular vesicles which cover
most of the anterior wall of the ganglion.
Family Vitreledonnelidae
Vitreledonnella richardi Joudin 1918
Vertical Distribution (Figure 46A)
Four specimens were captured. One small
specimen was taken in an oblique twilight tow
between the surface and 400 m. Two other small
specimens were taken between 600 and 650 m
during the day. One large specimen was captured
at 775 m during the night.
Photosensitive Vesicles
An organ consisting of a single spherical vesicle
is located on the posterior margin of each stellate
ganglion.
Order Vampyromorpha
Family Vampyroteuthidae
Vampyroteuthis iuferualis Chun 1903
Vertical Distribution (Figure 46B).
Eleven specimens were captured. Ten of the 11
were taken between depths of 800 and 1,200 m.
The remaining specimen came from an open ob-
lique tow that fished between 1,100 and 1.900 m.
Diel vertical migration does not occur.
607
FISHERY BULLETIN: VOL. 76, NO. 3
Photosensitive Vesicles
The vesicles have been described in detail by
Young (1972a). One set is present. They are lo-
cated in the dorsal wall of the mantle cavity at the
base of the funnel. Each organ consists of a small
cluster of spherical vesicles.
Order Sepioidea
Heteroteuthis hauaiiensis (Berry 1909)
Vertical Distribution (Figure 47)
The distribution of this species has been discuss-
ed by R. E. Young ( 1977). During the day, speci-
mens ssl7 mm ML were taken between 250 and
350 m; larger specimens were taken between 375
and 650 m. At night, most specimens <17 mm ML
came from depths between 150 and 200 m; larger
specimens were taken between depths of 110 and
550 m. Males and females mature at about 15-16
mm ML.
Photosensitive Vesicles (Figure 48)
Two sets of organs are present (R. E. Young
1977). The more dorsal set lies on the posterior
margin of the peduncle complex and consists of a
short and narrow string of tiny vesicles. An even
200
E
=E 400
a.
uj
600
SOOh
' 1
rau.
^10
••
-1 — ' — 1 — ' — r ' 1 ' 1
-1 —
9
1 '
•
T
6 9
-r- .
•
•
6
-r
0
' — T" '
• •
o
00
1 '
9
1
•
. O 0 * ^ •
O CD OOO 0
. . . %
-
"
' 1
1 , .1. . 1 , 1 .,. 1
1
1 i
1 1
1
1 ■
1 1
1
10 12 U 16 18
MANTLE LENGTH, mm
20
22
Figure 47.— Vertical distribution of
Heteroteuthis hawaiiensis. From R. E.
Young (1977). Symbols as in Figure 1.
24
26
Figure 48. — Photosensitive vesicles oi^ Heteroteuthis
hawaiiensis. In this figure the outline of portions of
the head and mantle are superimposed to give a clear
perspective of the peculiar arrangement of vesicles in
this species. EYE— eye; FUN.— funnel; MAN.—
mantle. Otherwise abbreviations as in Figure 2.
MAN
608
YOUNG: VERTICAL DISTRIBUTION AND PHOTOSENSITIVK VESICLES
narrower string of tiny vesicles extends ventrally
around the eyes and joins a rather large but ex-
tremely thin, loosely associated group of vesicles
that lies over the lateral base of the funnel.
DISCUSSION
Vertical Distribution
The numbers of cephalopod species taken in
different 100-m depth zones for the upper 1,400 m
showed a broad peak between 500 and 800 m dur-
ing the day (Figure 49). An abrupt increase in the
number of species near 400 m was obscured by the
method of analysis: eight species occurred for the
first time between depths of 375 and 450 m. To
indicate faunal change, the number of species
found for the first time in each zone (i.e., depth
zones containing species upper range limits) were
compared with zones where species found in lesser
depths were absent for the first time (i.e., depth
zones immediately below the lower range limits)
(Figure 49).
The peak at 700-800 m in the summed plot of
species added and species lost indicates that many
species dropped out in the 600-700 m zone and
many were added in the 700-800 m zone. The chart
also indicates that only two species were encoun-
tered for the first time at 800 m or below. One was
the poorly sampled Brachioteuthis sp. and the
other was deep-living Vampyroteuthis infernalis.
The data indicate peak species richness in the
upper few hundred meters with relatively little
change between 300 and 1,000 m during the night
(Figure 49).
Numbers of individuals in different depth zones
in the upper 1,400 m (exclusive of young individu-
als, captures in oblique tows, and contaminants)
were also examined (Figure 50). During the day,
the greatest abundance of individuals occurred
between 400 and 700 m. This peak reflects the
dominance of the enoploteuthids, especially
Pyroteuthis and Pterygioteuthis spp. The high rate
of capture in the 300- to 400-m zone was due in
part to a few species whose upper limits extended
slightly above 400 m. Nevertheless, an abrupt in-
crease in number occurred in the 400- to 500-m
zone. The rates of capture below 1,000 m were
unreliable due to the small amount of trawling.
The night data in the upper 400 m were lumped
into 50-m increments due to greater control over
trawling depths in near-surface waters. The
largest catches at night were made in the upper
200 m. In this region two peaks were apparent
(Figure 50). The peak in the 50- to 100-m zone was
largely due to Pterygioteuthis microlampas, the
NO A NO L A«L TOTAL NO
NIGHT
NO A NO L A' L TOTAL NO
E 600
200
E- 500
0 10 0 10 0 20 0 20
NO SPECIES
0 10 0 10 0 10 0 20
NO SPECIES
Figure 49. — Numbers of species versus depth. The histograms
were based on species ranges from midwater trawl data. Data
were lumped into 100-m depth increments and were not cor-
rected for unequal trawling times at different depths. Data for
some species were very meager. Young stages found in near-
surface waters that can be distinguished by an abrupt change in
habitat or by a metamorphosis have been eliminated from the
figures. No. A — number of species added (i.e., found for the first
time in a given depth zone). No. L. — number of species lost (i.e.,
absent from a given depth zone but present in the shallower
zone). A + L — sum of two previous histograms. Total No. — total
number of species in each depth zone.
-1 — I — I — I — ' — I
0 40 80 120 0 40 80 120
NO. SPECIMENS/ 1000 MIN. TRAWLING
Figure 50.— Total catch rate of numbers of cephalopod speci-
mens from both trawls.
609
FISHERY BULLETIN: VOL. 76, NO. 3
most abundant species in the collection. The peak
in the 150- to 200-m zone was largely due to
Pyroteuthis addolux and Heteroteuthis hawaiien-
sis; young Histioteuthis dofleini also contributed
considerably. YoungEledonella were also found in
this zone although they were excluded from the
figures as "larvae."
The zone between 200 and 700 m was sparsely
inhabited at night. The peak between 700 and
1,000 m represented the deep nonmigrating popu-
lation. The capture rate in this region was almost
identical to the day capture rate at the same
depths: deep-living migrators were few.
The total rate of capture for the water column
during the day was 459 specimens/1,000 min of
trawling. Surprisingly, the total capture rate at
night was only 309 specimens/1,000 min. This dif-
ference was largely due to smaller-than-expected
catches at night in the upper 200 m of the few most
abundant species. The reason for the low night
catches is unknown. Another estimate of the
number of animals in the upper 250 m at night was
obtained by assuming that the day peak from 300
to 700 m (minus the night catch at these depths)
shifted into this upper zone at night (see below).
On this basis, nearly 807c of the individuals occur-
red in the upper 250 m at night. If one considers
also the abundant ommastrephids which avoided
midwater trawls but occurred in near-surface
waters at night, then only a small percent of the
total number of individuals would remain below
250 m at night.
In many species, most of the population shifted
upward at night. Such day-night differences
existed in at least 25 of the 47 species examined,
based on present data and literature records.
Adequate data were lacking for ommastrephids
and Neoteuthis sp. Therefore, where the vertical
ranges are known, nearly 60% of the species ex-
hibited diel shifts in habitat. Species not exhibit-
ing diel migrations belonged primarily to the
Cranchiidae ( seven species ) and the Octopoda ( five
species).
At least 18 of the 25 species that exhibited diel
migrations occurred almost exclusively in the
upper 250 m at night. These included all 11 enop-
loteuthids, Liocranchia reinhardti, Phasmatopsis
fisheri, Ctenopteryx siculus, Octopoteuthis
neilseni, Brachioteuthis sp., Chiroteuthis sp.,
Onychoteuthis compacta, and young Heteroteuthis
hawaiiensis. Two species for which the data were
incomplete (Cycloteuthis serventyi and
Chiroteuthis picteti) probably belonged to this
category as well. Therefore, at least 80% of the
migratory species occurred in the upper 250 m at
night.
Amesbury ( 1975) examined vertical zonation of
midwater fishes during the day off Hawaii. He
concluded that the water column could be divided
into three regions: epipelagic, mesopelagic, and
bathypelagic zones. The boundary between the
epipelagic and mesopelagic zones occurred at
about 400 m and was marked by a sharp increase
in the numbers of individuals. This boundary ap-
peared to apply equally well to cephalopods. The
boundary between the mesopelagic and
bathypelagic zones occurred at about 1,200 m.
This boundary was marked by a noticeable de-
crease in fish numbers and represented the
greatest day depths of vertically migrating fish.
This lower boundary was not applicable to
cephalopods; there was no comparable decrease in
numbers of individuals; and this depth seemed to
be well below the range of vertical migrators.
Amesbury further divided the mesopelagic zone
into upper and lower zones with the boundary at
about 650 to 700 m. Cephalopods exhibited
maximum species turnover at about this depth, as
well as changes in light-related adaptations in
some species (Young 1975d). Although fish and
cephalopod distributions differed with respect to
the lower boundary, the distribution of
cephalopods generally supported Amesbury's zo-
nation.
In spite of the rather small size of the collection,
some evidence of vertical habitat separation
among closely related species emerged. Three of
the more abundant species belong to the subfamily
Pyroteuthinae: Pyroteuthis addolux, Pterygio-
teuthis microlampas. and P. giardi. In general
body proportions and armature, P. microlampas
was more similar to Pyroteuthis addolux than to
its congener. During the day, these two species
occupied the same depths around 500 m. At night,
their populations peaked at distinctly different
depths: Pterygioteuthis microlampas occurred
primarily between 50 and 100 m, while
Pyroteuthis addolux occurred primarily between
150 and 200 m. Although the data were less clear
for Pterygioteuthis giardi, this species seemed to
center around 400 m during the day and in the
upper 100 m at night with about half of the popula-
tion in the upper 50 m. Thus this species was
shallower than its two relatives during the day
and tended to be shallower at night, although
broad overlap occurred with its congener.
610
YOUNG; VERTICAL DISTRIBUTION AND PHOTOSENSITIVE VESICLES
Two species of the genus Abralia occurred off
Hawaii. A 6ra//a trigonura was a common vertical
migrator in the area sampled of the open ocean,
while A. astrosticta was never taken there. Ab-
ralia astrosticta seems to be a vertical migrator
that moves in close proximity to the ocean floor
(Roper and Young 1975).
Two species of Mastigoteuthis were taken. Both
species shared the same day habitat at 700 to 800
m. At night, the population of M. inermis spread
upward in the water column 400 or 500 m. Al-
though the data were few, M. famelica appeared to
spread little or not at all.
One of the clearest cases of habitat separation of
congeners occurred in Liocranchia . Liocranchia
valdiviae was taken in lower mesopelagic depths
during the day and it did not migrate. Liocranchia
reinhardti was taken in near-surface waters at
night and apparently occurred in upper
mesopelagic depths during the day.
Although the octopods Japetella diaphana and
Eledonella pygmaea are placed in separate genera,
they are very closely related. Both were taken in
deep waters and did not migrate. The adults (ex-
cept brooding females) were taken at distinctly
different depths: E . pygmaea occupied depths from
975 to 1,425 m while J. diaphana occupied depths
primarily from 700 to 950 m. Young stages prior to
descent were found in near-surface waters. In this
habitat E. pygmaea was captured primarily at 200
m or just above while J. diaphana was captured
primarily below 200 m. Young stages of both
species in the process of descent occupied depths of
about 400 to 800 m or more. The data indicated,
however, that at any given size, except for those
just beginning descent, the young of £■. pygmaea
occupied greater depths than the young of J .
diaphana.
In the geruis Abraliopsis three species were tak-
en: Abraliopsis sp. A and Abraliopsis sp. C form
the most closely related species pair. The available
data show no obvious habitat differences. Al-
though the more common Abraliopsis sp. A
reached a considerably larger size than species C
(43 mm ML vs. 33 mm ML), young individuals of
species A, however, apparently cooccurred with
species C of the same size. The day and night
habitats of Abraliopsis sp. B were not separable
from its two congeners.
Three other groups of congeners were taken
(i.e., in Enoploteuthis , Histioteuthis, and
Chiroteuthis). No differences in habitats were
found within these groups; however, the data were
extremely sparse.
Reproduction
Young ( 1972b) presented evidence for brooding
in Eledonella pygmaea (incorrectly reported as
Bolitaena microcolyla) and suggested that brood-
ing occurs in all pelagic octopods. Additional evi-
dence from the present study substantiated the
brooding habit for E. pygmaea. In addition, evi-
dence indicating brooding in the octopod Jape^e//a
diaphana was found. This species underwent
changes at maturity similar to E. pygmaea.
Further, newly hatched young have been found in
the same trawl with spent females, and in one case
the remnants of an egg string was found attached
to an arm sucker of such a female.
In both species, gravid or near-gravid females
were taken only at the lower limits of the species'
vertical range. Although mature males were not
taken, those nearest maturity were also taken in
the lower parts of the depth range. Apparently,
mating takes place at the lower depth limits of the
population. Brooding females, on the other hand,
were found only at the upper limit of the adult
population in J. diaphana and only well above the
upper limit of the remaining adult population in
E. pygmaea. The brooding females of both species
occurred around 800 m. Presumably the females
migrate upward to around 800 m either just before
or just after spawning. The increased risk of pre-
dation above 800 m probably pi-events the female
from further upward movement: the numbers of
fishes increase greatly above 800 m (Amesbury
1975), and visual detection of the large silhouette
presented by a brooding female should be possible
above about 750 to 775 m (Young and Roper 1977 ).
The upward movement must be unrelated to feed-
ing since brooding females do not feed (Young
1972b). The upward migration may simply de-
crease the distance the young must travel after
hatching to their larval habitat near 200 m.
A number of cephalopods may spawn at the
lower end of their depth range. Evidence for deep-
spawning was found in several vertically migrat-
ing species. A single spent female of Liocranchia
reinhardti was captured at 775 m at night, well
below its normal night habitat in the upper 200 m.
A single gravid, mated female of Brachioteuthis
sp. was captured at 1,125 m at night; its normal
night habitat is in the upper 200 m. Heteroteuthis
hawaiiensis migrated vertically and exhibited
611
FISHERY BULLETIN: VOL. 76. NO. 3
narrow vertical ranges day and night until sexual
maturity was reached; a poorly defined ontogene-
tic descent then ensued. Unfortunately, no other
clues to spawning depth are known.
The nonmigrating species that exhibit a
gradual ontogenetic descent would be expected to
spawn at the lower end of their range. Indeed, this
appeared to happen in Bathothauma lyromma.
The most dramatic example occurred in Leach i a
pacified. Young (1975a) demonstrated that this
species descends near the time of sexual maturity
from near-surface waters to depths of 1,000 to
2,000 m to mate and spawn.
Larvae of most pelagic oceanic cephalopods
occur in near-surface waters. Upward migration of
larvae would seem to be a formidable task for
species that spawn at great depths. The deep-
living octopods apparently carry their young
partway up presumably to lighten this task.
Perhaps squid egg masses are positively buoyant
and float to the surface. There are a number of
observations of egg masses of pelagic cephalopods
floating at or near the ocean surface (see Clarke
1966). However, these have yet to be shown to
belong to a deep-spawning species.
Photosensitive Vesicles
These vesicular organs were present in all
Hawaiian pelagic cephalopods and they occurred
in a great variety of shapes, sizes, and locations. In
many squids, the organs were subdivided into as
many as four sets of separate organs. In squid, the
organs were always found within the confines of
the cephalic cartilage and were located either on
the optic stalk ( central organs ) or dorsal, posterior,
or ventral to the optic stalk (dorsal, posterior, and
ventral organs, respectively). The separate organs
often faced different directions (i.e., their broadest
surface faced a dorsal, posterior, or ventral direc-
tion).
These separate organs were frequently as-
sociated with distinctive "windows" in the overly-
ing skin bearing few if any chromatophores. Such
windows seem to be unnecessary since most
cephalopods can become quite transparent by con-
traction of their chromatophores. The windows in
combination with the more pigmented surround-
ing skin, however, may restrict light to specific
receptors and thereby improve the directionality
of the organs. In a few cases (e.g., Phasmatopsis
fisheri and Ctenopteryx siculus ), the organ was not
subdivided but elongate and curved, allowing dif-
612
ferent portions of a single organ to face various
directions. A directional response of each portion
was insured either by heavy pigment (e.g., P.
fiaheri) or silvery iridophores (C siculus) which
shielded one surface of the organ. Not all species,
however, had vesicular organs that could dis-
criminate between dorsal, posterior, and ventral
sources of light. Some species had undivided cen-
tral organs (e.g., Sandalops melancholicus,
Taonius pavcj ) without apparent screening devices
which therefore are nondirectional. In others, the
total area surveyed by a nondirectional organ was
restricted by its cryptic position (e.g., Vam-
pyroteuthis). Clearly not all cephalopods use these
organs in the same way.
General trends between organ size and habitat
depth during the day occurred in these animals.
Teuthoids and sepioids found in the upper 400 m
(neritic species and young Heteroteuthis
hawaiiensis) generally had small organs. Species
found primarily between 400 and 700 m generally
had large, complex organs. These included most
enoploteuthids, histioteuthids, probably Dis-
coteuthis laciniosa, Liocranchia reinhardti, and
young Taonius povo. Between 700 and 800 m,
species with large, complex organs (i.e., Cteno-
pteryx siculus, Phasmatopsis fisheri, The-
lidioteuthis allessendrinii, Cycloteuthis serven-
tyi, probably large Octopoteuthis nielseni, adult
Taonius pavo, and Galiteuthis pacificus, and
Bathyteuthis abyssicola) cooccurred with species
which had small organs (i.e., chiroteuthids, Mas-
tigoteuthis, Grimalditeuthis bomplandii, large
Liocranchia valdiuiae, and probably Neoteuthis
sp.). Many of the small-vesicle species had ranges
that extended well below 800 m, where they were
joined by other small-organ species (i.e., Vam-
pyroteuthis infernalis and probably Joubiniteuthis
portieri).
The relationship of organ size to habitat depth
was especially marked in young Phasmatopsis
fisheri. The epipelagic larvae of P. fisheri, which
may grow to 40 and 50 mm ML had small central
vesicles. Upon metamorphosis and descent to the
adult day habitat (650-775 m), the organs became
greatly enlarged (Figure 34). As growth con-
tinued, however, a gradual positive allometric
growth of the organs occurred without a clear in-
crease in habitat depth.
A number of species did not follow these general
trends. Several cranchiids exhibited gradual on-
togenetic descent; one of these {Helicocranchia
beebei) had small organs, while the others {San-
YOUNG: VERTICAL DISTRIBUTION AND PHOTOSENSITIVE VESICLES
dalops melancholicus andBathothauma lyromma)
had large organs. Leachia pacifiva, which had
small organs, spent most of its life in epipelagic
waters and then descended to depths > 1,000 m.
Onychoteuthis compacta seemed to range widely
during the day and had rather large organs (its
habitat, however, is poorly known).
Brachioteuthis had similar organs but probably
occurred below 800 m during the day. The ommas-
trephids had a complex arrangement of organs,
yet these animals were primarily epipelagic. In
juveniles of many species (e.g., enoploteuthids),
the size of the organs (relative to the size of the
brain) may be large; yet their absolute size was
small when compared with adults occupying the
same depths.
Compared with squid, all octopods had small
organs. With the probable exception of the tubular
eyed Amphifretus pelagicus, octopods probably do
not occupy depths between 400 and 700 m during
the day except as juveniles in transit to greater
depths. Amphifretus pelagicus is the only pelagic
octopod that exhibited clear modification of its or-
gans. In contrast to the small organs, each consist-
ing of a single vesicle, of other octopods, this ani-
mal has a larger organ composed of many separate
vesicles.
Presumably the general trends with depth were
related to depth gradients in both downwelling
daylight and bioluminescent light. Downwelling
daylight decreases exponentially with depth.
Bioluminescent activity should increase from 400
to 600-800 m then decline rapidly if numbers of
midwater fishes at various depths (see Amesbury
1975) provide an index to bioluminescent activity
during the day. While many vesicles may detect
both downwelling skylight and bioluminescent
light, we will examine evidence for these two func-
tions separately.
The eyes of some mesopelagic animals can prob-
ably detect silhouettes at depths of 750 to 775 m
(Young and Roper 1977). Presumably some photo-
sensitive vesicles are at least as sensitive as the
eyes, especially when we consider the large size
and apparent high pigment density of some (see
Young 1972a). The large dorsal organs of squid
were positioned so they are exposed to downwel-
ling daylight. Large central organs appeared to be
exposed to this light in species lacking dorsal or-
gans.
Some experimental evidence indicates that
midwater cephalopods detect downwelling light
with these vesicular organs. A number of
cephalopods have been seen to conceal themselves
with bioluminescent light (Young and
Roper 1977). This counterillumination requires
that the intensity of downwelling light is precisely
determined by the animal, and the photosensitive
vesicles seem the likely photoreceptor (Young
1973, 1977). Recently R. E. Young, C. F. E. Roper,
and J. Walters (in manuscr.) covered the dorsal
organs oi Abraliopsis sp. B while it was counteril-
luminating and recorded a 909f drop in its
luminescence. They concluded that the dorsal or-
gans detect downwelling light. Since animals can
detect downwelling light with these organs for
counterillumination, they may use this photic in-
formation for other purposes as well.
Vertical migration in many midwater animals
is closely associated with changing light levels
(Boden and Kampa 1967). Since cephalopods mi-
grate during twilight periods, light cues received
by the vesicular organs may serve to trigger or
regulate their migrations. This view is supported
by three sources of evidence. First, nerves from the
vesicles pass into the peduncle complex of the
brain. Messenger (1967b) suggested, on the basis
of experimental evidence in Octopus, that this
complex is part of a visuomotor system: visual
information from the eyes enables this complex to
exercise control over locomotion. Secondly, ex-
perimental evidence on the function of the photo-
sensitive vesicles in neritic Octopus strongly
suggests that these organs regulate diurnal activ-
ity patterns (R. Houck pers. commun.). Finally,
most migrating cephalopods have large vesicular
organs positioned to detect downwelling light. The
only exceptions are species of Mastigoteuthis and
Chiroteuthis , whose migration patterns are not as
distinct as in other species.
If dorsal and central organs function primarily
in the detection of downwelling light, we may have
a clue to the peculiar arrangement of vesicular
organs in ommastrephids. The ommastrephids
were the only squids that had central organs on
the dorsal surface of the optic stalk as well as
dorsal organs. In Nototodarus hawaiiensis, these
two organs differ morphologically (the small cen-
tral organ has large component vesicles and the
large dorsal organ has small vesicles), but the
organs are adjacent to one another. The structural
differences suggest separate functions for the or-
gans, yet their close proximity indicates that both
will be exposed to the same source of light. The
same argument holds for these organs in other
ommastrephids, although the two organs are
613
FISHERY BULLETIN: VOL. 76, NO 3
somewhat further separated. The ommastrephids
have the unusual habit of usually living in
epipelagic waters but occasionally descending to
great depths (Roper and Young 1975). Perhaps the
central organs function in epipelagic waters while
the dorsal organs operate only in deep water. Cer-
tainly there are considerable problems associated
with a single organ functioning over such a wide
range of light intensities.
Certain photosensitive vesicles appear to detect
bioluminescent light rather than downwelling
skylight. The small vesicular organs in the deep-
living Vampyroteuthis infernalis are shielded
from downwelling skylight (Young 1972a). Ves-
icular organs are present in the blind bathypelagic
octopod Cirrothauma murayi (J. Z. Young in
Packard 1972) which lives in depths where detect-
able surface light is absent. Certain photosensi-
tive vesicles of many other species were shielded
from downwelling light. Such organs presumably
detect bioluminescent light. Young (1973, 1977)
demonstrated that certain vesicular organs in
some species were directly exposed to the animal's
photophores, presumably for counterillumination
purposes.
The detection of bioluminescence is not limited,
however, to the animal's own photophores. With
only a few exceptions all species examined had
some means of "viewing" various parts of the man-
tle cavity with their vesicular organs. In many
cases, the organs seemed precisely placed for this
purpose (see Figure 6). In most species with large
opaque livers (e.g., ommastrephids, enop-
loteuthids, histioteuthids, bathyteuthids, cy-
cloteuthids, octopoteuthids, and mastigoteuthids),
some organs extended laterally or ventrally past
the liver, enabling a "view" of the mantle cavity.
In other species with the liver far back in the
mantle cavity (e.g., cranchiids, Brachioteuthis,
Ctenopteryx), only central organs were present. In
Vampyroteuthis infernalis, the organs lay within
the mantle cavity and could only be exposed by
light originating within this cavity, the funnel, or
at the mantle opening. This animal, like most
other cephalopods, had no photophores in these
locations. The photosensitive vesicles in octopods
were also located within the mantle cavity. The
view of the mantle cavity is obscured only in
Onychoteuthis, Chiroteuthis , Joubiniteuthis, and
Chiroteuthidae gen. sp., although most of these
species could still detect light from within the fun-
nel and at the entrance to the mantle cavity.
Young (1972a) suggested that the photosensi-
tive vesicles in Vampyroteuthis detect small glow-
ing organisms that are carried into the mantle
cavity with the respiratory current. In the deep
sea, a glowing organism within the mantle cavity
could reveal the squid's location and have disas-
terous consequences. J. Z. Young ( 1977) extended
this idea to octopods. Nevertheless, this sugges-
tion seems unlikely to have broad application in
explaining the consistent relationships between
vesicle location and mantle cavity "visibility";
however, no alternative function has been found.
Some squid may detect bioluminescent light
originating outside the animal. The large vesicu-
lar organs in the deep-living Bathyteuthis abys-
sicola are not exposed to its own photophores and
probably detect bioluminescence from animals lo-
cated outside its restricted visual field (Young
1972a). In Ctenopteryx siculus, the elongate ves-
icular organs joined in the midventral line over
the funnel and were there shielded dorsally and
laterally by a thick layer of iridophores. The ven-
tral part of this organ would detect light originat-
ing within the funnel. Yet, the high organization
and sophisticated structure of the vesicles seem
overly matched for such a task. This organ proba-
bly "looks" ventrally through the funnel to the
area below the squid. Similar arguments could be
made for certain lobes in other squids.
The photosensitive vesicles in many cephalo-
pods apparently form an elaborate system for
monitoring bioluminescent light from their own
photophores, from within the mantle cavity, and
from the immediate vicinity of the animal that lies
outside the visual field
The great variety of photosensitive vesicles
found among the species of pelagic cephalopods off
Hawaii presumably reflects a variety of functions
for these organs associated with the detection of
both downwelling and bioluminescent light. The
morphology and placement of these organs have
provided some clues to these functions. A full un-
derstanding of this complex sensory system, how-
ever, must await experimental studies on living
animals.
ACKNOWLEDGMENTS
I wish to thank the Captain and the crew of the
RV Teritu and the many people that participated
in the "Teuthis" cruises. Of these, I especially
thank John Walters, Steven Amesbury, Sherwood
Maynard, and Fletcher Riggs. Most of the
614
YOUNG: VERTICAL DISTRIBUTION AND PHOTOSENSITIVE VESICLES
cephalopods taken by the IKMT came from a sam-
pling program by T. A. Clarke, University of
Hawaii. I thank J. Walters and S. Maynard for
their comments on the manuscript. This study was
supported by grants GB 20993 and GA 33659 from
the National Science Foundation. Hawaii Insti-
tute of Geophysics Contribution No. 916.
LITERATURE CITED
Amesbury, S. S.
1975. The vertical structure of the midwater fish commun-
ity off leeward Oahu, Hawaii. Ph.D. Thesis, Univ.
Hawaii, Hoholulu, 106 p.
Berry, S. S.
1914. The Cephalopoda of the Hawaiian Islands. Bull.
U.S. Bur. Fish. 32:255-362.
BODEN, B. P., AND E. M. KAMPA.
1967. The influence of natural light on the vertical migra-
tions on an animal community in the sea. Symp. Zool.
Soc. Lond. 19:15-26.
Clarke, M. R.
1966. A review of the systematics and ecology of oceanic
squids. Adv. Mar. Biol. 4:91-300.
1969. Cephalopoda collected on the SOND cruise. J. Mar.
Biol. Assoc. U.K. 49:961-976.
Clarke, M. R., and C. C. Lu.
1974. Vertical distribution of cephalopods at 30°N 23°W in
the North Atlantic. J. Mar. Biol. Assoc. U.K. 54:969-
984.
1975. Vertical distribution of cephalopods at 18°N 25°W in
the North Atlantic. J. Mar. Biol. Assoc. U.K. 55:165-
182.
Clarke, T. A.
1973. Some aspects of the ecology of lantemfishes (Myc-
tophidae) in the Pacific Ocean near Hawaii. Fish. Bull.,
U.S. 71:401-434.
GIBBS, R. H., Jr., AND C. F. E. ROPER.
1971. Ocean Acre: Preliminary report on vertical distribu-
tion of fishes and cephalopods. In G. B. Farquhar
(editor). Proceedings of an International Symposium on
Biological Sound Scattering in the Ocean, p. 1 19-133. Dep.
Navy, Washington, D.C., Maury Cent. Ocean Sci. Rep.
005.
LU, C. C, AND M. R. Clarke.
1975a. Vertical distribution of cephalopods at 40°N, 53°N
and 60°N at 20°W in the North Atlantic. J. Mar. Biol.
Assoc. U.K. 55:143-163.
1975b. Vertical distribution of cephalopods at 11°N, 20°W
in the North Atlantic. J. Mar. Biol. Assoc. U.K. 55:369-
389.
Mauro, a.
1977. Extra-ocular photoreceptors in cephalo-
pods. Symp. Zool. Soc. Lond. 38:287-308.
MAURO, A., AND F. BAUMANN.
1968. Electrophysiological evidence of photoreceptors in
the epistellar body o^Eledone moschata. Nature (Lond.)
220:1332-1334.
Messenger, J. B.
1967a. Parolfactory vesicles as photoreceptors in a deep-
sea squid. Nature (Lond.) 213:836-838.
1967b. The peduncle lobe: a visuo-motor centre in Oc-
topus. Proc. R. Soc. Lond., Ser. B, Biol. Sci. 167:225-251.
NISHIOKA, R. S., I. YASAMASU, A. PACKARD, H. A. BERN, AND
J. Z. YOUNG.
1966. Nature of vesicles associated with the nervous sys-
tem of cephalopods. Z. Zellforsch. Mikrosk. Anat.
75:301-316.
PACKARD, A.
1972. Cephalopods and fish: the limits of con-
vergence. Biol. Rev. Camb. Philos. Soc. 47:241-307.
PEARCY, W. G.
1965. Species composition and distribution of pelagic
cephalopods from the Pacific Ocean off Oregon. Pac. Sci.
19:261-266.
ROPER, C. F. E.
1969. Systematics and zoogeography of the world wide
bathypelagic squid Bathyteuthis (Cephalopoda: Oegopsi-
da). U.S. Natl. Mus. Bull. 291, 210 p.
ROPER, C. F. E., AND R. E. YOUNG.
1975. Vertical distribution of pelagic cephalo-
pods. Smithson. Contrib. Zool. 209, 51 p.
Walters, J. F.
1976. Ecology of Hawaiian sergestid shrimps (Penaeidea:
Sergestidae). Fish. Bull., U.S. 74:799-836.
Young, J. Z.
1977. Brain, behaviour and evolution of cephalopods.
Symp. Zool. Soc. Lond. 38:377-434.
Young, R. E.
1972a. Function of extra-ocular photoreceptors in
bathypelagic cephalopods. Deep-Sea Res. 19:651-660.
1972b. Brooding in a bathypelagic octopus. Pac. Sci.
26:400-404.
1973. Information feedback from photophores and ventral
countershading in mid-water squid. Pac. Sci. 27:1-7.
1975a. Leachia pacifica (Cephalopoda, Teuthoidea):
Spawning habitat and function of the brachial photo-
phores. Pac. Sci. 29:19-25.
1975b. A brief review of the biology of the oceanic squid,
Symplectoteuthis oualaniensis (Lesson). Comp.
Biochem. Physiol. 52B:141-143.
1975c. Function of the dimorphic eyes in the midwater
squid Histwteuthis dofleini. Pac. Sci. 29:211-218.
1975d. Transitory eye shapes and the vertical distribution
of two midwater squids. Pac. Sci. 29:243-255.
1977. Ventral bioluminescent countershading in midwa-
ter cephalopods. Symp. Zool. Soc. Lond. 38:161-190.
Young, R. E., and C. F. E. Roper.
1977. Intensity regulation of bioluminescence during
countershading in living midwater animals. Fish. Bull.,
U.S. 75:239-252.
Note added in proof: The correct name for the species listed here as Phasmatopsis
fisheri is Megalocranchia fisheri (N. Voss. In press. Studies on the cephalopod family
Cranchiidae. A revision of the genera, with a key for their determination. Bull. Mar.
Sci.)
615
SYSTEMATIC SAMPLING IN A PLANKTONIC ECOSYSTEM
E. L. Venrick'
ABSTRACT
Two sampling studies, computer simulation and field, investigated the consequences of applying
restricted systematic sampling (at predetermined depths) to estimate total chlorophyll in the water
column. Comparison was made with stratified random designs with one and two samples per strata.
Systematic sampling appeared more accurate than most stratified random designs. However, when
repeated over restricted spatial or temporal intervals, systematic designs tended to produce biased
estimates. In the central Pacific, an interval of several days, or 100-200 km, appeared necessary for
natural population fluctuations to average out the bias inherent in a restricted systematic sampling
design.
Underlying sampling theory is the assumption of
random collection of samples. This is the only
satisfactory method of assuring a representative
sample from an unknown population. In pelagic
ecology (and undoubtedly in other fields) this as-
sumption is generally neglected and surveys are
conducted at fixed geographic positions, at fixed
spatial or temporal intervals, and/or at fixed
depths, without recourse to randomization. The
implicit assumption is that the natural complex
variability of pelagic populations provides the
necessary element of randomization.
Two types of sampling strategies are frequently
called systematic. The present study is concerned
with the situation in which the sampling positions
are fixed according to some pattern determined by
the investigator and are not necessarily at equal
intervals; this will be termed restricted systematic
sampling (RSS) to distinguish it from the strategy
in which only the sampling interval is fixed and
the location of the first sample in the first interval
is determined at random (randomly located sys-
tematic sampling; Yates 1948). Among the alter-
nate sampling strategies which provide the requi-
site randomization, unrestricted random and
stratified random sampling ( SR ) have received the
most attention. In unrestricted random sampling,
samples are selected individually from the entire
population by some random process, such as by
numbering all sampling units and selecting from
them by means of a random numbers table. In SR,
the population is first divided into subpopulations
from each of which one or more samples are
'Scripps Institution of Oceanography, University of Califor-
nia, San Diego, La JoUa, CA 92093.
selected at random. SR is useful because it ensures
that the samples are distributed throughout the
entire population.
Three characteristics of sampling designs are of
interest (Figure 1): 1) bias, any consistent devia-
tion between the true population parameter and
repeated estimates based on the same sampling
design; 2) precision, the variability of successive
estimates about their mean when a sampling de-
sign is repeated on the same population; and 3)
PRECISION
BIAS
>-
<_)
-z.
UJ
o
3IASED BUT PRECISE
MORE ACCURATE
UNBIASED BUT IMPRECISE
LESS ACCURATE
Figure l. — Normal frequency distributions used to illustrate: a)
precision, the spread of observations about their mean value (x);
b) bias, the deviation of the mean of repeated observations from
the true parameter ( d)\ c) a distribution which is biased but
precise; and d) a distribution which is unbiased but imprecise.
Distribution c will be more accurate than distribution d, in spite
of the bias, if the average deviation of observations from B is
smaller.
Manuscript accepted Februar\' 1978.
FISHERY BULLETIN: VOL '76. NO. 3. 1978.
617 —
FISHERY BULLETIN: VOL. 76, NO. 3
accuracy, a concept including both freedom from
bias and high precision and which, in the absence
of bias, is equivalent to precision. The practical
determination of precision, in its strictest sense, is
restricted to quasi-static populations in which the
population remains unchanged between collec-
tions of replicate samples (forests or soil types or
mussel beds, etc.). In the case of RSS, the concept of
precision has no meaning in this type of popula-
tion because successive application of the same
sampling design to the same population will give
identical results. Such static populations do not
exist in a planktonic system because spatial and
temporal variability produce continual change.
Thus, in the present study, the concept of a popula-
tion is expanded to incorporate spatial and tem-
poral fluctuations in which case the precision of
RSS has a real value.
Theoretical aspects of systematic random sam-
pling strategies have been considered by many
(e.g., Yates 1946, 1953; Doming 1950; Cochran
1963; Sukhatme and Sukhatme 1970). Empirical
investigations have been restricted to terrestrial
systems, particularly to surveys of vegetation
types or timber volumes ( e.g., Hasel 1938; Osborne
1942; Finney 1948b, 1950; Numata and Nobuhara
1952; Bourdeau 1953; Milne 1959). The results
from these studies indicate that randomly located
systematic sampling often gives more accurate es-
timates than other procedures (Hasel 1938; Os-
borne 1942; Madow 1946; Yates 1946, 1948; Fin-
ney 1948a; Bordeau 1953; Milne 1959; Grieg-
Smith 1964) especially when the sampled popula-
tion has positive correlation between neighboring
units (Cochran 1946; Milne 1959; Sukhatme and
Sukhatme 1970). Because of the greater precision
and gi-eater convenience of systematic sampling,
some workers have recommended its use for ter-
restrial surveys (Hasel 1938; Yates 1946; Milne
1959). On the other hand, it has been shown that
irregular distributions or pronounced patterns of
variation, especially periodicity or linear trends,
may cause systematic designs to give biased esti-
mates or estimates of reduced precision (Madow
and Madow 1944; Finney 1950; Bourdeau 1953;
Sukhatme and Sukhatme 1970); nor does the pre-
cision necessarily improve with increasing sample
size (Madow 1946; Bordeau 1953).
Of the random designs, SR generally offers
greater precision than unrestricted random sam-
pling (Yates 1953; Milne 1959) and, with a con-
stant number of samples, this precision increases
as the number of strata increases (Yates 1953).
The most precise design is one with one sample per
strata, but this (like a systematic sample) offers no
internal estimate of error (Finney 1948a, b).
The success of systematic sampling clearly de-
pends upon the nature of the sampled population.
If individuals or properties in a population are
distributed at random, all strategies will be equi-
valent. Pronounced pattern, however, may in-
crease or decrease the effectiveness of systematic
designs. Thus, quite aside from the theoretical
objections to systematic sampling, uninformed
application of any systematic sampling is to be
discouraged.
Although Strickland (1968) warned that dis-
crete samples may give a poor representation of
the vertical distributions of highly stratified sub-
stances, such as chlorophyll, a thorough study of
the consequences of systematic sampling in the
ocean has not been conducted, even though most
populations have marked gradients, especially
along the vertical axis. This may be attributed to
the logistical difficulties of enumerating an
oceanic population in its entirety, in contrast to a
timber stand in which every individual may be
observed, counted, measured, and mapped.
The present study is restricted to the conse-
quences of applying RSS in the vertical direction.
The distribution investigated is that of
chlorophyll in an oligotrophic oceanic environ-
ment. Total chlorophyll in the water column is a
frequently used index of plant crop and it is most
often estimated from a series of restricted sys-
tematic samples. The major question is whether
such sampling produces any bias in the estimate of
total chlorophyll, or whether the temporal and
spatial heterogeneity of the chlorophyll distribu-
tion is sufficient to average out the biases of indi-
vidual determinations. Of secondary concern is
whether there is a significant difference in preci-
sion or accuracy between estimates derived from
RSS and those derived from SR.
The area of study is the North Pacific Central
Gyre in the vicinity of lat. 28°N, long. 155°W. The
region is one of relatively low spatial and temporal
variability (Venrick et al 1973; Gregg et al. 1973;
McGowan and Williams 1973; Eppley et al. 1973;
Haury 1976). Thus, it is an environment in which
any adverse characteristics of RSS are expected to
be magnified. The general features of the distribu-
tion of chlorophyll in the North Pacific Central
Gyre have been summarized (Venrick et al. 1973).
Most of the year, surface concentrations are low
(0.02-0.06 mg/m^), and there is a narrow subsur-
618
VENRICK: SYSTEMATIC SAMPLINC IN KCOSYSTKM
face maximum layer (0.10-0.20 mgm-M centered
between 90 and 120 m.
The present study was conducted in two parts.
In part A, a computer was used to sample nine
semiartificial populations derived from continu-
ous vertical profiles of chlorophyll fluorescence.
Changes in the fluorescence per extractable
chlorophyll unit with depth iKiefer 1973) and
smoothing of small-scale features during the
pumping procedure result in a profile which repre-
sents only the grosser features of the true distribu-
tion. From the vertical profiles, the total popula-
tion along the vertical axis was calculated,
allowing the accuracy of various sampling strate-
gies to be determined directly. Study B was con-
ducted in the field where restricted systematic and
stratified random samples were collected simul-
taneously from the population. In this study, a real
population was studied but the total population
could only be approximated.
METHODS
Analytical Procedures
offset the increase in fluorescence per unit of ex-
tractable chlorophyll with depth, one conversion
equation was used down to and including the
chlorophyll maximum and another below the
maximum. The conversion factors were deter-
mined by analysis of chlorophyll extracted from
discrete water samples collected periodically dur-
ing the cruise. The surface value of each continu-
ous profile was set to 0.03 mg/m-^ and the minimum
value below the maximum to 0.01 mg/m-^ these
were the mean values of extracted chlorophyll ob-
served at the surface and at 200 m, respectively.
The horizontal scale was adjusted to bring the
mean maximum value of all profiles to 0.156
mg/m'\ the average maximum of the discrete sam-
ples. A typical adjusted profile is presented in Fig-
ure 2.
These semi-artificial populations were sampled
with four stratified random designs ( Table 1 ). The
success of SR depends upon the extent to which the
strata can be made internally homogeneous. In an
attempt to achieve this, the stratum boundaries of
SR-1 and SR-2 were determined as much as possi-
ble by the hydrographic, biological, and chemical
Chlorophyll a was determined fluorometrically
according to the procedure of Yentsch and Menzel
(1963) as modified by Holm-Hansen et al. (1965).
Water for discrete, extracted chlorophyll samples
was obtained with Nansen bottles. Water for con-
tinuous vertical profiles was obtained with the
seawater pumping system described by Beers et
al. (1967) and was passed through a fluorometer
equipped with a flowthrough door.
Stud) A
The chlorophyll fluorescence profiles were taken
during September 1968, on 9 consecutive days
during which time the ship followed two drogues
which were set at 10 m depth to follow the mixed
layer. These were launched at lat. 27°00'N, long.
155°18'W and moved in a northwesterly direction
at speeds between 0.5 and 1.5 kn covering 345 km
in 9 days. The profiles were not made at the same
time of day. The closest two profiles were sepa-
rated by 13 h, the most distant by 40 h. Additional
aspects of these profiles and accessory data have
been published (Scripps Institution of Oceanog-
raphy 1974).
The fluorescence profiles were read at 1-m in-
tervals and translated into units of approximate
chlorophyll down to a depth of 180 m. In order to
50
a^ioo
Q
15
TEMPERATURE (°C)
20 25
1 1 1 \ I I I I 1 I 1 1 1 T"
CHLOROPHYLL (mg/m3)
05 iO 15 .20 .25 .30
"TT
- A
- A
- A
A
A
A
A
I M I I [ I M I I [ I I I
CHLOROPHYLL
M M M M I M I
TEMPERATURE "
150
L- hAh ■
Figure 2.— a typical population of chlorophyll values derived
from a continuous profile of fluorescence (27 September 1968)
and sampled in study A, together with the temperature values
from the associated hydrocast. Triangles indicate the location of
samples in restricted systematic design 3; bars represent the
boundaries of strata used in stratified random design 1.
619
FISHERY BULLETIN: VOL. 76, NO. 3
Table l. — Systematic and stratified random sampling designs used in studies A and B.
Systematic:
RSS-1
RSS-2
RSS-3
RSS-4
Stratified random:
SR-1
SR-2
SR-3
SR-4
0
0
0
0
0
0
0
0
20
10
20
45
35
75
15
45
40
25
40
65
55
95
45
90
Sample
60
35
60
80
Stratum
75
105
75
110
depths
80
50
80
90
boundaries
85
125
90
130
(m)
100
60
90
100
(m)
95
180
100
180
120
75
100
110
105
110
140
100
110
120
115
120
160
125
130
137
125
130
180
180
180
180
150
180
150
180
characteristics of the environment (Figure 2), and
larger strata were assigned to the layers in which
environmental gradients were small and several
narrow strata were placed in the region of the
chlorophyll maximum. The 35-m boundary
marked the average depth of the mixed layer; 95 m
was the approximate depth of penetration of 19( of
the surface radiation, and 125 m represented the
beginning of the nutricline. Design SR-1 consisted
of 10 strata, each with one sample; in design SR-2,
adjacent strata were lumped giving five strata
with two samples in each. Designs SR-3 and SR-4
were those used in study B (below) and were thus
based on environmental characteristics observed
at that time. Each of the nine populations was
sampled 20 times with each of the stratified ran-
dom designs. To facilitate comparison with the
systematic samples, for which there was only one
cast of each design per profile, it was desirable to
examine a series of unreplicated stratified random
samples. For this purpose, 10 subsets were
selected at random from the replicate casts, each
subset containing nine stratified random casts,
one from each population. Total chlorophyll was
calculated from the mean (arithmetic) concentra-
tion per strata times the width of that strata,
summed over all strata. This is the classical proce-
dure for summarizing data collected by SR.
Four RSS designs were employed: RSS-1, one
sample at the surface and every 20 m thereafter;
RSS-2, the design actually employed in September
1968 in which the cast was partially determined
by standard hydrographic depths; RSS-3, a design
which was based upon complete knowledge of the
vertical distributions which were being sampled
and which was derived from application of the
general rules of sample allocation, i.e., samples
were concentrated in the region of maximum var-
iability (the chlorophyll maximum layer); RSS-4,
a design based on stratified random design 1 (and
therefore more strictly comparable to it) with a
sample at the top of the upper stratum (0 m) and
bottom of the lowest stratum (180 m) and at the
center of all intermediate strata.
Two methods of calculating total chlorophyll
from systematic samples were investigated. In the
first, the layer between adjacent samples was rep-
resented by the arithmetic mean of the two sam-
ples (equivalent to integration with linear inter-
polation). In the second, the layer was represented
by the geometric mean of adjacent samples. This
latter procedure is sometimes recommended when
the population exhibits large, nonlinear changes
between adjacent samples. A comparison of the
two procedures was made in study A, on the basis
of which the method using geometric means was
rejected.
Study B
Study B, conducted in June 1977, combined two
10-sample designs, one restricted systematic
(RSS-1) and one stratified random (SR-3 or SR-4)
into a single 20-bottle cast. The strata boundaries
were primarily determined from two preliminary
18-bottle casts which defined the regions of
chlorophyll gradients and from a single STD trace
which defined hydrographic strata (Figure 3). As
in study A, narrower strata were established at
the depths of maximum gradients of chlorophyll
(the region of the maximum layer). The major
differences between designs SR-1 and SR-2 and
designs SR-3 and SR-4 were due to a shallower
mixed layer and broader, deeper maximum layer
observed in June 1977.
Over a period of 21 days, a total of 18 casts were
made, 9 employing RSS-1 and SR-3 and 9 employ-
ing RSS-1 and SR-4. All casts were located within
a rectangle bounded by lat. 28°21.6'N and
28°45.9'N, and by long. 155°14.0'W and
155°33.5'W. Fourteen casts were taken in con-
junction with another program between the hours
of 2200 and 0300 (with one exception, delayed by
winch failure until 0550). Twelve casts were
620
VENRICK: SYSTEMATIC SAMPI.IN'd IN ECOSYSTEM
TEMPERATURE (°C)
15 20 25
I — I — I — I — I — I — I — \ — \ — \ — \ — \ — I
CHLOROPHYLL (mg/m3)
■05 .10 .15 .20 .25
I I I I I I I I I M I I I I I M I I I I I I I
value and may be measured as a percent of the
true value ( 0):
50
CL 100
UJ
Q
,- kAh
- A
- A
- A
- A
hAH
hAH
- A
150
hAH
Figure 3. — Chlorophyll values observed in two 18-bottle casts
preliminary to study B, together with the temperature trace
from the associated STD lowering. Triangles indicate location of
samples in restricted systematic design 1; bars represent the
boundaries of strata used in stratified random design 3.
paired, taken within a few hours and within 3
n.mi. of each other. These have been considered
replicate casts.
When the combined systematic-stratified ran-
dom design called for bottles to be spaced more
closely than 3 m, it was necessary to use a mes-
senger heavier than the standard Nansen mes-
senger (such as a Niskin bottle messenger) in
order that it develop enough momentum to trip the
second bottle. When both sampling designs called
for the same depth, the extra bottle was arbitrarily
positioned, usually filling in the largest gap in the
region of the chlorophyll maximum layer. This
"free" sample was used only in the calculations of
total chlorophyll in the water column.
Statistical Procedures
Bias is evaluated by the consistency with which
n observations (.v, , / = 1, n) from a given sampling
design fall above or below the true population
2 iXi-9)
n
X 100%
le.
Precision is measured by the variance of a series of
n observations about their mean (x):
^j ^X j' X)
In an analogous way, accuracy is measured by the
mean square deviation of a series of observations
from the true population total:
2 {x,-Qy
n
Both accuracy and precision are inversely related
to their statistical measures, increasing as the
numerical value of the measure decreases. Since
most scientists are used to thinking in terms of
variances and sums of squares, it did not seem
desirable to invert these measures to achieve di-
rect correspondence.
In the analysis of the results, limited use was
made of the parametric analysis of variance. Most
statistical tests were nonparametric tests which
make few assumptions about the characteristics of
the data (e.g., Dixon and Massey 1957; Tate and
Clelland 1957; Conover 1971; Hollander and
Wolfe 1973). Unless stated otherwise, the prob-
abilities associated with conclusions in the text
are derived from the binomial distribution withp
= y-i.
In several analyses in these studies, the problem
of multiple testing arose, as when all four sys-
tematic designs were tested for bias. Unfortu-
nately, the tabulation on most nonparametric
procedures is not sufficiently complete to allow
correction for multiple testing to be made without
making the tests extremely conservative. Since
this was deemed undesirable, the probabilities
given for the statistical tests are uncorrected. It is
unlikely that this makes any real difference in the
outcome of these studies which gain most of their
force from the similarity of results in the two ap-
proaches.
621
FISHERY BULLETIN: VOL 76, NO. 3
RESULTS
Study A
Integration of values
The results of study A are summarized in Table
2. The total chlorophyll values derived from the
four systematic sample designs were calculated by
integration with linear interpolation (i.e., using
the arithmetic mean of adjacent samples to repre-
sent the average chlorophyll in the stratum be-
tween them). Use of the geometric mean ir. this
calculation resulted in the true total being under-
estimated 27 out of 36 times (P = 0.01). Nor was
there any increase in accuracy (the resultant ac-
curacies, based on use of the geometric mean, were
0.538, 0.987, 0.488, and 0.752 for RSS-1 through
RSS-4). The use of the geometric mean in the cal-
culation of total chlorophyll does not appear to be
justified.
Bias
The biases observed in the eight sampling
strategies are summarized in Table 3. Of the four
restricted systematic designs, only RSS-2 gave no
signs of bias. Design RSS-3, the "best informed"
design, overestimated the true population total in
eight of the nine trials (P<0.05). RSS-1 overesti-
Table 3. — Bias of systematic and stratified random sampling
designs, Study A.
Date
Systematic designs
Stratified random designs
(1968)
RSS-1
RSS-2
RSS-3
RSS-4
SR-1
SR-2
SR-3
SR-4
19 Sept.
-
-
+
-
-1-
_ _
-
+
20 Sept.
-
-
+
-»-
-
- -t-
'0
-
21 Sept
-
-
+
-
+
- -^
'0
+
22 Sept.
+
+
+
+
-
- +
'0
+
23 Sept.
-
-
-1-
-
-
-
-
24 Sept.
+
-
-1-
+
-1-
-1-
-1-
25 Sept.
+
+
-1-
'0
-
-
+
26 Sept.
+
-
-
-
-1-
h
-
+
27 Sept.
+
+
-1-
+
-1-
— -I
-1-
-
' Estimate =
= true value.
mated the population only fivir- out of nine times,
but the overestimates were clustered toward the
end of the series and the underestimates toward
the beginning. This temrjral trend lies just out-
side the usual level of significance (run test;
P<0.10) but it indicates that the time period
necessary for the population to provide "random"
variability of sufr.cient magnitude to eliminate
bias may be of tht- order of several days or 100-200
km. The magnitudes of the biases were —4.09^ for
the period 19-21 September and +3.77c for the
period 24-27 September. Similarly, the bias intro-
duced by using RSS-3 to estimate the true popula-
tion total for 19-25 September was -3.69c.
The peculiar periodicity of bias seen in RSS-4
also indicates a nonrandom interaction between
the sampling design and the sampled population
Table 2. — Restilts of study A, a computerized simulation sampling study. The estimated parameter ( 6) is total chlorophyll above 180
m; units are milligrams per square meter; time is local time.
Date
Time
True value
{0)
Systematic designs
One cast each
Means and
Stratified random designs
variances (in parentheses) of 20
replicates
(1968)
RSS-1
RSS-2
RSS-3
RSS-4
SR-1
SR-2
SR-3
SR-4
19 Sept.
1719
8.50
8.00
8.00
8.62
8.39
8.52
8.38
8.39
8.59
20 Sept.
2312
10.31
9.91
9.98
11.12
10.71
(0.345)
10.29
(2.460)
10.18
(0.642)
10.31
(0.050)
10.24
21 Sept
2335
7.35
7.21
6.69
7.57
7.33
(0.244)
7.36
(0.987)
7.16
(0.182)
7.35
(0.513)
7.36
22 Sept.
2351
6.89
7.23
7.72
7 45
718
(0.059)
6.85
(0.950)
6.88
(0.215)
6.89
(0.466)
6.92
23 Sept.
2025
8.83
7.97
7.51
9.03
8.47
(0.072)
8.62
(0.151)
8.68
(c.oei)i
F,.8.2
(0.129)
8.78
24 Sept.
0900
9.68
9.70
9.20
9.94
10.56
(0.801)
9.81
(1.780)
9.57
(0.341)
9.85
(1.227)
9.79
25 Sept.
0800
11.00
11 84
12.08
11.08
11.00
(0.178)
10.86
(1.490)
10.90
(0.406)
10.92
(1.841)
11.16
26 Sept
0830
13.85
14.23
13.14
13.36
13.06
(0.263)
13.92
(1.436)
13.23
(0.443)
1365
(0.687)
13.90
27 Sept.
2400
13.90
14.47
14.56
14.94
14.78
(1.123)
14.17
(0490)
(5.323)
13.60
(3,676)
(2.289)
14.06
(1.409)
(3.954)
13.83
(3.348)
Accuracy
n - 1
0.308
0.695
0.308
0.320
'0.574
'1.464
'0.779
'1.662
Precision
6.441
8.113
7.729
6.537
6.725
'8.004
'6.312
'7.036
'8.540
n - 1
'Mean values from 10 sets of unreplicated casts.
622
VENRICK: SYSTEMATIC SAMPLING IN ECOSYSTEM
(run test, P = 0.10). With the sampling interval
employed here, the biases of individual estimates
average out over the entire study. Had the inter-
val been twice as large, a consistent overestimate
or underestimate would have resulted, with re-
spective magnitudes of -i-5.8'7f and -1.9*^, until
25 September when the phase relationship ap-
pears to have shifted.
Tables 2 and 3 also present the results of the four
stratified random designs, based upon the means
of 20 replicates. The consistent underestimates
resulting from SR-2 were sufficiently unexpected
that a second series of 20 SR-2 samples were
drawn from each population. This series showed
no evidence of bias and, thus, it appears that the
initial results were the product of random chance.
Precision
Precision, in its strictest sense, could only be
examined in the case of the stratified random de-
signs, for which replicates were available. The
designs employing 10 strata, each with one sam-
ple, SR-1 and SR-3, offered greater precision than
designs with fewer strata. However, there was a
highly significant concordance (Kendell
coefficient, P<0.01i between the precisions of all
designs with respect to the profiles giving the most
precise result. Examination of the individual
profiles indicated that the precision of the results
was inversely related to the strength of the
chlorophyll maximum and to the amount of
small-scale variability along the vertical axis, or,
in other words, to the structural complexity of the
population. Later, the accuracy of the systematic
designs (discussed below) was found to show the
same relationship.
For all stratified random designs, the variance
between replicates was trivial compared with the
variance between the nine populations. Analyses
of variance gave /"s ^g ratios ranging from 54 to 344
(all P«0.01). When all nine profiles were consi-
dered to be replicates of the same population, the
variance between the nine estimates from each
systematic cast could be compared with the var-
iance between single stratified random casts, one
from each population (Figure 4A). On this scale,
there were no differences in precision between any
of the sampling designs. The large variation be-
tween populations masked any difference in per-
formance. Thus, when the concept of the sampled
population is expanded to include spatial and
temporal variations, RSS appears to offer neither
o
CL
cr
60
V
I-
<
Z5
O
o
<
40
20
30
140
120
100
80- X
A Precision:
I(xrx)
= \2
_ TRUE S^ BETWEEN
POPULATIONS
>
<
cr
Z)
o
o
<
20
V
Z(x,-0)^
B Accuracy: — —, —
4^
!V
r>
^
fv
^
^
^
Figure 4. — The results of the computer simulation sampling
study, study A, showing the relative precisions and accuracies of
the four restricted systematic sampling designs (RSS) and four
stratified random designs (SR).
advantages nor disadvantages with respect to pre-
cision of estimates.
Accuracy
The accuracy of the various designs was also
compared using sets of unreplicated stratified
random casts (Figure 4B). The greater accuracies
of stratified random designs SR-1 and SR-3 rela-
tive to SR-2 and SR-4 undoubtedly reflected their
greater precision; and perhaps the greater accu-
racy of SR- 1 relative to SR-3 was due to selection of
more appropriate strata. The systematic designs
were generally more accurate than the stratified
random designs. Only stratified random design
SR-1 achieved the accuracy of the systematic de-
signs.
623
FISHERY BULLETIN: VOL 76. NO. 3
Most of the chlorophyll work in the central
Pacific has been based upon 12 or more sampled
depths. Thus, it was encouraging to find that as
few as 10 depths, regardless of the sampling
strategy, gave a generally satisfactory picture of
the amount of chlorophyll in the water column. Of
nearly 400 estimates from individual casts, IGOi
fell within ±109? of the true value. This percent
increased to 859? for stratified designs SR-1 and
SR-3 and to 949r for the 36 systematic casts. How-
ever, to the extent that these fluorescence profiles
underestimate the structural complexity of the
true chlorophyll distribution, these results proba-
bly overestimate the accuracies of the designs.
Study B
The results of the field study were remarkably
similar to those of the computer study (Table 4).
Bias and accuracy were investigated by assuming
that the entire population was exactly represented
by the 20 samples in one cast (systematic samples
plus stratified random samples plus "free" sam-
ples). The results of study A indicate that the dis-
crepancy is not likely to be severe.
Table 4. — Results of study B, a field sampling study. The esti-
mated parameter is total chlorophyll above 180 m and the true
value ( 6) is estimated from the 20 combined samples of the two
designs; units are milligrams per cubic meter.
Date
Local
time
H
X, bias
(1977)
RSS-1
SR-3
SR-4
5 June"
6 June_
2345
1887
18.73-
18.25-
0220
1768
17.10-
18.69 +
8 June
0033
15.54
14.06-
18.21 +
9 June
0236
18 11
17.23-
17.55-
9 June"
2241
16.70
16.22-
17.19 +
10 June_
0550
15.47
14.75-
16.68 +
13 June"
2208
13.26
1329^
12.31-
14 June^
0035
13.42
13.47 +
13.33-
15 June
2203
14.53
15-9W
11.19-
19 June"
20 June.
2149
11 50
10 91-
9.03-
0100
1029
10.78 +
10.78 +
21 June
2246
14,25
14.73 +
11.42-
22 June
1107
10.67
11.00 +
10.17-
23 June"
2333
1360
13.41-
12.97-
24 June.
0133
1252
12.16-
12.09-
24 June"
1505
1683
17.11 +
18.11 +
24 June.
1632
1698
17.69 +
15.72-
26 June
0822
13.32
13.58 +
13.10-
Accuracy:
n - 1
0.45
283
2.31
Precision
(6/5-6/15)
0492
0.808
0.249
MS
(6/19-6/25)
0442
0.319
1.591
w/in pa
rs :
indicates pair of replicate casts.
Acciiracx
Bias
When the 18 casts are considered in chronologi-
cal sequence, it is evident that RSS tended to de-
viate from the true value in the same direction on
adjacent casts. The direction of bias was the same
within five of the six pairs of replicate casts
(0.0563-697. sand (2-7) >28% sand (23) >84-99%
sand (22-15) (see also Table 2). The same trend
with sediments is indicated (P<0.05) for slender
sole weight per square meter, but because catches
per square meter were more variable or mean dif-
ferences were less, differences were not significant
atP<0.01.
635
FISHERY BULLETIN: VOL 76, NO. 3
Table 4. — Results of an analysis of variance (ANOVA) of the numbers and weights of fishes caught per square meter or
per tow'; * indicates P<0.05; ** P<0.01. R^ (coefficients of determination) values are given below.
df
Alisp
ecies
Rex sole
Dover
sole
Slender sole
Pacific
No
sanddab
No.
Wt.
No.
Wt,
No.
Wt.
No
Wt.
Wt
Item
m^ Tow
m^ Tow
m^ Tow
m^ Tow
m^ Tow
m^ Tow
m^ Tow
m^ Tow
m-^ Tow
m^ Tow
Year
2
..
..
••
•
••
■•
Sediments
3
*
**
■
■
Seasons
1
*
'
Sediment ■ season
3
*
•
*
"
Depth
1
*
Depth ■ season
1
..
..
Error
95
Total
106
R^ (tow)
0 18
020
023
0.16
0 17
0.16
036
0.33
028
0.34
ft2 (m2)
039
0,51
047
0.40
0.46
043
024
0.28
036
037
'An ANOVA using a square root transformation for the data on all species combined per tow and per square meter gave similar significance effects
The ANOVA was unbalanced because of unequal numbers of observations per station, season, depth, etc Effects were tested using the extra sum of
squares principle (Searle 1971)
The biomass of Pacific sanddab. on the other
hand, showed opposite trends (P 0.05) and was
large on sandy sediments and small on silt or clay
sediments ( see also Day and Pearcy 1968; Barss et
al. see footnote 8). Since the effect of sediment was
not significant for total fish catch by numbers or
weight sandy stations with low percent organic
carbon, apparently did not support a markedly
lower abundance of demersal fishes (Table 2). Al-
though adult Dover sole show a strong preference
for mud or silt bottom (Barss et al. see footnote 8;
Demory see footnote 6), this trend was not appar-
ent for the small Dover sole caught inshore of
Heceta Bank in this study.
A sediment-season effect was indicated for slen-
der sole. They were caught in larger numbers per
tow and weight per square meter (P<0.05) at the
stations with a low percentage of sand (6-8) in the
winter than the summer.
Depth Effects
The slope of the regression between depth and
number and weight per tow of Pacific sanddab was
significant (P<0.01) and negative. Catches per
square meter on a number and weight basis gave
the same trends (P<0.05). Sanddab were most
abundant in shallow water. Weight of rex sole per
square meter and per tow and total fish numbers
and weight per tow also tended to decrease
(P<0.05) with depth.
Depth-season interactions were significant on a
square meter basis for all species combined
(number and weight) and for numbers of rex and
Dover soles. These effects were caused by appreci-
ably larger catches in deep water in winter than
summer. Seasonal differences were small in shal-
low water. This trend for lower catches on the
outer edge of the continental shelf during summer
than winter was obvious for Pacific sanddab. They
were completely absent from the deep stations (2,
6, 8) during the summer but were present at all
stations during winter. Seasonal bathymetric
migrations, with spawning migrations into deep
water in the winter and return to relatively shal-
low depths in the summer, have been described for
Dover sole and rex sole by Hagerman (1952),
Harry (1956), Alverson (1960), and Demory
(1971). Such movements could explain these
depth-season effects.
Seasons
No significant seasonal differences were de-
tected, indicating little seasonal variation in
catches of these species when all stations are com-
bined.
Year Effect
On the basis of numbers and weight per square
meter and per tow, more fishes were captured in
1969 and 1968 than in 1970 at all stations (Figure
3). This trend was significant (P<0.01) for all
species combined and for rex sole, Dover sole, and
Pacific sanddab. Year effects were also indicated
for slender sole (P<0.05). I have no cogent expla-
nation for these large annual variations. They
could represent actual variations in abundance or
availability, due to natural events or increased
fishing activity, or to undetected changes in sam-
pling efficiency. Dominant year classes have been
reported for these flatfishes off Oregon (Demory
and Robinson see footnote 9), which may contri-
bute to these annual differences, though changes
in length-frequency distributions were not obvi-
ous over this 2-yr period.
636
PEARCY niSTRIRl'TION AND ABUNDANCE OF SMALL FLATFLSHES
FiGL'RE 3. — Variations in the total num-
bers of fishes caught at each of the seven
stations, 1968-71. The two tows at each
station for each samphng period were av-
eraged.
OCT
1968
MAY AUG
1970
The amount of variability explained by the re-
gression {R'^) of all effects on catches ranged from
0.16 to 0.51 (Table 4). Values were larger for the
analysis based on catches per square meter than
catches per tow, except for slender sole. These low
values indicate that most of the variability was
not accounted for by the variables of sediment,
depth, year, and season. Large residual mean
squares indicate that sampling variability as-
sociated with catches at individual stations is ap-
preciable. Oviatt and Nixon (1973) completed a
multiple regression analysis of biomass and num-
bers of benthic fishes in Narragansett Bay, R.I.,
with 14 environmental variables. Depth and sed-
iment organic content contributed significantly to
the regression for total fish numbers and fish
biomass. But an R^ of only 0.21 was found. In both
of these studies, only a small fraction of the total
variability was explained by the environmental
factors included.
Size-Frequency Distributions
Differences in length-frequency distributions
were sometimes obvious among the stations lo-
cated at different depths or sediment types. For
example, the main length mode of rex sole at the
100- and 102-m stations was 125 mm, but at 190-
and 195-m depth there was a distinct bimodal dis-
tribution with peaks at 45 and 215 mm (Figure 4).
These differences imply that young-of-the-year
10
U 0
F,20
10
ff£X SOLE
100- 102 m
n=t004
— 1 1 r 1 1 —
50 100 150 200 250
STANDARD LENGTH (mm)
300
FIGURE 4.— Length-frequency data for rex sole at 100-102 m
stations (above) and 190-195 m stations (below).
637
FISHERY BULLETIN VOL 76, NO 3
(<50 mm) and age-groups III- V (200-250 mm) rex
sole (see Hosie and Horton 1977 for age-length
data) preferentially inhabit deep waters on the
outer edge of the continental shelf while inter-
mediate sizes (75-150 mm) inhabit shallower wat-
ers of the inner shelf. The peak of young-of-the-
year rex sole at 200 m corroborates the conclusion
of Pearcy et al. (1977) about the depth of larval
settlement and the nursery ground for early
benthic life. They concluded that rex sole larvae
settle to the bottom mainly on the outer continen-
tal shelf during the winter when they are >50 mm
SL. Powles and Kohler (1970) and Markle (1975)
believed that the nursery grounds ofGlyptocepha-
lus cynoglossus are also in deep waters off the east
coast of the United States.
Small Pacific sanddab (<70 mm) composed a
larger proportion of the catch at 102 m where sand
was 28'7f of the sediment (Station 23) than at 74
and 102 m where sand made up over 647f of the
sediment (Stations 22, 15) (Figure 5). Young
sanddab appear to inhabit deeper water with finer
sediments in early life and then aggregate on
sandy bottom areas in shallow water where they
often dominate the demersal fish fauna. Hence,
this trend of decreasing depth with increasing age
is similar to that found for rex sole.
SUMMARY
1. Demersal fishes were sampled at seven sta-
tions on Oregon's central continental shelf at vari-
15
PACIFIC SANDDAB
Stations 15822
84-99% Sand
10
^
/^^ 74 -102m
1 \ /7 - 138/
—
5
- /
-
1-
z
UJ o
^
r^ I
' 1 1 ^
01
Q.
100 150 200
STANDARD LENGTH (mm)
250
Figure 5.— Length-frequency data for Pacific sanddab at .sta-
tions with 84-99'7f sand (above) and 28% sand (below).
ous seasons of the year during a 2-yr period. A
fine-meshed, 3-m beam trawl was used in order to
quantitatively sample small flatfishes. The sta-
tions ranged from 74 to 195 m deep and had sedi-
ment types ranging from nearly lOO'/r sand to
clayey-silts with about 3''/< sand.
2. Stations were selected in an attempt to sepa-
rate the effects of depth and sediment on the as-
semblages of fishes and abundances of common
species. Three station-pairs were recognized that
had similar sediment types but were located at
different depths. Separation of sediment and
depth effects was complicated however by differ-
ences in measured (and possibly unmeasured) fac-
tors between station pairs.
3. Two general assemblages of fishes were re-
cognized on the basis of species composition of
fishes by numbers, biomass per square meter of
dominant species, and similarity indices among
the seven stations. These were a shallow water
(74-102 m) assemblage dominated numerically by
Pacific sanddab, and a deepwater ( 148-195 m) as-
semblage dominated by slender sole.
4. Species diversity iH) varied between 1.6 and
2.5 except at the shallow, sand station where it
was only 0.7. Dominance was pronounced at this
station: 86''^^ of all the individual fishes captured
were Pacific sanddab. The largest number of spe-
cies (34 or 35) was recorded for the three deep
stations. These values of H are similar to others
for temperate, demersal fish communities.
5. Similarity indices of the species composition
of fishes were high for two of the three station pairs
with similar sediments. However, indices were
also high among the four shallow stations of differ-
ing sediment types. Stations that were near each
other geographically were similar, indicating the
possibility of a proximity effect, but high similar-
ity was also found among deep stations, one of
which was over 65 km from the others.
6. An analysis of variance of the number and
weight per square meter and per tow of Dover, rex,
and slender soles. Pacific sanddab, and all species
combined indicates some effects of sediments and
depth. Largest catches of slender sole were at the
clayey-silt station pair, and largest catches of
Pacific sanddab were on sandy sediments. Catches
of Pacific sanddab were significantly larger at the
shallow stations. Catches of rex sole and all
species combined also tended to decrease with in-
creasing depth.
7. Differences in the length-frequency distribu-
tions of Pacific sanddab and rex sole were corre-
638
PEARCY: niSTRIBUTION AND ABUNDANCE OF SMALL FLATFISHES
lated with depth or sediment type. Small sanddab
predominated on the silty-sand station, whereas
large sanddab preferred sandy sediments.
Young-of-the-year rex sole were concentrated on
the outer edge of the continental shelf ( 190- 195 m).
8. Catches were sometimes larger in the winter
than the summer, especially at the deep stations.
This trend, which was noted for all four flatfishes
and for all species combined, is probably the result
of seasonal bathymetric movements.
9. A pronounced decrease in the catches of most
species and total catch per square meter occurred
during the 2 yr of this study. Reasons for this
decline are unknown.
10. The biomass of benthic fishes ranged from
0.9 to 2.4 g m "- at the seven stations. Biomass was
not appreciably lower at the pure sand stations,
which had about O.l'^^ organic carbon in the sedi-
ment. This is related to the fact that the Pacific
sanddab, the predominant species at this station,
is a pelagic feeder (see Pearcy and Hancock 1978).
11. The weight of fishes per square meter
caught in the 3-m beam trawl was several times
lower than that estimated from larger otter trawls
with coarser meshes. Although the beam trawl
caught many small flatfishes, large fishes and
nektobenthic species effectively avoided this
small beam trawl, resulting in low biomass esti-
mates.
ACKNOWLEDGMENTS
This research was sponsored by NOAA Office of
Sea Grant, No. 04-5-158-2. I am especially grate-
ful to D. L. Stein and G. L. Bertrand for their help
at sea; to D. L. Stein for identifying the fishes; to F.
L. Ramsey, W. L. Gabriel, and R. G. Petersen for
analysis of data; and to J. Dickinson, A. G. Carey,
Jr., and J. C. Quast for helpful comments on the
manuscript.
LITERATURE CITED
ALTON. M. S.
1972. Characteristics of the demersal fish fauna inhabit-
ing the outer continental shelf and slope off the northern
Oregon coast. In A. T. Pruter and D. L. Alverson (edi-
tors). The Columbia River estuary and adjacent ocean
waters, p. 583-634. Univ. Wash. Press, Seattle.
ALVERSON, D. L.
1960. A study of annual and seasonal bathymetric catch
patterns for commercially important groundfishes of the
Pacific Northwest coast of North America. Pac. Mar.
Fish. Comm.. Bull. 4, 66 p.
Ai.vERSON. 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 Lect. Fish., Inst.
Fish. Univ. B.C., Vancouver, 190 p.
Bertrand, G. A.
1971. A comparative study of the infauna of the central
Oregon continental shelf Ph.D. Thesis, Oregon State
Univ., Corvallis, 113 p.
Carey, a. G., Jr., and h. Heyamoto.
1972. Techniques and equipment for sampling benthic or-
ganisms. In A.T. Pruter and D.L. Alverson (editors). The
Columbia River estuary and adjacent ocean waters, p.
378-408. Univ. Wash. Press, Seattle.
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.
DEMORY, R, L.
1971, Depth distribution of some small flatfishes off the
northern Oregon-southern Washington coast. Res, Rep,
Fish, Comm, Oreg, 3:44-48,
Edwards, R, and J, H, Steele,
1968, The ecology of O-group plaice and common dabs at
Lock Ewe. I. Population and food. J. Exp. Mar. Biol.
Ecol. 2:215-238.
Emery, K. O.
1938. Rapid method of mechanical analysis of sands. J.
Sediment. Petrol, 8:105-111,
Gunther, F, J,
1972, Statistical foraminiferal ecology from seasonal
samples, central Oregon continental shelf Ph,D, Thesis,
Oregon State Univ,, Corvallis, 128 p,
Haedrich, R, L,, and S, O. Haedrich.
1974. A seasonal survey of the fishes in the Mystic River, a
polluted estuary in downtown Boston, Massachu-
setts. Estuarine Coastal Mar. Sci. 2:59-73.
Haedrich, R. L., G. T. Rowe, and P. T. Pollonl
1975. Zonation and faunal composition of epibenthic popu-
lations on the continental slope south of New Eng-
land. J. Mar, Res, 33:191-212,
Hagerman, F, B,
1952, The biology of the Dover sole, M/crostomwspaci/zcws
( Lockington), Calif Dep, Fish Game, Fish Bull. 85, 48 p.
Harry, G, Y„ III
1956, Analysis and history of the Oregon otter-trawl
fishery, Ph,D, Thesis. Univ, Washington, Seattle, 329 p.
HOSIE, M, J,, AND H, F, HORTON,
1977, Biology of the rex sole, Glyptocephalus zachirus, in
waters off Oregon, Fish, Bull,, U,S, 75:51-60,
INMAN, D, L,
1952, Measures for describing the size distribution of sed-
iments, J, Sediment, Petrol. 22:125-145,
KRUMBEIN, W, C, AND F. J. PETTLJOHN.
1938. Manual of sedimentary petrography. Appleton-
Century-Crofts, Inc., N.Y., 549 p.
KUIPERS, B.
1975. On the efficiency of a two-metre beam trawl for
juvenile plaice iPleuronectes platessa). Neth. J. Sea Res.
9:69-85.
KULM, L. D., R. C. ROUSH, J. C. HARLETT, R. H. NEUDECK, D.
M. Chambers, and E. J. Runge.
1975. Oregon continental shelf sedimentation: interrela-
tionships of facies distribution and sedimentarj' proces-
ses. J. Geol. 83:145-175.
639
FISHERY BULLETIN: VOL. 76, NO. 3
M.AKGALEF, R.
1968. Perspectives in ecological theor>-. Univ. Chicago
Press, Chicago, 111 p.
M.^KKLE, D. F.
1975. Young witch flounder. Glyptocephalus cynoglossus,
on the slope off Virginia. J. Fish. Res. Board Can.
32:1447-1450.
MClNTIRE. C. D., AND W. W. MOORE.
1977. Marine littoral diatoms: Ecological considera-
tions. In D. Werner (editor). Biology of diatoms, p.333-
371. Blackwell Scientific Publ. Oxf.
Merriman, d., and H. E. WARFEL.
1948. 5. Studies on the marine resources of southern New
England. VII. Analysis of a fish population. In A sym-
posium on fish populations, p. 131-164. Bull. Bingham
Oceanogr. Collect., Yale Univ. 11(4).
Oviatt, c. a., and S. W. Nixon.
1973. The demersal fish of Narragansett Bay: an analysis
of community structure, distribution and abun-
dance. Estuarine Coastal Mar. Sci. 1:361-378.
Pearcy. W. G. and D. Hancock.
1978. Feeding habits of Dover sole, Microstomus pacificus;
rex sole, Glyptocephalus zachirus; slender sole, Lyopsetta
exilis; and Pacific sanddab, Citharichthys sordidus, in a
region of diverse sediments and bathymetry off Ore-
gon. Fish. Bull., U.S. 76:641-651.
PEARCY, W. G., M. J. HOSIE, AND S. L. RICHARDSON.
1977. Distribution and duration of pelagic life of larvae of
Dover sole. Microstomus pacificus: rex sole, Glyptoce-
phalus zachirus; and petrale sole, Eopsetta Jordani, in
waters off Oregon. Fish. Bull., U.S. 75:173-183.
POWLES, P. M., AND A. C. KOHLER.
1970. Depth distributions of various stages of witch floun-
der {Glyptocephalus cynoglossus) off Nova Scotia and in
the Gulf of St. Lawrence. J. Fish. Res. Board Can.
27:2053-2062.
Richards, S. W.
1963. The demersal fish population of Long Island Sound.
I. Species composition and relative abundance in two
localities, 1956-1957. Bull. Bingham Oceanogr. Collect.,
Yale Univ. 18(2):5-31.
RILEY, J. D., AND J. CORLETT.
1966. The numbers of 0-group plaice in Port Erin Bay,
1964-66. Mar. Biol. Stn. Port Enn Annu. Rep. 78:51-56.
ROUSH, R. C.
1970. Sediment textures and internal structures: a com-
parison between central Oregon continental shelf sedi-
ments and adjacent coastal sediments. M.S. Thesis,
Oregon State Univ., Corvallis, 59 p.
Sanders, H. l., and r. r. hes.sler.
1969. Ecology of the deep-sea benthos. Science (Wash.,
D.C.) 163:1419-1424.
Searle, S. R.
1971. Linear models. John Wiley and Sons, N.Y., 532 p.
Shannon, C. E., and W. Weaver.
1963. The mathematical theory of communica-
tion. LTniv. 111. Press, Urbana, 117 p.
Thorson, G.
1957. Bottom communities (sublittoral and shallow
shelf). In J. W. Hedgpeth (editor). Treatise on marine
ecology and paleoecology. Vol. 1, p. 461-534. Geol. Soc.
Am. Mem. 67.
640
FEEDING HABITS OF DOVER SOLE, MICROSTOMUS PACIFICUS; REX SOLE,
GLYPTOCEPHALUS ZACHIRUS; SLENDER SOLE, LYOPSETTA EXILIS; AND
PACIFIC SANDDAB, CITHARICHTHYS SORDIDUS, IN A REGION OF
DIVERSE SEDIMENTS AND BATHYMETRY OFF OREGON
William G. Pearcy and Danil Hancock*
ABSTRACT
The feeding habits of the Dover sole and rex sole (mainly juveniles) and of slender sole and Pacific
sanddab were investigated at seven stations on the continental shelf off central Oregon. Dover sole had
a catholic diet, feeding on a large variety of infaunal and epifaunal invertebrates. The composition of
the diet varied among stations of different depth and sediment type indicating opportunistic feeding.
Pelecvpoda were the most important prey on a weight basis at the shallow station (74 m) of well-sorted
sand where they were the dominant macrofaunal invertebrate. Ophiuroids, sea pens, anemones, and
pelecypods were the most important prey at 100-102 m stations of silty sand or sandy silt. Polychaetes
composed over 90^^ of the diet at the deep stations ( 148-195 m) of clayey silt or silty sand. The average
standing stocks per square meter of Dover sole caught in beam trawl collections and polychaetes in
grab samples were positively correlated among stations.
Similarity of the food habits of Dover sole on the basis of food weight or frequency of occurrence was
generally higher among stations of similar depth than of similar sediment texture. Similar trends were
noted for assemblages of benthic fishes and invertebrates.
Dover sole collected during the winter had the highest percentage of empty stomachs, the fewest prey
taxa. and often the lowest frequency of occurrence of prey taxa within a size group. Because seasonal
variations were not observed in abundance of macrofaunal food in the sediments, availability of prey
may change with season, or more likely, Dover sole feed more intensely and less selectively during
summer.
Small ( <150 mm standard length) rex sole fed mainly in amphipods and other crustaceans. Large
(150-450 mm standard length) rex sole preyed chiefly on polychaetes. The diet of rex sole was less
diverse than that of the Dover sole and overlap of diet between the two species was not large.
Both the Pacific sanddab, numerically the most common species offish at the shallow sand station,
and the slender sole, the most common species at the three deep, soft-sediment stations, preyed
principally on pelagic crustaceans such as euphausiids, shrimps, and amphipods. Although the
biomass of mollusks in the sediments was large at the shallow sand station, they were not consumed by
Pacific sanddab. Fish were occasionally an important food for the sanddab.
The objectives of this study were: 1) to describe the
food habits of the four species of flatfishes that are
common in trawl catches on the central continen-
tal shelf off Oregon: Dover sole, Microstomas pac/-
ficus. rex sole, Glyptocephalus zachirus, slender
sole, Lyopsetta exilis, and Pacific sanddab,
Citharichthys sordidus; 2) to evaluate the possible
effects of depth and sediment, size of fish, and
season of capture on their food habits; and 3) to
compare the biomass and composition of fish food
from grab samples with feeding habits of fishes.
These species are among the most abundant
flatfishes in demersal communities of this region
of the Pacific Ocean ( Alverson et al. 1964; Day and
'School of Oceanography, Oregon State University, Corvallis,
OR 97331.
Pearcy 1968; Alton 1972; Demory and Hosie^).
They dominated the fish catches at the stations
where they were captured for this study ( Pearcy
1978). In order to know more about the role of
these fishes in their ecological communities, in-
cluding competitive-predatory relationships,
more data are required on their food habits.
Hagerman (1952) listed food items found in
Dover sole caught in California waters. Pearcy
and Vanderploeg ( 1973) listed general taxonomic
groups preyed upon by Dover, rex, and slender
soles and Pacific sanddab. Kravitz et al. (1977)
gave a detailed account, including species of prey
Manuscript accepted .January 1978.
FISHERY BULLETIN: VOL. 76 NO. .3. 1978.
^Demory, R. L., and M. J. Hosie. 1975. Resource surveys on the
continental shelf of Oregon. U.S. Dep. Commer., NOAA, Natl.
Mar. Fish. Serv., Commer. Fish. Res. Dev. Act, Annu. Rep. July
1, 1974 to June 30, 1975, 9 p.
641
FISHERY BULLETIN: VOL. 76, NO. 3
consumed by the rex sole and Pacific sanddab
caught in a single collection off the central Oregon
shelf. This study is, to our knowledge, the most
complete study of the food habits of these four
species.
METHODS
Fishes were collected during 115 tows with a
3-m beam trawl at seven stations on the continen-
tal shelf off central Oregon between August 1968
and August 1970. These stations are classified by
four depth categories and the percentage of sand in
the sediments in Figure 1 . Details on methods and
descriptions of the stations are given by Pearcy
(1978).
All fishes were preserved at time of capture with
Formalin,-' and the body wall of large (>150-200
mm SL) fishes was incised to insure preservation
of stomach contents. Fishes were identified and
measured (standard length, SL) in the shore
laboratory. Stomachs were removed from 326
Dover sole represented in the catches at all seven
stations; and from 614 rex sole, 1,109 slender sole,
and 723 Pacific sanddab captured at two or three
stations where each of these species was most
common.
Stomach contents were removed and empty
stomachs were noted. Food organisms were iden-
tified to species when possible. Annelids, crusta-
ceans, mollusks, echinoderms, coelenterates, and
remaining taxa (major taxa) were weighed to the
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
DEPTH (m)
100- 190-
74 102 148 195
<
CO
u
u
UJ
20
60
75
100
6
8
23
7
2
22
15
Figure l, — Classification of the seven stations, each indicated
by station number, according to depth of water and percent of
sand (0-20, 20-60, 60-75, and 75-100^^ ). The stations with similar
sediment types but different depths are 6 and 8, 7 and 2, and 22
and 15. (See Pearcy 1978 for additional information.)
nearest 0.01 g (wet-preserved weight). Usually
these weights were obtained for the contents of a
single stomach, but when the contents were in-
sufficient for accurate weighing, taxa from the
stomach contents of several fish of the same
species and size, and from the same tow, were
combined and weighed together to constitute an
observation. The number of observations for
Dover, rex, and slender sole and Pacific sanddab
were 325, 374, 607, and 392, respectively.
Results are reported as the a) percent that each
major food taxa constitutes of the total wet weight
of food found in stomachs for all seasons combined
and for winter and summer seasons separately;
and b) the frequency of occurrence (FO) of princi-
pal prey, i.e., species or taxa found in 59^ or more of
the observations for a species or size group of a
species for all seasons combined.
RESULTS
General Food Habits
Two general feeding types are indicated by dif-
ferences in the weights of major food taxa (an-
nelids, crustaceans, mollusks, echinoderms,
coelenterates, and other taxa) found in the
stomach contents of the four species (Table 1).
Dover and rex soles fed largely (64^ ) on annelids,
while slender sole and Pacific sanddab fed mainly
on crustaceans (75'7r ). Within these two apparent
feeding types, differences occurred among the
proportions of prey taxa of secondary importance.
For example, crustaceans were more abundant in
the diet of rex than Dover sole (31% vs. 11%),
whereas mollusks were more abundant in Dover
than in rex sole ( 18% vs. I9t ). Annelids composed
more of the stomach contents of slender sole than
Pacific sanddab (159^ vs. 7% ).
Based on the average frequency of occurrence of
principal prey (F0>5% ) from all sizes offish and
from all stations (Table 2), it is obvious that the
food habits within these two feeding types (Dover
sole-rex sole vs. slender sole-Pacific sanddab) are
not as similar as shown by Table 1. Principal prey
of Dover sole, for example, included 11 different
identified polychaetes. Rex sole preyed mainly on
three identified species of polychaetes. Only one
principal prey species of polychaete was common
to the diet of both Dover and rex soles. The shrimp
Pandalus Jordani, pelecypods, and ophiuroids
were principal prey in the food of Dover but not rex
sole, whereas crab larvae, cumaceans, and Oiko-
642
PKARCY and HANCOCK: FEEDING HAHITS OF FOUR FLATFISH SPECIES
Table l. — Percent by weight that major food taxa composed of
the diet of the four flatfishes, all stations and seasons combined.
Slender
Pacific
Taxa
Dover sole
Rex sole
sole
sanddab
Annelida
644
64,8
15.6
7.2
Crustacea
11,2
31 0
75.6
74,8
Mollusks
183
1,4
0.7
0
Echinoderms
34
0,1
<0.1
0
Coelenterales
26
0
0
0
Other taxa
0
2.8
8.2
18.0
T.\BLE 2. — Average frequency of occurrence of principal prey
(those occurring in S'c or more of the observations) in the four
species, all stations combined.
Rex
Dover
Slender
Pacific
Prey
sole
sole
sole
sanddab
Polychaeta:
Sternaspis fossor
12
14
Myriochele heeri
15
Nothna
geophiliformis
9
Goniada brunnea
10
Aricidea
neoseucica
6
Haploscoloplos
elongatus
g
Chaetozone setose
6
Terebellides
stroemii
6
Rhodine
bitorquala
5
Typosyllis hyalina
9
Lumbrinens sp.
6
Amphicteis sp.
5
Glyceridae
5
Unidentified
25
19
7
Crustacea:
Pandalus jordani
6
7
Euphausia pacifica
9
7
Crab larvae
or juvenile
7
13
Gammand amphipods
29 •
24
Copepods
12
Cumaceans
8
Unidentified
16
12
21
24
Mollusca
Pelecypoda
12
Ecfiinodermata
Optiiuroidea
12
Miscellaneous:
Oikopleura spp.
7
No, observations
347
325
607
392
No, fisfi
614
326
1.109
723
pleura were principal prey of rex but not Dover
sole. Gammarid amphipods occurred frequently in
stomachs of both rex and Dover soles; had they been
identified to species, the overlap between the diets
of the Dover and rex soles would appear even small-
er. Overlap in the diets of Dover and rex soles
of various size groups were also found to be small.
Gammarid amphipods, the major crustacean
prey of Dover and rex soles, were not principal
prey for slender sole and Pacific sanddab. Slender
sole and Pacific sanddab fed largely on pelagic
crustaceans, as indicated by the occurrence of
euphausiids in stomachs of both species, by crab
larvae and calanoid copepods in sanddab, and by
Pcimkdusjordani in slender sole. Slender sole and
Pacific sanddab are chiefly pelagic feeders. Pan-
dalus jordani is known to migrate off the bottom at
night (Pearcy 1970) and hence could have been
consumed on the bottom or in midwater by slender
sole. We have occasionally caught both slender
sole and Pacific sanddab in midwater trawls at
night. Barss"* caught in midwater trawls at night
Pacific sanddab that had been feeding heavily on
northern anchovy, Engraulis mordax. The only
good evidence for benthic feeding by either of these
two species is the presence of annelids in their
diets (Table 1).
Differences Among Stations
The proportions (by weight) of the major taxa in
the diet of Dover sole were sometimes markedly
different among stations (Table 3). Prey composi-
tion and availability may be functions of sediment
and/or depth. Annelids constituted over 907f of the
diet on a weight basis at the three deepest stations
(2, 6, and 8), but <13'^ at the shallowest station
(22) where the sediment was well-sorted sand. At
this shallow station, mollusks and crustaceans
were the major food items in the diet. Coelenter-
ates ( feeding polyps of sea pens and anemones) and
echinoderms (brittlestars) were minor food taxa
for Dover sole at all stations except Stations 15
and 23 ( 102 m depth), where together they com-
posed over one-half the diet. The proportion offish
with food in their stomachs was also higher at
these two stations than at any of the other sta-
tions.
To illustrate the similarities of the food habits of
Dover sole among these stations, we constructed a
station-station matrix (Table 4) using an index
( C ^) that Horn ( 1966) recommended for comparing
overlap in exploitation of alternative food sources
''Barss, W. H. (compiler). 1976. The Pacific sanddab. Oreg.
Dep. Fish Wildl., Inf Rep. 76-5, 5 p.
Table 3. — The average percentage composition of stomach con-
tents of Dover sole on a weight basis at each of the stations.
Station (depth in
meters
i in parentheses)
22
7
15
23
6
2 8
Taxa
(74)
(100)
(102)
(102)
(148)
(190) (195)
Annelids
12,6
59 9
250
305
91,9
93.4 91.0
Crustaceans
296
14 1
3,3
3,1
3 7
3.5 6.5
Mollusks
57 9
23,7
13,5
12,4
30
2.2 0.3
Ecfiinoderms
0
11
33-8
30-1
1,4
1 1 2-2
Coelenterates
0
1 2
24,3
238
0
0 0
Otfier taxa
0
0
0
0 1
0
0 0
No, fisfi examined
65
38
22
10
49
91 51
No, fisfi witti
stomacfi contents
28
19
20
10
22
61 32
643
FISHERY BULLETIN: VOL, 76, NO. 3
Table 4. — Similarity iCj^) in the diets of Dover sole at the seven
stations based on the percentage of major taxa in their diets on a
weight basis (below diagonal) and frequency of occurrence of
principal polychaete prey (above diagonal). Stations are ar-
ranged by depth.
Stations
22
Frequency of occurrence
15 23 6
22
7
15
23
6
2
8
0-04
0
0
0.09
0.06
0
058
064
0.57
0.24
0.54
005
0.34
0.56
0.59
0.21
0.38
0
0 34
0.64
0.99
0.09
0.42
0
0.22
0.88
044
052
037
042
0.21
0.87
043
0.52
1.00
0 45
0.21
088
043
0.53
1.00
1 00
within the same habitat. This measure of similar-
ity varies from 0 to 1.0. The percentages by weight
of the major taxa (lower half of Table 4) were
identical at the three deepest (148-200 m) Sta-
tions: 2, 6, and 8. Stations 15 and 23, both located
at 102 m depth, were also very similar. Station 7 at
100 m was fairly similar ( C^^ > 0.87) to Stations 2,
6, and 8 located in deeper water. The percent of
major taxa in the diet of Dover sole at the shallow,
sand location (Station 22) was not very similar to
any other station (C^^O.58).
The frequency of occurrence of principal prey of
Dover sole (Table 5) indicates fairly low similarity
among different stations for species of
polychaetes. Most species occurred at only one or
two stations and the assemblage of polychaetes
eaten by Dover sole appears to be different at each
station. As one would expect, similarity is higher
when higher taxa such as gammarid amphipods or
pelecypods are considered as a group. For this
reason comparisons of similarity among stations
should be confined to prey identified to the same
taxonomic level.
To examine differences in prey species among
stations we calculated the overlap in diet (Cj
based on polychaetes alone, the most common and
speciose prey animals of Dover sole (and the food
gi-oup that one of us (Hancock) was familiar with
taxonomically). The range in overlap of diets
based on frequency of occurrence of individual
taxa of polychaetes at these stations (Table 5) was
appreciably lower than that based on weight per-
centage of major taxa (upper half of Table 4). Sta-
tions 2, 6, and 8, which were very similar on the
basis of the weight of major taxa in the Dover sole
stomachs, overlapped only moderately on the
basis of frequency of occurrence of polychaetes ( C^^
= 0.37 - 0.45). Stations 15 and 23, similarly based
on major taxa, were the most similar iC , = 0.64)
stations based on frequency of occurrence of
polychaetes. Station 23 was the next highest in
Table 5.— Frequency of
occurrence
of principal
prey
of Dover
sole from the seven
stations.
Taxa
Stn 22
7
23
15
6
2
8
Polychaeta:
Sternaspis tossor
23
24
6
32
Myriochele heeri
31
14
6
8
27
Nothna geophiliformis
9
14
13
Chloeia pinnate
36
16
Melmna cnstata
14
Goniada brunnea
7
8
14
Ancidea uschakowi
5
9
Aricidea neosuecica
9
23
Haploscoloplos leongatus
25
18
8
Owenia fusilormis
10
18
Maldane sarsi
5
6
Chaetozone setosa
8
6
8
9
Terebellides stromenii
11
10
Tharyx multifilis
6
8
Rhodine bitorquata
13
Typosyllis hyalina
25
6
Ammonlrypane aulogasler
5
Nephtys cornuta
11
Anaitides groenlandica
10
Lumbnnens sp
12
6
Ammonlrypane spp
9
Amphitteis
9
18
Ancidea spp
5
Pherusa papillata
7
Tharyx spp.
9
Nephtys spp
5
8
8
Terebellidae
9
Ampharetidae
5
Maldanidae
9
Lumbrineridae
11
Glyceridae
10
8
Spinonidae
10
5
6
9
Hemipodus borealis
10
Laonice cirrata
6
Megelona sp.
5
Crustacea.
Pandalus lordani
30
7
16
23
8
gammarid amptiipod
70
22
22
24
14
45
26
copepod
20
cumacean
7
8
Diastylis spp.
8
Valvilera spp
10
Ostracoda
10
14
5
unidentified Crustacea
IVIollusca
Gastropoda
7
Solenogasters spp
7
8
Pelecypoda
20
20
10
16
8
Yoldia ensifera
8
Lucina sp
9
Megacrenella columbiana
10
10
Tellina salmonea
10
Acila castrensis
10
Ectiinodermata:
Ophiuroidea
11
27
16
23
sea pen
12
Miscellaneous
*"X^
Nematoda
18
No. observations
10
91
51
49
64
22
38
No. fish
10
91
51
49
65
22
38
similarity with Stations 7 and 15. Thus, the
polychaete prey of Dover sole were most similar
among these three 100-102 m stations.
Because stomachs of the other flounders were
examined for only two or three stations, few sta-
tion comparisons could be made. As with Dover
sole, the percentage of major taxa in the diets of
rex sole at Stations 2, 6, and 7 were similar, and
food habits were almost identical at Stations 2 and
644
PEARCY and HANCOCK; FEEDING HABITS OF FOUR FLATFISH SPECIES
Table 6. — The average percent by weight that major taxa com-
posed of the diet of rex sole, slender sole, and Pacific sanddab at
Stations 7, 6, and 2.
Rex sole
Slender sole
Pacific
sanddab
Taxa
Stn 7
6
2
7
6
2
7
6
Annelida
58.7
69.7
64.5
1.2
18.5
13.1
8.2
0.1
Crustaceans
392
24.8
29.6
92 2
72,3
77,8
71.3
99.9
Mollusks
2.1
10
05
09
0.8
0
0
0
Echinoderms
0
0
0
0
0
0
0
0
Coelenterates
0
0
0
0
0
0
0
0
Other taxa
0
4.5
55
58
8.5
9 1
20,5
0
No. fish
examined
376
210
28
68
844
197
690
33
No. fish with
stomach
contents
262
160
25
35
403
83
478
13
6 (Table 6). A larger percentage of crustaceans
(and the lowest percentage of annelids) was found
at Station 7 than at 2 or 6 for both rex and slender
soles. But crustaceans were more abundant in the
diet of sanddab at Station 6 than at Station 7.
Fishes (included as other taxa) were an appreci-
able part of the sanddab diet at Station 7. Again,
differences in availability of food taxa apparently
occurred among stations for the same predator
species, and different trends in the importance of
food taxa are evident for different species offish at
the same stations.
The principal prey were most similar for rex sole
at Stations 2 and 6, as were the major taxa by
weight. The polychaete Nothria geophilifonnis
and the larvacean Oikopleura occurred in over 59^
of the observations only at these two stations. It is
curious that the planktonic Oikopleura was so fre-
quent in the diet of this primarily benthophagus
fish. Other prey common at all these stations in-
cluded the polychaete Goniada brunnea, uniden-
tified polychaetes, gammarid amphipods, and
cumaceans.
PandaluH jordani was a principal prey species
for slender sole at Stations 6 and 7 but not at
Station 2. The shrimp Sp/ro/Jtoca/v's hiapinosa and
unidentified fish were found in over 5r/c of the fish
only at Station 6. Copepods were common only at
Station 7.
Euphausia pacifica was a principal prey for
Pacific sanddab at Stations 6 and 7. Pandalusjor-
dani occurred in 2&7( of the fish at Station 6, but
was uncommon at Station 7. Decapod crab larvae
and copepods, on the other hand, were common
prey only at Station 7.
Variations With Seasons or Size of Fish
Changes in the relative proportions of the major
taxa of food consumed by different sizes of the four
species of flatfishes are shown for "summer"
(May-September) and "winter" (October- April) in
Figures 2-5. Because food habits as well as sizes of
fishes vary among stations (Tables 3, 6; Pearcy
1978), geogi'aphic effects are confounded in these
figures.
Annelids usually dominated the diet of all size
groups of these juvenile Dover sole during both
seasons (Figure 2). Crustaceans appeared to de-
crease in importance with increasing size of fish
during the winter season, but reached peaks in the
summer. Mollusks iSolegasters spp., Yoldia ensif-
era, and unidentified pelecypods) attained peaks
in the diet of intermediate-sized (200-300 mm)
Dover sole, and echinoderms (mainly ophiuroids)
attained a peak at a larger size of fish.
100
NUMBER OF FISH
rej (3) (12) (27) 136)120) (11) (I) (7) (10) (5) (15) (22) (14) (I) (I)
I I I
0 100 200 300 400
1 I r
(WINTER)
/ . v6 / \
^-.
2/
/
;
/
\ .1
V / \ '
\ /
V
Figure 2. — The percent by wet weight of the
major food taxa for different length groups of
Dover sole for summer and winter. 1 = crusta-
ceans, 2 = annelids, 3 = other taxa, 4 = mollusks,
5 = echinoderms, and 6 = coelenterates.
100 200 300 400
STANDARD LENGTH (mm)
645
FISHERY BULLETIN; VOL. 76. NO. 3
The largest difference between seasons was for
coelenterates. Anemones and the feeding polyps of
sea pens were unimportant constituents of the
food during the summer (<29^ of diet by weight)
but were sometimes a major food >30'^ by weight)
during the winter. Anemones and sea pens are
probably available as prey during both seasons
but for some reason only consumed in significant
quantities during the winter.
Seasonal differences in the intensity of feeding
were also indicated by the higher frequency of
empty stomachs in winter than in summer (Table
7). The number of principal prey occurring in the
diet of Dover sole was consistently larger during
summer than winter regardless of fish size. Al-
though the smaller number of stomachs with con-
tents during the winter reduces sample size, and
hence the number of taxa found, the frequency of
occurrence of many of the individual taxa of
polychaetes, crustaceans, and mollusks (taxa
listed in Table 5) was higher in summer than
winter. Bertrand (1971) found no evidence for sea-
sonal variations in the numbers or biomass of
infauna sampled with grabs at these stations.
Therefore a more diverse assemblage of prey was
probably available to Dover sole during the sum-
mer or fish were usually less selective during the
summer than during the winter. The summer is
the season of most active growth of Dover sole
(Demory 1972) when intraspecific and possibly in-
terspecific competition for food may be most in-
tense. Decreased prey selectivity is known to occur
under conditions of low food abundance or avail-
ability (Ivlev 1961; Schoener 1971).
The number of principal prey taxa generally
increased with size of Dover sole (Table 7). This
trend may be related to sample size (number of
stomachs with food) and to the ability of large fish
to consume a larger range of prey sizes than small
fish. The less diverse diet of small fish resulted
from ingestion of only a few species of polychaetes.
Table 7. — Frequency of empty stomachs (no. empty stomachs/
no. fish) and the number of principal taxa (occurring in 5% or
more of at least 10 observations) of prey for different sizes of
Dover sole collected during summer and winter seasons.
Frequency of
Standard
empty
stomachs
Number of taxa
length (mm)
Summer
Winter
Summer
Winter
51-100
6/12
15/22
9
4
101-150
—
19/29
—
8
151-200
4/16
10/15
24
12
210-250
7/34
1 9/34
25
18
251-300
11/47
20/42
37
17
301-350
4/24
7/21
21
15
351-400
5,16
—
41
—
Those prey types eaten by a broad size range (50-
400 mm SL) of Dover sole include: Myrochele heeri,
Typosyllis hyalina, Lumhrlneris sp., Glyceridae,
gammarid amphipods, pelecypods, Megacrenella
Columbiana, ophiuroids, unidentified polychaetes,
and unidentified crustaceans.
Annelids and crustaceans were the major food
items for rex sole (Figure 3). (Most of the rex sole
represented here are juveniles.) Annelids in-
creased in importance with an increase in the sizes
of rex sole, up to 150-250 mm. This increase was
associated with a decrease in the proportion by
weight of crustaceans, the dominant food item for
small rex sole during both seasons. Euphausiids,
decapod crab larvae, copepods, and ostracods were
only found as principal prey of rex sole of <200
mm. Mollusks formed only a minor portion of the
diet. Differences in the FO of principal prey were
not pronounced. Some polychaetes (Sternaspis fos-
sor, Myriochele heerie ,Nothria geophiliformis , and
Chuelia pinnata) were found more frequently in
large (220-300 mm SL) rex sole.
Some seasonal differences in the diet of rex sole
were evident. Euphausiids were principal prey
only during the summer. Cumaceans and Oiko-
pleura were more common during the winter.
Principal prey that were commonly ingested by all
or most size groups during both seasons were:
Sternaspis fossor, Goniada brunnea, unidentified
polychaetes, gammarid amphipods, and uniden-
100
NUMBER OF FISH
f/9J (90) (33 J (41) (2) (6) (48) (63) (50) (82) (13)
1 I \ \
(SUMMER)
I no
200 300 0 100 200
STANDARD LENGTH (mm)
300
Figure 3. — The percent by wet weight of the major food taxa for
different length groups of rex sole for summer and winter. 1 =
crustaceans, 2 = annelids, 3 = other taxa, and 4 = mollusks.
646
PEARCY and HANCOCK: FEEDING HABITS OK FOUR FLATFISH SPF.CIKS
tified crustaceans. Kravitz et al. ( 1977) listed Not-
hria spp. as frequently occurring polychaetes and
Ampclisca macrocephola, Hippumedon wecomus,
Paraphoxus epistumusi?), and P. ubtusidens as
frequently occurring amphipod prey for rex sole.
Crustaceans composed the bulk of the diet of all
sizes of slender sole during both seasons (Figure
4). Annelids and "other taxa" were most important
in the diet of intermediate-sized (101-200 mm)
slender sole during either summer or winter.
Pelagic crustaceans such as copepods, eu-
phausiids, and crab larvae occurred frequently in
the diet of small (<150 mm) slender sole, whereas
polychaetes, the shrimps P. Jordan i and S. bi-
spinusa, and fishes were important for large slen-
der sole (>150 mm). Again, a larger number of
principal prey taxa occurred during the summer
than winter.
Crustaceans also were the most important taxa
in the diet of the Pacific sanddab, except for five
201-250 mm individuals during the summer,
when fishes composed 959^ of the food by weight
(Figure 5). Kravitz et al. (1977) found that all C.
sordidiis (90-377 mm total length) collected in
May off Oregon had been feeding intensively on
northern anchovy. Barss (see footnote 4) reported
NUMBER OF FISH
^2 J (15) (70) (41) (6) (7) (64) (179) (130) (7)
lOOn
X
o
>-
GQ
UJ
O
LJ
CL
(WINTER)
STANDARD LENGTH
Figure 4. — The percent by wet weight of the major food taxa for
different length groups of slender sole for summer and winter. 1
= crustaceans, 2 = annelids, 3 = other taxa, 4 = mollusks, and 5
= echinoderms.
(3) (96)(252)(55) (5)
lOOr-
NUMBER OF FISH
(32) (46) (4) (17) (I)
(WINTER)
A ,
100
200
300
STANDARD LENGTH (mm
Figure 5. — The percent by wet weight of the major food taxa for
different length groups of Pacific sanddab for summer and
winter. 1 = crustaceans, 2 = annelids, and 3 = other taxa.
that sanddab eat small fishes, squids, and oc-
topuses.
Crustaceans were the predominant prey during
both seasons and for most sizes of sanddab.
Euphausiids, copepods, and cumaceans occurred
more frequently in small than large individuals.
Pandaliis jordani, crangonids, and fishes were
most common in the diet of large Pacific sanddab.
DISCUSSION
The four common flatfishes caught in this study
compose two generalized feeding types. Dover and
rex soles feed almost exclusively on benthic inver-
tebrates, mainly polychaetes and amphipods,
while slender sole and Pacific sanddab prey
mainly on pelagic crustaceans. The food habits of
these two types are related to mouth structure and
digestive morphology. Flatfishes that feed on
benthos usually have asymmetrical jaws, small
stomachs, and long intestines, whereas pelagic
feeders have longer, symmetrical jaws with sharp
teeth and long serrated gill rakers, adaptations for
grasping and retaining animals that swim in
midwater (Hatanaka et al. 1954; Groot 1971).
Dover and rex soles belong to the benthos-feeding
type and sanddab and slender sole to the pelagic-
feeding type. Kravitz et al. (1977) also recognized
these two feeding types among five flatfishes off
647
FISHERY BULLETIN: VOL. 76. NO. 3
Oregon, and included rex sole as a benthophagus
species and Pacific sanddab as a piscivorous-
pelagic feeder.
Rae (1956, 1969) .studied the feeding habits of
the lemon sole, Microstomas kitt, and the witch,
Glyptocephalus cynoglossus, off Scotland. Some of
his results are remarkably similar to ours for the
congeneric Dover sole, M. pacificus, and rex sole,
G. zochirus. Both the witch and lemon sole, like
the Dover and rex soles, feed predominantly on
polychaetes. Crustaceans were next in importance
followed by other phyla such as mollusks,
echinoderms, and coelenterates. Ophiuroids and
anthozoans were also eaten by both lemon sole and
the witch. These similarities in diets indicate
common feeding specializations within pleuronec-
tid genera.
Although the major food of the lemon sole and
witch were very similar, these two species, like the
Dover and rex soles, preyed on different families or
different genera of the same family so that food
overlap, and presumably competition, are rare
(Rae 1956, 1969). As pointed out by Rae, these
differences in feeding habits reflect behaviorial
differences of the fishes as well as differences in
the composition of the benthic communities of
which these fishes are a part. The habitats of the
lemon sole and witch often differ, the lemon sole
preferring hard, rocky bottoms, and the witch soft,
muddy bottoms.
Both the lemon sole and witch fed most heavily
during the summer. Regional differences were
also marked. Polychaetes decreased in importance
as prey for the witch in shallow water ( < 100 m), as
they did in our study for Dover sole (Table 3). Rae
(1939, 1956) also believed that differences in the
types and quantities of food available between one
area and another resulted in different growth
rates of lemon sole. Sedentary polychaetes were
most common as prey in areas of rapid growth.
One of the objectives of this study was to learn if
differences in the availability of prey for flatfishes
occurred and how it may be related to sediment
types and water depth at our stations. The compo-
sition of prey of Dover sole clearly varies among
stations. Polychaetes were the main food at the
three deepest stations; echinoderms, coelenter-
ates, and polychaetes were similar on a weight
basis at the two 102-m stations; polychaetes, fol-
lowed by mollusks, were most important at the
100-m station; and mollusks and crustaceans were
most abundant at the 74-m station (Table 3).
Based on the percentage by weight of major food
taxa, higher similarities occurred among stations
at similar depths rather than with similar sedi-
ment types: Stations 15 and 23 at 102 m and the
deep stations 6, 2, and 8 at at 148-195 m (Table 4).
Sediment texture at Stations 15 and 23 were dis-
similar. (See Figure 1 for summary of depth and
sediments for the stations.) Although Station 2
had an average sediment texture that differed
from Stations 6 and 8, a thin layer of silt overlaid
coarse sand at Station 2, hence the surface sedi-
ment of Stations 6, 2, and 8 were probably more
similar than indicated in Figure 1.
The occurrence of individual species of poly-
chaetes consumed by Dover sole is probably a more
sensitive indicator of station differences than the
biomass of major taxa. Stations 7, 15, and 23, at
100-102 m, but with different sediment types,
were most similar in polychaete prey. Stations 2,
6, and 8 in deep water, at 148-195 m were again
similar. Thus, these similarities in prey for these
two groups of stations seem to be correlated with
depth. However, polychaete prey at Station 2 ( 190
m) was similar to that of Station 7 (100 m), which
had similar sediment type, as well as that at Sta-
tions 15 and 23 ( 102 m) with different sediments.
Stations 22 and 15 with similar sediment types,
but at different depths, had low similarity of
polychaete prey.
Based on 82 species of mollusks, cumaceans, and
ophiuroids sampled in O.l-m^ Smith-Mclntyre
grabs, Bertrand ( 197 1 ) calculated the similarity of
the fauna among the same seven stations included
in this study. He also found that Stations 2, 6, and
8 formed a deep-water group of high similarity.
Stations 7 and 23 (at 100-102 m) were similar, as
were Stations 7 and 8, with different depths and
sediment types. Gunther (1972) also calculated
similarities among these same stations based on
living benthic foraminifera and found that strong
faunal affinities crossed depth and sediment
boundaries. Again, Stations 2, 6, and 8 formed one
group. Stations 2 and 7, and 15 and 22, station
pairs based on sediments, were not very similar.
Similarities among the fishes caught were also
strong among Stations 2, 6, and 8. The remaining
stations (7, 15, 22, and 23) formed another group of
high affiinity (Pearcy 1978). These two species
associations agree with those described by Day
and Pearcy (1968) for the continental shelf off
central Oregon. They found a shallow (42-73 m)
water association on a sand bottom dominated by
Pacific sanddab and English sole, Parophrys vet-
iilus, and an association at 119-159 m on a silty-
648
PEARCY and HANCOCK: FEEDING HABITS OF FOUR FLATFISH SPECIES
sand bottom dominated by slender sole and rex
sole.
Shallow-water and deep-water associations are
therefore evident at these stations, based on pre-
vious studies of benthic invertebrates and verte-
brates, as well as the composition of the diet of
Dover sole in this study. Because surface sed-
iments were fairly similar at our three deep
stations, sediment vs. depth effects could not be
separated here. The lack of precise similarities of
sediment types for station pairs also weakens this
part of our study. Nevertheless, stations with the
most similar sediment types often had low similar-
ity of benthic fauna. We conclude that depth-
related factors may have greater influence than
sediment type on the composition of benthic fishes,
fish food, and invertebrate fauna within the
boundaries of our study area. This conclusion
must be tempered, however, by the realization
that other sediment parameters besides texture
and percent organic matter may be important, and
we simply did not study the proper sediment
characteristics. We agree with Peterson ( 1918): "It
is clear then that the character of the bottom is of
fundamental importance for the presence or ab-
sence of epifauna. Nevertheless, the succession of
the various types of epifauna and of the com-
munities belonging to the level bottom cannot be
explained by the character of the bottom alone."
Bertrand ( 1971) estimated the "edible" biomass
of infauna ( >1.0 mm) for demersal fishes (i.e., all
infauna less holothurians, echinoids, echiurids,
and burrowing anemones) at these stations from
0. 1-m'^ Smith-Mclntyre grab samples taken on the
same cruises. He detected no seasonal variations
in the wet or ash-free dry weight of this biomass
fraction. The ash-free dry weights per square
meter for polychaetes, mollusks, and crustaceans
given by Bertrand for the seven stations are shown
in Table 8. Crustacean biomass was consistently
low at all stations, probably because of the ineffec-
tiveness of the grab to sample epibenthic and
motile amphipods, major food items of Dover and
rex soles. There was no direct or consistent rela-
tionship between the biomass per square meter of
T.\BLE 8. — Ash-free dry weights in grams per square meter of
macro-infaunal fish food at the seven stations (from Bertrand
1971).
Taxa
Stn 22
23
15
8
Polychaetes
Mollusks
Crustaceans
Total
0.04 0 17 0.30 008 0.17 0.14 0.19
4.50 1.10 1.74 209 0.20 0.16 0.07
0.006 0.005 0 001 0004 0.008 0.004 0.003
4 55 1.28 2.04 2.17 0 38 0.30 0 26
"edible" fish food and the biomass of all fish or
Dover sole. Stations with similar standing stocks
of infaunal food had widely different standing
stocks of benthic fishes.
Station 22, the beach sand station — with the
lowest organic carbon in the sediment of all
stations — supported a fairly low biomass of fish,
but the largest biomass of edible fish food, 4.55
g/m^, and the largest biomass of invertebrate mac-
robenthos. Conversely, Wigley and Mclntyre
(1964) found the largest biomass in finer sedi-
ments off Massachusetts, and Lie and Kisker
(1970) found that the shallow-water sand com-
munities off Washington had a lower average
standing stock of infauna than deeper com-
munities on the shelf. The large biomass at Sta-
tion 22 is composed primarily of the bivalves Acila
castrefhsis and secondarily of Tellina salnwnea.
Both of these mollusks were principal prey of
Dover sole only at Station 22 (Table 5). Although
the frequency of occurrence of these two mollusks
in Dover sole stomachs was only 107^ , mollusks
composed 58^f by weight of the food of Dover sole
at this station. Thus Dover sole are versatile pred-
ators, changing their diets opportunistically in re-
sponse to changes of prey availability.
The dominant fish at Station 22 was Pacific
sanddab, primarily a pelagic feeder. Mollusks
were not principal prey. Acila and other burrow-
ing animals are unavailable as food for fishes
adapted for pelagic feeding, illustrating a basic
reason for the lack of any direct relationships be-
tween edible fish food and fish biomass.
The average biomass of Dover sole was directly
related to the biomass of their principal food,
polychaetes, at the seven stations (Figure 6). Sta-
tion 22, where Dover sole consumed principally
mollusks, had the lowest biomass of both
polychaetes and Dover sole; intermediate values
of biomass of both fish and food are found at the
three deep stations (2, 6, 8). The three stations at
about 100 m (2, 7, and 15) differed markedly in
standing stocks of both polychaetes and Dover
sole. This positive correlation (r = 0.73) of stand-
ing stocks of predator and prey implies that Dover
sole selected habitats within our study area where
their prinicpal preferred food was most abundant
regardless of depth and bottom type. Of more fun-
damental interest is the fact that standing stocks
of polychaetes may indicate the amount of food
available to Dover sole, and perhaps the produc-
tion rates of polychaetes at the different stations.
Similar direct relationships between standing
649
FISHERY EU'LLETIN: VOL. 76, NO. 3
CM
I
E
2 4
a>
o
>
o
Q
85'7f of their time in
water warmer than 20° C and < 10% in water col-
der than 18° C (Dizon et al. in press).
In the western South Pacific, skipjack tuna have
been caught in water near 15° C, off Tasmania
(Robins 1952) and off eastern Australia (G. I.
Murphy, Division of Fisheries and Oceanography,
CSIRO, New South Wales, Australia, 1977, pers.
commun.). These fish probably belong to a differ-
ent subpopulation (Fujino 1972) than fish found in
Hawaii.
Lower Dissolved Oxygen Limit
Gooding and Neill (see footnote 4) examined the
effects of low dissolved oxygen concentrations on
skipjack tuna. Their animals, habituated in open
tanks with circulating, essentially saturated sea-
water (4.5 ml 0.>/l, or 6.4ppm), were transferred to
tanks in which the concentration of dissolved oxy-
gen could be maintained at a preselected constant
subsaturation level. Temperatures in both sets of
tanks were ambient, 23° to 24° C. Dissolved oxy-
gen concentrations down to 1.0 ml/1 ( 1.4 ppm) were
used. Resistance times and swimming speeds were
measured, and general behavior was observed for
up to 4 h in each experiment. Under their experi-
mental conditions, Gooding and Neill concluded
that hypoxic stress was first manifest, through
changes in swimming behavior and speed, at
about 2.8 ml/1 (4.0 ppm), a value fairly typical for
fish. Lethal oxygen levels, leading to death in 4 h
or less, were found to be higher than those for any
other freshwater or marine fish thus far studied.
Only one fish (out of six) survived 4 h at 2.5 ml/1
(3.5 ppm), and none survived as long as 2 h at still
lower concentrations. At higher oxygen values,
above 2.5 ml/1, all skipjack tuna tested in this
study survived at least 4 h.
Because we sought to estimate the lowest dis-
solved oxygen concentrations that skipjack tuna
can tolerate indefinitely without significant
stress, we have chosen a conservative value of 3.5
ml/1 (5 ppm) as the lower limit to the skipjack
tuna's habitat, where temperature and other vari-
ables are not limiting.
Upper Temperature Limit
The case for an upper temperature limit to the
skipjack tuna's habitat is somewhat less direct.
Three small (30-35 cm) individual skipjack tuna
maintained in water warmed l°C/day survived
BARKLEY ET AT: SKIPJACK TUNA HABITAT
until the temperature reached 33 °C, when two
died; the other lived until the water reached 34° C
(Dizon et al. 1977). Skipjack tuna have a high
metabolic rate and a countercurrent heat ex-
changer in their circulatory system which dramat-
ically restricts heat loss through the gills. This
accounts for the fact that freshly caught wild skip-
jack tuna can have red muscle core temperatures
as high as 11° C above that of the surrounding
water (Stevens and Fry 1971).
Temperature excesses of this magnitude could
lead to dangerously high muscle temperatures if
they occur in the warmer parts of the ocean. To
examine this possibility we use a heat balance
model developed for skipjack tuna by Neill et al.
(1976) which yields an estimate of temperature
excess in the red muscle core as a function of size
and metabolic activity (Figure la). Actual muscle
core temperatures are found by adding the values
shown in this figure to the temperature of the
surrounding water, for any given size fish.
Clearly, large skipjack tuna in surface waters of
the tropics must either tolerate high muscle core
temperatures, or reduce their metabolic activity
substantially below the 3 mg Og g"^ h-^ level.
But skipjack tuna appear to avoid heating their
muscle tissue much above 35° C (Stevens and Fry
1971 ). This upper limit must place a similar upper
limit on the water temperatures which skipjack
tuna can inhabit, unless they can thermoregulate
physiologically or behaviorally. In Figure lb,
35°C is taken as the upper limit for the red muscle
core, and temperature excesses of Figure la are
subtracted from that value to arrive at an estimate
of the upper temperature limits for the habitat of
skipjack tuna, as a function of size. If the values
thus obtained are valid, these fish should be able to
live anywhere in the ocean when they are small,
but they should be limited to lower and lower
environmental temperatures as they grow. The
largest known skipjack tuna, weighing approxi-
mately 16 kg, would — if active enough — be
confined to water temperature near 18°C, which is
also their approximate lower limit.
SKIPJACK TUNA HABITAT
HYPOTHESIS
We hypothesize that skipjack tuna of the central
and eastern Pacific Ocean occupy a primary
habitat — a volume of water whose properties they
can tolerate indefinitely — which is 18°C or
warmer, but cooler than the upper limits for nor-
mally active animals shown in Figure lb, provided
that the dissolved oxygen concentration is at least
3.4 ml/I (5 ppm). Skipjack tuna can presumably
20
u
a:
IT
hJ
a.
o
v>
3
S
to
15
10
^
"^
1
^.^
,^
"^
■^
^
i
1
1 ' 1 '
^
K
^
^
^3mg«(
H
^^
|x
1
A
r
!
i
Y'
J
Y"
/
/
. —
/
' —
/
f
, — 1
■^
' '
/
■"^
^r,P» lmg«02g''hr'
1
/
..^
J
r^
>
10 15
WEIGHT OF SKIPJACK TUNA (Kg)
20
Figure la. — Calculated excess of internal temperature, over that of the surrounding water, in red muscle for skipjack tuna of all
known sizes. Values are shown for a measured minimum (anesthetized) level of metabolic activity (lower line) and our estimate of the
mean metabolic activity for normally active animals (upper line), triple the minimum level. (From Neill et al. 1976.)
655
FISHERY BULLETIN. VOL, 76, NO. 3
35
30
iij
Q.
2
25
20
15
N
— MUSCLE BREAK- DOWN TEMPERATURE (?)
\
\
r — -
p_
^Tm = 1 ma*0, a'hr"'
\
1 ^
/
\,
I 1
r— ______
s
s,
■ 1
\
1
^
1 1
N
' 1 1
X
X,
^
;
1
^3 mg'Og
^-<
1 1
"V.
■-^
■^,
^-^
^
,^^
"^
10 15
WEIGHT OF SKIPJACK TUNA (kg)
20
FIGL'RE lb. — Maximum tolerable water temperature for skipjack tuna as a function of size for the same two rates of metabolic activity
illustrated in Figure la. Based on calculated internal temperature excesses and the assumption that damage to red muscle tissue occurs
above 35 "C.
200
la'CU-SSml/LOg-i — «— I8°C^"^
-3.5ml/L02-
-I8°C-
5° S
200
Figure 2. — Upper panel: Temperature and dissolved oxygen (selected isopleths only) along long. 119' W, eastern Pacific Ocean,
August 1967 (Love 1972). Lower limits of the skipjack tuna habitat are assumed to be either 18° C or 3.5 ml/1 dissolved oxygen, as
indicated. The hatched layer should be warm enough for these fish, but oxygen deficient. Lower panel: Hypothesized habitat layers for
skipjack tuna of two sizes, 4 kg ( entire hatched area ) and 9 kg ( cross hatched area only) in the same section. Fish <4 kg could presumably
live anywhere between the sea surface and the lower limits, 18°C or 3.5 ml/1 of dissolved oxygen.
656
BARKLEY ET AL: SKIP.JACK TUNA HA15ITAT
leave their primary habitat only for limited
periods of time without suffering thermal or oxy-
gen stress. Prolonged excursions to colder water
would require increased activity and thus more
dissolved oxygen and food. Skipjack tuna could
stay in the warm upper layers only if they would
reduce their physical activity, or tolerate over-
heating. Hypothetical consequences of these con-
ditions are illustrated in Figure 2 which shows the
hypothetical layers along long. 119°W for skip-
jack tuna of two sizes. The 4-kg fish are those most
abundant in catches by the eastern Pacific fishery;
9-kg fish are the largest normally found there, and
then only in certain areas such as the Revil-
lagigedo Islands (ca. lat. 17° N, long. 112°W).
The deeper limit of the habitat should be the
same for skipjack tuna of all sizes. The upper limit
is deeper and more restrictive for larger fish,
which find essentially no habitable water between
lat. 5° and 12° N, a distance of more than 700 km
or 400 n.mi. Larger fish also have much less con-
tinuous access to the sea surface than those weigh-
ing 4 kg. Only skipjack tuna of the smallest size
commonly found in this area (<4 kg) could inhabit
all of the water above the lower limits in Figure 2.
Figures 3 to 6 are maps of a hypothetical skip-
jack tuna habitat for the entire central and east-
ern Pacific Ocean, based on oceanographic station
data used in preparing the Oceanographic Atlas of
the Pacific Ocean (Barkley 19681. For these maps,
Figure 3.— Hypothetical maximum depth (meters) of the skipjack tuna habitat in the eastern Pacific Ocean, as determined by the
depth of the 18°C isotherm (hatched area) or the 3.5 ml/1 (5 ppm) isopleth of dissolved oxygen (cross hatched area). Contour interval is
50 m except for a few areas near the coast, where a 25-m contour interval is used.
657
FISHERY BULLETIN, VOL. 76, NO. 3
Figure 4. — Hypothetical minimum depth of the skipjack tuna habitat in the Pacific Ocean east of long. 180°, for fish weighing about
6.5 kg ( 14 lb) which are limited to water cooler than 24° C. Contours show the depth, in meters, of the 24 °C isotherm.
station data were averaged within 2° areas of
latitude and longitude, for all months, to approxi-
mate annual mean conditions.
Figure 3 shows the hypothetical "floor" of the
skipjack tuna habitat, i.e., maximum depths (in
meters) for this species. In Figure 3, the unhatched
areas off the Pacific coast of the Americas, and at
latitudes higher than about 40° in both hemi-
spheres, indicate water which, at all depths, is
colder on average than 18° C; presumably skipjack
tuna would not normally be present.
Figure 4 shows the minimum habitat depth or
ceiling for 6.5-kg fish. Outside of the hatched area
annual mean surface temperatures are <24°C,
and 6.5-kg fish would normally have access to all
of the water column above the habitat floor (Fig-
ure 3), up to and including the sea surface.
Figure 5 shows the hypothesized habitat layer
thickness for 6.5-kg skipjack tuna. In some areas,
there is no water cooler than 24° C with more than
3.5 ml/1 dissolved oxygen, so in these areas there is
no habitat for 6.5-kg or larger fish; such areas are
double hatched on Figure 5. Extensive regions
around these areas have habitat layer thickness of
10 m or less, and large areas of the equatorial
Pacific Ocean have <25 m of habitat layer thick-
ness. This rather thin layer can lie beneath as
much as 1 50 m of water warmer than 24 ° C ( at lat.
4°N, long. 170°W, e.g.). North of the Hawaiian
Islands, the opposite situation is present: a 150-m
thick habitat layer lies under 25 to 50 m of water
658
BARKI.EY KT AI.: SKIPJACK TUNA HABITAT
Figure 5.— Thickness of the hypothetical habitat layer (meters), for 6.5-kg skipjack tuna in the eastern and central Pacific Ocean.
Contours were obtained by subtracting depths of the upper habitat limit (Figure 4) from the lower one (Figure 3). In the crosshatched
areas off Mexico and Peru, there should be no habitat suitable for fish of this or larger sizes. In the hatched area, water warmer than
24°C is present above the habitat layer. Immediately beyond this area the habitat (dashed depth contours) extends to the sea surface.
Outside of the 18° C surface isotherm, the water is probably too cold for skipjack tuna.
warmer than 24 C. Clearly, the South Pacific
Ocean offers the roomiest habitat to large skipjack
tuna, and in fact some of the largest known indi-
viduals of this species were caught slightly south
of Tahiti (lat. 17"S. long. 150"W), according to
catch records of longline fishing boats (R. A.
Skillman, Southwest Fish. Cent. Honolulu Lab.,
Natl. Mar. Fish. Serv., NOAA. Honolulu, HI
96812, 1977, pers. commun.).
An interesting feature of the habitat map for
6.5-kg fish I Figure 5) is a channel of relatively cool
and adequately oxygenated water some 200 km off
the coast of Mexico. This channel should allow
skipjack tuna as large as 6.5 kg to pass from the
Baja California fishery to the equatorial fishery, or
vice versa, when fish as small as 4 kg would find
stressful conditions for hundreds of kilometers on
either side of that channel. Seasonal and year-to-
year closure or shifting of this channel could readi-
ly explain puzzling variations in the distribution
of skipjack tuna catches in the eastern Pacific.
Figure 6 shows areas of presumably stressful
environment (zero habitat thickness) in the east-
ern tropical Pacific for fish of various sizes and
therefore temperature limitations. Skipjack tuna
>11 kg should find no habitat at all within the
shaded area. Fish 4 kg in size would find no habitat
within the smallest contour (temperatures above
659
FISHERY BULLETIN, VOL. 76, NO. 3
26° C). Fish weighing <4 kg should find some
thickness of habitable water, within and just
below the upper mixed layer, everywhere in the
eastern tropical Pacific.
DISCUSSION
Although we anticipate that temperature and
dissolved oxygen will prove to be primary deter-
minants of the habitat of skipjack tuna in all
oceans, it is possible that limiting values of these
variables may differ from one population or region
to another. The lowest temperature (ca. 15° C) at
which skipjack tuna are caught in Australian
waters (Robins 1952) is considerably lower than in
the eastern Pacific. The fish caught off Australia
may also differ in their ability to tolerate warm or
low-oxygen water.
The gross features of the distribution of skipjack
tuna in the eastern tropical Pacific, where only
small skipjack tuna are found in large numbers
(Williams 1970), agi-ee with the hypothesis. Those
areas where large skipjack tuna do occur (Ma-
tsumoto 1975) are outside of the hatched area in
Figure 6: the Revillagigedo Islands, e.g., are just
north of the hatched area, and Tahiti is well south
of it.
The hypothetical habitat proposed here ex-
plains why skipjack tuna leave the northern
fishery of the eastern Pacific when they reach a
certain size. To find cooler, better oxygenated
water as they grow, these fish must move out of the
eastern tropical Pacific toward higher latitudes in
the central Pacific. Also, they must then spend less
time at or near the sea surface, since the thermo-
cline, where they live, is generally much deeper in
the central Pacific, and the water above the ther-
mocline is too warm to permit normal activity.
This size-specific movement in response to the en-
vironment is consonant with Rothschild's (1965)
migration model for the eastern Pacific skipjack
tuna population. It also suggests a mechanism for
the evolution of migratory processes, an important
topic in marine ecology.
For several reasons, existing fishery data are
inadequate for making a refined judgment of our
skipjack tuna-habitat hypothesis: 1) Commercial
fishery data generally include neither information
Figure 6. — Average water temperature in the eastern Pacific Ocean at those depths where the concentration of dissolved oxygen is 3.5
ml/1. Deeper water is cooler and lower in oxygen, shallower water is warmer and has more oxygen. See Figure lb for the relationship
between skipjack tuna size and upper temperature limits.
660
BARKI.EY ET AL: SKIPJACK TUNA HABITAT
on individual sizes of skipjack tuna composing the
catch nor synoptic information on the vertical dis-
tribution of temperature and dissolved oxygen in
the fishing area. 2) The degree to which catch per
unit effort measures fish abundance may vary
greatly with gear type, fish size, and environmen-
tal conditions. For example, the habitat
hypothesis implies that purse seines (which fish
the upper 50 m or so of the water column ) should be
most effective in those parts of skipjack tuna
habitat with a shallow floor, hi fact, the eastern
Pacific purse seine fishery operates almost en-
tirely in waters with a skipjack tuna habitat-floor
at depths of 50 m or less (c.f. our Figure 3 and fig. 1
of Matsumoto 1975). Efforts to catch skipjack tuna
by purse seining in Hawaiian waters — where the
habitat-floor lies at depths near 200 m (Figure
3) — have been ineffectual (Murphy and Niska
1953 k Green ( 1967) reported strong positive corre-
lations between the success of purse seining for
eastern Pacific skipjack tuna and yellowfin tuna,
Thunnus albacares, and the presence of shallow
( «60 m to top), steep ( >0.55°C cm ' ) thermoclines.
3) Commercial fishermen naturally fish only
where they expect to find and catch fish; thus,
fishing effort tends to be very unevenly distri-
buted.
A partial test of the habitat hypothesis might be
achieved through experimental fishing in and
near the hatched areas of Figure 6; fishing effort,
the sizes of captured skipjack tuna, and vertical
distributions of temperature and dissolved oxygen
would need to be measured at each fishing loca-
tion. But, experimental fishing — even if con-
ducted in a thorough and systematic fashion —
might not yield a conclusive test of the hypothesis,
because there would still be no guarantee that
catch per unit effort accurately reflected flsh
abundance (reason 2 above).
We advocate, instead, the application of ul-
trasonic telemetry to test the skipjack tuna-
habitat hypothesis. Because skipjack tuna tagged
with ultrasonic transmitters tend to stay with
their school (Yuen 1966) and because skipjack
tuna schools tend to be homogeneous with respect
to fish size (Brock 1954), the track of a single
tagged fish could be taken as representative of a
large number of normally behaving, similarly
sized fish. Pressure-sensitive ultrasonic transmit-
ters ( like that described by Luke et al. 1973 ) would
permit continuous monitoring of fish position in
all three spatial dimensions through time.
Spatial-temporal coordinates offish could then be
compared with synoptic data on vertical distribu-
tions of temperature and dissolved oxygen. A few
dozens of such comparisons, for fish of various sizes
in waters with diverse vertical distributions of
temperature and dissolved oxygen, would consti-
tute a valid and sufficient test of the habitat
hypothesis. Toward this end, preliminary tele-
metry work is now underway at this Laboratory.
SUMMARY
Work with captive skipjack tuna at this
Laboratory has yielded information on the tem-
perature and dissolved oxygen requirements of
this species. If these laboratory results apply to
skipjack tuna in nature, they provide new insight
into the evolution of migration in skipjack tuna
populations, make it possible to account for the
geographic distribution of skipjack tuna on the
basis of environmental conditions, and provide
means for predicting their movements in major
fisheries such as those of the eastern tropical
Pacific.
In particular, we suggest that only young skip-
jack tuna can inhabit tropical surface waters, and
that the habitat of adult skipjack tuna in the
tropics is the thermocline and not the warmer
surface layer, as has generally been thought.
Since the thermocline in many areas is too
oxygen-poor to support these active fish and the
well-oxygenated surface layer is too warm for
adult skipjack tuna, only heat-tolerant young
skipjack tuna can live in those areas. As they
grow, these fish are forced to move into areas
where well-oxygenated water of the proper tem-
perature is more readily available.
Up to now, it has not been possible to trace the
movements of migrating skipjack tuna largely be-
cause they move through areas of many millions of
square miles, at unknown depths. Knowledge of
their temperature and dissolved oxygen require-
ments dramatically reduces the scope of the prob-
lem: the fish should be in a well-defined layer of
water, of directly and easily measured thickness,
whose geographic extent can be sharply defined
with either historical or current oceanographic
observations.
ACKNOWLEDGMENTS
The physiological studies on which this paper is
based were supported, in part, by the University of
Wisconsin Brittingham Foundation.
661
FISHERY BULLETIN, VOL. 76, NO. 3
We thank John J. Magnuson (University of
Wisconsin) and Garth I. Murphy (CSIRO, Aus-
traha) for reading this manuscript and providing
valuable suggestions.
LITERATURE CITED
Bakkley. R. a.
1968. Oceanographic atlas of the Pacific Ocean. Univ.
Hawaii Press, Honolulu, 20 p., 156 fig.
BROCK, V. E.
1954. Some aspects of the biology of the aku, Katfiuwonus
pelamix. in the Hawaiian Islands. Pac. Sci. 8:94-104,
DIZON, A, E.
1977, Effect of dissolved oxygen concentration and salinity
on swimming speed of two species of tunas. Fish. Bull.,
U.S. 75:649-653.
DIZON. A. E., R. W. BRILL. .AND H. S. H. YUEN
In press. Correlation between environment, physiology,
and activity and its effect on thermoregulation in skipjack
tuna, Katsuu'onus pelamis. In G. D. Sharp and A. E,
Dizon (editors). The physiological ecology of tunas.
Academic Press, N.Y.
DIZON, A. E., W. H. Neill, and J. J. Magnuson
1977. Rapid temperature compensation of volitional
swimming speeds and lethal temperatures in tropical
tunas (Scombridae). Environ. Biol. Fishes, 2:83-92.
FU.JINO, K,
1972. Range of the skipjack tuna subpopulation in the
western Pacific Ocean, In K. Sugawara (editor), The
Kuro.shio II, p. 373-384. Saikon Publ. Co. Ltd., Tokyo.
Green, r. e.
1967. Relationship of the thermocline to success of purse
seining for tuna. Trans. Am. Fish. Soc. 96:126-130.
Love, C. M. (editor).
1972. EASTROPAC atlas. Vol. 5. Physical oceanographic
and meteorological data from principal participating
ships. Second survey cruise, August-September
1967, U.S, Dep. Commer., NOAA Tech. Rep. NMFS
CIRC-330.
Luke, D. McG., D. G, Pincock, and a. B. Stasko.
1973. Pressure-sensing ultrasonic transmitter for track-
ing aquatic animals. J, Fish, Res. Board Can. 30:1402-
1404.
MATSUMOTO, W. M.
1975. Distribution, relative abundance, and movement of
skipjack tuna, Katsuwonus pelaniis, in the Pacific Ocean
based on Japanese tuna longline catches, 1964-67. U.S.
Dep. Commer., NOAA Tech. Rep. NMFS SSRF-695, 30 p.
MURPHY, G. I., AND E. L. NISKA.
1953. Experimental tuna purse seining in the central Pa-
cific. Commer. Fish. Rev. 15(4):1-12.
Neill. w. h., r. k. c. Chang, and a. e. dizon
1976. Magnitude and ecological implications of thermal
inertia in skipjack tuna, Katsuwonus pelamis (Lin-
naeus). Environ. Biol. Fishes. 1:61-80.
Robins, J. P,
1952, Further observations on the distribution of striped
tuna., Katsuwonus pelamis L., in eastern Australian wa-
ters, and its relation to surface temperature. Aust. J.
Mar. Freshwater Res. 3:101-110.
Rothschild, B. J.
1965. Hypotheses on the origin of exploited skipjack tuna
(Katsuwonus pelamis) in the eastern and central Pacific
Ocean. U.S. Fish. Wildl. Serv., Spec. Sci. Rep. Fish. 512,
20 p,
Stevens, E, D., and F. E. J. Fry.
1971. Brain and muscle temperatures in ocean caught and
captive skipjack tuna. Comp, Biochem, Physiol.
38A;203-211,
Williams, F,
1970. Sea surface temperature and the distribution and
apparent abundance of skipjack ( Katsuwonus pelamis) in
the eastern Pacific Ocean 1951-1968. Inter-Am. Trop.
Tuna Comm. Bull. 15:229-281.
Yabe, H., Y. Yabuta. and S, Ueyanagi
1963. Comparative distribution of eggs, larvae and adults
in relation to biotic and abiotic environmental factors. /n
H. Rosa, Jr. (editor). Proceedings of the World Scientific
Meeting on the Biology of Tunas and Related Species, p.
979-1009. FAO Fish. Rep. 6.
YUEN, H. S. H.
1970. Behavior of skipjack tuna, Katsuwonus pelamis. as
determined by tracking with ultrasonic devices. J. Fish.
Res, Board Can. 27:2071-2079.
662
HISTORICAL TRENDS AND STATISTICS OF THE
SOUTHERN OSCILLATION, EL NINO, AND INDONESIAN DROUGHTS
William H. Quinn, David O. Zopf. Kent S. Short, and Richard T. W. Kuo Yang'
ABSTRACT
A 116-yr Southern Oscillation index record was used in conjunction with environmental data and
reports from various authors on disturbances to the anchoveta fishery, marine bird life, etc. off the
Peruvian coast, to infer the occurrence of past El Nino type events and their intensities. The resulting
long time history substantiates our earlier report that certain Southern Oscillation index features are
excellent precursors of subsequent El Nino type events. We suggest that statistics derived from this
time history could be useful in the management of the Peruvian anchoveta fishery and for providing
long-range outlooks on El Nino type activity.
Anomalously heavy precipitation in the central and western equatorial Pacific and Indonesian
droughts were closely associated with El Nino type events.
In recent years the world demand for fishmeal has
continued to increase, as has the world population.
The Peruvian anchoveta fishery, which ordinarily
provides over half the world's supply of fishmeal,
has become a critical resource; and anything that
affects the output of this fishery is of world-wide
significance. Johnson and Seckel (1977) reported
that the catch in this fishery declined from a high
of over 12 million tons (about V.-> of the total world
catch of all fish) in 1970 to about 2 million tons in
1973. Although overfishing in 1970-71 may have
contributed heavily to this decrease in anchovy
catch, the strong El Nino of 1972-73 was undoubt-
edly also a major cause for the precipitous decline
in catch (Figure 1). However, the 1975 catch was
still only about 257c of the record 1970 catch, the
1976 catch remained low, and the target for 1977
has now been reduced to 2 million tons of an-
choveta and other fish such as sardines and hake.
Apparently the unfavorable environmental condi-
tions caused by the very weak event of early 1975
and the moderate El Nino of 1976-77 have not only
contributed to the delay in recuperation of the
fishery, but also are causing a further degradation
of it. In early October 1977 the Fisheries Ministry
of Peru said (according to a Reuters wire service
report) that the stocks were believed to be so low
that the anchoveta fishing, which was suspended
in May 1977, would not resume until the second
half of 1978.
'School of Oceanography, Oregon State University, Corvallis,
OR 97331.
Statistical information pertaining to the histor-
ical occurrence of El Nino type events is presented
to: 1) aid in long-term fishery assessment (Peru-
vian anchoveta fishery); 2) provide a basis for
speculative long-range outlooks on event occur-
rence (beyond a year in advance); and 3) guide
long-range predictions ( 1-12 mo in advance). Rela-
tionships between El Nino type events. Southern
Oscillation index trends, index component trends,
and Indonesian droughts are shown and discussed.
LD
14
Li_ IP _
0 '"^^
CO
1 10
_1
_)
^ 81-
X
o
I-
<
>
o
X
o
6-
4 -
0
I I I
±
J \ \ \ I I
1965
1970
YEARS
1975
Manuscript accepted January 1978.
FISHERY BULLETIN: VOL. 76, NO. 3. 1978.
Figure l. — The Peruvian anchovy catch for the period 1962-76
as obtained from the Industrial Fishery Products Market Review
and Outlook for June 1977 'National Marine Fisheries Service
19771. The 1976 figure is a preliminary value. The 1977 figure is
the Peruvian Fishery Ministry target value for anchovies and
other species such as sardine and hake, as reported by Reuters
wire service on 19 October 1977.
663-
FISHERY BULLETIN: VOL 76, NO. 3
Wooster (1960), Idyll (1973), Miller and Laurs
( 1975), and Caviedes ( 1975) furnished background
information on El Nino; Quinn (1974) discussed
monitoring and prediction; and Berlage (1957,
1966). Troup (1965), and Quinn (1971, 1976) pro-
vided background information on the Southern
Oscillation and how it relates to phenomena dis-
cussed in this paper.
Definitions for terms frequently used in this
paper follow: The Southern Oscillation was origi-
nally identified by Walker (1924). It was loosely
defined by Berlage (1966) as a fluctuation in the
intensity of the intertropical general atmospheric
and hydrospheric circulation over the Indo-Pacific
region. The fluctuation is dominated by an ex-
change of air between the South Pacific subtropi-
cal high and the Indonesian equatorial low.
The differences in sea level atmospheric pressure
between sites representing the South Pacific sub-
tropical high and sites representing the Indone-
sian equatorial low are used as indices to repre-
sent the Southern Oscillation (Quinn 1974).
The El Nino type event refers to the appearance
of anomalously warm sea surface temperatures
and abnormally heavy rainfall in the equatorial
Pacific and an invasion of anomalously warm sur-
face water off the coast of Peru and southern
Equador. This event, which is brought about by
relaxation from a prolonged period of strong
southeast trades, is represented by falling and low
Southern Oscillation indices (Quinn 1974). The
magnitude of the interannual relaxation and its
timing with relation to the regular seasonal relax-
ation (Southern Hemisphere summer) appear to
determine the strength of the El Nino invasion
along the Peruvian coast. Heavy central and west-
ern equatorial Pacific precipitation usually starts
a few or more months after El Nino initially sets
in, but this may not always be the case. By using
the term "El Nino type" we avoid arguments over
what is and what is not an El Nino and can then
account for events that evolve in a similar manner
but vary in timing, intensity, and extent.
The anti-El Nino refers to the contrasting situa-
tion when a strengthening and strong southeast
trade system prevails (represented by rapidly ris-
ing and high Southern Oscillation indices). At
such times we can expect strong upwelling (due to
the divergent equatorial flow under the influence
of strong southeast trades and equatorial easter-
lies), anomalously low sea surface temperatures,
and abnormally low amounts of rainfall over the
equatorial Pacific. Also, off the coast of Peru, we
find strong coastal upwelling, low sea surface
temperatures, lower than average sea level, and
generally favorable physical environmental con-
ditions for biological productivity (due to the up-
welling of nutrient-rich water from lower levels).
METHODS
Data Processing
Atmospheric pressure and much of the rainfall
data before 1961 were obtained from the World
Weather Records (Clayton 1927, 1934; Clayton
and Clayton 1947; U.S. Department of Commerce
1959, 1968). Data for 1961-76 were obtained from
Monthly Climatic Data for the World (U.S. De-
partment of Commerce 1961-76). We were primar-
ily interested in the large-scale interannual
changes. Therefore, we eliminated regular oscilla-
tions from the data, such as the diurnal cycle, by
using monthly mean values (or monthly amounts,
for rainfall), and the seasonal or annual cycle by
subtracting long-term average or normal monthly
values from the actual monthly values. Data so
processed show no particular regularity and no
apparent cycle (Panofsky and Brier 1965). The
filtered and unfiltered monthly anomalies were
used to detect, identify, and evaluate any unusual
changes that took place.
Our interests were focused on fluctuations of an
intermediate scale (Southern Oscillation), with
periods ranging between about 1 and 6 yr. The
remaining short period fluctuations in the
anomalies were eliminated by filtering with a low
pass filter. At the other end of the time scale, there
m.ay be a gradual change of the variate over many
years which is part of oscillations that are long
compared with the record. These extremely long,
gradual changes were not a factor in our study.
In earlier papers (e.g., Quinn 1974, 1976) the
12-mo running mean was applied directly to
monthly values of pressure, pressure differences
(indices), rainfall, etc. as a low pass filter. This
filter not only smoothed the data to some extent
but also eliminated the annual cycle. To more
clearly define the interannual fluctuations
(Southern Oscillation), we recently switched to
the use of the triple 6-mo running mean filter on
the monthly anomalies, which requires three suc-
cessive passes of the 6-mo running mean over the
data. It results in smoother plots and more clearly
defined peaks and troughs, which are of particular
assistance in establishing long-term trends. The
664
QUINN ET AL.: SOUTHERN OSCILLATION. EL NINO. AND INDONESIAN DROUGHTS
loss of 3 mo time with each application of the 6-mo
running mean is a drawback to its use in forecast-
ing, so we also use the 3-mo running mean and
monthly plots of anomalies for locating inflection
points and evaluating trends on a more immediate
basis in support of forecasts.
Anomaly trends for several indices were main-
tained in time section plots (Figure 2a, b) to
evaluate the Southern Oscillation and its expected
effects on the southeast trade system. Although
these limited records ( 25-30 yr) clearly showed the
close association of low indices with El Nino type
activity, and high indices with anti-El Nino condi-
tions (Quinn 1974, 1976), it was essential to
extend the study over a much longer period to
determine how frequently these climatic extremes
occurred
The World Weather Records were searched for
the longest and most complete atmospheric pres-
sure records which could be used to extend our
study into the past. Madras, India (1841-1976);
Bombay, India (1847-1976); Djakarta, Indonesia
(1866-1974); and Darwin, Australia (1882-1976)
were within the area noted by Berlage ( 1957,
1966) to reflect Southern Oscillation-related pres-
sure changes in the Indonesian equatorial low
pressure cell. Santiago, Chile ( 1861-1976) had the
only long pressure record that could possibly rep-
resent Southern Oscillation-related pressure
changes affecting the South Pacific subtropical
high pressure cell. Although Santiago is generally
to the east of the subtropical high, it does reflect
these pressure changes (Berlage 1957, 1966).
Correlations were run between the Tahiti-
Darwin index and the Santiago-Darwin index on
data for 1935-76 to further substantiate use of the
Santiago-Darwin index for representing the
Southern Oscillation and related El Nino type ac-
tivity. The Tahiti-Darwin index was used for this
comparison since it and the Santiago-Darwin
index showed similar amplitudes in their interan-
nual fluctuations. The similarity was due to the
fact that Tahiti and Santiago are separated by
analogous distances from the usual core of activity
in the subtropical high ( see fig. 10 in Berlage 1957,
or fig. 10 in Bjerknes 1969). At zero lag the correla-
tion coefficient between the two indices was 0.88.
The maximum correlation was 0.89 when the
Tahiti-Darwin index led the Santiago-Darwin
index by 1 mo.
Figure 3a-h shows the triple 6-mo running
mean plots of pressure anomalies for Madras
(1841-1976). Bombay (1847-1976), Djakarta
(1866-1974), and Darwin (1882-1976). They also
show similar plots of pressure index anomalies for
Santiago-Bombay ( 1861-81) and Santiago-Darwin
(1882-1976). The anomaly plots were used along
with other data in the evaluation of El Nino type
events reported over the past 135 yr.
Classification of Events
The classification of El Nino type events by in-
tensity is highly subjective since no two cases are
exactly alike with regard to time of onset, dura-
tion, areal extent, thermal departure, degree of
devastation, etc. Determinations concerning
event occurrence and intensity were primarily
based on: 1) reported disruptions of the anchoveta
fishery and marine bird life off the coast of Peru; 2)
scientific reports which discussed events that af-
fected the coastal regions of Peru and southern
Ecuador [e.g., Eguiguren ( 1894), Frijlinck ( 1925),
Murphy (1926), Hutchinson (1950), Sears (1954),
Schweigger (1961)1; 3) hydrological data for the
Peruvian coastal region; 4) sea-surface tempera-
ture data along the coasts of Peru and southern
Ecuador; 5) rainfall at coastal stations in Peru and
southern Ecuador; 6) height of preevent peaks and
depth of relaxation troughs in Southern Oscilla-
tion index trends; 7) related indications from
index component trends (when pressure compo-
nents from only one core of the Southern Oscilla-
tion were available); 8) sea-surface temperatures
over the equatorial Pacific; 9) rainfall data for
islands in the central and western equatorial
Pacific.
We categorized events as strong, moderate,
weak, or very weak, depending on the intensity of
the activity and the time of year that it occurred.
The true El Nino sets in during the first half of the
year. A symptom which is common to El Nirios is
the presence of anomalously high sea-surface
temperatures off the coasts of southern Ecuador
and Peru. Other frequently mentioned features
include a southward coastal current, heavy rain-
fall, red tide (aguage), invasion by tropical nekton,
and mass mortality of various marine organisms
including guano birds, sometimes with sub-
sequent decomposition and release of hydrogen
sulfide (known as El Pintor) (Wooster 1960).
Strong El Ninos are recognized as such by all
investigators; they involve positive sea-surface
temperature anomalies along the coast in excess of
3°C, they display most of the aforementioned fea-
tures, and the anchoveta fishery is seriously
665
FISHERY BULLETIN: VOL 76, NO. 3
I9«8 1909 1950 1951 195? 1953 1954 1955 1956 1957 1956 1959 I960 1961 1962 I96i
5
_
•■■■•.
1. -..1 -
5
2
-
.••■■■
...........
..•••■—]
*••••.,
.•
•_
2
JUAN FERNANDEZ -
DARWIN ( tibl
0
-2
2
-
■•■
.....
*
•
0
-2
2
-•...•
.••■•••.
., ...
..•••■■••
• ^
■':
0
'.
.••
.*■
•
.••■■•
1
0
EASTER -DARWIN
(mtJl
..'
,.."•
..*'
....
•■
-2
-
••..••■'
.. .•
-2
-3
2
-
r
-3
2
TOTEGEGIE- DARWIN
0
.•■■
■•.
.■■■■•■■.
..
1
0
-1
(mbl
u
..•■
•.
..••■
"...C
-2
2
-
-2
2
RAPA- DARWIN
0
-1
-
••
.■•••..
.' *'***
h*.
0
(mbl
-
•■'"■•..
.•■"■•.-.
...••'
'■•••'
-2
2
_
_.•■
.•*"
'■•..•'
..-
-2
2
0
~
*..
...
.-
..••-.......•
-:.-
0
(mbl
•-.
.. ..•
••••■
-2
-
-2
-3
-
-3
EN(W)
EN(M)
EN(S) »
EN(W)
1955 1956
YEARS
1957 1953
Figure 2a. — Triple 6-mo running mean plots of anomalies of the difference in sea level atmospheric pressure ( millibars) between Juan
Fernandez Is. (33°37'S, 78°50'W) and Darwin, Australia 1 12°26'S, 130"52'El, between Easter Is. (27°10'S, 109°26'W) and Darwin,
between Totegegie i23°06'S, 134°52'W) (Gambler Is.) and Darwin, between Rapa (27°37'S, 144°20' W) (Austral Is.) and Darwin, and
between Tahiti ( 17°33 'S, 149°20'W) (Society Is.) and Darwin for 1948-63. El Nino t.ype events (EN) are indicated in strong (S), moderate
(M), and weak or very weak (W) intensity.
i96a 1965 1966 1967 1968 1969 1970 1971
1972 1973 197a 1975 1976 1977
3
-
-
5
2
-
...
.
-
2
JUAN FERNANDEZ -
1
".*
•••■
-
1
DARWIN (mb)
0
-1
-2
..•"
•
0
-2
-
•..
-
2
2
1
0
r---..
...
.■
-
1
EASTER-DARWIN
(mb)
-1
-
..■
-1
-1
-2
-
—
-2
-3
-
•:.■•
■-.
....
-
-3
2
....
•
•
-
2
TOTEGEGIE -DARWIN
1
0
F •■
'••
•
"
(mb)
:
.
-2
-
■•■....
•'
...
■-...•
' ..
■•.
.•■■■•
•.•■
-
-2
2
2
1
— .■•
*..
.
'.
•
—
1
RAPA- DARWIN
0
-1
•'—'•-
." ■...
.
..
(mb)
,-.....*
••**
•'
.
_
-1
-2
:
'•..•■
■••"
"•...•■
.••
..•■•••
•. .•
—
-2
2
2
i
— ,•••
*•.
,
■..._
.
-
I
0
-.• — •
'"'■*. '.•—-.
.
•
■.
.
•
(mb)
-1
j-
",
.•■
'; '.-'
.'
—
-1
-2
-
■-•■
-
-2
-3
-
-
-3
EN(M1
EN(W)
EN(S) >
EN(W)
EN(M)
I96« 1965 1966 1967 1968 1969 1970 1971
YEARS
1972 1975 I97fl 1975
1977 1978
Figure 2b. — Triple 6-mo running mean plots of anomalies of the difference in sea level atmospheric pressure (millibars) between Juan
Fernandez Is. and Darwin, between Easter Is. and Darwin, between Totegegie and Darwin, between Rapa and Darwin, and between
Tahiti and Darwin for 1 964-76. El Nino type events (EN) are indicated in strong (S) , moderate ( M) , and weak or very weak ( W ) intensity.
666
QUINN ET AL.: SOUTHERN OSCILLATION. EL NINO, AN!) INDONESIAN DROUGHTS
IB4I
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
2
1
BOMBAY 0
(mb) .1
-2
-
-
2
-
-
-1
•2
2
1
MADRAS 0
(mb) .1
-2
1841
......■•■■
*••<
1850
1851
-
2
1
0
-1
•2
1842
1843
IS44
EN(S)-
1845
1846
1847
1848
1849
1852
2
1
BOMBAY 0
(mb) -1
-2
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
2
-
••.••
■•••■"
_-l
2
2
MADRAS 0
(mb) -1
-2
-
•■• ••
...•••■■■■
,..
■— ^._
•**r.
.••■■
■••-...-H
2
1
0
-1
■2
-
2
SANTIAGO -
0
BOMBAY ^
(mb)
■2
:
.■••■■■•
-
2
1
-
EN(S)
*■*.
-1
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
YEARS
Figure 3a. — Triple 6-mo running mean plots of sea level atmospheric pressure anomalies fmillibars) for Bombay ( 18'54'N, 72''49'E),
India ( 1847-65) and for Madras ( 1300'N, 80"11'E), India ( 1841-65); also triple 6-mo running mean plot of difference in atmospheric
pressure anomalies between Santiago ;33°27'S, 70°42'W), Chile and Bombay (1861-65). El Nino type events (EN) are indicated in
strong (S) or moderate (M) intensity.
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
2
_
2
BOMBAY 0
(mb) .,
T.
•■■
'■■-..
.—
•..
•■ ••
'■•.
..••••■■
**■•• ~
1
_
*■
..••*"
*'
*•„
-1
-2
-
2
2
_
2
1
MADRAS 0
(mb) -1
-
,•'
',
.•*
•-. -
1
_
1
'*••••
"
"•.
^.
,.••
-2
^
-2
2
1
..■••■■•••
..
•..^
2
t
•••
•,
,.•*
(mb) .1
-
••■
*•
..•••*
•1
-2
-
■2
2
1
-
■••.
..
2
SANTIAGO- 0
BOMBAY -1
*..
••••.
EN(S)-
-••—
••..
•••.
(mb) -2
-
*••••••*
*
2
-3
-
EN(M)
EN(M)
■••....•■
EN(M)
3
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1379
1880
1881
YEARS
Figure 3b. — Triple 6-mo running mean plots of sea level atmospheric pressure anomalies (millibars) for Bombay and Madras, India,
and Djakarta i06'irS, lOG'Sl'E), Indonesia (1866-81); also, triple 6-mo running mean plot of difference in atmospheric pressure
anomalies between Santiago and Bombay ( 1866-81). El Nino type events (EN) are indicated in strong (S) or moderate (M) intensity.
667
FISHERY BULLETIN; VOL. 76. NO. 3
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
2
_
-
2
BOMBAY 0
(mb) .1
-
..•■■
.■■••■•
—
_
■•*
-
-1
-2
_
-
-2
2
-
2
MADRAS 0
-
'-.
.■
'*•.. •■*'
(mb) -1
-
-1
■2
_
-2
2
2
DJAKARTA 0
-
■••.
n
'
,..-•
...•'
(mb) -1
_
--'
-2
^
--2
2
1 1
,
^2
1
-
_.'••■■.
1
DARWIN 0
•••'
(mb) -1
— ....
■1
-2
-
-2
2
-
.....
2
1
SANTIAGO -
~
0
DARWIN
■
•■....-•
.1
(mb) ^
-
EN(S)-
EN(M)-
—
EMS)
EMM)
2
1682
1883
1884
1885
1886
1887
I88S
1889
1890
1891
1892
1893
1894
1895
1896
1897
Figure 3c. — Triple 6- mo running mean plots of sea level atmospheric pressure anomalies (millibars) for Bombay, Madras, Djakarta,
and Darwin (12°26'S, 130"52'E), Australia (1882-97); also triple 6-mo running mean plot of difference in atmospheric pressure
anomalies between Santiago and Darwin ( 1882-97). El Nino type events (EN) are indicated in strong (S) or moderate (M) intensity.
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
2
1
-
•—
2
1
(mb) -1
-2
-
-2
2
MADRAS 0
(mb) -1
•2
-
2
1
0
-
2
2
DJAKARTA 0
(mb) -1
-2
-
2
1
-
-
•2
2
1
DARWIN 0
(mb) .1
•2
-
-
...•:!
2
1
0
-...••■■
...-•
-1
-2
2
SANTIAGO - 0
DARWIN -1
(mb) -2
•3
r
2
-
EN(S) —
.•■'
EN(M)
EMM)
EMS)—
2
3
1898
1899
' 1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
YEARS
Figure 3d. — Triple 6-mo running mean plots of sea level atmospheric pressure anomalies (millibars) for Bombay, Madras, Djakarta,
and Darwin ( 1898-1913); also, triple 6-mo running mean plot of difference in atmospheric pressure anomalies between Santiago and
Darwin (1898-1913). El Niifio type events (EN) are indicated in strong (S) or moderate (M) intensity.
668
ca'INN ET AL : SOUTHERN OSCILLATION. EL NINO. AND INDONESIAN DROUGHTS
1914
1915
1916
1917
1918
1919
19^0
1921
1922
1923
1924
1925
1926
1927
1928
1929
2
BOMBAY 0
(mb) -1
-2
-
*****.
•••^
-
2
1
'••.....
■■■■■••■..•
■**'
-
■2
2
MADRAS 0
(mb) .1
■2
-
".,
.«••••
•"*•.,
..••■•■
-
2
1
0
-
......
'"••.
•..••
.***••
*' '■*.
^
.1
2
2
DJAKARTA 0
(mb) -1
-2
••,.^
,
*'*'"*"
.••"*■•
—
■•••** — ^—
2
1
0
-
*'"'*'.
**••..
2
2
1
- ...■•
-••**
,.,
•-..
'**•*.
-■•■^^^
2
1
0
DARWIN 0
(mb) -1
-2
-
''
-1
-2
2
SANTIAGO-
-
..•
.■•■■••
••••■■••.
,,
....-
•..
,,•**
EN(M)-^
2
0
DARWM
(mb) 2
EN(M)
/
EN(S)—
..••"'*
■■■
•...•
EN(S)-
.•••■"
2
1914
1 1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
YEARS
Figure 3e.— Triple 6-mo running mean plots of sea level atmospheric pressure anomalies (millibars) for Bombay, Madras, Djakarta,
and Darwin ( 1914-29); also, triple 6-mo running mean plot of difference in atmospheric pressure anomalies between Santiago and
Darwin ( 1914-29). El Nino type events (EN) are indicated in strong (S) or moderate (M) intensity.
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
2
-
-
2
1
(mb) -1
•2
-
•■..••■*'*
•■'
-2
2
MADRAS 0
(mb) -1
-2
-
— ••
-
2
0
-
..•**'**•
■•■■ -^
2
2
—
-
2
0
DJAKARTA 0
(mb) -1
-2
-
' *..^ —
-
•2
1
_
••■—••■•
-
2
0
DARWIN 0
(tub) -1
■2
_
-
-2
2
SANTIAGO -
-
.••■
^
-
2
0
DARWIN
(mb) ^
EN(M)
EN(M)
EN(S)
-1
■2
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
YEARS
Figure 3f. — Triple 6-mo running mean plots of sea level atmospheric pressure anomalies (millibars) for Bombay, Madras, Djakarta,
and Darwin ( 1930-45); also, triple 6-mo running mean plot of difference in atmospheric pressure anomalies between Santiago and
Darwin < 1930-45). El Nino type events (EN) are indicated in strong (S) or moderate (M) intensity.
669
FISHERY BULLETIN: VOL. 76, NO. 3
1906
I9«7
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
z
1
BOMBAY 0
(mb) .1
-2
-
,.....:i^
-
2
1
0
-1
2
-■•••...
"••■"■
***
2
1
-
-
2
1
MADRAS 0
(mb) .1
-2
_
-
■ 1
2
2
DJAKARTA 0
(mb) -1
-2
-
-
2
-
'^
.1
2
2
DARWIN 0
(mb) .1
2
~
2
-
-
-2
2
1
SANTIAGO -
0
DARWIN
(mb) '^
-
-
2
-■• •
EN(M)
EN(S)-
:
-
-3
1946
,947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
YEARS
Figure 3g. — Triple 6-mo running mean plots of sea level atmospheric pressure anomalies (millibars) for Bombay, Madras, Djakarta,
and Darwin ( 1946-61); also, triple 6-mo running mean plot of difference in atmospheric pressure anomalies between Santiago and
Darwin (1946-61). El Nino type events (EN) are indicated in strong (S) or moderate (M) intensity.
1962
1963
1964
1965
1966
1967
1968
1969
1970
197!
1972
197 3
1974
1975
1976
1977
2
-
2
1
BOMBAY 0
..*
'"—'',—
0
-1
(mb) -1
••*'*
-2
-
2
2
1
-
2
1
MADRAS 0
(mb) .1
.,•*
=^^
0
_
'■—•■■
-2
-
2
2
-
2
DJAKARTA 0
(mb) -1
—
0
-1
■■■
■••••
•■••
■2
-
2
2
-
2
1
DARWIN 0
(mb) -1
—
.■■■■
•,
.•■'
_,
'*.
.*
0
\-
*•..•*'
*,.
.-•
■2
-
2
2
1
-
.._
.•••
.••■■"•
2
1
SANTIAGO-
0
DARWIN
■■...
.•
• EN(S)-
■• ►
(mb)
-
EN(M)'
EN(M)
2
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
YEARS
Figure 3h. — Triple 6-mo running mean plots of sea level atmospheric pressure anomalies (millibars) for Bombay, Madras, Djakarta,
and Darwin (1962-76); also, triple 6-mo running mean plot of difference in atmospheric pressure anomalies between Santiago and
Darwin (1962-76). El Niiio type events (EN) are indicated in strong (Sj or moderate (M) intensity.
670
(JUINN ET AI. : SOUTHERN OSCILLATION. KL NINO. AND INDONESIAN DROL'OHTS
affected (e.g., the 1957-58 and 1972-73 cases of re-
cent years). Moderate cases are recognized as El
Ninos by most investigators, and display typical
El Nino features to a lesser degree; maximum
monthly sea-surface temperature anomalies
along the coast usually peak in the 2.0°-3.5°C
temperature range (e.g., the 1953, 1965-66, and
1976-77 cases of recent years). The effects of a
moderate El Nino on the anchoveta fishery are
considerable, but less serious than for the strong
category.
Weak events may or may not be recognized as El
Nirios by investigators; maximum monthly sea-
surface temperature anomalies along the coast
usually peak in the 1.0°-2.5°C temperature range,
but may appear relatively late in the year ( e.g., the
1951 and 1969 cases of recent years). Very weak
events are not considered to be El Nirios;
maximum sea-surface temperature anomalies, if
they penetrate into the coast, are in the 0°-2°C
range (e.g., the 1963 and 1975 events). The weak
and very weak categories are included in this dis-
cussion because the difference between weaker
and stronger events depends not only on the
height of the preevent index anomaly peak and the
subsequent degree of relaxation reflected in the
southeast trade strength, but also on the timing of
this interannual relaxation. If the timing is in
phase with the regular annual relaxation (South-
ern Hemisphere summer and early fall), a moder-
ate or strong event is likely to occur; if they are out
of phase, a weak or very weak event is likely.
Relaxation troughs that occur near the end of the
year are usually associated with high Peruvian
coastal sea temperature anomalies in the latter
half of the year. The weak and very weak events
may not be of significance to the Peruvian an-
choveta fishery, but they do show up in the west-
ern equatorial Pacific rainfall and their larger
scale aspects may be significant from the
standpoint of associated global fluctuations. Fig-
ure 4 shows an example of how the recent events
were reflected in the Tarawa rainfall.
The weaker events were included as EN(W) in
Figure 2a, b, since we have a fairly large amount of
evidence available from 1950 on. They were not
included in Figure 3a-h due to the decreasing
availability of evidence as we reach further back in
time. However, these weaker events, ascertain-
ed to the best of our ability from availAle data
1946
19 J7
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957 1
1958
1959
i960
150
_
-
150
100
-
....
..
.".
.••..
•
'
-
100
TARAWA RAINFALL 50
_
•,
•.,.•
•..
•
-
50
ANOMALY (mm)
'.
,
*
■.
6MRM TRIPLE PASS °
*....*
■. .■
.
•....•
■•..
-50
-
.
*
...1
-50
-100
-
*,
.••"■••
...•■
-
-100
-ISO
—
•••■'
-
-150
3
3
2
—
.••■
■•.
-
2
RAPA-DARWIN PRESSURE ,
.■
.
-J
ANOMALY (mb)
.
...' '*.
6 MRM TRIPLE PASS °
•■
*.
.......
-2
-
■•...••■
-
-2
350
_
•••
-
550
300
_
,
-
300
250
_
,
-
250
200
_
.
-
200
TARAWA RAINFALL
ANOMALY (mm)
6 MRM TRIPLE PASS '°°
-
-■
■-.
.
• .
.•'-
ISO
100
50
•
■".
•'
•.
.
■ ■•. .
...
•
•
■■ ~
50
-
*.
0
-50
r"""'-
^
;
•.
....
.•
■-.
*.
\^''
!•
-SO
-100
-
••■
■•.•'
...1
Iff
•^
-
-too
2
•
RAPA-DARWIN PRESSURE
~~
■■■•••..
..*'.
.•"*.
*.
•
.•
^
ANOMALY (mb)
'~, T'
•
.
.....
•
1
,
6 MRM TRIPLE PASS -1
-
•
........
.
•
-2
-
**...*
•. .•
"....■
1961
1962
1963
1961
1965
1966
1967
1968
1969
19'0
1971
1972
1975
9^4
■975
9'6
YEARS
Figure 4. — Triple 6-mo running mean plot of anomalies of the difference in sea level atmospheric pressure (millibars) between Rapa
( 27°37 'S, 144 °20 ' W) ( Austral Is. ) and Darwin ( 1 2°26 'S, 130°52 'E), Australia compared with a similarly filtered plot of Tarawa ( 0r2 1 'N,
172°55'E) (Gilbert Is.) rainfall anomalies (millimeters).
671
FISHERY BULLETIN: VOL^ 76, NO. 3
and literature, are included in Table 1; and, the
typical index and index component trends as-
sociated with them can be noted in Figure 3a-h.
Eguiguren's (1894) data were evaluated to ex-
tend the record back even further. He classified
rainfall at Piura, Peru (lat. 5°5'S, long. 80°38'W)
into five categories: dry (0), light ( 1), moderate (2l,
good (3l, and extra ordinary (4). Considering the
distribution of events over the period 1891-1976 in
relation to Eguiguren's rainfall category distribu-
tion for 1791-1890, it appeared that we could re-
late his category 4 to a strong El Nino, category 3
to a moderate El Nino, and categories 2 and 1 to
weak and very weak events respectively. One
must realize that for this Peruvian desert area
long-term average rainfall values have little
meaning, since averages combine data from the
more-frequent drought years with data from the
smaller number of event years when significant
rainfall may occur. A year when an average
amount of rain fell is likely to have been a year
when an event occurred. Whereas the categories 3
and 4 rainfall situations were likely to have been
associated with El Nino, there is no assurance that
the categories 1 and 2 rainfalls were associated
with the oceanic events.
A study of presumed event occurrences (based
on Eguiguren's information) in relatiion to trends
of the Southern Oscillation indices and index
components for a period when overlapping data
records were available (1841-90) showed a high
degree of compatibility. Table 1 lists years in ac-
cordance with Eguiguren's rainfall classification
as well as our interpretation of event intensity
after considering his indications, the index and
index component trends, and the various data and
information sources listed early in this section.
After 1790 and prior to 1841, when no pressure
records were available for a cross-check, we avoid-
ed the weak event category, but accepted Eguigu-
ren's stronger categories 3 and 4 events. Events
occurring prior to 1791, as reported by Frijlinck
( 1925), were considered to be of the strong variety.
Sources for each event are listed in Table 1.
STATISTICAL STUDY OF
EL NINO TYPE EVENTS
From a practical standpoint, we were concerned
with the question of when the next event is likely
to occur and what its intensity might be. There-
fore, this study referred to the onset times for
separate events and the interval between onset
times. When the year immediately following the
year of onset reflected an event of equal or lower
intensity, it was assumed that the initial event
extended into this next year or that the effects of
the initial event held over into the early part of the
next year; and, the whole situation was treated as
a single event. However, when a weaker initial
event preceded a stronger event in a following
year they were treated as two separate events,
Table l. — Year of onset of El Nino type events, 1726-1976, as classified according to event intensity by Eguiguren ( 1894), left, and the
present authors, right (events below intensity 3 were not accepted prior to 1841 when pressure data became available). Numbers refer
to event intensity: 1, very weak; 2, weak; 3, moderate; and 4, strong. Asterisks indicate onset of events considered separate (see text).
Event
Key to
Event
Key to
Event
Key to
Event
Key to
Year
intensity
source
Year
intensity
source
Year
intensity source
Year
intensity
source
•1726
(3)
G
•1852
2
(2)
B, PC
•1896
(3)
PI. R
•1943
(2)
G. T, R
•1728
(4)
A,
G
1854
2
B
•1899
(4)
D. PI, R
1944
(2)
PI, T
•1763
(4)
A
•1855
(2)
PC
1900
(3)
PI, R
•1946
(1)
PI, R
■1770
(4)
A
•1857
2
(2)
B. PC
•1902
(3)
PI. R
■1948
(1)
PI,T, R
•1791
4
(4)
A,
B
1862
2
B
•1905
(3)
PI. R
■1951
(2)
G, PI,T
•1803
2
B
•1864
4
(4)
A. B, PI
•1911
(4)
F, E, PI
•1S53
(3)
G, E, T
•1804
4
(4)
B
1866
2
B
1912
(3)
F. PI
•1957
(4)
G, L, PI
•1814
4
(4)
B
■1868
1
(3)
B, C. PI
•1914
(3)
G, PI. R
1958
(4)
G, L, PI
•1817
3
(3)
B
•1871
4
(3)
A, B, C
■1917
(2)
H
•1963
(1)
PI, R
•1819
3
(3)
B
•1873
(2)
PI
■1918
(4)
C, D. PI
•1965
(3)
M, PI, T
•1821
3
(3)
8
•1875
1
(1)
B, PC
1919
(3)
D, PI, R
•1969
(2)
PI, T, R
•1824
3
(3)
B
•1877
4
(4)
A, B, PI
•1923
(2)
H. PI
■1972
(4)
N, 0, PI
•1828
4
(4)
A.
B
1878
4
(4)
A, B. PI
■1925
(4)
D. E. PI
1973
(4)
N, 0, T
1829
1
B
•1880
2
(3)
B. PI
1926
(4)
1. PI. T
■1975
(1)
P, PI, T
•1832
3
(3)
B
•1884
4
(4)
A, B, PI
•1929
(3)
G, 1. PI
■1976
(3)
Q, PI,T
•1837
3
(3)
B
1885
(3)
PI
1930
(3)
G. PI. T
•1844
3
(2)
B,
PC
•1887
2
(3)
B. PI
■1932
(2)
E, 1, J
•1845
4
(4)
B.
C. PC
1888
2
(3)
. B, PI
■1939
(3)
E, J, T
1846
2
(3)
B.
PC
1889
1
(1)
B. PI
1940
(2)
C, PI, T
•1850
2
(2)
B.
PC
•1891
(4)
D. E, PI
■1941
(4)
E, K, T
Key
Source
Key
Source
Key Sour
ce
A
Frijlinck (1925)
H
Lavalle (1917. 1924)
O Caviedes
(1975)
B
Eguiguren (1894)
1
SInepard (1930. 1933)
P Wyrtki et
al (1976)
C
Hutchinson
(1950)
J
Mears (1944)
Q Quinn (1
976)
D
Murphv (1923. 1926)
K
Lobell (1942)
R Rainfall (equatorial and;or Peruvian)
E
Sears (1954)
L
Wooster (1960)
T Sea-surfi
ice temperature off Peru
F
Forbes (1914)
M
G
jillen (1967)
PC Pressure
component of
G
Schweigger
(1961)
N
Idyll (1973)
Souttie
rn Oscillation ir
dex
PI Souttiern
Oscillation pressure index
672
gl'lNN KT AL SOUTHERN OSCILLATION, KL NINO, AND INDONESIAN DKOIIGHTS
since an additional contribution was introduced in
the following year. The foregoing assumptions
were based on findings from a study of the South-
ern Oscillation index trends and associated events
over recent decades when more data were avail-
able for case history studies.
Our study of strong events was limited to the
period 1763-present (Table 2, Figure 5), since the
break of 35 yr between 1728 and 1763 was 14 yr
longer than the longest subsequent break between
events, and there was no way of eliminating the
possibility that one or more strong events might
have gone unreported over the 35-yr gap. These
data indicate that given a strong El Niiio, there is
a 3d7c probability of having another strong event
in 7-8 yr, and an 82^7^ probability of having one
within the next 15-16 yr. Considering all available
data, the time between onsets of separate strong
events was never <7 yr.
For strong and moderate events (Table 3, Fig-
ure 5) the record was limited to the period
1791-present when data for both categories were
available. For strong, moderate, and weak events
(Table 4, Figure 5) the record was limited to
1842-present, so we would have at least one index
component trend available for cross-checking the
less prominent weak events. (Madras pressure
data became available in 1841.) With the addition
of very weak events (Table 5, Figure 5), we limited
our record to 1862-present in order to have an
0 50
3
o
>
Slrong Events
.(1763-19721
Slrong , Moderate
and Weak Events
0.50-
0 0 0
3 3
> I cvj I ^ T il ool ol *^
(NJ I ^ ! U3
fo >n r^
Strong and Moderate
(1844-19?
-
IS
9^
■
4 4
' ,.
'V
C\J
^
U3
00
Events
0 25 -
(1791-1976)
(1864-1976)
18
0 50
It
-
0 25
-
J_
4 4
0
4
,;i-
T^/il/u. 0 0 Jti,0
. . /
CM
1
in
CD
oIojI^Iu)I(dIoIc\jI
7 T T T 7 *^ f^
01 = Ki .n t cr> -
OJ
ro
ID CD 1
in ^-
Strong , Moderate, Weak
and Very Weali Events
CLASS LIMITS (years)
Figure 5. — Histograms of frequency distributions for El Nino
type events by intensity. Number of occurrences within class
intervals is indicated.
index trend available for cross-checking the more
obscure very weak events. (The Santiago-Bombay
index became available in 1861.)
Cases were noted where relaxation from a large
preevent index anomaly peak appeared to be a two
or more stage process. [This type development was
Table 2. — Strong El Ninos, with intervals between events from
onset to onset.
Years
Years
Onset
Onset
between
Onset
Onset
between
year
year
onsets
year
year
onsets
1763
1770
1791
1804
1814
1828
1845
1864
1877
1770
7
1884
1891
7
1791
21
1891
1899
8
1804
13
1899
1911
12
1814
10
1911
1918
7
1828
14
1918
1925
7
1845
17
1925
1941
16
1864
19
1941
1957
16
1877
13
1957
1972
15
1884
7
209 (cumulative years between onsets) - 17 (number of intervals) = 12.3 yr,
average time interval between onsets of strong El Ninos.
Table 3. — Strong and moderate El Nines with intervals be-
tween events from onset to onset.
Years
Years
Onset
Onset
between
Onset
Onset
between
year
year
onsets
year
year
onsets
1791
1804
13
1887
1891
4
1804
1814
10
1891
1896
5
1814
1817
3
1896
1899
3
1817
1819
2
1899
1902
3
1819
1821
2
1902
1905
3
1821
1824
3
1905
1911
6
1824
1828
4
1911
1914
3
1828
1832
4
1914
1918
4
1832
1837
5
1918
1925
7
1837
1845
8
1925
1929
4
1845
1864
19
1929
1939
10
1864
1868
4
1939
1941
2
1868
1871
3
1941
1953
12
1871
1877
6
1953
1957
4
1877
1880
3
1957
1965
8
1880
1884
4
1965
1972
7
1884
1887
3
1972
1976
4
185 (cumulat
ve years between onsets) - 34 (nu
Tiber of intervals)
- 5.4 yr.
average time
interval between onsets.
Table 4. — Strong, moderate, and weak El Ninos with intervals
between events from onset to onset.
Years
Years
Onset
Onset
between
Onset
Onset
between
year
year
onsets
year
year
onsets
1844
1845
1850
1852
1855
1857
1864
1868
1871
1873
1877
1880
1884
1887
1891
1896
1899
1902
132 (cumulat
average time
1845
1850
1852
1855
1857
1864
1868
1871
1873
1877
1880
1884
1887
1891
1896
1899
1902
1905
1
5
2
3
2
7
4
3
2
4
3
4
3
4
5
3
3
3
ive years between onsets)
interval between onsets.
1905
1911
1914
1917
1918
1923
1925
1929
1932
1939
1941
1943
1951
1953
1957
1965
1969
1972
36 (number
1911
1914
1917
1918
1923
1925
1929
1932
1939
1941
1943
1951
1953
1957
1965
1969
1972
1976
of intervals)
6
3
3
1
5
2
4
3
7
2
2
8
2
4
8
4
3
4
3.7 yr.
673
FISHERY BULLETIN: VOL. 76, NO. 3
Table 5. — Strong, moderate, weak, and very weak El Ninos
with intervals between events from onset to onset.
Onset
year
Onsel
year
Years
between
onsets
Onset
year
Onset
year
Years
between
onsets
1864
1868
1871
1873
1875
1877
1880
1884
1887
1891
1896
1899
1902
1905
1911
1914
1917
1918
112 (cumulati
average time
1868
1871
1873
1875
1877
1880
1884
1887
1891
1896
1899
1902
1905
1911
1914
1917
1918
1923
1923
1925
2
1925
1929
4
1929
1932
3
1932
1939
7
1939
1941
2
1941
1943
2
1943
1946
3
1946
1948
2
1948
1951
3
1951
1953
2
1953
1957
4
1957
1963
6
1963
1965
2
1965
1969
4
1969
1972
3
1972
1975
3
1975
1976
1
ive years between onsets)
nterval between onsets
35 (number of intervals) 3 2 yr.
first mentioned in Quinn and Zopf (1975).] In
some cases there was an initial fall from a large
preevent (primary) peak which was not fully in
phase with the seasonal relaxation (Southern
Hemisphere summer and early fall) and the result
was a relatively weak event; then, there was the
rise to a smaller secondary peak followed by relax-
ation to a secondary trough which was in phase
with the seasonal relaxation and resulted in a
stronger event. The length of time between the
two troughs was generally 18-22 mo and it is our
opinion that situations of this type may account
for many of the event-to-event intervals that fall
in the short 1-2 yr category. Examples of such
developments can be noted in 1950-53, 1962-65,
and 1973-76 (Figure 2a, b). Preevent peaks occur-
red in 1950, 1962, and late 1973-early 1974. The
first relaxation troughs following these peaks oc-
curred in late 1951, late 1963, and late 1974-early
1975, and weak or very weak events resulted in all
three cases. Then, there were rises to secondary
peaks by mid-1952, mid-1964, and late 1975, fol-
lowed by falls to troughs by early to mid-1953,
mid- 1965, and mid-1976, resulting in moderate El
Ninos for these latter years. We must be aware
that these situations can arise and should be par-
ticularly wary when a large preevent peak is fol-
lowed prematurely by a weak or very weak event.
(One must not lose sight of the fact that these
interannual fluctuations in the index anomaly
trends were used to represent the interannual
fluctuations in southeast trade and equatorial
easterly strength as affected by the Southern Os-
cillation,) Figure 6 demonstrates the similarity of
the three two-stage developments discussed
above; a particularly obvious index trend was
selected to represent each case.
The index trend between late 1872 and 1877
indicates a possible three stage development (Fig-
ure 3b), with a weak event in 1873, a very weak
event in 1875, and a strong event in 1877 (Table 1 ),
It is noteworthy that Indonesian droughts, which
are usually associated with El Nino, occurred in
1873, 1875, and 1877 (Berlage 1957),
The preevent index anomaly peak has been re-
ported to be a reliable indicator for subsequent El
Nino type activity, and our long index record sub-
stantiates this viewpoint. We compiled statistics
on the climb time from trough to peak and fall time
from peak to trough from our long index anomaly
record to provide some general guidance for event
predictions. Figure 7 shows the applicable statis-
tics. Events usually set in while the index is fall-
ing and prior to the index trough inflection point.
Therefore, the contents of Table 6 and Figure 8,
which pertain to time between index peak and
subsequent event onset, can be used to further
refine event predictions. We assumed a March
onset time for all cases in arriving at values in the
column headed "Peak to event onset" (Table 6),
This assumption was made since month of onset
was not available for most of the early cases, and a
study of recent cases showed onset times to range
from January to May,
INDONESIAN DROUGHTS
What happens over Indonesia relates to the
Southern Oscillation (Berlage 1957) and is, there-
fore, an integral part of the activity affecting the
E-D 1949
JF-D 1961
R-D 1973
3-
1950
1962
1974
1951
1963
1975
1952
1964
1976
1953
1965
1977
if) —
in
< CO
Q. LlI
_i
Q.
cr
2-
-
R-p
-
- ■■ / 1
: / 1
•■' / /
\ *•'
\ -^
V *
■ \ ^
f\
7
T J
E-D
\
YEARS
Figure 6. — Recent examples of two stage developments using
triple 6-mo running mean plots of: 1) Easter-Darwin (E-D) index
anomalies (1949-53); 2) Juan Fernandez-Darwin (JF-D) index
anomalies (1961-65); 3) Rapa-Darwin (R-D) index anomalies
(1973-77).
674
QUINN ET AL : SOUTHERN OSCILLATION, EL NINO. AND INDONESIAN DROUGHTS
equatorial Pacific and the oceanic region off
northwestern South America. In general, years
when the index is low and El Nino type activity
occurs are also years of drought in Indonesia (par-
ticularly during the east monsoon season, May-
October).
Using sea salt production on the Island of Mad-
ura (near Java), which is a very sensitive indicator
of drought and precipitation, as well as some ad-
ministration reports from Java estates, compiled
by Van Bemmelen ( 1916), Berlage ( 1957) drew up
a complete series of east monsoons drier than
normal, from 1830 to 1953. Although 937( of the
drought periods occurred during years when El
Niiio type events were under way (Table 7), only
77*^ of the periods when El Nino type events were
underway were also designated as periods of east
monsoon drought (Table 8). Nevertheless, the as-
sociation between occurrences of these two
FIGURE 7. -Frequency distributions of rise time from trough to phenomena and changes in the Southern Oscilla-
pealt and fall time from peak to trough lin months) for the triple i- • j i. j i u- -iu *-
^ . , r o .u /-> 11 ^- ■ J tion mdex trends are close enough in either case to
6-mo running mean plots of Southern Osculation index "
anomalies (see text). The number of cases falling within a class indicate common relationships With the large-
interval is entered at the top of the relevant histogram element, scale ocean-atmosphere changes over the Indo-
0 40
_ Pea* to Trough
13
(1862-19761
0 30
-
10
a
0 20
^
0 10
,
}
>.
.-.
3- 0
^0
(\j
m
^1 o
" '^ tr: 2) ;, A
> 0 40
Trough to Peak
o
^^ (IRfi4-iq7'i>
^0 30
3
020
—
"
*;
0 10
0
\—
J
Ifl
t\i
C
las
> L
imi
S (
mo
iths
)
Table 6.— Time (in months) between index peak and El Nino type event onset (assuming onset is in March
of indicated year), and time between index peak and associated Indonesian drought onset (assuming onset
is in May of associated year). Pressure indices (see text) used to determine time of preevent peak were: S-B,
Santiago-Bombay; S-D, Santiago-Darwin; JF-D, Juan Fernandez- Darwin; E-D, Easter-Darwin; T-D,
Totegegie-Darwin; and R-D, Rapa-Dai-win. In last column ND indicates no associated drought.
El Nino type event
Preevent index peak
Peak to event onset
No. of Monttis
Peak to drougfit onset
Year of onset
Month of peak
Index used
No. of months
1864
Aug -Sept 1862
S-B
18.5
20.5
1868
Mar-Apr 1867
S-B
115
ND
1871
Jan -Feb 1870
S-B
13.5
ND
1873
Jan 1873
S-B
20
4.0
1875
Aug -Sept. 1874
S-B
6.5
8.5
1877
Feb. 1876
S-B
13.0
15.0
1880
Nov 1878
S-B
16.0
30.0
1884
May 1882
S-B
22.0
12.0
1887
July-Aug 1886
S-B
7.5
21.5
1891
Nov.-Dec,1889
S-B
15,5
17.5
1896
Oct 1893
S-D
29 0
31.0
1899
Nov, -Dec. 1897
S-D
155
ND
1902
May 1901
S-D
10.0
12.0
1905
Sept.-Oct 1903
S-D
145
16.5
1911
Jan -Feb. 1910
S-D
13 5
ND
1914
Sept -Oct 1912
JF-D
175
7.5
1917
Aug. -Sept 1916
S-B
6.5
ND
1918
Aug. -Sept. 1917
S-D
65
8.5
1923
Aug. 1921
JF-D
190
21.0
1925
May-June1924
S-B
95
11.5
1929
Dec -Jan 1928 29
JF-D
25
4.5
1932
Aug -Sept 1931
S-D
6.5
8.5
1939
June-July 1938
JF-D
8.5
22.5
1941
Jan -Feb. 1940
JF-D
13.5
15.5
1943
July-Aug 1942
JF-D
7,5
21.5
1946
Dec-Jan. 1944 45
S-B
14,5
4.5
1948
Aug. 1947
JF-D
7.0
ND
1951
Apr-May 1950
E-D
105
ND
1953
Apr -May 1952
R-D
105
12.5
1957
July-Aug 1955
E-D
195
t
1963
May 1962
JF-D
10.0
1965
June 1964
JF-D
9.0
Data not
1969
Mar -Apr. 1967
T-D
23.5
available
1972
Oct -Nov. 1970
R-D
16.5
1
1975
Nov -Dec. 1973
R-D
15.5
1976
Sept.-Oct. 1975
R-D
5.5
7.0
675
0 40
0 30
Peak to Event Onset
15 (1862 -1976 >
o 1 ig I
FISHERY BULLETIN: VOL. 76, NO. 3
Pacific region. Based on the association between
El Nino type events, drought years and index fea-
tures, and also an assumption that the drought
will set in during May of involved drought years,
we arrived at values in the column headed "Peak
to drought onset" (Table 6). Figure 8 shows the
resulting statistics which could be applied to In-
donesian drought predictions.
040
~ Peak to Drougtit Onset
0 30
(1862-1953 a 1976)
f^J fO fO
Figure 8. — Frequency distributions of time (in months) be-
tween preevent peaks in triple 6-mo running mean plots of
Southern Oscillation index anomalies (see text) and: 1) the onset
of subsequent El Nino type events (assuming onset is in March of
involved years); 2) the onset of associated Indonesian droughts
(assuming onset is in May of involved years). The number of
cases falling within a class interval is entered at the top of the
relevant histogram event.
Class Limits (months)
Table 7. — Association of east monsoon droughts in Java with El Niiio type events.
Drought
years
El Nino type
event years
Notes
Drougtit
years
El Nino type
event years
Notes
1844
1844
1913»
1914(
1845
1845-46
1850
1850
1918»^
1853
None
Event in 1852
1919'
1855
1855
1923
1857
1857
1925\
1926/
1864
1864
1873
1873
1929
1875
1875
1932
1877
1877-78
1935
1881
1880
Index low 1880-81
1940
1883
1941
1884)
1884-85
1944
1885
1945*
1888
1887-89
1946/
1891
1891
1953
1896
1896
Drougt
1902
1902
1976
1905
1905
28 (separate events) - 30 (east
monsoon drought situations) =
= 0,93,
1914
1918-19
1923
1925-26
1929-30
1932
None
1939-40
1941
1943-44
1946
1953
Slight lowering of index
Drought data unavailable 1954-75
1976
93% of east monsoon droughts can be associated with El Nino type events
Table 8.— Association of El Nino type events with east monsoon droughts in Java.
El Nitio type
event years
Drought
years
Notes
El Nino type
event years
Drought
years
1844
1845-46
1850
1852
1855
1857
1864
1868
1871
1873
1875
1877-78
1880
1884-85
1887-89
1891
1896
1899-1900
1902
1844
1845
1850
None
1855
1857
1864
None
None
1873
1875
1877
1881
1883-85
1888
1891
1896
None
1902
Drought in 1853
Index low 1880-81
1905
1905
1911-12
None
1914
1913-14
1917
None
1918-19
1918-19
1923
1923
1925-26
1925-26
1929-30
1929
1932
1932
1939-40
1940
1941
1941
1943-44
1944
1946
1945-46
1948
None
1951
None
1953
1953
Drought data
unavailable 1954-75
1976
1976
28 (east monsoon drought situations) - 36 (separate events) '^ 0 78,
78°o of El Nirio type events can be associated with east monsoon droughts.
676
QUINN ET AL : SOUTHERN OSCILLATION. EL NINO, AND INDONESIAN DROUGHTS
DISCUSSION
Over the past 1 16 yi- ( 1861-1976), for which we
have adequate data on the occurrence of El Nino
type events of weaker intensity, there were de-
cades of minimal activity (e.g.. 1901-10, 1931-40,
1961-70), but no decade without such activity.
There is no reason to expect any significant change
in the amount of El Nino activity in the foresee-
able future over that experienced in the past cen-
tury. Therefore, it would appear that our data
(e.g.. Tables 2-5) might eventually be used in con-
junction with associated catch data and biological
findings for effective long-range planning in the
management of the anchoveta fishery. For exam-
ple, assessment of maximum sustainable yields
under various environmental conditions ranging
from the favorable extended anti-El Nino condi-
tion (when there are two or more consecutive years
with high Southern Oscillation indices) to the El
Nino situation (when there are rapidly falling and
low Southern Oscillation indices) might prove use-
ful for determining the optimum size and flexibil-
ity of the fishing fleet and fish processing facilities.
A key element to such assessments will be a know-
ledge of the required biological recuperation time
following cessation of an unfavorable physical en-
vironmental condition.
Such data could also be used for speculative
long-range outlooks. For example, if we had just
experienced a strong El Nino, our results suggest
that there is a near-zero probability that we would
experience another strong event in < 7 yr after the
onset of the recent situation. However, there
would be an 86Vir probability that an event in the
very weak, weak, or moderate category would
occur within 3-4 yr after the strong El Nino onset.
Considering the current situation, and recogniz-
ing that a moderate event set in during 1976 and
held over into early 1977, there is a dAVc probabil-
ity (based on our data) that another event of un-
known intensity would set in during 1980. It
would not be reasonable to go beyond statistical
estimates until we find we are approaching a peak
in the Southern Oscillation index anomalies.
When we are nearing a preevent peak and can
assess its height and time of occurrence, then we
can use peak to trough statistics (e.g., Figure 7) to
advantage in forecasting onset time and likely
intensity of the coming El Nino type event. The
intensity would be based on the height of the index
anomaly peak and the time of year when the sub-
sequent trough was expected to occur. Event onset
time can be further refined by considering Figure
8 statistics. It is also essential in the prediction
procedure to realize that some developments may
involve two or more stages. In cases of this type,
forecast lead times for the separate stages will
often be greatly reduced (to 1-6 mo in advance),
unless historical analogies lead to pattern recog-
nition as the situation evolves.
ACKNOWLEDGMENTS
We thank the Chief of the Naval Weather Ser-
vice and the Director of the Hydrographic Insti-
tute of the Armada de Chile; the Director of the
Civil Aviation Service and Chief of the
Meteorological Service of Polynesie, Francaise;
the Director of the Meteorological and Geophysi-
cal Institute, Djakarta, Indonesia; the President of
the Instituto del Mar del Peru; Ramon Mugica,
Cniversidad de Piura, Peru; the Director of the
Australian Bureau of Meteorology; and, the Na-
tional Climatic Center, Environmental Data Ser-
vice, NOAA, for their support of this study. We are
indebted to Forrest R. Miller of the Inter-
American Tropical Tuna Commission and Richard
Evans of the Southwest Fisheries Center, Na-
tional Marine Fisheries Service, NOAA, for their
timely information on sea temperatures and
weather conditions over the eastern tropical
Pacific. We also thank Clayton Creech for his sup-
port in data processing. Support by the National
Science Foundation under the North Pacific Ex-
periment of the International Decade of Ocean
Exploration through NSF Grant No. OCE 75-
21907 AOl, and under the Climate Dynamics
Program of the Division of Atmospheric Sciences
through NSF Grant No. ATM77-00870 is grate-
fully acknowledged.
LITERATURE CITED
Berlage, H. p.
1957. Fluctuations of the general atmospheric circulation
of more than one year, their nature and prognostic val-
ue. K. Ned. Meteorol. Inst., Meded. Verh. 69, 152 p.
1966. The Southern Oscillation and world weather. K.
Ned. Meteorol. Inst., Meded. Verh. 88, 152 p.
BJERKNES. J.
1969. Atmospheric teleconnections from the equatorial
Pacific. Men. Weather Rev. 97:163-172.
Caviedes. C. N.
1975. El Nino 1972: Its climatic, ecological, human, and
economic implications. Geogr. Rev. 65:493-509.
Clayton. H. H.
1927. World weather records. Smithson. Inst. Misc. Col-
lect. 79, 1199 p.
677
FISHERY BULLETIN: VOL 76. NO. 3
1934. World weather records. 1921-1930. Smithson.
Inst. Misc. Collect. 90, 616 p.
Ci-.-xYToN. H. H.. .•wn F. L. Cl.wton
1947. World weather records, 1931-1940. Smithson.
Inst. Misc. Collect. 105, 646 p.
ECUKU'REN. D. V.
1894. Las Nuvias de Piura. Bol. Soc. Geogr. Lima 4(7-
9):241-258.
FoKBE.S. H. O.
1914. Notes on Molina's pe\ican< Pelecanus thagus). Ibis,
Ser. 10, 2:403-420.
FRI.ILI.NCK. C. p. M.
1925. Bijdrage tot het probleem der Klimatwis-
selingen. Nat. Utr. 45:372-374.
Guillen. O.
1967. Anomalies in the waters off the Peruvian coast dur-
ing March and April 1965. Stud. Trop. Oceanogr.
(Miami) 5:452-465.
HUTCHINSON. G. E.
1950. Survey of existing knowledge of biogeochemistry 3.
The biogeochemistry of vertebrate excretion. Bull Am.
Mus. Nat. Hist. 96, 554 p.
Idyll, C. P.
1973. The anchovy crisis. Sci. Am. 228(6):22-29.
Johnson, J. H„ and G. R. Seckel
1977. Use of marine meteorological observations in fishery
research and management. U.S. Dep. Commer., NOAA,
Environ. Data Serv., p. 3-12.
L.av.alle. Y Garcia. J. A., De.
1917. Informe preliminar sobre la causa de la mortalidad
anormal de las aves ocurida en el mes de marzo del pre-
sente ano. Mem. Cia. Adm. Guano, Lima 8.
1924. Estudio de la emegracion y mortalidad de las aves
guaneras. Mem. Cia. Adm. Guano, Lima 15.
LOBELL, M. G.
1942. Some observations on the Peruvian coastal cur-
rent. Trans. Am. Geophys. Union 23:332-336.
MEARS. E. G.
1944. The ocean current called "The Child." Annu. Rep.
Smithson. Inst., 1943, p. 245-251.
MILLER, F. R., AND R. M. LAURS.
1975. The El Nino of 1972-73 in the eastern tropical Pacific
Ocean. Inter-Am. Trop. Tuna Comm. Bull. 16:403-448.
Murphy, r. C.
1923. The oceanography of the Peruvian littoral with ref-
erence to the abundance and distribution of marine
life. Geogr. Rev. 13:64-85.
1926. Oceanic and climatic phenomena along the west
coast of South America during 1925. Geogr. Rev. 16:26-
54.
National Marine Fisheries Service.
1977. Industrial fishery products, market review and out-
look, June 1977. U.S. Dep. Commer.. NOAA, Natl. Mar.
Fish. Serv., Curr. Econ. Anal. 1-29, 27 p.
Panofsky. H. a., and G. W. Brier.
1965. Some applications of statistics to meteorology. Pa.
State Univ. Press, University Park, 224 p.
QUINN, W. H.
1971. Late Quaternary meteorological and oceanographic
developments in the equatorial Pacific. Nature (Lond.)
229:330-331.
1974. Monitoring and predicting El Nifio invasions. J.
Appl. Meteorol. 13:825-830.
1976. Use of Southern Oscillation indices to assess the
physical environment of certain tropical Pacific fisheries.
In Proceedings of the NMFS/EDS Workshop on Climate
and Fisheries, Columbia, Mo., April 26-29, 1976, p. 50-
70. U.S. Dep. Commer., NOAA, Natl. Mar. Fish. Serv./
Environ. Data Serv.
QUlNN, W. H., AND D. O. ZOPF
1975. The Southern Oscillation, equatorial Pacific
anomalies and El Nino. Geofis. Int. 15:327-353.
SCHWEICCER, E. H,
1961. Temperature anomalies in the eastern Pacific and
their forecasting. Soc. Geogr. Lima. Bol. 78:3-50.
Sears. M.
1954. Notes on the Peruvian coastal current. 1. An intro-
duction to the ecology of Pisco Bay. Deep-Sea Res.
1:141-169.
SHEPARD. G.
1930. Notes on the climate and physiography of south-
western Ecuador. Geogr. Rev. 20:445-453.
1933. The rainy season of 1932 in southwestern
Ecuador. Geogr. Rev. 23:210-216.
Troup, A. J.
1965. The 'southern oscillation." Q. J. R. Meteor. Soc.
91:490-506.
U.S. DEPARTMENT OF COMMERCE.
1959. World weather records, 1941-1950. Wash., D.C.,
1361 p.
1961-76. Monthly climatic data for the world. Vol. 14-
29. Wash., D.C., var. pag.
1966-68. World weather records, 1951-1960, Vol. 3, 4, and
6. Wash., D.C., var. pag.
Van Bemmelen. W.
1916. Droogte-jaren op Java. Nat. Tijdschr. Ned. -Indie
75:157.
Walker. G. T.
1924. World Weather II. Mem. India Meteor. Dep.
24:275-332.
WoOSTER, W. S.
1960. El Nino. Calif Coop. Oceanic Fish. Invest. Rep.
7:43-45.
Wyrtki. K., E. Stroup. W. Patzert, R. Williams, and W.
QUINN.
1976. Predicting and observing El Nino. Science (Wash.,
D.C.) 191:343-346.
678
THE PRECISION OF SIMULATED TRANSECT SURVEYS OF
NORTHERN ANCHOVY, ENGRAULIS MORDAX, SCHOOL GROUPS
Paul C. Fiedler'
ABSTRACT
Simulated transect surveys of model anchovy populations were compared in terms of precision and
efficiency. The precision of systematic surveys varies inversely with the distance between transects.
Systematic surveys give more precise population estimates than random surveys, due to the large
positive correlation between closely spaced transects. The precision of stratified systematic surveys is
not significantly different from that of the unstratified surveys when the school groups are randomly
distributed in the survey area. However, stratified systematic surveys are more precise when the school
groups are clumped in one end of the survey area. The results of the simulations show that the patchy
distribution of anchovy schools can be a major source of error in population estimates.
Any sampling program intended to estimate the
size of a population is subject to a variety of errors
which may reduce the accuracy or precision of the
estimate. Precision is the reciprocal of the varia-
tion of replicate estimates. Successful manage-
ment of a northern anchovy fishery in California
will require the monitoring of changes in the popu-
lation size. Acoustic survey techniques are cur-
rently being developed to obtain population and
biomass estimates independent of the fishery
(Hewitt et al. 1976). As in the study of any biologi-
cal population, it will be important to avoid con-
fusing the variation of a series of estimates due to
sampling error with true fluctuations in the popu-
lation size.
Precision may be affected by 1) the manner in
which the sampled population is distributed in
space, and 2) variations within the sampling
method itself Several studies have shown that the
patchy distribution of individuals in a population
may cause considerable variation in replicate
population estimates and that the variation is re-
lated to sample design. Winsor and Clarke ( 1940)
studied the variation of catches in series of
plankton net tows. Although they did not separate
the components of between-tow variation due to
factors (1) and (2), it was observed that oblique
tows were more precise than vertical or horizontal
tows. Barnes and Marshall (1951) took an exten-
sive series of replicate pump samples and attri-
buted the considerable variation observed to the
■Scripps Institution of Oceanography, Universitv of Califor-
nia, San Diego, La Jolla, CA 92093.
nonrandom distribution of the zooplankton since
the volumes filtered were known accurately. Taft
( 1960) analyzed the variance of sardine egg counts
in a grid of closely spaced stations. The distribu-
tion of eggs was extremely patchy (the densities
between samples ranged over more than four or-
ders of magnitude) and the relative 95^}^ con-
fidence limits for an estimate of the egg population
in the area of the grid from a single sample were
represented by a factor of 62. A simulation study
by Wiebe (1971) showed that the precision of zoo-
plankton population estimates depends both on
the sampling design (net size and tow length) and
the distribution of the population (size and loca-
tion of patches).
Similar studies have investigated the precision
of sampling fish populations. Taylor ( 1953) discuss-
ed the implications of the patchy distribution of
fish for the optimum design of trawl surveys to
estimate population size. Cram and Hampton
( 1976) demonstrated that the patchy distribution
of pilchard schools can cause imprecision suf-
ficient to render a population estimate useless for
management.
The anchovy population is patchy on two levels:
individual fish are aggregated in schools and
schools themselves tend to be aggregated in school
groups. This patchiness, or nonrandomness, is ex-
pected to be a major source of variation in popula-
tion estimates. The present study simulated sur-
veys of model anchovy populations to determine
the effect of patchiness on the precision of popula-
tion estimates. Three transect survey designs
were compared: systematic, random, and strat-
ified systematic. These are merely different
Manu.-icnpt accepted .January 1978.
FISHERY BULLETIN: VOL. 76. NO. 3. 1978.
679
FISHERY BULLETIN: VOL. 76. NO. 3
methods of selecting transects, or allocating sam-
pling effort. The three types of simulated survey,
with a range of sample sizes (numbers of tran-
sects), were run on 15 model anchovy populations.
METHODS
Anchovy populations were modeled as arrays
with each element representing 1 n.mi.-. The
ari-ay dimensions were 180 x 75, approximately
the dimensions, in miles, of the Los Angeles Bight.
Since a school is the population unit detected in an
acoustic or aerial survey, the units of the model
populations were schools. One hundred fifty
thousand schools were distributed in the array
resulting in a mean density of 11.1 schools mi^.
Four acoustic surveys by the California Depart-
ment of Fish and Game- in 1975 and 1976 yielded
estimates ranging from 88,887 to 319,878 an-
chovy schools off southern California in the area
of the bight. Mais (1974) gave a range of 21,920-
343,070 ix = 150,996) schools off southern
California and northern Baja California, most of
which were within the bight.
The schools were placed in circular school
groups located at random. Schools were distrib-
uted uniformly within a school group. School
group radii and densities were chosen randomly
and independently from log-normal approxima-
tions of observed frequency distributions based on
52 school groups from six California Department
of Fish and Game Sea Survey acoustic surveys
(MacCall et al.-M (Figure 1). There was no sig-
nificant correlation between the density of schools
within a school group and the size of the school
group in these observations. Where school groups
overlapped, the densities were simply added to-
gether, although this effectively increased both
the mean radius and density. In one model popula-
tion illustrated in Figure 2, 16 school groups con-
taining 150,303 schools covered about 149^ of the
survey area. Fifteen model populations were used,
each with the same total number of schools, but
different locations, sizes, and densities of school
groups.
^S.J.Crooke. 1975. Cruise reports 75- A- 1 and 75- A-6. K.
F. Mais. 1976. Cruise reports 76-A-3 and 76-A-9. State of
California - The Resources Agency, Department of Fish and
Game, Marine Resources Region, Long Beach, CA 90802.
■■'MacCall, A., P. E. Smith, G. Stauffer, J. Squire, J. Zweifel,
and S. Crooke. Report of CalCOFI anchovy workshop working
group on methods of estimating anchovy abundance. Unpubl.
manuscr. Southwest Fisheries Center, National Marine
Fisheries Service, NOAA, La Jolla, CA 92038.
>-
o
5 030
o
Ul
tr
-
/
/
/
/
^ 020
/
x
>
/
\
5 QIO
UJ
q:
■>..
.
-
^
^
0
"*
T
r
1
1
1
2.2 2.6 3.0 3.4 38 4.2 4.6 5.0 54
Lege DENSITY (schools /miles^)
0.40
V
o
/
\
S O30
3
-
/
/
/
/
\
\
\
o
Ul
tr
\
\
^ Q20
-
/
/
\
Ul
>
.
/
\
\
h-
/
j 0.10
Ul
cc
-
/
'
\
/
\
/
\
N
^
0
r
I 1 1 I I I
-2.0
-1.0
0
1.0
2.0
3.C
4.0
5.0
Log e RADIUS (miles)
Figure l. — Comparison of distributions of northern anchovy
school group density and size observed in the California Current
during California Department Fish and Game surveys (solid
lines) to log-normal approximations used in simulations (dashed
lines).
A simulated survey consisted of a series of
transects across the survey area. There were 180
possible transects, each 1 mi wide. Acoustic sur-
veys currently run by the Southwest Fisheries
Center, National Marine Fisheries Service,
NOAA, used a transect width of 0.14 mi (250 m).
Aerial transect widths were typically 0.2 to 0.5 mi.
A larger transect width was used in the simula-
tions to hold the model population array down to a
reasonable size. We assumed that the general re-
sults of the simulations would not change by using
a smaller transect width. Since all schools were
coimted within a transect, the only source of error
in the survey estimate was the large variance in
the number of schools per transect. For instance,
in the model population in Figure 2, the mean
number of schools per transect was 835.0, while
SD was 920.4 (variance = 8.47 xlO^).
Systematic surveys were simulated by counting
the schools within a series of transects separated
by a constant transect interval. A population es-
timate was calculated simply by dividing the
survey count by the fraction of the survey area
covered by the transects. Transect intervals of 2, 3,
680
FIEDLER f'KECISION OF SIMl'LATEn TRANSECT Sl'RVFYS
N
180
Figure 2. — A model northem'anchovy population. Densities of
school groups in schools per square mile. Simulated survey
transects are oriented horizontally. The numbers on the axes are
the coordinates of the array and the dimensions, in miles, of the
survey area it represents.
4. 5. 7, 10, 12, 15, 20, 25, 30, and 40 mi were used. A
survey with a transect interval of d miles con-
sisted of 180 c/ transects. For each transect inter-
val, 20 replicate surveys were run by randomly
choosing the initial transect from the first d tran-
sects in the survey area. The replicate survey esti-
mates were used to calculate an unbiased mean
population estimate and a coefficient of variation
(standard deviation of the replicate estimates di-
vided by the mean), which is a measure of the
precision of the estimate (Wiebe 1971). This pro-
cedure was repeated on the 15 different model
populations.
Random surveys were simluated in an analo-
gous manner to allow a direct comparison of sam-
pling errors. For each of the transect intervals (c/i
of the systematic surveys, 20 replicate surveys
were run consisting of 180 (/ transects chosen at
random without replacement la transect was not
repeated within a survey). Coefficients of varia-
tion were calculated as a measure of precision.
Stratified systematic surveys were simulated
after dividing each of the model populations into
four 45-mi wide strata. The schools along three
transects in each of the four strata were counted to
obtain a preliminary estimate of relative popula-
tion sizes. Then a total of 60, 36, 18, 12, 9, 7, or 6
transects were divided among the strata according
to the estimated population fractions. For exam-
ple, if a stratum contained one-half of the schools
counted in the preliminary survey, one-half of the
total number of transects was allocated to that
stratum for the stratified survey. At least one
transect was allocated to each stratum to avoid
biasing the final population estimate. For each
total number of transects, 20 replicate systematic
surveys were run by randomly choosing the initial
transect and simulating a systematic survey
within each stratum with the allocated number of
transects. Once again, coefficients of variation of
the replicate population estimates were calculated
for each of the 15 model populations.
RESULTS
The results of the systematic survey simula-
tions indicate that the sampling error represented
e 5 F,
IT o q:
09
000
O ^ CO
CO o
o o
S 9 i"
o f^ —
iC o
o o
O O o
8°
c\j o
1 I 11 — 1 1 — I i I I 1^
2 34 5 7 10 12 15 20 25 30
TRANSECT INTERVAL-MILES
40
Figure 3. — Results of the simulations of systematic surveys of
the 15 models for northern anchovy populations. Relative ef-
ficiency is proportional to precision (the reciprocal of the
coefficient of variation) divided by relative cost (see text). Aver-
ages and 95% confidence limits are given for coefficients of varia-
tion.
681
FISHERY BULLETIN; VOL. 76. NO. 3
by the coefficient of variation increased as the
transect interval increased and sample size de-
creased (Figure 3). The cost of a survey was assum-
ed to be proportional to the distance covered along
the transects plus 360 mi to and from port. Relative
efficiency was 10-' times the reciprocal of the pro-
duct of the coefficient of variation (C.V.) and rela-
tive cost, i.e., precision divided by cost. Efficiency
generally decreased as the transect interval in-
creased, but peak efficiency was obtained at a
transect interval of 3 mi. By interpolation, it can
be seen that a population estimate may range 10
and 2U7c (2 x C.V.) from the true population size
when surveys are run with transect intervals of
8.5 and 16 mi. respectively.
Systematic sampling gave a consistently lower
coefficient of variation, or greater precision than
random sampling (Figure 4). The variability be-
tween model populations, indicated by the
confidence limits on the mean coefficient of varia-
tion, was greater for the random sampling error
(F,,^i^, ^5.44,P<0.05 12)for8ofthe 12 sample
sizes. Also represented in Figure 4 are the ex-
pected coefficients of variation for random sampl-
ing calculated from the model population
parameters (o-^ and yu.) by the following equation
with a finite population correction:
C.V. = \k- (^^)
[i^ n \ N /
where a- - the average variance of the number of
schools per transect in the 15 model
populations =1,154,636
At = the mean number of schools per tran-
sect = 835.3
n = the number of transects in the survey
N = total number of transects in the survey
area = 180.
0.5C
0.20
STRATIFIED
SYSTEMATIC
1 I M [—
2 34 5 7
90 45 26
60 36
10 12
18 15
20
9
25
30
6
— I 1 — TronseCt
40 Interval-miles
5 Number of Transects
Figure 4. — Comparison of the results of the simulations of
random, systematic, and stratified systematic surveys. For ran-
dom surveys, the curve represents the expected coefficients of
variation calculated from the parametric variance of the model
populations (see text).
For 1 1 of the 12 sample sizes, the expected value
was within 95''^ confidence limits of the mean ob-
served coefficient of variation. This close agree-
ment supports the validity of the method used to
obtain the coefficients of variation in the simula-
tions. There was apparently no significant differ-
ence between the coefficients of variation of the
systematic and stratified systematic surveys
(Figure 4). This was confirmed by analysis of var-
iance (Table 1,P>0.25 that there was no added
variance due to survey design). However, there
were significant interaction effects between sur-
vey design and model population (P<0.01 that
there was no added variance from this source) and
between survey design and the number of trans-
ects (P<0.05).
An attempt was made to elucidate the interac-
tions involving survey design by performing
analyses of variance on subsets of the data. It was
found that for large sample sizes (transect interval
^15 mi, or number of transects ^12) unstratified
Table l. — Analysis of variance of the coefficients of variation from simulated systematic and
stratified systematic surveys of model northern anchovy populations. This is a mixed model
analysis of variance (Sokal and Rohlf 1969): survey design and number of transects are fixed
treatment effects and model population is a random effect.
Source of variation
SS
df
MS
Significance level
Ivlain effects;
Sun/ey design
Number of transects
Ivlodel population
interactions;
Design-number of transects
Design-population
Number of transects-population
Error
Total
0.0157
1
00157
1 2872
6
0,2145
05270
14
00376
0 0748
6
00125
02418
14
00173
03179
84
00038
0.3988
84
0.0047
2.8633
209
(6 841 =
' M4S4)
, = 0.908. P -0 25
56.681. P<<0.001
= 8 000. P- 0.001
F (6,MI
F 114 84)
'164 84
= 2.627. P
= 3.681. P-:
= 0 819. P
;0-05
0,001
■0 50
682
FIEDLER: F^RECISION OF SIMll.ATED TKANSECT SIRVEYS
systematic surveys were significantly more pre-
cise than stratified systematic surveys iP<0.025).
although there is still a significant interaction
between survey design and model population
(P<0.001). For smaller sample sizes, there was no
significant difference between the precision of the
two designs.
In the model populations, school groups were
located randomly within the survey area. How-
ever, the distribution of schools between strata
was never random because of the wide range of
school group sizes and the small number of school
groups in a population. The 15 model populations
were divided into three groups (low, intermediate,
and high nonrandomness) based on the index of
dispersion of the number of schools per stratum.
Analysis of variance revealed that for highly non-
random populations, stratified systematic surveys
were significantly more precise than unstratified
surveys (P<0.025). On the other hand, there was
no significant difference between the survey de-
signs for populations of intermediate or low non-
randomness. The effect of the nonrandomness of
the populations, in the limited sense used here, is
illustrated more dramatically below.
In summary, these results indicate that both the
number of transects and the spatial distribution of
the population can affect the precision of a survey
estimate. The effect of survey design involves
complex interactions with the other two factors.
These factors should be considered, if possible,
when choosing the optimum design for a survey.
DISCUSSION
In general, systematic sampling may result in
considerable gains or losses in precision compared
with simple random sampling. The greatest in-
crease in precision occurs when there is a high
degree of correlation between adjacent sampling
units and the correlation decreases as the interval
between units increases. In this situation, sys-
tematic sampling resembles stratified sampling.
On the other hand, precision may be greatly re-
duced when there is a periodic variation in the
population and the sampling interval is equal to
this period or a multiple of it (Hansen et al. 1953).
Correlograms between sampling units (tran-
sects) in five of the model anchovy populations indi-
cated that transects <10 mi apart had a high posi-
tive correlation, while the correlation tended to be
slightly negative at distances >20 mi (Figure 5).
This autocorrelation structure was due to the fre-
FlGURE 5. — Autocorrelation of transect counts in five mode!
northern anchovy populations.
quency distribution of school group sizes. The
mean distance at which the autocorrelation func-
tion passed through zero was 15.0 mi, while the
mean diameter of the individual school groups in
the five model populations was 11.8 mi. Distribu-
tion of school groups within the model populations
was random. However, real populations are likely
to be nonrandom in this respect and additional
correlations would be expected from this factor.
The strong positive correlation between transects
separated by short distances explains why sys-
tematic surveys with small transect intervals
were more precise than random surveys with an
equivalent number of transects. As the transect
interval increased, the correlation between tran-
sects decreased to near zero and the imprecision of
systematic sampling approached that of random
sampling (Figure 4).
In order to reduce total sampling error, a com-
mon strategy is to allocate effort proportional to
the sampling error within parts of a sampling
program. The variation observed in the population
estimates of the simulated surveys was caused by
the large variance in the number of schools per
transect. It can be shown in the model populations,
as in many biological populations, that the stan-
dard deviation was positively correlated with the
mean number of schools per transect in a stratum.
Therefore, it was thought that the stratified sys-
tematic surveys would reduce the total sampling
error by allocating more transects where the var-
iance was large. The simulations failed to show
any gains in precision from this strategy. This
result was not expected, but is possibly due to the
random distribution of school groups. The model
populations may have been ideal in this sense, but
we had relatively little information on the dis-
tribution of school groups within the range of the
683
FISHKRY Hl'ia.KTIN: Vol, 7(S, NO. 3
northern anchovy. As slated above, stratified sys-
tematic sui'veys were significantly more precise
than unstratified surveys for the five model popu-
lations with the most nonrandom distribution of
schools between strata. If the school groups them-
selves are aggregated, it is reasonable to expect an
increase in precision by stratifying the survey.
To test this possibility, the simulations were
repeated on model populations in which the school
groups were limited to only one-half of the survey
area. An analysis of variance (Table 2) indicated
in this case that the stratified systematic surveys
were more precise than the unstratified systema-
tic surveys (P<0.005 that there was no added var-
iance due to survey design). The overall mean
coefficients of variation were 0.095 and 0.133, re-
spectively. However, there were significant in-
teraction effects involving survey design, indicat-
ing that the advantage of stratifying the survey
will depend on the number of transects and the
spatial distribution of the population. The addi-
tional cost of the preliminary survey in the strat-
ified design must also be considered when compar-
ing it with the unstratified design.
The results of the simulated systematic surveys
showed that the patchy distribution of schools was
an important source of error in estimates of the
anchovy population size. Acoustic surveys run by
the Southwest Fisheries Center have used tran-
sect intervals of 6.6 and 40 mi. The simulations
gave evidence that the population estimates from
these surveys could be expected to range at least 8
and 90''^ (2 x C.V.), respectively, from the true
population size. The most efficient simulated sam-
pling, in terms of precision per unit cost, occurred
at a transect interval of 3 mi. This would require a
cruise grid of 4,860 mi, equivalent to a 34-day
acoustic survey at 12 kn and 12 h per day, to
reduce the coefficient of variation (due to the
patchy distribution of schools) to 1.4^7^ . Maximiz-
ing efficiency is not a valid goal, however, when
the precision gained is greater than that required
for the problem of managing the fishery, when
other sources of error become more important, and
when there are absolute limits on cost. Anchovy
population estimates within 1W( of the true value
might be considered sufficient for management, at
least to allow confidence that a consistent change
observed over several years is real (pers. commun.,
P. E. Smith, Southwest Fisheries Center, National
Marine Fisheries Service, NOAA, La Jolla, Calif.,
Oct. 1977).
As stated before, the anchovy population is
patchy on two levels: individuals are aggregated
into schools and schools are aggregated into school
groups. The simulations have quantified the sam-
pling error due to the second level of patchiness
only. Although little is known about the distribu-
tion of anchovy school groups, it was also de-
monstrated that their aggregation is potentially
an important consideration in designing a survey.
The acoustic survey methods currently used by the
National Marine Fisheries Service and the
California Department of Fish and Game do little
more than count the number of anchovy schools
(Hewitt et al. 1976; Mais 1974). The Department
of Fish and Game calculates a biomass estimate by
multiplying the observed school area by a constant
factor thought to represent an average biomass
per unit area. More sophisticated methods of es-
timating biomass from the acoustic signal re-
ceived from a school are now being explored at the
Southwest Fisheries Center. For these reasons,
the problem of sampling error due to a varying
number offish per school (the first level of patchi-
ness) was not addressed here.
Many sources of error may be involved in an
anchovy biomass estimate. Patchiness is impor-
tant in any type of sampling program. Other
sources of error that may be important in an
Table 2. — Analysis of variance of the coefficients of variation from simulated systematic and
stratified systematic surveys of model northern anchovy populations when the model popula-
tion school groups are clumped in one-half of the survey area.
Source of variation
SS
df
MS
Significance level
Main effects
Survey design
0 0732
1
0 0732
F ,,,„ = 12.908, P<0.005
Number of transects
1 4995
6
02499
f ,6641 = 119.403, P<<0.001
Model population
0 1327
14
00095
'^,,4.64, = ''318, p. 0 001
Interactions
Design-number of transects
0 0476
6
00079
F,684, = 3,589, P<0,005
Design-population
0 0794
14
00057
Fi,4„, = 2,591, P< 0.01
Number of transects-population
0.1758
84
0,0021
F,e4e4, = 0,955, P 0 50
Error
0 1858
84
00022
Total
2.1941
209
684
FIEDLER: PRECISION OF SIMULATED TRANSECT SURVEYS
acoustic survey are as follows (Instituto del Mar
del Peru 1974; Cram and Hampton 1976; P. E.
Smith pers. commun.):
1) Failure to discriminate between anchovy
schools and other acoustic targets.
2) Unschooled fish and small schools not detected.
3) Vessel avoidance.
4) Inability to survey in shallow inshore waters.
5) Movement of school groups relative to the sur-
vey grid.
6) Fish in the top surface layer missed by the
acoustic beam.
7 ) Errors in the factor for conversion of the acous-
tic signal information to a biomass estimate.
8) Effect of varying hydrographic conditions on
the acoustic signal.
9) Blocking of signal to and from fish far from the
ship by fish nearer to the ship.
The magnitude of the error caused by these fac-
tors can now only be roughly estimated. They may
affect either or both the precision and accuracy of a
population estimate. Corrections to reduce the
biases are conceivable. The present study has de-
monstrated the magnitude of the error associated
with the patchiness of the anchovy population.
Although the model population distributions may
be crude approximations to the real distribution,
the general conclusions reached here are not
likely to be changed by adding further levels of
complexity to the model. The sampling error due to
patchiness can be reduced by properly designing a
survey, but never eliminated. Temporal and spa-
tial differences in population estimates must be
interpreted with an awareness that the error
exists.
ACKNOWLEDGMENTS
I thank Paul E. Smith for his guidance in com-
pleting this project, as well as Alec McCall, Keith
Parker, Jim Zweifel, and Roger Hewitt for
valuable discussions. Ian Hampton and two
anonymous reviewers provided helpful comments
on earlier drafts of the manuscript. This work was
carried out under a State of California Marine
Research Corpmittee contract at the Southwest
Fisheries Center, La Jolla, Calif.
LITERATURE CITED
Barnes, H., and S. M. Marshall
1951. On the variability of replicate plankton samples and
some applications of "contagious" series to the statistical
distribution of catches over restricted periods. J. Mar.
Biol. Assoc. U.K. 30:233-263.
CRAM, D. L., AND I. Hampton
1976. A proposed aerial/acoustic stragety for pelagic fish
stock assessment. J. Cons. 37:91-97.
Hansen, M. H., W. N. Hurwitz, and W. G. Madow.
1953. Sample survey methods and theory. Vol. 1. John
Wiley and Sons, N.Y., 638 p.
Hewitt, R. P., P. E. Smith, and J. C. Brown.
1976. Development and use of sonar mapping for pelagic
stock assessment in the California Current area. Fish.
Bull., U.S. 74:281-300.
Instituto del Mar del Peru
1974. Report of the fourth session of the panel of experts on
stock assessment on Peruvian anchoveta. Bol. Inst. Mar
Peru (Callao) 2:605-723.
Mais. K. F.
1974. Pelagic fish surveys in the California Cur-
rent. Calif Dep. Fish Game, Fish Bull. 162, 79 p.
SOKAL. R. R., AND F. J. ROHLF
1969. Biometry. W.H. Freeman and Co., San Franc,
776 p.
TAFT B. a.
1960. A statistical study of the estimation of abundance of
sardine (Sardinops caerulea) eggs. Limnol. Oceanogr.
5:245-264.
TAYLOR, C. C.
1953. Nature of variability in trawl catches. U.S. Fish
Wildl. Serv., Fish. Bull. 54:145-166.
WIEBE, P. H.
1971. A computer model study of zooplankton patchiness
and its effects on sampling error. Limnol. Oceanogr.
16:29-38.
WiNSOR, C. P., AND G. C. Clarke
1940. A statistical study of variation in the catch of
plankton nets. J. Mar. Res. 3:1-34.
685
NOTES
INTERSEX ANOMALIES IN SHRIMP OF THE
GENUS PliSAHOPSIS (CRUSTACEA: PENAEIDAE)
While examining a relatively large collection of
Penaeopsis (40 lots containing 196 specimens)
taken by the U.S. steamer Albatross during the
Philippine Expedition. 1907-10, I found three
specimens having external characteristics of both
males and females. Each specimen had a fully
developed thelycum, a moderately well-developed
petasma (about two-thirds the length of the
petasma of males of corresponding sizei, small ap-
pendices masculinae. and genital apertures on the
coxae of the fifth pair of pereopods. The shrimps
were poorly preserved — the exoskeletons were
soft, rostrums and telsons broken, and internal
organs macerated; however, most of the features of
the carapace were clearly distinct and the external
genitalia intact.
The discovery of these shrimp elicits several
questions: to which species do they belong? What
is their functional sex? Do they represent a transi-
tional stage in a protandrous hermaphroditic
species? If not, have their intersex-appearing
anomalies resulted from parasitism? Although
none of these questions is answered definitively,
all are discussed following a brief description of
the external genitalia of the shrimp.
The specimens are deposited at the National
Museum of Natural History: USNM 170581. 23
mm cl (carapace length), Bohol Strait, between
Bohol and Cebu, 291 m, 25 March 1909, Albatross
stn 5418. USNM 170582. 24.5 mm cl, Bohol Strait,
320 m, 25 March 1909, Albatross stn 5419. USNM
170583, 19 mm cl. Gulf of Davao, SE Mindanao,
247 m, 18 May 1908, Albatross stn 5247.
Description
Petasma (Figure IB-C) with length about two-
thirds that of petasma of males off*, rectaciita of
comparable size, and twice as long as endopod of
first pair of pleopods in females of P. rectaciita
(Figure 1 A) and all other congeners. Dorsomedian
lobule with distinct distomedian projection and
well-formed proximal plate. Dorsolateral lobule
with supporting rib (in two of three specimens)
ending proximally in subelliptical process. Ven-
tral costa tapering distally, forming free, inwardly
excavate, blunt projection directed dorsomesially
at broadly obtuse angle with shaft of petasma.
FIGURE 1.— A, Penaeopsis rectacuta, USNM 170586, 9 23.5 mm
cl (carapace length), off Palompon. Leyte, Philippines, endopod
of left first pereopod, dorsal view. B, Penaeopsis sp. USNM
170582, 24.5 mm cl, Bohol Strait, Philippines, Albatross stn
5419, petasma, dorsal view of left half. C, Ventral view of same.
D. Left appendix masculina, dorsal view, same specimen. E,
Penaeopsis rectaciita, USNM 170587, 6 13 mm cl, off Mindanao,
Philippines, A /6a?ross stn 5518, petasma, dorsal view of left half
F, Ventral view of same. G, Right appendix masculina, dorsal
view, same specimen. 0.5 mm indicated.
687
Appendix masculina (Figure ID) minute,
somewhat oval, bearing distal patch of setae;
thickening at its base, inconspicuous.
Thelycum (Figure 2A-C) with anterolateral
borders of plate of sternite XIV varying from
slightly concave to convex, and separated by pos-
teromedian projection of sternite XIII; plate
strongly slanting dorsomesially, its surface flat or
each side biconvex ventrally. Lateral borders
slightly concave, strongly converging posteriorly,
not reaching posterior ridge but separated from it
by deep depression, latter extending anteriorly
adjacent to median rib and merging with an-
teromedian depression; median rib broadest bas-
ally, gently tapering toward, hut not reaching,
posteromedian projection of sternite XIII. Median
plate of XIII trilobed to cordiform, slightly to pro-
nouncedly elongate, covered with setae (most lost
in specimen illustrated in Figure 2Cl except for
naked central concavity, setae directed anteriorly
except on base of posteromedian projection where
directed caudally; posteromedian projection short,
with caudal margin straight or shallowly emargi-
nate. Sternite XII bearing small posteromedian
tooth and pair of sharp ridges, extending pos-
terolaterally from base of tooth.
D
isCLis.sion
In several features the three specimens are
markedly similar to members of P. rectacuta ( Bate
1881). The rostrum (Figure 3) is straight and its
second tooth is located in line with the orbital
margin, the anteroventral angle of the carapace is
approximately 90°, and the moderately long bran-
chiocardiac carina is conspicuous, its anterior ex-
tremity not nearly reaching the posterior end of
the hepatic sulcus. Also the relative length of the
pereopods and — in the two smaller animals — the
shape of the median plate of sternite XIII are simi-
lar to those in P. rectacuta. Furthermore, the three
shrimps were collected together with specimens of
the latter species, all three in localities where the
only other Penaeopsis occurring in the area (P.
eduardoi Perez Farfante 1977) was not taken —
another indication that these shrimp probably be-
long to P. rectacuta.
The petasma and the thelycum of these shrimp
are different from those of other species of
Penaeopsis, including those of P. rectacuta. The
ventral costae of the petasma (Figure IB-C), ta-
pering distally into a short projection disposed at
an obtuse angle to the shaft, differ from those of
adult males of P. rectacuta in which the ventral
costae turn abruptly at right angles and bear a
thin marginal border that is bent inward. In the
three specimens the petasma somewhat resembles
that of large juveniles (with a carapace length of
about 13 mm) of P. rectacuta (Figure lE-F); how-
ever, in the latter the distomedian projections are
less distinct than they are in my specimens or in
adult P. rectacuta. Also, in juveniles of P. rectacuta
the ventral costae do not taper distally into free
projections, instead the tips are broad and turned
Figure 2.~Penaeopsis sp. Thelyca. A, USNM 170582, 24.5 mm cl, Bohol Strait. A/iofro.ss stn 5419. B, USNM 170581, 23 mm cl. Bohol
Strait, A /6a ^rosi- stn 5418. C, USNM 170580, 19 mm cl, Gulf of Davao, Mindanao, Philippmes, Albatross stn 5247. 2 mm indicated.
688
Figure 3. — Penaeopsis sp, USNM 170582, 24.5 mm cl, Bohol Strait, Philippines, Albatross stn 5419. Cephalothorax, lateral view. 5
mm indicated.
at right angles to the shaft. The appendices mas-
culinae (Figure ID) are considerably less well de-
veloped in the present specimens than in juvenile
males of P. rectacuta, in which they are circular
(Figure IG). and bear only marginal setae. The
thelyca of the three specimens differ from those of
P. rectacuta in that the lateral borders of the plate
of sternite XIV converge strongly (rather than
gradually) toward the posterior thoracic ridge and
are separated from the ridge by a deep groove, a
unique characteristic; the median plate of sternite
XIII, although trilobed in one specimen, is cor-
diform in the other two, the latter resembling that
of P. rectacuta.
The functional sex of the three specimens can-
not be ascertained because their gonads had disin-
tegrated. It is unlikely that they were hermaphro-
dites for each bears only one pair of gonopores.
Because they have a completely developed
thelycum one would expect ovipores to occur on
the coxae of the third pair of pereopods, but
whereas these coxae are similar in outline to those
of female P. rectacuta. they lack openings and are
covered by a hardened cuticle (Figure 2A-C) like
those of males; instead, the coxae ofthe fifth pair of
pereopods exhibit a membraneous cuticle with an
opening on the proximomesial border. Although
the latter aperture is situated on the last pereopod,
it occurs on the coxa (Figure 4A) rather than on
the bulging articular membrane as it does in typi-
cal males (Figure 4B). Furthermore, no terminal
ampullae — the ectal muscular region of the vasa
deferentia — appear to have been present, even
though the skeletal muscles are rather well pre-
served.
Many anomalies of the secondary sexual
characters of decapod crustaceans have been re-
coi'ded, e.g. in lobsters (Chace and Moore 1959,
among others) and crayfishes (Turner 1924, 1929,
1935). Recently Zongker (1961) described many
sexually aberrant individuals within a population
o{ Caniharus niontanus acuminatus Faxon 1884.
Among the aberrant individuals she found were
females (sex identified by examination of the
gonads) lacking ovipores on the coxae ofthe third
pair of pereopods, but with "male openings" on
those ofthe fifth, an anomaly similar to that exhi-
bited by my specimens. In these shrimp, the aper-
tures are not typical of penaeoid males because of
their location on the coxae rather than on the
articular membranes. Being present on the coxae,
they resemble female openings; however, ovipores
are typically subcircular rather than slitlike and,
furthermore, they are characteristically situated
on the mesial surface of the coxa, dorsal to the
coxal plate, instead of on the ventral face as in my
specimens.
Individuals ofthe superfamily Penaeoidea bear-
ing both a thelycum and a petasma have not been
recorded previously in the literature. Based on size
distribution and characters ofthe endopod ofthe
first pair of pleopods in females, Heegaard ( 1967,
1971, 1972) suggested the possibility that protan-
drous hermaphroditism occurs in Solenocera
memhranacea (Risso 1816) and also in Perjaeu.'i
kerathurus (Forskal 1977), but no individuals
with both petasma and thelycum were found by
him. The external genitalia in my three specimens
causes one to suspect that they might be transi-
tional forms and that therefore at least some
members ofthe genus Penaeopsis exhibit protan-
drous hermaphroditism (protandrous because at
their size the thelyca are fully developed whereas
the petasmata are relatively small). Their rather
689
Figure 4.— Right fifth pereopod. A, Penaeupsis sp. USNM 170582, 24.5 mm cl. Bohol Strait, Phihppines, A/ftarross stn
5419. B.Penaeopsis rectacuta , USNM 170586, 5 24.5 mm cl, off Palompon, Leyte, Philippines, A /6a?ross stn 5403. 1 mm
indicated.
large size, however, makes it unlikely that they
are in a transitional stage. Furthermore, herma-
phroditism has not been recorded in any species of
the genus Penoeopsis, consequently its occurrence
in these specimens would be exceptional.
Effects of parasitism in a species of
Metapenaeopsis, a genus closely allied io Penaeop-
sis, were reported by Hiraiwa and Sato (1939).
These authors observed conspicuous changes in
the petasmata of males and the gonopores of males
and females in the shrimp Penaeopsis akayebi
Rathbun 1902 i = Mctapenaeopsis barbata de
Haan 1850) parasitized by the bopyrid isopod
Epipenaeon japonica Thielemann 1910. In males,
the petasmata were considerably smaller than
those in normal individuals of corresponding size
and their two parts were unjoined; the gonopores
were barely noticeable, and the papillae, at the
tips of which the gonopores are situated in normal
individuals, were lacking. In the females, the ovi-
pores were obscured, but the thelycum was not
apparently affected by the presence of the para-
site. In the extensive material examined, how-
ever, none of the specimens bore both a petasma
and a thelycum. Among the specimens of P. rec-
tacuta collected in the waters of the Phillippines, I
found a few that were parasitized by bopyrids (one
of them was taken at Bohol Strait, in a locality
near those at which two of my three individuals
were obtained). The parasitized specimens had
normal external genitalia, thus lending no sup-
port to an assumption that the anomalies in the
genitalia of these three individuals were induced
by a bopyrid parasite. Nevertheless, parasitism
offers the only clue as to the possible origin of the
anomalies present in these shrimp.
Acknow ledgments
I am grateful to Horton H. Hobbs, Jr., for valu-
able suggestions during the study, and to Thomas
E. Bowman and Fenner A. Chace, Jr., for review-
ing the manuscript. My thanks are also due Maria
M. Dieguez for preparing the figures.
Literature Cited
CHACE, F. A., JR., AND G. M. MOORE.
1959. A bicolored gynandromorph of the lobster, //omari^s
americanus. Biol. Bull. (Woods Hole) 116:226-231.
Heegaard. p.
1967. On behaviour, sex-ratio and growth of Solenocera
memhranacea (Risso) (Decapoda, Penaeidae). Crus-
taceana 13:227-237.
1971. Penaetis kerathurus Forskal, a protandric her-
maphrodite. Bull. Inst. Oceanogr. Peche iSalambo)
2:257-266.
1972. Sexual dimorphism in some Mediterranean
penaeids and their spawning grounds. Thalassia Jugosl.
8:5-13.
690
HiRAiwA, Y. K., AND M. Sato.
1939. On the effect of parasitic Isopoda on a prawn,
Penaeopsis akayebi Rathbun, with a consideration of the
effect of parasitization on the higher Crustacea in gener-
al. J. Sci. Hiroshima Univ., Ser. B, Div. 1, 7:105-124.
Thielemann, M.
1910. Beitrage zur kenntnis der Isopodenfauna Ostasiens.
Beitr. Naturgesch. Ostasiens. Ahb. Math-Phys. Kl. K.
Bayer. Akad. Wiss. II. Suppl.-Bd. 3 Abh., 109 p.
TURNER, C. L.
1924. Studies on the secondary sexual characters of
crayfishes. I. Male secondary sexual characters in females
of Camharus propinquus. Biol. Bull. (Woods Hole)
46:263-276.
1929. Studies on the secondary sexual characters of
crayfishes, IX. Females of Cambarus with aberrant
female characters. Biol. Bull. (Woods Hole) 56:1-7.
1935. The aberrant secondary sex characters of the
crayfishes of the genus Cambarus. Am. Midi. Nat.
16:863-882.
ZONGKER, J.
1961. Monoecious tendencies in a population o{ Cambarus
montanus acuminatus Faxon. M.A. Thesis, Univ. Vir-
ginia. Charlottesville, 30 p.
ISABEL PEREZ FARFANTE
Northeast Fisheries Center Systematics Laboratory
National Marine Fisheries Service, NOAA
National Museum of Natural History
Washington. DC 20560
ON THE ROLE OF THE DIFFERENT FIBRE TYPES
IN FISH MYOTOMES AT INTERMEDIATE
SWIMMING SPEEDS
In most fishes the myotomal locomotor muscula-
ture is made up of two main fibre types: a super-
ficial layer of red fibres overlies the white fibres
which form the main mass of the myotome. A
spectrum of such differences as mitochondrial con-
tent, enzyme activities, blood supply, and innerva-
tion (as well as color) distinguishes these two fibre
types. The electrophysiogical properties of the two
fibre types have only been investigated in a few
species, but in all of these the white fibres have
been found to propagate muscle action potential,
whereas only local nonpropagated activity is seen
from red fibres (which are invariably multiply in-
nervated). In many (but not all) fishes, there are
also other less abundant fibre types in the
myotomes, in some respects intermediate between
the red and the white fibres (e.g., Patterson et al.
1975).
There is general agreement that at low sus-
tained swimming speeds only the red fibres are
employed and that the white fibres are active dur-
ing short bursts of maximum speed, which cannot
be long sustained. However, agreement has not
yet been reached about which fibres are active
during sustained swimming at speeds above the
minimum cruising speed. Indirect evidence from a
number of teleost species (e.g., Greer Walker and
Pull 1973) indicated that the white fibres are ac-
tive at these intermediate swimming speeds, as
did the direct electromyographic investigations of
Hudson (1973). More recently, several workers
have suggested that fibres of intermediate type
are recruited as swimming speed rises from the
minimal cruising speed, before white fibres are
activated and the fish attains its maximal
sustained speed. In this note, we report elec-
tromyographic observations on various teleosts
swimming at controlled speeds in a tunnel res-
pirometer, which show that the activity of the
myotomal fibre types during sustained swimming
is different in different fishes.
Material and Methods
We studied herring, carp, and trout. Juvenile
Pacific herring, Clupea harengus pallasi Valen-
ciennes, 15-17.5 cm FL (fork length) were caught
by seining in the Georgia Straits, B.C., and held in
circulating seawater at the Department of Zool-
ogy, University of British Columbia, until swum
in a tunnel respirometer (Brett 1964). Herring are
delicate fish and did not settle quietly in the res-
pirometer at flow lengths below 2-3 body lengths
per second (BL/s). Instead, they darted upstream,
and fell back again in an irregular manner, so that
it was necessary to force them to swim at such
speeds from their first entry to the apparatus,
without the acclimation period usual when work-
ing with other fishes.
Varnished copper wire ( 40 standard wire gauge)
electrodes bared at the tips were placed in the
postanal myotomes. The fish were anaesthetized
with MS-2221 (Sandoz) and the electrodes sutured
to the dorsal surface before being led downward
and backward to enter the myotomes. After recov-
ery for 30 min or so in a bucket of seawater, the fish
were introduced to the respirometer and muscle
potentials recorded on a Gould Brush 220 pen re-
corder via Tektronix 122 preamplifiers. It proved
difficult to record from electrodes whose tips lay
amongst the white muscle fibres, but activity from
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
691
these fibres was picked up by electrodes whose tips
lay in the thin lateral red muscle strip. Relative
proportions of red and white muscle fibres were
determined by dissection, and their innervation
patterns were examined in Formalin-fixed mate-
rial. Cryostat sections stained for lipid and for
succinic dehydrogenase by routine methods were
used to distinguish different fibre types.
Similar studies were carried out on carp, Cy-
prinus carpio Linnaeus, 25-30 cm FL, caught by
seining in the Fraser Valley, B.C., and rainbow
trout, Salmo gairdneri Richardson, 17-30 cm FL
obtained from a commercial supplier. Before trial
in the respirometer both species were held in cir-
cular tanks around which water was pumped to
give a constant flow of 30-40 cm/s around the cir-
cumference where the fish normally swam. For
these freshwater fish, much improved sig-
nal : noise ratios were obtained by adding small
amounts of sea water to the freshwater in the res-
pirometer; this did not affect the behavior of the
fish, which were allowed to acclimate for 18-20 h in
the apparatus before testing.
Results
Pacific Herring
The records of muscle activity shown in Figure
lA-C are from electrodes with their tips lying in
the lateral strip of red muscle. These records show
that bursts of irregular potentials around 200-300
/uV peak to peak are recorded from the red fibres
during slow sustained swiming, becoming more
synchronous and shorter as swimming speed in-
creases and tail beat frequency rises. At sustained
■••^M^^^^^
B ^Mf^^y^'^r^'^r^^
^^«/^|4.^^^
D
4/ftW^H4i
ii^f^fJ^aXt^W|^
Figure l, — Records from lateral red muscle of a juvenile Pacific herring ( 16.4 cm fork length) swimming at different speeds. A: 1.8
body lengths per second; B: 2.8 body lengths per second; C: 3.1 body lengths per second; and D: 4.3 body lengths per second. Note
gradually increasing synchrony of potentials as swimming speed and tail beat frequency rise (A-C), and the appearance of large
potentials in D < picked up from underlying white fibres) when the fish struggled to maintain position at high flow speed. Vertical bar:
500 ^V; time marker: seconds.
692
swimming speeds up to 4-5 BL/s, these are the only
kind of potentials recorded; no larger potentials
are observed. Because herring are delicate fish,
velocity/endurance experiments in respirometers
or water tunnels are likely to underestimate their
real capabilities, for it is probable that they slowly
deteriorate during their sojourn under experimen-
tal conditions. Our limited series of measurements
of sustained swimming speeds (Figure 2) showed
that juvenile herring were able to maintain speeds
around 4 BL/s for periods of at least 5 h, a per-
formance about double that previously observed
by Boyar ( 19611, but similar to that seen in large
circular tanks by Hempel (in Blaxter 1969).
Boyar's study was very much more extensive
than ours; some of his results are plotted in Figure
2 for comparison, where it can be seen that the
form of the velocityendurance curves we obtained
is similar to those found for other fish ( e.g. , Hunter
1971 ). It seems probable that 4-5 BL/s represents a
sensible upper value for continuous sustained
cruising by herring of this size.
If the speed of flow in the respirometer is in-
creased above this speed, or if the fish becomes
progressively exhausted, it intersperses periods of
steady swimming, as before, (during which it
slowly falls back to the downstream electrified
grid) with a few rapid tail beats, which drive it
upstream, and the cycle is repeated. During these
rapid beats (Figure ID), large potentials around 1
mV are observed. Similar potentials are seen
when the fish is struggling, and there can be no
doubt that (as in dogfish. Bone 1966) the electrodes
in the red fibre layer pick up these large potentials
from the underlying white fibres. White fibres in
o
5 -
tn 4 —
I
t-
3 -
o
o
m
-
o t
•
o«
-
o
-
o
1
1 1
o
1 o ,
5 10
MINUTES
30
1 2 5
HOURS
FIGL'RE 2. — Swimming speeds (in body lengths per second) plot-
ted against the time that the speeds were sustainable (abscissa).
Dots: present observations; open circles: data plotted for compar-
able fish from Boyar (1961). Note the different forms of the
velocity/endurance curves given by the two sets offish, probably
the consequence of damage to Boyar's fish in his apparatus.
herring are similar to those of dogfish in that they
are focal ly innervated (Bone 1964) and they must
therefore propagate action potentials. The white
fibre system in herring was rapidly exhausted, for
the fish could not swim at velocities above 5 BL/s
for more than 1-2 min (as indicated in Figure 2).
Thus there is good accord between our elec-
tromyographic observations and the values ob-
tained for maximum sustained swimming vel-
ocities: in herring only red muscle fibres are
employed during sustained cruising.
Histologically, the red and white fibres are dif-
ferent from each other. The red fibres are of more
or less uniform diameter, are multiply innervated,
and lipid and succinic dehydrogenase (SDH) posi-
tive. In contrast, the white zone of the myotome
contains both large fibres, and much smaller fibres
arrayed around them in a sort of lattice. Both
types contain little lipid, are SDH negative, and
there are no intermediate fibres either in the
juvenile herring which we examined in the res-
pirometer, or in adults. These histological ar-
rangements are summarized in Figure 3A.
Carp
The carp used were much more robust and
larger fish than the Pacific herring and it proved
possible to make simultaneous recordings of activ-
ity within white and red portions of the myotomes.
The results obtained were entirely different from
those seen in the herring. At speeds between 0.5
BL/s (the lowest speed at which the fish would
swim reliably) and the maximum speed used,
around 4 BL/s, electrical activity was always de-
tectable from both sets of electrodes in red and
white zones of the myotomes (Figure 4). As speed
increases from the lowest values, the bursts of
activity from each zone became more synchronous
and shorter and their amplitude increased. Occa-
sional spikes of greater amplitude were observed
from the white muscle zone (Figure 4B), these
were faster events than those composing the re-
mainder of the motor bursts. When the fish was
swimming near the maximum speed sustainable
in the respirometer ( Figure 4C ), these rapid poten-
tials formed the larger part of the motor bursts and
were always seen on both red and white record-
ings, though smaller from the former. Presuma-
bly, they represent spikelike activity from the
white zone of the myotome, picked up (as in her-
ring) by electrodes in the red zone. Since the red
and white electrodes did not lie in the same
693
^""^^"i^^^ r\ r^
— . - — ■'^
B
FIGURE 3. — Diagram summarizing the organization of white muscle fibres in the myotome of Pacific herring ( A), compared with those
of rainbow trout and carp (B). It is not known whether a single axon may supply both large and small white fibres in herring (though
both are focally innervated), nor is it known which of the two alternative innervation patterns for rainbow trout and carp are actually
present. An overlap of innervation between small and large fibres seems most likely. Note the presence ofintermediate fibres between
red and white zones of the myotome in carp and rainbow trout; they are absent m herrmg. (B partly after Patterson et al. 1975 and
Johnston et al. 1977.)
myotome (though on the same side of the fish and
fairly close to each other), the appearance of occa-
sional spikelike potentials in the white zone was
not always reflected directly in the record from the
red. At lower swimming speeds, when the elec-
trodes in the white zone did not pick up spikelike
potentials at every tail beat, higher recording
speed (Figure 4D) showed the variety of response
from the white zone of the same myotome at suc-
cessive tail beats.
Spikelike potentials were present (although
usually <0.5 mV in amplitude) and were often
reflected at lesser amplitude by the electrodes in
the red portion of the myotome, but there were also
much smaller irregular potentials from the white
region of the myotome, resembling the smaller
irregular potential bursts from the multiply in-
nervated red fibres. In carp, both red and white
muscle fibres are multiply innervated and there
694
are intermediate fibres lying between red and
white fibre zones ( Figure 3B ). The electrodes in the
white portion of the myotome were placed close to
the spinal column so that they did not lie near the
intermediate zone recently described by Johnston
et al. (1977).
Our results clearly indicated that the white
fibres were active even at low swimming speeds,
and that the activity at these speeds did not re-
semble the spikelike muscle potentials observed
when the fish are swimming faster.
Rainbow Trout
Rainbow trout were examined last of the three
fish studied and, to our surprise, gave results com-
parable with those from the herring, although in
salmonids the white fibres are multiply inner-
vated, as they are in carp. At speeds below 2 BL/s,
no activity was detectable from the white (mosaic)
w
->-**-
-Mm-
<— -itU
¥i»f
-Hth-
-^
R
w -^^Hf — f 'if "t 'j -^ t -^ — -^--f-^ -^ 4 '^ — ^ fl 'f
B
R
^^-^#-^^Hik^^-^4*^^-4^4~-^f^^^ 4
w
f/'^
-A'
- -^ -^^'v/^A/VV'^AMl Wl''
... . — 'r^^l
■ff^r-
D
" 1
Figure 4. — Records of activity from red (R) and white ( W) regions of a carp myotome at different swimming speeds. A: 0.75 body length
per second; B: 0.9 body length per second; C: 1 .26 body lengths per second; and D: 2.0 body lengths per second. Note that the white fibres
are active even at low swimming speed and that there is spikelike activity at each tail beat in C. The lowest record (D) taken at higher
chart speed shows the appearance of irregular bursts containing spikelike potentials. Note pick-up of these events by the overlying red
fibre electrode. Vertical bar; 500 /nV; time marker; seconds.
zone of the myotome: 1-200 /xV potentials of the
usual kind were obtained from the lateral red
musculature (Figure 5A). When startled, a few
rapid tail beats drove the fish forward and, under
these conditions, larger spikelike potentials
around 0.5 mV peak to peak were recorded from
the white zone of the myotome corresponding to
the rapid tail beats. As can be seen from Figure 5B,
these events were picked up at lower amplitude by
the electrodes whose tips lay in the lateral red
muscle layer. After a few rapid tail beats, the fish
coasted forward before dropping back and resum-
ing regular swimming: the normal rhythm of the
red fibre system was inhibited for a few cycles.
695
w
P — >V' . i> .ly ■ wi~^. .^^ ^ fi/ «l/.--~i\-. — .jjV . .^f. -y. ..^ ..ij — "V^"^'
w
■MMM«MlM«M««M^i^^MMaaM*M«**i>*MMll«i«*M«M««*«^»MWWM>M4
B
R ,y^,.«^,y..^vy^^^Hi,HVi^WHv-^^ ,«,♦W^-^^^^^-^-^^V,^A--M^"-'V'V-^*^^
w
R
-M~^ P"
'l*»W^**<^j
f^Hk^4"Hfrf
Figure 5. — Records of activity from red (R) and white (W) regions of the myotome of different rainbow trout swimming at various
speeds. A: 1 .8 body lengths per second; B: 4.6 body lengths per second; C: 2.2 body lengths per second. Note that during regular sustained
swimming involving red muscle activity, no activity is detected from the white zone of the myotome. Occasional spikelike activity from
the white zone of the myotome is seen in B and C (also picked up by the electrode lying in the overlying red zone of the myotome),
sometimes inhibits the red muscle bursts (B) and sometimes does not (C). Vertical bars: 1 mV; time marker: seconds.
A single rapid tail beat (to the right of the rec-
ord) interrupted the red muscle for a single cycle.
At higher sustained speeds, above 2 BL/s (as in
Figure 5C) this inhibition of red activity following
single rapid movement no longer took place.
Rather, the behavior was similar to that of the
herring in that the fish fell gradually back despite
the regular activity of the red system, until driven
forward again by a few rapid beats; to drop back
again and repeat the cycle until the white system
was exhausted. Under these conditions, the fish
did not "coast" following rapid tail movements.
No electrical activity was observed from the
white zone of the myotome (the so-called mosaic
zone) apart from the spikelike potentials shown in
Figure 5B and C, although particular pains were
taken to ensure that the electrodes were recording
satisfactorily. All the fish recorded fiom gave this
same result. We conclude from our observations
that this part of the motor system is not active at
696
speeds below 2-2.5 BL/s. Figure 3B summarizes
the structure of the system.
Discussion
Our observations have shown once again that
the lateral red musculature is used by fish for
sustained slow cruising, and that rapid move-
ments of the tail are brought about by the activity
of the white motor system, during which spikelike
potentials can be recorded from the white zone of
the myotome. At intermediate speeds, there are
manifest differences between different fishes.
The simplest situation is shown by the Pacific
herring, where sustained activity depends only on
the activity of the red motor system of the
myotome: the white fibres play no part in any
activity except rapid movements of short duration.
It is true that such movements can "top up," as it
were, the sustained activity of the red motor sys-
tern, but this process cannot be long continued: in
the respirometer flow velocities which overload
the red system and involve occasional activity
from the white system soon exhaust the fish. Pre-
sumably this artificial situation, where the fish
are forced to swim at such speeds, is not found in
nature.
The taxonomic position of clupeids is not yet
agreed upon (see Greenwood etal. 1966), but in the
organization of their myotomal motor system they
show the primitive pattern of focal innervation of
the white fibres (Bone 1970) found also in elas-
mobranches, Agnatha, and dipnoi, but in few
other teleosts.
We may surmise that in all fish where the white
motor system is innervated in this way, sustained
swimming will be the responsibility of the red
system alone, as it is in herring and dogfish. It is
important to notice that this is not to say that
gradation may not take place separately within
either system. For example, there are five fibre
types in the dogfish myotome (three slow and two
fast) distinguishable by histochemical and ultra-
structural criteria, and it is entirely reasonable to
suppose that the two fast fibre types are recruited
for movements of different rapidity as Kry vi and
Totland (1977) have suggested. At present, our
preliminary ultrastructural and histochemical
investigations of young and adult herring
myotomal fibres have only shown one type of red
fibre and two types of white fibre. The two white
fibre types may operate at different stages during
rapid swimming, but there is no direct evidence for
this assumption, and it may be more reasonable to
interpret the smaller white fibres as growth stages
in the development of the larger (see Bone in
press).
In carp, the situation during sustained swim-
ming at all speeds is entirely different. There is
inevitably some ambiguity in the interpretation of
electromyographic records since the position of the
electrode tip may not be certainly known, and the
records obtained may be from nearby small elec-
trical events or from distant larger ones, but it
certainly does not seem probable that the small
events recorded from the carp white muscle at
slow sustained swimming speeds can have been
picked up from the distant red muscle system. To
judge from our records taken deep within the
white muscle, as far as possible from the lateral
red strip, some fibres within the white zone are
active even at the slowest sustained speeds, and
this activity increases as the fish increases its
swimming speed. This kind of electrical activity at
the slower sustained speeds is very similar to that
of the red motor system, and presumably repre-
sents the activity of fibres which are not propagat-
ing muscle action potentials. Such records could
not, naturally, be obtained from the white system
of fish where the white fibres are focally inner-
vated, and in fact are not seen in herring or
dogfish. At higher sustained speeds, or when the
carp is disturbed, much larger rapid potentials are
observed from the electrode within the white zone.
Plainly, two alternative explanations are possible
for the variety of electrical response from a single
recording site within the white muscle. Either the
electrode tip lies close to fibres of two different
types, one of which is capable of propagating mus-
cle spikes and the other is not. In this situation,
the potentials observed simply reflect the fact that
the former system is only activated at higher
speeds, the latter operating during slow swim-
ming and so resembling the red motor system. In
other words, in the carp myotome, the arrange-
ment is essentially a mosaic one, in which red
fibres are intermingled with the usual fast fibres of
the white zone. Or, alternatively, the white zone
contains only a single muscle fibre type, which is
capable of local contractions not involving muscle
action potentials, but can also be stimulated to
twitch rapidly and, in this state, propagates mus-
cle action potentials. As pointed out earlier (Bone
1975) this would be an ingenious way of ensuring
for a single muscle fibre that it always operated at
the flattened upper part of the power curve, con-
tracting at very different rates whilst swimming
slowly and rapidly.
Our electromyographic records do not allow us
to distinguish between these two alternatives but
there is no evidence from the histochemical
studies by Patterson et al. (1975), or the recent
excellent paper by Johnston ( 1977), that there are
"red" fibres in the white zone of the carp myotome.
These authors have demonstrated clearly, how-
ever, that there is a zone of intermediate fibres
between the lateral red and deep white fibres of
the carp myotome. They have also shown that
these three fibre types are active at different
swimming speeds. At 1 BL/s only red fibres were
found to be active; at 1.3-1.5 BL/s both red and
pink fibres were active, whereas at 2.0 BL/s and
above, electrical activity appeared from the white
zone of the myotome. These results clearly indi-
cated the sort of recruitment of intermediate fibres
at intermediate sustained swimming speeds
697
which was implied by their accompanying
biochemical studies. Interestingly enough,
Johnston et al. ( 1977) observed the same kind of
electrical activity from the white zone of the
myotome that we observed at low speeds, and it
seems therefore extremely probable that such ac-
tivity (around 75 fxV in their records at 2.0 BL/s)
is indeed generated by muscle fibres in the white
zone. They did not observe spikelike activity from
the white zone, presumably because their fish
were not swimming sufficiently fast, i.e., they in-
vestigated only the lower sustained swimming
speed range.
It is then still an open question whether indi-
vidual fibres in the white zone can sometimes op-
erate producing only local potentials, at other
times generating muscle action potentials; or
whether there are two different fibre types in the
white zone, as yet not distinguishable histochemi-
cally. We incline to the former opinion, but to
settle the matter evidence from intracellular
studies will be essential.
In rainbow trout, our results were again differ-
ent. We obtained no evidence for activity of the
mosaic zone of the myotome during sustained ac-
tivity even at 4.5 BL/s (the maximum speed at
which we could swim the smaller fish). Consider-
ing Hudson's (1973) electromyographic evidence
from the same species, where he observed activity
from the mosaic zone at speeds above 3.0BL/s, this
seemed at first rather surprising.
However, the fish Hudson used came from a
stock of notoriously poor swimming performance
(see Webb 1971), and it is therefore quite possible
that we never attained the critical speed at which
the mosaic muscle became active in our fish. The
main muscle mass in rainbow trout consists of a
mosaic of small reddish fibres scattered amongst
larger pale fibres (Johnston et al. (1975) have
studied them histochemically), and it is thus un-
clear whether the low-level electrical activity which
Hudson (1973) recorded from this region (similar
to that which we found in carp white muscle)
comes from the same fibres as those generating
muscle action potentials during burst swimming.
In other words, the two kinds of electrical re-
sponses from the rainbow trout mosaic muscle
may result from the activity of two different kinds
of muscle fibres.
Fish are so diverse, and their patterns of life so
varied, that it is hardly surprising that there
should be differences on their locomotor muscula-
ture. We perhaps ought rather to be surprised at
698
the general uniformity of design of the locomotor
system imposed by the aquatic medium. It seems
probable, from the distribution of patterns of in-
nervation amongst different fish groups, and in-
deed amongst the teleosts alone, that focally in-
nervated, twitch fibres operating by anaerobic
glycolysis for short bursts of swimming represent
the primitive arrangement of the aquatic fast
motor system (see Bone 1970).
This fast-motor system contrasts markedly with
the universally found multiply innervated
nontwitch red fibre system for sustained move-
ment that operates aerobically. However, his-
tological and biochemical investigations of the
white myotomal zones of some specialized teleosts
such as tuna (Guppy et al. in press) or carp
(Johnston et al. 1977) have shown a definite
aerobic capacity in the white fibre system, and the
original simple dichotomy between anaerobic
white fibres and aerobic red fibres rather naively
suggested from elasmobranch studies (Bone 1966)
is plainly not a good description of the operation of
the myotome in all teleosts.
On the whole, it seems reasonable to assume
that in most teleosts where the white portion of the
myotome is multiply innervated, there will be
aerobic intermediate fibres for use during fast sus-
tained cruising, and that at the maximum cruis-
ing speed at least some fibres in the white zone of
the myotome will also be active aerobically. This
seems to be the situation in rainbow trout, and it
probably also obtains in most scombrids.
The situation in carp is less clear. The work of
Smitetal. ( 1971) has shown that goldfish (close to
carp) are able to sustain high speeds in a res-
pirometer apparently using the white muscle
system anaerobically. In line with this, Driedzic
and Hochachka ( 1975) were unable to detect other
energy sources than anaerobic glycolysis in carp
white muscle, and Johnston et al. (1977) only
found low values of aerobic enzymes in this sys-
tem. We have provided clear evidence that the
white motor system is operating over a wide speed
range, from the lowest speed at which the fish will
swim in the respirometer, and it seems bizarre
that a relatively inefficient anaerobic metabolism
should drive sustained activity. At low sustained
swimming speeds carp might keep in overall
aerobic balance by transferring lactate from the
white zone to other regions of the body, where
lactate could be completely metabolized (Bone
1975). Driedzic and Hochachka found only low
lactate levels in the white zone after severe
hypoxic stress, and suggested that this could be
explained by lactate transfer out of the system. It
is very hard to believe that such a process could
account for the extremely interesting results of
Smit and his colleagues (Driedzic and Hochachka
entitled their paper "The unanswered question of
high anaerobic capabilities of carp white muscle"),
and we agree with Johnston et al. ( 1977) in their
conclusion that carp would appear to be an ideal
species for studying the relationship between
muscle design and locomotor function.
Ackn<)\\ ledgements
We are grateful to D. J. Randall (Department of
Zoology, University of British Columbia, Van-
couver) for procuring the Pacific herring for us and
for allowing us to use his respirometer. One of us
(Q. B.) did part of this work during the award of a
Nuffield-NRC visiting lectureship, which is grate-
fully acknowledged, as is support (to D. R. J.) by
research grants from the National Research
Council of Canada and Fisheries Research Board
of Canada.
Literature Cited
BLAXTER. J. H. S.
1969. Swimming speeds offish. In A. Ben-Tuvia and W.
Dickson (editors), Proceedings of FAO conference on fish
behaviour in relation to fishing techniques and tactics, p.
69-100. FAO Fish. Rep. 62
BONE. Q.
1964. Patterns of muscular innervation in the lower chor-
dates. Int. Rev. Neurobiol. 6:99-147.
1966. On the function of the two types of myotomal muscle
fibre in elasmobranch fish. J. Mar. Biol. Assoc. U.K.
46:321-349.
1970. Muscular innervation and fish classification. In A.
de Haro (editor), I Simposio Intemacional deZoofilogenia,
p. 369-377. Alamanca V.P.
1975. Muscular and energetic aspects of fish swim-
ming. In T. Y-T. Wu, C. J. Brokaw, and C. Brennen
(editors), Swimming and flying in nature, Vol. 2, p. 493-
528. Plenum Press, N.Y.
In press. Locomotor muscle. In W. S. Hoar and D. J. Ran-
dall (editors), Fish physiology, Vol. 7. Academic Press,
N.Y.
BOYAR. H. C.
1961. Swimmingspeedofimmature Atlantic herring with
reference to the Passamaquoddy Tidal Project. Trans.
Am. Fish. Soc. 90:21-26.
BRETT. J. R.
1964. The respiratory metabolism and swimming per-
formance of young sockeye salmon. J. Fish Res. Board
Can. 21:1183-1226.
DRIEDZIC. W. R.. AND P. W. HOCHACHKA
1975. The unanswered question of high anaerobic
capabilities of carp white muscle. Can. J. Zool. 53:706-
712.
Greenwood. P. H., D. E. Rosen, S. H. Weitzman, and G. S.
Myers
1966. Phyletic studies of teleostean fishes, with a provi-
sional classification of living forms. Bull. Am. Mus. Nat.
Hi.st. 131:339-455.
Greer Walker. M., and G. Pull.
1973. Skeletal muscle function and sustained swimming
speeds in the coalfish Gadus virens L. Comp. Biochem.
Physiol. 44A:495-501.
GuppY. M., W. C. Hulbert. and p. W. Hochachka
In press. The tuna power plant and furnace. In G. D.
Sharp and A. E. Dizon (editors). Physiological ecology of
tuna. Academic Press, N.Y.
Hudson, R. C. L.
1973. On the function of the white muscles in teleosts at
intermediate swimming speeds. J. Exp. Biol. 58:509-
522.
HUNTER. J. R.
1971. Sustained speed of jack mackerel, Trachurus sym-
metricus. Fish. Bull., U.S. 69:267-271.
Johnston, I. A.
1977. A comparative study of glycolysis in red and white
muscles of the trout iSalmo gairdneri) and mirror carp
iCyprinus carpio). J. Fish Biol. 11:575-588.
Johnston. I. A., W. Davison, and G. Goldspink
1977 Energy metabolism of carp swimming muscles. J.
Comp. Physiol. 114:203-216.
Johnston, I. A., P. S. Ward, and G. Goldspink
1975. Studies on the swimming musculature of the rain-
bow trout. I. Fibre types. J. Fish. Biol. 7:451-458.
Kryvi, H., and G. K. TOTLAND
1977. Histochemical studies with microphotometric de-
terminations of the lateral muscles in the sharks Etmop-
terus spinax andGaleus melastomus. J. Mar. Biol. Assoc.
U.K. 57:261-271.
Patterson S., I. A. Johnston, and G. Goldspink.
1975. A histochemical study of the lateral muscles of five
teleost species. J. Fish. Biol. 7:159-166.
SMIT. H., J. M. AMELINK-KOUTSTAAL, J. Vl.rV'ERBER. AND
J. C. VON Vaupel-Klein
1971. Oxygen consumption and efficency of swimming
goldfish. Comp. Biochem. Physiol. 39A:l-28.
WEBB. P. W.
1971. The swimming energetics of trout. I. Thrust and
power output at cruising speeds. J. Exp. Biol. 55:489-
520.
Quentin Bone
Plymouth Laboratory
Marine Biological Association of the United Kingdorri
Citadel Hill
Plymouth PLl 2PB, England
JOE KICENIUK
David r. Jones
Department of Zoology
University of British Columbia
Vancouver, B.C., Canada V6T 1W5
699
SUMMER FOOD OF THE PACIFIC COD,
GADiS MACROCEPHALLS.
NEAR KODIAK ISLAND, ALASKA
1,2.3
The Pacific cod, Gadiis macrocephalus Tilesius,
was the target of the earliest United States com-
mercial fishery in the North Pacific (Buck'*). Its
fleet, organized in spring 1865 (Bean 1887), began
to fish along the Alaska Peninsula and the Aleu-
tian Islands and eventually expanded into the
Bering Sea (Cobb 1916). Dwindling stocks and
poor market prices ultimately resulted in the col-
lapse of this fishery shortly after World War II
(Ketchen 1961).
Growing pressures in recent years on domestic
fishing stocks, in addition to increased worldwide
protein demand, improved technological skills
and readily available investment capital, have re-
sulted in renewed interest in Pacific cod in the
United States (Jones 1977). A bottomfish survey
off the coast of Kodiak Island and throughout
Shelikof Strait by the National Marine Fisheries
Service in 1973 showed the Pacific cod to be one of
the most abundant fishes inhabiting the area and
the standing stock was conservatively estimated
to be about 36,363 t (Hughes and Parks 1975). A
small experimental trawl fishery for the Pacific
cod and other bottom fishes has been proposed for
the Kodiak region by Jones (1977).
Preliminary examination of G. macr^ocephalus
stomach contents by Alaska Department of Fish
and Game (ADF&G) biologist Guy C. Powell and
the author during ADF&G crab investigations off
Kodiak Island indicated a high frequency of oc-
currence of the commercially important snow
crab, Chionoecetes bairdi. In view of the probable
predation pressure on existing snow crab popula-
tions by G. macrocephalus and in view of the po-
tential commercial importance of the Pacific cod,
the summer food habits of this fish in the Kodiak
area were examined by me. Ancillary goals in-
cluded a comparison of food data from pot- and
trawl-captured cod.
'Contribution No. 339, Institute of Marine Science, University
of Alaska, Fairbanks, AK 99701.
^This study was partially supported under contract 03-5-022-
56 between the University of Alaska and NOAA, U.S. Depart-
ment of Commerce through the Outer Continental Shelf En-
vironmental Assessment Program to which funds were provided
by the Bureau of Land Management, U.S. Department of In-
terior.
■'Based on a thesis submitted in partial fulfillment of the re-
quirements for the M.S. degree. University of Alaska
■•Buck.E. H. 1973. Alaska and the law of the sea. National
patterns and trends of fishery development on the North Pa-
cific. Alaska Sea Grant Rep. No. 73-4, 65 p.
700
Methods
Specimens were taken near Kodiak Island,
Alaska, (Figure 1) in conjunction with the crab-
assessment studies of ADF&G and the surveys of
the International Pacific Halibut Commission.
Fishing gear consisted of commercial king crab
pots, measuring 203 x 203 x 76 cm (inside) and
weighing 340 kg; baited with chopped, frozen her-
ring. Webbing was #72 tarred nylon thread with
mesh stretched to 7.6 cm. The gear used on the
halibut-survey vessels in July 1975 and July 1976
was a standard 400-mesh Eastern otter trawl
(Greenwood 1958). Sampling by pots was from 26
June to 3 August 1973, 28 June to 31 July 1974,
and 30 June to 27 July 1975. Stations usually
consisted of 4-12 pots in a straight line, equally
spaced every 0.46 km. Gear was pulled every 18-24
h except when weather conditions prolonged
fishing time.
A haphazard sample of 3,933 of Pacific cod was
taken from 10,857 cod caught in pots (the number
sampled was contingent on the shipboard time
available for analysis of stomach contents). Food
items were identified to the lowest taxon practical
aboard ship, and unidentifiable contents were pre-
served for later laboratory examination. Analysis
of stomach contents was carried out using the fre-
quency of occurrence method in which the prey
organisms are expressed as the percent of
stomachs containing various food items from the
total number of stomachs analyzed. Cod were ar-
bitrarily divided into 33-52 cm, 53-72 cm, and
73-92 cm size (total length) groups for analysis.
The frequency of occurrence method was also
used for food data from trawl-caught Pacific cod.
The stomachs of 344 cod were examined from 24
trawl stations, which were located in the same
general area as the pot stations (Figure 1).
Results and Discussion
As determined from the pot data, the summer
diet of G. macrocephalus was fishes, crabs,
shrimps, and amphipods, in decreasing order of
occurrence (Table 1). The most frequently occur-
ring fish was walleye pollock, Theragra chalco-
gramma. Flatfishes (Pleuronectidae) and Pacific
sand \a.nce, A in modytes hcxapterus, were also fre-
quent. Suyehiro (1942:233-236), Moiseev (1953,
1960), and Mito (1974) also reported that Pacific
cod feed on these fishes.
Figure l. — Stations near Kodiak Is-
land. Alaska, where Pacific cod were
collected by pots and trawls during
summers of 1973-75.
... • ^ .. .. . .
-(11.9)
"1
O.ll- (9.2)
53
1
1
1
1
1
2
2
2
1
1
3
1
34
0.1
0.1
0.1
'♦(0
3
1
3) 9
4.5] 74 3.6]
0.9^.(5.4) 24 I.2L.
■(4.8)
0.1
0.1
0.2
0.1
0.1
1
0.1
1
0.1
1
0.1
1
0.1
2
0.1
1
0.1
1
0.1
4
0.2
36
1.8
5
0.2
1
0.1
3
0.1
1
0.1
1
0.1
0.2
2.6
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.2
0.1
0.1
0.2
0.2
0.1
1.7
164 8.0
127
49
l-(10.3)
0.08
0.03
0.2
L*(0.31)
1.2 l-
0.03
(4.4)
2
0.05
1
0.03
5
0.1
0.08
0.03
0.05
0.03
0.03
0.03
0.05
0.03
0.03
0.03
0.05
0.08
0.05
0.03
0.08
0.1
109
2.7
0.03
12
0.3
0.05
0.03
0.03
0.1
0.03
0.03
0.03
0.2
0.05
0.03
11
0.3
106
2.7
0.03
0.03
0.05
0.03
0.05
0.01
0.05
0.05
0.08
0.03
0.08
0.03
0.03
0.08
0.03
0.1
0.05
0.03
0.2
0.03
0.03
0.01
0.03
61
1.5
-(10.2)
326
8.3
702
TABLE 1. — Continued.
FOOD ITETIS
1973
N=689
7.
Freq. Freg.
1974
N=
1183
%
Freq
Fr€
q
34
2.
9
4
0.
3
195
16.
5
118
10.
cT
4
0.
3
1975
N=2061
%
Freq . Freq .
TOTAL
1973-75
N=3933
X
Freq. Freq.
Malacostraca
Euphausiacea (krill) and
Mysidacea (mysids)
Isopoda (pill bugs)
Amphipoda (sand fleas)
A^pelisaa macroaephala
Decapoda
Pandalidae (shrimps)
Pandalus hovealis
Pandalopsis dispar
Pandalus goniurus
Pandalus hypsinotus
Pandalus montagui tridens
Pandalus platyseros
Crangonidae (shrimps)
Argis arassa
Scleroemngon boreas
Hippolytidae
Spivontoaar^is sp.
Unidentified shrimps
Paguridae (hermit crabs)
Etassoahirus oavinanus
Elassochirus tenuinanus
Lithodidae (crabs)
Paralithodes camtsahatioa
Placetron wosnessenskii
Fhinolithodes uosnessenskii
Galatheidae (crabs)
Munida quadrispina
Cancridae (crabs)
Cancer sp.
Telmessus cheiragonus
Pinnotheridae (pea crabs)
Pinnixa sp.
Hajidae (spider crabs)
Chionoecetes bairdi
Hyas lyratus
Oregonia gracilis
Unidentified crabs
Echinodermata
Asteroidea (sea stars)
Ctenodiscus cnspatus
Echinoidea (sea urchins)
Holothuroidea (sea cucumbers)
Ophiuroidea (brittle scars)
Ophiura sarsi
Vertebrata
Osteichthyes
Clupeidae (herrings)
Clupea harengus pallasi
Osmeridae (smelts)
Mallotus villosus
Gadldae (codfishes)
Theragra chalaogramma
Gadus macrocephatus
Zoarcidae (eelpouts)
Lyaodes brevipes
Scorpaenidae (rockflshes)
Hexigrammidae (greenlings)
Pleurogramus monopterigius
Cottidae (bullheads)
Dasyaottus se tiger
Hemilepidotus jordani
Gyrmocanthus sp .
Agonidae (poachers)
Bathymasteridae (ronquils)
Bathymaster signatus
Trichodontidae (sandfishes)
Trichodon trichodon
Cyclopteridae (lumpsuckers)
Pleuronectldae (flatfishes)
Atheresth.es storrias
Hippog lossoides e lassodon
Hippoglossus stenolepis
Ammodytidae (sand lances)
Arrmodytes hexapterus
Stichaeidae (pricklebacks)
Crypacanthodidae (wrymouths)
Lynaoneotes aleutensis
Unidentified fishes
Stomachs empty
20
3
192
67
77
131
24
281
13
12
12
7
29
1
22
20
14
9
256
2.9
0.4
27.8
9.7
11.1
1
95
19.0L-(39.8) 82
3.41 21
0.3
0.5
0.1
0.7
40.7
1.8
1.7 I-
36
0.1
0.1
0.3
0.8
0.4
1.7
1.0
4.2
0.1
1.1
5
3
•-(0.4) -
0.1
3.2
2.9
2.0
1.3
37.1
1.6
32
13
9
27
4
1
21
20
0.1
8.0
6.9^^(25.3) 171
1.8
0.1
0.8
0.1
0.1
3.0
428 36.2
44 3.7
3 0.3
"(49.2) 3 0.3
2 0.2
23
•-(55.9) 476
59
0.4
0.3
0.1
0.2
2.7
0.9
0.8
0.1
2.3
0.3
0.1
0.3
0.1
1.8
1.7
0.3
40.2
5.0
'-(0.9)
1
1
1
10
3
2
181 8.8 235
10 0.5 17
407 19. "s] 794
52 2.5^(22.3) 52
185
166
23
4
7
8
4
458
3
5
166
8.1
19
0.9
4
0.2
7
0.3
8
0.4
3
0.2
286
13.9
3
0.2
5
0.2
55
2.7
2
0.1
3
0.2
31
1.5
2
0.1
1
0.1
1
0.1
13
0.6
0.2 5
a.3_'-(32.9) 384
100
2
4
42
3
1
1
14
4
3
64
0.1
1.1
735 35.6
42 2.0
6 0.3
-(46.4) 4 0.2
0.1
0.1
0.1
0.5
0.2
0.1
Hi.i)
4
1
2
17
6
2
2
0.1
4
0.2
1
0.1
109
5.3
3
0.2
7
0.3
3
0.2
2
0.1
6
0.3
2
0.1
1
0.1
6
0.3
17
0.8
0.1
2
0.1
5
0.2
40
1.9
2
0.1
12
0.6
2
0.1
9
0.4
10
0.5
6.0
0.4
20
0.2~l
1.3 L.<21.
5)
4.7
4.2
0.6
0.1
0.2
0.2
0.1
11.6
0.08
0.1
0.1
9.8 '-<31.8)
2.5
0.05
0.1
1.1
0.08
0.03
0.03
0.4
0.1
0.08
1.6
1444 36.7
99 2.5
9 0.2
K44.7) 19 0.5
'-<46.0)
0.1
0.03
0.05
0.4
0.1
0.05
■-(0.73)
9
0.2
9
0.2
1
0.03
153
3.9
23
0.6
45
1.1
3
0.08
2
0.05
2
0.05
41
1.0
2
0.05
1
0.03
6
0.1
20
0.5
0.05
6
0.1
7
0.2
83
2.1
2
0.05
12
0.3
2
0.05
49
1.2
24
0.6
4 0.2 17 0.4
-(51.9) 655 31.8L-(44.1) 1387 35.3 '-(48.2)
184 8.9 251 6.4
703
Table 2. — The importance of the snow crab, Chionoecetes bairdi, in the summer diet of Pacific cod.
Analysis based on specimens from pots. Crab incidence is given for total number of cod examined;
incidence as a percent of feeding cod given in parentheses.
Cod examined
Feeding cod
Incidence of crabs
Crabs
Average crab occurrence
Sampling date
(no.)
(%)
Number
Percent
(no.)
in cod feeding on crabs
26 June-3 August
689
98.8
281
40.7
1,022
3,6
1973
(41.3)
28 June-31 July
1,183
95.0
427
36.2
1.033
2.4
1974
(38,0)
30 June-27 July
2.061
91.0
734
35.6
2,682
3.6
1975
(39,1)
Total
3.933
93.6
1.442
36,7
(39.2)
4.737
3.3
Table 3. — Frequency and percent frequency of occurrence of food items in stomachs ofGadus mac-
rocephalus collected July 1975 and 1976 by otter trawl near Kodiak Island, Alaska. N = number of
stomachs examined. Subtotals in parentheses.
Ju
ly
July
Total
1975
1976
1975-
-1976
N =
150
N =
= 194
N =
344
Food items
Freq.
_%_
Freq
Freq
%_
Freq
Freq.
% Freq
Annelida
Polychaeta
2
1.3
3
1.5
5
1.4
Mollusca
Pelecypoda and Gastropoda
17
11.3
10
5.1
27
7.8
Cephalopoda
3
2.0
8
4.1
11
3.2
Arthropoda
Crustacea
Euphausiacea and Mysidacea
13
8.6
10
5.1
23
6.7
Isopoda
-
-
3
1.5
3
0.9
Amphipoda
14
9.3
15
7.7
29
8.4
Decapoda
Pandalidae
16
10.7
24
12.4'
40
11.6
Crangonidae
37
24.7
37
19.1
74
21.5
Unidentified shrimps
18
12.0
•(47
4)
24
12.4
-(43
9)
42
12.2
-(45
3)
Ma j idae
Chionoecetes bairdi
55
36.7
82
42.T
137
39.8
Unidentified crabs
13
8.7
^(45
4)
23
11.9
-(54
2)
36
10.5
-(50
3)
Echinodermata
1
0.6
-
-
1
0.3
Vertebrata
Osteichthyes
Cadidae
Theragra ahalaogranma
6
4.0
7
3.6
13
3.8
Pleuronec tidae
5
3.3
4
2.1
9
2.6
Ammodytidae
Armodytes hexapterus
20
13. 3
13
6.7
33
9.6
Unidentified fishes
66
44.0
-(64
6)
70
36.1
^(48
5)
136
39.5
- (55
5)
Stomachs empty
7
4.7
13
6.7
20
5.8
Table 4. — Comparison of percent frequency of occurrence of
summer food groups in male and female Gadua macrocephalus
caught by pots and trawls in the Kodiak Island area.
Percent frequency
of occurrence in
Pot-caught cod
Trawl-
caught cod
Food groups
Males
Females
Males
Females
Fishes
21 8
24.2
26.3
24,8
Crabs
22,0
19.3
24,2
20,9
Sfirimps
15,1
14,2
15,4
24,7
Amphipods
10,0
14,3
4,1
4,3
Gastropods and
pelecypods
5,0
4,7
3,3
4,5
Cephalopods
36
4,7
2,3
09
Eupfiausiids and
mysids
2,1
4,0
4,0
2,7
Polychaetous annelids
14
3,1
0,3
1,1
Echinoderms
0.4
0.4
0,1
0,2
Isopods
0.2
0.2
0,5
0,4
Empty stomachs
4.4
2.0
2,8
3,0
Stomachs
examined (no.)
2,106
1,827
188
156
other studies on Gadiformes (e.g., Romans and
Vladykov 1954; Wigley 1956; Powles 1958; Wigley
and Theroux 1965).
A significant difference ( x~, a = 0.05) was found
for occurrence of food groups between years for
each size group (Figure 2). The only similarity was
among 33-52 cm fish between 1973 and 1975 and
among 73-92 cm fish between 1974 and 1975.
Some trends in frequency of food groups by cod size
were apparent (Figure 2). Fishes and cephalopods
increased in frequency with increasing cod size
over all years while amphipods and polychaete
worms decreased. Daan (1973) investigated the
relative size of food items (crustaceans and fishes)
used by the Atlantic cod, G. morhua, and found
704
1973
1974
1975
100
50
"" 100
CL
IT
D
U
" 50
0
100
-I 1 r-
73 92 cm
-I ^ 1 r-
50
33 52 cm
N 66
-■ 1 1 1 1 1 1 1 r-
53 72 cm N - 424
—I \ \ r-
«= 199
S. fja.
100
50
0
100
50
100
73 92 cm
- 53 72 cm
n 1 \ —
33 52 cm
50 —
-\ 1 r-
«= 125
N = 719
N = 339
100
50
100
50
100
-I 1 1 r
73 92 cm
53 72 cm
yn
N = 212
T 1 r-
N = 1047
." 2 e
O O
D. a
E
<
Figure 2.
Food Items
-Percent frequency of occurrence of summer food items within three size groups of pot-caught Pacific cod by year of
coUection — 1973-75 — near Kodiak Island, Alaska.
that smaller crustaceans were more commonly
found in small cod while a gradual shift to a mixed
diet of larger prey (primarily fishes) was noted for
the larger fish. Arntz ( 1974) examined juvenile G.
morhua, and found the most important food to be
small crustaceans, mainly cumaceans (35.6% by
weight of the total food consumed); fishes ac-
counted for only 15.3% by weight of the total food
consumed. This trend of large cod being more pis-
civorous than small cod has also been dem-
onstrated by Powles (1958) and Rae (1967).
Acknowledgments
I am especially indebted to the ADF&G and Guy
C. Powell for their assistance in collection of data.
Special thanks to Howard M. Feder, University of
Alaska, for his editing suggestions. Thanks to
John R. Hilsinger, University of Alaska, for allow-
ing me to use his Pacific cod feeding data obtained
during International Pacific Halibut Commission
surveys, and to George Mueller and Kenneth
Vogt, both of the Marine Sorting Center, Univer-
sity of Alaska, for their taxonomic assistance. Mol-
lusc identifications were made by Rae Baxter,
ADF&G.
Literature Cited
Arntz, W. E.
1974. The food of juvenile cod iGadus morhua L.) in Kiel
Bay. [In Germ., Engl, summ.] Ber. dtsch. wiss. Komm.
Meeresforsch. 23:97-120.
BARR, L.
1970. Alaska's fishery resources — the shrimps. U.S. Fish
Wildl. Serv., Fish. Leafl. 631. 10 p.
BEAN, T. H.
1887. The cod fishery of Alaska. In G. B. Goode and staff
of associates, Fishery and fishery industries of the United
States, Sec. V. Vol. 1, p. 198-226. Wash.
BROWN, R. B., AND G. C. POWELL.
1972. Size at maturity in the male Alaskan tanner crab,
Chionoecetes bairdi, as determined by chela allometry,
reproductive tract weights, and size of precopulatory
males. J. Fish. Res. Board Can. 29:423-427.
COBB, J. N.
1916. Pacific cod fisheries. Rep. U.S. Comm. Fish., 1915,
append. 4, 111 p. (Doc. 830.)
DAAN, N.
1973. A quantitative analysis of the food intake of North
Sea cod, Gadus morhua. Neth. J. Sea Res. 6:479-517.
705
Greenwood, M. R.
1958. Bottom trawling explorations of southeastern
Alaska. 1956-1957. Commer. Fish. Rev. 20( 12):9-21.
ROMANS, R. E. S., AND V. D. VLADYK JV.
1954. Relation between feeding and the sexual cycle of the
haddock. J. Fish. Res. Board Can. 11:535-542.
HUGHES, S. E., AND N. B. PARKS.
1975. A major fishery for Alaska. Natl. Fisherman
55(13):34-40
JONES, W. G.
1977. Emerging bottomfish fisheries - potential ef-
fects. Alaska Seas Coasts 5:1-5.
KETCHEN, K. S.
1 96 1 . Observations on the ecology of the Pacific cod (Gadus
macrocephalus ) in Canadian waters. J. Fish. Res. Board
Can. 18:513-558.
MITO, K.
1974. Food relation in demersal fishing community in the
Bering Sea - walleye pollock fishing ground in October
and November 1972. Master's Thesis, Hokkaido Univ.,
Hakodate, 86 p.
MOISEEV, P. A.
1953. |Cod and flounders of far-eastern waters.] Izv.
Tikhookean. Nauchno-Issled. Inst. Rybn. Khoz.
Okeanogr. 40:1-287. (Tranl. 1956, Fish Res. Board Can.
Transl. Ser. 119, 576 p.)
1960. On the habits of the cod-fish Gadus morhiia mac-
rocephalus Tilesius in different zoogeographical re-
gions. [In Russ., Engl, summ.] Zool. Zh. 39:558-562.
POWLES, P. M.
1958. Studies of reproduction and feeding of Atlantic cod
(Gadus callarias L.) in the southwestern Gulf of St. Law-
rence. J. Fish. Res. Board Can. 15:1383-1402.
Rae, B. B.
1967. The food of cod in the North Sea and on the west of
Scotland grounds. Dep. Agric. Fish. Scotl, Mar. Res.
1967(1), 68 p.
RONHOLT, L. L.
1963. Distribution and relative abundance of commer-
cially important pandalid shrimps in the northeastern
Pacific Ocean. U.S. Fish Wildl. Serv., Spec. Sci. Rep.
Fish. 449, 28 p.
SUYEHIRO, Y.
1942. A study on the digestive system and feeding habits of
fish. Jpn. J. Zool. 10:1-303.
WIGLEY, R. L.
1956. Food habits of Georges Bank haddock. U.S. Fish
Wildl. Serv., Spec. Sci. Rep. Fish 165, 26 p.
WiGLEY, R. L., AND R. B. THEROUX.
1965. Seasonal food habits of Higlands Ground had-
dock. Trans. Am. Fish. Soc. 94:243-251.
Stephen C. Jewett
Institute of Marine Science
University of Alaska
Fairbanks, AK 99701
A COMPUTER SOFTWARE SYSTEM FOR
OPTIMIZING SURVEY CRUISE TRACKS'
Since 1972, the Southeast Fisheries Center, Na-
tional Marine Fisheries Service, NOAA, has been
conducting resource assessment surveys for
groundfish in the northern Gulf of Mexico. Ran-
dom sampling stations were selected and cruise
tracks plotted by hand requiring several man-
days of effort without assurance than an optimum
cruise track had been chosen. Consequently, a
computer routine was developed at the NMFS Na-
tional Fisheries Engineering Laboratory, Bay
Saint Louis, Miss., to satisfy two requirements:
Generate a set of randomly selected sampling sta-
tions from a preestablished station grid and
minimize the distance the vessel must travel to
sample each station once. This paper presents the
resultant routine, a comparison of results with
actual cruises, and a discussion of other possible
applications of the program.
Background
The problem of determining the optimum cruise
track to sample a given set of stations can be re-
stated as, "determining the shortest route from
one point to another which allows a vessel to visit
every station once." This problem is similar to one
in the field of operations research generally refer-
red to as "the traveling salesman problem." The
original formulation of the problem was to
minimize the time required by a traveling sales-
man to visit a number of cities and return home
(Bellmore and Nemhauser 1968). Several al-
gorithms have been developed which solve the
problem exactly; however, computer storage and
running time increase exponentially with the
number of points to be visited. Because the
groundfish surveys normally deal with station
numbers in excess of 100, an heuristic method of
solving the problern was selected. Lin and Ker-
nighan ( 1973) at the Bell Telephone Laboratories
(BTL) developed an approximate procedure for
solving traveling salesman problems with large
number of visitation points which appeared
applicable to cruise track optimization. ^ The Na-
tional Fisheries Engineering Laboratory obtained
'Contribution No. 78-19F from the Southeast Fisheries
Center, National Marine Fisheries Service, NOAA, NSTL Sta-
tion, MS 39529. MARMAP Contribution No. 154.
^To develop a feeling for the complexity of these problems, it
should be noted that for a given number of stations, n, there are
706
a Fortran program from BTL and converted it to
operate on a Univac^ 1108 system at the National
Aeronautics and Space Administration Computer
Complex, Slidell, La.
Modifications to the BTL algorithm were made
to satisfy requirements of the groundfish survey
program. Most internal modifications were fairly
general so that the program could be used for other
areas and purposes. Specifics of grid locations and
random selection requirements were stored on
magnetic tape in a separate master file. The pro-
gram, as presently configured, can handle up to
150 stations; however, 300 stations could be han-
dled using extended core storage.
Algorithm Description
Assume a number of stations (n) have been
selected, either randomly or specifically. There are
a total ofnin - l)/2 links between the n stations.
The object is to find an n -subset of these links such
that (a) each station is sampled exactly one time,
and (b) the total distance traveled is a minimum. A
sequence of links satisfying (a) is called a tour; if it
also satisfies (b), it is the optimum tour.
The optimization algorithm begins by comput-
ing all nin - l)/2 distances and storing them in a
matrix. A completely random tour is generated to
use as a starting point. An attempt is then made to
find two sets of links A' = .Vj , .V2 • • • ^'/,. o''<^ Y - v, ,
y.2 . . . V;, such that if the links in X are replaced
with the links in Y, the result gives a tour of a
shorter distance. This is done by identifying^ j and
yi as the "most-out-of-place" pair, setting them
aside, then proceeding with .t^ and y2, x^ and y^,
and so on.
A criterion is then used to determine how many
pairs of links are to be exchanged. This criterion
can be explained as follows: Let the length of.r, and
y, be dx, and dy,, and g, = dx, - dy^. This deter-
mines the gain (shorter distance) by exchanging x,
withy,. After examining a sequence of proposed
exchanges x^ , X2 ■ • . a:^ and y 1 , Vg . . . y^ with their
corresponding gains §1,^2 • • -Sk^ the actual value
of ^ that defines the number of sets to exchange is
the one for which §, + g2 + . . . +g^ is always zero
or negative. This indicates the solution is a local
(n - 1 ) factorial possible cruise tracks that satisfy the criterion of
sampling all stations once and returning to starting position
(e.g., if n = 101, the number of possible solutions is 9.3326 x
10'").
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
optimum based on the fact that if a sequence of
numbers has a positive sum, there is a cyclic per-
mutation of these numbers such that every partial
sum is positive. Hence, the algorithm looks for
sequences ofg/s whose partial sum is always posi-
tive, reducing the number of sequences that need
to be examined. This means that the value of ^,
which gives the number of links to be exchanged,
IS determined when G* = S g-^Q, i.e.
when the
partial sum of the gains fails to remain positive.
These links are then exchanged and the process of
selecting new links to be exchanged begins again
ati = 1. When all possibilities have been tried, the
tour length is recorded. The program generates a
new random initial tour and the entire process
begins again. Eleven distinct solutions are pro-
duced in this manner, and the tour with the short-
est length is considered the optimum solution.
Program operation can best be understood by a
simple example.
Assume that n stations are selected and a ran-
dom tour generated (Figure la5. The black dots
represent the stations and the circle represents
the random tour. Any station S 1 is selected and S2
is designated as an adjacent station in the tour.
The link connecting the two stations is designated
Figure 1. — Example of the algorithm operation (modified from
Lin and Kemighan 1973).
707
as .\\ as shown schematically in Figure lb. The
station closest to S., is designated as S3 and
Vj is the link joining S^ and S;^. The
link .\', is not permitted to be either of
the links already connected to S.,- The gain cri-
terion is then calculated as g^ = c/.v, - rA',. If
this is negative, S.2 is designated as the other
neighbor of S, in the tour. If^, is positive, S4 is
designated as one of the tour neighbors of S;j as
shown in Figure Ic. If y2 were chosen to join S^
with Si, the result would be a tour. The gain criter-
ion is then calculated asg? ^ dx2 - dy^. If^i + ^2
>0, the original tour could be improved by ex-
changing Xi and .V.2 with .v'l and y2, respectively.
This potential improvement, which results from
closing up the tour immediately (G'-' = g^ +^2'' is
then stored. Now S5 is chosen as the nearest
neighbor of S4, and _V2 is designated as the link
connecting the two stations. Station S5 is not per-
mitted to be either of the stations already con-
nected to S^. Figure Id shows there is only one
choice for station Se and the link X3 such that if Sg is
connected toSj, a tour remains. If S^ were chosen
as the other neighbor of S5 in the original tour,
closing up Sg to Sj would result in a tour of two
disconnected pieces (Figure le). The gain as-
sociated with closing up immediately (connecting
Sg with S, ) is then compared with that obtained by
joiningS4toS] (G* ). The link connecting Sg to S, is
designated asy j. The gain criterion is then calcu-
lated as §3 = dx-f, - dy:i. Ifg^ + g-z + gs^G'' G*, however, a new station S^ and link .V4 are
selected and the process is continued.
A limited backtracking feature of the program
is included for the case when G- = 0 (i.e., no im-
provement can be made). The link ^2 was chosen
( Figure Id ) to join S-, to S4 as the closest station to
S4. When no improvement is made at some stage
(G* = 0), new links y2 are considered in order of
increasing length to a maximum of five choices. If
still no improvement is found, the fiveV] links are
examined in order of increasing length. When G"
cannot be improved, and the valued determined, a
new initial station S, is selected and the process
repeated. The procedure ends when all n stations
have been examined. A new random tour is gener-
ated, and each station is examined as an Sj again
in the same manner. This limited backtracking
significantly increases program effectiveness.
The computational procedure has other features
that improve the calculations and reduce running
time; such as limited foresight to the next links to
be broken, allowance for nonsequential link ex-
changes, and elimination from computation of
those links previously recorded in good tours. For
a more complete description of the algorithm, see
Lin and Kernighan (1973).
Results
Station Description
Separate station grids lere used for areas east
and west of the Mississippi River Delta. A station
consisted of a rectangle, lat. 2'30" by long. 2'30",
within which three trawl tows were made. Sta-
tions were identified and located at the center
point of the rectangle.
The station grid for the West Delta area con-
sisted of an area extending from long. 89°30'W to
91°30'W (Figure 2). The station gind for the East
Delta area consisted of a primary and secondary
zone extending between long. 88°00'W and
89^^30'W and long. 79°30'W and 88"00'W (Figure
3). Each area was limited by the 9.2-m (5-fm) and
92-m (50-fm) depth contours. Stations were
excluded from random selection in both areas be-
cause of navigation and trawling hazards, and
areas of known low groundfish densities.
Random Selection
Station number, latitude, and longitude were
stored in a master grid file for each area. Input to
station selection for the West Delta region was the
number of stations to be sampled. This region had
780 stations. For the East Delta area, the number
of stations must be specified separately for the
primary and the secondary zones — there were 555
stations in the primary zone and 139 in the sec-
ondary region. Station selection was performed by
a random number generator which selected sta-
tions based on the number required for each area.
Crujse Track Optimizatit)n
Requirements for an optimized cruise track
were different for the areas east and west of the
delta. A round-trip track was desired for the West
Delta area, while a one-way calculation was de-
sired for the East Delta area. The latter consisted
of the shortest route from a designated starting
point near Pascagoula. Miss., through each
selected station and ending at a point near the
mouth of the Mississippi River.
708
Figure 2.— Master station grid for groundfish survey sampling in northern Gulf of Mexico - West Delta area. Dot labeled 00-00 is start
and end point for round-trip cruise track optimization.
Since the grid used in the calculations was
square, a coefficient was included to account for
differences in absolute distance for one unit of
longitude vs. one unit of latitude. The coefficient
used for optimizing groundfish survey tracks is
52.10/59.85, which is the ratio of the distance in
nautical miles for 1° of longitude to that for 1° of
latitude at lat. 30°N. All longitudinal Cartesian
coordinate distances were multiplied by this
coefficient before calculations began.
For the West Delta area, the cruise track was
optimized from a point located just east of the
primary survey area (Pascagoula station number
00-00) through all randomly selected stations, re-
turning to the starting point. The optimization
program computed 11 solutions and the best route
in terms of the shortest distance was selected.
Output consisted of a listing of stations in proper
sampling order, and a plot of the stations and
optimum cruise track with every fifth station
labeled.
The starting point of the cruise track was south
of Pascagoula for the East Delta area. Optimiza-
tion was done for a cruise track that visited all
709
randomly selected primary and secondary stations
and terminated at a point near the Mississippi
River Delta designated 99-99 (Figure 3). Outputs
were the same as for the West Delta except for
treatment of the stations randomly selected which
appear in blocks 45, 46, 47, and 48. These were not
included in the optimized cruise track, but were
listed at the end of the optimized cruise track list-
ing. The stations in these blocks were added to the
end of the optimized cruise track and plotted as
individual points labeled with their Pascagoula
number.
Figure 3. — Master station grid for groundfish survey sampling in northern Gulf of Mexico - East Delta area. Dot labeled 00-00 is
starting point and dot labeled 99-99 is end point for one-way cruise track optimization. Primary and secondary areas are indicated by
arrows at top of figure.
710
30.0
29.5
29.0
28.5
28.0
92.0 91.5
WEST ORIGINAL TOUR
91.0
90.5
90.0
89.5
89.0
Figure 4. — Actual cruise track followed for West Delta area, FRV Oregon II cruise 55. Every fifth station is labeled.
30.0
29.5
29.0
28.5
28.0
92.0 91.5
WEST TOUR 10
91.0 90.5 90.0 89.5 89.0
Figure 5.— Optimized cruise track for West Delta area, FRV Oregon II cruise 55. Ever>- fifth station is labeled.
711
Test Case and Sample Products
The optimization program was tested to com-
pare computational results with a cruise track
actually fol lowed during a survey — FR V Oregon II
cruise 55, 5-29 November 1974.
West Delta
The 126 stations sampled during cruise 55
for the West Delta area were entered in the order
they were sampled (Figure 4), and the total dis-
tance (in grid units) was calculated to be 254. Each
grid unit was equivalent to approximately
4.6 km; thus, the total distance was about 1,176
km.
Eleven computations were performed on these
stations by the optimization program, and a
minimum length of 233 grid units (approximately
1,078 km) occurred three times. It can be said with
confidence the optimum tour (Figure 5) rep-
resented an 8.3% improvement over the actual
cruise track. Distances were calculated from the
center of each subsquare; therefore, the actual dis-
30.5
30.0
•04
29.5
29.0
28.5
89.5
EAST ORIGINAL TOUR
89.0
88.5
88.0
87.5
Figure 6.— Actual cruise track followed for East Delta area, FRV Oregon II cruise 55. Every fifth station is labeled. Station numbers
listed at lower left are those not included in optimization calculations.
712
tance would be decreased by the vessel cutting
corners of the subsquares. Calculations for the 126
stations on the Univac 1108 system used about
60K of core storage and required 2 min of Central
Processing Unit (CPU) time.
East Delta
Cruise 55 was used to test the program for the
East Delta area also. Of 1 16 stations sampled, 105
were included in the computation of an optimum
one-way cruise track. The other 11 stations were
located in blocks 45, 46, 47, and 48. They were,
however, added to the end of the optimized listout,
plotted, and labeled on the cruise track plot. The
actual cruise track distance for the 105 stations
was 229 grid units (approximately 1,061 km)
(Figure 6). The optimized one-way path was calcu-
lated to be 216 grid units ( 1,000 km), an improve-
ment of 5.8'7f (Figure 7). Calculations for the 105
stations used 60K of core storage and required 66 s
of CPU time.
30.5
30.0
29.5
29.0
28.5
89.5
EAST TOUR 10
89.0
88.5
88.0
87.5
Figure 7. — Optimized cruise track for East Delta area, FRV Oregon II cruise 55. Every fifth station is labeled. Station numbers listed
at lower left are those not included in optimization calculations.
713
Discussion
The basic optimization program has the capabil-
ity and inherent versatility to be utilized for a
wide range of applications. The round-trip capa-
bility can be modified to a one-way path calcula-
tion as was done for the East Delta portion of the
groundfish survey by manipulating the distance
matrix. Cartesian integrity of the start-stop points
is kept intact but the distance between the two
stations is set equal to zero in the distance matrix.
The program then calculates the optimum tour as
if the start-stop points were very close together
when, in fact, they are not.
There is no requirement that distance be the
optimization parameter. Factors such as cost,
time, or suitable weighted combinations of other
variables could be used to compute a cruise track
considered optimum for specific user require-
ments. Also, there is no requirement that the prob-
lem be symmetric or Cartesian in nature. For
example, the distance (cost, time, etc.) in going
from station A to station B need not be equal to
that from station B to station A. Applications of
these characteristics and other distance matrix
manipulations include:
1) The "cost" in going from station to station in
the presence of strong currents, such as the
Gulf Stream, could be adjusted. "Downstream"
directions from station to station would be
given preferential status for computing the op-
timum cruise track.
2) In some situations, it may be desirable to group
selected stations to be sampled preferentially
as a subset or subsets of the total station pat-
tern. This might occur if certain sampling
areas had a higher priority than others because
of biological and/or environmental considera-
tions.
3) Actual curvilinear distances between stations
could be entered into the distance matrix when
sampling in areas near the coast. This would be
done for station pairs connected by a straight
line that passes across land.
4) If the number of stations exceeds the present
150 maximum allowable (300 with extended
core storage), and it is possible to divide them
into subgroups, the problem is limited only by
CPU restrictions.
Many variations of the optimum cruise track
theme could be solved with this program and the
requirements are usually unique to a particular
problem or investigation.'* The examples demon-
strate the types of problems that could be solved.
Simple problems, such as those solved for the
groundfish survey, can be improved about 7% over
manually produced cruise tracks.
Improvements obtained using the optimized
cruise track for the cited application are not
dramatic, but would be significant over a long time
period and/or extensive cruising distance. The
program eliminates selecting stations from ran-
dom number tables and hand plotting the cruise
track, which may require several man-days
Literature Cited
Oper.
BELLMORE, M., AND G. L. NEMHAUSER.
1968. The traveling salesman problem: A survey.
Res. 16:538-558.
Lin, S., and B. W. Kernighan.
1973. An effective heuristic algorithm for the traveling-
salesman problem. Oper. Res. 21:498-516.
Thomas D. Leming
hillman j. holley
Southeast Fisheries Center
National Fisheries Engineering Laboratory
National Marine Fisheries Service, NOAA
NSTL Station, MS 39529
■^Inquiries regarding possible uses and applications of this
system should be directed to the Director, Southeast Fisheries
Center National Fisheries Engineering Laboratory, National
Space Technology Laboratories, NSTL Station, MS 39529.
714
Notices
NOAA Technical Reports NMFS published during the first 6 mo of 1978.
Circulars
409. Marine flora and fauna of the northeastern United
States. Copepoda: cyclopoids parasitic on fishes. By
Ju-Shey Ho. February 1978. iii + 12 p.. 17 fig.
410. The 1976 Cerotium tripos bloom in the New York
Bight: causes and consequences. By Thomas C.
Malone. May 1978. iv + 14 p. 17 fig., 1 table.
411. Systematics and biology of the tilefishes (Per-
ciformes: Branchiostegidae and Malacanthidael, with
descriptions of two new species. By James K. Dooley.
April 1978, v + 78 p., 44 fig., 26 tables.
412. Synopsis of biological data on the red porgy, Pa-
grus pagrus (Linnaeus). By Charles S. Manooch III
and William W. Hassler. May 1978, iii + 19 p., 12 fig., 7
tables. Also FAO Fisheries Synopsis No. 1 16. For sale
by the Superintendent of Documents. U.S. Govern-
ment Printing Office. Washington, DC 20402 Stock
No. 003-017-00418-0.
413. Marine flora and fauna of the northeastern United
States. Crustacea: Branchiura. By Roger F. Cressey.
May 1978, iii -i- 10 p., 15 fig. For sale by the Superin-
tendent of Documents. U.S. Government Printing
Office, Washington, DC 20402 Stock No. 003-017-
00419-8.
Special Scientific Report — Fisheries
719. Seasonal description of winds and surface and bot-
tom salinities and temperatures in the northern Gulf
of Mexico, October 1972 to January 1976. By Perry A.
Thompson, Jr. and Thomas D. Leming. February
1978, iv -t- 44 p., 43 fig., 2 tables. For sale by the
Superintendent of Documents, U.S. Government
Printing Office, Washington, DC 20402 Stock No.
003-017-00414-7.
720. Sea surface temperature distributions obtained
off San Diego, California, using an airborne infrared
radiometer. By James L. Squire, Jr. March 1978, iii -i-
30 p., 15 fig.. 1 table, 90 app. fig. For sale by the
Superintendent of Documents, U.S. Government
Printing Office, Washington, DC 20402 Stock No.
003-017-00415-5.
721. National Marine Fisheries Service survey of trace
elements in the fishery resource. By R. A. Hall, E. G.
Zook, and G. M. Meaburn. March 1978, iii + 313 p., 5
tables, 3 app. fig., 1 app. table.
722. Gulf menhaden, fireroor//a patron ;/s, purse seine
fishery: catch, fishing activity, and age and size com-
position, 1964-73. By William R. Nicholson. March
1978, iii -H 8 p., 1 fig., 12 tables.
723. Ichthyoplankton composition and plankton vol-
umes from inland coastal waters of southeastern
Alaska, April-November 1972. By Chester R. Mattson
and Bruce L. Wing. April 1978, iii + 11 p., 1 fig., 4
tables, 1 app. table.
724. Estimated average daily instantaneous numbers
of recreational and commercial fishermen and boaters
in the St. Andrew Bay system, Florida, and adjacent
coastal waters, 1973. By Doyle F. Sutherland. May
1978, iv + 23 p., 31 fig., 11 tables.
725. Seasonal bottom-water temperature trends in the
Gulf of Maine and on Georges Bank, 1963-75. By Clar-
ence W. Davis. May 1978, iv + 17 p., 22 fig., 5 tables.
NOAA Technical Reports NMFS are available free in limited numbers to Federal and State government agencies.
They are also available in exchange for other scientific and technical publications in the marine sciences. Individual
copies, if available, may be obtained by purchase from the Superintendent of Documents or by writing to User Services
Branch (D822), Environmental Science Information Center, NOAA, Rockville, MD 20852.
Data on fisheries subjects accessioned through NMFS by NODC during the first 6 mo of 1978.
Drift bottle, northwestern Gulf of Mexico, February
1962 to December 1963. Fifty data sheets in manuscript
form. By R. F. Temple and John R. Martin, Gulf
Fisheries Center, NMFS. Ref: NAPIS 78-0035.
This material is available from the National Oceanographic Data Center (D7514), National Oceanic and Atmospheric
Administration, Washington, DC 20235.
715
ERRATA
Fishery Bulletin. Vol. 76, No. 2
Fletcher, R. Ian., "Time-dependent solutions and efficient parameters for stock-production models," p.
377-388.
1) Page 377, right column, line 4, correct line to read:
growth rate/? and B^^., Graham's formula for latent
2) Page 378, left column. Equation (2), correct equation to read:
P{B) = c^B + c^B\ (2)
3) Page 378, left column, line 14, correct line to read:
antecedents of this analysis appear there.
4) Page 378, right column, line 7, correct line to read:
by average effort f on the assumption that F =
5) Page 379, left column, the equation that immediately follows Equation (la), correct equation to
read:
k =
Am
'b~
Br
6) Page 381, right column, line 4, correct line to read:
Equation (6), B = 0,F =Fj and B =B,. If we now
7) Page 381, right column, line 28, correct line to read:
F = 2mlB~^\ stock size Bft) -^ p (p being
8) Page 383, right column. Equation (15), correct equation to read:
B = ym
~ B
— ym
B~
FB. (15)
9) Page 383, Figure 5, caption under right figure, line 3, correct line to read:
Pniax in Equation (12)].
10) Page 385, left column, line 14, correct line to read:
then Bit)-^p and Y-*m, irrespective of initial con-
716
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 affiliation, and
i 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 Publi-
cation No. 6, A List of Common and Scientific
Names of Fishes from the United States and
Canada, Third Edition, 1970.
Text footnotes should be typed separately
from the text.
Figures and tables, with their legends and
headings, should be self-explanatory, not requir-
ing reference to the text. Their placement should
be indicated in the right-hand margin of the
manuscript.
Preferably figures should be reduced by photog-
raphy 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. Photo-
graphs should be printed on glossy paper.
Do not send original drawings to the Scientific
Editor; if they, rather than the photographic re-
ductions, are needed 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 re-
ferred 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 literature cited.
The accuracy of the literature cited is the re-
sponsibility of the author. Abbreviations of names
of periodicals and serials should conform to Bio-
logical Abstracts List of Serials with Title Abbrevi-
ations. (Chemical Abstracts also uses this system,
which was developed by the American Standards
Association.)
Common abbreviations and symbols, such as
mm, m, g, ml, mg, °C (for Celsius), %, %o and so
forth, should be used. Abbreviate units of mea-
sure 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.
FORM OF THE MANUSCRIPT
The original of the manuscript should be typed,
double-spaced, on white bond pap)er. Please triple
space above headings. We would rather receive
good duplicated copies of manuscripts than car-
bon copies. The sequence of the material should
be:
TITLE PAGE
ABSTRACT
TEXT
LITERATURE CITED
APPENDIX
TEXT FOOTNOTES
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. Jay C. Quast, Scientific Editor
Fishery Bulletin
Northwest and Alaska Fisheries Center
Auke Bay Laboratory,
National Marine Fisheries Service, NOAA
P.O. Box 155, Auke Bay, AK 99821
Fifty separates will be supplied to an author
free of charge and 100 supplied to his organiza-
tion. No covers will be supplied.
Contents-continued
JEWETT, STEPHEN C. Summer food of the Pacific cod, Gadus macrocephalus , near
Kodiak Island, Alaska 700
LEMING, THOMAS D., and HILLMAN J. HOLLEY. A computer software system
for optimizing survey cruise tracks 706
Notices
NOAA Technical Reports NMFS published during the first 6 mo of 1978 715
Data on fisheries subjects accessioned through NMFS by NODC during the first 6 mo
of 1978 715
■is GPO 796-049
^
V
Fishery Bulletin
^o
^^ATES O^ ^
Vol. 76, No. 4
SINDERMANN, CARL J. Pollution-associated diseases and abnormalties of fish
and shellfish: a review 717
FROST, BRUCE W., AND LAWRENCE E. McCRONE. Vertical distribution, diel
vertical migration, and abundance of some mesopelagic fishes in the eastern subarc-
tic Pacific Ocean in summer 751
SMITH, T. D., and T. POLACHECK. Analysis of a simple model for estimating
historical population sizes 771
GORE, ROBERT H. Larval development of Galathea rostrata under laboratory
conditions, with a discussion of larval development in the Galatheidae (Crustacea
Anomura) 781
LENARZ, WILLIAM H., and JAMES R. ZWEIFEL. A theoretical examination of
some aspects of the interaction between longline and surface fisheries for yellowfin
tuna, Thunnus albacares 807
PARRACK, MICHAEL L. Aspects of brown shrimp, Penaeus aztecus, growth in the
northern Gulf of Mexico 827
EHRLICH, KARL F., J. MYRON HOOD, GERALD MUSZYNSKI, and GERALD E.
McGOWEN. Thermal behavioral responses of selected California littoral fishes 837
BROWN, B. E., J. A. BRENNAN, and J. E. PALMER. Linear programming simula-
tions of the effects of bycatch on the management of mixed species fisheries off the
northeastern coast of the United States 851
Notes
ROBERTS, JOHN L., and JEFFREY B. GRAHAM. Effect of swimming speed on the
excess temperatures and activities of heart and red and white muscles in the
mackerel. Scomber japonicus 861
EHRLICH, KARL F., JOHN S. STEPHENS, GERALD MUSZYNSKI, and J. MYRON
HOOD. Thermal behavioral responses of the speckled sanddab, Citharichthys
stigmaeus: laboratory and field investigations 867
SHULTZ, CYNTHIA D., and BERNARD M. ITO. Mercury and selenium in blue
marlin, Makaira nigricans, from the Hawaiian Islands 872
PERSCHBACHER, PETER W., and FRANK J. SCHWARTZ. Recent records of
Callinectes danae and Callinectes marginatus (Decapoda: Portunidae) from North
Carolina with environmental notes 879
(Continued on back cover)
Seattle, Washington
U.S. DEPARTMENT OF COMMERCE
Juanita M. Kreps, Secretary
NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION
Richard A. Frank, Administrator
Terry L. Leitzell, Assistant Administrator for Fislieries
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 docimients 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.
EDITOR
Dr. Jay C. Quast
Scientific Editor, Fishery Bulletin
Northwest and Alaska Fisheries Center
Auke Bay Laboratory
National Marine Fisheries Service, NOAA
P.O. Box 155. Auke Bay, AK 99821
Editorial Committee
Dr. Elbert H. Ahlstrom Dr. Merton C. Ingham
National Marine Fisheries Service National Marine Fisheries Service
Dr. Bruce B. Collette Dr. Reuben Lasker
National Marine Fisheries Service National Marine Fisheries Service
Dr. Edward D. Houde Dr. Jerome J. Pella
University of Miami National Marine Fisheries Service
Dr. Sally L. Richardson
Gulf Coast Research Laboratory
Kiyoshi G. Fukano, Managing Editor
The Fishery Bulletin is published quarterly by Scientific Publications Office. National tVIanne Fisheries Service. NOAA. Room 450,
1107 NE 45th Street, Seattle. WA 98105. Controlled arculation postage paid at Tacoma, Wash.
Although the contents have not been copyrighted and may be repnnted freely, 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 'aw of this Department Use of funds for printing of this periodical has tjeen approved by the Director of the
Office of Management and Budget through 31 December 1978.
Fishery Bulletin
CONTENTS
Vol. 76, No. 4
SINDERMANN, CARL J. Pollution-associated diseases and abnormalties of fish
and shellfish: a review 717 ^^
FROST, BRUCE W., AND LAWRENCE E. McCRONE. Vertical distribution, diel
vertical migration, and abundance of some mesopelagic fishes in the eastern subarc-
tic Pacific Ocean in summer 751
SMITH, T. D., and T. POLACHECK. Analysis of a simple model for estimating
historical population sizes 771 -Vr~
GORE, ROBERT H. Larval development of Galathea rostrata under laboratory
conditions, with a discussion of larval development in the Galatheidae (Crustacea
Anomura) 781 "^
LENARZ, WILLIAM H., and JAMES R. ZWEIFEL. A theoretical examination of
some aspects of the interaction between longline and surface fisheries for yellowfin
tuna, Thunnus albacares 807
PARRACK, MICHAEL L. Aspects of brown shrimp, Penaeus aztecus, growth in the
northern Gulf of Mexico 827 -V
EHRLICH, KARL F., J. MYRON HOOD, GERALD MUSZYNSKI, and GERALD E.
McGOWEN. Thermal behavioral responses of selected California littoral fishes 837
BROWN, B. E., J. A. BRENNAN, and J. E. PALMER. Linear programming simula-
tions of the effects of bycatch on the management of mixed species fisheries off the
northeastern coast of the United States 851
Notes
ROBERTS, JOHN L., and JEFFREY B. GRAHAM. Effect of swimming speed on the
excess temperatures and activities of heart and red and white muscles in the
mackerel. Scomber Japonicus 861
EHRLICH, KARL F. , JOHN S. STEPHENS, GERALD MUSZYNSKI, and J. MYRON
HOOD. Thermal behavioral responses of the speckled sanddab, Citharichthys
stigmaeus: laboratory and field investigations 867
SHULTZ, CYNTHIA D., and BERNARD M. ITO. Mercury and selenium in blue
marlin, Makaira nigricans, from the Hawaiian Islands 872
PERSCHBACHER, PETER W., and FRANK J. SCHWARTZ. Recent records of
Callinectes danae and Callinectes marginatus (Decapoda: Portunidae) from North
Carolina with environmental notes 879
(Continued on next page)
Seattle, Washington
1979
For sale by the Superintendent of Documents. U.S. Government Printing Office. Washington.
DC 20402— Subscription price per year: $12.00 domestic and $15.00 foreign. Cost per single
issue: $3.00 domestic and $3.75 foreign.
Contents-continued
STOUT, VIRGINIA F., and F. LEE BEEZHOLD. Analysis of chlorinated hydrocar-
bon pollutants: a simplified extraction and cleanup procedure for fishery products . 880
RENSEL, JOHN E., and EARL F. PRENTICE. Growth of juvenile spot prawn,
Pandalus platyceros, in the laboratory and in net pens using different diets .... 886
LAURENCE, GEOFFREY C. Larval length-weight relations for seven species of
northwest Atlantic fishes reared in the laboratory 890
CRADDOCK, DONOVAN R. Effect of thermal increases of short duration on sur-
vival of Euphausia pacifica 895
MAY, ROBERT C, GERALD S. AKIYAMA, and MICHAEL T. SANTERRE. Lunar
spawning of the threadfin, Polydactylus sexfilis, in Hawaii 900
PEEBLES, JOHN B. The roles of prior residence and relative size in competition for
shelter by the Malaysian prawn, Macrobrachium rosenbergii 905
COLTON, JOHN B., JR., WALLACE G. SMITH, ARTHUR W. KENDALL, JR.,
PETER L. BERRIEN, and MICHAEL P. FAHAY. Principal spawning areas and
times of marine fishes. Cape Sable to Cape Hatteras 911
WADE, LAWRENCE S., and GARY L. FRIEDRICHSEN. Recent sightings of the
blue whale, Balenoptera musculus, in the northeastern tropical Pacific 915
BAGLIN, RAYMOND E., JR. Sex composition, length- weight relationship, and
reproduction of the white marlin, Tetrapturus albidus, in the western North Atlan-
tic Ocean 919
BECKER, C. DALE, and DENNIS D. DAUBLE. Records of piscivorus leeches
(Hirudinea) from the central Columbia River, Washington State 926
SMIGIELSKI, ALPHONSE S. Induced spawning and larval rearing of the yellow-
tail flounder, Limanda ferruginea 931
YOUNG, DAVID R., and TSU-KAI JAN. Trace metal contamination of the rock
scallop, Hinnites giganteus , near a large southern California municipal outfall . 936
INDEX, VOLUME 76 941
Vol. 76, No. 3 was published on 16 November 1978.
The National Marine Fisheries Service (NMFS) does not approve, rec-
ommend or endorse 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 pro-
motion 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.
POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
OF FISH AND SHELLFISH: A REVIEW
Carl J. Sindermann'
ABSTRACT
The relationship of disease and environmental stress is becoming increasingly well established with
time. Human activities — particularly those that result in chemical additions to the coastal/estuarine
environment — have increased the potential stresses on fish and shellfish inhabiting those areas.
Circumstantial evidence for associations of pollutants with certain fish and shellfish diseases and
abnormalities is accumulating.
This paper attempts to review and evaluate existing information about associations of diseases and
marine environmental degradation. Emphasis has been placed on: diseases caused by contaminant
stress and related facultative pathogens; stress-provoked latent infections; environmentally induced
abnormalities; genetic abnormalities associated with mutagenic and other properties of contaminants;
experimentally induced lesions; contaminant effects on resistance and immune responses; and
pollutant-parasite interactions.
There are several diseases, particularly fin erosion and ulcers in fish and shell disease in crustaceans,
for which a relationship with pollution seems evident, and there are a number of other diseases or
abnormalities (such as certain neoplasms and skeletal anomalies) for which a relationship with
pollution is indicated. Furthermore, there is some evidence that certain latent viral infections may be
provoked into patency by environmental stress.
Disease is a constant concomitant of life for any
species, normally removing individuals from the
population continuously. Marine animals are, of
course, subject to a wide spectrum of diseases of
infectious or noninfectious etiology ("disease" can
be defined in the broad sense as "any departure
from normal structure or function of an animal" or
as "the end result of interaction between a noxious
stimulus and a biological system").
Disfunction and death due to the activity of in-
fectious agents constitute the narrower, but often
predominant concept of disease. Infectious
diseases — caused by viruses, bacteria, fungi, pro-
tozoa, and other pathogenic organisms — are usu-
ally prime suspects in searches for causes of mor-
talities, often to the exclusion of other possible
causes. Noninfectious diseases include such
phenomena as environmentally induced skeletal
anomalies, genetic abnormalities, physiological
malfunctions caused by chemical environmental
factors, metabolic disorders resulting from nutri-
tional deficiencies, many forms of neoplasia, and a
host of others (Sparks 1972). In many instances, it
is probably the combination of an infectious agent
'Northeast Fisheries Center Sandy Hook Laboratory, Na-
tional Marine Fisheries Service, NOAA, Highlands, NJ 07732.
and environmental stress that eventually causes
mortality.
The distinction between "infection" and "dis-
ease" must be kept in mind. Most organisms are
constantly hosts to potentially pathogenic micro-
organisms, but disease results from imbalance of
the interactive system which includes virulence of
the pathogen, resistance of the host, and effects of
environmental stresses.
Infectious disease usually exists in an enzootic
form, weakening or disabling individuals and
rendering them more susceptible to predators or
other environmental stresses. Occasionally,
though, epizootics and mortalities comparable to
the great plagues of the Middle Ages may sweep
through animal populations. In marine species we
have seen such massive epizootics result in the
great herring mortalities of the mid-1950's in the
Gulf of Saint Lawrence (Sindermann 1958), and
the extensive oyster mortalities of the 1960's in
the Middle Atlantic states (Sindermann 1968).
These epizootics are triggered by a complex
interplay of pathogen, environment, and host pop-
ulation. Considering only the environmental as-
pects of such outbreaks, any departure from nor-
mal conditions produces a degree of stress on the
population, and may contribute to an increase in
prevalence of a pathogen, or of facultative invad-
Manuscript accepted Mav 1978.
FISHERY BULLETIN; VOL. 76, NO. 4, 1979.
717
ers. Some of these environmental factors are dras-
tic changes in temperature, lack of adequate food,
or overcrowding. Resistance of the host animal to
the disease is, of course, intimately related to
these stresses (Snieszko 1974).
Environmental stresses have been implicated in
a number of fish and shellfish diseases, but are
difficult to quantify. Even a definition of stress can
be elusive. Selye ( 1950, 1952) defined stress as the
sum of all the physiological responses by which an
animal tries to maintain or reestablish a normal
metabolism in the face of a physical or chemical
force. Brett ( 1958) defined it as "A state produced
by any environmental or other factor which ex-
tends the adaptive responses of an animal beyond
the normal range, or which disturbs the normal
functioning to such an extent that, in either case,
the chances of survival are significantly reduced."
Another definition which identifies stress as the
product and not the cause of homeostatic change is
thatofEschetal. (1975): "Stress is the effect of any
force which tends to extend any homeostatic or
stabilizing process beyond its normal limit, at any
level of biological organization."
Human activity has introduced or has increased
environmental stresses for fish in estuarine and
coastal waters. We have, for instance, added pes-
ticides and other synthetic chemicals which can,
even in low concentrations, drastically affect the
physiology of fish and shellfish, and with which
the species may have had no previous evolution-
ary experience. We have added heavy organic
loads, in the form of sewage sludge and effluents,
which can produce anaerobic or low-oxygen envi-
ronments and which are often accompanied by
other contaminants such as heavy metals, that
can interfere with enzymes of the fish and the food
organisms they consume.
During the past decade, several diseases and
abnormalities offish and shellfish have been de-
scribed that seem associated with pollutant stres-
ses. These can be categorized and discussed as:
1. Diseases caused by contaminant stress and
related pathogens;
2. Stress-provoked latent infections;
3. Environmentally induced abnormalities;
4. Genetic abnormalities associated with
mutagenic and other properties of contam-
inants;
5. Experimentally induced lesions;
6. Contaminant effects on resistance and im-
mune response; and
FISHERY BULLETIN: VOL. 76, NO. 4
7. Pollutant-parasite interactions.
In the first and second categories a synergistic
activity of chemical contaminants (or other form of
pollutant stress) and an infectious agent seems to
be a plausible explanation for at least some of the
observed effects. In categories three and four, it is
sometimes difficult to determine conclusively
whether environmental contaminants act directly
on target tissues or biochemical pathways, or if the
genetic material is first affected, with subsequent
changes in structure and/or function.
During the past several years there have been
signs of increasing interest in relationships be-
tween marine fish and shellfish diseases and en-
vironmental pollution. Several conferences have
been held recently, including the 1974 Symposium
on Tumors in Aquatic Animals, held in Cork, Ire-
land; the 1975 Symposium on Sublethal Effects of
Pollution on Aquatic Organisms, held as part of
the 13th Pacific Science Congress in Vancouver,
B.C.; and the 1976 Conference on Aquatic Pollu-
tants and Biological Effects with Emphasis on
Neoplasia, held in New York. The amount of rel-
evant literature available for consideration
within the title "pollution-associated diseases and
abnormalities of fish and shellfish" is somewhat
overwhelming. Even the list of books containing
pertinent material is impressive (Dawe and
Harshbarger 1969; Snieszko 1970; Ruivo 1972;
Vernberg and Vernberg 1974; Koeman and Strik
1975; Ribelin and Migaki 1975; Dawe et al. 1976;
Lockwood 1976; Kraybill et al. 1977; Vernberg et
al. 1977). Additionally, significant recent reviews
have appeared, for example, Rosenthal and Alder-
dice (1976) and Mclntyre^.
This paper attempts to summarize the present
state of knowledge about possible associations of
fish and shellfish diseases (infectious and nonin-
fectious) with estuarine and coastal pollution.
Much of the evidence for such associations is still
circumstantial and is presented as such. The orig-
inal literature on this subject, as for any
pollution-related subject, is voluminous. The ref-
erences cited here constitute only a small but, I
hope, a representative fraction of the published
information available. It should also be pointed
out here that this paper does not consider
^Mclntyre, A. D. (Convenor). 1976. ICES working group on
pollution baseline and monitoring studies in the Oslo Commis-
sion and ICNAF areas. Report of the subgroup on the feasibility
of effects monitoring. Int. Counc. Explor. Sea, Doc. CM1976/
E:44, 36 p.
718
SINDERMANN: POLLUTION-ASSOCIATED DKSEASES AND ABNORMALITIES
physiological and behavioral disorders, which
might be included in a broad definition of disease.
Finally, in these introductory comments, it
should be noted that to make any firm association
of a disease with environmental pollution there
are several basic requirements: 1) knowledge of
the history of occurrence of the disease in a par-
ticular species in the geographic area of concern;
2) knowledge of the history of occurrence and
levels of particular pollutants in that area; 3) a
review of the biology, life history, and occurrence
of the disease in other areas, in other species, and
under different environmental conditions; 4) an
intensive baseline survey of the current disease
and pollution situation, with attention to statisti-
cal reliability of sampling; 5) laboratory and field
experimentation with the principal objective of
reproducing the disease by exposure to known
levels of contaminants; and 6) resurveys of the
disease and pollution levels over several years,
looking for changes or trends. As will become ap-
parent in this paper, these requirements have
been fully satisfied for few if any of the diseases
discussed.
DISEASES CAUSED BY
CONTAMINANT STRESS AND
RELATED FACULTATIVE PATHOGENS
Fin Erosion
Probably the best known but least understood
disease offish from polluted waters is a nonspecific
condition known as "fin rot" or "fin erosion" (Fig-
ures 1, 2), a syndrome which seems rather clearly
associated with degraded estuarine or coastal en-
vironments. Fin rot has been reported from the
New York Bight (Mahoney et al. 1973; Ziskowski
and Murchelano 1975; Murchelano 1975),
California (Young 1964; Southern California
Coastal Water Research Project-"^; Mearns and
Sherwood 1974), Puget Sound (Wellings et al.
1976), Biscayne Bay and Escambia Bay in Florida
(Couch 1974a; Sindermannetal. 1978), the Gulf of
Mexico ( Overstreet and Howse 1977 ), the Irish Sea
(Perkins et al. 1972), and the Japanese coast
(Nakai et al. 1973).
Fin rot seems to occur in at least two types: one
^Southern California Coastal Water Research Proj-
ect. 1973. The ecology of the Southern California Bight: Im-
plications for water quality management. Ref. No. SCCWRP
TR 104, El Segundo, Calif.
in bottom fish, where damage to fins seems site-
specific and related to direct contact with con-
taminated sediments, and another in pelagic
nearshore species, characterized by more
generalized erosion, but with predominant in-
volvement of the caudal fin.
Recent quantitative surveys along the Middle At-
lantic coast have disclosed high prevalence (up to
38*^ ) of fin rot in samples of trawled marine fishes
from the New York Bight. Thus far, 22 affected
species have been found. While bacteria of the
genera Vibrio, Aeromonas, and Pseudomonas
were frequently isolated from abnormal fish, a
definite bacterial etiology has not been estab-
lished. Fin rot disease was significantly more
abundant in the New York Bight Apex, the area of
greatest environmental damage, than in any com-
parable coastal area from Block Island, R.I., to
Cape Hatteras, N.C. (Murchelano and Ziskowski
1976). An association between high fin rot preva-
lence and high coliform counts in sediments is
emerging (Mahoney et al. 1973), as is an associa-
tion between high fin rot prevalences and high
heavy metal levels in sediments (Carmody et al.
1973). The disease signs can be produced experi-
mentally by exposure offish to polluted sediments.
Fin erosion has also been observed in striped bass,
Morone saxatilis, overwintering in heated efflu-
ents of power plants in the Middle Atlantic States.
The histopathology of fin erosion in winter
flounder, Pseudopleuronectes americanus, from
the New York Bight was examined by Murchelano
(1975). Significant descriptive findings were
epidermal hyperplasia accompanied by dermal
fibrosis, hyperemia, and hemorrhage. Bacterial in-
fections were not found, nor was pronounced in-
flammatory response. However, reference was
made to acute fin lesions seen in summer flounder,
Paralichthys dentatus, in which bacteria were
readily demonstrable. The absence of pronounced
inflammatory response in either species of floun-
der led Murchelano to suggest that the necrotic
process is not primarily microbial and that ac-
tivities of a chemical irritant may be involved.
Another histopathological and bacteriological
study of fin rot in winter flounder from Narragan-
sett Bay, R.I., by Levin et al. (1972) described
acute ulcerative lesions as well as fln erosion,
thought to be produced by Vibrio anguillarum.
Acute inflammatory response was observed, and
ulcerations were reproduced in fish exposed ex-
perimentally to V . anguillarum isolates. It is pos-
sible that several poorly defined disease entities or
719
FISHERY BULLETIN: VOL. 76. NO. 4
Figure l. — Site-specific fin erosion concentrated in the midportion of fins in winter flounder (anterior dorsal fin is folded over in this
picture). Note melanism in areas of erosion. (Photograph courtesy of J. O'Reilly, Northeast Fisheries Center Sandy Hook Laboratory,
NMFS, NOAA, Highlands, N.J.)
generalized disease signs (one of which is fin ero-
sion) may be responsible for the disparate nature
of histopathological findings in this report, as
compared with those of Murchelano ( 1975).
Fin rot, with associated mortalities, was re-
ported by Couch and Nimmo (1974b) in Atlantic
croaker, Micropogon undulatus, and spot, Leio-
stomus xanthurus, from Escambia Bay, Fla. The
disease syndrome and mortalities were observed
for several years during periods of high tempera-
ture and low dissolved oxygen. Escambia Bay has
been polluted by the PCB (polychlorinated
biphenyl), Aroclor^ 1254, for a number of years
(Duke et al. 1970).
Information from southern California (South-
ern California Coastal Water Research Project,
see footnote 3) also indicates an association of fin
rot with degraded habitats; relevant statements
are: "The incidence of fin erosion was high in areas
with high concentrations of waste water con-
stituents in the sediments . . . ." "Although there
is a definite association between fin erosion and
waste water discharges, the causal factors are un-
known." "Nearly half of the 72 species caught off
the Palos Verdes Peninsula were affected with
this syndrome" (eroded fins). It is interesting that
a histopathological study of fin erosion in Dover
sole, Microstomus pacificus, from the California
coast (Mearns and Sherwood 1974; Klontz and
Bendele^) produced findings similar to those of
Murchelano (1975) — hyperplasia, fibrosis, ab-
sence of inflammation, and absence of microbial
infection.
Some species either seem more resistant to fin
erosion or are exposed differentially to toxic sub-
stances in water or sediments. A recent study by
Wellings et al. (1976) in a heavily polluted arm of
Puget Sound (the Duwamish River) in which over
6,000 fish of 29 species were examined, disclosed
fin erosion only in starry flounder, Platichthys stel-
latus, and English sole, Parophrys vetulus. Av-
erage incidences were 8 and 0.5% respectively.
Histopathological findings were similar to those
for east coast and California flatfishes — epidermal
hyperplasia, fibrosis, resorption of fin rays, aggre-
gation of melanophores, mucus cell changes, and
absence of bacterial invasion. The authors de-
scribed briefly what may be highly relevant obser-
vations of liver pathology in starry flounder from
the area where fin erosion was common. His-
topathology included increased fat deposition in
hepatic cells, fibrosis, and vascular distension.
Recent Japanese publications have mentioned
fin erosion in fish from polluted bays. Nakai et al.
■"Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
^Klontz, G. W..andR. A.Bendele. 1973. Histopathological
analysis of fin erosion in southern California marine fishes.
Southern Calif Coastal Water Res. Proj., El Segundo, Calif,
Rep. TM203, 8 p.
720
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
Figure 2. — Generalized fin erosion in weakUsh, Cynoscion regalis, (above) and in summer flounder (belowl. Note that in the weakfish
the anal, caudal, and pelvic fins are eroded, while the dorsal fins are not usually damaged. In contrast the summer flounder shows
erosion of wide areas of the fin fringes.
( 1973) found that as many as 60^^ of all stargazer,
Uranoscopusjaponicus, sampled from Suruga Bay
had evidence of disintegration of caudal and pec-
toral fins. Six other species also had abnormal fins.
An increase in occurrence of fin erosion and
other epidermal lesions (ulcers and lymphocystis)
in flatfish from the Irish Sea since 1970 was re-
ported by Perkins et al. (1972). Fin damage,
unknown before 1970, was observed in
plaice, Pleuronectes platessa, and dab, Limanda
limanda, taking the form of erosion or total loss of
caudal and lateral fins. Ulcers were described that
"did not have the typical appearance of bacterial
ulcers . . . ." The authors pointed to ocean dump-
ing of toxic wastes, particularly of PCB's, as a
possible factor contributing to observed preva-
lences of epidermal lesions, but no clear relation-
ship was demonstrated. Another study conducted
in the Irish Sea in 1972 (Shelton and Wilson 1973)
did not identify fin erosion in plaice or dab, but did
find a low incidence of "healed fin damage (proba-
bly caused by previous capture and rejection or by
passage through the cod-end mesh)."
The possible role of environmental chemical
contamination in the etiology of fin erosion
emerges more clearly as additional studies are
reported. Fish from the New York Bight, reported
in studies by Mahoney et al. (1973), Murchelano
(1975), and Ziskowski and Murchelano (1975),
exist in a highly contaminated area, with chemi-
cals such as heavy metals and petroleum residues
in sediments far above background levels. In
California, McDermott and Sherwood*^ fount DDT
to be significantly higher in fish with fin erosion,
and PCB levels slightly higher in such fish than in
normal individuals. Both contaminants were sig-
nificantly higher in Palos Verdes fish than in fish
''McDermott, D. J., and J. Sherwood. 1975. Annual re-
port. Dep. Fish. Mar. Fish. Program, Coastal Water Res. Proj.,
El Segundo, Cahf , p. 37.
721
FISHERY BULLETIN: VOL. 76, NO. 4
from a distant control area ( Dana Point). Wellings
et al. (1976) found abnormally high concentra-
tions of PCB's in English sole and starry flounders
from the Duwamish River in Washington.
Several authors have postulated that fin erosion
in flatfish may be initiated by direct contact of
tissues with contaminated sediments. Mearns and
Sherwood (1974) and Sherwood and Mearns
(1977), for example, suggested that toxic sub-
stances (sulfides, heavy metals, chlorinated hy-
drocarbons, etc.) could remove or modify the pro-
tective mucus coat and expose epithelial tissues to
the chemicals. Sherwood and Bendele^ reported
that Dover sole from the California coast with
severe fin erosion produced much less mucus than
normal fish.
It seems quite likely that the "fin erosion" syn-
drome in fish includes chemical stress, possibly
acting on mucus and/or epithelium; stress result-
ing from marginal dissolved oxygen concentra-
tions, possibly enhanced by a sulfide-rich envi-
ronment; and secondary bacterial invasion in at
least some instances. Some recent experimental
information tends to support this hypothesis.
A series of experiments at the Gulf Breeze (Fla.)
Environmental Research Laboratory of the U.S.
Environmental Protection Agency, using the spot,
resulted in experimental production of fin rot dis-
ease following exposure to 3-5 ^tg/l of Aroclor 1254
(Couch 1974a). Mortalities of up to 80% were re-
ported.
Minchew and Yarbrough (1977) exposed Mugil
cephalus in brackish water ponds ( 12%o) to 4-5 ppm
crude oil and found that fin erosion developed in
most of the exposed fish within 6-8 days. Lesions
were often hemorrhagic, and a tentative Vibrio sp.
was isolated consistently from surfaces of diseased
fish, but was rarely found systemically. Fin regen-
eration characterized most experimental fish 2
mo after exposure. This experiment should be re-
peated and extended.
Experimental induction of fin erosion has fol-
lowed exposure to several other contaminant
chemicals. Chronic exposure of fingerling rainbow
trout, Salmo gairdneri, to lead caused a variety of
grossly visible abnormalities, including fin ero-
sion (Davies and Everhart^); and chronic exposure
of minnows (Phoxinus phoxinus) to zinc and cad-
■'Sherwood, M. J., andR. A. Bendele. 1975. Mucous produc-
tion in Dover sole. Annu. rep., Coastal Water Res. Proj., El
Segundo, Calif., p. 51.
8Davies, P. H.,andJ.H. Everhart. 1973. Effects of chemi-
cal variations in aquatic environments. III. Lead toxicity to
mium resulted in similar abnormalities
(Bengtsson 1974, 1975).
A recent report by Overstreet and Howse ( 1977)
pointed to fin erosion and other abnormalities as
indicators of gradually increasing pollution stress
on the Mississippi gulf coast. Among other disease
conditions noted by Overstreet and Howse was
"red sore," characterized by hemorrhagic lesions
beneath scales, occasional hyperplasia, and ac-
companying ciliate (Epistylis sp.) infestation of
the body surfaces. The authors indicated that red
sores now occur in many of the fish in some fresh-
water and low salinity areas of the gulf coast of
Mississippi, a striking similarity to recent obser-
vations in Biscayne Bay, Fla., where many fish of
many species now exhibit hemorrhagic lesions be-
neath the scales, a condition which was unknown
a decade ago (Sindermann 1976). Red sores and
associated mortalities have also been described by
Rogers (1970, 1972) and Esch et al. (1976) from
centrarchid fishes in freshwater reservoirs of the
southeastern United States. The disease condition
in freshwater seems clearly related to Epistylis
infestation, probably abetted by secondary bacte-
rial infections, particularly by Aeromonas, al-
though there is still some question about which
organism is the primary invader.
It seems likely that generalized disease signs,
such as fin rot and red sores (and probably other
epidermal lesions such as ulcerations, papillomas,
and lymphocystis), may be characteristic of fishes
resident in degraded habitats, where environmen-
tal stresses of toxic chemicals, low dissolved oxy-
gen, and high microbial populations exist. The
extent and nature of these external manifesta-
tions are probably variable with resistance of the
particular species and the extent and nature of
environmental degradation.
Ulcers
Next to fin erosion, probably the commonest ab-
normality reported from fish taken in polluted
waters can be identified as "ulceration of bacterial
etiology," even though precise bacterial etiology
has not been demonstrated in every case. Where
bacterial isolations have been made from ulcer-
ated tissue. Vibrio anguillarum has been by far
the most predominant organism, with pseudo-
monads and aeromonads in lesser abundance.
rainbow trout and testing application factor concept. EPA-
R3-73-011C, 80p.
722
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
The report on ulcerations and fin rot in winter
flounders from Narragansett Bay, by Levin et al.
(1972) has been mentioned in the previous section.
The acute ulcerative lesions were thought to be
caused by V. anguillarum infections, and the ul-
cerative phase was reproduced in fish exposed ex-
perimentally to cultured V. anguillarum isolates.
A more recent report by Robohm and Brown
(1977) described systemic bacterial infections and
ulcerative lesions of the tail and dorsal muscles in
summer flounder from Connecticut waters. A
highly pathogenic Vibrio sp. was isolated, and ex-
perimental infections were produced by sub-
cutaneous inoculation and by seeding holding
tanks with bacteria at levels of 360/ml. Ulcers at
the inoculation site and subcutaneous hemor-
rhages along the bases of fins characterized ex-
perimental infections (Figure 3). These observa-
tions resemble those of Levin et al. (1972) in
winter flounder.
Ulcerations, probably of bacterial etiology, have
been reported in fish of several species from the
Irish Sea. Perkins et al. (1972) and Shelton and
Wilson (1973) reported ulcers from European
flounders (Platichthys flesus), dab, and plaice.
Prevalences were low (1-49^) in most instances.
An "ulcer syndrome" in cod, Gadus morhua,
from Danish coastal waters has been studied for
several years and seems associated with localized
areas of severe pollution (Jensen and Larsen 1976,
1977; Larsen and Jensen 1977a, b). Vibrio anguil-
larum and anAeromonas species have been impli-
cated (S^rensen 1977).
Ulcerations or external lesions on fish may, of
course, have a number of causes other than bacte-
rial infection. They may be due to net damage or
other surface abrasions, or to predator attacks.
Some protozoa (Myxosporida and Microsporida)
can infect muscle or skin tissue and multiply to
produce gross cysts. These infections mature to
produce many characteristic microscopic spores,
and in the process the overlying epidermis may be
sloughed, producing ulcers with usually smooth
borders (Figure 4). However, it seems to be a
Figure 3. — Ulcers and fin erosion in summer flounder produced by experimental inoculation oiVibrio sp. (Photograph courtesy of
R. Robohm, Northea.st Fisheries Center Milford Laboratory, NMFS, NOAA, Milford, Conn.)
723
FISHERY BULLETIN; VOL. 76. NO. 4
Figure 4. — Ulcer with smooth margins in Atlantic herring, resulting from infection by the myxosporidan protozoan, Kudoa
clupeidae.
reasonable generalization that many of the infec-
tions that produce grossly visible ulcerations in
fish are bacterial, and are often due to pathogens of
the genera Vibrio, Pseudomonas, or Aeromonas
(Lamoletetal. 1976). Ulceration often begins with
scale loss or formation of small papules, followed
by sloughing of the skin, exposing the underlying
muscles, which may also be destroyed. Bacterial
ulcers may have rough or raised irregular mar-
gins, and will often be hemorrhagic. Ulcers may or
may not be associated with fin erosion.
Shell Disease of Crustacea
Also associated with badly degraded estuarine
and coastal waters is a disease condition in Crus-
tacea commonly referred to as "shell disease" or
"exoskeletal disease" or "shell erosion." This can
be considered in some ways as the invertebrate
counterpart of fin erosion.
Homarus americanus and rock crabs (Cancer
irroratus ) from grossly polluted areas of the New
York Bight were found to be abnormal, with ap-
pendage and gill erosion a most common sign, by
Young and Pearce (1975). Skeletal erosion occur-
red principally on the tips of the walking legs,
ventral sides of chelipeds, exoskeletal spines, gill
lamellae, and around areas of exoskeletal articu-
lation where contaminated sediments could ac-
cumulate. Gills of crabs and lobsters sampled at
the dump sites were usually clogged with detritus,
possessed a dark brown coating, contained
localized thickenings, and displayed areas of ero-
sion and necrosis. Similar disease signs were pro-
duced experimentally in animals held for 6 wk in
aquaria containing sediments from sewage sludge
or dredge spoil disposal sites. Initial discrete areas
of erosion became confluent, covering large areas
of the exoskeleton, and often parts of appendages
were lost. The chitinous covering of the gill fila-
ments was also eroded, and often the underlying
tissues became necrotic.
Dead and moribund crabs and lobsters have
been reported on several occasions by divers in the
New York Bight Apex, and dissolved oxygen con-
centrations near the bottom during the summer
often approach zero (Pearce 1972; Young 1973).
Low oxygen stress, when combined with gill foul-
ing, erosion, and necrosis, could readily lead to
mortality.
In a related study, Gopalan and Young (1975)
examined "shell disease" in the caridean shrimp,
Crangon septemspinosa, an estuarine and coastal
food chain organism common on the east coast of
North America and important in the diets of
bluefish, weakfish, flounders, sea bass, and other
economic species. Examinations of samples of
Crangon from the New York Bight disclosed high
prevalences (up to 15*7^ ) of eroded appendages and
blackened erosions of the exoskeleton. The disease
condition was only rarely observed at other col-
lecting sites (Beaufort, N.C., and Woods Hole,
Mass.). Histological examination of diseased
specimens produced findings similar to those of
724
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
Young and Pearce ( 1975) with crabs and lobsters.
All layers of the exoskeleton were eroded; affected
portions were brittle and easily fragmented;
cracking and pitting of calcified layers occurred;
and underlying tissues were often necrotic.
Laboratory experiments using seawater from the
highly polluted inner New York Bight resulted in
appearance of the disease in 509f of individuals.
Erosion was progressive, crippled individuals
were cannibalized, and eroded segments of append-
ages did not regenerate after ecdysis. No disease
signs developed in control animals held in arti-
ficial seawater.
A German study of the effects of industrial
wastes on the brown shrimp, Crangon crangon
(Schlotfeldt 1972), disclosed high prevalence of
so-called "black spot disease," with signs very
similar to those seen in C. septemspinosa from the
New York Bight. Juvenile and adult shrimps from
the Fohr Estuary had black areas of erosion on the
carapace and appendages, with necrosis of under-
lying tissues, and, frequently, missing terminal
segments of appendages. The disease condition
varied in prevalence seasonally, with a peak of
8.9^f in summer. Lesions persisted and worsened
after ecdysis, and experimental exposure to deter-
gent accelerated the course of the disease.
Shell disease of Crustacea has been observed in
many species and under many conditions, both
natural and artificial (Rosen 1970; Sindermann
1970). Actual shell erosion seems to involve activ-
ity of chitinoclastic bacteria, with subsequent
secondary infection of underlying tissue by facul-
tative pathogens. Initial preparation of the exo-
skeletal substrate by mechanical, chemical, or mi-
crobial action probably is significant; thus high
bacterial populations and the presence of contam-
inant chemicals in polluted environments, as well
as extensive detrital and epibiotic fouling of gills,
could combine to make shell disease a common
phenomenon and a significant mortality factor in
crustaceans inhabiting degraded environments.
There is much room for study in this cloudy
territory at the boundary between infectious and
noninfectious disese processes, as exemplified by
fin and shell erosion. This is the area where en-
vironmental stress and facultative microor-
ganisms exert their impacts; where high bacterial
populations in eutrophic waters interact with ex-
posed, or injured, or chemically modified surface
membranes; where epibiotic fouling organisms
can assume pathogenic roles; and where
nonspecific lesions such as fin rot and skeletal
erosions can occur in epizootic proportions.
Lymphocystis
While fin erosion, ulcers, and shell disease seem
to have reasonable associations with degraded en-
vironments, it is difficult to find additional good
examples in the category of "Diseases caused by
facultative pathogens." Probably the most likely
candidate (in an obviously poor field) would be
lymphocystis, a virus disease which causes ex-
treme hypertrophy of fibroblast cells in a large
number of freshwater and marine fishes, and
which has been postulated to be associated with
environmental stresses. Perkins et al. (1972)
found in a 1971 survey that three diseases —
lymphocystis, epidermal ulcers, and fin
erosion — were abundant in plaice and dab from
the Northeast Irish Sea. Lymphocystis infection
levels in individual trawl catches ranged from 0 to
2b^( in plaice and from 0 to 17^^ in dab. The au-
thors pointed out that the Irish Sea has been used
recently for dumping of toxic wastes, particularly
PCB's, but their concluding statement is
". . . there is insufficient evidence to be certain
whether the increased incidence of the diseases
noted in 1971 is the result of an outbreak of
epidemics of purely biological origin or if the
dumping of toxic wastes is responsible."
Another survey of lymphocystis in the Irish Sea,
this one in 1972, was reported by Shelton and
Wilson ( 1973). They found lymphocystis to be the
most abundant of observable pathological condi-
tions, with highest prevalence ( 14.6*^ ) in flounder,
Platichthys flesus, and lesser prevalences in other
flatfish (1.97f in plaice and 1.1% in dab). Unlike
Perkins et al. ( 1972), Shelton and Wilson consid-
ered recent pollution of the Northeast Irish Sea to
be the least likely explanation for high .levels of
lymphocystis — pointing out that the disease has
been known from that area for 70 yr, having been
described early in the century by Woodcock ( 1904)
and Johnstone ( 1905) from flounders taken in the
Irish Sea. Van Banning ( 197 1 ) studied lymphocys-
tis in North Sea plaice (Figure 5) and also con-
cluded that pollution was not a likely cause of high
prevalences.
A recent lymphocystis epizootic with over 50%
prevalence was reported from flatfish in the North
Sea by Mann (1970) and earlier epizootics have
occurred in Europe (Weissenberg 1965). Temple-
man (1965) reported an epizootic in American
725
FISHERY BULLETIN; VOL^ 76. NO. 4
Figure 5. — Lymphocystis in European plaice, Pleuronectes platessa. (Photograph courtesy of P. Van Banning, Rijkinstituut voor
Visserijonderzoek, IJmuiden, Netherlands.)
plaice, Hippogiossoides platessoides, from the
Grand Banks of Newfoundland. He suggested sev-
eral possible explanations for the outbreak, in-
cluding the possibility that the disease is enzootic
in the population and may increase in intensity
periodically. Earlier, Awerinzew (1911) found an-
nual lymphocystis prevalences of 11% in P. flesiis
from the Murmansk coast, and Nordenberg ( 1962)
found infections as high as 12% in the same species
from the Oresund, with some indication of higher
prevalence in the warmer months of the year.
None of these outbreaks seems to have any ap-
parent association with environmental contami-
nation.
Lymphocystis has been reported recently in
Baltic herring (C/;//xY/ harengus var. membras ) by
Aneer and Ljungberg (1976). Of the 2,629 indi-
viduals examined, 14 had gross signs of the dis-
ease. The authors pointed out that a number of
infections were slight and might easily have been
overlooked. It is quite likely that this is the case
with other species also.
The presence of lymphocystis cells in the viscera
of herring was noted by Aneer and Ljungberg, and
there are several other reports of systemic lym-
phocystis infections, particularly that of Dukes
and Lawler (1975) in which lymphocystis cells
were found in and behind the eyes and in the
kidney, spleen, liver, heart, ovaries, and mesen-
teries of silver perch, Bardiella chtysiira, from the
Mississippi coast.
Lymphocystis has also been recognized in 4.3%
ofyellowfin sole, Limanda aspera, sampled in the
Bering Sea by Alpers et al. (1977a) and in 68% of
winter flounder sampled in 1975 from Casco Bay
in the Gulf of Maine (Murchelano and Bridges
1976).
Despite inconclusive attempts to relate lym-
phocystis epizootics in flatfish to specific environ-
mental factors, including pollutants, there are re-
cent observations of the disease in fishes of the
Gulf of Mexico that reopen the issue. Christmas
and Howse (1970) found lymphocystis in Atlantic
croaker and sand seatrout, Cynoscion arenarius,
from the Mississippi coast of the Gulf of Mexico
and observed that "The pollution load was much
gi'eater in estuarine systems where lymphocystis
was encountered." However, only 12 infected fish
were found in a 10-mo trawling survey with
monthly collections at 35 stations, which is not
overwhelming evidence for a relationship of the
disease to pollution. In a later study, Edwards and
Overstreet (1976) reported marked increases in
lymphocystis incidences in Atlantic croakers from
the Mississippi coast, with as high as 50% infected
fish in some trawl collections. Increased preva-
lences of another strain of lymphocystis were also
observed in silver perch. In a later paper Over-
726
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
street and Howse (1977) stated that (with refer-
ence to the silver perch strain) "prevalence
appears to relate to rainfall, suggesting that toxi-
cants, salinity, or enriched water could play a
major role in infections."
Lymphocystis in striped bass, Morone sa.xatilis,
on the U.S. east coast seems to have some tenuous
association with heated effluents. Recent unpub-
lished observations by staff members of the Sandy
Hook Laboratory (J. S. Young, Fishery Biologist,
Northeast Fisheries Center Sandy Hook Labora-
tory, National Marine Fisheries Service, NOAA,
Highlands, NJ 07732. Pers. commun., September
1975), pointed to high prevalence of lymphocystis
disease (Figure 6) in limited samples of striped
bass overwintering in the heated effluent of a Long
Island generating station (Northport, N.Y.). This
disease is considered rare in striped bass
(Anonymous 1951; Krantz 1970), and its unusual
abundance in a localized population may well be
related to the abnormally high winter tempera-
ture regime in which the population exists, or to
abnormal crowding, with consequent increase in
stress and ease of transfer of the pathogen. The
high temperature may promote survival or trans-
fer of the pathogen, or lower resistance of the host,
or provoke latent infections into patency, result-
ing in grossly recognizable stages of infection.
Lymphocystis is considered to be highly infec-
tious; initial lesions often develop where injuries
to the fish have occurred; and lymphocystis virus
reaches peak infectivity when water temperatures
are high (Midlige and Malsberger 1968). Some or
all of these factors may be important in fostering
the high prevalences observed in striped bass. An
important concern about fish diseases such as
lymphocystis in populations overwintering in
heated effluents is that a focus of infection will be
provided for incoming spring migrants.
STRESS-PROVOKED LATENT
INFECTIONS
A number of microbial diseases offish have been
shown to be provoked into patency by environ-
mental stress (Wedemeyer 1970; Snieszko 1974).
This seems to be true for kidney disease and
furunculosis of salmonids, which often exist in
carrier or latent states that can develop into active
infections if fish are stressed. It is also probably
true for anaerobic bacterial (Eubacterium sp.) in-
fections of mullet and 10 other species offish from
Biscayne Bay (Udey et al. 1977). A report of vib-
riosis in eels held in freshwater (Reidsaether et al.
1977) suggested that latent infections with Vibrio
anguillarum produced disease and mortalities
when eels were exposed experimentally to 30-60
ixgl\ copper for 50 days in freshwater. Similarly an
epizootic oi Aeromonas liquefaciens (= A. hydro-
phila) in Atlantic salmon, Salmo salar, and the
sucker, Catostomus commersoni, in the Miramichi
River, Canada, seemed to be related to combined
stresses of copper and zinc pollution and high
water temperatures (Pippy and Hare 1969).
Figure 6. — Lymphocystis disease in striped bass from heated effluent of a power plant.
727
FISHERY BULLETIN: VOL. 76. NO 4
Snieszko ( 1962) stated, concerningA. liquefacicns
that ". . . fish may have latent infections that
flare up when the flsh are exposed to stress."
There are recent published accounts of two viral
diseases of marine invertebrates which also indi-
cate that latent infections may be provoked into
patency by environmental stress. One, a
Baculovirus infection of pink shrimp, Pcnaeus
duororum, was first recognized in stressed
laboratory populations (Couch 1974b, 1976). The
other, a herpes-like viral infection of oysters, was
discovered in a population held in a heated power
plant effluent in Maine (Farley et al. 1972).
An association of shrimp virus disease and low-
level chronic exposure to pollutant chemicals is
being explored at the Gulf Breeze Environmental
Research Laboratory of the U.S. Environmental
Protection Agency (Couch 1974a, 1978). In this
work a virus disease of pink shrimp caused by B.
penaei reached patent levels and caused mor-
talities of dO-SO'Vo in shrimp exposed to the PCB
Aroclor 1254 and to the organochlorine insecticide
Mirex (Couch and Nimmo 1974a, b; Couch 1974a,
b, 1976). Other experiments in which the shrimp
were crowded, but not exposed to chemicals, re-
sulted in similar enhancement of virus infections,
indicating that environmental stress may be an
important determinant of patent infections. The
virus infection has been found subsequently in
brown and white shrimp (Overstreet and Howse
1977; Couch 1978).
Couch and Courtney (1977) have recently pro-
posed an elaborate and unique conceptual scheme
to utilize the shrimp virus for interactive bioas-
says for chronic sublethal effects of contaminants.
The authors point out that there are a number of
possible interactions of host, pathogen, and chem-
ical stressors — change in resistance of shrimp to
the virus, enhancement of widespread latent in-
fections in the shrimp population, change in viru-
lence of the virus, and losses of diseased shrimps
by cannibalism. Criteria developed by Couch and
Courtney for interaction include increased viral
prevalence in stressed populations (as indicated
by numbers of inclusion bodies), increased infec-
tion intensity in stressed individuals, increased
mortality in stressed populations, and greater
cytopathic effects in infected and stressed indi-
viduals. The shrimp virus infection has great po-
tential for elucidating effects of pollutants on
host-pathogen relationships.
An association of high environmental tempera-
tures with high disease prevalence (or disease en-
hancement) in molluscan shellfish sampled from
thermal effluents has been made recently. Farley
et al. (1972) described a lethal herpes-type virus
disease of oysters held in heated discharge water
in Maine. The disease, which apparently existed
at a low enzootic level in oysters growing at nor-
mal low environmental temperatures (12°-18°C
summer temperatures), seemed to proliferate in
oysters maintained at elevated temperatures
(28°-30°C) and to produce mortalities in those
populations. Intranuclear inclusion bodies, con-
taining viral particles, characterized advanced in-
fections. Mortalities of oysters held at higher
temperatures were correlated with greater preva-
lence of the viral inclusions. Elevated water tem-
peratures were considered by the authors to favor
spread of the infection or to activate latent infec-
tions, or both.
This evidence for a possible role of environmen-
tal stress in activating latent viral infections could
hardly be termed overwhelming, since it is possi-
ble that new infections produce the effects discus-
sed. However, the two viral diseases may provide
an insight into the total effect of pollutant and
other environmental factors on disease prevalence
and disease-caused mortalities. The carrier state
is often difficult to diagnose, but it may play a
much larger role in the epizootiology of marine
disease than can be demonstrated at present.
ENVIRONMENTALLY INDUCED
ABNORMALITIES
Neoplasms (Tumors)
The terms "neoplasia" and "neoplasms," par-
ticularly as they concern lower animals, are
difficult to define precisely. The Oxford Dictionary
definition of neoplasm is "a new formation in some
part of the body; a tumor." More elaborate defini-
tions exist. Warren and Meissner (1971) defined a
neoplasm as "a disturbance of growth charac-
terized primarily by an unceasing, abnormal, and
excessive proliferation of cells." Prehn (1971)
defined neoplasia as "that form of hyperplasia
which is caused, at least in part, by an intrinsi-
cally heritable abnormality in the involved cells."
Although neoplasia has been studied most exten-
sively in humans and laboratory mammals, the
existence of tumors in fish and shellfish has been
recognized for almost a century (the first oyster
tumor, for example, was reported by Ryder in
728
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
1887, and Bonnet mentioned thyroid hyperplasia
in fish due to iodine deficiency in 1883).
Circumstantial evidence associating environ-
mental contamination with neoplasms (tumors) in
fish has accumulated from a number of studies:
1. Lucke and Schlumberger (1941) described
166 catfish (Ameiurus nebulosus) with
epitheliomas of lips and mouth, taken from
the Delaware and Schuylkill Rivers near
Philadelphia. The rivers were grossly pol-
luted. Tumors of this type may result from
mechanical, infectious, or chemical irrita-
tion. Catfish from other areas did not have a
high prevalence of tumors. The authors did
not exclude the possibility that the lesions
were induced by chemical carcinogens in the
water. The lesions developed into epidermoid
carcinomas, some of which were invasive.
2. Russell and Kotin (1957) found 10 of 353
white croakers, Genyonemus lineatus, from
Santa Monica Bay, Calif., with papillomas of
lips and mouth. Fish were taken 2 m from an
ocean outfall. No tumors were found in 1,116
croakers from unspecified nonpolluted wa-
ters 70 km away.
3. Cauliflower disease (epidermal papilloma)
has been increasing in prevalence in eels
(Anguilla anguilla) from the Baltic since
1957. The pattern of spread and high preva-
lence indicates an infectious process (viral
arrays have been seen) or progressive ac-
cumulation of industrial contaminants such
as fuel oil and smelter wastes (known to con-
tain carcinogenic hydrocarbons such as ben-
zopyrene and heavy metals such as arsenic).
4. Cooper and Keller ( 1969) reported that 12'7f
of nearly 16,000 English sole from San Fran-
cisco Bay had epidermal papillomas, with as
many as 33 tumors per fish. Incidence of
tumorous fish in the northern part of the Bay
was twice that in the southern part. The
greatest concentration of industrial waste
discharge, especially petrochemicals, existed
in the northern part of the Bay. A later sur-
vey (Kelly 1971) failed to confirm the areal
difference in tumor abundance.
5. Young (1964) found many small (10-15 cm)
Dover sole from Santa Monica Bay with
tumors. Fish above 15 cm did not have
tumors. According to Young, numerous
white croakers from Santa Monica and Los
Angeles-Long Beach were found with papil-
lomas of the lips, and papillomas were ob-
served on tongue soles, cusk eels, and Pacific
sanddabs. Such tumors were not seen by
Young on fish from unpolluted areas, but
Dover sole with epidermal papillomas have
since been collected off Baja California as far
south as Cedros Island (Sherwood and
Mearns 1976). The prevalence of lip tumors
in white croakers from Santa Monica and the
Palos Verdes shelf has been <¥'/< since 1970
(Mearns and Sherwood 1977).
6. Carlisle (1969) found "growths" frequent on
white croakers and Dover sole from Santa
Monica.
7. Sindermann (1976) found wartlike tumors
histologically resembling fibromas in Mugil
cephalus from Biscayne Bay in 1969-70 (Fig-
ure 7). Other fibrous tumors have been re-
ported since then by Lightner (1974) and
Edwards and Overstreet (1976) in mullet
from the Gulf of Mexico.
From the foregoing, it is apparent that much
attention has been given, and continues to be giv-
en; to the common occurrence of epidermal papil-
lomas in a number of Pacific flatfishes ( Wellings et
al. 1964, 1965; Wellings 1969a, b). The tumors of
English sole from the Pacific coast, for example,
have been studied for almost half a century (Pacis
1932; McArn et al. 1968; Good 1940; Angell et al.
1975). Stich et al. (1976) in their review offish
tumors and sublethal effects of pollutants, found
highest prevalences to occur in young-of the-year
fish. Maximum prevalences reported in the litera-
ture were 587c in English sole (Stich and Acton
1976); 55^7^ in starry flounder (McArn and Wel-
lings 1971); 15% in Ratheadsole, Hippoglossoides
elassodon (Miller and Wellings 1971); and over
40'7f in sand sole, Psettichthys melanostictus (Nig-
relli et al. 1965). A relationship of high frequen-
cies of such papillomas with coastal pollution is
still uncertain. Stich et al. (1976) stated "There
seems to be a higher skin tumor frequency among
English sole inhabiting areas of urban contamina-
tion (Vancouver) than among fish populations in
regions remote from human activities . . . ."
In an extension of this study, Stich et al. (1977)
reported prevalences of skin neoplasms in 1-yr-old
English sole of from 20 to 70% in samples taken
near eight cities on the Pacific coast, while preva-
lences did not exceed 0.17c in several samples
taken on the British Columbia coast more distant
from cities. However, Oishi et al. (1976) examin-
729
FISHERY BULLETIN: VOL. 76, NO. 4
Figure 7. — Wartlike fibrous tumors in Mugil cephalus from Biscayne Bay, Fla.
ing prevalences of similar epidermal papillomas
in flatfish from relatively unpolluted waters of
northern Japan felt that a possible association
existed between high tumor occurrence ( up to 20%
in certain samples) and parasitization of the flesh
by a nematode, Philometra marine, but then they
suggested that the involvement of naturally oc-
curring chemical contaminants as well as man-
made pollutants must be considered in the etiol-
ogy of flatfish neoplasms.
Wellings et al. (1977) found 1.0% of rock sole,
Lepidopsetta bilineata, sampled in the still-
unpolluted Bering Sea, with epidermal papil-
lomas. Infections were widely distributed geo-
graphically, mostly in older individuals. The age
distribution of infection was quite different from
that in Puget Sound flatfishes, where predomi-
nantly younger fish are involved.
The etiology of skin tumors in English sole from
the Pacific coast of North America was reviewed in
a recent paper by Angel et al. (1975), with the
conclusion that the cause is unknown, and may be
multifactorial. Three stages of tumorigenesis
were described in young-of-the-year English sole,
beginning with angioepithelial nodules and pro-
gressing to epidermal papillomas and an-
gioepithelial polyps. No conclusive role of an en-
vironmental carcinogen has been demonstrated;
there seem to be subpopulation differences in dis-
ease prevalences; and electron microscopy has dis-
closed the presence of viruslike particles in cells of
papillomatous fish (Wellings and Chuinard 1964),
but attempts to isolate a viral agent have been
unsuccessful.
To further complicate the story, an unknown
cell type, called an "X cell," has been found in all
three tumor stages in English sole. The cells may
be parasitic, as was suggested by Brooks et al.
(1969), and Alpers et al. (1977b), or they may be
transformed host cells, analogous to lymphocystis
cells, as was suggested by Angell et al. (1975).
Angell et al. concluded by stating that "given the
pervasiveness of certain pollutants, experimental
evidence and further field studies will be neces-
sary to clarify the relationship between tumorous
flatfishes and pollution."
Another observation on the possible relation-
ship of flatfish tumors and pollutants has been
supplied by Mearns and Sherwood (1974). The dis-
tribution and abundance of skin tumors and fin
730
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
erosion were studied simultaneously in Dover sole
from the California coast. Fin erosion was more
common in specimens collected near major sewer
outfalls, whereas tumorous fish were distributed
more evenly throughout southern California
coastal waters. The authors concluded that "The
spatial and temporal distribution of tumor-
bearing Dover sole suggest that initiation of the
disease was not related to [municipal] wastewater
discharges [in southern California]."
A recent report of neoplasms in the Atlantic
hagfish, Myxine glutinosa, by Falkmer et al. (1977)
suggested a possible relationship of PCB contami-
nation and tumor prevalence. During a 5-yr
(1972-76) study in Gullmar Fjord, Sweden, neo-
plasm prevalences, particularly hepatomas, de-
creased from 5.8 to 0.6'^ . PCB levels in livers of
hagfish were appreciable (5 ppm), but the use of
PCB was prohibited in 1971. Liver PCB levels in
hagfish caught inside the Qord were five times
higher than in those caught outside. However, the
association of PCB contamination with liver
tumors must be considered to be tenuous. Earlier
reports of neoplasms in hagfish (Fange etal. 1975;
Falkmer et al. 1976) described remarkably high
frequencies in Gullmar Fjord, but only low con-
centrations (0.5-1.0 ppm) of PCB in livers, and low
environmental levels of PCB and other contami-
nants.
The role of environmental chemical factors in
induction of neoplasms in shellfish is even less
clear than for fish, but there is some limited infor-
mation. Yevich and associates (Barry and Yevich
1975; Yevich and Barszcz 1976, 1977) have for a
number of years examined the occurrence of neo-
plastic growths in the soft-shell clam, Mya
arenaria, in relation to petroleum contamination.
Gonadal and hematopoietic neoplasms were ob-
served in animals collected from two chronically
contaminated sites on the Maine cost, with pre-
valences up to 29% in certain samples. Yevich and
Barszcz (1976) stated that "no tumors similar to
those described [from the petroleum contaminated
area] have been encountered in animals collected
from any other area." They described the scope of
their study as "several thousand animals from all
coastal areas of the United States." Additional
samples of clams from a number of other coastal
locations are needed, as is a more precise descrip-
tion and confirmation of the neoplastic condition.
It is interesting that a counterpart study of
soft-shell clams from Rhode Island and Mas-
sachusetts (Brown et al. 1976) reported occur-
rences of neoplasia, apparently of hematopoietic
origin, in up to 26% , with the highest frequency in
samples from a 1975 oil spill area near Bourne,
Mass. A later report (Brown et al. 1977) included
additional samples from other geographic areas.
Neoplasms of gonadal origin, similar to those re-
ported by Yevich and associates, were found in
clams from an oil-contaminated site at Searsport,
Maine. The highest prevalence of neoplasms of
hematopoietic origin was 64% , in a small sample
from Bourne. The authors pointed out, however,
that clams from some oil-contaminated sites had
no neoplasms, and stated that "These results
suggest that the type and degree of hydrocarbon
pollution are possibly related to the frequency of
neoplasms and other lesions in Mya, but they are
by no means the only causative factors."
Other types of cellular abnormalities have been
reported from soft-shell clams. In earlier studies
by Yevich and associates (Barry et al. 1971)
atypical epidermal hyperplasia in gills and kidney
was reported in up to 40% of clams sampled near
Providence, R.I. Lesions occurred more frequently
in large individuals, and seasonal changes were
not observed. Lower prevalences were found in
limited samples from Maine, Maryland, and
California. Unlike the oil spill studies, no associa-
tion with environmental factors was made by the
authors.
Yevich and associates (Yevich and Barry 1969;
Barry and Yevich 1972) have also described
gonadal neoplasms in quahogs, Mercenaria mer-
cenaria, from Narragansett Bay. Samples col-
lected in 1968, 1969, and 1970 had tumor frequen-
cies of 0.2, 2.3, and 2.7% respectively.
Epizootic neoplasms with a possible environ-
mental etiology were reported from several mol-
luscan species of Yaquina Bay, Oreg. (Farley
1969b; Farley and Sparks 1970; Mix et al. 1977).
Blue mussels, Mytilus edulis, native oysters,
Ostrea lurida, and two species of Macoma were
affected, and winter mortalities were associated
with the disease. Neoplasms have not been found
in bivalve molluscs sampled elsewhere on the
Oregon coast (Mix et al. 1977).
In another study (Christensen et al. 1974) simi-
lar epizootic neoplasms (up to 10% prevalence)
were found in a localized population of the clam
Macoma balthica from a tributary of Chesapeake
Bay. The neoplasms were invasive and systemic,
with initial foci in the gill epithelia. Holding ex-
periments indicated that the disease was usually
731
FISHERY BULLETIN: VOL. 76. NO 4
fatal. The authors suggested, but did not dem-
onstrate, an environmental contaminant etiology,
possibly associated with bottom detritus. Other
bivalve molluscs in Chesapeake Bay contain neo-
plasms. American oysters, Crassostrea uirginica,
were found with hematopoietic neoplasms (Farley
1969a; Couch 1969, Frierman 1976), and indi-
vidual oysters have been reported to contain other
types of neoplasms (Pauley 1969; Couch 1970).
Much of the evidence associating certain neo-
plasms offish and shellfish with pollutants should
be considered as circumstantial but provocative
(Rentchnick 1976). Many of the neoplasms have
been reported from bottom-feeding fish and de-
tritus or filter-feeding bivalves, as was pointed out
by Harshbarger.^ Chemical carcinogens such as
certain heavy metals and hydrocarbons can be
concentrated in surficial layers of bottom sedi-
ments and can thus be readily available to ani-
mals inhabiting that zone. It should be noted,
though, that a number of recent studies of neo-
plasms in fish and shellfish have found no obvious
relationship between neoplasms and specific en-
vironmental factors.
Skeletal Anomalies
Skeletal anomalies, particularly those of the
spinal column, are commonly observed in fish and
are the subject of an extensive literature (see Rick-
ey 1972, for a recent summary and Dawson 1964,
1966, 1971, and Dawson and Heal 1976 for a com-
plete bibliography).
Such anomalies may be genetic, resulting from
mutations or recombinations; epigenetic, acquired
during embryonic development; or postembryonic,
acquired during larval development, at metamor-
phosis, or during juvenile development (Hickey
1972). Spinal flexures and compressions, as well as
vertebral fusions, have been observed in many
teleost species, as have head and fin abnor-
malities. Evidence exists for a hereditary basis for
some skeletal anomalies (Gordon 1954; Rosenthal
and Rosenthal 1950), but other evidence points to
effects of environmental factors such as tempera-
ture, salinity, dissolved oxygen, radiation, dietary
deficiencies, and toxic chemicals. For example, in-
creased percentages of abnormal embryos and lar-
vae of Atlantic herring, Clupea harengus, resulted
'Harshbarger, J.C. 1974. Activities report (of the) registry
of tumors in lower animals 1965-1973. Smithson. Inst., Wash.,
D.C., 141 p.
from experimental exposures to sulfuric acid
waste water (Kinne and Rosenthal 1967) and to
the algicides 2,4- and 2,5 dinitrophenol ( Rosenthal
and Stelzer 1970).
Recently, increased prevalences of skeletal de-
formities and anomalies, considered to be
pollution-associated, have been recognized in a
few fish species from southern California, the
British Isles, and Japan. In studies carried out in
California, skeletal deformities occurred with
greater frequency in samples from areas with sig-
nificant pollutant stress (Valentine and Bridges
1969; Valentine et al. 1973). Exposure of fry to
very low concentrations of DDT ( <1 ppb) produced
anomalies in fin rays (Valentine and Soule 1973).
Probably the most convincing observational
evidence for environmental influences on induc-
tion of skeletal abnormalities in marine fish is
that presented by Valentine (1975). Examining
samples of barred sand hass, Paralabrax nebulifer,
Valentine found significantly higher prevalences
of anomalies, particularly gill raker deformities,
in fish from the southern California coast (Los
Angeles and San Diego) than from the Baja
California coast. The anomalies increased in fre-
quency and severity with increasing size of the fish
and an association with disturbed calcium
metabolism was suggested. The author pointed to
the high chlorinated hydrocarbon and heavy
metal levels which characterize the California
coastal area (Schmidt et al. 1971; Galloway 1972),
but emphasized that a causal relationship with
increased prevalence of anomalies had not been
established. However, Valentine's suggestion of a
possible causal relationship between high en-
vironmental levels of chlorinated hydrocarbons
and heavy metals, both of which are known to
interfere with calcium metabolism, and skeletal
anomalies in fish seems reasonable, in view of
experimental evidence from a wide range of ver-
tebrates (Ferm and Carpenter 1967; Lehner and
Egbert 1969; Peakall and Lincer 1970; Pichirallo
1971; McCaull 1971; Galloway 1972).
Valentine (1975) referred briefly to additional
observations on two other Pacific coastal
species — California grunion, Leuresthes tenuis,
and barred surfperch, Amphistkhus argenteus —
in which gill raker anomalies increased in fre-
quency with age, and were "virtually restricted to
[samples from] fishes from Southern California."
This finding in three species reduces the likelihood
that frequency differences could be attributable to
732
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
inherited subpopulation differences in one of the
three species studied.
While the deformed gill rakers were the most
prevalent anomalies observed in southern
California barred sand bass by Valentine, other
abnormalities (pugheadedness, cranial asymme-
tries, deformed vertebrae, and fin anomalies) oc-
curred and were associated directly in frequency
and severity with gill raker deformity.
An analysis of vertebral deformities in herring
taken in waters around the British Isles (van de
Kamp^") indicated a slight but significant increase
in prevalences from 1960 to 1975. The predom-
inant abnormality was a cluster of two or three
incomplete vertebrae located near the pelvic fins
or anus. The highest percentages of abnormalities
were found, according to the author, in areas
"which probably had the highest degree of pollu-
tion." It was in these areas where prevalences also
showed slight increases during the study period,
supporting the author's hypothesis that vertebral
deformities in herring can be related to "unusual
substances" in the environment. However, van de
Kamp concluded by stating that more experimen-
tal work on the causal relationship between pollu-
tion and deformities will be required.
Several reports from Japan refer to high and
increasing occurrences of skeletal anomalies in
fish. Komada (1974) and Ueki and Sugiyama
( 1976) observed increasing numbers of malformed
sweetfish or ayu, Plecoglossus altivelis, in rivers
and culture farms. Skeletal abnormalities in mul-
let and eight other species from the Inland Sea of
Japan were reported by Matsusato (1973).
Deformed fin rays (Figure 8) and associated
skeletal abnormalities have been observed re-
peatedly in winter flounders from the highly pol-
luted waters of the New York Bight (Ziskowski et
al. in press), and a summarization of observations
on skeletal anomalies and related developmental
defects has been published recently (Sindermann
et al. 1978).
There is some evidence from studies of a few
other fish species for an involvement of various
kinds of environmental stress in the occurrence
of skeletal anomalies. Gabriel (1944) noted
anomalies in vertebrae of Fundulus heteroclitus
due to temperature changes, and Mottley (1937)
found anomalies in vertebral numbers of trout due
'"van de Kamp, G. 1977. Vertebral deformities in herring
around the British Isles and their usefulness for a pollution
monitoring programme. Int. Counc. Explor. Sea, Fish. Improv.
Comm.. Doc. CM1977 E:5, 9 p.
to temperature (and possible oxygen). Hubbs
(1959) found high prevalences of vertebral abnor-
malities in mosquitofish, Gambusia affinis, from
Texas warm springs and concluded that the high
temperature was responsible.
There is also an appreciable literature con-
cerned with induction of skeletal injuries in fish
by exposure to contaminants. Vertebral damage
following experimental exposure to aquatic con-
taminants has been reported for a number of
freshwater fishes (Bengtsson 1975). Long-term (10
wk) exposure of minnows {Phoxiniis phoxinus) to
sublethal concentrations of zinc and cadmium
resulted in hemorrhaging, spinal curvatures,
and vertebral fractures, particulary in the caudal
region, in up to 70^^ of individuals. Spinal curva-
tures and muscle atrophy were produced in rain-
bow trout by chronic exposure to lead. It is in-
teresting that caudal fin erosion was also observed
in these experiments. In earlier studies, sum-
marized by Bengtsson, exposure to sublethal con-
centrations of the chlorinated hydrocarbon Toxa-
phene as well as to Malathion, parathion, and
certain other organophosphorus pesticides pro-
duced vertebral damage or spinal flexures in sev-
eral fish species. Vertebral damage was consid-
ered to have a neuromuscular origin, or, in the
case of long-term exposure, to be a consequence of
demineralization.
John Couch and associates at the Gulf Breeze
Environmental Research Laboratory of the U.S.
Environmental Protection Agency are developing
experimental evidence for induction of skeletal
abnormalities by exposure to environmental con-
taminants. Couch et al. (1977) reported severe
scoliosis and associated pathology in the sheeps-
head minnow, Cyprinodon variegatus, exposed to
the organochloride pesticide Kepone. The authors
concluded that scoliosis was a secondary effect of
Kepone toxicity, with the nervous system or cal-
cium metabolism as the primary target.
Couch and associates (J. A. Couch, Research
Pathologist, Environmental Research Labora-
tory, U.S. Environmental Protection Agency, Gulf
Breeze, PL 32561. Pers. commun., June 1977)
have also found that trifluralin (Treflan) induced
extensive osseous hyperplasia in vertebrae of
sheepshead minnows when life history stages
from zygote to 28-day juveniles were exposed to
25-50 ppb trifluralin. Centra of vertebrae, thick-
ened by active osteoblasts and fibroblasts, in-
creased in size up to 10-30 times their normal
dimensions — a striking sublethal effect.
733
FISHERY BULLETIN: VOL. 76, NO. 4
Figure 8. — Deformed fin rays in winter flounder from New York Bight. External appearance (above) and radiograph (be-
low i
(From Sindermann et al. 1978.
GENETIC ABNORMALITIES
The mutagenic properties of a number of chem-
ical contaminants including heavy metals, pes-
ticides, and petroleum-derived polycyclic hydro-
carbons have been demonstrated in experimental
studies with terrestrial animals (Huberman 1975;
Longwell"). Fish eggs can be vulnerable to con-
taminant effects from the body burden of the par-
ent female and from exposure to contaminants in
surface water and/or sediments (depending on
where in the water column spawning and de-
velopment occur). Sperm cells are sensitive to
contaminants, and eggs are especially sensitive
during meiosis and early cleavage stages. Fur-
thermore, chemical mutagens can reduce the rate
ll
"Longwell, A. C. 1975. Mutagenicity of marine pollutants
as it could be affecting inshore and offshore marine fisheries.
734
Middle Atl. Coastal Fish. Cent., Natl. Mar. Fish. Serv., Inf. Rep.
79, 72 p.
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
of cell division and can damage the mitotic spindle
apparatus. Pelagic eggs may be most severely
damaged, since the surface film of the ocean has
been found to contain high concentrations of con-
taminants such as petroleum components,
halogenated hydrocarbons, and heavy metals
(Maclntyre 1974).
Some experimental evidence is available. Fish
larvae incubated in cadmium-polluted water ac-
cumulated the metal (Westernhagen et al. 1974;
Rosenthal and Sperling 1974), and eggs incubated
in as little as 1 ppm cadmium produced low per-
centages of viable larvae (Westernhagen et al.
1975; Westernhagen and Dethlefsen 1975).
Some relevant experimental research on
radionuclide-induced mutagenesis (Romashov
and Belyayeva 1966; Ivanov 1967; AEC-TR-
7299'"^) has disclosed that many fish embryos with
severe chromosomal damage died during the
transition from blastula to gastrula. Abnormal
postgastrula embryos contained higher numbers
of chromosomal aberrations than normal em-
bryos, and the abnormal embryos had high mor-
tality just before hatching. However, even the
normal-appearing embryos with radiation expo-
sure (and consequent genetic disturbances) had
low viability and high mortality at hatching and
subsequent to hatching.
Recently, Longwell (1976a, b) reported high
prevalences of chromosomal anomalies in Atlantic
mackerel , Scorn ber scotfi brus , eggs and embryos in
certain samples taken from the New York Bight.
All degrees of chromosomal damage were found,
including failure to align at the metaphase plate,
incomplete spindle formation, translocation
bridges, chromosomal "stickiness," losses of por-
tions of chromosomes and "pulverization." Eggs
with at least one chromosome or mitotic abnor-
mality varied from 13 to 79^r . Higher percentages
seemed associated generally with degrees of en-
vironmental degradation. In addition to
chromosomal anomalies, one station (the one with
highest prevalence of anomalies) was also charac-
terized by significant {26'7c ) egg mortality.
The techniques developed by Longwell (1976b)
permitted examination of historical collections of
eggs and embryos for chromosomal damage. A
limited collection taken in 1966 from the same
geographic area disclosed a lower incidence of
cytogenetic abnormalities than that found in the
1974 collection.
Samples examined to date from normal and de-
graded waters are still insufficient, as Longwell
( 1976b) pointed out, to make definitive statements
about the relationship of pollutants and extent of
damage to genetic material, but the data pre-
sented so far indicates that such a relationship
may exist. Because of the implications of these
findings in survival and abundance of economic
marine species, it is particularly important that
this kind of research be pursued vigorously. It may
well be that a new and significant mortality factor
for estuarine and coastal populations — increased
genetic damage — may have been introduced with
increasing chemical pollution.
It is likely that marine organisms will respond
to mutagens in species-specific ways and with dif-
fering sensitivities. Some indication of this can be
found in a recent paper by Vandermeulen and
Lee^-^ in which cultures of the alga Chlainydomo-
nas reinhardtii were exposed to crude and refined
oils (Kuwait crude, Saran Gach crude, diesel 25,
and bunker C). No enhanced mutation rates (as
detected by streptomycin resistance) were found
after 3 wk of exposure (40-50 generations), a sur-
prising finding, since the alga is susceptible to cer-
tain other known mutagens and since the test oils
contain various polycyclic aromatics which are
known mutagens. No cytological examinations
were reported. The authors pointed out that con-
centrations of mutagenic components in the test
oils may be low compared with concentrations
used in cell and tissue culture to elicit enhanced
mutation rates, and that extrapolation of labora-
tory results to the marine environment should be
done very conservatively.
An indirect test for the presence of mutagens in
the marine environment has been reported re-
cently by Parry et al. (1976). Mytilus edulis were
sampled from polluted and unpolluted waters of
the United Kingdom, and extracts of their tissues
were tested for ability to induce genetic changes in
bacterial and yeast cultures. Significant increases
in mutation rates for specific gene loci charac-
terized cultures exposed to extracts of mussels
from polluted waters, but not those from clean
waters — providing evidence for the presence of
mutagens that had been concentrated in the tis-
■2AEC-TR-7299. Marine radioecology. 1972. (Distrib-
uted by NTIS, U.S. Department of Commerce, 5285 Port Royal
Road, Springfield, VA 22151.)
i^andermeulen, J. H., and R. W. Lee. 1977. Absence of
mutagenicity due to crude and refined oils in the alga
Chlamydomonas reinhardtii. Int. Counc. Explor. Sea,
Plankton Comm., Doc. CM1977/E:69, 5 p.
735
FISHERY BULLETIN: VOL. 76. NO. 4
sues of mussels from polluted areas. The chemical
nature of the mutagens was not identified, except
that the mussels came from areas with heavy in-
dustrial pollution.
EXPERIMENTALLY INDUCED LESIONS
There is a vast and almost unmanageable
amount of published information about the induc-
tion of various lesions in fish by experimental ex-
posure to chemical contaminants (see for example
Ribelin and Migaki 1975). A "lesion" may be
defined generally as "any localized abnormal
structural change in the body." Such a definition
obviousl)' includes too much, so the term can be
reduced to encompass "those cellular and tissue
changes, demonstrable histologically, that result
from a disease process." Histopathology offish and
shellfish is still a developing science, and as such it
still draws from human and veterinary (mamma-
lian) patholog}' for its concepts and much of its
terminology. Histopathology has been a basic tool
in human medicine for some time, and a large
amount of information is available about cellular
responses to toxicants. A similar core of knowl-
edge is being developed for fish and shellfish —
relating cell and tissue changes to kinds and
amounts of contaminants.
Early experimental exposures of estuarine and
marine animals to contaminants usually had the
purpose of determining lethal dosages, either from
acute or chronic exposures. More recently, atten-
tion has been redirected to sublethal toxic
effects — expressed in behavioral, physiological, or
cytological responses to specific contaminants. An
extensive literature exists concerning cell and tis-
sue damage resulting from experimental exposure
to contaminant chemicals. Generalizations that
can be made are almost predictable: 1) increas-
ing dosages, beyond a threshold level, produce in-
creasingly severe tissue abnormalities; 2) particu-
lar contaminants often exert effects on specific
target tissues; 3) principal target tissues seem to
be gill epithelium, liver (or in the invertebrate, the
hepatopancreas), and neurosensory cells; 4)
specific lesions cannot usually be described as
characteristic of any group or class of chemicals;
and 5) effects that may be of chemical origin can be
obscured by stress-provoked infections with facul-
tative pathogens. Some information about ex-
perimental induction of fin erosion and skeletal
abnormalities has been included in earlier sec-
tions of this paper, but because of the sheer volume
736
of published information about other types of ex-
perimental lesions, it seems worthwhile to sum-
marize some of the observations here.
Couch (1975) published a recent and excellent
review of the histopathological effects of pesticides
and related chemicals on the livers of fishes. The
liver and fatty tissues of fish from natural waters
are known to accumulate a number of chlorinated
hydrocarbons (Duke and Wilson 1971), and ex-
perimental exposures offish to pesticides result in
high concentrations and greatest effects on the
liver (Johnson 1968; Eisler and Edmunds 1969;
Hansen et al. 1971; Eller 1971). Some of the ob-
served liver histopathology includes:
Chlorinated hydrocarbon pesticides: Focal
areas of parenchymal cell vacuolation and de-
generation (Eller 1971), inflammation, and
loss of glycogen and fat (Lowe 1965).
Clorinated hydrocarbon herbicides: Increase
in connective tissue, massive focal necrosis
(Cope et al. 1969), and loss of glycogen (Cope
et al. 1970).
PCB's: Focal degenerative regions, paren-
chymal cell vacuolation and pleomorphism
(Eller''*), lipid accumulation in hepatic cell
vacuoles, and leucocytic infiltration (Couch
1975).
Organophosphates: Edema, hyperemia, vacuo-
lation, and necrosis of parenchymal cells (El-
ler, see footnote 14).
Carbamates: Hypertrophy and vacuolation of
acinar cells (Couch 1975).
It should be noted that not all experimental
exposures to pesticides, even for prolonged
periods, necessarily caused demonstrable tissue
pathology, but in many instances additional expo-
sure experiments are needed (even though the lit-
erature as summarized by Couch (1975) seems
voluminous). Couch pointed out that over 900,
commercial pesticide formulations are in general
use, and of these fewer than 30 have been tested
for pathological effects on livers of fishes.
Pesticides can, of course, affect fish tissues other
than liver. A summarization of general his-
topathological effects of pesticides on fish was pub-
lished by Walsh and Ribelin (1975). Data from
their own studies with coho salmon, Oncorhyn-
chus kisutch, and lake trout, as well as from other
'••Eller, L. L. 1970 and 1971. Annual reports. U.S. Bur.
Sport Fish. Wildl., Fish Pestic. Lab., Columbia, Mo.
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
published work, led them to the conclusion that
tissue changes observed as a result of exposure to
an array of common pesticides were largely
nonspecific, and therefore of limited diagnostic
value. Their attempts to identify specific lesions as
characteristic of any group or class of pesticides
were described by them as "futile," but the amount
of histopathological information presented in the
paper is substantial, and their summarization of
pathology pi'oduced by exposure to widely used
pesticides is instructive.
DDT: Necrosis of hepatic cells; lymphocytic
infiltration of intestinal lamina propria; pos-
sible degeneration of kidney tubules.
Carbaryl (Sevin): Intramuscular hemor-
rhages adjacent to vertebral column; atrophy
of the lateral line musculature; myxomatous
degeneration of fat; vacuoles within the optic
tectum of the brain.
Malathion: Subcutaneous hemorrhages at the
bases of pectoral fins.
Endosulfan (Thiodan): Hyperemia of intestine
and brain; adrenal cortical hyperplasia.
2,4-D: Striking degree of brain hyperemia;
hyperemia of intestine.
Atrazine: Marked edema of all tissues;
changes in skin pigmentation.
It is interesting that Walsh and Ribelin (1975)
(unlike Couch 1975) found liver changes in fish
exposed to pesticides ". . . minimal and diagnosti-
cally unimportant . . . ." They also considered gill
epithelial hyperplasia, gill hemorrhages, and
lymphocyte reduction in the spleen to be
nonspecific responses to stress and/or infection.
They further pointed out that rapid autolysis of
fish tissues after death rather than direct effects of
pesticides might account for some reported his-
topathological findings. These are all points of im-
portance in evaluating histological findings after
exposure to any contaminant.
There are still other histopathological studies of
the effects of pesticides on fish that disclose dam-
age to neurosensory tissue. Epithelial necrosis
was found in lateral line canals of killifish, F»/?-
dulus heteroclitus. that survived 96-h exposures to
the chlorinated hydrocarbon methoxychlor at 25
mg/1 (Gardner 1975). No damage to the
mechanoreceptors was evident, but the radius of
the canal lumina was reduced.
Pesticides can produce tissue pathology in in-
vertebrates as well. Oysters exposed chronically to
3 ppb DDT, Toxaphene, and parathion exhibited
variable lesions, including leucocytic infiltration
or hyperplasia of the gonadal germinal
epithelium, necrosis of digestive tubule epithelium,
and edema (Lowe et al. 1971). In another study,
chronic exposure of oysters to 5 ppb PCB produced
atrophy of digestive epithelium, leucocytic in-
filtration, and degeneration of vesicular connec-
tive tissue (Lowe et al. 1972). Gill edema and pro-
gressive necrosis of filaments in the crustacean
Gammarus oceanicus resulted from exposure to
sublethal concentrations of PCB (Wildish 1970).
Examination of pink shrimp, exposed experimen-
tally to PCB's, disclosed a variety of nonspecific
tissue changes, especially in the hepatopancreas
(Couch et al. 1974). Histological changes included
lysis of hepatopancreatic epithelium, nuclear pye-
nosis, vacuolization of secretory cells, and a vari-
ety of ultrastructural changes in absorptive cells.
The literature on experimentally induced le-
sions in estuarine/marine fish caused by exposure
to heavy metals was reviewed recently by Gardner
( 1975) in a paper which also presented significant
new information. His general conclusion was that
sensory organ systems of some species are vulner-
able to copper, mercury, and silver. Short-term
exposure of the killifish to sublethal concentra-
tions of copper resulted in degeneration of anterior
lateral line and olfactory sensory tissues (Gardner
and LaRoche 1973). Prolonged exposure to copper
(copper chloride) resulted in hyperplasia or necro-
sis of sustentacular epithelium of the olfactory
organs and necrosis of the epithelial lining of ol-
factory pits. Mercury (mercuric chloride) also pro-
duced severe degenerative changes in cells of the
lateral line canals and olfactory organs of killifish,
but without associated necrosis of supporting tis-
sues. Exposure to silver produced histopathologi-
cal changes very similar to copper. Cadmium
(cadmium chloride), however, did not seem to af-
fect the sensory tissues discussed above, at least in
terms of causing demonstrable tissue changes.
Cadmium exposure did result in transient thyroid
hyperplasia and altered blood cell ratios in long-
term exposures.
Experimental exposure of the cunner,
Tautogolabrus adspersus, to cadmium caused
pathological changes in kidney, intestine,
hemopoietic tissue, epidermis, and gills (Newman
and MacLean 1974). Necrosis of tubular
epithelium of the kidney, sloughing of intestinal
epithelium, hypertrophy and hyperplasia of gill
epithelium, and decrease in mucus secretion were
737
FISHERY BULLETIN: VOL. 76. NO. 4
the principal histopathologic findings. Mortality
following acute exposures was attributed to renal
failure. These results were similar in most re-
spects to cadmium-induced pathology in killifish,
reported earlier by Gardner and Yevich (1970).
Experimental cadmium exposures can cause
gill lesions in shrimp, as reported in recent papers
by Nimmo et al. (1977) and Couch (1977). Expo-
sure of pink shrimp to 763 /tg/l of cadmium for 15
days resulted in a "black gill" condition charac-
terized by necrosis of all cell types in the distal gill
filaments, with coincident appearance of black
granules in the cytoplasm, and some hemocyte
infiltration at the bases of the necrotic filaments.
Couch suggested that the black deposits could be a
metallic sulfide or even cadmium. He further
pointed out that the distal filament tissue has been
postulated to have detoxifying, as well as os-
moregulatory and respiratory functions, so that
cell death could result from cadmium filtration
and accumulation as part of a detoxification pro-
cess.
Zinc has been shown to be toxic for fish (see
reviews by Skidmore 1964, 1970). Gill tissues can
be destroyed in acute exposures, while chronic
levels induce stress which may result in mortality
and may also produce severe degenerative
changes in the liver and kidneys (Crandall and
Goodnight 1963). Synergistic activity of zinc with
a wide range of environmental variables — other
contaminant heavy metals, low dissolved oxygen,
and temperature — has been demonstrated for a
number of fish species (Doudoroff 1957; Lloyd
1960, 1961a, b). Resistance to zinc poisoning var-
ies with individuals, with age, with degree of
acclimatization, and with species (Jones 1938,
1940).
Histopathological effects of sublethal concen-
trations of copper on the winter flounder were de-
scribed by Baker ( 1969). At dosages of 1,000-3,200
/Lig/1, the kidney hemopoietic tissue became necro-
tic; gill epithelium became disoriented; chloride
cells increased in number and size; gill lamellar
fusions occurred; and fatty metamorphosis of the
liver was observed. Experimental concentrations
were far above those levels expected in most
marine environments (concentrations in polluted
waters have been reported to reach 300/Ltg/l by
Fujiya (I960)).
An interesting study of pathology in American
lobsters was made following disclosure of severe
yellow phosphorus industrial contamination of
Placentia Bay, Newfoundland (Aiken and Byard
738
1972). Experimental lobsters, exposed to phos-
phorus contaminated sediments in aquaria, ex-
hibited degenerative changes in antennal glands
and in all cell types in the hepatopancreas, as well
as massive coagulation of hemolymph.
Experimental exposure to petroleum compo-
nents and residues may also induce histopatholog-
ical changes in fish. Hyperplasia of the olfactory
sustentacular epithelium and degeneration of the
olfactory mucosa of the Atlantic silverside,
Menidia menidia, resulted from exposure to crude
oil (Gardner 1975). Additionally, degeneration of
the ventricular myocardium of the heart and
pseudobranch secretory cells was seen. Soluble
components of the crude oil also caused epithelial
metaplasia, replacing the sensory epithelium of
the olfactory organs by poorly defined cell types
( Gardner 1975). Liver damage occurred in fish fed
cyclopropenoid fatty acids (Malevski et al. 1974),
but Brocksen and Bailey (1973) found no his-
topathology in chinook salmon and striped bass
exposed to sublethal concentrations of benzene.
Histopathological effects of petroleum on
bivalve molluscs are varied in the extreme. Ef-
fects, particularly on gill epithelium, have been
observed by Barry et al. (1971), Jeffries (1972),
LaRoche (1972), Clark et al. (1974), and Gardner
et al. ( 1975). Fries and Tripp ( 1976) found damage
to gill epithelium in hard (hard-shell) clams, Mer-
cenaria mercenaria, exposed to as little as 1 ppm
phenol. Vaughan^'', however, found little his-
topathology after chronic exposures of oysters to
No. 2 fuel oil. Stainken ( 1975) found that exposure
of soft-shell clams to No. 2 fuel oil at winter seawa-
ter temperatures (4°C) for 28 days had little his-
topathological effect, beyond signs of starvation
(glycogen depletion and vacuolization of digestive
diverticula cells), and a generalized leucocytosis,
even at 100 ppm. No mortalities occurred, and
exposure concentrations dropped rapidly, possibly
because much of the oil was trapped in mucus as
part of the mucociliary feeding mechanism, and
ejected from the clam.
Experimental lesions are instx'uctive in iden-
tifying target organs and tissues for particular
contaminants, but they have numerous flaws
when attempts are made to relate experimental
findings to events in the natural (polluted) envi-
ronment: 1) dosage levels are often beyond
'^Vaughn.B.E. (editor). 1973. Effects of oil and chemically
dispersed oil on selected marine biota - a laboratory study. Am.
Pet. Inst. Publ. 4191.
SINDKRMANN: fOLLl'TION-ASSOCIATKl) OISEASKS AND ABNOKMAI.ITIKS
maximum observed environmental levels; 2) usu-
ally single chemicals are tested, ignoring possible
synergisms and antagonisms; 3) tests are often
static acute rather than chronic exposures in
flow-through systems; and 4) experimental ani-
mals are often under stress from the mere act of
confinement.
These and other limitations of experimental
studies degrade the evidence obtained to cir-
cumstantial when attempts are made to extrapo-
late findings to natural populations in polluted
habitats. Despite this handicap, there is a large
and useful literature on experimental lesions in
fish and shellfish produced by chemicals which
occur as contaminants in the coastal environment.
The presence of specific pollutants cannot be
recognized by the occurrence of specific lesions,
but a general description of pathological responses
can be useful. Categories of pathological responses
which should be considered in experimental
studies are; 1 (inflammation (acute and chronic);
2) degeneration (including edema, necrosis, and
metaplasia); 3 ) repair and regeneration (prolifera-
tion, hyperplasia, and scar formation); 4) neo-
plasia (including consideration of cell origin,
stage, and type — whether benign or malignant);
and 5) genetic derangement (including
chromosomal changes and skeletal abnor-
malities).
CONTAMINANT EFFECTS ON
RESISTANCE AND IMMUNE RESPONSES
Suppression of immune responses by toxicants
such as heavy metals and pesticides has been
demonstrated repeatedly in mammals (Kolom-
iitseva et al. 1969; Hemphill et al. 1971; Khan and
Hill 1971; Jones et al. 1971; Roller 1973; Street
and Sharma 1975). Therefore, it might be expected
that environmental pollutants could influence the
ability of fish and shellfish to resist infection by
reducing the effectiveness of external and internal
defense mechanisms, and indeed there is some
evidence that this is so. Changes in the principal
external defenses — mucus secretion offish and the
epicuticle of Crustacea — have already been men-
tioned in connection with fin erosion and exo-
skeletal erosion. Some specific information is
available about contaminant influences on inter-
nal defenses, principally through suppression of
immune responses. Environmental stress from
contaminants can affect internal resistance to in-
fection in fish by causing a decrease in phagocytic
activity (Wedemeyer 1970) or a decrease in anti-
body synthesis (Goncharov and Mikyakov 1971).
Both mechanisms have been demonstrated ex-
perimentally.
One of the best pieces of supporting infor-
mation about suppression of host responses was
derived from a recent multidisciplinary experi-
mental study of the effects of short-term sublethal
exposures to cadmium on the teleost Tautogolab-
rus adspefsus (Calabrese et al. 1974). The study
included chemical analyses of tissue uptake,
physiological and biochemical effects, his-
topathological changes, and effects on the immune
system. Robohm and Nitkowski ( 1974), who were
responsible for the immunology, found that expo-
sure offish to 12 ppm cadmium affected phagocyte
response to foreign antigen, but not the humoral
response. The rate of bacterial uptake in phago-
cytes of liver and spleen was increased, but the
rate of bacterial destruction within the phagocytes
was decreased significantly. No change was ob-
served in the antibody response of immunized con-
trol and experimental fish as determined by
hemagglutination techniques. The authors postu-
lated that cadmium may prevent delivery of
lysosomal substances to the phagocytic vacuole, or
may inhibit the action of these substances on bac-
teria, but that cadmium does not seem to inhibit
antibody synthesis by lymphocytic cells. The au-
thors further suggested that cadmium and possi-
bly other pollutants may affect fish populations by
causing phagocytic dysfunction, reducing the re-
sistance offish to facultative and other pathogens.
The effect of sublethal copper exposure on the
immume response of juvenile coho salmon, On-
corhynchus kisutch, was examined by Stevens.'^
At copper levels of 18 )U,g/l, agglutinin titers in
fingerlings injected intraperitoneally with Vibrio
anguillarum bacterin were significantly lower
than those of controls. Copper exposure also re-
duced survival of coho salmon fingerlings during
saltwater acclimation.
Reduction in immunological competence may
well have been involved in observed outbreaks of
vibrosis (V . anguillarum ) in eels exposed to copper
(R0dsaether et al. 1977) and in epizootics of
Aeromonas liquefaciens (= A. hydrophila) in
salmon and suckers exposed to copper and zinc
pollution (Pippy and Hare 1969), although in
'^Stevens, D. G. 1977. Survival and immune response of
coho salmon exposed to copper. Environ. Prot. Agency - 600/3-
77-031, 37 p.
739
FISHKRY BULLKTIN: VOL. 76. NO. 4
neither instance were antibody titers determined.
In the latter instance, A. Uquefaviens is an
ubiquitous water bacterium, but only causes dis-
ease and mortalities in fish with lowered resis-
tance (Snieszko 1962).
Reduction in antibody response to injected virus
was demonstrated by Perlmutter et al. (1973) in
blue gourami, Trkhogaster trichopteri/s, to result
from overcrowding. The authors postulated that
stressed fish released a pheromonelike im-
munosuppressive factor under crowded condi-
tions. It is reasonable to expect that other types of
environmental stresses could result in a similar
response.
Among the invertebrates, indirect evidence for
reduction of disease resistance caused by conta-
minant exposure is available and has already been
discussed in previous sections on crustacean shell
disease and shrimp virus disease. Direct experi-
mental evidence however, is scarce. Fries and
Tripp (1976) exposed hard (hard-shell) clams to
phenol and found damage to gill and digestive
tract epithelia — tissues which are considered im-
portant components of internal defense
mechanisms. The authors suggested, but did not
demonstrate, that phenol-treated clams may be
more susceptible to microbial infections than
normal ones. In other studies with invertebrates,
Telford (1968, 1974) demonstrated that environ-
mental stress affected blood glucose levels in
Homarus americanus and crayfish, Cambarus
clarkii.
POLLUTANT-PARASITE
INTERACTIONS
Much has been said and much documentation
exists about the role of environmental stress in
induction, severity, and persistence of disease.
Some of the best information about stress and dis-
ease in fish comes from studies concerned with
aquaculture — where environmental factors such
as temperature, oxygen, water quality, salinity,
and diets clearly influence the course of disease
and the impact of disease on cultured populations.
There is also a developing body of information,
from experimental work as well as from field ob-
servations and surveys, about the possible rela-
tionship of parasitism and pollution. The relation-
ship is not simple, and in essence involves a
double-edged phenomenon, in which pollutant
stress may result in an increase (or in some in-
stances decrease) in the prevalence of certain
740
parasites, or in which parasitization may decrease
host resistance to toxic pollutants. Subsidiary is-
sues quickly emerge however, such as the effects of
pollutants on intermediate or alternate hosts in
parasite life cycles, possible effects of pollutants on
free-living life cycle stages of parasites, and effects
of pollutants on host defenses against parasite in-
vasion.
Thus far in this review, the role of microbial
infectious agents, principally viruses and bac-
teria, has been emphasized, but there is some li-
mited evidence that environmental pollution may
change the relationships among animal parasites
and their fish hosts (Esch et al. 1975).
Looking first at the influence of parasites on
host susceptibility to contaminants, several recent
papers (principally from studies in freshwater)
offer significant insights. Boyce and Ydmada
(1977) found in laboratory experiments that sock-
eye salmon, Oncorhynchus nerka, smolts with
preexisting parasitization by the intestinal
pseudophyllidean cestode Eiihothrium salvelini
were more susceptible to zinc poisoning than un-
parasitized siblings. Similarly, Pascoe and Cram
(1977) found that survival times of the threespine
stickleback, Gasterosteiis aciileatus, exposed to
various concentrations of cadmium, were much
shortened if the fish were parasitized by the larval
cestode Schistocephalus solidiis. Perevozchenko
and Davydov (19741 found that juvenile carp
parasitized by the intestinal cestode Both rioceph-
alus goivkongensis were more susceptible to DDT
poisoning than were nonparasitized individuals.
These results are not surprising, since fish already
weakened by parasites would undoubtedly be less
able to tolerate other environmental stresses. The
nature and degree of parasitization offish clearly
must be considered in bioassays and in studies of
effects of contaminants on fish and shellfish
species.
Looking next at the reverse viewpoint, the
influence of contaminants on parasite prevalence,
definitive information is less readily available for
marine species, but some information is available
for freshwater species. Thermal loading was as-
sociated with changes in the distribution and
abundance of two larval trematodes in mos-
quitofish (Aho et al. 1976). Similarly, thermal
loading from a nuclear power plant was directly
correlated with incidence of the ciliate Epistylis
sp. and the bacterium Aeromonas liquifaciens
( = A. hydrophila) in six species of centrarchids in
South Carolina (Esch et al. 1976). Effects of ther-
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
mal effluents on parasitism of largemouth bass,
Micropterus salmoides, by the acanthocephalan
Neoechinorhynchus cylindratus were examined by
Eure and Esch ( 1974). Parasite densities were sig-
nificantly higher in fish from heated water during
the winter months, a possible reflection of greater
densities of larval parasites and intermediate host
populations in the effluent. River pollution from
domestic and industrial sources was considered to
be a contributing factor in increased parasite bur-
dens found in fish from areas of heaviest pollution
in Poland (Dabrowska 1974).
For marine species, good evidence relating pol-
lutants with changes in parasite abundance is
scarce. Results of an extensive survey of external
parasites and disease conditions in North Sea fish
(Moller''') did not disclose clear-cut relationships
between parasitism and pollution, although the
higher prevalence of vibriosis and lymphocystis in
southern sectors which are most polluted indi-
cated a possible influence of pollution. Other fac-
tors seemed responsible for differential abun-
dances reported for the larger external parasites.
Several parasites of estuarine fishes from the
Gulf of Mexico were examined by Overstreet and
Howse (1977) in a search for associations with
environmental pollution. Samples of Atlantic
croaker were collected in 1970-72, and again in
1975. Large variations in prevalences of helminth
parasites occurred, but clear associations with pol-
lutants and changes in pollutant levels were not
established. A myxosporidan protozoan seemed to
be more promising. Infections of sheepshead min-
nows by Myxobolus lintoni were very abundant in
one polluted bayou of Mississippi, but were absent
in seemingly healthy habitats.
The stalked peritrich ciWdiie Epistylis sp., men-
tioned in an earlier section in connection with fin
erosion and red sores, seems to be related to high
organic content and possibly other stresses in
freshwater and brackish water habitats. The
ciliate, together with secondary bacterial invaders
(principally Aeromonas liquifaciens (= A. hy-
drophila), produces a hemorrhagic hyperplastic
condition beneath the scales that is referred to as
red sore (Overstreet and Howse 1977). The ciliate
infests a wide range offish species in low salinity
waters of Mississippi, especially centrarchids,
"Moller, H. 1977. Distribution of some parasites and dis-
eases of fishes from the North Sea in February, 1977. Int.
Counc. Explor. Sea, Fish. Improv. Comm., Doc. CM1977/E:20, 16
P-
sheepshead, and black drum (the drum is a marine
invader in brackish water). Secondary bacterial
infections associated with the ciliate m.ay become
systemic, and mortalities may result.
In addition to field observations, there is some
experimental evidence for a causal relationship
between specific pollutant chemicals and fungus
parasitization offish and shellfish. In one study,
oysters exposed to pesticides (DDT, Toxaphene,
and parathion) became infected with a mycelial
fungus that caused lysis of the mantle, gut,
gonads, gills, visceral ganglion, and kidney
tubules (Lowe et al. 1971). None of the control
oysters became infected, indicating a role for one
or several of the pesticides in altering the host-
parasite relationships of the oysters and the fun-
gus. Presence of fungus infections made it difficult
to differentiate histopathological effects of pes-
ticide exposure from those due to the parasite.
CONCLUSIONS
In considering pollution-associated diseases of
fish and shellfish, a number of conclusions seem
warranted:
1. Environmental stress from pollutants seems
to be an important determining factor in sev-
eral fish and shellfish diseases. Effects in-
clude direct chemical-physical damage to cell
membranes or tissues, modification of
physiological and biochemical reactions, in-
creased infection pressure from facultative
microbial pathogens, and reduced resistance
to infection.
2. The multifactorial genesis of disease in
marine species is becoming apparent, involv-
ing environmental stress, facultative patho-
gens, resistance of hosts, and latent infec-
tions.
3. Some circumstantial evidence for the role of
environmental carcinogens in the etiology of
neoplasms offish and shellfish is accumulat-
ing, but at present definitive conclusions are
not justified.
4. The presence of marginal or degraded
estuarine/coastal environments may be sig-
nalled by the appearance of, or the increase
in prevalence of a number of diseases, includ-
ing fin erosion, "red sores," ulcers, and possi-
bly lymphocystis in fish; by "shell disease" in
crustaceans; and by certain neoplasms in
bivalve molluscs, but an absolute cause and
741
FISHERY BULLETIN: VOL. 76, NO. 4
effect relationship has not yet been de-
monstrated for most of these diseases.
5. Among the most severe and persistent prob-
lems in establishing pollutant-disease rela-
tionships are: the absence of baseline in-
formation about the organisms and their
habitats prior to pollution, the existence of
multiple pollutants in many badly degraded
waters, and the circumstantial nature of
much of the evidence linking pollution and
disease.
6. A number of viruses have been found in crus-
taceans and molluscs in recent years, and the
pathogenic role of two of them (shrimp
Baculovirus and oyster Herpesvirus) has
been demonstrated by exposure to increasing
environmental stress. Other latent virus in-
fections of invertebrates may be identified by
similar experimental methods.
The evidence for an association of pollution and
disease presented in this paper (except for results
of experimental studies) is largely circumstantial.
When confronted with the hard question "Can you
state positively that the disease condition seen in
natural populations is caused by specific environ-
mental contaminants?", the answer at present has
to be "No." However, the weight of this cir-
cumstantial evidence, particularly for diseases
such as fin erosion and ulcers, is such that it leads
to the conclusion that associations do exist be-
tween pollutants and disease.
ACKNOWLEDGMENTS
A number of people read drafts of this paper, and
many changes and additions have resulted from
their comments. I would like to acknowledge the
assistance of R. Overstreet, A. Sparks, M. Sher-
wood, R. Wolke, J. Couch, J. Pearce, A. Farley, A.
Rosenfield, and R. Murchelano — without neces-
sarily implying their agreement with any or all of
the interpretations and conclusions in this paper. I
would also like to thank K. Melkers for maintain-
ing the continuity and accuracy of the manuscript
through a series of revisions.
LITERATURE CITED
Aho, J. M., J. w. Gibbons, and g. w. esch.
1976. Relationship between thermal loading and
parasitism in the mosquitofish. In G. W. Esch and R. W.
McFarlane (editors), Thermal ecology II, p. 213-218.
Technical Information Center, Energy Research and De-
velopement Agency (CONF-750425).
Aiken, D. E., and E. H. Byard.
1972. Histological changes in lobsters {Homarus
americanus) exposed to yellow phosphorus. Science
(Wash., D.C.) 176:1434-1435.
ALPERS, C. E., B. B. McCain, M. S. Myers, and S. R. Well-
INGS.
1977a. Lymphocystis disease in yellowfin sole (Limanda
aspera) in the Bering Sea. J. Fish. Res. Board Can.
34:611-616.
ALPERS, C. E., B, B. McCain, M. Myers, S. R. Wellings, M.
POORE, J. BAGSHAVv', and C. J. DAWE.
1977b. Pathologic anatomy of pseudobranch tumors in
Pacific cod, Gadus macrocephalus. J. Natl. Cancer Inst.
59:377-398.
Aneer, G., and O. Ljungberg.
1976. Lymphocystis disease in Baltic herring iClupea
harengus var. membras L.). J. Fish. Biol. 8:345-350.
angell, C. L., b. S. Miller, and s. R. wellings.
1975. Epizootiology of tumors in a population of juvenile
English sole (Parophrys vetulus) from Puget Sound,
Washington. J. Fish. Res. Board Can. 32:1723-1732.
Anonymous.
1951. Diseased stripers in Connecticut are safe to
eat. Salt Water Sportsman, June 22, p.l.
AWERINZEW, S.
1911. Studien liber parasitische Protozoen. V. Einige neue
Befunde aus der Entwicklungsgeschichte von Lymphocys-
tis johnstonei Woodc, Arch. Protistenkd. 22:179-196.
Baker, J. T. P.
1969. Histological and electron microscopical observa-
tions on copper poisoning in the winter flounder (Pseudo-
pleuronectes americanus). J. Fish. Res. Board Can.
26:2785-2793.
BARRY, M. M., AND P. P. YEVICH.
1972. Incidence of gonadal cancer in the quahaug {Mer-
cenaria mercenaria). Oncology 26:87-96.
1975. The ecological, chemical and histopathological
evaluation of an oil spill site. Part III. Histopathological
studies. Mar. Pollut. Bull. 6:171-173.
Barry, M. M., P. P. Yevich, and N. H. Thayer.
1971. Atypical hyperplasia in the soft-shell clam Mya
arenana. J. Invertebr. Pathol. 17:17-27.
BeNGT-SSON, B.-E.
1974. Vertebral damage to minnows Phoxinus phoxinus
exposed to zinc. Oikos 25:134-139.
1975. Vertebral damage in fish induced by pollutants. In
J.H. KoemanandJ. J. T. W. A. Strik (editors), Suble-
thal effects of toxic chemicals on aquatic animals, p. 23-
30. Elsevier Sci. Publ. Co., Amsterdam.
Bonnet. R.
1883. Studien zur Physiologie und Pathologie der
Fische. Bayer. Fisch. Z. 8:79-81.
Boyce, N. p., and S. B. YAMADA.
1977. Effects of a parasite, Eu both rium salvelini (Cestoda:
Pseiidophyllidea), on the resistance of juvenile sockeye
salmon, Oncorhynchiis nerka ,tozmc. J. Fish. Res. Board
Can. 34:706-709.
Brett, J. R,
1958. Implications and assessments of environmental
stress. In P. A. Larkin (editor), The investigation of
fish-power problems, p. 69-97, H. R. MacMillan Lee.
Fish., Univ. B.C., Vancouver.
742
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
Brocksen, R. W., AND H. C. Bailey.
1973. The responses of juvenile chinook salmon and
striped bass exposed to benzene, a water-soluble compo-
nent of crude oil. Proceedings of the 1973 Conference,
Prevention and Control of Oil Spills, Wash., D.C., p. 783-
791.
Brooks, R. E., G. E. McArn, and S. R. Wellings.
1969. Ultrastructural observations on an unidentified cell
t}T)e found in epidermal tumors of flounders. J. Natl.
Cancer Inst. 43:97-109.
Brown, R. S., R. E. Wolke, and S. B. Saila.
1976. A preliminary report on neoplasia in feral popula-
tions of the soft-shell clam, Mya arenaria: prevalence,
histopathology and diagnosis. Proc. First Int. Colloq.
Invertebr. Pathol., Kingston, Ontario, Can., p. 151-159.
BROWN, R. S., R. E, Wolke, S. B. Saila, and C. W. Brown.
1977. Prevalence of neoplasia in 10 New England popula-
tions of the soft-shell clam {Mya arenaria). Ann. N.Y.
Acad. Sci. 298:522-534.
Calabrese, a., R. S. Collier, and J. E. Miller.
1974. Physiological response of the cunner.Taf/to^o/aftr;/.?
adspersus, to cadmium. I. Introduction and experimental
design. U.S. Dep. Commer., NOAA Tech. Rep. NMFS
SSRF-681, p. 1-3.
Carlisle, J. G., Jr.
1969. Results of a six-year trawl study in an area of heavy
waste discharge: Santa Monica Bay, California. Calif.
Fish Game 55:26-46.
Carmody, d. J., J. B. Pearce, and W. E. Yasso.
1973. Trace metals in sediments of New York Bight.
Mar. Pollut. Bull. 4:132-135.
Christensen, D. J., C. A. Farley, and F. G. Kern.
1974. Epizootic neoplasms in the clam Macoma balthica
(L.) from Chesapeake Bay. J. Natl. Cancer Inst.
52:1739-1749.
Christmas, J. Y., and H. d. howse.
1970. The occurrence of lymphocystis in Micropogon un-
dulatus and Cynoscion arenarius from Mississippi es-
tuaries. Gulf Res. Rep. 3:131-154.
Clark, R. C, Jr. J. S. Finley, and G. G. Gibson.
1974. Acute effects of outboard motor effluent on two
marine shellfish. Environ. Sci. Technol. 8:1009-1014.
Cooper, R. C, and C. a. Keller.
1969. Epizootiology of papillomas in English sole, Paro-
phrys vetulus. Natl. Cancer Inst. Monogr. 31:173-185.
Cope, O. B., J. P. McCraren, and L. Eller.
1969. Effects of dichlobenil on two fishpond environments.
Weed Sci. 17:158-161.
Cope, O. B., e. m. wood, and g. H. Wallen.
1970. Some chronic effects of 2,4-D on the bluegill
iLepomis macrochirus). Trans. Am. Fish. Soc. 99:1-12.
Couch, j. a.
1969. An unusual lesion in the mantle of the American
oyster, Crassostrea virginica. Natl. Cancer Inst.
Monogr. 31:557-562.
1970. Sarcoma-like disease in a single specimen of the
American oyster. [Abstr.] In R. M. Butcher (editor).
Comparative leukemia research 1969. p. 647. Bibl.
Haematol. 36.
1974a. Free and occluded virus, similar to Baculovirus, in
hepatopancreas of pink shrimp. Nature (Lond.)
247:229-231.
1974b. An enzootic nuclear polyhedrosis virus of pink
shrimp: Ultrastructure, prevalence, and enhance-
ment. J. Invertebr. Pathol. 24:311-331.
1975. Histopathologic effects of pesticides and related
chemicals on the livers of fishes. In W. E. Ribelin and G
Migaki (editors). The pathology of fishes, p. 559-
584. Univ. Wis. Press, Madison.
1976. Attempts to increase Baculovirus prevalence in
shrimp by chemical exposure. Prog. Exp. Tumor Res.
20:304-314.
1977. Ultrastructural .study of lesions in gills of a marine
shrimp exposed to cadmium. J. Invertebr. Pathol.
29:267-288.
1978. Diseases, parasites, and toxic responses of commer-
cial penaeid shrimps of the Gulf of Mexico and south
Atlantic coasts of North America. Fish. Bull., U.S.
76:1-44.
Couch, J. A., and L. Courtney.
1977. Interaction of chemical pollutants and virus in a
crustacean: a novel bioassay system. Ann. N.Y. Acad.
Sci. 298:497-504.
Couch, j. a., and d. R. Nimmo.
1974a. Ultrastructural studies of shrimp exposed to the
pollutant chemical, polychlorinated biphenyl (Aroclor
1254). Bull. Pharmacol. Environ. Pathol. 11:17-20.
1974b. Detection of interactions between natural patho-
gens and pollutants in aquatic animals. In R. L. Am-
borski, M. A. Hood, and R. R. Miller (editors), Proc. Gulf
Coast reg. symp. diseases of aquatic animals, p. 261-
268. La. State Univ., Cent. Wetland Resour. Publ. No.
LSU-SG-74-05.
Couch, J., G. Gardner, J. C. Harshbarger, M. R. Tripp, and
P. P. Yevich.
1974. Histological and physiological evaluations in some
marine fauna. In Marine Bioassays Workshop Proceed-
ings 1974, p. 156-173, Mar. Technol. Soc, Wash., D.C.
Couch, J. A., J. T. Win.stead, and L. R. Goodman.
1977. Kepone-induced scoliosis and its histological con-
sequences in fish. Science (Wash. D.C.) 197:585-587.
Crandall, C. a., and C. j. Goodnight.
1963. The effects of sublethal concentrations of several
toxicants to the common guppy, Lebistes re-
ticulatus. Trans. Am. Microsc. Soc. 82:59-73.
Dabrowska, H,
1974. An attempt to evaluate the state of health of fish
from the Lyna and Walsza Rivers in connection to their
pollution. [In Pol] Przegl.Zool. 18(3):390-395. Abstract
by S.F.Snieszko,1977,/« Fish Health News 6(3):156-157.
Dawe, C. j., and j. C. Harshbarger (editors).
1969. Neoplasms and related disorders of invertebrate and
lower vertebrate animals. Natl. Cancer Inst. Monogr.
31, 772 p.
Dawe, C. J., D. G. Scarpelli, and S. R. Wellings (editors).
1976. Tumors in aquatic animals. Prog. Exp. Tumor Res.
20,411 p.
Dawson, C. E.
1964. A bibliography of anomalies of fishes. Gulf Res.
Rep. 1:308-399.
1966. A bibliography of anomalies of fishes. Suppl. 1,
Gulf Res. Rep. 2:169-176.
1971. A bibliography of anomalies of fishes. Suppl. 2,
Gulf Res. Rep. 3:215-239.
Dawson, C. E., and e. Heal.
1976. A bibliography of anomalies of fi.shes. Suppl. 3,
Gulf Res. Rep. 5:35-41.
743
FISHERY BULLETIN: VOL. 76, NO. 4
DOUDOROFF, P.
1957. Water quality requirements of fishes and effects of
toxic substances. In M. E. Brown (editor), The physiol-
ogy of fishes, Vol. II, p. 403-4.30. Acad. Press, N.Y.
DUKE, T. W., .AND A. J. Wilson, Jr.
1971. Chlorinated hydrocarbons in livers of fishes from the
northeastern Pacific Ocean. Pestic. Monit. J. 5:228-232.
Duke, T. W., J. I. Lowe, and A. J. Wilson, Jr.
1970. A polychlorinated biphenyl (Aroclor 1254*) in the
water, sediment, and biota of Escambia Bay, Flori-
da. Bull. Environ. Contam. Toxicol. 5:171-180
DUKES, T. W., AND A. R. LAWLER.
1975. The ocular lesions of naturally occurring lym-
phocystis in fish. Can. J. Comp. Med. 39:406-410.
Edwards, R. H., and R. M. Overstreet.
1 976. Mesenchymal tumors of some estuarine fishes of the
northern Gulf of Mexico. I. Subcutaneous tumors, proba-
bly fibrosarcomas, in the striped mullet, Mugil
cephalus. Bull. Mar. Sci. 26:33-40.
Eisler, R., and p. H. Edmunds.
1966. Effects of endrin on blood and tissue chemistry of a
marine fish. Trans. Am. Fish. Soc. 95:153-159.
ELLER, L. L.
1971. Histopathologic lesions in cutthroat trout (Salmo
clarki) exposed chronically to the insecticide en-
drin. Am. J. Pathol. 64:321-336.
EscH, G. W., J. W. Gibbons, and J. E. Bourque.
1975. An analysis of the relationship between stress and
parasitism. Am Midi. Nat. 93:339-353.
E.SCH, G. W., T. C. HAZEN, R. V. DIMOCK, JR., AND J. W. GIB-
BONS.
1976. Thermal effluent and the epizootiology of the ciliate
Epistylis and the bacterium Aeromonas in association
with centrarchid fish. Trans. Am. Microsc. Soc. 95:687-
693.
EURE, H. E., AND G. W. ESCH.
1974. Effects of thermal effluent on the population
dynamics of helminth parasites in largemouth bass. In
J. W. Gibbons and R. R. Sharitz (editors). Thermal ecol-
ogy, p. 207-215. U.S. At. Energy Comm. Symp. Ser.
(Conf. 730505).
Fange, R., C. Rappe, and S. FALKMER.
1975. Liver tumours in the Atlantic hagfish (Myxine
glutinosa L.). In J. H. Koeman and J. J. T. W. A. Strik
(editors). Sublethal effects of toxic chemicals on aquatic
animals, p. 103-109. Elsevier Sci. Publ. Co.. Amster-
dam.
FALKMER, S., S. O. EMDIN, Y. dSTERBERG, A. MATTISSON, M.
L. JOHANSSON SJOBECK, AND R. FANGE.
1976. Tumor pathology of the hagfish, Myxine glutinosa,
and the river lamprey, Lampetra fluviatilis. A light-
microscopical study with particular reference to the oc-
currence of primary liver carcinoma, islet-cell tumors,
and epidermoid cysts of the skin. Prog. Exp. Tumor Res.
20:217-250.
FALKMER, S., S. MARKLUND, P. E. MATTSSON, AND C. RAPPE.
1977. Hepatomas and other neoplasms in the Atlantic
hagfish (Myxine glutinosa): a histopathologic and chemi-
cal study. Ann. N.Y. Acad. Sci. 298:342-355.
Farley, C. A.
1969a. Probable neoplastic disease of the hematopoietic
system in oysters, Crassostrea virginica and Crassostrea
gigas. Natl. Cancer Inst. Monogr. 31:541-555.
1969b. Sarcomatoid proliferative disease in a wild popula-
tion of blue mussels iMytilus edulis). J. Natl. Cancer
Inst. 43:509-516.
Farley, C. A., and A. K. Sparks.
1970. Proliferative diseases of hemocytes, endothelial
cells and connective tissue cells in mollusks. In R. M.
Dutcher (editor). Comparative leukemia research 1969, p.
610-617 Bibl. Haematol. 36.
Farley, C. A., W. G. Banfield, G. Kasnic, Jr., and W. S.
Foster.
1972. Oyster herpes-type virus. Science (Wash., D.C.)
178:759-760.
Ferm, V. H., AND S. J. Carpenter.
1967. Developmental malformations resulting from the
administration of lead salts. Exp. Mol. Pathol. 7:208-
213.
Frierman, E. M.
1976. Occurrence of hematopoietic neoplasms in Virginia
oysters (Crassostrea virginica). Mar. Fish. Rev.
38(10):34-36.
Fries, C, and M. R. Tripp.
1976. Effects of phenol on clams. Mar. Fish. Rev.
38(10):10-11.
FUJIYA, M.
1960. Studies on the effects of copper dissolved in sea water
on oyster. Bull. Jpn. Soc. Sci. Fish. 26:462-468.
Gabriel, M. L.
1944. Factors affecting the number and form of vertebrae
in Fundulus heteroclitus. J. Exp. Zool. 95:105-147.
Galloway, J. N.
1972. Man's alteration of the natural geochemical cycle of
selected trace-metals. PhD. Thesis, Univ. California,
San Diego, 151 p.
Gardner, G. R.
1975. Chemically induced lesions in estuarine or marine
teleosts. In W. E. Ribelin and G. Migaki (editors), The
pathology of fishes, p. 657-693. Univ. Wisconsin Press,
Madison.
Gardner, G. R., and G. LaRoche.
1973. Copper induced lesions in estuarine teleosts. J.
Fish. Res. Board Can. 30:363-368.
Gardner, G. R., and P. P. Yevich.
1970. Histological and hematological responses of an es-
tuarine teleost to cadmium. J. Fish. Res. Board Can.
27:2185-2196.
Gardner, G. R., P. P. Yevich, and P. F. Rogerson.
1975. Morphological anomalies in adult oyster, scallop,
and Atlantic silversides exposed to waste motor oil. In
Proceedings of the 1975 Conference, Prevention and Con-
trol of Oil Pollution, San Francisco, Calif., p. 473-
477. Am. Pet. Inst.
Goncharov, G. D., and V. R. Mikyakov.
1971. The effect of low concentrations of phenol on anti-
body formation in carp, Cyprinus carpio. Probl. Aquatic
Toxicol. 1970, p. 171-175. (Transl. U.S. Bur. Sport Fish.
Wildl., Wash., D.C.)
GOOD, H. V.
1940. A study of an epithelial tumor of Parophrys vet-
ulus. M.S. Thesis, Univ. Washington, Seattle, 48 p.
GOPALAN, U. K., AND J. S. YOUNG.
1975. Incidence of shell disease in shrimps in the New
York Bight. Mar. PoUut. Bull. 6:149-153.
744
SIN'nERMANN POLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
GORDON, M.
1954. The genetics offish diseases. Trans. Am. Fish. Soc.
83:229-240.
Hansen, D. J., P. R. Parrish. J. I. Lowe, A. J. Wil.son. Jr.,
AND p. D. Wilson.
1971. Chronic toxicity, uptake, and retention of Aroclor "
1254 in two estuarine fishes. Bull. Environ. Contam.
Toxicol. 6:113-119.
Hemphill, F. E., M. L. Kaeberle, and W. B. Buck.
1 97 1 . Lead suppression of mouse resistance to Salmonella
typhimurium. Science (Wash., D.C.) 172:1031-1032.
HICKEY, C. R.. JR.
1972. Common abnormalities in fishes, their causes and
effects. N.Y. Ocean Sci. Lab. Tech. Rep. 0013, 20 p.
HUBBS, C.
1959. High incidence of vertebral deformities in two
natural populations of fishes inhabiting warm springs.
Ecology 40:154-155.
HI'BERMAN, E.
1975. Mammalian cell transformation and cell-mediated
mutagenesis by carcinogenic polycyclic hydrocarbons.
Mutat. Res. 29:285-291.
IVANOV, V. N.
1967. Effects of radioactive substances on the embryonic
development of fish. In Problems of biological oceanog-
raphy, p. 185. AEC-TR-6940.
Jeffries, H. P.
1972. A stress syndrome in the hard clam, Mercenaria
mercenaria. J. Invertebr. Pathol. 20:242-251.
Jensen, N. J., and J. L. Larsen.
1976. Forskningsprojet vedrorende marine vib-
riopopulationers for hold til dratforurening i marine
biotoper. Dans. Vet. Tidsskr. 59:521-524.
1977. Sarsygdom hos torsk i danske kystnaere omrader
med forskellig vandkvalitet. Nord. Veterinaermed.
29(Suppl. 11:4.
JONES, J. R. E.
1938. The relative toxicity of salts of lead, zinc and copper
to the stickleback iGasterosteus aculeatus L.I and the ef-
fect of calcium on the toxicity of lead and zinc salts. J.
Exp. Biol. 15:394-407.
1940. A study of the zinc-polluted river Ystwyth in north
Cardiganshire, Wales. Ann. Appl. Biol. 27:368-378.
JONES, R. H., R. L. WlI.l.IA.MS. AND A. M. JONES.
1971. Effects ofheavy metal on the immune response. Pre-
liminary findings for cadmium in rats. Proc. Soc. Exp.
Biol. Med. 137:1231-1236.
JOHNSON, D. W.
1968. Pesticides and fishes — A review of selected litera-
ture. Trans. Am. Fish. Soc. 97:398-424.
JOHNSTONE, J.
1905. Internal parasites and disease conditions of fishes.
Rep. Lanes. Sea-Fish. Labs. 1904. No. 13, p. 98-120.
Kelly, D. L.
1971. Epidermal papilloma in the English sole. Parophrys
vetulus. Ph.D. Thesis, Univ. California, Berkeley. 139 p.
KHAN, A., AND J. M. HILL.
1971. Immunosuppression with (/.s-platinum (II) diam-
minodichloride: effect on antibody plaque-forming spleen
cells. Infect. Immun. 4:320-321.
KiNNE. O., AND H. Rosenthal.
1967. Effects of sulfuric water pollutants on fertilization,
embryonic development and larvae of the herring, Clupea
harengus. Mar. Biol. (Berl.i 1:65-83.
KOEMAN, J. H., AND J. J. T. W. A. STRIK (editors).
1975. Sublethal effects of toxic chemicals on aquatic ani-
mals. Elsevier Sci. Publ. Co., Amsterdam, 234 p.
Koi.LER. L. D.
1973. Immunosuppression produced by lead, cadmium,
and mercury. Am. J. Vet. Res. 34:1457-1458.
Kolomiitseva, M. G., F. M. Voznesenskaya, and E. a.
I.SAEVA.
1969. Immunobiological reactivity of the animal organism
in association with different copper content in the ra-
tions. J. Microbiol. Epidemiol. Immunobiol. 46(8i:99-
104.
Komada, N.
1974. Studies on abnormality of bones in anomalous
"Ayu", Plecoglossus altivelis Gyobo Kenkyu 8:127-135
(In Jpn., Engl. Summ.).
KRA.NTZ, G. E.
1970. Lymphocy.stis in striped bass, Roccus saxatilis, in
Chesapeake Bay. Chesapeake Sci. 11:137-139.
Kraybill, H. F., C. J. Dawe, J. C. Harshbarger, and R. G.
TARDIFF (editors).
1977. Aquatic pollutants and biologic effects with em-
phasis on neoplasia. Ann. N.Y. Acad. Sci. 298, 604 p.
La.molet, L., C. Chepeau, and B. Rea.
1976. Etude sur les pollutions le long du littoral francais
de sud de la merdu nord. Inst. Sci. Tech. Peches. Marit.,
Dunkerque, 81 p.
L.\R0CHE, G.
1972. Biological effects of short-term exposures to hazard-
ous materials. In Proc. 1972 Natl. Conf Contr. Hazard-
ous Mater. Spills, U.S. Environ. Prot. Agency, p. 199-206.
Larsen, J. L., and N. J. Jen.sen.
1977a. An Aeromonas species implicated in ulcer-disease
of the cod iGadus morhua\. Nord. Veterinaermed.
29:199-211.
1977b. De mikrobiologiske forholdved "ulcus-syndromet"
hos torsk og relationeme til mikrobiologien i kystnaere
omrader. Nord. Veterinaermed. 29(Suppl. 1):5.
LEHNER, p. N., and A. EGBERT.
1969. Dieldrin and eggshell thickness in ducks. Nature
(Lond.) 224:1218-1219.
Levin, m. A.. R. E. wolke, and V. J. Cabelli.
1972. Vibrio anguillaritm as a cause of disease in winter
flounder, iPseudopleuronectes americanus). Can. J. Mi-
crobiol. 18:1585-1592.
Lk;htnkr, D. V.
1974. Case reports of ossifying fibromata in the striped
mullet. J. Wildl. Dis. 10:317-320.
LOOGKWOOD, A. P. M. (editor).
1976. Effects of pollutants on aquatic organisms. Camb.
Univ. Press, Camb. 193 p.
Lloyd. R.
1960. The toxicity ofzinc sulphate to rainbow trout. Ann.
Appl. Biol. 48:84-94.
1961a. Effect of dissolved oxygen concentrations on the
toxicity of several poisons to rainbow trout {Salmo
gairdnerii Richardson). J. Exp. Biol. 38:447-455.
1961b. The toxicity of mixtures of zinc and copper sul-
phates to rainbow trout {Salmo gairdnerii
Richardson). Ann. Appl. Biol. 49:535-538.
Longwell, a. C.
1976a. Chromosome mutagenesis in developing mackerel
eggs sampled from the New York Bight. Am. Soc. Lim-
nol. Oceanogr., Spec. Symp. 2:337-339.
745
FISHERY BULLETIN: VOL. 76, NO. 4
1976b. Chromosome mutagenesis in developing mackerel
eggs sampled from the New York Bight. U.S. Dep.
Commer., NOAA Tech. Memo, ERL-MESA-7, 61 p.
LoWK, J. I.
1966. Some effects of endrin on estuarine fishes. South-
east Assoc. Game Fish Comm., Proc. 19th Annu. Conf , p.
271-276.
Lowe, J. 1., P. R. P.xrrish, J. M. Patrick. Jr., .^nd J. Fores-
ter.
1972. Effects of the polychlorinated biphenyl Aroclor"
1254 on the American oyster, Crastiostrca virgini-
ca. Mar. Biol. (Berl.) 17:209-214.
Lowe, J. L,P. D. Wilson, A. J. Rick, and A. J. Wilson, Jr.
1971. Chronic exposure of oysters to DDT, Toxaphene and
Parathion. Proc. Natl. Shellfish. Assoc. 61:71-79.
LUCKE, B., AND H. SCHLUMBEROER.
1941. Transplantable epitheliomas of the lip and mouth of
catfish. J. Exp. Med. 74:397-408.
MacIntyre, F.
1974. The top millimeter of the ocean. Sci. Am.
230(5):62-77.
MAHONEY, J. B., F. H. MlDLIOE, AND D C. DEUEL.
1973. A fin rot disease of marine and euryhaline fishes in
the New York Bight. Trans. Am. Fish. Soc. 102:596-605.
MALEVSKI, Y., J. H, WALES, AND M. W. MONTGOMERY.
1974. Liver damage in rainbow trout {Salmogairdneri I fed
cyclopropenoid fatty acids. J. Fish. Res. Board Can.
31:1397-1400.
MaNN, H.
1970. Uber den Befall der Plattfische der Nordsee mit
Lymphocystis. Ber. Dtsch. Wiss. Komm. Meeresforsch.
21:219-223.
MATSUSATO, T.
1973. On the skeletal abnormalities in marine fishes. \.
The abnormal marine fishes collected along the coast of
Hiroshima Pref Bull. Nansei Reg. Fish. Res. Lab. 6, p.
17-58.
MCARN, G. E., AND S. R. WELLINGS.
1971. A comparison of skin tumors in three species of
flounders. J. Fish. Res. Board Can. 28:1241-1251.
McArn, G. E., R. G. Chuinard, B. S. Miller, R. E. Brooks,
and s. r. wellings.
1968. Pathology of skin tumors found on English sole and
starry flounder from Puget Sound, Washington. J. Natl.
Cancer Inst. 41:229-242.
McCaull, J.
1971. Building a shorter life. Environment (St. Louis)
13(7):2-15.
Mearns, a. J., AND M. J. Sherwood.
1974. Environmental aspects of fin erosion and tumors in
southern California Dover sole. Trans. Am. Fish. Soc.
103:799-810.
1977. Distribution of neoplasms and other diseases in
marine fishes relative to the discharge of waste wa-
ter. Ann. N.Y. Acad. Sci. 298:210-224.
MiDLiGE, F. H., Jr., and R. G. Malsberger.
1968. In vitro morphology and maturation of lymphocystis
virus. J. Virol. 2:830-835.
Miller, b. s., and s. R. Wellings.
1971. Epizootiology of tumors on flathead sole (Hippoglos-
soides elassodon) in East Sound, Orcas Island,
Washington. Trans. Am. Fish. Soc. 100:247-266.
Minchew, C. D., and J. D. Yarbrough.
1977. The occurrence of fin rot in mullet (Miigi! cephalus)
associated with crude oil contamination of an estuarine
pond-ecosystem. J. Fish. Biol. 10:319-323.
MI.X, M. C, H. J. PRIBBLE, R. T. RILEY, AND S. P. TOMASOVIC.
1977. Neoplastic disease in bivalve mollusks from Oregon
estuaries with emphasis on research on proliferative dis-
orders in Yaquina Bay oysters. Ann. N.Y. Acad. Sci.
298:356-373.
Mottle Y, C. McM.
1937. The number of vertebrae in trout (Sa/77io). J.Biol.
Board Can. 3:169-176.
MURCHELANO, R. A.
1975. The histopathology of fin rot disease in winter floun-
der from the New York Bight. J.Wildl.Dis. 11:263-268.
MURCHELANO, R. A., AND D. W. BRIDGES.
1976. Lumphocystis disease in the winter flounder,
Pseudopleuronecten amencanus. J. Wildl. Dis. 12:101-
103.
MURCHELANO, R. A., AND J. ZlSKOWSKI.
1976. Fin rot disease studies in the New York Bight. Am.
Soc. Limnol. Oceanogr., Spec. Symp. 2:329-336.
NAKAI,Z.,M. KOSAKA,S. KUD0H,A.NAGAI,F. HAYASHIDA.T.
KUBOTA, M. OGURA, T. MIZUSHIMA, AND I. UOTANI.
1973. Summary report on marine biological studies of
Suruga Bay accomplished by Tokai University
1964-72. J. Fac. Mar, Sci. Technol. Tokai Univ. 7:63-
117.
NEWMAN, M. W., AND S, A. MACLEAN.
1974. Physiological response of the cunner, Tautogolabrus
adspersus, to cadmium. VI. Histopathology. U.S. Dep.
Commer., NOAA Tech. Rep. NMFS SSRF-681, p. 27-33.
NIGRELLI, R. F,, K. S. KETCHEN, AND G. D. RUGGIERI.
1965. Studies on virus diseases of fishes. Epizootiology of
epithelial tumors in the skin of flatfishes of the Pacific
coast, with special reference to the sand sole iPsettichthys
melanosticus) from northern Hecate Strait, British Co-
lumbia, Canada. Zooligica (NY.) 50:115-122.
NiMMo, D. R., R. R. Blackman, a. J. Wilson, Jr., and J.
FORESTER.
1971. Toxicity and distribution of Aroclor *1254 in the
pink shrimp, Penaeus diiorarum. Mar. Biol. (Berl.)
11:191-197.
NiMMO, D. R., D. V. LIGHTNER, AND L, H. BAHNER,
1977, Effects of cadmium on the shrimps, Penaeus
duorarum, Palaeomonetes pitgio and Palaemonetes vul-
garis. In F. J, Vernberg, A, Calabrese, F. P. Thurberg,
and W. B. Vernberg (editors). Physiological responses of
marine biota to pollutants, p. 131-183. Acad. Press, N.Y.
NORDENBERG, C.
1962. Das Vorkommen der Lymphocystiskrankheit bei
Scholle und Plunder in Oresund. K. Fysiogr. Sallsk.
Lund Forh. 43:17-26.
OisHi, K., F. Yamazaki, and T. Harada.
1976. Epidermal papillomas of flatfish in the coastal wa-
ters of Hokkaido, Japan. J. Fish. Res. Board Can.
33:2011-2017.
OVERSTREET, R. M., AND H. D. HOWSE.
1977. Some parasites and diseases of estuarine fishes in
polluted habitats of Mississippi. Ann. N.Y. Acad. Sci.
298:427-462.
PACIS, M. R.
1932. An epithelial tumor of Parophrys vetulus. M.S.
Thesis, Univ. Washington, Seattle, 94 p.
Parry, j. m., d. j. Tweats, and m. a. j. al-mossawi.
1976. Monitoring the marine environment for mutagens.
Nature (Lond.) 264:538-540.
746
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
Pascoe, D., AND P. Cram.
1977. The effect of parasitism on the toxicity of cadmium to
the three-spined stickleback, Gasterosteus aculeatus
L. J. Fish. Biol. 10:467-472.
Pauley, G. B.
1969. A critical review of neoplasia and tumor-like lesions
in mollusks. Natl. Cancer Inst. Monogr. 31:509-539.
Peakall, D. B., and J. L. Linger.
1970. Polychlorinated biphenyls, another long-life wide-
spread chemical in the environment. Bioscience
20:958-964.
PEARCE, J. B.
1972. The effects of solid waste disposal on benthic com-
munities in the New York Bight. In M. Ruivo (editor),
Marine pollution and sea life, p. 404-411. Fishing News
(Books) Ltd., Lond.
Perevozchenko, L L, and C. N. Davydov.
1974. DDT and its metabolites in some cestodes in fishes.
[In Russ.] Gidrobiol. Zh. 10(61:86-90.
Perkins, E. J., J. R. S. Gilchrist, and O. J. Abbott.
1972. Incidence of epidermal lesions in fish of the North-
East Irish Sea area, 1971. Nature (Lond.) 238:101-103.
PERLMUTTER, a., D. a. SAROT, M.-L. YU, R. J. FlLAZZOLA, AND
R. J. Seeley.
1973. The effect of crowding on the immune response of the
blue gourami, Trichogaster trichopterus, to infectious
pancreatic necrosis (IPN) virus. Life Sci. 13:363-375.
Pichirallo, J.
1971. PCB's: Leaks of toxic substances raises issue of ef-
fects, regulation. Science (Wash., D.C.I 173:899-904.
PiPPY, J. H. C, AND G. M. Hare.
1969. Relationship of river pollution to bacterial infection
in salmon (Salmo salar) and suckers iCatostomus com-
mersoni). Trans. Am. Fish. Sec. 98:685-690.
PREHN, R. T.
1971. Neoplasia. In M. F. LaVia and R. B. Hill, Jr.
(editorsi. Principles of pathobiology, p. 191-241. Oxford
Univ. Press, N.Y.
RENTCHNICK, P. (editor).
1976. Neoplasms in aquatic animals as indicators of en-
vironmental carcinogens. UICC (Union Int. Cancer)
Bull. 14(2):4-5.
RIBELIN, W. E., and G. MIGAKI (editors).
1975. The pathology of fishes. Univ. Wis. Press, Madi-
son, 1004 p.
ROBOHM, R. A., AND C. BROWN.
1977. A new bacterium (presumptive Vibrio species) caus-
ing ulcers in flatfish. Abstracts of the 2d Annual Eastern
Fish Health Workshop, Easton, Md., Abstr. No. 3, 2 p.
ROBOHM, R. A., AND M. F. NITKOWSKI.
1974. Physiological response of the cunner, Tautogolabrus
adspersus, to cadmium. IV. Effects on the immune sys-
tem. U.S. Dep. Commer., NCAA Tech. Rep. NMFS
SSRF-681, p. 15-20.
R0DSAETHER, M. C, J. OLAFSEN, J. RAA, K. MYHRE, AND J. B.
Steen.
1977. Copper as an initiating factor of vibriosis {Vibrio
anguillarum) in eel (Anguilla anguilla). J. Fish. Biol.
10:17-21.
Rogers, w. a.
1 970 . A summary of fish disease cases received a five-year
period at the Southeastern Cooperative Fish Disease
Laboratory. Proc. 23rd Annu. Conf Southeast Assoc.
Game Fish Comm., p. 353-358.
1972. Disease in fish due to protozoan Epistylis (Ciliata:
Peritricha) in the Southeastern U.S. Proc. 25th Annu.
Conf. Southeast Assoc. Game Fish Comm., p. 493-496.
ROMASHOV, D. D., AND V. N. BELYAYEVA.
1966. About constancy of radiation damage chromosomes
in fish embryogenesis. Genetika 4:4-15.
Rosen, B.
1970. Shell disease of aquatic crustaceans. In S. F.
Snieszko (editor), A symposium on diseases of fishes and
shellfishes, p. 409-415. Am. Fish. Soc. Spec. Publ. 5.
Rosenthal, H., and D. F. Alderdice.
1976. Sublethal effects of environmental stressors,
natural and pollutional, on marine fish eggs and lar-
vae. J. Fish. Res. Board Can. 33:2047-2065.
Rosenthal, H., and K.-R. Sperling.
1974. Effects of cadmium on development and survival of
herring eggs. In J. H. S. Blaxter (editor). The early life
history offish, p. 383-396. Springer-Verlag, Berl.
Rosenthal, H., and R. Stelzer.
1970. Wirkungen von 2,4- und 2,5-dinitrophenol auf die
Embryonalentwicklung des Herings Clupea haren-
gus. Mar. Biol. (Beri. I 5:325-336.
Rosenthal, H. L., and R. S. Rosenthal.
1950. Lordosis, a mutation in the guppy. J. Hered.
41:217-218.
RUIVO, M. (editor!
1972. Marine pollution and sea life. Fishing News
(Books) Ltd., Lond., 624 p.
Russell, F. E., and P. Kotin.
1957. Squamous papilloma in the white croaker. J.Natl.
Cancer Inst. 18:857-861.
Ryder, J. A.
1887. On a tumor in the oyster. Proc. Acad. Nat. Sci.
Phila. 39:25-27.
Schlotfeldt, H.-J.
1972. Jahreszeitliche Abhangigkeit der "Schwarzflecken-
krankheit" bei der Garnele, Crangon crangon L. Ber.
Dtsch. Wiss. Komm. Meeresforsch. 22:397-399.
Schmidt, T. T., R. w. Risebrough, and F. Gress.
1971. Input of polychlorinated biphenyls into California
coastal waters from urban sewage outfalls. Bull. Envi-
ron. Contam. Toxicol. 6:235-243.
SELYE, H.
1950. Stress and the general adaptation syndrome. Br.
Med. J. 1:1383-1392.
1952. The story of the adaptation syndrome. Acta. Inc.,
Montreal Que., 225 p.
Shelton, R. G. J., AND K. W. Wilson.
1973. On the occurrence of lymphocystis, with notes on
other pathological conditions, in the flatfish stocks of the
north-east Irish Sea. Aquaculture 2:395-410.
Sherwood, M. J., and A. J. Mearns.
1976. Occurrence of tumor-bearing Dover sole (Micros-
tomas pacificus ) off Point Arguello, California and ofFBaja
California, Mexico. Trans. Am Fish. Soc. 105:561-563.
1977. Environmental significance of fin erosion in south-
em California demersal fishes. Ann. N.Y. Acad. Sci.
298:177-189.
SINDERMANN, C. J.
1958. An epizootic in Gulf of Saint Lawrence fishes. Trans.
23rd North Am. Wildl. Conf, p. 349-360.
1968. Oyster mortalities, with particular reference to
Chesapeake Bay and the Atlantic coast of North America.
U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 569, 10 p.
747
FISHERY BULLETIN: VOL. 76, NO. 4
1970. Principal diseases of marine fish and shellfish. Acad.
Press, N.Y., 369 p.
1976. Effects of coastal pollution on fish and fisheries —
with particular reference to the Middle Atlantic
Bight. Am. Soc. Limnol. Oceanogr., Spec. Symp. 2:281-
301.
SINDERMANN,C. J., J. J. ZISKOWSKI, AND V.T.ANDERSON, Jr.
1978. A guide for the recognition of some disease condi-
tions and abnormalities in marine fish. Natl. Mar. Fish.
Serv., Northeast Fish. Cent., Tech. Ser. Rep. 14, 65 p.
SKIDMORE, J. F.
1964. Toxicity of zinc compounds to aquatic animals, with
special reference to fish. Q. Rev. Biol. 39:227-248.
1970. Respiration and osmoregulation in rainbow trout
with gills damaged by zinc sulphate. J. Exp. Biol.
52:481-494.
SNIESZKO, S. F.
1962. The control of bacterial and virus diseases of fishes.
In Biological problems in water pollution, 3rd Semin., p.
281-282. U.S. Pub. Health Serv. Publ. 999-WP-25.
1974. The effects of environmental stress on outbreaks of
infectious diseases of fishes. J. Fish. Biol. 6:197-208.
SNIESZKO, S. F. (editor).
1970. A symposium on diseases of fishes and shellfishes.
Am. Fish. Soc. Spec. Publ. 5, 526 p.
S0RENSEN, U. S.
1977. Praeliminaere serologiske unders0gelser over
Aeromonas-og V;6no-stammer associeret med ulcus syn-
dromeincod. Nord. Veterinaermed.29(Suppl. 1):10-12.
SPARKS, A. K.
1972. Invertebrate pathology. Noncommunicable dis-
eases. Acad. Press, N.Y., 387 p.
STAINKEN, D. M.
1975. The effect of oil pollution on the physiology of A/va
arenaria, with special reference to the effect upon respira-
tion and mode of accumulation. Ph.D. Thesis, Rutgers
Univ., Newark, N.J. 160 p.
Stick, H. F., and A. B. Acton.
1976. The possible use of fish tumors in monitoring for
carcinogens in the marine environment. Prog. Exp.
Tumor Res. 20:44-54.
Stich, H. F., a. B. Acton, and C. R. Forrester.
1976. Fish tumors and sublethal effects of pollutants. J.
Fish. Res. Board Can. 33:1993-2001.
Stich, H. F., A. B. Acton, K. Oishi, F. Yamazaki, T. Harada,
T. HIBINO, AND H. G. MOSER.
1977. Systematic collaborative studies on neoplasms in
marine animals as related to the environment. Ann.
N.Y. Acad. Sci. 298:374-388.
STREET,J. C, AND R. P. SHARMA.
1975. Alteration of induced cellular and humoral immune
responses by pesticides and chemicals of environmental
concern: quantitative studies of immunosuppression by
DDT, Aroclor 1254, carbaryl, carbofuran, and methyl-
parathion. Toxicol. Appl. Pharmacol. 32:587-602.
Telford, M.
1968. The effects of stress on blood sugar composition of
the lobster, Homarus americanus. Can. J. Zool. 46:819-
826.
Telford, M.
1974. Blood glucose in crayfish — II. Variations induced by
artificial stress. Comp. Biochem. Physiol. 48A:555-560.
TEMPLEMAN, W.
1965. Lymphocystis disease in American plaice of the
eastern Grand Bank. J. Fish. Res. Board Can. 22:1345-
1356.
Udey, L. R., E. Young, and B. Sallman.
1977. Isolation and characterization of an anaerobic bac-
terium, £t;6ac?en(/w tarantellus sp.nov., associated with
striped mullet (Mugil cephalus) mortality in Biscayne
Bay, Florida. J. Fish. Res. Board Can. 34:402-409.
Ueki, N., and T. Sugiyama.
1976. On abnormality of bones in artificially produced
ayu, Plecoglossus altevelis (Preliminary Note).
[In Jpn.] Bull. Fish. Exp. Stn. Okayama Pref.
50:360-362.
Valentine, D. W.
1975. Skeletal anomalies in marine teleosts. In W. E.
RibelinandG. Migaki (editors). The pathology of fishes, p.
695-718. Univ. Wis. Press, Madison.
Valentine, D. W., and K. W. Bridges.
1969. High incidence of deformities in the serranid fish,
Paralabrax nehulifer, from southern California. Copeia
1969:637-638.
Valentine, D. W., and M. E. Soule.
1973. Effect of p,p'-DDT on developmental stability of pec-
toral fin rays in the grunion, Leuresf/iesfenuis. Fish. Bull.,
U.S. 71:921-926.
Valentine, D. W., M. E. soule, and p. Samollow.
1973. Asymmetry analysis in fishes: A possible statistical
indicator of environmental stress. Fish. Bull., U.S.
71:357-370.
Van Banning, p.
1971. Wratziekte bij platvis. Visserij 24:336-343.
VERNBERG, F. J., AND W. B. VERNBERG (editors).
1974. Pollution and physiology of marine organisms.
Acad. Press, N.Y., 492 p.
VERNBERG, F. J., A. CALABRESE, F. P. THURBERG, AND W. B.
VERNBERG (editors).
1977. Physiological responses of marine biota to pollu-
tants. Acad. Press, N.Y., 462 p.
Walsh, a. H., and W. E. Ribelin.
1975. The pathology of pesticide poisoning. In W. E.
Ribelin and G. Migaki ( editors). The pathology of fishes, p.
550-594. Univ. Wis. Press, Madison.
Warren, S., and W. a. Meissner.
1971. Neoplasms. 7n W. A. Anderson (editor), Pathology.
(6th ed.) Vol. 1, p. 529-561. C. V. Mosby Co., St. Louis.
WEDEMEYER, G.
1970. The role of stress in disease resistance of fishes. In
S. F. Snieszko (editor), A symposium on diseases of fishes
and shellfishes, p. 30-35. Am. Fish. Soc, Spec. Pub. No. 5.
WEISSENBERG, R.
1965. Fifty years of research on the lymphocystis virus
disease of fishes (1914-1964). Ann. N.Y. Acad. Sci.
126:362-374.
WELLINGS, S. R.
1969. Environmental aspects of neoplasia in fishes. In O.
W. Neuhausand J. E. Halver (editors). Fish in research, p.
3-22. Acad. Press, N.Y.
1969. Neoplasia and primitive vertebrate phylogeny:
Echinoderms, prevertebrates, and fishes — A re-
view. Natl. Cancer Inst. Monogr. 31:59-128.
WELLINGS, S. R., C. E. ALPERS, B. B. MCCAIN, AND B. S. MIL-
LER.
1976. Fin erosion disease of starry flounder (Platichthys
stellatus) and English sole (Parophrys vetulus) in the es-
748
SINDERMANN: POLLUTION-ASSOCIATED DISEASES AND ABNORMALITIES
tuary of the Duwamish River, Seattle, Washington. J.
Fish. Res. Board Can. 33:2577-2586.
Wellings, S. R., C. E. Alpers, B. B. McCain, and M. S.
Myers.
1977. Fish disease in the Bering Sea. Ann. N.Y. Acad.
Sci. 298:290-304.
Wellings, S. R., and R. G. Chuinard.
1964. Epidermal papillomas with virus-like particles in
flathead sole, Hippoglossoides elassodon. Science
(Wash., D.C.) 146:932-934.
Wellings, S. R., R. G. Chuinard, and M. Bens.
1965. A comparative study of skin neoplasms in four
species of pleuronectid fishes. Ann. N.Y. Acad. Sci.
126:479-501,
Wellings, S. R., R. G. Chlunard, r. T. Gourley, and R. a.
Cooper.
1964. Epidermal papillomas in the flathead sole, Hippo-
glossoides elassodon, with notes on the occurrence of simi-
lar neoplasms in other pleuronectids. J. Natl. Cancer
Inst. 33:991-1004.
Westernhagen. H. von, and V. Dethlefsen
1975. Combined effects of cadmium and salinity on de-
velopment and survival of flounder eggs. J. Mar. Biol.
Assoc. U.K. 55:945-957,
Westernhagen, h. von, V. Dethlefsen, and H.
Rosenthal.
1975. Combined effects of cadmium and salinity on de-
velopment and survival of garpike eggs. Helgol. wiss.
Meeresunters. 27:268-282.
Westernhagen, H. von, H. Rosenthal, and K.-R.
Sperling.
1974. Combined effects of cadmium and salinity on de-
velopment and survival of herring eggs. Helgol. wiss.
Meeresunters. 26:416-433.
WlLDISH, D. J.
1970. The toxicity of polychlorinated biphenyls (PCB) in
sea water to Gammarus oceanicus. Bull. Environ. Con-
tam. Toxicol. 5:202-204.
WOODCOCK, H. M.
1904. Note on a remarkable parasite of plaice and floun-
ders. Rep. Lanes. Sea-Fish. Labs. 1903, No. 12:63-72.
YEVicH, P. P., AND M. M. Barry.
1969. Ovarian tumors in the quahog, J, Invertebr.
Pathol. 14:266-267.
YEVICH, P. P., AND C. A. BARSZCZ.
1976. Gonadal and hematopoietic neoplasms in Mya
arenana. Mar. Fish. Rev. 38( 10):42-43.
1977. Neoplasia in soft-shell clams {Mya arenaria) col-
lected from oil-impacted sites. Ann. N.Y. Acad. Sci.
298:409-426.
Young, J. S.
1973. A marine kill in New Jersey coastal waters. Mar.
Pollut. Bull. 4:70.
Young, J. S., and j. b. pearce.
1975. Shell disease in crabs and lobsters from New York
Bight. Mar. Pollut. Bull. 6:101-105.
Young, p. h.
1964. Some effects ofsewer effluent on marine life. Calif
Fish Game 50:33-41.
ZiSKOwsKi, J., V. Anderson, and R. Murchelano.
In press. A bent fin disease of winter flounder (Pseudo-
pleuronectes americanus) from Sandy Hook-Raritan Bays,
New Jersey, and Lower Bay, New York. J. Fish. Res.
Board Can.
ZISKOWSKI, J., AND R. MURCHELANO.
1975. Fin erosion in winter flounder. Mar. Pollut. Bull.
6:26-28.
749
I
I
VERTICAL DISTRIBUTION, DIEL VERTICAL MIGRATION,
AND ABUNDANCE OF SOME MESOPELAGIC FISHES IN
THE EASTERN SUBARCTIC PACIFIC OCEAN IN SUMMER*
Bruce W. Frost and Lawrence E. McCrone^
ABSTRACT
Vertical distributions of myctophid fishes and other components of the mesopelagic micronekton were
determined during the summers of 1973-75 at two stations in the eastern subarctic Pacific Ocean.
Stratified samples were collected with a multiple net Tucker trawl so that the entire water column
extendmg to between 385 and 460 m could be sampled during a daytime or nighttime period; two to four
day and night vertical series of samples were obtained each summer. Four species of myctophids made
up 87^^ of the total fish catch: Stenobrachius leucopsarus and Diaphus theta, which performed diel
vertical migrations of 300 m vertical extent: andProtomyctophum thompsoni andS. nannochir, which
exhibited only slight diel variation in vertical distribution. Populations of each myctophid species
tended to be vertically stratified by age or size with larger individuals occurring in samples taken
progressively deeper. Two other major components of the micronekton were euphausiids and decapod
shrimps, chiefly Euphausia pacifica and Sergestes stmilis; both species were conspicuous diel vertical
migrators. Samples collected in horizontal hauls immediately following sunset showed that three
migratory species, the two migratory myctophids and E. pacifica. were closely associated with the
single migratory sound-scattering layer (12 kHz); S. similis lagged the ascent of the migratory
scattering layer. A single, deep, nonmigratory sound-scattering layer corresponded closely to the
distribution of P. thompsoni during both day and night. As in other subpolar oceanic waters, abun-
dance and standing stock of myctophids were high — 0.9 fish/m^ and 0.37 g dry weight/m^.
In 1973 we began a field study of some small
mesopelagic fishes of the family Myctophidae,
commonly known as lanternfishes or myctophids,
in the eastern subarctic Pacific Ocean. The objec-
tives of the study were to determine the vertical
distribution and migration characteristics of the
numerically dominant species, to document their
feeding behavior, and to ascertain if the distribu-
tions of fish were in any way influenced by the
distribution of their preferred prey. Myctophids
are major components of the mesopelagic fauna
throughout the world ocean, and in most areas
they are sufficiently abundant and stratified in the
water column to cause deep sound-scattering
layers (Baird et al. 1974; McCartney 1976). In-
deed, study of these fishes has been heavily
oriented toward aspects of their distribution in
relation to sound-scattering layers (e.g.. Tucker
1951; Barham 1966; Taylor 1968; Holton 1969;
Farquhar 1971; Baird et al. 1974), although some
investigations emphasized aspects of biological
'Contribution No. 1039 from the Department of Oceanog-
raphy, University of Washington. Seattle, WA 98195.
^Department of Oceanography, University of Washington,
Seattle, WA 98195.
Manuscript accepted .Mav 1978.
FISHERY BULLETIN: VOL. 76, NO. 4. 1979.
and ecological significance, such as individual
growth rates, seasonal changes in abundance, and
association among species (e.g., Pearcy and Laurs
1966; Harrisson 1967; Lavenberg and Ebeling
1967; Smoker and Pearcy 1970; Badcock 1970;
Clarke 1973; Pearcy et al. 1977). Much of the
research on myctophids has, in addition, stressed
description of the prominent diel vertical migra-
tions which are apparently undertaken by almost
all species.
In the few species studied in detail, both the
occurrence and pattern of vertical migration vary
with age or ontogeny. Larval myctophids are
nonmigratory, spending day and night in near-
surface waters (Ahlstrom 1959). Diel vertical
migration is first evident at or shortly after
metamorphosis and usually persists throughout
the remaining life of the fish, although in very old
fish, migrations may differ substantially in char-
acter from those of younger fish and may even be
supressed (Nafpaktitis 1968). Apart from this
variation with age, diel vertical migrations of
myctophids seem to be relatively regular, on a
day-to-day basis, and exhibit little or no seasonal
variation (Pearcy and Laurs 1966; Halliday 1970;
Pearcy et al. 1977). Among some species, however,
751
therfe may be a portion of the population which
does not migrate, while other members of similar
size and age do migrate (Clarke 1973; Badcockand
Merrett 1976; Pearcy et al. 1977).
Virtually nothing is known about the biological
causes or consequences of these diel vertical mi-
grations, either with respect to the myctophids or
their environment. Marshall (1954) suggested
that myctophids migrate into the surface layer
each night in order to feed on zooplankton, which
is usually most abundant in surface waters (Vino-
gradov 1968). As pointed out above, larval myc-
tophids spend both day and night in the
zooplankton-rich surface layer, but as the larvae
grow they perhaps become more conspicuous to
visual predators and, after metamorphosis, they
descend to greater depths, returning to the surface
layer only at night, if at all. Vertical migrations
may indeed have evolved as a means of avoiding or
minimizing predation, but it is unlikely that this
hypothesis can be tested in the ocean.
On the other hand, it is practicable to investi-
gate the feeding ecology of myctophid fish in rela-
tion to their migrations; for example, what types of
prey the fish utilize, when and where in the water
column the fish feed, and whether the vertical
distributions of the fish are affected by the vertical
distribution and abundance of their preferred
prey. As necessary background for such a study, in
this paper we present details of the vertical dis-
tributions of the numerically dominant species of
myctophids in the eastern subarctic Pacific Ocean.
METHODS
Study Area
We conducted the investigation during three
summer cruises in areas centered at lat. 50°N,
long. 145°W (July-August 1973 and July-August
1975; Station P in Figure 1 ) and at lat. 51°N, long.
137°W (July 1974; Station Q in Figure 1). These
stations lie within the hydrographic province des-
ignated the Central Subarctic Domain by
Dodimead et al. (1963). We chose the subarctic
region for ease of sampling and identifying the fish
and zooplankton. For example, in an earlier
meridional cruise from Kodiak, Alaska, to Hon-
olulu, Hawaii (August-September 1972), we found
that deep sound-scattering layers are fewer in
number, shallower, and more intense in the sub-
arctic region than in transition and subtropical
waters (Frost unpubl. data). Apparently related to
752
FISHERY BULLETIN; VOL. 76, NO. 4
1
M'^.c^^^
TRANSITIONAL
_j I I ; ; I L
Figure l. — Sampling stations in the eastern subarctic Pacific
Ocean. Representative hydrographic domains for summer condi-
tions after Dodimead et al. ( 1963).
this, the subarctic myctophid fauna is a simple
one; only a few species are abundant, and they are
relatively shallowly distributed in the daytime
(Taylor 1968). Further, the study area is an open
ocean environment, outside the potentially com-
plicating influences of coastal and transitional
waters (cf. McGowan 1971) and is roughly in the
middle of the latitudinal range of several species of
myctophids. Finally, the zooplankton assemblage
in subarctic waters is also less diverse than in
lower latitudes, it is well known taxonomically,
and relatively few species are abundant.
Sampling Gear
Nekton samples were collected with a modified
Tucker trawl ( Tucker 1951) described by Frost and
McCrone (1974). Briefly, the trawl had a rigid
rectangular mouth with a A-vcr area when inclined
forward at a 45° angle from vertical, and carried
five separate nets ( 6. 35-mm stretch mesh, knotless
nylon ace netting) stacked one on top of another
(much like fig. 4 in Harding et al. 1971). The net
shape followed the design of Clarke (1969). The
trawl carried an electronics package containing a
strain gage pressure transducer (range 0-1,500
Ib/in^) for determination of depth and a precision
pendulum-type tilt transducer (range 0°-90° from
vertical) for determination of angle of inclination
of the trawl mouth. A TSK (Tsurumi-Seiki
Kosakusho)-' flowmeter fitted with a magnetic
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
FROST and McCRONE: MESOPELAGIC FISHES IN THE EASTERN SUBARCTIC PACIFIC
reed switch was mounted on the top heam of the
trawl.
The trawl was towed on a two-conductor coaxial
cable and its depth, angle of inclination of the
mouth, and revolutions of the flowmeter were
monitored continuously during trawling by means
of a shipboard display unit. The nets were opened
and closed at the mouth by means of a net-tripping
assembly which was controlled electronically from
the ship. The bottom net ( without cod end ) was left
open during deployment to eliminate kiting of the
trawl when a net was first opened (Clarke 1969;
Badcock and Merrett 1976); thus, four sequential
samples could be collected in one haul. The volume
of water filtered by each net (assuming 10O7f fil-
tration efficiency) was calculated from flowmeter
revolutions and average angle of inclination of the
net mouth.
To determine vertical distributions of fish and
other components of the nekton, we towed the
trawl obliquely and collected samples on the up-
ward leg of a haul. We monitored the speed of the
ship during trawling by reference to a Doppler
ship's speed indicator.
Recordings of deep sound-scattering layers were
obtained using an Edo-Western transducer
operating at 12 kHz (pulse length 3-10 ms, beam
width about 33°) and a Precision Depth Recorder
(PDR) operating on the 0-400 fm (0-732 m) depth
range scale. At the beginning of each cruise, echo-
sounder characteristics (pulse length, power out-
put) and recorder gain were set to give optimal
resolution of the sound-scattering layer and were
not varied thereafter.
Sampling Program
Taylor (1968) found a close correlation between
distribution of abundant species of myctophid
fishes and distribution of deep sound-scattering
layers in the eastern subarctic Pacific near the
Queen Charlotte Islands. Relying on this correla-
tion, at each station we designed our sampling
program after observing the depth and migrations
of deep sound-scattering layers. The number,
depth, and migration of scattering layers were
virtually identical at Stations P and Q. We ob-
served no differences between years at Station P,
and our observations do not differ substantially
from those of Bary ( 1967) who also used a 12-kHz
echosounder in summer at Station P. In the day-
time, a single, diffuse, sound-scattering layer ex-
tended from about 275 to 375 m depth (Figure 2 A).
In the late afternoon and early evening, this scat-
tering layer became broader, chiefly by upward
movement of the top of the layer, and it persisted
with relatively little further change throughout
the night. At about 2130 h (local time), a single,
upwardly migrating layer became evident, and
within half an hour it merged completely with the
surface reverberation (Figure 2B). This migratory
scattering layer began descent at about 0530 h and
merged with the deep nonmigratory layer shortly
after 0600 h. Slight variations in times of ascent
B
0 -I
100 H
200-
300 -I
400 -
1200
2100
2130
2200
Figure 2. — 12-kHz echograms typical ofthesummer period (July-August) in the sampling areas in the northeastern Pacific Ocean. A.
Daytime record about noon, local time. B. Evening record taken on the same day showing the upward movement of the migratory
sound-scattering layer and the persistence of the nonmigratory layer at depth. Local time, depth in meters. The dark areas above 100 m
are due to surface reverberation.
753
FISHERY BULLETIN: VOL. 76, NO. 4
and descent of the migratory sound-scattering
layer depended on weather conditions; also, year-
to-year differences are attributed to slight varia-
tions in time of cruises. We usually set the lower
limit of nekton sampling at least 50 m below the
depth of the deep nonmigratory scattering layer.
With the exceptions noted below, nighttime sam-
pling was confined to the time period between as-
cent and descent of the migratory scattering layer.
Somewhat different sampling programs were
carried out in different years (Table 1 ). At Station
P in 1973 the objective was to obtain information
on vertical distributions offish and zooplankton to
aid in developing an optimal sampling strategy for
studying diet and feeding behavior of myctophids.
The 0-440 m water column was sampled in 55-m
depth strata, and seven successive vertical series
of samples, 4 night and 3 day series, were
obtained. A shallow haul (0-220 m) and a deep
haul (220-440 m) were required for each complete
vertical series. The first nighttime series was not
completed before the descent of the migratory
sound-scattering layer. In order to obtain both the
shallow and deep hauls during one night, the
hauls were of relatively short duration, and con-
sequently the nekton samples were relatively
small.
At Station Q in 1974, the objectives were to
confirm the vertical distributions found in 1973 at
Station P and to document the feeding chronology
of the common myctophids. The sampling for ver-
tical distributions ( Table 1 ) extended from the sur-
face to between 400 and 460 m, depending on the
depth of the deep sound-scattering layer, and usu-
ally included one sample collected below the scat-
Table 1 . — Sampling data for vertical series of nekton samples in
the northeastern Pacific. The three lower entries for Station P
(1975) represent data for; first, the routine day-night vertical
series (0-400 m); second, a single shallow (0-60 m) night vertical
series; and, third, a single deep (440-782 m) daytime vertical
series.
Mean
(range)
Mean (range)
Ship
duration
volume filtered
speed
of samples
per sample
Stn.
Dates
(km h)
(min)
(m^)
P 5-9 Aug.
1973 7.2 = 0.5 29(16-45) 8.412 (4,863-13,357)
O 18-22 July
1974 6.2±0.5 45(24-66) 14.229 (6,833-20,361)
P 26-28 July
1975 6,9 = 0.5 32(19-43) 10.984 (7.664-14,731)
27 July
1975 6.9=0.5 21 (16-29) 7.638 (6,184-10,052)
31 July
1975 6.9±0.5 68(58-79) 23,232 (20,012-28.614)
tering layer. Complete daytime vertical series
(both shallow and deep hauls) of samples were
collected on 2 days; as no fish were collected in the
upper 200 m, two additional daytime vertical
series were made with only one haul extending
from below the depth of the sound-scattering layer
to about 200 m. In order to achieve adequate sam-
ple size, the duration of each haul was long, but
because there were so few hours of darkness, only
one nekton haul could be made each night. Data
from a shallow haul and a deep haul on successive
nights were therefore combined to give a single,
complete, night vertical series; two such night
series were obtained, and all sampling was per-
formed between ascent and descent of the migra-
tory sound-scattering layer.
In addition to vertical series of nekton samples
taken at Station Q, we utilized two types of hori-
zontal hauls. To identify components closely as-
sociated with the migratory sound-scattering
layer, on two evenings the trawl was launched
near sunset and towed horizontally at 125 m. The
12-kHz echosounder was operated continuously
during the hauls. Approximately 30 min before
the scattering layer began to ascend from its day-
time depth, a trawl net was opened and sampling
began. The second net was opened just as the scat-
tering layer reached 125 m, and the net was closed
after the layer had passed that depth. The trawl
was towed at 125 m for an additional 30 min,
taking a third sample, then closed and retrieved.
As part of the study of diel variations in feeding
intensity of myctophids, a series of three horizon-
tal hauls, each yielding four samples of 30 min
duration, was made in the upper layer (40-m
depth) throughout one night.
The sampling program at Station P in 1975 was
similar to that at Station Q, although the nekton
sampling for vertical distributions (Table 1) was
much less extensive than in the previous two
cruises. We obtained only one complete nighttime
vertical series (400-0 m) and two deep daytime
vertical series (385-220 m) to check on vertical
distributions of myctophids. We collected one shal-
low vertical series in the 0-60 m layer in 15-m
depth strata to examine the vertical distribution
of myctophids within the surface layer at night,
and we took one very deep daytime vertical series
(782-440 m) to determine the distribution of myc-
tophids below our usual sampling depths.
All samples obtained with the nekton trawl
were preserved in a 47f formaldehyde seawater
solution buffered with sodium borate.
754
FROST and McCRONE; MESOPELAGIC FISHES IN THE EASTERN SUBARCTIC' PACIFIC
Analysis of Samples
All organisms in the nekton samples were
counted. Fish were identified from descriptions in
Hart (1973) and Wisner (19761. The standard
length (SL) (distance from the tip of the snout to
the end of the vertebral column) of each fish was
measured to the nearest millimeter. Among the
invertebrates collected in nekton samples, only
euphausiids and decapod shrimps were consis-
tently captured in substantial numbers. The
numbers offish and shrimp were standardized to
number per 10,000 m-' of water.
Myctophid fish, preserved for about 3 yr, were
sorted from samples for determination of body
length and dry body weight. Intact, undamaged
specimens were dried to constant weight at 65°C
on glass slides in a drying oven. Drying usually
took 3-4 days, but up to 10 days for some of the
largest fish. Dried fish were weighed to the nearest
milligram. The relationship between standard
body length and dry body weight for each species of
myctophid was determined by linear regression
analysis of logarithmically transformed mea-
surements.
RESULTS
The vertical series of nekton samples collected
at Stations P and Q yielded nine species of myc-
tophids, one abundant species of chauliodontid,
and one relatively raremelanostomiatid. All other
families combined made up only 2-Wi of the total
catch by number (Table 2). In addition to fish, the
samples contained considerable numbers of
euphausiids and decapod shrimps. Other inver-
tebrate groups, such as siphonophores and squids,
were only sporadically captured. The more com-
mon fishes had similar relative abundances at the
two stations. More than 80^- of the myctophids
consisted of three species (Stenobrachius leucop-
sarus, Protoniyctophum thompsuni , and Diaphus
thetu ) whose vertical distributions were generally
well bracketed by the sampling. The other species
of myctophids were either rare or appeared to be
distributed below the usual range of sampling;
therefore, emphasis in this study was placed on
the above three, abundant, relatively shallowly
distributed species.
Vertical Distribution of Fish
The most abundant fish in our samples was S.
leucopsarus. Its only congener, S. nannuchir. was
rarer (Table 2). As discussed later, with the excep-
tion of the very deep vertical series ( 782-440 m ) in
1975, only small specimens ( <35 mm) of S. nan-
nochir occurred in the vertical series. The fish
caught at Station P in 1973 were in such poor
condition that it was not possible to discriminate
the smaller specimens of the two species of Steno-
brachius. However, there is evidence (presented
below) that S. nannochir was extremely rare at
Station P in 1973, much rarer than at Station Q or
Station P in 1975 (Table 2). Redesign of the cod
ends, after the 1973 cruise, provided us with good
specimens which permitted discrimination of the
two congeners.
T.ABLE 2. — Composition of the total fish catch in vertical series of nekton samples in the
northeastern Pacific. Data for each station combine all vertical series.
Family and species
Myctophldae:
Stenobrachius leucopsarus
S nannochir
Protomyctophum thompsoni
Diaphus Iheta
Tarletonbeania crenulans
Lampanyclus regalis
L. ritteri
Notoscopelus japonicus
Symbolophorus californiense
Chauliodontidae:
Chauliodus macouni
Melanostomiatidae:
Tactostoma macropus
Others
Totals
Station P,
1973
Station Q,
. 1974
Station P.
1975
No,
O.
No.
°o
No.
%
720
638
1.038
454
461
49.3
(')
(')
268
11.7
92
9.8
125
11.1
413
18.1
210
22.5
111
9.8
304
13.3
61
65
16
1.4
26
1.1
3
0.3
9
0.8
11
0.5
0
0
3
0.3
11
05
0
0
1
0.1
0
0
0
0
1
0.1
0
0
0
0
57
1.129
5.0
147
6.4
2.288
84
935
9.0
21
19
15
0.7
3
03
64
5.7
55
2.4
21
22
'Due to the poor condition of the fish caught at Station Pin 1973. it was impossible to discriminate the smaller
specimens of the two species of Stenobrachius It is possible that some of the fish listed here as S leucopsarus
•Mere in fact S, nannochir. but for reasons described in the text, we do not believe this to be the case.
755
FISHERY BULLETIN; VOL 76, NO, 4
At Station P in 1973, S. Icmopsarus occurred in
largest numbers in the surface layer (0-55 m) at
night and at 275-330 m during the day (Figure
3A). This pattern could occur if most specimens
undertook a diel vertical migration over a depth
range of 250-300 m. To be certain that the data
reflect a vertical migration and not simply light-
aided avoidance of the net by fish in the daytime, it
was necessary to compare day and night total
catches of fish integrated over the water column
sampled (Table 3 ). Assuming that the entire verti-
cal range of S. leiicopsarus was sampled (this as-
sumption is qualified below), then it is clear that,
since the total catches for day and night series
were statistically indistinguishable (Table 3),
there was no evidence of daytime avoidance of the
trawl by fish. Further, judging from the results of
replicate sampling of zooplankton with nets
(Wiebe and Holland 1968), the day and night to-
tals in Table 3 are well within the range of vari-
ability expected for repeated samples from a
pelagic population. Thus, the observed diel differ-
n
Nl 01 N2 D2 N3 D3 N4
c
B
FIGLIRE 3. — Vertical distribution of Stenobrachius leucopsarus
at Stations P and Q in the northeastern Pacific Ocean. A. 1973,
Station P: four night (Nl and three day (D) vertical series. B.
1974, Station Q: four day and two composite night vertical series.
C. 1975, Station P: two composite night and two day vertical
series. The profiles for each year are presented in the chronologi-
cal order in which they were taken. Scales represent 100
individuals/10'' m^.
D 1 D 2 N 1.2 D 3 N 3,4 D 4
Table 3. — Day and night total catches for the water column sampled, of selected fish and crustacean
species at Stations P and Q in the northeastern Pacific; means and ranges (parentheses) as number/100
m^. Ratios given of largest to smallest estimate of abundance for a station.
Species
Station
Day
Nigtit
Ratio
Stenobrachius leucopsarus
Diaphus Iheta
Protomyctophum thompsoni
Chauliodus macouni
Euphausia pacifica
Sergestes similis
p,
Q,
P.
1973
1974
1975
677
101 1
28 0
(28,6-820)
(48 4-153 6)
(20 3-358)
59 8
62 1
37,6
(47,3-73,7)
(56,5-67 7)
(29 2-46 1)
2.9
3.2
23
P.
Q,
P.
1973
1974
1975
6,8
226
36
(1 1-13,8)
(15,2-34,4)
(3 1-4.0)
12,4
'11.5
54
(3,9-18.2)
(11.0-12.0)
(3.6-10.4)
16.5
3.1
34
P,
Q,
P
1973
1974
1975
156
26 7
173
(12 7-20 9)
(21 6-39 1)
(17 2-17 3)
7.9
23.6
47,0
(0.0-13.8)
(20.0-27.3)
(28.9-65.2)
24.8
2.0
38
Q,
1974
10,1
(3,2-15,7)
8,2
(7.5-90)
4.9
P,
Q.
P
1973
1974
1975
1697
4174
3225
(242-323,9)
(272 0-531,5)
(223 8-421,3)
'1.075 3
154 0
423 7
(214 5-5.825 0)
(175-2905)
(294 5-532 9)
2407
304
25
P,
Q
P.
1973
1974
1975
19,8
330
10,2
(0-5-31,3)
(14,0-47,6)
(9,3-11.0)
'50.3
'72,9
111
(28 0-80.8)
(57.5-883)
(9.6-12.6)
161.6
6.3
1.4
'Day and night abundance significantly different (P-
^Estimate based on smallest nonzero catcfi.
0.1) by rank test (Tate and Clelland 1957).
756
FROST and McCRONE: MESOPELAGIC FISHES IN THE EASTERN SUBARCTIC PACIFIC
ences in the vertical distribution of S. leucupsarus
(Figure 3A) indicate that the majority of individu-
als do perform a diel vertical migration.
The occurrence of some S. leucopsarus in the
deepest samples at night (Figure 3A) indicates
that the entire population was not participating in
the vertical migration described above. Differ-
ences in migratory pattern appear to be largely a
function of size or age of individual fish. The
length-frequency histogram for our entire catch of
S. leucopsarus at Station P in 1973 (Figure 4A)
indicates that several size classes of fish were
sampled. Since S. leucopsarus metamorphoses to
the juvenile stage at 18 mm (Smoker and Pearcy
1970), the abundant 19-35 mm size class (Figure
4A) probably represented the youngest juvenile
fish. The largest specimens caught at Station P
attain the maximum size expected for S. leucop-
sarus, 85-111 mm (Kulikova 1957; Smoker and
Pearcy 1970).
To determine the effect of size on vertical migra-
tion, we examined the three obvious size classes of
S. leucopsarus: 19-35 mm, 38-82 mm, and 90-112
mm. The smallest size class ( 19-35 mm) performed
a clear diel vertical migration from 275-330 m in
the daytime to 0-55 m at night (Figure 5A). The
anomalously low density of fish on the third day
must be attributed to horizontal patchiness offish.
Note especially that only on one night ( N3, Figure
5A) was one small-sized Stenobrachius captured
below 275 m. The medium size class (38-82 mm)
shows a similar migration (Figure 5B), though
these fish seemed to be more dispersed vertically,
both at night and in the daytime, than the small-
est size class. The high density of medium-sized
fish at 275-330 m on the first night, not apparent
on the other three nights, probably reflected the
fact that this sample was collected between 0554
and 0613 h, a time period when the migratory
sonic scatterers, and presumably myctophids,
were descending. Some of the medium-sized fish
probably had already descended into the 275-330
m layer at the time this sample was collected. The
largest size class of S. leucopsarus (90-112 mm)
had a pattern of vertical distribution totally dif-
ferent from those of the two smaller size classes.
Individuals of the largest size class were not cap-
tured at all in the first two daytime series and were
caught only in the two deepest samples in the third
daytime series ( Figure 5C ). They were captured in
all four night series, but never in the surface layer
(0-55 m), and in three of the four night series, the
greatest density of large-sized fish occurred be-
tween 330 and 440 m. It is tempting to conclude
from these data that the individuals of the largest
size class also perform a diel vertical migration,
moving from daytime depths below our lower limit
of sampling (440 m) into our sampling range at
night. Of course, a similar vertical distribution
pattern could be obtained if the largest fish avoid
the trawl in the daytime, although it seems un-
likely that all fish of this size class could effec-
tively do so. Nevertheless, with the data from
Station P (1973), it is impossible to discriminate
between these two possibilities for the largest
fish.
Stenobrachius leucopsarus had a very similar
pattern of distribution and diel vertical migration
at Station Q (Figure 3B). The length-frequency
distribution of the species was strongly skewed to
juvenile fish 1 19-31 mm), which made up 88. 79^ of
the total catch of the species. There was only one
relatively distinct secondary mode, consisting of
very large fish (81-108 mm), which composed 3.9%
of the total catch (Figure 4B). Fish in the smallest
mode and also the rarer intermediate sizes offish
(32-79 mm) were clear vertical migrants, closely
following the pattern described above for Station
P, and there was no difference in vertical distribu-
tion between the small- and medium-sized fish.
Also, as at Station P, representatives of the largest
size class offish were captured, with the exception
of 1 fish ( out of 45 caught ) in the deepest sample on
day 4, only in the night hauls and almost always
(43 out of 44 fish) below 50 m.
At Station P in 1975, the same patterns of diel
vertical migration ( Figure 3C ) and size-dependent
variation in vertical distribution of S. leucopsarus
were evident, though far fewer fish were collected,
both because of the fewer vertical series taken and
an apparent decrease in abundance of the species
compared with the previous 2 yr (Table 3). This
decrease appears due partly to reduced abundance
of the smallest size class (17-32 mm) which made
up only 47.2% of the total catch in 1975 (Figure
4C ), compared with 62.6% at Station P in 1973 and
88.7% at Station Q. In the one deep daytime verti-
cal series (782-440 mm) at Station P, large S.
leucopsarus were captured between 440 and 740 m
(Table 4), thus supporting our earlier hypothesis
that the largest fish caught at night above 440 m
migrated in the daytime below our usual range of
sampling. However, extensive day and night sam-
pling over the entire vertical range of the large
fish is required to completely rule out daytime
avoidance of the trawl.
757
FISHERY BULLETIN; VOL. 76, NO. 4
70 -1
60-
_i
<
>
40 -
30 -
LlJ
I 20
Z)
10
A
Mbjy Uhl MmE
50 60 70
LENGTH (mm)
lruD_, 7
80 90
irm rin
T
100 no
200 -
<
5l50
>
z
100 -
cc
UJ
CD
5
3 50
B
n r
20 30
In I n.l n hji-i n ip — cam a " "^ — I" ■ — • f ^ " n -r-TTl-w-t I -
40 50 60 70 80
LENGTH (mm)
90 100
n — I
no
35
30 -
<
g 25H
>
5 20-1
0 15 H
cr
UJ
1 10 H
3
Z
c
M
20
r
30
Myrm ^[]mn n ^mlllNJlTTi
I ' I ' I ' ' 1 1'" I" I ""I I I — I — I
40 50 60 70 80 90 100 110
LENGTH (mm)
Figure 4. — Length-frequency distributions of
Stenobrachius leucopsarus from all vertical
series. A. 1973, Station P.N = 720. B. 1974,
Station Q, TV = 1,038. C. 1975, Station P,iV =
461.
758
FROST and McCRONE: MESOPELAGIC FISHES IN THE EASTERN SUBARCTIC PACIFIC
fl
Nl 01 N2 D2 N3 D3 N4
B
Nl 01 N2 02 N3 03 N4
c
Nl 01 N2.D2 N3 03 N4
Figure 5. — Vertical distribution of three sizes of Steno-
brachius leucopsarus at Station P, 1973. A. 19-35 mm, scale
represents 100 individuals/lO^m^. B. 38-82 mm, scale represents
50 individuals/lO'' m^. C. 90-112 mm, scale represents 25
individuals/10'* m^. Sequence of vertical series as in Figure 3A.
The one shallow night vertical series (60-0 m) at
Station P indicated that S. leucopsarus were dis-
tributed throughout the surface layer but were
concentrated between 15 and 30 m (Figure 3C).
Examination of sizes of fish caught in this series
suggested very fine-scale vertical stratification by
age or size (Table 5). Recall that in all other verti-
cal series taken at night very large fish ( >80 mm)
were always captured (except for one fish) below
50 m. Because we took only one such shallow ver-
tical series, we cannot evaluate the frequency of
occurrence or temporal persistence of this appar-
ent stratification offish by age in the surface layer
at night.
The third most abundant myctophid in our sam-
ples, Diaphus theta, also performed a diel vertical
migration (Figure 6); there was no consistent dif-
ference between day and night catches (Table 3).
At night, D. thcta ranged over the upper 1 65 m but
was concentrated near the surface (0-55 m), while
during the day most of these fish were collected
below 275 m. As stated above for 5. leucopsarus,
the occurrence of D. theta at 275-330 m the first
night at Station P ( 1973 ) is misleading because the
sample was probably collected after the downward
vertical migration of myctophids had begun.
The size range for the total catch of D. theta was
36-88 mm in 1973 and 46-84 mm in 1975 at Station
P, and 33-76 mm at Station Q. The size-frequency
distributions were similar in all 3 yr. Considering
only Station Q, for which we have the largest col-
lection, the size-frequency distribution (Figure
7A) was quite different from S. leucopsarus (Fig-
ure 4B). Small and large fish were rare and the
samples contained primarily intermediate sizes
(45-58 mm). Distinguishing, somewhat arbi-
trarily, three classes in the size-frequency dis-
tribution, there is indication of size-dependent
Table 5. — Vertical distribution of size classes ofStenobrachius
leucopsarus in the shallow night vertical series at Station P
(1975) in the northeastern Pacific, as number/10,000 m^. Data
based on a single haul with a single sample at each depth.
Depth (m)
17-32 mm
37-82 mm
•85 mm
0-15
37,2
16
0
15-30
1405
136
0
30-45
7.8
468
0
45-60
15.9
42,8
1.0
Total no captured
138
89
1
Table 4. — Deep daytime vertical distribution of selected species of micronekton at Station P (1975)
in the northeastern Pacific, as number/10,000 m^. For Stenobrachius leucopsarus, the numbers in
parentheses are abundances of large fish (91- 1 12 mm SL). Data based on a single haul with a single
sample at each depth.
Stenobrachius Stenobrachius Prolomyctophum Lampanyctus Chauliodus Sergestes
Depth (m) leucopsarus nannochir thompsoni ritten macouni similis
440-540
4.5 (3,5)
49,6
0.7
0
2.4
10.8
540-640
8,5 (7.5)
70
0
0
0
7.5
640-740
4,4 (1,3)
4,9
0
0.9
0.4
0
740-782
0
2,7
0.5
0.5
0
0
Total no
captured
40 (28)
173
3
3
8
45
759
FISHERY BULLETIN: VOL. 76. NO. 4
n
Nl Dl N2 D2 N3 D3 N4
B
0 1 D 2 N 1 .2 D 3 N 3.4 D 4
U
100
: "^
T
n
200
X
li 1
300
1
(_)
1
4UU
t;nn
c
N 1
N 2
D 1
D 2
FIGURE 6.— Vertical distribution oWiaphus theta. A. 1973, Sta-
tion P. B. 1974, Station Q. C. 1975, Station P. Scales represent 25
individuals/10'' m^. Sequence of vertical series as in Figure 3.
variation in vertical distribution and vertical mi-
gration. The smallest sizes ( 35-44 mm ) offish were
consistently shallower than larger sizes both dur-
ing the day and at night (Table 6); although the
numbers offish are small, they do indicate a possi-
ble trend. Diaphus theta was not captured in the
very deep (782-440 m) daytime vertical series at
Station P in 1975.
The second most abundant myctophid, Pro-
tomyctophum thompsoni , did not perform an ex-
tensive die! vertical migration similar to that of S.
leucopsarus or D. theta; it remained below about
200 m both day and night (Figure 8). Neverthe-
less, the species tended to be somewhat more shal-
lowly distributed at night than in the daytime.
This is best demonstrated by the data from Station
Q where the largest catches of this species were
made. At Station Q,P. thompsoni ranged from 16
30-,
25-
20-
15 -
jR
A
30
1 ' r
50 60
LENGTH ( mm )
80
70 -1
50 -
30 -
20 -
3
10 -
r
B
r
Jl^nrrdV
" flr^^
1 1
0 20
1 1 ' ' 1
30 40
50 6
LENGTH (mm )
Figure 7. — Length-frequency distributions of Diaphus theta
{A),N = 304, andProtomyctophum thompsoni {B),N = 413, from
all vertical series at Station Q, 1974.
to 53 mm SL and the length-frequency distribu-
tion of the population was bimodal (Figure 7B).
Calculations of mean depths of the two size classes
showed that the smaller fish were always slightly
more shallowly distributed that the larger fish
(Table 6). Moreover, both size classes tended to be
deeper in the daytime than at night, although the
average change in depth (30-40 m for both size
classes) was relatively small (Table 6). Protomyc-
tophum thompsoni was rare below 440 m at Sta-
tion P in 1975 (Table 4). The size range of the
species at Station P was 18-51 mm (1973) and
16-50 mm ( 1975), and the size-frequency distribu-
tion was similar to that of Station Q.
The above three species of myctophids had ver-
tical distributions which were, with the possible
exception of the rare large specimens of S. leucop-
sarus, well bracketed by our vertical series of
samples. Two other relatively abundant species of
fish seemed to have vertical distributions which
760
FROST and McCRONE: MESOPKLACMf FISHES IN THE EASTERN SUBARCTR- PACIFIC
Table 6. — Mean depth (meters) of size classes ofDiaphus theta
and Protomyctoph urn Ihompsoni in day (D1-D4) and night (Nl.
N2) vertical series at Station Q in the northeastern Pacific. Mean
depth/) was calculated from the equation D = i;!,Z,/i/!, , where
n I is the population density (number/ 10,000 m^) of a size class in
sample; andZ, is the midpoint of the depth range of sample i.
Size class
(mm)
D1
D2
D3
D4
Nl
N2
Total no
captured
Diaphus theta
35-44 342
357
325
336
25
25
40
45-58 394
382
377
351
32
30
224
58 390
400
404
344
76
69
40
Protorrtyctophum thompsoni:
16-35 332 316 330 301 294 257
36-53 381 338 340 340 307 309
354
59
fl
Nl Dl N2 D2 N3 D3 N4
B
0 2 N 1.2 D 3 N 3.4 0 4
E 200
c
300
400
500
N 2
N 1
D 1
D 2
Figure 8. — Vertical distribution of Protomyctophum
thompsoni. A. 1973, Station P. B. 1974, Station Q. C. 1975,
Station P. Scales represent 25 individuals/10^ m^. Sequence of
vertical series as in Figure 3.
extended deeper than our usual range of sampling.
Stenobrachius nannochir was only captured below
275 m in the routine vertical series at Stations Q
(1974) and P (1975). As noted earlier, due to the
poor condition of the catch, small specimens of S.
nannochir and S. leiaupsarus were not distin-
guished in samples from Station P ( 1973). Half of
the total catch ofS. nannochir was from below 400
m, and all of the specimens caught above 440 m
were <35 mm SL. It is for this reason that we
think that the species must have been extremely
rare in the 0-440 m layer at Station P in 1973, for
we caught almost no small Stenobrachius in the
deep samples at night (Figure 5A). The virtual
restriction of catches of S. nannochir to our
deepest samples, day and night, indicates that its
distribution probably extended below our range of
sampling. Indeed, it was the most abundant fish in
the one very deep daytime vertical series at Sta-
tion P ( 1975); it occurred down to 782 m and was
concentrated in the 440-540 m layer (Table 4).
Furthermore, an interesting vertical stratifica-
tion by size was evident in this series, with the
smallest fish dominating the shallowest sample
and largest fish dominating the deepest two sam-
ples (Table 7). Note that we captured only small
specimens ( <35 mm) in all of the other, shallower
vertical series. Stenobrachius leucopsarus and S.
nannochir of similar body size tended to be verti-
cally well separated in the water column at all
times (Tables 4, 7; Figure 5).
Table 7. — Vertical distribution of size classes of Stenobrachius
nannochir in the deep daytime vertical series at Station P ( 1975)
in the northeastern Pacific, as number/10,000 m^.
Depth (m)
22-37 mm
38-70 mm
85-113 mm
440-540
37.0
12.2
0.3
540-640
0.5
65
0
640-740
0
22
2.7
740-782
0
09
1.8
Total no. captured
107
55
11
The only other moderately abundant fish was
the chauliodontid Chauliodus macouni, and only
at Station Q was it captured in sufficient numbers
to warrant description. Chauliodus macouni al-
ways occurred below 150 m, and there was no
conclusive evidence of change in its vertical dis-
tribution during the day-night cycle (Figure 9,
Table 3). However, in contrast to P. thompsoni,
whose range of vertical distribution apparently
was well sampled day and night ( Figure 8B, Table
4), it appears from the abrupt truncation of histo-
grams in Figure 9 that the deepest portion of the
population of C. macouni was not sampled either
in the daytime or at night. Indeed, in the very deep
vertical series at Station P ( 1975), a number of C.
macouni were captured in the 440-540 m layer
761
FISHERY Bl'LLETIN: VOL. 76, NO. 4
500
D 1
D 2
N 1,2
D 3
N 3.4
D 4
Figure 9. — Vertical distribution ofChauliodus macouni at Sta-
tion Q, 1974. Scale represents 25 individuals/10'* m^. Sequence of
vertical series as in Figure 3B.
Table 8. — Abundance fnumber'10,000 m^) of Stenohrachius
leucopsarus and Diaphus thcta in three series of half-hour sam-
ples collected in horizontal tows at 40-m depth during one night
at Station Q in the northeastern Pacific. Sampling commenced
after the migratory scattering layer had merged with the surface
reverberation and terminated after the scattering layer had
descended below the surface reverberation (0425). Time is when
net was opened.
S
D
S.
0.
Time
leucopsarus
theta
Time
leucopsarus
theta
2200
157
31
0130
262
42
2230
89
15
0200
329
36
2300
102
10
0300
188
72
2330
109
8
0330
178
66
0030
80
21
0400
146
28
0100
197
54
0430
8
0
(Table 4), indicating that the distribution of this
fish probably extended below the normal limit of
sampling in the routine vertical series. At Station
Q, specimens of C. macouni ranged from 29 to 189
mm SL. Very large fish ( >100 mm) were usually
captured at night in the deepest samples, but for
fish <100 mm there was no clear trend of size-
dependent variation in vertical distribution.
Other fish species (Table 2) occurred sporadi-
cally in the samples and were caught primarily at
night: the only daytime catches were below 300 m
(e.g., Lampanyctus ritteri in Table 4). Included in
the category "Others" in Table 2 were members of
the families Bathylagidae, Gonostomatidae,
Melamphaeidae, Opisthoproctidae, Paralepidi-
dae, and Scopelarchidae.
Variability in Abundance of
Myctophids in Replicated Samples
With a few exceptions, the estimates of abun-
dance of myctophids integrated over the water
column sampled did not vary by more than a factor
of 4 between vertical series within cruises (Table
3). At Station Q, three series of half-hour horizon-
tal hauls were made at 40 m throughout one night
(Table 8). Excluding the sample (0430) collected
after the scattering layer had descended, concen-
trations of S. leucopsarus varied by a factor of
about 4, those for D. theta by a factor of about 9. For
both species, there was a significant trend (P =
0.05, run test, Tate and Clelland 1957) toward
increased abundance during the night, and their
abundances were strongly correlated (rank differ-
ence correlation coefficient 0.74, P ~ 0.01, Tate
and Clelland 1957). Myctophids were abundant in
the surface layer until the migratory scattering
layer descended.
Estimated Abundance and
Standing Stock of Fishes
Our data for mean abundance of all fishes cap-
tured for the 3 yr ranged from 0.78 to 1.61/m^ for
the water column extending to between 385 and
460 m (Table 9). The three most abundant species
of myctophids combined accounted for 77-859r by
number of all fish collected. There was no consis-
tent difference between day and night estimates of
concentrations offish.
Equations for the regression of dry body weight
on body length (Table 10) were used in conjunction
with the lengths and abundance offish from each
sample to calculate the population standing stocks
of S. leucopsarus, D. theta, and P. thompsoni for
Table 9. — Estimated mean abundance and standing stock of
mesopelagic fishes at Stations P and Q in the eastern subarctic
Pacific Ocean. Myctophids includes only the three most abun-
dant species, Stenohrachius leucopsarus, Diaphus theta, and
Protomyctnphum thompsoni . Estimated mean abundance and
standing stock are based on average of all day and night vertical
series; values in parentheses are means for night vertical series
only.
Abundance
(no 'm^)
Standi
ig stock (g dry wt/m^)
Station
Myctophids
All fishes
Myctophids
P, 1973
Q. 1974
P, 1975
0,85
1.24
0.61
1-00
1.61
0.78
0.53 (0.77)
0-.27 (0.39)
0.34 (0.55)
Table 10. — Equations for the regression of dry body weight, W
(grams), on body length, L (centimeters), for three species of
myctophids.
Regression
Range of
Species
equation
SL (cm)
N
Stenohrachius
leucopsarus
W = 0,00125/.^"*
2.0-11-8
92
Diaphus theta
W = 0.00537 1^*'^
3.0-7,4
79
Protomyctophum
thompsoni
W = 0.00212/.""
1.7-4.9
54
762
FROST and McCRONE: MESOPELAGIC FISHES IN THE EASTERN SUBARCTIC PACIFIC
each vertical series of nekton samples at each sta-
tion. A slight (90 mm), specimens of S. leucopsarus which
were caught chiefly at night.
Vertical Distribution of Crustaceans
The most abundant organisms in the vertical
series of nekton samples were euphausiids, pre-
dominantly large individuals (>12 mm total
length). At Station P in both 1973 and 1975,
Euphausia pacifica made up more than 80% of the
euphausiid catch by number. At Station Q, 51% of
the total euphausiid catch was E. pacifica; other
species were Thysanoessa spinifera (30%), Tes-
sarabrachion occulatum (9%), Thysanoessa lon-
gipes (8% ), and Stylocheiron maximum (2% ). All of
these species also occurred at Station P, but were
rare. Consequently, only the data for E. pacifica
are presented here.
During the day, large E. pacifica occurred in
greatest concentration between 275 and 400 m,
while at night they were usually concentrated in
the upper 55 or 60 m (Figure 10). No consistent
difference between day and night total catches
was evident, but sporadic, extraordinarily large or
small catches of £. pacifica were obtained in both
1973 and 1974. Variations such as these are com-
mon in euphausiid catches (Brinton 1962b) and
are usually attributed to horizontal patchiness.
Our ranges of estimated abundances were con-
sequently very large (Table 3). The other four
species of euphausiids were too rare to draw
definite conclusions about their distributions.
The penaeid decapod shrimp, Sergestes similis,
was the only other abundant invertebrate in our
nekton samples. At Station P (1973) and Station
0
100-
5 200-
6 300 f
400
500
0
100
5 200
n
fc 300
CD
400
500
■
1 L i
1
n
Nl Dl N2 02 N3 03 N4
+
1
B
0 1 0 2 N 1.2 0 3 N 3.4 0 4
500
c
N 1
N 2
0 1
D 2
Figure lO.— Vertical distribution of Euphausia pacifica. A.
1973, Station P. Scale represents 1,000 individuals/lO" m^. (The
0-55 m sample on the fourth night represents 10,447
individuals/lO-* m^*.) B. 1974, Station Q. Scale represents 500
individuals/ia» m^. C. 1975, Station P. Scale represents 500
individuals/10^ m^. (The 15-30 m sample on the first night repre-
sents 3,086 individuals/10^ m^*.) Sequence of vertical series as in
Figure 3.
763
FISHERY BULLETIN; VOL. 76, NO. 4
Q, the species appeared to be performing an exten-
sive diel vertical migration (Figure 11 A, B); how-
ever, the average daytime catches at both stations
were a bit less than half the average nighttime
catches, though only in 1973 and 1974 were there
statistically significant differences (Table 3). Ex-
cept for the largest size class of Stenobrachius
leucopsarus (Figure 5C), Scrgestes similis is the
only species for which we found such a prominent,
repeated, day-night difference in catches. Either
S. similis is a diel vertical migrator and descends
below our usual range of sampling in the daytime
or it is capable of avoiding the nekton trawl in the
daytime. Our very deep daytime vertical series
taken at Station P (1975) bears on this question.
Although the species seemed considerably less
abundant in 1975 (Table 3), this was probably
n
Nl Dl N2 02 N3 D3 N4
B
01 02 N1.2 03 N3.4 04
100
1
E 200
1
n
S: 300
a
II
1
1
400
son
c
N 1
N 2
0 1
0 2
Figure ll.— Vertical distribution ofSergestes similis. A. 1973,
Station P. B. 1974, Station Q. C. 1975, Station P. Scales repre-
sent 50 individuals/10' m''. Sequence of vertical series as in
Figure 3.
partly due to the shallower depth to which the
routine vertical series extended in the daytime. In
the very deep daytime vertical series, S. similis
occurred in considerable numbers between 440
and 640 m (Table 4 ). Thus it probably was a migra-
tor and in the daytime ranged well below the
greatest depth of sampling on routine vertical
series.
At both stations, S. similis tended to be rather
broadly distributed over the 0- 1 50 m layer at night
and often was more abundant below 50 m than
above ( Figure 11). In this respect its diel migration
differs from that of the two migratory myctophid
fishes and E. pocifica, which tended to aggregate
strongly above about 60 m at night.
In addition to S. similis several other types of
malacostracans were collected in the samples: the
caridean decapods Hymenodora frontalis, Noto-
stomiis japoniciis, and Pasiphaea sp.; the penaeid
decapod Bentheogennema borealis; and the my-
sids Gnathophausia gigas, Boreomysis sp., and
Eucopia sp. All were rare, were collected only at
night, and almost always occurred below 200 m.
Micronekton Associated With
Sound-Scattering Layers
In the daytime, the position of the scattering
layer corresponded closely with the daytime depth
of occurrence of the smaller size classes of Steno-
brachius leucopsarus and the populations of D.
theta and Protomyctophum thompsoni (Figure
12A). For example, in the profiles shown in Figure
12A, the 300-400 m stratum contained an average
concentration of 136 fish/10,000 m^ of the three
species combined. Sergestes similis is distributed
too broadly and deeply in the daytime to contrib-
ute to the observed scattering layer (Figure IIB,
Day 3). Excluding euphausiids, in our samples no
other potential sound-scattering organism (e.g.,
physonect siphonophores) consistently had its
center of abundance between 275 and 400 m in the
daytime. The large E. pacifica collected with the
nekton trawl had a pattern of vertical distribution
(Figure lOB, Day 3) very similar to that of the
migratory myctophid fishes.
Comparison of the vertical distribution and diel
migration of Stenobrachius leucopsarus with the
echosounder trace indicates a correlation between
the fish and the migratory sound-scattering layer
(Figures 2, 3). The correlation is best for individu-
als of the small and medium size classes (Figure 5).
Similarly, the vertical distribution and diel mi-
764
FROST and McCRONE: MESOPELAGIC FISHES IN THE EASTERN SUBARCTIC PACIFIC
300
+ #T
,■,.;, v*;
On
100-
200-
B
T
200tish/IO''m3
S leucopsarus D theta
f
p. thompsont
Figure 12. — Vertical distribution of three species of myctophids
relative to a sound-scattering layer recorded with the 12-kHz
echosounder. A. Midday distribution of fish and an echogram
showing the position of the scattering layer at the time of sam-
pling (Day 3 at Station Q). B. Nighttime distribution offish and
an echogram showing the position of the scattering layer at the
time of sampling (Nights 3, 4 at Station Q).
gration of D. theta closely parallel the behavior of
the migratory sound-scattering layer (Figures 2,
6). To examine this relationship more closely, at
Station Q two series of three horizontal samples
each were collected at 125 m in the periods preced-
ing, during, and after ascent of the migratory
sound-scattering layer past that depth. In the first
series (Table 11, 17 July) bothS. leucopsarus and
D. theta were most abundant in the sample col-
lected as the scattering layer was passing 125 m.
Euphausiids (predominantly E. pacifica) and
Sergestes similis were also abundant in the sam-
ples; however, maximum concentrations of each
were obtained either in the sample collected before
or after the scattering layer had passed 125 m
(Table 11). Results from the second series (Table
11, 18 July) were similar except that euphausiids
were not as abundant in the first sample of the
series, and Stenobrachius leucopsarus was most
abundant in the sample collected after the scatter-
ing layer had passed 125 m. The results, therefore,
indicate that both migratory myctophids and
euphausiids are associated with the migratory
sound-scattering layer, whereas sergestid
shrimps are not.
Table ll. — Occurrence of migratory myctophids and crusta-
ceans (number/10,000 m'') in two series of three samples col-
lected in horizontal hauls at 125 m depth before, during, and
after ascent of the migratory sound-scattering layer past that
depth.
17 July
18 July
Species
Before
During
After
Before
During
After
Stenobrachius
leucopsarus
0
38
1
0
13
27
Diaphus
theta
1
37
5
0
25
0
Euphausia
pacifica
87
64
1
5
68
2
Sergestes
similis
3
13
53
0
4
82
Position of the nonmigratory portion of the deep
sound-scattering layer which was present at night
was strongly correlated with the distribution of P.
thompsoni, particularly the small size class. The
scattering layer was broader and more diffuse at
night, and so was the distribution of P. thompsoni
(Figures 2, 8, 12B). Over the 200-300 m stratum,
the average concentration of P. thompsoni was
16.3 fish/10,000 m^ for the profile shown in Figure
12B. The day-to-night persistence of the nonmi-
gratory scattering layer (Figure 2) cannot be
explained by reference to the distribution of either
S. leucopsarus or D. theta. The two smaller size
classes of S. leucopsarus and all D. theta have
migrated into the surface layers at night, and the
largest S. leucopsarus are not only rare but
broadly distributed over 50-450 m. There are no
other abundant potential sound-scattering or-
ganisms concentrated in the 200-300 m stratum at
night.
DISCUSSION
Previous work on myctophids in open waters of
the subarctic Pacific dealt chiefly with systematics
and biogeography (Wisner 1976). However, Aron
(1962) and Taylor (1968) considered aspects of the
distribution of myctophids in eastern subarctic
waters. Aron's (1962) results are qualitative due
to the nature of the sampling gear used (unme-
tered, nonclosing nets of variable mesh size). Dif-
ferences between results of our study and those of
Taylor's (1968) comprehensive investigation are
probably attributable to the different sampling
gear employed rather than to fundamental varia-
tions in behavior of fish in different parts of the
subarctic Pacific. For example, Taylor's use of very
course-meshed nets probably accounts for both his
finding of different relative abundances of myc-
tophid species and for somewhat different patterns
765
FISHERY BULLETIN: VOL 76, NO. 4
of vertical distribution of species. Thus in Taylor's
study, carried out not far from Station Q, P.
thonipsojti and D. thcUi were more abundant than
S. leiHopsarus, but this was probably because
Taylor's net either did not efficiently catch
juvenile ( <35 mm) S. lei/copsarus which were the
numerically dominant size class of that species in
our samples (Figure 4), or they were much less
abundant during the time he sampled. Further,
Taylor obtained some of the largest catches of D.
theta and S. Icucopsariis below 90 m at night. This
probably also reflects the sampling bias of his net
for larger sizes offish, which at night tend to be
more broadly spread over the water column than
smaller fish (Figure 5C: Table 6, night series).
Unfortunately, Taylor did not report the sizes of
fish captured. Except for probable sampling bias
toward larger sizes offish, Taylor's results on ver-
tical distribution of the nonmigratory P.
thompsoni and other species of fish agree with
ours.
Pearcy et al. ( 1977) described patterns of verti-
cal distribution of mesopelagic fishes and crusta-
ceans off the coast of Oregon. The mesopelagic
assemblage there is essentially subarctic in
faunistic affinity and the vertical distributions of
species are similar to those observed at Stations P
and Q. The only notable departure from our re-
sults was the finding by Pearcy et al. that sig-
nificant numbers of all sizes of S. leucupsarus did
not participate, at least on a regular basis, in the
diel vertical migrations. Our observations at both
Stations P and Q indicate that virtually all S.
leucopsarus smaller than about 80 mm performed
extensive diel vertical migrations (Figure 5).
However, in our studies, S. leucopsarus was also
very rare below 400 m (Figure 5, Table 4), whereas
Pearcy et al. found large concentrations below
that depth. Thus there may be major differences in
the vertical distribution and migration behavior
of S. leucopsorus in different parts of its geograph-
ical range (Paxton 1967). Significantly, Pearcy et
al. ( 1977) detected no seasonal variations in verti-
cal distributions and migrations for any species,
which may also be true for subarctic waters to the
north (Taylor 1968).
Perhaps the most remarkable feature of the
mesopelagic fauna of the area sampled was its
simplicity. Only four species of myctophid fishes
were abundant in the upper 700 m. Two of these
species, S. leucopsarus and D. theta, undertook
diel migrations of substantial vertical extent; the
other two, P. thompsoni andS. nannochir, did not.
766
Other taxonomic groups also showed low diver-
sity. Among the micronektonic crustaceans there
were single species of abundant euphausiid, E.
pacifica, and decapod shrimp, Sergestes similis,
and both were vertical migrators. The contrast
between this relatively simple mesopelagic mi-
cronekton fauna and that, for example, in the sub-
tropical North Pacific (Brinton 1962a; Clarke
1973; Walters 1977) or subtropical North Atlantic
(Badcock 1970; Foxton 1970a, b) is striking, but
not atypical. Low taxonomic diversity of the
mesopelagic micronekton is found in other subpo-
lar oceans, such as the Boreal Atlantic (e.g.. Back-
us et al. 1971; Zahuranec and Pugh 1971).
Associated with the taxonomic simplicity of the
mesopelagic fauna herein reported, was a rela-
tively simple structure of the sound-scattering
layers. Generally, both the number and depth of
sound-scattering layers change with latitude in
the deep ocean; fewer and shallower layers are
found in subpolar oceans than in tropical-
subtropical oceans (Haigh 1971; Cole et al. 1971;
Donaldson and Pearcy 1972; Tont 1976). Our un-
published observations on deep sound-scattering
layers ( 12-kHz echosounder), taken in September
1972 along long. 155°W between Alaska and
Hawaii, showed this trend. Subarctic waters had
the relatively simple sound-scattering structure
illustrated in Figure 2, with single migratory and
nonmigratory layers occurring shallower than
400 m. In the subtropical waters near Hawaii, at
least three sound-scattering layers were observed
in the daytime at depths ranging from 260 to 625
m, and three to four migratory layers were re-
corded.
It is unlikely that the correlation between
taxonomic diversity of the mesopelagic micronek-
ton and complexity of the sound-scattering struc-
ture in the water column was fortuitous. Attempts
to causally relate deep sound-scattering layers to
aggregations of mesopelagic organisms were
stimulated by hypotheses advanced more than
three decades ago (for a review see Hersey and
Backus 1962). However, field studies based on net
samples taken simultaneously with echosounder
records tend to be inconclusive for a variety of
reasons. A major difficulty is that different
taxonomic groups tend to occur together at the
same depths and may even show similar migra-
tory behavior. For example, all four of the migra-
tory mesopelagic species in our study (Stenobrach-
ius leucopsarus , D. theta , E. pacifica , and Sergestes
similis) ascended towards the surface layer after
FROST and McCRONE: MKSOPEI.AGIC FISHES IN THE EASTERN SUBARCTIC PACIFIC
sunset, and only from fine temporal spacing of
samples did it become apparent that some species
were more closely associated with the scattering
layer than others (Table 11). Similarly, in the day-
time some of these migratory species cooccurred at
the depth of the sound-scattering layer together
with the nonmigratory P. thonipsoni, and any or
all could have contributed to the daytime sound-
scattering layer. Despite extensive cooccurrence
of several types of potential sound-scattering or-
ganisms, the most reasonable hypothesis is that
myctophids were primarily responsible for both
the migratory and nonmigratory sound-scattering
layers in the eastern subarctic Pacific.
Taylor (1968), also working in the subarctic
Pacific, found the best correlation between deep
sound-scattering layers and those mesopelagic
fish which possessed gas-filled swim bladders. Al-
though Taylor grouped Stenuhrach ius leucopsarus
and D. theta among fish with fat-invested swim
bladders, gas is present in the swim bladders of
immature individuals ( <30 mm SL) of both
species (Capen 19671. Taylor made no mention of
the size of the fish caught in his study; however, in
view of the very coarse-meshed nets he used, it is
probable that he did not quantitatively sample
immature fish. At Stations P and Q, some indi-
viduals of S. leucopsarus and D. theta were
theoretically the right size to resonate at 12 kHz
while at their daytime depths (Capen 1967), and
the abundance of either species was probably
sufficient to produce deep sound-scattering in the
daytime (Hershey and Backus 1962). This pre-
sumably holds also for P. thompsoni, which has a
gas-filled swim bladder throughout life (Taylor
1968: Butler and Pearcy 1972). Indeed, concentra-
tions of either D. theta or P. thompsoni alone in
Figure 12 were comparable with the concentration
of D. taaningi, which Baird et al. (1974) believe
was responsible for the migratory sound-
scattering layer over the Cariaco Trench.
As pointed out above, Sergestes si m His may be
excluded as a potential sound scatterer; it was
distributed too broadly and deeply in the daytime
and lagged the ascent of the migratory sound-
scattering layer at sunset (Figure 11, Table 11).
Although E. pacifica (Figure 10) was about five
times more abundant in the depth of the daytime
sound-scattering layer than all myctophid fishes
combined, it did not approach concentrations
necessary for it to be an effective scatterer of
12-kHz sound (Hersey and Backus 1962; Bary
1966; Beamish 1971).
In conclusion, we suggest that the nonmigratory
deep sound-scattering layer (Figure 2B) in the vi-
cinities of Stations P and Q in the eastern subarc-
tic North Pacific was caused by P. thompsoni, and
that the migratory sound-scattering layer (Figure
2B) recorded the migrations of smaller size classes
of Stenohrachius leucopsarus and D. theta. Pro-
tomyctophum thompsoni may have been largely
responsible for the deep scattering layer observed
in the daytime, with possible lesser contributions
from the two migratory myctophid species. Pearcy
(1977) found similar general correspondence be-
tween vertical distributions of the same three
species of myctophids and deep sound-scattering
layers off Oregon, but he pointed out that quan-
titative correlation between abundance of poten-
tial sound-scatterers and distribution of volume
scattering was not always strong. A more defini-
tive analysis, similar to that of Baird etal. (1974),
is required; that is, simultaneous observations
should be obtained on distribution of volume scat-
tering and abundance and acoustical properties of
suspected sound-scatterers.
In single hauls, we observed concentrations of
myctophids, all species combined, which regularly
exceeded 100 fish/10'* m'^ in the region of the deep
sound-scattering layer in the daytime and in the
surface layer at night. Similar concentrations of
myctophids are found in other oceans (e.g., Kash-
kin 1967). Further, the maximum concentrations
of myctophids observed by us in the surface layer
at night (365 fish/ 10-* m^ Table 8) and at depth in
the daytime (874 fish/ 10"* m'K horizontal haul at
327-333 m. Station Q) equal or exceed maximum
concentrations inferred from the apparently high
catch rates of single hauls reported by Halliday
(1970) and Backus et al. (1971) for the western
Boreal Atlantic, where one species of myctophid,
Benthosema glaciale , predominates. The very low
concentrations of myctophids found by Pearcy et
al. (1977), using a 2.4 m Isaacs-Kidd midwater
trawl, are puzzling and seem to indicate that myc-
tophids are about Vio as abundant off the Oregon
coast as in the open subarctic Pacific. However, the
data of Pearcy et al. ( 1977) differ from the earlier
results of Pearcy and Laurs (1966), in which re-
ported concentrations of myctophids were much
higher and similar to concentrations observed by
us; the difference could be due to year-to-year var-
iability (Pearcy 1977).
There is relatively little variability between
years in our estimates of abundance of myctophid
fishes (the three most abundant species. Table 2)
767
FISHERY BULLETIN: VOL. 76, NO. 4
in the water column extending to 385-460 m. We
estimate 0.61-1.24 myctophid fish/m^ based on av-
eraged day and night series (Table 10). No quan-
titative study comparable to ours has been made
in the open subarctic Pacific, but Pearcy and Laurs
(1966) provided data on the abundance of
mesopelagic fish near the Oregon coast. In two
cruises (August 1963), Pearcy and Laurs found
about 0.78 myctophid fish/m^ in the 0-500 m water
column at night; this estimate is based upon the
three numerically dominant myctophid fish cap-
tured (Pearcy and Laurs 1966, fig. 4), two of which
ranked 1 and 3 in abundance among myctophids
in our study. The average standing stock of all
mesopelagic fish found by Pearcy and Laurs ( 1966)
was 2.9 g wet weight/m^ in the 0-500 m water
column at night. Using a factor of 0.3 to convert
wet weight to dry weight, the average nighttime
standing stock is 0.87 g/m^, a value probably not
significantly different from our estimates based on
night samples (Table 9), especially since the
Pearcy and Laurs estimate is based on all
mesopelagic fish captured. Similar concentrations
of myctophids (about 0.6-0.8 fish/m^) are found in
the subtropical Pacific near Hawaii (Clarke 1973;
Maynard et al. 1975). However, many more
species of myctophids (47) occur there, and the
standing stock of myctophids (about 0.3-0.7 g wet
weigh t/m^) is somewhat less than our estimates
(0.23-0.53 g dry weight/m^, Table 9), probably be-
cause the fish are considerably smaller in average
size (Clarke 1973).
With regard to sampling bias, we found no evi-
dence of light-aided avoidance of the nekton trawl
by either myctophids or other types of micronek-
ton occurring in the upper 385-460 m during the
daytime (Table 3). Consistent day-night differ-
ences in catches of organisms, such as those ob-
served for the largest size class ( >80 mm SL) of S.
leucopsarus and for Sergestes similis, were proba-
bly due to migration of these organisms below the
depth range of daytime sampling. The results of
the single very deep vertical series at Station P
(Table 4) support this interpretation. Further-
more, very deep vertical migrations of both species
are well documented in other parts of their geo-
graphical ranges in the North Pacific (Omori et al.
1972; Pearcy et al. 1977).
In addition to determining vertical distributions
and vertical migrations of myctophid fishes, on
each cruise we also sampled zooplankton with a
smaller trawl (Frost and McCrone 1974). Analyses
of the zooplankton samples, together with data on
stomach contents of the three most abundant myc-
tophids, are the subject of a report (in preparation)
on the feeding behavior of myctophids in relation
to their vertical distribution and the vertical dis-
tribution of their zooplankton prey.
ACKNOWLEDGMENTS
We extend special thanks to David Thoreson for
his assistance in all phases of the research. Bruce
Davies participated in the cruises and was primar-
ily responsible for nearly flawless operation of the
trawl. We were fortunate to obtain help and advice
on systematics of myctophid fishes from Richard
McGinnis. Karl Banse made many useful sugges-
tions on the manuscript. Others whom we wish to
thank for participating in the cruises or assisting
with the analysis include Gene Anderson, Arthur
Griffiths, Louise Hirsch, Jeffrey Napp, Bruce Nes-
set, Mary Nirini, Layne Nordgren, Scott Ralston,
Wesley Rowland, Gary Shigenaka, and Steve
Spyker.
This research was supported by the Office of
Naval Research (Contract N00014-75-C-0502,
Project NR 083-012). Early development of the
trawl was supported by National Science Founda-
tion Grant GA-25385.
LITERATURE CITED
Ahlstrom, E. H.
1959. Vertical distribution of pelagic fish eggs and larvae
off California and Baja California. U.S. Fish Wildl.
Serv., Fish. Bull. 60:107-146.
ARON, W.
1962. The distribution of animals in the eastern North
Pacific and its relationship to physical and chemical con-
ditions. J. Fish. Res. Board Can. 19:271-314.
Backus, R. H., J. E. Craddock, R. L. Haedrich, and D. L.
Shores.
1971. The distribution of mesopelagic fishes in the equato-
rial and western North Atlantic Ocean. In G. B. Far-
quhar (editor), Proceedings of an international sym-
posium on biological sound scattering in the ocean, p.
20-40. Maury Center for Ocean Science, Wash., D.C.
BADCOCK, J.
1970. The vertical distribution of mesopelagic fishes col-
lected on the SOND Cruise. J. Mar. Biol. Assoc. U.K.
50:1001-1044.
BADCOCK, J., AND N. R. MERRETT.
1976. Midwater fishes in the eastern North Atlantic - I.
Vertical distribution and associated biology in 30°N,
23°W, with developmental notes on certain myctophids.
Prog. Oceanogr. 7:3-58.
Baird, R. C, D. F. Wilson, R. C. Beckett, and T. L. Hopkins.
1974. Diaphus taaningi Norman, the principal component
of a shallow sound-scattering layer in the Cariaco Trench,
Venezuela. J. Mar. Res. 32:301-312.
768
FROST and McCRONE: MESOPELAGIC FISHES IN THE EASTERN SUBARCTIC PACIFIC
BARHAM, E. G.
1966. Deep scattering layer migration and composition:
observations from a diving saucer. Science (Wash., D.C.)
151:1399-1403.
BARY, B. MCK.
1966. Back scattering at 12 kc/s in relation to biomass and
numbers of zooplanktonic organisms in Saanich Inlet,
British Columbia. Deep-Sea Res. 13:655-666
1967. Diel vertical migrations of underwater scattering,
mostly in Saanich Inlet, British Columbia. Deep-Sea
Res. 14:35-50.
Beamish, P.
1971. Quantitative measurements of acoustic scattering
from zooplanktonic organisms. Deej>Sea Res. 18:811-
822.
BEAUCHAMP, J. J., AND J. S. OLSON.
1973. Corrections for bias in regression estimates after
logarithmic transformation. Ecology 54:1403-1407.
BRINTON. E.
1962a. The distribution of Pacific euphausiids. Bull.
Scripps Inst. Oceanogr., Univ. Calif 8:51-269.
1962b. Variable factors affecting the apparent range and
estimated concentration of euphausiids in the North Pac-
ific. Pac. Sci. 16:374-408.
BL'TLER, J. L., AND W. G. PEARCY.
1972. Swimbladder morphology and specific gravity of
myctophids off Oregon. J. Fish. Res. Board Can.
29:1145-1150.
Capen, R. L.
1967. Swimbladder morphology of some mesopelagic
fishes in relation to sound scattering. U.S. Navy Elec-
tron. Lab., Res. Rep. 1447, 32 p.
CLARKE, M. R.
1969. A new midwater trawl for sampling discrete depth
horizons. J. Mar. Biol. Assoc. U.K. 49:945-960.
CLARKE, T. A.
1973. Some aspects of the ecology of lantemfishes (Myc-
tophidaei in the Pacific Ocean near Hawaii. Fish. Bull.,
U.S. 71:401-434.
Cole, H. p., G. M. Bryan, and A. L. Gordon.
1971. The deep scattering layer: patterns across the Gulf
Stream and the Sargasso Sea. In G. B. Farquhar (editor).
Proceedings of an international symposium on biological
sound scattering in the ocean, p. 281-293. Maury Center
for Ocean Science, Wash., D.C.
Dodimead, a. J., F. Favorite, and T. Hirano.
1963. Salmon of the North Pacific Ocean. Part II. Review of
oceanography of the subarctic Pacific region. Int. North
Pac. Fish. Comm. Bull. 13, 195 p.
Donaldson, h. a., and w. G. pearcy.
1972. Sound-scattering layers in the northeastern Pa-
cific. J. Fish. Res. Board Can. 29:1419-1423.
Farquhar, G. B. (editor).
1971. Proceedings of an international symposium on
biological sound scattering in the ocean. Maury Center
for Ocean Science, Wash., D.C, 629 p.
FOXTON, P.
1970a. The vertical distribution of pelagic decapods [Crus-
tacea: Natantia] collected on the SOND Cruise 1965.
I. The Caridea. J. Mar. Biol. Assoc. U.K. 50:939-960.
1970b. The vertical distribution of pelagic decapods [Crus-
tacea: Natantia] collected on the SOND Cruise 1965.
II. The Penaeidea and general discussion. J. Mar. Biol.
Assoc. U.K. 50:961-1000.
Frost, B.W., and l. e. mcCrone.
1974. Vertical distribution of zooplankton and myctophid
fish at Canadian weather station P, with description of a
new multiple net trawl. Proceedings of the Interna-
tional Conference on Engineering in the Ocean Environ-
ment, Vol. I, p. 156-165. Halifax, N.S.
haigh, K. K. R.
1971. Geographic, seasonal, and annual patterns of mid-
water scatterers between latitudes 10° and 68° North in
the Atlantic. In G. B. Farquhar (editor). Proceedings of
an international symposium on biological sound scatter-
ing in the ocean, p. 268-280. Maury Center for Ocean
Science, Wash., D.C.
Halliday, R. G.
1970. Growth and vertical distribution of the glacier lan-
temfish, Benthosema glaciale, in the northwestern Atlan-
tic. J. Fish. Res. Board Can. 27:105-116.
Harding, D., E. Shreeve, D. S. Tungate, and D. Mummery.
1971. A net-changing mechanism for the Lowestoft mul-
tipurpose sampler. J. Cons. 33:483-491.
Harrisson, C. M. H.
1967. On methods for sampling mesopelagic fishes. Symp.
Zool. Soc. Lond. 19:71-126.
Hart, J. L.
1973. Pacific fishes of Canada. Fish. Res. Board Can.,
Bull. 180, 740 p.
Hersey, J. B., AND R. H. Backus.
1962. Sound scattering by marine organisms. In M. Hill
(editor). The sea: Ideas and observations on progress in the
study of the seas. Vol. 1, p. 498-539. Interscience, N.Y.
Holton, a. a.
1969. Feeding behavior of a vertically migrating lan-
temfish. Pac. Sci. 23:325-331.
KASHKIN, N. I.
1967. On the quantitative distribution of lantern fishes
(Myctophidae s.l.) in the Atlantic Ocean. In T. S. Rass
(editor). The pelagic and bathypelagic fishes of the world
oceans, p. 206-264. Tr. Inst. Okeanol. Akad. Nauk SSSR
84. (Translated from Russ. by the Translation and In-
terpretation Division, Institute of Modern Languages,
Inc., 1969.)
KULIKOVA, E. B.
1957. Growth and age of deep- water fishes. /nB. N.Niki-
tin (editor), Marine biology, p. 284-290. Tr. Inst. Okeanol.
Akad. Nauk SSSR 20. (Translated from Russ. by Am. Inst.
Biol. Sci., 1959.)
LaVENBERG, R. J., AND A. W. EBELING.
1967. Distribution of midwater fishes among deep-water
basins of the southern California shelf. In R. N. Phil-
brick (editor), Proceedings of the symposium on the biol-
ogy of the California Islands, p. 185-201. Santa Barbara
Botanic Garden, Inc., Santa Barbara, Calif.
Marshall, N. B.
1954. Aspects of deep sea biology. Hutchinson's Scien-
tific and Technical Publications. Lond., 380 p.
MaYNARD, S. D., F. V. RIGGS, AND J. F. WALTERS.
1975. Mesopelagic micronekton in Hawaiian waters:
Faunal composition, standing stock, and diel vertical
migration. Fish. Bull., U.S. 73:726-736.
McCartney, b. s.
1976. Comparison of the acoustic and biological sampling
of the sonic scattering layers: R.R.S. "DISCOVERY"
769
FISHERY BULLETIN: VOL. 76, NO. 4
SOND Cruise, 1965. J. Mar. Biol. Assoc. U.K. 56:161-
178.
MCGOWAN, J. A.
1971. Oceanic biogeography of the Pacific, /n B. M. Fun-
nell and W. R. Riedel (editors), The micropaleontology of
oceans, p. 3-74. Camb. Univ. Press, Camb.
NAFPAKTITIS, B. G.
1968. Taxonomy and distribution of the lantemfishes,
genera Lobianchia an&Diaphus, in the North Atlantic.
Dana Rep. Carlsberg Found. 73, 131 p.
Omori, M., A. Kawamura, and Y. AIZAWA.
1972. Sergestes similis Hansen, its distribution and impor-
tance as food of fin and sei whales in the North Pacific
Ocean. In A. Y. Takenouti (editor). Biological oceanog-
raphy of the northern North Pacific Ocean, p. 373-391.
Idemitsu Shoten, Tokyo.
Paxton, J. R.
1967. A distributional analysis for the lantemfishes (Fam-
ily Myctophidae) of the San Pedro Basin, California
Copeia 1967:422-440.
PEARCY, W. G.
1977. Variations in abundance of sound scattering ani-
mals off Oregon. In N. R. Anderson and B. J. Zahuranec
(editors), Oceanic sound scattering prediction, p. 647-666.
Plenum Press, N.Y.
PEARCY, W. G., E. E. KRYGIER, R. MESECAR, AND F. RAMSEY.
1977. Vertical distribution and migration of oceanic mi-
cronekton off Oregon. Deep-Sea Res. 24:223-245.
PEARCY, W. G., AND R. M. LAURS.
1966. Vertical migration and distribution of mesopelagic
fishes off Oregon. Deep-Sea Res. 13:153-165.
Smoker, W., and W. G. Pearcy.
1970. Growth and reproduction of the lantemfish Steno-
brachius leucopsarus. J. Fish. Res. Board Can.
27:1265-1275.
Tate, M. W., and R. C. Clelland.
1957. Nonparametric and shortcut statistics in the social,
biological, and medical sciences. Interstate Printers and
Publishers, Inc., Danville, 111., 171 p.
Taylor, F. H. C.
1968. The relationship of midwater trawl catches to sound
scattering layers off the coast of northern British Colum-
bia. J. Fish. Res. Board Can. 25:457-472.
Tont, S. a.
1976. Deep scattering layers: patterns in the Pacif-
ic. Calif. Coop. Oceanic Fish. Invest., Rep. 18:112-117.
TUCKER, G. H.
1 95 1 . Relation of fishes and other organisms to the scatter-
ing of underwater sound. J. Mar. Res. 10:215-238.
Vinogradov, m. E.
1968. Vertical distribution of the oceanic zooplank-
ton. Inst. Okeanol. Akad. Nauk SSSR, Moscow, 339 p.
(Translated from Russ. by A. Mercado and J. Salkind, Isr.
Program Sci. Transl., Jerusalem, 1970.)
Walters, J. F.
1977. Ecology of Hawaiian sergestid shrimps (Penaeidea:
Sergestidae). Fish. Bull., U.S. 74:799-836.
WiEBE, P. H., and W. R. Holland.
1968. Plankton patchiness: effects on repeated net tows.
Limnol. Oceanogr. 13:315-321.
Wisner, R. L.
1976. The taxonomy and distribution of lantemfishes
(Family Myctophidae) of the eastern Pacific Ocean. Navy
Ocean Res. Dev. Act., Rep.-3, U.S. Gov. Print. Off., Wash.,
D.C., 229 p.
Zahuranec, B. J., and W. L. Pugh.
1971. Biological results from scattering layer investiga-
tions in the Norwegian Sea. In G. B. Farquhar (editor).
Proceedings of an international symposium on biological
sound scattering in the ocean, p. 360-380. Maury Center
for Ocean Science, Wash., D.C.
770
ANALYSIS OF A SIMPLE MODEL FOR
ESTIMATING HISTORICAL POPULATION SIZES
T. D. Snqth' and T. Polacheck=^
ABSTRACT
Estimates of historical abundance of animal populations are important in many management deci-
sions. Historical estimates based on a simple model of population growth have been made for several
populations of dolphin involved with the yellowfin tuna purse seine fishery. We used the data for the
bridled dolphin, Stenella attenuata. to investigate the behavior of the model by which these historical
estimates were calculated. For populations with low net reproductive rates, the effect of bias in the
estimates of the input parameters on the estimated historical abundances was approximately linear
and additive. When all the input parameters were independently estimated, the variances of the
historical abundance estimates were dominated by the variance of the initial abundance estimate and
the coefficient of variation of the historical estimate was less than the largest coefficient of variation of
any parameter.
Many decisions about the management of animal
populations are based on the estimates of abun-
dance of the population relative to its historical or
preexploitation size. These estimates are basic to
any application of the theory of maximum sus-
tained yield as incorporated in several interna-
tional marine mammal management agreements
such as the North Pacific Fur Seal Treaty and the
International Whaling Convention. Similarly, the
concept of "optimum sustainable populations" as
specified in the recent Marine Mammal Protection
Act of 1972 (MMPA) has been defined in terms of
comparing the present size of a population with its
original size (Southwest Fisheries Center^).
Schools of dolphin of several species (primarily
Stenella attenuata and S. longirostris) have been
used by purse seine fishermen in the eastern tropi-
cal Pacific to locate yellowfin tuna, Thunnus alba-
cares, since 1959, as described by Perrin (1969).
Significant numbers of dolphin have been killed
by becoming entangled in the purse seines. In
order to make management decisions under the
MMPA about these dolphin populations, the Na-
tional Marine Fisheries Service (NMFS) needed
'Department of Zoology, University of Hawaii, Honolulu,
Hawaii; present address: Southwest Fisheries Center, National
Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA
92038.
^Department of Zoology, University of Hawaii, Honolulu,
Hawaii; present address: Department of Biology, University of
Oregon, Eugene, OR 97477.
^Southwest Fisheries Center. 1976. Report of the Workshop on
Stock Assessment of Porpoises Involved in the Eastern Pacific
Yellowfin Tuna Fishery. Southwest Fish. Cent. La Jolla Lab.,
Natl. Mar. Fish. Serv., NOAA, Admin. Rep. LJ-76-29, 53 p.
estimates of the preexploitation abundance of the
various populations. The NMFS convened a work-
shop of scientists to obtain the estimates based on
a simple model of population change (see footnote
3). This paper evaluates the behavior of estimates
of abundance obtained from their approach. This
is important in order to be able to evaluate the
degree of confidence to be placed in such estimates,
and hence in management plans based on them.
METHODS AND MATERIALS
The model used to estimate preexploitation
abundance is based on a common discrete model of
population growth:
AT^^j = N-K^HN-K^) (b-d) (1)
where N^. = the abundance at time r
b = the birth rate
d = the natural death rate
K^ = the number of animals killed, as-
sumed to occur at the beginning
of time interval t
N^^ 1 = the abundance 1 time unit later.
Reversing the procedure (i.e., solving the above
equation for N,. ) results in the expression
^ l+i?_ ^
(2)
Manuscript accepted May 1978.
FISHERY BULLETIN; VOL. 76. NO. 4, 1979.
where Nr now is the estimate of abundance 1 yr
771
FISHERY BULLETIN: VOL. 76. NO. 4
earlier andi?. is the net reproductive rate ib-d).
The above model was modified in the procedure
used by NMFS to account for situations when the
kills occur throughout the time interval instead of
instantaneously at the end of the interval, as:
r+i "• T
(3)
This equation can be repeatedly applied to give
estimates any number of years it) into the past.
When rearranged to explicitly display the popula-
tion size t years earlier, and relabeling so that the
initial abundance is N„, one obtains
1974 and the annual incidental kills and repro-
ductive rates from 1959 to 1974. Several se-
quences of estimated annual kills and reproduc-
tive rates were considered, incorporating the
uncertainty in the data.
In the present paper the sequences of annual
kills and net reproductive rates given in Table 1
are used to illustrate several general aspects of the
behavior of Equation (4). These correspond to the
"high kill" and "central reproductive rate" se-
quences for the bridled dolphin, Stenella at-
tenuata, in the Workshop report. The estimate of
1974 abundance used by us and the Workshop was
3.5 million.
A^„
A^. =
t
n
(1+^p
t
X.(l+i?./2)
(4)
Note in this form that the time-index t runs back-
wards from zero. As is apparent in this form, the
estimation of abundance t years earlier involves
2t+\ parameters. The sequences of annual kills
and net reproductive rates can be termed the kill
and the net reproductive rate vectors, each com-
posed of t elements.
The data used here to explore this estimation
procedure is from the report of NMFS Workshop
discussed above (see footnote 3).^ From existing
unpublished data and reports the Workshop par-
ticipants used estimates of the population size in
''It should be noted that the estimates used here are based on a
number of assumptions currently under investigation and that
these estimates are subject to significant change in the near
future (I. Barret, Director, Southwest Fisheries Center, La Jolla,
CA 92038, pers. commun. April 1978).
Estimation of Bias
A sensitivity analysis was done to examine the
effects of biased parameter estimates on the
backcalculated abundance. A new population size
1 yr earlier, from Equation ( 3 ), when each parame-
ter is changed by a specified amount is
N^{n, k, r) =
N^{l+n)+0.bK^{l+k)
and in general for t years earlier.
(5)
iV;(n, k, r) =
N^il+n)
U^ (l+/2.(l+r))
, KiUk) (l-H(/2 (l+r)/2))
-H_2— ^-^ (6)
S.(l+/?.(l+r))
Table l. — Estimates used for kill and reproductive rate vectors
of Stenella attenuata in the eastern Pacific.
Kill
f
Year
(thousands)
Net reproductive rate
1
1973
120
0,040
2
1972
273
.040
3
1971
185
.040
4
1970
308
.036
5
1969
331
.032
6
1968
164
.028
7
1967
194
.024
8
1966
281
.020
9
1965
297
.016
10
1964
255
.012
11
1963
133
.008
12
1962
106
.004
13
1961
446
.000
14
1960
534
.000
15
1959
129
.000
where Nq, R^, and K^ are defined as above, and
n = the proportion that A^q deviates from its
estimate
r = the proportion that all elements of the net
reproductive vector deviate from their es-
timates
k = the proportion that all elements of the kill
vector deviate from their estimates.
N'j in,k,r) was then compared with A^, from Equa-
tion (4) or equivalently N ', (0,0,0). As a measure of
the sensitivity of the basic model, S, {n,k,r) is
defined to equal the percent that A^', (n,k,r) de-
viates from A^^
772
SMITH AND POLACHECK: ANALYSIS OF SIMPLE MODEL
S^ (n, k, r) =
N^ (n, K r)-N^
N,
• 100 . (7)
Estimation of Variance
The variance of the backcalculated estimate of
A^, from Equation (4) was approximated using the
delta method (Sober 1973). This method is based
upon a Taylor series expansion for a function in
which quadratic and other higher order terms are
ignored. If/" is a function of the random variables
x^, X2,X2 . . . ,Xn then the expression for the vari-
ance of/" by the delta method is
v(r(x^,x,,X3...,xj)= iv(xp(^]'
+ 2S2 Gov (X.,X.) I-J^ ■ -^1 . (8)
i^j
^i' v\ax. bx.i
1 J/
In applying this expression to Equation (4), it is
necessary to be able to define which of the
parameters should be considered as random vari-
ables, and to give reasonable estimates for value of
the variances and covariances of these variables.
For the purpose of exploring the behavior of Equa-
tion (4), we assumed that the estimates of all the
parameters in Equation (4) are independent ran-
dom variables. The covariance terms in Equation
(8) are then zero. This approach provides a picture
of the variance of the back estimate of abundance
if in fact independent estimates of the kills and the
net reproductive rates were available for each
year. A generalized expression for the variance
using this approach is
V(iV^)
V(iV,)
dN. ^2
3A^.
t
+ S
V{K.) I
dN.
dN.
+ 2 V(i?.)
y=i ^ r
'dN.
(9)
where all parameters are defined as for the basic
model [Equation (4)]. For detailed expressions for
each of the right hand terms see Appendix I.
As noted the method used for approximating the
variance of a function depends on the higher order
terms in the Taylor's series expansion being small.
The higher order terms in the delta method ex-
pression for the variance of A/^, are composed of the
second and higher order derivatives of N, with
respect to A^n- K^, and Rf, and the higher order
central moments of the probability distributions of
the estimates of N„, K,, and R, (i.e., skewness,
kurtosis, etc.). The second and higher derivatives
with respect to A^^ andKf are zero. Thus the terms
involving/?, are the only higher order terms not
equal to zero. The higher order derivatives of Nf
with respect to i?, involve i?,+ j to increasing nega-
tive powers. The three higher order moments ofR,
are always decreasing since /?, is much less than
one. Thus each of the higher order terms in the
delta method expression for the variance of N, are
each less than the first order term in R; (iii of
Appendix I). The contribution of this first order
term in Rt to the variance ofN, is small, as shown
below. Thus the error induced by ignoring the
higher order terms in the Taylor's series appears
small.
The objective in doing the variance calculations
was to understand the behavior of the variance of
the population size when estimated by the basic
back projection model [Equation (4)]. Thus a range
of variances was calculated for a range of reason-
able values of the variances of the estimated
parameters. However, in our example of bridled
dolphin estimates of the variance of many of the
parameters were not available. Many of the kill
estimates were not independently estimated and
hence have large unknown covariances (Smith
and Polacheck^). Estimates of net reproductive
rate were obtained by extrapolation from other
populations and fi'om assumptions about density
dependence. It is not clear that the uncertainty in
these estimates can adequately be described by
the notion of variance. Thus, the variances that we
used and that we calculated for N, should not be
interpreted as actual estimates of variance for this
population.
RESULTS
Bias
The results of the sensitivity analysis of the
basic model will be presented by examining the
effects of varying each of the variables «, k, and r of
Equation (7), separately, and then in combina-
tions.
The sensitivity of the back projected estimates
^Smith, T. D., and T. Polacheck. 1977. Uncertainty in estimat-
ing historical abundance of porpoise populations. Contract Rep.
MM 7A C006, 39 p. Marine Mammal Commission, 1625 Eye
Street, Washington, DC 20006.
773
FISHERY BULLETIN: VOL. 76, NO. 4
(S,) for a fixed number of years / into the past is
linear with respect to A? or/? (Figure D.ThisHnear-
ity can be seen in Equation (6) since n and k enter
only as linear terms in the numerator. Positive
30i
Figure l.— Sensitivity of the model Sffn,;fe,H in 1959 for a range
of deviations in the initial number (n ), for a range of deviations in
the kills (k), or for a range of deviations in the net reproductive
rate (r), for Stenella attenuata in the eastern tropical Pacific.
values of either n ov k yield positive deviations in
the back estimates. However, the farther back the
population is projected in time, the smaller the
contribution of N,, to the back estimate becomes
relative to the contribution of the kills. Thus the
effect of bias in the estimate of the initial numbers
(n) becomes progressively smaller the farther
back in time the population is projected, while the
consequence of a consistent bias in the kill esti-
mates (k) becomes larger (Figure 2). Since the
annual kills have no simple relationship to time,
the effect of a particular value of « or ^ over time
(Figure 2) cannot be described by any simple func-
tion. This trade off in the sensitivity of the back
projected estimates between n and k is exact in the
sense that for any decrease over time in the slope
of S with respect to n there is an equivalent in-
crease in the slope of S with respect to k. This can
be seen by evaluating the partial derivates of S
with respect to n and with respect to k and noting
that they sum to 1.
The effects of bias in the estimates of the net
reproductive rate vector are more complicated
than for the other two factors. Positive deviations
in the net reproductive rates (r) yield negative
deviations in the back projected estimate (Figure
2). The effect of r tends to increase over time (Fig-
ure 2). S approaches being linear with respect to r
for any particular year, but unlike the relation-
ship for k and n, this result is not exact ( Figure 1 ).
The approximate linearity of the sensitivity ofN^
£. 10-
.^ 5
-5-
,o-
.- 0-'
.-O'
,o'
.-cr
,0'
* A-
^
— *-
*■
—A—
0 1
I
2
1
3
1
4
1
5
r
6
1 T
7 8
Time (t)
I
9
10
1
11
— I —
12
r ■
13
14
15
1973
1971
1969
1967
Year
1965
1963
1961
1959
FIGURE 2.— Sensitivity of the model St
(n,k,r} over time to a 30% deviation in
the initial number (n = 0.3), in the kill
vector (k = 0.3), and in the net reproduc-
tive rate vector (r = 0.3) when all factors
are held constant (or Stenella attenuata
in the eastern tropical Pacific.
774
SMITH AND POLACHECK; ANALYSIS OF SIMPLE MODEL
to r appears to be a general feature of this proce-
dure when r is small. This can be seen by examin-
ing S, expressed as a function of r, which can be
obtained explicitly by substituting the definitions
of N, [Equation (4)] and iV',[Equation (6)] into
Equation (7) and simplifying.
The consequences of having two factors varying
simultaneously are shown in the series of contours
of equal values of S from Equation (7) (Figures
3-5). These contour plots present a visual picture
of the sensitivity of the back projection to the dif-
ferent factors. From this set of contour maps, it
can be seen that the surface generated by S [Equa-
tion (7)] tends to be nearly linear. Since S has no
nonlinear terms with respect to n and k, the sur-
face described by S in these two dimensions is
simply a plane (Figure 4). There are nonlinear
effects between the net reproductive rate and both
initial abundance and the sequence of kills. For
the example examined here, the nonlinearity be-
tween k and r is insignificant. For instance, if r and
k both equal 0.50, S deviates from a linear model
by 60%), the CV
of the back estimate does not exceed the CV of A^q.
1 +
and the basic model [ Equation (3)] for the dolphin
population examined here is given in Table 6. The
simpler model always gives a slightly higher es-
timate for the size of the back projected population
but the increase in the estimate is always <1%.
The sensitivities of the two models are nearly
equivalent. When the values for the parameters in
these models deviate as much as SO'/t the differ-
ence between sensitivities of the two models is
<1%. The approximate variances of the back es-
timates of the two models are also similar.
That the difference between the original and the
simpler model is small can be shown by analyti-
cally comparing the two models. If the projections
are made only 1 yr into the past, the ratio of the
estimate from Equation (2) to the estimate from
Equation (3) is
1 +
O.bR^K^
N^+K^+O.bR^K^-
Only if the value ofR^K^ is large relative to Ao +^i
can this ratio deviate significantly from 1. This is
only possible if /?i is relatively large. The general
formula for the ratio of the two models is
S 0.5KJi.( fl (l+R, J)
N^+ 1 0.5K. (n (l+R^ J)+ 2 O.bK.in (l+RJ)
Comparison of Equations (2) and (3).
A comparison of the estimated back abundance
as calculated by the simpler model [Equation (2)]
As in the case for projecting back only 1 yr, it can
be seen that unless the RjKj terms are large rela-
tive to Nq and unless the net reproductive rate is
also large, the ratio of the two models will be close
to 1.
Table 6. — Comparison of the back estimate of the abundance of
bridled dolphin as calculated by the basic model [Equation (3)]
and the simpler model [Equation (2)].
Simple model
Basic model
Year
('10«)
(xlO^)
Simple/basic
1974
3.500
3.500
1.000
1973
3.485
3.483
1.001
1972
3.624
3.617
1.002
1971
3.670
3.659
1.003
1970
3.850
3.835
1.004
1969
4.062
4.0416
1.005
1968
4.115
4.093
1.005
1967
4.214
4 190
1.006
1966
4.412
4.386
1.006
1965
4.640
4.612
1.006
1964
4.840
4.811
1.006
1963
4.934
4.905 .
1.006
1962
5.021
4.991
1.006
1961
5.467
5.437
1.005
1960
6.001
5.971
1.005
1959
6.130
6,100
1.005
DISCUSSION AND CONCLUSIONS
The results of this analysis indicate that errors
in the input parameters do not compound in this
procedure for estimating historical abundance. In
fact, a systematic bias in the procedure for the
estimation of a single set of parameters (either A^o
or R^'s or K,'s) always induces a bias in the back
projected estimate which is less than the bias of
the estimated parameters. This conclusion follows
directly from the linear or near linear relation
between St and n, k, or /• with small rates of
change. Moreover, the effects of bias in two or
more sets of parameters are nearly additive. The
interaction effects of bias in estimates of kills, net
reproductive rates, and the initial number tend to
777
FISHERY BULLETIN: VOL 76. NO. 4
be small or nonexistent. This will be globally true
for the relationship between k and n, but will be
true for the relationship between /?, r, and /; only
when the net reproductive rate is small. The rela-
tive importance of bias in A'/s, /?/s, or 7V„ on A^,
depends upon the actual values of the parameter.
In the bridled dolphin example, after 15 yr, the
back estimates were most sensitive to bias in the
kill estimate, slightly less sensitive to bias in N„,
and considerably less sensitive to bias in the net
reproductive rate. However, the importance of
bias in A/^o will diminish with the number of years
in the back estimate with a proportionate increase
in the importance of bias in the kills.
The sensitivity analysis developed in this paper
will include the extremes of a complete sensitivity
analysis of the model. The values forS/ (0,/?,0) are
limiting values to a complete sensitivity analysis
of the individual elements of the kill vector on N,.
Similarly S, ( 0,0,/') is a limit to complete sensitiv-
ity analysis of the individual elements of the net
reproductive rate. Given the additivity of S, with
respectto«,r, and^, the surface S^ («,/?,r) contains
the extremes of a sensitivity analysis in all 2t+\
dimensions. If in fact the elements within the kill
vector and within the reproductive vector are
highly interdependent (as is the case for the data
used here), then the sensitvity analysis used to
look at the effects of bias in this paper approaches
a total sensitivity analysis of the back projected
estimate given these constraints.
The variance approximations also indicate that
variability in the parameter estimates does not
result in compounding uncertainty in the back
projected estimates. When estimates of the
parameters are independent and the net reproduc-
tive rate is low, the CV of the back estimate will be
smaller than the CV of the input parameters. In
our example if all the CV's were equal, the vari-
ance of A^o would make the largest contribution to
the estimated variance oiN,. In general this will
be true as long as the kills in any one year do not
approach the initial abundance. This is a direct
consequence of the basic additivity of the model
when the net reproductive rate is small.
In Smith and Polacheck (see footnote 4), an al-
ternative probability structure was considered in
which the elements within the kill vector and
within the net reproductive rate vector were
highly interdependent. In this situation, the vari-
ance of Nt is not completely dominated by the
variance of N^^. The variances of N, calculated
using this interdependent probability structure
are larger than the variances presented here in
which all the parameters are assumed indepen-
dent. However, the CV of TV, for the dolphin data
within this interdependent probability structure
is still less than the CV of the parameters if all
parameters have equal C V. It appears that even in
the situation in which a high degree of inter-
dependence exists within the kill estimate or the
net reproductive estimates, the variability in the
parameter estimates does not induce compound-
ing uncertainty in the back projected estimate.
The comparison of the results from the basic
model [Equation (3)] with the simpler model
[Equation (2)] indicate that there are no sig-
nificant differences between the two models as
long as the net reproductive rate is small. Thus it
appears that there is no reason to favor the more
complex model over the simpler.
In conclusion, it appears that this back projec-
tion procedure (either model) has reasonable
statistical properties, at least when the net repro-
ductive rates are small. However, Equation ( 1 ) is a
simplified description of how the abundance of a
population changes through time, especially in
not accounting for changes in age structure. The
authors feel that caution should be used in apply-
ing estimates from this procedure to the manage-
ment of long-lived species since changes in the age
structure for long-lived species are likely to be
important.
ACKNOWLEDGMENTS
Financial support for this study was supplied by
the U.S. Marine Mammal Commission (Contract
MM74C006). We wish to acknowledge the journal
editor and an anonymous reviewer for their help-
ful comments.
LITERATURE CITED
Perrin, W. F.
1969. Using porpoise to catch tuna. World Fish.
18(6):42-45.
Seber, G. a. F.
1973. The estimation of animal abundance and related
parameters. Hafner Press, N.Y., 506 p.
778
SMITH AND POLACHECK: ANALYSIS OF SIMPLE MODEL
Appendix I. — Expressions for the variance components of N,.
Expression for the right hand terms of Equation (9) are:
^ »/ \.n (i+R,),
2 V(X)(^ = 2 V(if) -^ ^
(ii)
;?, ^<«>' br = ,?, v<«/' — -i-> — ) • "">
\ '/ \(i+R.)^^n.^^(i.R,)/
Appendix II. — Coefficient of variation of a sum of random variables.
The following is a proof that the coefficient of variation of a sum of two independent random variables is
smaller than the greatest CV for either of the random variables if the expected value of the random
variables is greater than zero.
If A and 5 are independent random variables such that
E(A) = a>Q E(B) = &>Oand
then
CV(A) = ^^^>«=CV(B)
V(A) ^ V(B)
V(A)(62 + 2ab) > YiB)a^ ,
V{A)ib^ + 2ab) + YiA)a^ > YiB)a^ + Y{A)a^ ,
V(A)(a + bf > [V(B) + V(A)]c2,
V(A) Y{B) + V(A) ^ V(A + B)
a^ (a + 6)2 [E(A + B)f '
CV(A) > CV(A + B) .
779
LARVAL DEVELOPMENT OF GALATHEA ROSTRATA UNDER
LABORATORY CONDITIONS, WITH A DISCUSSION OF
LARVAL DEVELOPMENT IN THE GALATHEIDAE (CRUSTACEA ANOMURA)^
Robert H. Gore^
ABSTRACT
The complete larval development of the western Atlantic anomuran crab, Galathea rostrata, consists of
four or five zoeal stages, and a single megalopal stage, based on larvae cultured under laboratory
conditions. Variation in the duration and number of zoeal stages appears to be temperature-dependent,
with larvae reared at 15°C developing through five zoeal stages and attaining megalopa in 52 days,
whereas larvae cultured at 20°C passed through four or five zoeal stages, reaching megalopa in 18 or 23
days, respectively. At 20°C some third stage zoeae molted to a "regular" fourth zoeal stage, without
pleopods, which was followed by a subsequent fifth stage before reaching megalopa. Other zoeae molted
to an "advanced" fourth stage, possessing pleopods, which subsequently molted directly to megalopa,
bypassing stage V completely. The variation noted in larval development in other galatheid genera is
briefly discussed, and a provisional s3Tiopsis of morphological characters of systematic value is pro-
vided for their identification.
The anomuran crab genus Galathea is presently
represented in the western North Atlantic by two
species, G. agassizii andG. rostrata (A. Milne Ed-
wards 1880). Galathea agassizii, primarily tropi-
cal and insular in distribution, is a deepwater
species known from 166 to 490 fm (304-897 m) off
St. Augustine, Fla., and from Cuba, St. Vincent, St.
Lucia, and Barbados in the Caribbean Sea. In the
eastern Atlantic the species is found from 82 to 898
fm (150-1,643 m) in the vicinity of both the Cape
Verde and Canary Islands, and off northwestern
Africa (Chace 1942; Miyake and Baba 1970).
Contrarily, G. rostrata appears to be a warm-
temperate or tropical/subtropical species, primar-
ily continental in distribution. The species is re-
corded from the North American continental shelf
at Cape Hatteras, N.C., to southeastern Florida,
and in the Gulf of Mexico from western Florida,
the Mississippi Delta, and southward to Islas Jol-
bos, north of the Yucatan Peninsula. There is a
questionable record from off Rhode Island (Wil-
liams 1965). Galathea rostrata is also found in
shallower water than G. agassizii and has been
collected from 10 to 50 fm (18-92 m), with the
exception of the possible depth record of 1,178 fm
'Scientific Contribution No. 100, from the Smithsonian
Institution-Harbor Branch Foundation, Inc., Scientific Consor-
tium, Link Port, Ft. Pierce, Fla. This report is Article IX.
Studies on Decapod Crustacea from the Indian River Region of
Florida.
^Smithsonian Institution, Ft. Pierce Bureau, Ft. Pierce, FL
33450.
Manuscript accepted June 1978.
FISHERY BULLETIN: VOL. 76, NO. 4, 1979.
(2,156 m) from off Rhode Island. The only distribu-
tional record of the species for the entire eastern
Florida coast was that of Haig (1956) who reported
a single specimen collected from 21 fm (38 m) off
Hillsboro Lighthouse (Broward County) in south-
eastern Florida. However, recent collections show
that the species is not uncommon in the Indian
River region of the central eastern Florida coast,
especially on deeper water (60+ m) coquinoid
limestone ledges and reefs of the ivory tree coral,
Oculina varicosa Leseuer.
The few studies made on the larval development
of new world galatheid crabs (e.g., Rayner 1935;
Boyd 1960; Fagetti 1960; Boyd and Johnson 1963;
Fagetti and Campodonico 1971) have all been
made on eastern Pacific species, and the larvae of
Atlantic American galatheids, including the
genus Galathea, remain undescribed.
This paper provides the first description and
illustration of the complete larval development of
G. rostrata, as well as the first report on any
species of Galathea reared totally under labora-
tory conditions, from hatching to megalopal stage.
The larvae and postlarvae are compared with lar-
val stages known from other members of the
Galatheidea throughout the world, and shared
features are briefly summarized.
MATERIALS AND METHODS
Eight ovigerous females of G. rostrata were ob-
tained on 15 April 1977 by lockout diver from the
781
FISHERY BULLETIN: VOL. 76, NO. 4
Research Submersible Johnson-Sea-Link II, of the
Harbor Branch Foundation, Inc. The adult
galatheids inhabited a large clump of ivory tree
coral which grew in 80 m of water on Jeffs Reef,
lat. 27°32.8'N, long. 79°58.8'W, located about 17 n.
mi. (27 km) northeast of Ft. Pierce Inlet, Fla. The
entire coral colony was collected and returned to
the surface inside of a 500-/u,m mesh cloth bag.
Ambient seawater temperature on Jeffs Reef was
12°C at time of collection. The galatheids were
immediately placed in compartmented plastic
trays containing recently collected neritic seawa-
ter previously chilled to 10°C. Upon return to the
laboratory each adult specimen was transferred to
individual 100 x 80 mm covered glass laboratory
dishes filled with approximately 340 ml of seawa-
ter previously chilled to 15°C. Each isolated
female was maintained at this temperature, pro-
vided a change of chilled seawater, and fed freshly
hatched Artemia salina nauplii, daily. All speci-
mens were exposed to a 12-h light-12-h dark il-
lumination program in a controlled temperature
unit (CTU) until hatching occurred. Five females
survived in this regimen and yielded larvae over a
period from 16 April to 6 May 1977.
Seven larval series were initiated. Using
methodology previously described by Gore
(1968), five such series were cultured in 24-
compartmented plastic trays. These consisted of
two series of 8 and 24 larvae, held in the CTU at
15°C ( ±0.5°C), and three series of 24 larvae each,
maintained at cool laboratory room temperature
(ca. 20°C, ±1°C). Two mass culture series of about
30 larvae each were also established in individual
100 X 80 mm glass dishes at cool laboratory room
temperature, which was controlled by reverse-
cycle air conditioning, and was monitored daily
with a 7-day recording thermometer. Fresh
surfzone seawater (35.5-36%o) was collected week-
ly, filtered through glass wool, stored in 14-gal (ca.
56-1) polypropylene carboys, and used throughout
the rearing period.
All larval series were checked daily, and any
molts or dead individuals were recorded and pre-
served in IWc ethanol. Specimens were examined
microscopically, slides prepared, and drawings
made as described in previous studies by Gore
(1968). Measurements given below are the arith-
metic average of all specimens examined in any
particular stage. A complete series of larvae, or
their molts, is deposited in the National Museum
of Natural History, Washington, D.C. (USNM
170862); the Allan Hancock Foundation, Univer-
782
sity of Southern California, Los Angeles (AHF
1028-01); the British Museum (Natural History),
London (BMNH 1978:103); and the Rijksmuseum
van Natuurlijke Historie, Leiden (D 31735).
RESULTS AND DISCUSSION OF
THE REARING EXPERIMENT
Galathea rostrata passes through four or five
morphologically distinct zoeal stages and a single
megalopal stage, before completing development
in the laboratory. Culture temperature undeni-
ably affects duration of development, and perhaps
larval survival as well. While the duration of the
zoeal and megalopal stages differed at each rear-
ing temperature, it was nevertheless generally
consistent within each of the temperature series,
as will be discussed below.
At 15°C five morphologically distinct zoeal
stages were observed for those larvae surviving to
metamorphosis. The minimum time required to
pass through these stages and attain megalopal
stage was 52 days. Most larvae remained in each
zoeal stage approximately 9-11 days through the
first four stages. Only two stage V zoeae survived,
and they remained as such 14 and 16 days before
molting to megalopa. However, neither of these
specimens survived longer than 6 or 7 days as
megalopae, so the mean duration of the postlarval
stage at 15°C remains unknown (Table 1). With
the minimum noted period of 6-7 day duration for
megalopae at this temperature, completion of de-
velopment and metamorphosis to first crab stage
Table l. — Duration of larval and postlarval development in
Galathea rostrata under laboratory conditions at the indicated
temperatures.
Temp
and
Days required to attain next stage
stage
Mm
Mean
Mode
Max
15°C:
1
10
10.8
10
14
II
8
9.7
10
M6
III
9
10.5
9
17
IV
8
9.4
9
11
V
14
—
—
16
Mg
6-7
(Both megalopae died in stage)
20 C:
1
5
5.8
6
7
II
4
4.2
4
5
III
4
4.1
4
5
IV (regular)
5
5.7
6
6
V (regular)
5
5.8
6
6
IV (advanced)
3
33
3
4
Mg (combined)^
12
12.6
13
13
III- IV (inter-
mediate)
7
—
—
8 (Died in stage)
'One zoea remained 30 days in stage II, dying 13 days later in stage III.
^Combined megalopae data include stages obtained from both IV (ad-
vanced) and IV (regular).
GORE: LARVAL DEVELOPMENT OF GMATHEA ROSTRATA
is conservatively estimated to take well over 60
days (Figure 1).
At 20°C either four or five morphologically dis-
tinct zoeal stages occurred. The minimum time
required to complete larval development and
reach megalopa was 18 days. Most larvae re-
mained in each zoeal stage from 3 to 6 days and as
megalopae from 12 to 13 days. The total duration
of development from hatching to first crab stage
spanned a minimum of 30 days at 20°C, if only four
zoeal stages were required, but took at least 37
days with five zoeal stages (Figure 1, inset).
The larvae generally fared well at both culture
temperatures. Although the larvae at 15°C took
longer to complete their development, they ini-
tially appeared to survive better than their coun-
terparts at 20°C (Figure 1). At 15°C, at least 50^7^
larval survival occurred through stage IV, before a
rapid decrease occurred in stage V and megalopa.
Larvae reared at 20°C exhibited a steep decline
after stage 1, to about 35'??^ survival, and showed a
continual decline thereafter. The precipitous de-
cline in larval survival at this temperature from
stage I to stage II was the result of an almost
complete mortality in one culture tray, for un-
known reasons.
At 15°C ecdysis in the earlier zoeal stages (I-III)
generally was a less critical period than at 20°C,
although the larvae at the latter temperature
were still able to complete most molts. The larvae
at 20°C attained subsequent stages more rapidly
than did those at 15°C, and some were able to
complete zoeal development, although overall lar-
val mortality was relatively higher. On the other
hand, at 15°C larval survival may have been en-
hanced by lower temperature, but the major
difficulty then seemed to be the attainment of
stage V and megalopa. Only two megalopae were
obtained in the 15°C program and neither was able
to molt to the succeeding first crab stage. In con-
trast, four megalopae survived at 20°C, and
molted to crab stage I; three of these specimens
were maintained in the laboratory to crab stages
XII and XIV.
Ecdysial and Sequential Variation
in Galathea rostrata
Two modes of developmental variation were
noted in G. rostrata at 20°C. In one mode, some
zoeae III molted to an instar which, for purposes of
discussion, is labelled "regular" stage IV. This
stage was characterized, among other features, by
a reduced number of antennular aesthetascs and
was always without well-developed pleopod buds
on the abdominal somites. Zoeae remained in this
stage for 3-4 days before molting to stage V, an
instar possessing distinct, well-developed, pleopod
buds and an increased number of antennular aes-
thetascs. The duration of stage V lasted 5-6 days
and was followed by the molt to megalopa. One of
these postlarvae subsequently molted to first crab
stage.
In the second mode of variation, some zoeae III
molted to an "advanced" stage IV, with some, but
not all, of the features as noted above for stage V.
Zoeae remained longer in the advanced stage (5-9
days) before molting directly to megalopa. Three
of these megalopae went on to attain first crab
stage. The two types of development are compared
in the inset of Figure 1.
Two other stage III zoeae, which remained in
stage III 7-8 days (instead of the usual 4-5), molted
to what appeared to be an intermediate stage IV.
These zoeae exhibited some stage V zoeal features
in size, maxillipedal setae numbers, and in posses-
sing pleopod buds, although the latter were only
rudimentaiy. A reduced number of antennular
aesthetascs similar to that of regular stage IV
zoeae was also seen. The two specimens survived
only 4-5 days in this stage before dying. This mode
of variation was not considered as important as
the previous two modes and will not be discussed
further.
Remarks
The regular and advanced fourth stages cannot
be equated to an early and late fourth stage, nor to
substages IVa and IVb, because no molt occurred
from one fourth stage or substage to another. If the
molt from stage III was to regular stage IV, this
was invariably followed by an ecdysis to stage V,
and then a subsequent molt to megalopa. If the
molt from stage III produced an advanced stage
IV, this in turn molted directly to megalopa, skip-
ping stage V altogether. At 15°C the regular stage
IV and stage V appear to be necessary plateaus in
larval development, whereas at 20°C development
may proceed in some zoeae without resorting to
either of these instars.
The regular fourth zoeal stage (as defined
above), therefore, appears to be a true sequential
stage of development, inasmuch as it was seen in
larvae at both 15°C and 20°C programs. However,
it is also a stage which can occasionally be skipped
783
FISHERY BULLETIN: VOL. 76, NO. 4
100
5 10
DAYS IN STAGE
784
Figure l.— Percentage survival and duration of stages in larvae of Galathea rostrata reared under laboratory
conditions at 15°C (upper 2 graphs) and 20°C (lower graph). Inset at 20°C gives duration and survival of regular
(IVr, dashed line) and advanced (IVa, solid line) stages; number of days the same as in larger graphs.
GORE: LARVAL DEVELOPMENT OF GALATHEA ROSTRATA
by some 20°C larvae, and thus could be thought of
as an intercalated stage, if the advanced stage IV
be considered more indicative of the developmen-
tal sequence. Other features shared between the
advanced fourth and regular fifth zoeal stages
(besides the presence of well-developed pleopod
buds noted earlier) include increased numbers of
antennular aesthetascs, a remarkable elongation
of the antennal endopodite, the appearance of a
mandibular palp, and slight changes in setae
number on maxillulae, maxillae, maxillipeds, and
telsonal uropods (see section on Description of the
Larvae). Moreover, the advanced stage IV zoeae
were always larger than the regular stage IV
zoeae.
It will probably remain a question of semantics
whether the regular stage IV is considered an in-
tercalated stage or one that occasionally may be
skipped. It could just as well be asked whether the
advanced stage IV was an intercalated stage be-
cause it embodies many of the features of regular
stage IV, plus some seen only in stage V zoeae in
the developmental sequence. What is of more im-
portance in the development of G. rostrata is that
the substitution of an advanced stage IV and the
subsequent elimination of the regular stages IV
and V allows earlier postlarval metamorphosis.
The resultant early benthic crab stages may be
reached in a shorter period of time by the species,
thereby reducing the time spent in" the plankton.
Discussion
It is, of course, conjectural as to whether the
larvae of G. rostrata skip stages in their develop-
ment in the natural environment or are ever sub-
ject to constant low (e.g., 15°C) or intermediate
(20°C) seawater temperatures. The adults of the
species, found in deeper continental shelf waters,
presumably are often exposed to cool seawater
temperatures, as was noted, e.g., during the time
the adult females for this study were collected. It is
not unreasonable to assume that developmental
stages may occasionally be subjected to relatively
constant cool temperatures as well, either im-
mediately after hatching or just prior to postlarval
metamorphosis when the megalopae settle to the
sea floor. In addition, should the larvae become
entrained in cyclonic cold core rings of Gulf
Stream origin (see Richardson 1976; Wiebe 1976;
Wiebe et al. 1976), they would presumably be sub-
jected to relatively constant cold water (at least
17°C) for at least part of their developmental
period. Delayed metamorphosis provides an alter-
native hypothesis against the more traditional
"stepping-stone" idea, to account for the rather
extensive distribution of the species along the
Middle and North American continental shelves.
There is some evidence that larvae of other
species of Galathea may skip stages in the
plankton (Lebour 1930, 1931) and that other
galatheids may intercalate substages (e.g., Boyd
and Johnson 1963). For example, the larvae of four
of the five British galatheids described by Lebour,
viz. Munida rugosa (Fabricius 1775 [as M. banffica
= M. bamffica (Pennant 1777)]), Galathea inter-
media Lilljeborg 1851, G. squamifera Leach 1814,
and G. strigosa (Linnaeus 1767) developed
through four zoeal stages, whereas G. dispersa
Bate 1859, exhibited four or five stages. Lebour
( 1930) considered five stages in the latter species
as "probably normal" but pointed out that the
megalopa could be obtained from the fourth
[numerical] stage, and "the normally fifth
[numerical] stage has been seen to emerge from
the third stage." She stated that the fourth or fifth
stage may therefore be omitted in G. dispersa, but
made no mention of intercalated stages or sub-
stages.
The developmental situation in G. dispersa is
quite similar to that noted in this report for G.
rostrata, in which an advanced fourth stage re-
places the regular fourth and fifth stages, thereby
causing them to be omitted from the developmen-
tal sequence. Lebour's (1930) "fifth stage. . . from
third" is probably equivalent to what is termed in
this report the advanced fourth stage. Her state-
ment that long, unjointed pleopods appear in the
"last" stage of G. dispersa indicates that either the
fourth stage (or advanced) or fifth stage (or regu-
lar) possess these appendages, depending on
whichever stage is "last." It also indicates that the
molt to megalopa does not occur without the ap-
pearance of pleopods in the "last" larval stage.
However, North Sea species of Galathea differ
from G. rostrata in possessing pleopod primordia
"in the third stage" which are "long but unjointed
in the last stage" (Lebour 1930). In addition, Sars
(1889) had also noted and illustrated pleopod de-
velopment in the "last" stage of larvae attributed
to G. intermedia, Munida rugosa, and Munidopsis
[as Galathodes]tridentata (Esmark 1857). The lat-
ter species will be considered further below.
Rayner (1935), using planktonic stages from
Argentinian waters, described the larvae he at-
tributed to Munida gregaria (Fabricius 1793) and
785
FISHERY BULLETIN: VOL. 76. NO. 4
M. subrugosa (White 1847). Rayner did not note
any substages or skipped stages in the five instars
he descritjed for the two species, and was not cer-
tain whether additional stages followed. By anal-
ogy with M. rugosa [as M. bamffica] he thought it
possible that the next stage would be postlarval. In
this he was probably correct, but it seems strange
in retrospect that Rayner did not attach impor-
tance to the well-developed pleopods on the larvae
before him, a feature by which he earlier charac-
terized the fifth zoeal stage. These appendages in
other galatheid larvae are quite obviously de-
veloped at stage V (see Lebour, Sars, Boyd and
Johnson, and others), and Sars (1889) even drew
attention to them when describing his "last zoeal
stage."
Intercalation of substages, however, is known in
the genus Pleuroncodes, as was specifically discus-
sed by Boyd and Johnson ( 1963) in the larvae of P.
planipes Stimpson 1860.^ Five zoeal stages had
been initially noted in this species (Boyd 1960),
but a sixth stage, apparently unnatural and not
known to occur in the plankton, could be induced
in the laboratory. Boyd and Johnson thought this
stage was due to the presence of penicillin pills or
to the CaCOg buffer in the pills, used to control
bacterial growth in the cultures. These authors
also stated that numerical stage IV could be sub-
divided into a complex of from four to nine sub-
stages, each represented by a molt, all without
pleopods, but otherwise morphologically similar
to each other. Although no sequential substages
were skipped (e.g., a molt from substage IVa to
IVh), one or more substages could be omitted ter-
minally, with a subsequent molt to the morpholog-
ically discrete stage V, which possessed pleopods
(Boyd 1960). Boyd and Johnson suggested that in
P. planipes the number of substages in stage IV
was probably influenced by temperature, with
higher culture temperatures (e.g., 16°-20°C) pro-
ducing faster development but causing more sub-
stages to occur before the molt to stage V. They
noted, however, that other factors such as food
supply or crowding of larvae might also exert an
effect on the number of substage instars, but ne-
glected to consider the possibility that the large
number of induced substages in stage IV might
also be due to the use of antibiotics in the cultures,
as suggested by Fagetti and Campodonico ( 1971 ).
^'Both Stimpson (1860) in his original description of Pleuron-
codes planipes and Haig (1955) have suggested that the species
may prove to be only a northern Pacific form of the Chilean P.
Monodon.
786
Figure 2. — Galathea rostrata, zoeal stages in lateral and dorsal
view: (A, a) First zoea; (B, b) second zoea; (C, c) third zoea; (D, d)
fourth zoea (regular); (E, e) fifth zoea. Scale line equals 1.0 mm.
The Chilean congener, Pleuroncodes monodon
Milne Edwards 1837, also was found to have inter-
calated substages (Fagetti and Campodonico
1971). At 15°C, substage IVa-d were followed by a
molt to stage V, possessing pleopods; at 20°C a fifth
substage (IVe) was attained instead of zoeal stage
V. Whether stage IVe would be followed by ecdy-
sial stage V is not known because all larvae in
stage IVe died. However, the lack of pleopods in
stage IVe implies that stage V should occur, with
pleopods, before the molt to megalopa takes place.
Whether such substages occur in the plankton is
conjectural , but they certainly would present some
difficulty in separation because of their great simi-
larity to each other in samples collected from the
plankton.
Abbreviated larval development is also known
to occur in at least two galatheids. Sars ( 1889), in
describing the prezoel, "first" and "last" zoeal
stages of Munidopsis tridentata from Norwegian
waters suspected that development time was
shorter than that seen in Galathea , but came to no
conclusion as to the total number of stages. He
commented on the remarkably advanced features
exhibited in the early zoea, an observation later
supported by Samuelsen (1972). Samuelsen de-
termined that only three zoeal stages exist for M.
tridentata and further suggested that the
megalopal stage followed stage III because the
latter stage was in the same relative state of de-
velopment as some fourth zoeae which preceded
the megalopae in other galatheids. Samuelsen
noted that the presence of a mandibular palp,
pleopod primordia, antennular aesthetascs, an-
tennal setae, and scaphognathite setae in the
early zoeal stages were all advanced features usu-
ally restricted to later zoeae in other galatheid
larvae. The relatively nonsetose feeding append-
ages and endopodites of the natatory appendages
indicate that the larvae may not feed, although
they can swim well.
Al-Kholy (1959) described and figured larvae
attributed to a "Galathea sp." which apparently
developed through only three zoeal stages. How-
ever, no methodology was given, nor indication as
to whether the larvae were cultured in the
laboratory or collected from the plankton. It is
doubtful whether the species will ever be identi-
GORE: LARVAL DEVELOPMENT OF GALATHEA ROSTRATA
787
FISHERY BULLETIN: VOL. 76, NO. 4
fiable based on his incomplete descriptions and
rather stylized illustrations.
Advanced development is implied in but one
galatheid, the cave-dwelling Munidopsis
polymorpha Koelbel 1892. This species is pres-
ently known only from a littoral cave formed by
lava tunnels which connect to the sea in the Ca-
nary Islands (Fage and Monod 1936). These au-
thors never found more than five, extremely large
(1.5-1.8 mm in diameter) eggs on an individual
female. No larval stages were described, but it was
hypothesized that the young Munidopsis was well
advanced in development inside the egg and prob-
ably hatched into a form nearly like the adult.
Given the rather unique habitat for a Munidopsis,
advanced development in M. polymorpha would
not be surprising. The vast majority of other
species of Munidopsis are deep-sea forms, most of
which occur below 500 m (Mayo 1974) in the At-
lantic Ocean, although some species occur in shal-
lower waters on the continental shelf.
In summary, it is apparent that larval develop-
ment in the Galatheidae is quite diverse, includ-
ing advanced development (i.e., with imminent
metamorphosis) in the cave dwelling M.
polymorpha , abbreviated development with as few
as three larval stages (M. tridentata, Al-Kholy's
Galathea sp.?), to "normal" development of four-
five zoeal stages (e.g., Munida, Galathea). Sub-
stage intercalation is known in the genxx^Pleuron-
codes, but seems to be restricted to the fourth, or
penultimate, ecdysial stage. Intercalation of a
sixth zoeal stage, perhaps only a laboratory ar-
tifact, is also known in one species of this genus.
Skipped stages appear only in two species of
Galathea, and perhaps one of Munida, at present,
and these result in the elimination of regular zoeal
stages IV and V and their replacement by an ad-
vanced stage IV which subsequently molts di-
rectly to megalopa.
Developmental variation such as that just dis-
cussed allows some interesting speculation as to
its evolutionary consequences in view of the fact
that the phylogenetically closely related anomu-
ran family Porcellanidae generally undergo a re-
■"The term "direct" development is restricted in this paper to
those larvae which hatch from the egg in a form morphologically
similar to the adult and undergo no further metamorphosis.
Larvae exhibiting "advanced" development usually hatch in the
penultimate or ultimate zoeal stage and thus may undergo addi-
tional ecdysis prior to metamorphosis. Larvae with "ab-
breviated" development hatch as early zoeae (often with a pre-
zoeal or first zoeal stage present), but may dispense with one or
more intermediate stages in completing their larval develop-
ment.
duced developmental sequence of usually no more
than two zoeal stages. These stages appear to be
morphologically equivalent in most respects to
Galathea stages I and IV, sensu lato. Substages
have been postulated for some porcellanid larvae,
notably Indo-Pacific species, but are not positively
known to occur in Atlantic and eastern Pacific
species. Previously postulated substages in Atlan-
tic species have been shown to be the result of
accelerated morphological development without
an ensuing molt and have been seen primarily in
larvae collected from the plankton (Gore 1968 and
others). However, the larvae of the western Pacific
genus Petrocheles apparently do reflect their
galatheid ancestry by undergoing five zoeal stages
during development. Morphological features of
the telson, uropods, and antennal scale in these
larvae all resemble, to a greater or lesser degree,
their counterparts in larvae of Galathea and
Munida (Wear 1965). Further studies along these
lines should be most interesting and productive.
DESCRIPTION OF
THE LARVAE
First Zoea
Carapace length: 1.0 mm.
Number of specimens examined: 10.
Carapace: (Figure 2A, a). Typically galatheid,
somewhat inflated; rostral spine horizontal, little
expanded proximally, straight, extending to level
of scapherocerite spine, or slightly beyond, about
0.5 X carapace length(CL), unarmed; posterolat-
eral carapace margins armed with a series of
about 15 small denticles placed before large, pos-
terior spine; latter slightly more than 0.1 x CL;
dorsomedial carapace margin excavated, with
about 13 small denticles along sinus margin. Two
small setae medially above eyes; latter sessile.
Antennule: (Figure 3A). A simple rod, both en-
dopodite and exopodite fused to protopodite;
former with 1 elongate plumose seta, latter with 3
aesthetascs and 3 setae.
Antenna: (Figure 3B). Endopodite rodlike,
about 0.4 X scaphocerite length, fused to protopo-
dite, a single distinct spine at its tip, plus a long
plumose seta; scaphocerite usually with 9 setae
along margin, tip produced into long daggerlike
spine about 0.3 x total scale length; protopodite
788
GORE: LARVAL DEVELOPMENT OF GALATHEA ROSTRATA
Figure 3. — Galathea rostrata, first zoeal appendages: (A) Antermule; (B) antenna; (C) mandibles, lower view rotated interiorad to
zoea to show dentition; (D) maxillule; (E) maxilla; (F) maxilliped 1; (G) maxillped 2; (H) maxilliped 3; (I) telson. Scale lines total 0.3
mm.
789
FISHERY BULLETIN: VOL. 76, NO. 4
with sharply pointed spine ventrally, armed along
either side with distinct acute spinules; this spine
falling short of distal tip of endopodite; scattered
setae basally on protopodite.
Mandibles: (Figure 3C). Asymmetrical den-
tate and spined processes, as shown.
Maxillule: (Figure 3D). Endopodite seg-
mented, 3 terminal, 1 subterminal seta. Basal en-
dite with 2 large, widely separated strong spines,
plus 3 setae; coxal endite with 4 spines, 3 strong
setae.
Maxilla: (Figure 3El. Endopodite setae, pro-
gressing subterminally, 3-4. 3. plus 3 laterally,
and additional fine hairs as illustrated. Basal en-
dite proximal and distal lobes each with 3 regular
and 1 spinelike seta; coxal endite proximal and
distal lobes with 8, and 4 spinelike setae, respec-
tively. Scaphognathite with 4 lateral, 1 stout
elongate apical seta.
Maxilliped 1: (Figure 3F). Coxopodite with 2
setae. Basipodite setae formula progressing dis-
tally 2, 3, 3, 3. Endopodite five-segmented, setae
progressing distally 3, 2, 1, 2, 4 +1 (Roman nu-
meral denotes dorsal setae); all endopodal and
basipodal setae heavy, spikelike. Exopodite two-
segmented, 4 natatory setae.
Maxilliped 2: (Figure 3G). Coxopodite naked.
Basipodite setae 1, 2, progressing distally. En-
dopodite four-segmented, setal formula 2, 2,2,4-1-
I; all spikelike. Exopodite two-segmented, 4
natatory setae.
Maxilliped 3: (Figure 3H). A small, unseg-
mented amorphous bud.
Pereiopods: Appear as small and undifferen-
tiated buds, gradually enlarging as stage progres-
ses.
Abdomen: (Figure 2A, a). Five somites; last 2
with large lateral spines; somites 2-5 each with
paired setae dorsally, plus a series of small distinct
spinules along posterior margin of somite; somite
6 fused to telson. Pleopods absent.
Telson: (Figure 31). Setal formula on margin 7
+ 7; all plumose setae (= processes 3-7) with
small, hooklike spinules progressing down their
790
length; other setae and hairs as illustrated. Anal
spine absent.
Color: Zoea transparent; frontal region of
carapace diffused with orange, brighter orange
dorsally on midgut region. Chromatophores as fol-
lows: orange on protopodite of antennule, faintly
orange on scaphocerite of antenna; red-orange
around inner oral region; mandibles and labrum
outlined in red, interiorly orange; basipodites of
maxillipeds 1 and 2 red-orange along dorsal and
ventral margins; red spiderlike chromatophores
dorsally in longitudinal line on abdominal somites
3-5; orange chromatophores ventrally placed in a
similar manner. Eyes black, with bluish high-
lights in reflected light.
Second Zoea
Carapace length: 1.2 mm.
Number of specimens examined: 8.
Carapace: (Figure 2B, b). More inflated; ros-
tral spine more or less knifelike in lateral view,
noticeably expanded proximally in dorsal view;
about 0.5 X CL, overreaching distal tip of antennal
scaphocerite spine in several specimens, unarmed;
posterolateral margins of carapace with about 14
small denticles or spinules, dorsomedial margin
possessing only scattered nubs or with denticles
totally absent; posterior spines remain slightly
more than 0.1 x CL; eyes now stalked.
Antennule: (Figure 4A). Incipient segmenta-
tion seen at junction of exopodite with protopodite;
former usually carrying 4 aesthetascs and setae,
with 4 small thick setules on junction with pro-
topodite. Endopodite retains single long plumose
seta.
Antenna: (Figure 4B). Endopodite thickened,
drawn into point distally, appearing conical, about
0.3 X scaphocerite length, incompletely fused to
protopodite, now lacking elongate plumose seta
seen in first stage. Scaphocerite usually with 10
marginal setae, plus numerous small spinules
ventrally along outer margin; distal spine about
Figure 4. — Galathea rostrata, second zoeal appendages: (A)
Antennule; (B) antenna; (C) mandibles; (D) maxillule; (E)
maxilla; (F) maxilliped 1; (G) maxilliped 2; (H) maxilliped 3; (I)
telson. Scale lines total 0.3 mm.
GORE; LARVAL DEVELOPMENT OF GALATHEA ROSTRATA
A-H y
791
0.2 X scale length. Protopodite now carries second
sharp spine ventrally, armed as first along outer
margins; larger ventral spine now shorter than
endopodite; ventral spinulelike setae inconspicu-
ous or lacking.
Mandibles: (Figure 4C). Dentition now larger,
more complex. No palp.
Maxillule: (Figure 4D). Endopodite unchanged
from stage I. Basal endite with 4 large spines, 3
setae: coxal endite processes stronger, but number
unchanged, from stage I, appearing to be 5 spines,
2 strong setae.
Maxilla: (Figure 4E). Endopodite setal formula
progressing subterminally 4, 2, plus 3 lat-
erally, and fine hairs as shown. Basal endite
proximal and distal lobes with 4-5, 6 processes,
respectively, former as 3-4 spinelike and 1 thin
seta, latter as 1 strong and 1 regular spine, 4 thin
setae. Coxal endite distal lobe with 3 spines, 1
strong seta, proximal lobe with 8 spines or strong
spinelike setae. Scaphocerite with 6 lateral, plus
usual elongate apical seta.
Maxilliped 1: (Figure 4F). Coxopodite and
basipodite setae unchanged from stage I. Endopo-
dite setal formula now 3, 2 -^ I, 1 + I, 2, 4 + I.
Exopodite remains two-segmented throughout
later development, now with 7 natatory setae.
Maxilliped 2: (Figure 4G). Coxopodite and
basipodite as in stage I. Endopodite setal formula
2, 2 -f- 1, 2, 5 + I. Exopodite as above and for later
stages, carrying at this stage 7 natatory setae.
Maxilliped 3: (Figure 4H). Remarkably de-
veloped; incompletely two-segmented exopodite
with 6 natatory setae; endopodite poorly calcified,
originating about half way up basipodite, two-
segmented, with 2 terminal setae.
Pereiopods: (Figure 2B). Undifferentiated, but
enlarging buds throughout stage.
Abdomen: (Figure 2B, b). Five somites, sixth still
fused to telson; lateral spine on somite 5 distinct,
that of somite 4 reduced, even vestigial; paired
dorsal setae on posterior dorsal margins of somites
2-5 remain, and are present throughout later zoeal
stages; posterior marginal spinules much reduced
in size and number.
792
FISHERY BULLETIN: VOL. 76, NO. 4
Telson: (Figure 41). Marginal setal formula 8
+ 8, additional pair added in medial sinus; latter
reduced from distinct U-shaped notch seen in
stage I. Armature on plumose processes as before,
but distal tips with hooklike processes more dis-
tinct; other setae and hairs as shown.
Color: Similar to stage I, but with less diffusion
of orange frontally; internal midgut region, man-
dibles and maxillipedal basipodites retain red-
orange color, mandibles showing noticeable red
outline, maxillipedal color appearing more dif-
fused than stage I; abdominal somites 4-5 with red
dorsal and lateral chromatophore lines, plus
orange line ventrally, all connecting to single
orange ring of spiderlike chromatophores around
each anterior margin of somites 4 and 5. Eyes
electric blue to black in reflected light.
Third Zoea
Carapace length: 1.3 mm.
Number of specimens examined: 8.
Carapace: (Figure 2C, c). Proximal margins of
rostral spine more developed laterally when seen
dorsally, in this and subsequent stages: length
remains about 0.4-0.6 x CL, distal tip reaches to
about tip of scaphocerite spine or slightly beyond;
posterolateral margins of carapace with denticles
much reduced, becoming irregular nubs; dor-
somedial margin with only poorly developed, rag-
ged nubs, almost totally obsolete; posterior
carapace spines considerably shortened, less than
0.1 X CL. Eyes much enlarged, basal peduncles
elongate.
Antennule: (Figure 5A). Exopodite segmented
from protopodite, bearing 2 lateral aesthetascs in
addition to 3 terminal, plus 3 or 4 setae, at tip.
Endopodite slightly enlarged, retaining long
plumose seta. Protopodite carries single long lat-
eral seta distally, plus 2 short fine setae, placed
medially, and basally, and 4 short stout setae dis-
tally.
Antenna: (Figure 5B). Endopodite continues to
develop, but remains incompletely segmented
from protopodite, now about 0.5-0.6 x scaphocerite
length, a thin seta just below spinous tip. Scapho-
cerite with 9-11 marginal setae, number some-
what variable on left and right appendages in
same specimen, plus additional shorter spinules
GORE; LARVAL DEVELOPMENT OF GALATHEA ROSTRATA
A-H ►
Figure b.—Galathea rostrata, third zoeal appendages: (A) Antennule; (B) antenna; (C) mandibles; (D) maxillule; (E) maxilla; (F)
maxilliped 1; (G) maxilliped 2; (H) maxilliped 3; (I) telson. Scale lines total 0.3 mm.
793
FISHERY BULLETIN: VOL. 76, NO. 4
along ventral outer margin; distal spine shortened
to about 0.1 X scale length. Protopodite retains 2
sharp ventral spines, larger about 0.6 x endopo-
dite length, smaller about 0.3 x length of larger.
Mandibles: ( Figure 5C ). Incisor and molar pro-
cesses more developed; no palp.
Maxillule: (Figure 5D). Endopodite un-
changed. Basal and coxal endites both with 5
spines, 3 setae.
Maxilla: (Figure 5E). Endopodite unchanged.
Numbers and form of processes on either endite
little changed from earlier stage, with exception of
basal endite distal lobe; latter now with 1 spine, 4
strong setae, 1 thin seta. Scaphognathite with 10
marginal setae and usual thick apical seta.
Maxillipeds 1 and 2: (Figures 5F, G). Coxal,
basipodal, dorsal and ventral endopodal, and
exopodal natatory setae as in previous stage.
Maxilliped 3: (Figure 5H). Endopodite bud
now subequal to basipodite length, incipient seg-
mentation more prominent in some specimens
than others; exopodite with 7 natatory setae.
Pereiopods: (Figure 2C, and detailed in-
set). More developed, many with incipient seg-
mentation; partial chelation of protochela often
visible.
Abdomen: (Figure 2C, c). Six somites present,
sixth divided from telson; distinct lateral spine
remains only on somite 5; spinules on dorsal poste-
rior margins vestigial, ragged and irregular. In
some specimens small, amorphous swellings occur
ventrally on somites 2-5, signifying future posi-
tion of pleopod buds.
Telson: (Figure 51). Marginal setal formula
remains 8-1-8; fourth pair of processes elongate
spines fused to telson; processes 3, 5-8 retain
noticeable hooklike spinules distally; other dorsal
setae as illustrated. Uropods present at junction of
abdominal somite 6 and proximal margin of tel-
son; exopods of same well developed, with variable
number of marginal plumose setae (usually about
8); endopods, if present, merely foreshortened
naked buds.
Color: More distinctly colored than stage II.
794
Orange chromatophores: dorsally on interior
margin of eyestalks, a single orange spot on
carapace laterally, just above each maxilliped 1,
another small grouping laterally on abdominal
somite 2; diffused orange on antennular peduncle,
ventrolaterally on carapace below eyes, interiorly
on mouthparts and within gut region, and on en-
dopodites of maxillipeds 1-3. Red chromatophores:
on cutting edge of mandibles, dorsomedially and
laterally on abdominal somite 4, laterally on so-
mite 5, the latter appearing as if small drops of
blood.
Fourth Zoea (Regular)
Carapace length: 1.4 mm.
Number of specimens examined: 10.
Carapace: (Figure 2D, d). Rostral spine with
noticeably raised lateral margins, greatly ex-
panded basally at point of attachment to carapace,
slightly overreaching scaphocerite spine and an-
tennular exopodite; carapace posterolateral and
dorsomedial margins unarmed; posterior spines
quite short, hooked downward. Eyes large, on
elongated stalks.
Antennule: (Figure 6A). Exopodite with three
rows of lateral aesthetascs, numbering distally
2-3, 3, 2-3, in addition to usual 3, plus 3 setae, at
tip. Endopodite about 0.5 x length of exopodite,
plumose seta absent. Protopodite retains distal
lateral setal, plus usual 4 stout setae at junction of
exopodite; 3 medial, 2 basal setae now also pres-
ent.
Antenna: (Figure 6B). Endopodite elongate sub-
equal to scaphocerite length; latter bearing 10-12
(numbers variable on left and right appendages in
same specimen) marginal setae plus numerous
ventral spinules on outer margin. Larger pro-
topodal ventral spine about 0.3 x endopodite
length, smaller about half size of larger; armature
of both remains as in earlier stages.
Figure 6. — Galathea rostrata, fourth zoeal appendages: (A)
Antennule , regular stage; ( a ) same , advanced stage; ( B ) antenna,
regular stage; (b) same, advanced stage; (C) mandibles, regular
stage; (c) same, advanced stage; (D) maxillule; (E) maxilla; (F)
maxilliped 1; (G) maxilliped 2; (H) maxilliped 3; (I) telson; all
regular stage; (i) detail, fifth telsonal process, 40 x objective. See
test for discussion. Scale lines total 0.3 mm.
GORE: LARVAL DEVELOPMENT OF GALATHEA ROSTRATA
A-H ^
795
FISHERY BULLETIN: VOL. 76, NO. 4
Mandibles: (Figure 6C). Molar and incisor pro-
cesses acutely spinous, otherwise unchanged from
earlier stages; no palp.
Maxilliile: (Figure 6D). Endopodite un-
changed, but may have small seta at base of seg-
ment. Basal endite with 7 stout spines, 3 setae;
coxal endite with 5 or 6 long spines, 3 strong setae,
and occasional small tooth.
Maxilla: (Figure 6E). Endopodite unchanged.
Basal endite distal lobe with 6 spines, 2 setae,
proximal lobe with 7 or 8 spines and strong setae
intermixed; coxal endite distal lobe with 4 termi-
nal spines, 1 lateral seta, proximal lobe with 11
spines, placed 5 terminally, 4 subterminally, 2
laterally. Scaphognathite with 17-20 marginal
setae, including 2 or 3 anteriorly near base, plus
usual long apical plumose seta, as shown.
Maxilliped 1: (Figure 6F). Basipodite adds a
single small seta proximally, ventral formula now
3, 3, 3, 3. Endopodal and coxal setae unchanged; 8
exopodal natatory setae.
Maxilliped 2: (Figure 6G). Coxal and basipodal
setae unchanged. Endopodite setal formula 2,2 +
I, 2 + I, 5 + I. Exopodal natatory setae 8.
Maxilliped 3: (Figure 6H). Endopodite over-
reaches basipodite, bearing 3 setae. Exopodite
with 8 natatory setae.
Pereiopods: (Figure 2D). Chelation and segmen-
tation more or less apparent, progressing rapidly
throughout stage; entire pereiopodal mass hangs
from beneath posterolateral carapace region in
later stage.
Abdomen: (Figure 2D, d). Lateral spine present
only on somite 5; dorsal spinulation on posterior
margins of somites nearly absent; paired dorsal
setae remain. A short, sharp spine on pos-
terolateral margin of somite 6, just above inser-
tion of uropodal basipodite. Pleopod primordia
may be present in some specimens, but develop-
ment is weak and occurs slowly, if at all, through-
out stage.
Telson: (Figure 61, i). Uropods completely de-
veloped, exopodite distal outer tip produced into
long spine, 8-11 long marginal setae present; en-
dopodites with 4 or 5 setae; with shorter setae on
796
both rami. Telson marginal setal formula 8 + 8,
fused fourth process now heavily spinulose, other
movable processes (except process 2, which, as in
other anomurans, remains a simple seta) carry
distinctive, sharp, separated, spinules along their
length (Figure 6i; 40 x objective), these spinules
much more hooklike distally, more spinous prox-
imally. Other dorsal and ventral setae on telson as
illustrated.
Color: Similar to stage III; quite developed and
noticeable along anterior and internal margin of
eyestalks; interiorly on midgut, and bases of
maxillipeds; single red chromatophores now bas-
ally on antennular protopodite, on posterior mar-
gin of maxillipedal basipodites, and laterally on
abdominal somites 3-5; eyes blue, reflecting green
highlights.
Remarks: This stage, with limited aesthetasc
numbers, reduced antennular endopodite and un-
segmented protopodite, lacking mandibular palps,
and with developing pereiopods and usually only
pleopodal primordia, always molted to zoeal stage
V.
Fourth Zoea (Advanced)
Carapace length: 1.6 mm.
Number of specimens examined: 12.
Carapace: Differs little from regular stage IV
except being larger, more inflated; armature simi-
lar to regular stage zoea.
Antennule: (Figure 6a). Endopodite about
0.75 X to nearly equal to length of exopodite; latter
with four rows of aesthetascs laterally, as 1, 3, 3, 2,
plus usual 3, plus 3 setae terminally. Other setae
as illustrated.
Antenna: (Figure 6b). Endopodite distinctly
overreaches (1.2x) scaphocerite; latter with 12-14
marginal setae; larger ventral propodal spine
about 0.25 X endopodite length, remaining half
again as long as smaller propodal spine.
Mandibles: (Figure 6C). Large, heavily
toothed processes, distinguished now by undivided
simple palp.
Maxillule: May add one more process on basal
endite; tooth on coxal endite usually distinct.
GORE: LARVAL DEVELOPMENT OF GALATHEA ROSTRATA
Pereiopods: Well formed, completely segmented
and chelated, protruding almost totally from
under posterolateral carapace margins.
Abdomen: Somites 2-5 each with a pair of undi-
vided pleopod buds, gradually lengthening as
stage progresses, but never becoming bifid.
Color: Similar to regular stage IV zoeae.
Remarks: The zoeae in this stage are much more
developed morphologically, possessing a different
arrangement of antennular aesthetascs, a well-
developed antennal endopodite, mandibular
palps, segmented and chelated pereiopods, and
distinct (but undivided) pleopod buds. These zoeae
molt directly to megalopae, bypassing stage V
completely.
Fifth Zoea
Carapace length: 1.6 mm
Number of specimens examined: 8.
Carapace: (Figure 2E, e). Rostral spine with
lateral margins appearing somewhat embossed at
posterolateral angle of zoeal orbit; carapace lat-
eral margins deeply rounded, convex posterolat-
erally, unarmed; posterior spine recurved ventral-
ly in some specimens, nearly straight in others,
inner margin of same curving regularly inward to
deeply excavated dorsomedial margin of carapace;
latter entirely without armature. Eyes large,
ovoid, on well-developed elongate stalks.
Antennule: (Figure 7A). Exopodite with five
rows of aesthetascs laterally: 2, 3, 3, 3, 2, plus 3
and 3 setae at tip. Endopodite from about 0.75 x to
just subequal in length to exopodite. Protopodite
segmented into elongate basipodite and truncated
coxopodite; former with a single long plumose seta
distally, plus 4 stout setae terminally at exopodite
junction, 3 more medially; latter with 2 stout setae
ventrally near line of segmentation.
Antenna: (Figure 7B). Endopodite very notice-
ably longer than scaphocerite ( 1.3-1.4 x); latter
bearing 12-14 plumose marginal setae plus addi-
tional ventral marginal spinules as in earlier
stages. Larger propodal ventral spine less than
0.2 X endopodite length, smaller remains about
half the size of larger, both armed similarly as
illustrated. Toward end of larval stage trans-
parent endopodite reveals distinctly segmented
megalopal antennal flagellum within endopodal
sheath.
Mandibles: (Figure 7C). Noticeably dentate,
each with simple, distinct palp.
Maxillule: (Figure 7D). Endopodite un-
changed from regular stage IV; basal setule may
not be present. Basal endite with 8 stout spines, 3
setae; coxal endite with 6 long spines, 3 strong
setae, and small tooth, placed as illustrated.
Maxilla: (Figure 7E). Endopodite unchanged.
Basal endite distal lobe with 8 spines and strong
setae, 2 thin setae terminally, one regular seta
laterally; proximal lobe with 6 terminal, 2 sub-
terminal, 2 lateral processes, most appearing to be
strong setae and spines. Coxal endite distal lobe
with 2 spines, 2 strong apical setae, 2 thinner
subapical or lateral setae; proximal lobe with
about 13 spines and strong setae, progressing ter-
minally to laterally as 7, 4, 2. Scaphognathite with
about 22-25 marginal setae, including enlarged
plumose seta apically; 2 small setules present,
positioned laterally.
Maxilliped 1 and 2: (Figures 7F, G). Little
changed from previous stage.
Maxilliped 3: (Figure 7H). Little changed in
form from previous stage, except endopodite now
much larger, longer, extending well past distal
margin of basipodite; 3 setae as before.
Pereiopods: (Figure 2E). Extremely large, ap-
pearing to be nearly functional, protruding be-
neath, and forcing posterolateral margins of
carapace, outward; walking leg segmentation and
cheliped chelation distinctly visible.
Abdomen: (Figure 2E, e). Lateral spine on so-
mite 5, and that on posterodistal angle of somite 6,
the only armature. Pleopods present as
well-developed, bifid, buds.
Telson: (Figure 71, i). Uropods well-developed,
both endopodite and exopodite with variable
number of marginal setae, usually 8-10, and 10-13
or occasionally 14, respectively. Telsonal fused
and movable processes as illustrated; fourth pro-
cess distinctly spinulose; occasionally an extra
797
FISHERY BULLETIN: VOL. 76, NO. 4
798
GORE: LARVAL DEVELOPMENT OF GALATHEA ROSTHATA
FlClRE 7. — Galathea rostrata. fifth zoeal appendages: (A) An-
tennule; (B) antenna; iCi mandibles, lower view rotated ex-
teriorad ofzoea to show dentition; (Dl maxillule; (E) maxilla; (F)
maxilliped 1; iGi maxilliped 2; (H) maxilliped 3; (Ii telson; (ii
detail i,40 x objective), fifth telsonal process. Scale lines total 0.3
mm.
plumose process appears making telson marginal
setal formula 8 + 9 as shown.
Color: Chromatophores as follows: Red, on an-
terior margin of eyestalks, paired on carapace dor-
sally just behind eyes on frontal region, single
ventrolaterally beneath each eyestalk just above
mandibular region, ventrally on both antennular
and antennal peduncles at junction with carapace,
laterally on carapace above insertion of maxil-
liped 2; interiorly on mouth region on outer mar-
gin of mandible, posterior to mandible on maxil-
lule, and on midgut; abdominal somites 3-5 with
several groups laterally, plus a reddish-orange
line above hindgut of same. Orange
chromatophores in elongate streaks longitudi-
nally on basipodite of maxillipeds 1-3, more dif-
fused on maxillipedal endopodites, and in lateral
groupings on abdominal somites 3-5. Eyes blue-
green in reflected light, corneas dark, probably
black.
Remarks: This stage followed the regular stage
IV and invariably molted to megalopal stage.
Megalopa
Carapace length x width: 1.7 x 1.2 mm.
Number of specimens examined: 4.
Carapace: (Figure 8A, B). Resembling minia-
ture adult; rostrum triangular proximally, drawn
into sharp point distally, armed along lateral
margins with 4 distinct spines, some smaller
spinules occasionally interspersed; frontal region
with additional spinules as illustrated; 2 elongate
thickened setae on gastric region, plus other setae
and spinules as shown; lateral margins with 4
large spines, including 1 at epibranchial angle, 2
placed about equidistant behind, and the fourth at
junction with cervical groove; a variable number,
usually 3, smaller spines laterally between larger
spines; a fifth large spine on posterolateral mar-
gin, followed by another, smaller, dorsally and
posteriorly. Numerous small setae scattered over
entire carapace; eyes each with 2 large, feathery
setae on anterodorsal margin.
AnU'iuutle: ( Figure 9A, a). Biramous; peduncle
large, three-segmented; basal segment inflated,
with 2 large forward-directed spines dorsally,
another, smaller, distoventrally; other setae as
shown; remaining two segments nearly smooth,
sparsely setose. Lower ramus three-segmented,
tip with 2 spinules (see detail. Figure 9a), other
setae as shown. Upper ramus six- or occasionally
indistinctly seven-segmented; aesthetascs on
segments two through five in the following se-
quence of rows and numbers: one row (2), two rows
(3, 3, + 2 setae), two rows (3, 2,-1-1 seta), one row
< 2 ); sixth segment with a single elongate terminal
seta plus other smaller setae.
Antenna: (Figure 9B). Peduncle three-
segmented, heavily spined; flagellum with 2 or 3
fused segments plus a variable number ( about 24)
shorter segments each bearing 5 or 6 setae dis-
tally; terminal segment with 7 longer setae, as
illustrated.
Mandible: (Figure 9C). Symmetrical, scoop-
shaped process, chitinized along leading margin; a
three-segmented palp, basal segment of which
bearing 2 short, spinelike setae, third segment
with about 13 or 14 stout, toothed spines.
Maxillule: (Figure 9D). Endopodite now pos-
sessing but a single short, terminal seta. Basal
endite with 4 strong terminal setae, followed by 16
short, stout spines, 4 subterminal and 3 lateral
setae; a single seta basally as shown; coxal endite
lower portion extended into elongate, weakly
chitinized, lobe fringed with fine hairs; 3 basal
setae, 3 lateral setae, followed by 11 stout spines
and 8 strong setae terminally.
Maxilla: (Figure 9E, e). Endopodite with a
single, long subterminal seta. Coxal and basal en-
dites heavily spinose and setose, numbers and po-
sition difficult to discern, but approximately as
follows: basal distal lobe, about 14 terminally, 4 +
2 subterminally, 2 laterally; proximal lobe, about
6 terminally, 3 + 1 subterminally, 1+2 laterally;
coxal distal lobe, about 3 each, terminally and
subterminally, 2 + 8 in irregular row laterally;
proximal lobe, about 11 placed more or less termi-
nally, 8 subterminally, 22 in a row encircling lobe
laterally, 1+2 beneath these; for exact position-
ing refer to outer (Figure 9E) or inner (Figure 9e)
views of lobes. Scaphognathite with about 40
799
FISHERY BULLETIN; VOL. 76, NO. 4
yj
ff:l
//
,/V/\
\
V:
'r^
^
y/
_^
rf
K
J
/■
/
800
B
Figure 8. — Galathea rostrata, megalopal stage: (A) lateral
view; (B) dorsal view. Scale line equals 1.0 mm.
GORE: LARVAL DEVELOPMENT OF GALATHEA ROSTRATA
Figure 9. — Galathea ros^rato, megalopal appendages: (A) An-
tennule; (a) detail, tip of lower ramus; (B) antenna; (C) mandible;
(D) maxillule; (E) maxilla, exterior surface; (e) same, interior
surface showing only inner setation; (F) maxilliped 1; (G) maxil-
liped 2; (H) maxilliped 3; (I ) pereiopod 4; (J) cheliped, pereiopod 5;
(K) pleopod 1 (right) and 4 (left); (L) tail fan. Scale lines total 0.3
mm.
AAi.iai-
a.C-H.J I-
801
FISHERY BULLETIN: VOL. 76, NO. 4
marginal setae plus finer setules on either side of
upper lateral surface.
Maxilliped 1: (Figure 9F). Exopodite and en-
dopodite weakly chitinized; former two-
segmented, with 3 more or less terminal setae;
latter naked. Protopodite with about 27 and 15
setae on basal and coxal endites, respectively,
placed as illustrated.
Maxilliped 2: (Figure 9G). Exopodite two-
segmented, 8 terminal, plus other setae, as shown.
Endopodite four-segmented, proximal two with 4
and 2 setae, respectively, distal two each with
about 12 processes, including 5 daggerlike spines
terminally. Setae on basipodite and coxopodite as
illustrated.
Maxilliped 3: (Figure 9H). Exopodite two-
segmented, 8 terminal setae. Endopodite five-
segmented; ischium and merus each with strong,
sharp triangular spine, plus a shorter spine at
anterodistal angle; ischium also with prominent
crista dentata; last three segments (carpus, prop-
odus, dactylus) with 3, about 15, about 18 long
daggerlike spines plus numerous longer setae in-
terspersed among them. Several setae on coxopo-
dite and basipodite.
Pereiopods: (Figures 8 A, B; 91, J). Chelipeds
rounded, equal, elongate, heavily spined, covered
with long, stiff bristlelike setae, these more prom-
inent in gape of fingers and on outer surface of
manus; fingers of each hand trifid at tips. Merus
and carpus with marginal spines. Walking legs
thin, elongate; merus, carpus, and propodus cov-
ered with setae plus small spinules ventrally
along margins, these often difficult to discern ex-
cept under higher (40x objective) magnification;
propodus with 2 larger spinules ventrally; dac-
tylus with 3 large movable spinules plus one fixed
triangular tooth on ventral margin; a second, very
small, almost vestigial tooth may appear about
midway between larger fixed tooth and dactylar
tip. Pereiopod 5 chelate, 1 long serrated seta, 3
scythelike pectinate setae quite noticeable, plus
additional numerous setules on manus; 2 vei'y
small, spinulelike teeth on distal tip of dactylus.
Pleopods: (Figures 8B, 9K). Occur on somites
2-5; biramous, greatly elongate; exopodal setae
progressing toward telson 8, 8, 8, 7, with minor
variation of 1 or 2 occasionally seen on left or right
802
side in same specimen; endopodites not as long as
exopodites, thin, naked, but each with appendix
interna of 2 or 3 small hooks developed at tip.
Tail Fan: (Figure 9L). Telson with 6 or 7 long
plumose setae, and several shorter marginal setae
interspersed among these, numbers of latter in-
consistent in same specimen; 1 or 2 small toothlike
spines laterally, as shown. Uropods biramous,
each with 4 widely separated marginal spines,
that on outer lateral margin of endopodite the
strongest; exopodite with about 18-20, endopodite
blade with 11-14 plumose marginal setae, num-
bers again variable in same specimen. Smaller
setae on dorsal and ventral surfaces of tail fan as
illustrated.
Color: Megalopa beautifully colored. Carapace
and abdomen overall red-orange, dorsolateral
carapace margins and spines darker red; an ir-
regular longitudinal white, or semitranslucent
stripe extends dorsally from just behind frontal
region along entire length of carapace and ab-
domen; this stripe bordered with darker orange-
red chromatophores along its length; a similar
white stripe appears laterally, below which
carapace becomes translucent, but covered with
numerous red spiderlike chromatophores; a third
stripe appears ventrally on sternum, extending to
junction of abdomen. Numerous pale blueish-
white dots interspersed over dorsal surface of
carapace, especially on either side of previously
noted longitudinal stripe. Eyestalks orange-red,
with regular white band longitudinally, this meet-
ing second band which encircles distal margin of
eyestalk just before cornea; latter black, overlain
with dark red maculations. Chelipeds with distal
margin of mei'us, entire carpus, propodus, and all
but distal tip of dactylus ivory white; merus prox-
imally red-orange; cheliped finger tips orange.
Walking legs translucent, speckled with many
red-orange chromatophores, these coalescing to
form irregular bands on outer segments; dactyli of
latter clear, or light horn color.
DISCUSSION
In the western North Atlantic Ocean the family
Galatheidae is represented by four genera:
Galathea (2 species), Munida (31 species),
Munidopsis (48 species), and Phylladiorhynchus
(1 species). With the exception of the present re-
port, the larval development of the remaining 81
GORE: LARVAL DEVELOPMENT OF GALATHEA ROSTRATA
galatheid species known to occur in the western
Atlantic is unknown.
The majority of our knowledge on galatheid lar-
vae comes from studies conducted on species from
the eastern Atlantic and Pacific Oceans, and as-
sociated seas such as the Red, Mediterranean, or
North Seas. Lebour ( 1930) first characterized lar-
vae in the family Galatheidae, and Gurney ( 1942)
was the first to provide a synopsis of larval fea-
tures based on Lebour's work and studies he made
on western Pacific galatheid larvae. As might be
expected, only some of the characters considered
important by Gurney in 1942 remain valid today,
and the lack of detailed descriptions in earlier
studies on galatheid larval morphology prevents
comparative statements to be made among most of
the species for which the larvae are known.
Nevertheless, morphological differences in ros-
tral, carapacial, antennal, abdominal, and tel-
sonal features continue to be of some value in
distinguishing the larvae of at least five galatheid
genera.
In general, the larvae so far described for species
of Munida share several features with those
known from Galathea and Pleiironcodes, and are
thus somewhat indicative (as seems true for the
adults) of close relationships among the three gen-
era. Pleuroncodes, an eastern Pacific genus, is
morphologically very similar to Munida in several
larval features, more so than are larvae of
Galathea as presentlydescribed. As noted in the
following synopses, the larvae of the three genera
can be easily separated. The adults, based on pres-
ent taxonomic criteria, are distinct and generic
status is undoubtedly warranted.
The genus Munidopsis, on the other hand, is a
heterogeneous grouping of forms, some adults
bearing little resemblance to others in the taxon
(see Mayo 1974, for discussion). The larvae from
the sole species so far described, however, are cer-
tainly distinctive and do not resemble those from
other genera. The genus Miinidopsis, as presently
constituted, would seem to provide an ample
example of a taxon wherein the relationships
among the various species (and perhaps their ele-
vation to generic status) might be clarified on the
basis of morphological relationships among their
larvae.
The first zoeal larvae of the eastern Pacific
species Ceruimunida johni (Fagetti 1960) are
quite spinose but could perhaps be confused with
either Munida or Pleuroncodes larvae (Fagetti
and Campodonico 1971). It remains to be seen
whether later larval stages would be more distinc-
tive. The presence of a single ventral antennal
spine (instead of two as seen in other genera) is of
limited value, because Galathea and Munida
exhibit a single spine in stage I and two spines in
later stages (see below).
In the genus Galathea, larvae are principally
known from northeastern Atlantic species de-
scribed by Lebour (1930, 1931) and Sars (1889).
Live specimens of these species may be separated
from larvae of G. ro.strata by chromatophore color
and position, but unfortunately no further de-
tailed comparison is possible until the former
species are completely redescribed and illustrated.
This holds true for most of the studies by the 19th
and early 20th century authors which were listed
in Gurney (1942). The "Galathea sp." briefly de-
scribed and illustrated by Al-Kholy (1959) from
the Red Sea agrees in several respects with "typi-
cal" Galathea larvae, but differs in others.
Whether it may be equated with Gurney's (1938)
G. longimana remains uncertain as the brief de-
scriptions and illustrations of both authors pro-
hibit meaningful comparison between the two
studies, and those on other Galathea larvae.
In order to facilitate comparison between the
two western Atlantic Galathea species a summary
of larval features exhibited by G. rostrata is pro-
vided in Table 2. These may be applied both to G.
agossizii, when its larvae become known, and to
other Galathea larvae when expanded or more
complete descriptions are provided. In addition, a
provisional synopsis of larval characters for the
five genera discussed above is also presented. The
summaries have been extracted from the more
reliable larval descriptions, as so noted, and may
allow distinction among the more typical larvae in
each genus. As our knowledge increases further
modification may be required.
SYNOPSES OF GALATHEID LARVAE
In the following section, emphasis is placed on
the setal-spinal formulae of the larval telson.
Conventionally, this formula may be expressed
thusly: 8+8, indicating that eight telsonal pro-
cesses, consisting of fixed and movable spines,
setae, and thin hairs, occur on each side of the
telsonal midline. It is apparent now that the type
of these processes may provide a useful reference
feature in distinguishing between the various
galatheid larvae. Accordingly, spines (whether
movable or fixed) are herewith denoted by Roman
803
FISHERY BULLETIN: VOL. 76, NO. 4
Table 2. — Summary of zoeal features in the larval stages oi Galathea rostrata.
Zoea 1
Zoea II
Zoea III
Zoea IV (regular)
Zoea IV (advanced)
Zoea V
Rostral spine
Not expanded
Expanded
Expanded
Raised lateral
As in regular
Expanded
proximally
proximally
proximally
margins
stage
proximally
Posterior cara-
pace spines
Elongate
Elongate
Reduced
Slightly hooked
Slightly hooked
Hooked
Eyes
Sessile
Stalked
Stalked.
Stalked,
Stalked.
Stalked, greatly
enlarged
elongate
elongate
developed
Antennule
Simple rod.
As in stage 1
Exopodite seg-
3 lateral rows
4 lateral rows
5 lateral rows
no lateral
Endopod more
mented. 2 lat-
aesthetascs
aesthetascs
aesthetascs
aesthetascs
developed
eral aesthetascs
Endopod "2 exo-
Endopod subequal
Endopod subequal
Endopod reduced
Protopod lacks
Protopod with 1
pod length
to exopod
to exopod
lateral seta
lateral seta
Protopod unseg-
mented
Protopod segmented
Antenna
Exopod with seta
Exopod lacks seta
Exopod "2 scapho-
Exopod subequal
Exopod longer
Exopod much longer
Scaphocerile
Scaphocerite
cerite length.
to scaphocerite,
than scaphocerite
than scaphocerite
spine elongate
spine reduced
with apical seta
with apical seta
with apical seta
with apical seta
Mandibles
Without palp
Without palp
Without palp
Without palp
Palp present
Palp present
Maxillipeds
Endopod (1)
3.2,1,2,4 + 1
3.2 + 1.1 4 1.2.4 . 1
As in previous stage
and thereafter
Endopod (2)
2.2,2,4 + 1
2.2 + 1.2.5 + 1
2.2 + 1.2.5 + 1
2,2 + 1.2 + 1.5+1
As in previous stage
and thereafter
Endopod (3)
Bud
Less than basipod.
Subequal to basi-
Longer than basi-
As in regular
Much longer than
2 seta
pod, 2 seta
pod. 3 seta
stage
basipod, 3 seta
Exopods 1 , 2)
4
7
7
8
8
8
3)
0
6
7
8
8
8
Pereiopods
Amorphous buds
Developing buds
Developing buds
Well-formed buds
Segmented, che-
lated buds
Large, nearly func-
tional
Alxlomen
5 somites, later-
5 somites, spine
6 somites, spine
6 somites, spine
6 somites
6 somites
al spines 4, 5
on 4 reduced
on 4 vestigial
on 4 absent
Pleopods absent
Pleopods absent
Pleopods absent
Pleopod primordia
may be present
Pleopods present,
undivided
Pleopods present,
bifid
Uropods
Absent
Absent
Exopods present
Endopods rudimen-
tary
8 + 8 setae
Exopods and endo-
pods developed
As in previous Stage
and thereafter
Telson
7 + 7 setae
8+8 setae
8 + 8 setae
8 + 8 setae
8+8 setae
4th process
4th process
4th process
As in previous stage
and thereafter
movable
movable
fused
numerals, setae by Arabic numerals, and fine
hairs by lower case Roman numerals. It should
also be remembered that previously movable setae
may, in a subsequent stage, become fixed spines
and the setal formulae will change accordingly.
Thus, a setal configuration proceeding medially of
a fixed spine (I), a thin hair ( ii ), a regular seta (3), a
previously movable seta now a fixed spine (IV),
followed by four movable setae (5-8) results in the
telsonal formula of I + ii + 3 + IV + 5-8. While
somewhat more ponderous than the previously
used formula of 8 -1-8, it does provide a clearer
picture of the type of processes and their changes
throughout subsequent larval development.
Cervimunida (Fagetti I960)
Rostrum elongate, needlelike, noticeably den-
ticulate; carapace posterolateral and posterior
margins dentate; posterior spines extremely elon-
gate, reaching fifth abdominal somite; antennal
scaphocerite elongate, aciculate, distinctly spined
along outer margin, and upper surface, basal seg-
ment with a single dorsal spine, unarmed ven-
trally (thus differing noticeably from other
galatheids where the situation is exactly the re-
verse); abdominal somites spined dorsally, so-
mites 4 and 5 with large lateral spines; telson
deeply bifurcate, furcae heavily armed; setal for-
mula I -H ii + 3-7 (based on first stage zoeae).
Presumably four or five larval stages.
Galathea (Sars 1889; Lebour 1930, 1931)
Rostrum acute, often expanded at base, may be
armed distally; carapace posterolateral margin
usually spinulate or denticulate, posterior spine
rarely exceeding third abdominal somite; anten-
nal scaphocerite broad, flattened, basal segment
with single spine ventrally in stage I, two spines in
all other stages; posterodorsal margins of abdomi-
nal somites minutely denticulate, but may become
unarmed in later stages; distinct posterolateral
spines on somites 4, 5, or both but may be absent
later; no median dorsal spine on somite 6; telson
triangular, not deeply bifurcate in early stages,
becoming more elongate and truncately triangu-
lar in later stages; lateral spines may be denticu-
late; marginal setal formula in stages I and II of I
+ ii -(- 3-7, 3-8, respectively, and in all later stages
I -I- ii -f 3 -I- IV + 5-8. Four or five larval stages,
pleopods present in last stage, as primordia in
penultimate stage on occasion.
Munida (Sars 1889;
Lebour 1930, 1931; Rayner 1935)
Rostrum elongate, needlelike, spinulate on dis-
804
GORE: LARVAL DEVELOPMENT OF GALATHEA ROSTRATA
tolateral margins and tips in early stages, but may
be unarmed in later stages; a serrated posterolat-
eral carapace margin with noticeable posterior
spine, latter often extending to about fourth ab-
dominal somite; antennal scaphocerite elongate,
thin or even noticeably aciculate, often spined;
basal segment with a single ventral spine in first
stage, 2 in later stages; abdominal somites 2-5
with two or more spines or spinules dorsally, mar-
gin of somite 6 with a single larger median spine
from stage III onward; telson originally deeply
bifurcate in early stages of development, but be-
coming more triangularly truncate later, thus ap-
pearing similar to that in Galathea in later stages;
telson furcae often spined; telson setal formula I +
ii + 3-7, 3-7 or -8 in stages I and II and I + ii + 3 +
IV + 5-9, 5-10, 5-11 or -12 in stages III-V, respec-
tively. Four of five larval stages, pleopods present
in last stage.
Munidopsis (Sars 1889; Samuelsen 1972)
Rostrum broad, flattened, nearly spatulate in
all zoeal stages, profusely armed about outer mar-
gins; carapace with a large, forward-directed spine
on anterolateral margin; entire ventral and pos-
terolateral margins noticeably spinulate, posterior
margin rounded, lacking elongate posterior spine
otherwise typical of larvae in the family; antennal
scaphocerite a flattened blade, two spines ven-
trally; posterodorsal margins of abdominal so-
mites unarmed, a posterolateral spine present on
somite 5; telson broadly spatulate, posterior mar-
ginal setal formula 1 + 2 -l-iii + IV -I- 5-15 in stage
I, and I + 2 -t- iii -h IV + 5-15 in stages II and III;
other smaller hairs interspersed among setae
5-15. Three larval stages, pleopods present in
each.
Pleuroncodes (Boyd I960;
Fagetti and Campodonico 1971)
Rostrum flattened basally, expanded similarly
to that oi Galathea, distal portion acute, margins
noticeably spinulate, especially at tip; posterolat-
eral carapace margins serrated, elongate posterior
spines usually extending to fourth abdominal so-
mite; antennal scaphocerite narrow, not as acicu-
late as in some Munida , basal segment with either
1 or 2 ventral spines; abdominal somites 1-5 heav-
ily spined dorsally on posterior margins, becoming
somewhat reduced in spination in later stages;
somite 6 with a median dorsal spine in stage III
and later; telson deeply bifurcate in stages I and II,
becoming more truncately triangular in later
stages as in Munida and Galathea; furcae may be
denticulate; marginal setal formula I -I- ii + 3-7,
3-8 in stages I and II, and I + ii + 3 + IV + 5-9, 5-10,
and 5-12 in stages III-V, respectively. Five larval
stages, including up to eight substages in stage IV;
stage VI in laboratory culture; pleopods present in
stage V. The genus is presently restricted to the
eastern Pacific Ocean.
ACKNOWLEDGMENTS
I thank my laboratory assistants Kim A. Wilson
and Nina Blum for their aid in field collections and
in laboratory culture of the larvae. Robert M.
Avent and the Coral Biology Section of the Harbor
Branch Science Foundation Laboratory obtained
the ovigerous females which provided the larvae
used in this study. Susan Bass and Karen Rodman
recorded data and helped in the laboratory during
their tenure on Harbor Branch Foundation and
Smithsonian Institution Fellowships.
LITERATURE CITED
Al-Kholy. a. a.
1 959. Larval stages of three anomuran Crustacea (from the
Red Sea). Publ. Mar. Biol. Stn., Al Ghardaqa 10:84-89.
Boyd, C. M.
1960. The larval stages ofPleuroncodesplanipes Stimpson
(Crustacea, Decapoda, Galatheidae). Biol. Bull. (Woods
Hole) 118:17-30.
Boyd. C. M., and M. w. Johnson.
1963. Variations in the larval stages of a decapod crusta-
cean, Pleuroncodes planipes Stimpson (Galatheidae).
Biol. Bull. (Woods Hole) 124:141-152.
CHACE, F. a., Jr.
1942. Reports on the scientific results of the Atlantis ex-
peditions to the West Indies, under the joint auspices of
the University of Havana and Harvard University. The
Anomuran Crustacea. I. Galatheidea. Torreia (Havana)
11, 106 p.
Fage, L., and Th. MONOD.
1936. Biospeologica. LXin. La faunne marine due jameo de
agua. Lac. souterrain de I'lle de Lanzarote (Canaries).
Arch. Zool. Exp. Gen. 78:97-113.
Fagetti, E.
I960. Huevos y el primer estadio larval del langostino
(Cervimunida johni Porter 1903). Rev. Chil. Hist. Nat.
55:33-42.
Fagetti, E., and L Campodonico.
1971. Larval development of the red crab Pleuroncodes
monodon (Decapoda Anomura: Galatheidae) under
laboratory conditions. Mar. Biol. (Berl.) 8:70-81.
Gore, R. H.
1968. The larval development of the commensal crab
Polyonyx gibbesi Haig, 1956. (Crustacea: Decapo-
da). Biol. Bull. (Woods Hole) 135:111-129.
805
FISHERY BULLETIN: VOL. 76, NO. 4
GURNEY. R.
1938. VIII. The larvae of Galathea longimana Paul-
son. In Notes on some decapod Crustacea from the Red
Sea— VI.-VIII., p. 82-84. Proc.Zool. Soc. Lond., Ser. B, 108.
1942. Larvae of decapod Crustacea. Ray Soc. (Lond.)
Publ. 129, 306 p.
HAIG, J.
1955. Reports of the Lund University Chile Expedition
1948-49. 20. The Crustacea Anomura of Chile. Lands
Univ. Arsskr. (N.F. Avd. 2) 51(121:1-68.
1956. The Galatheidea ( Crustacea Anomura) of the Allan
Hancock Atlantic Expedition with a review of the Porcel-
lanidae of the western North Atlantic. Rep. Allan Han-
cock Atl. Exped. 8:1-45.
LEBOL'R, M. V.
1 930. The larvae of the Plymouth Galatheidae. I. Munida
hanffica, Galathea strigosa and Galathea dispersa. J.
Mar. Biol. Assoc. U.K. 17:175-187.
1931. The larvae of the Plymouth Galatheidae. II.
Galathea squamifera and Galathea intermedia. J. Mar.
Biol. Assoc. U.K. 17:385-390.
Mayo, B. S.
1974. The systematics and the distribution of the deep-sea
genus Munidopsis (Crustacea, Galatheidae) in the west-
em Atlantic Ocean. Ph.D. Thesis, Univ. Miami, Miami,
432 p.
MIYAKE, S., AND K. BABA.
1970. The Crustacea Galatheidae from the tropical-
subtropical regionof West Africa, with a list of the known
species. Atl. Rep. 11:61-97.
RAYNEK. G. W.
1935. The Falkland species of the crustacean genus Muni-
da. Discovery Rep. 10:209-245.
Richardson. P.
1976. Gulf Stream rings. Oceanus 19(3):65-68.
SAMUELSEN, T. J.
1972. Larvae of Munidopsis tridentata (Esmark) (De-
capoda, Anomura) reared in the laboratory. Sarsia
48:91-98.
Sars, G. 0.
1889. Bidrag til Kundskaben om Decapodernes For-
vandlinger. II. Lithodes — Eupagurus — Spiropagurus —
Galathodes — Galathea — Munida — Porcellana — Neph-
rops). Arch. Math. Naturvidensk. 3:133-201.
STIMPSON, W.
1860. Notes on North Amercian Crustacea in the museum
of the Smithsonian Institution. No. 11. Annals Lyceum
Nat. Hist. N.Y. 7:176-246.
Wear, R. G.
1965. Larvae of Petrocheles spinosus Miers, 1876 (Crus-
tacea, Decapoda, Anomura) with keys to New Zealand
porcellanid larvae. Trans. R. Soc. N.Z., Zool. 5:147-168.
WIEBE. P.
1976. The biology of cold-core rings. Oceanus 19(3):69-76.
WiEBE, P. H,, E. M. HURLBURT, E. J. CARPENTER, A. E. JAHN,
G. P. KNAPP III, S. H. BOYD, P. B. ORTNER, A.ND J. L. COX.
1976. Gulf Stream cold core rings: large-scale interaction
sites for open ocean plankton communities. Deep-Sea
Res. 23:695-710.
WH.LIAMS, A. B.
1965. Marine decapod crustaceans of the Carolinas. U.S.
Fish Wildl. Serv., Fish. Bull. 65:1-298.
806
A THEORETICAL EXAMINATION OF SOME ASPECTS OF
THE INTERACTION BETWEEN LONGLINE AND SURFACE FISHERIES
FOR YELLOWFIN TUNA, THUNNUS ALBACARES
William H. Lknarz' and James R. Zweifel^
ABSTRACT
This paper explores several aspects of a dual fishery (surface and longline) on yellowfin tuna, Thunnus
albacares. The work is exploratory in nature and results, though indicative, are not conclusive for any
specific fishery. Our results indicate that the yield per recruit is higher for the longline fishery than for
surface gear if all fish are available to both gears and higher for the combined gears than for either gear
fishing alone. Theeffect of fishing by one gear on yield to the other gear and the effect of the fishery on
stock fecundity is shown to be greater for the often assumed 1:1 sex ratio than for the ratios usually
observed. A simulation model was used to examine the interrelations of pattern of movement offish,
pattern of recruitment, and fishing strategy. It was assumed that movements were random and
recruitment occurred either only along the coast or throughout the fishing area. The results indicated
that either of these patterns of recruitment could allow for increased catch as the surface fleet moved
offshore. However, location or pattern of recruitment is shown to be important when measuring
natural mortality and for examining the potential of a localized fishery, primarily on younger fish,
relative to a fisherj' exploiting the full range of the stocks or to one taking primarily older fish. Tagging
and fecundity studies are suggested for further investigation of the questions examined in this paper.
An unsolved problem common to many of the tuna
fisheries of the world is the nature of the interac-
tion between longline and surface (i.e., seining,
pole and line, and occasionally trolling and shal-
low handline) fisheries for the same species.
Fisheries for yellowfin tuna, Thunnus albacares;
albacore, T. alalunga: bluefin tuna, T. thynnus;
southern bluefin tuna, T. maccoyii; and bigeye
tuna, T. obesus, are prosecuted by both types of
gear in the Pacific, Atlantic, and Indian Oceans.
Although there can be considerable overlap of
sizes of fish taken by the two types of gear, in
general, longline gear takes larger (older) fish.
Exploitation of a tuna stock by the two types of
gear presents management with the problems of
determining the effect of various combinations of
fishing effort by the two gears on both yield per
recruit to the two gears and recruitment to the
stocks. In order to make these determinations, it is
necessary to estimate 1 ) availability of the stock at
each age to each of the two gears [The available
portion of the stock is subject to both other mortal-
' Southwest Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service, NOAA, La Jolla, Calif; present ad-
dress: Southwest Fisheries Center Tiburon Laboratory, NMFS,
NOAA, 3150 Paradise Drive, Tiburon, CA 94920.
^Southwest Fisheries Center La Jolla Laboratory, NMFS,
NOAA, P.O. Box 271, La Jolla, CA 92038.
Manuscript accepted Mav 1978.
FISHERY BULLETIN: VOL. 76. NO. 4. 1979.
ity (any mortality not caused by gear of concern)
and fishing mortality caused by the gear of con-
cern. The unavailable portion of the stock is sub-
ject only to other mortality.], 2) fishing mortality
of the available portion of the stock caused by each
gear, 3) natural mortality, 4) growth, 5) fecundity,
and 6) the relationship between egg production
and recruitment.
The aim of this paper is to examine the interac-
tions between longline and surface fisheries for
yellowfin tuna and to determine the impact such
interactions may have on the assumptions often
made in assessment of yellowfin tuna fisheries and
thus on the assessment calculations themselves.
The paper is divided into three major sections. The
first section examines the relationship between
availability of the stock(s) of yellowfin tuna to
surface and longline fishing and yield per recruit
to the two gears. This is an important, and to our
knowledge unexamined, aspect of all tuna fish-
eries exploited by both types of gear; the subse-
quent sections examine two asepcts of the biology
of tuna that can affect the catch by each type of
gear. The second section examines the effect of age
specific sex ratios of yellowfin tuna on yield per
recruit to the two types of gear and on egg produc-
tion. The third section examines the effect of
807
FISHERY BULLETIN: VOL. 76, NO. 4
random migration or dispersal and location of re-
cruitment of yellowfin tuna on estimates of mor-
tality and yield per recruit to each gear. We have
restricted our analysis to yellowfin tuna but be-
lieve that the concepts that we develop apply to the
other species as well.
MATERIALS AND METHODS
While stocks of yellowfin tuna are subjects of
important fisheries in all tropical oceans, infor-
mation on vital parameters is sketchy and
nonuniform. For example, tagging information
available in the Pacific is lacking for the Atlantic
stocks. On the other hand, regulation of the Pacific
fishery makes interpretation of the catch informa-
tion more difficult. Hence it is necessary to pick
and choose from the available information that
which is most relevant to the problems at hand.
Although the parameters are likely to differ for
fish from different oceans, if not fish from different
areas of the same ocean, few studies have conclu-
sively demonstrated that such differences exist. In
addition, several (e.g., Lenarz et al. 1974) have
found that conclusions from studies such as de-
scribed in this paper are often insensitive to the
likely range of values of parameters such as
natural mortality, fishing mortality, and growth.
In the first and second sections, we have used data
primarily from the eastern Atlantic because his-
torically catches have been more equally shared
by longline and surface fisheries than in the east-
ern Pacific; in the third section we have modelled
the eastern Pacific since information on migration
patterns is more extensive. In both instances, the
results are intended to be general rather than
specific. Data extracted from one area and used in
another is thought to be the best available and the
question of real differences is left for further inves-
tigation.
With a noted exception, the growth equation L
= 194.8 X (1 - e-0 42u 0.67)) estimated by Le Guen
and Sakagawa ( 1973) and length-weight equation
W = 0.0000214L2^^36 estimated by Lenarz (1974)
are used for yellowfin tuna where L is fork length
in centimeters, t is age in years, and W is weight in
kilograms. Unless otherwise stated, we assumed
that the annual instantaneous coefficient of
natural mortality (M) is 0.8 (Hennemuth 1961).
We estimated age-specific fecundity from two indi-
ces derived by Hayasi et al. ( 1972) (Table 1). Their
index I was obtained from longline data and their
index II was obtained from surface data. The
Table 1. — Indices of fecundity of yellowfin tuna as interpolated
from Hayasi et al. ( 1972), for fish caught in the Pacific calculated
by multiplying average ova counts by percentage of mature
female fish for each age and then dividing each product by the
product calculated for age 3 fish.
Midpoint of size interval
Fecundity
Fecundity
(cm)
index 1
Index II
80
0.04
0.07
85
0.04
0.14
90
0.05
0.21
95
0.08
0.27
100
0.15
0.36
105
0.23
0.42
110
0.33
0.51
115
0.42
0.61
120
0.55
0.70
125
0.70
0.81
130
0.88
0.92
135
1.12
1.04
140
1.40
1.15
145
1.80
1.26
150
2.30
1.37
155
2.77
1.50
160
3.20
1.62
165
3.57
1.76
170
4.05
1.91
175
4.42
2.06
180
4.82
2.23
5.01
2.43
fecundity indices were calculated by Hayasi et al.
( 1972) for fish caught in the Pacific by multiplying
mean ova counts by percentage of mature female
fish for each age and then dividing each product by
the product calculated for age 3 fish. For much of
our work, we used estimates of the 1967-71 aver-
age size (age) composition of the Atlantic yellowfin
tuna fishery made by Lenarz et al. (1974) (Table 2).
Use of length-age key assumes that length and age
are equivalent. Sex composition shown in Table 2
is based on data from the Pacific.
Estimates of the size- (age-) specific instantane-
ous coefficient of fishing mortality (F,) on an an-
nual basis were made using the Gulland (1965)
and Murphy (1965) method. The computer pro-
gi-am COHORT, written by W. W. Fox, Jr., of the
Southwest Fisheries Center, was used to obtain
estimates of F, for each 5-cm size interval, begin-
ning at 32.5 cm. The estimation procedure was
initiated with a trial value of F, for the largest size
interval (Input F).
Estimates of F, were obtained from the average
1967-71 catch composition data (Table 2) as was
done by Lenarz et al. (1974). When feasible it is
more desirable to estimate F, from individual
cohorts. This was not done because of the small
number of years in the data series and belief that
estimates from the average composition would
adequately reflect conditions of the fishery. In a
latter study, Fonteneau and Lenarz (1974) esti-
mated F; for individual cohorts from a longer time
808
LENARZ and ZWEIFEL: INTERACTION BETWEEN LONGLINE AND SURFACE FISHERIES
Table 2. — Composite catch in numbers of yellowfin tuna by gear, sex, and size. Length composition by gear is based on
data from Lenarz et al. ( 1974) on the Atlantic fishery. Sex composition is based on data from the Pacific (Murphy and
Shomura 1972).
Age at
IVlidpoint
beginning of
interval
of size
interval
(cm)
IVIale
Female
(yr)
Surface
Longline
Total
Surlace
Longline
Total
1 0579
35
1.179
0
1,179
1,179
0
1,179
1 1325
40
14.528
0
14,528
14,528
0
14,528
1 2039
45
61.563
0
61,563
61,563
0
61.563
1 2888
50
186.611
4
186,615
186.611
4
186.615
1.3710
55
237.622
11
237,633
237,622
11
237,633
1 4562
60
210.711
226
210,937
210,711
226
210,937
1.5445
65
121.824
324
122,148
121,824
324
122.148
1 6363
70
137.389
1,076
138,465
137,389
1,076
138,465
1.7317
75
102.046
2,718
104.764
102.046
2.718
104,764
1 8310
80
90,710
2,847
93,557
90.710
3.847
93,557
1 9348
85
67,060
6,013
73,073
67.060
6,013
73,073
20432
90
52.541
6,525
59,066
52,541
6,525
59,066
2 1568
95
51,366
5,833
57,199
51,366
5,833
57,199
22761
100
56.714
7,537
64.251
56,714
7,537
64,251
24017
105
52.752
17.036
69.788
52.752
17,036
69.788
25343
110
51,497
20.105
71,602
51.497
20,105
71.602
26748
115
35,981
22,017
57,998
35,981
22,017
57.998
28240
120
26.167
21,430
47,597
26.167
21,480
47,597
2 9832
125
30,779
28,679
59,458
30.779
28,679
59,458
3 1538
130
26,001
29,272
55,273
26,001
29,272
55,273
33376
135
21.975
22,345
44.320
21,975
22.345
44,320
35368
140
16.749
26.035
42,784
16,749
26.035
42,784
37542
145
26.919
38,782
65,701
11,661
16.800
28,461
39935
150
31,942
36,099
68.041
8,450
9,549
17.999
4 2595
155
24,727
33.933
58.665
3,767
5,170
8,937
45590
160
18,701
22,644
41.345
1,524
1.845
3,369
49017
165
14,497
13.140
27.637
573
519
1,092
53021
170
5.621
6,162
11,783
94
103
197
57838
175
3,703
240
3,943
21
1
22
6 3883
180
1,836
55
1,891
3
0
3
Total
1,781,711
371.093
2,152,804
1.679.858
254.020
1,933.878
span and obtained results similar to Lenarz et al.
(1974).
The computer program MGEAR, written by W.
H. Lenarz, was used to obtain estimates of yield
per recruit using the Ricker ( 1958) yield equation.
A description and listing of MGEAR is available
from its author. The program was slightly mod-
ified to calculate indices of egg production using
the following equation
Ef,.,, = 0.5 (t2 - t,)N,^ {FI>^ + FI,fi -'^'. ^ '^'i'"^-''')
where E,^,,^ = index of egg production between
age ^1 and t.2,
FI,^ = index of fecundity for age ^. ,
A^,_ - number of females in population
of age ^,
F,^ = coefficient of instantaneous fish-
ing mortality between age t^ and
age t2, and
Mi^ = coefficient of instantaneous
natural mortality between age ^i
and age ^2-
For this equation it is assumed that the estimates
of FI are proportional to egg production per
female, which is assumed to be continuous, and
that the rate of egg production is linear over the
interval (^i, ^2*-
A computer program MIGR was written by J. R.
Zweifel to perform the calculations used for the
third section of this paper. Since new methodology
is developed, a description of the calculations will
be given in that section.
AVAILABILITY OF THE STOCK(S)
OF ATLANTIC YELLOWFIN TUNA
TO SURFACE AND LONGLINE GEAR
In previous works on yield per recruit, yellowfin
tuna of all ages in either the entire Atlantic (e.g.,
Hayasi and Kikawa 1970; Wise 1972; Hayasi et al.
1972; Lenarz et al. 1974), or in the eastern Atlan-
tic (e.g., Fonteneau and Lenarz 1974) were as-
sumed to be equally available to both longline and
surface gear. However, since the surface fishery
for yellowfin tuna occurs very close to the west
African coast (Fox and Lenarz 1973) while the
longline fishery for yellowfin tuna is distributed
throughout the tropical Atlantic, it seems possible
that the longline fishery is exploiting some fish
that are not available to the surface fishery. It is
809
FISHERY BULLETIN; VOL. 76. NO. 4
also possible that some stock(s) which are avail-
able to surface fishing are never available to the
longline fishery. Since significant tagging efforts
have begun only recently in the Atlantic and the
results of these studies have not been published,
data are not available to evaluate the availability
of yellowfin tuna to both gears.
However, there is evidence from the Pacific that
yellowfin tuna are not equally available to long-
line and surface gears. With the permission of W.
H. Bayliff of the Inter-American Tropical Tuna
Commission (lATTC), we examined yellowfin
tuna tag return data from the eastern Pacific dur-
ing 1963-66 in an attempt to evaluate the avail-
ability offish to both gears in that area. We tabu-
lated the number of tag returns for fish larger than
100 cm at return by 10-cm size groups (Table 3).
All of the fish hadbeenatliberty for at least 10 mo.
Although all of the tagged fish were measured
when released, not all were measured when recov-
ered. Bayliff recommended the relationship
167 (1 - e
-0,fii/-0.833i -
Again at the suggestion of Bayliff, we estimated
the expected return of tags from longline-caught
fish when all fish are equally available to both
gears. Assuming tag recoveries were independent
of each other, recovered tags were reported at the
same rate by both components of the fishery, and
tagged fish were equally available to both gears:
then the expected returns of tagged fish of size / by
gear / in year k is given by
E(R„,) =R,.,N„,JN,^k (1)
'1, when size is between 101 and 110 cm
,6, when size is between 151 and 160
. _ fl, when fish are caught by surface gear
l2, when fish are caught by longline gear
'1, when fish are caught in 1963
, 2, when fish are caught in 1964
3, when fish are caught in 1965
,4, when fish are caught in 1966
estimated by Davidoff ( 1963) for growth of yellow-
fin tuna in the eastern Pacific as the best equation
to estimate the size of unmeasured fish. All of the
returns were surface-caught fish, even though
longliners captured a considerable number of yel-
lowfin tuna in the eastern Pacific (east of long.
130"^ W) (Kume and Joseph 1969). In fact for many
of the 10-cm size groups, the longliners caught
more yellowfin tuna than the surface gear
operators (Table 4).
Table 3. — Number of returns of tagged yellowfin tuna from the
eastern Pacific Ocean by size interval and year iW. H. Bayliff,
pars, commun.).
Size interval (cm)
1963
1964
1965
1966
101-110
111-120
121-130
131-140
141-150
151-160
16
7
0
0
0
0
where /?,;/; = number of returns and
N ,ji; = number of fish caught.
A dot in the position of a subscript signifies sum-
mation of the variable over the subscript, e.g., X, .^
2
= 1 X,
./ = 1
,jh ■
Forty fish were returned by the surface gear
during 1963-66 (Table 3). Using the statistics of
Tables 3 and 4 and the three assumptions, a return
of 5.4 of these tags would have been expected
from the longline fishery and 34.6 from the sur-
face fishery. The chi-square value, corrected for
discontinuity, for the observed and expected re-
turns (Equation 1) is 5.13, with probability
slightly less than 0.025. The power of the test of
the hypothesis of independence, equal reporting
rate, and equal availability was reduced because
we combined the year and size strata to avoid
Table 4. — Catch of yellowfin tuna from the eastern Pacific Ocean (east of long. 130°W) in hundreds of
fish by size and gear iKume and Joseph 1969).
Size interval
(cm)
1963
1964
1965
1966
Surface
gear
Longline
gear
Surface
gear
Longline
gear
Surface
gear
Longline
gear
Surface
gear
Longline
gear
101-110
111-120
121-130
131-140
141-150
151-160
653
473
508
237
240
212
336
455
390
751
541
144
4,082
2.245
720
448
320
102
173
465
1.078
804
469
104
3.386
2.211
1.895
905
498
194
30
93
444
758
466
205
2.926
2.044
1.312
718
536
204
54
116
304
515
575
200
810
LENARZ and ZWEIFEL: INTERACTION BETWEEN LONGLINE AND SURFACE FISHERIES
strata with low expected values. The probability
under Equation ( 1) of a returned tag being from a
surface-caught fish (P,ik) is
P.u = N.u/Nik
(2)
The exact probability of all returns during the
1963-66 period being from surface-caught fish,
given the distribution of returns among year and
size categories, is
p.i. = n n (Pnk)"'"
1=1 * = 1
(3)
Our estimate of P j is 0.00152, which is very low
and indicates that Equation (1) does not hold.
Thus we may conclude that 1) tag returns are not
independent (e.g., fish that were captured from a
school and tagged may remain in the same school
until recaptured), and/or 2) longline recoveries are
reported at lower rates than surface recoveries,
and/or 3) the fish were not equally available to
both gears. Since all fish were at liberty for more
than 10 mo before being recovered, the assump-
tion of tag returns being independent seems likely
to be valid. The independence of tag returns would
seem to be a desirable subject for further research
since the assumption is so often made in analyses
of tag returns. A considerable number of southern
bluefin tuna have been recovered and returned by
longliners (Shingu 1970), indicating longline
fishermen do cooperate in tagging programs. Dur-
ing the period of the study, the surface fishery was
only beginning to move offshore (Calkins and
Chatwin 1971), while the longline fishery was dis-
tributed throughout the area (Kume and Joseph
1969). Also, the fish that were released were
caught by surface gear, tagged, and released in
nearshore areas. Thus, tagged fish were probably
more representative of fish exploited by the sur-
face fishery than those that were exploited by the
longline fishery, if two groups offish existed. Thus
it seems plausible that the tagged fish were not
equally available to longline and surface gears.
This is further evidence of unequal availability
of yellowfin tuna to the two gears in the Pacific.
Previously, Hisada (1973) showed that yellowfin
tuna caught near the surface using handlines
were of the same size as those caught by longliners
at the same time and in the same area of the
western Pacific. However, the surface-caught fish
tended to be more sexually mature except in areas
in which the 26°C isotherm occurred at depths
fished by longliners. He attributed this phenome-
non to a preference for warmer waters by sexually
mature fish and noted that larvae of yellowfin
tuna tend to be found at water temperatures ex-
ceeding 26°C. Thus, some yellowfin tuna evidently
behave in a fashion that makes them available to
surface fishing but not to longline fishing. Further
evidence along these lines is provided by Shingu
and Tomlinson (Patrick K. Tomlinson, Inter-
American Tropical Tuna Commission, La Jolla,
Calif. Pers. commun., 1974) who found that the
length-weight relationship estimated by Lenarz
(1974) for surface-caught yellowfin tuna in the
Atlantic was more representative of the longline
catch in the eastern Pacific than was the relation-
ship estimated by Chatwin (1959) for surface-
caught yellowfin tuna in the eastern Pacific.
With the above in mind, we considered three
hypothetical stock structures for the Atlantic yel-
lowfin tuna fishery: 1) the same stock(s) are
equally available to both gears, 2) half of the catch
of the longline fishery comes from stock(s) not
available to the surface fishery, and 3) the entire
catch of the longline fishery comes from stock(s)
not available to the surface fishery. The effects of
the three hypotheses on estimates of fishing mor-
tality and yield per recruit to the gear were
examined.
Using the data in Table 2, we estimated the
vector F of size-specific instantaneous mortality
rates F, under the three hypotheses which are
identified by the proportion ) of the longline catch
which comes from the stocks exploited by the sur-
face fishery as 0 = 1.0, 0.5, and 0.0 respectively.
For 4) = 1.0, all of the data in Table 2 was used to
estimate the F vector. For > = 0.5, the surface
catch plus 50% of the longline catch was used and
for (t> - 0.0 only the surface catch was used for
estimating F. When (/> - 0, an additional F vector
was estimated for a longline fishery operating
without the presence of a surface fishery by using
only the longline catch. The F vectors were then
used to calculate yield per recruit to the two gears.
Estimation of a vector of size-specific F requires an
estimate of natural mortality and size-specific F
for one size category. In all instances, we chose to
use an estimate of size-specific F for the fish
> 177.5 cm. This estimate will be referred to as
Input F. The final value of size-specific F was set at
0.2 following Lenarz et al. (1974). The estimates
(Figure 1) indicate that values of F for large fish
are directly related to the portion of the longline
catch that comes from the stock(s) exploited by the
811
FISHERY BULLEITN: VOL. 76. NO. 4
80 100 120 140
FORK LENGTH (cm)
Figure l. — Estimates of size- specific fishing mortality of Atlan-
tic yellowfin tuna as a function of porportion of catch ((/>) by
longline fishery that comes from stock(s) exploited by surface
fishery.
surface fishery. The relative values of yield per
recruit within a hypothesis are not significantly
affected by the portion of the longline catch that
comes from the stock(s) exploited by the surface
fishery (Figure 2). Therefore, the three hypothet-
ical stock structures do not seem to have much
bearing on decisions concerning minimum size
regulations.
Estimates of yield per recruit were also plotted
as functions of fishing effort (mortality), size at
recruitment, and portion of longline catch that
comes from stock(s) exploited by the surface
fishery. Again the relative values of the results are
not significantly influenced by the stock structure
(Figure 3a, b). We note that Figure 3 is in agree-
ment with the conclusion of Fox and Lenarz
(1974), ". . . that the Atlantic yellowfin fishery is
approaching or has obtained a plateau where sub-
stantially increased sustainable average yield of
yellowfin tuna will not be obtained by increasing
fishing effort without some concomitant change in
the constitution of the fishery. . . ." They used the
production model approach under the alternative
assumptions that either the longline or surface
gear exploits the same or separate stock(s).
The effect of the surface fishery on the longline
fishery was examined by estimating yield per re-
5 4r
9
i 5.0
q:
o
111
a:
4.6
4.2
3.8
J L
J L
J.
J L
32.5 52.5 72.5 92.5 112.5 132.5
FORK LENGTH AT RECRUITMENT (cm )
Figure 2. — Yield per recruit ( kilograms) ofAtlantic yellowfin to
surface and longline gear as a function of size at recruitment and
proportion of catch (>) by the longline fishery that comes from
stock! s ) exploited by surface fishery . The vector of fishing mortal-
ity is equal to the value at the time of study.
cruit to the longline fishery in the presence and in
the absence of a surface fishery (Figure 4). The
results suggest that if the two gears exploit the
same stock(s), the surface fishery reduces the po-
tential yield per recruit to the longline fishery by
about twofold at the position of the fishery during
the study period (i.e., multiplier of effort = 1) and
about fivefold for a threefold increase in effort. The
same procedure was used to examine the effect of
the longline fishery on the surface fishery (Figure
5). The results indicate that at the level of fishing
effort at the time of study, the yield per recruit to
the surface fishery would be increased by 25% if
the longline fishery ceased.
Although the presence of each fishery reduces
the yield per recruit of the other, the yield per
0 04 08 12 16 20 2'»
MULTIPLIER OF EFFORT
04 08 12 16 20 24 28
MULTIPLIER OF EFFORT
Figure 3. — Yield per recruit of Atlantic yellowfin tuna as a
function of fishing effort and proportion of catch i4>) by longline
fishery that comes from stock(s) exploited by surface fishery: (a)
size at recruitment is 32.5 cm, (b) size at recruitment is 77 cm.
812
LENARZ and ZWEIFEL: INTERACTION BETWEEN LONGLINE AND SURFACE FISHERIES
tr.
o
u
o
0.8 1.2 16 2.0
MULTIPLIER OF EFFORT
Figure 4. — Estimates of yield per recruit of Atlantic yellowfin
tuna to the longline fishery as a function of effort and presence ( 4>
= 1.0) or absence (({> = 0.0) of a surface fishery.
o
llJ
q:
o
^^^— ^
•— •
— •
•
Jj^ • — •
o — o
0 = 0.0
- Jd o — 0
f \ \ 1
0 = LO
1 1 1
1 1
J 1
0.4 0.8 1.2 1.6 2.0 2.4
MULTIPLIER OF EFFORT
2.8
Figure 5. — Estimates of yield per recruit of Atlantic yellowfin
tuna to the surface fishery as a function of effort and presence (<^
= 1.0) or absence {^ = 0.0) of longline fishery.
recruit of the combined fisheries is higher than the
yield per recruit of either fishery alone. The re-
sults suggest that if a catch quota system is im-
posed on the Atlantic yellowfin tuna fishery, all
components should be included unless it is shown
that different stock(s) are being exploited by the
gear.
The above results (Figures 4, 5) suggest that a
stock of yellowfin tuna will produce a potentially
higher yield per recruit to a longline fishery than
to a surface fishery, if the fish are equally avail-
able to the two gears. However, until the question
of availability is settled, it is not possible to predict
the potential production to the two gears. We point
out here that gear-specific availability is not well
known for any tuna fishery and would be difficult
to determine. Thus, we are faced with the prospect
of probably being forced to determine empirically
the production potential for each gear in each
fishery. After a fishery is established, an analysis
of the type conducted on the Atlantic yellowfin
tuna fishery could be used to examine the effects of
availability to the two gear types, and a tagging
study could be designed to provide the required
answers.
EFFECTS OF AGE-SPECIFIC SEX
RATIOS OF ATLANTIC YELLOWFIN
TUNA ON YIELD PER RECRUIT TO
THE TWO TYPES OF GEAR AND
STOCK FECUNDITY
While a number of authors have noted that the
ratio of females to males appears to be less than 1:1
for catches of larger tunas, none to our knowledge
has incorporated these observations into calcula-
tions of yield per recruit or stock fecundity.
Beardsley ( 1971) reported that the ratio of female
to male Atlantic longline-caught albacore was
233:365 during the December 1969-September
1970 period. Males increasingly dominated at
sizes >100 cm. Females slightly outnumbered
males between 92 and 100 cm. One explanation for
the catch curves estimated by Beardsley is a 1:1
sex ratio at small sizes, a slightly slower growth
for females for fish >90 cm, and beyond 100 cm,
either a higher rate of natural mortality for
females or a change in behavior that makes
females less available than males to longline
fishing. Other explanations exist, e.g., a combina-
tion of low sex ratio and slow growth of females
throughout their life. Sakamoto ( 1969) noted for
Atlantic bigeye tuna, ". . . males predominated in
areas of higher water temperature. Proportion of
females increase as the water temperature gets
lower." His data indicate that as size increases the
proportion of females decreases and females may
grow slower than males in waters between lat. 30°
to 50°N, but not in equatorial waters. Data pre-
sented by Kikawa (1964) indicate that southern
bluefin tuna >150 cm are predominantly males,
while females often outnumber males at smaller
sizes. Thus, female southern bluefin tuna may
grow more slowly than males.
Since there is considerable evidence for age-
specific changes in the sex ratio of tunas, we be-
lieve that the effects of such changes on estimates
of yield per recruit to each gear type and fecundity
should be investigated. We have assumed sex
ratios to be the same as with Pacific yellowfin tuna
because no extensive studies of age-specific sex
ratios for Atlantic yellowfin tuna have been pub-
lished. We used results from a study by Murphy
and Shomura (1972), who found that beyond 140
cm male yellowfin tuna greatly outnumbered
females (Figure 6). The data in Figui'e 6 do not
show a large excess of females in any size interval
and thus no evidence of sex-specific growth is
exhibited. Using their data and the age-length
813
FISHERY BULLETIN: VOL. 76, NO. 4
60 70 80 90 100 110 120 130 140 150 160 170
LENGTH (cm)
Figure 6. — Length distribution by sex of longline-caught yel-
lowfin tuna in the central Pacific Ocean (Murphy and Shomura
1972).
relationship of LeGuen and Sakagawa (1973), we
estimated that beyond 140 cm
]nR = 6.74 - 1.96^ (4)
where R = ratio of females to males
t = age in years.
One interpretation of the above result (assum-
ing that males have a coefficient of instantaneous
natural mortality of 0.8 on an annual basis as do
all fish <145 cm) is that female yellowfin tuna
>140 cm have a coefficient of apparent natural
mortality of 2.76. Assuming that the results of
Murphy and Shomura apply to the Atlantic and
that all yellowfin tuna are equally available to
both gears, we separated the catch of yellowfin
tuna into males and females using Equation (4)
and Table 2, and estimated F for the males using
Input F values of 0.2 and 0.8 for fish >177.5 cm
(Lenarz et al. 1974). An alternative method would
be to use the same Input F for the three hypotheses
at the smallest size interval. This was attempted
and resulted in either estimates of F, which, based
on the results of other studies, appeared to be too
low under the 1:1 hypothesis or too high under
the other hypotheses. The estimates of size-
specific F are similar except for very large yel-
lowfin tuna (Figure 7). Since the deviations in sex
ratio from 1:1 occurs only at large sizes, we used
both sets of estimates of F.
o
35 55
75 95 115
SIZE (cm)
T — r
135 155 175
HIGH M
Female
BEH Female
1 — f~— I — r
SIZE (cm)
Figure 7. — Estimates of size and sex specific coefficient of in-
stEintEineous fishing mortahty on annual basis (F) for Atlantic
yellowfin tuna for 1:1, BEH and HIGH M hypotheses (see text):
(a) low Input F, B) high Input F.
For females, three hypotheses were examined
for estimating F: 1 ) the observed differences in sex
ratios are artifacts, and consequently females
have the same values of F and M as males (de-
noted 1:1); 2) females >140 cm have a higher
natural mortality rate than males but are
exploited at the same rate as males for all sizes
(denoted as HIGH M); and 3) females have the
same natural mortality rate as males but become
less subject to fishing mortality beyond 140 cm
(denoted as BEH for behavior changes). Under the
BEH hypothesis, F, for females >140 cm is equal
to the ratio of the catch of females to the catch of
males times F, estimated for males. The alterna-
tive hypotheses considerably affected the esti-
mates of size-specific F (Figure 7).
In the following analyses, we found that the
BEH and HIGH M hypotheses produce similar
results. To save space, we refer to only the one
hypothesis that produced results which showed
the greatest difference from the 1:1 hypothesis.
Also, when not specifically indicated, size of re-
cruitment and effort are assumed to be those at the
time of the study, i.e., 1967-71 where the multi-
plier of effort is equal to unity.
Estimates of yield per recruit as a function of
fishing effort are shown in Figure 8. The choice of
Input F has little effect on the relative values of
814
LENARZ and ZWEIFEL: INTERACTION BETWEEN LONGLINE AND SURFACE FISHERIES
ENTIRE FISHERY
ENTIRE FISHERY
HIGH M
3
o
UJ
q:
Q
-BEH
-BEH
-1 — I — r
4 8 12
~ 1 I I I I I I r
16 20 24 28 32
1
3 6
MULTIPLIER OF EFFORT
Figure 8. — Estimates of yield per recruit of Atlantic yellowfin
tuna at size of recruitment at time of the study as a function of
fishing effort and sex hypothesis: (a) high Input F, (b) low In-
put F.
yield per recruit. Yield per recruit is closer to the
maximum under high Input F than low Input F.
The curves are considerably more dome-shaped
when a 1:1 sex ratio is assumed than under the
other two hypotheses. Under high Input Fand the
1:1 hypothesis only a 3^c increase in yield per
recruit could be obtained by increasing fishing
effort. Under the BEH hypothesis, a 20% increase
in yield per recruit could be obtained by doubling
the effort.
Estimates of yield per recruit as a function of
size at recruitment are shown in Figure 9. Again
the choice of Input F has little effect on the relative
values of yield per recruit. A slightly greater de-
pendence of yield per recruit on minimum size is
obtained when the high Input F is used. Under
high Input F, and the 1:1 hypothesis a 107c in-
crease in yield per recruit could be achieved by
increasing size at recruitment. Under the BEH
hypothesis, only a d'^c increase would occur.
Eumetric fishing occurs when size at recruitment
is raised from the current 32.5 to 82.5 cm under the
1:1 hypothesis and 72.5 cm under the BEH
hypothesis.
Estimates of yield per recruit as a function of
fishing effort were also calculated for each gear
H M
1 — I — 1 — \ r
J2 5 52 5 72 5 92 5 112 5 132 5
32 5 52 5 72 5 92 5 112 5 132 5
MINIMUM SIZE (cm)
Figure 9. — Estimates of yield per recruit of Atlantic yellowfin
tuna at level of fishing effort at the time of the study for 1:1, BEH
and HIGH M hypotheses as a function of size at recruitment: (a)
high Input F, (b) low Input F.
(Figure 10). The results show that the curves are
more dome-shaped for the longline fishery than for
the surface fishery under all three hypotheses.
Furthermore, the longline fishery is more sensi-
tive to fishing effort under the 1:1 hypothesis than
under the other two. The curves for the surface
fishery are dome shaped under the 1:1 hypothesis,
but appear to approach an asymptote under the
other two.
We also estimated yield per recruit for each gear
when the other gear is not exploiting the stock
(Figure 11). A comparison of Figures 10 and 11
reveals that yield per recruit to the longline
fishery would increase by about 115*^ if surface
fishing were eliminated under high Input F and
the 1:1 hypothesis and 76*?^ under high Input F
and the BEH hypothesis. Yield per recruit to the
surface fishery would increase by about 30% if the
longline fishery were eliminated under high Input
F and the 1:1 hypothesis and 229c under the BEH
hypothesis. Thus, the nature of age-specific sex
ratio has a greater effect on that of the longline
fishery than on the relative success of the surface
fishery. The curves for a longline fishery in the
presence of a surface fishery are dome-shaped
(Figure 10), while the curves in the absence of a
surface fishery are not (Figure 11). This again
points out the importance of not treating the two
fisheries as separate entities unless it is shown
that they exploit separate stocks.
Stock fecundity (egg production per recruit)
relative to an unfished stock was estimated as a
function of fishing effort. Stock fecundity was con-
siderably affected by the choice of fecundity index
815
FISHERY BULLETIN: VOL. 76, NO. 4
4i-
OC_l I I I I I I I I I I I — I — I — I 1 1 — I
0 04 08 12 16 20 24 2 8 3 2 36
MULTIPLIER OF EFFORT
Figure 10. — Estimates of yield per recruit of Atlantic yellowfin
tuna when both gear fish at size of recruitment at the time of the
study as a function of sex ratio hypothesis, fishing effort, and
gear: (a) surface gear with high Input F, (b) longHne gear with
high Input F, (c) surface gear with low Input F, and (d) longline
gear with low Input F.
and sex ratio hypothesis but only slightly affected
by the choice of Input F (Figure 12). At the level of
fishing effort at the time of the study under high
Input F and 1:1 hypotheses, the relative fecundity
is 0.28 when the fecundity index I is used and 0.39
when fecundity index II is used. Under the HIGH
M hypothesis, relative fecundity is 0.55 when
fecundity index I is used and 0.61 when fecundity
index II is used. Thus, at the level of fishing effort
at the time of the study, the choice of fecundity has
a 10 to 30% effect on estimates of relative fecun-
dity, while the choice of sex ratio hypothesis has a
30 to 50% effect. The two choices, fecundity index
and sex ratio hypothesis, also have considerable
effect on relative fecundity when plotted as a func-
tion of size at recruitment (Figure 13).
The relationship between stock fecundity and
recruitment has not been demonstrated for any
tuna. As shown above, one of the difficulties in
demonstrating such a relationship is obtaining a
reasonably accurate estimate of stock fecundity.
Even if stock fecundity could be accurately deter-
mined, the recruitment process is likely to be so
complex that much more research would be re-
quired before a reliable predictor of recruitment
could be developed.
It is interesting to note that similar estimates of
yield per recruit and relative fecundity are ob-
tained under the HIGH M and BEH hypotheses.
Thus it appears that research should be directed
toward determining whether or not the 1:1
hypothesis or one of the other two are valid rather
than distinguishing between the HIGH M and
BEH hypotheses. This research should be a fairly
simple matter. The choice of fecundity index is
also of significance for estimating relative fecun-
dity. The difference between the two indices is
caused mainly by different maturity schedules
(Hayasi et al. 1972). The surface-caught fish ap-
peared to mature at an earlier age than longline-
caught fish, and could be an artifact related to the
phenomenon noted by Hisada ( 1973); i.e., mature
fish tend to prefer warm water. It should also be a
fairly simple matter to determine the cause of the
difference between the two indices.
SIMULATION MODEL OF PATTERNS
OF DISPERSAL AND RECRUITMENT
OF YELLOWFIN TUNA
Factors that could cause groups of tuna to not be
available to all components of a fishery include
nonrandom movements, random movements but
nonrandom distribution of fishing gear or effort,
and recruitment that is nonrandom in a geo-
graphical sense.
Extensive tagging experiments have not pro-
duced any clear-cut evidence of a definite migra-
tion pattern for yellowfin tuna in the eastern
Pacific. Bayliff and Rothschild (1974) recently
found evidence for both random dispersal and di-
rected movements. They were not able to remove
the effects on their data of lack of fishing effort in
some time-area strata and of the coastal boundary.
The evidence for directed movements indicated
that such movements were generally parallel to
the coast, suggesting that the presence of the coast
influenced their results. Fink and Bayliff (1970),
in a synthesis of extensive tagging data, proposed
that recruitment to the nearshore surface fishery
is not random in a geographical sense, but tends to
take place off Mexico and in the Panama Bight.
816
LENARZ and ZWEIFEL: INTERACTION BETWEEN LONGLINE AND SURFACE FISHERIES
5-
0
1 1
4-
BEH
3-
/y/^
2-
1/
1—
1 -
0-
"T r r r 1
1 — 1 — 1 — 1 — 1 — 1 — 1 — r —
-I — I — I — 1 — \
32 36
MULTIPLIER OF EFFORT
Figure ll. — Estimates of yield per recruit at size at recruitment at the time of the study as a function of fishing effort, sex ratio
h>'pothesis, and fishing gear when only one gear is fishing: (a) high InputF and surface gear, (b) high InputF and longline gear, (c) low
Input F and surface gear, and (d) low Input F and longline gear.
3
<
HIGH M
MULTIPLIER OF EFFORT
Figure 12. — Estimates of relative stock fecundity at size at recruitment at the time of the study as a function of fishing effort, fecundity
index, and sex ratio hypothesis: (a) high Input F and fecundity index I, (b) high Input F and fecundity Index II, (c) low Input F and
fecxmdity index I, and (d) low input F and fecundity index II.
With the above results in mind, we developed a
computer simulation model to examine the inter-
relationships of: 1 ) patterns of movement offish; 2 )
patterns of recruitment (i.e. by area), and 3)
fishing strategy for two gear types (surface and
longline) fishing alone or together on the same
population.
The model is general in that it allows the user to
817
FISHERY BULLETIN: VOL 76, NO. 4
325 525 725 92 5 112 5 132 5 32 5 62 5 72 5 92 5 1125 1325
SIZE AT RECRUITMENT (cm)
Figure 13. — Estimates of relative stock fecundity at level of
fishing effort at the time of study as a function of size at recruit-
ment, fecundity index, and sex ratio hypothesis: (a) high Input F
and fecundity index I, (b) high Input F and fecundity index II, (c)
low Input F and fecundity index I, and (d) low Input F and
fecundity index II.
specify the nature of movements, locations of re-
cruitment, parameters of growth, and natural
fishing mortality.
We crudely represented the eastern Pacific
Ocean with the grid of 5° square areas shown in
Figure 14. The number offish of a specific age in
each cell at time t is given by the vector
N, = AS,N,_,
(5)
where A^, (112 x 1) has elements (n,), equal to the
numberof fishincelU attimef, S;(112 x 112) is a
diagonal matrix with elements (s,,); equal to the
survival rate offish in cell ;' from time ^ - 1 to time
t, A (112 X 112) is a probability transfer matrix
with elements (Oy) equal to the probability of a fish
incellj moving to cell J, and where A'^o^ 112 x l)has
elements {n,)^ equal to the number of recruits in
cell i. Five consecutive year classes are in the
system at a time.
For our work we specified A, the transfer ma-
trix, by the assumption that for any cell the prob-
abilities of fish remaining stationary and moving
to each of eight adjacent cells is the same, i.e., 1/9.
Any other transfer has zero probability. This gen-
eral rule is modified as follows:
1) Probabilities of remaining stationary in cells
adjacent to the shore are augmented by the sum of
probabilities of those movements which would
otherwise put fish on land and the probability of
occurrence on land is zero.
818
1
2
3
4
5
COLUMN
6 7 8
9
10
//
12
13
14
/
1
2
3
4
5
6
7
8
9
10
II
12
13
14)
2
15
16
17
18
19
20
21
22
23
24
25
26
" *' \\\\\
27 28;
^\~^~-
3
29
30
31
32
33
34
35
36
37
38
39
40
41
42;
s\S\\N
5 4
O
Q: 5
43
44
45
46
47
48
49
50
51
52
53
54
55 56
57
58
59
60
61
62
63
64
65
66
67
68
69
;70;
\\\\N
6
71
72
73
74
75
76
77
78
79
80
81
82
83
84
7
85
86
87
88
89
90
91
92
93
94
95
96
97
^98^
8
99
100
101
102
103
104
105
106
107
108
109
no
III
Figure 14. — Representation of eastern Pacific Ocean. Each cell
represents a 5° square area. Hatched cells represent land. Col-
umn 1 is western boundary and Column 14 is eastern boundary.
Row 1 is northern boundary and row 8 is southern boundary.
2) Probabilities projecting beyond the northern
and southern edges are similarly absorbed on the
boundaries.
3) In cells of rows 2 and 7, probabilities of mov-
ing toward rows 1 and 8 are decreased by half with
the probability of remaining stationary increased
by a like amount. This is an attempt to simulate a
stock encountering increasingly marginal condi-
tions as the northern and southern boundaries are
approached.
4) Probabilities of remaining stationary on the
western edge are augmented by the probability of
returning from beyond the boundary in a single
time interval. The remainder of the fish that move
beyond the western boundary are lost to the sys-
tem.
The speed of dispersion is controlled both by A and
the time interval. The time interval was 3 mo for
this study. The combination of A as defined and
time interval of 3 mo allows a fish to travel a
maximum of 1,200 mi in a year. Only 1 out of 820
surviving fish that begin the year in the center of
the grid travel 1,200 mi in a year. These relatively
slow random movements seemed reasonable,
based on the results shown in Bayliff and
Rothschild (1974) and recent results of lATTC
tagging studies (Inter- American Tropical Tuna
Commission^).
Two alternative recruitment models were
examined. For the first, denoted as inshore re-
^Cited with permission of M. Clifford Peterson, Acting Direc-
tor of the Inter- American Tropical Tuna Commission. From the
Inter-American Tropical Tuna Commission Bi-monthly Report,
March- April 1976:8-13.
LENARZ and ZWEIFEL: INTERACTION BETWEEN LONGLINE AND SURFACE FISHERIES
cruitment, recruits are divided equally among the
five cells 51, 52, 69, 83, and 84, which resemble the
recruitment areas proposed by Fink and Bayliff
(1970). For the other alternative, denoted as uni-
form recruitment, recruits are divided equally
among all cells except those on the boundaries or
on land. Total annual recruitment is 100 fish. We
assumed 1 ) that fish are 1 yr old when recruited, 2 )
growth proceeds according to the von Bertalanffy
curve of LeGuen and Sakagawa (1973), and 3) the
coefficient of instantaneous natural mortality is
0.8 on annual basis and is independent of time and
location. Fish >6 yr old (175 cm) were removed
from the system. Consequently, under constant
conditions the fishery reaches equilibrium in 5 yr.
The system was always run for 5 yr before an
experiment was begun.
We first examined the effects of sampling loca-
tion, dispersal, and location of recruitment on age
distribution and the resulting apparent rate of
natural mortality obtained from unbiased sam-
ples from an unfished population. Mortality was
estimated with the standard linear regression
model {\n Nt = In Nq - Mt) from the age distribu-
tion offish in each cell. It is assumed that mortal-
ity is constant after full recruitment, and that the
modal age represents first age of full recruitment.
The results reveal that M is usually overestimated
as would be expected when fish emmigrate from a
sampled area (Figure 15). Estimates of M tend to
be relatively high near areas of spawning with
inshore recruitment. In the case of uniform re-
cruitment, estimates of M tend to be highest on the
western boundary where fish are lost to the sys-
tem. Modal age tends to increase in a westerly
direction for inshore recruitment and stay rela-
tively constant for uniform recruitment (Figure
15). The modal size of actual catches of surface-
caught yellowfin tuna in the eastern Pacific in-
creases in a westerly direction (Figure 16). Al-
though the surface fishery probably does not take
an unbiased sample of the size distribution of the
population, the data are suggestive of reduced re-
cruitment in the western areas.
We simulated a 20-yr hypothetical yellowfin
tuna fishery to examine interactions among a
longline fishery, inshore surface fishery, ocean-
wide surface fishery, and ocean-wide surface
fishery that does not heavily exploit young fish as
follows:
0--0- -0--0--
. MODAL AGE
ROW
ROW
COLUMN
COLUMN
Figure 15. — Estimates of coefficient of instantaneous natural mortality on annual basis ( M) and modal age
of yellowfin tuna by row and column: (A) inshore recruitment, and (B) uniform recruitment.
819
FISHERY BULLETIN: VOL. 76. NO. 4
S^
o J
■'4
^ o
i£
< -
UJ f^
< ro
q: io i
K U>
< in /
5-OD
X 1
O "
^ * /
UJ *
z „
O
cfl
^
>-
OS
U
3
c
1
(u
0
+j
0
0
c
0)
c
0
*->
to
01
01
Tf
J3
C8
CO
B
E
2
CO
CO
(1)
u
ca
s-
^^
a<
01
^
nt-i
>H
C)
T3
OJ
C
n
0)
CO
.M
JS
.*->
«
0
u
c
3
4-^
c
CO
0
OJ
fTl
C
be
1
^
lo
0
-*->
m
n
01
>>
c
01
E
0
c
0
01
u
01
•E
-M
*J
m
r/l
,
r
ca
3
.«-!
0)
-Q
k.
C8
CS
■ — ■
0>
to
^
J=
■o
t^
>>
fT)
C
u
T—t
X
C
3)
r.
3
0
•s
1)
m
to
i
E
6
01
0
3
c
0)
c
3
0)
c
•—1
00
(Tl
u
0
c
Tl
H
ca
3
4^
c
(M
to
0
<
§
Ih
f-H
ID
<
CO
0
.2
CO
0.
E
0
CO
0
be
o>
N
0)
m
E
0
1
rn
0
H
0
01
1
y=
^
1
:*
3
T
CO
.— t
fc
01
w
e
0)
CO
&
Iw^
ffl
V
Uh
01
-a
820
LENARZ and ZWEIFEL: INTERACTION BETWEEN LONGLINE AND SURFACE FISHERIES
1) For the first 5 yr only longliners fished and
only in rows 5 to 8.
2) For the next 5 yr, this longline fishery was
augmented with surface gear in all cells adjacent
to the coast.
3) Next, exploitation by the surface gear was
expanded to include all cells for 5 yr.
4) Finally, for the last 5 yr, age specific surface
fishing mortality was reduced by 75% for fish <2.5
yr of age because much of the surface catch of
yellowfin tuna in offshore areas of the eastern
Pacific comes from schools associated with por-
poise. Typically, porpoise schools contain few yel-
lowfin tuna <2.5 yr of age (Calkins 1965).
Steps 1,2, and 3 resem.ble the sequence of events in
the eastern Atlantic fishery for yellowfin tuna.
Yellowfin tuna first were exploited in a significant
fashion by longliners in a 10° band along the
equator, then a nearshore surface fishery became
significant, and in recent years some exploitation
by surface gear in offshore areas has occurred. To
our knowledge, step 4 has not occurred in the At-
lantic. Age-specific fishing mortality rates similar
to those by surface gears estimated by Lenarz et al.
( 1974) for the Atlantic yellowfin tuna fishery were
used (Table 5). The Ricker yield equation was used
to calculate yield for each time-area stratum.
Total yields per recruit were calculated and are
shown in Figure 17. Yields per recruit are quite
similar for both recruitment models except near
shore, where yield per recruit was considerably
higher for the inshore recruitment model than for
the uniform recruitment model. The difference in
yield per recruit betweeti the two models decreases
slightly as time increases. Yield per recruit closely
approached equilibrium yield within 3 yr after a
change was made in the fishery. Total equilibrium
yield per recruit with an inshore surface fishery
Table 5. — Estimates of age-specific F on an annual basis used
as baseline for simulation. See text for modifications of mortality
rates during simulation.
Age
Surface gear with
(yr)
Longline gear
Surface gear
reduced F
1.0
0.00
0.30
0.08
1.5
0.00
0.30
0.08
2.0
0.05
0.22
0.06
2.5
0.15
0.20
0.20
3.0
0.25
0.18
0.18
3.5
0.35
0.30
0.30
4.0
0.45
0.35
0.35
4.5
0.40
0.42
0.42
5.0
0.40
0.27
0.27
5.5
0.20
0.20
0.20
6.0
0.05
0.15
0.15
6 00-
\
400-
J^
2 00-
:v_^:rH
1 1 1 1 1 1 1 1 1 1
Uniform RecruitmenI
Inshore Recruitment
04-
0
200-
FlGURE 17. — Yield per recruit of hypothetical yellowfin tuna
fishery: (a) total, (b) longliners in all areas, (c) surface gear in all
areas, (d) longliners in cells 71 and 85, (e) surface gear in cells 71
and 85, (f) longliners in cells 69, 84, and 97, and (g) surface gear
in cells 69, 84, and 97.
and longline fishery was about 179c higher than
with a longline fishery alone, 54% higher with a
uniform surface fishery than with only a longline
and inshore surface fishery, and increased by 9%
when F for small fish was reduced by 75% . Under
the assumption that the catchability coefficient is
independent of area, the surface fishery increased
its equilibrium yield per recruit about fourfold by
increasing its effort about 12-fold when it ex-
panded into offshore waters. The same action de-
creased yield per recruit to the longliners by about
55%r.
We next examined the potential yield per re-
cruit to longliners in rows 5, 6, 7, and 8 by starting
a longline fishery with the age-specific F vector
multiplied by the scalar 0.3 and then multiplying
by 1.3 each year afterward. Yield per recruit ap-
pears to approach an asymptote of about 6 kg for
inshore recruitment and 5 kg for uniform recruit-
ment (Figure 18). The reduction in catch per re-
cruit per effort by fishing is not significantly af-
fected by choice of recruitment model. Even
though catch per recruit per effort at high levels of
effort was only about 20% of that at the beginning
of exploitation, overfishing in a yield-per-recruit
sense did not occur. Average size of fish in the
catch was not significantly affected by the re-
cruitment model, and decreased from about 50 to
30 kg with increased fishing effort (Figure 18).
A simulation for an inshore surface fishery indi-
cated an asymptotic production curve with a
821
FISHERY BULLEITN: VOL. 76, NO. 4
c
2 3 4 5 6 7
MULTIPLIER OF EFFORT
2 3 4 5 6
MULTIPLIER OF EFFORT
Figure 18. — Yield per recruit, shield per recruit per effort, and average size of catch for hypothetical longline fishery: (a) total yield per
recruit, (b) total yield per recruit per effort, (c) yield per recruit per effort in cells 7 1 and 85, (d) yield per recruit per effort in cells 69, 84,
and 97, and (e) average size in all squares.
maximum yield per recruit of about 1.4 kg for
uniform recruitment and 2.2 kg for inshore re-
cruitment (Figure 19). Catch per recruit per effort
was reduced by about 75% under both alterna-
tives. The ratio of maximum yield per recruit for a
longline fishery to an inshore surface fishery was
about 2.7 for inshore recruitment and 3.4 for uni-
form recruitment. Average size offish in the catch
was about 2 kg higher for uniform recruitment
than for inshore recruitment and decreased from
16 or 18 kg to 8 or 11 kg with increased fashing
effort (Figure 19).
Simulation of a uniform surface fishery revealed
that choice of recruitment model had an insig-
nificant effect on yield per recruit, catch per re-
cruit per effort, and average size of catch, except
that catch per recruit per effort in the nearshore
area was relatively high for inshore recruitment
(Figure 20). A 75% reduction in F for fish <2.5 yr
old had considerable effect on the results. Maxi-
mum yield increased from about 5. 1 to 6.9 kg when
F was reduced. Both yield curves are dome-
shaped. Catch per recruit per effort became rela-
tively higher at high levels of effort when F was
reduced. As expected, average size was consider-
ably higher for reduced F.
With inshore recruitment, maximum yield per
recruit changes from about 2.2 kg for an inshore
fishery (Figure 19) to about 5.1 kg for a uniform
„ 4
t 3
3
o
K 2
>-
o
Z3
O
10
08
06
04
02
UNIFORM RECRUITMENT
INSHORE RECRUITMENT
c
rT
30
j£
•—^
UJ
M
20
-
CO
^i^ZT — —
UJ
in
^ '" — " °~-— m-
o
oc
X — "
LlJ
>
<
1 1 1 1 1 — 1 1 1 1
01 23456789
MULTIPLIER OF EFFORT
Figure 19. — Yield per recruit, yield per recruit per effort, and
average size of catch for hypothetical inshore surface fishery: (a)
yield per recruit, (b) yield per recruit per effort, and (c) average
size of fish in catch.
822
LENARZ and ZWEIFEL: INTERACTION BETWEEN LONGUNE AND SURFACE FISHERIES
0.5-
CE
O
^^ 0.4 H
5 0.3-
q:
o
UJ
. (2b)
The Gompertz function
S^ = S.exp[G(l-exp[-^(a-c.)])] (3a)
becomes
S^. = S.[expG][S,J(S.expG)]^''P[-^<^°>l (3b)
A linear function of size upon age
S = b + ka (4a)
a
is written
S = S +k{Aa) . (4b)
r rn
Definitions of symbols employed above are:
Sa = size at age a,
Sr = size at recapture,
S,„ = size when marked.
Si = size of the smallest animal in the data,
a, = age of the smallest animal in the data,
Table l. — Brown shrimp mark- recapture experiments, northern Gulf of Mexico.
Length range (mm)
Length range (mm)
Number
Release area
Release date
of
released shnmp
of recovered shrimp
recovered
Galveston Estuary
May 1967
66-175
71-124
13
50 mi east of
Galveston, Tex.
June 1967
83-147
86-178
301
60 mi southeast of
Freeport, Tex
Sept. 1967
122-181
124-196
40
Biloxi Bay, Miss.
May 1 968
90-122
90-181
4,218
40 mi southeast ot
Freeport, Tex.
Feb-Mar 1969
109-168
136-185
69
Galveston Estuary
June-July 1969
90-128
91-182
257
50 mi southeast of
Freeport, Tex.
Nov 1969
145-203
141-213
593
828
PARRACK: ASPECTS OF BROWN SHRIMP GROWTH
Sx= an equation parameter, the asymptotic
size,
b = an equation constant related to the size
at birth,
Aa = a^ - o„, = time at large,
Gr - age of an individual on the date recap-
tured,
a,„ = age on the date marked and released,
and G, g, and k are equation parameters.
Equation parameters S^, k, G, and g were esti-
mated by utilizing the Marquardt algorithm to
minimize the residual sum of squares:
y:iS'-s f
\ f J./
where n = the number of individuals marked
and recaptured,
S'r = the observed size at recapture, and
Sr = the size at recapture as estimated by
the growth equation.
The remaining equation constants, b in Equa-
tions (la) and (2a) and a, in Equation (3a), are
respectively computed:
6 = (sjs^)-i
b = is^-s^)is
a.
In
1-
In (S^ IS.)
g
(5a)
(5b)
(5c)
where S^ is the size at birth and other symbols are
as before. The parameter b in Equation (4a) is
simply the size at birth.
Studies of the early development of brown
shrimp indicate that newly hatched larvae are
0.35 mm total length (Cook and Murphy 1971).
Estimates of the equation constant b in the logistic
and von Bertalanffy models were based on that
length at birth.
Shrimp eggs are 0.26 mm in diameter (Cook and
Murphy 1971) and about the density of water
(Cook and Lindner 1970) so that the weight at
hatching is about 0.000009 g. Brown shrimp un-
dergo metamorphosis 11 to 15 days after hatching
(Cook and Lindner 1970) and are 0.0008 g at that
time (Wheeler 1969). The weight at birth was cal-
culated as the midpoint between that weight and
the egg weight. Calculations of b and a, in the
various models were based on that weight at birth.
RESULTS
Growth in Length
In anticipation that differences in growth be-
tween sexes may exist, equations were fit for
males and females separately. Estimated equa-
tion parameters (Table 2) are quite different be-
tween sexes. The fitted models indicate that
females are much larger than males of the same
age. The estimates of the growth coefficient k do
not differ greatly between sexes for both the lo-
gistic and the von Bertalanffy models; the 90%
probability support plane confidence intervals
(Conway et al. 1970) extensively overlap for both
models. The estimates of asymptotic length are,
however, greatly different and such confidence in-
tervals on those estimates are very disjoint. Pool-
ing all data together to estimate overall growth
functions for both sexes combined was therefore
judged unrealistic.
The relative abilities of the von Bertalanffy,
logistic, and linear models to correctly reflect
growth was judged by comparing residual sums of
squares (Table 3). The von Bertalanffy function
produced the smallest residual and the linear
model the largest. The residual sum of squares for
the linear model was well over three times that of
the von Bertalanffy and logistic models for both
males and females. The difference between the
two nonlinear models was much smaller; the re-
sidual of the logistic was but 8% larger than that of
Table 2. — Growth models for brown shrimp. Lengths (L) in millimeters, weights {W) in grams, and ages (a) in months.
Model
Males
Females
Logistic
L
= 162.8/(1 + 464.14296 ^0.56643)
L
= 187.5/(1 + 534.71436 "0-6^ ^^3)
von Bertalanffy
Linear
Gompertz
L
L
W
= 168.7(1 - 0.9979e"° 33573)
= 0.35 + 4 21813
= 5.07(exp( 1.9996(1 - exp[-0.3735(a
- 4.6688)])])
L
L
W
= 193.6(1 - 0.99826 -0-33633)
= 0.35 + 7.82093
= 3.55(exp[2.8359(2 - exp[-0.4410(3
- 3.2549)])])
Monomolecular
Linear
w
w
= 43.51(1 - 0.99996^°^^"*^^)
= 0.0004045 + 1.8018a
W
W
= 74.32(1 - 0.99996 ~°^'*^63)
= 0.0004054 + 3.9013
829
FISHERY BULLETIN: VOL 76. NO. 4
the von Bertalanffy in the case of males and 5%
larger for females.
Table 3. — Residual sums of squares for six brown shrimp
growth models.
Model
Males
Females
Length:
Von Berlalantfy
Logistic
44.161 65
47.661.96
155.797.40
163,278.00
Linear
162.661 15
599,677.13
Weight:
Monomolecular
5.548.57
33,930.69
Gompertz
Linear
7,027.72
12,335.42
38.751.26
67,526.07
Number of observations
1.536
3.588
The difference in growth between sexes and the
ability of the von Bertalanffy model to fit the ob-
servations is visible from plots of the observed
lengths about the growth models. Data points
were plotted by first calculating the age at release
from the fitted model, adding time at large to com-
pute the age at recapture, then plotting that age
and the recapture size. The plots (Figure lA, B)
show that sex specific growth does exist and that
the differences are of significant magnitude.
Further, the von Bertalanffy model does visibly fit
the observed data. Although the observed data do
o
C\J
CNJ
00
• ■-.^^^^■^-r^'^r'^ .■■••"
■. .J^i^r^r^ ;■•
^^
'j/^- - ". '
ID-
^^^-^^""^"^""^
./■M.-'. ■ .
-^
• •• ■/^:".-
■■ f '
CM
ro-
■■^ ■ ■
M-
S
■'.f': ■ ■ .
■■/■■•
-2-
^
m--
LENGTH
88 1
f-y A
p: B
/
/
/
/
C\J
CNJ"
o-
1- r 1 » — -r F 1 ? 1 >
7 10 12 14
AGE (MONTHS)
7 10 12 14
AGE (MONTHS)
Figure L— Brown shrimp growth models. A) von Bertalanffy growth model, males; B) von Bertalanffy growth model, females;
C) monomolecular model, males; D) monomolecular model, females.
830
PARRACK: ASPECTS OF BROWN SHRIMP GROWTH
not in general fall close to the modeled line, the
scatter is not severe.
Growth in Weight
The Marquardt algorithm (Marquardt 1963)
was employed to estimate parameters of weight-
length relations used to transform individual re-
lease and recapture lengths into weights so that
growth in weight could be modeled. Plots of the
estimated relations (Figure 2) indicate them to be
sex specific. Support plane confidence intervals
(Conway et al. 1970) on equation parameters (90%
probability) for males did not overlap those for
females further indicating that the functions dif-
fer between sexes. In addition the data were log-
ged to linearize the relation and covariance
analysis techniques applied to test for differences
between sexes. The probability that the linearized
functions are the same is small (P,. <0.00 1 ) further
indicating the sex specificity of these relations.
Further inspection of the plots shows the scatter of
observations to be restricted and that the models
effectively fit. These sex specific models were
therefore employed to transform the data.
The magnitude of residual sums of squares (Ta-
ble 3) indicates the monomolecular model is the
best predictor of weight at age and the linear
model the poorest. The residual term for the linear
fit is about twice as large as that for the
monomolecular model and about 1.8 times that of
the Gompertz for both sexes. The reduction in re-
siduals of the monomolecular model as compared
with the Gompertz was much smaller, 25*^ in the
case of males and 14% in the case of females.
As in the case of growth in length, estimated
growth parameters indicate that growth in weight
is sex dependent. Both the Gompertz and the
monomolecular model estimate females to be
much larger than males of the same age. Asymp-
totic weight (monomolecular model) is estimated
to be 75 g for females and 46 g for males; support
plane confidence intervals (90% probability) on
these estimates do not overlap. Estimates of the
parameter k in the monomolecular model appear
to be about the same for both sexes and in fact the
support plane confidence interval for males com-
pletely includes that interval for females.
The differences in growth between sexes and the
degree of fit of the monomolecular model is shown
in Figure 1. Although appreciable scatter is ap-
o
(NlT
o.
o
MALES
WEIGHT = 3.931587 x IQ-^ LENGTH 3. 152658
0 24 48 72 96 120 144 168 192
LENGTH (MM)
0 24 48 72 96 120 144 168 192 216 240
LENGTH (MM)
Figure 2. — Weight-length relationships for brown shrimp.
831
FISHERY BULLETIN: VOL. 76, NO. 4
parent, systematic departure of the observed
points from the model is not evident so that the
model does reflect the data.
CONCLUSIONS
The relative abilities of prediction of the differ-
ent models can be judged by comparison of their
residual sum of squares. The comparison strongly
suggests that the linear function was by far the
poorest model of brown shrimp growth both in
length and weight. Although the size-age relation
does appear linear for small young individuals,
the rate of increase in size decreases with age, a
phenomenon documented for many organisms
both terrestrial and aquatic. A nonlinear function
is therefore required to model brown shrimp
growth throughout their entire life span.
The residual sum of squares for the von Ber-
talanffy equation was smaller than the logistic
equation when modeling weight; however, these
differences were not large. It is therefore not com-
pletely evident that the von Bertalanffy equation
is vastly superior to the logistic and Gompertz in
the modeling of brown shrimp growth. The von
Bertalanffy equation did, however, constantly fit
these data best for both sexes in the modeling of
both length and weight. This study does therefore
show the von Bertalanffy model to be slightly
superior to the logistic and the monomolecular
model superior to the Gompertz for both sexes.
The difference in the size-age function between
sexes was found to be large. This phenomenon was
previously reported for brown shrimp in the
southern Gulf of Mexico (Chavez 1973) and
northwest Atlantic (McCoy 1972) and for many
other marine organisms. This study indicates that
male brown shrimp apparently grow to approxi-
mately only three-fifths the weight and five-sixths
the length of females; however, the coefficients of
growth, as indexed by /j in the monomolecular and
von Bertalanffy models, are roughly equivalent. It
is interesting to note that the rate of increase in
size tends to fall off at an earlier age for males than
for females (see Figure 1). Since, in general, a
decrease in that rate roughly conforms to the age
of maturity and sexual activity, it is not unreason-
able to assume that males mature at a younger age
than do females.
Comparison of growth functions derived herein
with those generated by other workers indicate
that brown shrimp growth in the northern Gulf of
Mexico is very different than that in the southern
gulf and in U.S. Atlantic coastal waters. Growth
functions derived from populations off Mexico
(Chavez 1973) demonstrated a faster and pro-
longed growth compared with growth observed in
this study. That trend was consistent for both
males and females. Studies off North Carolina
(McCoy 1972) showed growth in Atlantic waters
to be very rapid although a smaller asymptotic
size was realized. As before, that trend was the
same for both sexes. The kinds of data used and the
methods employed to fit the growrth models dif-
fered in all three studies; therefore, some dis-
agreement in results may be expected. The
magnitude of the differences observed, however,
indicated truly different rates of growth may well
exist in the three geographical locations. The
growth of wild populations of white shrimp,
Penaeus setiferus, a similar species, is correlated
with water temperature (Gaidry and White 1973)
in the shallow estuarine and nearshore areas they
inhabit throughout their entire life span. Since
the temperature of seasonally homothermic deep
offshore waters where brown shrimp spend their
adult life may be assumed to increase with de-
creasing latitude, the differences in growth be-
tween northwest Atlantic, northern gulf, and
southern gulf brown shrimp populations are likely
positively correlated with gross water tempera-
ture.
ACKNOWLEDGMENTS
The staff at the Southeast Fisheries Center Gal-
veston Laboratory, National Marine Fisheries
Service, NOAA, provided assistance in this study.
Susan Brunenmeister and Edward Klima con-
tributed helpful advice in the reviewing of this
manuscript. Patricia Phares and Scott Nichols
provided valuable advice as to applicable statisti-
cal procedures.
LITERATURE CITED
BERTALANFFY, L. VON.
1938. A quantitative theory of organic growth (Inquiries
on growth laws. II). Hum. Biol. 10:181-213.
ChAVEZ, E. a.
1973. A study on the growth rate of brown shrimp
(Penaeus aztecus aztecus Ives, 1891) from the coasts of
Veracruz and Tamaulipas, Mexico. Gulf Res. Rep.
4:278-299.
CLARK, S. H., D. A. EMILIANI, AND R. A. NEAL.
1974. Release and recovery data from brown and white
shrimp mark-recapture studies in the northern Gulf of
832
PARRACK: ASPECTS OF BROWN SHRIMP GROWTH
Mexico, May 1967 - November 1969, U.S. Dep. Com-
mer., NOAA, Natl. Mar. Fish. Serv., Data Rep. 85, 152 p.
CONWAV, G. R., N. R. Glass, and J. C. Wuxox.
1970. Fitting nonlinear models to biological data by Mar-
quardt's algorithm. Ecology 51:503-507.
COOK, H. L., AND M. J. Lindner.
1970. Synopsis of biological data on the brown shrimp
Penaeus aztecus aztecus Ives, 1891. FAO Fish. Rep.
57:1471-1497.
COOK, H. L., AND M. A, MURPHY.
1966. Rearing penaeid shrimp from eggs to postlar-
vae. Proc. 19th Annu. Conf Southeast Assoc. Game
Fish. Comm., p. 283-288.
1971. Early developmental stages of the brown shrimp,
Penaeus aztecus Ives, reared in the laboratory. U.S.
Fish. Wildl. Serv., Fish. Bull. 69:223-239.
Fabens, a. J.
1965. Properties and fitting of the von Bertalanffy growth
curve. Growth 29:265-289.
Fontaine, C. T., and R. A. Neal.
1971. Length- weight relations for three commercially im-
portant penaeid shrimp of the Gulf of Mexico. Trans.
Am. Fish. Soc. 100:584-586.
Gaidry, W. J., and C. J. White.
1973. Investigations of commercially important penaeid
shrimp in Louisiana estuaries. La. Wildl. Fish. Comm.
Tech. Bull. 8, 154 p.
George, M. J.
1962. Preliminary observations of the recruitment of post-
larvae and growth of juveniles of the brown shrimp
Penaeus aztecus Ives in Barataria Bay. La. Wildl. Fish.
Comm. 9 Bienn. Rep., p. 160-163.
GOMPERTZ, B.
1825. On the nature of the function expressive of the law of
human mortality, and on a new mode of determining the
value of Life Contingencies. Philos. Trans. R.Soc. Lond.,
Ser. B, Biol. Sci. 115(11:513-585.
Jacob, j. S.
1971. Observations on the distribution, growth, survival
and biomass of juvenile and subadult Pe^iaews aztecus in
southern Louisiana. M.S. Thesis, Louisiana State Univ.,
Baton Rouge, 68 p.
Knudsen, E. E., W. h. Herke, and J. M. MACKLER.
1977. The growth rate of marked juvenile brown shrimp,
Penaeus aztecus. in a semi-impounded Louisiana coastal
marsh. Gulf Caribb. Fish. Inst., Proc. 29th Annu. Sess.,
p. 144-159.
LOESCH, H.
1965. Distribution and growth of penaeid shrimp in
Mobile Bay, Alabama. Publ. Inst. Mar. Sci., Univ. Tex.
10:41-58.
McCoy, E. G.
1972. Dynamics of North Carolina commercial shrimp
populations. N.C. Dep. Nat. Econ. Res. Spec. Sci. Rep.
21, 53 p.
Marquardt, D. W.
1963. An algorithm for least-squares estimation of non-
linear parameters. Siam. J. Appl. Math. 11:431-441.
Medawar, p. B.
1945. Size, shape, and age. In W. E. Le Gros Clark and P.
B. Medawar (editors). Essays on growth and form pre-
sented to D'Arcy Wentworth Thompson, p. 157-187. Oxf.
Univ. Press, Lond.
Pearl, R., and L. J. reed.
1920. On the rate of growth of the population of the United
States since 1790 and its mathematical representa-
tion. Proc. Natl. Acad. Sci. U.S.A. 6:275-288.
RINGO, R. D.
1965. Dispersion and growth of young brown
shrimp. U.S. Fish Wildl. Serv., Circ. 230:68-70.
Rose, C. D., a. H. Harris, and B. Wilson.
1975. Extensive culture of penaeid shrimp in Louisiana
salt-marsh impoundments. Trans. Am. Fish. Soc.
104:296-307.
St. Amant, L. S., J. G. Broom, and T. B. Ford.
1966. Studies of the brown shrimp, Penaeus aztecus. in
Barataria Bay, Louisiana, 1962-1965. Gulf. Caribb.
Fish. Inst., Proc. 18th Annu. Sess., p. 1-17.
St. Amant, L. S., k. C. Corkum, and j. G. Broom.
1963. Studies on growth dynamics of the brown shrimp,
Penaeus aztecus. in Louisiana waters. Gulf Caribb. Fish.
Inst., Proc. 15th Annu. Sess., p. 14-26.
Silliman, R. p.
1967. Analog computer models of fish populations. U.S.
Fish Wildl. Serv., Fish. Bull. 66:31-46.
Welker, B. D., S. H. Clark, C. T. Fontaine, and R. C. Ben-
ton.
1975. A comparison of Petersen tags and biological stains
used with internal tags as marks for shrimp. Gulf Res.
Rep. 5(1): 1-5.
WENGERT, M. W.
1972. Dynamics ofthe brown shrimp, Penaeus oz/ecMs Ives
1891, in the estuarine area of Marsh Island, Louisiana in
1971. M.S. Thesis, Louisiana State Univ., Baton Rouge,
93 p.
Wheeler, R. S.
1969. Culture of penaeid shrimp in brackish-water ponds,
1966-67. Proc. 22d Annu. Conf Southeast Assoc. Game
Fish. Comm., p. 387-391.
833
FISHERY BULLETIN: VOL. 76, NO. 4
APPENDIX
The linear, logistic, and Gompertz functions were expressed in terms of size at release age, size at
recapture age, and change in age (time at large) following the rationale presented by Fabens ( 1965) for
the von Bertalanffy function (as follows).
Each individual was of some unknown age (a,„) upon the date marked (t,„) and released. Upon the
recapture date it^) the individual was of an unknown older age (a,) so that the difference between the
release and recapture date (AO is equivalent to the increase in age (Aa) of that individual:
/sa = /st = t — t = a —a . (Al)
r m r m
That equality can be substituted into the von Bertalanffy function when expressed in terms of the size at
recapture (S^) and the age at recapture. Therefore the von Bertalanffy equation
S^ = S^[l-6exp(-/2a^)] (A2)
becomes S = S^{1 — b exp{—ka ) exp[—k{Aa)]). (A3)
The equation, when expressed in terms of the size when marked (S,„ ) with rearrangement, is:
bexp{-ka ) = 1-(S /S ). (A4)
That expression is substituted into Equation (A3) to yield the required function:
S = S -{S -S )e-fe(^«). (A5)
r "o ^ CO ui '
That form can then be employed to estimate the equation parameters k andS from mark-recapture data.
The final parameter ib) can be calculated directly by first rearranging terms of the original function:
S = S {1-be-'''') (A6)
a °°
SO that b = [l-{SJSJ]le-'''' (A7)
where S,, is the size at age a. If the size at birth, i.e., at age 0, is known, then:
b = 1-iSJSJ (A8)
where S-^_ is estimated from Equation (A5) and the size at birth (S/, ) is derived from life history studies.
That same rationale was applied to the logistic function. The size of a recaptured individual is
expressed:
S^ = S^/[l + b exp{-ka^)] . (A9)
Since Ur = a + a,n substitution gives:
S^ - S^I{l + bexp[-k(Aa)]bexpi-ka^)]. (AlO)
Expressing the logistic equation in terms of the size marked and rearrangement of terms gives:
bexpi-kaj + [{S^-SJISJ. (All)
834
PARRACK: ASPECTS OF BROWN SHRIMP GROWTH
Substitution yields:
S = S /(I + ((e~''^'''^)(S -S )IS )]. (A12)
Since S,., S,„, and Aa were all directly observable from mark-recapture data, the logistic equation
parameters S^ and k may be estimated from the data set. The remaining equation constant was calculated
by rearrangement of terms:
b = {(SJSJ-l) le-"^"'> . (A13)
From life history studies the size at birth, Sf,, was determined. Since at birth age is zero (a = 0) the
expression can be written:
b = {SJSJ-1. (A14)
The Gompertz function was likewise expressed in terms of the mark-recapture data. From Equation
(3a), the size at the time of marking is:
S^ = S. exp[G(l-exp[-^(a^-G.)])] (A15)
and by substitution becomes
S^ = S. exp[G -G(exp[-^(a^^^ -c.)]) (exp[-^(Aa)])] . (A16)
Writing in terms of the size at recapture and rearrangement of terms gives:
exp(-G exp[^(a^ -a.)]) = S^ /(S. exp G). (A17)
Substitution yields the expression required to estimate the constants G and^ from the mark-recapture
data:
S^. = [S.exp(G)][S^/(S. exp(G))J^''P[ '^^"^^ (A18)
where S, was the smallest size observed in those data. The remaining equation constant, a,, was then
calculated by writing Equation (3a) in terms of the size at birth:
S^ = S. exp[G( l-exp[-g(a-a.)])]. (A19)
Since at birth age is zero (a = 0) the expression can be written as:
a. = ln(l-[ln(S^/S.)/G]/^) (A20)
where S , the size at birth, was determined from natural history studies. The linear function:
S = b + ka (A21)
a
requires a much simpler derivation. Expressed in terms of the size at recapture:
S = b + ka . (A22)
r r
Substitution gives:
S = b + k {t.a+a ) . (A23)
r ni
835
FISHERY BULLETIN: VOL. 76. NO. 4
The function expressed in terms of the size at release
can be rearranged to
^,n = ^ -^ f'^.n ^A24)
°m ^ ^^m ~^^/^ erated on a counter-current principle.
Alternation of hot and cold water entering the
experimental chamber between replicate runs
eliminated any potential rheotactic interference.
The inner experimental trough ' 1.75 m long x 5
cm diameten 0.8 mm wall thickness) contained
DEFINITIONS
The term "preferred temperature" has been
used in various contexts in the literature 'e.g..
Brett 1952: Javaid and Anderson 1967: McCauley
and Tait 1970: Tatyankin 1972; McCauley and
Read 1973 ». Much of the variation in the use of this
T0= /EW *
h ®6 ^
■ 7- V ^^~^
V
csOST VIEW WITH COVER
*'^ri_
®-^)LJ-^^
#
i-
^
Figure 2. — &nall experimental dbamber for temperature selection measurements in larval fishes: aj experimental chamber, b) water
jaiiet, c) air line, d) drains, ei seawater input line, f > freshwater input line, gi daylight-simulating fluorescent light, h; light diffuser, i;
viewing slits, jj thermistor probes, ki door on lightproof cabinet, 1> supports for water jacket, m; water flow control valves, n> nylon
screen on ends of e^)erimental chamber.
840
EHRLICH ET AL THERMAL BEHAVIORAL RESPONSES OF FISHES
term can be attributed to different species exhibit-
ing various behavioral patterns in a gradient
tank, just as they do in their natural habitats. This
makes it impossible to use only one procedure to
determine the preferred range for all species
under all conditions. We determined the preferred
range and final perferendum by evaluating, on a
case-by-case basis, the behavioral responses of a
species subjected to known conditions such as ac-
climation temperature, feeding patterns, and cap-
tivity environment.
Experiments with larvae lasted 5-6 h, but
juveniles and adults were tested for approxi-
mately 7-8 h. An "experiment," in this study, con-
sisted of a set of individual runs, with each "run"
being the observation of the position and water
temperature selected by each fish in the gradient
at a fixed point in time. We employed constant
time intervals between runs for any experiment: 5
min for larvae and 15 min for juveniles and adults.
Run selected temperatures were the primary data
source, and we calculated their mean, mode, and
variance prior to combining them with data from
other runs to determine the preferred tempera-
ture.
DATA ANALYSES
The frequency of occurrence of all experimental
temperatures was not uniform due to the shifting
of the gradient as well as having a variable
number of degree intervals per run and generally
fewer than the 21 compartments. This caused a
bias in the number offish observed at a particular
temperature when summed over an entire exper-
iment. To compensate for this, prior to calculation
of the mean and modal selected temperatures, we
adjusted the data by using the number offish per
total occurrence of a particular temperature in all
experimental compartments rather than the ac-
tual number of fish at each temperature.
We defined the "initial selected temperature" as
that chosen by the fish immediately following es-
tablishment of a gradient of lO'^C. The modal
selected temperature was determined from the
percent occurrence frequency distribution derived
from adjusted data. After an initial time of appar-
ent searching and testing of water conditions the
experimental animals selected a temperature or
range of temperatures at which they remained for
the duration of the experiment. We called this the
final selected temperature (or temperature range)
and determined it from plots of selected tempera-
ture against time. The mean selected temperature
was derived by methods presented in Appendix
Table 1.
Reynolds (1977) reported that skewness of
temperature preference frequencj' distributions
required a complete description of the distribu-
tion. We examined the following parameters to
delineate thermal behavioral responses, the ini-
tial, mean, modal, and final selected tempera-
tures, standard deviation about the mean,
coefficients of skew^ness and kurtosis, and
coefficient of dispersion. The first four parameters
defined the preferred temperature range. The
standard deviation about the mean selected temp-
erature quantified movement through a range of
temperatures and gave a measure of the degree of
eury- or stenothermal preference. We used
coefficients of skewness and kurtosis (Sokal and
Rohlf 1969) in testing for normality and then to
help define the shape of the temperature-specific
fish frequency of occurrence distribution and to
refine interpretation of behavioral types. The
coefficient of dispersion quantified the tendency of
a species to aggi-egate or school and gave the per-
centage of use of the experimental chamber by all
fish within one standard deviation of the run
selected temperature.
EXPERIMENTAL TECHNIQUES AND
BEHAVIORAL RESPONSES
Our experimental techniques and data in-
terpretation methods are useful for a wide variety
of behavioral tj-pes. There are three salient fea-
tures of this methodology- 1) the shifting and re-
versal of the temperature gradient to partition
position preference from thermal preference, 2)
the extended duration of the experimental period
and its relationship to the thermal histon,- of the
test organisms, and 3) the criteria for behavior
evaluation.
Shifting and Reversal of
Temperature Gradient
Hasler ( 1956) pointed out that fishes in experi-
mental gradients can position themselves accord-
ing to small deformities in the tank structure. We
employed two methods to eliminate this factor:
shifting the position of a given isotherm in the
gradient during an experiment, and reversing the
hot and cold ends between replicate experiments.
841
FISHERY BULLETIN: VOL. 76, NO. 4
Shifting the isotherm position during an exper-
iment required the fish to thermoregulate ac-
tively, similar to those studied by Beitinger ( 1976,
1977) in his temporal gradient. This technique
demonstrated that the fish could follow an
isotherm and did not necessarily arbitrarily select
a position in the experimental tank. The precision
with which a group of fish followed an isotherm
varied between species and was related to the size
of their preferred temperature range. Juvenile
surfperch, Danialichthys vacca, for example,
which preferred a narrow range of temperatures.
closely followed an isotherm ( approximately 1 1°C)
(Figure 3). In contrast, juvenile topsmelt, Atheri-
nops affinis, after initially selecting approximate-
ly 22°C, remained within that compartment, and
shifting the gradient did not cause them to move
until the temperature reached 26°-27°C. This
isotherm was then tracked (Figure 3). Topsmelt
are physiologically eurythermal, at least during
embryonic stages, and the upper limit for hatching
of topsmelt eggs is 26.8°C (Hubbs 1965). Brett
(1956) suggested that the preferred temperature
may not be a strong enough directing force to move
I I I
1 I I I I T
I • r
180 240
TIME (minutes)
Figure 3. — Changes in fish and isotherm position in the experimental gradient. Juvenile Damalichthys vacca followed the
10°- 1 1°C isotherms. Juvenile Atherinops affinis remained in the position they initially selected and moved very little until the
temperature reached 26°-27°C. They then followed the temperature range of 25°-27°C. The small numbers indicate isotherms.
Large dots indicate the mean temperature selected by nine individuals.
842
EHRLICH ET AL: THERMAL BEHAVIORAL RESPONSES OF FISHES
fish with wide temperature tolerances from a par-
ticular area until stress-inducing conditions are
reached.
Temperatures Selected and
Their Relationship to Thermal History
We classified the behavioral responses of the 16
species surveyed into four groups based on
changes in temperatures selected throughout an
experiment: 1) immediate response — no general
shift in selected temperature over time, 2) fast
response — a shift in selected temperature not ex-
ceeding the first 2 h of the run, 3 ) slow response — a
shift in selected temperature over more than 2 h,
and 4) positioned response — a broad preference
and a tendency to remain in a given position in the
gradient until conditions become extreme.
Shiner surfperch, Cymatogaster aggregata; pile
surfperch, Damalichthys vacca; black surfperch,
Embiotoca jacksoni; and black croaker, Cheilo-
trema saturnum, showed the first behavioral pat-
tern of immediate response (Figure 4). These
fishes moved most quickly from their preexperi-
mental acclimation temperature to their final
selected one or range, remaining there for the du-
ration of the experiment. These fishes generally
had the narrowest selected temperature ranges
and also aggregated tightly (Table 1).
Fishes with a fast response to the temperature
gi'adients included speckled sanddab, Citharich-
thys stigmaeus; seiiorita, Oxyjulis californica;
spotted sand bass, Paralabrax maculatofasciatus;
25
20
— 15
Ol5
o
II). I
• ♦
* ♦ 4 • ♦
Cymatogaster aggregota
,.l..()t,'*l
I • ' • I ' • ' I
I • « • I I I • I
UJ
|_20H
<
cr I5H
.♦♦*♦♦♦♦•♦♦.♦♦♦ 4 ♦♦ ♦
Damalichthys vocco
T
4 4
I I I
UJ
• + 4
Embiotoca jacksoni
I • • « I
1 M
• ♦ . .
I I
T
T
25
20
4 * ♦ . 4 ' N ♦ *
4 4
..t
Cheilotrema saturnum
• I ' ' ' I ' ' • I
•••'i.'t'*
'"'"
' 1 ' ■
• 1 ' ' ' 1 ' ' ' 1 ' ' ' 1 ' '
' ' 1 '
•-n
60
120 ISO 240 300
TIME (minutes)
360
420
Figure 4. — Immediate response to temperature change. These
species showed no trend in selected temperature with time. Dots
are mean selected temperatures and vertical lines are 1 SD about
the means. Results are for duration of one experiment.
and sculpin Scorpaena guttata (Figure 5). These
species required up to 2 h to home in on a selected
temperature and also generally did not aggregate
as tightly, nor select as narrow a temperature
range as those fishes that showed an immediate
response to the temperature gradients (Table 1).
All larvae studied responded slowly under ex-
perimental conditions. These included topsmelt,
Atherinops affinis; California grunion, Leuresthes
tenuis; rockpool blenny, Hypsoblennius gilberti;
and painted greenling, Oxylebius pictus (Figure
6). Four older fishes also showed this behavior:
kelp bass, Paralabrax clathratus; olive rockfish,
Sebastes serranoides; California halibut. Para-
TABLE 1. — Behavioral responses of larval and juvenile fishes in temperature selection experiments. Initial and final selected
temperatures are taken from Figures 4-6 and other similar experimental data.
Mean
Mean
Coeffi-
Coeffi-
Coeffi-
Standard
Acclimation
Selected temperatures ( = C)
cient of
cient of
cient of
Experimental
date
No of
length
(mm)
temperature
skewness
kurtosis
disper-
Species
fish
Mean
SD
Mode
Initial
Final
(g.)
(92)
sion (%)
Immediate response:
Cymatogaster aggregata
12 June 1975
8
109
18.2
19.9
2.1
19
18.7
21
0.22
2.87
24
Damalichthys vacca
6 June 1975
9
69
18.1
10.5
0.9
10
11.4
11
0.41
4.6r
12
Embiotoca jacksoni
13 Dec 1974
7
118
16.7
18.0
1.6
18
17.0
18
-0.61-
4.54-
5
Cheilotrema saturnum
6 Oct. 1975
7
42
17.0
27.6
2.0
28
26.6
28
-0.79-
3.12
13
Fast response:
Citharichthys stigmaeus
22 Dec, 1975
9
90
18.9
10.1
2.6
10
14.8
10
0.43
3.44
33
Oxyjulis californica
28 May 1975
6
120
17.2
15.5
1.9
15
15.0
16
-069-
5.44-
11
Paralabrax maculatofasciatus
31 July 1975
6
179
20-6
24.2
3.1
27
21 2
25
-0.83-
2.70
29
Scorpaena guttata
24 Nov. 1975
6
64
17.6
17.5
4.2
19
17.2
17
-0.63-
4.01"
35
Slow response:
Atherinops affinis (larvae)
31 July 1975
6
14.5
21.5
252
2.9
27
21.9
27
-1 07-
4.12-
26
Leuresthes tenuis (larvae)
9 May 1975
6
8.1
16-5
25.2
4-1
26
19.2
27
-0.12
2.86
37
Hypsoblennius gilberti (larvae)
2 July 1975
6
4.4
19-4
222
3-1
19
19.7
26
0.39
2.10
36
Oxylebius pictus (larvae)
14 May 1975
6
3.4
16.0
26.8
33
27
19.7
29
082-
3.24
30
Paralabrax clathratus
28 July 1975
6
196
21.0
13.5
3.1
14
17.2
15
0.13
2.70
47
Sebastes serranoides
11 Dec 1974
8
82
17.0
180
1.3
18
16.2
17
-0.21
2.44
4
Paralichthys californicus
15 Oct. 1975
5
94
20.5
20.8
6.6
24
20.3
22
-0.10
1.91-
65
Pleuronichthys coenosus
3 Dec. 1975
4
134
10.0
7.5
2.5
7
10.8
7
1.40"
5.15-
23
Positioned response:
Atherinops affinis
14 Jan. 1976
9
60
15-0
23.3
32
26
16.4
26
-0.71-
2.41
4
•P<0.05.
843
15
10'
\\
Cithorichthys stiqmaeus
1 1 " ■' I M I H i f * * M * ♦ M H H I
5*11
20i
I I I — I I I I I I • — r-
3
• ihl.
O 10 I
• • ♦ 4 •
f ♦
I I I I I I I I I I
Oxyjulis colifornica
T
T
T
tt: 15
UJ 35
a.
2 30
UJ
t-25
20
15
HI
W
lii^
Porolabrox maculatofasciotus
I I I I I I I I
T
T
T
Scorpoena guttata
l|tiMlf|||)
10 ' I I I I I I I I I I I I I I I I 1 I I — I I • « I I • • I
60 120 180 240 300 360 420
TIME (minutes)
Figure 5. — Fast response to temperature change. These species
changed their selected temperatures over the first 2 h of the
experiment only- Symbols as in Figure 4.
lichthys californicus; and C-0 turhot, Pleuronich-
thys coenosus (Figure 6). Members of this group
required more time to stabilize their response
than either the immediate or fast responders. The
temperature selection acuity and aggregating
tendencies of these fishes were similar to those of
the second group (Table 1).
Juvenile topsmelt were the only species ob-
served that showed a positioned response (Figure
3). Ehrlich et al. (in press) discussed this behavior
in detail. California grunion are closely related to
topsmelt, and we have observed them together, in
the field, throughout larval, juvenile, and adult
stages. Possibly juvenile California grunion,
which were not tested, may show similar re-
sponses.
Preexperimental acclimation temperatures
showed the greatest effect on thermal selection of
the fishes during the first 2 h after establishment
of the gradient ( Figures 4-6). The short duration of
the influence of thermal history on temperature
selection has also been reported by Doudoroff
(1938). Clearly, trying to determine a preferred
temperature for these species or others with simi-
lar responses, during the transition period, would
make data interpretation difficult. After this ini-
tial period, the fishes, in most cases, chose a final
selected temperature, which may be synonymous
with what Fry (1947) termed the "final preferen-
dum." He defined this as the temperature range
30
25-
20 •
35 1
30
25
20
FISHERY BULLETIN: VOL. 76, NO. 4
Atherinops offinis larvae
.. 1^ M p* t f I M JM* *H
15
25'
20
O 30'
o
""' 25
LU
(T 20
Z)
I
Hilt
(■■It'ltMil
Leurestl\as tenuis larvae
' ' ' I
.,|nHtt).|||t-*t*'t't'
Hypsoblennius gilbert i larvae
T
T
T
.1
tllH
ttt)tHtlM(tt*l
Oxylebius pictus larvae
< 20-j
UJ 15-
a.
Ill iiiiiii|
I • ' ' I
M|,j|MtH|llt(.Mli.n*'
Poralabrax ctothrotus
20'
15 ■
10'
35'
30'
25'
20'
15'
I I I i I I I I I I I I I I I I I I I — I I I I
♦ . •
♦ ♦ ^ M • ♦
Sebastes serranoides
-r-^
\.^
Paralichttiys californicus
n ,1 4 t "
10* I I I I I I I I I I I I I I I I — I I I I ■ — I — I I I I I
10
Pleuronictithys coenosus
*''l]IlT'*»»**+***»»***»*
1 I I I I I I I I I I I I I I I — I I I I I — I I I I I I I
60 120 180 240 300 360 4 20
TIME (minutes)
Figure 6. — Slow response to temperature change. These fishes
changed their selected temperature over more than 2 h in the
experiments. Symbols as in Figure 4.
that the fish would eventually select, independent
of their acclimation temperatures. Topsmelt,
however, did not show this pattern, for their initial
selected temperature gave a good indication of
their preference and was independent of their ac-
climation temperature (Ehrlich et al. in press).
Doudoroff (1938) also found that fishes did not
select the temperatures to which they had been
acclimated but rather selected a common range of
temperatures, which he suggested must have
some physiological significance.
Figures 4-6 show that the final preferendum
was reached within several hours after the estab-
lishment of the gradient. This is considerably less
time than the approximately 24 h reported by
844
EHRLICH ET AL: THERMAL BEHAVIORAL RESPONSES OF FISHES
Reynolds and Thomson (1974) or Reynolds and
Casterlin (1976). The differential, however, be-
tween the acclimation and the final preferendum
must be considered. Reynolds and Thomson ( 1974 )
tested fish acclimated 17°C below their final pref-
erendum. Crawshaw (1975) used a range of ac-
climation temperatures from 22°C below to 3°C
above the final preferendum and found that as the
temperature differential diminished so did the
time required to reach the final preferendum. Dif-
ferentials of 5°C required as little as 1 h and 3°C
only 0.5 h (Crawshaw 1975). Based on the temper-
ature differences between acclimation and the
final preferendum (Table 1), Reynolds' and our
results generally fit the pattern described by
Crawshaw.
Behavioral Criteria
Most studies pertaining to behavioral responses
of fishes to thermal gradients have been concerned
with only one factor: the final preferendum. Addi-
tional information, however, can be obtained from
examination of parameters associated with the
frequency distribution of the selected tempera-
tures, particularly: skewness (degree of distortion
from symmetry) and kurtosis (peakedness). Ivlev
and Leizerovich (1960) compared the percent of
the area under the curve of number of fish per
temperature against the mode of the distribution
as well as the percentage of the curve on either
side of the mode. Reynolds and Casterlin (1976)
and Reynolds (1977) discussed the relationship
between various measures of central tendency
(mean, mode, and median) and skewness. They
also improved descriptions of thermal behavioral
responses by quantifying skewness but did not
state the statistical significance of the skewness.
Sokal and Rohlf (1969) stated that the absolute
value of coefficients of skewness and kurtosis have
little meaning and that they must be tested for
statistical significance. We identified distinct be-
havioral types with respect to the frequency dis-
tribution of selected temperatures by examining
skewness and kurtosis. The responses were, in
part, species-specific but also varied with on-
togenetic stage and nutritive condition. Reynolds
and Casterlin (1976) showed that skewness also
varied diurnally. Kurtosis can be used to assess
whether the test organisms display eury- or steno-
thermal behavioral responses (Ivlev and Leizero-
vich 1960). A narrow preferred temperature range
will be overly peaked about the mean (leptokur-
tic), and a broad range of preferred temperatures
will show no obvious mode or only a very slight one
(platykurtic). The coefficient of kurtosis is particu-
larly useful for quantifying the strength of the
temperature selection response in populations
that are not normally distributed where normal
parameters such as mean and standard deviations
are inappropriate.
A normal bell-shaped frequency distribution is
representative of species with a wide preferred
temperature range that is not close to lethal or
avoided temperatures. Speckled sanddabs dis-
played this type of behavior (Table 1, Figure 7).
Newly hatched larvae, however, of species such as
California grunion showed little temperature
selection acuity and preferred an even wider range
of temperatures (^i = 0.003, 0.50.05) nor lepto- or
platykurtic (^2 ~ 2-5, 0.2
19
0 0
0
5
9
9
0 0
0
STEP 7. Determine the mean
n
MST = (0.20) (12°C) +
6
10
29
0 0
0
selected temperature
MST
E P
i/lOOT,
(0.60)(13°C) +
7
11
10
0 0
0
(MST), uhen TjtR.
1=1
(0.20)(14''C) = 13.0°C
8
12 f
R 30
29 0.97
20%
(See Data Analyses for
9
13
10
30 3.00
60%
Artificial Data Set.)
10
U
12
13
14
15
16 X
30
10
, 25
30 1.00
0 0
0 0
20%
0
0
17
7
0
-
14
18
13
0
-
15
19
3
0
-
16
2n
-
0
if .=4.97=f
ARTIFICIAL DATA SET
RUN
SELECTED
TEMP.(°C)
MEAN SD
TIME
COMPARTMENTS
cd
21
(min) 1 2 3
^
5
6
7
8 9 10 11 12
13 14
15
16 17
18
19 20 21
000
TEMP. (°C) 5.0 5.5 6.0
6.5
7.0
7.5
8.0
8.5 9.0 9.5 10.0 10.5
1.0 11.5
12.0
12.5 13.0
13.5
14.0 14.5 15.0
NO. OF FISH 0 0 0
0
0
0
0
0 0 0 0 0
0 0
1
2 3
2
1 0 0
13.0 0.9
8
015
TEMP. ( C) 6.0 6.5 7.0
7.5
8.0
8.5
9.0
9.5 10.0 10.5 11. n 11.5
2.0 12.5
13.0
13.5 14.0
14.5
15.0 15.5 16.0
NO. OF FISH 0 0 0
0
0
0
0
0 0 0 0 0
1 2
3
2 1
0
0 0 0
13.0 0.9
8
030
TEMP. (°C) 7.0 7.5 8.0
8.5
9.0
9.5
10.0
10.5 U.O 11.5 12.0 12.5
3.0 13.5
14.0
14.5 15.0
15.5
16.0 16.5 17.0
NO. OF FISH 0 0 0
0
0
0
0
0 0 0 1 2
3 2
1
0 0
0
0 0 0
13.0 0.9
8
045
TEMP. (°C) 8.0 8.5 9.0
9.5
10.0
10.5
U.O
11.5 12.0 12.5 13.0 13.5
4.0 14.5
15.0
15.5 16.0
16.5
17.0 17.5 18.0
NO. OF FISH 0 0 0
0
0
0
0
0 12 3 2
1 0
0
0 0
0
0 0 0
13.0 0.9
8
060
TEMP. (°C) 9.0 9.5 10.0
10.5
u.o
11.5
12.0
12.5 13.0 13.5 14.0 14.5
5.0 15.5
16.0
16.5 17.0
17.5
18.0 18.5 19.0
NO. OF FISH 0 0 0
0
0
0
1
2 3 2 10
0 0
0
0 0
0
0 0 0
13.0 0.9
8
075
TEMP. (°C) 10.0 10.5 11.0
11.5
12.0
12.5
13.0
13.5 14.0 14.5 15.0 15.5
6.0 16.5
17.0
17.5 18.0
18.5
19.0 19.5 20.0
NO. OF FISH 1 0 0
0
0
2
3
2 10 0 0
0 0
0
0 0
0
0 0 0
13.0 0.9
9
090
TEMP. (°C) 9.0 9.5 10.0
10.5
11.0
11.5
12.0
12.5 13.0 13.5 K.O 14.5
5.0 15.5
16.0
16.5 17.0
17.5
18.0 18.5 19.0
NO. OF FISH 0 0 0
0
0
0
1
2 3 2 10
0 0
0
0 0
0
0 0 0
13.0 0.9
8
105
TEMP. C°C) 8.0 8.5 9.0
9.5
10.0
10.5
11.0
11.5 12.0 12.5 13.0 13.5
4.0 14.5
15.0
15.5 16.0
16.5
17.0 17.5 18.0
NO. OF FISH 0 0 0
0
0
0
0
0 12 3 2
1 0
0
0 0
0
0 0 0
13.0 0.9
8
120
TEMP. (°C) 7.0 7.5 8.0
8.5
9.0
9.5
10.0
10.5 U.O 11.5 12.0 12.5
3.0 13.5
14.0
14.5 15.0
15.5
16.0 16.5 17.0
NO. OF FISH 0 0 0
0
0
0
0
0 0 0 1 2
3 2
1
0 0
0
0 0 0
13.0 0.9
8
135
TEMP. (°C) 6.0 6.5 7.0
7.5
8.0
8.5
9.0
9.5 10.0 10.5 U.O 11.5
2.0 12.5
13.0
13.5 14.0
14.5
13.0 15.5 16.0
NO. OF FISH 0 0 0
0
0
0
0
0 0 0 0 0
1 2
3
2 1
0
0 0 0
13.0 0.9
%_
849
LINEAR PROGRAMMING SIMULATIONS OF THE EFFECTS OF
BYCATCH ON THE MANAGEMENT OF MIXED SPECIES FISHERIES
OFF THE NORTHEASTERN COAST OF THE UNITED STATES
B. E. Brown, J. A. Brennan, and J. E. Palmer^
ABSTRACT
We evaluated the results of using historic bycatch (incidental catch) ratios in adjusting fishing
regulations by linear programming techniques. We used both 197 1 and 1973 bycatch ratios separately
to assess the sensitivity of the results to the reported changes in bycatch ratios in estimating the total
1975 catch of countries fishing in the northwest Atlantic. For 4 of the 1 1 countries for which data were
examined, the difference between the percentage of a country's species total allowable catches (i.e.,
those catches allowed a country by regulation) using the 1971 and 1973 bycatch ratios, was at least
20% . Only four countries were predicted to catch at least 807f of their species total allowable catches.
The predicted total catches of all countries and all species was only 60% of the total species quotas. The
simulated directed fisheries constituted only 70% of the total catch using 1971 bycatch ratios and only
73% using 1973 bycatch ratios. Examination of the reported 1975 catches indicated that the total
allowable catches for herring were most frequently limiting a country's catch. Except for U.S.S.R., the
differences between reported and simulated catches were less than 50 metric tons, with the difference
less than 10 metric tons for 6 of the 11 countries. There was little difference in reported versus
simulated catches between the schemes using the 1971 and 1973 bycatch ratios.
The control of fishing mortality by means of indi-
vidual species catch quotas is difficult in a mixed
fishery, i.e., where a significant proportion of the
fishing mortality on a given species is generated as
a result of the incidental catch, or bycatch, of that
species in fisheries directed toward other species.
Moreover, if a country is allowed to catch a spec-
ified amount of a given species by means of a di-
rected fishery for that species, the total species
catch may exceed that amount because of the as-
sociated bycatch of that species in the other
fisheries.
The International Commission for the North-
west Atlantic Fisheries (ICNAF) modified its
regulatory measures several times in attempts to
account for bycatches of species under quota re-
strictions. The initial haddock quota regulations
(Subarea 5 and Division 4X, Figure 1) stated that
the directed fishery should cease when the ac-
cumulated catch (directed catch plus bycatch) re-
ported to ICNAF biweekly reached 809f of the
quota, anticipating that the catch after closure (a
bycatch by definition) would be 209f of the quota
(ICNAF 1969). When yellowtail flounder was
added to the list of species under quota, the closure
'Northeast Fisheries Center Woods Hole Laboratory, Na-
tional Marine Fisheries Service, NOAA, Woods Hole, MA 02543.
procedures were changed. The Assessments Sub-
committee of ICNAF estimated the expected
monthly bycatch after closure of directed fisheries
and the decision to cease directed fishing was then
made when the accumulated total catch reported
to ICNAF on a biweekly basis plus the expected
bycatch during the remainder of the year equalled
the quota (ICNAF 1970). With the introduction of
national quota allocations in 1972, the procedure
again changed, requiring each country to control
its directed fishery so that the sum of its directed
catch and the estimated bycatches would not ex-
ceed its quota allocation (ICNAF 1972a).
The bycatch problem was acknowledged by
ICNAF in its decision to establish a TAC (total
allowable catch, i.e., that catch allowed a country
by regulation) for all species combined that was
less than the sum of the individual species TAC's
for 1974 and 1975 (ICNAF 1974a). Linear pro-
gramming simulations utilizing bycatch ratios
from directed fisheries for all countries combined
substantiated this policy (Brown et al. 1973; An-
thony and Brennan 1974).
Since 1974, TAC's were set for all species (either
singly or in groups) and for national catches (IC-
NAF 1974a, 1975a). Under this regime, it was
possible to utilize linear programming more
realistically to investigate the extent to which the
Manuscript accepted July 1978.
FISHERY BULLETIN: VOL. 76. NO. 4, 1979.
851
FISHERY BULLETIN: VOL 76. NO 4
64°40 A
Figure l. — Northwest Atlantic Ocean partitioned into ICNAF areas.
regulations in ICNAF were adequate to account
for the bycatch. Simulations of 1975 catches were
made utilizing bycatch ratios from both 1971 and
1973 to assess the sensitivity of the technique to
differences in historic bycatch ratios. Brennan
( 1975) found little evidence of a decline in bycatch
ratios when examined on a country-gear level over
the years 1970-73. We compared the simulated
catches and the reported catches on a species basis
and on a country basis and examined the results to
determine for which countries and species the
simulations were successful.
METHODS AND MATERIALS
Data Base
Almost all countries fishing in Subarea 5 and
Statistical Area 6 (Figure 1) submitted data on
nominal catch (i.e., that reported landed (adjusted
to live weight) by the country, not necessarily that
actually caught — it is the term used in the ICNAF
Statistical Records following standard United Na-
tions Food and Agricultural Organization proce-
dures) and effort for main species (or a species)
sought. These data are published each year in
tables 4 and 5 in the annual ICNAF Statistical
Bulletins. The data of 1971 and 1973 (ICNAF
1972b, 1975b) were the sources of the bycatch
ratios. Data of these years were reported according
to the species categories given in Table 1. The
nominal catches do not include fish caught and
discarded at sea.
The nominal catch and effort (days fished) for
1971 and 1973 for finfish were summed over
months for each target fish of the fishery (the
"main species sought") categories reported in ta-
bles 4 and 5 of the ICNAF Statistical Bulletin
(1972b and 1975b, respectively). Catches made
with fixed gear as well as catches of Atlantic
menhaden, Atlantic halibut, and large pelagic
fishes, i.e., tunas, billfishes, and sharks (other
852
BROWN ET AL.: LINEAR PROGRAMMING SIMULATIONS
than dogfishes), were excluded. Most of these were
not covered by the regulations and have <1 1 (met-
ric ton) per 100 t of directed species caught. In in-
stances where no "main species sought" category
was indicated or where landings were attributed
to a mixed fishery, the monthly landings by vessel
classification and gear were assigned to "species
sought" categories according to the species which
formed a simple plurality of the catch. The United
States of America often reported mixed fisheries
on groundfish species. The Union of Soviet
Socialist Republics (U.S.S.R.), Poland, Japan,
and German Democratic Republic (G.D.R.) typi-
cally reported their pelagic and/or squid fishery
catches as mixed.
The term "fishery" as used in this paper refers to
the vessels and associated catch on these "main
species sought" categories. The term "species" re-
fers to both individual species and species groups.
All reported landings were thus identified by two
factors: species and fisheries. Such tabulations
were prepared for all nations for which data were
available. For Romania, which has had an Atlan-
tic herring fishery but did not report a directed
Atlantic herring fishery in 1973, bycatch ratios for
1972 (ICNAF 1974b) were used for that species
fishery. The only countries with an allocated na-
tional quota for which 1971 and 1973 data were
not available and thus could not be analyzed were
Italy (1971 and 1973) and France (1971).
In this paper, all catch restrictions described
below will all be referred to as "quotas." To apply
linear programming techniques to the bycatch
problems restraints on the total catches for each
species by country need to be set. For countries and
for species categories reported in ICNAF Statisti-
cal Bulletins, we used restraints in linear pro-
gramming (ICNAF 1974a). For countries and/or
species for which ICNAF had not set specific quota
allocations (but for which the quota was included
in, say, "other countries" under ICNAF regula-
tions— a country not giv^ a specific catch quota
could fish in competition with other similar coun-
TabLE 1. — Species categories as reported to ICNAF, 1971 and
1973.
1971
1973
1973
Atlantic cod
Atlantic cod
Yellowtall flounder
Haddock
Haddock
Other flounder
Redfish
Redfish
Atlantic herring
Atlantic halibut
Silver hake
Atlantic mackerel
Silver hake
Red hake
Other pelagic
Atlantic herring
Pollock
Other groundfish
Other pelagic
Amencan plaice
Other fish
Other groundfish
Witch flounder
Squids
Other fish plus squids
tries from an "other country" allocation or "other
flounder" category), we estimated these re-
straints by the following procedures. These were
chosen so that the categories of quota allocations
matched the species categories (Table 1) by which
the catches were reported. We proportioned the
"others" allocation category for each individual
species to countries based on the 1973 nominal
catch for each particular species and the catch of
that species of all of the countries that did not have
a national quota for the species. We proportioned
the quota for "other groundfish" and "other
pelagic" from the "other fish" TAC for each coun-
try. The quotas for American plaice and witch
flounder were subtracted from the "other floun-
der" TAC for each individual country. Since the
quota for pollock was set by ICNAF for Division
4VWX plus Subarea 5, national quota allocations
were estimated as an average percent of the nomi-
nal pollock catches during 1971, 1972, and 1973 in
Subarea 4VW and 5.
Analysis Methods
Linear programming is a optimization method
for which the effectiveness of an allocation scheme
distributed over several variables is measured by
the maximum or minimum value of some linear
function of those variables, when those variables
are subject to linear constraints. The problem con-
sidered here was to determined = (Xj, X2, . . . , x„)
such that
1 = 1
(1)
is maximized, where for each /, c, was the weight-
ing coefficients of the variable x^. In the present
context,
X, = catch of species / to be taken in directed
fishery for species /,
= catch of species / in all fisheries divided by
catch of species i taken in directed
fishery for species i (c, ^ 1.00),
= number of directed fisheries considered,
and
z = total catch of all species.
Solutions (Xj, X2, . . . , Xn ) of Equation (1) were sub-
ject to the constraints for each /
c,
n
(2)
853
FISHERY BULLETIN: VOL. 76, NO. 4
0
(3)
where d,^ = catch of species j taken in directed
fishery for species //catch of
species i in directed fishery for
species i
bi = constraint on total catch of speciesj,
for J = 1 . . . m.
The estimates of d^ for each country for 1973 are
presented in Appendix Table 1. Analogous tables
for the 1971 data are in Brown et al. (1973).
The solution used in this paper was devised by
using the Simplex Algorithm (Hadley 1963: 132f)
which was computed by using a Honeywell^ com-
puter program LINPRO; a description of this use
of linear programming is given in appendix II of
Brown et al. (1973). In this analysis the linear
constraints were that no country would exceed its
national allocation for any species (6,). The output
of the LINPRO program includes the vector X of
directed catches of the species along with the re-
sultant total catches of the species and the overall
total catch.
RESULTS AND DISCUSSION
The results of each country's simulation are
given in Appendix Table 2. In each case the sum of
the species quota allocations exceeded the coun-
try's maximum possible catch (without violating
single species constraints) as determined by the
linear programming model. Table 2 lists the ratios
of the simulated catches to the TAC's using 1973
and 1971 bycatch ratios. For 4 countries (Bul-
garia, Canada, G.D.R., and Japan) of the 11, the
percentages derived from 1971 bycatch ratios dif-
fered from those derived from 1973 fishing pat-
terns by at least 0.20. More detailed reporting of
catches (i.e., by species rather than groups) in
1973 than in 1971 and, therefore, in the analysis
contributed to this change. Poland, United States,
France, and Federal Republic of Germany (F.R.G.)
were the only countries which could have taken
>80'^ of the sum of their species TAC's based on
1971 or 1973 bycatch rates. The United States,
however, has a significant discard of fish which is
not taken into consideration in this analysis. Of
the other countries considered, the effect of unre-
TABLE 2. — Comparison of maximum catches from linear pro-
gramming simulation using 1971 and 1973 bycatch ratios, with
sum of species "quotas" for the ICNAF area.
Maximum catch — sum of
species quota using:
1 973 bycatch
1971 bycatch
Country
ratios
ratios
Bulgaria
0.64
0.83
Canada
.54
.78
France
.52
—
Federal Republic of Germany
.97
82
German Democratic Republic
.40
.64
Japan
.57
.17
Poland
.94
.93
Romania
.08
.05
Spam
.72
.72
USSR.
.25
.35
United States
.90
.93
ported discard would be expected to be greatest in
the Spanish squid fisheries.
Closer inspection of Appendix Tables 2 and 3
reveals the species which were the limiting factors
in a country's inability to take the sum of its
species quotas at present. These are the species
which were caught in significant amounts as
bycatch and directed catch and for which a species
quota was met. The species whose catch was most
frequently limiting was herring, when either 1971
or 1973 bycatch ratios was used. The next major
species using 1973 ratios were pollock and "other
pelagic" and using 1971 ratios were "other fish,"
"other pelagic," and haddock. Pollock was less
limiting when 1971 ratios were used because it
was combined with the "other groundfish" cate-
gory, which had not been limiting.
The sum of the linear programming estimates
over countries using 1971 and 1973 data are pre-
sented in Tables 3 and 4, respectively. In each case
the sum of the expected maximum catches deter-
mined by the linear programming runs was only
about 60^^ of the sum of the species quota. The
simulated directed fisheries catch levels composed
only 707c using 1971 bycatch ratios and 73% of the
Table 3. — Sum of individual country's linear programming
simulation of 1975 catches in the ICNAF area, maximizing total
catch (1,000 t) and using 1971 bycatch ratios. Catches of France
assumed to be those using 1973 bycatch ratios.
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
Total allowable
Directed
Total
Species sought
catch restraint
catch
catch
Atlantic cod
45.00
1.7
18.53
Haddock
6.00
0.0
5.23
Redfish
25.00
6.60
22.20
Silver hake
175,00
43.65
62.68
Flounders
41.00
1.32
36.25
Other groundfish
152.00
64.08
84.49
Atlantic herring
175.00
140 14
176.69
Other pelagic
311.90
18907
210.48
Other fish plus squids
127.40
26.08
67.25
Total
1,058.30
482.64
683.81
854
BROWN ET AL : LINEAR PROGRAMMING SIMULATIONS
Table 4. — Sum of individual country's linear programming
simulation of 1975 catches, maximizing total catch (1,000 1), and
using 1973 bycatch ratios for the ICNAF area.
Total allowable
Directed
Total
Species sought
catch restraint
catch
catch
Atlantic cod
45.00
16.39
31.48
Haddock
6.00
000
5.25
Redfish
25.00
1824
22.25
Silver hake
175.00
74 69
85.72
Red hake
65.00
11.83
26.51
Pollock
21.30
9.57
20.28
American plaice
2.70
—
1.15
Witch flounder
4.30
—
1.70
Yellowtail flounder
1600
11.02
15.06
Other flounder
18.00
—
6.54
Other groundfish
65.70
27 38
40.96
Atlantic hernng
175.00
107.38
120.01
Atlantic mackerel
28500
127.51
150.60
Other pelagic
26.90
16.97
26.45
Other fish
56.40
9.33
33.35
Squids
71.00
25.93
40.30
Total
1 ,058.30
456.24
626.75
total using 1973 bycatch ratios, the rest being
taken as bycatch. The highest percentage of
TAC's, which were caught in directed fisheries,
were for other pelagics (9(K^), Atlantic herring
(797r), other groundfish (76%), and redfish (75%)
using 1971 bycatch ratios, and for Atlantic her-
ring (89% ), silver hake (87% ), Atlantic mackerel
(85%), and redfish (82%) using 1973 bycatch
ratios.
Referring to the individual country linear pro-
gramming output tables in the Appendix, it is
obvious that under 1971 and 1973 bycatch ratios,
national patterns ran the gamut from almost a
total mixed fishery by the U.S.S.R., and to a some-
what lesser extent by the G.D.R., to very specific
fisheries of the F.R.G. and Poland.
As noted earlier, the species which was most
frequently limiting to the total reported 1975
catch was Atlantic herring (6 out of 11 countries),
and the countries which had the most limiting
species TAC's were United States (5) and U.S.S.R.
(4). Except for the catches of U.S.S.R., United
States, G.D.R., and Poland, there was little differ-
ence in reported total catch minus simulated re-
ported catch, when 1971 and 1973 bycatch ratios
were used. Moreover, only for U.S.S.R were these
differences > 50,000 t, and for six of the countries
the differences were < 10,000 t for both schemes.
The species for which the simulated and reported
total catches differed most varied by country. At-
lantic herring and Atlantic mackerel were the
species most frequently differing in simulated vs.
reported catches, but Atlantic mackerel and silver
hake contributed most in metric tons to the differ-
ences. In general, and in view of the findings of
Brennan ( 1975), the differences between schemes
using 1971 and 1973 bycatch ratios were minimal,
and more likely due to the different grouping of
the data.
A summary of the 1975 TAC's, the 1975 re-
ported catches, and the linear program estimates
of total catch by country, is presented in Table 5. It
is obvious that the overall TAG of 850,000 t for
1975 would not be attained without exceeding cer-
tain species TAC's unless bycatch was reduced,
according to the simulations. The expected catches
of 626,750 t using 1973 bycatch ratios and of
681,050 t using 1971 bycatch ratios are only 74%
and 80% , respectively, of the 1975 total TAG On a
country basis, and using the results derived from
the 1973 bycatches, it can be seen that the country
total TAC's were set for 1975 at approximately
appropriate levels for France and Spain (based on
Table 5. — Comparison of linear programming estimates of maximum total catch by overall country's total allowable catches (TAC's)
in the ICNAF area. Figures in 1,000 t.
1973 nominal
Sum of
Linear programming
estimate of
Actual 1975
catch of species
species
1975
total catch
nominal catch of
regulated by
TACs
total
1973 bycatch
1971 bycatch
species regulated
on total TAG
Country
the total TAG
for 1975
TAG
ratios
ratios
Bulgaria
37.29
34.40
24.65
2222
28.74
24.69
Canada
16.80
26 32
26.00
14.24
20.51
14.00
France
3.62
529
2.95
276
2.76
3.36
Federal Republic of
Germany
38.28
30.89
24.85
3005
2531
25.10
German Democratic
Republic
150 85
100 98
82.85
40.52
64.17
82.74
Italy
3 92
4.15
(')
(')
4.40
Japan
32 90
45.35
21.25
26.05
7.59
20.84
Poland
190 55
153.94
129.25
144.87
144.37
127.05
Romania
7.14
5.71
3.85
0.46
0.27
1.80
Spain
22.20
20 98
14.80
15.06
15.10
14.65
U.S.S.R.
449.04
36664
301.80
93.10
127.02
313.78
United States
203.09
26237
211.60
237.42
245.21
221.04
Total
1,155.68
21,052.87
3850.00
62675
"681.05
853.45
'No estimate available.
^Six thousand metric tons of other species not prorated to other species.
^Includes 2,000 t allocated to others.
"Due to the absence of bycatch ratios for 1971 data, estimate of France's total catch is derived from the 1973 bycatch ratios.
855
nSHERY BLLLETTN: VOL. 76. NO. 4
reported statistics', too low for the F.R.G.. Japan.
Poland, and United States, and too high for the
other countries. In fact, summing the national
total TAC's rather than the linear program esti-
mate-; • "'■ c • : - - r>- catch, when the former are limit-
ing. : " overall estimated catch, results in
an expe :atch of 575.000 t. only 68^ of
the TAG. The analogous expected total
cater, derived from 1971 bycatch ratios was
627.470 t. only 74'~c of the overall TAG. Bycatch
may be reduced through actions initiated by
fishing fleets or by regulations such as the closure
to bottom trawling by larger vessels in the south-
em New England. Middle Atlantic, and Georges
Bank areas IGNAF 1975' for 1975 and by the
similar closure on Georges Bank for 1976. The
reduction of the overall TAG to 650.000 t in
1976 IGXAF 1976 ' and 525.000 1 in 1977 > IGXAF
1977 1 was designed to reduce the bycatch problem.
It should be noted, however, that despite the
above potential for change as well as the in-
a^T _ -lies of the rep>orting to IGXAF. which may
c:~ : -r -.ore than one directed fisherv" under a
mixed category-, there were other factors which
worked in the opposite direction. The first was the
inadequate recording of bycatch noted during in-
ternational inspections. Some of this was dis-
carded and not rep>orted. and some was apparently
utilized but not accurately reported on logbooks.
Both the lack of reporting and any underestimates
of bycatch can cause the bycatch ratios used in this
anal\-sis to be underestimated.
In mixed species fisheries, bycatch mxist be con-
sidered in the allocation of quotas to species and to
elements of the fisherj- 'in this example the ele-
ments are countries, but under different cir-
cumstances they cotild be otherwise — e.g.. ports i.
Lack of attention to attendant bycatch may result
in an unexpected overhar\"est of selected species or
conversely the wastage of large quantities of pro-
tein depending on whether or not the directed
fishen.' ceased when a small amount of bycatch
had been taken. Linear programming pro%ides a
suitable technique for examing this problem.
However, to have a refined analysis, accurate
statistics as to main species sought and the com-
position of the bycatch including discards must be
available. Lacking these, the inferences as in this
pap>er. are directional. The specific indi\-idual es-
timates can be interpreted for policy decisions only
when the user has the understanding of the fishen.-
to qualitatively account for the appropriate re-
porting inadequacies.
LITERATURE CITED
ANTHONY. V. C. .^NT> J. .\. BRENN.^N.
1974. An example of the by-catch piroblem on directed
fisheries for 1975. Annu. Meet. Int. Comm. Northwest
Atl. Fish.. Summ. Doc. 7-i 47 'Re%is€d i. Ser. No. 3386. 5 p.
BREN'N.\N. J. A.
1975. By-catch trends of selected fisheries operating in
ICNAF Subareas 5 and 6. Annu. Meet. Int. Comm.
Northwest Atl. Fish.. Res. Doc. 75 70. Ser. No. 3554, 14 p.
BROWN. B. E.. J. A. BREN-N.\N. E. G. HEYERD.\HL, .4NT) R. C.
Hen-n-emlth.
1973. Effect of by-catch on the management of mixed
specie fi.gheries in Subarea 5 and Statistical area 6. Int.
Comm. Northwest Atl. Fish.. Redb. 1973. Part HI. p. 217-
231.
H.\DLEY. G. F.
1962. Linear pn^ramming. Addison- Wesley, Reading.
Mass.. 520 p.
INTERNATIONAL COESOSSION FOR THE NORTHWEST ATIAN-
TTC FISHERIES.
1969. Int. Comm. Northwest Atl. Fish. Annu. Proc. 19. 55
P-
1970. Int. Comm. Northwest Atl. Fish. Annu. Proc. 20. 47
P-
1972a. Int. Comm. Northwest Atl. Fish. Annu. Proc. 22, 94
P-
1972b. Int. Comm. Northwest Atl. Fish.. Stat. Bull. 21, 135
P-
1974a- Int. Comm. Northwest Atl. Fish.. -Annu. Proc. 24.
128 p.
1974b. Int. Comm. Northwest Atl. Fish., Stat. Bull. 22, 239
P-
1975a. Int. Comm. Northwest AtL Fish. Annu. Rep. 25.
116 p.
1975b. Int. Comm. Northwest Atl. Fish.. Stat. Bull. 23. 277
P-
1976. Int. Comm. Northwest Atl. Fish. Annu. Rep. 26, 139
P-
19 / / . Int. Comm. Northwest Atl. Fish. Annu. Rep. 27, 84
P-
856
BROWN ET AL.: LINEAR PROGRAMMING SIMULATIONS
•E
CO
tC
C
bo
3
o
v
a.
JZ
u
"S
0)
S5 c
a> o
< 1s
2 iS
'^ Q.
ce X
•^ 0)
CO u
-*^ o
in a,
^
-^
C
3
o
be
C
c
ca
CO
c
E
o
c
CO
Ed
<
E-
X
3
z
w
a.
a.
<
(uro
y O)
= 0)
OB
Og
^1
2 o
a) o
§g
J58
CO
o
CD
p
d
I I I
I I I I I II
o o o
O I CD O
in I o o
I I
o in
5 I 8
b C3
-■ O o
I C^ O o
cb 6 ^
cj) r-
38
CD d
en o
■* o
CM o
-
in 1-
o o
o o
d 6
II I I
CO
I I I 8 I
d
I I I 8 I
un <3^ o
CM in O I I
- o o I I
d o -^
<- -t a>
o q 5 I I
d d d
o p II I
d o
c\j T in
o o o I I
o o o I I
d d d
CO in cy
o o o I I
d d d
•^ (D CM
GO CM r-^ I I
O - 'J- I I
d o CO
I I
I I
o
° I
S ■* S
I I o o) c6
I I p o o
odd
CO o o 00
I CO >- o r-
I o o o ■■-
o o ^' o
I I
CO CD
2 o
o o
f^ CM in ^ f^ o
CO •- r^ CO ^ o
p p 00 CO Tt o
d d d d d '-^
GO CM in o *-
CO CD m o r^
p o p o o
d d ^ f~ c^
r^ CO o f^ m
in -- o o ^
O CO O T CM
d d -^ d d
I I
CM CM
o o
o
o
o
»- o O) CD r^ o r^
CO m CO in CO o o>
O CM o o <- o '-
d o d d o T- d
- I
■- d d d
^ -^ CO CM
•^ 2 I o 1 o
op I o IP
d d d d
00 O CM ^- T-
I I in o in ■»- CO
I I cu O CO CO CM
d y- d d d
f^ o in in CO o
■^ I o r^ CM CO CO
p I o o o o o
d -^ d o d d
CM CM cj> rs. r^
I I ^ ^ CO »- in
I I o o o o o
d d d d d
I I
I I
MM II II I I I I I II II
II II II I I M II II II
MM II I II I II I I I II
II I I I II
Mill M
MM
I II II
O -- CD
O O I O
o o I O
■^ d d
CO
d
I II I II
■- o
CO o
o o
o
o
o
MSI
CM
O
d
II I 8
CO '- CD
I O O O
I o p O
O CM O
. CM ^ CM
O I O O O
I 5 I
I I I
CM r~
■■j- CO
CM Tt
I 2 CO
I O O
O
d
I I
CD --
O O
O O
d d
C35 CM 1^
O O CM I I
o p o I I
odd
^ o o
.- O O I I
CM o r^ I I
o >- d
o o^ f^
O "^ 00 I I
o m o I I
— d --
I I
I I
881 1
1 5
in
5
d d
d
d
II I II
8 II I
d
I II I I II II I II M II
O O
O O
O
d
I II I 8
I I I
O O
o o
d d
I I I
857
FISHERY BULLETIN: VOL. 76, NO. 4
u
o
0)
a.
0)^
o-
OS
£ c
Og
0) p
is
E Q.
<
X
S
z
u
a.
a.
<
o
o
o
d
in CO
m in
o o
u ^
c!¥
n
< >
CM
o
ra o
CVJ
o
- d o
O) CO "^ o <-
(D 1- CD O O
O ^ ■- O CO
o o o »- o
CO CO
88
CM
8
f^
o
o
o
o ..-
o in
>- o
in
ID
o o
o
o
T-
o o
O
1- Tt CM C\J CO
O O '^ O O
o o o o o
o o o o o
■^ CO C\i CO CM
o o o o o
o o o p o
o o o o o
CO O) 1^ CD
O O I I IT I C^
o o ! I o I T-
o d> o cri
CD o '- o -^ a> lo
CO o o ■^ a^ en -^
CM O ^ ■^ P CM CM
o ^ o o o o o
o o '~ r^ 1^ CM h-
o ^ r^ 00 ^ CD Tj
o ^ \n y- -- <:> ■•-
^' CD CD CD O O O
■a- Ol
CO T- 1
o o 1
1 o
o
5
o o
o
o
I I I 8 I
in o -^ '- o
o CM CT) r: r: I
o o T o o I
o o d o d
co.^mr.-incoCM»-
OOCMf^-OOO"!
oooooooo
doododdo
o
o
C75 CD *-
I O .- O
I O CM O
in en
I S 'D
o o
o d o o o
(D CO I O
- - I o
o d '-
■^ r^ o in
rvi o o o
o T- o o
o o ■- o
CO CD
11188
o o
I I
CM ■-
I O O
O O
^cOf.-cor^ooin
oo^ocoocdcm
opoooo--p
dddoO'-dd
CO CM O Tt
I I o o O CO
o o o o
I I I
Tj- ^ t- CO o O^
■^ '- O I I OJ O IT) I
o p p I loop I
O O O O ■^ CD
o
o
o
CO
in
o
o
o
in
o
o
CD
in
o
CO
00
o
in
o
o
o
o
o
o
o
o
o
o
o
o
in
CO
o
s
o
o
o
CD
o
o
in
CO
o
CO
in
o
o
o
o
o
o
o
o
o
cDr^cDcoi^o-^cMo coo o
in'-';j-Lnincoor-~o i -^co i »-
OOOOCOt-OOO I "-p I "-
ddoddodO'- (d c> d
COCU'^COOOO CDCD >-
^cooino.* I ■•- o I 2
O'-oop^ Ipp IP
dodo*-o c> ci o
.^ CO r^ o CO o
o ■^ o o CM t^ I
o >- o o -^ o I
d d d <- d d
■-
d o p p
I T- o o
o
d d d d
I I I
ocoa)'^cocO'-CM-<^i- COO
0«3'COir)C\J(DC7)01^0 I O'-
OCOOOO'-O'^COP Ipp
'-^OCDcbOCDOOoO (DO
0)
ro
-^s
•p -O DC
V 3 © ^
SZ (0 CJl O CJ Q. .^ ".'
^- ~ ~ o o'-^
6 -^ ^
C ^ -« C3. £
3 .S B E ^
o .c
= c
_ D
o
•= 0 M
0)01 —
T 5 ■*
^ -2 Q.
w ^-CJZ ro^-!r= C3)U
5g-«I^IS5l£KS^^I^Bl££5S££&
<(f>(/iincr.O<<00^00<<00(n
858
BROWN ET AL.: LINEAR PROGRAMMING SIMULATIONS
Appendix Table 2. — Linear programming simulation by country in ICNAF Subarea 5 and Statistical Area 6, 1975 catches to
maximize total catch (1,000 t). Simulated using 1973 bycatch ratios. Actual directed and total catches are included also.
Total
Total
allowable
catch
Simulated
Actual
allowable
catch
Simu
ated
Actua
Directed
Total
Directed
Total
Directed
Total
Directed
Total
Species sought
constraint
catch
catch
catch
catch
Species sought
constraint
catch
catch
catch
catch
BULGARIA
POLAND
Atlantic cod
0.07
—
0.03
—
—
Atlantic cod
0.49
—
0.37
—
0.48
Redfish
0.50
—
0.03
—
—
Redfish
0.40
—
—
—
<0.01
Sliver hake
200
—
0.92
1.02
1.92
Silver hake
5.30
—
0 13
024
0.38
Red hake
5,41
—
0.23
—
0.03
Red hake
2.20
2.12
2.20
—
—
Yellowtail flounder
0.14
—
0.06
—
<0.01
Pollock
0.35
0.28
0.35
—
0.02
Other groundfish
0.65
—
0.13
—
0.34
Other groundfish
1.40
—
1.40
—
1.11
Atlantic herring
1 20
0.47
1.20
—
0.42
Atlantic herring
38.40
32.14
38 40
33.05
38.46
Atlantic mackerel
18.75
1864
1875
18.47
18.75
Atlantic mackerel
90.00
81.45
90.00
68.45
74.28
Other pelagic
075
—
0.15
—
0.39
Other pelagic
220
0.15
220
0.17
3.77
Other fish
2.60
—
0.48
—
2.63
Other fish
6.40
0.34
6.40
—
1.71
Squids
1.70
—
024
—
0.21
Squids
6.80
0.45
3.42
3.25
6.84
Total
34 40
2222
24.70
Total
153.94
144.87
127.05
CANADA
ROMANIA
Atlantic cod
4 82
0.55
1.31
1.10
1.93
Haddock
0.01
—
<0.01
—
—
Haddock
1.20
—
0.60
0.44
1.44
Redfish
0.34
—
<0.01
—
0.01
Redfish
0.50
—
0.02
0.01
0.06
Silver hake
0.50
—
<0.01
—
0.12
Pollock
2.46
—
2.46
4.13
4.74
Yellowtail flounder
0.01
—
<0.01
—
—
American plaice
<0.01
—
<0.01
—
0.02
Other groundfish
0,15
—
0.01
—
<0.01
Witch flounder
<0 01
—
<0.01
—
0.01
Atlantic herring
0.20
0.20
0.20
1.54
1.54
Yellowlail flounder
0.02
~-
<0.01
—
0.01
Atlantic mackerel
3.75
0.05
0.10
—
0.07
Other flounder
0.03
—
0.02
—
0.05
Other pelagic
0.13
—
0.13
—
—
Other groundfish
0.78
0.70
076
030
0.66
Other fish
0.02
—
0.02
—
—
Atlantic herring
9.00
9.00
9.00
5.08
5.08
Squids
0.60
—
<0.01
—
0.05
Atlantic mackerel
7.50
—
0.06
—
<0.01
Total
5.71
0.46
1.79
Other pelagic
0.01
001
0.01
—
—
SPAIN
Total
26.32
14.24
14.00
Atlantic cod
7.09
1 49
1.49
407
4.07
FRANCE
Haddock
0.30
—
0.10
—
0.07
Other groundfish
0.02
—
002
—
—
Red hake
0.07
—
<0.01
—
0.01
Atlantic herring
1.87
1.87
1.87
3.34
3.34
Pollock
0.42
—
0.42
—
0.10
Squids
3.40
087
087
—
—
Other groundfish
0.10
—
0.05
—
0.42
Total
5.29
2.76
3.34
Squids
Total
13.00
20.98
13.00
13.00
15.06
9.90
9.90
14.57
FEDERAL REPUBLIC OF GERI^ANY
Atlantic cod
009
—
0.01
—
0.02
USSR.
Silver hake
0.50
—
0.04
—
0.04
Atlantic cod
2.50
—
0.24
—
2.43
Pollock
1.60
1.60
1 60
0.10
0.15
Haddock
0.05
—
0.05
—
0.01
Other groundfish
0.90
0,48
090
—
0.02
Redfish
1.44
—
1.44
—
1.37
Atlantic herring
24.50
24 50
24.50
22 99
2301
Silver hake
113.30
40.20
41.22
71.38
88.88
Atlantic mackerel
1.40
0.99
1.40
0.08
0.47
Red hake
44.40
—
11.18
4.50
26.12
Other pelagic
0.51
—
0.35
—
1.46
Pollock
1.26
—
0.20
—
0.19
Other fish
0.39
—
025
—
—
American plaice
0.20
—
0.05
—
0.18
Squids
1.00
0.68
1.00
—
003
Witch flounder
0.20
—
0.05
—
0.20
Total
3089
30.05
25.20
Yellowtail flounder
0.84
—
—
—
0.08
Other flounder
0.60
—
0.20
—
0.56
GERtVIAN DEfVlOCRATIC REPUBLIC
Other groundfish
16.70
—
2.79
—
2.86
Atlantic cod
1.30
—
0.03
—
0.03
Atlantic herring
42.10
1.91
5.28
37.08
40.95
Redfish
0.63
—
0.02
—
0.01
Atlantic mackerel
101.25
1.96
14.80
99.91
106.31
Silver hake
3.10
—
0.06
—
0.04
Other pelagic
4.40
4.15
4.40
—
0.68
Pollock
3.50
3.49
350
<0.01
0.10
Other fish
28.90
—
8.20
5.99
34.08
Other groundfish
<0.01
—
—
—
0.07
Squids
8.50
—
3.00
3.53
8.94
Atlantic herring
31.90
13.00
13.75
27.00
30.90
Total
366.64
93.10
313.84
Atlantic mackerel
5625
20.00
20.14
47.95
48.34
Other pelagic
0.06
0.06
0.06
UNITED STATES
Other fish
294
—
2.90
0.12
2.18
Atlantic cod
28.00
14.35
28.00
12 46
23.41
Squids
1.30
—
0.06
—
0.90
Haddock
4.50
—
4.50
086
5.09
Total
100.98
40.52
82.63
Redfish
20.62
18.24
20.62
7.07
896
Silver hake
43.00
34.49
43.00
17.79
2059
JAPAN
Red hake
12.90
9.71
12.90
0.11
2.43
Atlantic cod
0.05
—
—
—
—
Pollock
11.50
4.19
11.50
3.80
806
Redfish
0.50
—
0.12
—
0.02
American plaice
2.50
—
1.10
0.26
2.19
Silver hake
7.30
—
0.35
—
<0.01
Witch flounder
4.10
—
1 65
0.36
2.03
Red hake
0.03
—
—
—
<0.01
Yellowtail flounder
15.00
11.02
15.00
14.99
19.32
Pollock
0.25
—
0.25
—
—
Other flounder
17.30
—
6 28
11.81
19.39
Other flounder
0.06
—
0.04
—
—
Other groundfish
44.88
26.20
34 80
10.34
19.11
Other groundfish
0 10
—
0.10
0.33
1.13
Atlantic herring
2465
23.20
24.65
3576
36.09
Atlantic herring
1.16
1 09
1 16
1.88
1.88
Atlantic mackerel
4.70
4.11
4.70
0.54
1.65
Atlantic mackerel
0.80
0.31
065
008
0.20
Other pelagic
9.52
5.95
9.52
1961
23.40
Other pelagic
9.30
6,71
9.30
265
3.62
Other fish
13.60
8.62
13.60
17 02
27.65
Other fish
1.50
0,37
1.50
—
—
Squids
5.60
1.04
5.60
0.21
1.67
Squids
24.30
9.89
12.58
13.25
1399
Total
262.37
237.42
221.04
Total
45.35
26.05
20.84
859
~3iZRY BULLETrS: VOL. 76. NO 4
APfEKTHX Table 3. — Linear TjroGT=
T in ICNAF St-
=tic6. Actual (Lr
.Area 6 of catches to i
rs are included also.
scares
Z -e-r?:
--• =
: ?: e:
--■ =
. . .- = '
:s-~
:i"r'
s:-
:=::'
0.49
^.^
0-46
; -:
—
I Cv
—
<0-01
: ;:
—
: :-9
--2-
0.38
: rf
—
_ i^
—
1-13
.-: -.
2t :'
35 -iC'
33-05
38.46
--_ ^ .
= - ^:
92^
68.62
78-05
' z 1'.
' -:
13.20
3.25
8-55
zl ^
144.37
127.05
0.01
_
<0JJ1
_
_
0-34
—
—
—
0.01
Z5Z
—
0.01
—
0.12
—
C 31
—
—
<0 C"
—
<0.01
. _ .
—
Z-I-i
V54
1.54
3 t'z
;.'4
C-'5
—
0-07
0-62
—
0-06
—
0-05
5.71
0^
1.79
7.09
1.71
1 T1
4.07
4-07
^ -3^
—
030
—
0.07
: -.1
—
0 09
0-53
'':'.'.
13.00
■3 :c
5-9C
9-90
22 9>c
15.10
1457
2 iC'
_
* ";:
_
2.43
0-05
—
0.05
—
0.01
1 4.4
—
1 44
—
1-37
- - -,-
' — -
71 38
88.88
' i-
—
■ :-
—
1-02
D^JC
C.",4
5-14
4j0
29-17
45 10
34,46
42.10
37.08
40.95
39-85
47-33
99.91
106.99
: " -;
—
10.01
9.52
43.02
3cc 5^
127 02
313.84
2?:-:
15.54
1246
23.41
- : :
—
4.50
0-86
509
21 t-
16.59
20.62
7.07
8.96
43-00
32-77
43.00
17.79
20-59
35 5:
"32
34.20
27.42
4293
" r :: 1
: "r c-
69^
14.25
29-60
iL- 11
• .11
24-65
35.76
36.09
14-22
12-56
14.22
20.15
25.05
19.20
9-86
19^
17.23
29.32
262-37
245^1
221.04
Tec*
0.70
0-01
0-50
2Xi
0.14
ace
-2C
19o0
43C
34.41
--2C
OjC
0-06
9.00
751
25-32
2-50
245C
1-91
18-;
-2C
—
0.42
=-50
ia47
19.15
4-30
—
2-84
5 ~i.
24.:'0
: OC 5 06
14.00
054 O-O8
'.33
130
—
053
—
3.10
—
350
0-54
31.90
3C-53
5631
2'.Zi
424
—
54-17
3 08
£2-63
050
—
" " :
7-X
—
0 'A
—
'_-
—
0 35
—
: V:
: .-.
- - -
1.16
•-15
1-16
--88
" 55
10-10
4.49
450
273
3-32
25-30
—
-I SI
13.24
-3 yz.
860
NOTES
EFFECT OF SWIMMING SPEED ON THE
EXCESS TEMPERATURES AND ACTIVITIES
OF HEART AND RED AND W HITE MUSCLES IN
THE MACKEREL, SCOMBER JAPOSICUS
Body temperatures of most fish t^-pically are about
the same as the water in which they swim for
much of the heat generated by muscular activity is
ducted away via the circulating blood and lost b\'
convection at the gills and body surface.
Some scombrids and lamnid sharks consei'\'e
muscle heat using countercurrent vascular heat
exchangers tretia mirabiliai so that temperatures
are maintained significantly above ambient in the
brain, eyes, red and white swimming muscles, and
viscera (Carey et al. 1971; Stevens and Frj' 1971;
Linthicum and Carey 1972; Graham 1973). In
other fishes lacking these heat conserving devices,
only small temperature excesses above ambient
have been recorded, but rarely more than VC
(Stevens and Fry 1974). Since heat production
must depend primarilj' on work output by the
locomotor musculature, we have examined effects
of swimming speed on the magnitude of the small
temperature excesses in a "cool" scombrid not
equipped with the retia exchangers, the mackerel.
Scomber Japojjicus (locally the Pacific mackerel =
chub mackerel).
Another important question concerning scom-
brid locomotion is how contractions of red and
white muscle fibers are staged as swimming speed
increases. It is generally thought that red muscle
provides power for cruise swimming and that
white muscle functions in "burst" swimming
(Rayner and Keenan 1967). Red muscle is pre-
dominately aerobic and utilizes fatty acids as the
major energy source whereas white muscle (which
uses glycogen) usually functions anerobically
(Gordon 1968; Bilinski 1974). The second objective
of our study was to determine how heart rate and
red and white muscle activity of S.japonicus are
affected by swimming speed. For this purpose,
electrodes were implanted into the pericardial
space and in swimming muscles of fish so that
simultaneous records of electrocardiograms
(ECG's) and red and white electromyographs
(EMG's) could be obtained.
The genus Scomber is a primitive member of the
familv Scombridae (Kishinouve 1923). It has a
fusiform shape, is less heavilj' bodied than the
skipjack tuna. Katsuwonus pelamis. and other
tunas, but shares several characteristics with
warm-bodied species; they swim continuously
(swim bladders are reduced or absent), have high
rates of oxygen consimiption (Baldwin 1923: Hall
1930). and have high blood hemoglobin levels
(Greer-Walker and Pull 1975 1. They are also ob-
ligatorily dependent upon ram gill ventilation as
adults (Roberts 1975) and have large gill siirface
areas with a high diffusion efficiency (Hughes
1966; Steen and Berg 1966).
Materials and Methods
Surgical Procedures and Swimming Experiments
The general procedure was to implant either
thermocouples or cardiac (ECG) and muscle
(EMG) electrodes into mackerel which were then
placed in a Blazka-Fry tunnel respirometer ( 12 cm
i.d.) to swim at controlled velocities. Fifteen
specimens (35-40 cm fork length (FL); 0.38-0.62
kg) were obtained from regularly replenished and
maintained mackerel stocks at the Southwest
Fisheries Center La Jolla Laboratory. National
Maiine Fisheries Service. NOAA. After netting,
each fish was anesthetized in a large basin of
oxygenated seawater containing 0.2 g 1 of tricaine
methanesulfonate (Crescent Research Chemical,
Inc.)^ and placed on an operating table where its
gills were perfused continuousl.v with a fast flow of
oxygenated seawater containing a small amount
of the same anesthetic (0.08 gl). Thermocouples
(0.127 mm in diameter copper constantan,
polyvin3i chloride insulation) or electrode pairs
(hooked. 0.07 mm in diameter stainless-steel,
epoxy insulated) were implanted within the
pericardial cavity just posterior to the ventricle,
and in red and white muscles just under the lead-
ing edge of the second dorsal fin.
The white muscle thermocouple tip was placed
midway between the vertebral column and the
lateral edge of the body at the level of the horizon-
tal midline. Preliminary dissections confirmed
that red muscle in S. japonicus occurs in bands
'Reference to trade names does not imply endorsement by the
National Marine Fisheries Ser\ice. NOAA.
FISHERY BULLETIN; VOL. 76. NO. 4. 1979.
861
that are concentrated below the skin along the
lateral midline and become thicker posteriorly
(see also Kishinouye 1923, fig. 16; Braekkan 1959,
fig. 1). To ensure that the tip of the red muscle
thermocouple would remain in place, the wire was
passed from near the second dorsal fin obliquely
through white muscle and then into the thin red
muscle band. Once inserted, its position was easily
verified by gentle fingertip probing.
To facilitate positioning of the two muscle ther-
mocouples, 3-4 cm deep holes were tapped with a
20-gage hypodermic needle. The heart ther-
mocouple was passed into the pericardial cavity
through a 17-gage needle that was subsequently
withdrawn. All wires were anchored in place by
skin sutures. Wire leads ( 1 m long) to the recorder
were lap wound together, passed posteriorly, and
sutured to the dorsal midline near the finlets to
prevent tangling around the tail. Implanting re-
quired about 15 min after which the fish was
transferred to the respirometer swimming tube
where aerated water was circulated over the gills
by the driving impeller at a slow speed.
Two hours recovery from anesthesia and a brief
period of swim training was required before a fish
could maintain station in the tube and regulate
swimming speed in response to water flow. This
time delay also allowed stabilization of tissue
temperature at ambient conditions following
surgery.
Adaptation to the swimming chamber was car-
ried out at a basal swimming speed which is 1.5
BL/s (body lengths per second) for S. Japan icus
(Magnuson 1973). This speed is alsojustabove the
velocity required for sustained ram gill ventila-
tion(Roberts 1975). Flow rates in the respirometer
were calibrated with a ducted flowmeter (Marine
Advisors, Inc. model B-7C ) and controlled by alter-
ing the applied armature voltage to the impeller
pump motor. Eight fish were used for excess tem-
perature measurements and seven were used to
monitor EMG (4) and EGG (3) patterns.
Calibration Procedures
Thermocouples were made by soldering to-
gether the twisted bared tips of the copper and
constantan wires and sealing them with epoxy
cement. The three tissue thermocouples and a ref-
erence thermocouple (for respirometer water
temperature) were each connected in series (con-
stantan leads) to an ice-bath reference couple (0°G)
and to an RS Beckman Dynograph (copper leads)
through a high-quality, shorting rotary-switch.
This arrangement permitted rapid switching be-
tween thermocouples without opening the recorder
circuit. Thermocouples were standardized in a
water bath at 20°±0.05°G before and after each
trial.
Paired electrodes for recording ECG's and
EMG's were prepared and implanted (in the same
sites used for thermocouples) as described by
Roberts (1975). The EGG and EMG signals were
preamplified using high impedence, probe am-
plifiers (Grass, P511DR) to improve the frequency
response of the RS Dynograph.
Seawater was kept continuously flowing
through the respirometer tube and ambient tem-
perature was maintained within 2.0°G in each ex-
periment by mixing warm and cold seawater at
the outlet taps of the laboratory seawater system.
Over the 2-mo course of experiments, respirome-
ter temperatures ranged from 16° to 22°G.
Results
Changes in excess tissue temperatures that ac-
company increased swimming speed in the mack-
erel are best seen in a particularly successful trial
with fish number 6 (Figure 1). Similar, but some-
what variable records of heart, and red and white
muscle temperatures were obtained for all fish
(Table 1).
While cruising at low speeds, excess tempera-
tures reached a maximum of about 0.3°G in the red
and white muscles, but doubled within 3 min
swimming at enforced higher speeds (3.2-4.5
BL/s). Excess temperatures recorded in the heart
averaged about one-half of the excess developed in
muscles at all swimming velocities. When swim-
ming speeds were reduced once again to slow
cruising, excess temperatures returned to pre-
burst levels within 8-15 min.
During bouts of prolonged high-speed swim-
ming (5-6 min), water in the swimming tunnel was
warmed about 1°C due to frictional heating even
though a continuous exchange of seawater was
maintained from the supply tap (about 15 1/min).
This thermal error was minimized by rapidly ac-
celerating the fish from slow cruising to its pre-
determined, burst-swimming velocity. In Figure 1
for example, the fish was accelerated from 1.4 to
3.9 BL/s in about 5 s followed by sustained swim-
ming for 3 min, and then rapidly decelerated to 1.4
BL/s. Equilibration of tissue thermal excess (i.e.,
generation minus dissipation) occurred in most
862
white muscle
FK'iURE 1. — Temperature excess in the
heart and in red and white muscles re-
corded from Scomber japonicus no. 6(35
cm FL, 0.45 kg) swimming at speeds
from 1.4 to 3.9 BL/s. Arrows indicate
timing and direction of speed changes.
Ambient temperature, 19.5°-19.6°C.
r*! I I ' I I
400 600 1000
Table l. — Temperature excesses as AT (°C) recorded for seven Scomber japonicus swimming at basal and
moderately fast speeds in body lengths per second (BL/s).'
Fish number
Item
1
2
3
4
6
7
8
Mean
Fork length (cm)
35.6
39.2
389
38 1
350
36 3
34,3
368
Weight (kg)
0,54
0,62
059
0.55
045
0-58
039
0 53
Highest AT at basal speed
(1 3-1,9 BUS) in:
Red muscle
0.2
0.1
02
02
02
0,3
05
0,24
White muscle
0.2
0.3
03
02
04
03
0.3
0.29
Heart
0.0
0.3
0.1
0.1
0,2
02
{')
0.15
Highest AT and swimming
speeds (BLs) in:
Red muscle
0.9
0.75
0,55
0,85
03
0,5
0,8
0.66
(4.2)
(3.2)
(3 7)
(42)
(39)
(38)
(43)
(3,9)
White muscle
0.75
0.8
0,65
06
065
0,3
0,7
064
(4.2)
(3.2)
(3,7)
(3,9)
(39)
(3,8)
(4.3)
(3,9)
Heart
0.45
{')
{')
0,4
0,25
0,25
{')
0.34
(4.2)
(4,5)
(3,9)
(38)
(4.1)
Maximum trial speed (BL/s)
4.2
3,2
3,7
45
39
38
43
39
Water temperature, range (°C)^
16.1-17,0
16.5-17.0
16-8-17,8
17,1-17.5
19.5-196
20,5-21,0
21 1-21,8
'Fish no 5 omitted because it would not swim in the respirometer tube,
^Thermocouple malfunction
^Starting temperature is that of the seawater supply from mid-June to mid-July.
cases within the 3-min swimming bouts. Although
the thermal excess was greater in white muscle of
fish number 6 ( Figure 1 ) , mean maximum temper-
ature excesses recorded in red and white muscles
of the seven mackerel were about the same (Table
1).
Variability observed in excess temperature
measurements seems attributable to different per-
formances of individual fish. Some specimens had
more body fat than others and did not swim stead-
ily. Others were affected by the trailing ther-
mocouple cable as evidenced by their tail-beat pat-
terns. The cable also added drag which reduced
speed but probably increased total heat production
at a specific speed. None of the fish trailing ther-
mocouple cables could swim steadily above 5 BL/s,
whereas fish trailing the thinner EGG and EMG
cables could maintain a speed of 6 BL/s. Some of
the variability in recorded thermal excesses may
have also been due to the slightly differing loca-
tions of thermocouples in each fish. In addition,
trauma due to thermocouple insertion, which
probably interrupts normal blood flow locally may
have been a factor influencing thermal convection.
In a few cases, thermocouple signals changed
abruptly possibly because of insulation failure at
the tip due to rapid body flexing of fishes at higher
swimming speeds.
A wide range was found in heart rates of mack-
erel cruising at 1-1.5 BL/s (mean, 106; range,
80-140 beats/min). With acceleration to 4-5 BL/s,
the mean heart rate increased by 549f , (mean, 130;
range, 112-150), but rapidly returned to the rest-
ing rate within a few minutes of deceleration.
863
2iiri:^sr=L ~wi Ji^ EI H SL ~ Yisutb 2 .. A: sji^wsr
-csunf-rnog f^vfii. OBiPsr Ti?^^i^ ^)eerL been T^ fmr
-w^~g- TTtygf'^ IPSTE: a".. V— rii.-Tig' — •ar'r "rpn — Tv^ar ly-
T-eC Tr~-=.r":T-
■ 2
5£iIT5r^=iC ttt.-t-i r r-^t rr^ yrr.-j-
2d£ 4 3L = . "Whis iinsii& EM^'5 nf :iie fiTs: inaii-
irr in; xt xrtr "'^■grt^r skst bi "^ra^i m=- "^tst- smi'C
aSDuen: r^ji-nr^ toidiSE fnuL lirse icijsr^s: ^in
merr Tries. T£rj£3tQii= :x ±IM3 PTrrp-TTTirio;: gf
smEt
KjTLH T5S1 ?rrir «*t. i/f
irnBC±B^ HI £_ -Wt :TTir--T<£ S»S5E£ ■ ^T^ :i
lie: 1^ rsEor^ iar iij£ "n^r -r J^^nr^ Jl ir ~rm= ^ase-
ceari sna^ oirrss or :r=jc irnsr;^„ iJin; scirnnxiHiJi
nf ijESTciEST^ *rtr;'..-^Tr"'t^ ^t-SUS IE rsi- y^ir "Brrir
Tfc^ snaSj" 3ii?w5 t^MT Scr»"n;.r>e'
."! u 15 i ..c
'WHS ziD'. - _- ~~~.fr7r- r-^' - T ^ _-^. ^.-
T_&_ sorrresijvenr or ijt cdt
3£n X 1 -
TS23aiir "RXJ ~r~
nj TsacKE iuDCfd.. J~rsi- : z ' - . -
lOISliSE "ITDrLLli be £r>DJ=d 2£ it •n-iT^-qs v " " ■ :
rnr^^ if^^ -^^-^^
-''ill
is
Aj^ii^ ,}^J}}L U^J^JL :jJaV..V^
0.x ~,
iMA"'*** *M
' : 1
t
HI - • ft
1/ . . . i
t4in i;,,,,, .:im Ht}'^-
' 1 '^/^if I
BL%
2 0
2 5
3J
4>4
50
^0
5^
187
5tt vBiin* EJft'S^ S5ft»!t Anttrjeatv
5S
200
$&$
Wis- graerqjpgr rj^r me-
?tr~S IZ r^?fi ' _»-<" «Z- ij^ "^O'JJ"— iTl- jlf^
rr. 5v : : -.-;i^tL his
-iix ice^c ^^: ^ itctn -5 i±Ls ii_ ^^ »rea
id
c^rr
dtses -re:
Wt
-:cxuT
r^kAAA?^ ,t»i ••^-*^ ^
- ~- in tttos s£Uki> c^
" ~ - - - -"-ted wiuii tail o^3.is
• ery Sio* speed* . -:t 2 BL. s-
-
rni?
"
:.cl^
-^— rr.-"—!
^T^r .
-
. s-
rl^^iSt
SvCatr
litkr
.f^r--Tvr
^SK^
K i
«u nsc
.::.se rsT::*s? ,c 5^
r^ r*
r-r tfoe
5»55
For scombrids, which swim continuously and
rely upon forward motion to ventilate their gills,
the existence of a relatively high speed for the
division of labor between red and white muscles,
has been assumed primarily on the basis of work
done by Rayner and Keenan (1967). These inves-
tigators concluded that in the skipjack tuna, red
muscle alone powered cruise swimming and white
muscle only became active at burst velocities. The
initial objective of Rayner and Keenan's study was
to demonstrate contractile properties of red mus-
cle, and to this end they blocked white muscle
activity (pentobarbital) and worked exclusively
with tranquilized (propriopromazine) or sedated
fish. Moreover, their specimens were restrained in
a fixed position and artificially ventilated by per-
fusion tubes in the mouth. Thus the movements by
these skipjack tunas that were identified as "low
frequency swimming," were in fact only casually
related to the swimming requirements for gill
ventilation and hydrostatic equilibrium; both are
controlling factors in normal swimming (Magnu-
son 1973; Roberts 1975).
Our results with S. japonicus contrast in that
they show both red and white muscles function in
low-speed swimming. Also, Dizon and Brill (A. E.
Dizon, Southwest Fisheries Center Honolulu
Laboratory, National Marine Fisheries Service,
NOAA, Honolulu, HI 96812. Pers. commun., Sep-
tember 1977) recorded red and white EMG's from
yellowfin tuna, Thunnus albacares, and found
that white muscle activity begins at swimming
velocities of <3 BL/s — a speed only slightly above
the minimum for hydrostatic equilibrium and well
below maximal burst capabilities (Magnuson
1973). These observations indicate that in fast-
swimming scombrids, patterned staging of red and
white muscle activity may differ in that activity
begins in white fibers at very low speeds, and that
both red and white muscle remain active through-
out a wide range of sustainable speeds as well as at
burst velocities. Implicit in this idea is the pre-
sence of a high scope for aerobic activity in scom-
brid white muscle which has been recently dem-
onstrated for the skipjack tuna (Guppy et al. in
press). Also required by the hypothesis are
specializations in red muscle for high-speed con-
traction which is supported by the findings of
Johnston and Tota (1974) that high levels of
myofibrillar ATPase occur in the red muscle of
bluefin tuna, T. thynnus.
What physiological advantage might be gained
by a 1°C thermal excess during fast swimming?
Assuming a Qjo of 2 then a 10% increase in
metabolism would afford about a 2-3% rise in
swimming speed, but an insignificant change in
overall swimming efficiency (Webb 1971). An in-
teresting speculation is that the extensive heat-
exchanging vascular network used for en-
dothermy in the scombrids may have initially
evolved to meet the high oxygen requirements of
red and white myotomal muscle. More metabolic
heat is produced during aerobic respiration and
natural selection may have proceeded toward a
vascular design that maximized oxygen delivery,
yet augmented muscle function by conserving
heat and insulating the swimming musculature
from ambient conditions.
Acknowledgments
This work was conducted at the Southwest
Fisheries Center La Jolla Laboratory, National
Marine Fisheries Service (NMFS), NOAA where
John L. Roberts was supported by a NOAA Senior
Research Associateship. We thank J. R. Hunter,
R. Lasker, and G. D. Sharp for advice and many
stimulating discussions. H. T. Hammel of the
Scripps Institution of Oceanography advised us on
the preparation and use of thermocouples. Q. Bone
and J. R. Hunter critically read drafts of this paper
and made many suggestions. Useful technical ad-
vice and assistance were provided by J. Brown and
R. Leong of NMFS.
Literature Cited
Baldwin, F. M.
1923. Comparative rates of oxygen consumption in marine
forms. Proc. Iowa Acad. Sci. 30:173-180.
BlLINSKl, E.
1974. Biochemical aspects offish swimming. /nD.C.Ma-
lins and J. R. Sargent ( editors), Biochemical and biophysi-
cal perspectives in marine biology. Vol. 1, p. 239-288.
Academic Press, N.Y.
BONE, Q.
1966. On the function of the two types of myotomal muscle
fibre in elasmobranch fish. J. Mar. Biol. Assoc. U.K.
46:321-349.
1975. Muscular and energetic aspects of fish swim-
ming. In T. Y.-T. Wu, C. J. Brokaw, and C. Brennen
(editors), Swimming and flying in nature. Vol. 2, p. 493-
528. Plenum Press, N.Y.
BRAEKKAN, O. R.
1959. A comparative study of vitamins in the trunk mus-
cles of fishes. Fiskeridir. Skr. Ser. Teknol. Unders.
3(8): 1-42.
Carey, F. G., J. M. Teal, J. W. Kanwisher, K. D. Lawson,
AND J, S, Beckett,
1971. Warm-bodied fish. Am. Zool. 11:137-145.
866
George, J. C.
1962. A histophysiological study of the red and white mus-
cles of the mackerel. Am. Midi. Nat. 68:487-494.
GORDON, M. S.
1968. Oxygen consumption of red and white muscles from
tuna fishes. Science (Wash., D.C.) 159:87-90.
GRAHAM, J. B.
1973. Heat exchange in the black skipjack, and the blood-
gas relationship of warm-bodied fishes. Proc. Natl.
Acad. Sci. U.S.A. 70:1964-1967.
1975. Heat exchange in the yellowfin tuna, Thunnus alba-
cares, and skipjack tuna, Katsuwonus pelamis. and the
adaptive significance of elevated body temperatures in
scombrid fishes. Fish. Bull., U.S. 73:219-229.
Greer Walker, M.
1971. Effect of starvation and exercise on the skeletal
muscle fibres of the cod (Gadus morhua L.) and the coal
fish iGadus virens L.) respectively. J. Cons. 33:421-427.
Greer- Walker, M., and G. A. Pull.
1975. A survey of red and white muscle in marine fish. J.
Fish Biol. 7:295-300.
Guppy, M., W. C. Hulbert, and p. W. Hochachka.
In press. The tuna power plant and furnace. In G. D.
Sharp and A. E. Dizon (editors). The physiological ecology
of tunas. Academic Press, N.Y.
Hall, F. G.
1930. The ability of the common mackerel and certain
other marine fishes to remove dissolved oxygen from sea
water. Am. J. Physiol. 93:417-421.
HUDSON, R. C. L.
1973. On the function of the white muscles in teleosts at
intermediate swimming speeds. J. Exp. Biol. 58:509-
522.
HUGHES, G. M.
1966. The dimensions offish gills in relation to their func-
tion. J. Exp. Biol. 45:177-195.
JOHNSTON, I. A., AND B. TOTA.
1974. Myofibrillar ATPase in the various red and white
trunk muscles of the tunny iThunnus thynnus L.) and the
tub gurnard iTrigla lucerna L.). Comp. Biochem.
Physiol. 49B:367-373.
JOHNSTON, I. A., W. Davison, and G. Goldspink.
1977. Energy metabolism of carp swimming muscles. J.
Comp. Physiol. 114:203-216.
KISHINOUYE, K.
1923. Contributions to the comparative study of the so-
called scombroid fishes. J. Coll. Agric, Imp. Univ. Tokyo
8:293-475.
LINDSEY, C. C.
1968. Temperatures of red and white muscle in recently
caught marlin and other large tropical fish. J. Fish. Res.
Board Can. 25:1987-1992.
LINTHICUM, D. S., AND F. G. CAREY.
1972. Regulation of brain and eye temperatures by the
bluefin tuna. Comp. Biochem. Physiol. 43A:425-433.
MAGNUSON, J. J.
1973. Comparative study of adaptations for continuous
swimming and hydrostatic equilibrium of scombroid and
xiphoid fishes. Fish Bull., U.S. 71:337-356.
Rayner, M. D., and M. J. Keenan.
1967. Role of red and white muscles in the swimming of the
skipjack tuna. Nature (Lond.) 214:392-393.
Roberts, J. L.
1975. Active branchial and ram gill ventilation in fishes.
Biol. Bull. (Woods Hole) 148:85-105.
STEEN, J. B., AND T. BERG.
1966. The gills of two species of haemoglobin-free fishes
compared to those of other teleosts — with a note on severe
anaemia in an eel. Comp. Biochem. Physiol. 18:517-526.
STEVENS, E. D., AND F. E. J. FRY.
1971. Brain and muscle temperatures in ocean caught and
captive skipjack tuna. Comp. Biochem. Physiol.
38A:203-211.
1974. Heat transfer and body temperatures in non-
thermoregulatory teleosts. Can. J. Zool. 52:1137-1143.
STEVENS, E. D., AND A. M. SUTTERLIN.
1976. Heat transfer between fish and ambient water. J.
Exp. Biol. 65:131-145.
WEBB, P. W.
1971. The swimming energetics of trout. II. Oxygen con-
sumption and swimming efficiency. J. Exp. Biol.
55:521-540.
JOHN L. ROBERTS
Department of Zoology
University of Massachusetts
Amherst, MA 01003
Department of Zoology
San Diego State University
San Diego, CA 92182
JEFFREY B. Graham
THERMAL BEHAVIORAL RESPONSES OF
THE SPECKLED SANDDAB,
CITHARICHTHYS STIGMAEUS:
LABORATORY AND FIELD INVESTIGATIONS
The speckled sanddab, Citharichthys stigmaeus, is
a small bothid flatfish that is common in southern
CaliforniaiFord 1965; Stephens etal. 1974). These
authors and Helly' have suggested that tempera-
ture may have a significant effect on localized
population abundances and distributions of speck-
led sanddabs. No studies to date, however, have
examined in detail the relationship between
temperature and fish behavior and distribution.
We designed this work to study the speckled
sanddab population in King Harbor, Redondo
Beach, Calif. This harbor (Figure 1), which re-
ceives the thermal effluent from an electricity
generating station as well as cold upwelled water
from the adjacent Redondo Submarine Canyon,
contains a highly diversified thermal environ-
ment (Stephens 1972).
'Helly, J. J., Jr. 1974. The effects of temperature and tem-
perature selection on the seasonality of the bothid flatfish,
Citharichthys stigmaeus. Honors Thesis, Occidental Coll.,
Los Ang., 34 p.
FKSHERY BULLETIN: VOL. 76. NO 4. 1979.
867
Figure l. — Location of field sampling
stations for speckled sanddabs in King
Harbor, Redondo Beach, Calif
PACIFIC OCEAN
300 m
Methods
We collected adult speckled sanddabs between
August 1975 and January 1976 with a 3-m otter
trawl. The fish were transported to the laboratory
in aerated seawater and acclimated to a range of
normally occurring temperatures (10.0°-19.7°C)
according to the methods of Ehrlich et al. (1979).
Prior to acclimating the speckled sanddabs, we
removed the gill isopod Liuonica vulgaris indi-
vidually with forceps. During holding and accli-
mation, we fed the fish to satiation daily with live
and frozen Artemia salina. The behavioral re-
sponses of the fish to temperature were studied
using a 3.6-m long horizontal gradient and
employing the techniques of Ehrlich et al. (1978).
Each experiment lasted for 7-8 h with observa-
tions every 15 min. We shifted isotherm positions
during each experiment to separate selection of
temperature from preference for a given position
within the experimental chamber.
Speckled sanddab abundance and distribution
were studied using timed diver transects at six
stations (Figure 1). Two divers swimming side by
side for 5 min traversed each 6-m wide transect.
They recorded the species and number of indi-
viduals observed in the same area. The transects
by each pair of divers were run in duplicate on a
monthly basis at each station from September
1974 through February 1976 and quarterly there-
after. In the analyses, we used the largest number
of individual fish counted by either diver, but the
average count of the two independent observa-
tions was used for estimates of large groups of
fishes. The divers recorded the temperature at
least twice during each transect, with thermome-
ters readable to 0.5°C.
Results and Discussion
We examined the effects of acclimation temper-
ature, size, and sex of speckled sanddabs on their
temperature selection during 11 experiments (Ta-
ble 1). The presence of some skewed temperature-
specific frequency distributions (Table 1) pre-
cluded comparison of the results with parametric
statistics. We tested these distributions for homo-
geneity using a Kruskal-Wallis test (Steel and Tor-
TABLE 1. — Temperatures selected by speckled sanddabs in laboratory experiments.
No. test
No fish
Standard
lengtti (mm)
Acclimation
temperature
Selected
temperature
(°C)
Coefficent
of skGwness
Coefficient
of kurlosis
Date
animals
observations
Mean
SD
Sex
Mean
SD
Mode
(5,)
(92)
21 Aug. 1975
9
265
963
06
not noted
14.0
10.5
3.4
10
0.493-
2.788
15 Dec
8
230
91.5
1.3
not noted
10.0
123
4.5
9
0.543-
2.824
18 Dec.
6
218
88.5
3.2
M
19.7
10.5
4.1
8
0673-
2.928
22 Dec.
9
220
903
3.5
F
18.9
10 1
26
9-10
0427
3.438
5 Jan. 1976
9
210
77.1
4,0
M
15.2
109
2.6
11
0,220
3.512
8 Jan.
9
251
82.0
6.7
F
15.2
10.4
3.2
8-9
0,465
3.036
9 Jan.
9
250
82.0
67
F
15.2
11.6
3.6
8
0.573-
2,673
26 Jan.
6
162
763
5,6
M
12.0
9.9
4.3
8
0400
0.592-
27 Jan.
6
167
71.5
27
M
12.0
10.5
2,5
9-10
0.291
2.372
29 Jan.
6
157
73.2
2.6
M
120
11,1
30
9
0-724-
2.538
30 Jan.
6
160
75.5
5.1
M
12,0
98
2.5
8
0.422
2.078
P<0.05.
868
rie 1960) and detected no significant differences
(X^odf = 15.99, P>0.05). The overall mean
selected temperature from pooled data was 10.8°C
(SD = 3.1°C), and the mode was 9°C; 709^ of all
occurrences were in the range of 8°-13°C (Figure
2). The frequency distribution, however, was sig-
nificantly skewed towards warmer temperatures
(Figure 2).
Brett (1971) showed that the preferred tempera-
ture coincided with the optimal temperature for
growth of sockeye salmon, Oncorhynchus nerka.
Crawshaw (1977) found that physiological re-
sponses are often optimized in a zone of efficient
operation rather than at a peak, and temperature
preference reflects this. We are not aware of any
work on the effects of temperature on the physiol-
ogy of speckled sanddabs.
Considering the work of Brett (1971) and Craw-
shaw ( 1977), it is not unreasonable to suspect that
speckled sanddabs may have a range of tempera-
tures (approximately 8°-13°C) for efficient growth.
The skewness may partially be due to activity
increasing with temperature that could have re-
sulted in occasional excursions into a greater
number of compartments (temperatures) than at
colder temperatures. Ehrlich et al. (1978) also
suggested that skewness of a temperature-specific
frequency distribution could result from preferred
temperatures approaching lethal limits. These
limits are not known for speckled sanddabs. De-
Witt (1967) suggested that skewness of distribu-
tion could result from the regulation of body tem-
15-1
(ij
o
z
UJ
(T lo-
ir
O
o
o
o
5-
■f-f
T — I — I — I T T T T T T — r — I — 7~~l — I — I — I — I
4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 202!
TEMPERATURE CO
Figure 2. — Temperature-specific occurrences of speckled
sanddabs, based on pooled data from 2,290 fish observations. The
frequency distribution was significantly skewed toward warmer
temperatures (^, = 0.571, fgsdf = 2.33, 0.0180% of the total, a finding consistent
with other mercury values reported (Kamps et al.
1972; Westoo 1973). In muscle and liver tissues of
blue marlin, Makaira nigricans Lacepede, how-
ever, only a small portion of the total mercury was
found to be organic mercury. Additional studies on
marlin landed during fishing tournaments in 1972
(Schultz et al. 1976) and 1973 (Schultz and Crear
1976) revealed low levels of organic mercury in six
other tissues. These studies also showed that the
difference between total and organic mercury was
indeed inorganic mercury. G. Westoo (National
Swedish Food Administration, Stockholm. Pers.
commun., 1972) had previously identified the or-
ganic fraction as methyl mercury.
An assessment of mercury is complicated by the
presence of selenium. Selenium has been shown to
reduce the toxicity of mercuric chloride and
methyl mercury in laboratory animals when given
as selenite, selenomethionine, or as selenium pre-
sent in tuna (Pah'zek et al. 1971; Ganther and
Sunde 1974). The presence of selenium in tuna, a
principal food item of marlin (NaughtonM, indi-
cates that it should also be present in marlin.
For this report, nine tissues from blue marlin
were analyzed for selenium, total mercury, and
organic mercury.
Materials and Methods
Samples of muscle, liver, kidney, spleen, pyloric
caecum, stomach, gill, gonad, and blood were col-
lected from 46 marlin landed during a fishing
tournament in Kailua-Kona, Hawaii, during Au-
gust 1974. The tissues were ground with Dry Ice^
in a blender and stored in acid-washed plastic
vials.
The organic extraction was carried out as de-
scribed by Rivers et al. (1972), i.e., a benzene ex-
traction of the methyl mercury was reextracted
with cysteine, oxidized with permanganate, and
reduced to elemental mercury with stannous ion
prior to being volatilized into the flameless atomic
absorption apparatus. Total mercury digestions
were performed (Rivers et al. 1972) but with 10 ml
of concentrated nitric acid instead of 30 ml. All
analyses were made with a Perkin-Elmer 303
atomic absorption spectrophotometer equipped
with a vapor chamber (Manning 1970).
Selenium was determined by a fluorometric
technique (Watkinson 1966), as modified by S.
Nishigake (Tokyo Metropolitan Research Labo-
ratory of Public Health, Tokyo, Japan. Pers. com-
mun., 1975), i.e., following sample digestion with
nitric and perchloric acids, the selenium was com-
plexed with 2,3-diaminonaphthalene and this
fluorescent compound then extracted into cyclo-
hexane. All analyses were made using a Turner
Model 110 fluorometer equipped with a primary
filter at 369 nm and a secondary filter at 522 nm.
'Naughton, J. J. 1973. To all billfishermen. (Summary report
of 15th Hawaiian International Billfish Tournament, 27-31 Au-
gust 1973), 9 p. Southwest Fisheries Center Honolulu Laborato-
ry, National Marine Fisheries Service, NOAA, Honolulu, HI
96812.
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
872
FISHERY BULLETIN: VOL. 76, NO. 4, 1979.
Results
A summary of mercury and selenium concentra-
tions in the marlin is given in Table 1. Average
total mercury and selenium values were greatest
in kidney (26.33 mg/kg Hg, 23.42 mg/kg Se) and
least in blood (0.18 mg/kg Hg, 1.29 mg/kg Se) and
gill (0.32 mg/kg Hg, 1.29 mg/kg Se). Average
methyl mercury was highest in muscle (0.40)
mg/kg) and lowest in blood (0.04 mg/kg) and gill
(0.06 mg/kg). The percentage of organic to total
mercury ranged from 19f in kidney to 279f in
gonad. The molar ratio of mercury to selenium
ranged from 0.06 in blood to 0.62 in muscle. (Molar
ratio is computed here using sample average
statistics of combined male and female data.)
An analysis of variance revealed significant dif-
ferences (P<0.05) in total mercury between males
and females for all tissues except gill and blood. A
similar pattern was found for selenium. The or-
ganic mercury levels were not statistically differ-
ent (P>0.05) between the sexes. In earlier studies
on marlin caught from the same area during
fishing tournaments in 1971, 1972, and 1973, mer-
cury concentrations in both sexes were found to be
similar.
Table 2 presents correlation coefficients for body
weight and tissues. In most cases, the relation-
ships are positive and highly significant. Figures
1-6 illustrate the dependence of mercury and
selenium on weight and on each other.
Discussion
The data clearly demonstrate that methyl mer-
cury concentrations are low relative to total mer-
cury in blue marlin. Westoo (pers. commun.) and
Nishigaki (pers. commun.) confirmed this low
percentage in our samples based on subsamples
sent to them. Nishigaki has also confirmed our
selenium results.
In a study of 37 Pacific blue marlin (50-238 kg,
average 109 kg) from Japanese waters, Nishigaki
(pers. commun.) found total mercury levels rang-
ing from 0.02 to 13.0 mg/kg (average 2.83 mg/kg)
and methyl mercury ranging from 0.02 to 1.28
mg/kg (average 0.57 mg/kg). Selenium values for
11 marlin ranged from 0.52 to 1.99 mg/kg, averag-
ing 0.97 mg/kg. These values are similar to our
findings for mercury and selenium in marlin from
Hawaiian waters.
Table l. — Summary of mercury and selenium data in 46 blue marlin from the Hawaiian Islands by sex.'
Total Hg (mg kg) Organic Hg (mg, kg) o„ ^.^a^.r Hg Se (mg/kg)
Hg/Se
Mean Range Mean Range of total Hg Mean Range
ratio
Tissue
M&F M F M&F M&F M F M&F M&F M&F M F M&F
M&F
Muscle
Liver
Kidney
Spleen
Stomach
F>ylonc
caecum
Gill
Gonad
Blood
3.12
11.58
26 33
693
1.27
1 83
0.32
0.40
0.18
2.83
12.53
2625
6.34
1 33
1.74
0.26
030
0.18
3 76
9.46
26.53
821
1,16
2.03
0.44
0.68
0.18
0.09-10 00
0.13-39 20
0.18-77.00
0 08-17 60
0.06-3.00
0.17-4.50
008-096
0.03-2.15
0.02-0.53
040
026
0 22
0 18
0.12
030
006
0.11
0.04
032
0 21
0 16
0 12
0.08
022
003
006
004
0.58
038
0.35
030
021
0.49
0.11
025
0.05
0.02-1.02
009-0.76
004-086
002-0.66
003-0.47
008-0.98
0.01-0.34
0.03-0.65
;0.01-0.11
13
2
1
3
9
16
19
27
22
198 1.88 2.21 0.63-5.32
17 47 20 36 1106 2 50-61 12
23.42 2533 1906 2 63-56 25
9.31 912 973 0 63-24 25
291 316 239 135-4 03
4 92
1.29
2.14
1.29
5.10
1.31
1.97
1.33
4 52
1.26
2.60
1.19
2.28-10.10
0.71-2.21
1 .24-3.80
0.72-2.30
0.62
026
0.44
0.29
0.17
0.15
0.10
0.07
0.06
'Weights of 32 males ranged from 58 to 112 kg (average 80); 14 females weighed 39. 75, and 115-342 kg (average 166) Average for all samples was 106 kg.
Table 2. — Correlation coefficient of mercury, selenium, and weight in 46 blue marlin from the
Hawaiian Islands by sex.
Total Hg/wt
Organic
Hg/wt
Organic
total Hg
Sewt
Total
Hg/Se
Organic
Hg/Se
Tissue
M&F
M
F
M&F
M&F
M&F
M
F
M&F
M&F
Muscle
069"
090"
083"
0.79"
072"
068"
0.88"
0.85"
0.93"
061"
Liver
023
076"
069"
0.76"
0 13
0.00
0.68"
0.38
0.80"
-008
Kidney
0.49"
083"
079"
087"
042"
032-
0,79"
0.73"
0 91"
0.30-
Spleen
06r-
088"
075"
0,84"
0.51"
0.45"
072"
0.61-
0.87"
0.41"
Stomach
0.38-
0.77"
082"
0.81"
0.35-
-0.27
0 18
0.15
0.32-
-033-
Pylonc
caecum
0.59"
083"
082"
086"
059"
-0.11
032
-0.10
025
-0.14
Gill
086"
070"
090"
0.88"
0.77"
004
0.08
016
021
-0.14
Gonad
085"
0.75"
0.95"
082"
0.76"
015
-0.10
-0.53
010
0.25
Blood
0.28
0.32
057-
0.36-
0.44"
-004
-0.17
049
0.44"
009
•P<0.05.
••P<0.01.
873
12
>-
q:
O
II -
10
o«-^rs"oi'--i'»!r"'
150 200
WEIGHT (kg)
350
3
50
100
150 200
WEIGHT (kg)
250
350
874
12
Fk;L'RK 1— The relationship
of total and organic mercury
in muscle tissue of 46 blue
marlm from the Hawaiian Is-
lands to fish weight.
10
FIGURE 3— Relationship be-
tween total mercury and
selenium in muscle tissue of 46
blue marlin from the Hawaiian
Islands.
a:
3
O
ir
UJ
g
a FEMALES
X MALES
2 3 4
SELENIUM mg/kg (ppm)
FIGURE 2.— Relationship be-
tween selenium in muscle tis-
sue and weight of 46 blue
marlin from the Hawaiian Is-
lands.
FIGURE 4.— Relationship be-
tween organic mercury and
selenium in muscle tissue of 46
blue marlin from the Hawaiian
Islands.
1.2
II
10
i 8
ir
O
(T
UJ
S
(J
z
<
l PCB. (Incompletely washed
glassware and new batches of reagents are com-
mon sources of high blanks.)
Recovery
Carry out the complete procedure with solutions
of standards of known concentration in the range
anticipated for the samples to assure quantitative
recovery. (Some loss of chlorinated hydrocarbons
always occurs in the absence of proteins and lipids,
which act as keepers. A minimum recovery of 80^f
is essential. Doping samples originally containing
very low levels of chlorinated hydrocarbons gives
results which better reflect the accuracy of the
method: 859'f or higher recovery of DDE, TDE,
DDT, and PCB, and 80-85'7r recovery of dieldrin
and endrin.)
Extraction
1) Weigh approximately 10 g of the material
(see Procedure Variations for exceptions), to be
analyzed into a Virtis flask and record exact
weight to the desired degree of accuracy.
2) Add 20 ml of a 1:1 IPA/benzene mixture to
the flask.
3) Homogenize at about 23,000 rpm for 5 min.
4 ) Rinse homogenizer blade and Teflon cap with
hexane so that the hexane drips into the Virtis
flask. Fill the flask with hexane nearly to the bot-
tom of the flask neck.
5) Place the Virtis flask in a hot-water bath (ca.
85°C) or sand bath. (An electric fry pan with a
layer of sand covered with water provides an
economical heating bath.) Boil moderately for at
least 45 min, adding hexane whenever the level
falls to about one-third of the flask capacity. (If the
solution boils too rapidly, material will be lost in
the spray. The rate of boiling must be adequate to
distill all the H^O, IPA, and benzene from the
flask, because they interfere with the cleanup.)
When adding hexane, do it so as to rinse down the
sides of the flask as well. Keep the water level in
the bath and the hexane level in the flask adjusted
so that the flask does not become buoyant and tip
over. After 45 min of boildown, reduce the vol-
ume of the solution to about 20 ml (ca. 1 cm from
the bottom of the flask). Cool. If a layer of water
separates, add Na.^ SO^ and allow to stand 1-3 min.
6) Filter through a funnel plugged with glass
wool into a 50-ml graduated centrifuge tube. Rinse
the flask with three or four 5-ml portions of hexane
and pour the rinse solutions through the filter into
the centrifuge tube.
7) Concentrate the extract in a hot- water bath
to desired volume (20-25 ml), record volume, and
pour most of extract into a 23-ml borosilicate
screw-cap bottle (with Teflon-lined cap) contain-
ing 1 g Na^SOj. (The extract can be stored in this
condition for extended periods.) (If the extract,
prior to concentration, still contains traces of H2O
or IPA, as indicated by cloudiness, add hexane
while concentrating in order to remove the H2O or
IPA.)
Oil Determination
8) Pipette exactly 1 ml of the extract into a
tared aluminum weighing dish and allow it to
882
evaporate 4-6 h at room temperature to minimum
weight. (Since marine oils oxidize, the weight of'oil
begins to rise again after a few hours.) Weigh the
residue, which contains the oil in 1 ml of extract.
(,"lcaiuip
9) Prepare.' a Flurisil (.olumn by filling a 9 mm
i.d. X 150 mm glass tube, plugged with glass wool,
with ca. 5 cm of Florisil. Wash the column with at
least 15 ml of hexane added 1 ml at a time. Allow
the hexane level to drop to 1-2 mm, but not to
dryness. Pipette 1 ml of the extract onto the col-
umn. For samples with very high oil content (refer
to step 8 for the amount of oil in 1 ml of the ex-
tract), adjust the volume placed on the Florisil so
that no more than 0.1 g, and preferably no more
than 0.08 g, of oil is placed on the Florisil. Elute
with 1-ml portions of hexane; collect the first 12-13
ml of the eluate in a 13-ml graduated centrifuge
tube. I Note: Once the Florisil has been wetted, it
must always have solvent above it.)
10) If DDT and PCB are not going to be sepa-
rated, concentrate (in a tube heater) the eluate to
an appropriate volume for gas-liquid chromato-
graphic analysis."* If separating DDT and PCB.
evaporate the eluate to slightly less than 1 ml.
Separation of DDE, TDE, and DDT from PCB
Quantitation of TDE and DDT is often difficult
and quantitation of the PCB is usually impossible
unless the DDT family is separated from the PCB.
Separation is achieved by chromatography on
silica gel. The behavior of DDT and PCB during
solid-liquid chromatography is very similar, and
obtaining optimal separation requires careful con-
trol of all the parameters of the procedure. Even
so, DDE does not separate entirely from PCB.
Therefore, the DDE in the PCB fraction is quanti-
tated'' and included with that in the DDT fraction.
Evaluate the degree of separation of DDE, TDE,
and DDT from PCB by chromatographing stan-
dard solutions of these compounds according to the
procedure described below. Adjust the time and
"•For a detailed de.scription of gas chromatography of chlori-
nated hydrocarbon pollutants, see the Pesticide Analytical
Manual il977), available from Management Methods Branch,
DMS, ACA, HFA-250, 5600 Fishers Lane, Rockville. MD 20857.
or National Technical Information Service iNTIS), Springfield.
Va. The manual provides extensive background on residue
analysis.
•'^Although DDE elutes from the gas-liquid chromatograph at
the same time as one of the PCB peaks, measurement of the other
five intense PCB peaks provides accurate quantitation of PCB.
temperature of activation, the degree of rehydra-
tion, the amount of sil ica gel, and the volume of the
pentane fraction to obtain the optimum separation
of DDT and TDE from PCB: that is, to maximize
the amount of PCB in the pentane fraction and the
amount of DDT and TDE in the benzene fraction.
1 1 ) Activate the silica gel by heating at 215°C
for 16 h. (The time and the temperature are ad-
justed to obtain an arbitrarily, but consistently
activated product with suitable separating
characteristics, since complete dehydration occurs
over a long period of time.) Cool to room tempera-
ture in a desiccator. Rehydrate by placing 98 g
silica gel in a glass-stoppered bottle and adding 2 g
distilled water. Stopper the bottle and shake and
tumble until the water is evenly distributed.
Allow the silica gel to equilibrate for 2-4 h before
use.
12) Place a portion of this prepared silica gel in
a beaker and cover with pentane. Let stand 5-10
min to return to room temperature. (Because the
chromatographic columns do not contain stop-
cocks, a special technique is required for packing.)
Quickly transfer silica gel to a glass-wool-
stoppered column, 9 mm i.d. x 250 mm long, wet
with pentane. (A disposable transfer pipette with
the narrow part of the tip removed works quite
well for transferring the silica gel slurry. ) Tap the
column gently to facilitate packing. Make sure
that there is always enough pentane above the
column to allow the silica gel to settle slowly in
order to eliminate air bubbles and prevent the top
of the column from running dry. Pack the column
to a height of ca. 8 cm. (Throughout the whole
separation procedure the silica gel must always
have .solvent above it and must be free of bubbles
and cracks, which interfere with the desired sep-
aration. If the column runs dry or cracks, discard
it.) Rinse the column with 15-20 ml of pentane.
13) Allow the pentane level to descend to 1-2
mm (not dry) and place the Florisil eluate (ca. 1
ml) on the column with a Pasteur capillary
pipette. Rinse the Florisil eluate tube with three
or four 1-ml portions of pentane, and transfer each
rinse successively to the column. After the sample
and rinses have been adsorbed, fill the tube with
Use the amount of PCB in those peaks to determine the size of the
peak overlapping the DDE and correct the apparent total DDE
I actually DDE plus PCB) to obtain the tme DDE concentration.
The electron-capture detector is so much more sensitive to DDE
than PCB, that the correction affects the accuracy of DDE de-
temiination only to a small extent. Consequently the variability
in amount of DDE in the pentane fraction does not markedly
affect the accuracy of DDE analysis.
883
pentane. Collect 42 ml of pentaneeluateina 50-ml
graduated centrifuge tube. (This fraction contains
PCB and some DDE.)
14) After the appropriate volume of pentane
eluate has been collected, place a second 50-ml
graduated centrifuge tube under the silica gel col-
umn. Then fill the tube with benzene. Collect 35
ml of benzene eluate. (This fraction contains most
of the DDT complex.) (DDE elutes very rapidly
with benzene. If benzene is added to the column
before the second centrifuge tube is in place, the
DDT complex will often be found in the pentane
fraction.)
15) Concentrate each fraction in a boiling
water bath to less than the desired final volume
and quantitatively transfer with hexane to a vol-
umetric flask of the desired final volume. (The
50-ml centrifuge tubes are not very accurate vol-
umetiic containers.) Proceed with gas-lic|uid
chi-omatographic analysis (see footnote 4).
Notes on DDT/PCB Separation Proccdiirc
1) Elution with hexane instead of pentane dur-
ing the silica gel chromatography fails to provide
the necessary separation of DDT from PCB.
Hexane is reported to contain variable amounts of
benzene, which would obviously affect an already
delicate separation. Use of UV-quality pentane or
hexane has been recommended by others, and
might allow use of hexane in warm weather.
2) For high residue level samples, evaporation
of the Florisil eluate to ca. 1 ml is not necessary;
instead an appropriate aliquot is used. However,
no more than 1 ml of the eluate should be placed on
the silica gel column because the hexane may con-
tain benzene.
Procediirt' Variations
During the 45-min boildown, scrape the mate-
rial on the bottom of the flask. Pile up the solids to
leave areas of the flask bottom in direct contact
with the solvent to improve boiling action and
prevent bumping. Cool. If water separates, add
Na.SO,.
Filter through glass wool and proceed as usual
(step 6 under Extraction).
In oi'der to compensate for the low residue level
usually found in plankton, place 2 ml of the extract
on the Floi'isil column for cleanup.
I'islimcals or Dr\ leeds
If the standard extraction procedure is used for
meals and animal feeds, the finely ground meal
forms a layer on the bottom of the Virtis flask
which causes bumping and loss of solvent during
the 45-min boildown. Extraction with hexane pro-
vides as good recovery as IPA/benzene. This sub-
stitution allows omission of the boildown.
1) Homogenize sample with 20 ml of hexane.
Wash down Virtis blades and Teflon top with min-
imal amount of hexane. Add 10 g Na^SO^.
2) Immediately filter through glass wool
tightly wadded to remove as much of the solids as
possible. Wash flask and funnel with a minimal
amount of hexane so that the volume of the cen-
trifuge tube is not exceeded (35-40 ml).
3) Stir to mix and centrifuge at 1,500 rpm for
45-60 min at 10°C. (There should be about 35 ml of
clear solution with less than 1 ml of solids.)
4) Record volume, subtracting 50% of the vol-
ume occupied by the solids. (Although this in-
volves an approximation, the error involved
should be no more than 2^/c.)
5) Decant the supernatant liquid into a storage
bottle containing 1 g Na.,S04. Proceed with
Florisil cleanup (step 9 under Cleanup).
Plankton and Other High-Moistiire Samples
Plankton do not homogenize well using the
standard procedure. They also contain water in
excess of the amount that 20 ml of 1: 1 IPA/benzene
can absorb. Addition of Na.2S04 before homogeni-
zation overcomes both difficulties. Na2S04 not
only absorbs water, but also serves as a grinding
aid.
To the weighed sample of plankton add 25 ml of
1 : 1 IPA benzene; then add 25 ml hexane and 10-15
g Na^SOj. Homogenize 5 min and proceed as
usual.
Fish Oil
Homogenization and extraction of oil are un-
necessary.
1) On an analytical balance weigh accurately
about 2 g of oil into a 50-ml graduated centrifuge
tube.
2) Dilute to about 20 ml with hexane and swirl
to dissolve the oil completely.
3) Record the volume.
4) Place in a storage bottle, containing 1 g
Na., SO4 , and proceed with the usual cleanup ( step
9 under Cleanup).
884
NOTE: For greater accuracy in oil analysis, weigh
accurately about 2.0-2.5 g oil into a 25-ml vol-
umetric flask. Dilute to volume with hexane.
Shake thoroughly. Place sample in storage bottle
and proceed at step 9 under Cleanup.
Paper
This procedure is included because occasionally
fishery samples come in contact with contami-
nated packaging and labelling materials, such as
carbonless carbon paper and cardboard. Although
the procedure has not been validated by collabora-
tive studies, it provides guidelines for an analysis
relevant to fishery studies.
Cut the paper (or cardboard* into small pieces,
approximately 1 cm on a side, with a scissors or
office paper cutter, which has been cleaned
thoroughly with iso-octane or hexane. Mix the
paper thoroughly and weigh ca. 7 g into a Virtis
flask. Add 15 ml distilled water and mix
thoroughly with the paper. Allow the mixture to
stand 2-5 min, stir again, and dilute with portions
of 1:1 IPA/benzene to a total of 70 ml. Homogenize
the mixture briefly at low speed. Push the paper
down with the Virtis blades, then homogenize
briefly. Repeat the process until the paper is com-
pletely homogeneous, approximately 10 min total
homogenization time. Follow the usual boildown
and Florisil procedures.
Proccdiiri.' For Dieldrin and Kndrin
Saponification'' and Extraction
1) Weigh 10 g of material to be analyzed into a
250-ml Erlenmeyer flask. For oil, use only 2 g.
2) Dissolve 10 g of KOH in 6 ml distilled water.
Slowly dilute with 34 ml of ethyl alcohol (95 or
100%). Swirl until clear.
3) Pour the alcoholic KOH over the sample and
heat in a water bath without boiling for 20 min;
the exact temperature is not critical.
4) Allow the mixture to cool. Pour the liquid
portion into a 250-ml separatory funnel and rinse
out the Erlenmeyer flask with 50 ml of water,
divided into 4 or 5 portions. Avoid pouring any
solids into the separatory funnel. For finely pow-
dered samples like meal, filter the sample through
^Saponification in the pre.sence of some proteinaceous mate-
rials has been reported to cause degradation of dieldrin. As
stated above, recovery studies are always important.
a wad of glass wool and rinse the glass wool care-
fully with each rinse.
5) Add 15 ml hexane to the separatory funnel
and shake for 2 min. Open the stopcock several
times during shaking to relieve the pressure
buildup. Allow the layers to separate completely,
usually about 30 min.
6) Drain off the aqueous layer into the Erlen-
meyer flask from which it came. Pour the hexane
layer into a 30-ml beaker. Do not let any water
escape into the beaker. Cover the beaker tightly
with aluminum foil.
7) Pour the aqueous layer back into the
separatory funnel and repeat step 5.
8) Drain off the aqueous layer and discard it.
9) Return the hexane extract in the beaker to
the separatory funnel. Rinse the sides of the
beaker with 1 ml hexane. Add the hexane wash to
the extract in the separatory funnel.
10) Wash the hexane extract with 10 ml water
by rotating the separatory funnel gently to avoid
emulsion formation. Do not shake. Allow the
layers to separate and discard the aqueous layer.
11) Pour the hexane layer into a 50-ml
graduated cylinder. Do not transfer any water to
the cylinder. Record the volume. Pour the extract
into a 23-ml borosilicate screw-cap bottle (with
Teflon-lined cap) containing 1 g NagSO^.
Cleanup
12) Prepare a Florisil column by filling a 9 mm
i.d. X 150 mm glass tube, plugged with glass wool,
with ca. 4 cm of Florisil. Or use a 7 mm i.d. x 150
mm tube containing ca. 5 cm of Florisil. (The
longer column of adsorbent may give slightly bet-
ter cleanup.) Wet the Florisil with benzene and
wash it with 4 to 5 ml benzene added 1 ml at a time.
Wash it next with 10 ml hexane (or more) added 1
ml at a time. Pipette 2 ml of the hexane extract
onto the column. Put a 12-ml graduated centrifuge
tube under the column. Flute with hexane added 1
ml at a time. Collect the first 12 ml of hexane
eluate, which contains DDMU (the dehydro-
chlorination product of TDE), DDE, and PCB.
(They may be quantitated if desired.)
Place a second 12-ml centrifuge tube under the
column. Change eluant to benzene; add it 1 ml at a
time. Collect the first 10 ml of eluate. Place a
modified micro Snyder column on the centrifuge
tube to prevent loss of residues during evapora-
tion. Concentrate the eluate containing dieldrin
and endrin to an appropriate volume (1-3 ml) for
885
gas-liquid chromatographic analysis (see footnote
4).
Ac know Icclgments
We thank the following laboratories for helping
to establish the validity of these methods by par-
ticipating in verification studies: Seattle, Detroit,
and New Orleans District Laboratories of the U.S.
Food and Drug Administration; the Gulf Breeze,
Fla., Field Station of the Environmental Protec-
tion Agency; the Wisconsin Alumni Research
Fund Institute, Madison, Wis.; and the Iowa De-
partment of Agriculture, State Chemical Labora-
tory, Des Moines.
Robert Reinert willingly shared his methods
with us before they were published. Daniel B.
Menzel, formerly of the Institute of Marine Re-
sources, University of California, Berkeley, Calif.,
generously shared his knowledge of the intricacies
of chlorinated hydrocarbon analysis. Laura G.
Lewis provided technical support in developing
this procedure.
Literature Cited
Horowitz, W. (editor).
1970. Fat-containing foods. /n Official methods of analysis
of the Association of Official Analytical Chemists. 11th
ed., p. 480. Assoc. OfT. Anal. Chem., Wash., D.C.
Pksticide Analytical Manual.
1977. Gas-liquid chromatography. In Pesticide analytical
manual. Vol. I, Chapter 3. U.S. Dep. Health, Educ, and
Welfare, Food and Drug Admin., Wash., D.C.
PORTKR, M. L., S. J. V. Young, and J. A. Burke
1970. A method for the analysis offish, animal, and poul-
try tissue for chlorinated pesticide residues. J. Assoc. Off.
Anal. Chem. 53:1300-1303.
Reinert. R. e.
1970. Pesticide concentrations in Great Lakes fish. Pestic.
Monit. J. 3:233-240.
Snyder. D., and R. reinert
1971. Rapid separation of polychlorinated biphenyls from
DDT and its analogues on silica gel. Bull. Environ. Con-
tam. Toxicol. 6:385-390.
Virginia F. Stout
Northwest and Alaska Fisheries Center
National Marine Fisheries Service, NOAA
2725 Mont lake Boulevard East
Seattle. WA 98112
F. Lee Beezhold
Northwest and Alaska Fisheries Center
Present address: Food Chemical and Research Laboratories. Inc.
4900-9th N.W., Seattle, WA 98107
GROWTH OF JUVENILE SPOT PRAWN,
PANDALUS PLATYCEROS, IN
THE LABORATORY AND IN NET PENS
USING DIFFERENT DIETS
Floating net pens have been used to culture Pacific
salmon, genus Oncorhynchu.'^. in the marine wa-
ters of the West Coast since 1969 ( Mahnken 1975).
Although it has been a monoculture effort to date,
use of a companion crop species such as the spot
prawn, Pandalus platyceros Brandt, could diver-
sify and enhance this industry.
In 1975 the National Marine Fisheries Service
selected the spot prawn to examine as a potential
companion species to net pen-reared salmon. The
spot prawn was selected as a candidate for several
reasons; 1 ) it has a rapid growth rate and large size
compared with other pandalids (Butler 1964); 2) it
can be successfully cultured to maturity in captiv-
ity (Prentice 1975); 3) it will reproduce in captiv-
ity, often for two consecutive years (Rensel and
Prentice 1977); 4) it is gregarious and is normally
not cannibalistic; 5) it adapts to vertical or hori-
zontal substrates; and 6) it scavenges for, and ac-
cepts, a wide variety of foods (Wickins 1972).
Coincident to investigating the prawn as a com-
panion crop to salmon, several prawn diets were
evaluated with prawns held in tanks and net pens
at the NMFS Aquaculture Experiment Station on
Puget Sound near Manchester, Wash. These ex-
periments wer^ conducted using diets made up of
underutilized marine species or fishery byprod-
ucts that are available to most salmon farmers.
Materials and Methods
The spot prawns used in the experiments were
laboratory-reared progeny of females captured in
Hood Canal, Wash. Three concurrent experiments
with juvenile prawns ( <1 yr of age) began 10 July
1975 (Table 1).
Experiment A was conducted in the laboratory
where prawns were held in flowing seawater tanks
at 110 animals/m^ of immersed substrate. Four
diets were evaluated: 1) steamed mussel, Mytilus
ediilis. meat; 2) chopped salmon that had died in
nearby net pens; 3) feces and pseudofeces from the
Pacific oyster, Ci-assostrea iii^as (eight oysters per
replicate having a mean weight (total) of 153 g);
and 4) no food (control). Diets 1 and 2 were fed
every other day while diet 3 was always present in
varying amounts. A sample of 10 prawns for each
of four replicates was measured during each of
886
fishery BULLETIN: VOL 76, NO, 4, 1979.
Table 1. — Growth and survival of juvenile Pandal us platyceros during three tests.
Location
Diet
No of
replicates
Start of experiment
No of prawns Mean
in eactn weight
replicate (g)
60 days after
start of experiment
Mean Mean
survival weight
(°°) (g)
c
nd of experiment
Experiment
No
of
days
Mean
survival
(°o)
Mean
weight
(g)
A (prawns alone)
Laboratory
tanks
Mussel
Salmon
Oyster wastes
No food
4
4
4
4
25
25
25
25
0.72
0.62
069
0.64
82
83
64
26
252
2.27
1.06
1 07
90
90
(')
(')
74
71
(')
(')
3.14
2.61
(')
(')
B (prawns alone)
Net pens
Mussel
Salmon^
2
2
200
200
0.64
063
98
98
3,14
279
365
365
69
64
10.30
10.61
C (prawns and
salmon)
Net pens
Variety of
feeds^
1
100
064
93
3,42
206
93
860
'Terminated at 60 days
^Includes net-fouling organisms.
^Includes salmon mortalities, uneaten fish feed (Oregon moist pellets), salmon feces, and net-fouling organisms
three 30-day sampling periods. Carapace lengths'
and individual wet weights (nonblottedi were
measured to the nearest 0.5 mm and 0.01 g, re-
spectively. In all experiments growth data were
analyzed by one-way analysis of variance and sur-
vival by chi-square tests.
In Experiment B, prawns were held in two net
pens measuring 1.2 x 1.8 x 1.8 m and constructed
of knotless nylon web (6.8-mm stretched measure
mesh I. Each pen was vertically divided into three
equal compartments (6.5 m-^ of substrate each)
with only the outer two being stocked with
prawns. The prawns were stocked at a density of
30.8/m^ of immersed substrate. The net pens were
covered with black plastic to prevent bird preda-
tion and to reduce light intensity. Two dietary
treatments were evaluated, mussel meat and salm-
on. Each treatment consisted of two replicates
and was fed exclusively on one of the two diets.
Feed was to excess every other day.
A sample of 50 prawns/replicate (100/
treatment) was measured for length and weighed
during each of eight sampling periods, except dur-
ing the last three periods in which all survivors
were measured and weighed. The prawns were
sampled at the beginning of the experiment and
32, 60, 88, 146, 205, 292, and 365 days later.
In Experiment B, in addition to evaluating the
diets, we studied the net cleaning ability of the
prawns and the value of the organisms on the net
(i.e., the net fouling organisms) as a supplemen-
tary food source. Samples of net fouling organisms
were taken from inside and outside of a compart-
ment containing prawns (test) and inside and out-
side of one not containing prawns (control). The
nets were selected at random, and each sample
' Carapace length is defined as the distance from the base of the
eyestalk to the posterior middorsal edge of the carapace.
consisted of all the organisms on the net within the
area of two 20-cm-diametei" circles; one circle was
at 0.25-m depth, and the other was at 1 .0-m depth.
The material collected was identified, enumer-
ated, and measui'ed volumetrically. The nets were
sampled during November 1975 and March
1976.
In Experiment C, juvenile prawns were stocked
in a net pen with coho salmon, Oncorhynchus
kisutch, (age-group 0) that averaged 20 g each. A
single net pen (without dividers) having a sub-
strate area of 10.8 m^ was used. Prawn density was
9.3/m^ of substrate, and salmon density was 82'm''
of water. The salmon were fed Oregon moist pel-
lets at 37f body weight/day; however, no feed was
provided for the prawns other than what they
could scavenge. All the prawns were measured at
each of seven periods: at the beginning of the ex-
periment and 15, 33, 60, 89, 146, and 206 days
later.
Care was taken to standardize culture condi-
tions such as lighting, substrate type, and water
temperature within each experiment. This was
not practical between experiments because of in-
herent differences between laboratory and net pen
work.
Stocking density differed between experiments,
but its impact on growth and survival (agnostic
behavior and feeding dominance) was minimized
by distributing an excess of food throughout the
rearing enclosure. In several years of behavioral
observations we have rarely seen overtly agres-
sive or cannibalistic behavior in spot prawns.
Results and Discussion
Mussel-fed juvenile prawns had the best survi-
val and growth rate of all the prawns raised in the
laboratory tanks (Experiment A); 74% of the
887
prawns survived, and they had a final mean
weight of 3. 14 g (Table 1 ). The prawns fed salmon
were significantly smaller (P<0.01) than the
mussel-fed group. However, the growth of prawns
in both the mussel and salmon diet groups equaled
or exceeded the growth reported for a natural
population in British Columbia (Butler 1964).
Growth was similar between oyster waste and
no food diet groups (Table 1), but survival was
significantly different (P<0.01), 64'/f for oyster
waste and 269^ for no food supplement. Prawns fed
oyster wastes or receiving no food grew at a slower
rate than the prawns in the other two diet groups;
this portion of the study was terminated after 60
days.
The poor growth of prawns fed oyster wastes
contrasts with the good growth of lobster //o/??a/7/.s
cifucricanus fed algae and oyster wastes in a sew-
age enriched raceway system (Mitchell 1975). In
that study intermediate organisms that fed on the
solid wastes of oysters were also available as a food
source for juvenile lobsters. In our study, raw
unfiltered seawater was used, but cleaning the
tanks twice weekly prohibited the establishment
of intermediate organisms.
Prawns in the net pens (Experiment B) grew
significantly faster than those in the laboratory
(Experiment A) for both mussel- (P<0.01) and
salmon-fed (P>0.01) treatments (Table 1).
Further, there was no significant difference in the
growth (P>0.10) or survival (P>0.25) of prawns
fed mussel or salmon in the pens as there had been
in the laboratory tanks.
In the net pens, the presence of net fouling or-
ganisms as an additional food source for the
prawns could explain the improved growth over
that seen with the same basic diets used in the
laboratory. Net fouling organisms could have pro-
vided nutritional requirements that were
deficient in the basic diets as provided in the
laboratory.
Prawns in both diet groups were observed re-
moving organisms that were on both the inside
and outside surfaces of the net pens using their
second periopods. The amount of net fouling or-
ganisms was reduced by the prawns in these en-
FlGURE 1. — Webbing of a three-chambered net pen showed reduced fouhng in the right and left chambers (spot prawns present)
compared with the center chamber without prawns.
888
closures ( Figure 1 , Table 2). The largest amount of
organisms occurred at the shallower depths of the
nets as in Moring and Moring's (19751 study of
salmon net pens at the same site. Mussels, asci-
cliaiis. and t uliicolous polychactt's, Spirarhis sp.,
c(iiiti'il)utc'(l lln' most fouling iti the control cham-
bers (without prawns). Except for a few Spirorbis
sp.. the net pen chambers with spot prawns were
completely clean. Little algal growth was present
due to the reduction of light by the black plastic
covers.
Rearing prawns and salmon in the same net pen
(Experiment C) proved encouraging. After 6'j mo
of culture the growth of these prawns (Figure 2)
exceeded that of the monoculture Experiment B
(P<0.01t and that reported for a natural popula-
tion (Butler 1964). Survival was 93' f and not sig-
nificantly different (P >0.95) from that of Experi-
ment B.
There was no evidence of adverse salmon/prawn
interaction. The types of food available to the
prawns when reared with salmon included: dead
fish, uneaten fishfood pellets, fish feces, and net-
fouling organisms. The relative contribution of
each was unknown.
A limiting factor to stocking juvenile prawns in
commercial salmon net pens is the requirement
that the prawns must be large enough to prevent
them from going through meshes of the net.
Smaller "nursery" nets of reduced mesh size could
be hung inside the main salmon nets until the
prawns reach a suitable size (about 4 g, or 3 mo of
age).
Several advantages might accrue from using a
scavenger, such as the spot prawn, as a companion
crop in salmon culture. In Experiment B a reduc-
tion in net fouling was seen in net pens with
prawns (Table 2. Figure 1). This reduction will aid
salmon culture because it would allow greater
water circulation within the enclosure, thus in-
creasing dissolved oxygen levels and flushing of
CO
5 10
<
cr
Q_
o
8
S^ 6
UJ
Ijj
<
UJ
>
<
DIET
A Experiment C (voriety of foods)
• Experiment B (mussel)
■ Experiment B (salmon)
British Columbia
natural population
(Butler 1964)
X
_L
3 6 9
NUMBER OF MONTHS
12
FlClKK 2.
-Growth of juvenile Pandali/s plalyrcros in net pens
compared with a natural population.
metabolic wastes from the pens. A reduction in net
maintenance cost might also be realized. Experi-
inents B and C demonstrated that fish in the pres-
ence of net-fouling organisms was an acceptable
food for the prawn (Figure 2); the utilization of
dead salmon by a scavenger would be a valuable
conversion of an otherwise unused protein and
would reduce the labor needed to remove salmon
mortalities from the system.
Further experiments are needed to determine
proper stocking densities of prawns and salmon to
maximize growth rates, survival, and to make op-
timum use of the net cleaning activities of the
prawns. Further, while our studies showed that a
single diet fed to prawns in the laboratory was not
adequate for rapid growth, other studies (Kelly et
al. 1976) have shown that combination raw diets
could produce adequate growth. These combina-
tion diets need to be evaluated in the net pen
system.
T.AHl.K 2. — Displacement volumes i milliliters) of fouling or-
ganisms on the inside and outside surfaces of net pens with and
without Pandcihis platycerog from July 1975 to March 1976. All
pen chambers were clean at the start of the experiment. Sample
area was 314.2 cm^ of vertical mesh.
Location of
sample
Depth of
sample
(m)
Net pen compartment
with prawns
Net pen compartment
without prawns
Nov. 1975 Mar 1976
Nov, 1975
IVIar, 1976
Inside
0 25
0 40
0 00
2640
41 55
of net
1 00
0 00
0 00
1050
12 03
Outside
025
0 25
0 50
9 00
17 00
of net
1.00
0,10
0,40
600
11,53
Literature Cited
Butler, T. H.
1964. Growth, reproduction, and distribution of pandalid
shrimps in British Columbia. J. Fish. Res. Board Can.
21:1403-1452.
KKLLY, R. O. a., a. W. H.\SKI.TINK. .-\Nn E. E. EliKRT.
1976. Mariculture potential of the spot prawn, Pandalus
platyceros Brandt. Aquaculture 10( 1 1:1-16.
Mahnken, C. V. W.
1975. Status of commercial net pen farming of Pacific
salmon in Puget Sound. In J. W. Avault, Jr.. and R. Miller
889
(editors), Proc. 6th Annu. Meet. World Mariculture Soc,
p. 285-298.
MITCIIKI.I,. J. R.
1975. A polyculture system for commercially important
marine species with special reference to the lobster.
Honuiriisaiueritaruis. In J. W Avault, Jr., and R. Miller
(editors). Proc. 6th Annu. Meet. World Mariculture Soc.
p. 249-259.
MOKIN'C;, J. R., .AM) K. A. MOKINC.
1975. Succession of net biofouling material and its role in
the diet of pen-cultured chinook salmon. Prog. Fish-
Cult. 37(l):27-30.
PRE.NTRK, E. F.
1975. Spot prawn culture: status and potential. /« C. W.
Nyegaard i editor). Proceedings of a seminar on shellfish
farming in Puget Sound, Oct. 7, 1975, Poulsbo, Wash., p.
1-11. Wash. State Univ., Coll. Agric, Coop. Ext. Serv.,
Pullman.
RKN.SKI., J. E., .^\\^ E. F. PKKNPKK.
1977. First record of a second mating and spawning of the
spot prawn, Pandalus platycerox, in captivity. Fish.
Bull., U.S. 75:648-649.
WlCKl.WS, J, F.
1972. Experiments on the culture of the spot prawn Pan-
dalus platyccros Brandt and the giant freshwater prawn
Macrobrachium roscnhcrgti ide Man). Fish. Invest.,
Minist. Agric, Fish. Food (G.B.)., Ser. II, 27(5), 23 p.
John E. Rk.wskl
Squaxin Island Tribe
Route 1 , Box 25 7
She! ton. WA 98584
E.AKl, F. PKKNTKK
Northwest and Alaska Fisheries Center
National Marine Fisheries Service, NOAA
2725 Montlake Boulevard East
Seattle, W A 98112
LARVAL LENGTH-WEIGHT RELATIONS FOR
SEVEN SPECIES OF NORTHVC EST ATLANTIC
FISHES REARED IN THE LABORATORY
Growth is an important connecting link in the
functional influence of biotic and abiotic factors on
the dynamics of fish populations. Length-weight
relations are used by fishery scientists to describe
the growth characteristics of species or popula-
tions and as a basis of evaluating the consequences
of environmental influences on growth. Length-
weight relations are also used in assessing produc-
tion when combined with age and growth informa-
tion and in determining length or weight in a
situation where either one or the other is unknown
due to sampling procedures.
Studies of the early life of fishes are receiving
increasing emphasis, particularly with regard to
growth and survival in the larval stage. Survival
during this period is thought to be minimal and
potentially variable from year to year. Small
changes of tenths of a percent in mortality have
the potential to produce orders of magnitude dif-
ferences in eventual adult populations. Larval
growth can be influenced due to food limitations
and varying abiotic factors (Houde 1974; Lasker
1975; Laurence 1977). Because of these facts,
fishery scientists are particularly concerned with
two aspects: 1 ) quantifying variable larval growth
and survival, relating it to subsequent year-class
recruitment, and applying it to traditional stock-
recruitment relationships where recruitment has
often been considered constant; and 2) the poten-
tial use of this type of information in evaluating
the increasing effects of pollution or other en-
vironmental perturbations because of the fragility
and sensitivity of larvae to changing or altered
environmental variables.
Solutions to these problem areas require quan-
titative knowledge of growth parameters of larval
fishes, and length-weight relations can be helpful
in providing information or establishing relation-
ships between pertinent sets of data. It is gener-
ally thought that weight' is a better measure of
absolute growth offish larvae than length as well
as the prime determinant of condition when com-
bined with length. Many species exhibit allo-
melric or dispi'oportionate length-weight growth.
This is especially true during the period of
metamorphosis when some species display vary-
ing or unusual body proportions with age (Blaxter
1969) and length does not increase in proportion to
increasing weight. Additionally, recent attempts
to construct models of larval survival, as
influenced by environmental variables and den-
sity dependent feeding relationships, require
weight detei'minations for estimates of biomass
and caloric tui-nover between larval and prey
trophic levels.
There is an extensive data base to asses larval
fish growth and survival based on ichthyoplank-
ton collected on survey cruises during the last 75
yr by marine laboratories throughout the world.
Unfortunately, almost all of these data are in
standard or total length measurements as they are
much more easily and rapidly taken than dry
weights. The difficulty involved in obtaining dry
'Weight for species in this research refers to dry weight. Dry
weight is the most accurate for fish larvae because accurate wet
weights are difficult to obtain and yield variable results on or-
ganisms as small as fish larvae.
890
FISHKKV Bl'I.I.KTlN Vol. 7fi. NO 4. 1979.
weights of young stages has been overcome to a
certain degree with the advent of experimental
laboratory programs at a few research facilities
during the last 10 yr.
The experimental larval fish program at the
Northeast Fisheries Center Narragansett
Laboratory, National Mai'ine Fisheries Service,
NOAA, has been studying grow th, metabolic, and
trophodynamic factors for a number of important
commercial and sport species, and it is the object of
this report to present larval length-weight rela-
tions for seven species including Atlantic cod,
Gadus ruorhua: haddock, Melanogrammus
aeglefinun: scup, Stcnotonuis chrysops: Atlantic
herring, Cliipea harengiis; winter flounder,
Pseudoplci/ronectes omericanus\ summer floun-
der, Paralicbthys dentatus\ and yellowtail floun-
der, Limanda ferruginea . The larval length-
weight relations presented here are previously
unreported in the literature for six of the seven
species with the exception of the Atlantic herring,
which is included because it represents the only
data available for western North Atlantic stocks.
Materials and Methods
All larvae were obtained from experimental
spawning of adults in the laboratory and reared by
techniques reported by Smigielski ( 1975a, b) and
Laurence ( 1975). The length-weight data were col-
lected coincident with a variety of experimental
studies on larval growth, survival, metabolism,
and feeding reported by Laurence (1974, 1977,
1978, and as yet unpublished).
In all cases the data were collected from larvae
reared at prey concentrations in the range of 0.5 to
3.0 organisms ml. Concentrations of 0.5 and above
have been shown to be adequate for normal grow th
in the studies cited above. Rearing temperatures
were optimum for growth and survival or within a
3°C nonlethal range about the optimum depend-
ing upon the experiment from which the data were
taken. Optimum temperatures determined in
laboratory studies for rearing the seven species
were 7°C for cod and haddock, 8°C for winter
flounder, 10°C for herring and yellowtail flounder,
16°C for summer flounder, and 18°C for scup.
Length measurements were taken from the tip
of the snout to the end of the notochord in the
preflexion stage. During flexion of the notochord
measurements were taken to a line vertically per-
pendicular to the tip of the notochord until the
hypural bones became prominent or exceeded the
line vertically perpendicular to the notochord tip.
At this time, a standard length measurement to
the posterior end of the hypural plate was re-
corded. Since the original experiments were not
designed for developmental anatomy purposes,
the different flexion stages were not recorded coin-
cident with the length and weights.
Lengths were recorded to the nearest 0.1 mm
with a filar ocular micrometer. Dry weights were
determined after rinsing larvae in distilled water,
pipeting onto a glass Petri dish, and drying to a
constant weight at 60°-90°C for 24 h. Individual
dry weights were recorded to the nearest 0.1 /xg on
a gram electrobalance.
All measurements were made on post yolk-sac
larvae that were freshly sacrificed and unpre-
served. The data points for each species represent
lengths and weights for individual larvae except
for winter flounder and haddock. The data for
these two species are the means of lengths and
weights for samples of 10-25 larvae collected on a
weekly basis during different experiments. The
experimental procedures precluded the matching
of individual lengths with weights for these two
species.
Regression equations and associated parame-
ters were calculated as geometric mean, func-
tional regressions using log base 10 transformed
data according to the methods of Ricker (1973)
rather than using the previously standard predic-
tive, regression techniques. Ricker demonstrated
the advantages of using functional rather than
predictive regression calculations to reduce bias in
length-weight conversions where the populations
of measurements are typically open ended, where
only a portion of the length and weight distribu-
tions are represented, and where the variability
may be more inherent in the biological material
itself rather than the means of measuring length
and weight.
Results and Discussion
The exponential relation between length and
weight for all seven species are presented in
linearized form by logarithmic transformation in
Figures 1-7. The larvae studied in this research
are from different taxonomic families (Clupeidae,
Gadidae, Sparidae, Bothidae, and Pleuronec-
tidae), represent different adult life styles (pelagic
and demersal), develop in a range of different
temperatures, and demonstrate different patterns
of metamorphosis from larval to juvenile stages.
891
•4. 0
3.0
ID
a.
2. 0
1. 0
STRNDHRD LENGTH (MM)
10. 0
SUMMER FLOUNDER
L06r--0. 763+3. 780LO6X
100.0
10000. 0
1000. 0
100. 0
10. 0
0. 1 1.1
LOG STflNDHRO LENGTH (MM;
FiGl'RE 1. — Standard length-dry weight relationship of larval
summer flounder.
STRNDRRD LENGTH CMM)
10. 0
4.0
3.0 -
^
CD
o
3.
3.
(-
I—
I
U3
|J3
UJ
Ul
rz
3
V
V
q:
or
Q
a
o
o
J
2.0 -
1. 0
0.1 1.1
LOG STRNDRRD LENGTH (MM.)
100. 0
r 10000.0
1000. 0
(SI
a.
I
01
Q
100. 0
10. 0
Figure 3. — Standard length-dry weight relationship of larval
cod.
4. 0
3.0
2. 0
I.O
STRNDHRD LENGTH (MM)
10.0
100. 0
10000. 0
1000. 0
a.
100. 0
10. 0
0. 1 1.1
LOC S^RNDflRD LENGTH (MM)
Figure 2.— Standard length-dry weight relationship of larval
haddock. Points represent means for length and weight of sam-
ples of 10-25 larvae.
892
4. 0 -
a.
::. 3.0 -
>-
tr
a
o
J
2. 0
1. 0
STRNDRRD LENGTH (MM)
10.0 100.0
rELLOWTRiL FLOUNDER
L06r = -1. 017+3. 909LO6X
- 10000.0
ID
a.
1000.0 7r.
01
Q
100. 0
10. 0
0. 1 1.1
LOG STRNDRRD LENGTH (MM)
Figure 4. — Standard length-dry weight relationship of larval
yellowtail flounder.
STRNDRRD LENGTH (MM)
10. 0
4.0
3.0
a.
X
2.0
kl INTER FLOUNDER
L0GY = -1. 347t-4. 769LOGX
10000. 0
1000. 0
1.0
Q
100.0
10. 0
0. i
LOG STRNDRRD LtNGTH fMM)
Figure 5. — Standard length-dry weight relationship of larval
winter flounder. Points represent means for length and weight of
samples of 10-25 larvae.
In spite of these differences, a visual examination
of the length-weight regression equation
coefficients and associated parameters for all
species reveals no obvious correlations with the
differences (Table 1). It would not be prudent to
statistically test for differences or associations be-
tween the species because data for haddock and
winter flounder were averaged. Ricker ( 1973) cau-
tions that averaging changes the variances as-
sociated with the variables, particularly the inde-
pendent variable, so that a comparison between
STRNDRRD LENGTH (MM)
10.0
100. 0
4.0
3.0
3.
o
J
2. 0 -
1.0
10000. 0
1000.0
I
C3
a
100.0
10. 0
0. 1 1.1
LOG STRNDRRD LENGTH (MM)
Figure 6. — Standard length-dry weight relationship of larval
Atlantic herring.
STRNDRRD LENGTH (MM)
10.0 100.0
10000.0
CD
3.
1000.0 5
a
100. 0
1. 0
10.0
0.1 1.1
LOG STRNDRRD LENGTH (MM)
FIGURE 7.— Standard length-dry weight relationship of larval
scup.
893
Table l. — Regression parameters for length-weight relations of seven species of laboratory-reared larval north-
west Atlantic fishes.
Coefficient
Standard error
95°o CI about
Number
Correlation
of
Regression
of regression
regression
Larval species
sampled
coefficient
determination
coefficient
coefficient
coefficient
Summer flounder
57
0997
0 994
3,780
0 039
3702-3858
Yellowtail flounder
80
0.995
0 990
3.909
0 044
3821-3953
Herring
98
0,997
0 993
4.295
0037
4,221-4 369
Soup
100
0-997
0 993
3.756
0028
3692-3820
Cod
104
0 997
0 995
4.081
0029
4 023-4 104
Haddock'
23
0997
0 995
4476
0,071
4 328-4,624
Winter flounder'
36
0991
0 982
4 769
0 110
4 545-4 993
IDala represent means for length and weigfit of samples of 10-25 larvae
averaged and unaveraged data is not valid: al-
though he does not discredit the use of averaged
data by itself. Also, seven species, some of which
are closely related taxonomically, probably do not
constitute enough cases for drawing conclusions
about functional differences. Consequently, these
length-weight relations should properly be con-
sidered individually as empirically derived rela-
tions for each particular opecies.
The length-weight relation of fishes usually ap-
proximates the cube law relationship in which the
weight is proportional to the cube of the length
(Beckman 1948; Rounsefe'l and Everhart 1953).
This is usually true for adult fishes; however, re-
sults of this research imply that it is not necessar-
ily .so for larvae. All the length exponents for the
species investigated in these studies were >3.6
with a mean value of 4.152. It would seem then
that the dry weight of larval fishes may be more
closely proportional to length to the fourth power
rather than cubed. Length-weight relations for
fish larvae are scarce in the literature. Examina-
tion of the data available (log,,, formulation)
seems to substantiate that the length exponent is
always greater than three and more closely ap-
proximates four. Marshall et al.( 1937) presented a
total length-dry weight equation for larval her-
ring, the only species with data available to com-
pare with this study, equivalent to log W =
-5.6990 + 4.52 log L. The length exponent is >4
and similar to the value of 4.295 for herring in this
research. Ehrlich et al. ( 1976) also presented a simi-
lar standard length-dry weight relation for Firth
ofClyde herring larvae ( log W = -5.7052 -(-4.5710
logL) as well as a relationship of log W = -4.3043
-i- 3.9155 log L for larvae of plaice, Pleuronectcs
plati'ssa. Stepien (1976) reported a standard
length-dry weight relationship for larval sea
bream, Anhosargus rhonihoidcilis, of log W =
-0.5144 + 4.2816 log L, and Lasker et al. ( 1970)
reported a standard length-dry weight relation-
ship for northern anchovy larvae, Engrau lis nior-
dax. of log H' = -3.8205 + 3.3237 log L.
It is acknowledged that variables such as tem-
perature and feeding conditions can influence
growth and complicate length-weight relations.
These factors may have contributed to some var-
iability in the present study. However, it is felt that
these influences were minimized by the experi-
mental feeding levels and temperatures which
were within ranges for adequate growth and sur-
vival, and any changes in length or weight were
most likely mitigated together causing little effect
on the form of the length-weight relation. This is
supported in studies of haddock larvae (Laurence
1974) where condition factors were similar and
randomly associated with prey concentrations
>0.5 organisms/ml.
The use of larval length-weight relations for
extrapolation may result in some underestimation
or overestimation at the smallest and/or largest
sizes due to changes in growth rates for yolk-sac or
metamorphosing larvae. Farris (1959) suggested
that growth rates of larval marine fishes could be
separated into three different phases; the first two
prior to yolk absorption and the third following.
Zweifel and Lasker ( 1976) presented a mathemat-
ical interpretation of larval growth with age
defined by the Laird-Gompertz growth function.
They noticed two growth cycles; one extending
from hatching to yolk absorption and the other
following yolk absorption. This variability in the
small sizes is probably not inherent in the data of
this study because larvae were not included until
yolk was absorbed and active feeding had com-
menced. Some variability may be present in the
upper range of sizes in these length-weight rela-
tions. In some cases data for larger larvae are not
as extensive as for smaller larvae. Also, the major-
ity of the largest individuals for each species were
either undergoing or had completed metamor-
phosis where changes in growth rates of length or
894
weight might cause allometry. Zvveifel and Lasker
11976) briefly considered the length-weight rela-
tion in terms of a modified Gompertz-type relation
and noted overestimation problems in extrapola-
tion at the largest sizes.
Length-weight relations have merit, but their
usefulness is greatly enhanced when combined
with other studies, particularly those on age.
Length- weight by itself does not necessarily imply
rate of change because of the potential influence
the environment may have on changing growth
with time. However, when correlated with age and
compensated for change in rate due to biotic and
abiotic influences, length-weight studies can be an
important component in estimating growth, sur-
vival, and population production.
Acknowledgments
I wish to thank J. B. Colton, Jr. and K . Sherman
for their critique of the manuscript, and B. R.
Burns, T. A. Halavik, and A. S. Smigielski for
their assistance in rearing the larvae and collect-
mg measurements.
Literature Cited
BECKMAN. W. C.
1948. The length-weight relationship, factors for conver-
sions between standard and total lengths, and coefficients
of condition for seven Michigan fishes. Trans. Am. Fish.
Soc. 75:237-256.
BL.A.XTER. J. H. S.
1969. Development: eggs and larvae. In W. S..Hoar and D.
J. Randall (editors). Fish physiolog>'. Vol. 3. p. 177-252.
Academic Press, N.Y.
EHRLICH. K. F., J. H. S. BLAXTER. AND R. PEMBERTON.
1976. Morphological and histological changes during the
growth and starvation of herrmg and plaice larvae. Mar.
Biol. (Berl.i 35:105-118.
FARRIS. D. A.
1959. A change in the early growth rates of four larval
marine fishes. Limnol. Oceanogr. 4:29-36.
HOUDE. E. D.
1974. Effects of temperature and delayed feeding on
growth and survival of larvae of three species of subtropi-
cal marine fishes. Mar. Biol. (Berl.) 26:271-285.
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.
Lasker, R., H. M. Feder. G. H. Theilacker, and R. C. May.
1970. Feeding, growth, and survival ofEngraulis mordax
larvae reared in the laboratory. Mar. Biol. (Berl.) 5:345-
353.
Laurence. G. C.
1974. Growth and survival of haddock Melanogrammus
aeglefinus larvae in relation to planktonic prey concentra-
tion. J. Fish. Res. Board Can. 31:1415-1419.
1975. Laborator>' growth and metabolism of the winter
flounder Pseudopleuronectes americanus from hatching
through metamorphosis at three temperatures. Mar. Biol.
(Berl.) 32:223-229.
1977. A bioenergetic model for the analysis of feeding and
survival potential of winter flounder, Pseudopleuronectes
americanus, larvae during the period from hatching to
metamorphosis. Fish. Bull., U.S. 75:529-546.
In press. Comparative growth, respiration and delayed
feeding abilities of larval cod [Gadus morhua) and had-
dock (Melanogrammus aeglefinus) as influenced by tem-
perature during laboratory studies. Mar. Biol. (Berl.)
Marshall. S. M., A. G. Nicholls. and A. P. Orr.
1937. On the growth and feeding of the larval and post-
larval stages of the Clyde herring. J. Mar. Biol. Assoc.
U.K. 22:245-268.
RlCKER, W. E.
1973. Linear regressions in fishery research. J. Fish. Res.
Board Can. 30:409-434.
Rounsefell, G. a., and W. H. EVERHART
1953. Fishery science: its methods and applications. John
Wiley and Sons Inc., N.Y., 444 p.
Smigielskl a. S.
1975a. Hormone-induced spawnings of the summer floun-
der and rearing of the larvae in the laboratory. Prog.
Fish-Cult. 37:3-8.
1975b. Hormonal-induced ovulation of the winter floun-
der, Pseudopleuronectes americanus. Fish. Bull., U.S.
73:431-438.
Stepien, W. p., Jr
1976. Feeding oflaboratory-reared larvae ofthe sea bream
Archosargus rhomboidalis (Sparidae). Mar. Biol. (Berl.)
38:1-16.
Zweifel. J. R., AND R. Lasker
1976. Prehatch and posthatch growth of fishes— a general
model. Fish. Bull., U.S. 74:609-621.
Geoffrey C. Laurence
Northeast Fisheries Center Narragansett Laboratory
National Marine Fisheries Service. NOAA
R.R. 7 A. Box 522 A
Narragansett, RI 02882
EFFECT OF THERMAL INCREASES OF
SHORT DURATION ON
SURVIVAL OF EUPHAUSIA PACIFICA
Euphausiids are an important source of food for
many valuable species of fish including herring,
cod, pollock, and salmon. Cooney (1971) reported
that Euphausia pacifica was the most abundant
species associated with the diffuse scattering layer
at all locations in Puget Sound, Wash. He found
that during the day euphausiids are most abun-
dant between depths of 50 and 100 m and that at
night most of the population migrates into the
upper 50 m. Cooney's findings indicate that great
numbers of euphausiids could be drawn through
FISHERY BULLETIN: VOL. 76. NO. 4, 1979.
895
the condenser cooling systems of thermal nuclear
power plants where they would encounter a siz-
able thermal shock.
Zooplankton entrained in a power plant cooling
system located in a saltwater environment could
be subjected to an average temperature increase
ranging from 12° to 16°C (Coutant 1970). In some
plants, the increases are as high as 19°C.
Maximum temperatures would be reached in <1
min in the condenser and would be maintained for
at least 9 min in a diffuser discharge system and at
almost maximum temperature, for possibly up to
21 min, in a discharge canal system. Other factors
that could cause damage to euphausiids in a cool-
ing system include pressure changes, abrasion,
and toxic substances.
I simulated the thermal conditions encountered
in a cooling system to determine the temperature
increases that E. pacifica could resist for short
periods (15 and 30 min). This information can be
applied to the design and operation of cooling sys-
tems to protect zooplankton.
These studies were conducted at the National
Marine Fisheries Service's Mukilteo Field Sta-
tion, Washington, during 1971-74.
Methods
Euphausiids for these experiments were cap-
tured during daylight hours in Port Gardner of
northern Puget Sound, Washington, between
Mukilteo and Gedney Island. A 10-m net with
333- /Lim aperture Nytex^ netting was towed at a
depth of about 60 m at a rate of 4.6 km/h. Tows
were usually of a 5-min duration. A 946-ml glass
bottle was used as a collection receptacle to protect
the animals.
As soon as the net was retrieved, the catch (con-
sisting mostly of euphausiids) was divided be-
tween two or three 18.9-1 Nalgene carboys filled
with fresh seawater and covered with black
polyethylene sheeting to exclude light. The catch
was taken immediately to the laboratory, usually
<2 km away, where the euphausiids were sepa-
rated from other organisms in the catch and placed
in 5-1 battery jars (23 x 14 x 17 cm) of fresh
seawater. They were then placed in a dark, low-
temperature incubator set at their previous am-
bient seawater temperature where they were held
before and after testing.
'Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
The test apparatus consisted of a series of 5-1
battery jars filled with seawater that were main-
tained at specific temperatures by immersion
heaters activated by temperature controllers
(Craddock 1976). The jars were in a primary bath
of running seawater at ambient temperature and
air was continuously bubbled into the jars to
eliminate stratification.
Test containers for holding the euphausiids
were polyvinyl chloride boxes of 5 cm^ with two
opposing sides having 4-cm diameter cutouts cov-
ered with 333-/xm aperture Nytex netting to allow
free water circulation. Styrofoam glued to the
boxes provided flotation (Figure 1).
The temperature-time regime to which the
euphausiids were subjected was designed to simu-
late their passage through a condenser cooling
system. Coutant (1970) depicted a hypothetical
temperature-time course for organisms entrained
in condenser cooling water and discharged by dif-
fuser or by discharge canal. An animal could be
subjected to maximum temperature increases for
up to about 10 min in a diffuser and up to about 20
min in a discharge canal system. Relative to his
study, I chose 15- and 30-min exposure tests to
represent the longest exposure that might be en-
countered. To simulate these conditions, test
euphausiids were subjected to a given tempera-
ture ranging from 14° to 29°C for 15 or 30 min,
starting from temperatures of 11° or 9°C.
Euphausiids used as controls were always kept at
the prevailing ambient temperature (approxi-
mately the same as the subsurface temperature of
Puget Sound). Five 15-min tests were conducted
during June- July 1971, four 30-min tests were run
during June-August 1971, and two 15-min tests
were made during March-April 1974.
The euphausiids were held 18 h or longer before
testing to eliminate handling mortality and were
then counted into test containers in seawater
while the secondary baths were being raised to the
test temperatures. Either 5 or 10 euphausiids
were tested in each container, depending upon the
numbers available for that particular test.
When all secondary baths became equilibrated
at the test temperatures, the boxes containing the
euphausiids along with a small amount of water
were placed in the test baths. Water in the test
containers was within 0.5°C of the test tempera-
ture in an average of 28 s after introduction. At the
end of the exposure period, the test boxes contain-
ing the euphausiids were removed from seawater
at the test temperature and placed in fresh seawa-
896
Figure l. — Test chambers and apparatus for testing thermal effects on Euphausia pacifica.
ter at the acclimation temperature and main-
tained in the low temperature incubator. All lots
were checked for mortality at 5, 10, and 15 min
after introduction to the test temperature when
the duration of the exposure was 15 min and also
at 20 and 30 min after introduction for the 30-min
exposure. All were again checked at 1, 24, and 48 h
after testing. The 48 h survival was taken as diag-
nostic for the TL5o's.
Temperature effects were evaluated on the basis
of mortality during tests and 48 h after testing.
Forty-eight hours was assumed to be a reasonable
holding period to check for delayed mortalities —
yet not long enough to cause mortality due to
confinement and lack of food. A euphausiid was
considered dead if no movement of the thoracic
appendages, pleopods, or antennae could be de-
tected using 3 X magnification. I modified the term
median tolerance limit (TL50) to indicate the
maximum 15- or 30-min exposure temperature
survived by at least 50'7f of the experimental ani-
mals 48 h after testing. This should be considered
the maximum temperature-time combination re-
sisted.
Lengths of the test animals between the ex-
treme tips of the rostrum and telson were taken at
the end of each test. The mean-lengths of the
euphausiids tested at the various seasons ranged
from 12.11 to 18.37 mm (Table 1). The actual
range was from 9 to 27 mm. Those tested in the
early part of June were the largest; they exceeded
those tested later in June by an average of 6.26
Table 1.— Sizes (milHmeters) of Euphausia pacifica tested.
Dates
Mean
Range
Dates
Mean
Range
1971;
1974:
June 2-9
18,37
14-26
Mar, 12
14.15
10-20
July 21-30
12.11
9-16
Apr. 12-16
13.85
10-23
Aug 4-11
13.14
10-19
897
mm. Those tested in August, March, and April had
an average spread of only 1.01 mm.
Results
Controls in the different tests suffered no mor-
tality during the exposure period except the
June-July tests, where the 15-min group lost 7% of
controls by the end of 48 h and the 30-min group
lost 10% by the end of 48 h. The data were cor-
rected to reflect the loss of the controls in the
June-July tests, using the method of Tattersfield
and Morris ( 1924) as reported by Sprague ( 1969).
Acclimation temperature influenced resistance.
The TL50 of euphausiids given a 15-min exposure
to elevated temperature was 25°C for those accli-
mated to 11°C; it was 23°C for those acclimated to
9°C. Exposure to 26°C resulted in survivals of 32%
and exposure to 27°C resulted in almost im-
mediate death ( <15 min). In the 15 min 9°C accli-
mation test (March-April 1974), the TL50 was at
23°C and 47% were still surviving 48 h after expo-
sure at 24°C. However, 15 min after exposure to
25°C, only 13% remained alive and all were dead
in <15 min at 26°C. Figure 2 depicts the survival
after a 15- and 30-min exposure to elevated tem-
peratures and after a 48-h holding period.
Increasing the duration of exposure to test
temperatures from 15 to 30 min when the ambient
temperature was 11°C decreased the TL50 by 1° to
24°C. Of those tested at 25°C, only 44% survived 48
h after testing. At 26°C, only 2.5% survived the
30-min test period. None survived the test period
at 27°C.
The logistic model was fitted to the data from
the three different thermal shock tests. The prob-
ability of survival was taken to be the form
P [survival at temperatures] = l/d+e"*^''^ where
e = 2.718. This is the so-called logistic model and a
and b are parameters which are estimated using
the data. In the 15-min exposure of June-July
1971, a = 0.6544 and h = -16.4138; in the
March-April 1974 exposure, a = 0.9568 and b =
-22.2860; whereas in the 30-min exposure July-
August 1971, d = 0.5173 and b = -12.2572. The
estimates of TL50 and an approximate 95%
confidence interval for it follow for the three tests:
1) 25.08°C, 24.51°-25.65°C; 2) 23.29°C, 22.76°-
23.82°C; and 3) 23.69°C, 22.95°-24.44°C.
There was no obvious difference in the effect of a
I5MIN
100
90
80
^ 70
1 60
JUNE - JULY
CO 50
50%
S^ 40
-
30
15 MIN EX
20
-
o
CO
I I I I i_l 1 l_i I . I ■ l_i I 1 I V I . I
10 12 14 16 18 20 22 24 26 28 30
TEST TEMP (°C)
30 MIN
/^
48 H ""^ /
V
JUNE - JULY
50
40 -
30 -
20 -
10 -
0 V/^
8
\ /
50% ^
30 MIN EX
-1— J — l-j 1—1 1 ] I I I . I
I . J
10 12 14 16 18 20 22 24 26 28 30
TEST TEMP t°C)
100
90
80
I 60
a: 50
I5MIN
30 -
20 -
10 -
0 Ly/
48 H
MARCH -APRIL
50% —
15 MIN EX
I . I
I ■ I ■ I ■ I ■ I ■ I
I I ]
8 10 12 14 16 18 20 22 24 26 28 30
TEST TEMP (°C)
Figure 2. — Survival (including mean and range) ofEuphausia
pacifica after 15- or 30-min exposures to elevated temperatures
and subsequent holding at 9° or \\°C ambient temperatures for
48 h.
898
short exposure to increased temperature on the
largest or the smallest euphausiids tested. Two
out of three groups tested for 15 min in early June
(the largest euphausiids) exceeded 50'7r mortality
after exposure to 26"C as did the two groups tested
in late July (the smallest euphausiids).
D
ISC us-sion
The intake of a condenser cooling system may
entrain large quantities of euphausiids —
depending to some extent on the depth of the in-
take, the season of the year, and even the time of
the day. During the summer, fall, and winter, the
young euphausiids make diurnal vertical migra-
tions from the 50- to 100-m strata, rising daily to
the surface during the dark hours. After sexual
maturity in the early spring they descend even
deeper until they inhabit depths over 200 m dur-
ing their second winter. The following spring they
rise to the surface for the second time to breed. The
young euphausiids thus spend much of their first
year at depths above 50 m, and older adults are
again near the surface in the spring (Ponomareva
1963).
Gilfillan (1972) pointed out that E. pcuifica is
widely distributed and is abundant in water hav-
ing differing temperature characteristics. His
studies showed that E. pcuifica from the Pacific
Ocean were more easily stressed by changes in
temperature and salinity than those from the west
entrance of Strait of Juan de Fuca — which, in
turn, were more readily stressed than those from
Saanich Inlet. His results indicate that E. pcuifica
from inner Puget Sound would be among the most
resistant to thermal stress of these different
groups.
Temperatures encountered by euphausiids in
Puget Sound normally vary only slightly from the
surface to 100 m and deeper. From October
through about May there is usually no change in
temperature from the surface to 100 m, whereas in
the summer the surface to 10 m or less may be a
few degrees warmer (Lincoln and Collias^).
Seasonal temperature variations in most of
Puget Sound are also small, ranging from a low of
7°or 8°C in February to 11°C in late July, August,
and September. Even considering their vertical
migrations in summer, euphausiids are normally
^Lincoln, J. H., and E. E. Collias. 1970. Skagit Bay Study
Progress Report No. 3. Univ. Wash. Dep. Oceanogr., Seattle,
Ref. M70-111, 88 p.
subjected to only slight temperature fluctuations
and, therefore, the mortalities observed at simu-
lated condenser cooling temperatures are not sur-
prising.
Once entrained in a condenser cooling system,
the euphausiids would encounter an abrupt tem-
perature increase of 12°-16"C (Coutant 1970),
which could increase temperatures above the am-
bient temperature of Puget Sound to the critical
range for survival. There are periods from July
through September when surface temperatures
may reach or exceed 15°C in portions of Puget
Sound (Lincoln and Collias see footnote 2). Nor-
mally, surface temperatures do not exceed 14°C.
Cooney ( 1971) noted high surface temperatures in
June of 16.7°- 19°C. These temperatures could re-
sult in condenser cooling temperatures of 27°C and
above, which this study found to be 1007f lethal in
a very short time.
Data from this study indicate that even a short
passage time through a condenser (15 min) at
temperatures of 23°-24°C could kill from 11 to 53V.
of the euphausiids by thermal causes alone. The
added loss due to abrasion, pressure, and toxic
substances is unknown.
To minimize damage to the euphausiid popula-
tions, condenser cooling system intakes should be
located deep enough to take advantage of the cold-
est cooling water available to minimize tempera-
tures in the system. A very deep intake (just below
100 m) would probably minimize the entrainment
of euphausiids. A surface intake would be espe-
cially harmful because of the higher surface tem-
peratures and because of the swarming of
euphausiids on the surface. Plant lights at night
could cause the surface swarming.
Literature Cited
Cooney, R. T.
1971. Zooplankton and micronekton associated with a dif-
fuse sound-scattering layer in Puget Sound, Washing-
ton. Ph.D. Thesis, Univ. Washington, Seattle, 208 p.
COL'TANT. C. C.
1970. Entrainment in cooling waters: Steps toward pre-
dictability. Proc. 50th Anna. Conf. West. Assoc. State
Game Fish Comm,, p. 90-105.
CR.'SiDDOCK. D. R.
1976. Effects of increased water temperature on Daphnia
putex. Fish. Bull., U.S. 74:403-408.
GILFILLAN E.
1972. Reactions of Euphausia pad fica Hansen (Crustacea)
from oceanic, mixed oceanic-coastal and coastal waters of
British Columbia to experimental changes in tempera-
ture and salinity. J. Exp. Mar. Biol. Ecol. 10:29-40.
899
PONOMAREVA, L. A.
1963. Evfauziidy sevemoi poloviny Tikhogo okeana, ikh
rasprostranenie i ekologiya massovykh vidov (Euphausi-
ids of the North Pacific, their distribution and ecolo-
gy). Moscow. Izd. Akad. Nauk SSSR, 142 p. (Translated
Isr. Program Sci. Transl., 1966. 154 p.; available U.S. Dep.
Commer., Natl. Tech Inf. Serv.. Springfield, Va., as TT
65-50098.1
SpR.'^GUE. J. B.
1969. Review paper: Measurement of pollutant toxicity to
fish. I. Bioassay methods for acute toxicity. Water Res.
3:793-821.
T.-XTTERSFIELD. F., AND H. M. MORRIS
1924. An apparatus for testing the toxic values of contact
insecticides under controlled conditions. Bull. Entomol.
Res. 14:585-590.
species (Lowell 1971; Morris and Kanayama'^l in-
dictates that spawning takes place close to shore.
The larvae and juveniles lead a pelagic existence
for about 3 mo, juveniles moving to shallow in-
shore areas at fork lengths (FL) between about 50
and 100 mm. The fish become sexually mature
males at 20-25 cm FL and subsequently undergo a
sex reversal, passing through a hermaphroditic
stage and becoming functional females between
30 and 40 cm FL. Adults inhabit inshore rocky and
sandy areas, frequently in zones of turbulence.
Methods
Donovan R. Craddock
Northtvest and Alaska Fisheries Center-
National Marine Fisheries Service. NOAA
2725 Montlake Boulevard East
Seattle. WA 98112
LUNAR SPAWNING OF THE THREADFIN,
POLYDACrVLL'S SEXFIUS, IN HAWAII'
Recent evidence indicates that lunar spawning
rhythms are more common in fishes than was once
thought. Johannes ( 1978) listed 50 species of tele-
ost fishes with lunar spawning rhythms, most of
them tropical and all of them marine or catadro-
mous. In the course of developing methods for cul-
turing the threadfin, Polydactylus sexftlis (Cuvier
and Valenciennes), we found that this species dis-
played a lunar spawning rhythm (May 1976). The
lunar pattern of spawning had been indicated by a
previous field study (Lowell 1971) and is consis-
tent with fishermen's lore (Hosaka 1944), but
proof was lacking and details of the rhythm were
unknown. In this paper we present detailed infor-
mation on the lunar spawning of P. se.x fills along
with observations of spawning behavior, using re-
sults from captive fish.
Polydactylus sex fills is a much sought-after food
fish in Hawaii and supports an important sport
fishery as well as a small commercial fishery
(Rao^). Information on the life history of this
'Contribution No. 552, Hawaii Institute of Marine Biology.
^Rao, T. R. 1977. Enhancement of natural populations of
moi (Polydactylus sexfilis) in Hawaii through the release of
hatchery-reared juveniles - a feasibility study of sea ranch-
ing. Univ. Hawaii, Hawaii Inst. Mar. Biol., Tech. Rep. 33, 46 p.
Hawaii Institute of Marine Biology, P.O. Box 1346, Kaneohe, HI
96744.
Juvenile P. sexfilis were captured by seine on
reef flats along windward Oahu in September and
October 1970 and reared to sexual maturity in
tidal ponds at Coconut Island, in Kaneohe Bay,
Oahu. The fish were daily fed chopped squid or
smelt, commercial trout chow (409r protein), or
trout chow supplemented with chopped squid. In
May 1973, 30 mature fish (18 females and 12
males) were transferred to a 18-m^ nylon net en-
closure suspended from Styrofoam'* floats and an-
chored off the leeward (southwest) side of Coconut
Island. In June and July 1973, a small number of
these fish were removed to laboratory tanks and
used in experiments on hormone-induced spawn-
ing. During this work, ovarian biopsy samples
were examined which contained residual eggs and
indicated that the fish had been spawning spon-
taneously. In order to monitor any such spawning,
an airlift egg collector was installed (May et al.^)
in the center of the net in July 1973 and operated
continuously (except for a few days when equip-
ment malfunctioned) between 14 July 1973 and 31
December 1975. Polydactylus sexfilis produces
pelagic eggs, so that the collector obtained a sam-
ple of eggs at each spawning. Every morning the
entire contents of the collector were harvested and
examined under a dissecting microscope, and the
number of P. sexfilis eggs was estimated by sub-
^Morris, D. E., and R. Kanayama. 1964-69. Life history
study of the moi, Polvdactvlus sexfilis. Job Completion Rep.,
Projects No. F-5-R-li to F-5-R-17, Div. Fish Game, State of
Hawaii. Division of Fish and Game, Honolulu, Hawaii.
■•Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
■■^May, R. C, G. 8. Akiyama, and M. T. Santerre. 1976. A
simple method for monitoring the spawning activity offish in net
enclosures. International Milkfish Workshop-Conference,
May 19-22, 1976, Tigbauan, Iloilo, Philipp. Working Pap. 10, p.
133-138. Southeast Asian Fisheries Development Center,
Kalayaan Building, Dela Rosa corner Salcedo Sts., Makati,
Metro Manila, Philipp.
900
FISHERY BULLETIN: VOL. 76. NO 4, 1979.
sampling. In 1976 and 1977, the collector was op-
erated only during the spawning periods, which
were predictable on the basis of data collected dur-
ing the previous years. Eggs of P. sexfilis were
distinguished from occasional eggs of other species
by their diameter (800-825 ^im) and by comparing
hatched larvae with larvae obtained from
hormone-induced spawnings; the identification
was further corroborated by rearing larvae to the
juvenile stage on several occasions. Since P.
sexfilis undergoes a male-female sex reversal, ad-
ditional males were added to the population each
year to maintain a female to male sex ratio be-
tween 1:1 and 1.5:1 during the spawning season.
In 1976, additional males were not added until the
third spawning' month, so data on spawning were
not available for the first two spawning months of
that year. At the end of this study in December
1977, the population of spawners numbered 57.
In order to determine the exact time of day when
spawning occurred, water was sampled from the
holding net continuously with a centrifugal pump
at 10 1/min and passed through a small collecting
basket (500-/i,m nylon mesh) on a barge anchored
next to the net. The collecting basket was moni-
tored visually, and the time when eggs first ap-
peared was noted. The superstructure of the barge
provided a barrier between the observer and the
holding net, so that activities associated with
monitoring the basket did not disturb the spawn-
ers in the net.
Results
Lunar Spawning Rhythm
Eggs of P. sexfilis were first observed in the
collector from 23 to 25 July 1973. Subsequently
the fish were found to spawn at night over a period
of 3-7 (in one instance, 10) days once each month,
always in proximity to the last quarter phase of
the moon (Figure 1). and in a spawning season
that extended from May or June to October ( Table
1). Because the lunar month is 29.5 days long and
does not coincide exactly with the calendar month,
the calendar dates of spawning (Table 1) were
generally 1-3 days earlier in each succeeding
month. The first spawning of each monthly series
was usually preceded by 1 or 2 days on which the
fish fed less actively than normal. Counts of P.
sexfilis eggs from the collector were made in Au-
gust 1973, and thereafter. Judging from the sam-
ples obtained by the collector, relatively few eggs
10
5
r
1 1 ■ ' ' 1 ' ' 1 ■ ' I 1
- 1973 1
1 T —
0
,ll
1., 1
1"
2 20
3
X 15- 1974
,
—
Ul 5
b.
O 0
ill
1
1
■
a:
UJ
m 25
Z
Z 20
1
MO)
15
1975
-
10
-
5
0
1
1
1 1
1 1 ^ 1 i M r r , 1 1 1 , - 1 . . , . 1 "i . . . . /i , , , /I
Figure l. — Numbers of Polydactylus sexfilis eggs obtained by
the airlift egg collector between August 1973 and October 1975.
In July 1973, eggs were noted on 3 days but were not counted.
Carets indicate the time of the last quarter phase of the moon.
Table l. — Dates on which eggs were produced by captive
Polydactylus sexfilis during study period. For each spawning
month, the upper date is the first observed day of spawning, the
lower date the last day.
Spawning monttn
Year
III
IV
VI
1973
(')
(')
(')
22 Aug
25 Aug
19Sept.2
22 Sept
18 0ct.2
21 Oct.
1974
13 [Way
9 June
9 July2
7 Aug
5 Sept
6 Oct.
15 May
14 June
14 July
13 Aug
10 Sept
9 Oct
1975
1 June2
29 June
29 July
27 Aug
25 Sept 2
25 Oct.
5 June
3 July
1 Aug.
30 Aug
29 Sept
27 Oct.
1976
(')
V)
17 July
15 Aug
1 3 Sept 2
1 3 Oct
20 July
18 Aug
16 Sept,
16 Oct
1977
9 May
7 June^
6 July
4 Aug
3 Sept
1 Oct
14 May
1 1 June
1 1 July
13 Aug
7 Sept
5 Oct
'Data not available
^Collector malfunctioned on previous day: initial spawning day possibly ear-
were produced on the first and last days of spawn-
ing, with a peak usually on the second or third day
(Figure 1). However, because the eggs collected
represented only a sample of the total number of
eggs produced, it is possible that the peaks
reflected some sampling variability.
The time of spawning relative to the lunar cycle
appeared to change as the spawning season pro-
gressed. The first spawning series of the year al-
ways began (Figure 2) on the day of the last quar-
ter (in one case the egg collector malfunctioned
prior to the last quarter, and it is possible that the
first day of spawning occurred earlier). In sub-
sequent months the series began 1-4 days prior to
901
the last quarter (Figure 2). In April 1974 the col-
lector obtained several thousand eggs over a 3-day
period about 12 days before the last quarter; al-
though the eggs were of the same diameter as P.
sexfilis eggs, the larvae were not reared for posi-
tive identification. In view of the consistency of
subsequent spawning data, and the low ovarian
weights which two previous studies found in P.
sexfilis in April (Lowell 1971; Morris*'), we believe
these were eggs of another species.
Some of the eggs found in the collector were
transparent, buoyant, and developing normally,
while others sank and were opaque and obviously
not viable. Among the spawnings in which 1,000
or more eggs were collected, an average of 52*%^ of
the eggs were viable (range, 0-989^). The airlift
collector apparently damaged the eggs to some
extent. For example, on 13 June 1974, when 75^^
of the eggs from the collector were viable, eggs
obtained by dip net at the time of spawning
showed over 90^'r viability. It is not known how
^Morris, D.E. 1964. Life history study of the moi, Po/vrfac-
tylus sexfilis. Job Completion Rep., Project No. F-5-R-11, Div.
Fish Game, State of Hawaii, 15 p. Division of Fish and Game,
Honolulu, Hawaii.
Days Before
4 3 2 1 3 1 2
Days After
3 4 5 6 7
■T I I ;
I
1
■ ' '. ■ '. 1
n
1
1
r— ' .
■ ■ ■ 1
x:
d TTT
^^^^^
1
o HI
\///>//////x
c
a nz:
1
1
V/////////i
1 ■11
CO
■ 1974
lU 1975
7A 1976
D 197 7
1 1 > 1 1
2:
1
1
\/yyyyyyyyy'\
\ ' 1
Hj^^^^^l
BHI
3ZI
1
V/////////^
i 1 1 1
,
Figure 2. — Duration of spawning cS. Poly dactyl us sexfilis rela-
tive to the time of the last quarter ((J) for the six spawning
months in 1974, 1975, and 1977. Data are given only for the last
four spawning months of 1976.
many of the fish in the net participated in each
spawning or whether the same fish spawned each
month, although Kanayama'^ believed that indi-
vidual P. sexfilis spawned more than once in a
season.
Time of Spawning
The developmental stage of eggs found in the
collector in the morning indicated that spawning
had occurred shortly after sunset. Visual monitor-
ing of water sampled continuously from the net on
34 spawning nights showed that with few excep-
tions the fish spawned between 2030 and 2130 h
(Figure 3, Table 2). The times recorded were those
when eggs were first observed in the collector; it is
possible that additional spawnings took place
slightly later on the same night, and behavioral
observations (see below) indicated that this may
have been true on at least some nights. The time of
spawning did not vary with the time of sunset
(Figure 3) and appeared unrelated to the time of
moonrise (which occurred generally between 2300
and 0400 h during the spawning season).
''Kanayama, R. 1967. Life history study of the moi,
Polydactylus sexfilis. Job Completion Rep., FVoject No. F-5-R-
15, Div. Fish Game, State of Hawaii, 9 p. Division of Fish and
Game, Honolulu, Hawaii.
06
02
22
-
§
LiJ
2
h-
14
10
06
-
_l l_
I II 111 IV V VI
SPAWNING MONTHS
Figure 3. — Times of .spawning oi^ Polydactylus sexfilis during
the six spawning months iroman numerals) of 1974 in relation to
the tidal cycle. Dots indicate observed times of spa wning in 1974,
and the horizontal lines delineate the time of spawning as indi-
cated by data from 1974, 1975, and 1977 (see Table 2). Dotted
horizontal lines show the times of sunset in 1974. Vertical lines
for each spawning night indicate time between the evening high
and low tides, i.e., the duration of the outgoing tide, as measured
by a tide gage at Coconut Island in 1974.
902
Table 2. — Times of first spawning in a captive population of
Polydactylus sexfilis on nights in 1975 and 1977. Where a single
time is given, the egg collector was examined continuously; in
other cases, the collector was examined at intervals of 5-15 min.
Numbers in parentheses indicate times of peak fish activity,
presumed to be the exact time of spawning.
Disc
ussion
Time of first
Time of first
Dale
spawning
Date
spawning
1975:
26 Sept
2045-2100(2052)
2 June
2115-2118
27 Sept
2054-2100
3 June
2110-2115
28 Sept
2050-2054 (2053)
4 June
2110-2115
29 Sept
2135-2140 (2138)
5 June
2115-2120
26 Oct
2050-2100 (2056)
30 June
2115-2120
27 Oct.
2105-2110 (2107)
2 July
2125-2130(2127)
1977.
3 July
2140-2145
7 June
2105
30 July
2045-2100(2058)
9 July
2147
31 July
2120-2125
10 July
2140
28 Aug.
2115-2118 (2117)
6 Aug
2050
29 Aug.
2129
4 Sept
2103
30 Aug.
2115-2118 (2117)
3 Oct.
2038
Data from a tide gage located at Coconut Island
were available for 1974 and showed that spawning
nearly always took place on the outgoing tide
(Figure 3). Although tide gage data were not
available for subsequent years, tides predicted
from tide tables showed patterns similar to the
1974 tide gage data. For 1975, 1976, and 1977, the
time of spawning (i. e., 2030-2130 h) was compared
with predicted tides and again found to occur
mostly during the outgoing tide. Spawning occur-
red on the outgoing tide in 739r of the spawning
nights during 1974-77.
Spawning Beha\ ior
Observations of spawning behavior were made
initially by watching bioluminescence caused by
the fish's movement; later, direct observations
were made by shining lights on the water at the
time of spawning. The level of activity of the fish
gradually increased beginning around 2015 h and
culminated in the spawning act as determined by
the appearance of eggs in the centrifugal pump
samples. Occasionally the fish broke the surface of
the water during the period of increased activity.
During courtship the fish swam rapidly around
the net in a circular manner in groups of two or
three. They appeared to be chasing one another,
and often one fish would contact another from be-
hind, either dorsally or ventrally, with snout.
Spawning appeared to take place between pairs
rather than among larger groups offish. Increased
activity usually continued for 20 or 30 min after
eggs were first noted.
The captive population of P. sc.xftlis was clearly
spawning with a well-defined lunar rhythm.
Other evidence implies that this is a natural be-
havior for this species. Lowell (1971) set gill nets
weekly in certain shoal areas of Oahu from April
to August 1970 and reported that exceptionally
large catches of P. se.xfilis per effort occurred "af-
ter the full moon and continuing until the last
quarter (1 week duration)," and because of the
stage of gonadal development among fish in such
catches, he termed them "spawning runs." In July
1970, female fish caught 3 days before the last
quarter all had well-developed ovaries, but fish
caught 4 days after the last quarter had spent
ovaries. Fishermen seem to have been aware of the
habits of P. scxfilis; for a long time: Hosaka
(1944:117) stated, "Moon light nights are best for
moi (= P. sexfilis) fishing, and this is especially
true when the moon is in the last quarter phase."
Spawning at the time of the last quarter phase of
the moon appears to be rare among fishes. Of the
50 lunar spawners which Johannes (1978) listed,
only two besides P. sexfilis spawned on the last
quarter; both of these are species of Amphiprion,
and one spawned on the first as well as the last
quarter. Since the various species covered in
Johannes' list occurred in different geographic lo-
cations, the variations in spawning days, taken
together with variations in spawning times, could
reflect local adaptations such as would occur if egg
or larval survival were related to tides or currents.
The coincidence of spawning in P. sexfilis with
the outgoing tide indicates that the remarkably
precise timing of spawning may act as a
mechanism for offshore dispersal of eggs and lar-
vae. Lowell (1971) noted that there was a strong,
oceanward current at his sampling site during
falling tide, when he estimated spawning occur-
red, and results of ichthyoplankton surveys indi-
cated that P. sexfilis eggs and larvae are not found
in inshore waters in Hawaii (Leis and Miller 1976;
Miller et al.»; Watson and Leis»). Johannes ( 1978)
^Miller, J. M., W. Watson, and J, M. Leis. 1973. Larval
fishes. In S. V. Smith (editor), Atlas of Kaneohe Bay, a reef
ecosystem under stress, p. 101-105. Univ. Hawaii Sea Grant
Tech. Rep. 72-1. Sea Grant College Program, University of
Hawaii, Honolulu, HI 96822.
^Watson, W., and J. M. Leis. 1974. Ichthyoplankton of
Kaneohe Bay, Hawaii: a one-year study of the fish eggs and
larvae. Univ. Hawaii Sea Grant Tech. Rep. 75-1, 178 p. Sea
Grant College Program, Universitv of Hawaii, Honolulu, HI
96822.
903
pointed out that spawning on outgoing tides is a
common phenomenon among coastal marine
fishes in the tropics, and he believed it evolved as a
strategy for ensuring that eggs and larvae are
transported away from the heavy concentration of
predators in shallow water. Johannes noted that
nocturnal spawning is also common in tropical
reef fishes and serves to reduce predation both on
the eggs and on the spawners.
The first P. sexfilis spawning of the year appears
to be anomalous in that it occurs relatively late
with respect to the last quarter. If offshore trans-
port confers an important selective advantage on
P. sexfilis, the lateness of the first spawning is
maladaptive because it results in release of eggs
early relative to the outgoing tide (see Figure 3).
The initial phase of the spawning season may thus
result in few viable offspring and could represent a
gradual initiation of the main spawning season,
delayed perhaps by the lower water temperatures
which usually prevail during the first spawning
month (Batheni").
No observations on the spawning behavior of a
polynemid fish have been published previously. In
P. sexfilis the sexes apparently pair and spawn
after a brief courtship involving rapid following
and nosing of one fish by another. The spawning
behavior of this species is similar in many respects
to that of the Pacific bonito, Sarda chiliensis, in-
cluding behaviors described by Magnuson and
Prescott (1966) as "circle swimming," "tail nos-
ing," and "following." The circling behavior noted
among P. sexfilis may have been imposed by the
confinement of the net enclosure, but S. chiliensis
also showed tight circling behavior at the time of
gamete release in a very large tank at Marineland
of the Pacific, and circling prior to spawning occurs
naturally in mullets ( Helfrich and Allen 1 975) and
some (perhaps many) other tropical fishes (R. E.
Johannes, Hawaii Institute of Marine Biology,
P.O. Box 1346, Kaneohe, HI 96844. Pers. com-
mun., December 1977). The circling behavior dur-
ing spawning observed in captive P. sexfilis thus
may not be abnormal for this species but may, as
Magnuson and Prescott (1966) theorized for S.
chiliensis, serve to enhance the probability of fer-
tilization.
'-^ Acknowledgments
We wish to thank R. E. Johannes for reading
and commenting on the manuscript, and Lloyd
Watarai, Steven Shimoda, Michael Muranaka,
and Michael Matsukawa for their help in main-
taining fish and collecting data. This work was
supported by the University of Hawaii Sea Grant
College Program, NOAA, Office of Sea Grant, U.S.
Department of Commerce, under Grant No. 04-3-
158-29, 04-5-158-17, 04-6-158-44026, and 04-6-
158-441 14, and by the State of Hawaii through the
Hawaii Insitute of Marine Biology, the Office of
the Marine Affairs Coordinator, and the Depart-
ment of Planning and Economic Development.
Literature Cited
Helfrich, P., and P. M. Allen.
1975. Observations on the spawning of mullet, Creni.*7iu^j/
crenilabis (Forskal), at Enewetak, Marshall Islands. Mi-
cronesica 11:219-225.
HOSAKA, E. Y.
1944. Sport fishing in Hawaii. Bond's, Honolulu, 198 p.
Johannes, R. E.
1978. Reproductive strategies of coastal marine fishes in
the tropics. Environ. Biol. Fishes 3:65-84.
Leis, J. M., AND J. M. Miller.
1976. Offshore distributional patterns of Hawaiian fish
larvae. Mar. Biol. (Berl.) 36:359-367.
Lowell, N. E.
1971. Some aspects of the life history and spawning of the
moi {Polydactylus sexfilis). M.S. Thesis, Univ. Hawaii,
45 p.
MAGNU.SON, J. J., AND J. H. PRESCOTT.
1966. Courtship, locomotion, feeding and miscellaneous
behaviour of Pacific bonito iSarda chiliensis). Anim.
Behav. 14:54-67.
May, R. C.
1976. Studies on the culture of the threadfin, Polydactylus
sexfilis, in Hawaii, FAO Technical Conference on
Aquaculture, Kyoto, Jpn., 26 May-2 June 1976. FIR:AQ/
Conf/76/E.5. FAO, Rome, 5 p.
ROBERT C. May
Hawaii Institute of Marine Biology
Kaneohe, Hawaii
Present address: Asian Development Bank
P.O. Box 789, Manila, Philippines
Gerald S. Akiyam.a
Michael T. Santerre
Hawaii Institute of Marine Biology
P.O. Box 1346, Kaneohe, HI 96744
'"Bathen, K. 1968. A descriptive study of the physical
oceanography of Kaneohe Bay, Oahu, Hawaii. Univ. Hawaii,
Hawaii Inst. Mar. Biol., Tech. Rep. 14, 353 p. Hawaii Institute of
Marine Biology, P.O. Box 1346, Kaneohe, HI 96744.
904
THE ROLES OF PRIOR RESIDENCE AND
RELATIVE SIZE IN ( OMPEII LION FOR
SHELTER BY LHE MALA^ SLAN I'RAW N,
MAC ROliltKLlULM KOSISBHRGIl^
Behavorial dominance, tenitoiiality. and their re-
lationship to survival and population density have
been the subject of extensive research (reviewed
by Brown and Orians 1970; Ito 1970; Brown 1975).
Generally dominance (behavioral) hierarchies
imply some form of ranked order (reviewed by
Marler and Hamilton 1966; Eibl-Eibesfeldt 1970;
Ito 1970) whereby the alpha animal(s) has prefer-
red access to food, shelter, or mates. Dominance
may develop within a short time after an initial
encounter (Dingle and Caldwell 1969), is partially
controlled by differences in relative size (Marler
and Hamilton 1966). and in some species is mod-
ified by relative location in space (Brown 1963).
This latter modification is related to Noble's
( 1939) original concept of territory. Noble referred
to territory as "any defended area." This area
could serve as a "retreat" ( in contrast to a sexual or
nesting area) that "is occupied because it is famil-
iar and defended because any newcomer is irritat-
ing to the resident."
Such space-related aggressive behavior has
been reported in numerous animals (Brown and
Orians 1970). Territorial behavior can be related
to "defense" of 1 ) a breeding area ( Buechner 1 96 1 ;
Watson 1964); 2) a renewable resource such as
food (Stimson 1970); or 3 ) a physical shelter
I Crane 1958; Reese 1964; Fielder 1965; Hughes
1966; Dingle and Caldwell 1969). Often the out-
come of such a defensive action is exclusion of the
intruder by the resident. Since this area is "famil-
iar" to the resident and unfamiliar to the "new-
comer," it follows that the resident has some type
of advantage. This "prior resident effect" has been
observed in a number of species (Braddock 1945,
1949; Miller 1958; Hughes 1966; Baird 1968;
Dingle and Caldwell 1969; Selander 1970). Thus
in many animals, spacing behavior is a powerful
mechanism that can regulate resource utilization
and influence distribution patterns.
Many of the above-mentioned studies and re-
views dealt with animal populations in natural
open systems subject to both immigration and
emigration. In contrast, aquaculture systems are
'Contribution No. 544 from the Hawaii Institute of Marine
Biology, University of Hawaii, Kaneohe, Hawaii.
closed and deal with confined high-density popula-
tions. In the case of Macrobrachi ion rosenbergii,
ponds are stocked with postlarvae, and harvesting
of adults begins 9-12 mo later. The same space-
related behavioral mechanisms observed in open
systems may be operating in these high-density
ponds. Circumstantial evidence indicates that this
is occurring in ponds containing M. rosenbergii.
Animals of the same age exhibit large variation in
size at the end of several months of growth
(Fujimura nad Okamoto 1970). Malecha (1977)
reported that small M. rosenbergii can greatly in-
crease their size when larger animals are absent.
This has been called the "Bull Effect" by Fujimura
and Okamoto (1970). Similar observations have
been reported for carp (Nakamura and Kasahara
1955, 1961; Wohlfarth and Moav 1972), trout
(Brown 1946), and salmon (Symons 1971). One
hypothesis advanced by Nakamura and Kasahara
( 1961) is that the larger animals are outcompeting
the smaller subordinates for food.
Maerobraehiiini rosenbergii is a large freshwa-
ter prawn. Its native distribution ranges from
Pakistan to Papua, New Guinea, and Palau
(Johnson 1960; McVey 1975). Usually it is found
in fresh and brackish streams and pools. The eggs
hatch near ocean waters, and the adults are found
up to 200 km from the coast (Ling 1969). Generally
males are thought to stay in upstream waters
while the females undergo a seasonal migration,
moving downstream and into brackish waters
(Raman 1967). Relatively little is known of M.
rosenhergii's behavioral ecology but Raman ( 1964)
reported juveniles "hiding in crevices or among
submerged plants along river banks." In order to
understand how social behavior affects resource
utilization by M. rosenbergii. three experiments
were conducted in which shelter was the limiting
resource, and relative size and prior residence
were measured as variables.
Methods
The three experiments consisted of: a prior resi-
dent experiment, a simultaneous introduction ex-
periment, and a control experiment. The prior res-
ident experiment was used to test for the role of
prior residence and relative size in competition for
shelter. The simultaneous introduction experi-
ment tested for the role of relative size on competi-
tion in the absence of a "prior resident effect." The
control experiment tested for the effect of handling
and capture.
FISHERY BULLETIN: VOL. 76. NO. 4. 1979.
905
Water conditions were maintained via an air lift
filter. The animals were fed a dry pellet diet ap-
proximately every other day (see diet #5, Balazs et
al. 1973). A 12-12 photoperiod with one-half hour
twilight lighting at "sunrise" and "sunset" was
employed.
Prior RLsiticnt 1 xpcrmicni
Eai'lier experiments revealed a foi'm of shelter
preference or selection operating in M. rosenhcrgii
(Peebles 1977). The shelters used in this experi-
ment were identical to those most frequently
selected by animals in the earlier experiments.
One shelter was placed in each experimental tank.
A shelter consisted of six concrete bricks arranged
into a double open ended square tunnel ( 19.3 x 19
X 11.4 cm tall).
Refuge othei" than the shelter was eliminated by
the use of oblong experimental tanks ( 137 x 75 x
92 cm deep) and the suspension of the air lift filters
just below the water surface ( the usual position for
these filters was on the bottom i. Water depth was
34 cm.
Adults from commercial ponds were placed in
two separate holding tanks, where they were kept
for no longer that 1 wk. Two animals were re-
moved, one each from the separate holding tanks.
Three body characteristics were measured: stan-
dard length (tip of telson to orbit of eye), and
lengths of left and right chelae. The animals were
tagged by means of a small plastic "bread bag
twist-tie" that was color coded and tied around the
tail. It took about 15 s to attach. Following tagging
the two animals were placed separately in ex-
perimental tanks. Three observations were made
before the introduction of the "immigrant" and
four observations were made after the introduc-
tion. The preintroduction observations were made
on the second, third, and seventh days after the
animals were placed in their separate experimen-
tal tanks. There were three observations per ani-
mal, each lasting 3 min. After the preintroduction
week a coin was flipped to determine which animal
would be the immigrant. The immigrant was des-
ignated as the introduced specimen and was
moved via a dip net from its tank to the resident's
tank. The resident was the animal that was not
moved from one experimental tank to another.
The postintroduction observations were made on
the day of introduction, and the second, third, and
seventh days after introduction. The observation
performed on the day of introduction was 15 min
and designed to monitor agonistic interactions as-
sociated with the initial encounters of the paired
animals. The remaining three postintroduction
observations were 3 min each and designed to re-
cord the animal's position within the tank. All
observations were made between 1000 and 1530.
Since these animals are nocturnal, movement and
behavioral interactions were minimal during the
daytime.
A total of 36 animals (18 immigrants, 18 resi-
dents) were used. Paired animals were of the same
sex. This controlled for the possible confounding
effect heterosexual courtship behavior might have
on competition foi- shelter occupancy.
Simultaneous Introdut tion Fxpcrinicnt
The treatment of the simultaneous introduction
experiment differed from the prior resident exper-
iment in four ways: 1 ) only males were used; 2) the
animals were simultaneously introduced into the
oblong tanks; 3) two additional body characteris-
tics were measured (body weight and carapace
length); and 4) the animals were not separately
observed prior to introduction.
Fifteen trials were run employing a total of 15
pairs or 30 animals. Observations were made on
the day of simultaneous introduction, and the sec-
ond, third, and seventh days after introduction.
The observation performed on the day of simul-
taneous introduction was 15 min and designed to
monitor agonistic interactions associated with ini-
tial encounters of the paired animals. The remain-
ing three postintroduction observations were 3
min each and designed to record the animal's posi-
tion within the tank.
Control Experiment
Eleven controls were run to test the effect of
handling. Animals were selected, measured, tag-
ged, and placed individually in experimental
tanks. One week later the control was netted, held
in the air, and reintroduced into the same experi-
mental tank. Observations were made for the
week before and the week after netting (mock im-
migration).
Operation.il Definitions
Successful: an animal that was in a shelter at
the end of the 7-day period following immigration.
Unsuccessful: an animal that was not in a shel-
906
ter at the end of a 7-day period following immigra-
tion.
Push: an aggressive act where one animal
pushes one of its chelae against the body of
another animal.
Nip: an aggressive act where one animal closes
down the tips of its chela on the body part of
another animal.
Tete-a-tete: a type of aggressive act charac-
terized by a head to head confrontation with at
least one nip or one push. The tete-a-tete appeared
to be difficult enough in orientation from the push
and the nip to be placed in a separate category.
Further observation and analysis might not sup-
port this separation.
Shove: an aggressive act where one animal
holds both chela forward and parallel while charg-
ing into the flanks of another animal.
Bout: an agonistic exchange between two ani-
mals where at least one aggressive act occurred. A
bout was considered terminated when aggressive
acts stopped or one animal moved away and was
not chased. Bouts were measured in units of ag-
gressive acts.
Bout length: the number of aggressive acts that
occurred during a bout.
Body characteristics: standard length (cen-
timeters*, right and left chelae length (centime-
ters), weight (grams), and carapace length (cen-
timeters).
Body size index: the number of body characteris-
tics in which an animal was larger. It was derived
as follows: animal A larger than animal B in stan-
dard length and right chela length, then A's body
size index is two. In the Prior Resident Experi-
ment three body characteristics were measured,
thus the maximum body size index in this experi-
ment was three. In the Simultaneous Introduction
Experiment five body characteristics were mea-
sured, thus in this experiment the maximum body
damage index was five.
Results
Control Experiment
Ten out of 11 animals were in the shelter on
every observation period before mock immigra-
tion. The remaining animal was in the shelter on
one of the three observation periods. The same 10
were in the shelters on all observations following
mock immigration, while the same remaining one
was never observed in a shelter after immigration.
It was concluded that the act of netting had no
effect on shelter use.
Prior Resident Experiment
Shelters were occupied on every observation by
every animal during the preimmigration week.
Following immigration all shelters were occupied
on every observation period. On several occasions
more than one animal was in a shelter during the
first two observation periods following immigra-
tion. However, by the end of the week, observation
period 4, one animal was in a shelter while the
other was usually at the opposite end of the tank.
When the data were examined by immigrant ver-
sus resident for shelter use over the 7-day period,
an interesting change became apparent (Figure
1). On the day of immigration, residents were oc-
cupying shelter significantly more often than im-
migrants (Binomial Test,P = 0.044, Siegel 1956).
By the second observation period and for the re-
maining two observations there were no sig-
nificant differences between residents and immi-
gi-ants in frequency of shelter use (Binomial Test:
day 2 after immigration,P = 1.0, day 3,P = 0.814;
day 7,P = 0.814).
Examining the data for the effect of size (Figure
2) revealed that successful animals were sig-
nificantly larger than their unsuccessful paired
12 3 4
OBSERVATION PERIOD
Figure l. — Shelter usage by observation period for 18 pairs of
Macrobrachium rosenhergii. The data from prior resident exper-
iments are summed for the 18 pairs. During observation period 1,
18 residents (circles) and 7 immigrants (dots) were inside shel-
ters. On observation period 1 there were seven cases of double
occupancy; observation period 2, two cases; and observation
period 3, none.
907
=> < -7
u <
u
Z) ->
Ll_ 1/5
o ^
□£ U
z) z 1
Z 3 '
BODY SIZE I NDEX
Figure 2. — Frequency of relative body size for successful (open
bar) and unsuccessful (solid bar) prawns in the prior resident
experiment. A body size index of one indicates one Macro-
hrachium rosenbergii was larger than the other in one body trait
but smaller in the other two body traits.
partners (Kolmogorov-Smirnov Two Sample Test:
D max = 11,/? = 18,P = 0.01).
SiniiiltaneoLi,>> Introduction Experiment
A similar effect of size on shelter use was ob-
served in the simultaneous introduction experi-
ment (Figure 3; r, = 0.579, P<0. 001). Once again
larger animals used the shelters more often than
their smaller partners.
Aggressive behavior was observed only on the
day of introduction. The nip and push occurred
more often than the shove or tete-a-tete ( Figure 4).
Generally aggressive interactions were limited to
a few (one to three) bouts per 15 min (Figure 5),
and these bouts were usually one or two aggres-
sive acts long (Figure 6).
Prior Resident Experiment by Simultaneous
Introduction Experiment
A Kolmogorov-Smirnov chi-square approxima-
tion (Goodman 1954; Siegel 1956) revealed that
animals of the simultaneous introduction experi-
ment were more aggressive on the day of introduc-
tion than were animals in the prior resident exper-
iment on the day of immigration (x^ = 15.54,
P<0.002 for number of bouts/animal per 15-min
period; x^ - 13.877, P<0.002 for number of ag-
40n
3 5-
N = NIP
P = PUSH
T=TETE-A-TETE
>30-
S= SHOVE
Z25-
LJ
320-
O
^15-
^10-
5-
1
N
T
49
to Qi
m ^!^3
o
LU CO
Z)
O
>2
LU LU
-^ CO
1-20-
U
O
Li.
5-
1 2 3 4 5 6 7 8
BOUT LENGTH
Figure 6. — Frequency of bout length ( number of aggressive acts
per bout per animal) for the simultaneous introduction of 14
pairs of male Macrobrachium rosenbergii.
gressive acts/animal per 15-min period). Simul-
taneous introduction animals exhibited a total of
68 aggressive acts occurring in 43 bouts in =28
animals), while the prawns from the prior resident
experiment exhibited only five aggressive acts in
five bouts in = 32 animals).
Discussion
The results indicate that when M. rosenbergii
compete for shelter at least three factors, i-elative
size, prior residence, and length of time contes-
tants are paired, play important roles in determin-
ing who occupies a shelter. It has long been recog-
nized that in crustaceans relative size plays a
large role in determining dominance (Allee and
Douglas 1945; Bovbjerg 1953, 1956, 1960; Lowe
1956). More recent observations have confirmed
the size dominance relationship (Hughes 1966;
Crane 1967; Griffin 1968; Hazlett 1968; Dingle
and Caldwell 1969; Warner 1970; Rubenstein and
Hazlett 1974; Jachowski 1974; Molenock 1976;
Sinclair 1977). However, relative size does not ap-
pear equally important in all species (Hazlett and
Estabrook 1974).
In prawns, relative size strongly influences the
outcome of competition. When two prawns en-
counter one another in an area new to both, the
larger animal usually has the advantage. Often
these encounters are characterized by a limited
series of pushes with one or the other chela. The
function of the pushing might be threefold: 1) to
test their opponent's weight (rest inertia), 2) to
determine the opponent's molt state, and 3) to see
if the opponent is capable of pushing back (has
chelae). Other crustaceans appear to measure
their opponent's physical strength by means of
physical interactions involving the chelae (Griffin
1968; Schone 1968). In Cambarellus shufeldtii,
claw removal causes dominant animals to drop in
rank (Lowe 1956). InM. rosenbergii deaths related
to agonistic behavior usually occurred near ec-
dysis and often the first appendages lost during an
agonistic encounter were the chelae (Peebles
1977).
Smaller animals have been observed success-
fully defending shelters from attempted occupa-
tion by larger congeners (Bovbjerg 1953; Griffin
1968; Sinclair 1977). This is related to the prior
resident phenomenon and it is central to Nobel's
(1939) definition of territory. Resident M. rosen-
bergii, regardless of their relative size, success-
fully retained their shelters. The mechanism the
residents employed apparently was not limited to
direct physical interaction. Immigrants and resi-
dents seldom fought. Generally immigrants were
inactive upon placement into a tank housing a
resident. The immigrant's aggressive behavior
was well below its counterpart in the simultane-
ous introduction group. Only occasionally (Figure
1) did the immigrant seek out the shelter. This
latter behavior is in direct contrast to the control
gi'oup. A control group animal was usually back in
its shelter within 1 min after reintroduction. Pos-
sibly an exocrine was an agent of communication
between resident and immigrant prawns, since a
novel environment did not inhibit exploration in
animals of the simultaneous introduction experi-
ment; and animals from the control experiment
reintroduced into tanks contaminated with their
own exocrines, rapidly entered their shelter.
The advantage conferred upon resident M.
rosenbergii appears to disappear within a short
period of time. The smaller resident can defend its
shelter against intrusion for no longer than a few
days (Figure 1). Apparently relative size can over-
come the prior resident effect if resident and im-
migrant continue to encounter one another. Simi-
lar observations were reported by Lowe ( 1956). In
the case of the C. shufeldtii, a dominance hierar-
chy was established before shelters were intro-
duced. Dominant C. shufeldtii displaced subordi-
nates from occupied shelters. In my experiments,
M. rosenbergii first exhibited territoriality as de-
termined by the presence of the prior resident ef-
fect. Territoriality then broke down, due to con-
tinued encounters, into simple dominance.
909
The important point addressed in this paper is
not who wins or loses the encounter but which
animal gains access to the I'esource. Investigators
whose observations w^ere limited to the first en-
counter might suggest that residents almost al-
ways outcompete intruders foi- shelter. However. 1
have shown that in a closed system the prior resi-
dent effect breaks down into simple size-related
dominance. These results offer a behavioral ex-
planation for the known and i-ecognized bull effect
in praw n aquaculture ponds. Larger animals have
preferential access to food and shelter, two impor-
tant resources which are often dispersed in a
clumped or patchy fashion.
Acknow Icdgnicnts
I wish to thank E. Reese, A. Kinzie, S. Malecha,
R. May, T. Smith, and the editors and reviewers of
Fishery Bulletin for the advice they gave during
the development of this paper. Most of the work
was carried out at the Hawaii Institute of Marine
Biology with funds awarded by University of
Hawaii Sea Grant College Program (through In-
stitutional Grant No. 04-3-158-29 from NOAA
Office of Sea Grant) and administered through the
Department of Genetics at the University of
Hawaii.
Literature Cited
allee, w. C, and M. B. DOUGLIS.
1945. A dominance order in the hermit crab, Pagurus lon-
gicarpus Say. Ecology' 26:411-412.
B.AL.AZS, G. H., E. Ross, .AND C. C. BROOKS.
1973. Preliminary studies on the preparation and feeding
of crustacean diets. Aquaculture 2:369-377.
BAIRD, R. C.
1968. Aggressive behavior and social organization in Mol-
lienesia latipinna Le Sueur. Texas J. Sci. 20:157-176.
BOVBJERG. R. V.
1953. Dominance order in the crayfish Orconectes virilis
(Hagen). Physiol. Zool. 26:173-178.
1956. Some factors affecting aggressive behavior in
crayfish. Physiol. Zool. 29:127-136.
1960. Behavioral ecology of the crab, Pachygrapsus cras-
sipes. Ecology 41:668-672.
BRADDOCK, J. C.
1945. Some aspects of the dominance-subordination rela-
tionship in the fish Platypoecilus maculatus. Physiol.
Zool. 18:176-195.
1949. The effect of prior residence upon dominance in the
fish Platypoecilus 7Tiaculatus. Physiol. Zool. 22:161-169.
Brown, J. L.
1963. Aggressiveness, dominance and social organization
in the Stellar Jay. Condor 65:460-484.
1975. The evolution of behavior. W. W. Norton & Co..
N.Y., 761 p.
BKOWN, J. L., AND G. H. ORIANS.
1970. Spacing patterns in mobile animals. Annu. Rev.
Ecol. Syst. 1:239-262.
Brown, M. E.
1946. The growth of brown trout tSalmo Truita Linn.) I.
Factors influencing the growth of trout fry. J. Exp. Biol.
22:118-129.
BUECHNKR. H. K.
1961. Territorial behavior in Uganda Kob. Science
(Wash., DC.) 133:698-699.
Crank, J.
1958. Aspects of social behavior in fiddler crabs, with spe-
cial reference to Uca maracoani (Latreille). Zoologica
(N.Y.) 43:113-130.
1967. Combat and its ritualization in fiddler crabs
• Ocypodidae) with special reference to Uca rapax
(Smith). Zoologica (N.Y.) 52:49-76.
Dingle, H., and R. L. Caldwell.
1969. The aggressive and territorial behaviour of the man-
tis shrimp Gonodactylus bredini Manning (Crustacea:
Stomatopoda). Behaviour 33:115-136.
ElBL-ElBESFELDT, L
1970. Ethology, the biology of behavior. Holt, Rinehart
and Winston, N.Y., 530 p.
FIELDER, D. R.
1965. A dominance order for shelter in the spiny lobster
Jasus lalandei (H. Milne-Edwards). Behaviour 24:236-
245.
Fujlml'ra, T., and H. Okamoto.
1970. Notes on progress made in developing a mass cultur-
ing technique for Macrobrachium rosenbergii in
Hawaii. Indo- Pacific Fisheries Council (FAO), Thai-
land.
G<:)ODMAN, L. A.
1954. Kolmogorov-Smirnov tests for psychological re-
search. Psychol. Bull. 51:160-168.
Griffin, D. j. G.
1968. Social and maintenance behaviour in two Austra-
lian ocypodid crabs (Crustacea: Brachyura). J. Zool.
(Lond.) 156:291-305.
Hazlett, B. a.
1968. Size relationship and aggressive behavior in the
hermit crab Clibanarius vittafus. Z. Tierpsychol.
25:608-614.
Hazlett, B. A., and G. F. Eastbrook.
1974. Examination of agonistic behavior by character
analysis. I. The spider crab Microphrys bicornutus. Be-
havior 48:131-144.
HI'(;HES, D. a.
1966. Behavioural and ecological investigations of the
crab Ocypode ceratophthalmus (Crustacea: Ocyp-
odidae). J. Zool. (Lond.) 150:129-143.
ITO, Y.
1970. Groups and family bonds in animals in relation to
their habitat. In L. R. Aronson, E. Tobach, D. S.
Lehrman, and J. S. Rosenblatt (editors). Development and
evolution of behavior, p. 389-415. W. H. Freeman, San
Franc.
JACHOWSKI, R. L.
1974. Agonistic behavior of the blue crab, Callinectus
sapidus Rathbun. Behaviour 50:232-253.
Johnson, D. S.
I960. Some aspects of the distribution of freshwater or-
ganisms in the Indo-Pacific area, and their relevance to
the validity of the concept of an Oriental region in
910
zoogeography. Proc. of the Centenary and Bicentenary
Congress of Biology, Singapore, 1958, p. 170-181.
Ll.\(;, S. \V.
1969. The general biology and development of Macru-
hrachium rosenhi'rgii ( de Mani. FAO Fish. Rep.
57:589-606.
Lowe, M. E.
1956. Dominance-subordinance relationships in the
crawfish Camharellus shufeldtit. Tulane Stud. Zool.
4:139-170.
M.M.KCH.A, S. R.
1977. Genetics and selective breeding of M. rusenber-
gii. In J. A. Hanson and H. L. Goodwin (editors), Shrimp
and prawn farming in the Western Hemisphere, p. 328-
355. Dowden, Hutchinson & Ross, Inc., Stroudsburg, Pa.
M.ARLKR, P., .AND W. J. H.AMILTON III.
1966. Mechanisms of animal behavior. John Wiley &
Sons, N.Y., 771 p.
McVey, J. p.
1975. New record of Minn)brachium rosenbergii ( De Mani
in the Palau Islands (Decapoda, Palaemonidaei. Crus-
taceana 29:31-32.
MILLER, R. B.
1958. The role of competition in the mortality of hatchery
trout. J. Fish. Res. Board Can. 15:27-45.
MOLEN'OC'K, J.
1976. Agonistic interactions of the crab Pe/ro/;.s7/!es (Crus-
tacea, Anomura). Z. Tierpsychol. 41:277-294.
N.AK.A.ML'R.A, N., .AN'D S. K..\.S,AH.-\R.\.
1955. A study on the phenomenon of the tobi-koi or shoot
carp — I. On the earliest stage at which the shoot cai-p
appears. Bull. Jpn. Soc. Sci. Fish. 21:73-76.
1961. A study on the phenomenon of the tobi-koi or shoot
carp — IV. Effects of adding a small number of larger indi-
viduals to the experimental batches of carp fry and of
culture density upon the occurrence of shoot carp. Bull.
Jpn. Soc. Sci. Fish. 27:958-962.
Noble, G. K.
1939. The role of dominance in the social life of birds. Auk
56:263-273.
Peebles, J. B.
1977. Movement, dispersion and mortality in Macro-
brachium rosenbergii. Ph.D. Thesis. University of
Hawaii, Honolulu, 220 p.
Raman, K.
1964. On the location of a nursery ground of the giant
prawn Macrobrachium rosenbergii (de Man). Curr. Sci.
(India) 33:27-28.
1967. Observations on the fishery and biology of the giant
freshwater prawn Macrobrachium rosenbergii de
Man. In Symposium on Crustacea, Pai-t I, p. 649-669
Mar. Biol. Assoc. India, Ernakulum.
Reese, E. S.
1964. Ethology and marine zoology. Oceanogr. Mar.
Biol. Annu. Rev. 2:455-488.
RLBENSTEIN, D. I., AND B. A. HAZLETT.
1974. Examination of the agonistic behaviour of the
crayfish Orconectes virilis by character analysis. Be-
haviour 50:193-216.
SCHONE, H.
1968. Agonistic and sexual display in aquatic and semi-
terrestrial brachyuran crabs. Am. Zool. 8:641-654.
SELANDER, R. K.
1970. Behavior and genetic variation in natural popula-
tions. Am. Zool. 10:53-66.
SlEGEL, S.
1956. Nonparametric statistics for the behavioral
sciences. McGraw-Hill Book Co., N.Y., 312 p.
SLNCLAIR, M. E.
1977. Agonistic behaviour of the stone crab, Menippe mer-
cenaria (Say). Anim. Behav. 25:193-207.
Stlmson, J.
1970. Territorial behavior of the owl limpet, Lottia
gigantea. Ecology 51:113-118.
symons, p. E. K.
1971. Behavioural adjustment of population density to
available food by juvenile Atlantic salmon. J. Anim.
Ecol. 40:569-587.
WARNER, G. F.
1970. Behaviour of two species of grapsid crab during in-
traspecific encounters. Behaviour 36:9-19.
W.ATSON, A.
1964. Aggression and population regulation in red
grouse. Nature (Lond.) 202:506-507.
WOHI.FARTH, G. W., AND R. MOAV.
1972. The regression of weight gain on initial weight in
carp. I. Methods and results. Aquaculture 1:7-28.
John B. Peebles
Hawaii Institute of Marine Biology
University of Hawaii
Kaneohe. HI 96744
PRINCIPAL SPAWNING AREAS AND TIMES
OF MARINE FISHES,
CAPE SABLE TO CAPE HATTERAS
The purpose of this compendium is to summarize
spawning areas and seasons of the more abundant
marine fishes of the continental shelf between
Cape Sable, N.S., and Cape Hatteras, N.C., as an
aid to the identification offish eggs and larvae and
planning and scheduling ichthyoplankton sur-
veys. We have used the term "marine" to encom-
pass fishes which spawn at sea (in contrast to es-
tuarine spawners), although some of the species
included spawn in both environments contingent
on geographic location ( e.g., winter flounder which
spawn exclusively in estuaries in the Middle At-
lantic Bight and offshore in the Gulf of Maine and
Atlantic menhaden which spawn in estuaries
along southern New England and in the New York
Bight and offshore in the lower Middle Atlantic
Bight and in the South Atlantic Bight).
The Gulf of Maine is defined as the oceanic bight
bounded by Nantucket Shoals and Cape Cod on
the west (long. 70°W) and Cape Sable on the east
(long. 65°W) including Georges and Browns
Banks and waters out to the 200-m contour (Col-
ton 1964). The Middle Atlantic Bight is the area
FISHERY BULLETIN VOL. 76, NO. 4. 1979. 911
inshore of the continental slope bounded by Cape
Cod and Nantucket Shoals to the east (long. 70°W )
and Cape Hatteras to the south (lat. 35°N). The
New York Bight, as defined in the MESA New
York Bight Atlas Monograph Series (Bowman and
Wunderlich 1977), is the offshore waterarea in the
bend of the Atlantic coastline from Long Island
(long. 7r30'W) to New Jersey (lat. 38°30'N). A
chart of the Gulf of Maine and Middle Atlantic
Bight and the names of places and areas referred
to in the spawning summary are given in Fig-
ure 1.
76*
74*
72*
70'
68'
66'
46*
44'
42<
40*
38*
36*
76** 74** 72** 70" 68"
FIGURE 1.— Orientation chart of the Gulf of Maine and Middle Atlantic Bight.
46*
44'
42*
40*
38*
36*
6 6*
912
In this summary (Table 1) we have treated the
Gulf of Maine and the Middle Atlantic Bight sepa-
rately for there is an abrupt general division be-
tween the biological and physical properties of
water east and west of Cape Cod. The boreal wa-
ters over most of the Gulf of Maine are well mixed
by strong tidal currents, while the circulation of
the warmer shelf waters west of Cape Cod is more
sluggish, and its chemical and physical properties
are less complex (Colton 1964). The offing of Cape
Cod also appears to be a definite transition zone
(probably thermal) for some northern and south-
ern species of fishes and invertebrates, both
pelagic and benthic (Colton 1964). The species
composition and abundance of fishes vary marked-
ly between the two regions, with boreal, non-
migratory species dominating the Gulf of Maine
and warmwater, migratory species prevailing in
the Middle Atlantic Bight. The bulk or total
spawning of many species of fishes is restricted to
areas east (e.g., haddock, pollock, redfish) or west
(e.g., bluefish, menhaden, anchovies) of Nantuck-
Table 1. — Principal spawning areas and times of marine fishes. Cape Sable to Cape Hatteras.
Family
Species
Comon
Name
Gulf of Maine
Sub Area J
.lantic Right
Sub Area
J
F
M
A
M
J
J
A
S
0
N
D
F M
A
M
J
J A
b 0
N
u
Clupeidae
Brevoortia
tyrannus
Atlantic
menhaden
t
* *
*
Clupea harenqus
harenqus
Atlantic
herring
Georges Bank
*
*
N. of Delaware
i
j
*
1
Western Nova
Scotia
!
*
1
1
i i
i
1
1 \ 1
1
Jeffries Ledge & ; i ' '
Stellwagen Bank i
i , K-
!
1
i
1
1
1
Engraulidae
Anchoa hepsetus
striped
anchovy
Nantucket Shoals
' !
1
j
i
1
1
*
• 1
1
' j
•
Engraulis
eurystole
silver
anchovy
1
1 ,
Gadidae
Brosme brosme
Enchelyopus
cimbrius
cusk
fourbeard
rockling
*
*
*
*
1
Gadus morhua
! Melanograiimus
aeqlefinus
Atlantic
cod
haddock
Georges Bank
Browns Bank
Nantucket Shoals
Georges Bank
*
1
—
k-
--
*
*
!
1 I
1
1 '
i
* i * *
i : . . ,
,
Browns Bank
' *
■ *
1 .
' ; !
,.,...
I
South Channel
* ,
I
^ '
1
1
I
Merlucdus
albidus
offshorp
hake
1
i
i
!
"TT^
-i ■
1
Merluccius
bilinearis
silver
heke
NE Georges &
Cent. Gulf
;
* *
Nant. Shoal s-
|Virginia
j
-
*
1 !
1
1
Southern
Georges
■ * , *
i
i • ' ; ■ !
1
1 ' i
1
1
Pollachius
virens
1
pollock
Mass. Bay _^
Stellwagen
South Channel
* .
• • :
i
i
1
' Urophycis
1 Chester!
long
finned
hake
1
. ■ i : 1 ■ NY Bight ,,,
[ '. . 1
! * i : i
j ' i ' i
'l
1
1 1
1 1 1
1 "
j Urophycis chuss
red hake
S. Georges
Nant. Shoals
;
; * *
1* {
,'
* '*
1
■ Urophycis reqius
spotted
hake
i NY Bight-
, 1 C. Hatteras
*
i Urophycis tenuis
Pomatomus
saltatrix
white hake
bluefish
I Cont. Slope *
' *
___ * J
i ■ . !
1 i
* *
1
1 Scianidae
Leiostomus
xanthurus
spot
j , i Ches. Bay-
1 1 Cape Hatteras
. i : : *
Micropogon
undulatus
Atlantic
croaker
1
Ches. Bay-
Cape Hatteras
* ' *
1
Cynoscion
1 regal is
weakfish
; Ches. Bay- I
1 , 1
1 ; ! 1 Montauk, LI 1
1 ' 1 1 ill 1
1
913
Table L— Continued.
Family
Species
Common
Name
Gulf of Maine
Middle Atlantic Biqht
Sub Area
J
F
M
A
H
J
J
A
5
0
N
D
Sub Area
J
F
M
A
M
J
J
A
S
0
N
D
Labridae
Tautoqa onitis
tautog
cunner
Mass. Bay
S. Georges
Nant. Shoals
*
--
*
\
Tautogolabrus
■k
.
adspersus
i
Scombri dae
Scomber
scombrus
Atlantic
mackerel
W. Gulf
Cape Cod Bay
1
i
*
Cape Cod-
Chesapeak Bay
' * i
Scorpaenidae
Sebastes
marinus
redfish
Scotian Shelf &
* *
] 1
i
!
Cent. Gulf i '
Triglidae
Prionotus
carolinus
northern
searobin
i ' 1 ,
' 1
Block Island-
1
j
1
Cape Hatteras
Cottidae
Hyoxocephalus
octodecem-
spinosus
longhorn
sculpin
* ' 1 i
: 1
1
i *
1
i
i ;
Amnodytidae
Stromateidae
Ammodytes sp.
Peprilus
triacanthus
sand lance
* * J ' '
u ^
*
butterfish
SW Georges
Nant. Shoals :
* *
1
* *
j
1
1
1 1
Bothidae
Citharichth^s
arctitrons
Gulf
Stream
flounder
SW Georges '
Nant. Shoals .
.„-ui-4 1 :■ 1-
1
1 i*i
1
i
Hippoglossina
oblonqa
fourspot
flounder
Nant. Shoals-
' ' ' 1
: *
1
South ' ;
j
■ I
1
;
I
1
Paralichthys
dentatus
summer
flounder
Nant. Shoals- ' ' _^_
South '
1
¥ 1
1
1* : i 1
'
Scophthalinus
aquosus
windowpane
Georges Bank |
Nant. Shoals- !
South !
--
--
"
--
' i
! i
i ' '
*
Pleuronec-
tidae
Glyptocephalus
cynoglossus
witch
flounder
1
*
J
! :
Cape Cod-
Delaware Bay
*
*
1
!
1
1
Hippoqlossoides
platessoides
American
plaice
1
i
*
*
South of
Martha's Vine-
yard
-
--
1 1
1
Limanda
ferruqinea
yellowtail
flounder
Browns Bank | j
! 1 . '
: 1 '
*
1 ;
*
*
1
1
f^"
t
-
1
Georges Bank
Nant. Shoals-
South
1
*
*
1 '
1
1
1
1
1
1 1
! ;
1
I
Pseudopleuro-
nectes
americanus
winter
flounder
Georges Bank
1
1
1
i
j
1 '
r
1
1
Known spawning season.
Uncertain spawning season.
*Peak spawning.
et Shoals, although there are exceptions to this
general rule (notably, yellowtail flounder and
silver hake).
The spawning summaries are based primarily
on published data collected on Bureau of Commer-
cial Fisheries (now National Marine Fisheries
Service) ichthyoplankton surveys of the Gulf of
Maine and Middle Atlantic Bight made in the
1950's and 1960's and listed in the References.
Published data from earlier studies (e.g., Fish
1929; Walford 1938; Pearson 1941; Sette 1943;
Bigelow and Schroeder 1953) and some unpub-
lished information from more recent National
Marine Fisheries Service ichthyoplankton sur-
veys have also been utilized. We have not at-
tempted to make the bibliography encyclopedic.
However, the papers cited include references to all
pertinent spawning summaries. Spawning areas
and seasons were determined on a basis of the
occurrence of eggs and/or early stage (yolk-sac)
larvae. The families are arranged in phyletic
sequence (Greenwood et al. 1966) and the species
are listed in alphabetical order. Common names
follows those recommended by Bailey et al. ( 1970).
References
Bailey, R. M., J. E. Fitch, E. S. Herald, E. A. Lachner, C. C.
LINDSEY, C. R. ROBINS, AND W. B. SCOTT.
1970. A list of common and scientific names of fishes from
the United States and Canada. 3d ed. Am. Fish. Soc. Spec.
Publ. 6, 150 p.
914
BIGELOVV, H. B., AND W. C. SfHROKDKK.
1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv.,
Fish. Bull. 53, 577 p.
Bowman, M. J., and L. D. Windkki.ich.
1977. Hydrographic properties. MESA N.Y. Bight Atlas
Monogr. 1, 78 p.
COLTON, J. B., Jr.
1964. History of oceanogi-aphy in the offshore waters of the
Gulf of Maine. U.S. Fish Wildl. Serv., Spec. Sci. Rep.
Fish. 496, 18 p.
CoLTON, J. B., Jr., and J. M. St. Onck.
1974. Distribution of fish eggs and larvae in continental
shelf waters. Nova Scotia to Long Island. Ser. Atlas Mar.
Environ., Am. Geogr. Soc. Folio 23.
Fahay, M. p.
1974. Occurrence of silver hake, Merluccius hilinearis,
eggs and larvae along the Middle Atlantic continental
shelf during 1966. Fish. Bull., U.S. 72:813-834.
FISH, C. J.
1929. Production and distribution of cod eggs in Mas-
sachusetts Bay in 1924 and 1925. U.S. Bur. Fish., Bull.
43(21:253-296.
Greenwood, P. H.. D. E. Ro.sen, S. H. Weitzman, and G. S.
Myers.
1966. Phyletic studies of teleostean fishes, with a provi-
sional classification of living forms. Bull. Am. Mus. Nat.
Hist. 131:339-455.
HiGH.\M, J. R., AND W. R. Nicholson.
1964. Sexual maturation and spawning of Atlantic
menhaden. U.S. Fish Wildl. Serv., Fish. Bull. 63:255-
271.
Kendall, A. W., Jr., and J. W. Reintjes.
1975. Geographic and hydrographic distribution of Atlan-
tic menhaden eggs and larvae along the Middle Atlantic
coast from RV Dolphin cruises, 1965-66. Fish. Bull.,
U.S. 73:317-335.
Marak, R. R., and J. B. Colton, Jr.
1961. Distribution offish eggs and larvae, temperature,
and salinity in the Georges Bank-Gulf of Maine area,
1953. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 398,
61 p.
Marak, R. R., J. B. Colton, Jr., and D. B. Foster.
1962. Distribution of fish eggs and larvae, temperature,
and salinity in the Georges Bank-Gulf of Maine area,
1955. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 411,
66 p.
Marak, R. r., J. B. Colton, Jr., D. B. Foster, and D. Miller.
1962. Distribution of fish eggs and larvae, temperature,
and salinity in the Georges Bank-Gulf of Maine area,
1956. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 412,
95 p.
NoRCROss, J. J., S. L. Richardson, W. H. Massman, and E. B.
Joseph.
1974. Development of young bluefish iPomatomus salta-
trix) and distribution of eggs and young in Virginian
coastal waters. Trans. Am. Fish. Soc. 103:477-497.
Pearson, J. C.
1941. The young of some marine fishes taken in lower
Chesapeake Bay, Virginia, with special reference to the
gray sea trout, Cynoscion regalis (Bloch). U.S. Fish
Wildl. Serv., Fish. Bull. 50:79-102.
Richards, S. W., and a. W. Kendall, Jr.
1973. Distribution of sand lance, A/?!woc/y/t's sp., larvaeon
the continental shelf from Cape Cod to Cape Hatteras
from RV Dolphin surveys in 1966. Fish. Bull., U.S.
71:371-386.
Richardson, S. L., and E. B. Joseph.
1973. Larvae and young of western north Atlantic bothid
flatfishes Etropus microstomus and Citharichthys arcti-
frun^ in the Chesapeake Bight. Fish. Bull., U.S.
71:735-767.
SE'H'E, O. E.
1943. Biology of the Atlantic mackerel wScomber scom-
brus) of North America. Part 1: Early life history, includ-
ing growth, drift, and mortality of the egg and larval
populations. U.S. Fish Wildl. Serv., Fish. Bull. 50:149-
237.
S.MITH, W. G.
1973. The distribution of summer flounder, Paralichthys
dentatus, eggs and larvae on the continental shelf be-
tween Cape Cod and Cape Lookout, 1965-66. Fish. Bull.,
U.S. 71:527-548.
S.MITH, W. G., J. D. SIBUNKA, AND A. WELLS.
1975. Seasonal distributions of larval flatfishes (Pleuro-
nectiformes) on the continental shelf between Cape Cod,
Massachusetts, and Cape Lookout, North Carolina,
1965-66. U.S. Dep. Commer., NOAA Tech. Rep. NMFS
SSRF-691, 68 p.
JOHN B. Colton, Jr.
Northeast Fisheries Center Narragansett Laboratory
National Marine Fisheries Service. NOAA
Narragansett. RI 02882
Wallace G. Smith
Arthur W. Kendall, Jr.
Peter L. Berrien
Michael P. Fahay
Northeast Fisheries Center Sandy Hook Laboratory
National Marine Fisheries Service. NOAA
Highlands. NJ 07732
RECENT SIGHTINGS OF THE BLUE WHALE,
BALENOPTERA MUSCULUS, IN
THE NORTHEASTERN TROPICAL PACIFIC
The blue whale, Balenoptera rnusculus, in the
North Pacific, migi'ates to the Gulf of Alaska and
Aleutians in the summer for feeding (Nishiwaki
1966). It is believed to migrate to tropical waters
in winter for calving, but sightings of blue whales
in lower latitudes are rare (Tomilin 1957). In
mid-July 1928, Cruikshank reported seeing
". . . several blue whales . . ." at lat. 11°32'N and
long. 91°58'W (Kellogg 1929). A Peruvian fishery
reported taking 247 blue whales between De-
cember 1925 and March 1926 (Ingebrigtsen 1929).
Potentially these were from a North Pacific stock,
since the Southern Hemisphere blue whale is most
numerous in the Antarctic at this time. Volkov
and Moroz (1977) noted an abundance of baleen
fishery BULLETIN: VOL. 76, NO. 4, 1979.
915
whales between lat. 7° and 10°N. Although indi-
vidual species of baleen whales were not enumer-
ated by Volkov and Moroz, two sightings of blue
whales were made on 29 March 1975 and are pre-
sented here (Table 1, Vni/shitelnyi cruise).
Typically blue whales are seen along the Baja
California coast in October while migrating
southward, and subsequently reappear off Baja
California in large numbers in March-June on
their northward migration (Rice 1974). The
whereabouts of the North Pacific blue whales dur-
ing the winter months is completely unknown, but
this is probably due to the lack of sighting effort.
For instance, Japanese whale scouting has been
carried out systematically since 1965, but their
effort has been restricted to the Pacific waters
north of lat. 20°N (Wada 1977).
Two theories have evolved regarding the win-
tering grounds of the blue whale. Wheeler ( 1946),
suggested that blue whales winter within a lim-
ited area of the subtropics. He maintained that
whales congregate in large groups in areas not
frequented by vessels. A second theory maintains
that wintering blue whales disperse between the
feeding grounds and the tropics (Harmer 1931;
Mackintosh 1942). Presented in this note is a 3-yr
record of blue whales sighted in the northeastern
tropical Pacific. Reference is made to migration,
whale groupings, behavior, and to the oceano-
graphic features of the sighting area. These recent
sightings were made by trained observers aboard
vessels involved with the National Marine
Fisheries Service Tuna/Porpoise Research Pro-
gram. Sighting information from other experi-
enced observers has also been contributed.
Shipboard identification of rorquals is difficult
and this problem was compounded by the fact that
most of the sightings mentioned in this paper were
incidental to ship's activities. However, the blue
whale is easily discerned from other rorquals by
the recognition of the following combination of
characteristics:
1. Mottled blue-grey coloration. All other ror-
quals are uniformly steel grey on the dorsal
surface.
2. A small dorsal fin of varying shapes located in
the posterior third of the body. The dorsal fin of
the sei, fin, and brydes whales is larger, falcate
shape, and placed farther anterior than the
blue whale dorsal fin.
3. A U-shaped rostrum. The rostrum shape of
other balenopterids is more pointed.
4. Tall, dense, disperse blows. Generally, the blow
of the sei and brydes whales is low and dissi-
pated, while the fin whale has a tall conical-
shaped blow.
A total of 1 1 cruises are discussed in this report,
covering the period from January to May for 1971,
1 975, 1 976, and 1 977. 1 The area of effort and sight-
ings of blue whales are reported in Figure 1. The
'No blue whale sightings were made in 1977. although two
cruises have been included in Figure 1 to complement survey
effort.
Table l. — Annotated listofblue whale sightings by National Marine Fisheries Service observers in the northeastern tropical Pacific,
January-May 1971-76.
Date
Lat., Long.
No. of whales
Observations
Cruise/Observer
8 Jan, 1971
07 54N, 095 52W
1
Waufz/us Leatherwood
23 Jan, 1975
01 30N.083 04 W
1
Dove for 7 mm
Pan Pacific Friedrichsen
3 Feb 1975
07=29 N. 093 48 W
cow and calf
3 dives, 5 57 min/dive: lengtfi cow 27 m, calf
8-10 m
Aquanus.Wade
4 Feb. 1975
07 48'N, 097 40W
1
Surfaced in middle of tuna scfiool
F/nesferre Walker
7 Feb. 1975
07 45'N, 098 2rW
1
5 dives, 10.1 1 mm dive
AquariusiWade
7 Feb. 1975
07 47'N, 098 24'W
1
2 dives, 8 19 mm dive
Aquarius Wade
7 Feb, 1975
07 52'N. 098 47 W
2
Small unidentified wfiale, 10 m, witfi visible
blow, swimming with a large blue whale
AquariusiVJade
7 Feb 1975
07 47'N. 09900W
1
3 dives. 1 1 39 mm dive, exposed tail fluke
on all dives
Aquarius \Nade
9 Feb 1975
08 50'N, 096 04 W
4-6
Paired groups, length estimate: 27 m
Pan Pacrf/c/Friedrichsen
10 Feb. 1975
0833N, 096 47'W
1
Pan Pac///c Friedrichsen
15 Feb 1975
0836'N, 096 29W
1
Pan Pac/l/c/Friedrlchsen
17 Feb 1975
08°58'N, 096°54W
8-10
Mostly pairs dispersed over several square
miles; all headed northeast
Pan Pac/ftc/Frledrichsen
17 Feb 1975
08 53N, 096 36W
1
Pan Pac/ftc, Friedrichsen
29 Mar 1975
08 55N, 093 34'W
10-13
Vnushitelnyi Rice
29 Mar 1975
09 07N. 093 55' W
6-7
Vnushitelnyi Rice
13 Feb 1976
09 44'N, 092 25'W
2
Cromwell Friedrichsen et al.
13 Feb 1976
09'44'N. 092 31 'W
1
Length estimate: 23 m
Cromwell Friedrichsen et al.
13 Feb. 1976
09 44'N, 092"35'W
4-5
Whales dispersed over 4-5 mi^
Cromwell Friedrichsen et al.
13 Feb. 1976
09 44'N. 092 52 'W
1
Exposed tail flukes prior to sounding
Cromwell Friedrichsen et al.
28 May 1 976
10 31'N,092 45'W
1
Length estimate: 20 m
Martinac Friedrichsen
916
(PO'w no**.'
I- -T 1 1 1 — h 1 1 n~l 1 1 1 1 1 1 — I 1 1 1
— t
_,OOV
-1 —
— 1 —
— ,
dO-w
— 1 — 1
1 — 1
I
80-~
I — 1
— 1
i~ 3 3 1 S^^V
i
H—
-T- 2 1Z^T\
.•si
5^
-r- 1 1 M Ik \ \
1
1 1 1 1 1 216 Yk\ '^
\
z 1 II iM z 1) H; >x
/
—
. _i_.
-
-
-■
-
--
-■
-
--
__!_-
-
--
---
---
-
-
--
--
_^ 2 1 II Tl 3 4VJ^ \
MEXICO
i 11 II 1 1 1 4 4 ixSi \
I
'2 II 1 1 1 2 Z 3Z 2 n
\
X
I2 1 1 1 1 1 12 2 2 3 k
)
V
-f^
1 — 1
1 zll 1 1 1 1_ 1 112 13 1
1
N
\
1
L
1 ' 12 < 2 .3 1 1 1 2
\
V.
t
"11' 1 1 1- 2 Z 2 2 2 1
3
\]
\
==%iH-v
4
11 1 IIZ ; Z1M2
1
3
2
1
V i
2 1 '' 1 Z 1 \ ' 1 1 1 3 1
1
1
2
I
2
z
21
_^ ,
/
1 1 1 111 I'lll Z123I
1
1
1
1
"r
I
I
\
--
--
---
-
■-
-■
• --
T""^"! "1 1 1 2 122224
1
1
Z
2
1
3
1
1
zTz
1
\
K i
1 2'l 1 ■ 13 2 13 1
4
4
1
1
2
\ 1
1
1
1
3
2:
^
p-^
^ £\ NU'ARACtUA
111^ 11 ' Z 1 Z
Z
1
1
Z
3
2
1
1
l|2
3
3
3
\
321 M2l 211 ' IZZ\33
1
2
3 2|Z
z
ilz
M
4
1
1
1
\
2 3 Z 1 ' 1 1 1 1 i 1 1 3.a-P^..TDM
2
3
3
1
1
1
5 i
5 1
1
4|2
,^.B
2
•^
2
\
J
Ws
V
2I11M2 1 ll 226feZ32
3
2
3
3
2
2
Z 2
5 5
6
2{5
^
4
2
2
z
\
1
rt"*
SU
^
u!
\ '\ 1 J_;_L 11 M 7 M 2 1
1
3
1
2
2
2
3
6H
2M^3
1
1
2^
\
1
2|
v.
NN/ U
1 1 ll 11 ' 2 3 12 2
1
3
2
5
4
3
5 !
VlZRiHi
2
3
z
2
3|
M
~\
N ^
111:1 1 Z
2
1
2
1
2
2
1
1 il
2
1
1
4
^
M
2
z
1
1
2 1 1 : 1 111' 2
1
[
2
2
1
1 12
1
1
1
.^.^i4
zlz
1
-■
"' 1" ■ ^ 1 1 1 1 ^ "2
1
1
1
|2
1
1
1
4
1
zll
1
2 ll||2l2l 11.1111131
1
1
1 1 1
1
1
1
1
3
1
1
2
1
111 2' II 1 1 1 3 Zll 1 2111
1
1
1
1
1
1
1
2
1
1
2
1
z
1
3
z
/^
1222! |2|i\ 11
1
1
2
1
Z
M
z
Z
D
z
i2IZl'lllMl 1 ' II 2433
3
2
1
1
1
1 ' 1
1
1
1
2
2
Z
1 1 1
1
Z 1 1 1 M 1 1 1 1 1 1 2 1 1 1 1 1 1
2
1
1
2
1
1
1
I
OAl^p-VtO^ r^'^^o.
2
1
1
111
1
1
/
^ ;
'■ 1 1 1 11 2 1 1
1
1
1
1
z:
i
1 1
1
1
r
•
1
1
1 1
1
z
]
1 2 2| 1 1
1
1
1
1
1
1
c
I 1
Iz
2
I
1
1
I
2
2
1
z
<
1 ' 1 1 1 1 111
Z
'
1
1
c
J
1 1
1
1
1 1
i ' 1 1 1 2 1
1
1
1
1 1
1
1
1
(
1
1
l(
1 1
1
2
1
1
1
2
2
1
ll
1
1
1 >
1 1
1
Z
1
1
1
1
1 1
1
' i '
1
Z
1
1
1 ll
1
1
1
\
1 ' 1
1
1
1
2I
1
1
1
^
: ! Ill
1
I
1
1
z
1
1 ll
1
1
1 1
i
1
I 1
;l
1
1 ]l
\
11
1
ITT
1
1 1
+-3^
- O
-icr
Figure 1 . — Survey effort for blue whales in the northeastern tropical Pacific over 1 1 cruises from January to June in 197 1 , 1975, 1976,
and 1977. Each numeral represents the total number of times that vessels entered a 1° block. Darkened blocks indicate blue whale
sightings.
fact that blue whales have been sighted in the
same general area for three winters indicates that
North Pacific blue whales may have a distinct
wintering ground to which they migrate each
year.
The location of the suggested wintering grounds
indicated by the sighting data are lat. 7°29'-
10°31'N and long. 95°25'-99°00'W.2 Only 1 of the
20 blue whale sightings (lat TSO'N and long.
83°04'W, 23 January 1975, a solitary individual)
was outside of these bounds. The equatorial sight-
ing location of this whale may indicate that either
it was not from the North Pacific population, or
that North Pacific blue whales do not restrict mi-
gration to the hypothetical wintering grounds.
^Cruikshank's observations, in 1928, were also in this area.
Whale Groupings and Behavior
Blue whales are believed to be found singly or in
pairs (Leatherwood et al. 1976). In fact, Nemoto
( 1964) reported that blue whales observed on the
summer feeding grounds were solitary. However,
five of the sightings reported here were aggrega-
tions of whales dispersed over several square
miles. Many of the whales were paired. The multi-
ple sightings on 7 February 1975 and 13 February
1976 appear to be mostly of solitary animals (Ta-
ble 1). However, on both days, no less than 40 n. mi.
separated the first and last whales sighted. Also,
four out of the five whales observed on 7 February
1975 were headed in a northeasterly direction.
This information may indicate that these appar-
ently solitary whales were part of a large dis-
persed group.
917
At least one cow and calf were observed and
possibly a second pair (Table 1). This is the first
actual record of a blue whale calf in the tropics,
although historically it has been believed that
blue whales have their calves in the warm tropical
waters (Mackintosh 1966:126).3
Steve Leatherwood, W. A. Walker, and D. W. Rice,
who contributed blue whale sighting data and
Karen J. Rice for technical assistance. Last, we
thank Robert Schoning of the National Marine
Fisheries Service for releasing this data for publi-
cation.
Oceanoyraphic Features
Literature Cited
There are several unique oceanographic fea-
tures which relate to the sighting location of the
blue whales. Cromwell (1958) and Wyrtki (1964)
discussed the Costa Rican Dome which is located
at approximately lat. 9°N, long. 89°W. The dome is
apparently a permanent topographic feature (150
km X 300 km) and is formed by the convergence of
several major current systems. These currents
typically create an area of nutrient transport or
upwelling. High standing stocks of zooplankton in
the area near the Costa Rican Dome (lat. 7°25'N-
10°N) has been reported by several authors (Reid
1962; Blackburn et al. 1970; Holmes"*). Volkov and
Moroz (1977) suggested that the high stable food
base of the area creates a habitat suitable for
nonmigratory populations of baleen whales.
North Pacific blue whales may also use this area
for their winter feeding grounds.
In conclusion, the recent sightings of blue
whales in the tropics indicates that North Pacific
blue whales have a wintering area to which they
return each year. Since most of the cruises have
occurred largely during the winter months, more
information must be collected to determine if
whales are found in this area the year round. The
high standing stock of zooplankton in this area
may indicate that this is a winter feeding area, as
well as a calving ground.
AckiiDw ledgments
We are particularly indebted to E. D. Mitchell,
Fisheries Research Board of Canada, for his criti-
cal evaluation of the manuscript. D. W. Rice, Na-
tional Marine Fisheries Service; K. S. Norris,
University of California at Santa Cruz, and Ron
Garrett, Wilderness Research Institute, also made
helpful suggestions on the draft. We also thank
^The average water temperature for 1 1 sightings was 26.5°C.
■'Holmes, R. W. 1970. A contribution to the physical, chem-
ical, and biological oceanography of the northeastern tropical
Pacific. (Unpubl. manuscr.) Institute of Marine Resources.
Scripps Institute of Oceanography, Univ. Calif La Jolla, Calif
AEC-UCSD-.34P99-4.
BLACKBURN, M., R. M. LAURS, R. W. OWEN, AND B.
ZEITZSCHEL
1970. Seasonal and areal changes in standing stocks of
phytoplankton, zooplankton, and micronecton in the east-
ern Tropical Pacific. Mar. Biol. (Berl.) 7:14-31.
Cro.mwell. T.
1958. Thermocline topography, horizontal currents and
"ridging" in the Eastern Tropical Pacific. [In Engl, and
Span.] Inter-Am. Trop. Tuna Comm. Bull. 3:133-164.
HAR.MER. S. F.
1931. Southern whaling. Proc. Linn. Soc. Lond. 142:85-
163.
INGEBRIGTSEN, A.
1929. Whales caught in the North Atlantic and other
seas. Rapp. P.-V. Reun. Cons. Perm. Int. Explor. Mer
56:1-26.
Kellogg, R.
1929. What is known of the migrations of some of the
whalebone whales. Smithson. Inst., Annu. Rep. 1928, p.
467-494.
LEATHERWOOD, S., D. K. CALDWELL, AND H. E. WiNN
1976. Whales, dolphins, and porpoises of the western
North Atlantic. A guide to their identification. U.S. Dep.
Commer., NOAA Tech. Rep. NMFS CIRC-396, 176 p.
Mackintosh, N. A.
1942. The southern stocks of whalebone whales. Discov-
ery Rep. 22:197-300.
1966. Distributionof southern blue and fin whales. InK.
S. Norris (editor), Whales, dolphins, and porpoises, p.
125-144. Univ. Calif Press, Berkeley.
Nemoto, T.
1964. School of baleen whales in the feeding areas. Sci.
Rep. Whales Res. Inst. Tokyo 18:89-110.
NlSHlWAKL M.
1966. Distribution and migration of the larger cetaceans
in the North Pacific as shown by Japanese whaling re-
sults. In K. S. Norris (editor). Whales, dolphins, and por-
poises, p. 170-191. Univ. Calif. Press, Berkeley.
Reid, J. L.
1962. On circulation, phosphate-phosphorus content, and
zooplankton volumes in the upper part of the Pacific
Ocean. Limnol. Oceanogr. 7:287-306.
RICE, D. W.
1974. Whales and whale research in the eastern North
Pacific. In W. E. Schevill (editor). The whale problem: a
status report, p. 170-195. Harvard Univ. Press, Cambr.,
Mass.
TOMILIN, A. G.
1957, Zveri SSSR i prilezhashchikh stran Vol. IX Kitoob-
raznye (Mammals of the U.S,S,R, and adjacent countries
Vol. IX Cetacea). Akad, Nauk SSSR, Mosk,, 756 p, (Trans-
lated by Isr. Program Sci. Transl., Jerusalem, 1967, 717
p.)
918
VOLKOV, A. F.. AND I. F. MOROZ.
1977. Oceanological conditions of the distribution of
cetacea in the Eastern Tropical part of the Pacific
Ocean. Int. Whaling Comm. Rep. 27:186-188.
W.-\n,A. S.
1977. Indices of abundance of large-sized whales in North
Pacific in the 1975 whaling season. Int. Whaling Comm.
Rep. 27:189-192.
WHEELER. J. F. G.
1946. Observations on whales in the South Atlantic Ocean
in 1943. Proc. Zool. Soc. Lond. 116:221-224.
WYRTKI. K.
1964. Upwelling in the Costa Rica Dome. U.S. Fish
Wildl. Serv.. Fish. Bull. 63:355-372.
Lawrence S. Wade
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service. NOAA
La Jolla. Calif.
Present address: P.O. Bo.x 4455
Areata. CA 95521
Gary L. Friedrichsen
Southwest Fisheries Center La Jolla Laboratory
National Marine Fisheries Service, NOAA
La Jolla. Calif.
Present address: P.O. Box 890
Areata. CA 95521
A substantial sport fishery exists for white mar-
lin in the Atlantic off North and South America.
In the United States, the major sport fisheries
occur along the Middle Atlantic States, from New
Jersey to North Carolina, off southeast Florida,
and along the Gulf Coast States. Important sport
fisheries also occur in the Bahamas, off Havana,
Cuba, and along the coast of Venezuela ( Mather et
al. 1972). Another important sport fishery re-
cently developed off eastern Brazil (Anonymous
1976).
The white marl in is also an incidental catch of
commercial longline vessels fishing for tuna in the
Atlantic and Gulf of Mexico (Mather et al. 1975).
The marlin is highly prized as a food item in some
countries (Kume and Joseph 1969).
My review of the literature on white marlin
shows that there is a need for additional informa-
tion on sex composition and length-weight rela-
tionships. Until recently, no information was
available regarding its reproductive potential
(Baglin^). In this paper I update reproductive and
sex ratio data presented by Bagliij (see footnote 2)
and include length-weight relationships.
Materials and Methods
SEX COMPOSITION, LENGTH-WEIGHT
RELATIONSHIP, AND REPRODUCTION OF THE
WHITE MARLIN, TETRAPTURUS ALBIDUS, IN
THE WESTERN NORTH ATLANTIC OCEAN'
In the Atlantic, white marlin, Tetrapturus al-
bidus. range from lat. 35°S to 45°N with concen-
trations in the western Atlantic, including the
Gulf of Mexico, and the Caribbean Sea ( Mather et
al. 1975). Tag returns show that some white mar-
lin migrate seasonally from the U.S. Middle At-
lantic Bight (the coastal area between Cape Cod
and Cape Hatteras) in the summer to the south-
eastern Caribbean Sea in the winter ( Mather et al.
1972). Commercial catches by Japanese longline
vessels support the tagging results, but the
catches also indicate that a second group of white
marlin moves from a wintering area in the south-
eastern Caribbean to summer grounds in the Gulf
of Mexico (Ueyanagi et al. 1970; Mather et al.
1972; Wise and Davis 1973).
White marlin from the northern Gulf of Mexico
(hereafter referred to as the gulf), the Florida
Straits, the western Bahamas, and the Middle At-
lantic Bight of the western North Atlantic (hereaf-
ter referred to as the Atlantic) were sampled from
anglers' catches at sport fishing tournaments and
at Pflueger Marine Taxidermy, Inc., Hallandale,
Fla. One marlin was collected by longline in the
Windward Passage between Cuba and Hispaniola
during RV Oregon Cruise 66.
Sex data were obtained from 1,128 white marlin
captured by anglers in the gulf ( 1971-77) and from
720 white marlin caught by anglers from the At-
lantic (1972-77).
Lengths and weights were obtained from 904
white marlin captured in the gulf (1971-76) and
from 489 white marlin captured in the Atlantic
( 1972-76). Body lengths ( straight distance from tip
of lower jaw to tips of midcaudal rays) were mea-
sured in centimeters (Rivas 1956); weights were
recorded to the nearest pound and converted to
kilograms.
'Contribution No. 78-44M, Southeast Fisheries Center Miami
Laboratory-, National Marine Fisheries Service, NOAA, Miami,
Fla.
^Baglin, R. E., Jr. 1977. Maturity, fecundity and sex composi-
tion of white marlin I Tetrapturus albidusl . Collective Volume of
Scientific Papers 6(79):408-416. International Commission for
the Conservation of Atlantic Tunas, Madrid, Spain.
FISHERY BULLETIN: VOL 76. NO. 4, 1979.
919
Ovaries from 186 females caught from 1972
through 1976 were examined. Fresh ovaries were
either blotted dry and weighed in grams or stored
in 107f Formalin'^ and weighed later. No sig-
nificant difference was found between the mean
weight of fresh and preserved ovaries (F = 0.0001;
df = 1, 16;P>0.75). The gonosomatic index (GSI),
ovary weight as a percentage of total body weight,
was used as an indicator of maturity.
Preserved eggs 0.56 mm in diameter and larger
were counted when estimating fecundity. Eggs
<0.56 mm were less spherical in shape and in an
earlier stage of development. The 0.56-mm size
was determined by measuring the diameters of
3,912 eggs from mature, partly spent, and spent
fish. Small transparent ova were stained with
aceto-carmine to facilitate measuring. Egg
diameters were measured with an ocular microm-
eter at 30 X magnification and the orientation of
egg diameters was assumed to be random. Thin
cross sections were taken from the anterior, mid-
dle, and posterior parts of one ovary of a mature
fish and subdivided into three subsamples, repre-
senting the center, midregion, and periphery of
the ovary (Otsu and Uchida 1959).
Fecundity was defined as the potential number
of mature eggs (yolked ova in the most advanced
size mode) that could be spawned during one re-
productive season and was estimated using a dry
weight method. For six fish in which entire ovaries
were saved for fecundity analysis, subsamples
consisted of a thin cross section taken from the
anterior, middle, and posterior parts of each ovary.
The eggs in these subsamples were separated from
the ovarian tissue, enumerated, dried, and
weighed according to the procedure described by
Baglin (see footnote 2). For six other fish, only the
ovary weight and a single cross section from the
middle of the ovary were taken; these cross sec-
tions comprised the subsamples. The eggs were
separated, counted, dried, and weighed. A dry/wet
weight regression was used to estimate the total
dry weight of the eggs in these ovaries, which were
not saved. Before the eggs in the subsample were
counted, 25 eggs 0.30 mm and larger from the two
most advanced modes were randomly selected and
measured. Eggs in this second most advanced
mode were included to give an indication of the
percentage of eggs in both modes because future
histological studies may indicate that these
smaller eggs undergo further development and
are also spawned. Fecundity estimates, rounded to
the nearest 0.1 million eggs, were calculated from
the relationship: C = iAD/B) + A, where A is the
number of mature ova in the subsample, B is the
weight of the ova in the subsample, C is the
number of mature ova, and D is the weight of ova
from both ovaries.
Results and Discussion
Sex Q)mposition
From 1971 through 1977, sex was determined
from 1,128 white marlin from the gulf (Table 1).
The deviation from an expected 1:1 sex ratio was
significant from May through October. Sampling
was inadequate for the remaining months.
Females were more prevalent than males for each
month studied.
From 1972 through 1977, sex was determined
for 720 white marlin from the Atlantic (Table 1).
There were 323 sex determinations from the
Florida Straits (March through May) and 397 from
the Middle Atlantic Bight (June through Sep-
tember). Sampling was inadequate from October
through February. No significant difference from
an expected 1:1 sex ratio was found for March,
May, July, August, and September, but a sig-
nificant difference was found for April and June.
For the months in which the sex ratio was sig-
nificantly different from the expected 1:1 ratio,
females were more prevalent.
deSylva and Davis (1963) found a significant
difference from an expected 1:1 sex ratio (60%
females) when they combined their data from the
Middle Atlantic Bight for the summers of 1959
and 1960. They presented monthly sex composi-
TABLE 1. — Monthly sex ratios for white marlin from the north-
ern Gulf of Mexico! 197 1-77), Florida Straits and Middle Atlantic
Bight (1972-77).
^Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
Number of
Sex ratio
Location
f^ontfi
white marlin
(females/males)
Gulf of (Vlexico
May
21
4.25'
June
85
4.00*
July
374
3.16*
August
444
1.63*
September
150
1.50*
October
54
1.84*
Florida Straits
March
103
0.87
April
172
1.96*
May
48
1.40
Ivliddle Atlantic
June
55
3.23*
Bigtit
July
56
1.67
August
219
0.80
September
67
0.97
'Significant departure from null fiypothesis at 0.05 level (chi-square)
920
tion data for 1960 only. My analysis of their data
shows a significant difference for June and July
but no significant difference for August and Sep-
tember. Their findings for June, August, and Sep-
tember agree with those in this study. The ex-
treme difference in se.x ratios found in the present
study for the gulf (May through October) and for
the Florida Straits in April has not been reported
previously. The above findings suggest that some
white marlin segregate into distinct areal groups
according to the predominating sex and that sex
ratios may change with season. A similar occur-
rence has been noted for the blue marlin, Makuira
nigricans (Kume and Joseph 1969).
Length- Weight Relationship
The average length of females is greater than
that of males from both the gulf and the Atlantic
(Figure 1). It is also apparent (Figures 2, 3) that
the average length of females is greater than that
of males from each area for each month studied.
This difference may be due to faster growth of
females or higher mortality of males and should be
considered in future growth studies of the white
marlin.
The length- weight relationship by sex was de-
termined for white marlin taken in the gulf from
1971 through 1976 (Figure 4) and in the Atlantic
from 1972 through 1976 (Figure 5). Analysis of
covariance (Table 2) indicated that length-weight
regression coefficients were significantly different
between gulf females and males (F = 16.0; df = 1,
900;P<0.001), gulf males and Atlantic males (F =
19.2; df= 1, 514; P<0.001), and gulf females and
Atlantic females (F = 10.8; df = 1, 871;P<0.001).
The adjusted means were also significantly differ-
ent between Atlantic females and males (F = 13.4;
df = 1, 486; P<0.001). These findings agree with
those of Lenarz and Nakamura (1974), who found
a significant difference between sexes in the rela-
tionship between weight and eye-fork length for
white marlin from the gulf during 1971.
Analysis of covariance was conducted for the
length-weight relationship, on a monthly basis,
for which sufficient samples were available: gulf
females versus Atlantic females in May, June,
August, and September, and gulf males versus
Atlantic males for June, August, and September.
A significant difference in the regression
coefficients was found only for the August males (F
= 13.7; df= 1,211;P<0.001). A significant differ-
ence in adjusted means was found for females dur-
215-
210-
205-
200-
E 195-
u
190-
o
185-
180-
5 175-
o
it.
I 170-
< 165-
I 155J
150-
145-
140-
135-
■r
?
N=277
31%
N=248
51%
0*'
N = 627
69%
N=241
49%
GULF
ATLANTIC
Figure 1. — Comparison of length between female and male
white marlin collected in the northern Gulf of Mexico (1971
through 1976) and in the Atlantic (1972 through 1976). The
number, percent, mean (horizontal line), range (vertical line), 1
SD on each side of the mean ( open box), and 2 SE on each side of
the mean (shaded box) are shown.
ing June (F = 9.3; df = 1, 74;P<0.005) and August
(F = 12.0 df = 1, 286;P<0.001), and for males
during September (P = 7.6; df = 1, 73; P<0.01).
Differences between length-weight relation-
ships of white marlin from the gulf and the Atlan-
tic suggest the possibility of separate groups in-
habiting the two areas. Tag returns, however,
showed there is at least some migratory move-
ment from the Middle Atlantic Bight to the gulf.
To date, tag return data have not shown white
marlin migrations in the reverse direction, al-
though one fish tagged in the gulf was recaptured
off Cuba, giving some support to the likelihood
that they do migrate in the opposite direction
(Chester C. Buchanan, Southeast Fisheries
Center, National Marine Fisheries Service,
NOAA, 75 Virginia Beach Drive, Miami, FL
33149, pers. commun.)
921
215-
210-
205-
200-
J 190-
o
Z 185-
lii
_l
180-
ae
u
175-
I 170-
< 165-
160-
O 155-
150-
145-
140-
135
N = 17
N=75
?
N = 123
36%
?
N = 231
75%
N^222
64%
N = 74
61%
II
N=30
62%
N=18
38%
N = 47
39%
MAY
JUNE
JULY
AUGUST
SEPTEMBER
OCTOBER
Figure 2. — Monthly comparisons of length between female and male white marlin collected in the northern Gulf of Mexico during
1971 through 1976. The number, percent, mean (horizontal line), range (vertical line), 1 SD on each side of the mean (open box), and 2
SE on each side of the mean (shaded box) are shown.
205
200
195-
190-
185-
Ul
K
u.
' 160-
5 1551
Uj I50H
O 145-
140-
135-
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
Figure 3. — Monthly comparisons of length between female and male white marlin collected in the Atlantic during 1972 through
1976. The number, percent, mean (horizontal line), range (vertical line), 1 SD on each side of the mean (open box), and 2 SE on each
side of the mean (shaded box) are shown.
922
H I 4S
X
u
J|40-
o
GULF OF MEXICO
L0G,oW=-4 49978 •2 65546 LOG,oL
N:627
/— L0CnoW=-3 10422* 2 01104 LOC.qL
— r-
2 15
2 20 2 25 2 30
LOG,o LOWER JAW- FORK LENGTH (cm)
Figure 4. — Length-weight relationship (log transformation) for
female and male white marlin from the northern Gulf of Mexico.
w I 40'
5
0
u
O I 35-
1 30'
ATLANTIC
LOC,nW= -5 52016 'S 13550 LOCoL
N=248
LOG,„W=-4 96728 •2 87607 LOG,^ L
—I 1 1 1 1
2 15 2 20 2 25 2 30 2 35
LOG,o LOWER JAW- FORK LENGTH (Cm)
Figure 5. — Length-weight relationship (log transformation) for
female and male white marlin from the Atlantic.
Reproduction
A significant difference in egg diameter was
found among the anterior, middle, and posterior
sections of an ovary from a mature fish {F — 1.1; df
= 2, 2,676; P<0.001). There was no significant
Table 2. — Regression equations, number, sum of squares of jc,
and mean square calculated for the length-weight relationship
(logio transformations) of white marlin from the northern Gulf of
Mexico and the Atlantic.
S.v^-blx-j
y + b(X - X)
N
Xx'
N-2
Gulf and Atlantic females:
1.41704 + 288186(X - 2.22302)
Gulf and Atlantic males:
1 .34355 + 2.37655(X - 2.20228)
Gulf females:
1.39996 + 2.65546(X - 2.22174)
Gulf males:
1.32735 + 2 01104(X - 2.20363)
Atlantic females:
1.46021 + 3.13550(X - 2.22624)
Atlantic males:
1.36217 + 2.87607(X - 2.20073)
May gulf females:
1 48714 + 2.60014(X - 2.24572)
May Atlantic females:
1.44276 + 2.96929(X - 2.21669)
June gulf females:
1,44221 + 3.22066(X - 2.23317)
June Atlantic females:
1.40502 + 2.60180(X - 2.23216)
June gulf males:
1 35686 + 3.55091(X - 2.21054)
June Atlantic males:
1.24868 + 3.01101(X ~ 2.17815)
August gulf females:
1.38976 + 2.72170(X - 2.21831)
August Atlantic females:
1.41118 + 2.81615(X - 2.21783)
August gulf males:
1.32575 * 1.93866(X - 2.20207)
August Atlantic males:
1.35384 + 2.91546(X - 2.20279)
September gulf females:
1.40689 + 3.01922(X - 2.22399)
September Atlantic females:
1 42732 ^ 3,10221(X - 2.22137)
September gulf males:
1.32709 + 1.39787(X - 2.20530)
September Atlantic males:
1.34390 + 2.01746(X - 2.19767)
875
0.604832
0.00336706
518
0.254956
0.00299720
627
0.396458
0.00252756
277
0.128593
0.00250737
248
0.204773
0.00377638
241
0.125280
0.00244554
17
00164014
0.00365206
16
0.0085322
0.00226977
51
0.0210122
0.00162473
26
0.0163375
0.00323226
11
0.00624798
0.00085387
10
0.00707686
0.00165193
222
0 121132
0.00226799
67
0.0391569
0.00206003
123
0.524543
0.00179221
92
0.0484607
0.00169370
74
0.0440863
0.00317564
17
0.0150308
0.00160237
47
0.0199974
0.00220416
29
0.0105484
0.00144724
difference in mean diameter among the center,
midregion, and periphery within each of the three
sections. Because some heterogeneity occurred,
estimates of fecundity were based, when possible,
on eggs from each section of both ovaries.
Heterogeneity of egg size within an ovary has also
been shown for albacore, Thunnus alalunga (Otsu
and Uchida 1959), and swovdfish., Xiphias gladius
(Uchiyama and Shomura 1974).
The left ovary (X = 25.0 cm, S^ = 0.732) was
significantly longer (F = 35^7; df = 1, 196;
P<0.001) than the right ovary (X = 19.4cm,S.v =
0.561). Eldridge and Wares ( 1974) reported differ-
ential growth in the size of ovaries for striped
marlin, Tetrapturus audax, and for sailfish, Is-
tiophorus platypterus. Both were similar to the
white marlin in having larger left ovaries.
Well-developed ovaries were present only in 12
white marlin collected during April and May in
the Florida Straits. These fish had a GSI of about
923
&'Jc or greater and were used for estimating fecun-
dity. The mean GSI showed that ovarian weights
were lowest during October and increased from
November through May (Figure 6). The mean GSI
of 2.6 for April and May is lower than the 4.5 mean
GSI found by Krumholz (1958) for late April. The
GSI of 9.3 (Table 3) agrees with the highest GSI of
9.76 found by Krumholz. The high mean GSI val-
ues determined by me for April and May, with the
sudden decrease in June, indicated that spawning
probably occurred during April and May (Figure
6). Therefore, only one spawning season per year
was indicated for the Florida Straits.
White marlin may also spawn in other areas.
One fish captured in April 1976 in the Windward
Passage had ripe eggs measuring 1.16 mm. Hayasi
et al. (1970) found white marlin with mature
gonads during April-June in the northern Carib-
bean. Erdman (1956) found well-developed
Figure 6. — Seasonal variation of mean gonosomatic index in
186 white marlin collected from 1972 to 1976 (number offish
indicated above histograms).
Table 3.— Weight, length, and gonadal data for 12 female white
marlin from the Florida Straits collected during 1972, 1974, and
1975. The mean and standard error of the mean are given at the
bottom of the columns.
Estimated number of eggs
Body
weigtit
(kg)
Body
lengthi
(cm)
Ovary
wet weight
(g)
Gono-
somatic
index
0,55 mm in
diameter
(millions)
■0,29 mm in
diameter'
(millions)
26.8
160
21.600
6.0
5.4
26.8
169
22,050
7,6
4.8
30.4
168
22,324
7,6
70
30.4
176
21,700
5.6
38
31.3
168
2,908
93
10.4
32.7
166
2,150
66
7 1
32.7
166
2,693
8,2
10 1
33.6
167
2,161
6,4
7.6
35.0
169
22,250
6,4
6.5
35.4
170
22,320
6,6
7.5
36.3
171
2,488
68
105
37.2
179
22,050
5.5
8 1
324
169
2,224
69
7.4
0.98
14
107
032
0.62
10.4
8.0
117
7.3
18.6
11.8
16.8
11.9
10.2
14.4
20.2
14.5
13.0
1 16
'Estimated using actual percent from 0 30 to 0 55 mm in diameter from 25
eggs measured for each fish.
2Entire ovaries available.
ovaries in white marlin caught off Puerto Rico in
April and found well-formed eggs in a fish taken in
June from the same locality.
The smallest fish approaching a ripe condition
with large ovaries weighed 26.8 kg (Table 3).
Ueyanagi et al. ( 1970) reported that white marlin
reach sexual maturity at 130 cm eye-fork length.
Using the conversion equation of Lenarz and
Nakamura (1974), 130 cm eye-fork length would
be equal to about 20.3 kg.
Frequency distributions of white marlin ovum
diameters were made from measurements on
3,912 ova from spent, partly spent, and mature
fish (Figure 7). Spent fish caught during May and
June contained mostly eggs 0.15 mm in diameter
and smaller. Eggs from a partly spent fish caught
during June had a frequency mode of about 0.35
mm, with few eggs larger than 0.60 mm. Some of
the larger eggs appeared to be undergoing absorp-
tion. Jolley ( 1977), in his histological examination
of spent sailfish, found degeneration and absorp-
tion of advanced unovulated eggs common. Mer-
rett (1970), studying several species of billfish
from the Indian Ocean, suggested that there also
may be at least a partial resorption of resting
5
o
S300
Z 250-
150-
100-
50
1
0.05
0 15 0.25 0.35
045 0 55 0 65 0 75 0 85 0 95 105
DIAMETER (mm)
1
1.15
Figure 7. — Frequency distribution of white marlin ovum
diameters for: A, four spent fish (827 ova) in May and June; B,
one partially spent fish (406 ova) in June; C, one mature fish (2,
679 ova) in April.
924
oocytes. I found frequency modes of about 0.35 mm
and 0.65 mm in a mature fish caught in April.
Only eggs measuring 0.56 mm and larger were
included when estimating fecundity. Because
there were two frequency modes present in mature
fish, an estimate of the number of eggs 0.30 mm in
diameter and larger is also presented (Table 3).
Fecundity, based on the number of ova in the
most advanced size mode, ranged from 3.8 to 10.5
million eggs iX = 7.4, Sy = 0.62) for white marlin
weighing 26.8 to 37.2 kg (Table 3). The number of
mature ova per gram of body weight ranged from
125 to 332 (X = 227, %- = 16.76). The average
number of eggs measuring 0.30 mm in diameter
and larger was estimated as 13 million (S^ =
1.16).
Fecundity was based on the number of fully
yolked eggs, forming a group distinct from another
group of developing eggs. Fecundity would vary
depending on whether smaller eggs develop
further or are absorbed. If fractional spawning
occurs, as reported for sailfish by Jolley ( 1977), the
eggs in the next distinct group should be included
in seasonal fecundity estimates.
Acknowledgments
I thank the boat captains and sport fishermen
who provided samples, and Pflueger Marine
Taxidermy, Inc., Hallandale, Fla., for allowing
National Marine Fisheries Service (NMFS) per-
sonnel to sample available specimens. I thank
Grant Beardsley, Chester Buchanan, Eugene
Nakamura, William Richards, Luis Rivas, and
James Tyler of the Southeast Fisheries Center
(NMFS), and John Jolley of the Florida Depart-
ment of Natural Resources for their helpful com-
ments on the manuscript.
Literature Cited
Anonymous.
1976. Brazil: A sleeping giant awakens. Int. Mar. Angler
38(6):6-7.
DE Sylv.a, d. p., and W. p. Davis.
1963. White marlin, Tetrapturus albidus, in the Middle
Atlantic bight, with observations on the hydrography of
the fishing grounds. Copeia 1963:81-99.
Eldridge, M. B., and p. G. Wares.
1974. Some biological observations of billfishes taken in
the eastern Pacific Ocean, 1967-1970. In R. S. Shomura
and F. Williams (editors). Proceedings of the Interna-
tional Billfish Symposium, Kailua-Kona, Hawaii, 9-12
August 1972. Part 2. Review and contributed papers, p.
89-101. U.S. Dep. Commer., NOAA Tech. Rep. NMFS
SSRF-675.
ERDMAN, D. S.
1956. Recent fish records from Puerto Rico. Bull. Mar.
Sci. Gulf Caribb. 6:315-340.
Hayasi, S., T. Kato, C. Shinou, S. Kume, and Y. MORITA.
1970. Status of the tuna fisheries resources in the Atlantic
Ocean, 1956-1967. [In Engl, and Jpn.] In Resources
and fisheries of tunas and related fishes in the Atlantic
Ocean. Far Seas Fish. Res. Lab. (Shimizu), S. Ser. 3:1-
72.
Jolley, J. W., Jr.
1977. The biology and fishery of Atlantic Sailfish, Is-
tiophorus platypterus , from southeast Florida. Fla. Mar.
Res. Publ. 28, 31 p.
krumholz, l. a.
1958. Relative weights of some viscera in the Atlantic
marlins. Bull. Am. Mus. Nat. Hist. 114:402-405.
KUME, S., AND J. Joseph.
1969. Size composition and sexual maturity of billfish
caught by the Japanese longline fishery in the Pacific
Ocean east of 130°W. Bull. Far Seas Fish. Res. Lab.
(Shimizu) 2:115-162.
LENARZ, W. H., AND E. L. NAKAMURA.
1974. Analysis of length and weight data on three species
of billfish from the western Atlantic Ocean. In R. S.
Shomura and F. Williams (editors), Proceedings of the
International Billfish Symposium, Kailua-Kona, Hawaii,
9-12 August 1972. Part 2. Review and contributed papers,
p. 121-125. U.S. Dep. Commer., NOAA Tech. Rep. NMFS
SSRF-675.
M..\THER, F. J., Ill, H. L. Clark, and J. M. Mason, Jr.
1975. Synopsis of the biology of the white marlin, Tetrap-
turus albidus Poey ( 1861 ). InR. S. Shomura and F. Wil-
liams (editors). Proceedings of the International Billfish
Symposium, Kailua-Kona, Hawaii, 9-12 August 1972.
Part 3. Species synopses, p. 55-94. U.S. Dep. Commer.,
NOAA Tech Rep. NMFS SSRF-675.
Mather, F. J., Ill, A. C. Jones, and G. L. Beardsley, Jr.
1972. Migration and distribution of white marlin and blue
marlin in the Atlantic Ocean. Fish. Bull., U.S. 70:283-
298.
MERRETT, N. R.
1970. Gonad development in billfish (Istiophoridae) from
the Indian Ocean. J. Zool. (Lond.) 160:355-370.
Otsu, T.. and R. N. UCHIDA.
1959. Sexual maturity and spawning of albacore in the
Pacific Ocean. U.S. Fish Wildl. Serv., Fish. Bull.
59:287-305.
RiVAS, L. R.
1956. Definitions and methods of measuring and counting
in the billfishes (Istiophoridae, Xiphiidae). Bull. Mar.
Sci. Gulf Caribb. 6:18-27.
Uchiyama, J. H., AND R. S. Shomura.
1974. Maturation and fecundity of swordfish, Xiphias
gladius, from Hawaiian waters. In R. S. Shomura and F.
Williams (editors), Proceedings of the International
Billfish Symposium, Kailua-Kona, Hawaii, 9-12 August
1972. Part 2. Review and contributed papers, p. 142-148.
U.S. Dep. Commer., NOAA Tech. Rep. NMFS SSRF-675.
Ueyanagi, S., S. Kikawa, M. Uto, and Y. Nishikawa.
1970. Distribution, spawning, and relative abundance of
billfishes in the Atlantic Ocean. [In Jpn., Engl, ab-
str.] Bull. Far Seas Fish. Res. Lab. (Shimizu) 3:15-55.
925
WISE, J. P., AND C. W. Davis.
1973. Seasonal distribution of tunas and billfishes in the
Atlantic. U.S. Dep. Commer., NOAA Tech. Rep. NMFS
SSRF-662, 24 p.
Raymond E. Baglin, Jr.
Southeast Fisheries Center Miami Laboratory
National Marine Fisheries Service, NOAA
75 Virginia Beach Drive
Miami. FL 33149
RECORDS OF PISCIVORUS LEECHES
(HIRUDINEA) FROM THE CENTRAL
COLUMBIA RIVER, WASHINGTON STATE
No records of leech infestations on fish of the Co-
lumbia River exist in the published literature. As
a whole, the freshwater hirudinean fauna of the
Pacific Northwest remains a relatively unsur-
veyed, little known, and neglected biotic group.
This is due, in part, to problems in leech identifica-
tion as well as in obtaining representative collec-
tions.
We obtained leeches from the external surface,
oral cavity, and gill chambers of fish during a
continuing environmental assessment program
on the central Columbia River above Richland,
Wash. (Benton and Franklin Counties), from 1975
through 1977. This paper identifies four piscivo-
rous species, provides new host and distribution
records, and reviews some recent taxonomic
changes for the species encountered. Ecological
observations are included.
The leeches recorded herein are Myzobdella
lugubris Leidy 1851, Piscicola salmositica Meyer
1946, Placobdella montifera Moore 1906, and Ac-
tinobdella inequiannulata Moore 1901.
Methods and Site Description
Fish were collected at monthly or bimonthly
intervals by a variety of gear (gill nets, trammel
nets, hoop nets, beach seines, and electroshocker)
from January 1975 to December 1977. Over
20,000 fish, representing nearly 40 species, were
examined during this period (Gray and Dauble
1977). Leech specimens were preserved in 10%
Formalin^ solution, either when captured or after
being examined alive in the laboratory.
Our leech collections were more qualitative
than quantitative because leech-fish associations
in nature are normally periodic and facultative
despite the nutritional requirement of piscivorous
leeches for fish blood. Also, piscivorous leeches can
readily detach from fish captured by most types of
fishing gear, particularly from fish recovered
when moribund or dead.
Occurrence of many freshwater leech species
can be correlated with characteristic aquatic
habitats. Water quality parameters vary season-
ally in the central Columbia River, as follows:
dissolved oxygen, 8.0-12.0 mg/1; pH, 7.4-8.6; phos-
phate (as PO4), 0.03-0.04 mg/1; ammonia-
nitrogen, 0.01-0.2 mg/1; hardness (Ca, Mg), 55-75
mg/1; and alkalinity (CaCOg), 50-67 mg/1. Water
temperatures range from 1° to 3°C in midwinter to
about 21°C in late August and early September.
There are no significant quantities of organic and
inorganic pollutants (our data). The water carries
minimal silt loads.
The central Columbia River in the Hanford
Reach where our collections were made (river km
550-629) survives as the last free-flowing section
of the main channel above Bonneville Dam. Dec-
ades of hydroelectric development have trans-
formed other sections into a consecutive series of
river-run reservoirs. River flows in the study area
usually range from about 2,000 m^/s over much of
the year to over 12,000 m^/s during the annual
spring spate, when surplus runoff is passed down-
river over spillways from reservoirs (Nees and
Corley^).
Additionally, Hanford flows are now regulated
at Priest Rapids Dam in response to daily and
weekly power demand peaks, causing water levels
in the river to fluctuate widely. This periodically
exposes and inundates a rocky or muddy shoreline
zone, apparently restricting development of a di-
verse leech fauna along the river margins. Water
levels in Wanapum Reservoir behind Priest
Rapids Dam (river km 639) and in Umatilla Res-
ervoir behind McNary Dam (river km 470) are
relatively stable, although subject to controlled
summer drawdowns. Substantial populations of
such common omnivorous leeches as Erpobdella
punctata (Leidy 1870), Helobdella stagnalis (Lin-
naeus 1758), and T/ieromjzon spp. occur along the
'Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
^Nees, W. L., and J. P. Corley. 1974. Environmental sur-
veillance at Hanford for CY-1973. Unpubl. manuscr., 56 p.
R&D Rep., BNWL-1881, Battelle, Pacific Northwest
Laboratories, Richland, WA 99352.
926
FISHERY BULLETIN: VOL, 76, NO 4, 1979.
margins of these mid-Columbia River reservoirs
(our observations).
Results and Discussion
Four leech species were recovered from Colum-
bia River fish (Table 1). About 907^ of the speci-
mens were Myzobdella lugubris. Two families
belonging to the order Rhynchobdella were rep-
resented, Glossiphoniidae and Piscicolidae. Mem-
bers of this order typically possess a small pore on
the anterior sucker for a mouth, from which a
muscular pharyngeal proboscis can be protruded,
and lack biting jaws or denticles. Relatively few
glossiphoniids are piscivorous (Klemm 1975).
However, the piscicolids characteristically are ec-
toparasites offish and feed on fish blood. The mus-
cular proboscis of piscicolids is effective in pene-
trating epidermal layers of fish wherever scales
are reduced or absent, although gills are favored
feeding sites.
Myzobdella lugubris
Myzobdella lugubris has not been recorded pre-
viously as a common ectoparasite of Columbia
River fish. Its distribution and host records are
included in previous publications that refer to the
genxis Illinobdella in which M. lugubris was for-
merly placed.
Myzobdella lugubris and M. (syn. Illinobdella)
moorei (Meyer and Moore 1954) were until re-
cently believed to be two distinct species. The
former was considered characteristic of brackish
and marine waters, while the latter was consi-
dered characteristic of freshwater. The distinction
was primarily ecological since anatomical fea-
TabLE 1. — Piscivorous leeches (Hirudinea) collected in this sur-
vey from teleost fishes in the central Columbia River,
Washington State.
Species
Host infected
Piscicolidae:
Myzobdella
Norttiern squawfish, Ptychocheilus
lugubris
oregonensis
Ctiiselmouth, Acrocheilus alutaceus
Brown bullhead, Ictalurus nebulosus
Largescale sucker, Catostomus macrocheilus
Bridgelip sucker, C columbianus
Piscicola
Chinook salmon, Oncorhynchus tshawytscha.
salmositica
fry
Sucker, Catostomus sp., fingerling
Glossiphoniidae:
Placobdella
montifera
Sucker, Catostomus sp , fingerling
Actinobdella
inequiannulata
Largescale sucker, C macrocheilus
tures of both species were remarkably similar.
However, M. lugubris and M. moorei are now con-
sidered to be a single euryhaline species (Sawyer
et al. 1975).
Further, it now appears that all members of the
related piscicolid genus Illinobdella are synony-
mous with M. lugubris. Thus species reported in
the literature as Illinobdella alba, I. elongata, and
/. richardsoni, as well as M. ( =/.) moorei, all prey-
ing on fish in North American waters (Meyer
1940, 1946b), are junior synonyms ofM. lugubris,
which holds taxonomic priority. Locality and host
records of the ubiquitous M. lugubris under these
synonyms are given by Hoffman (1967), Klemm
( 1972a, b, 1977), and Sawyer et al. (1975). Studies
on M. lugubris infesting the blue crab, Callinectes
sapidus, and the white catfish, Ictalurus catus , in a
South Carolina tidal river support this synonymy
(Daniels and Sawyer 1975).
Myzobdella lugubris was recovered from a wide
size range of adult chiselmouth, Acrocheilus
alutaceus, and northern squawfish, Ptychocheilus
oregonensis, in our collections, and less frequently
from adult suckers, Catostomus macrocheilus and
C columbianus. The associations were apparently
facultative. Myzobdella lugubris occurred primar-
ily in the oral cavity of chiselmouth (Figure 1 ) and
northern squawfish, where they were retained
during the struggle of hosts captured in overnight
net sets. Leeches were recorded and counted the
next day when fish were recovered. Many leeches
on the external surfaces of moribund or dead fish
may have detached before net recovery. Myzob-
della lugubris were never found in the mouth of
suckers but only in the axila of pelvic or pectoral
fins, on fin rays, or in the gill chambers.
Myzobdella lugubris was also a fairly common
ectoparasite of brown bullhead,/, nebulosus, col-
lected by angling in backwater sloughs of the cen-
tral Columbia and lower Snake Rivers during the
summer. Infestations on bullheads usually con-
sisted of one or two small leeches attached to the
pectoral or pelvic fins.
The incidence of M. lugubris on adult chisel-
mouth and northern squawfish (Table 2) shows
infestations only during June, July, and August
when C6lumbia River water temperatures ranged
from 13° to 21°C. The leeches were primarily sexu-
ally mature. Collections from the oral cavity of
chiselmouth in October 1975, 1977 and November
1977 contained numerous small, immature
leeches that had apparently hatched within the
preceding 1 or 2 mo.
927
>-#■
»*^^
0 mm
Figure l. — Four sexually mature Myzobdel I a lugubns in the oral cavity of chiselmouth collected in the central Columbia River. The
small subterminal mouth of the host is bordered by a cartilaginous upper and lower lip for grazing on sessile algae. The leeches occupy
most of the available space in the oral cavity when the mouth is closed.
Table 2. — Incidence of infestation of chiselmouth and northern
squawfish by the piscicolid leech, Myzohdel la luguhris, indicated
by infestation ratio.'
Ctiiselmouth
Norttie
rn squawfi
sh
Month
1975
1976
1977
1975
1976
1977
Jan.
0/5
0/4
0/5
0/2
0/1
Feb,
0/1
—
—
0/3
0/2
0/1
Mar.
—
—
—
0/1
—
—
Apr.
0/5
0/3
0/4
0/14
0/9
0/7
May
0/22
0/5
0/10
0/34
0/7
0/14
June
0/32
0/16
6/27
0/33
0/3
0/19
July
1/32
8/19
1/12
3/38
0/16
6/25
Aug.
1/19
0/4
1 9/42
0/13
0/2
0/8
Sept
0/46
0/9
0/13
0/18
0/15
0/4
Oct
27/37
0/10
^3/13
0/23
0/2
0/1
Nov
0/7
0/2
27/9
0/4
—
0/3
Dec.
0/6
0/7
0/2
0/2
—
0/1
'Infestation ratio = number of fish infested/number of fish examined.
^Numerous small leeches, recently hatched, were attached to some hosts.
According to Sawyer etal. ( 1975), M. lugubris is
a relatively warm-water species encountered most
often at 21°-30°C, occasionally at 16°-20°C, and
less often in colder water. They reported that the
leech appeared to be injured if the water was sud-
denly cooled to 10°-15°C in laboratory experi-
ments. Obviously some M. lugubris survive over
winter at low temperatures, but it must remain
inconspicuous due to dormancy in temperate re-
gions of North America. We have never recovered
M. lugubris during winter in the central Columbia
River, either from fish or from benthic samples
928
designed to quantitatively collect invertebrates.
Large, adult M. liigubns from Columbia River
fish were characterized by a green background
coloration, superficially suggesting that they fed
on algal cells ingested by the host. Chiselmouth
and suckers commonly feed on sessile, green-
colored diatoms from bottom substrates. However,
microscopic examination revealed that this col-
oration was due entirely to pigments in the adult
leech's musculature and not to the presence of
algae in their digestive tract. Myzobdella lugubris
fed entirely on fish blood cells and plasma.
Feeding on fish blood is clearly required by M.
lugubris for growth and reproduction. Copulating
M. lugubris were noted on fish. But since deposi-
tion of a cocoon requires hard substrates, sexually
mature leeches must eventually detach, thus free-
ing fish of infestations. The breeding, growth, and
reproductive cycle of piscivorous leeches may ac-
count for the periodic infestations of fish so fre-
quently documented in the literature.
In the Columbia River, the cycle in M. lugubris
is correlated with a seasonal change in water
temperatures, with peak activity occurring in late
summer and fall.
In tidal estuaries on the east coast of the United
States, M. lugubris has a life cycle that involves
tw^o hosts, a fish and a crustacean. It engorges on
fish blood before detaching to deposit cocoons on
crabs (Daniels and Sawyer 1975). This indicates
possible involvement of a freshwater crustacean
in the life cycle of M. lugubris in the Columbia
River. The only large crustacean available is
Pacifasticus leniusculus, but extensive collections
of this crawfish in previous years by the senior
author disclosed no attached leeches. Therefore,
stones are probably used as cocoon deposition sites
in the Columbia River.
Piscivorous leeches are often vectors of
hemoflagellates (genera Trypanoplasma and
Trypanosoma) found in the blood of freshwater
and marine fishes iKhaibulaev 1970; Becker
1977). Although we have occasionally detected
Trypanoplasma in Columbia River fish, we found
no hemoflagellates in the digestive tract of over 20
M. lugubris taken from various hosts.
We did not examine histopathology of leech at-
tachment and feeding sites in the oral cavity of
infested Columbia River fish, although petechiae
were evident during the fall on some chiselmouth.
Inflammatory conditions and hyperplasia were
described previously from a massive infestation of
M. lugubris (misidentified as Cystobranchus vir-
ginicus) on white catfish in Virginia (Paperna and
Zwerner 1974).
Piscicola salmositica
The salmonid leech, P. salmositica, was de-
scribed from specimens taken, in part, from sea-
run steelhead trout, Sal mo gairdneri, transferred
from the Columbia River to Mason Creek, Chelan
County, Wash. (Meyer 1946a). The leeches infest-
ing the fish originated either in the Columbia
River or from Mason Creek. Thus the salmonid
leech has previously been recorded from the upper
Columbia River system. This species is usually
associated with fall spawning runs of adult salmo-
nid fishes in coastal streams, but it occurs
elsewhere in the Pacific Northwest (Becker and
Katz 1965a).
We collected several P. salmositica at various
times from chinook salmon, Oncorhynchus
tshawytscha , fry and once from a fingerling
sucker. Each infestation consisted of a solitary
leech attached to the dorsal surface of its host and
feeding on blood. All specimens were taken in
April and June as water temperatures (10°-14°C)
increased and consisted of small leeches that pre-
sumably had hatched from cocoons the previous
fall or winter. Three specimens contained develop-
ing trypanoplasms among their intestinal con-
tents, evidence of prior feeding on infected fish.
Therefore, P. salmositica is confirmed as a vector
transmitting trypanoplasms among various fish
in the central Columbia River. The salmonid leech
is the only known vector of the piscine hemoflagel-
lateTrypanoplasma salmositica (Katz 1951) in the
Pacific Northwest (Becker and Katz 1965b).
Piscicola salmositica requires meals of fish blood
before detaching to deposit cocoons on bottom sub-
strates (Becker and Katz 1965a). Thus salmonid
leeches presumably occur among and infest popu-
lations of anadromous chinook salmon that spawn
each fall in the central Columbia River near our
fish collection sites. However, we have not de-
tected P. salmositica on transient adult fall
chinook salmon returning from the sea to spawn or
from downstream drifting, spawned out salmon
carcasses. Neither have we found salmonid
leeches on several adult steelhead trout and
spring-run chinook salmon examined during the
summer at the Ringold Hatchery (Washington
State Department of Game) above Richland. On
the basis of our observations, P. salmositica is not
an abundant leech in the central Columbia River.
929
Placobdella montifera
One immature P. montifera was recovered from
the dorsal surface of a fingerling sucker in early
October 1976. We also collected one adult speci-
men from beneath shoreline rocks at Umatilla
Reservoir during June where it was depositing a
cocoon. The species is not a common ectoparasite of
fish. It probably occurs mostly along reservoir
shorelines where water levels remain relatively
stable, rather than along the margins of the free-
flowing Columbia River above Richland.
Placobdella montifera has been reported to at-
tack aquatic worms, insect larvae, mussels, frogs,
toads, and fish, but the only specific host records
are fish (Hoffman 1967; Klemm 1972a, 1975, 1976;
Sawyer 1972; and others). This leech, as do most
glossiphoniids, broods its cocoon and carries its
young. An uncommon but widely distributed
species, P. montifera is listed as having been re-
ported previously from Washington (Klemm
1972b). Distributional records probably valid in-
clude British Columbia, Saskatchewan, Ontario,
and the northern states east of the Mississippi
River southward to Georgia (Sawyer 1972; Klemm
1977).
The host relationship for glossiphoniids is gen-
erally considered to be less obligatory than for
piscicolids, and most are omnivorous feeders. Ap-
parently A. inequiannulata, P. montifera, and P.
pediculata Hemingway 1908 are the only three
American glossiphoniids consistently reported to
parasitize fish. Several authors have reported P.
pediculata from the freshwater drum, Aplodinotus
grunniens, and Sawyer ( 1972) has indicated a high
degree of host specificity; it has not been reported
from the Pacific Northwest, nor would it be ex-
pected in this region due to its narrow host prefer-
ence.
Acknowledgments
Donald J. Klemm, Research Aquatic Biologist,
U.S. Environmental Protection Agency, collabo-
rated with us on leech identification and literature
survey, and critically reviewed the manuscript.
Fish collections were supported largely by en-
vironmental assessment programs for
Washington Public Power Supply System under
Contract 2311201335 between United Engineers
and Constructors and Battelle, Pacific Northwest
Laboratories.
Actinobdella inequiannulata
Six A. inequiannulata were collected from the
axila of the pelvic and pectoral fins of one adult
largescale sucker in mid-August 1975. According
to Sawyer (1972), this glossiphoniid is known from
Illinois, Minnesota, and Ohio; Klemm (1972a, b,
1977) adds Michigan, Pennsylvania, and New
York; and Daniels and Freeman (1976) add On-
tario. Actinobdella triannulata Moore 1924, a
name common in earlier literature, is now consid-
ered a junior synonym of A. inequiannulata
(Daniels and Freeman 1976; Klemm 1977).
Daniels and Freeman (1976) provide a rede-
scription of A. inequiannulata on basis of speci-
mens collected from two species of suckers (genus
Catostomus) and preserved material from the U.S.
National Museum. The species was earlier consid-
ered as free-living with no known hosts (Sawyer
1972; Klemm 1972a). Since its synonym A. trian-
nulata displayed a predilection for suckers
(Hoffman 1967), the host preference of A. in-
equiannulata is now partially resolved. Little is
known of its ecology and life cycle. We did not
examine our specimens for ingested fish blood.
Literature Cited
Becker, C. D.
1977. Flagellate parasites offish. /« J. P. Kreier (editor),
Parasitic protozoa. Vol. 1, Taxonomy, kinetoplastids, and
flagellates offish, p. 357-416. Academic Press, N.Y.
Becker, C. D., and M. Katz.
1965a. Distribution, ecology and biology of the salmonid
leech, Piscicola salmositica (Rhynchobdellae: Pis-
cicolidae). J. Fish. Res. Board Can. 22:1175-1195.
1965b. Transmission of the hemoflagellate Cryptobm
salmositica Katz, 1951, by a rhynchobdellid vector. J.
Parasitol. 51:95-99.
Daniels, B., and R. S. Freeman.
1976. A review of the genus Actinobdella Moore, 1901
(Annelida:Hirudinea). Can. J. Zool. 54:2112-2117.
Daniels, B. a., and R. T, Sawyer.
1975. The biology of the leech Myzobdella lugubris infest-
ing blue crabs and catfish. Biol. Bull. (Woods Hole)
148:193-198.
Gray, R. H., and D. D. Dauble.
1977. Checklist and relative abundance of fish species
from the Hanford reach of the Columbia River. North-
west Sci. 51:208-215.
HOFFMAN. G. L.
1967. Parasites of North American freshwater fishes.
Univ. Calif Press, Berkeley, 486 p.
KHAIBULAEV, K. KH.
1970. The role of leeches in the life cycle of blood parasites
of fishes. [In Russ., Engl.^summ.] Parasitologya
(Lenningr.) 4:13-17.
930
\
Klemm. D. J.
1972a. The leeches (Annelida: Hirudinea) of Michi-
gan. Mich. Acad. 4:405-444
1972b. Freshwater leeches (Annelida: Hirudinea) of North
America. U.S. Environ. Prot. Agency, Biota of Freshwa-
ter Ecosystems, Ident. Man. 8, 53 p.
1975. Studies on the feeding relationships of leeches (An-
nelida: Hirudinea) as natural associates of mollusks.
Sterkiana 58:1-50, 59:1-20.
1976. Leeches (Annelida: Hirudinea) found in North
American mollusks. Malacol. Rev. 9:63-76.
1977. A review of the leeches ( Annelida-Hirudinea) in the
Great Lakes region. Mich. Acad. 9:397-418.
MEYER, M. C.
1940. A revision of the leeches (Piscicolidae) living on
fresh-water fishes of North America. Trans. Am. Mi-
crosc. Soc. 59:354-376.
1946a. A new leech, Piscicola sahnositica n. sp. (Pis-
cicolidae), from steelhead trout iSatnio gairdneri
gairdneri Richard.son, 1838). J. Parasitol. 32:467-476.
1946b. Further notes on the leeches (Piscicolidae) living
on fresh-water fishes of North America. Trans. Am. Mi-
crosc. Soc. 65:237-249.
Papern.'\. l, and D. E. ZWERNER.
1974. Massive leech infestation on a white catfish (/c-
talurus catus): a histopathological consideration. Proc.
Helminthol. Soc. Wash. 41:64-67.
Sawyer. R. T.
1972. North American freshwater leeches, exclusive of the
Piscicolidae, with a key to all species. 111. Biol. Monogr.
46, 154 p.
Sawyer. R. T., A. R. Lawi.er. a.vd R. M. Overstreet
1975. Marine leeches of the eastern United States and the
Gulf of Mexico with a key to the species. J. Nat. Hist.
9:633-667.
C. Dale Becker
Dennis D. Dauble
Ecosystems Department
Battelle. Pacific Northwest Laboratories
Richland, WA 99352
INDUCED SPAWNING AND LARVAL REARING
OF THE YELLOWTAIL FLOUNDER,
LIMANDA FERRUGINEA
The yellowtail flounder, Limanda ferruginea
(Storer), a commercially important flatfish, occurs
in North American continental waters from the
north shore of the Gulf of St. Lawrence southward
to the lower part of Chesapeake Bay (Bigelow and
Schroeder 1953). The yellowtail flounder spawns
from March through August where water temper-
atures over its range vary from about 5° to 12°C
(Colton 1972). The eggs are pelagic and lack an oil
globule; diameter of the live eggs (range 0.79-1.01
mm I averages 0.88 mm (Colton and Marak 1969).
A program to obtain viable yellowtail eggs
through hormone induction, to rear larvae through
metamorphosis, and to determine the mechanisms
of survival of early life stages under controlled
laboratory conditions was undertaken. The suc-
cessful induction of yellowtail flounder and sub-
sequent rearing of the larvae through metamor-
phosis marks the first time the early life history of
this flatfish has been completed in the laboratory.
Materials and Methods
Adult yellowtail flounder were captured by otter
trawling in Block Island Sound in the winters of
1974, 1975, and 1976 and transported to the Nar-
ragansett Laboratory in a 380-1 live car equipped
with an aerator. In the laboratory the fish were
held in a 28,000-1 aquarium. A continual supply of
filtered seawater was pumped to the aquarium
from Narragansett Bay.
Individuals presumed to be sexually mature
were selected by length. Available length-weight
data (Lux 1969) indicated that yellowtail flounder
in southern New England waters mature when
they attain a length near 35 cm or an age of 3 yr
(Lux and Nichy 1969). After acclimating in the
laboratory, the fish were segregated by sex, mea-
sured and weighed, and tagged with numbered
plastic pennants secured through the caudal pe-
duncle. Yellowtail flounder were sexed by holding
the white underside to the light and looking
through the flesh. The outline of the ovary extend-
ing posteriorly from the mass of viscera can read-
ily be seen even in immature females (Royce et al.
1959). Yellowtail flounder are delicate and excit-
able. To minimize injury, the fish were anes-
thetized in a solution of tricane methanesulfonate
(MS-222') at a concentration of 1:20,000 (Leitritz
and Lewis 1976) during each examination.
While the fish were held in captivity, a photo-
period of 1 1 h of light and 13 h of dark simulated
spawning light conditions. Four banks of fluores-
cent lights (each bank composed of 16 40-W bulbs)
were suspended 4 m from the ceiling and mechani-
cally timed. The light banks were sequentially
turned on and off in the morning and evening at
15-min intervals to simulate dawn and twilight.
Prior to receiving hormones the fish were fed a daily
diet of chopped frozen hake, whiting, or squid. Dur-
ing the trials the fish were not fed.
'Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
FISHERY BULLETIN: VOL. 76. NO. 4. 1979.
931
The effectiveness of the pituitary preparations
was evaluated by monitoring the gonosomatic
index (GSI), ovulation, success of egg fertilization,
and hatching success. Hormones were prepared on
the day of injection, and dosages were established
by the weight of each individual fish. A saline solu-
tion of isotonic sodium chloride was used as a car-
rier. All injections were administered (2-cm'^
syringe, 20 gage 3.85-cm needle) intramuscularly
into the back below the dorsal fin. Inserting and
withdrawing the needle slowly aided in retaining
most of the fluid in the flesh. After injection, the
flesh of the fish was massaged to diffuse the fluid
into the muscles.
Sexually mature fish were hand stripped and
the eggs fertilized in a polyethylene pan. Several
thousand eggs were collected at each spawning,
and the sperm of two males was used to fertilize
the eggs from each female. Yellowtail flounder are
nonsynchronous spawners (Bigelow and
Schroeder 1953), and multiple spawnings occur-
red among most induced fish. The fecundity of
yellowtail flounder increases with age and body
length, and an individual female may yield from
350,000 to more than 4,000,000 eggs during the
spawning season (Pitt 1971).
The state of ova maturation of the experimental
fish was observed at the start and termination of
each experiment. Before injecting, a polyethylene
cannula was inserted into the oviduct and oocyte
samples were orally withdrawn. The oogenesis of
oocytes was divided by microscopic observation
into three general histological stages:
Stage I - the primary oocyte stage, oocytes con-
tained cytoplasmic vacuoles and mea-
sured between 0.1 and 0.25 mm.
Stage II - the yolk globule stage, cytoplasm of
oocytes was filled with dense yolk
granules and measured up to 0.6 mm.
Stage III - ripe stage, hyaline oocytes present
and measured 0.75-1.00 mm in size.
Fertilized eggs were incubated in static, aer-
ated, black-sided aquaria that had been inocu-
lated with the green algae Dunaliella sp. A single
application of penicillin (25 international units
iIU]/ml) and streptomycin (0.02 mg/ml) at the con-
centration of 50 mg/1 was effective in controlling
bacterial contamination of the aquaria in almost
all cases.
Three series of experiments were undertaken to
determine the effectiveness of the hormone injec-
tions (Table 1). The first trial was conducted in
winter 1975 to determine if induced spawning
would occur at low winter water temperatures.
The second and third were conducted in the
springs of 1976-77 and coincided with the yellow-
tail flounder's natural spawning season.
Hormone dosage levels of 2, 5, and 10 mg/kg fish
and frequency of injecting were dictated by previ-
ous successful results obtained with the summer
flounder, Paralichthys dentatus, (Smigielski
1975a) and winter flounder, Pseudopleuronectes
americanus, (Smigielski 1975b). After each trial
the female fish were killed and reweighed, the
ovaries were examined, and gonosomatic indices
were recorded. Prior to receiving hormone injec-
tions, all the female test fish in the first trial were
in Stage I of oocyte oogenesis, and most females
prior to the second and third trials were in Stage
II. Males were not injected in the second and third
trials because they were sexually ripe.
Rfsults and Discussion
First Trial
In the first trial (Table 1), most females in the
group receiving 10 daily injections of 2 mg pitui-
tary were refractory with low GSI values (7-13%).
One fish hydrated but did not ovulate, and a small
number of Stage II ova were found in the ovaries.
Hydration is an increase in total body weight. The
weight gain is due mostly to water intake and is
reflected by higher GSI values as most of the water
appears to go into the gonads. Excessive hydration
is manifested by grossly bloated fish which in some
instances can hydrate to the point of death without
ovulating.
Table l. — Hormone dosages, water temperatures, and number of yellowtail flounder in each trial.
Trial 1 - January 1975
Water temperatures. 3 -6 C
(Mean 510)
Wate
Tnal 2 ■ April 1976
>r temperatures. 7 -
(Mean 9.2 C)
IOC
Water
Trial 3 • April
temperatures.
(Mean 10.1
1977
8,5'-
C)
12,5'C
Dally carp
pituitary
dosages
Number
of
females
Number
of
males
Uninjected
controls
Number
of
females
Number
of
males
Sham
injected
controls
Number
of
females
Number
of
males
Sham
Injected
controls
2.0 mg/kg fish
5.0 mg/kg fish
10.0 mg/kg fish
6
6
6
4
4
4
3
3
3
8
8
7
0
0
0
4
4
4
9
0
4
932
In the group of females receiving 10 daily injec-
tions of 5 mg, three fish were refractory with low
GSI values (10-15'^>f ), and three hydrated but did
not ovulate. Two of the latter fish contained a
small number of Stage II ova; the other developed
a cloacal plug of membranous tissue and Stage I
ova.
In the group receiving 10 daily injections of 10
mg, all were refractory with low GSI values (9-
1 17( ), except for one fish that hydrated but did not
ovulate. A membranous plug developed in the
cloaca of this fish, and it was bloated. A very small
number of Stage II ova were found in the ovaries.
There was no indication of sexual ripening in the
uninjected control fish, and their GSI values were
low dO-lS'^). Copious semen was obtained from
the males injected at all three dosages; however,
fertilization was not attempted. It was reasoned
from the first trial that low GSI values (7-159r ) of
females coupled with low water temperatures
(3°-6°C, mean 5.1°C) that were less than optimum
inhibited the effectiveness of the hormones, for
although some fish hydrated, they did not ovulate.
Second Trial •
The results obtained from the second trial were
variable (Table 2). All but one fish receiving injec-
tions of 2 mg hydrated and ovulated. Two fish died
during the trial; one, after yielding spawn on two
occasions, developed a membranous plug and be-
came grossly bloated.
In the group receiving injections of 5 mg, two
fish ovulated but the eggs obtained were not fer-
tile. Three other fish developed plugs and hy-
drated to the point of death. Injections were dis-
continued at the first sign of abnormal hydration,
but the fish continued to imbibe water.
In the group that received 10 mg, five fish ex-
perienced excessive hydration manifestated by
bloating, plug formation, and, in two instances,
death. The membranous plugs were identical to
those that developed in the test group that re-
ceived hormone dosages of 5 mg. The controls had
four fish with signs of hydration but no Stage III
ova were found in their ovaries.
Third Trial
The results of the third trial paralleled those of
the second trial at a dosage of 2 mg. Seven of the
experimental females hydrated normally and ovu-
lated (Table 3). Fertilization and hatching of these
eggs were satisfactory and the larvae were nor-
mal. The remaining two fish died during the trial,
and their ovaries had a small number of Stage III
ova. The control fish neither hydrated nor ovu-
lated; GSI values were fairly high, but Stage III
ova were absent.
The anomalous hydration with bloating and
formation of membranous plugs during hormonal
induction is not unique. Clemens and Grant ( 1964)
injected female goldfish, Carassius auratus, with
carp pituitary and observed that the gonadal
water content increased, apparently in association
with ovulation. The hormone regulating the hydra-
tion process appeared to be a gonadotropin.
Shehadeh and Ellis (1970) reported the forma-
tion of plugs in the cloaca in striped mullet, Miigil
cephaliis, treated with a combination of salmon
pituitary and Synahorin. Sinha ( 1971 ) studied the
gonadal hydration response of Puutis gonionotus
using the second fraction of molecular seived carp
pituitary extract and suggested that the second
fraction is involved in osmoregulation, since an
injection of an additional amount enhances the
rate of water transport resulting in maturation.
Hirose and Ishida (1974) studied the effects of
Cortisol and human chorionic gonadotropin
(HCG) in ayu., Plecoglossus altiuelis, and reported
that the water content of the ovary from
hormone-treated fish increased by 6%. Smigielski
(1975b) reported a similar response in winter
flounder injected with pregnant mare serum
(PMS) and HCG. Hirose ( 1976) demonstrated that
gonadotropin-treated ayu imbided a greater quan-
tity of water than control ayu. He suggested that
gonadotropin may act on the sodium and po-
tassium system or permeability of the egg mem-
brane.
Hydration is a normal and necessary prelude to
maturation and ovulation. The cases of abnormal
hydration experienced with yellowtail flounder
may be attributed to an adverse reaction to hor-
mone dosage. Most of the test fish that hydrated
abnormally and became bloated had an increase in
body weight of more than 10%. The increase in
body weight appeared to be a result of the fish
imbibing an excess amount of water. An excessive
amount of introduced hormone may upset the
water transport or sodium potassium systems, re-
sulting in more water bing imbibed.
In conclusion, it appears that water tempera-
tures higher than 6°C and GSI values approaching
209^^ coupled with carp pituitary injections ap-
proximating 2 mg/kg offish is an effective combi-
933
Table 2. — Effects of carp pituitary on yellowtail flounder receiving daily injections. All fish were exposed to 11L:13D photoperiod and
water temperatures of 8.5°-12.5°C (Mean 10. TC). Symbols: + = did, 0 = did not hydrate or ovulate.
Dosage
No. of Total Initial
injec- length body weight Weight change GSI
tions (mm) (g) (°o initial wt) (% final wt)
Effect
Hydrated Ovulated
Date of
spawning
1976
Ferti-
lization
(%)
Hatch
(%)
2.0 mg/kg fish
Controls
5.0 mg/kg fish
Controls
10 mg/kg fish
Controls
5
6
4
5
10
5
10
5
4
5
5
7
10
10
348
340
353
435
280
350
396
392
421
817
263
381
+ 6.94
- 1 84
-330
-0.96
-049
1.12
19,7
5.6
5.1
4.3
7,0
8.4
450
882
+ 954
240
387
721
+ 0.98
11,2
420
667
t2.12
18.6
392
611
+ 2.63
19.6
361
409
+ 1.09
10.5
381
562
+ 0.44
9.8
401
683
.+2.11
18.6
501
1,489
+ 9.62
29.5
307
566
+ 11.69
26.1
435
1.131
+ 13.17
249
348
762
-019
4.6
351
491
+ 8.47
30.8
336
418
^0.96
11.0
361
521
+ 2.92
19.3
392
587
+ 2.87
19.5
406
702
+ 2.19
15.1
339
367
+ 1 87
129
467
1,259
+ 11.15
286
346
593
+ 13.57
27,6
352
463
+ 15.11
29,6
433
1,245
+ 10.76
25,3
341
485
+ 14.06
29.1
360
501
+ 1.16
126
348
437
+ 1.84
14.2
343
428
+ 1.23
15.1
389
672
+ 2.63
17.9
430
891
+ 4.78
19.3
369
551
+ 1.63
14.3
+
+
0
0
+"
+ 2
0
0
0
+
+
0
0
+
Apr
6
80
Apr
7
70
+
Apr
6
75
Apr
7
70
Apr
8
80
Apr
9
90
Apr
11
70
+
Apr
4
75
Apr
5
80
Apr
6
70
Apr
7
70
Apr
9
75
+
Apr
5
80
Apr
7
75
Apr
8
80
+
Apr
8
80
Apr
9
80
Apr
11
70
+2
Apr
1
80
Apr
3
75
+
Apr
5
80
Apr
6
75
0
0
0
0
+ 3
0
0
0
0
+3
0
0
0
0
0
0
0
+ 3
0
0
0
0
0
0
0
0
70
75
55
60
70
75
60
80
75
60
50
75
75
70
70
60
50
80
75
70
80
65
'Died, -' 10% Stage III ova in ovaries, not fertilized.
^Plug formed, fish became bloated and hydrated to point of death.
^Stage III ova in ovaries, not fertilized.
"Plug formed, fish became bloated.
nation for inducing spawning of yellowtail floun-
der.
Larval Rearing
Fertilized yellowtail flounder embryos were in-
cubated in 64-1, rectangular, black-sided, static,
well-aerated aquaria at a density of approxi-
mately 80 embryos/1. The incubating and rearing
temperature was 10°C and the salinity 32%o.
Banks of 40-W timed fluorescent lights suspended
1 m over the aquaria simulated a day and night
regimen of 15 h light and 9 h dark ( 15L:9D). The
aquaria were inoculated with green algae
(Dunaliella sp.) which may have aided in the re-
moval of metabolic waste products produced by the
larvae. The algae also served to sustain the zoo-
plankton introduced as food. A single application
of penicillin (25 lU/ml) and streptomycin (0.02
mg/ml) was effective in controlling bacterial con-
tamination in almost all cases.
At 10°C, hatching occurred 6-7 days after fertili-
zation. Yolk absorption occurred 4-5 days after
hatching, which coincided with first feeding. The
larvae averaged 2.75 mm long upon hatching and
possessed a completely formed gut. The eyes were
934
Table 3. — Effects of carp pituitary onyellowtail flounder receiving daily injections. All fish were exposed to llL:13Dphotoperiod and
water temperatures of 8.5°-12.5°C (Mean 10.1°C). Symbols: + = did, 0 = did not hydrate or ovxilate.
Dosage
No of
Injec-
tions
Total
length
(mm)
Initial
body weight
(g)
Weight change
("o initial wt)
GSI
(°o final wt)
Date of Fertili-
fcftect spawning zation Hatch
Hydrated Ovulated 1977 (%) (%)
2.0 mg/kg fish
Controls
5
6
5
3
391
552
453
898
474
1.298
424
878
412
733
364
486
396
598
405
694
437
945
392
601
446
891
409
667
359
472
+ 0.56
+ 9.13
+ 470
+ 7.16
+ 2.04
+ 0.97
+ 1.62
+ 2.79
+ 3.17
+ 2.16
+ 2.79
+ 1,92
*2 19
46
21.2
6.9-
20.8
5.3
10.7
B.8
7.7
93
18.0
18.9
167
18.4
+ '
+
+
+
+
0
0
0
0
Apr 12
75
0
+
Apr 12
80
Apr 14
80
0
+
Apr. 16
50
Apr 17
60
Apr 18
60
+
Apr 17
60
Apr 18
70
+
Apr. 17
70
Apr 18
70
+
Apr 16
85
+
Apr. 14
70
0
0
0
0
80
70
75
60
40
45
55
65
60
55
60
80
'Died: Stage III ova in ovaries
pigmented at 1 day and the mouth was functional
at 1-3 days after hatching. No abnormahties were
observed in hormone-induced larvae.
Twenty larvae were sampled weekly (Table 4).
The specific growth rates for the 63-day period
from first feeding averaged 9.9T7(,day for dry
weight (micrograms) and 2.759^ standard length
(millimeters). The first fish metamorphosed 54
days after hatching at 17.00 mm standard length.
By the 63d day after hatching, all the larvae had
completed metamorphosis, and average length
was 17.40 mm.
Wild copepod nauplii were collected daily from a
nearby estuary and fed to the larvae after being
sieved to obtain the proper particle size. Larvae of
the yellowtail flounder required small food or-
ganisms (<100/xm in largest dimension) to in-
itiate feeding. The most difficult aspect of rearing
the larvae was the problem of obtaining enough
food organisms in the size range required. Larval
mortality was high for the first 2 wk of feeding,
possibly caused by starvation. However, yellow-
tail flounder larvae are able to survive for consid-
T.ABLE 4. — Size of yellowtail flounder larvae reared artificially
at 10°C from hac ing to metamorphosis. Average sizes of 20
larvae are followed by standard deviation.
Days after
l^ean length
Mean dry weight
first feeding
(mm)
(MQ)
1
3.08^0.20
.16.2=4.3
7
3.39*0.25
19.8=43
14
5 16i047
560=201
21
5.92-079
899 = 574
27
682±0,51
126.5 = 340
34
668 + 089
161 5 = 664
41
8 95=1 03
608 6 = 340 2
48
10.53±4.38
1.133-3 = 1.267 8
55
14.73=398
5.576.6 = 2.694.1
63
17.40±2.33
8,635.9 = 3,058.4
erable periods of days without exogenous food.
Some larvae were maintained at 8°C and fed suc-
cessfully and survived after being deprived of food
for 10 days after hatching (Smigielski unpubl.
data). As the larvae increased in size through
metamorphosis, larger food organisms such as
adult copepods, the rotifer Branchion us plicatilis,
and the brine shrimp, Artemia sp., were offered.
Acknow ledgments
The author expresses his appreciation to Hugh
Poston of the Tunison Laboratory of Fish Nutrition,
Cortland, N.Y., and Geoffrey C. Laurence of
the Northeast Fisheries Center Narragansett
Laboratory, National Marine Fisheries Service,
NOAA for their many helpful criticisms of the
manuscript, and Kathy Dorsey and Thomas
Halavik for their technical assistance.
Literature Cited
BIGELOW, H. B., AND W. C. SCHROEDER.
1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv.,
Fish. Bull. 53, 577 p.
Clemens, H. P., and F. B. Grant.
1964. Gonadal hydration of carp [Cyprinus carpio) and
goldfish iCarassius auratus) after injections of pituitary
extracts. Zoologica (N.Y.) 49:193-210.
COLTON, J. B., Jr.
1972. Temperature trends and the distribution of
groundfish in continental shelf waters. Nova Scotia to
Long Island. Fish Bull., U.S. 70:637-657.
Colton, J. B., Jr., and R. R. Marak.
1969. Guide for identifying the common planktonic fish
eggs and larvae of Continental Shelf waters, Cape Sable to
Block Island. U.S. Bur. Commer. Fish., Biol. Lab.,
Woods Hole, Mass., Lab. Ref. 69-9, 43 p.
935
HIROSE, K. -
1976. Endocrine control of ovulation in medaka iOryzias
latipes) and ayu iPlecoglossus altivelis). J. Fish. Res.
Board Can. 33:989-994.
HiROSE, K., AND R. ISHin.A.
1974. Effects of Cortisol and human chorionic gonadotrophin
(HCG) on ovulation in ayuPlecoglossus altivelis (Temminck
& Schlegel) with special respect to water and ion bal-
ance. J. Fish Biol. 6:557-564.
Leitritz, E., and R. C. Lewis.
1976. Trout and salmon culture (hatchery methods). Calif.
Dep. Fish Game, Fish Bull. 164, 197 p.
LUX, F. E.
1969. Length-weight relationships of six New England
flatfishes. Trans. Am. Fish. Soc. 98:617-621.
Lux, F. E., AND F. E. NICHY.
1969. Growth of yellowtail flounder, L/wo/!(/o ferruginea
(Storer), on three New England fishing grounds. Int.
Comm. Northwest Atl. Fish. Res. Bull. 6:5-25.
Pitt, T. K.
1971. Fecundity of the yellowtail flounder (Limanda fer-
ruginea) from the Grand Bank, Newfoundland. J. Fish.
Res. Board Can. 28:456-457.
ROYCE, W. F., R. J. BULLER, AND E. D. PREMETZ.
1959. Decline of the yellowtail flounder {Limanda fer-
ruginea) off New England. U.S. Fish Wildl. Serv., Fish.
Bull. 59:169-267.
Shehadeh, Z. H., and J. N. Ellis.
1970. Induced spawning of the striped mullet Mugil
cephalus L. J. Fish Biol. 2:355-360.
SINHA, V. R. P.
1971. Induced spawning in carp with fractionated fish
pituitary extract. J. Fish Biol. 3:263-272.
Smigielski, a. S.
1975a. Hormone-induced spawnings of the summer floun-
der and rearing of the larvae in the laboratory. Prog.-
Fish Cult. 37:3-8.
1975b. Hormonal-induced ovulation of the winter floun-
der, Pseudopleuronectes americanus. Fish. Bull., U.S.
73:431-438.
Alphonse S. Smigielski
Northeast Fisheries Center Narragansett Laboratory
National Marine Fisheries Servece, NOAA
R.R. 7A, Box 522 A
Narragansett. RI 02882
TRACE METAL CONTAMINATION OF
THE ROCK SCALLOP, HINNITES GIGANTEUS,
NEAR A LARGE SOUTHERN
CALIFORNIA MUNICIPAL OUTFALL'
Los Angeles County's submarine discharge of
municipal wastewater from the Joint Water Pol-
lution Control Plant (JWPCP) off Palos Verdes
Peninsula is the single largest anthropogenic
source of trace metals to the marine ecosystem off
southern California. The 1974 annual mass emis-
sion rates of chromium, copper, and zinc via this
discharge (4.8 x 10" 1/yr, which underwent pri-
mary treatment only) were about 400, 300, and
850 t, respectively; these were approximately 10
times the corresponding inputs measured in
1971-72 surface runoff from southern California
( Young et al. 1973). As a result, bottom sediments
around this submarine outfall system are highly
contaminated by a number of trace metals (Gallo-
way 1972; Young et al. 1975). Here we report ab-
normal levels of seven metals in three tissues of
the filter-feeding rock scallop, Hinnites gigan-
teus,^ that was collected in the discharge zone and
thus had been exposed to suspended wastewater
particulates. (The adductor muscle of this bivalve
mollusc is considered to be a delicacy, and scallops
near the discharge are sought by sport divers.)
Procedures
During 1974, divers collected eight scallops
within the size range generally consumed ( 10 to 25
cm in diameter) from depths of about 20 m at three
stations in the discharge zone between Whites
Point and Point Vicente: these stations were <1
km off Palos Verdes Peninsula. Six scallops in the
same size range also were taken from control sta-
tions at similar depths off Santa Catalina and
Santa Barbara Islands (Figure 1). To check our
1974 results, during 1976 eight specimens within
this size range were again collected from this re-
gion in the discharge zone. However, we were not
able to obtain additional island samples; there-
fore, five specimens were collected from each of
two coastal stations located approximately 50 km
to the north and south of Palos Verdes Peninsula.
The samples were frozen in plastic bags after col-
lection. Later, digestive gland, gonad, and adduc-
tor muscle tissues were excised from each speci-
men before it was fully thawed, using a new
carbon steel scalpel and a cleaned Teflon^ sheet;
the tissues were placed in cleaned polyethylene
vials. Care was taken to avoid contaminating the
gonadal or muscle tissue samples with sediments
or juices from the digestive glands.
Following dissection, each sample ( 1 to 2 g wet
weight) was digested in 10 ml of a 1:1 nitric acid
'Contribution No. 85 of the Southern California Coastal Water
Research Project.
'^Formerly Hinnites multirugosus (Roth and Coan 1978).
■''Reference to trade names does not imply endorsement by the
National Marine Fisheries Service, NOAA.
936
FISHERY BULLETIN: VOL. 76. NO. 4. 1979.
1 ■ ■ ■
I 1
LOS ANGELES
l°N
\ PALOS VERDES
) PENINSULA
POINT ,* ^r.- — -"""x
VICENTE rVwHITES \
40-
.
POINT
JWPCP
OUTFALLS
•0
fi — V
SANTA
^--^^^
BARBARA i
) \
20'
f
^
N
SANTA CATALINA 1
0 15
1 1
30
1
K \1
1 1 1
119°W
40-
20'
118°W
FliilRK 1. — Outfall and island control sites ofT Los Angeles,
Calif., for collection of rock scallops.
solution (ultrahigh-purity reagent grade) until
the remaining volume was about 3 ml. This proce-
dure was repeated once, and the final residue was
filtered through an acid-washed Whatman No. 40
filter. The filtrate was then diluted to an appro-
priate volume, and the treated sample was
analyzed by atomic absorption spectrometry.
Silver, chromium, copper, nickel, and lead were
measured by injecting 2.5 /u,l of sample into a
graphite furnace; cadmium and zinc levels were
determined by aspirating the sample into an air-
acetylene flame.
Process blanks were analyzed with all samples.
Typical blank corrections were